EUR-Lex & EU Commission AI-Powered Semantic Search Engine

Modern Legal

Query in any language with multilingual search
Access EUR-Lex and EU Commission case law
See relevant paragraphs highlighted instantly

We Found Similar Cases for You

Home

UNIVERSAL MUSIC GROUP / EMI MUSIC

M.6458

UNIVERSAL MUSIC GROUP / EMI MUSIC

September 20, 2012

With Google you find a lot.
With us you find everything. Try it now!

I imagine what I want to write in my case, I write it in the search engine and I get exactly what I wanted. Thank you!

Valentina R., lawyer

This text is made available for information purposes only. A summary of this decision is published in all EU languages in the Official Journal of the European Union.

Case No COMP/M.6458 - Universal Music Group/ EMI Music

Only the English text is authentic.

REGULATION (EC) No 139/2004 MERGER PROCEDURE

Article 8 (2) Date: 21/09/2012

Brussels, 21/09/2012 C(2012) 6459 final

PUBLIC VERSION

COMMISSION DECISION

of 21/09/2012

(Only the English text is authentic)

1. THE NOTIFICATION ................................................................................................. 6

2. THE PARTIES ............................................................................................................. 7

3. THE PROPOSED CONCENTRATION...................................................................... 7

4. UNION DIMENSION ................................................................................................. 8

5. PROCEDURE .............................................................................................................. 8

6. MARKET DEFINITION ............................................................................................. 9

6.1. Introduction to the recorded music industry ................................................................ 9

6.1.1. Background .................................................................................................................. 9

6.1.2. The recorded music sector ......................................................................................... 12

6.1.2.1. Major record companies and independent record companies .................................... 12

6.1.2.2. The recorded music value chain ................................................................................. 13

6.1.3. Other activities related to recorded music .................................................................. 25

6.1.3.1. Ancillary services ....................................................................................................... 26

6.1.3.2. Music publishing ........................................................................................................ 27

6.2. Relevant product markets ........................................................................................... 29

6.2.1. A&R ........................................................................................................................... 29

6.2.1.1. The Notifying Party's view ........................................................................................ 29

6.2.1.2. The Commission's assessment ................................................................................... 29

6.2.2. Wholesale of recorded music ..................................................................................... 30

6.2.2.1. The Notifying Party's view ........................................................................................ 30

6.2.2.2. The Commission's assessment ................................................................................... 33

6.2.3. Retailing of recorded music ....................................................................................... 44

6.2.3.1. The Notifying Party's view ........................................................................................ 44

6.2.3.2. The Commission's assessment ................................................................................... 45

6.2.4. Music-related activities outside the recorded music sector........................................ 45

6.2.4.1. The Notifying Party's view ........................................................................................ 45

6.2.4.2. The Commission's assessment ................................................................................... 46

6.2.5. Piracy.......................................................................................................................... 48

6.2.5.1. The Notifying Party's view ........................................................................................ 48

6.2.5.2. The Commission's assessment ................................................................................... 51

6.3. RELEVANT GEOGRAPHIC MARKETS ................................................................ 55

6.3.1. A&R ........................................................................................................................... 55

6.3.1.1. The Notifying Party's view ........................................................................................ 55

6.3.1.2. The Commission's assessment ................................................................................... 55

6.3.2. Wholesale of recorded music ..................................................................................... 55

6.3.2.1. The Notifying Party's view ........................................................................................ 55

6.3.2.2. The Commission's assessment ................................................................................... 56

6.3.3. Retail of recorded music ............................................................................................ 58

6.3.3.1. The Notifying Party's view ........................................................................................ 58

6.3.3.2. The Commission's assessment ................................................................................... 58

6.3.4. Music-related activities outside the recorded music sector........................................ 58

6.3.4.1. Ancillary services ....................................................................................................... 58

6.3.4.2. Music publishing ........................................................................................................ 58

7. THE COMPETITIVE ASSESSMENT IN THE RELEVANT MARKETS.............. 59

7.1. HORIZONTAL NON-COORDINATED EFFECTS ................................................ 59

7.1.1. Analytical framework................................................................................................. 59

7.1.2. Market shares ............................................................................................................. 59

7.1.2.1. IFPI data ..................................................................................................................... 60

7.1.2.2. GfK data ..................................................................................................................... 61

7.1.2.3. OCC data .................................................................................................................... 62

7.1.2.4. Platform data .............................................................................................................. 62

7.1.2.5. Chart and radio airplay data ....................................................................................... 62

7.1.2.6. Market shares according to the different data sources ............................................... 63

7.1.2.7. The Notifying Party's view ........................................................................................ 78

7.1.2.8. The Commission's assessment ................................................................................... 83

7.1.3. EMI's competitive position ...................................................................................... 105

7.1.3.1. The Notifying Party's view ...................................................................................... 105

7.1.3.2. The Commission's assessment ................................................................................. 108

7.1.4. Customers' limited possibilities of switching supplier ............................................. 118

7.1.5. Likely effects of the proposed concentration ........................................................... 119

7.1.5.1. The Notifying Party's view ...................................................................................... 124

7.1.5.2. The Commission's assessment ................................................................................. 124

7.1.5.3. Comparison of Universal and EMI's agreements with online platforms ................. 125

7.1.5.4. Formal economic analysis of commercial conditions for digital customers ............ 134

7.1.5.5. Results of the market investigation .......................................................................... 137

7.1.5.6. Evidence from the merging parties .......................................................................... 138

7.1.5.7. EEA-wide market for the wholesale of digital recorded music ............................... 139

7.1.5.8. National markets for the wholesale of digital recorded music ................................. 140

7.1.5.9. Control share analysis .............................................................................................. 141

7.1.5.10. Overall conclusion on commercial conditions for digital customers ....................... 156

7.1.5.11. Likely impact on consumers .................................................................................... 156

7.1.5.12. Commercial conditions for physical customers and foreclosure of competitors ..... 168

7.1.6. Piracy........................................................................................................................ 169

7.1.6.1. The Notifying Party's view ...................................................................................... 169

7.1.6.2. The Commission's assessment ................................................................................. 170

7.1.7. Countervailing factors .............................................................................................. 177

7.1.7.1. Buyer power ............................................................................................................. 177

7.1.7.2. Entry ......................................................................................................................... 181

7.2. COORDINATED EFFECTS ................................................................................... 184

7.2.1. The view of the Notifying Party............................................................................... 184

7.2.2. The Commission's assessment ................................................................................. 186

7.3. VERTICAL AND CONGLOMERATE EFFECTS ................................................ 191

7.3.1. The Notifying Party's view ...................................................................................... 191

7.3.2. The Commission's assessment ................................................................................. 192

7.3.2.1. Digital retail distribution of recorded music ............................................................ 193

7.3.2.2. Online music video services..................................................................................... 195

7.3.2.3. Mobile telephony and broadband services ............................................................... 196

7.3.2.4. Physical retail distribution of recorded music .......................................................... 198

7.3.2.5. Pay TV services........................................................................................................ 198

7.3.2.6. TV and film production and distribution ................................................................. 199

7.3.2.7. Games and interactive entertainment ....................................................................... 200

7.3.2.8. Ticketing services..................................................................................................... 202

COMMISSION DECISION

of 21/09/2012

(Only the English text is authentic)

THE EUROPEAN COMMISSION,

Having regard to the Treaty on the Functioning of the European Union,

Having regard to the Agreement on the European Economic Area, and in particular Article 57 thereof,

Having regard to Council Regulation (EC) No 139/2004 of 20 January 2004 on the control of 1 concentrations between undertakings, and in particular Article 8 thereof,

Having regard to the Commission's decision of 23 March 2012 to initiate proceedings in this case,

Having given the undertakings concerned the opportunity to make known their views on the objections raised by the Commission,

2 Having regard to the opinion of the Advisory Committee on Concentrations,

3 Having regard to the final report of the Hearing Officer in this case ,

Whereas:

1. THE NOTIFICATION

(1) On 17 February 2012, the Commission received a notification of a proposed concentration pursuant to Article 4 of Council Regulation (EC) No 139/2004 of 20 January 2004 on the control of concentrations between undertakings (the

1 OJ L 24, 29.1.2004, p. 1. With effect from 1 December 2009, the Treaty on the Functioning of the European Union ("TFEU") has introduced certain changes, such as the replacement of "Community" by "Union" and "common market" by "internal market". The terminology of the TFEU will be used throughout this decision. 2 OJ C ...,...200. , p.... 3 OJ C ...,...200. , p....

4 "Merger Regulation"), by which the undertaking Universal Music Holdings Limited ("UMHL") acquires, by way of purchase of shares, control within the meaning of Article 3(1)(b) of the Merger Regulation of the recorded music assets of EMI Group Global Limited ("EMI Group").

2. THE PARTIES

(2) UMHL is a wholly-owned subsidiary of Universal International Music B.V., which is the parent company of the Universal Music Group ("Universal") (the "Notifying Party"), the world's leading music recording company. As such, Universal is active both in discovering, developing and promoting recording artists, as well as the recording of their music (so-called "A&R", which stands for "artists and repertoire") and in the wholesale of recorded music. It also has activities in other fields, such as online music retail, music publishing, artist management, merchandising, event management, and event venue services.

(3) Vivendi S.A. ("Vivendi") is Universal's ultimate parent company. Vivendi is an international media company whose activities include telecommunications, the creation and distribution of content and TV channels, digital music retail and videogames.

(4) EMI Group is active in discovering, developing and promoting recording artists and in the wholesale of recorded music. It also has activities in other fields, such as music retail, music publishing, artist management and merchandising.

3. THE PROPOSED CONCENTRATION

(5) On 11 November 2011, Vivendi and UMHL, on the one hand, and EMI Group, on the other hand, entered into a Share Purchase Agreement pursuant to which UMHL would acquire EMI Group's activities in A&R and in the wholesale of recorded music, as well as EMI Group's retail recorded music activities, certain limited publishing rights (mainly those held by EMI Christian Music Group) and artist management and merchandising activities. For the purpose of this decision the assets subject to this proposed concentration are together referred to as "EMI".

(6) As a result of the proposed concentration, EMI would be solely controlled by Universal. The operation therefore constitutes a concentration within the meaning of Article 3(1)(b) of the Merger Regulation.

4 Council Regulation (EC) No 139/2004 of 20 January 2004 on the control of concentrations between undertakings, OJ L 24, 29.01.2004, p. 1.

4. UNION DIMENSION

(7) The undertakings concerned have a combined aggregate worldwide turnover of ∗ 5 more than EUR 5 000 million (Universal: EUR […]; EMI: EUR […]*) . Each of them has a Union-wide turnover in excess of EUR 250 million (Universal: EUR[…]*; EMI: EUR[…]*). They do not achieve more than two-thirds of their aggregate Union-wide turnover within one and the same Member State. The proposed concentration therefore has a Union dimension.

5. PROCEDURE

(8) Based on the results of the first phase market investigation, the Commission raised serious doubts as to the compatibility of the proposed concentration with the internal market. On 23 March 2012, the Commission adopted a decision to initiate proceedings pursuant to Article 6(1)c of the Merger Regulation (the "Article 6(1)(c) decision").

(9) Following a request by the Notifying Party, non-confidential versions of certain key submissions of third parties collected during the first phase investigation were provided to the Notifying Party on a rolling basis from 23 March 2012.

(10) On 29 March 2012, at the request of the Notifying Party, the time limit for taking a final decision in this case was extended by 20 working days.

(11) The Notifying Party submitted its written comments to the Article 6(1)(c) decision on 13 April 2012 (the "Comments to the Article 6(1)(c) decision").

(12) On 19 June 2012 the Commission issued a Statement of Objections (the "Statement of Objections") to the Notifying Party. The Notifying Party responded to the Statement of Objections on 6 July 2012.

(13) On 25 July 2012, the Commission issued a Letter of Facts (the "Letter of Facts") to the Notifying Party, in which it presented further evidence collected since the adoption of the Statement of Objections supporting the preliminary conclusions reached by the Commission in the Statement of Objections. The Notifying Party responded on 1 August 2012, essentially stating that it considered the Letter of Facts to modify the theory of harm on which the Commission based its concerns as articulated in the Statement of Objections rather than simply presenting new evidence. The Commission replied on 3 August 2012 confirming that the evidence contained in the Letter of Facts did not alter the theory of harm but simply corroborated the Commission's preliminary findings made in the Statement of Objections.

∗ Parts of this text have been edited to ensure that confidential information is not disclosed; those parts are enclosed in square brackets and marked with an asterisk. 5 Turnover calculated in accordance with Article 5(1) of the Merger Regulation and the Commission Consolidated Jurisdictional Notice under Council Regulation (EC) No 139/2004 on the control of concentrations between undertakings, (OJ C 95, 16.04.2008, p. 1).

(14) In order to address the competition concerns identified in the Statement of Objections the Notifying Party submitted commitments on 27 July 2012. The time limit for the adoption of a decision pursuant to Article 8 of the Merger Regulation was therefore extended by 15 working days pursuant to Article 10(3) of that Regulation. The Notifying Party offered a new set of commitments on 13 August 2012, which were subsequently amended on 25 August 2012.

6. MARKET DEFINITION

6.1. Introduction to the recorded music industry

6.1.1. Background

(15) The industry sector on which the possible impact on competition should be assessed is the recorded music sector, as this comprises the core of the activities of each of Universal and EMI.

(16) The total size in terms of sales of the EEA recorded music sector was around EUR 3 613 million in 2010. Out of these total sales, EUR 2 913 million were sales of physical music and EUR 700 million were sales of digital music. Therefore, sales of physical music still account for around 80% of total music sales in the EEA. Five Member States (the United Kingdom, Germany, France, Italy, and the Netherlands) account together for around 70% of total music sales in the EEA. There are, however, significant differences in music consumption patterns across different Member States. For example, in Germany digital music accounts for 13% of the total music sales, while in Sweden it represents 28%.

(17) The recorded music industry has experienced a significant decrease in music sales in the last decade both globally and in the EEA. By way of illustration, over the period from 2001 to 2010, total music revenues in the EEA have declined by 51%. The key drivers of this decline appear to be the reduction in physical music sales, which has not yet been compensated by the growth in digital music sales and piracy, that is to say the widespread consumption of music from "illegal" sources, or in a manner infringing copyright law, whether in physical form (unlicensed copies of CDs or DVDs available for free or at a discounted price) or, even more so, in digital form (unlicensed tracks, albums or even full discographies available for free download or streaming over the Internet).

(18) Sales of digital music are expected to significantly increase over the next years and to reach a point where they will exceed sales of physical music. Based on 6 data from the International Federation of the Phonographic Industry ("IFPI"), the decline of the global recorded music trade revenue is slowing down, with a 2.9% decrease from 2010 to 2011. In some countries, such as the United States and Germany, recorded music trade revenues are stabilising and in other countries,

6 IFPI's report "Recording Industry in Numbers - The recorded music market in 2011". IFPI represents the music recording industry worldwide, with a membership comprising some 1400 record companies in 66 countries and affiliated industry associations in 45 countries. According to its website, its mission is to promote the value of recorded music, safeguard the rights of record producers and expand the commercial uses of recorded music in all markets where its members operate.

such as Australia, Canada, Brazil and South Korea, the growth of digital music already outweighs the decline of physical music. In the Union, digital music revenues grew by 21.6% in 2011. Most major EEA markets, including France, Germany, Italy, Netherlands and the United Kingdom experienced double-digit sales increases. In the United States, digital sales have overtaken physical sales in the United States market in 2011 and account for 52% of the total market of record music which amounts to approximately USD 6 thousand million. Conversely, physical music sales are expected to continue to decrease (albeit at different paces depending on the Member State and, possibly, on the music genre).

7 (19) Several Universal internal documents [...]* .

(20) Most recently on 23 January 2012, at the presentation of the Digital Music Report 2012, Rob Wells, President of Global Digital Business at Universal, stated that: "[t]he future is looking extremely bright. Has the industry turned a corner? I'm definitely more positive now than I've ever been 3 I think that 2013 8 is probably a safe bet".

(21) In any event, as is clear from Table 1 and

7 3]*. 8 Available at: http://www.smh.com.au/digital-life/mp3s/music-sales-fall-again-in-2011-but-optimism- grows-20120124-1qekc.html (last accessed on 1 June 2012).

Table 2, [...]*.

Table 1 6 Universal's recorded music yearly earnings before interest, taxes, depreciation, and 9 amortization ("EBITDA") margins (EMEAand global) (2001-2011)

EMEA

Global

% of Total

EUR million

Net Sales

[...]*

2001

[...]*

2002

[...]*

2003

[...]*

2004

[...]*

2005

[...]*

2006

[...]*

2007

[...]*

2008

[...]*

2009

[...]*

2010

[...]*

2011

Source: Universal

9 Europe, the Middle East and Africa.

Table 2– EMI's recorded music yearly EBITDA margins (EEA and global) (April 2001-March 2011)

EEA

Global

% of Total

EUR million

Net Sales

[...]*

FY 00-01

[...]*

[…]*

FY 01-02

[…]*

FY 02-03

[…]*

FY 03-04

[…]*

FY 04-05

[…]*

FY 05-06

[…]*

FY 06-07

[…]*

FY 07-08

[…]*

FY 08-09

[…]*

FY 09-10

[…]*

FY 10-11

Source: EMI

6.1.2. The recorded music sector

6.1.2.1. Major record companies and independent record companies

(22) The recorded music sector in the EEA is currently characterised by the presence of four worldwide record companies (Universal, Sony Music ("Sony"), EMI and Warner Music Group ("Warner")), the so-called "majors" or ''major record company(ies)'', and a large number of significantly smaller record companies, the so-called "independents" or "indies".

(23) The four majors are characterized, among other things by the following:

(a)a worldwide presence and therefore a world-wide representation for international artists;

(b)a presence not only in the upstream "A&R" and in the downstream wholesale of recorded music, but also, to some extent, at retail level, as well as in the provision of so-called "ancillary services", such as artist management, merchandising and live shows and event management services, and in music publishing;

(c)significant financial strength, making the major record companies less vulnerable to risks and enabling them to offer artists more attractive financial benefits; and

(d)the large, diversified artist roster and repertoire, both in terms of new releases (i.e., songs that are less than 18 months old based on their release date) and of back catalogue (songs that are more than 18 months old based on their release date).

(24) Conversely, the independents consist of a large number of smaller players on the market that are typically characterised by the following:

(a)a much smaller organisation;

(b)a focus on A&R and the wholesale of recorded music;

(c)operations taking place mainly on a national scale, although a limited number of successful independents also have international operations which, however, only cover a limited number of Member States (for international representation of their artists, they generally need to conclude licensing agreements with the majors or other independents; for international distribution they often depend on the distribution networks of the majors);

(d)a much more limited budget for promotional and marketing expenditure, which makes them less attractive for international artists;

(e)a frequent focus on a particular genre;

(f)a limited access to mass media, in particular radio and television; and

(g)a dependency on new releases, more than on a large back catalogue.

6.1.2.2. The recorded music value chain

(25) A&R. The value chain in the recorded music sector begins with A&R, i.e., the discovery, development and promotion of artists, as well as the recording of their music, with a view to exploiting the copyright in the resulting sound recordings. Success or failure in A&R is key to success in the markets for recorded music.

(26) In more detail, A&R encompasses the following:

(a)talent scouting: the traditional manner is for A&R scouts to attend small-scale music venues, listen to demonstrations and meet artists’ managers in order to discover new artists (although these types of scouting activities are decreasing in today's increasingly digital world);

(b)identifying artists who have developed a following on the Internet or social media sites, such as YouTube, Facebook, MySpace and Twitter, but who have not yet signed with a record company;

(c)seeking to persuade established artists who are already signed to other record companies to switch to a rival record company once their recording contract expires;

(d)negotiating artist contracts;

(e)working with the artist to find producers and providing artistic and creative guidance to create a master recording; and

(f)promoting the artist.

(27) All four majors and most of the indies are active at this level of the value chain. Their counterparts are artists and their managers.

(28) Competition for artists. Competition between record companies to sign artists is intense. There is, however, a difference in the ability of the different record companies to compete for artists. While both majors and indies are able to compete effectively to discover and sign artists that are not yet famous and are therefore less "contested", only the majors seem to have the ability to effectively compete to sign the most "contested" artists, including both new "hot" artists and established artists and superstars .

(29) When choosing to sign with a record company an artist takes into account a wide range of considerations, including, obviously, the financial component of the offer, but also other factors, including the label's ability to secure access to retailers and to adequately promote and advertise the artist, as well as the label's creative vision, track record and reputation, existing roster and catalogue, as well as the artist's personal connection with the label's A&R personnel. The fact is, however, that the financial component of the offer (both in terms of advance payments and of expected future revenues), as well as the related aspect of the label's ability to "break" artists (and its track record in this respect), through the label's ability to secure access to retailers and medias, constitute very important factors for any artist's decision to sign with a certain label .

(30) Agreements with artists. When signing to a record company, an artist will typically agree to record exclusively with that record company for a period of time or deliver a certain number of albums. Pursuant to such agreements, the record company would either obtain the ownership of (by assignment), or an exclusive license for the exploitation of the recorded music master rights over the artist's recordings. In exchange for this commitment artists generally receive an advance on royalties at the start of each recording contract, the royalties normally varying in accordance with format (albums or singles), price category and expected commercial success of the artist. An artist agreement will invariably consist of one or more “firm” albums (that is to say albums the record company

is contractually committed to release) and a number of subsequent "options" (that is to say albums the record company can choose to release). The precise number of firm and option albums will be negotiated on a case-by-case basis.

(31) Options are critical to the business model of any record company because of the risk and cost involved in signing artists. On signing an artist, the record company will typically pay an advance to the artist, cover the recording costs for an album (likely to include hiring a recording studio and musicians, paying an advance to a producer, mixing the tracks on the albums), make videos for one or more tracks, manufacture, distribute and market the album. This can cost anywhere between a few thousand and millions of Euros.

(32) Since the industry norm is that five to nine out of any ten albums typically lose money, in order to sustain this level of investment, a record company needs to know that, for the few artists which become successful, it has the rights (but not the obligation, if not successful) to release a number of future albums on pre-agreed terms. The option mechanism is the tool that is used to achieve this goal and the guaranteed revenue flow of such future albums provides the basis for the initial investment to be made.

(33) Generally, a record company will have a period of time (normally a few months) from releasing one album to decide whether to exercise its option for the next. If it exercises its option, it will usually pay the advance for the next album, and again commits to pay the recording and other costs involved. If it does not exercise the option, then the artist is freed from any future delivery commitments and is free to sign to another record company.

(34) Record companies also increasingly seek to secure a number of ancillary rights over an artist in addition to a recording agreement, for example in relation to merchandising, music publishing, endorsements, live music, social media or an artist’s website. These rights can be "active", when they are actually actively exploited by the record company or one of its affiliates, or "passive", when the record company does not directly exploit these rights, but receives a share of the relevant revenues regardless of how the artist decides to exploit these rights. The nature and extent of such rights to be acquired, and whether the rights are active or passive, varies by artist and territory.

(35) In practice negotiation of ancillary rights is simply one element of the negotiation between artists and record companies and record companies are often unable to secure the full or a significant part of such rights, that is to say, record companies are often unable to secure so-called "360 degree" deals with artists. […]*

(36) Development and promotion of artists. After signing a contract with an artist, a record company’s A&R team will work closely with the artist to develop his/her repertoire and brand. While the level of expenditure is likely to differ between record companies and countries, recorded music is a sector with significant levels of A&R expenditure (the industry’s equivalent of Research & Development expenditure) in which the successful artists have to pay for the unsuccessful

artists. According to the recent IFPI report "Investing in Music" from March 2010 : "[m]usic companies internationally invest around US$5 billion annually in developing and marketing artists – around 30 per cent of their turnover. This is an exceptional level of spending in comparison to the proportionate research and development spending of other industries. Commercial success is expensive: it can take US$1 million in investment to break a new act in major markets, with marketing and promotion needed across a range of channels" . The same report further explains that the above 30% figure comprises around 16% of A&R discovery expenditure, essentially consisting of advances to artists, and around 13% of marketing expenditure.

(37) Access to sales, advertising and marketing outlets, including, but not limited to, access to prominent shelf space at physical retailers, adequate placement on the web pages of digital music providers, inclusion in promotional campaigns of physical and digital customers, editorial coverage by retailers and medias, radio airplay, participation at TV shows and traditional paid-for advertising are all key (together with the underlying quality of the sound recording) to ensuring the success of an artist.

(38) According to David Guetta, one of the most successful artists currently in EMI's roster: "[t]o reach the public, music needs to be known; it needs to be available to buy and it needs to be talked about".Similarly, according to The Cardigans and A Camp lead singer Nina Persson, one of the major reasons for signing and remaining with a label is their marketing and promotional muscle: “[t]hey have all the resources and contacts needed to get my albums out and tell people to go and get it. They have people there who understand how to present me and my music in a way that’s right, and since I don’t have their infrastructure I’m happy to be helped with these things” . The same message is confirmed by a representative of Universal in Sweden, according to whom: "[m]any think they can manage without marketing, but nothing could be more wrong. With the vast selection of music you have today, you must have a strategy to succeed, you must invest time and money and you must be daring” .

(39) While, at least in theory, all these sales / marketing / promotional / advertising channels are equally available to all record companies, in practice larger record companies, that is to say the majors and, among the majors, those, such as Universal, who have the largest artist roster and control the largest repertoire, are able to secure better "access" for their artists to many of these channels. This, in turn, increases the appeal of larger record companies in relation to artists, both new artists and established ones.

(40) Self-releasing. Finally, it should be noted that, in addition to signing with a record label to launch one’s career, artists today also have, at least in theory, the option to launch themselves. The rapid and substantial digitisation of social networks and recorded music distribution channels over the past decade has created a wide range of low-cost opportunities for artists to release, promote and distribute their recorded music content.

(41) Those authors that are able to do so can use social networking sites (such as Facebook Twitter, MySpace YouTube and Google+) or other types of self-release distribution services. So-called "Artist aggregator" services, such as Google Music Artist Rooms (currently available only in the United States), cdbaby and TuneCore, not only distribute music directly, but also act as onward distributors to more mainstream online retailers. These services typically act on behalf of a significant number of artists or small labels and usually charge either an upfront fee and/or a percentage of earnings made through digital or physical distribution services.

(42) The services provided by “artist aggregators” to their clients are their ability to encode recorded music files into the correct format demanded by mainstream online retailers; ensure that digital releases are timed to coincide with physical releases; market their clients’ music to online retailers; and consolidate royalty information across different online stores, as broken down by track and between à-la-carte and subscription models.

(43) Despite the artists' increased ability to self-release, record companies continue to be key to the discovery, development and promotion of artists. According to Martin Mills, chairman of the Beggars Group, one of the largest independent record companies, "[y]ou can't just step out of MySpace and talk to retailers or concert promoters". In its response submitted during the market investigation, another […]* record company, […]* underlines that "there is not one single example of success of an artist being able to release his own record and break it. Even the huge superstars, who theoretically would be able to do it alone, all have a record company behind them. They need their full range of services (A&R, radio, Promo, marketing, digital, commercial, syncs...) Even the so called "self-release" by Radiohead has been done via Naive/Beggars for physical and Warner for Digital ... and is far from being a successful release". Accordingly, the Independent music companies association ("IMPALA")highlights that "a record company does not just make music; its activities are far wider including, for example, seeking out talent and the full A&R function, signing artists (and being asked to pay increasing advances), making professional recordings, manufacturing CDs or having them manufactured, distributing music in multiple

territories, negotiating with digital companies, many of which are global operators, making sure that artists are heard by music critics, are played on the radio and appear on TV" .

(44) The IFPI report referred to in recital 36 clearly explains the key role that record companies still today play in A&R as follows: "[o]ne of the biggest myths about the digital age is that artists no longer need record labels. The internet allows them to reach their public directly, the myth goes. Live music and other revenue sources, like merchandising and advertising, will do the rest. Yet the reality is in fact completely different. A very small minority of artists, mostly well known, established acts, are achieving success through this DIY route – they deserve good luck. But the vast majority are not. The truth is that artists are generally much better served by a record deal. They want the funding and the specialist support that indie and major record labels provide. Put another way, whilst the direct route afforded by the internet is open to all, mixing the talents of business and creativity is often a minefield, with creativity often compromised by the challenges of running a business, which requires totally different skills. Artists generally prefer to leave the complex administration of a rights based business to someone else. A few years into the digital revolution, it has now become clear that the internet is by itself no guaranteed route to commercial success. MySpace has more than 2.5 million registered hip hop acts, 1.8 million rock acts, 720,000 pop acts, and 470,000 punk acts. The gulf between acclaim and anonymity, where record labels do their essential work, has never been greater than today. Investment in music covers many diverse areas. There is the belief and the support in the artist; the expertise to plot a road to success; and the money, provided upfront, and always at great risk, that is needed to allow an artist to create their work. There is the choosing of the songs and the recording work in the studio. Artwork is needed and videos may be needed. There are manufacturing and distribution costs, both for online and CDs. Marketing teams campaign, in one market or several. Hundreds of digital channels and outlets are created and serviced. All this takes money, time, people, knowledge and skills. These are some of the essential qualities of record companies as long term investors in music" . The report continues by explaining that: "[r]ecord companies do their business in a virtuous circle of investment by which successful collaboration with existing artists generates the revenues to invest in the artists of the future. This is what enables companies to sustain large active rosters of artists and provide advances to a wide community of artists, many of whom will not be commercially successful" .

(45) Wholesale of recorded music. The next level of the recorded music value chain comprises the sale or licensing by record companies of their repertoire to wholesalers and retailers.

(46) From the point of view of a record company, the exploitation of the recorded music produced by the artists in all formats (CDs, DVDs, download, streaming) and across all product types (singles, new albums, re-releases, compilations) is inextricably linked. This reflects the fact that, given the record companies' very high A&R expenditure, record companies only have an incentive to make these investments if they also have the ability to recoup these investments (at least for the successful artists) by exploiting the relevant artists' sound recordings across all formats and distribution channels, as well as on a global basis.

(47) Physical and digital products. Recorded music includes a wide variety of physical and digital products.

(48) Physical products mainly comprise audio CDs and DVDs (including audio and video). CDs, in turn, could include only one or few songs from the same artist (singles) - although these products are progressively disappearing - or more songs (albums). Albums, in turn, may include songs from only one artist or from several artists (so-called "multi-artist" compilations). There also exist "single-artist" compilation albums, typically consisting of a single artist's "greatest hits" or any other collection of different songs previously included in separate albums and then re-packaged and re-released as a new album.

(49) Most multi-artist compilations (and some single-artist compilations, when the recording rights for the different songs of the relevant artist are controlled by different record companies) consist of tracks from a number of different record companies. The company issuing the compilation will therefore license some repertoire from third party labels and use other repertoire from its own artists. The terms of the licence from third party labels are subject to individual negotiations. Licensing repertoire to compilations helps record companies in recovering the cost of investment in the initial recording and in introducing an artist to a wider audience. However, many artist recording agreements require artist approval for inclusion in a compilation .

(50) Digital products also include single tracks and albums (including compilations). Record companies sell these products in digital form (that is to say electronically, without any physical support) to digital retailers, such as Apple/iTunes and Amazon, which, in turn, resell them to end users.

(51) In addition, record companies can also license their repertoire, that is to say essentially all or the vast majority of the songs for which they control the recording rights, to audio or video streaming digital service providers, such as Spotify or YouTube, which, as opposed to download service providers, would not "sell" these tracks or albums to end users, but would rather grant end users access to the various songs or videos subject to certain restrictions and/or against the payment of a fee.

(52) Distribution deals. Record companies are typically not active in the manufacturing and distribution of CDs and DVDs, which are generally manufactured and distributed by third party sub-contractors. Moreover, record companies can either distribute and market their physical and digital repertoire themselves (or through sub-contractors as far as the actual distribution is concerned) or enter into so-called distribution agreements with other record companies. While majors typically directly distribute their repertoire in all countries where they operate (with the exception of […]* for some smaller countries), indies often rely on one or more of the majors to distribute their physical products in a number of countries.

(53) As noted by the Commission in Sony/BMG, there are a number of different categories of these distribution arrangements, namely “pure distribution deals”, under which the distributor is typically only responsible for physical distribution and the principal retains control over list prices, discounts, marketing, publicity and strategy, and under which the distributor retains a fee and returns the balance to the principal; full “licence deals”, under which the licensee manufactures, distributes, sells, markets and promotes on behalf of the principal; and "sales and distribution deals", where the distributor may have some degree of influence over list prices and discounts, assumes sales functions, and usually collects sales revenues before remitting the proceeds, less a sale and distribution fee, to the principal.

(54) As far as both Universal and EMI are concerned […]*.

(55) Distribution agreements typically have a three-year duration and grant the distributor the exclusive rights to (manufacture and) distribute the principal's physical or digital repertoire in one or more territories. These rights may also include the right to license the principal's master recordings to other third parties, for example for synchronisation use in film, television, video games, and for use in soundtracks, compilations and other non-record exploitation.

(56) Physical retail channels. Record companies promote, market and sell physical products to wholesalers and different categories of retailers. The record companies' counterparts on the demand side include different categories of retailers, of which the main types are hypermarkets/supermarkets ("mass merchants"), e-tailers and specialist retailers. There are also wholesalers of music, which act as intermediaries between record companies and retailers.

(57) Digital retail channels. The principal means of online music dissemination are through downloading and streaming. The same repertoires are typically licensed for both forms of online music dissemination.

(58) Downloading involves the purchase and storage of a digital copy of a musical work on one or more computers or electronic devices. Prototypical examples of download platforms are Apple's iTunes service and Amazon's MP3 service.

(59) With a streaming service, the user does not download music files and no permanent copy is stored on the user's computer or electronic device. Instead, the audio file is delivered in small data packets over the Internet and playback commences as soon as the Internet streaming is initiated. Streaming services can comprise a basic, advertisement-sponsored service that is available free of charge and a premium, paid-for service. Premium services are typically advertisement-free, offer additional functionality (for instance unlimited plays of songs, a larger music library or support of mobile devices) and may grant users the possibility to download and store tracks in order to listen to them offline during the period of the subscription (so-called "tethered downloads") . There is also a difference between so-called "lean-forward services", which the user accesses with a view to listening to one or more specific tracks of its choice, and so-called "lean-back services", which are more similar to Internet radios, and which the user accesses with a view to listening to tracks belonging to one or more music genre(s) of its choice, without, however, selecting the specific tracks to be played (which are chosen by the service).

(60) Important examples of streaming services available in the EEA include Spotify and Deezer, as well as Google's YouTube, which is an important video streaming platform funded by advertising.

(61) Streaming technology has also allowed for new platforms and business models to develop over the last few years. Internet Service Providers ("ISPs") and Mobile Network Operators ("MNOs") increasingly offer music streaming services, either by developing their own branded services (often bundled with telecom subscriptions) or via partnerships with existing streaming platforms. Some streaming services have also struck partnerships with social networking sites (for instance, Spotify with Facebook).

(62) Digital platforms are attempting to bring to the market other innovative ways to access and listen to musical content.

(63) For instance, cloud music services that offer users the ability to store their acquired music on remote servers have emerged. An example of this type of service is Apple's iTunes Match, which was launched in the EEA at the end of 2011. iTunes Match scans a user's music library, matches those songs with the iTunes music catalogue hosted on Apple's servers and allows users to download or stream a song from a customer's "cloud locker" to the device.

(64) There have also been attempts to provide combined ‘hybrid’ services which offer subscribers unlimited streaming under an integrated tariff and a certain number of permanent downloads per month (for example a user’s five favourite tracks). Examples include Sky Songs and Nokia's Comes with Music service. While services of this type are currently nascent and evolving, these types of hybrid services may play a larger role in the future as the digital music segment matures.

(65) Downloads currently continue to account for the large majority of online revenues. Streaming services are, however, increasing and many market

26 After termination of the subscription, the tracks are no longer available.

participants expect streaming revenues to grow significantly in the future. Within three years of its launch, the streaming service Spotify has grown into the second largest online music platform after iTunes. It is now the largest digital retailer in Norway, Finland and Sweden and, having launched in Austria, Belgium, Denmark and Switzerland in autumn 2011 it is now available in 11 EEA countries.

(66) Finally, TV broadcasters (for example Virgin Media in the United Kingdom or MTV across the EEA) also offer on-demand and online audio and video music services. For these purposes, they also negotiate digital licences with record companies, which, depending on the service offered, may be akin to a download or a streaming licence.

(67) Pricing and other key elements in commercial negotiations. The starting point for price negotiations between record companies and physical retailers is typically the published price to dealers (“PPD”), that is to say the gross wholesale unit price of any given CD or DVD. However, the net wholesale price paid by customers, and which may significantly vary from customer to customer, is the result of a variety of discounts applied to the PPD, which are agreed bilaterally and confidentially between the record company and each physical customer.

(68) Universal and EMI, place each of their physical products into a PPD category, as do most record companies. The number of PPD categories is not fixed and the determination of a PPD category for a given release will depend on a number of factors, including the type of product (for example CD, double CD, number of tracks) and the price band (full price, mid-price, low price). The price band will in turn depend on a number of factors, including the nature of the artist (for example superstar or newly discovered) and the expected commercial success. The PPD of a product is not fixed but changes (often frequently) during the life cycle of the release, in particular as it increases or decreases in popularity.

(69) The same PPD category for a given product will apply across all retailers in a given territory. Net wholesale prices are then negotiated on a confidential basis with each customer. Consistently with the commercial practice of retailers in relation to other physical non-music products, retail customers and record companies will usually agree on a broad variety of discounts. These discounts may include for example the following:

(a)file discounts, negotiated annually and affecting all albums sold to a particular customer;

(b)campaign discounts, offered for particular promotional campaigns, tied to a particular album or albums, and applicable for a short period of time;

(c)volume discounts, occasionally provided on pre-order for particular new releases in order secure high retail volumes, and usually playing a similar role to campaign discounts;

(d)retrospective discounts, paid at the end of the year and either fixed or negotiated on the basis of the turnover generated by that customer during the preceding year;

(e)co-operative marketing payments, negotiated in relation to sales of a particular album or albums, whereby the record company provides the products and the promotional funds and the retailers provide the promotional programmes;

(f)free goods, provided typically after the purchase of a large volume of albums;

(g)returns, agreed on an ad hoc basis or annually, whereby a record company agrees to take back certain unsold CDs free of charge;

(h)early payment discounts;

(i)range discounts, provided in return for the customer stocking a wider range of products; and

(j)retail space discounts, provided in return for the customer devoting a certain amount of space to that particular record company's products.

(70) Other key commercial terms of these agreements include, for example, guarantees of shelf space or promotional space.

(71) In relation to the pricing of digital recorded music, download pricing (for customers such as Apple/iTunes and Amazon MP3) must be distinguished from pricing for streaming services customers (such as Spotify).

(72) For digital downloads, the PPD is typically set out (as in the case of physical music) in rate cards, which determine the starting point for negotiations with digital customers. The rate cards set out the prices for the different album and track pricing tiers, as well as the varying other digital products such as ringtones, ring backs and video downloads. Actual wholesale prices will then be negotiated between the record company and the customer on the basis of the PPDs. The resulting contract will typically contain wholesale prices or price caps that differ by customer and differ for a given customer by the same sort of categories and price bands outlined in relation to physical music above, but also certain digital music specific categories, such as whether a download is digital rights 27 management ("DRM") free .

(73) Discounts are also agreed based on obligations for digital platforms in relation to sales reporting towards record companies, the achievement of sales targets and advertisement and marketing of record companies' artists' repertoire.

(74) Other key commercial terms of these agreements include advance payments, minimum guaranteed revenue, advertising commitments and most favoured nation ("MFN") clauses (see Section (241)).

27 DRM is a class of access control technologies that are used by hardware manufacturers, publishers, and copyright holders with the intent to limit the use of digital content and devices after sale. A download is DRM-free when the end user, who purchases it is not subject to any restriction on the use of the track or the album (for example the end user can listen to the music on any device and not only on the device to which the music was originally downloaded).

(75) For streaming services, the record company will typically negotiate a blanket deal setting the terms for all content provided. The record company’s starting point on price will be considered by the amount of free usage permitted for the service in order to solicit subscribers, marketing packages on offer, and the mix of minimum guarantees and advances that the digital streaming customer will provide. The fee payable by the customer may depend on one or more of (defined for example by the greater of several amounts) a variety of parameters and most notably, a share of the net revenue generated by the record company’s pro rata share of music streamed; a minimum fee per stream or per subscriber; a multiple of the number of subscribers to the service pro rata to the record company’s share of music streamed; a multiple of the number of streams accessed pro rata to the record company’s share of music streamed; and a pro rata share of advertising revenue generated by the service.

(76) Record companies may offer discounts to streaming customers, typically based on reaching certain turnover thresholds, as well as advance payments, guaranteed revenue, advertising commitments and MFN clauses.

(77) Retailing of recorded music. Mirroring the structure of the recorded music market at wholesale level, the retail supply of recorded music covers the sale of recorded music in physical and digital formats to end customers.

(78) The main categories of retail providers of both physical and digital music and the different services offered by each such provider were already discussed in Recitals 56 to 66.

(79) Licensing of music to end customers. There are certain specific categories of end customers, namely TV broadcasters when offering linear services, radio broadcasters and other businesses such as bars, nightclubs, shops, hotels and gyms, which do not acquire recorded music from record companies, but from producers’ music licensing companies.

(80) In order to broadcast a given song, these customers typically need to take a licence of the relevant music publishing performance right and recorded music right. In relation to recorded music rights, there are producers’ music licensing companies engaged in the collective licensing of recorded music to end users such as broadcasters. For example, in the United Kingdom, Phonographic Performance Limited (“PPL”) licenses audio recorded music rights on behalf of approximately 5 000 record companies, as well as approximately 45 000 performers. PPL's sister company, Video Performance Limited (“VPL”), licenses music video rights. Almost all United Kingdom record companies are members of PPL. Such producers’ music licensing companies grant licences directly to customers, such as the BBC, BSkyB, Channel 4, ITV and Five.

(81) Compared to retail consumers, the terms and conditions of licensing recorded music to broadcasters are different. Record companies normally license their recorded music to retail customers directly by means of bilateral negotiations. In the case of TV and radio stations and other broadcasters, pricing will in most cases be negotiated by the producers’ and performers’ music licensing companies in the form of royalties contained in different types of blanket licence. For example, in the United Kingdom, PPL operates on the basis of around 60

different royalty tariffs, which are negotiated with broadcasters or which have been established by means of a reference to the Copyright Tribunal. The royalties for a recorded music licence will usually reflect a proportion of the revenues (usually advertising) generated by the licensee during the licence period and vary depending on the importance of music in generating those revenues. In the case of public broadcasters, alternative commercial terms (not based on advertising revenues) are negotiated collectively. For example, all other things being equal, the royalty rate is likely to be higher for a music-based radio station than a talk-based radio station. The producers’ music licensing companies then remit the royalties (less their administrative expenses) to the record companies and performers, whose rights are being licensed.

6.1.3. Other activities related to recorded music

(82) Universal and/or EMI are also active in a number of other areas related to recorded music, including the provision of ancillary services.

6.1.3.1. Ancillary services

(83) Record music companies are often active, to a greater or lesser degree, in providing one or more of the following services, which are ancillary or complementary to their recorded music activities.

(84) Manufacturing and logistics (physical music). These activities include the production and dispatching of CDs and DVDs. As explained, record companies do not typically engage in these activities directly, but rather sub-contract them to third party manufacturers and logistic service providers.

(85) Artist management. Artist management services comprise a range of activities designed to assist musicians in developing and promoting their work. These services include arranging recording facilities, booking live venues, liaising with publicists and the media, and dealing with finances. The activity levels and income of artist management companies can fluctuate significantly, reflecting the activity levels of their artists and in particular their live events or touring commitments.

(86) The services provided by artist management services are typically based on strong personal relationships and there are few corporate players in this field: rather, artists tend to work with particular individuals and when these move from one firm to another they tend to move with the manager. Artist management is mostly performed by individuals that represent either one or a few acts.

(87) Merchandising. Merchandising comprises the licensed production, supply and sale of branded products which bear the name, logos and trademark of right holders. In its simplest form, merchandising is the exploitation of a name or mark, usually a copyright or trademark, although the subject matter of the right granted may not necessarily be covered by a trade mark or any other intellectual property rights (''IPRs''). Merchandising in the entertainment, sports and event-related sectors can be described as the production, distribution and sale of products using the name, likeness or image of a character, sportsman or team, film, band, individual artist or celebrity. Merchandising covers a wide range of sectors including film, animation, television, sports and music.

(88) Merchandisers compete for the licensing of rights across these sectors. The owner of the relevant right will generally ask for bids from several merchandisers. Alternatively, it is common for merchandisers to contact licensors and offer their services (if they consider a successful merchandising opportunity exists). Once the deal is negotiated, the merchandiser will produce with the approval of the licensor a range of products, for example, printed products (such as posters and calendars), T-shirts, branded apparel, toys, wallets, key chains and footwear, to be distributed and sold at events or live performances and through retailer channels including high street, internet, and mail order catalogues.

(89) Event merchandising includes live performances or tours for leading artists. Retail merchandising may involve the exploitation of the name and image rights of performers and artists.

(90) Live show and event management services. Live show and event management services comprise services such as booking venues, organising logistics, arranging catering and security facilities, negotiating event sponsorship deals and liaising with ticketing agents.

6.1.3.2. Music publishing

(91) Universal and, to a limited extent, EMI, are also active in music publishing. A distinction needs to be made between the artists recording the rendition of the song and the authors (writers and composers) that create the song to be recorded. Authors deal with music publishers and their works are subject to music publishing rights. Universal, EMI and other majors are active as record companies, as well as music publishers. Music publishers are active on two market levels. Upstream, they are active in the supply of publishing services to authors. These services include signing authors and providing them with financial, marketing and career support. As a counterpart to these services, authors transfer the rights in their musical work to the publisher or grant that publisher an economic interest in the musical work by providing the publishers the right to obtain a certain portion of the royalties collected. Downstream, music publishers are active in the exploitation of works of authors under contract or for a certain period of time following the expiration of the contract (''retention period''). On that level, they either directly grant licences for use of the musical works to right users against the payment of royalties, or receive a part of the royalties collected by collecting societies (for licences issued by the societies).

(92) The Commission previously identified the following different categories of publishing rights:

(a)Mechanical rights: the right to reproduce a work in a sound recording (for example CDs);

(b)Performance rights: the right for commercial users such as broadcasters (TV or radio stations), concert halls, theatres, night clubs, restaurants to divulge a work to the public;

(c)Synchronisation rights: the right for commercial users such as advertising agencies or film companies to synchronise music with a visual image;

(d)Print rights: the right to reproduce a work in sheet music;

(e)Online rights: a combination of mechanical and performance rights for online applications, such as music downloading and/or streaming services.

(93) To be able to legally sell or stream a given song digital customers require both the recording rights in relation to that song and the copyrights relating to the authors' works. Mechanical and performance licences are required when a sound recording is downloaded; the download services provider first makes a digital copy (mechanical right) and then publishes that copy via the internet to the end consumer (performance right). The dissemination of musical works via online or mobile services through streaming arguably also involves both mechanical and performance rights insofar as streaming relies on the creation of temporary copies (caching) and the transmission of the digital data to the end user. In addition,

many online and some mobile distribution platforms allow end consumers to listen (at least to a limited extent as a "teaser") to tracks for free. Again this involves both mechanical reproduction and performance rights.

(94) These mechanical and performance rights are typically licensed and collected by collecting societies on the basis of standard rates set by each relevant collecting society. However, with respect to mechanical rights for Anglo-American repertoire, publishers have the ability to license these rights outside the collecting society system. In Universal/BMGand its recent decision in Sony/Mubadala/EMI Music Publishing, the Commission confirmed that the publishers could in essence withdraw these rights from the collecting societies, since they are the rights that are regularly assigned to music publishers by the authors. All other rights (performance rights as well as mechanical rights for Continental European repertoire) are normally assigned by the authors to collecting societies. Any withdrawal of those rights by the publishers would normally require approval by the authors. The Commission considered that given that the approval of thousands of authors would be necessary, a withdrawal of offline performance rights and offline mechanical rights for Continental European repertoire is not likely in the foreseeable future.

(95) In light of these factors, the licensing of publishing rights for online use has indeed undergone significant changes in the past decade. Prior to 2005, only national collecting societies issued licences (and determined the rates at which these rights were licensed) for online mechanical and performance rights. Those licences were limited to the home territory of the collecting society. An online music platform wishing to set up a service across several EEA countries therefore had to negotiate with the national collecting societies for each of those countries in addition to negotiating with record companies.

(96) Following the Commission Recommendation of 18 May 2005 on collective cross-border management of copyright and related rights for legitimate online music services(''Commission's 2005 Recommendation on the cross-border collective management of copyright for online use''), which favoured a business model whereby publishers should be free to choose among collecting societies to manage their works on a pan-European basis (so-called ''Option 3 model''), all the major music publishers have reorganised the licensing of the online rights for their Anglo-American repertoire. This reorganisation has taken the form of music publishers appointing individual collecting societies or collective rights management entities owned by one or several collecting societies to administer centrally the licensing of their Anglo-American repertoire for online dissemination. For the practical and legal reasons identified in Recitals 94 and 95, these initiatives have concerned only Anglo-American repertoire.

28 Commission's decision of 22 May 2007 in Case No COMP/M.4404 - Universal/BMG Music Publishing.

29 Commission's decision of 19 April 20012 in Case No COMP M. 6459 - Sony/Mubadala development/EMI Music publishing.

30 OJ L 276, 21.10.2005.

(97) As of today, all major publishers have withdrawn their mechanical online rights for Anglo-American repertoire from collecting societies. Universal has appointed the French collecting society Société des Auteurs, Compositeurs et Editeurs de Musique (''SACEM'') to administer and license its mechanical online rights for Anglo-American repertoire and French repertoire on a pan-European basis. EMI with regard of its publishing activities has appointed CELAS GmbH (''CELAS'') - a joint venture between the UK and German collecting societies, the Mechanical-Copyright Protection Society (''MCPS'') (within Performing Right Society (''PRS'') for Music) and the German collecting society Gesellschaft für musikalische Aufführungs- und mechanische Vervielfältigungsrechte (''GEMA'') – as its rights manager. Sony/ATV Music Publishing and various indies have assigned similar rights to the MCPS (within PRS for Music). Warner/Chappell Music has entered into analogous arrangements for the pan-European licensing of mechanical rights to their Anglo-American repertoire for digital exploitation.

(98) The Commission considers that the publishing rights that are relevant for the competitive assessment in this case are these online rights. This reflects the fact that online rights are the publishing rights which digital customers need (together with recording rights) to be able to legally sell or stream a given song and in relation to which, at least as far as mechanical rights for Anglo-American repertoire are concerned, music publishers have the ability to control rates and other licensing terms, as they can be licensed outside the collecting societies model. As a result, these are the only type of publishing rights which an integrated recorded music and music publishing company, such as the merged entity, could use, together with its presence in the recorded music sector, to exercise market power in relation to customers, and, in particular, digital 31 customers .

6.2. Relevant product markets

6.2.1. A&R

6.2.1.1. The Notifying Party's view

(99) The Notifying Party does not take a view as to the exact scope of the relevant product market in relation to A&R activities.

6.2.1.2. The Commission's assessment

31 For the sake of completeness, the Commission notes that music publishers are also able to license synchronisation rights outside the collecting society system and that customers for such rights (mainly advertising agencies, computer game developers and TV and film production companies) equally need both these rights and recorded music rights to use a given song in their works. As a result, an integrated recorded music and music publishing company may in theory have the ability and the incentive to leverage its combined position on recorded music and also on music publishing in relation to these customers. As the Commission's investigation in this case did not reveal this as a possible competition concern, the Commission does not further discuss synchronisation rights in this decision.

32 (100) In its previous decisions relating to the recorded music sector, the Commission never defined an upstream market for A&R. It is, however, the case that A&R services are provided to different market players (artists) and entail a different organisation within record companies compared to the downstream wholesale activities. There are also certain competitive dynamics, which are specific to A&R.

(101) The market investigation was inconclusive as to whether A&R activities should be viewed as a separate product market from the wholesale of recorded music. Several respondents to the market investigation pointed out that there is a close interrelation between these two activities since record companies typically leverage their strength in the downstream wholesale market to attract new and established artists to their labels (a strong presence at the wholesale level typically allows record companies not only to achieve revenues that can be re-invested in A&R, but also access to key retailers and to promotional opportunities) and a strong artist roster including both new and established artists allows record companies to offer an attractive repertoire to wholesalers and 33 retailers and therefore reinforces their market position downstream .

(102) The close interrelation between A&R and wholesale of recorded music is also confirmed by the fact that the majors’ respective market position is by and large similar for both activities (the Notifying Party submits that a record company's market position in the wholesale of recorded music is a good proxy for its position in relation to A&R activities).

(103) It is not necessary to take a view as to whether A&R activities should be considered and analysed as a separate product market or even whether the recorded music market should be viewed as a two-sided market, where the strength of a record company on one side of the market (A&R) has a positive influence over its market position on the other side of the market (wholesale of recorded music) and conversely. This reflects the fact that, regardless of the position on this point, the competitive assessment of the proposed concentration will not be affected, as the strength of record companies in A&R will be taken into account in the competitive assessment as one of the key factors contributing to their market position in the market for wholesale of recorded music.

6.2.2. Wholesale of recorded music

6.2.2.1. The Notifying Party's view

(104) First, the Notifying Party claims that there should be a single market encompassing sales of recorded music in both physical and digital formats. In this regard, the Notifying Party also notes a growing trend among its traditional physical customers to also provide music in digital format, whether through their own service or through, for example Apple/iTunes Gift Cards. The Notifying Party also claims that, since retail prices for digital music strongly constrain retail

32 Commission's decision of 3 October 2007 in Case No COMP/M.3333 – Sony/BMG; Commission's decision of 15 September 2008 in Case No COMP/M.5272 – Sony/SonyBMG.

33 Competitors' reply to the Commission Request for information of 17 February 2012, question 25.

prices for physical music, price negotiations with its digital customers are 34 influenced by physical music price levels and conversely .

(105) In its Response to the Statement of Objections, the Notifying Party argues that, even if wholesale for digital and for physical recorded music should be considered as separate product markets, the Commission should take into account the significant competitive interactions that exist between the two. In particular, the Notifying Party considers that the Commission has not sufficiently considered in its assessment the competitive constraints imposed by piracy not only on the digital recorded music market, but also on the physical recorded music market given the pricing interaction between these markets. The Notifying Party refers to the following replies from competitors and a customer in support of that view:

–“[t]he increasing popularity and relatively lower prices of digital music has 35 had a very significant impact upon … retail prices of physical music” ;

–“there is a direct correlation between physical and digital pricing. We monitor both and bring digital into line with physical where possible. Equally, we have to be mindful of pricing for digital only retailers as we are in competition 36 across the board" ;

–“the digital market is increasingly moving towards a model for on demand products (whether downloads or streams) where the total wholesale price including publishing royalties parallels traditional wholesale/retail splits in the 37 physical world”; and

–"[a]s regards digital downloads and physical CDs evidence suggests that there 38. is a clear correlation in consumer price"

(106) In its Response to the Statement of Objections, the Notifying Party furthermore claims that the Commission failed to consider the fact that major retail players are active in both physical and digital markets. In this regard, it emphasises that […]* is Universal's […]* EEA physical customer and […]* most important digital customer as well as EMI's […]* largest physical customer and […]* largest digital customer. Moreover, the Notifying Party underlines that Universal's top […]* EEA physical customers all offer digital download services. According to the Notifying Party, Tesco's recent acquisition of the streaming service We7 is illustrative of the trend towards retailers straddling between physical and digital markets.

(107) In addition, the Notifying Party claims that a segmentation of the digital segment between downloading and streaming services is not relevant given that digital

music services can in principle offer both download and streaming services and 39 that there are no significant barriers or costs for such supply-side substitution .

(108) Second, according to the Notifying Party, the market for recorded music should not be further delineated by genre, between albums and singles or between single 40 artist and compilation formats .

(109) As regards the distinction by genre, the Notifying Party claims that genres cannot be meaningfully defined, artists’ works typically cut across a number of different 41 categories, these categories vary by record company and over time.

(110) In its Response to the Statement of Objections, the Notifying Party also refers to a number of statements from respondents to the market investigation, which are contrary to the Commission's assessment in relation to jazz and classical music. Moreover, according to the Notifying Party, the Commission's assessment is based on a number of distinctions, which would be equally applicable to every form and genre of music. In particular, the Notifying Party considers that each genre appeals to a specific customers' group and that all genres present differences in the A&R process and the musical instruments required. In addition, the Notifying Party submits that many independent labels are currently producing orchestral recordings. Finally, the Notifying Party considers that the Commission overlooked the fact that its top customers for jazz and classical music are broadly the same as the customers of its overall repertoire.

(111) As regards the distinction between compilation and single artist albums, the Notifying Party points out that consumers buy both single artist recordings and compilations and that single artist recordings and compilations are offered alongside one another in online music and in physical retailer stores. Moreover, according to the Notifying Party, such a distinction has further diminished, given the availability of single tracks in digital format, which allows consumers to create their own individual "playlists".

(112) In its Response to the Statement of Objections, the Notifying Party rejects the Commission's finding that compilations are subject to different competitive dynamics. In particular, according to the Notifying Party, the Commission's finding that compilations have little relevance in streaming services is misplaced, as illustrated by the existence of a "NOW" application on the Spotify platform, which is used to browse and discover music. Moreover, the Notifying Party refutes the Commission's finding that compilations have different products characteristics and a lower price and refers to the following quotes from respondents, which would contradict that view:

–“Compilations are not provided for in […]* recorded music licensing deals in any different way than any other music product”;

39 Paragraphs 2.9-2.13 of the Comments of the Notifying Party to the Article 6(1)c Decision.

40 Form CO, pp.111 to115.

41 Form CO, p.112.

42 Form CO, pp. 112 and 113.

–“Compilations are not provided for in […]* recorded music licensing deals in any different way than any other music product”;

39 Paragraphs 2.9-2.13 of the Comments of the Notifying Party to the Article 6(1)c Decision.

–“[…]* does not view compilations as being materially different from other albums, as far as their characteristics and consumption are concerned. As a result, it does not view it as meaningful to seek to distinguish compilations from albums”;

–[…]* notes: “[c]ompilations certainly now compete within our limited sales space for placement alongside or instead of albums";

–“[…]* generally believes that compilations in digital format compete with other digital albums and are purchased by the same types of consumers or end users"; and

–[…]* notes: “[c]ompilations certainly now compete within our limited sales space for placement alongside or instead of albums";

–[…]* notes: “[t]here is no differentiation between the way that compilation albums and other albums are presented to users within our services".

(113) Third, in the Notifying Party’s view the wholesale licensing of music video rights form part of the overall recorded music market. Music videos are just a different medium for monetising the same recorded music content. Where music videos are downloaded from platforms such as Apple/iTunes, they are essentially another type of premium product with a higher PPD on record companies’ rate cards. Where music videos are distributed freely (for example on MTV), they are used in order to sell advertising space and as a promotional means to drive sales of audio products, while audio sales may encourage consumers to buy the premium video product. Furthermore, the Notifying Party claims that Universal’s principal customers generally purchase both audio and video products for onward retail.

–“[…]* generally believes that compilations in digital format compete with other digital albums and are purchased by the same types of consumers or end users"; and

43 Universal's Reply to the Commission Request for information of 4 April 2012.

(114) Fourth, the Notifying Party claims that a separate product market should not be identified for the licensing of recorded music to TV, radio broadcasters and other businesses, which license these rights from producers’ music licensing companies. Although the terms on which recorded music is licensed differ as between retailers and broadcasters, the underlying recorded music rights stem from the same competitive A&R process. On the demand side, there is also a large degree of competitive interaction between the two channels given that greater air-time devoted to particular songs can lead to an increase in the popularity of such songs amongst consumers, which in turn increases sales in the retail channels. In addition, the rise of streaming music services and Internet-based radio stations demonstrates a structural convergence of the broadcasting and retail channels.

–[…]* notes: “[t]here is no differentiation between the way that compilation albums and other albums are presented to users within our services".

44 Form CO, pp. 113, 114 and 115.

6.2.2.2. The Commission's assessment

45 Commission's decision of 3 October 2007 in Case No COMP/M.3333-Sony/BMG.

6.2.2.2. The Commission's assessment

46 music . Within the digital music market, the Commission left open the question whether digital music delivered for online applications and mobile applications, or whether music delivered for downloading and streaming, are part of one and the same relevant product market . The Commission did not consider it appropriate to further segment these relevant product markets based on genre (for example pop, classical, jazz etc.) or to identify separate product markets for single artist albums and compilations .

47 Commission's decision of 3 October 2007 in Case. No. COMP/M.3333, Sony/BMG, Recital 34.

(115) The Commission has assessed concentrations in the recorded music sector in the past, most recently in Sony/BMG. In that case, the Commission identified separate product markets for the wholesale of physical music and of digital music. Within the digital music market, the Commission left open the question whether digital music delivered for online applications and mobile applications, or whether music delivered for downloading and streaming, are part of one and the same relevant product market. The Commission did not consider it appropriate to further segment these relevant product markets based on genre (for example pop, classical, jazz etc.) or to identify separate product markets for single artist albums and compilations.

48 Commission's decision of 3 October 2007 in Case No COMP/M.3333 - Sony/BMG, Recital 17.

(116) In this case, the Commission also examined whether it is appropriate to distinguish separate markets for physical recorded music versus digital recorded music, within digital recorded music for download and streaming and for both physical and digital recorded music markets, further delineations by genre, compilations versus albums, singles versus albums, audio versus video music. The market investigation has not revealed any new material element which would justify departing from the approach adopted in Sony/BMG. The only exception to this concerns the licensing of recorded music to certain end users.

(117) Physical vs. digital. The majority of the respondents to the market investigation (both record companies and customers) confirmed that the segmentation of the wholesale market for recorded music between physical and digital (both downloads and streaming) recorded music is still appropriate.

(118) The majority of competitors (13 competitors out of 14) also consider that, from the consumer perspective, physical and digital music are different. Customer responses were more mixed, with 22 customers indicating that physical music is different from digital music, and 27 expressing the opposite view.

(119) The Commission also notes that customers on the wholesale markets for physical and digital music are generally different, with limited examples of physical retailers, such as Amazon and Media Markt, which have also developed a digital offering and hence offer both physical and digital music. These customers are, however, exceptions in the sense that even though other physical retailers are trying to develop a digital offering, their activities remain marginal. In fact, [….]* of Universal's [….]* EEA physical customers […]* is listed amongst Universal's […]* EEA digital customers. As an illustration, […]* which is Universal's […]* EEA physical customer and generates […]* revenues for Universal […]* is not part of Universal's […]* EEA digital customers; […]*, which is Universal's […]* EEA physical customer and generates […]* revenues for Universal […]* is Universal's […]* EEA digital customer and generates […]* revenues for Universal […].

(120) In any event, even where some retailers sell both physical and digital music, it clearly emerges from the market investigation that at the wholesale level, which is the level at which the Commission needs to assess the competitive effects of the proposed concentration, negotiations with record companies typically take place separately for physical and digital recorded music.All competitors and customers confirmed this.

50 Customers' reply to the Commission Request for information of 17 February 2012, question 7; Competitors' reply to the Commission Request for information of 17 February 2012, question 12.

51 Only one competitor indicates that it conducts joint negotiations for physical and digital music.

(121) The majority of customers (27 customers out of 42) and competitors (8 competitors out of 14) also confirm that at least large record companies tend to have separate sales organizations for physical and digital music.

52 Customers' reply to the Commission Request for information of 17 February 2012, question 4.3; Competitors' reply to the Commission Request for information of 17 February 2012, question 9.3.

(122) Moreover, the majority of customers (32 customers out of 45) and competitors (9 competitors out of 14) confirm that production and distribution costs, as well as distribution channels, are different. While marketing budgets increasingly cover both digital and physical products, the cost structures remain different, in particular as regards manufacturing costs, which are non-existent for digital music, and publishing rights costs, which are paid by the record companies for physical music and by the digital service providers for digital music.

53 Customers' reply to the Commission Request for information of 17 February 2012, question 4.5; Competitors' reply to the Commission Request for information of 17 February 2012, question 9.5.

(123) Further, the majority of respondents in the Commission's market investigation indicate that accordingly, the technical and commercial conditions for sales differ for physical recorded music and digital recorded music.

54 Customers' reply to the Commission Request for information of 17 February 2012, questions 4.4 and 6; Competitors' reply to the Commission Request for information of 17 February 2012, questions 9.4 and 11.

(124) As regards the Notifying Party's claim that the Commission did not take sufficient account of the pricing interaction between digital and physical markets, the Commission notes that three out of four quotes put forward by the Notifying Party refer to retail prices, not to wholesale prices. In paragraph 123 of the Statement of Objections the Commission does acknowledge that retail prices for physical and digital recorded music might exercise a certain influence on each other at the retail level. Nevertheless, even if there were to be a certain degree of price correlation at the retail level, this would not necessarily be the case also at the wholesale level, which is the focus of the Commission's assessment in this case. The margins of retail distributors of digital recorded music differ from those of retail distributors of physical recorded music. Moreover, the commercial conditions under which negotiations take place between wholesale distributors of recorded music and their customers differ from the conditions under which retail prices to end consumers are set.

50 Customers' reply to the Commission Request for information of 17 February 2012, question 7; Competitors' reply to the Commission Request for information of 17 February 2012, question 12.

51 Only one competitor indicates that it conducts joint negotiations for physical and digital music.

52 Customers' reply to the Commission Request for information of 17 February 2012, question 4.3; Competitors' reply to the Commission Request for information of 17 February 2012, question 9.3.

53 Customers' reply to the Commission Request for information of 17 February 2012, question 4.5; Competitors' reply to the Commission Request for information of 17 February 2012, question 9.5.

(126) In addition, the market investigation did not confirm that such a pricing interaction exists at wholesale level. The large majority of competitorsand customers considered that the sale of physical recorded music and the sale of digital recorded music present different technical and commercial conditions. Moreover, the majority of competitorsreplied that there is no, or only partial, pricing correlation at wholesale level.The majority of customers confirmed this.Moreover, several respondents observed increasing divergences in pricing between digital and physical markets with the emergence of streaming.

(125) In other words, irrespective of any possible interaction between physical and digital music at retail level, such interaction does not provide any indications on the competitive dynamics when customers negotiate with record companies at wholesale level. For instance, should the merged entity increase prices or apply other commercial terms that are more disadvantageous to its digital customers, it seems unlikely that customers which are predominantly active in the physical market would be in a position to exert any significant countervailing competitive constraint given their marginal presence, in terms of revenues for the merged entity, on the digital market; the decreasing trend of physical music sales; and the fact that prices on the physical market tend to be twice as high as prices on the digital market.

(127) Finally, recorded music in physical format has different characteristics to recorded music in digital format, such as a different sound quality, the absence of any usage restriction and the art work (covers, booklets and lyrics). Hence, physical music is generally listened to by users on different devices than digital music. Moreover, physical music can be collected and stored in a way that is similar to books. On account of their distinctive features, both physical and digital music seem to continue to appeal to different categories of consumers. While collectors and older users prefer the physical characteristics of CDs, the accessibility of digital music is attractive to younger generations of consumers. As pointed out by a respondent:"[t]he multi-access capacities of digital music (from your home, work, car, etc.) and the on demand features (the user can listen to an extensive music library, on a track by track basis or via customizable radios, via playlists either edited by himself or by the digital provider, etc) is now important to users and it also corresponds much more to the culture of the younger generations".

61 Customers' reply to the Commission Request for information of 17 February 2012, question 4.1; Competitors' reply to the Commission Request for information of 17 February 2012, question 9.1.

(126) In addition, the market investigation did not confirm that such a pricing interaction exists at wholesale level. The large majority of competitors and customers considered that the sale of physical recorded music and the sale of digital recorded music present different technical and commercial conditions. Moreover, the majority of competitors replied that there is no, or only partial, pricing correlation at wholesale level. The majority of customers confirmed this. Moreover, several respondents observed increasing divergences in pricing between digital and physical markets with the emergence of streaming.

62 Customers' reply to the Commission Request for information of 17 February 2012, question 4.1 (ID: 6368).

(128) Therefore, the Commission considers for the purpose of this case that the wholesale of physical recorded music and the wholesale of digital recorded music constitute separate product markets.

(129) Download vs. streaming services. The market investigation revealed certain specificities of the wholesale of digital recorded music for both download and streaming services which may support the view that these services belong to separate product markets or, at least, separate segments within a single differentiated product market.

(130) First, on the demand side, the customers that deal with record companies are different. For instance, download and streaming services are generally offered to end consumers by different providers. Where a provider of music download services has developed streaming services, or conversely a provider of music streaming services has developed download services, the latter offerings are generally rather marginal and complementary to the main type of services provided. By way of illustration, […]* of Universal's […]* EEA download customers is also listed in its […]* EEA streaming customers; […]* of EMI's […]* EEA download customers are also listed in its […]* EEA streaming customers; […]*.

(128) Therefore, the Commission considers for the purpose of this case that the wholesale of physical recorded music and the wholesale of digital recorded music constitute separate product markets.

63 Customers' reply to the Commission Request for information of 17 April 2012, question 2.

64 Universal's and EMI's reply to the Commission Request for information of 4 April 2012, question 1; Competitors' reply to the Commission Request for information of 17 April 2012, question 2.

(131) Second, wholesale pricing models and other key commercial conditions differ between the two types of services. For download services, pricing is generally established on the basis of a price per unit (track and albums PPD) whereas, for streaming services, pricing is more complex and primarily established on the basis of a share of the retail revenue or on a fixed amount per subscriber. The other key commercial conditions, notably related to advertising, marketing, advance payments and MFN clauses tend to be more comparable between download and streaming platforms.

65 Competitors' reply to the Commission Request for information of 17 April 2012, question 4.

(132) Also from the perspective of end users, pricing between these two services is very different. End users of download services can purchase one track or one album at a given price on a transactional basis. End users of streaming services can either access typically a limited number of tracks for free or, by paying a monthly subscription fee, access and temporarily store on their devices as much music as they would like to, whether chosen by them (lean-forward services) or proposed to them by the service provider (lean-back services).

66 Customers' reply to the Commission Request for information of 17 April 2012, question 3.1.

(133) Third, a large majority of respondents stress that, at least today, these services are not fully substitutable from an end user's perspective given the different characteristics of each type of service. Most importantly, portability and ownership of the music seem to be still significant differentiating factors. Whereas downloaded tracks are accessible on any device and are owned by the user who can listen to the music for an unlimited period of time, including when it is offline, streaming services are tethered to a limited number of devices and the ability to listen to a given song, including when offline, is limited in time and conditional on the payment of a subscription fee. In addition, advertising

63 Customers' reply to the Commission Request for information of 17 February 2012, question 10.

(134) Fourth, both types of services present different delivery processes and technical constraints. Whereas for streaming services, network capabilities are critical, they do not require significant storage capacity. The opposite is typically true for download services. Furthermore, a number of respondents indicate that sound quality of downloads is superior to the sound quality of streaming services, with compression rates of respectively 320 kbps and 192 kbps

68.

(135) Finally, certain respondents make a distinction between end users of streaming services, who tend to be regular consumers of music, and end users of download services who tend to be more sporadic consumers.

69 Customers' reply to the Commission Request for information of 17 April 2012, questions 3.3 and 4.

(136) Conversely, the market investigation confirmed that, on the supply side, the same sales and marketing people at the different record companies deal with wholesale of recorded music for both download and streaming services.

(137) Moreover, new cloud-based services tend to blur the distinction between download and streaming services.

70 Competitors' reply to the Commission Request for information of 17 April 2012, question 3.

(138) Furthermore, the market investigation also confirmed that, from the point of view of end users, there is a certain competitive interaction between download and streaming services.

71 Competitors' reply to the Commission Request for information of 17 February 2012, question 10; Customers' reply to the Commission Request for information of 17 April 2012, question 3.2.

(139) It is therefore likely that, in the future, any differences that may exist today between these two types of services would become increasingly less marked.

(140) In light of the above, the Commission considers for the purpose of this case that it is not necessary to segment the market for the wholesale of digital music between digital download services and digital streaming services. In any event, the different competitive dynamics, which characterise each of these two segments, will be taken into account in the context of the competitive assessment.

(137) Moreover, new cloud-based services tend to blur the distinction between download and streaming services.

72 Customers' reply to the Commission Request for information of 17 February 2012, question 8; Competitors' reply to the Commission Request for information of 17 February 2012, question 14.

73 of a song or an album into a specific genre is entirely subjective. Third, a vast majority of customers indicated that they purchase and sell all genres of music.

(138) Furthermore, the market investigation also confirmed that, from the point of view of end users, there is a certain competitive interaction between download and streaming services.

74 Customers' reply to the Commission Request for information of 17 February 2012, question 9.

(142) A number of competitors have stressed the specificities of classical music and jazz as a genre. The Commission therefore assessed these genres in further detail.

(139) It is therefore likely that, in the future, any differences that may exist today between these two types of services would become increasingly less marked.

75 Competitors' reply to the Commission Request for information of 17 February 2012, question 16.

(143) The Commission considers that there are indeed certain elements emerging from the market investigation, which suggest that, in particular, classical music may have certain specificities compared to other music genres.

(144) It emerges from the market investigation that classical music attracts specific consumer groups with specific purchasing patterns. According to one respondent, typical classical music "customers are fewer, older and very discerning". According to another respondent: "traditional classical music consumers are generally adults, culturally experienced and expecting top quality (this is a niche market)".

76 Competitors' Reply to the Commission's Request for information of 17 April 2012, question 8 (ID 7769).

(141) Genre. The market investigation showed that, by and large, a segmentation of the recorded music market based on genre is not appropriate. First, the borders between genres are often blurred and artists and songs can fit within several genres at the same time. Second, several customers also underline that placing of a song or an album into a specific genre is entirely subjective. Third, a vast majority of customers indicated that they purchase and sell all genres of music.

77 Competitors' Reply to the Commission's Request for information of 17 April 2012, question 8 (ID 7272).

(145) In addition, sales of classical music are still overwhelmingly in physical format and specialist retail outlets still play an important role. In this regard, physical retail of classical music requires a broad offering, specific purchase and sales teams and separate shop departments. Distribution on digital services also entails technical constraints, in terms of the following: "search engine (the research of classical tracks is more complex as it can be done on more elements than just artist-name of the track, it can also be done on author, orchestra, maestro, etc); length of the tracks: the streaming of such tracks is more demanding in terms of bandwidth, storage capacity of the terminal devices of end users (in other terms, basically, it is much harder to download or stream classical music over mobile devices); cost of the rights: because of the length of the tracks which commonly exceed the average 3 minutes of track, the right to be paid to record companies are much more expensive; quality of sound: a digital service cannot generate a great audience on this type of music unless it can ensure a good sound quality".

(142) A number of competitors have stressed the specificities of classical music and jazz as a genre. The Commission therefore assessed these genres in further detail.

78 Customers' Reply to the Commission's Request for information of 17 February 2012, question 10 (ID 6368).

79 Customers' Reply to the Commission's Request for information of 17 February 2012, question 10.

80 Competitors' Reply to the Commission's Request for information of 17 February 2012, question 16.

and/or caters to the preferences of a specific group of consumers. Sales of classical (and jazz) music are not hit or new release driven and there are virtually no singles released for those genres. While sales are partly driven by critical review, promotion on radio and TV is not essential.

81 Customers' reply to the Commission Request for information of 17 February 2012, question 10; Competitors' reply to the Commission Request for information of 17 April 2012, question 7.

(147) On the supply side, all responding competitors also note that the production costs of classical music are significantly higher than production costs for other genres.According to one respondent, "[c]lassical music requires special expertise on the A&R side with a specific workforce holding great knowledge of repertoires – classical music is targeted to a demanding niche audience that is very knowledgeable. The complexity of the A&R for classical music also lies in the fact that a solid knowledge of numerous specific music instruments is necessary. […] As for classical music, jazz is targeted to a very specific and knowledgeable audience that requires having a highly skilled A&R workforce. New jazz music is very difficult to develop and significant investment is required to develop new artists in that genre". Moreover, the recording process for classical music often involves orchestra and soloists together with specific venues or studios.

82 Competitors' reply to the Commission Request for information of 17 April 2012, question 7.

83 Competitors' reply to the Commission Request for information of 17 April 2012, question 7 (ID: 7750).

(146) Moreover, contrary to the Notifying Party's claims, 20 customers out of 37 have identified classical music as a genre with distinctive product characteristics (and 9 out of 14 responding competitors have outlined jazz as a genre of music with specific product characteristics). These respondents also indicate that classical and jazz music are marketed and/or sold differently from other genres, and/or caters to the preferences of a specific group of consumers. Sales of classical (and jazz) music are not hit or new release driven and there are virtually no singles released for those genres. While sales are partly driven by critical review, promotion on radio and TV is not essential.

(148) The Notifying Party claims that all genres present differences in the A&R process and the musical instruments required are actually irrelevant from a market definition perspective. However, it is important to take into account the fact that A&R and production costs for at least classical music tend to be much higher than for other genres. This means in practice that record companies that are currently not present in the classical segment are unlikely to be able to produce and market classical music in the short term without incurring significant additional costs or risks in response to small and permanent changes in relative prices.

(149) The Commission considers for the purpose of the present case that the wholesale of recorded music (whether in physical or in digital form) should not be further segmented by genre. However, the Commission will take into account the different competitive dynamics, which are specific particularly to the classical music, in the context of the competitive assessment.

(147) On the supply side, all responding competitors also note that the production costs of classical music are significantly higher than production costs for other genres. According to one respondent, "[c]lassical music requires special expertise on the A&R side with a specific workforce holding great knowledge of repertoires – classical music is targeted to a demanding niche audience that is very knowledgeable. The complexity of the A&R for classical music also lies in the fact that a solid knowledge of numerous specific music instruments is necessary. […] As for classical music, jazz is targeted to a very specific and knowledgeable audience that requires having a highly skilled A&R workforce. New jazz music is very difficult to develop and significant investment is required to develop new artists in that genre". Moreover, the recording process for classical music often involves orchestra and soloists together with specific venues or studios.

(150) Compilations vs. albums. In the physical market, while a majority of respondents anticipate a decrease in the sales of compilations, certain respondents underline that, in some large EEA countries, such as the United Kingdom, France and Germany, there is still high demand for compilations, particularly during certain times of the year (Valentine's day, Summer or Christmas periods). In the digital market, the availability of single tracks and the unbundling of albums, together with the possibility to create individual playlists, have made compilations

81 Customers' reply to the Commission Request for information of 17 February 2012, question 10; Competitors' reply to the Commission Request for information of 17 April 2012, question 7.

redundant to some extent. However, at least as far as download services are concerned, some compilations continue to be moderately successful.

(151) It also emerges from the market investigation that compilations have different product characteristics and a lower price (at least compared to new releases) than albums. According to one respondent: "[o]ften compilation albums will have more recordings than an album of new release material. It is not uncommon for a multi-artist compilation album to comprise 30 or 40 tracks in a 2-CD box set (sometimes even more) and the equivalent in digital format where a new non-compilation album will have around 10-12 tracks. Compilation albums are usually marketed at lower price points than new release albums to increase their attractiveness to consumers: so a 40-track compilation album will often be found at the same price point as a new-release album containing a quarter as many tracks".

(152) The lower price of compilations is made possible by their lower cost of production. As underlined by a respondent: "[t]he record company is able to maintain its margins notwithstanding the greater number of tracks being sold because the origination costs are usually lower – the tracks have already been recorded, the significant artist advances have already been paid and, sometimes for older material, the royalty rates payable to artists are lower". In fact, as explained, the production process of compilations is different from that of other albums. Compilations do not require new recordings and new A&R expenses.

(153) Several respondents to the market investigation also point out that compilations and albums are aimed at different consumer categories. According to several respondents, compilations have a mainstream clientele, "without tighter attachment to specific artist or music, who enjoy in a variety of music rather than its specificity."Whereas a competitor considers that the target consumers of compilations may be larger in scope ("[c]ompilations appeal to a broad range of consumers – from the younger consumer wanting to rapidly acquire many current hits at a relatively low cost, to the hardcore aficionado looking to complete a collection of an artist’s body of work by seeking out previously unreleased material, to the occasional music buyer that is only interested in the best known hits of a particular artist. Compilations are very popular as gifts particularly at certain times of the year – Christmas, Mother’s Day, Valentine's Day and as impulse purchases – for example, the driver with a long journey ahead buying a “sixties” compilation in a service station"), the same competitor concludes that "[c]ompilations therefore appeal to a demographic group that may not regularly purchase new releases but have a general affinity to music".

85 Customers' reply to the Commission Request for information of 17 February 2012, question 18; Competitors' reply to the Commission Request for information of 17 February 2012, question 12. 86 Customers' reply to the Commission Request for information of 17 April 2012, question 6.1; Competitors' reply to the Commission Request for information of 17 April 2012, question 12.1. 87 Competitors' reply to the Commission Request for information of 17 April 2012, question 12 (ID: 7750). 88 Competitors' reply to the Commission Request for information of 17 April 2012, question 12. 89 Competitors' reply to the Commission Request for information of 17 April 2012, question 12 (ID: 7750). 90 Competitors' reply to the Commission Request for information of 17 April 2012, question 11. 91 Customers' reply to the Commission Request for information of 17 April 2012, questions 6.2 and 6.3. 92 Customers' reply to the Commission Request for information of 17 February 2012, question 11 (ID: 3291).

(154) These differences in target consumers are reflected in marketing and distribution. According to a number of respondents, compilations enjoy a broader distribution and higher visibility. According to one respondent: "outlets for compilations may be more varied – for example non-traditional retail is sometimes used to reach consumers who may not be regular visitors to physical record stores, such as coffee stores or cafes, as free-to-the-consumer releases in newspapers or through mail order".

(155) Moreover, certain respondents indicate that compilations are prominently displayed in supermarkets and hypermarkets so as to favour impulse purchases. A number of respondents refer in particular to Universal and EMI's "Now that's what I call Music" compilation (the "NOW compilation business"), which has a strong brand name. It seems that those types of compilations benefit from significant marketing budgets and extensive media campaigns compared to most albums As summarised by one respondent: "[c]ompilation albums are frequently impulse purchase products; rarely does a consumer await the release of a new compilation album in the way that he or she might await the release of an album from his or her favorite artist. As a result of this tendency to be purchased on impulse, compilation albums are heavily reliant upon TV marketing and high profile retail positioning in order to achieve their sales. Because compilation albums seek to appeal to as broad a commercial sector as possible, they are more successfully marketed through non-traditional music retail channels, for example supermarkets". However, according to a number of respondents, distribution and marketing of compilations and of albums are similar.

(156) On the other hand, it appears that, from the point of view of physical and digital retailers (at least with regard to download services), compilations are substitutable and compete for shelf space and visibility with new albums. They are typically included in the sales and licensing deals entered into between record companies and their customers and they do not form the subject-matter of any specific negotiation to a greater extent than albums.

93 Competitors' reply to the Commission Request for information of 17 April 2012, question 12 (ID: 7750). 94 Competitors' reply to the Commission Request for information of 17 April 2012, question 11; Customers' reply to the Commission Request for information of 17 April 2012, question 6.4. 95 Competitors' reply to the Commission Request for information of 17 April 2012, question 12 (ID: 7750). 96 Customers' reply to the Commission Request for information of 17 April 2012, question 6.4. 97 Competitors' reply to the Commission Request for information of 17 April 2012, question 12.2 (ID: 7291). 98 Customers' reply to the Commission Request for information of 17 April 2012, question 6.4. 99 Customers' reply to the Commission Request for information of 17 February 2012, question 11. 100 Customers' reply to the Commission Request for information of 17 April 2012, question 5.

(157) Streaming service providers represent a possible exception to this, as compilations in themselves are not a particularly important product for them. It is important for streaming service providers to have the repertoire of as many record companies as possible available on their service, that is to say, as many songs from different artists as possible. As a result, from their perspective, compilations are a duplicate, as they would already have the recording rights for each of the songs included in the compilations from their licensing deals with each of the record companies, whose songs are included in the compilations.

(158) The Commission therefore considers for the purpose of this case that the wholesale of compilations does not constitute a separate product market within the wholesale of each of physical and digital music.

(159) Singles vs. albums. The market investigation confirmed that this distinction retains little relevance today in the physical market, where singles have almost completely disappeared. In the context of limited shelf space, physical retailers favour the sale of albums which is more profitable.

(160) In the digital market, a majority of respondents consider that albums and single tracks compete for the same shelf space (whether physical or digital) and consumer attention. Several respondents also submitted that the price of albums is constrained by the price of single tracks.

(161) The Commission therefore considers for the purpose of this case that the wholesale of recorded music (in digital form) should not be further segmented between singles and albums.

(162) Music videos. The market investigation confirmed that audio and video music are different products because the demand side and the supply side of the market are organised differently from audio music. As outlined by one respondent: "video is its own product, has its own commercial structure and in turn its own commercials deals. […] There are also different partners who specialize in videos only. Separate negotiations therefore need to be had for videos." Moreover, as indicated by one respondent to the market investigation, the repertoires available are different: "promotional videos are produced only for a small percentage of audio tracks and, in addition, video content is increasingly produced in respect of which there is no audio only analogue (e.g., live performance videos or behind the scenes video)".

(163) On the other hand, music video rights, like audio music rights, are licensed (other than to linear TV broadcasters) directly by the record companies and some respondents point out that there is no material difference between the record companies' wholesale revenue models for audio music and for video music. […]* . In addition, the same record company typically controls both the audio music rights and the music video rights (when the video exists) over a given song. Accordingly, the competitive position of record companies in relation to music videos seems similar to their competitive position in relation to audio music rights. The Commission therefore considers that audio music and video music belong to the same product market for the wholesale of recorded music (whether physical or digital).

(164) Music licensing to end users. In addition, it appears that a separate product market for the provision of recorded music to end users licensing recorded music from producers’ music licensing companies should be identified. This reflects the fact that these customers do not typically purchase the recorded music that they use in their programmes directly from record companies, but from national music licensing companies, to which record companies grant the right to license their repertoire to these specific end users and the terms and conditions at which these customers license music are also very different from the prevailing ones among music retailers.

(165) It is nonetheless noted from the outset that, since record companies do not appear to have any material influence over the terms and conditions at which producers’ music licensing companies license music to end customers, the proposed concentration is not likely to have any anti-competitive effects on this market, which therefore will not be further discussed in the competitive analysis.

(166) Conclusion. The Commission takes the view that a distinction should be made between the wholesale of physical recorded music and the wholesale of digital recorded music.

(167) The Commission also takes the view that the wholesale market for physical and digital recorded music should not be further segmented on the basis of genre, compilations versus albums, singles versus albums and audio music versus video music and that the market for the wholesale of digital recorded music should not be segmented between download and streaming services.

(168) The Commission further concludes that a separate product market should be identified for the wholesale of recorded music to end users licensing recorded music from producers’ music licensing companies.

(194) Seventh, according to the Notifying Party, the analysis of the effects of three "shocks", namely the implementation of the Directive 2004/48/EC of the European Parliament and of the Council of 29 April 2004 on the enforcement of intellectual property rights("IPRED Directive") in Sweden, the introduction of the legislation enacted by the Haute Autorité pour la diffusion des œuvres et la protection des droits sur internet legislation ("HADOPI legislation") in France and the shutdown of LimeWire in the United States, demonstrates a significant degree of substitutability between pirated music and legitimate online music consumption, insofar as anti-piracy measures seem to result in an increase, at least transitory, in music sales

(195) In sum, in its Response to the Statement of Objections, the Notifying Party claims that the evidence available to the Commission supports the conclusion that piracy is the main cause for the dramatic decline of recorded music revenues. The Notifying Party lists quotes from respondents to the market investigation […]*, in support of the claim that legal music services compete directly with pirate music services. In addition, the Notifying Party refers to the Commission decision in case Oracle/Sun Microsystems,in which the Commission accepted the principle that a free product can exercise an important and growing competitive constraint on a paid-for product. The Notifying Party argues that the closeness of competition between pirated and legal music copies is even greater given that both products are functionally and qualitatively identical – which was not the case in Oracle/Sun Microsystems.

(196) Finally, in its Response to the Article 6(1)(c) decision, the Notifying Party rejects the point made by the Commission that legal and illegal sources of music do not appear to be substitutes from the point of view of the record companies' direct customers, that is to say physical and digital music retailers. In this regard, it considers that the Commission did not take into account the competitive interactions between the wholesale and the retail level of the music supply chain

and the fact that consumer demand at retail level would shape wholesale negotiations between record companies and music services. The Notifying Party refers to a number of internal documents from Universal and EMI and to several replies from competitors to the Commission market investigation that would show the direct and material constraining effect of piracy on wholesale commercial negotiations .

6.2.5.2. The Commission's assessment

(197) The Commission has never considered in its past decisions that legitimate and unauthorised supplies of music belong to the same product market. However, the market investigation in the Sony/BMG caserevealed that piracy exerts a competitive constraint on record companies in certain territories.

(198) The Commission considers that, contrary to the Notifying Party's claim, assessing the degree of substitution between legal and illegal music and the extent to which consumers mix and match between legal and illegal music is not relevant to determine whether illegal music should be considered as part of the market for the wholesale of recorded music. This reflects the fact that, regardless of whether legal and illegal music are to be considered as substitutes from the point of view of end users, they do not appear to be substitutes from the point of view of the record companies' direct customers, that is to say physical and digital music retailers. From the demand side, physical or digital music retailers do not have the choice to switch to illegal music if the price for legal music were to rise by 5-10%, since, even in this scenario, they would still need to procure legal music from the record companies. Similarly, from the supply side, record companies are not active in the supply of illegal music (nor could they be expected to start engaging in any such activity if the price for legal music were to decrease by 5-10%).

(199) By the same token, pirate services do not compete with record companies. They are not active in A&R and they are not active in the wholesale supply of recorded music to physical and digital customers. They merely make available music that has been produced by record companies.

(200) The Commission also notes that the claim that consumer demand at retail level, which may be influenced by piracy, may shape the wholesale negotiations between record companies and music service, even if proven, would not alter the conclusion that legal and illegal music are not substitutable from the point of view of music retailers. If anything, it would be an element to be taken into consideration as part of the competitive assessment to assess the constraint exercised by piracy on record companies' market power.

(201) The Notifying Party has not put forward any argument to demonstrate the opposite view in its Response to the Statement of Objections.

(202) In any event, the market investigation in the present case was also inconclusive as to whether legal and illegal sales of music are substitutable from the point of view of end users.

(203) It remains unclear whether and, if so, to what extent, consumers substitute or mix-and-match illegal and legitimate sources of music. While a large number of responding customers consider that consumers mix-and-match, several respondents stress that only certain services are directly substitutable by illegal music from the point of view of final consumer and no common view emerges as to which services are substitutable. For certain respondents streaming services are a substitute for illegal music consumption, while according to others the user experience offered by streaming services cannot easily be replicated within illegal services.

(204) According to several respondents, there may be three categories of consumers: those who only consume illegally obtained digital music for budget reasons, those who only consume legally obtained music for moral reasons or fear of viruses or sanctions and those who both pay for digital music and obtain it illegally to discover new music, “check out” recordings before purchasing them or for hits and seasonal songs .

(205) In addition, a respondent refers to an overlap analysis between customers who obtain illegal and legal music for France, Ireland, New Zealand and Canada, carried out by IFPI, which would indicate that only a minority of users currently access both legal and illegal sources. The estimates made by the Notifying Party from the IFPI study on the share of consumers that would use illegal sources of recorded music do not challenge the existence of three categories of consumers. Those are only relevant in order to determine the degree of competitive constraint exercised by piracy on retailers of recorded music. However, those are not informative of the degree of competitive constraint, if any, exercised by piracy on record companies.

(206) In support of that view, one respondent refers to a September 2011 pricing study carried out on the basis of an analysis of the weekly iTunes sales data for a particular set of music from one of the “Big 4” record companies around the period when Apple moved to tiered pricing for single-track downloads on iTunes (April 2009). This study would find a low degree of price elasticity for digital singles at -0.45, which means that a 1% price increase resulted in a 0.45% decline in quantity sold. Although the report does not explicitly model piracy, the effect of piracy is implicit in the estimates, because piracy was and is an available option to consumers .

(207) The Commission takes good note of the methodological comments made by the Notifying Party on this study. However, the Commission considers that these comments do not address the crucial issue of whether and how the pressure exercised by piracy on retail pricing and sales is transmitted to wholesale pricing and sales. In this regard, the Commission concludes that these comments are irrelevant. Moreover, the Commission finds that the distinction made in that study between high-piracy and low-piracy tracks is an indication that piracy does not affect all types of recorded music to the same extent. This is a further indication that illegal and authorised sources of recorded music are not part of the same product market.

(208) Economic literature also supports this categorisation to some extent. While it is considered that pirated material may amount to foregone sales, it is also considered that illegal downloads may perform a promotional role, more or less in the same way one would discover new artists or new releases through 30 second free streams or listening to them in record shops .

(209) Moreover, a large number of respondents to the market investigation consider that piracy does not affect all demographics and that typical users of unauthorised sources of music are in the 16-25 year old age group .

(210) As regards the argument made by the Notifying Party that replies of certain respondents to the market investigation would contradict their public statements, the Commission shares the view of […]* that statements that piracy is a problem for the industry cannot be considered as evidence that piracy forms part of the same relevant product market as legitimate sales. The question whether digital piracy and legitimate online music form part of the same market does not relate to the impact of piracy on legitimate sales as such, but on the degree of substitutability between both products from a customer’s perspective. In this regard, the Frontier BASCAP study makes it clear that it does not attempt to estimate business losses, which would involve making assumptions about the degree of substitution between illegitimate and legitimate products. Moreover, the projections in the Frontier BASCAP study of an increase in the total volume of digital piracy do not imply that the share of music consumers who listen to pirated music will also increase. In addition, […]* ’s statements in the pleadings opposing […]* and claiming that the service directly competed with the legal digital music services should also be placed back in the specific context of the […]* 's website, which effectively offered the ability for end users to find links to “free music” without indicating that these were links to illegal websites and infringing content .

(211) As regards the results of the […]* study put forward by the Notifying Party, the Commission considers that the results of this study are undermined by a number of elements, including the uncertainties relating to the representativeness of the samples and the specific conditions of the surveys. Importantly, the results of the surveys are undermined by the lack of information on the type of content being downloaded or streamed and the fact that some sites allow access to a wide range of digital content, such as movies, computer games and electronic books. The conclusion that there is a strong correlation between the time spent on a digital content site and downloading or streaming at least some of that content cannot therefore be ascertained.

(212) The Commission also considers that the arguments submitted by the Notifying Party do not address the issue of the representativeness of the samples used for the […]* study, in particular in terms of demographics. In any case, the Commission considers that the […]* study is only informative of the competitive constraint exercised by piracy on retail pricing and sales.

(213) Contrary to the claim of the Notifying Party in its Response to the Statement of Objections, the Commission shares the view of the Notifying Party, supported by the market investigation, that piracy has an adverse impact on retail prices and volumes and that digital retailers are constrained in their ability to raise retail prices by the threat of piracy. However, the Commission considers that this observation does not alter the conclusion that there is no substitution between illegal and authorised sources of music at wholesale level.

(214) As a result, the Commission considers for the purpose of this case that it is not appropriate from the demand or from the supply side, to consider legal and illegal music as part of the same relevant product market for the wholesale of recorded music. This conclusion is, however, without prejudice to the analysis of piracy as an out-of-market constraint on record companies' pricing and output decisions in the context of the competitive assessment of the proposed concentration.

6.3. RELEVANT GEOGRAPHIC MARKETS

6.3.1. A&R

6.3.1.1. The Notifying Party's view

(215) Universal does not take a view as to the exact scope of the relevant geographic market in relation to A&R activities.

6.3.1.2. The Commission's assessment

(216) In its previous decisions relating to the recorded music sector, the Commission never defined an upstream market for A&R.

(217) It is not necessary to take a view on the geographic scope of this possible market. Indeed, regardless of the position taken on this point, the competitive assessment of the proposed concentration will not be affected, as the strength of record companies in A&R will be taken into account anyway in the competitive assessment as one of the key factors contributing to their market position in the market for wholesale of recorded music.

6.3.2. Wholesale of recorded music

6.3.2.1. The Notifying Party's view

(218) The Notifying Party submits that today the geographic market for the wholesale of recorded music (both physical and digital) is EEA-wide

(219) As to the wholesale of music in physical format, the Notifying Party submits that many major retailers are becoming increasingly international, that there are significant levels of parallel trade between Member States, and that total distribution costs are generally low at around […]*% of a record company’s net receipts from the sale of a CD.

(220) As regards the wholesale of digital music, the Notifying Party submits that, even if licensing agreements with online customers include territorial restrictions, the advent of online retailing has facilitated cross-border purchases where attractive price differentials exist.

(221) In its Response to the Statement of Objections, the Notifying Party claims that the Commission did not take into account the pan-European dimension of the market and failed to assess the proposed concentration on an EEA-wide market.

6.3.2.2. The Commission's assessment

(222) In Sony/BMG, the Commission concluded that the market for the wholesale of physical recorded music was national in scope.Regarding the market for the wholesale of digital recorded music, the Commission concluded that this market was national in scope, with indications that it could develop into a multi-territorial market in the future .

(223) The Commission's market investigation in this case confirmed that it is still appropriate to define the relevant geographic markets in this manner. In particular, with regard to the wholesale market for digital recorded music the Commission considers that, although there are elements that militate in favour of a broader geographic market, these are not sufficient to justify a broader market definition. However, the Commission will take these elements, which are relevant for all (national) markets, into account in the competitive assessment.

(224) The market investigation revealed that, while the market for the wholesale of physical recorded music is national in scope, the market for the wholesale of digital recorded music may be broader than national.

(225) As regards the market for the wholesale of physical recorded music, the majority of respondents confirmed that national repertoire continues to be appealing to a large audience only in the home country, that prices of recorded music significantly differ across Member States, that marketing campaigns are generally customized according to the local demand for music in each country, that record companies and, in particular, the majors, have a physical presence in the largest Member States to distribute their own, and often also the indies’, music repertoire in the country/region, and that the smallest physical music retailers are mainly active within one Member State and, even those larger retailers which are active across multiple Member States typically negotiate their agreements separately at national level .

(226) Therefore, the Commission takes the view that the wholesale of physical recorded music is national in scope.

(227) As regards the wholesale market for digital recorded music, the Commission found in Sony/BMG that although the geographic scope of this market was still national, this market was nonetheless evolving towards an EEA-wide oneThe Commission's market investigation in this case confirmed that this evolution is

still very much on-going. It confirmed that large digital customers increasingly tend to conclude agreements for the wholesale of recorded music on an EEA-wide basis. However, there are also important digital customers, such as MNOs, but also important download and streaming services, which conclude agreements on a national basis or conclude agreements that are effectively bundles of national agreements. Even the largest digital customers, Apple and Spotify, […]*, are not yet active in all EEA countries. Customers and competitors confirm these findings .

(228) As a result, it emerged, on the one hand, that negotiations between record companies and the largest digital customers are increasingly conducted simultaneously for all the territories where these services are active and often result in agreements covering several Member States. On the other hand, however, there are still important digital customers, who conclude their agreements on a national basis. Moreover, for other customers that have EEA-wide or multi-territorial agreements, key terms, notably related to prices, often differ across territories. Hence competitive dynamics at the national level have a significant impact on the commercial conditions agreed for those markets .

(229) In light of the results of the Commission's market investigation, the Commission takes the view that the exact geographic market definition for the wholesale of digital recorded music should be left open. The Commission leaves this open, as its competitive assessment would be the same irrespective of the market definition chosen. In particular, the Commission has undertaken its competitive assessment of the proposed concentration at both EEA-wide and national market levels.

(230) In this context, the Commission disagrees with the Notifying Party's claim that during its investigation, the Commission did not assess the proposed concentration on an EEA-wide market.As it emerges from Section 7 of this decision and paragraphs 221 to 792 (including) of the Statement of Objections, the Commission analysed the likely impact of the proposed concentration both at EEA and at national level.

(231) The Commission therefore concludes that the market for the wholesale distribution of physical recorded music is national in scope.

(232) The Commission concludes that the market definition for the wholesale distribution of digital music is also still national in scope, and in any event, the

competitive assessment remains the same even if the market is considered to be EEA-wide.

6.3.3. Retail of recorded music

6.3.3.1. The Notifying Party's view

(233) The Notifying Party submits that the advent of online retail has increasingly led to pricing transparency for end consumers across national borders. Moreover, the prevalence of increasingly multi-territorial agreements at the wholesale level would reflect the multi-territorial retail operations of Universal's retail customers .

6.3.3.2. The Commission's assessment

(234) The market investigation provides strong indications that the retail market for each of physical and digital recorded music remains national in scope .

(235) In any event, for the reasons explained in Section 6.2.3.2, the Commission does not consider it necessary to take a view on the definition of the relevant geographic market.

6.3.4. Music-related activities outside the recorded music sector

6.3.4.1. Ancillary services

(236) The market investigation was inconclusive as regards the geographic scope of the possible relevant market(s) for ancillary services .

(237) In any event, for the reasons explained in Section 6.2.4.2, the Commission does not consider it necessary to take a view on the definition of the relevant geographic market.

6.3.4.2. Music publishing

(238) The market investigation revealed the existence of a variety of models for the licensing of online rights. In particular, it seems that online customers increasingly obtain EEA-wide licences for online rights to Anglo-American repertoire given that all major music publishers and some independent music publishers have appointed collecting societies or rights management entities as their agents to offer EEA-wide or multi-territory licences. In case that collecting societies or rights management entities have not been appointed as agents of music publishers, publishers operate across Member States through their network of sub-publishers .

(239) Customers, however, indicate that prices and royalty rates tend to vary between Member States. It is also the case that customers obtain a large share of publishing rights on a national basis .

(240) In any event, for the purposes of the analysis of the proposed concentration, the Commission does not consider it necessary to take a view on the exact scope of the relevant market for online music publishing rights.

7. THE COMPETITIVE ASSESSMENT IN THE RELEVANT MARKETS

7.1. HORIZONTAL NON-COORDINATED EFFECTS

7.1.1. Analytical framework

(241) Through its control of mergers, the Commission prevents mergers that would be likely to deprive consumers of the benefits that effective competition brings by significantly increasing the market power of firms.What is meant by increased market power is the ability of a firm to profitably increase prices, reduce output, choice or quality of goods and services, diminish innovation or otherwise influence parameters of competition.A merger would inter alia significantly impede effective competition if it would create or strengthen a dominant position of a single firm, which typically would have an appreciably larger market share than the next competitor post-merger .

(242) In line with the Guidelines on the assessment of horizontal mergers under the Council Regulation on the control of concentrations between undertakings ("Horizontal Merger Guidelines"), the Commission focussed on a number of factors to determine whether the proposed concentration in this case is likely to significantly impede effective competition, in particular as a result of the creation of a dominant position and in the light of the merged entity's ability to behave to a significant extent independently of its customers and competitors.

7.1.2. Market shares

(243) According to the Horizontal Merger Guidelines, market shares and concentration levels provide useful first indications of the market structure and of the competitive importance of both the merging parties and their competitors. According to well-established case law, very large market shares - 50 % or more - may in themselves be evidence of the existence of a dominant market position. This is however not a decisive threshold; pursuant to the Horizontal

–Merger Guidelines, a merger involving a firm whose market share will remain below 50 % after the merger may also raise competition concerns in view of other factors such as the strength and number of competitors that would remain after the merger. The Commission has thus in several cases considered that mergers resulting in firms holding market shares between 40 % and 50%, and in some cases below 40%, create a significant impediment to effective competition.

Available market share sources in the recorded music sector

(244) Within the recorded music sector, the general industry practice is to refer to market share data as provided by various industry sources such as IFPI, Music & Copyright, Nielsen SoundScan, GfK, the Official Charts Company ("OCC") or Media Control. The data collected by each of these organisations may vary by methodology and may be collected at different levels of the market (wholesale vs. retail).

(245) The Notifying Party and third parties have provided data gathered by IFPI, GfK and OCC, as well as, for the digital recorded music markets, certain revenue data on digital platforms, such as iTunes and [digital platform]*.

(246) The Notifying Party also provided its own estimates of the total market size for countries for which other industry data sources were not available.

(247) There also exist certain chart and radio airplay data that were provided to the Commission by several respondents to the market investigation.

7.1.2.1. IFPI data

(248) IFPI gathers data and provides published estimates of the total market size of the recorded music industry on a country-by-country as well as on an EEA-wide basis. It publishes its estimates both for physical and for digital markets. IFPI estimates both the trade (that is to say wholesale) value and the retail value of these relevant markets.

(249) In estimating the total market size values, it obtains the estimated total revenues generated within each country for both physical and digital formats from the local affiliate industry associations ("National Group"). According to IFPI, in collecting data on the total market size and market statistics, each national group reports via a standard online system, the Market Research Online. IFPI asks all National Groups to follow two sets of international guidelines issued by IFPI, the Market Reporting Guidelines and the Data Definitions Guidelines. Both sets of guidelines cover both physical and digital sales data and differ in some respects between digital sales and physical sales. The IFPI Data Definitions Guidelines are designed to ensure consistency of data collected across different National Groups and provide clarity to National Groups as to what should be included and excluded in the sales figures at national level also bearing in mind to the extent possible the different business practices in different countries. Despite the guidelines, IFPI recognises that there are some differences in the methodology used to compile the market statistics. The members of National Groups in any

country almost always include the local operating companies of the majors as well as local or international independent record companies.

(250) In addition to this, each National Group estimates the percentage of its total market that is represented by non-reporting companies, for example small independent record companies that are not members of the National Group or who are members but have not reported their sales figures. This estimate is ordinarily based on chart information and general market knowledge and is often referred to as the "coverage factor". According to IFPI, whilst there is inevitably some inherent uncertainty in determining the coverage factor and thus the total size of the national markets, in the majority of countries the National Group members account for a large proportion of the market.

(251) The IFPI has, over the years, adapted its template and Data Definitions Guidelines to reflect international changes in the market. In particular, the Commission notes that the process of updating the Data Definition Guidelines involves the IFPI Market Research Steering Committee, where representatives of the four majors review, comment on and sign off the proposed guidelines. IFPI also collects questions and clarification sought by the National Groups, and discusses those with the Steering Committee. Some of the updates to the Data Definition Guidelines are a result of these discussions.

(252) IFPI also collects, through a separate exercise, revenue data directly from the four majors in order to calculate the majors' market share. This exercise is not carried out via the National Groups. Rather, IFPI centrally collects country specific revenue data from each of the majors (namely, Universal, Sony, Warner and EMI). IFPI then assigns the residual market share in respect of each territory, and each format, to the indies (collectively). For completeness, it is noted that the majors' market shares are not published but are distributed to the majors.

(253) The purpose of this exercise is to allocate correctly the majors' financial contributions to the IFPI given that the financial contributions are relative to each major's market position. IFPI is funded from three main sources of income, membership subscriptions, National Group contributions and deficit funding. Membership subscriptions and National Group contributions are insufficient to cover the annual costs of IFPI, and the shortfall each year is made up by deficit funding received from the majors. The amount of deficit funding payable by each major is calculated as to fifty per cent in equal shares, and as to the other fifty per cent by reference to the market share of each company.

7.1.2.2. GfK data

(254) GfK is an international data source for the entertainment industry. GfK collects and reports point-of-sale data.

(255) The Notifying Party provided data for the retail sales of physical recorded music compiled by GfK for two Member States (France and Germany) in the Form CO and for a further five Member States (Belgium, Ireland, Italy, the Netherlands and Spain) in the Response to the Statement of Objections. According to the Notifying Party, GfK estimates its coverage to be between 74% and 90% of the physical retail market (the coverage varying from Member State to Member State).

(256) In respect of the digital recorded music market, the Notifying Party provided data for retail sales compiled by GfK for Germany in the Form CO and for a further five Member States (Belgium, France, Ireland, Italy, the Netherlands and Spain) in its Response to the Statement of Objections. According to the Notifying Party, the GfK surveys cover a wide range of digital platforms and based on GfK's market estimates, cover between 90% and 99% of the total digital markets they relate to (the coverage varying from Member State to Member State).

7.1.2.3. OCC data

(257) The OCC gathers industry data in the United Kingdom. According to the OCC's website, it gathers sales data from retailers. The data collection is carried out via a variety of electronic methods depending on the type of retailer (general, specialist, supermarkets, independents, digital).

(258) The Notifying Party provided data for the sales of both physical and digital recorded music compiled by OCC for the United Kingdom in the Form CO. OCC estimates its coverage to be in excess of 90% of each of the physical and digital retail markets in the United Kingdom.

7.1.2.4. Platform data

(259) Another source of data in respect of the digital markets is the digital platforms themselves. The Notifying Party has provided retail revenue data from Apple/iTunes and […]* data from [digital platform]* for a number of countries.

(260) In addition, the Commission gathered wholesale revenue data for a number of countries from six digital platforms ([…]*), that is to say data showing the revenues paid by each of these platforms to record companies.

7.1.2.5. Chart and radio airplay data

(261) There are various top download, radio airplay and chart data, both published and non-published, which show a record company's position within certain set universe of songs. Depending on the compiler, different methodologies and samples may be used which explain varying results.

7.1.2.6. Market shares according to the different data sources

Physical recorded music

(262) According to IFPI data, on the market for the wholesale of physical recorded music, Universal would, post-merger, hold a market share of [40-50]*% (with an increment of [10-20]*%) in the EEA in 2010, with Sony being about half its size (with [20-30]*%) and Warner even smaller (with [10-20]*%). In 2011, on the basis of IFPI estimates, Universal would, post-merger, hold an EU-wide market share of [40-50]*% across both physical and digital (with an increment of [10-20]*%), with Sony being less than half its size [10-20]*% and Warner even smaller [10-20]*%).

(263) Post-merger, based on IFPI data, Universal would hold market shares exceeding 50% in four Member States in 2010 (Belgium, the Czech Republic, France and Greece), exceeding 40% in four Member States (Ireland, Poland, Spain and the United Kingdom) and Norway and between [30-40]* % in a further six Member States (Denmark, Germany, Italy, the Netherlands, Portugal and Sweden).

(264) In 2011, Universal would have even higher post-merger market shares in a number of countries compared to 2010. Universal's market shares would exceed 50% in three Member States (Belgium, Czech Republic and Greece), exceed 40% in eight Member States (France, Ireland, Italy, Poland, Portugal, Spain, Sweden and the United Kingdom) and Norway and between [30-40]*%in a further three Member States (Austria, Germany and the Netherlands).

Table 3 - Wholesale of Physical Music 2010 (revenues) for IFPI countries

Universal Universal EMI Sony Warner Indies + EMI

Austria

[20-30]*% [5-10]*% [30-40]*% [20-30]*% [5-10]*% [30-40]*%

Belgium

[30-40]*% [10-20]*% [50-60]*% [10-20]*% [5-10]*% [20-30]*%

Czech Republic [30-40]*% [10-20]*% [50-60]*% [20-30]*% [0-5]*% [20-30]*%

Denmark

[20-30]*% [10-20]*% [30-40]*% [20-30]*% [5-10]*% [30-40]*%

Finland

[10-20]*% [10-20]*% [20-30]*% [10-20]*% [20-30]*% [20-30]*%

France

[30-40]*% [10-20]*% [50-60]*% [20-30]*% [10-20]*% [10-20]*%

Germany

[20-30]*% [5-10]*% [30-40]*% [20-30]*% [10-20]*% [30-40]*%

[20-30]*% [60-70]*% [10-20]*% [0-5]*% [20-30]*%

Greece

Hungary

[10-20]*% [5-10]*% [10-20]*% [10-20]*% [0-5]*% [60-70]*%

Ireland

[30-40]*% [10-20]*% [40-50]*% [20-30]*% [10-20]*% [10-20]*%

Italy

[20-30]*% [10-20]*% [30-40]*% [30-40]*% [20-30]*% [10-20]*%

Netherlands

[20-30]*% [10-20]*% [30-40]*% [10-20]*% [5-10]*% [40-50]*%

Poland

[20-30]*% [10-20]*% [40-50]*% [10-20]*% [5-10]*% [30-40]*%

Portugal

[20-30]*% [10-20]*% [30-40]*% [10-20]*% [0-5]*% [40-50]*%

Spain

[20-30]*% [10-20]*% [40-50]*% [20-30]*% [10-20]*% [10-20]*%

Sweden

[20-30]*% [10-20]*% [30-40]*% [20-30]*% [10-20]*% [10-20]*%

United Kingdom [20-30]*% [10-20]*% [40-50]*% [10-20]*% [10-20]*% [20-30]*%

[20-30]*% [10-20]*% [40-50]*% [20-30]*%

Norway

[20-30]*% [5-10]*% [30-40]*% [10-20]*% [10-20]*% [20-30]*%

[20-30]*% [10-20]*% [40-50]*% [20-30]*%

EEA

Source: IFPI (including distribution deals)

(265) As regards the countries for which IFPI provides no estimates, namely Bulgaria, Cyprus, Estonia, Iceland, Latvia, Liechtenstein, Lithuania, Luxembourg, Malta, Romania, Slovakia and Slovenia, the Notifying Party provided its own best estimates of the total market size. The Commission also obtained from both Universal and EMI their actual revenues received from these countries and on this basis, calculated Universal's and EMI's market shares. The market share estimates for the wholesale of physical music for 2010 and 2011 are set out in Table 5.

Table 5 - Wholesale of Physical Music 2010 and 2011 (revenues)

Physical 2011

Physical 2010

Universal EMI Universal Universal Universal EMI + EMI + EMI

[30-40]*% [20-30]*% [0-5]*% [10-20]*% [30-40]*% [0-5]*% 40]*%

Bulgaria

[40-50]*% [20-30]*% [10-20]*% [30-40]*% [10-20]*% [20-30]*% 50]*%

Cyprus

[30-40]*% [30-40]*% [5-10]*% [0-5]*% [5-10]*% [0-5]*% 40]*% 40]*%

Estonia

[40-50]*% [40-50]*% [0-5]*% [0-5]*% [0-5]*% [0-5]*% 50]*% 50]*%

Latvia

[10-20]*% [20-30]*% [30-40]*% [0-5]*% [0-5]*% [0-5]*% 20]*%30]*%40]*%

Lithuania

[40-50]*% [40-50]*% [20-30]*% [10-20]*% [30-40]*% 80-90]*% 50]*%50]*% 30]*%20]*%40]*%

Luxembourg

Malta

[0-5]*% [0-5]*% [0-5]*% [0-5]*% [0-5]*% [0-5]*%

[10-20]*% [30-40]*% [30-40]*% [0-5]*% [10-20]*% [0-5]*% 20]*% 40]*% 40]*%

Romania

[30-40]*% [20-30]*% [20-30]*% [10-20]*% [40-50]*% 50]*% 40]*%30]*% 30]*%20]*%50]*%

Slovakia

[10-20]*% [10-20]*% [20-30]*% [10-20]*% [30-40]*% 20]*%20]*%20]*% 30]*%20]*%40]*%

Slovenia

[10-20]*% [30-40]*% [40-50]*% [0-5]*% [10-20]*% 5-10]*% 20]*% 40]*% 50]*%

Iceland

[30-40]*% [30-40]*% [30-40]*% [0-5]*% [30-40]*% [0-5]*% 40]*% 40]*% 40]*%

Liechtenstein

Source: Commission analysis based on data provided by the Notifying Party and EMI.

(266) On the basis of the Notifying Party's own estimates, Universal would, post-merger, have market shares close to or exceeding 40% in 2011 in a further ten countries (Bulgaria, Cyprus, Estonia, Latvia, Lithuania, Luxembourg, Romania, Slovakia, Iceland and Liechtenstein) in addition to those countries reported by IFPI.

(267) As explained in Sections 7.1.2.2 and 7.1.2.3, GfK and OCC data are based on music sales at the retail level (that is to say, from physical and digital music retailers to end users) rather than the wholesale level, where physical and digital music retailers pay revenues to record companies. According to GfK and OCC

data, as provided by the Notifying Party, in 2011, Universal and EMI would have a combined market share of the physical markets on the retail level of between 40% and 50% in five out of the eight Member States for which data has been provided (namely Belgium, France, Ireland, Italy and the United Kingdom).

Table 6 – Retail Sales of Physical Music 2011 (retail revenues, including distribution)

Physical 2011

Universal Sony Warner + EMI

[10-20]*% [40-50]*% [5-10]*% 20]*%

Belgium

Not provided Not provided [40-50]*%

France

[20-30]*% [30-40]*% [5-10]*% 30]*%

Germany

[20-30]*% [10-20]*% [40-50]*% 30]*% 20]*%

Ireland

Not provided Not provided [40-50]*%

Italy

[10-20]*% Netherlands [30-40]*% [5-10]*% 20]*%

[20- [10- [30-40]*% 30]*%20]*%

Spain

Not Not United Kingdom [40-50]*% providedprovided

Source: Notifying Party Response to the Statement of Objections

Digital recorded music

(268) According to IFPI data, on the market for the wholesale of digital recorded music post-merger, Universal would hold a high market share of [40-50]*% (with an increment of [10-20]*%) in the EEA in 2010 (with Sony having [20-30]*% and Warner [10-20]*%). In 2011, on a Union-wide level, Universal would hold a [40-50]*% market share across both physical and digital recorded music (with an increment of [10-20]*%).

(269) Post-merger, based on IFPI estimates, Universal's market shares would exceed 50% in 2010 in six Member States (Belgium, the Czech Republic, Greece, Poland, Sweden and the United Kingdom) and would be close to or above 40% in a further ten Member States (Austria, Denmark, Finland, France, Germany, Ireland, Italy, the Netherlands, Portugal and Spain) and Norway.

(270) In 2011, based on IFPI estimates, Universal would have a market share in excess of 50% in six Member States (Belgium, the Czech Republic, Greece, Poland, Portugal and Sweden) and would be close to or above 40% in a further ten Member States (Austria, Denmark, Finland, France, Germany, Ireland, Italy, the Netherlands, Spain and the United Kingdom) and Norway.

176 Table 7 - Wholesale of Digital Music 2010 (revenues) for IFPI countries

Universal Universal EMI + EMI

Sony Warner Indies

[20- Austria [30-40]*% [10-20]*% [40-50]*% [10-20]*% [10-20]*% 30]*%

Belgium [30-40]*% [10-20]*% [50-60]*% [20-30]*% [10-20]*% [5-10]*%

Czech Republic [60-70]*% [10-20]*% [70-80]*% [20-30]*% [0-5]*% [-0]*%

[30- Denmark [20-30]*% [10-20]*% [30-40]*% [10-20]*% [5-10]*% 40]*%

[10- Finland [30-40]*% [10-20]*% [40-50]*% [20-30]*% [10-20]*% 20]*%

[20- France [30-40]*% [10-20]*% [40-50]*% [10-20]*% [10-20]*% 30]*%

[20- Germany [20-30]*% [10-20]*% [30-40]*% [20-30]*% [10-20]*% 30]*%

[10- [40-50]*% [10-20]*% [50-60]*% [20-30]*% [5-10]*% 20]*%

Greece

+100% [0-5]*% +100% [20-30]*% [0-5]*% [-0]*%

Hungary

[20- Ireland [20-30]*% [20-30]*% [40-50]*% [10-20]*% [10-20]*% 30]*%

[10- [20-30]*% [10-20]*% [30-40]*% [30-40]*% [10-20]*% 20]*%

Italy

[30- Netherlands [20-30]*% [10-20]*% [30-40]*% [10-20]*% [5-10]*% 40]*%

Poland [60-70]*% [5-10]*% [70-80]*% [20-30]*% [0-5]*% [5-10]*%

176 Based on IFPI information on the total size of the market in Hungary and Universal's own sales data, it appears that Universal holds a share [bove 100% and the indies a share [below 0%]*. On the basis of alternative estimates provided by Universal as to the total market size in Hungary (see Form CO, Annex7.1.A, Table 7.1.A.1), post-merger, Universal and EMI would have a combined market share of around [10-20%]*. This appears to be more in line with the IFPI market share estimates for Hungary for 2011. As regards the Czech Republic, the 2010 digital market shares also seem unreliable given the indies result with a share below 0%.

Hungary Number of entries in […]*% […]*% […]*% […]*% […]*% […]*% Top 40

Italy Number of entries in […]*% […]*% […]*% […]*% […]*% […]*% Top 100

Portugal Number of entries in Top 100 […]*% […]*% […]*% […]*% […]*% […]*%

Slovakia Number of entries in […]*% […]*% […]*% […]*% […]*% […]*% Top 100

Spain Number of entries in […]*% […]*% […]*% […]*% […]*% […]*% Top 100

Sweden Number of entries in […]*% […]*% […]*% […]*% […]*% […]*% Top 100

United Kingdom Number of entries in […]*% […]*% […]*% […]*% […]*% […]*% Top 100

Source: The Notifying Party's submission of 24 April 2012.

7.1.2.7. The Notifying Party's view

(278) The Notifying Party submits that, in the case at hand, market shares are not a meaningful proxy of the merged entity's market power after the proposed concentration. In the Notifying Party's view, any traditional market definition

exercise and the resulting share data will not provide the Commission with any meaningful empirical insights into the prevailing competitive dynamics in the recorded music market. According to the Notifying Party, the principal reasons for this are twofold: first, Universal and EMI are not particularly close competitors due to EMI's limited A&R spending in recent years and focus on back catalogue, as opposed to Universal's significant A&R spend and focus on new releases; and second, market shares underplay the constraints that would be exercised over the merged entity by the presence of increasingly powerful buyers (of both physical and digital music), the fierce competition from rival record companies (both majors and indies), low barriers to entry, a declining market and the threat of piracy over legal sales of music, which in turn, strongly influences record companies' price and output decisions.

(279) In its Response to the Statement of Objections, the Notifying Party further submits that the Commission has failed to explain why, at least as a matter of principle, market shares are a useful first indication of market power in the recorded music industry. The Notifying Party claims that it is well recognised that where products are differentiated, market shares are of limited relevance to understanding the degree of competition between the merging parties. Finally, the Notifying Party claims that were the Commission to rely on any market shares to support a finding of a significant impediment to effective competition, the Commission must identify accurate market share data for any such assessment.

IFPI data

(280) The Notifying Party contests the reliability of the IFPI market share data . According to the Notifying Party, especially in relation to digital recorded music, IFPI data understates the market position of the independent labels and overstates, often to a very significant extent, the combined market share of Universal and EMI. The Notifying Party highlights certain limitations to the data.

(281) First, the IFPI data only covers some EEA countries and does not cover the following 12 EEA countries: Bulgaria, Cyprus, Estonia, Iceland, Latvia, Liechtenstein, Lithuania, Luxembourg, Malta, Romania, Slovakia and Slovenia.

(282) Second, the IFPI data clearly contains some anomalies. For example, in the data for 2010, Warner Music's share of the recorded music revenues was […]*% in both Hungary and Portugal and […]*% in the Czech Republic. While it may be the case, for example, that Warner Music has licensed the exploitation of its repertoire in some of these countries, it is also possible that these shares understate Warner Music's position and therefore overstate the shares of other

184 The Notifying Party refers to various Commission precedents such as MAN/Auwärter, Case COMP/M.2201, Recital 28, Carnival Corporation/P&O Princess, Case COMP/M.2706, HP/Compaq, Case COMP/M.2609 and IBM/Telelogic, Case COMP/M.4747.

185 This is despite stating clearly at paragraph 4.3 of Annex 6.2A to the Form CO that […]* . For the Notifying Party's arguments against the use of IFPI data, see Form CO, in particular paragraphs 373 to 380, 385 to 394 and Annex 6.2.A, the Notifying Party's Comments to the Article 6(1)(c) decision, paragraphs 3.1 to 3.16 and the Notifying Party's supplementary Submission of 10 May 2012.

record companies. Furthermore, the Notifying Party highlights other anomalies in its Response to the Statement of Objections, for example, the IFPI digital shares for the majors alone in each of Hungary and the Czech Republic in 2010 added up to over 100%.

(283) Third, the methodology used by IFPI to collect the data has limitations. Since the IFPI obtains the estimated total revenues and the estimated revenues of each major from two separate sources, there is an opportunity for a mismatch between definitions, allocations and calculations, especially in relation to digital revenues, in which revenues may be split between lump sums, such as advances, and regular on-going payments. These inconsistencies are likely to be greatly magnified in respect of smaller territories where even small absolute differences in the treatment of revenues might give rise to very significant differences in the relative magnitude of reported revenues. These errors are likely to be compounded in respect of very small segments within the overall market .

(284) In particular, the Notifying Party argues that since the IFPI's data collection cannot cover the totality of the market (and the sales excluded from the scope of the surveys are likely to be made by indies), the "coverage factor" used by the National Groups to estimate the total size of the market is nothing but a subjective guess by a limited number of IFPI staff. Furthermore, the total market size estimates do not take account of self-released music, whether in physical or digital format.

(285) This is, in the Notifying Party's view, exemplified by IFPI's recent increase in the figure for total physical sales revenues in France (by over 18%) and the attribution of the entire increase to the indies as well as in IFPI's recently revised digital sales revenues figure in Denmark (by around 25%) and the attribution of the entire increase to the indies.

(286) The Notifying Party highlights that the data published by IFPI is simply intended to show trends rather than absolute numbers and used for "benchmarking exercises" and to inform general industry commentary.

(287) As regards the "majors' market share" data collected by IFPI, the Notifying Party claims that the purpose of IFPI's data collection (calculation of financial contributions and deficit payments to IPFI) and the fact that these market shares

186 The Notifying Party has gone to great lengths in setting out the limitations of the IFPI data collection in Annex 6.2A to the Form CO and has reiterated some of these points in its Response to the Statement of Objections. In its Response to the Statement of Objections, the Notifying Party further refers to a statement from […]*

187 In its Response to the Statement of Objections, the Notifying Party quotes from IFPI's reply to the market investigation as evidencing the unreliability of IFPI figures: "IFPI collects data only from its National Groups and member record companies. No data is collected from physical or digital music retailers. National Groups report to IFPI estimated physical retail sales, based on an estimated mark-up (which includes the retailer mark-up and all taxes). The retailer mark-up is often estimated based on chart information and market knowledge. For digital sales, IFPI estimates the digital retail value for each market. […] IFPI clearly states in its reports that all retail values are estimates only, since no sales are collected directly from physical or digital retailers" (emphasis added by the Notifying Party). However, this quote refers to the retail market value estimates made by IFPI and not the trade, that is to say wholesale, value estimates. The Notifying Party also refers to a statement made by […]* .

are not published mean that these market shares are unsuitable for any market share assessment in this case. The Notifying Party submits that given these factors, neither IFPI nor its members have any rational incentives to invest significant amount of resources and time to ensure that these shares are in fact relatively precise. According to the Notifying Party, what matters is that they are broadly of the right order of magnitude so as to ensure that a major which appears to grow its revenues also contributes a bit more to the IFPI's deficit funding requirements.

(288) In support of its arguments, the Notifying Party further relies on a statement from […]* attached to its Response to the Statement of Objections.

GfK and OCC data

(289) The Notifying Party submits that for those countries for which GfK and OCC data are available, those data should be used. In the Notifying Party's view, the GfK and OCC data are based on a comprehensive assessment of all recorded music sold through an extensive and representative range of both physical and digital retailers. According to the Notifying Party, for those countries for which GfK or OCC is not available, the Commission cannot rely on IFPI figures but must reconstruct the markets.

(290) When comparing IFPI figures with those provided by GfK and the OCC, the Notifying Party highlights France as an example where sources other than IFPI give wholly different market share estimates. In 2010, for example, IFPI estimated Universal's physical market share (excluding distribution deals) to be [30-40]*% whilst GfK estimated it to be [20-30]*%. Similarly the Notifying Party refers to the difference between the OCC estimate [30-40]*% and the IFPI estimate [30-40]*% in the United Kingdom for the digital recorded music market in 2010 (with the indies share being [10-20]*% according to the OCC as opposed to [10-20]*% according to IFPI). For Germany, the Notifying Party refers to the GfK estimate showing Universal as having a digital market share of [20-30]*% in 2010 as opposed to [20-30]*% according to IFPI (however, the indies' share is exactly the same according to both GfK and IFPI).

Platform data

(291) The Notifying Party submits that as regards the wholesale of digital music, a more relevant data set by which market shares should be calculated is data derived directly from digital platforms, in particular from Apple/iTunes and [digital platform]*. The Notifying Party claims that those data are far more reliable than IFPI estimates since they reflect actual sales rather than estimates.

(292) The Notifying Party has provided iTunes retail sales data and [digital platform]* data (that is to say, data based on […]*) on the basis of which Universal's and EMI's market shares are significantly lower than according to IFPI data and the indies' market share, by comparison, significantly higher than according to IFPI data.

(293) The Notifying Party claims that the iTunes and [digital platform]* data are all the more compelling given that these two services accounted for […]*% of

Universal's EEA-wide digital revenues in 2011. The Notifying Party considers that the iTunes and [digital platform]* data are also the best proxy for revenue share figures for other […]* services. In its Response to the Statement of Objections, the Notifying Party again highlights the fact that, in its view, the starting point for calculating any digital market shares across the EEA must be iTunes as iTunes is active in a significant and diverse range of Member States, accounts for […]* of Universal's and EMI's total digital revenues in most EEA markets in which it is active and is a long-established retail platform. To ignore the iTunes share data would be to disregard the single largest empirically robust dataset available to the Commission and in the absence of compelling reasons and a better dataset, any such decision would, in the Notifying Party's view, not be justifiable.

(294) Any argument that the iTunes and [digital platform]* datasets (or indeed the datasets of GfK and OCC) do not cover a sufficiently large part of the overall market could, in the Notifying Party's view, be dealt with by the Commission by building up these datasets with data received from third parties via requests for information, that is to say, by reconstructing the market from the bottom up. The Notifying Party notes that the Commission has indeed collected data from six digital platforms which in principle are capable of extending the "actual" sales based iTunes dataset in that it covers additional and diverse digital platforms. However, in the Notifying Party's view, the GfK and OCC data remains more comprehensive and therefore more robust.

(295) The Notifying Party also makes the point that if the Commission were to consider IFPI data as reliable because they cover a large proportion of the market, iTunes data are equally representative since they also cover a large proportion of the market.

(296) As regards [digital platform]* data, the Notifying Party argues that […]* shares on [digital platform]* in fact allow a cross-check of the robustness of other market share data. In particular, the Notifying Party notes that […]*. Furthermore, shares […]* are a good proxy for shares […]* in the digital space since there is a close correlation between average shares […]*, and shares […]* on the iTunes service. According to the Notifying Party, the shares on [digital platform]* are most relevant as it is […]* digital music service in a number of […]* countries.

(297) When comparing the iTunes and [digital platform]* data with the other data sources, the Notifying Party notes the following:

(a)There is a very close correlation between the GfK and OCC share data, on the one hand, and stand-alone iTunes data, on the other. More specifically, […]*;

(b)A comparison of the closeness of the correlation between GfK and OCC data with iTunes and IFPI data respectively clearly establishes a much

closer correlation of GfK and OCC data with the former (iTunes) than the latter (IFPI);

(c)As regards the shares calculated by the Commission across the six digital platforms, the Notifying Party notes that […]*;

(298) In the Notifying Party's view, GFK and OCC datasets are therefore a reliable proxy for total market shares in the Member States covered by these surveys. As regards those countries for which GfK and OCC data are not available, the GfK and OCC data strongly supports the use of iTunes data as the second best but still reliable proxy for total market size shares. In conclusion, the Notifying Party submits in its Response to the Statement of Objections that the Commission should place no reliance on IFPI market shares in the assessment of the proposed concentration but should, in the first instance, rely on GfK and OCC share data, and, on a secondary level, iTunes and [digital platform]* data.

Chart and radio airplay data

(299) The Notifying Party argues that other proxies such as chart success or radio airplay fail to give any indication of market power since chart success is unpredictable, constantly changing and often short lived and there is no clear correlation between chart success and sustained commercial success. As regards airplay, record companies have no control or influence over airplay and as such, airplay data cannot be a measure of record companies' market power.

7.1.2.8. The Commission's assessment

Introduction

(300) The Commission reiterates the point that market shares are but one proxy for market power. As stated in the Horizontal Merger Guidelines, market shares and concentration levels provide useful first indications of the market structure and of the competitive importance of both the merging parties and their competitors but other factors must also be taken into account .

(301) The Notifying Party's arguments as to why market shares are not a meaningful proxy in this case will be dealt with in Sections 7.1.3, 7.1.6 and 7.1.7. In this Section, the Commission focuses on what the market shares unequivocally show.

The reference to market shares throughout the case

(302) In its Article 6(1)(c) decision, the Commission set out the majors' market shares on the basis of IFPI data and for non-IFPI countries, on the basis of the total market size estimates provided by the Notifying Party and actual revenue figures received from each major. The Commission also highlighted the fact that it was

189 Horizontal Merger Guidelines, paragraph 17. See also, for example, Case T-282/06, Sun Chemical Group and Others v Commission, [2007] ECR II-2149, paragraphs 134, 135 and 136, Case T-342/07, Ryanair Holdings v Commission, [2010] ECR II-3457, paragraph 41 and Case T-282/02, Cementbouw Handel & Industrie v Commission, [2006] ECR II-319, paragraphs 200 and 201.

clear from the replies to the market investigation that currently the main competitors of Universal are the three other major record companies (with the indies only exerting a limited competitive pressure mainly due to their smaller repertoire) but that post-merger, the largest remaining major, namely Sony, would be approximately half the size of the merged entity, and the other major, Warner, would be considerably smaller than Sony in both the EEA and in the majority of Member States. Ultimately, for the purposes of the Article 6(1)(c) decision, the Commission considered that the proposed concentration could lead to competition concerns on the wholesale markets for physical recorded music in several Member States and for digital recorded music in the EEA and in several Member States, where the merged entity's market shares would be 40% or more.

(303) In the Statement of Objections, the Commission further emphasised the inferences it drew from market shares as to Universal's market position if the proposed concentration were to go ahead. Although the Commission was of the preliminary view that IFPI was, overall, the most comprehensive, reliable and the most widely used dataset and there were no reasons to discard IFPI market shares, the Commission highlighted the fact that since market shares are used as an indication of market power rather than proof of such power, the exact market share (whether or not above or below a certain level) is not determinative in the analysis of the effects of the proposed concentration. In line with the Horizontal Merger Guidelines, a significant impediment to effective competition is likely to arise in cases where the merger would create or strengthen the dominant position of a single firm, one which, typically, would have an appreciably larger market share than the next competitor post-merger. It is therefore the indication that market shares give of the overall post-merger strength of the merged entity and its relative position to its competitors after the merger that drives the Commission's competitive assessment in this case.

(304) The Commission considers the relevance of the different data sources for this assessment to be as follows.

The relevance of the various data sources

(305) IFPI, GfK and OCC data relating to physical recorded music. The Commission considered all the various market share data sources and their reliability and relevance. As regards the IFPI data, the Commission concludes that, despite certain limitations, the IFPI data is, overall, a comprehensive and widely used dataset. Although some anomalies exist, there are several reasons why this is not sufficient to discard all IFPI data.

(306) First, contrary to the Notifying Party's claim, IFPI data is widely used by all market participants, including the Notifying Party itself, many different contexts, including the following:

190 Horizontal Merger Guidelines, paragraph 25.

(a)All the majors, including Universal, seem by and large content with using the market shares calculated by IFPI in the context of apportioning the amount of the financial contributions they are to make to IFPI;

(b)Market participants, rely on IFPI market shares when arguing for their share of industry-wide piracy settlements ;

(c)Internal documents show that Universal and EMI rely on IFPI market shares, both in relation to the proposed concentration and in the normal course of business. […]*

– […]* ; and

– […]* ;

(d)Finally, and most critically, the majors tend to refer to IFPI market shares when negotiating with their customers.

(307) Regarding the reliability of the data itself, representatives of the Notifying Party have themselves approved the Data Definition Guidelines that guide the calculation of IFPI market shares and, according to […]*, each major also approves its own IFPI data prior to the publishing of such data. Furthermore, as far as IFPI is aware, it has never received any complaint that its published industry numbers are generally unreliable .

192 See http://www.vivendi.com/wp-content/uploads/2011/11/umg-investor-presentation-november-2011-final-2.pdf , in particular slides 4, 6 and 7.

193 – […]* ;

194 See […]* submission of 31 May 2012, (ID: 7182 ).

195 […]* submission of 1 May 2012, in response to Commission's question 6 to […]*.

(308) The Commission considers that the fact that Universal and other participants would also use other data sources and would not solely refer to IFPI in their business decisions does not demonstrate that IFPI market shares are without any relevance. The Commission considers that this shows that market participants will consider their own market positions on the basis of a range of data available given that ultimately, all industry sources are estimates. If anything, it shows that IFPI data are also used to determine the market positions of the different players on the relevant markets.

(309) The Commission is also not convinced by the Notifying Party's argument that the fact that Universal signs off on the various IFPI guidelines for the calculation of market shares or the absence of complaints on their reliability have no relevance when judging the general reliability of these shares. If these figures were completely inaccurate (including the published total market size estimates), it is, in the Commission's view, highly unlikely that these would be approved. It is also clear from the statements of […]* that the Notifying Party quotes that if data is publicly released, there are rigorous checks and audit processes that must be applied.

191 See […]* submission of 31 May 2012, (ID: 7182).

(310) Furthermore, the purpose of the IFPI published total market size figures, even were it simply to show trends and be used as a benchmarking tool as the Notifying Party claims, does not automatically undermine the accuracy of that data. Equally, the statements of […]* must be balanced against the views of the third parties who use and believe in the general accuracy of the IFPI data.

(311) Importantly, the Notifying Party does not contest that […]*. The Notifying Party cannot credibly claim that IFPI data can be used for all other purposes but not for assessing Universal's and EMI's market position in the context of the competitive assessment of the proposed concentration.

(312) The Commission has relied on IFPI market share data in all of its previous decisions in the recorded music sector, most recently in Sony/SonyBMG. The Sony/BMG decision also relies on IFPI's Recording Industry in Numbers for both physical and digital recorded music market shares. No market participant has ever complained against or challenged the use of IFPI data before.

196 Commission's decision of 8 August 2008 in Case No COMP/M.5272, Sony/SonyBMG, Recitals 52 and 53.

197 Commission's decision of 3 October 2007 in Case No COMP/M.3333, Sony/BMG; see, for example, Recitals 84, 156, 409, 442 and footnote 96.

198, (313) The Commission considers it irrelevant that in Sony/BMG IFPI figures were used in a decision that focussed on coordinated rather than non-coordinated effects. Since market shares give an indication of market structure, a factor which is equally relevant for assessing coordinated effects, the fact that the Sony/BMG decision did not focus on non-coordinated effects is not sufficient to argue that the use of IFPI market shares in that decision is irrelevant.

(314) Finally, the results of the market investigation support the use of IFPI data. In response to a specific question by the Commission to competitors as to whether the IFPI data accurately reflects the performance of the major record companies, a wide majority of respondents considered this to be the case. Of the two that did not, one considered the IFPI data to underestimate the majors' market performance, rather than overestimating their performance as the Notifying Party claims.

199 Competitors' reply to the Commission Request for information of 17 April 2012, question 25. Nine out of eleven respondents who provided a non-confidential version of their reply confirmed that IFPI data accurately reflect the majors' market performance. These companies are […]*, […]*, […]*, […]*, […]*, […]*, […]*, […]* and […]*. […]* considered that the IFPI data do not accurately reflect the majors' market performance, but rather underestimate the performance of the majors. Only one company […]* considered that the IFPI data do not accurately reflect the performance of the majors (but it did not consider that the data underestimate or overstate the majors' position).

(315) The Commission is not convinced by the Notifying Party's argument that the endorsement of the use of IFPI data by third parties in the market investigation is of no value unless those third parties are aware of all the deficiencies in collecting those data. During the Commission's investigation, it emerged that […]*, […]*, IMPALA and […]* all support the use of IFPI data. The

Commission cannot simply disregard the fact that major industry participants all state that IFPI figures are generally the best available estimates of market shares. According to […]* the arguments put forward by the Notifying Party are nothing but a campaign to discredit IFPI figures in order to "argue away the inconvenient truth of its dominant market shares" .

(316) […]* stated that it believes "IFPI figures to be generally the best available estimates of market shares. […]* would agree that IFPI figures are the most comprehensive, reliable and the most used market share figures currently available and confirms that it is not aware of any better data available. As far as […]* is aware, IFPI’s data collection guidelines and market shares are approved by all the majors" .

(317) IMPALA further points out that IFPI market shares are used in many instances to calculate revenue, or shares of shelf space, or editorial space in relation to retailers and IMPALA would be surprised if IFPI data were not therefore accurate.

(318) The Commission considers that, even if the IFPI data were not accurate, they are perceived as accurate enough by the industry to be used widely in all commercial contexts. Therefore, at the very minimum, the IFPI data show what the markets believe the merging parties' market share position is. This is the key factor in the assessment of the merged entity's post-merger market position.

(319) Further, the Commission considers that it would not be appropriate to use the data put forward by the Notifying Party such as the GfK and OCC data for the eight Member States for which the Notifying Party has been able to provide this information, to assess Universal's post-merger market position. Both the GfK and OCC data are retail sales data and not wholesale data. As the relevant markets for the Commission's assessment of the proposed concentration are the wholesale markets for the sale of physical recorded music, the GfK and OCC data do not address the correct market level. The Commission considers that higher wholesale revenues for a given selection of recorded music products may not translate into higher retail revenues for those same products. The retailers may make different margins on different products and, as such, the share of a record company of retail sales revenues does not necessarily reflect its share of wholesale revenues.

(320) In any event, for the majority of the eight Member States, the difference between the IFPI data on the one hand and the GfK and OCC data on the other is not material. Moreover, even for the Member States for which these differences are more marked, the data submitted by the Notifying Party still confirms that, post-merger, Universal would be the market leader and would be significantly larger than any of its competitors. Since as explained, the Commission's competitive assessment in this case is not premised on a certain market share threshold, these differences in market share levels would not undermine the soundness of the Commission's competitive assessment.

200 See […]*'s comments to the Statement of Objections (ID 8375).

201 (ID 9554).

(321) Summarising the differences in respect of the eight Member States in terms of the combined share of Universal and EMI, the Commission notes the following:

(a)Belgium: both the IFPI [50-60]*% and GfK [40-50]*% estimates give a high market share for the combined entity;

(b)France: there is one percentage point of difference between the IFPI [40-50]*% and the GfK [40-50]*% estimate;

(c)Germany: there is one percentage point of difference between the IFPI [30-40]*% and the GfK [30-40]*% estimate;

(d)Ireland: there is one percentage point of difference between the IFPI [40-50]*% and the GfK [40-50]*% estimate;

(e)Italy: both the IFPI [40-50]*% and GfK [40-50]*% estimates give a high market share for the combined entity;

(f)The Netherlands: there is no difference in the IFPI [30-40]*% and GfK [30-40]*% market share estimates;

(g)Spain: there is a difference of two percentage points between the IFPI [40-50]*% and the GfK [40-50]*% estimate; and

(h)The United Kingdom: there is no difference in the IFPI [40-50]*% and GfK [40-50]*% market share estimates.

(322) In light of the above, the Commission considers that there are no reasons to discard the IFPI market shares when carrying out its competitive analysis of the proposed concentration in the markets for the wholesale of physical recorded music. Especially given that market shares are merely used as an indication of market power and the relative position of the merged entity in relation to its competitors, the Commission considers that IFPI data are sufficiently reliable and that it not necessary to take a final position on the relevance of single-digit differentials in market share estimates.

(323) IFPI, GfK, OCC and platform data relating to digital recorded music. As regards IFPI data, although some anomalies exist, that is not a sufficient reason to discard all IFPI data relating to the digital recorded music markets for the same reasons as set out in Section 7.1.2.8 in relation to the physical recorded music data, including, in particular, the fact that record companies use IFPI shares when negotiating with digital platforms as to their share of, amongst others, a relevant digital platform's advertising inventory or revenues.

(324) Regarding the data provided by Gfk and OCC, as is the case regarding physical recorded music, the data relate to retail sales and do not therefore relate to wholesale revenues. Given that the relevant markets are the wholesale markets for the sale of physical and digital recorded music, the GfK and OCC data therefore do not address the correct market level.

(325) In any event, similar to the situation in relation to physical recorded music, apart from Belgium and Spain, in the Member States for which digital data has been

provided the difference between the IFPI data on the one hand and the GfK and OCC data on the other is not material.

(326) Summarising the differences in respect of the seven Member States in terms of the combined share of Universal and EMI, the Commission notes the following:

(a)Belgium: although the IFPI [60-70]*% estimate is materially higher than the GfK [30-40]*% estimate, even the GfK estimate gives a relatively high market share for the combined entity, in particular when compared to GfK's estimate for the next largest players Sony [10-20]*% and Warner [10-20]*%;

(b)Germany: there is no difference in the IFPI [30-40]*% and the GfK [30-40]*% estimates;

(c)Ireland: there is one percentage point of difference between the IFPI [40-50]*% and the GfK [40-50]*% estimate;

(d)Italy: there is a difference of three percentage points between the IFPI [40-50]*% and the GfK [30-40]*% estimates;

(e)The Netherlands: there is one percentage point of difference between the IFPI [30-40]*% and the GfK [30-40]*% estimates;

(f)Spain: although the IFPI [40-50]*% estimate is materially higher than the GfK [30-40]*% estimate, even the GfK estimate gives a relatively high market share for the combined entity, in particular when compared to GfK's estimate for the next largest players Sony [20-30]*% and Warner [10-20]*%; and

(g)The United Kingdom: there is a difference of three percentage points between the IFPI [40-50]*% and GfK [40-50]*% market share estimates.

(327) Turning to platform data, Table 17 sets out the proportion of digital revenues that Universal and EMI each receive from iTunes.

Table 17 - iTunes share of total Universal and EMI recorded music digital revenues (%) in 2011/FY 2012

iTunes share of totaliTunes share of totaliTunes share of totaliTunes share of total Universal RM digitalUniversal RM digitalEMI RM digitalEMI RM digital revenues (%) forrevenues (%) forrevenues (%) forrevenues (%) for 2010 2011 FY2011 FY2012

Austria

[…] *

Belgium

[…] *

Bulgaria

[…] *

Cyprus

[…] *

Czech Republic

[…] *

Denmark

[…] *

Estonia

[…] *

Finland

[…] *

France

[…] *

Germany

[…] *

Greece

[…] *

Hungary

[…] *

Ireland

[…] *

Italy

[…] *

Latvia

[…] *

Lithuania

[…] *

Luxembourg

[…] *

Malta

[…] *

Netherlands

[…] *

Poland

[…] *

Portugal

[…] *

Romania

[…] *

Slovakia

[…] *

Slovenia

[…] *

Spain

[…] *

Sweden

[…] *

United Kingdom

[…] *

Iceland

[…] *

Norway

[…] *

Liechtenstein

[…] *

EEA

Source: the Notifying Party and EMI

(328) From Table 17, it emerges that, for the EEA, as well as for the vast majority of Member States, it would be misguided to attempt to calculate a record company's market share on the relevant national market for the wholesale of recorded music based on such record company's share of the total sales on iTunes alone. This reflects the fact that iTunes does not represent the totality, or even near the totality, of the sales of each of Universal and EMI and, therefore, these companies' share of sales on iTunes cannot constitute a good proxy for their market position on the total digital market in the relevant Member State.

(329) Moreover, even in those Member States where iTunes does represent close to the totality of the sales of either Universal or EMI, namely […]*, this is only the case for one of the two merging parties. By way of example, in […]*, where iTunes accounts for more than […]*% of Universal's total digital revenues, EMI does not appear to achieve any revenues from iTunes (or those data have in any event not been provided). Conversely, in [...]*, where iTunes accounts for […]*% of EMI's total digital revenues, it only accounts for […]*% of Universal's digital revenues. Universal's largest digital customer in Hungary is […]* (representing […]*% of Universal's digital revenues), whereas EMI […]* in Hungary. Similarly, in Ireland, EMI only represents […]*% of Universal's total digital revenues, while it accounts for […]*% of EMI's such revenues. As a result, even in these Member States, a calculation of the merging parties' market share based on iTunes data would be misguided and inevitably flawed.

(330) Indeed, in order for iTunes to constitute a good proxy for record companies' market position on the total digital market in the relevant country, iTunes would have to account for the vast majority of the digital market in that country, which, in turn, cannot be the case if it does not account for the vast majority of the digital sales of the merging parties' in that country. For iTunes to represent the vast majority of the digital market in any given country without also accounting for the vast majority of the sales of the merging parties in that country, there would have to exist in each such country other digital services carrying only the repertoire of one or both of the merging parties and accounting for a significant percentage of their sales, but, at the same time, not being big enough to have a significant presence potentially threatening the representativeness of iTunes data. This assumption is simply not borne out by the facts.

(331) The only possible exception to the findings in Recitals 328 to 330 is Belgium, where iTunes consistently accounts for more than […]*% of both Universal's and EMI's total digital revenues. Even based on the Notifying Party's iTunes data for Belgium, the merged entity would have a market share of [40-50]*%. Moreover, based on the data gathered by the Commission, it appears that the merged entity would be the uncontested market leader after the proposed concentration in all Member States, including Belgium.

(332) In any event, regardless of the precise figures, what Table 17 clearly shows is that iTunes is not representative of the whole digital market at EEA level and in almost all Member States and Norway. Furthermore, it is clear that the share of revenues that iTunes represents for each record company varies from year to year, […]*. On an EEA-wide level, iTunes' share of EMI's digital revenues […]*. Therefore, with the possible exception of Belgium, any "market shares" calculated on the basis of a company's revenue share of iTunes are not comparable given that for one company iTunes may represent a much larger (or smaller) proportion of its total digital revenues.

(333) In light of the above, the Commission disagrees with the Notifying Party's claim that iTunes data are equally representative as IFPI data merely because they also cover a significant proportion of the market.

(334) In this respect, the Commission considers that IFPI data nevertheless aim to cover a much larger proportion of the digital market than iTunes data alone aim to do, by including sales made by all record companies to all digital platforms. Furthermore, there is no evidence that iTunes represents a similar percentage of competitor's revenues as it does for Universal or EMI (the fact that iTunes represents different proportions of revenue for both Universal and EMI shows that its relevance to a record company will vary from company to company).

(335) As regards the [digital platform]* data submitted by the Notifying Party, the Commission notes that they are […]* data, that is to say, […]*. As stated in the Commission Notice on the definition of the relevant market for the purposes of Community competition law ("Commission Notice on Market Definition") : "[i]*n cases of differentiated products [and recorded music certainly is a differentiated product from the point of view of retailers and end users], sales in value will usually be considered to better reflect the relative position and strength of each supplier" .

(336) The Commission also takes the view that, for the wholesale of recorded music, royalty-based market shares are a much better proxy of recorded music companies' market power (of the downstream market) than […]* market share. This reflects the fact, among other things, that a recorded music company's market power typically translates into its ability to extract revenues from its customers. These revenues could be extracted either by increasing the number of its songs played by the service or by securing payments from digital services (for example, in the form of advances or minimum guarantees), which are not proportionate to the number of songs played on the service. The […]* data provided by the Notifying Party for [digital platform]* fails to capture this dimension and therefore does not constitute a meaningful proxy of a recorded music company's market power when dealing with […]* platforms at wholesale level.

(337) Furthermore, the same concerns that are raised regarding the representativeness of iTunes' market shares relate to [digital platform]* figures since despite being […]* service in the EEA, with few exceptions, [digital platform]* still accounts for a limited percentage of both Universal's and EMI's total digital sales in all the Member States for which data is provided. Moreover, [digital platform]* accounts for […]* different percentages of total digital sales of both Universal and EMI in the vast majority of Member States.

(338) The Commission also collected and analysed certain revenue data from digital platforms for the purposes of testing its bargaining theory of harm (see Section 7.1.5). This dataset includes royalty data, that is to say, the actual revenues paid by the relevant digital platform to each recorded music company, for several platforms, which makes it a more representative sample of the whole digital market in the relevant territory than the Notifying Party's revenue data, which was only based on iTunes figures (as explained, the [digital platform]* figures

202 OJ C 372, 09.12.1997, p. 5.

203 Commission Notice on Market Definition, paragraph 55.

204 See Section (241).

submitted by the Notifying Party are […]*). Therefore, the royalty shares that can be computed on the basis of the Commission's own collected dataset can be a useful test to verify the reliability of the IFPI data for the wholesale of digital recorded music relative to the parties' iTunes figures.

(339) The Commission notes that the royalty share data it computed based on the available dataset, which were made available in full to the Notifying Party's economic advisors in the context of the data room exercise following the issuing of the Statement of Objections, as well as to the Notifying Party itself in a redacted non-confidential version on 2 July 2012 and 4 July 2012, clearly show that the inclusion in the dataset of additional platforms (in particular streaming

205 In this context, it is important to note that the Notifying Party itself submits that non-ownership models, including streaming and subscription services, most accurately reflect the likely future format of the market given their current strong and sustained growth. See paragraph 376 of the Form CO.

and mobile platforms)to iTunes increases the merged entity's share compared to the iTunes figures to a level that is closer to the IFPI numbers. In the Commission's view, this observation shows that the market shares estimated based on iTunes data provided by the Notifying Party are likely to underestimate the merged entity's market position.

(340) The Notifying Party notes that in the majority of countries included in the Commission's dataset, the merged entity's share is still closer to the iTunes shares than to the IFPI shares or is at least half- way between the two. In this respect, the Commission notes that the purpose of its exercise was not to calculate precise market shares on which to rely but rather, to show that adding other platforms to the iTunes data consistently increases the market share of the combined entity from that seen purely on iTunes. It must be remembered that for most countries, the Commission did not receive data from all six platforms and as such, its dataset by no means purports to cover the totality of the digital markets but rather shows a trend whereby the iTunes shares increase consistently when platforms are added.

(341) Chart and radio airplay data. The Commission has also considered a number of other proxies for market power, such as chart success or radio airplay, which support the market share data available from other sources. Whilst the various top download, airplay and chart data sets all give slightly varying results (which is normal given that slightly different samples and methodologies have been used in compiling these data), these data sets nevertheless consistently show a combined Universal and EMI having a high share of either chart hits or radio airplay or of both.

(342) The Notifying Party argues that other proxies such as chart success or radio airplay fail to give any indication of market power . In the Notifying Party's view, chart success is unpredictable, constantly changing and often short lived and there is no clear correlation between chart success and sustained commercial success. As regards airplay, record companies have no control or influence over airplay and as such, airplay data cannot be a measure of record companies' market power.

206 Response to the Statement of Objections.

(343) Contrary to the Notifying Party's arguments, chart success and radio airplay also constitute a proxy for market power in the recorded music sectors. This essentially reflects the fact that retailers, whether physical and digital, need to have chart hits to attract consumersand radio airplay is both a proxy to measure the share of hits held by a record company and itself a means to drive the success of an artist or a song.

207 Customers' reply to the Commission Request for information of 17 February 2012, question 43; Customers' reply to the Commission Request for information of 17 April 2012, question 23; Competitors' reply to the Commission Request for information of 17 February 2012, question 77.

(344) Available data from different sources that are summarised in Section 7.1.2.6 clearly show that the merged entity would hold shares between [40-50]*% and [60-70]*% based on these indicators in most of the Member States, where, based on IFPI data, it would already hold a high market share in the wholesale of physical or digital recorded music, namely: the Czech Republic, Denmark, France, Germany, Italy, Portugal, Slovakia, Spain, Sweden and the United Kingdom, as well as at EEA level.

The different market players' relative position

(345) Pursuant to the Horizontal Merger Guidelines, the merging parties' market shares must be considered in the context of other factors, notably the strength and number of competitors . The Commission follows this approach in this case in that it treats the merged entity's and its competitors' market share levels as giving first indications of the post-merger strength of the merged entity and its position in relation to its competitors.

208 Horizontal Merger Guidelines, paragraph 17.

(346) In this context, the Commission considers that the indies should not be viewed as a single entity and therefore their market share cannot be considered as that of one and the same market player. Independent record companies are a fragmented group including hundreds of different labels, they typically negotiate individually with retailers and therefore their negotiating power in relation to retailers cannot be equated to that of a major. Moreover, retailers are perfectly able to operate a viable offering without the content of many of the indies, while many of them would likely be unable to survive without the content of all majors. In other words, the indies' marginal utility to retailers cannot be compared to that of a major.

(347) The Commission considers that these considerations equally apply to the physical and the digital wholesale recorded music markets, despite the Notifying Party's argument that, in the digital markets, Merlin should be considered akin to a "fifth major". Indeed, the Commission considers that Merlin cannot be viewed akin to a fifth major for the following reasons:

(a)Merlin does not have agreements with the largest music download services in the EEA, namely iTunes and Amazon MP3. In addition, it has not been able to gain access for its members to a number of digital services it has approached, such as those operated by Vodafone, Telefonica and Nokia;

(b)Merlin’s activities are carried out on behalf of its members only, so any benefits it provides are not for all independent labels, but Merlin member labels only;

(c)due to the ability of its members to opt in or out of each agreement Merlin proposes to enter into, Merlin can only operate on a deal by deal basis, and therefore it can only provide its members with any benefits on a case by case basis;

(d)Merlin’s resources are extremely limited. As such, it does not have the resources to enter into every business opportunity with which it is presented or which may be attractive; and

(e)many aspects of the independent record companies' dealings with digital services, including marketing and promotion, are carried out by each of the indies on an entirely individual basis, independently of Merlin.

(348) Against this background, the Commission compared the post-merger market shares of the merged entity with those of the main competitors that would remain after the proposed concentration. For this purpose, the Commission carried out a relativity exercise between the four majors on the basis of revenues reported by each major for each country in question (that is to say, on a hypothetical market consisting only of the four majors) in order to assess Universal's and EMI's market position in relation to the other majors. This exercise was carried out separately in respect of both physical recorded music and digital recorded music and in respect of IFPI and non-IFPI countries.

(349) This exercise proved particularly useful in relation to those countries for which no data sources were available other than the Notifying Party's own estimates. While the Commission considers that the Notifying Party would not have provided completely unrealistic total market size estimates, the Commission nonetheless considers that these market share estimates are likely to be completely accurate .

209 The key issues surrounding the Notifying Party's market share estimates and their comparability from year to year can be summarised as follows (i) Country-by-country revenue data may be flawed: Universal has local operating companies in Bulgaria and Romania. Universal does not have local operating companies in Cyprus, Estonia, Latvia, Lithuania, Luxembourg, Malta, Slovakia, Slovenia, Iceland and Liechtenstein. The revenues generated from sales to customers located in these countries are generated through activities managed by local operating companies in other countries […]*. Universal has sought to allocate its revenues to these smaller countries to the extent possible but in some cases has been unable to do so on any meaningful basis. EMI has similarly allocated its revenue data to every EEA Contracting Party to the best of its ability. There may therefore be anomalies already in the revenue data of the parties for any given country. Finally, EMI […]* which means that whatever revenue EMI has recorded for these countries is not a true reflection of EMI's repertoire sales in these countries; (ii) Total market size estimates not comparable for 2010 and 2011: For 2010, for Bulgaria, Romania and Slovakia, the Notifying Party has applied its market share in a neighbouring country to its revenues in order to extrapolate the total market size estimate. For the other countries, the Notifying Party has, via various steps, used its market share in a neighbouring country as a proxy, but at the same time taking into account the fact that it is not generally active in domestic repertoire in countries where it does not have a local operating company (see Form CO, Annex 7.1.A., Table 7.1.A.1. for a more detailed

(350) The relative position of the Notifying Party in the physical recorded music market. Tables 18 and 19 set out the comparison of the market shares of the merging parties and its major competitors in relation to physical recorded music.

Table 18 –Relative position of the majors in the markets for the wholesale of physical recorded music in countries reported by IFPI (2011)

Sony

Warner

[40-50]%

[0-10]%

[30-40]%

[0-10]%

Austria

[40-50]% [20-30]% [20-30]%

[0-10]%

Belgium

[30-40]% [20-30]% [30-40]%

[0-10]%

Czech Republic

[40-50]% [10-20]% [30-40]%

[0-10]%

Denmark

[30-40]% [10-20]% [20-30]% [20-30]%

Finland

[40-50]% [10-20]% [20-30]% [10-20]%

France

[30-40]% [10-20]% [30-40]% [10-20]%

Germany

[40-50]% [10-20]% [30-40]%

[0-10]%

Greece

[30-40]% [10-20]% [40-50]%

[0-10]%

Hungary

[20-30]% [20-30]% [30-40]% [10-20]%

Ireland

[20-30]% [10-20]% [30-40]% [10-20]%

Italy

[40-50]% [20-30]% [20-30]% [10-20]%

Netherlands

[30-40]% [30-40]% [20-30]% [10-20]%

Poland

[30-40]% [30-40]% [30-40]%

[0-10]%

Portugal

[20-30]% [10-20]% [30-40]% [20-30]%

Spain

[30-40]% [10-20]% [20-30]% [10-20]%

Sweden

United Kingdom [30-40]% [20-30]% [20-30]% [10-20]%

[30-40]% [10-20]% [20-30]% [10-20]%

[40-50]% [10-20]% [20-30]% [10-20]%

Norway

EEA

[30-40]% [10-20]% [20-30]% [10-20]%

explanation) As regards the year 2011, however, the Notifying Party changed its methodology. Instead of using neighbouring countries as a proxy, it used its share on iTunes as a proxy for estimating the total market size (for both physical and digital) in Bulgaria, Cyprus, Estonia, Latvia, Lithuania, Luxembourg, Malta, Romania, Slovakia and Slovenia. IFPI does give total market size estimates for Bulgaria but Universal does not consider these reliable and has given its own market size estimate. For Iceland and Liechtenstein, the neighbouring country's iTunes share was taken as proxy.

(374)The Notifying Party submits that the proposed concentration would combine the two least similar traditional majors. Universal's focus is on new signings and new releases whereas EMI would be more focused and dependent on its back catalogue (that is to say, releases that are at least 18 months old). According to the Notifying Party, for the financial year 2011, EMI derived […]*% of its revenues from back catalogue sales, whereas Universal derived only […]*% of its revenues from back catalogue sales in 2010. EMI has faced a steady decline in its new release revenues from […]*% in financial year 2007 to […]*% in financial year 2011. Universal, on the other hand, has maintained a high proportion of new release revenues, down from […]*% in 2006 to […]*% in 2010 .

211 See Case T-282/02, Cementbouw Handel & Industrie v Commission [2006] ECR II-319, paragraph 201. 212 The typical characteristics of majors and independents are explained in Section 6.1.2.1 of this Decision. 213 See paragraphs 3 and 901 of the Form CO.

105

(375)According to the Notifying Party, EMI's greater focus on exploiting its back catalogue at the expense of investing in new artists and new sound recordings is also demonstrated by EMI's reduction of its roster of artists by around […]*% between 2007 and 2010 and the significant decrease in its spending on A&R in recent years. For example, EMI's gross spend on A&R (gross advances spend) was around […]* during the financial year ending 31 March 2008 and fell to around […]* for the financial year ending 31 March 2011. According to the Notifying Party, with its reduced focus on a limited number of new signings and given the overall reduction of its A&R spend, […]*, EMI would today be closer in its competitive proposition and strategy to the leading indies in Europe than to any of the other three traditional majors .

(376)In this context, the Notifying Party submits that in recent years, EMI has developed its creative and brand management credentials to offer a compelling "value-add" proposition in the absence of a general willingness to pay significant advances for new signings. This competitive strategy is very much aligned with that of many mid-size and smaller record companies, which pitch for new artists on the basis of the strength of their creative capabilities, the collaborative nature of their work with artists and the greater amount of management attention which these companies can give to individual artists. According to the Notifying Party, it is for this reason that EMI today is the least close competitor to Universal among the traditional majors.

(377)The Notifying Party also submits declining compilation revenues as evidence of EMI's declining market position. […]*.

(378)The Notifying Party emphasises that EMI Group's precarious financial position and resulting management distractions have adversely impacted EMI's ability to invest in and to attract and develop new talent, and has diminished its competitive presence in the market in recent years. […]*.

(379)The Notifying Party also refers to the derogation request made under Article 7(3) of the Merger Regulation in the context of the takeover of EMI by Citigroup Inc ("Citi"). The Notifying Party highlights that the Commission granted this derogation request in the light of […]* . According to the Notifying Party, Citi's ownership of EMI did not last for long enough to materially improve EMI's weakened competitive position on the market.

(380)Overall, the Notifying Party concludes that there is a great likelihood that EMI's market position would most probably have declined further in the absence of the proposed concentration given EMI's precarious financial position, the problems it has encountered both in retaining and in signing successful artists […]* and the decline of compilations as a commercial distribution format.

(381)The Notifying Party finally submits that the relevant counterfactual to the proposed concentration is not one of a stand-alone EMI as it existed prior to the

214 See paragraph 4 of the Form CO. 215 See paragraph 5 of the Form CO. 216 See paragraphs 601 to 603 of the Form CO.

106

merger, but of an EMI stripped of its publishing arm and in the hands of a highly-leveraged new owner.

107

7.1.3.2. The Commission's assessment

EMI is a major record company

(382)The Commission addresses each of the Notifying Party's claims in turn.

(383)With regard to the position of EMI as a major, the Commission notes that EMI, like Universal, is active in the discovery, signing and development of artists as well as in the wholesale of recorded music. EMI, like Universal, has traditionally been considered as one of the four majors.

(384)The characteristics that, amongst other things, differentiate EMI as one of the majors from the indies comprise: a worldwide presence and therefore a world-wide representation for international artists; a presence not only in A&R and the wholesale of music but also, to some extent, at retail level, as well as ancillary services (and prior to the sale of EMI, in music publishing); significant financial strength, making majors less vulnerable to risks and enabling them to offer artists more attractive financial benefits; and a large, diversified artist roster and repertoire, both in terms of new releases and back catalogue. Indies, on the other hand, tend to: be much smaller in terms of organisation; focus on A&R and the wholesale of recorded music (often focused on particular genres); operate on a national rather than international or worldwide basis (with some exceptions of multi-territory presence); have a more limited budget for promotional and marketing activities and a more limited access to mass media; and have a higher dependency on new releases than on a large back catalogue.

(385)Contrary to the Notifying Party's claims, EMI's profile is by no means comparable to that of an indie and is much more similar to that of the other majors. First, the Notifying Party's claim that EMI has reduced some of its global 217 presence in recent yearsdoes not detract from the fact that it continues to be a record company with a worldwide footprint, which achieves more than half of its turnover outside the EEA, and which has the capability to represent its artists on a worldwide basis. Second, the fact that EMI does not undertake all ancillary services and, as of very recently, no longer has a music publishing arm does not make EMI more similar to an indie than to a major. Its sheer size, its wide range of products and the geographic scope of its activities, as well as the fact that it does engage in some ancillary services, such as merchandising, militate against this conclusion. Third, even prior to Citi's takeover, but even more so after that takeover, EMI's profile in terms of resources and ability to attract artists was by no means comparable to that of an indie, however large such an indie may be.

(386)The Notifying Party's claims with regard to EMI's competitive position are also inconsistent with its own description of EMI when announcing the proposed concentration. In the press release that accompanied that announcement, the Notifying Party described EMI as: "one of the world’s most prominent music companies, possessing a rich musical legacy. Its recorded music division, EMI Music, operates around the world and represents artists spanning all musical

217 […]*.

108

tastes and genres through record labels including Angel, Astralwerks, Blue Note, Capitol, Capitol Latin, Capitol Records Nashville, EMI Classics, EMI CMG, EMI Records, EMI Records Nashville, Manhattan, Parlophone, Virgin Classics 218 and Virgin Records".

(387)With regard to EMI's alleged focus on back catalogue, the Commission considers the Notifying Party's figures to be misleading.

(388)First of all, the comparison that the Notifying Party has drawn between the 219 revenues that Universal and EMI each derive from new music releases , new 220 221 releases of back catalogueand core catalogueare not for the same geographic region. The breakdown of EMI’s revenue used by the Notifying Party is at the global level, whereas Universal's figures are for the geographic region that encompasses the EEA, the Middle East and Africa. The Commission considers that for the competitive assessment in this case, the relevant base for comparison is the breakdown between the different categories of catalogue at EEA-level. When analysing the breakdown of revenue at the correct level, EMI's position is not significantly different from that of Universal. The data for digital releases show that on an EEA-wide basis, core catalogue represented only […]*% of EMI's EEA-wide digital revenues. New (re-) release of catalogue represented […]*% of EMI's EEA-wide digital revenues and new music represented […]*%. As regards physical sales, new music represented […]*% of EMI's EEA-wide revenues, new release of catalogue represented […]*% and core catalogue […]*% of EMI's EEA-wide physical revenues.

(389)Overall (that is to say, aggregating EEA-wide sales for digital and physical music), EMI's EEA-wide revenues were divided as follows: […]*% new music, […]*% new release catalogue and […]*% core catalogue. EMI provided an updated table for 2010 which is more in line with the way in which Universal provided its figures (although some differences in methodology remain). On the basis of those data (relating to physical recorded music only), EMI received […]*% of its revenues from catalogue on an EEA-wide basis and […]*% from new releases. At an EEA-wide level EMI is clearly not focused on catalogue, but received around […]* of its revenue from new releases. These figures do not compare unfavourably with those of Universal, according to which Universal received […]*% of its 2010 recorded music revenues from back catalogue and […]*% from new releases.

(390)In relation to EMI's investments in A&R, the Commission notes that, contrary to the Notifying Party's submission, the comparison of the A&R spend of each of Universal and EMI was not the "key" arguments relied upon by the Commission

in the Statement of Objections to demonstrate that Universal and EMI are close competitors. It was simply one of a number of factors that allowed the Commission to reach this conclusion. In relative terms, Universal spent around […]*% of its net EEA-wide sales on A&R. EMI spent around […]*% of its net 223 sales in the EEA on A&R for the financial year ending 31 March 2011. The Commission retains its conclusion that when viewed in those terms, the A&R spend of EMI shows that EMI has still been a significant competitive force. Given EMI's far higher revenues than any indie in absolute terms, these figures confirm EMI's position as a far closer competitor to Universal than any indie would be.

(391)With regard to EMI's alleged reliance on revenues from compilations, the Commission considers that the fact that these revenues have fallen in both absolute and relative terms in no way indicates that EMI is a less close competitor to Universal than the other majors. In fact, EMI's compilations revenues as a proportion of revenues from compilations, tracks and albums are similar to those of Universal. In 2011, on an EEA-wide basis, compilations revenues represented […]*% of EMI's revenues whereas the equivalent figure for Universal was […]*%. In respect of the digital market, on an EEA-wide basis, the proportion of revenues derived from compilations was […]*% for EMI and […]*% for Universal. In respect of the physical market, on an EEA-wide basis, the proportion of revenues derived from compilations was […]*% for EMI and 224 […]*% for Universal .

(392)The Notifying Party argues at length that EMI is less of a competitive force because its compilations depend on third party tracks. The Notifying Party refers to the EMI presence on the NOW compilations to make this point. When the presence of the various majors on the NOW compilations is considered, EMI appears to have the same amount of tracks on the NOW compilations as Warner, which is clearly considered by the Notifying Party to be a major competitor to Universal. In that sense, therefore, EMI is just as strong a competitor as Warner.

(393)Contrary to the Notifying Party's claims, the Commission does not purport to establish that EMI is Universal's closest competitor. Rather, the Commission considers that the results of its market investigation, including those referred to by the Notifying Party, show that EMI is a major record company and a strong and close competitor of Universal. Respondents to the market investigation do not appear to consider EMI as a lesser competitor to Universal than at least Warner. In fact, a number of respondents (depending on the country in question) 225 consider EMI to be a closer competitor to Universal than Sony or Warner . Under all relevant categories (back catalogue, hits, superstars, established artists, new artists, local repertoire, international repertoire, jazz labels, classical labels and compilations) cited by the Notifying Party in its Response to the Statement of Objections, EMI is considered by market participants as a close competitor to Universal.

223 See paragraph 200 of the Form CO. 224 Form CO, pp. 288 and 289. 225 Competitors' replies to the Commission Request for information of 17 February 2012, question 50.

110

(394)This conclusion is also borne out by other data sources. If the Commission considers, for example, the Global Top 50 Albums for 2010, EMI and Warner had exactly the same number of albums in the Global Top 50.

Table 22 - Albums listed in IFPI’s “Global Top 50 Albums 2010”

N° of Albums

Percentage %

Company

46%

12%

Sony

26%

Warner

12%

Others

Source: IFPI. Figures include physical and digital sales

(395)The same conclusion can be drawn from other sources, such as the performance of EMI's artists and catalogue at the IFPI Platinum Europe Awards. According to IFPI, Platinum Europe Awards "honour the elite community of artists who have achieved sales of one million copies of an album in Europe. The Awards are 226 recognised as the established hallmark of success for artists in Europe" . If one considers, for example, the IFPI Platinum Europe awards for 2010 and 2011, EMI performed better than either Sony or Warner. In fact, in 2011, EMI had exactly the same number of albums as Universal that achieved sales of at least one million units.

Table 23 - IFPI Platinum Europe awards 2010 and 2011 (albums which achieved sales of at least one million units) – 44 albums listed in 2010/29 albums listed in 2011

N° of Albums 2010

N° of Albums 2011

% 2010

% 2011

Company

57%

31%

18%

31%

Sony

14%

Warner

11%

28%

Others

10%

2010 Source: IFPI - 2011 Source: IFPI Platinum Europe Awards

226 Source: IFPI website.

111

(396)As regards the particular segments of classical and jazz music, the Commission notes that, according to […]*, the merged entity's market share in classical and jazz music in a number of Member States, including, in particular the United Kingdom and France, would be as high as 70% to 80%, leaving its competitors 227 far behind .

(397)According to an anonymous industry participant, Universal and EMI will combine some of the best classical music catalogues. A combined Universal/EMI would be particularly strong in the classical music genre in at least France and the United Kingdom. A combined Universal/EMI would have 28 of the top 30 classical music products in France based on 2011 sales. As regards the United Kingdom, Universal/EMI's share of the UK top 20 classical music chart would be 228 as high as 85-95% (weeks 48-52 of the year 2011) .

(398)Universal and EMI will combine some of the best classical music catalogues. Universal owns Deutsche Grammophon and Decca, which are two of the largest classical music labels in the world. Deutsche Grammophon has a unique roster of classical music artists, including Claudio Abbado, Roberto Alagna, Martha Argerich, Daniel Barenboim, Berliner Philarmoniker, Pierre Boulez, Placido Domingo, Vladimir Horowitz, Vadim Repin and Maurizio Pollini. Decca is also one of the leading recording labels for classical music. It has landmark artists and composers on its roster such as Vladimir Ashkenazy and Benjamin Britten, as well current stars such as Cecilia Bartoli, Sir Colin Davis, Renée Fleming and Kiri Te Kanawa. EMI Classics and Virgin Classics are the main competitors to the Universal labels. The only other meaningful competitor in this music genre post-merger would be Sony, with its label Sony Masterworks.

(399)Moreover, as regards classical music, after the issuing of the Statement of Objections, the Commission has collected data on classical music and performed the same relativity exercise as described in Section 7.1.2.8. The Commission informed the Notifying Party of these additional factual elements on which it intended to rely in the decision by means of a letter of facts of 25 July 2012. Despite being offered the opportunity to do so, the Notifying Party did not submit any specific observation on these additional factual elements.

(400)The data set available to carry out this exercise was limited by the fact that Universal was unable to provide the revenues for classical digital sales due to the 229 nature of its internal financial reporting procedures . Nonetheless, by comparing the revenues of the other majors generated by the wholesale of classical music in digital format with the revenues generated by the wholesale of classical music in physical format, the Commission concludes that the vast majority of the total sales of classical music (around 90%) are in physical format and that, therefore, a comparison of the different major's position based on their

227 Competitors' reply to the Commission Request for information of 17 April 2012, question 20 (ID: 7750). 228 Anonymous submission (ID 7767). 229 See Universal's reply to the Commission Requests for information of 17 February 2012 and 13 July 2012.

112

sales of classical music in this format would be meaningful to assess their relative position in relation to this genre.

(401)Table 24 shows the relative position of the four majors in the wholesale of classical music in physical format:

Table 24 – Relative position of the majors in the wholesale of classical music in physical format (2011)

Sony

Warner

[60-70]%

[10-20]%

[0-10]%

Austria

[50-60]%

[30-40]%

[0-10]%

Belgium

Czech Republic

[70-80]%

[10-20]%

[0-10]%

[80-90]%

[10-20]%

[0-10]%

Denmark

[40-50]%

[30-40]%

[0-10]%

Finland

[40-50]%

[30-40]%

[10-20]%

[0-10]%

France

[50-60]%

[10-20]%

[30-40]%

[0-10]%

Germany

[50-60]%

[10-20]%

[20-30]%

[0-10]%

Greece

[40-50]%

[20-30]%

[0-10]%

Hungary

[60-70]%

[20-30]%

[0-10]%

Ireland

[50-60]%

[10-20]%

[20-30]%

[0-10]%

Italy

[60-70]%

[20-30]%

[10-20]%

[0-10]%

Netherlands

[50-60]%

[20-30]%

[10-20]%

[0-10]%

Poland

[20-30]%

[30-40]%

[0-10]%

Portugal

[50-60]%

[20-30]%

[10-20]%

[0-10]%

Spain

[60-70]%

[10-20]%

[0-10]%

Sweden

United Kingdom

[70-80]%

[10-20]%

[0-10]%

[50-60]%

[10-20]%

[0-10]%

[40-50]%

[0-10]%

Norway

[10-20]%

EEA

[50-60]%

[10-20]%

[0-10]%

Source: Commission analysis based on data provided by the Notifying Party, EMI, Sony and Warner.

(402)Table 24 confirms that in Belgium, the Czech Republic, Denmark, Finland, France, Ireland, the Netherlands, Poland, Spain, the United Kingdom and Norway in 2011 EMI is Universal's strongest competitor, while in Hungary,

113

Portugal and Sweden it is the strongest competitor to Universal together with Sony. In the remaining countries, EMI is still a strong competitor to Universal. At the overall Union and EEA levels, EMI is just as strong a competitor to Universal as Sony.

(403)Moreover, post-merger, the merged entity would be by far the largest market player in this genre, with essentially only one major competitor left (namely Sony), since Warner's presence in the classical music genre is extremely limited.

(404)The Commission considers that this data confirms the conclusion that Universal and EMI are particularly close competitors with respect to classical music.

(405)For jazz, EMI owns the Blue Note label and Universal owns Verve, which are the two leading jazz labels worldwide. A combination of these labels would bring together the recordings of some of the world’s best-known past and present jazz musicians such as Louis Armstrong, John Coltrane, Ella Fitzgerald, Herbie Hancock, Count Basie, Miles Davis and Wynton Marsalis. Universal also distributes Concord Music Group that has a well-established roster of jazz artists 230 .

(406)The Commission therefore concludes that EMI is one of the four major record companies currently active in the recorded music sector worldwide, with a significant artist roster and music repertoire comprising a formidable back catalogue and strong new releases. It is also likely to be Universal's closest competitor at least in relation to classical music.

(407)The Commission now turns to the Notifying Party's claims in relation to EMI's financial position, and the correct counterfactual to apply in this regard.

(408)The Notifying Party submits that, when assessing the proposed concentration, the relevant counterfactual should not be the pre-merger situation, but rather a scenario in which EMI would have changed ownership. The Notifying Party indeed submits that Citi's ownership of EMI following its takeover of EMI from Terra Firma Capital Partners Limited ("Terra Firma") was never meant to be long-term. The Notifying Party notes that absent the proposed concentration, EMI's most likely purchasers would have been either Warner or a private equity investor. In the Notifying Party's view, both alternative transactions would have led to EMI being highly leveraged and therefore in a situation similar to that where EMI was under the Terra Firma ownership. As a result, according to the Notifying Party, the relevant counterfactual would not be the pre-merger situation with a debt-free EMI following its takeover from Citi, but rather EMI's situation prior to its takeover from Citi. The Notifying Party also submits that the relevant counterfactual should feature an EMI without the publishing business, which has been separately sold by Citi to the Sony Corporation/ Mubadala Development 231 Company PJSC consortium .

230 Competitors' reply to the Commission Request of information of 17 April 2012, question 20 (ID: 7750). 231 Paragraph 4.7 of the Response to the Statement of Objections.

114

(409)The Commission notes that, absent the proposed concentration, it is likely that EMI would have changed ownership and that EMI would have been purchased either by Warner or by a private equity investor. The Commission also notes that, in the relevant framework of analysis, EMI does not have a music publishing business, since the Commission approved the acquisition of that business by a 232 consortium led by Sony/Mubadala in April 2012and that transaction has been closed in the meantime. This however does not mean that when assessing EMI's competitive position as a recorded music company, the Commission should take account of any possible development that is not the situation that existed pre-merger.

232

(410)Absent exceptional circumstances (for example, a failing firm situation), this pre-merger situation is the relevant framework against which to assess the likely effects on competition of the proposed concentration. This means that, absent such exceptional circumstances, the relevant proxy to assess EMI's competitive position is not EMI's position under the Terra Firma ownership, that is to say, prior to its takeover from Citi, but rather EMI's pre-merger situation, that is to say, the situation of EMI following its takeover from Citi. The Notifying Party did not put forward convincing arguments or evidence as to why the Commission should deviate from this principle. Indeed, while it is possible that, as submitted by the Notifying Party, a potential purchaser would finance the acquisition of EMI in a similar way to Terra Firma in 2007, at the Commission considers it more plausible that any such potential purchaser would seek to avoid committing the same errors as Terra Firma and structure the acquisition differently.

(411)The Commission therefore considers that it should not speculate about the ways in which another potential purchaser would have financed its acquisition of EMI and about EMI's capital structure following such acquisition. As a result, the Commission concludes, in line with the general principles of its Horizontal Merger Guidelines, that the relevant framework of assessment of the proposed concentration is the pre-merger situation, that is to say, EMI's situation following Citi's takeover and following the sales to the Sony/Mubadala consortium of its publishing arm.

(412)As a final remark, the Commission notes that, consistently with the Notifying Party's arguments and with the Commission's approach in this decision, the fact that, in the relevant framework of analysis, EMI would no longer have a publishing arm is irrelevant to the assessment of the competitive force of EMI in the recorded music sector, which is the centre of gravity of the proposed concentration. This reflects the fact that, while not having a publishing business may prevent a record company from using its combined presence in recorded music and music publishing to extract even better terms from digital customers, it does not in any way diminish such record company's competitive strength in the recorded music sector.

(413)The Commission considers that the evidence demonstrates that EMI was a viable, competitive and financially viable business, with a very strong back catalogue

232Commission decision of 19 April 2012 in Case No COMP/M.6459, Sony/Mubadala/EMI Music Publishing.

115

and some successful active artists, which, absent the proposed concentration, would have continued to constitute a close competitor to Universal and a strong competitive force in the relevant recorded music markets.

(414)The Commission already demonstrated that in terms of its alleged focus on back catalogue as opposed to new releases, a proper assessment of EMI's revenues from each of these activities shows that EMI was a close and strong competitor to Universal. In any event, even if EMI would have a particularly strong back catalogue, this does not show that its importance as a competitor has diminished. A strong catalogue brings in a steady revenue stream to a record company, whereas new release revenue is less certain. Back catalogue is therefore as important as new releases and hits, especially in relation to digital customers. As is stated in a presentation to […]*. According to the same document,[…]* .

(415)Universal's own internal documents confirm that […]* These documents state the following:

234 235 […]* and […]* .

(416)The Notifying Party suggests that EMI is no longer a competitive force in signing artists and that artists have been leaving EMI because of EMI's financial difficulties. Whilst that may be the case for some artists, it is common for artists to move from time to time between record companies for varying reasons. Nevertheless, as the Notifying Party also highlights, a recorded music company does not necessarily have to pay high advances to sign up or retain successful 236 artists. It gives as examples […]*. The Commission recalls that competition not only takes place in relation to signing new artists but also to nurturing those who are already part of the existing roster and may become or continue to be successful frontline artists. EMI's success in the Global Top 50 albums selling over one million copies, referred to in Recital 394, also suggests that EMI is a company that is still able to bring out top new releases.

(417)In any event, even if EMI has been a relatively weak competitor in A&R due to the budgetary constraints imposed by Terra Firma over its A&R spending, this would be unlikely to be true going forward. Following Citi's takeover, EMI became a debt-free company, which, at least on the basis of certain internal 237 documents, would appear to have the ability to compete strongly with the other record companies, including in A&R.

(418)The Commission does not agree with the Notifying Party's claim that these internal documents lack probative value because they were "investor marketing documents prepared by Citigroup", not by EMI, "which were clearly designed to show EMI in the best possible light, describing only the positive aspects of the

238 business, in order to maximise the sales price". While the Commission acknowledges the fact that these documents are not EMI internal documents, it fails to understand how, short of implying that EMI's owner Citi would be untruthful in its presentation of EMI to potential investors, the mere fact that these documents were prepared by Citi in view of the sale of EMI could undermine their reliability in assessing the financial situation of EMI pre-merger. EMI's own internal document titled […]* states that […]* which shows that also in these terms, EMI is, at the very least, just as close, strong and effective a competitor to Universal as […]*.

(419)In any event, as explained, absent the proposed concentration, EMI would likely have been acquired by players other than Universal. In this regard, Universal’s 239 internal documents show that […]* .

240 (420) Finally, the Notifying Party itself referred to […]* .

(421)In this regard, the Notifying Party seems to argue that EMI's high EBITDA level in the short term is not a proxy for a good business performance. There seems to be an inherent inconsistency in the Notifying Party's claims in this regard, because low EBITDA levels at the time that Citi acquired EMI would be a good indicator of a poor business performance, whilst currently high EBITDA levels are not a good indicator of a good business performance by EMI in recent years. The Commission considers that either EBITDA levels are a good proxy for business performance, in which case EMI's high EBITDA levels in recent years is a proxy for its strong competitive position, or they are not, in which case the Notifying Party's argument that EMI's low EBITDA levels are an indicator of its poor performance cannot be true.

(422)In relation to its recorded music business, and discussing business performance 241 for recorded music alone, EMI states the following about itself:

242 "[…]*" .

(423)The Notifying Party has tried to undermine the probative value of the statements included therein by submitting that the document was "clearly designed to be a management presentation to potential buyers in order to maximise the sales price and, as such, it picks out a limited number of EMI artists successes in certain 243 limited territories" . Regardless of whether the document is selective in highlighting EMI's successes and of the representativeness of these successes, the fact remains that EMI did achieve such successes in the course of 2011 and cannot therefore be considered as a lesser competitor in the recorded music sector. It is equally difficult to disagree, and the Notifying Party does not do so,

238 Response to the Statement of Objections, paragraph 4.7(a). 239 […]*. 240 […]*. 241 This invalidates the Notifying Party's view that business performance comments were inflated at that time. EMI was still an integrated business with a recorded music and music publishing business. The comments concern EMI Music alone. 242 […]*. 243 Response to the Statement of Objections, paragraph 4.7(c).

117

with the following statement included in the document:

244 "[…]*" .

245 (424) In a letter […]* states the following:

[…]*.

(425)The Notifying Party claims that during his investigational hearing before the Federal Trade Commission, Mr […]* explained that […]* The Commission considers that, even if it were to be accepted that an explanation of a contemporaneous internal document after the fact, which contradicts the natural meaning of the words used in the said document, has any probative value […]* explanation seems to do little to further the Notifying Party's cause. This reflects the fact that, in his explanation, relied on by the Notifying Party, […]* essentially only says that the document […]*, which, in itself, seems to strengthen instead of weaken its probative value.

(426)To conclude, the Commission does not consider EMI to be, in any sense, a failing business that would, absent the merger, cease to be a close and strong competitor to Universal. On the contrary, EMI is a formidable competitor in the market for recorded music and is as close a competitor to Universal as Sony or Warner, or at the very least, as strong a competitor as Warner.

(427)The Commission's overall conclusion is therefore that EMI is a strong competitor of Universal in the wholesale of physical and digital music, as well as in A&R, in the EEA and that the proposed concentration leads to the elimination of an important competitive force from the relevant markets.

7.1.4.Customers' limited possibilities of switching supplier

(428)Pursuant to the Horizontal Merger Guidelines, customers that have limited possibilities of switching supplier are particularly vulnerable to price increases by 246 the merged entity. The merger may affect these customers' ability to protect themselves against price increases. As in other parts of the Horizontal Merger Guidelines, "increase of prices" is used as shorthand for any competitive harm that may result from a proposed merger, which includes price increases, reduced 247 output, choice or quality of goods and services and the reduction of innovation .

(429)The Commission considers that both the customers on the physical and the digital markets face difficulties in switching suppliers.

(430)For most customers, the repertoire of all majors, including that of Universal, and perhaps of some indies is already a "must have" today so that they already face 248 difficulties in switching supplier. Respondents to the market investigation also

244 […]*. 245 […]*. 246 Horizontal Merger Guidelines, paragraph 31. 247 Horizontal Merger Guidelines, paragraph 8. 248 Customers' responses to the Commission Request for information of 17 February 2012, question 29.

118

confirmed that the combination of Universal and EMI would indeed create a ''super major'', which would represent such a large part of the recorded music 249 market that it would be able to effectively dictate its conditions to customers . Indeed, any retailer wishing to have a credible and relevant offer would need to stock the merged entity's repertoire, since no other competitor would be able to offer such a comprehensive volume of repertoire.

(431)As a result, all categories of physical customers, including those customers which only offer a limited stock of music, such as mass merchants, and those customers which offer a broader selection of products, such as specialised retailers, as well as all categories of digital customers, including those customers which strive to offer as much as possible of the existing music, such as download or (lean-forward) streaming services, and those which do not need to do so, such as (lean-back) music services, will have to carry at least some of the merged entity's products.

(432)The Commission therefore concludes that, post-merger, the merged entity's repertoire would (continue to) be "must have" and customers would therefore likely have limited (if any) ability to switch supplier and the merged entity's ability to exploit them will likely significantly increase.

(433)The likely effects of the size increase of Universal if the proposed concentration were to go ahead needs to be assessed within this context. In particular, the Commission needs to assess whether, given the limited possibilities of switching supplier, the size increase for Universal would enhance its ability and incentive to apply licensing conditions that are more disadvantageous to its customers. The Commission also needs to assess whether against this background, the proposed concentration would enhance the merged entity's ability to impose commercial terms and conditions that would hinder the expansion of its competitors, with customers having limited opportunities to resist the imposition of such terms and conditions.

7.1.5.Likely effects of the proposed concentration

(434)250 (434) In Universal/BMG,the Commission considered that the size of record companies matters greatly in commercial negotiations with digital customers:

"The (digital) music providers which reported (…) rate differences (between large and small music companies) referred to the higher bargaining power of those majors with higher market shares as their catalogues are of greater importance for an online and mobile music provider. (…) Several (digital) music providers stated that record companies have refused or delayed wholesale licences in the past (…) in order to achieve higher rates. It is noteworthy that such behaviour has only been possible with respect to recording rights as publishing rights have so far been licensed by collecting societies."

249Competitors' responses to the Commission Request for information of 17 April 2012, question 110.

250Commission Decision of 15 September 2008 in Case No COMP/M.4404 – Universal/BMG Music Publishing.

119

(435)During the Commission's market investigation in this case, the vast majority of digital customers (thirteen out of sixteen) confirmed that the size of a record company increases its bargaining power in relation to digital customers. According to these customers, this high bargaining power translates into the imposition of various commercial terms that are more disadvantageous to them, including higher levels of remuneration for musical content, higher levels of advance payments, guaranteed revenues, guaranteed share of promotional opportunities and MFN clauses. These digital customers for instance note the following:

"Les gros labels étant indispensables à une offre compétitive de musique, ils sont en position de demander et de sécuriser leurs contrats beaucoup plus que les petits labels. Cela se traduit notamment par des minimum garantis, des exigences de promotion marketing, et des restrictions d'usage accrus par rapport aux petits labels (Exemple : les majors imposent des DRM avec licence de 30 jours, tandis que les indépendants acceptent le format MP3, le téléchargement est possible 251 depuis un mobile pour les indépendants mais pas pour certaines majors...)" ;

"Major labels demand and secure higher prices e.g EMI […]* for a premium 252 release, […]*" ;

"As a matter of course major record labels relying on their strong market power are able to secure better terms and conditions for themselves than smaller labels. The main factors are the size of their catalogues and a high proportion of Top 20 253; artists that are very important for a content provider to have in stock"

254 "Yes - in general, larger labels have more negotiating leverage."

"Usually they do, either by pushing for that (i.e. extorting), or by simply offering 255 a better incentive" ;

"Ja, Major Labels können aufgrund ihrer Verhandlungsposition (größeres und zielgruppenrelevantes Repertoire) grundsätzlich bessere Terms durchsetzen als kleinere, insb. in den Bereichen - Detailed reporting - Up front (and/or minimum) payments - Greater Of Modelle (Vereinbarung und Höhe von Per Play 256 Fees)" ; "The major labels (Universal, Warner, Sony, and EMI) consistently secure better terms than smaller “independent” labels. Because they have large repertoires, they have sufficient market power to demand higher royalties and advances to access their catalog, and to obtain preferential marketing and promotional opportunities on an MFN basis. Larger major labels have an even greater advantage than do smaller major labels. For example, unlike major labels,

independent labels rarely obtain any form of advance or minimum guarantee payment. They also obtain fewer marketing opportunities because the major labels require unique marketing placement on an MFN basis that consumes the 257 vast majority of our open marketing opportunities." ; "Larger record companies (in this case the Majors) always secure better terms and conditions, obviously because of their market share. The music that is successful on our service are the Charts content, and those are dominated by the Majors, which puts further pressure on us to sign deals with them on their terms. In our case, we even had to pay advances in the past to Majors. Smaller record companies (Indies) don't request such things. All in all, because all the most successful content are owned by the Majors, one either have to submit to their 258 requests" ;

259 "Typically yes since they have a big negotiation power" ;

"From our point of view larger record companies offer worse terms and 260 conditions than the smaller ones" ;

"Size obviously matter for bargaining position just like in any other line of 261 business. I don't think that the music business is any different" ;

262 "Negotiation strength follows naturally from market share"; and

"The larger companies demand minimum guarantees and take higher 263 revenue" .

(436)Likewise, competitors to the merging parties confirm that larger record companies have higher bargaining power and are able to negotiate better commercial terms with digital customers. These competitors for instance note the following:

"Being a big player in the music market, regardless of overall turnover of products, gives you leverage on record labels, and more leverage over the smaller companies. However, if that portion of the customer's business is small in 264 relation to the overall music market then that power is diminished" ;

"The fact that music may not account for a significant proportion of a customer’s revenues does not confer market power on that customer vis-à-vis the record companies if music is key or important to its business for reasons other than the revenue streams it directly generates. In fact, in this case, its bargaining power will be rather limited, particularly in relation to record companies holding must-have repertoire, since the customer will not be able to threaten to switch to other 265 suppliers"; and

"Music has brand value, driving traffic to stores where consumers may purchase other higher value goods as a result of a visit initially to purchase a music product. The simple interface of the iTunes music store with Apple’s iPhone, iPod and iPad dramatically enhances the appeal of those products. For these reasons, […]* believes that these customers will remain active in the market for recorded music and will not be able to resist a dominant record company seeking to increase prices or otherwise improve trading terms. These customers may be forced to yield to the demands of the largest record companies but seek to 266 recover lost margins from its smaller suppliers" .

(437)Competitors also consider that this discrepancy in commercial terms offered to digital customers is not levelled by MFN clauses in commercial agreements with digital customers. They indicate that smaller size record companies are generally not able to negotiate MFN clauses.

(438)[…]* states the following:

"in […]* experience: - MFNs are not applied systematically to all recorded music companies; and - where MFNs are given, they are most likely given to the largest companies and not for example, to […]*. Therefore, rather than providing any level of equality, we have found MFNs have often acted as a block 267 to our achievement of favourable terms in negotiations" .

(439)[…]* states the following:

"[…]* believes that if a dominant record company imposes higher prices or more onerous terms on a particular platform and that platform is unable to resist such a move, it will inevitably seek to recover its lost margin on those new arrangements by squeezing its other smaller suppliers to provide it with lower prices or better terms. While other record companies could theoretically protect themselves through an MFN, in reality smaller record companies would not be able to obtain such MFN protection or do not request them as a matter of policy or practice.

In themselves MFNs do not prevent a dominant player from imposing higher prices or more onerous terms, they just result in the beneficiaries of the MFN getting the same higher prices or better terms to the detriment of the service and ultimately the end-consumer. It is […]* understanding from discussions with its digital partners that Universal is particularly aggressive in its imposition of “blanket” MFN provisions covering all aspects of a commercial agreement, thus

ensuring that it always gets the best terms in all elements of a deal. This results in the “double whammy” of Universal both being able to leverage its existing market power and ensuring that no other record company can ever get a better 268 deal, resulting in an upwards harmonization of pricing" .

(440)These submissions confirm the Commission's findings in Universal/BMG concerning the relevance of the size of a record company in its negotiations with digital customers. The Commission therefore assessed whether the increase in Universal's size and market power that would result from the proposed concentration, is likely to significantly impede effective competition on the markets for the wholesale distribution of digital music.

(441)In this context, the Commission reiterates that the purpose of Union merger control is to prevent mergers that would deprive consumers of the benefits that effective competition brings. The Commission's Horizontal Merger Guidelines provide that effective competition brings a wide variety of benefits to consumers, such as low prices, high quality products, a wide selection of goods and services, 269 as well as innovation. Accordingly, the Horizontal Merger Guidelines specify that what is meant by increased market power of merging undertakings is the ability of one or more firms to profitably increase prices, reduce output, choice or quality of goods and services, diminish innovation or otherwise influence 270 parameters of competition. The expression "increased prices" is used as shorthand for these various ways in which a merger may result in competitive 271 harm .

(442)In line with this framework, the Commission's assessment of the effects of the proposed concentration has focussed on all consumer benefits from effective competition and all relevant competitive parameters on recorded music markets. In particular, the Commission has assessed whether, due to the increase in the post-merger size of Universal and its market power, the proposed concentration is likely to lead to the following:

(a)More disadvantageous commercial terms for digital customers (including higher wholesale prices and licensing fees), likely resulting in higher prices for end users of digital music services; and

(b)A reduction of consumer choice and innovation with respect to digital music services.

(443)The Commission considers that if that were the case, the proposed concentration would also have a negative impact on cultural diversity. Article 167 (4) of the TFEU requires the Union to take cultural diversity aspects into account in its actions under the other provisions of the Treaties, including Union competition rules. Moreover, the UNESCO Convention on the protection and the promotion

of the diversity of cultural expressions to which the Union is a Party sets out a comprehensive set of guiding principles relating to the diversity of cultural expressions. The case-law has confirmed that other Treaty objectives may be 272 taken into account in the context of the assessment of a concentration .

7.1.5.1.The Notifying Party's view

(444)The Notifying Party claims that the proposed concentration would not have any impact on price and other commercial terms for digital customers.

(445)The Notifying Party claims that if the Commission's investigation were to reveal that Universal, which is currently already the largest record music company, currently obtains commercial terms that differ from those that its competitors obtain, this finding would be irrelevant. The Notifying Party points out that to the extent that a supplier such as Universal offers better products or services or other efficiencies to customers (for example lower negotiation, transaction or reporting costs), one would normally expect such a supplier to get better distribution terms. In the Notifying party's view, there is nothing a priori anticompetitive about that result.

(446)The Notifying Party claims that in any event, the facts do not support the notion that Universal, by virtue of its size or for any other reason, is able to bargain for better commercial terms from online platforms than its rivals. The Notifying Party considers in turn that its increased size as a result of the proposed concentration would therefore not significantly impede effective competition.

(447)To support this argumentation, the Notifying Party has compared the commercial terms of Universal and EMI with four digital customers, namely the download platforms [digital platform]* and [digital platform]*, and the streaming platforms [digital platform]* and [digital platform]*. The Notifying Party claims this comparison is conclusive as these customers collectively account for […]*% of Universal's digital revenues. The Notifying Party claims that this comparison shows that the commercial terms that Universal and EMI obtain are not materially different.

7.1.5.2.The Commission's assessment

(448)The Commission's investigation has not confirmed the Notifying Party's claims. Various categories of evidence show that currently Universal already obtains commercial terms from digital customers that are more favourable to Universal as compared to […]*, and less favourable to digital customers. These categories of evidence are the following:

(a)A comparison of Universal and EMI's agreements with online platforms ;

(b)Detailed quantitative analysis undertaken by the Commission;

(c)Evidence from the market investigation; and

(d)Evidence from the merging parties.

7.1.5.3.Comparison of Universal and EMI's agreements with online platforms

(449)The Commission has conducted a review of the commercial agreements of both of Universal and EMI with a selection of key digital customers. The Commission considers that this comparison is informative, as Universal is currently a far larger supplier of recorded music to digital customers than EMI is. Accordingly, this comparison provides indications as to the current ability and incentive of Universal to obtain favourable commercial terms on the basis of its larger size. If this is confirmed, it also in turn provides indications as to the likely effects on commercial terms for digital customers if the proposed concentration were to go ahead as the proposed concentration would increase Universal's size and power and hence its bargaining position in commercial negotiations with digital customers.

(450)In a first step of its assessment, the Commission identified the key digital customers of the merging parties that should be included in the comparison of the commercial terms that each of Universal and EMI obtain. In a second step, the Commission identified the key commercial terms to include in the comparison.

(451)Selection of digital customers. On the basis of the data provided in the Form CO, Universal's 10 largest digital customers are the following:

(451)Table 25 - Universal's top 10 digital customers based on 2010 revenues in the EEA

Share of Universal's totalGroup customer's name EEA digital revenues

Ranking

[…] *

Source: Notifying Party

(452)EMI's 10 largest digital customers are the following:

(452)Table 26 - EMI's top 10 digital customers based on 2010 revenues in the EEA

125

Share of EMI's total EEAGroup customer's name digital revenues

Ranking

[…] *

Source: EMI

(453)Considering their importance for the merging companies' digital revenues, the Commission has focused its analysis of commercial terms on eleven digital customers: Apple, Spotify, Amazon MP3, Vodafone, Nokia, Orange, Napster, Deutsche Telekom, Deezer, Virgin Media and YouTube.

(454)The Commission considers this selection to be representative and relevant for two main reasons: first, they represent […]*% of Universal's 2010 digital revenues in the EEA and […]*% of EMI's 2010 digital revenues in the EEA. Second, they represent a wide variety of digital platforms present of the market: electronic communication operators, e-tailers and "pure" digital music services, recent services and more established services, streaming and download platforms.

(455)This comparison is likely to be more informative than the comparison that the Notifying Party has made. The selection by the Notifying Party of 4 key customers only includes the […]* download platforms ([…]* and […]*) and the […]* streaming platforms ([…]* and […]*). The Commission considers that this selection is likely to understate the actual impact that the proposed concentration would have on the overall digital recorded music markets for two reasons: first, these customers have the largest market share and are therefore likely to have relatively more bargaining power in relation to record companies. Second, this selection does not encompass the wide range of nascent digital music services that have launched operations in various EEA countries and that bring important consumer benefits in terms of wide consumer choice and innovative ways to consume digital music.

(456)The Notifying Party itself recognises that, according to the widely held view in the music industry, the model of streaming and particularly subscription services is the strongest growing business model, will continue to grow strongly and will prove particularly significant in the future. Both online customers and

273competitors express similar projections. […]* Generally speaking, […]* has become the second largest source of digital revenue for record companies across the EEA. It is already the largest digital music retailer in certain EEA countries, such as Finland, Sweden and Norway. The current proportions that digital customers currently hold in Universal's and EMI's digital revenues, in particular the proportions of the four digital customers that the Notifying Party relies upon, are by no means informative of the likely proportions they would hold in the next three years. Any analysis of commercial terms would need to take particular account of the online platforms that are important for the future.

(457)In its own analysis of commercial terms granted to online platforms, the Notifying Party makes a distinction between download services and streaming services. The Commission has followed this approach and has classified these customers into the following main categories:

(a)Download services: Apple, Amazon MP3, Vodafone, Nokia, Orange, Deutsche Telekom/T-Mobile and Virgin Media; and

(b)Streaming services: Spotify, Deezer, Napster and YouTube.

(458)The Commission notes that this categorisation is based on the revenues of these customers from their main digital music service offerings. It is not exhaustive as digital customers can offer a combination of different digital music services. For instance, Apple also offers the iCloud service, which is a hybrid between a download and streaming service. As another example, Nokia's "Comes with Music" service is a subscription download service, and is therefore a hybrid between the traditional à-la-carte download services such as Apple and Amazon, and subscription-based streaming services such as Spotify and Deezer. Where relevant, the Commission refers to these different digital music offerings that each digital customer offers when analysing the commercial terms that Universal and EMI obtain.

(459)Selection of key commercial terms. In a second step of its assessment, the Commission identified the key commercial terms that need to be compared.

(460)The Notifying Party itself has identified the following key commercial terms in its comparison of Universal's and EMI's commercial terms: […]*. The Notifying Party made separate observations on MFN clauses.

(461)The Commission's market investigation indicates that the majority of digital customers consider these commercial terms negotiated with record companies to be key: these include price terms including PPD and applicable discounts, advance payments and marketing terms .

273Competitors' reply to Commission Request for information of 17 April 2012, question 49; and Customers' reply to Commission Request for information of 17 April 2012, question 43.

274Form CO, paragraphs 726 and 770.

275Customers' reply to the Commission's Request for information of 17 February 2012, question 37.

(462)Contrary to the Notifying Party's view, the vast majority of customers do not consider these MFN clauses to be a protective shield and a tool that assists them 276 in launching and expanding their onlin e service offerings. Instead, they see MFN clauses as having a direct negative impact on their price and marketing terms for the licensing of recorded music content and hence on their ability to launch and expand their operations.

(463)In this context, an important digital customer notes that:

"In our opinion, the MFN has never been proposed as a protective shield for the service provider. To the contrary, our view is that a label proposing a MFN clause wants to be sure that nobody else receives better terms than the terms the service provider grants (the label proposing the MFN clause) (…) The objective 277 of the MFN is to always apply the highest condition to any label. .

When asked about the role of MFN clauses as a protective shield, other digital customers gave the following replies:

"Non, pas vraiment. La tentation est au contraire de s'aligner sur les contrats les 278 plus chers" .

"On the contrary, MFN clauses are utilised by the majors to force a licensee to apply the highest secured by any of them to all of them, as they know a licensee needs to secure repertoire from all in 4 in order to offer a commercially viable service. It is not a shield for customers, it is a sword of the record companies to 279. ensure they receive the same (and maximum) price point or royalty rate"

(464)Considering the direct impact on over material commercial terms, the Commission has included MFN clauses in the key commercial parameters to compare.

(465)Accordingly, the Commissions' review of the commercial terms will focus on the following parameters: remuneration and advance payments, advertising commitments and MFN clauses.

(466)The results of the Commission's analysis are the following.

Download platforms

(467)Revenue payments and advances. With regard to download platforms, the Commission has compared the remuneration terms and conditions that each of Universal and EMI obtain for single track download services.

276Customers' reply to the Commission Request for information of 17 April 2012 questions 32 and 46.

277Customers' reply to the Commission Request for information of 17 April 2012, questions 32 and 46 (ID: 6441).

278Customers' reply to the Commission Request for information of 17 April 2012, question 32 (ID: 5873).

279Customers' reply to the Commission Request for information of 17 April 2012, question 32 (ID: 6005).

(468)A comparison of terms for albums is complicated by the fact that there is no 280 281 harmonisation in the number of different price tiers. […]* […]* . Hence, it is not simple to draw firm conclusions from the comparison of the commercial clauses that set these tiers. This is, however, not the case for tracks, where a harmonised three price-tier model is widely used. […]*. The Commission considers that the comparison of track remuneration terms does provide a relevant picture as it represents the majority of the merged entity's digital download revenues. The Commission also notes that its assessment for albums will be undertaken on the basis of the other sources of evidence, notably the quantitative analysis undertaken by the Commission. In any event, the comparison of the other terms, namely advance payments, advertising commitments and MFN clauses, applies to both albums and tracks.

(469)As regard Apple/iTunes, […]*.

(470)[…]*:

282a) […]* , […]*.

283284 285 b) […]* .[…]* […]* .

286(471) […]* . […]*.

(472)[…]* the review of Amazon's agreements indicates that Universal and EMI 287 288 […]* . […]* . […]*:

289[…]* .

(473)A similar picture emerges for other download platforms. For [digital platform]*, 290 291 292 293 […]* . […]* . […]* . […]* . […]*.

(474)As regards the commercial deal between Universal and [digital platform]*, […]*. For the commercial deals that Universal and EMI have concluded with other MNOs, such as [digital platform]* and [digital platform]*, […]*.

294(475) As regards advances, […]* .

(476)Advertising commitments. […]*.

295296 (477) Amazon […]* . […]* . […]*.

297(478) Nokia […]* . […]*.

(479)Vodafone […]*.

(480)Virgin Media […]*.

(481)The Commission disagrees with the Notifying Party's suggestion that the promotional and marketing requirements that many download platforms face do not matter. […]*.

(482)[…]*. The Commission considers that the wide availability of music needs to be secured across the entirety of the digital retail channels. Any further restriction of marketing and promotion of that music on digital platforms, including the smaller ones, would deprive consumers of the benefits they derive from that wide availability.

(483)MFN clauses. The review of agreements indicates that […]* key customers […]* have extensive MFN obligations towards Universal.

(484)[...]*:

298a) […]* ; and

299b) […]* .

300(485) […]* .[…]*.

(486)[…]*:

301[…]* .

(487)[…]*:

302[…]* .

303(488) […]* . […]*.

(489)Conclusions of the review of download platforms' agreements. The Commission's conclusions from the analysis of the commercial terms that Universal and EMI obtain from download platforms are threefold.

(490)First, the analysis disproves the Notifying Party's claim that the commercial terms that Universal and EMI obtain from […]* and […]* are the same.

(491)Second, the comparison shows that across the download platforms subject to the Commission's analysis, Universal obtains better commercial terms on various key parameters of negotiations with online platforms. This includes […]*.

(492)Third, the commercial terms that Universal obtains from smaller download platforms obtain are […]* than the commercial terms that Universal obtains from larger download platforms. […]*.

(493)The Commission disagrees with the Notifying Party's claim that, in its assessment, the Commission has attached undue weight to commercial terms that would be of marginal relevance whilst not giving due weight to those commercial terms that would matter most from a commercial point of view and from the point of view of competition in the market. The Commission has compared exactly those commercial terms that the Notifying Party itself has identified as being the commercial terms that are of key importance for digital customers. The comparison of these commercial terms shows that Universal generally obtains better commercial terms than its smaller competitor EMI does. […]* The latter customers are important in terms of bringing innovation to the entirety of the digital markets.

(494)Finally, other categories and pieces of evidence support the Commission's position that Universal generally obtains better commercial terms than its competitors.

Streaming platforms

(495)Remuneration and advances. The review of agreements shows that […]* streaming services […]* have more onerous remuneration terms with Universal as compared to EMI.

(496)Irrespective of the Notifying Party's claims that commercial terms for […]* and […]* are not readily comparable, the Commission has observed that the main difference between the commercial terms of Universal and EMI with these two customers is that […]*.

(497)The Commission's investigation has not confirmed the Notifying Party claims that minimum guarantees are one element of the commercial negotiations that actually supports the growth of a platform. The Notifying Party claimed that these guarantees allow it to assume the eventual success of the service from day one and therefore provide more favourable commercial terms based on this assumption of success. Digital customers themselves oppose this view. They actually see the minimum guarantees […]* as a sign of […]* larger bargaining power and as a main obstacle to their scope to expand or launch their service offerings.

(498)[…]*.

(499)With regard to advance payments, […]* streaming customers […]* have to pay 304 305 […]* payments to Universal as compared to EMI. […]* . […]* .

(500)Advertising commitments. […]* streaming platforms […]* have obligations related to marketing and advertising towards Universal that are […]* to the actual shares of Universal and EMI in revenues generated by these online 306 platforms. They do not have similar advertising commitments in relation to EMI.

(501)MFN clauses. […]*.

(502)First, the analysis disproves the Notifying Party's claim that the commercial terms and conditions that Universal and EMI obtain from […]* are the same.

(503)As regards […]*, there are three main differences in the commercial terms that Universal and EMI obtain:

(a)Remuneration: […]*;

(b)Advance payments: […]*;

(c)MFN clauses: […]*.

(504)As regards […]*, there are also three main differences in commercial terms that Universal and EMI obtain:

a)Guaranteed remuneration: […]*;

b)Advance payments: […]*;

c)Advertising commitments: […]*.

(505)Second, the comparison shows that across the streaming platforms subject to the Commission's analysis, Universal obtains better terms on various key parameters of negotiations with online platforms. This includes […]*.

(506)Third, the commercial terms that Universal obtains from smaller streaming platforms obtain are […]* than the terms that Universal obtains from larger streaming platforms. The divergence in commercial terms between Universal and EMI is inversely correlated with the size of the streaming platform. […]*.

(507)The Commission disagrees with the Notifying Party's claim that in its assessment of the commercial terms and conditions for streaming services, the Commission has undertaken too static an analysis, not sufficiently taking into account that these services represent new business models. According to the Notifying Party,

(508)Contrary to the Notifying Party's claims, the evidence collected through the review of agreements indicates that customers' terms with Universal […]*. In particular:

307a) […]* ;

308309 b) […]* […]* […]*;

c)[…]*;

310d) […]* .

(509)Other terms of the contracts between Universal and each of […]*, as well as the agreements with […]* do not show a significant evolution over time. There has also been not a similar change in the commercial terms that EMI obtains from these customers. This evolution in commercial terms and conditions is counter-indicative of the claim that the differences between Universal and EMI relate to different projections and expectations for the evolution of these services, as these would likely be levelled out over timer rather than being aggravated. The Commission considers that it is instead indicative of the relative differences in the bargaining positions of Universal and EMI.

(510)In any event, other categories and pieces of evidence support the Commission's position that Universal generally obtains better terms than […]* from streaming platforms. The same evidence also refutes the Notifying Party's claim that commercial terms change over time and are difficult to compare, so that any difference in these terms is ultimately commercially irrelevant.

(511)Overall conclusion of the comparison of commercial terms. The comparison of the commercial terms between each of Universal and EMI and their largest digital customers reveals that there are generally significant differences in the commercial terms that record companies obtain. Universal is able to secure more favourable terms with digital customers than EMI. […]*.

(512)The Commission finds that the differential in the commercial terms that Universal and EMI obtain constitutes an indication of the negative effects on competition that the proposed concentration would produce. Indeed, an extension of these terms from Universal to EMI's repertoire would directly make digital customers worse off after the proposed concentration. For these customers, the conditions at which they license musical content from EMI would be

307[…]*.

(513)During the Commission's investigation, the Notifying Party submitted that the Commission did not compare the terms obtained by Universal with the terms obtained by all of its competitors. The Commission disagrees with the suggestion that, for that reason, the comparison of the Universal and EMI terms only cannot be informative for projecting the likely effects of the proposed merger between these two parties. The Commission reiterates that this comparison is informative, as Universal is currently a far larger supplier of recorded music to digital customers than EMI is. Accordingly, this comparison provides indications as to the current ability and incentive of the Universal to obtain favourable licensing terms on the basis of its larger size. It also supports the conclusion that, due to the increased size of Universal after the proposed concentration, digital customers would likely face even more disadvantageous terms from the merged entity than those which they currently face from each of Universal and EMI.

(514)Confidentiality restraints preclude the Commission from including the terms and conditions of all of Universal's competitors in the comparison and from disclosing those details to the Notifying Party. In any event, as the market investigation and the quantitative analysis undertaken by the Commission demonstrates that the correlation between the size of record companies and favourable commercial terms for digital platforms is not limited to Universal and EMI but instead applies across the whole recorded music industry, there is no need to analyse the commercial terms and conditions of the competitors that create these discrepancies. The Commission therefore assesses the implications of the competitive position of the other competitors to Universal on the basis of that evidence. As set out below, that evidence supports the Commission's position.

7.1.5.4.Formal economic analysis of commercial conditions for digital customers

(515)The Commission put forward a conceptual framework, which predicts that larger companies are able to extract better terms. It also undertook a quantitative analysis, which confirms that due to the increased size of the merged entity, the proposed concentration would likely increase its bargaining power in relation to digital customers (or platforms), resulting in higher prices for these customers and therefore leading to a significant impediment to effective competition.

(516)The Notifying Party made several claims with regard to the theoretical economic framework that would be relevant in this case.

(517)The Notifying Party claims that the desire of digital platforms to offer a broad coverage of titles and their pricing model create a complementarity between the parties' music repertoires on a particular platform. The implication of this "complementarity" claim is that, following the proposed concentration, the prices charged by the merged entity are likely to decrease rather than increase. From a recorded music company's perspective, the claim suggests that the sale of part of a recorded music company's repertoire on a given platform creates demand for its remaining repertoire due to the complementary nature. As a result, a recorded music company would be (more) willing to sell the entire repertoire at a lower

price to benefit from an overall increase in the demand generated for its repertoire.

311The Notifying Party presents three theoretical models that are based on such complementarity. The first two setups are similar in that they introduce fixed costs for an end consumer when they use a particular platform (which may be monetary, such as the fixed subscription fee that these users pay for streaming services or non-monetary, such as the cost that a user incurs when organising his or her music library on a platform). In the first model the consumers incur a fixed cost when they adopt a platform which may reflect transaction cost or the monetary cost of listening to music from a given platform. In the second model, the source of the fixed cost affects the outside option of illegal downloading (a moral cost or the expected cost of getting caught). Given these fixed costs, any user would value a wider coverage of the repertoire offered on these platforms. In these setups retail demand for purchases at a digital platform is increasing in the coverage of the platform which introduces complementarity in the pricing decision of record companies. In the third setup complementarity is created through the prevalence of uniform (retail) pricing on digital platforms such as Apple's iTunes. The underlying assumption is that due to lack of price differentiation at retail level, an increase in a record company's wholesale price is not going to be translated into higher retail price for the given record company's portfolio but would instead be spread across all record companies. As a result the incentives to increase wholesale prices are reduced. The Commission considers that the Notifying Party's claim that retail pricing for digital music is uniform does not reflect the facts. For example, for iTunes there are three retail price tiers for tracks (as the Notifying Party acknowledges). Therefore, even if there are only limited choices for retail prices for tracks in certain platforms, platforms are still able to choose amongst them. Furthermore, as iTunes also has three wholesale price tiers it becomes unlikely that a wholesale price increase for the tracks of a given record company is not passed on into retail prices. See confirmation of this in […]* response to the Commission's request for information quoted in Section 7.1.5. For albums, there is typically an even greater flexibility as regards the setting of retail prices.

(518)The Commission has considered an alternative "substitution claim" premised on the decreasing marginal utility of additional repertoire and of consumer preference for variety. According to this claim, access to additional repertoire becomes less and less valuable for the user, the wider the repertoire available on the platform is. Technically, this creates a concave function of profits for the platform relative to the size of the repertoire ("sub-additivity") in the sense that the increase in profits for the platform becomes progressively smaller as the size of the repertoire increases ("decreasing returns"). In such a setting, the recorded music company bargains with the platform anticipating that the negotiations with other record companies will be successful at equilibrium . In such a bargaining setting, the recorded music company credibly threatens to withhold its whole repertoire if the negotiations break down. As a result of: (1) the concavity of the platform's profit function; and (2) the bargaining setting, the wholesale terms that a recorded music company can extract from platforms will depend on the size of its own repertoire relative to the size of the repertoire available to a platform (at equilibrium). The value of the additional repertoire to the platform determines the terms that will result from the negotiation. With decreasing returns relative to the size of the repertoire, it is proportionately more costly for the platform to replace a significant part of its portfolio that is provided by a "large" recorded music company than to replace a significant part of its portfolio that is provided by a "smaller one". When the addition to the value of the repertoire is marginal then the increase in the value of the additional repertoire is marginal and therefore the platform will only be willing to offer a relatively low remuneration. With a larger

312Technically, it is common in the bargaining setting to assume that negotiations takes place in anticipation that other deals with other parties will be concluded. Therefore, each record company is seen by the platforms as the last negotiator.

contribution to the platforms' value of repertoire the average wholesale terms (that is to say net price) per unit will increase and will be greater than the terms for the marginal unit. Therefore, an implication of this sub-additivity claim is a "size effect" by which larger recorded companies will be in a position to obtain better commercial terms.

(519)The implication of the sub-additivity claim is that, post-merger, the merged entity would be in a stronger bargaining position as it would be in a position to offer a greater share of a platform's repertoire value. As a result, it would be expected to be able to extract higher rents in the negotiations with that platform.

(520)The Notifying Party itself does not rule out the possibility that there is sub-additivity of music repertoires but invites the Commission to demonstrate it. According to it, "in order to rely on decreasing marginal utility to identify alleged non-coordinated effects for the proposed concentration, the Commission would therefore have to provide concrete evidence that the parties' repertoires are substitutes at the platforms level" .

(521)The Commission has to provide evidence that the larger record companies are able to extract better terms, thereby leading to a significant impediment to effective competition, and that the substitution claim itself is correct. The Commission underlines that the substitution claim constitutes the conceptual and theoretical framework under which it undertook its quantitative analysis. It is a theoretical framework that is consistent with the other evidence that larger record companies are in a position to extract better terms and that hence a "size effect" exists. The Commission does not as such deny that larger companies may be in a position to extract better terms under different theoretical settings, as the Notifying Party itself claims . Indeed some bargaining setups are consistent with the above predictionsand others may predict multiple equilibria (that is to say where an equilibrium may exhibit either sub-additivity or super-additivity) . Ultimately, the relevant question is whether there is a size effect irrespective of the underlying sub-additivity or super-additivity of the value function. While theoretical models may predict sub-additivity or super-additivity (or both), what is important is that the Commission's investigation has confirmed that the size effect exists on recorded music markets. This evidence includes the results of the Commission's market investigation, the analysis of the commercial terms that Universal and EMI currently obtain, the merging parties' internal business strategy documents, as well as the results of the quantitative analyses undertaken by the Commission that will be set out in the following paragraphs and explained in detail in Annex 1 and 2.

(522)The Commission has carried out quantitative exercises (statistic and econometric tools), in order to test whether, in line with the qualitative evidence supporting this conclusion, the ability of a recorded music company to extract rents depends on the size of the recorded music company. To this end, the Commission has collected data from two sources. First, it has collected data from six major digital customers, which include the main digital customers across several Member States. Second, the Commission collected data from six record companies .

(523)The analysis shows that larger firms have a greater ability to extract higher rents/better terms out of the total retail revenues of the platforms (even after considering the possible negative impact of piracy on the total size of these revenues). In particular, the quantitative analysis has shown that: (1) the rent that record companies can extract increases with their size (measured in terms of the share of revenues that a recorded music company's repertoire offers contributes to a digital platform); (2) this effect is particularly pronounced for relatively small online platforms, but is also present in platforms with relatively larger bargaining power; and (3) there is a positive relation between the size of a recorded music company's repertoire and the wholesale price it negotiates with online platforms.

(524)This analysis is set out in more detail in Annexes I and II to this Decision, which are integral parts thereof, and which explain the methodology used by the Commission to analyse the available competitors' and customers' data, detail the reasons for the different robustness checks carried out by the Commission, respond to the Notifying Party's comments on the analysis undertaken and report in detail the different results obtained.

(525)The Commission notes that the results of the Commission's quantitative analysis are consistent with the qualitative evidence and in particular with the above analysis of Universal and EMI's contracts with platforms, that is to say that larger record companies are able to extract a larger rent from digital customers than smaller record companies particularly with respect to the smaller digital customers.

7.1.5.5.Results of the market investigation

(526)The Commission's position on the basis of its assessment of the commercial terms that digital customers obtain is corroborated by submissions that digital customers and other third parties have made throughout the Commission's investigation.

(527)The large majority of customers (13 out of 16) and competitors confirmed that the size of a recorded music company matters greatly for its bargaining position and leverage in commercial negotiations with digital customers.

(528)The majority of respondent digital customers anticipate an extension of Universal's terms to EMI's repertoire as a result of the proposed concentration . This would produce an immediate negative effect on the licensing terms that digital customers obtain. They anticipate that it would lead to increase the

merged entity's bargaining power and hence that it would lead to less favourable terms, notably higher prices, higher advance payments and higher advertising commitments. The majority of digital customers anticipate that they would be confronted with less favourable terms and conditions if the proposed concentration were to go ahead.

(529)One digital customer confirms the following:

(530)"(The merger) would mean that close to 50% of the music offer would be dominated by 1 supplier which is unacceptable. The financial requirements from such a company if based on current business terms asked by Universal would deeply hurt our business model and damage the current user experience" .

(531)Virtually all competitors expect that the proposed concentration would increase the ability and incentive of the merged entity to increase revenue payments and impose licensing terms that are more disadvantageous to online platforms . None of the competitors believe that digital customers can avoid these effects. In particular, no competitor believes that digital customers could credibly negotiate first with Universal's competitors and subsequently exert commercial pressure on Universal to improve its licensing terms .

(532)The results of the Commission's market investigation are therefore twofold:

(a)They confirm the outcomes of the comparison of Universal's and EMI's agreements, namely that large record companies are able to negotiate more onerous terms with digital customers; and

(b)They indicate that this link between size of record company and onerous terms and conditions is not limited to Universal and EMI, but applies across the whole recorded music industry. This result is corroborated by the quantitative analysis undertaken by the Commission.

7.1.5.6.Evidence from the merging parties

(533)Internal documents confirm that Universal […]*.

(534)[…]*:

(535)[…]*.

(536)[…]*:

(537)[…]*.

(538)[…]*.

(539)[…]*:

(540)In light of all the above, the Commission concludes that, due to its increased size in the wholesale of recorded music to digital customers, the merged entity will likely apply commercial conditions that are less advantageous to digital customers, including higher prices, if the proposed concentration goes ahead.

7.1.5.7.EEA-wide market for the wholesale of digital recorded music

(541)The Commission also considers that these effects would be likely to materialise both at the EEA level and at the national level.

(542)At the EEA level, the merged entity would significantly increase its size and market share and become the market leader by some distance in relation to its next competitor.

(543)Moreover, the Commission's analysis of the commercial terms that Universal and EMI obtain from digital customers, which shows that, already today, Universal manages to achieve substantially better terms from these customers than EMI, covers the eleven digital customers, which account for the majority of each of Universal's and EMI's sales in the EEA.

(544)In addition, the findings of the Commission's quantitative analysis, according to which there is a positive relationship between a recorded music company's size and its ability to extract revenues from digital customers, apply to thirteen EEA countries.

(545)The Commission concludes that its findings, according to which due to its increased size in the wholesale of recorded music to digital customers, the merged entity would likely apply commercial conditions that are less advantageous, including higher prices, to digital customers apply to the relevant market if it were to be considered as EEA-wide in scope.

7.1.5.8.National markets for the wholesale of digital recorded music

(546)At the national level, the Commission notes that, in Austria, Belgium, Bulgaria, the Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece, Iceland, Ireland, Italy, Latvia, Lithuania, Luxembourg, Malta, the Netherlands, Norway, Poland, Portugal, Romania, Slovakia, Spain, Sweden and the United Kingdom, the merged entity would significantly increase its size and market share and become the market leader by some distance in relation to its next competitor.

(547)Moreover, for twenty Member States, namely Austria, Belgium, Czech Republic, Finland, France, Germany, Ireland, Italy, Latvia, Lithuania, Luxembourg, Malta, the Netherlands, Poland, Portugal, Romania, Slovakia, Spain, Sweden, the United Kingdom, as well as for Norway, the eleven key digital customers for which the Commission carried out a comparison of the relevant commercial terms account for the majority of Universal's and EMI's local revenues.

(548)Similarly, a comparison of the key commercial terms between Universal and its largest customers in Bulgaria , Denmark and Greecewith the key commercial terms between EMI and its largest customers in those Member States, confirms the Commission's findings that Universal also achieves better terms than EMI in relation to each of those Member States.

(549)Finally, in Estoniaand Iceland there are significant national players, Musikaa24 and Tonlist.com, which given their small size and operations are unlikely to have a strong bargaining power in relation to record companies. It can therefore be reasonably expected that the market position of these players would be affected in the same way as other smaller digital platforms.

(550)The Commission's quantitative analysis confirmed the likely effect of the proposed concentration on digital platform's margins in thirteen Member States, namely Austria, Belgium, the Czech Republic, Finland, France, Germany, Greece, Italy, the Netherlands, Poland, Spain, Sweden and the United Kingdom, as well as in Norway.

(551)The Commission considers that these factors confirm the Commission's general position that the proposed concentration would have an impact on those markets where post-merger, Universal would increase its size to such an extent that it would become the clear market leader, with considerable distance in relation to its next competitor.

(552)In light of the above, the Commission concludes that its findings, according to which due to its increased size in the wholesale of recorded music to digital customers, the merged entity would likely apply commercial conditions that are

less advantageous to digital customers, including higher prices, if the proposed concentration were to go ahead, apply to the relevant market if it were to be considered as national in scope in Austria, Belgium, Bulgaria, the Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece, Ireland, Italy, Latvia, Lithuania, Luxembourg, Malta, the Netherlands, Poland, Portugal, Romania, Slovakia, Spain, Sweden, the United Kingdom, as well as in Iceland and Norway.

7.1.5.9.Control share analysis

(553)The Commission has explained the role of music publishers and the relevance of music publishing rights in Recitals (91)-(98) and (179)-(183) of this decision.

(554)In Universal/BMG, the Commission observed that, in order to offer any song for download or streaming, digital music retailers need to acquire a license for both the relevant recording rights, that is to say the right covering a specific singer's interpretation of the musical work, and the relevant publishing rights, including the copyright belonging to the author of the musical work and covering its reproduction (“mechanical right") and its commercial use (“performance right"). The Commission also noted that, while recording rights are typically controlled by the record companies, publishing rights were traditionally controlled by the collecting societies, which were under the obligation to license these rights on a non-discriminatory basis. The Commission, however, noted that certain publishers had started withdrawing the online mechanical rights for their Anglo-American repertoire from the various collecting societies around the EEA with a view to re-acquiring control over the exploitation of these rights in relation to digital music retailers and being able to license these rights directly, or via rights management entities or collecting societies appointed as agents, on a pan-European basis. The Commission therefore took the view that, in the near future, digital music retailers willing to offer any given song for download or streaming would have to enter into licensing agreements not only with record companies for the relevant recording rights and with collecting societies for the relevant publishing rights (generally granted under blanket licences), but also with publishers for the online Anglo-American mechanical rights.

(555)It is important to underline that following the Commission's approach in Universal/BMG, revenue based market shares were not considered the most relevant proxy for the market power of the merged entity whether on the music publishing markets alone or across music publishing and recorded music. This stemmed from the fact that the Commission considered that it was the overall size and the characteristics of the repertoire that were of major importance for the bargaining position of the publisher. The Commission therefore assessed the post-merger market power of Universal on the basis of so-called "control shares".

(556)Control shares can be used to assess market power purely on a music publishing level by looking at the number of songs within a given selection over which the publisher "controls" at least part of the publishing rights; and across recording and music publishing rights by looking at the number of songs within a given selection over which the company "controls" the recording rights or at least part of the publishing rights. This is the number of songs in the given selection over which it therefore enjoys the right to at least veto the licensing to digital music retailers.

(557)In the Universal/BMG case, after carrying out its analysis, the Commission concluded that the proposed concentration would bring Universal's control share in a number of Member States to over 50% (that is to say that Universal would control either the recording or at least part of the publishing rights in more than half of the songs included in the sample analysed by the Commission in the relevant Member States), which the Commission viewed as the threshold above which a record and publishing company would be "incontournable" from the point of view of digital music retailers, therefore enjoying significant market power and having the ability and incentive to increase prices .

(558)Given the above, the Commission considers that, in order to assess the degree of market power held by the merged entity vis-à-vis digital music retailers following the proposed concentration, it is also necessary to consider the merged entity's combined market position in the recorded music and in the music publishing sectors.

(559)As explained, the Commission considers that the proposed concentration would give rise to a significant impediment to effective competition in several relevant markets based on the overlap between the merging parties' activities in the recorded music sector alone. The Commission, however, also considers that the fact that a recorded music company is also active in music publishing confers upon it an additional leverage in its bargaining with digital customers. This reflects the fact that, consistent with the theory of harm developed in this case with respect to recorded music, having certain publishing rights which cover songs that are different from those songs in relation to which it holds the recording rights, further extends a recorded music company's product line. This, in turn, increases a record company's size and its bargaining power in relation to digital customers.

(560)In other words, in this case, the Commission considers control shares as an "aggravating factor", which exacerbates the anti-competitive effects deriving from the proposed concentration based on the combination of the merging parties' activities in the recorded music sector (which, in and by themselves, is sufficient to give rise to competition concerns).

The Notifying Party's view

(561)The Notifying Party strongly contests the applicability of any control share theory of harm in this case, arguing, among other things, that the Commission’s prediction formulated in 2007 as to how the licensing of mechanical and performance rights to digital music retailers would evolve proved wrong. In more detail, the Notifying Party claims that, contrary to what the Commission predicted in 2007, collecting societies continue to play a key role in the licensing of publishing rights for the Anglo-American repertoire. The Notifying Party therefore submits that publishers do not have the ability to dictate or otherwise influence the terms at which at least the performance rights related to their Anglo-American repertoire are licensed or to veto the licensing of any such rights.

(562)Furthermore, the Notifying Party highlights the following facts:

(a)publishing and recorded music rights are licensed and negotiated entirely separately;

(b)the negotiations take place at different points in time – i.e. digital service generally negotiates licences over publishing rights after a licence for the recorded music rights has been agreed; and

(c)the separate commercial arrangements entered into for the licensing of these different rights are entirely independent of each other, for example there is no conditionality between these arrangements. […]*

(563)The Notifying Party also challenges in several respects the methodology that the Commission used in Universal/BMGto calculate control shares. According to

the Notifying Party, the Commission should not limit its analysis to a sample of the top 100 singles chart hits in each Member State, but it should also include catalogue songs and albums, which represent a significant percentage of the revenues of digital services. In its Response to the Statement of Objections, the Notifying Party also argues that the sample used for the purposes of calculating control shares should also include back catalogue music, as well as tracks that do not make the weekly chart hits. The Notifying Party also argues that it would be a fundamental error of assessment on the part of the Commission to attach any relevance to Universal's "control share" data in the absence of equivalent data from other market participants such as Sony, Warner and the indies and in the absence of any evidence of past effects resulting from previous mergers in the industry.

(564)Moreover, in the Notifying Party's view, in its analysis the Commission should not rely on weekly single chart hits, given the nature of chart hits which are not a meaningful measurement of market power (as opposed to revenues) and given the absence of a direct and consistent relationship between chart presence and commercial success. The Notifying Party claims that an assessment based on historical chart hits is not relevant in any ex ante assessment of the likely future effects of a merger, as past chart hits cannot reliably be expected to result in future chart hits, nor translate into revenue based market share.

(565)In the Response to the Statement of Objections, the Notifying Party re-iterates the argument that, since recorded music and music publishing belong to separate product markets, the combination of Universal's and EMI's recorded music activities with Universal's music publishing activities should be analysed under the framework provided for by the Guidelines on the assessment of non-horizontal mergers under the Council Regulation on the control of concentrations between undertakings of 18 October 2008 ("Non-Horizontal Merger Guidelines"), rather than with reference to the notion of control shares. In this context, including in the Charles River Associates' consulting firm ("CRA") economic report attached to the Response to the Statement of Objections, the Notifying Party notes that the proposed concentration, to the extent that it leads to a reduction in the fragmentation of rights, would likely lead to efficiencies.

(566)The Notifying Party also further elaborates on the reasons why, in its view the merged entity would not have the ability and incentive to leverage recorded music and music publishing rights. In more detail, the Notifying Party argues that, in order to engage in any such leveraging, the merged entity would have to be able to credibly threaten digital services to "hold out", that is to say not license its repertoire. The Notifying Party submits that any such threat would not be credible either in a so-called "full rights" scenario, where the merged entity holds both the recording and the publishing rights to a given track or album or in a so-called "partial rights scenario", where the merged entity holds either the recording rights or part of the publishing rights to a given track or album. In this regard, the Notifying Party further maintains that the results of the market

investigation support its views and that the internal documents relied upon by the Commission do not have sufficient probative value.

(567)Finally, including in the CRA economic report attached to the Response to the Statement of Objections, the Notifying Party claims that the Commission did not present any evidence, similar to that which it collected in relation to the recorded music theory of harm articulated above, to support its control share theory of harm.

The Commission's assessment

(568)The Commission notes that, for the reasons set out at recitals 553 to 560, it considers control shares to be a viable tool to assess the market power of an integrated recorded music / music publishing company in relation to digital customers. While the relationship between recorded music and music publishing could also be assessed under the analytical framework provided for by the Non-Horizontal Merger Guidelines, any such analysis would be aimed at assessing the possible foreclosure effects that may result from any attempt by the merged entity to leverage its market power in recorded music into music publishing or vice versa. Conversely, the control share analysis is aimed at assessing the market power enjoyed by an integrated recorded music / music publishing company in relation to digital customers.

(569)The Commission also notes that, while the re-unification in a single owner of certain recording and music publishing rights may lead to some efficiencies, the fact remains that, to the extent that the proposed concentration would lead to an increase of the repertoire that is controlled by the merged entity, by combining Universal's existing recorded music and music publishing activities with EMI's recorded music activities, it will result in an extension of the merged entity's product line and an increase in its bargaining power.

(570)Ability and incentive to engage in joint negotiations. The Commission considers that the merged entity would likely have the ability and incentive to engage in joint negotiations covering both recorded music and music publishing.

(571)The Commission notes that contrary to the Notifying Party's claim, it is likely that the merged entity would be able to present a credible threat to digital customers not to license the combined recorded music / music publishing repertoire that it would control following the proposed concentration. This reflects the fact that the Commission has found that the ability to make such a threat is likely to be the key reason larger record companies (that is to say record companies which control a larger repertoire) are able to secure better terms and conditions than smaller record companies. The Commission therefore considers that these bargaining dynamics, which already take place in the recorded music sector, would also likely apply in the context of the negotiations between an integrated recorded music / music publishing company and its digital customers.

(572)The Commission further notes that none of the arguments put forward by the Notifying Party on Universal's lack of incentives in engaging in such conduct post-merger is founded. This reflects the following:

(a)that some of the arguments apply to both recorded music and music publishing and the Commission's analysis has demonstrated that the factors referred to by the Notifying Party probably do not prevent larger record companies from credibly threatening to hold out their recorded music repertoire to obtain better terms;

(b)that other arguments, such as the one based on the complex cost/benefit calculation in which an integrated recorded music / music publishing company would have to engage before deciding whether it has an incentive to threaten to hold out, do not seem to prevent record companies from doing the same in a scenario where only full rights exist and where therefore holding out is potentially more costly; and

(c)that the only argument that is specific to music publishing (and is related to the possible negative effects of holding out on the relationship between a music publisher and collecting societies) is not supported by sufficient evidence showing that these possible negative effects are such as to change a company's incentives.

(573)The market investigation in this case confirmed that, contrary to the Notifying Party's claims, an integrated recorded music / music publishing company, such as the merged entity, would have the ability to engage in joint negotiations for its recorded music and music publishing repertoire.

(574)In more detail, the results of the market investigation suggest that publishers (or at least the major ones, which have withdrawn the mechanical rights for their Anglo-American repertoire from collecting societies) have the ability to influence the licensing terms of their repertoire to digital customers whether by directly participating in the negotiations with the major customers or by approving the agreements negotiated by collecting societies or rights management entities, which they have appointed as rights administrators. This finding is consistent with the conclusions reached by the Commission in the recent Sony/Mubadala/EMI Publishing decision .

(575)Hence, contrary to the Notifying Party's claims, record companies, including the merged entity, do not need the cooperation of collecting societies to block the licensing of, at least, the mechanical rights for their Anglo-American repertoire.

(576)The Commission also disagrees with the Notifying Party's claim that, for the control share theory to work, the whole system by which recorded music and digital publishing rights are currently licensed in the industry would have to be restructured to allow record companies / music publishers to essentially pick and choose the tracks to license (the ones for which they have full rights) and those to (threaten to) hold back (the ones for which they have partial rights). As explained, the way an integrated recorded music company / music publisher would leverage its combined presence in the two sectors in relation to digital

customers is similar to the way a recorded music company does so in recorded music, that is to say by leveraging its increased size to obtain better terms and conditions. This negotiation strategy does not require any prior re-organisation of the way in which the tracks are licensed, particularly in light of the fact that it is used in the recorded music sector, where all tracks are fully owned by record companies and partial rights do not exist.

(577)As a result, the Notifying Party's claim that the merged entity would lack the ability to leverage its combined presence in recorded music and music publishing is rejected.

(578)This conclusion is supported by […]* . […]*.

(579)In its Response to the Statement of Objections, the Notifying Party submits that the fact that […]* does not mean that Universal has the ability and incentive to leverage its combined presence in recorded music and music publishing. The Notifying Party also submits, including by reference to a statement by […]* and to SACEM's reply to the market investigation, that any such leveraging does not take place in practice.

(580)The Notifying Party also submits that the analysis of the responses of digital customers to the market investigation did not reveal any instance of joint negotiation or leveraging between recorded music and music publishing rights. The Notifying Party further submits that digital customers' replies to the market investigation support the view that Universal does not have the ability and incentive to engage in a leveraging of its recorded music and music publishing rights.

(581)The Commission rejects the Notifying Party's arguments.

(582)First, the provisions of the agreements between Universal and SACEM […]*.

(583)Second, […]*.

(584)Third, the market investigation revealed that, according to some market participants, in the future, integrated recorded music/music publishing companies would be able to use their leverage over a wider recorded music and music publishing repertoire to increase their remuneration. Although not yet systematic, this type of leveraging is already happening in the market place. For the sake of clarity, such leveraging does not necessarily entail joint negotiations but can also include various "hold out" strategies, essentially consisting of delaying the licensing of the recorded music repertoire while a deal for the music publishing repertoire is being negotiated or vice versa.

(585)For example, […]* streaming service provider in the EEA, submitted that it has occasionally faced joint negotiations between recorded music and music publishing rights, the reason clearly being that it creates a strengthened bargaining position allowing the integrated recorded music and music publishing

companies to leverage the collective bargaining power across the label and publishing arm .

(586)Moreover, competitors […]* explained that […]*. According to […]*, this conduct is driven by a rational profit maximising strategy. Any single economic unit, such as Universal or the merged entity, will pursue what is the most profitable for the firm as a whole and this necessarily implies a strong coordination between the recorded music and music publishing businesses. […]* submits that on the upstream A&R level, majors are increasingly seeking to attract recording artists who are also songwriters by presenting themselves as a one-stop-shop for recorded music and music publishing services. On the downstream level, deals with the recorded music and music publishing divisions are negotiated, parallel with one another, with completion of one subject to the satisfactory completion of the other. According to […]*, the majors can also engage in "negative" forms of coordination, that is to say, the recorded music division of a major music company might, for example, postpone a deal as long as no agreement is reached with its music publishing arm and vice versa.

(587)[…]*. According to […]*, one division will often assess the ability to leverage the other division during on-going negotiations.

(588)[…]* further submits that any attempts by Universal to deny the value of coordination between a major music company's recorded music and music publishing businesses would be contrary to its own statements made in previous Commission investigations. In the context of the proposed joint venture between Time Warner and EMI and the merger between AOL Inc. ("AOL") and Time Warner in 2000, Universal made several submissions […]* .

(589)In its Response to the Statement of Objections, the Notifying Party submits that […]*'s submission is unsubstantiated in that […]* and that therefore the Commission should attach no weight to this submission.

(590)The Commission considers that […]*’s statements, together with the additional elements referred to by the Commission in this Section, support the conclusion that the merged entity would have the ability and incentive to engage in joint negotiations covering both recorded music and music publishing.

(591)The Commission considers that the examples referred to by the Notifying Party in its Response to the Statement of Objections do not support the proposition that Universal (or the merged entity) would not have the ability to engage in joint negotiations or leveraging. If anything, they show that, until today, Universal has not engaged in such conduct, which does not detract from the merged entity's ability to do so in the future.

(622)The majority of digital customers, that is to say 15 out of 26 digital respondents, have confirmed that, if their licensing costs were to go up, they would pass on these costs at least in part to their end-users in the form of retail price increases. These customers, who run their day-to-day operations on markets for the retail sales of digital music, therefore consider that despite any alleged constraints posed by piracy or the price of physical music, they would nonetheless pass on their increased licensing costs to end users on in the form of full or partial price increases.

(623)In any event, the Commission has demonstrated that the proposed concentration is likely to result in the imposition of commercial terms that are more disadvantageous to digital customers that offer digital music services, and this is particularly likely for those digital customers that are currently smaller than Apple/iTunes, but that are important because they bring innovative ways to consume music to the relevant markets. In this manner, the proposed concentration would cause consumer harm in the form of reduced innovation and consumer choice for digital music. The Commission reiterates that this would also have a negative impact on cultural diversity, as sufficient access to a wide-ranging music repertoire for as many digital platforms is crucial to preserve that diversity. Article 167 (4) TFEU requires the Union to take cultural diversity aspects into account in its actions under the other provisions of the Treaties, including the Union competition rules. Moreover, the UNESCO Convention on the protection and the promotion of the diversity of cultural expressions to which the Union is a Party sets out a comprehensive set of guiding principles relating to the diversity of cultural expressions. The case-law has confirmed that other Treaty objectives may be taken into account in the context of the assessment of a concentration .

(624)The Notifying Party contests this conclusion. The Notifying Party claims that any negative impact on consumer choice on innovation must be based on a theory of harm that the ability and incentive for a merged entity to bring innovative "pipeline" product would be reduced as a result of a proposed concentration. In that context, the Notifying Party refers to paragraph 38 of the Horizontal Merger Guidelines, which inter alia provides that a merger may significantly impede effective competition if it involves two important innovators, for instance two companies with "pipeline" products related to a specific product market. The

Notifying Party suggests that any theory of harm that is not based on this part of paragraph 38 of the Horizontal Merger Guidelines is not supported by those same Guidelines nor by the Commission's decisional practice. The Notifying Party claims that this part of paragraph 38 does not apply here and that even if Universal and EMI were considered to have important pipeline products, there is nothing to suggest that alternative, innovative rivals to the merged firm will not step in to fill any gap that is created.

(625)The Notifying Party submits that in any event post-merger, Universal would continue to have the incentive to license all online platforms widely and at low prices. It would have the incentive to do so as it would favour the launch and expansion of additional online platforms in order to reduce piracy […]* .

(626)In this context, the Notifying Party submits the following:

(a)that Universal has no power to stop or decelerate the irreversible decline of the physical music market. Hence, Universal's incentive is to develop new digital platforms and entice customers away from illegal content;

(b)that streaming services in particular are a strong driver of digital revenues growth rate; and

(c)that the concern that Universal has any incentive to delay or block the launch and development of innovative services is inconsistent with the Commission's recognition that Universal’s strategy is to license […]*.

(627)The Notifying Party claims that rather than Universal having the ability and incentive to shape the operation of new digital music services, the fundamental commercial interests of Universal and such services are fully aligned . In relation to download services […]* . In relation to streaming services, Universal's incentive would be […]* .

(628)To further support its claims, the Notifying Party submits that Universal has licensed its repertoire to over […]* digital services in the EEA and some [...]* of which in 2011. The Notifying Party refers to the following examples:

(a)Universal was the first record label to sign a deal with Spotify;

(b)Universal was the first record label company to license the services "Nokia Comes with Music" involving unlimited music downloads to a mobile phone;

(c)Universal licensed the world’s first unlimited DRM free subscription service in partnership with […]* but the venture failed because other companies would not license their repertoire;

(d)Universal has supported the development of various music-related apps on mobile devices;

(e)Universal has been at the forefront of developing innovative digital music video services including VEVO;

(f)Universal has keenly supported the integration of subscription and streaming services within TV set top boxes and home audio systems such as the Sonos system; and

(g)Universal has been at the forefront of introducing digital services to telecom companies in order to develop new mobile/music packages for consumers.

(629)The Commission disagrees with these claims.

(630)As a first observation, the Commission notes that, by claiming that any negative impact on consumer choice or innovation and hence cultural diversity on digital music markets must be based on the demonstration that Universal would have a decreased ability and incentive to bring important pipeline products to those markets, the Notifying Party makes an assessment on the wrong market level.

(631)Innovation in online music is mainly driven by digital platforms, which develop new ways of distributing music. Spotify, for instance, makes use of cloud storage and broadband capabilities in order to give access to music from multiple devices.

(632)Record companies generally do not develop the new digital applications themselves. Instead, they license innovative services launched by third parties such as Apple/iTunes' Match, Spotify's and Deezer's streaming services, Nokia Comes With Music. As record companies provide these customers with the music they wish to offer, they have a decisive influence over the development of these services insofar as they have different policies towards the licensing of new digital services. Hence, it is the ability and incentive of record companies to license their music to innovative digital customers that counts, and not the pipeline products of record companies themselves. Any negative impact on consumer choice and innovation on digital music markets is therefore best assessed on the basis of the impact that the proposed concentration would have on the ability and incentive of the merged entity to license new digital services on terms that are as attractive as the terms that are offered on the markets pre-merger.

(633)The Commission notes that this approach is fully in line with its Horizontal Merger Guidelines. The key question that the Commission needs to answer in this case is whether the proposed concentration would increase the market power of Universal to such an extent that a negative impact on prices can be expected. The Commission reiterates that the Horizontal Merger Guidelines make it clear that what is meant by "increased market power" is the ability and incentive of the merged entity to profitably increase prices, reduce output, choice or quality of goods and services, diminish innovation or otherwise influence parameters of competition. The expression "to increase prices" is consistently used in the Horizontal Merger Guidelines as shorthand for the various ways in which a

merger may result in competitive harm. Accordingly, the Horizontal Merger Guidelines clearly provide that the Commission can assess whether the proposed concentration would increase the ability and incentive of Universal to reduce consumer choice and to diminish innovation as a result of its licensing practices towards digital customers, especially those customers that bring innovative ways to consume music to the various digital music markets. This approach is also in line with the Commission's obligation to preserve cultural diversity in the EEA.

(634)The Notifying Party's claim that Universal would continue to have the incentive to license digital customers subject to terms and conditions that are as attractive as those that are offered on the digital music markets pre-merger is not conclusive. The launch and growth of new digital services involves possible changes of the market structure and as a consequence, possible changes in the market shares of record companies. The Notifying Party itself submits that […]*. Accordingly, it is not clear from the outset that a market leader, such as Universal, always and by necessity has a strong incentive to support new digital services.

(635)Importantly, the evidence contradicts the Notifying Party's claims as it shows that its commercial terms for those smaller platforms that bring innovative ways to consume music to the relevant markets are significantly […]* than those it extracts from more established services such as Apple/iTunes. This is counter-indicative of the Notifying Party having a primary interest in supporting the entry and expansion of these services. The Commission would then expect to see more attractive terms being offered to those customers.

(636)According to the results of the Commission's qualitative and quantitative analysis, margin effects of the proposed concentration would vary widely depending on the size of digital customers. In particular, the Commission found that the Notifying Party's ability to extract more onerous terms is even stronger in the case of small digital customers with less bargaining power. The digital customers with less bargaining power are the emerging, innovative online platforms, such as streaming services and MNOs. The Commission also found that even for an innovative service of a customer that otherwise has a relatively strong bargaining position […]*. Such onerous licensing terms are likely to have a negative impact on the development of these customers and their ability to geographically expand to the various digital markets. This is turn likely to negatively affect innovation and consumer choice for the end users of such services, in particular for customers who are not yet covered by these services.

(637)The Commission disagrees with the Notifying Party's suggestion that consumer choice and innovation and hence cultural diversity could only be impacted in this manner, if the Commission were to have solid evidence that as result of Universal's licensing practices, many digital customers have failed to launch their services or have collapsed after a short time. These consumer benefits can already be negatively impacted by licensing practices that are likely to significantly hamper digital customers' ability to expand their offerings or to shape their business model freely, without resulting in their total collapse or withdrawal from digital music markets.

(638)The Commission's investigation has produced examples of digital customers indicating that their cost of licensing, in particular for the content of large record companies such as Universal, has been the main barrier for this geographical expansion. A digital customer makes the following comments:

"This level of music rights prevented [Digital Customer] from expanding in other european countries earlier, since it first had to generate high revenues in its domestic country" .

(639)Negotiations for geographical expansion are very demanding from a financial point of view, particularly since major companies demand upfront payment and guaranteed minima. Therefore, securing rights on a multi-territorial basis requires an extremely high financial investment (payment of advances in particular). This has not been possible for this digital customer at the time of its creation and for the following years. Hence, the geographical scope of its service was limited.

(640)Other digital customers indicate that such licensing practices have restrained them from offering new products and services in Member States other than their own. […]*, a […]* digital customer, expresses the following view: "we have been exploring the possibility of launching a Streaming Service. So far the minimum guarantees are unrealistic". To limit the geographical expansion of digital music services within the EEA reduces consumer choice and innovation in all the territories where end-users do not benefit from these services.

(641)Digital customers and other third parties have confirmed that the impact of Universal's licensing practices goes far beyond occasional disagreements on the business models for digital music services that are sustainable.

(642)Online platforms consider the availability of Universal content in innovative products and services that require a license to be very important. However, online platforms overwhelmingly confirm that one of the main barriers for entry in digital recorded music markets is the licensing terms and conditions they obtain for their music content. In particular, high advance payments and minimum revenue guarantees, requested by large record companies, significantly increase the risk for nascent digital services. By extension, these licensing terms are one of the main obstacles to the emergence of innovative business models that offer comprehensive consumer choice for digital music.

(643)Accordingly, online platforms note that the following are the main barriers for launching a successful digital music service:

“enormous label advance payments"

“The key issue has not been lack of general support on the service, the issue has

always been the revenue guarantees from Universal, Sony and Warner (…) Amazon's and especially Google's fallout with the labels on scan and match cloud locker rates was well publicised in the press. Consequently, both have now launched a less innovative dumb cloud locker, which does not need the labels' meta-data"

Record companies “consistently demand a minimum guarantee per user which stifles innovation and which is not reflective of their risk profile n these types of business models" .

"[The] cost of licensing, including the size of possible advances and/or minimum revenue guarantees" .

"It is generally difficult to propose innovative business models to record companies due to licensing issues (i.e. a new model needs certain rights in order to be attractive) and due to high demands (margins, advances and/or market share guarantees)" .

"The main barrier to launch a music service is the lack of flexibility of the business models. Industry (record labels, publishers and collecting societies) difficult standard terms and conditions lead to an almost lack of margins that result in the critical need of massive consumer volumes to achieve a certain level of profit. This, in connection with extremely high advance payments, makes impossible for almost every entrepreneur the launch of any service, making business extremely risky" .

(644)Online platforms confirm the link between the merged entity's increased ability and incentive to impose other more onerous licensing terms and the negative impact on innovation and consumer choice. The vast majority of online platforms expect that post-merger, the commercial terms and conditions of Universal would be applied to the combined Universal and EMI repertoire and that the merger will lead to less innovation and consumer choice with the removal of a competitor which applied more reasonable terms and conditions to new digital services. Digital customers make the following comments:

"[T]he merged entity will have the power to increase prices and to control and shape the new services that will be launched on the market. The price increases could endanger the presence on the market of digital music distributors as this would further reduce their very thin margin" .

"Les labels ont été peu supportives de l’innovation. Toute innovation nécessite de longues explications et doit être assortie de garanties financières importantes pour être acceptée par les gras labels. Et elles ne le sont pas toujours, ce qui pénalise fortement alors notre compétitivité par rapport aux offres illégales. Cela est particulièrement vraie des principaux labels, les indépendants étant plus audacieux et ouverts à l’innovation" .

"Major companies have never been supportive of innovative services ([...] the launch of innovative service is the occasion to negotiate huge advances or minimum guarantee and sometime a stake in the company providing the service). Universal clearly tried to stop free on demand streaming services […]. Warner Music tried to impose outrageous obligations to license its catalog [...]. On the opposite indies and aggregators are often very supportive" .

"Potential effects that the merger might have are a reduction in choice of music for consumers and higher prices for consumers. There would also be a potential chilling effect in innovation by digital music services due to a greater concentration of licensing power and the removal of the smallest and arguably most innovative of major labels" .

"The creation of a merged Universal/EMI would create a major with even greater market power that could extract even more advantageous terms from retailers of its music while eliminating an innovative partner" .

"it is [Digital Customer]’s experience that EMI has tended to be more supportive of innovation. Conversely, Universal tends to be more reluctant to license new products and services, and has sought to impose terms on those products and services beyond licensing content rights. [Digital customer] views the availability of Universal content in innovative products and services that require a license as very important. Were Universal to succeed in acquiring EMI, [Digital customer] would expect there to be a major negative impact upon its ability to obtain music rights on acceptable terms for such products and services post merger and hence a reduction in innovation across the sector" .

"The first negative aspect arising from the merger could the increase of the price which is the main factors that influence in a decisive way the purchase of goods and services. Keep as open as possible the market is essential to ensure competition and fight against piracy while putting the music market in the hands of fews would prevent the evolution of the market with the result of push up prices and handle it in an unfair way for consumers and music itself" .

"We believe that the merger could lead to less diversity in terms of offer range, and could possibly have an impact on retail price, should the price of the

catalogue rise" .

"Generally, record companies are not very supportive due to their strong market position and they often blocked innovative models. Universal’s approach with respect to innovative ideas is particularly poor" .

(645)The Commission also finds that the proposed concentration would in addition reduce consumer choice and innovation and hence cultural diversity as a consequence of the increased ability and incentive of the merged entity to shape the business model of digital customers and increase retail prices. The Commission's investigation confirms that Universal already uses its bargaining power today in order to […]* on free services offered by digital customers to their end users.

(646)An online platform observes that record companies are, in general, more supportive of paying services than advertising-funded services where access to music is free for end users. In particular, Universal has strongly opposed unlimited and free streaming services.

(647)A digital platform describes the following precedent:

"Universal is also very demanding when a new service is to be launched on the market. Universal is clearly aware that its catalogue is a "must have" catalogue necessary for every new service. […]* recently negotiated the launch of a new service based on bundles of song downloads. To accept to license its catalog, Universal demanded modifications of the main features of the new service including the two main aspects of it: the public price and the number of song included in the bundles. Universal retained its approval until the very last moment. The cost and the difficulties encountered to launch innovative services are the main reason why […]* stopped to exploit directly such service" .

(648)[…]* describes the following post-merger scenario:

"Universal would serve as the gatekeeper determining which new and innovative digital music services will be allowed to survive. It would have the power to unilaterally withhold support from new digital services or to license them only at supra-competitive rates and terms. Universal also would have the ability to tailor the structure of fees in a way that it prefers. […] Even if another label tried to promote innovation by being the first to reach an agreement with a start-up service, the larger post-merger size of Universal and its status as absolute “must have” likely would give it enough leverage to be able to drive its own terms, thus discounting the precedential effect of any deal a service may have already concluded. Whatever economics might be obtained by another label in exchange for being first to sign on to a new service, Universal might very well demand a premium on those economics. Anticipating these dynamics, digital music service

providers would have little incentive to exert the effort to enter into such early deals with smaller companies in the first place, recognizing that they must go to Universal to determine what models Universal will support and how much it will charge to license its catalogues before proceeding with any other discussions. Once a new service reaches terms with Universal, it would be largely committed to the design and terms of its service, and the design and terms of its service would be shaped to Universal ’s preferences. This means that Universal would be able to control or influence the business models of new services, in addition to raising the prices of their music rights" .

(649)[…]* also observes the following:

“Universal has already shown a propensity to hold out, and its actions may have led to the bankruptcy of a service called Beyond Oblivion, as well as forced the sale to Apple of a service called lala. […] Deezer was also required to litigate in France (Judgement of the Tribunal de Grande Instance de Paris of 5th September 2011) in order to ensure it could remain in business and continue to offer Universal’s catalogue. If Universal were to be allowed to acquire EMI, the leverage over Universal that could be achieved by a service signing up the two remaining majors would be even further reduced”.

(650)The decision n°11/55727 of the Tribunal de Grande Instance de Paris of [read: 5 September] 2011 follows Universal's attempt to impose certain limitations on Deezer's free streaming service in France and to prohibit Deezer from distributing its recorded music if it did not comply with such limitations. These limitations of use consisted of:

(a)The user being required to subscribe to the Deezer service after having listened to five consecutive tracks during the same visit to the website, or after having listened to the same sound recording 3 times in the same month;

(b)A sound recording may not be listened to more than five times following a period of six months during which it may be listened to without limitation; and

(c)Revenues for listening to a sound recording lasting longer than fifteen seconds may not be less than EUR 0.003.

(651)Faced with Deezer's refusal to renew the licensing agreement for the distribution of Universal's recorded music content at these terms and with Deezer not ceasing to exploit its content in its service, Universal brought a legal action before the Tribunal de Grande Instance de Paris. In its decision of [read: 5 September] 2011, the French Court ruled on an interim basis in favour of Deezer, finding that:

(a)Universal represents between 35% and 45% of the online music market. It is not only the largest catalogue in size, but also the most important in terms of hits since it includes 50% of the Top 100 hits. It may be considered as an unavoidable and indispensable element for online platforms in terms of size and coverage. Universal is therefore in a dominant position in France ;

(b)Deezer has adequately established that imposing limited listening of five listens per track on its five million active users would cause a discriminatory competitive advantage in favour of its competitor Spotify ;

(c)Furthermore, if Deezer were to accept Universal's limitations on its service, Deezer's 3.6 million non-subscriber users would divert from the website and would go to more general websites such as You Tube. This would endanger the very existence of Deezer ;

(d)Universal's refusal to supply its catalogue due to failure to accept its terms, which differ from those in the previous contracts, amounts to an abuse of dominant position. Such refusal is likely to harm consumers ; and

(e)Therefore, the French Court dismissed Universal's claims and ordered Universal to pay Deezer the sum of EUR 10 000 .

(652)This example litigation illustrates Universal's ability to shape business models in order to impose limitations on free services and thus to indirectly increase retail prices.

(653)Finally, internal documents from Universal confirm […]*.

(654)[…]* , […]*.

(655)[…]* […]*.

(656)[…]*

(657)[…]*.

(658)[…]* .

(659)[…]*.

(660)[…]* .

(661)An internal e-mail report of negotiations with [digital platform]* also contains the following observations:

[…]* .

(662)All of this evidence confirms that Universal currently has the ability and incentive to shape the business models of online platforms to suit its own commercial interests, to the detriment of innovation.

(663)The Commission notes that the impact of any increased ability and incentive of the merged entity to shape the business models of online platforms in this manner would be felt beyond Universal and EMI repertoires only. Online platforms and competitors overwhelmingly confirm that due to the technical features of online services, any usage restriction or other type of restriction that a large recorded music company imposes is extended to the repertoire of its competitors as well.

(664)[…]* makes the following observation:

"If one company imposes a limit on the functionality of a service could provide consumers (e.g. as regards the number of times a track could be played per month), than if that limit was imposed as a service-wide restriction the effects on all labels would be equal for all companies. [….] By way of example, the continuing imposition of limits on the Spotify free to the user tier in the UK and France (and the previous imposition of the same restrictions in the rest of the EEA) clearly seems to make sense to at least one company's business (at least in its opinion) whereas these limitations do not make sense to […]*'s business (because we believe these restrictions are damaging to Spotify's business. [….] This worsening of licensing business terms [….] does have a disproportionate effect, because it tends to direct the development of the market towards business models which that company supports, due of course to the fact that these models deliver it a competitive advantage (for example by replicating physical world limitations on shelf space)" .

(665)The negative effects this would have on consumer choice and innovation would therefore be propelled onto the entirety of the music available on the online platforms that are concerned.

(666)In light of the above, the Commission concludes that the proposed concentration would harm consumers by increasing Universal's bargaining power and ability to impose onerous licensing terms on digital platforms, in particular small and emerging innovative music platforms. This would likely result in retail price increases for consumers, as well as a reduction in innovation and consumer choice. As a result, the proposed concentration would also have a negative impact on cultural diversity.

400[…]*.

401[…]*.

402Competitors' response to the Commission's request for information of 17 April 2012, question 34 (ID: 7763).

167

(667)These anti-competitive effects would be felt across the EEA and in all the countries identified above. This also reflects the evidence collected through the market investigation further indicates that high licensing costs, in particular high advance payments and revenues guarantees required by record companies, hamper the territorial expansion of digital services within the EEA.

(668)The geographic scope of the activities of the largest digital customersof the merging parties shows that, apart from […]*, none of the digital platforms, which were initially launched in the EEA, was able to expand enough to cover the whole EEA territory. The market investigation indicates that the cost of licensing, in particular for the content of large record companies, is the main barrier for this geographical expansion. High licensing costs therefore restrict EEA consumers' choice as the digital services remain limited to one or several countries and hinder innovation as it has a negative effect on the development of new digital services launched in the EEA.

(669)Hence, the Commission concludes that the proposed concentration is likely to significantly impede effective competition on the EEA-wide market for the wholesale of digital recorded music, as well as in Austria, Belgium, Bulgaria, the Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece, Iceland, Ireland, Italy, Latvia, Lithuania, Luxembourg, Malta, The Netherlands, Norway, Poland, Portugal, Romania, Slovakia, Spain, Sweden and the United Kingdom.

7.1.5.12.Commercial conditions for physical customers and foreclosure of competitors

(670)The market investigation provided some indications that, due to its increased size, the merged entity would also likely result in the imposition of commercial conditions that are more disadvantageous to physical customers.

(671)Moreover, a number of competitors, including […]*, IMPALA and […]*, indicated that the proposed concentration may also result in the merged entity being able, due to its increased size, to guarantee to its artists better access to promotional opportunities (from certain physical retailers, as well as digital retailers) and to radio airplay than those offered by any other recorded music company, by securing for itself a share of such opportunities that will be significantly larger than that of its competitors. As a result, the merged entity would have the ability to sign the best artists, including at better conditions than those offered by other record companies, by offering those artists greater access to promotion. Competitors would find it increasingly difficult to compete with the merged entity with regard to the signing of new artists. This, in turn, would weaken these competitors' position in relation to customers as their repertoire would become less attractive and, conversely, strengthen the merged entity's position, thereby giving rise to a self-perpetuating vicious circle, which may lead to the foreclosure of at least some of the merged entity's competitors.

(672)The market investigation also indicated a number of different strategies that, the merged entity could implement in order to achieve the result of securing

403[…]*.

404[…]*.

168

exposure for its artists at certain key physical retailers (mass merchants) and at digital retailers, as well as at radio stations. The market investigation further provided indications supporting the claim that artist value exposure when deciding what recorded music company to sign up with.

(673)The Commission, however, considers that, for the purposes of this Decision, it is not necessary to examine whether these possible effects, whether in terms of commercial conditions that are more disadvantageous to physical customers or in terms of foreclosure, would translate into a significant impediment of effective competition in the relevant markets. This reflects the fact that the commitments offered by the Notifying Party are such as to remove the cause which would have led to the unfolding of these effects, namely a significant increase of Universal's size post-merger. Moreover, as explained, the commitments are also such to ensure that, in the future, the merged entity would continue to face sufficient competition in the relevant markets, which would render the possible anti-competitive effects discussed in this Section even more unlikely.

7.1.6.Piracy

7.1.6.1.The Notifying Party's view

(674)In its Comments to the Article 6(1)(c) Decision, the Notifying Party reiterates its claim that legal and illegal sources of music are substitutes from the point of view of the record companies' direct customers, that is to say digital music retailers. In this regard, it considers that the Commission did not take into account the competitive interactions between the wholesale and the retail level of the music supply chain and the fact that consumer demand at retail level would shape wholesale negotiations between record companies and music services. The Notifying Party refers to a number of internal documents from Universal and EMI and to several replies from competitors to the Commission market investigation that in its view would show the direct and material constraining effect of piracy on wholesale commercial negotiations. The Notifying Party in particular claims that if relatively large digital platforms such as iTunes and Spotify perceive that piracy constrains record companies negotiating position, regardless of the views expressed by other digital services, piracy constrains negotiations with these other digital services as well. According to the Notifying Party, the commitment by record companies to offer additional discounts for telecom companies which agree to specific anti-piracy requirements illustrates the constraining power of piracy on wholesale prices.

(675)Moreover, the Notifying Party argues that many of the vertically integrated customers of record companies are ultimately agnostic as to whether their retail customers access music legally or illegally and that it would increase their bargaining power. According to the Notifying Party, […]* would be the most illustrative example given that the sale of its devices or the use of its new […]* service would not rely on the availability of music from […]*. The same point

407Comments to the Article 6(1)(c) decision, pp. 41-42.

(676)In addition, given that the recorded music industry is characterised by fixed-costs, piracy would have created an incentive for record companies to offer music on an unbundled and low cost basis to support the emergence of new retail services. In this regard, the Notifying Party claims that piracy creates compelling commercial incentives for Universal to license all its customers at terms which allow them to compete in a consumer market place which is dominated by piracy. The Notifying Party refers to an internal document which would support this view. In addition, given the high concentration of its customer base, Universal would […]* .

408Comments to the Article 6(1)(c) decision, p. 21.

(677)In its Response to the Statement of Objections, the Notifying Party claims that the Commission did not take sufficient account of piracy as a competitive constraint on all market players. It refers to the Commission Article 6(1)(c) and Article 8(2) decisions in the Sony/BMG case , where the Commission acknowledged that piracy exerts a competitive constraint in certain national markets and creates an incentive for record companies to license their repertoire to some new digital services at attractive conditions. In particular, the Notifying Party refers to the results of the market investigation in the Sony/BMG case, which show that 75% of customers replied that illegal on-line download and piracy are a competitive pressure to majors. In addition, the Notifying Party submits that anti-piracy legislative measures will not have a material and sustained effect on piracy levels in the foreseeable future. In this regard, it refers to a very recent study which shows that peer-to-peer ("P2P") traffic is increasing again .

409Commission Decision of 19 July 2004 in Case No COMP/M. 3333 – Sony/BMG.

410http://www.paloaltonetworks.com/news/press/2012/New-Report-Shows-P2P-Filesharing-Streaming-Media-Use-Exploding-Worldwide.html

(678)The Notifying Party finally claims that the concept of "derived demand" is relevant in this case given that music is not transformed as it passes down the supply chain from record companies via retailers to consumers. Accordingly, following the Commission's Guidelines on Vertical Restraints, the behaviour of final consumers should also be taken into account when assessing the competitive constraints on suppliers upstream.

(679)As a first observation, the Commission notes that any quantification of the impact of piracy on competition in the digital recorded music markets is in itself fraught with difficulties and uncertainties.

407Comments to the Article 6(1)(c) decision, pp. 41-42.

408Comments to the Article 6(1)(c) decision, p. 21.

409Commission Decision of 19 July 2004 in Case No COMP/M. 3333 – Sony/BMG.

410http://www.paloaltonetworks.com/news/press/2012/New-Report-Shows-P2P-Filesharing-Streaming-Media-Use-Exploding-Worldwide.html

411Non-Horizontal Guidelines, paragraph 39.

170

(680)Extensive economic academic studies have tried to identify the impact of piracy on the recorded music sector without reaching consistent outcomes.

412Form CO, paragraphs 142 to 167.

413Comments to the Article 6(1)(c) decision, Section 4.

414Danaher et al (2012) reported in the Form CO (§151).

415Oberholzer-Gee and Strumpf (2007) find that the effect of piracy is "statistically indistinguishable from zero". This study is presented by Danaher et al (2012) as an exception in an empirical literature where "nearly all of the academic studies find that media piracy has a significant negative impact on sales". Liebowitz's (2011) view that this paper must be considered as "unreliable" (as stressed by the Notifying Party in footnote 30 of its reply to the Article 6(1)(c) decision) is not shared for example by Hammond (2012). This latter article finds instead "that file sharing does not have a large, negative effect on sales. In no specification, using no instrument, do I find any evidence to the contrary. A slightly positive effect of file sharing on sales is consistent with Oberholzer-Gee and Strumpf (2007) and Blackburn (2006)". It then contrasts these findings with Liebowitz's conclusion (as quoted in Hammond (2012) itself) that "the majority of studies find results supportive of the thesis that file sharing is causing harm".

416See Frontier Economics (2011), page 25.

(682)Any quantification of the total value of unlicensed digital files available online is already in itself subject to a significant margin of error. Moreover, and importantly, it does not equate to "an estimate of the business losses associated with digital piracy, and should not be interpreted as doing so" . In other words, only a portion of piracy that is perpetrated is likely to correspond to foregone sales, and it is far from clear that most of it would have translated into sales of those same songs and albums.

(683)Piracy may also have a different impact on different types of music and artists. On this issue, economic literature provides conflicting answers. According to one study, for less popular artists, higher availability of their music on illegal download networks is associated with increased legal sales. In contrast, the same study finds that, for more established artists, higher availability of their music on illegal P2P networks leads to lower legal music sales. Another studyreaches the exact opposite result that file sharing benefits more established and popular artists at the expense of newer and smaller artists.

419See Belleflamme and Peitz (2010).

(684)A recent surveycompares a basic scenario characterising most of the early seminal papers on digital piracy and file sharing with some variants addressing specific issues. As the following excerpt from this survey shows, the conclusion that piracy has an unambiguous negative impact on record companies is not a straightforward one: "[i]t was suggested that piracy also had the potential to increase the right-holders' profits, thereby counteracting its negative competition effect; a number of ways were described by which pirated copies could increase the value of original products either by solving an experience good problem (sampling), by increasing network benefits, or simply because the ability to make copies raises the consumers' willingness to pay for originals (indirect appropriation). It was then shown that there exist circumstances where right-holders favor piracy and where fighting piracy could prove welfare-detrimental".

(685)Accordingly, the only conclusion that can possibly be drawn from the existing economic literature is that piracy may be one of the causes of the decline in recorded music sales. Importantly, empirical references to piracy do not and cannot answer the question which the Commission is called to answer in this case, namely what the proposed concentration would change in Universal's bargaining position and its use thereof in relation to digital platforms that offer legal music.

(686)As stated in Recital 678, the Notifying Party claims that the concept of "derived demand" is relevant in this case given that music is not transformed as it passes down the supply chain from record companies via retailers to consumers. Accordingly, following the Commission's Guidelines on Vertical Restraints, the behaviour of final consumers should also be taken into account when assessing the competitive constraints on suppliers upstream. The Commission considers this reference to its Guidelines on Vertical Restraints to be misguided. In the context of the competitive assessment in this horizontal merger case, the fact remains that for digital retailers the choice of the source of recorded music is not a matter of end consumer preferences. Digital retailers do not have the choice to substitute legal music with illegal music when purchasing music for inclusion on their digital platforms. The reference to the impact of "derived demand" is therefore irrelevant. The relevant question is whether the existence of piracy has any impact on the wholesale negotiations between record companies and digital customers.

(687)Similarly, the Commission considers that the Notifying Party's reference to the Sony/BMG decision cannot assist it in its claim that the proposed concentration would not likely result in a significant impediment to effective competition due to the constraints posed by piracy.

(688)The Commission's analysis in this case confirmed that Universal currently uses its bargaining position to extract commercial terms and conditions that are more disadvantageous to digital platforms. This in itself does not support the Notifying Party's claim that piracy constrains Universal's bargaining position to a significant extent, but rather supports the conclusion that, despite the existence of piracy, Universal is able to exercise market power. In other words, even if piracy may have the effect of reducing the total size of the recorded music market, it equally affects all market players (as acknowledged by the Notifying Party itself), and does not alter their respective competitive positions (or, if anything, affects smaller players, whose scale is closer to some minimal level allowing to be viable, more than the larger ones). Against that background, the Commission's analysis confirmed, that irrespective of any constraint posed by piracy, the proposed concentration would likely result in of the imposition of commercial terms and conditions that are more disadvantageous to digital customers.

172

(689)The Commission also obtained specific evidence that piracy would not likely have any significant impact on wholesale negotiations between record companies and digital customers so as to alter this conclusion.

(690)In this context, the Commission considers that the Notifying Party has misrepresented the replies from customers to the Commission's question whether piracy has any impact on their negotiations with record companies. The Notifying Party quotes statements from […]* regarding the impact of piracy on its retail business. It omits to mention that […]* reply to the question on the impact of piracy on negotiations was: "not to a great enough extent."The reply of […]* to the same question, also quoted by the Notifying Party, was the following: "piracy has not been part of any discussion or negotiation between ourselves and the music companies"; […]* reply to the same question was: "no" . The statement by an anonymous customer that piracy is "very seriously"taken into account in negotiations also referred to by the Notifying Party actually goes on to explain the following: "[r]ecord Labels have always tried to impose certain obligations (i.e. graduated response) as a condition to license their contracts. In several occasions ( i.e. graduate response) as we have serious concerns about the legality of such requested measures it ended in a label/labels negative to license their catalogues and we were not able to build the desired music service, so that the initiative was cancelled". Hence, the latter respondent, when stating that piracy has an impact on its negotiations with record companies, refers to the record companies trying to impose anti-piracy requirement rather than to piracy having an impact on price and other commercial conditions. By taking customer statements out of their context in this manner, the Notifying Party is trying to create the impression that customers share the view that piracy has an impact on negotiations with record companies. However, this is not the case.

(691)As far as telecom companies are concerned, the most direct impact of piracy on their negotiations seems to be the need to support certain anti-piracy measures, such as a graduated response. Contrary to the claim of the Notifying Party in its Response to the Statement of Objections, none of the telecom companies considers that piracy affects price negotiations with the record companies.

(692)The Commission also disagrees with the Notifying Party's claim that if […]* and […]*, given their leadership on digital retail of recorded music, perceive that piracy constrains the record companies' negotiating position with them. A fortiori piracy should have the same impact on the negotiating position of record companies with other retail services. However, […]* has clearly stated that its

420Customers' response to the Commission's request for information of 17 April 2012, question 24 (ID: 7526)

421Customers' response to the Commission's request for information of 17 April 2012, question 24 (ID: 7419)

422Customers' response to the Commission's request for information of 17 April 2012, question 24 (ID: 7630)

423Customers' response to the Commission's request for information of 17 April 2012, question 24 (ID: 8122)

424Customers' response to the Commission's request for information of 17 April 2012, question 24 (ID: 8122)

173

negotiations with record companies are not affected by piracy to a meaningful extent. As regards […]*, the Notifying Party's claim ignores the fact that that these digital customers have a stronger bargaining position and are hence in a better position to use any argument based on piracy when negotiating with record companies. This of course does not provide any indication in relation to the ability of other digital customers, whose bargaining power is weaker, to exploit piracy in their negotiations with record companies.

(693)More generally, the vast majority of customers consider that piracy does not affect their negotiations with record companies. Out of 68 responding customers, only seven consider that piracy affects their price negotiations with record companies. Amongst those customers, which responded that piracy has an impact on their dealings with record companies, a number of them refer to the impact of piracy on sales volumes, not on prices. While one customer indicates that "there is a general understanding in the business that there are certain price points that it would not be possible to go above in order to be able to compete with piracy", another customer explains that its "profit margin is squeezed because it pays licence fees and it is increasingly difficult to compete with those businesses which do not so can offer a cheaper and arguably more consumer-driven service" . Another customer indicates that record companies have increased their wholesale rates in order to compensate for revenues lost through piracy .

425Customers' reply to the Commission Request for information of 17 April 2012, question 24 (ID: 6210).

426Customers' reply to the Commission Request for information of 17 April 2012, question 33 (ID: 7404).

427Customers' reply to the Commission Request for information of 17 April 2012, question 24 (ID: 7595).

174

(696)The Commission reiterates that the fact that Universal has in the past licensed its music to a high number of platforms across the EEA cannot be used to rebut the existence of the ability and incentive of the merged entity to extract more onerous terms from digital platforms after the proposed concentration. What matters for these purposes is not the fact that Universal has licensed one or more services, but the conditions at which these services were licensed. As amply demonstrated in Section 7.1.5, it clearly emerges from the analysis of these conditions and the level of revenues that record companies are able to extract from their customers, that larger record companies have the ability and the incentive to extract more revenues from their customers than smaller record companies and that Universal's ability and incentive to do so would be significantly increased by the proposed concentration. Moreover, what clearly emerged from the Commission's investigation is that Universal's terms and conditions for smaller, innovative platforms that have relatively weaker bargaining positions are […]*. This finding runs counter Universal's claim that it necessarily seeks to ensure that these smaller platforms thrive as much as possible in relation to more established players such as Apple/iTunes. If this were to be the case, the Commission would expect Universal's terms and conditions for the smaller, innovative platforms to be more attractive than, or, at least, comparable to, the ones offered to Apple/iTunes.

(697)Similarly, the Notifying Party has also misrepresented the submissions of certain competitors during the Commission's market investigation. In fact, out of the fifteen responding competitors, only three consider that piracy increases the bargaining power of customers. These three customers are indie record labels, whose bargaining position is far weaker than that of Universal, EMI and the other majors. It emerges from the competitors' replies to the market investigation that the impact of piracy on negotiations would be limited. These competitors also consider that the primary impact relates to anti-piracy requirements that the record companies would try to impose upon the ISPs and telecom companies .

428Competitors' reply to the Commission Request for information of 17 April 2012, question 36.

(698)Most competitors responding to the market investigation confirm the Commission's position. In particular, they disagree with the claim made by the Notifying Party that the wide availability of illegal music would confer bargaining power upon their customers, because customers are agnostic as to whether end customers access legal or illegal music provided they purchase their devices, use their search engine or their bandwidth to access the internet. Record companies remain indispensable for a legal music offering. According to one competitor: "[w]hen you decide to have a service, whether or not you make money elsewhere, you still need to deal with the labels and that gives them power. […] Some services hide behind the legislative protection given to them in the information society directive (mere conduit, etc), but that doesn't give them bargaining power. They still have to deal with labels if they want a legitimate service and because you can't get an artist cheaper elsewhere, and the big labels have the key artists or hits, the majors are essential" .

429Competitors' reply to the Commission Request for information of 17 April 2012, question 36 (ID: 7472).

175

(699)Most competitors outline that distributing legal music is a core, or at least a desired, aspect of their customers' commercial strategy. According to one respondent: "Apple's iTunes store remains the most successful source of legal music downloads and continues to help drive sales of Apple's hardware devices. ISPs themselves have an interest in promoting legal over illegal music to reduce the clogging of bandwidth by large-scale users of illegal P2P services. A number of major businesses, from Apple, Amazon and Google to the ISPs, are developing cloud storage services that in turn enhance the attractiveness of purchased music"

430.

(700)These competitors indicate that the nature of device manufacturers' business model, built on an entire eco-system, would make them dependent on content owners. A respondent considers that "the recorded music industry has a very strong bargaining position with Apple, due to the enormous importance of music to Apple's business, where content and devices are integrated via a seamless 'end to end' eco-system, which is thus reliant on content rights." According to […]*, its success would have in turn influenced the rest of the market: "other hardware and software manufacturers now seek to replicate the integrated hardware, software, content eco-system pioneered […]*". […]* would have a similar position in relation to the record companies […]* .

(701)Likewise, customers almost unanimously reject the Notifying Party's claim that they are agnostic with regard to the source of music and that piracy would increase their bargaining power towards record companies In particular, […]* . According to […]* telecom customer: "the fact that we close deals with the labels proves that we need their content. If illegal music would be enough for our purposes, then we would have no need to negotiate" .

432.

(702)Finally, respondents to the Commission's market investigation have pointed out that piracy is becoming less of an issue. A number of respondents to the market investigation underlined that, while the fight against piracy is far from being won, the development of innovative new music services, which address the needs of consumers on a commercially compelling basis, together with the enactment of good legislation, which is actually enforced, seems to be the best way to fight piracy. According to one respondent, the wide availability of services aimed at the distribution of content is key: "[m]ultiple studies have shown that copyright infringement is, in part, an access problem. […] it is necessary to ensure a well-functioning marketplace for digital goods to drive users to legitimate sources. Simple, efficient licensing procedures; legal certainty in limitations and exceptions to copyright; and the creation of a pro-investment environment are all keys to enabling the growth of new businesses to meet consumer demand and to

436achieve the ideal of a single digital market". Another respondent indicates that "[r]ecent trends have shown that when the market 'meets' the consumer, i.e. develops new offerings that suit an individual's listening patterns and preferred ownership they will pay for music" . According to another responding competitor: "the emergence of Spotify and other streaming services with a low 'eat as much as you can' offer is clearly attractive to all and those who are price-sensitive on download services have got a very reasonable alternative. So the price issue is less of a problem, at least in the Nordic territories. It has also been highlighted that there are substantial security risks with illegal file-sharing, so all in all we are moving towards less competition from illegal services" . […]* .

(703)These replies are corroborated by an internal Universal document which states the following: […]* .

(704)Some respondents, as well as a recent IFPI report, also indicate that the cooperation between record companies, ISPs and Internet search engine operators may also constitute an effective tool to fight piracy . According to one respondent: "the hard-core users of pirate sources is now being controlled and may be in decline. There is growing evidence that litigation, government initiatives to stem piracy – such as the HADOPI law in France – coupled with the availability of new 'freemium' and low-cost subscription models are beginning to have an impact in reducing piracy" .

(705)Accordingly, the Commission concludes that, while the existence of piracy may reduce the size of the overall market, it would not deprive the merged entity of its ability and incentive to exercise its bargaining power in relation to digital retailers after the proposed concentration.

7.1.7.Countervailing factors

7.1.7.1.Buyer power

The Notifying Party's view

(706)The Notifying Party claims that in the negotiations with its key digital retailers it is facing a very significant buyer power. According to the Notifying Party, the balance of power between record companies and music retailers has increasingly

(707)shifted in favour of the latter due to the very concentrated distribution base for digital music, which is dominated by Apple/iTunes, the decreasing importance of music for digital retailers and the exponential growth of piracy to the detriment of legal sales. In the Notifying Party's view, as a consequence of these factors, post-merger, the merged entity's behaviour would be disciplined by these retailers, which would be able to counteract any attempt by the merger entity to exert market power .

(708)(707) In its submissions, the Notifying Party explains that a limited number of key customers account for the vast majority of its recorded music revenues in the EEA. The Notifying Party notes that each of the three digital distribution channels, that is to say e-tailing, à-la-carte download services and streaming services, is dominated today by a single company, that is to say Amazon, Apple/iTunes and Spotify respectively. According to the Notifying Party, by leveraging their market power, these powerful customers are able to demand better terms from Universal and EMI .

(709)(708) The Notifying Party also submits that customer buyer power is now strengthened due to the decreasing importance of recorded music for its digital customers. According to the Notifying Party the decreasing importance of recorded music in the digital market is due to the rapid diversification and fragmentation of digital entertainment formats. In the Notifying Party's view, this trend is driven by the increasing convergence of hardware devices, which allow consumers to access a wide range of entertainment content, namely movies, TV series, eBooks, social networking links, video games and a huge variety of applications, which all compete for the consumers' attention and time .

(710)(709) Moreover, the Notifying Party submits that the exponential growth of piracy to the detriment of legal sales has contributed to the shifting of power from record companies to music retailers. According to the Notifying Party, several key customers enjoy an even greater bargaining position by being able to also monetise the consumption of illegal content .

(711)(710) In its Response to the Statement of Objections, the Notifying Party also submits that the frame of reference adopted by the Commission in the Statement of Objections to assess the existence of buyer power is misguided. In other words, the Commission should not have adopted a "binary" approach, according to which buyer power only exists if the specific examples mentioned in the Horizontal Merger Guidelines are met, but should instead have analysed the extent to which customers would be able to counter any price increase by the merged entity.

(712)(711) As regards digital customers' buyer power, the Notifying Party notes that the Commission failed to acknowledge the fact that these customers, in particular the

(713)larger ones, such as Apple and Spotify, exercise very considerable bargaining power as indispensable trading partners of Universal. According to the Notifying Party, some of these digital customers, such as Apple, could also easily switch away from music, which only accounts for a small percentage of their total revenues. Moreover, the Notifying Party submits that the case file includes a lot evidence of the strong bargaining position of Universal's customers.

(714)(712) Finally, the Notifying Party submits that there is no economic or commercial justification for Universal to offer smaller digital customers worse terms than those offered to Apple/iTunes or other large digital platforms, as […]*.

The Commission's assessment

(715)(713) The Horizontal Merger Guidelines explain that: "[c]ountervailing buyer power […] should be understood as the bargaining strength that the buyer has vis-à-vis the seller in commercial negotiations due to its size, its commercial significance to the seller and its ability to switch to alternative suppliers" .

(716)(714) The Horizontal Merger Guidelines also explain that customers will be in a position to counter the increase in market power that a merger would otherwise be likely to create, that is to say to effectively exercise countervailing buyer power, if they could credibly threaten to resort, within a reasonable timeframe, to alternative sources of supply should the supplier decide to increase prices or to otherwise worsen the quality or the conditions of delivery, by immediately switching to other suppliers, credibly threatening to vertically integrate into the upstream market, or sponsoring upstream expansion or entry .

(717)(715) The Horizontal Merger Guidelines further explain that: "[i]t is more likely that large and sophisticated customers will possess this kind of countervailing buyer power than smaller firms in a fragmented industry"and that "[c]ountervailing buyer power cannot be found to sufficiently off-set potential adverse effects of a merger if it only ensures that a particular segment of customers, with particular bargaining strength, is shielded from significantly higher prices or deteriorated conditions after the merger. Furthermore, it is not sufficient that buyer power exists prior to the merger, it must also exist and remain effective following the merger" .

(718)(716) Against this background, the Notifying Party appears to confuse the notion of countervailing buyer power within the meaning of the Horizontal Merger Guidelines with the typical bargaining dynamics, which take place in any commercial negotiation. In other words, the fact that certain customers may be particularly aggressive in their commercial negotiations does not itself constitute evidence that they enjoy countervailing buyer power within the meaning of the Horizontal Merger Guidelines. As explained, in their negotiations with the merged entity, these customers will lack the outside option, consisting in essence

(719)of the credible threat to switch to an alternative supplier, which constitutes a pre-requisite for any finding of countervailing buyer power.

(720)(717) When applying the principles set out in the Horizontal Merger Guidelines correctly, the Commission considers that the Notifying Party's claim that the merged entity's physical and digital customers would enjoy countervailing buyer power is unfounded for a number of reasons.

(721)(718) First, the Notifying Party did not explain, by reference to the Horizontal Merger Guidelines, how the merged entity's digital customers could concretely exercise their buyer power. In other words, the Notifying Party did not explain whether, in order to counter any price increase from the merged entity or any other anti-competitive effects arising from the proposed concentration, customers would switch to other suppliers, threaten to vertically integrate into the upstream market and sponsor upstream expansion or entry.

(722)(719) Moreover, based on the evidence in the file, the Commission concludes that any such countervailing buyer strategy is unlikely to succeed in this case. Indeed, as already discussed in Section 7.1.4, due to the size and strength of the merged entity, post-merger, there would be limited possibilities for the merged entity's customers to credibly threaten to switch to competing suppliers. Moreover, based on the indication in the file, there is no indication that the merged entity's digital customers could credibly threaten to vertically integrate upstream or to sponsor entry. Furthermore, based on the evidence in the file, the Commission considers that the digital customers of the merged entity would not enjoy a sufficient degree of bargaining power, which would allow them to counter any price increase by the merged entity following the proposed concentration. This reflects the fact that, as explained in detail in Section (241), already today, Universal, due to its size, is able to extract better terms than […]* from, at least, most of its digital customers. This fact shows that these customers do not enjoy countervailing buyer power and are unlikely to enjoy such power after the proposed concentration.

(723)(720) Second, the Commission also considers that the fact that, as also explained in detail in Section 7.1.5, smaller customers, including, in particular, the most innovative new streaming services, appear to be more exposed to the larger record companies' market power (which manage to extract higher rents from these customers than from larger and more established customers, such as Apple and Spotify). As it clearly emerges from Section 7.1.5, the evidence in the file shows that, contrary to the Notifying Party's claims, Universal does impose […]* terms and conditions to smaller digital platforms than those which it manages to extract from larger customers […]*. While it may be the case that Universal has the incentive to support smaller digital customers […]*, it also appears from the evidence in the file that Universal is not ready to do so at all costs, but only subject to conditions which, at the same time, allow at least some of the services to survive, […]*. The Commission notes that even in relation to overall large and sophisticated customers […]*. The proposed concentration would likely increase Universal's ability to do so even further.

(724)(721) Third, as regards the importance of recorded music, the market investigation has shown that music is still important for both Apple and Spotify. In the digital

(725)market, recorded music is a key input and it is at the core of digital retailers' offer that crucially depends on it. All online platforms confirm that recorded music remains a critical input for their overall service offerings and hence disagree with the Notifying Party's view that recorded music would have become an ancillary part of their business to such an extent that they would be able to exert significant bargaining power in relation to record companies.

(726)(722) Finally, the vast majority of the respondents to the market investigation did not confirm the Notifying Party's argument according to which the availability of illegal music would confer increased bargaining power upon some digital customers . The Commission therefore concludes that, consistent with the conclusions reached in Section (241), while the phenomenon of piracy may have an impact on the negotiations between record companies and digital platforms, this impact is not such as to deprive record company of their ability to exercise market power in relation to at least some of these customers.

(727)(723) Hence, the Commission therefore concludes that customers will not be in a position to exert their buyer power as a countervailing factor to an increase of the merged entity's market power.

(728)(724) In this regard, the Commission also considers that, contrary to the Notifying Party's arguments, it did not adopt a binary approach to assess whether the merged entity's customers would not have sufficient buyer power to counter any price increase by the merged entity. On the contrary, the Commission: first assessed the Notifying Party's arguments on buyer power based on the analytical framework provided for by the Horizontal Merger Guidelines and concluded that the Notifying Party was unable to explain what concrete strategy among those outlined therein the merged entity's customers would have adopted to counter any price increase by the merged entity. The Commission then analysed the degree of bargaining power exercised in practice by the merged entity's customers and concluded that the merged entity's customers would not have sufficient buyer power to counter any price increase by the merged entity.

7.1.7.2.Entry

The Notifying Party's view

(729)(725) The Notifying Party claims that there is ample evidence of new entry to the market, since artists continue to emerge and record companies are founded, either from new A&R and marketing teams, or by experienced teams from larger record companies. The Notifying Party provides a list of fifteen record companies which entered the market in the last five years in the following three Member States: France (My Major Company, launched in 2007), Germany (Tonpool, a distribution company, in 2008, BMG Rights Management in 2009) and the United Kingdom (eleven independent labels, including Dreambrother set up by Placebo in 2009, Jazz/Soul label Tri-Sound in 2009, BPM Entertainment created

(730)by a former BMG executive in 2010, or Two Black Cats created by the advertising agency Rainey Kelly).

(731)(726) In its Response to the Statement of Objections, the Notifying Party also submits that the Commission did not give due weight to the opportunities for artists to bypass established record companies by self-releasing and forming their own record company. The Notifying Party also puts forward the largely unsubstantiated argument that, in today's recorded music market, barriers to entry are much lower for record companies than they were only five years ago due, among other things, to the role played by aggregators and label services, the unlimited shelf space at digital retailers and the low cost of promotion through social networks. Furthermore, the Notifying Party considers that the Commission did not give due weight to recent the market entry, including, in particular, that of Google and Facebook.

The Commission's assessment

(732)(727) According to the Horizontal Merger Guidelines, entry analysis constitutes an important element of the overall competitive assessment. However, "[f]*or entry to be considered a sufficient competitive constraint on the merging parties, it must be shown to be likely, timely and sufficient to deter or defeat any potential anti-competitive effects of the merger" .

(733)(728) As regards likelihood of entry, the Commission examines whether entry is likely or whether potential entry is likely to constrain the behaviour of incumbents post-merger. For entry to be likely, it must be sufficiently profitable taking into account the price effects of injecting additional output into the market and the potential responses of the incumbents. As regards timelines of entry, the Commission examines whether entry would be sufficiently swift and sustained to deter or defeat the exercise of market power. As regards sufficiency of entry, entry must be of sufficient scope and magnitude to deter or defeat the anti-competitive effects of the merger. Small-scale entry, for instance into some market "niche", may not be considered sufficient .

(734)(729) Against this framework, the Commission concludes that entry in the relevant markets is unlikely to counter the likely anti-competitive effects arising from the proposed concentration.

(735)(730) During its investigation, the Commission did not find any significant evidence of the possibility that, post-merger, despite the allegedly low barriers to entry in the recorded music sector, if the merged entity were to increase its prices, new players would enter the relevant markets, both upstream in A&R and downstream in the wholesale of recorded music to physical or digital customers, in such a timely and sufficient manner as to counteract the anti-competitive effects likely arising from the proposed concentration.

(736)(731) The examples of market entry referred to by the Notifying Party in its submissions cannot be considered as indicative of the likelihood that a timely and sufficient entry by competitors could counteract the anticompetitive effects of the proposed concentration. To name a few, the company My Major Company, which, as the Commission understands it, is now operating in France and in the United Kingdom is a "fan-funded" music label, which allows individuals to produce artists. This company is operating a new, innovative business model, which is limited to few artists and two countries at the moment, and would therefore be unable to exercise any meaningful constraint over the majors in the relevant markets. Similarly, the other new entrants referred to by the Notifying Party are small indie labels, which address niche market segments and would therefore be equally unable to exercise any meaningful constraint over the majors in the relevant markets. Some of the other record companies mentioned by the Notifying Party as examples of new entrants are directly or indirectly linked to the majors (for example Popjustice Hi Fi is a joint-venture between Popjustice and Virgin Records, owned by EMI; Mute Records is licensed from EMI the right to use the trademark "Mute" and part of the former Mute back catalogue, with EMI also holding a minority equity interest in Mute Records)and would therefore be equally unable to constrain them. By the same token, the fact that the advertising company Saatchi & Saatchi decided to start producing some recorded music and that certain artists decided to set up their own record label, whether solely or together with the other examples mentioned by the Notifying Party, do not appear to constitute evidence of entry of the necessary scale to constitute a meaningful counterweight to the anti-competitive effects arising from the proposed concentration.

(737)(732) The Commission also does not consider Facebook or Google as a new entrant in the recorded music market. As a matter of fact, neither company has entered the market as a recorded music company. Moreover, while Google is active via YouTube (and may in the future expand its activities via its new service Google Music) in the downstream market for the wholesale of digital music, Facebook is not even active at this downstream level of the recorded music sector, but simply entered into a partnership with certain digital music services, such as Spotify, pursuant to which Facebook users can essentially share among each other the tracks that each of them listens to on these services. The Commission therefore considers that the fact that Facebook and Google are directly or indirectly present in the recorded music sector does not support the conclusion that these companies are new entrants in the relevant markets where the proposed concentration would likely have anti-competitive effects within the meaning of the Horizontal Merger Guidelines.

(738)(733) The Commission considers therefore that it is unlikely that a timely and sufficient entry by competitors in the relevant recorded music markets would counter the anti-competitive effects arising from the proposed concentration.

(739)(734) Finally, as regards the Notifying Party's argument concerning the artists' ability to bypass established record companies by self-releasing and forming their own record company, the Commission limits itself to making reference to Section 6.1.2.2, where it explained at length why, while self-release may be an option, at least for some time, for an artist at the very beginning of his or her career or for a superstar artist, record companies remain nonetheless essential for artists to be successful.

7.2.COORDINATED EFFECTS

7.2.1.The view of the Notifying Party

(740)(735) The Notifying Party submits that the Commission's findings on coordinated effects in Sony/BMG are still valid. In that case, the Commission concluded that there was no risk of coordination among major record companies in the wholesale of recorded music given the dynamics of competition on the relevant markets, as well as in the light of the absence of any evidence of past coordinated behaviour on those markets.

(741)(736) The Notifying Party claims that these findings are currently all the more valid in the light of the intermittent reshaping of the recorded music industry that makes it even less plausible that record companies coordinate their commercial behaviour. According to the Notifying Party, the structural change in the market resulting from the merger would not create any additional opportunities or commercial incentives for record companies to reach, monitor and sustain terms of coordination.

(742)(737) The Notifying Party also underlines that that the Commission’s assessment of coordinated effects in Sony/BMG case concerned the actual effects of the notified concentration in that case rather than its prospective effects as that concentration had already been implemented after the Commission’s first clearance decision. In its Comments to the 6(1)(c) decision, the Notifying Party underlines the absence of factual evidence of any past coordination on the relevant markets and the competition dynamics that are not prone to giving rise to any such effects.

(743)(738) According to the Notifying Party, the rapid and permanent shift of music consumption from physical to digital distribution channels has continued unabated. This has in turn led to further diversification in prices and business models and increasingly complex and dynamic competitive interactions on the relevant markets.

(744)(739) The Notifying Party submits that the digital distribution channel has become even more varied and complex since 2007. Many new music download services have been launched and new successful business models, such as advertising funded and subscription based streaming services, like Spotify and Deezer, have emerged.

(745)(740) In the Notifying Party's view, these developments have not only introduced new and different but also very complex pricing models and strategies on the relevant markets, thus further increasing the lack of transparency and inherent difficulties of reaching terms of coordination between record companies. In this regard, the Notifying Party underlines that specific discounts are offered by record

companies to support “partnership deals”, for instance where streaming services are offered for free to a purchaser of a new mobile telephony contract for a period of time. In addition, the Notifying Party claims that pricing in relation to digital music downloads has become more complex with the introduction by Apple in 2009 of a three retail pricing points system which would have reduced any pricing homogeneity at retail level and would exclude the possibility of iTunes providing a focal point for price coordination at wholesale level.

(746)(741) The Notifying Party submits that not only there is a wider dispersion of titles wholesale prices but also there is a more complex pricing structure of titles along their life cycle. Indeed, while there are more categories of wholesale price tiers than at the time of the Sony/BMG decision, there is also more fluidity with tracks and albums being moved across different wholesale price tiers during their life cycle.

(747)(742) As regards digital streaming services, the Notifying Party submits that those services are still nascent and operate on the basis of a range of complex and varied consumer propositions and wholesale pricing formulae with different and changing permutations.

(748)(743) The Notifying Party submits that the analysis of commercial terms is strong evidence of the non-transparent and complex nature of contractual terms between record companies and digital music services. In its view, the diversity of terms across contracts indicates that there is no focal point for any coordination across the digital segment.

(749)(744) According to the Notifying Party, physical customers have become more powerful and sophisticated and display an extremely high level of price sensitivity. The Notifying Party argues that there has not been any simplification of the pricing structures and models used at both the wholesale and retail level of the distribution chain for physical formats with the increasing use of discounts and promotional campaigns. Finally, the rapid shift from physical to digital modes of consumption and supply have created stronger and more complex competitive interactions between sales of digital and physical recorded music.

(750)(745) The Notifying Party submits that there are various other factors than those identified in the Commission decision in the Sony/BMG case that would further destabilise and therefore render unsuccessful any attempt at coordination. According to the Notifying Party, the exponential growth of piracy would create strong commercial incentives for record companies to license their music content as widely as possible at attractive prices. In addition, the shift of music distribution from physical to digital formats would progressively eliminate any remaining barriers to entry and expansion for the indies and defeat any attempt of coordination among the majors. Finally, the ability of and increasing ease with which new and established artists can undertake the promotion, production, and distribution of their own music by bypassing record labels altogether would also render any attempt at coordination unstable and hence ineffective.

7.2.2.The Commission's assessment

(751)(746) According to well-established case lawand the Horizontal mergers Guidelines, for coordinated effects to be identified it needs to be established that a proposed concentration will make coordination more likely, more effective or more sustainable. The analysis needs to focus in particular on the following:

(a)the ability of record companies to reach terms of coordination;

(b)the ability to monitor deviations;

(c)the existence of a credible deterrent mechanism if deviation is detected and

(d)the reactions of outsiders such as potential competitors and customers.

(752)(747) In Sony/BMG, the Commission excluded the risk of coordination between the majors for the sale of physical music based on the following factual elements:

(a)At the level of budgets: in addition to the confidential nature of budgets and the lack of evidence that they are distributed amongst the majors, budgets are instruments to forecast sales and discounts, but they are not appropriate tools to control discounts granted to customers on a regular basis;

(b)At the level of title pricing: identifying the precise net wholesale price of an album is not reproducible at the level of each album on a systematic basis, considering the information that is publicly available. In addition, PPDs, campaign discounts and retail prices evolve constantly so that regular monitoring of these prices would require the collection of an extraordinary amount of information. Such a database would be highly costly, whilst remaining, in all likelihood, largely incomplete and becoming quickly out of date considering the constant evolution of prices and discounts. This would significantly limit interest in trying to produce such a database;

(c)At the level of pricing policy (stabilization of current business model): such coordination would leave sufficient margin for different majors to develop many alternative pricing schemes, which could not be distinguished from competition on the merits. For example, a new pricing scheme applied by one major to a limited number of albums would not be considered as a deviation from the "commonly agreed policy";

(d)On album prices at and shortly following their release date: the market investigation found that majors use different PPDs. Furthermore, a study of the stability of the discounts granted by the majors suggested that, even based on the conservative assumption of full transparency of PPDs, a significant number of sales transactions did not follow a simple and stable discount pattern from which a knowledgeable could make sufficiently meaningful inferences. Finally, there were also indications that it is not

460 Case T-342/99, Airtours plc v Commission, Case T-464/04, Impala v Commission. 461 Horizontal Merger Guidelines, paragraphs 39-57.

186

(e)possible to deduce net wholesale prices from retail prices with the necessary degree of certainty for coordination to be sustainable;

(f)(e) At the level of non-price items, such as access to retailers: the Commission's analysis suggested that majors do compete for access to retail space with discounts. In addition, hits released by independent record companies benefit from similar conditions to hits released by major companies, with shelf space being allocated independently by retailers, on the basis of their own evaluation of the potential sales of each album;

(g)(f) At the level of non-price items, such as access to airplay: all record companies compete to be registered on radio playlists as regular and sustained airplay is one of the factor supporting sales;

(h)(g) At the level of non-price items, such as influence on physical chart rules or access to charts: even if charts can increase market transparency by providing regular information on the sales performance of songs to all observers, including major record companies, the Commission did not find any evidence that majors use chart rules as an opportunity to deny, in a coordinated manner, access to the charts to artists of independent record companies;

(i)(h) At the level of non-price items, such as an album's release date: even if decisions with regard to release dates of albums may be influenced by the release dates of competitors’ albums, the Commission found that this could not be qualified as collusion, but rather as competition following the rule "first come, first served"; and

(i) At the level of non-price items, such as publishing and recording activities: even if mechanical copyrights are an input for recording companies to produce albums, collecting societies, and not publishing companies, determine the prices and license conditions provided to customers in the physical market, so that publishers cannot therefore exercise influence on the market for recorded music.

(748)Moreover, the Commission found that the threat of exclusion of a deviating competitor from compilation joint ventures or joint activities and the termination of the tacitly coordinated behaviour with respect to prices and releases of albums did not constitute sufficient deterrent mechanisms to ensure compliance with the common behaviour.

(749)Finally, the Commission found that a sizeable proportion of customers could destabilise coordination by majors by reducing the amount of products and advertising purchased from them.

(750)As regards the sale of digital music, the Commission excluded the risk of coordination between the majors based on the following factual elements:

(a)Record companies apply different prices, different price structures and use different business models, with the main distinctions concerning the platform (for instance online versus mobile), the types of service (for instance pay-per track, subscription), the product types (for instance singles or albums), frontline and catalogue tracks, premium albums versus standard and mid-price albums, the product packages (3-packs, 6-packs, digital EPs, mini-albums), usage rights and DRM restrictions and the duration and early termination rights for subscription services;

(b)The Commission's investigation confirmed that digital wholesale prices are negotiated with each service provider on a confidential and bilateral basis. Digital retailers do not know the licensing rates of their competitors and do not pass such information to the majors in the light of their confidentiality clauses;

(c)The majors cannot reconstruct wholesale prices and cannot identify any deviation regarding wholesale pricing on the basis of observed retail prices given the diversity and complexity of the existing wholesale pricing structures, including discounts and volume related prices, a two tier pricing system that charges the greater of a minimum price per download or a percentage of the retail price, advance payments, usage restriction variations, bundled content, the fact that pricing can depend on contract duration and the fact that majors increasingly use different prices per retailer in the different EEA countries;

(d)The iTunes pricing model cannot constitute the focal point for coordination for other online download contracts since the majors have an incentive to expand their activities in the digital markets and there are important wholesale price differences between the majors that fluctuate over time;

(e)MFN clauses are not applied systematically in the EEA;

(f)There is no credible retaliation mechanism to ensure the coordination among the majors sustainable, including a temporary returning to competitive behaviour, excluding the deviator from joint agreements in the digital market, or excluding of the deviator from joint ventures and agreements in the physical market; and

(g)The balance of power between the majors and iTunes, and the increased number of digital players such as telecom operators, could jeopardise to a certain extent the benefits that any coordination could bring to the majors.

(751)The majority of competitors who responded to the Commission's market investigation in this case confirmed that the Commission's conclusions on coordinated effects in Sony/BMG are still valid today .

462 Competitors' reply to the Commission request for information of 17 February 2012, question 61. Coordination unlikely Confirmed Did not confirm At the level of budgets 8 2 At the level of each title pricing 9 1 At the level of pricing policy 8 1 At the level of album prices 8 2

188

(752)That being said, several competitors indicated that, as a result of the proposed concentration and of the elimination of EMI as a competitor, the market structure would change in that two market leaders, Universal and Sony, would face little competition from one trailing major (Warner) and a fragmented group of independents .

463 Competitors' reply to the Commission Request for information of 17 February 2012, question 63.

(753)According to certain competitors, the market structure resulting from the proposed concentration would increase the scope for coordination between the majors in the digital wholesale market, where price structures would be less complex than in the physical wholesale market. In this regard, one responding competitor indicated that: "[a]lthough different price structures and different business models apply to different types of digital music services, the structures/models for each type of service are relatively simple and the more complicated structures of the physical environment do not apply in the digital world. Given that negotiations in respect of digital services are not conducted on a release by release basis, but across a record company's entire catalogue and future output and often on a multi-territorial basis, there is considerably simpler and far more fixed environment in which effective co-ordination may take place as compared to the physical world" . With regard more specifically to downloads, some competitors identified iTunes price as a possible focal point for coordination between majors .

464 Competitors' reply to the Commission Request for information of 17 April 2012, question 56. 465 […]* 's reply to the Commission Request for information of 17 February 2012, question 62.1.1 (ID: 4000). 466 Competitors' reply to the Commission Request for information of 17 February 2012, question 62.4.1.

189

(754)However, these views were not shared by a majority of customers . Most customers indicated that the structures of digital services are very complex. Amongst these respondents, customers indicate the following:

"Digital services are much more complex and sophisticated than the physical services. Not only because the business itself, but also because of the incredible amount of administrative tasks that have been now imposed by the labels to the e-tailers that do not exist in the brick and mortar retail industry, including the responsibility to license, report and pay the rights from the collecting societies, authors and publishers, which have never ever been responsibility of the physical retailers (the labels handle that responsibilities in the physical business), and not to mention the level of risk that most e-tailers must face due to the guarantees required by the labels, which do not exist in the physical retail" .

"In our view the structures of digital music services are complex. Key reasons [include]* complex commercial models resulting in complex financial reporting, anti-piracy requirements resulting in complex reporting and partly conflicting with privacy objectives, complex and expensive distribution platforms, which need high volumes to be economical, complex billing environment/price plans, etc" .

(755)Moreover, several respondents point out that the digital music business is still evolving and that the emergence of streaming services has increased the complexity of wholesale digital recorded music. The majority of responding competitors do not expect a convergence between wholesale price points for download and streaming services and/or between the business models of the providers of these two services. One of these competitors notes:

"[t]raditional models based on wholesale pricing or percentages of net revenue, but with the advent of streaming services models have become increasingly complex. In a new "standard" deal we are presented with revenue shares, per user minimums, per stream minimums, subscriber guarantees, bundle offerings, locker services, content aggregation and tiered pricing. These models greatly increase the complexity of proposed licensing agreement" .

(756)In addition, the market investigation was inconclusive regarding the question whether there is less scope to negotiate discounts and rebates based on unit volumes in the digital wholesale environment than in the physical wholesale environment .

467 Customers' reply to the Commission's Request for information of 17 April 2012, question 49. 468 Customers' reply to the Commission Request for information of 17 April 2012, question 49 (ID: 6441). 469 Customers' reply to the Commission Request for information of 17 April 2012, question 49 (ID: 7469). 470 Customers' reply to the Commission Request for information of 17 April 2012, question 49; Competitors' reply to the Commission Request for information of 17 April 2012, question 55. 471 Competitors' reply to the Commission Request for information of 17 April 2012, question 54. 472 Competitors' reply to the Commission Request for information of 17 April 2012, question 54 (ID: 6233). 473 Customers' reply to the Commission Request for information of 17 April 2012, question 50.

190

(757)Most customers consider that record companies are not able to anticipate with a degree of precision and certainty the negotiating behaviour of other record companies. On this account, the market investigation clearly indicated that the on-time reporting provided by customers to record companies would not enable the latter to monitor the evolution of their competitors' competition variables in the wholesale of digital music. This view was confirmed by a number of competitors .

474 Customers' reply to the Commission Request for information of 17 April 2012, question 53. 475 Competitors' reply to the Commission Request for information of 17 April 2012, question 62. 476 Customers' reply to the Commission Request for information of 17 April 2012, question 47. 477 Customers' reply to the Commission Request for information of 17 April 2012, question 48. 478 Customers' reply to the Commission Request for information of 17 April 2012, question 46.

(758)Furthermore, the Commission's market investigation confirmed that, whereas record companies may have similar negotiation targets with regard to certain points, they tend to focus on specific terms and have different approaches, some being more flexible than the others. EMI also did not emerge as a strongly disrupting competitor .

(759)The Commission investigated whether the inclusion of MFN clauses in licensing agreements with record companies resulted in practice in the equivalence of terms across all licensors. It emerged that, MFNs are not systematically included in licensing agreements with all customers. In addition, one customer indicated that "in practice the inclusion of MFN, at the time of signature of the deals, does not necessarily mean that all licenses exactly need to have the same structure, specifically in terms of business model (but if and only if MFN are to be applied on a "whole basis agreement")". Finally, it emerged that the scope of the MFNs that the different record companies conclude, […]*, differs widely.

(760)As set out in, the proposed concentration would increase the asymmetry of market shares between the merged entity and its competitors. The Commission considers that this is not indicative of the proposed concentration resulting in an alignment of the competitive strategies of the various competitors.

(761)In light of the above, the Commission concludes that the majors active on markets for on wholesale of recorded music are not likely to easily reach a common understanding on the terms of coordination, given their complexities and the relative lack of transparency on the market. The Commission also concludes that the proposed concentration would not significantly impede effective competition due to coordinated effects.

7.3.VERTICAL AND CONGLOMERATE EFFECTS

7.3.1.The Notifying Party's view

(762)The Notifying Party submits that Universal and EMI only have marginal activities in downstream markets which could be regarded as vertically related to recorded music. In its submissions, the Notifying Party explains that Universal’s parent company Vivendi is active in a number of markets which use recorded music as an input (for example mobile telephony, internet broadband services, pay-tv services, games/interactive entertainment, content distribution), while EMI's retail activities are extremely limited. In the Notifying Party's view, the proposed concentration would create neither an ability nor an incentive for Universal to engage in a strategy of input or customer foreclosure as any such attempt would significantly adversely affect the merged entity's retail services' commercial attractiveness to consumers, resulting in a loss of revenues, which would vastly outweigh any possible increase of revenues as a result of increased physical sales, downloads or streams of Universal and EMI recorded music.

(763)Moreover, as regards possible conglomerate competition concerns, the Notifying Party submits that Universal has neither the ability nor the incentive to seek to leverage its position in neighbouring markets to the recorded music market.

7.3.2.The Commission's assessment

(764)Universal is already active in the retail of recorded music, in the provision of ancillary services and in music publishing. EMI's activities in these sectors are rather limited. The proposed concentration would therefore not alter to a significant extent the market position of Universal and/or its parent company Vivendi on the downstream markets. Moreover, the acquisition of EMI's music publishing business by a consortium led by Sony and Mubadala was cleared conditionally by the Commission on 17 April 2012. This acquisition has been completed on the date of the decision in the present case. Therefore, the proposed concentration would not alter the market position of Universal on the music publishing markets to a significant extent.

(765)As a result, any vertical concerns that the proposed concentration may raise, particularly, as a result of input foreclosure in the retail of recorded music or any segment thereof would be associated with the merged entity's ability to do so due to its increased market power on the upstream market for the wholesale of physical and digital recorded music.

(766)Similarly, any conglomerate concerns that the proposed concentration raises, particularly due to the possible marginalisation of competitors in the provision of ancillary services or in music publishing would also be associated with the merged entity's ability to do so due to its increased market power on the market for the wholesale of physical and digital recorded music.

(767)All of the above concerns are primarily of a horizontal nature and were the focus on the Commission's competitive assessment on the markets for the wholesale of physical and digital recorded music. These concerns would be removed by the commitments accepted in this case.

(768)However, for the sake of completeness, the Commission has undertaken an analysis of the vertical and conglomerate links in this case In this context, The Commission notes that apart from recorded music, music publishing and ancillary services, the Notifying Party is also active in online music video services (through the joint venture VEVO), digital music retail distribution (through zaOza in France and Germany and eCompil in France), telecommunications (through La Société Française du Radiotéléphone -"SFR"- in France, Maroc Telecom Group in Morocco and GVT S.A. in Brazil), pay-TV services (through Canal+ Group S.A. in France and Poland), games (through Activision Blizzard Inc.) and events ticketing services (through Digitick S.A. in France and See Group Limited in the United Kingdom).

(769)EMI is also active in physical music retail distribution and has a music store in Vienna (Austria), and in digital music retail distribution through the e-commerce web site www.nowmusicstore.com.

(770)Activities related to digital retail distribution, physical retail distribution, music video services, electronic communications and games are downstream activities of recorded music markets. Hence, the Commission will assess the risk of vertical effects related to the proposed concentration. Any such effect would either be based on input or on customer foreclosure according to the Non Horizontal Merger Guidelines. In order to assess these risks of vertical effects, the Commission will examine the ability and incentive of the merged entity to foreclose access to inputs and access to downstream markets. In respect of pay TV services and events ticketing services, the Commission will assess the risks of both vertical and conglomerate effects.

7.3.2.1.Digital retail distribution of recorded music

(771)Universal is active in the digital retail distribution of recorded music with Vivendi Mobile Entertainment S.A. (“VME”) (trading as “zaOza”). zaOza is a subscription based exchange and legitimate content sharing website. zaOza has operated in France since 2008 and in Germany since mid-2010. This subscription-based service offers downloading and streaming services of selected film, TV, video, games and music content. zaOza enables consumers to access content from different devices such as a mobile phone or a personal computer. Users may share stored files with other subscribers. In 2010 zaOza’s turnover was […]* in France and […]* in Germany. zaOza’s market share of digital retail recorded music markets is approximately [0-5]*% in France and [0-5]*% Germany.

(772)Universal is also active in the music retail sector in France through its eCompil music service, which offers both à-la-carte download and subscription services. eCompil’s total revenues for 2010 were approximately […]*, which would equate to less than [0-5]*% share of the digital wholesale market in France in 2010.

(773)EMI operates an e-commerce website (www.nowmusicstore.com) aimed at offering physical and digital formats of EMI’s “Now That’s What I Call Music” compilations in the United Kingdom. The current website was launched in early 2011 and its sales from its re-launch to date have been […]*. This would equate to less than [0-5]*% share of the digital wholesale market in the United Kingdom.

(774)The Notifying Party points out that the merged entity's incentive to engage in an input foreclosure strategy would be limited by the following factors:

(a)Universal's lost revenues from third party digital and physical retailers (such as iTunes, Amazon and Spotify) would vastly outweigh the increased revenues attributable to the merged entity’s own retail music activities;

(b)A strategy of input foreclosure would materially undermine Universal’s commercial relationships with its key digital customers who often operate both globally and across a number of key markets in the EEA; and

(c)Such a strategy would also materially undermine Universal’s collaborative relationships with all artists on its roster and would fundamentally undermine its ability to sign new artists and retain existing artists.

(775)As regards the input foreclosure, the Non Horizontal Merger Guidelines provide that “the merged entity faces a trade-off between the profits lost in the upstream market due to a reduction of input sales to (actual or potential) rivals and the profit gain, in the short or longer term, from expanding the sales downstream or, as the case may be, being able to raise prices to consumers” . The Commission considers that, in the light of the limited revenues of the merged entity's retail platforms (less than [0-5]*% of EEA-wide revenues and less than [0-5]*% of any national revenues) and their limited geographic scope (only in France for eCompil, only in France and Germany for zaOza and only in the UK for www.nowmusicstore.com), the merged entity will likely have no incentive to engage in any strategy of input foreclosure. Any such attempt would lead to a loss in wholesale digital revenues that is not commensurate to any possible gain in the retail digital revenues.

(776)In relation to customer foreclosure, the Notifying Party submits that it would likely have no incentive to foreclose record companies competing with Universal as customer foreclosure may occur when a supplier is integrated with an important customer in the downstream market. Conversely, the merged entity’s retail music services are limited when compared to the retail level of the supply chain of recorded music across the EEA. Furthermore, any attempt to foreclose Universal's recorded music competitors from access to its retail services would significantly adversely affect the merged entity’s retail services' commercial attractiveness to consumers, resulting in a loss of revenues, which would vastly outweigh the increased revenues as a result of increased downloads or streams of Universal and EMI recorded music.

(777)The evidence collected through the market investigation indicates that a majority of competitors of the Notifying Party believe that Universal could favour its own distribution channels and could benefit from the convergence between digital music retail distribution and recording of music. However, none of the digital customers considers eCompil, zaOza or www.nowmusicstore.com as a credible competitor. Similarly, none of the digital customers considers that there is any significant risk of foreclosure due to the vertical integral of Universal or EMI with these three platforms. The Commission also considers that given the limited size of the merged entity's three retail platforms, foreclosing the merged entity's

competitors from access to these platforms would be likely to have a negative effect on the Notifying Party's retailing revenues, whilst it would not have any positive impact on the merged entity's wholesale revenues as its competitors would still have access to the vast majority of the retail platforms that compete with the merged entity's platforms.

(778)Hence, the Commission considers that the proposed concentration would not be likely to result in a significant impediment to effective competition due to non-horizontal effects in the market for retail distribution of digital music.

7.3.2.2.Online music video services

(779)VEVO is an advertising funded music video service, offering consumers the ability to watch music videos for free and is available both online and across a range of mobile devices through the VEVO Application.

(780)Universal holds a […]*% shareholding in VEVO. VEVO is a Delaware limited liability company and is jointly owned by Universal ([…]*%), Sony Music Entertainment ([…]*%) and a wholly owned affiliate of the Abu Dhabi Media Company, ADMC US Holdings, Inc. ([…]*%).

(781)According to its audited accounts, VEVO’s global revenues for 2010 were […]*. Its global revenues for 2011 were […]* but its accounts have not yet been audited. VEVO was launched in the United Kingdom late in 2011 and is expected to be rolled out in other EEA countries during 2012, notably in France. The Notifying Party submits in the Form CO that it received […]* received in 2010 and […]* in 2011 from VEVO.

(782)VEVO’s content is available on YouTube, VEVO.com, and through a VEVO- embedded media player. YouTube steers its traffic for music videos to VEVO such that there is a VEVO channel on YouTube. […]*

(783)The majority of respondents to the market investigation consider that the proposed concentration would not lead to any risk of anti-competitive effects due to the link between the merged entity's activities in the wholesale of recorded music and its activities in online music video services .

481 Customers' reply to the Commission Request for information of 17 April 2012, question 56.

195

(785)Accordingly, the Commission finds that the proposed concentration would not be likely to result in a significant impediment to effective competition due to the vertical links between the merged entity's activities in the wholesale of recorded music and its music video services.

7.3.2.3.Mobile telephony and broadband services

(786)SFR is a wholly-owned subsidiary of Vivendi. Vivendi bought out Vodafone’s minority shareholding in SFR in April 2011. SFR is active in the markets for mobile telephony and broadband internet services in France.

(787)The market for mobile telecommunication services has been found by the Commission in the past to be divided into wholesale and retail products markets with a national geographic scope. The market for the provision of broadband internet access has also been divided into wholesale and retail product markets. The retail market has previously been found to have a national geographic scope and the wholesale market’s geographic scope has been left open.

(788)SFR is the second largest mobile telecommunications operator and the second largest high-speed internet operator within France. SFR is active in both the retail and wholesale markets for mobile telecommunications services. Likewise, SFR is active in both the retail and wholesale provision of broadband internet access. SFR has an estimated market share of approximately [30-40]*% in the French market for mobile telecommunication (with around […]* mobile customers) and [20-30]*% in the French market for high-speed internet provision (with around […]* broadband internet customers and […]* WiFi hotspots).

(789)The Notifying Party submits that SFR faces vigorous competition in mobile telecommunications from the market leader Orange France (with an estimated [40-50]*% market share) and Bouygues Telecom (“Bouygues”) ([10-20]*%), as well as from Mobile Virtual Network Owners ([5-10]*%) such as Virgin Mobile, Carrefour Mobile, Auchan Mobile and Simyo. Furthermore, a fourth mobile phone operator with its own network, Free Mobile, entered the market on 12 January 2012. SFR’s principal competitors in the provision of broadband internet are the incumbent and market leader France Télécom (with an estimated share of [40-50]*%), Iliad ([20-30]*%) and Bouygues Télécom ([0-5]*%). Other broadband operators account for around [5-10]*% of the retail market in France.

(790)As regards music services, SFR offers a range of services to its customers: first, it offers ringtone and ring-back tones as well as a wider MP3 download service. Secondly, in July 2011, SFR launched a streaming service that integrates Spotify into three of its tariff plans for mobile customers. As an alternative, SFR's mobile customers can also subscribe to Spotify on an "add on" basis, paying a monthly fee which is billed and collected by SFR.

(791)Regarding input foreclosure, the Notifying Party submits that Universal would not have the ability or incentive to foreclose downstream competitors of SFR by

482See Commission's decision of 1 March 2010 in Case No COMP/M.5650 T-Mobile/Orange.

196

(791)reducing access by the downstream competitors of SFR to its music repertoire for the following reasons:

(a)[…]*;

(b)Any attempt from Universal to restrict access to its music to SFR’s competitors would not be profitable as this strategy would mean that Universal would forgo [60-70]*% of the French mobile telephony market and [70-80]*% of the French market for high-speed internet provision. Furthermore, customers of SFR's competitors could still access music on their mobile handset, tablet computers or personal computers through a variety of other well-established channels such as iTunes, Amazon and Spotify; and

(c)The combined revenue, generated by SFR from both ring tones and full track downloads, was […]* in 2010 and […]* in 2011. This represents less than […]* of SFR’s total revenues (for example music revenues for 2010 count for […]* of total revenue). Similar conclusions can be drawn for SFR's competitors. For instance, Orange's 2010 revenues were EUR 45.503

483http://orange.com/en/group/key-figures-and-activities/latest-key-figures-2012-first-half-results.

484See paragraph 716 of the Form CO.

(792)Regarding customer foreclosure, the Notifying Party submits that SFR would not have the ability or incentive to foreclose Universal's competitors in the wholesale of recorded music from access to the markets for mobile telephony and Internet broadband by exclusively purchasing music from Universal or by reducing its purchases from Universal's competitors. Such attempt would result in SFR losing access to a significant repertoire of other record companies. Such a foreclosure strategy would also have very little impact on Universal’s competitors because any revenue they would lose from SFR is likely to be small in comparison with their revenues from other major customers in France, such as iTunes, FNAC, Amazon and Deezer.

(793)The evidence collected through the market investigation indicates that competitors of Universal believe that the proposed concentration could favour

485Competitors' reply to the Commission Request for information of 17 February 2012, question 91.

(793)SFR in the French electronic communication markets. However, they do not detail how the merged entity could benefit from the convergence between Universal and SFR. The market investigation also indicates that competitors of SFR have not experienced, at this stage, any explicit conduct from Universal aimed at favouring SFR. Competitors of SFR believe that a possible effect on the French mobile telephony and broadband retail markets would be based on the exclusive distribution of Universal's music through SFR. This would mean that Universal would stop distributing its music through other digital platforms, such as iTunes or Spotify. If Universal continues to distribute its music on the digital market, then competitors of SFR do not believe that there will be any effect on the French broadband and mobile telephony markets.

197

(794)The Commission considers that in the light of SFR limited revenues related to retail distribution of digital music (less than […]*% of EEA-wide revenues) and its limited geographic scope (only in France), any attempt at input foreclosure would lead to a loss in wholesale digital revenues that is disproportionate to any possible gain in the French electronic communication markets. On account of SFR's limited activities, the Commission also considers that there is no risk of customer foreclosure. Hence, the Commission considers that the proposed concentration would not be likely to result in a significant impediment to effective competition due to the vertical links between Universal's activities in the wholesale of recorded music and SFR's activities in mobile telephony and Internet broadband.

7.3.2.4.Physical retail distribution of recorded music

(795)EMI operates a music store in Vienna (Austria) which had revenues of around […]* in its last financial year. This represents around [0-5]*% market share of the Austrian market of physical music distribution.

(796)The evidence collected through the market investigation does not indicate any risk of vertical effect related to EMI's activity in the retail distribution of physical music. Considering the limited geographic scope and the low share of the Austrian physical retail distribution market, the merged entity will have no incentive to engage in any strategy of input foreclosure or of customer foreclosure. Hence, the Commission considers that the proposed concentration would not be likely to result in a significant impediment to effective competition due to the vertical links between Universal's activities in the wholesale of recorded music and EMI's activities in the retailing of music in Austria.

7.3.2.5.Pay TV services

(797)Vivendi solely controls Canal+ Group SA (“Canal+ Group”), which is active in the pay-TV markets through its 80% owned subsidiary, Canal+ France SA (trading as Canal+/CanalSat), in France and, through its 75% owned subsidiary, CANAL+ Cyfrowy Sp. z o.o. (trading as Cyfra+), in Poland. Vivendi has also recently agreed to acquire a minority stake in the Polish TV broadcaster TVN. Canal+ Group also operates both regionally through Canal+ Regie and in a number of other territories including Monaco, Switzerland, French overseas territories, French speaking African nations and Vietnam through Canal+ Overseas. In 2010, Canal+ Group had an annual turnover of EUR 4.7 billion.

(798)At the upstream level, Canal+ Group acquires programming rights and distributes the channel Canal+ and multiple thematic channels (including cinematic, sports, information, documentaries, lifestyle, series and youth channels) offered on the pay TV retail market. Canal+ Group also offers Video on Demand services (CanalPlay) through cable, ADSL and online (using a computer) and, through a partnership with Microsoft, on the Xbox 360. Finally, Canal+ Group is involved in aggregating activities, assembling channels produced by Canal+ Group and third parties into packages to offer to various retailers and customers.

(799)Canal+ Group does not acquire any music recording rights. Instead recorded music rights are acquired by those supplying Canal+ Group with programming content. There is therefore no vertical relationship between Universal recorded music activities and Vivendi’s pay TV activities. The Commission therefore considers that these activities are not relevant for the assessment of the proposed concentration.

7.3.2.6.TV and film production and distribution

(800)Canal+ Group is also active in the co-production of TV content (through its Canal+ TV business) and film content (through StudioCanal SA).

(801)The Notifying Party submits that the share of Canal+ TV co-production for all TV fiction content shown in France is below [0-5]*% and that the share of Canal+ TV co-production for all TV fiction content produced in French and shown in France is around [5-10]*%.

(802)StudioCanal SA (“StudioCanal”), which is a wholly owned subsidiary of Canal+ Group, is active in the production, acquisition and distribution of films to various media outlets (cinema theatres, video, audio-visual and Video-on-Demand). Its principal activity is the distribution of film content mainly in France, the United Kingdom and Germany. It manages and licenses film rights to cinematographic and audio-visual works and on average StudioCanal distributes around 50 new films a year in movie theatres throughout Europe. StudioCanal owns 100% of the share capital of Kinowelt, a German distributor of feature films and supplier of DVDs which has limited activities in co-production as well. StudioCanal also acts as a co-financer (or co-producer) for a limited number of films each year.

(803)The Notifying Party submits that StudioCanal never acts as sole producer and that it never has any de jure or de facto control over the music chosen for a particular film it co-produces. This decision typically rests with the producer and director. As a result StudioCanal does not acquire music directly and has no control over the choice of music acquired for particular film projects it co-produces. The Notifying Party submits that the estimated market share of StudioCanal’s film co-production for all films distributed in France is around [0-5]*% and below [10-20]*% of all films produced in French and distributed in France. Studio Canal’s shares are below the above estimates in all other EEA countries. Kinowelt’s limited co-production activities do not have any material impact on these co-production share estimates. The exploitation by StudioCanal of recorded music rights (for example, in sound track albums or tracks specifically composed for a film) represents only a negligible part of its business, that is to say […]* in 2010 and […]* in 2011. Any links with the recorded music market are therefore extremely limited. Hence, the Notifying Party submits that it has no incentive to implement any foreclosure strategy related to TV and film production and distribution.

(804)The evidence collected through the market investigation did not indicate any concern related to TV and film production and distribution. Considering the limited scope and revenues and the results of the market investigation, the Commission concludes that the proposed concentration would not be likely to result in a significant impediment to effective competition due to the links between Universal's activities in the wholesale of recorded music and the activities of the Canal+ Group in in TV and film production and distribution.

199

7.3.2.7.Games and interactive entertainment

(805)Vivendi acquired control over Activision Blizzard, Inc. (“Activision Blizzard”) in 2008 and reduced its equity stake to around 60% in November 2011. The company has two divisions, Activision Publishing, Inc. and Blizzard Entertainment, Inc. Activision Publishing, Inc. publishes interactive software products and peripherals and develops and publishes games for personal computers, for a wide range of video games consoles (including Sony Playstation 3, Nintendo Wii and Microsoft Xbox 360), for handheld platforms and for mobile phone devices through franchise and licensing arrangements. Blizzard Entertainment, Inc. is a developer and publisher of entertainment software and a developer and publisher of games for Windows and Apple personal computers and specialises in online multi-player games such as “World of Warcraft”. Activision Blizzard is active in video games publishing across the EEA. Activision Blizzard is also active in the EEA with its wholesale game distribution business. Activision Blizzard is also a supplier of logistics services to game publishers in the United Kingdom.

(806)In 2010, Activision Blizzard had an annual turnover of EUR 3.3 thousand million. The principal competitors for Activision Blizzard across the EEA include vertically integrated companies, which have a substantial games hardware business, such as Nintendo, Sony, Microsoft and games developers and publishers which typically create games for multiple platforms, such as Electronic Arts, Ubisoft and Take2.

(807)Activision Blizzard uses music tracks to enhance its video games so that its games may incorporate music based on rights owned by record companies or music publishers. However, some musical content used by Activision Blizzard is created in-house, for instance […]*. Alternatively Activision Blizzard commissions sound tracks directly from authors. […]*.

(808)In the decision COMP/M.5008 - Vivendi/Activision of 16 April 2008, the Commission left open the question of market definition, including whether there was a hypothetical downstream market for music games and whether the markets were national, EEA-wide or wider. However, the Commission considered that that concentration in that case did not raise vertical concerns as:

(a)the hypothetic downstream market for music games only represents a small part of the overall demand for music rights, and therefore a minimal proportion of the revenues generated by the licensing of such rights;

(b)the cost of the music rights only represents a small part of the overall cost of production of games; and

(c)Competing licensors of music rights are able to license rights to music works substitutable to those held by UMG.

(809)This is in line with the Notifying Party's claims. According to the Notifying Party, the value of recorded music rights licensed from record companies can be considered low, amounting to approximately […]*% of total annual purchases made by Activision Blizzard. On average, the costs associated with obtaining the required music rights for use in games represent not more than […]*% of the

200

total production cost of a game (and even for games with substantial music content this percentage is generally not higher than […]*%).

(810)The Notifying Party submits that the games software publishing sector can be divided into the market for the publishing of offline games for PCs, consoles and handhelds, (the market for the publishing of (offline) games for mobile handsets and the market for the publishing of interactive online games. The Notifying Party submits that these markets are at least EEA-wide. The Notifying Party submits that the market shares for Activision Blizzard are in all cases below [20-30]*% (and in many cases significantly below [20-30]*%) for each of the three principal platform segments of personal computers, handheld and console, both on a national basis and an EEA-wide basis. Hence, the Notifying Party claims that the proposed concentration does not give rise to any vertical concerns considering the limited market shares of Vivendi’s activities in the video games sector. The Notifying Party also considers that it has no ability to foreclose downstream competitors since they have many alternative suppliers, including the merged entity's competitors, in-house production and the commissioning of musical works from authors directly.

(811)Finally, the Notifying Party submits that the fact that the revenues from producers and sellers of music games only represents a minimal proportion of the revenues generated by the licensing of recorded music rights excludes the possibility of any significant risk of customer foreclosure.

(812)During the market investigation in this case, a majority of publishers of computer and video games confirmed these claims, by supporting that:

(a)The hypothetic downstream market for music games only represents a small part of the overall demand for music rights (all respondents confirmed it) ;

(b)Access to specific music works is not critical for the video games publishers (all respondents to the market investigation confirmed that); and

(c)Competing licensors are able to licence rights to music works that are substitutable to those held by Universal (2 out of 3 respondents confirmed that).

(813)In short, the market investigation in this case confirmed that the majority of

486487Competitors' reply to the Commission Request for information of 17 February 2012, question 94.

respondents agree with the Commission's findings in Vivendi/Activision .The Commission therefore considers that its analysis in that case also applies in this case. Accordingly, the Commission concludes that the proposed concentration would be unlikely to result in a significant impediment to effective competition due to the links between the merged entity's activities in the wholesale of recorded music and Activision Blizzard's activities in video games.

7.3.2.8.Ticketing services

(814)Vivendi acquired Digitick SA (“Digitick”), a French ticketing business in December 2010 and See Group Limited (trading as “See Tickets”), a United Kingdom ticketing business in September 2011. Digitick provides event ticketing services in France to organisers of shows, cultural and sporting events. These services include marketing, sales monitoring, hosting, SMS positioning and the provision of hotlines. Digitick offers tickets for these events directly to the public on its www.digitick.com website. See Tickets carries out similar activities in the United Kingdom. For the year to September 2011, Digitick and See Tickets’ joint turnover was […]* . Competitors for See Tickets include Ticketmaster, Eventim, Ticketline and Stargreen.

(815)The Notifying Party estimates Digitick’s market share in France to be around [10-20]*% given that the two main competitors in the French market are FNAC (with an estimated market share of [50-60]*%) and Tickenet (with an estimated market share of [30-40]*%).

(816)Neither Digitick nor See Tickets acquire any recorded music rights directly. The evidence collected through the market investigation does not raise any concern about vertical effect related to the merged entity's activities in ticketing services. Accordingly, the Commission concludes that the proposed concentration would not be likely to give rise to a significant impediment to effective competition due to the links between the merged entity's activities in the wholesale of recorded music and Vivendi's ticketing activities.

7.3.2.9.Conclusion

(817)The Commission concludes that the proposed concentration, particularly as modified by the commitments, would not be likely to result in any significant impediment to effective competition due to non-horizontal effects related to the links between the merged entity's activities in the wholesale of recorded music and Universal's and EMI's activities in digital retail distribution, physical retail distribution, music video services, electronic communications, TV and film production and distribution, games and interactive entertainment and events ticketing services.

7.4.GENERAL CONCLUSION OF THE COMPETITIVE ASSESSMENT IN THE RELEVANT MARKETS

(818)The Commission considers that the proposed concentration would likely result in a significant impediment of effective competition, in particular as a result of the creation of a dominant position, in the markets for the wholesale of digital music at EEA level and in 24 Member States, namely Austria, Belgium, Bulgaria, the Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece, Ireland, Italy, Latvia, Lithuania, Luxembourg, Malta, the Netherlands, Poland, Portugal, Romania, Slovakia, Spain, Sweden and the United Kingdom, as well as Iceland and Norway.

202

8.COMMITMENTS

(819)In order to address the competition concerns identified by the Commission in the Statement of Objections, the Notifying Party on 27 July 2012 submitted commitments pursuant to Article 8(2) of the Merger Regulation (the "Commitments of 27 July 2012"). The Commission launched a market test of this set of commitments on the same day in order to gather the views of relevant market participants on their effectiveness and their ability to restore effective competition in the markets where competition concerns were identified (the "market test").

(820)In light of the results of the market test, the Notifying Party submitted a revised set of commitments on 13 August 2012, which was subsequently amended on 25 August 2012 (the "Final Commitments").

(821)The Commission's assessed these commitments and their ability to eliminate the competition concerns in line with its Notice on remedies acceptable under Council Regulation (EC) No 139/2004 and under Commission Regulation (EC) No 802/2004 (the "Remedies Notice") .

8.1.Analytical Framework

(822)Where a concentration would likely result in a significant impediment to effective competition, in particular as a result of the creation or strengthening of a dominant position, the parties may seek to modify the proposed concentration in order to resolve the competition concerns.

(823)Under the Merger Regulation, it is the responsibility of the Commission to show that a concentration would be likely to significantly impede effective competition. By contrast, it is for the parties to the concentration to put forward appropriate commitments. The Commission only has the power to accept commitments that are deemed capable of rendering the concentration compatible with the internal market so that they will prevent a significant impediment to effective competition in all relevant markets where competition concerns were identified. To this aim, the commitments have to eliminate the competition concerns entirely and have to be comprehensive and effective from all points of view .

8.2.Description of the Commitments of 27 July 2012

(824)The Commitments of 27 July 2012 consisted of the divestment of a number of EMI and Universal assets (the "Divestment Commitment"), the termination of certain licensing / distribution agreements (the "Commitment to terminate licensing / distribution agreements") and a number of behavioural commitments (the "Behavioural Commitments").

488OJ C 267, 22.10.2008, p.1.

489Remedies Notice, paragraph 9.

203

(825)In terms of revenues, the total size of the Commitments of 27 July 2012 was […]*. This accounted for approximately [60-70]*% of EMI's pre-merger

490EMI's recorded music revenues (including distribution deals) amounted to […]* and accounted for about […]* of its total revenues in the EEA in the last fiscal year (2011/2012).

8.2.1.Divestment Commitment

(826)The Divestment Commitment consisted of a combination of the following EMI and Universal assets (hereinafter referred to as "Divestment Business") as follows:

(a)the following assets of EMI Records Limited ("EMIRL"): all Parlophone-branded artists, such as Coldplay, Lily Allen, Tinie Tempah, Kylie Minogue, Eliza Doolittle, Gorillaz, Blur, Radiohead, Pet Shop Boys and Tina Turner (but excluding the rights to the Beatles catalogue collectively and individually); the rights to David Guetta’s future recordings pursuant to his current recording agreement with EMIRL; the rights to the works of the EMI Classics artists, including Sir Simon Rattle, Herbert von Karajan, Daniel Barenboim, Yehudi Menuhin, Vanessa Mae and Maria Callas; the rights to the works of Virgin Classics branded artists, such as Veronique Gens, Natalie Dessay, Piotr Anderszweski, Emmanuelle Haim, Philippe Jarousky and Artemis Quartet; and the Pink Floyd catalogue;

(b)other EMI labels: Chrysalis Records Limited ("Chrysalis") and its sub-label Ensign Records Limited ("Ensign") (featuring The Ramones, Jethro Tull, Spandau Ballet, Specials, Runrig, Proclaimers and Sinead o'Connor, but excluding Robbie Williams), and Mute Records Limited ("Mute") (featuring Depeche Mode, Moby, Richard Hawley, Goldfrapp and Nick Cave & the Bad Seeds);

(c)EMI Music Belgium BVBA ("EMI Belgium") (featuring Clouseau, Ozark Henry, Buscemi, Novastar, Das Pop and Flemish folk singers such a no s Bart Peeters and Raymond van het Groenewoud);

(d)EMI Czech Republic s.r.o. ("EMI Czech Republic") (featuring Tri sestry, Orlik, Kabat, Lucie Bila and Daniel Landa);

(e)EMI France SAS ("EMI France") (featuring the David Guetta catalogue, no Julien Clerc, Edith Piaf, Alain Souchon, Renaud, Diam's and Matt Pokora), but with the exclusion of EMI's France 51% controlling stake in Play.on SAS ("Play.on");

(f)EMI Music Norway AS ("EMI Norway") (featuring Madrugada and Silvert Hoyem, Lene Marlin and Morten Abel);

(g)EMI Music Poland sp. z.o.o. ("EMI Poland") (featuring Gregorz Ciechowski, Republika, Bajm, Manaam and Budka Suflera);

490EMI's recorded music revenues (including distribution deals) amounted to […]* and accounted for about […]* of its total revenues in the EEA in the last fiscal year (2011/2012).

204

(h)EMI Music Portugal lda ("EMI Portugal") (featuring Madredeus, Mariza, Da Waesel, GNR and the international fado artists Carmane and Caminho);

(i)EMI Music Sweden AB ("EMI Sweden") (featuring Lasse Stefanz, Magnus Uggla and Eric Amarillo);

(j)EMI's share in the "NOW" compilation business;

(k)the following Universal's assets: Sanctuary Records Group Limited ("Sanctuary") (featuring Black Sabbath, Iron Maiden and the Kinks, and the rights to individual albums of The Who, Status Quo and Brian Wilson); Co-op Music Limited ("Co-op") (a label licensing business with also some own A&R activities featuring Garbage, Fleet Foxes, Mumford & Sons and Two Door Cinema Club); King Island Roxystar Recordings AB ("Roxy") (featuring Swedish artists such as Agnes, Eric Saade and Erik Aasle); and MPS Newco ("MPS") (featuring Oscar Peterson, George Duke and Monty Alexander);

(l)Universal's 40% share in Jazzland Recordings ("Jazzland") (featuring Norwegian artists such as Bugge Wesseltoft, Sidsel Endresen and Ola Kvernberg); and

(m)Universal Music A.E. ("Universal Greece") (featuring Eleftheria Arvanitaki, The Going Through Group, no Notis Sfakianakis, Nikos Vertis and Petros Gaitanos).

(827)According to the Commitments of 27 July 2012, only the EEA rights associated with the Divestment Business would be transferred to the purchaser(s) of the assets making up the Divestment Business. The purchaser(s) would have to enter into an exclusive licensing agreement with the Notifying Party to license back any existing and future rights for exploitation outside the EEA. The Commitments of 27 July 2012 covered both the physical and digital exploitation rights for the Divestment Business.

(828)To the extent that the transfer of one or more of the relevant artist agreements required the consent of the artist, the Notifying Party committed to take all reasonable steps to procure such consents. […]* not to re-sign such artist(s) for a period of ten years.

(829)The Notifying Party further committed not to acquire direct or indirect influence over the whole or part of the Divestment Business without the Commission's prior approval for a period of ten years as of the date of the adoption of this decision.

(830)In terms of revenues, the size of the Divestment Business amounts to approximately […]*% in the EEA. This is the equivalent of […]*% of EMI's EEA-wide recorded music revenues in the last fiscal year (2011/2012) from

491This excludes distribution deals.

205

8.2.2.Commitment to terminate licensing / distribution agreements

(831)In addition to the Divestment Commitment, the Notifying Party committed to terminate certain licensing / distribution agreements relating to the content of third party record companies. The Notifying Party also committed not to enter into any licensing / distribution agreement relating to any repertoire covered by such agreements in the EEA or parts thereof for a period of ten years from the adoption of this Decision. More precisely, Notifying Party committed to terminate the following agreements:

(a)EMIRL's licensing agreement with […]* and […]*;

(b)a licensing agreement between […]* and […]*; and

(c)a sales and distribution agreement between Universal Music Operations Limited and Ministry of Sound Recordings Limited (the "MoS distribution agreement").

(832)Moreover, the Notifying Party committed not to bid for a period of […]* years for the exclusive physical and digital distribution rights for the […]*.

(833)This commitment accounted for about […]* of revenue in 2011/2012.

8.2.3.Behavioural Commitments

(834)The Notifying Party also offered to enter into a number of binding supplementary behavioural commitments. Such commitments were offered to the benefit of a 492 limited number of new digital customers , which have adopted a certain pre- 493 defined business model .

(835)Specifically, the Notifying Party proposed to undertake for a period of five years not to require the following from new digital services:

(a)MFN clauses on remuneration and promotional space to the extent that excluding any such clause from the agreement between Universal and the relevant digital service would not unfairly discriminate against Universal in respect of any and all terms;

(b)[…]*;

(c)[…]*; and

(d)[…]*.

8.3.Results of the market test of the Commitments of 27 July 2012

(836)The Commission launched a market test of the Commitments of 27 July 2012 on the same day. The Commission sent 238 questionnaires to both customers and

492[…]*.

493[…]*.

206

494competitors of the Notifying Party. The purpose of the market test was to gather the views of the relevant market participants on the adequacy of the proposed commitments to remedy the competition concerns that the Commission identified in this case and on the quality and viability of the various assets proposed for divestment. The market test also sought to assess the effectiveness of the Commitment to terminate licensing / distribution agreements and of the Behavioural Commitments. Finally, the market test aimed at evaluating whether the Commitments of 27 July 2012 could be implemented in practice in particular in view of the strictly personal nature of some of the assets that would be subject to the divestment (that is to say, artist contracts).

(837)Overall, the market test showed that the Commitments of 27 July 2012 presented a number of significant shortcomings.

8.3.1.The scope of the remedies package and its adequacy to remedy competition concerns in the case at hand

(838)The market investigation clearly showed that divestiture of recorded music assets is an appropriate measure to decrease the market power of the merged entity.

(839)With regard to the size and content of the Commitments of 27 July 2012, a majority of customers confirmed that, subject to certain improvements, the scope of the proposed commitments was generally sufficient to restore competition that 495 would be lost as a result of the proposed concentration. They also confirmed that some of the assets included in the Commitments were significant ones with 496 good future revenue making potential. To the contrary, some respondents indicated that, in light of Universal's pre-existing strong market position it was questionable whether the competition issues arising from the proposed concentration could be remedied at all.

(840)A large number of customers, which expressed an informed view on the 497 Commitments of 27 July 2012 , highlighted a number of elements, which needed to be amended / included in any suitable commitments to address the competition concerns raised by the proposed concentration. These customers for instance highlighted that it needed to be ensured that as many top revenue generating artists as possible are included in the divestiture and that pursuant to the Commitments of 27 July 2012, the merged entity might have still retained a large number of such artists. They also underlined that the assets included in the commitments needed to have sufficient revenue potential in the future, including from artists with future delivery obligations. Finally, these customers referred to a number of risks for the implementation of divestiture commitments in this industry, such as the uncertainties for artists when the divestiture is to an unknown purchaser and takes considerable time to be completed, the duration of

494The Commission sent 176 to customers (63 replies) and 62 to competitors (30 replies).

495Customers' reply to the Commission Request for information of 27 July 2012.

496Customers' reply to the Commission Request for information of 27 July 2012, question 1.

497A number of customers provided very rudimentary (and therefore not very meaningful) replies to the market test questionnaire, essentially limited to answering yes or no to specific questions without providing any explanation of the reasons supporting their reply.

207

(841)the rights to songs that are ultimately included in the assets and the need to 498 procure consents from individual artists .

(841)These customer's comments went in the same direction as those of the vast majority of competitors to the merged entity, which also indicated that the Commitments of 27 July 2012 needed to be substantially improved in various 499 ways . In light of these results, the Commission had to make a balanced assessment as to which improvements to the Commitments of 27 July 2012 were needed to conclude that the likely significant impediment to effective competition caused by the proposed concentration would be removed in its entirety.

(842)The market test generally pointed to the fact that the size and scope of the Divestment Commitment had to be assessed from the point of view of its quality and lasting viability. In this context one customer explained that "[i]ndeed, even if the divestment represents […]*% of EMI recorded music revenues in 2011, the relevant question is according to us, what value the divestment package will 500 represent in 5 to 6 years from now" .

(843)With regard to the composition of the remedies package, respondents indicated that divestiture of owned content is generally more effective to eliminate the significant impediment to effective competition than the severance of licensing / distribution agreements. These respondents indicated that the divestiture of owned content is of a more lasting and structural nature and confers stronger bargaining power in commercial negotiations. Conversely, respondents indicated that the commitment to terminate some licensing / distribution agreements is not of the same structural nature as divestment of owned repertoire and therefore is 501 not really suitable to remedy competition concerns in this case. This is mainly because licensing / distribution agreements are of a temporary nature and may therefore not get re-assigned to Universal. Indeed, according to the market participants Universal can easily replace licensing / distribution agreements by other similar agreements. In this context, IMPALA also noted that licensing / distribution agreements "do not have the same value as the owned content in the divestment catalogue in terms of long term market power. Plus there is nothing in the current package which would prevent Uni simply acquiring other similar 502 licensing and distribution arrangements" . Moreover, respondents indicated that licensing / distribution agreements generally generate lower profitability than 503 the sales and distribution of owned content. Finally, respondents pointed out that such a commitment would affect the independent label's freedom to choose a distribution partner and therefore would reduce distribution options for these third parties.

498Customers' reply to the Commission Request for information of 27 July 2012, question 1.

499Competitors' reply to the Commission Request for information of 27 July 2012, question 1.

500Customers' reply to the Commission Request for information of 27 July 2012, question 1 (ID: 9195).

501Competitors' reply to the Commission Request for information of 27 July 2012, questions 32 and 33.

502Competitors' reply to the Commission Request for information of 27 July 2012, question 1 (ID: 9409).

503Competitors' reply to the Commission Request for information of 27 July 2012, questions 32 and 33.

208

(844)The Commission therefore concluded that the Commitments of 27 July 2012, while including certain valuable assets, required a number of improvements, in particular in terms of their size and content (both in terms of owned content included and of ratio between owned content and licensing/distribution agreements), to be suitable to address the competition concerns raised by the proposed concentration. The Commission also concluded that the quality and lasting viability of the assets included in the Divestment Business are key elements to assess its suitability to address any such concerns.

8.3.2.Quality and viability of the assets contained in the Divestment Business

(845)The market test also sought to assess whether the assets contained in the Divestment Business are such that the market share that the Notifying Party would divest would likely remain stable over time. It was therefore necessary to assess whether the artists included in the Divestment Business constitute an adequate mix between back catalogue generating steady revenues and contemporary artists which are likely to continue generating hits in the future.

(846)In this context the market test generally indicated that the quality of the each of the assets included in the Divestment Business varies. While the Parlophone-branded artists (including artists such as Coldplay) and EMI France (containing David Guetta) were considered as high quality assets capable of generating revenues in the future, other assets were considered of lesser value either because their revenues would be on a downward path […]* or because they cannot be sufficiently operated on a stand-alone basis, for instance because they can only be operated with additional assets (for example the local entities included in the Divestment Business).

(847)With regards to the mix between back catalogue and contemporary artists, the market test revealed a mixed view. While competitors tended to consider that the Commitments of 27 July 2012 are weighted towards back catalogue, customers 504 indicated that the package was adequately balanced. The fact that artists like Coldplay or David Guetta were included in the remedies package was considered as important by both customers and competitors who all predicted that these artists are likely to continue generating hits in the future. In this context, one market participant noted that "[l]a coprésence d'artistes tels que Pink Floyd et 505 David Guetta garantit un bon équilibre entre fond et nouveauté" , while another stated that "[t]he repertoire contains a good balance between back catalogue of great artists such as Pink Floyd while having the future potential of 506 artists such as Coldplay and David Guetta". Conversely, another competitor indicated that: "t]he package appears to be heavily weighted towards back catalogue (with the exception of Coldplay and David Guetta, the repertoire 507 offered appears to be very limited in terms of future hit-making potential)" .

504Competitors' and customers' reply to the Commission Request for information of 27 July 2012, questions 4.

505Customers' reply to the Commission Request for information of 27 July 2012, question 4 (ID: 9256).

506Competitors' reply to the Commission Request for information of 27 July 2012, question 4 (ID: 8912).

507Competitors' reply to the Commission Request for information of 27 July 2012, question 4 (ID: 9198).

209

(848)The quality of the assets, whether back catalogue or contemporary artists, ultimately depends on the sustainability of the sales. As was explained convincingly by one market participant: "[t]he quality of recorded music catalogues is assessed on the longevity of sales, i.e., the ability to achieve long- term returns given limited amount of marketing and promotion expense to keep the repertoire relevant. The quality of a portfolio of active artist is assessed on the consistency of sales achieved by the artists’ on-going releases, i.e., the sustainability of the artists’ relevance in the market place. Anglo-American contemporary repertoire, both for back catalogue and active artists, has proven to be must have content which continues to attract demand when new technologies and distribution channels emerge as well as when new younger and international audiences are reached the first time (in the other core markets, e.g., France and Germany, local contemporary repertoire alongside Anglo-American repertoire is required as well). Ownership of such must have content, both for active artists and for back catalogue, is required to gain direct access with major online and physical retailers and distributors. Major online and physical retailers and distributors do not take into account classical and jazz repertoires when determining which rights organization with which to have 508 direct relationships" .

(849)It was generally considered that Anglo-American repertoire has more future revenue potential. Some market participants indicated that the remedies package may not include a sufficient amount of American repertoire which was considered to have the largest hit-making potential and to be most widely sold. For instance, one customer indicated that there appears to be "little or no 509 American repertoire at all", while another stated that "there is very little 510 American repertoire". According to these respondents American repertoire also represents a substantial amount of EMI's current sales.

(850)It was also pointed out that some valuable assets were excluded from the remedies package, namely The Beatles and its individual members, part of the repertoire of EMIRL, which included significant artists like David Bowie, Kate Bush and Kylie Minogue, Robbie Williams and Play.on. This was labelled by 511 some respondents as "cherry picking" .

(851)Doubts were also raised as to the significance of the package and its lasting nature in the sense that it was unclear how many of the artists signed to the legal entities being proposed for divestment had contracts that would be expiring 512 soon . Some respondents indicated that some rights of the Divestment Business will soon be reverting to the artists and therefore will not decrease Universal's

508Competitors' reply to the Commission Request for information of 27 July 2012, question 5 (ID: 8848).

509Customers' reply to the Commission Request for information of 27 July 2012, question 2 (ID: 9059).

510Customers' reply to the Commission Request for information of 27 July 2012, question 2 (ID: 8906).

511Competitors' reply to the Commission Request for information of 27 July 2012, question 6. "[w]*e note that in respect of the UK, Universal sought to "cherry pick" for retention certain key artists such as the Beatles and Robbie Williams." (ID: 9413).

512See, for example: […]* reply to the Commission Request for information of 27 July 2012, (ID: 9059); […]* reply to the Commission Request for information of 27 July 2012, (ID: 9195); […]* reply to the Commission Request for information of 27 July 2012, (ID: 9018); and […]* reply to the Commission Request for information of 27 July 2012, (ID: 9198).

210

catalogue in the same way as artists who still have some future delivery obligations.

(852)The market test highlighted that the package may not contain sufficient local repertoire, in particular in relation to a number of countries where the proposed 513 concentration would be likely to significantly impede effective competition. In this context one competitor pointed that "they are not offering to divest any Spanish repertoire at all, which I think is significant in a market which is 514 famously dominated by local artists" .

(853)The Commission therefore identified concerns as to the composition (in terms of adequate mix between back catalogue and active artists, as well as of amount of local repertoire included), quality (in terms of ability to generate future sales) and lasting nature (in terms of residual duration of some of the artist agreements included) of the Divestment Business.

8.3.3.Geographic scope of divested rights

(854)The market test also aimed at assessing the implication of splitting divested rights between the EEA (which would be included in the Divestment Business) and the rest of the world (which would remain with Universal).

(855)Respondents to the market test overwhelmingly confirmed that the splitting of the divested rights is a significant shortcoming in the Commitments of 27 July, which is likely to affect the viability and attractiveness of the Divestment Business. These respondents explained that for viability and competitiveness purposes worldwide rights would have to be divested (that is to say, there should not be any licence back to Universal for the exploitation of the divested rights outside the EEA) . For instance, IMPALA explained that "while the package contains some good content, there is a license back to UMG for world rights outside the EEA. Given the importance of many of the recorded music markets outside the EEA – e.g. the US, Asia – and the fact that these markets, especially

513See, for example: […]* reply to the Commission Request for information of 27 July 2012 (ID: 9161); […]* reply to the Commission Request for information of 27 July 2012 (ID: 8848); […]* reply to the Commission Request for information of 27 July 2012 (ID: 9023); […]* reply to the Commission Request for information of 27 July 2012 (ID: 9177); […]* reply to the Commission Request for information of 27 July 2012 (ID: 9189); […]* reply to the Commission Request for information of 27 July 2012 (ID: 9062); […]* reply to the Commission Request for information of 27 July 2012 (ID: 8892); […]* reply to the Commission Request for information of 27 July 2012 (ID: 9198); […]* reply to the Commission Request for information of 27 July 2012 (ID: 8949); […]* reply to the Commission Request for information of 27 July 2012 (ID: 8980); […]* reply to the Commission Request for information of 27 July 2012 (ID: 9059); IMPALA reply to the Commission Request for information of 27 July 2012 (ID: 9203); and […]* reply to the Commission Request for information of 27 July 2012 (ID: 9413).

514[…]* reply to the Commission Request for information of 27 July 2012, question 3 (ID: 9189).

515The majority of customers who replied to the market test indicated that splitting the rights geographically would affect the viability of the divestment package. The vast majority of competitors also concluded that an exclusive licence back to universal for rights outside the EEA would have an impact on the viability of the package.

211

(856)that in the US, are linked to that in the EEA, this will impact the effectiveness of 516 the divestment package in the EEA" .

(856)Respondents also indicated that the geographical split of divested rights would be at odds with how the industry is organized. In this context, respondents noted that the marketing and promotion of artists tends to be global and that record companies typically represent artists on a global basis, with superstars having global distribution agreeements. One market participant explained that "[A divestiture] restricted to EEA assets is nonsense. Record labels operate in a global market, artists need global representation, and dealings with digital services are global. Artists cannot be properly represented these days on a split 517 territory basis" . Furthermore, as one competitor noted: "[… ] digital services operate on a global basis and require worldwide licences. As a result, to ensure that an artist can maximise his or her income from these sources, rights management firms require worldwide licences. If the rights management firm does not have worldwide rights, the streaming service, cloud operator or film company may choose not to license an artist or repertoire that is split among multiple territories because of clearing rights, collecting the royalties and tracking income is much more 518 difficult for split rights". This was also confirmed by customers, which explained that from the customers' perspective "geographically fractured rights 519 do not make licensing easier, faster and more transparent" . Another customer commented that the EEA geographical coverage foreseen in the commitments was "insufficient to balance competition of a global operator where license 520 arrangements are negotiated on a global (and not regional) basis" .

(857)Similarly, it was highlighted that split distribution agreements may have an adverse impact on artists. Respondents indicated that from the artists' perspective 521 "cross-territorial marketing of one artist is more complex"and "it could also cause tension if different territories saw very different directions and marketing 522 plans for a particular artist" . Another customer pointed out that "[i]t can be frustrating for an artist and management if the release and promotional strategies of the different labels clash or are inconsistent, or if one label is not supporting a release or chooses not to release at all. This circumstance can in turn impact the ability of the labels to generate critical mass of buzz around a 523 release or an artist". Yet another respondent stated: "[r]ecord labels generally seek to obtain worldwide agreements as this allows them to ensure a coherent commercial and marketing strategy. Artists generally have the same interest in having a single record label. A global agreement also has the benefit of reducing

516IMPALA reply to the Commission Request for information of 27 July 2012, question 1 (ID: 9203).

517[…]*.

518[…]*.

519[…]*.

520[…]*.

521[…]*.

522[…]*.

523[…]*.

212

(858)the costs associated with marketing and promotion as certain common costs can 524 be recouped on a worldwide basis" .

(858)Respondents explained that for the purchaser(s) of the assets in the Divestment Business, the splitting of rights would be likely to create peculiar situations where Universal, if it were to retain the majority of the rights, would determine commercial strategy, thus enabling it to impact the viability of the Divestment Business. One respondent noted that "[… ] the package would be a marginalized business operating in a global digital world. When licensing digital services 525 globally UMG will dictate the terms as they own most of the territories" . Another respondent stated that "[i]t puts UMG in position of strength globally. As the US is the key territory for the world, it could also mean that the important decision making on an artist still comes directly from UMG. The main profits 526 would also be made from the US". It could even result in a situation where Universal could block global deals. In this respect, […]* noted the following: "[…] given that the US market is the largest digital market in the world and that arrangements with digital services typically cover worldwide rights (and in the vast majority of cases at least North American and EEA rights), then: […] in securing for itself a licence back of rights outside of the EEA, the package would in terms of negotiating position effectively retain for Universal the value of the entire EMI business to leverage in its worldwide negotiations with digital services – i.e. in terms of licensing digital services for the EEA, Universal would under these proposals still be able to leverage the value of the entire EMI 527 business […]" .

(859)The Commission therefore concluded that splitting the divested rights between the EEA (which would be included in the Divestment Business) and the rest of the world (which would remain with Universal) would likely have had very negative implications for the viability and competitiveness of the Divestment Business.

8.3.4.Suitability and effectiveness of the Behavioural Commitments

(860)Respondents to the market test, in particular customers, welcomed in principle the idea of behavioural commitments as a further means to limit the merged entity's post transaction market power in negotiations with customers. However, it emerged from the market test that the Behavioural Commitments were too restrictive and vague and that this would seriously affect their effectiveness and leave room for circumvention which in turn would create disputes. For instance it was explained that "[d]efinitions are circular and the meaning of them is 528 unclear"or that definitions were not sufficiently broad to be effective in practice.

524[…]*.

525[…]*.

526[…]*.

527[…]*.

528[…]*.

213

(861)Respondents also indicated that it would very difficult to effectively monitor and implement these commitments. In terms of the duration of the Behavioural Commitments, five years was generally considered as short and respondents stated that it should be extended to ten years.

(862)The Commission therefore concluded that the Behavioural Commitments as included in the Commitments of 27 July 2012 would likely be difficult to monitor and to implement in practice.

8.3.5.Likelihood of effective implementation of the Commitments of 27 July 2012

(863)Respondents to the market test highlighted a number of issues with the effective implementation of the Commitments of 27 July 2012.

(864)Some respondents considered that the transfer of the Divestment Business could be complicated by the uncertainty associated with a divestment process where the purchaser is unknown. This could incentivize the artists to renegotiate their contracts or to otherwise abandon their contractual obligations, even in cases where their consent is not formally required. One customer explained that "[i]t would obviously mean an uncertainty to an artist. He or she signed with one label and is suddenly with another label. This may result in artists wanting to get out of the deal they are under to go to a label of their choice if they are able 529 to" .

(865)In order to avoid these issues, respondents underlined the need to transfer entire legal entities as opposed to individual artist contracts. When legal entities are transferred in their entirety, artist consents are typically not needed. According to the experience of one record company "transfers of individual agreements out of a legal entity can require prior approval of the artists if the transfer is not part of a transfer of all or substantially all of the assets of that entity". Unless legal entities are being transferred artist contracts often contain provisions limiting the right to assign these contracts which may have an impact on their transferability and therefore affect the effectiveness of the proposed package. Some artists may even have included in their contracts so-called "change of control" provisions, which would limit the transferability of the contract even in situations where the legal entity to which the artist is assigned is being sold. As explained by one market participant: "[s]ome of the key currently active artists included in the remedies package will have change of control provisions in their contracts so 531 there is no guarantee they will be included in the remedies package" . Should this be the case, the implementation of the commitments could be jeopardised. The market test thus stressed the need for the Commission to analyse the artist contracts included in the Divestment Business to verify that there is no significant risk for the effective implementation of the commitments.

(866)Respondents also considered that there was an implementation risk in relation to the commitment to divest EMI's 50% share in the NOW joint venture, as the

529[…]*.

530[…]*.

531[…]*.

214

(867)purchaser of this share would have to rely on its joint venture partner – Universal – for content to be included in compilations. This was considered to be problematic if the purchaser were not to have equally attractive repertoire on offer. One market participant explained that "it is difficult to understand how another entity (which may have no repertoire to contribute to the albums and no 532 manufacturing capability) would simply step in EMI's shoes". Therefore, the market participants convincingly explained that if not supplemented by an adequate behavioural commitment whereby Universal would commit to stay in the joint venture and continue licensing its repertoire to the joint venture at the pre-merger conditions this commitment could turn out to be ineffective as Universal would have incentives to leave the joint venture and not to license its repertoire.

(867)The Commission therefore identified significant risks, which could jeopardise the effective implementation of the Commitments of 27 July 2012.

8.3.6.Other issues associated with the Commitments of 27 July 2012

(868)Respondents to the market test pointed to issues associated with the transfer of the individual assets of the Divestment Business to a large number of purchasers. In this context, some respondents questioned the stand-alone viability of some assets, in particular the local entities included in the Divestment Business. They pointed to the fact that the operations of these local entities largely depend on the exploitation of rights of international / classical catalogues (in addition to the exploitation of the local repertoire), the link to which would be lost if the divestment were to be made in a fragmented manner. For instance, one market participant highlighted the fact that "a key strength of major record companies is the reciprocal arrangements they enjoy with their international affiliates. In the case of EMI, this enables its European companies to benefit from the investment, risks and successes of their ex-European affiliates. From year to year this provides them with readymade hit repertoire, which can provide an element of 533 certainty or 'backbone' to their business". Similarly […]* pointed out that "[i]f UMG is permitted to sell the individual assets in the divestment package piecemeal, the competitive impact of the divestment package will be largely dissipated and the remedy will be rendered ineffective". In the same context, another market participant stated that there "[d]oesn't need to be one single purchaser but at the same time it shouldn't be singled out on more than 2 or 3 companies. If the Divestment Business is purchased by a large number of different companies we would still have the same situation with one huge record company and a lot of very small".

(869)On the other hand, for reasons associated with their financial capabilities, independents tended to consider that the divestment should be to several purchasers and consist of multiple small packages so as to enable many independents get a share of the Divestment Business. In this context, IMPALA

532 […]*. 533 […]*. 534 […]*. 535 […]*.

215

(870)noted that "the independent sector should have the opportunity to acquire some of the divested assets. This will not only benefit competition but also consumer choice and cultural diversity […] Clearly, the financial resources involved may rule Indies out of bidding for the larger parts of the package but they may be able 536 to bid for smaller entities".

(871)The market test indicated that certain other issues would also have to be addressed to ensure the effectiveness and viability of the Divestment Business. For instance, the divested legal entities would have to include the relevant key executives and A&R personnel. As artists have close relationships with the relevant A&R personnel, the transfer of such personnel would incentivise the artists to also transfer their contracts and not to seek renegotiation or severance of their contract. This transfer would also ensure that the Divestment Business has the necessary human resources to nurture and develop the artists. As one customer explained, "music is people business. Of course artists often have personal bindings with the record company people they are working with for 537 years".

(872)Respondents also indicated that the non-solicitation clause of key personnel should be of a longer duration than […]* stipulated in the Commitments of 27 July 2012.

(873)Similarly, the market test highlighted the fact that it is of utmost importance that Universal be prohibited from re-signing any artists that belong to the divested legal entities for a period of ten years.

(874)Finally, the market test indicated that all trademarks related to the Divestment Business should be included for viability reasons.

(875)The Commission concluded that these additional issues raised by market participants should have been addressed in any revised remedy package submitted by the Notifying Party.

8.3.7.Purchaser criteria and interest from potential purchasers to acquire the Divestment Business

(875)The Commission also sought market participants' views on the nature of the purchaser that would be best positioned to acquire and develop the Divestment Business or parts of it. The market test clearly showed that the purchaser should be a record company or a company already otherwise active in the music industry .

(876)Many respondents, including customers and competitors, generally considered the remedies package to be attractive and likely to generate interest from

(877)potential purchasers. Several market participants, including large and small 539 competitors, expressed an interest in all or part of the Divestment Business .

(877)The Commission therefore concluded that the purchaser(s) of the Divestment Business should be a company which is or was previously active in the music industry.

8.4.Final Commitments

(878)The Commission informed the Notifying Party of the results of the market test during a meeting on 9 August 2012. To address the shortcomings identified by the Commission in the Commitments of 27 July 2012 taking into account the results of the market test, as well as further information provided by the Notifying Party on the proposed remedy package, the Notifying Party offered a new set of commitments on 13 August 2012, which was subsequently amended on 25 August 2012 (the "Final Commitments").

(879)The following are the main additional features of the Final Commitments compared to the Commitments of 27 July 2012:

(880)First, the following owned content was added to the Divestment Business:

(a)the repertoire of EMIRL that was excluded from the Commitments of 27 July 2012, with the exception of the Beatles and their individual members, as well as of those artists signed to EMIRL, but branded "Virgin". Additional artists thus included in the Divestment Business include: Iron Maiden, David Bowie, Kylie Minogue, Cliff Richard, Kraftwerk, Whitesnake, Marillion and Kate Bush. These additional assets increase the size of the Divestment Business by […]* of EEA recorded music revenues;

(b)EMI Music Denmark A/S ("EMI Denmark");

(c)EMI Music Spain SL ("EMI Spain"); and

(d)Play.on.

(881)Second, the geographic scope of the remedies was extended to worldwide rights.

(882)Third, the Notifying Party committed:

(a)to sell at least two-thirds of the Divestment Business (EMIRL, the Pink Floyd catalogue and the local entities) to a single purchaser, with that purchaser being a company in a group which is either currently or was previously active in the recorded music or music publishing industry;

(b)not to re-sign artists whose repertoire is included in the package for a period of 10 years; and

539 Competitors reply to the Commission Request for information of 27 July 2012, question 31.

(c)to replace the Behavioural Commitments, which were considered by the respondents to the market test and the Commission as too complex and vague, with a simple commitment by Universal not to enter into any agreement including an MFN clause in Universal's favour (i.e., any MFN clause, pursuant to which the digital customer would be required to also grant Universal any commercial terms that it would grant to other record companies if such term were to be more favourable than the comparable one included in the agreement with Universal) in any agreement with any digital service provider that is active in the EEA for a period of ten years.

(883)In terms of revenues, the size of the revised commitment package was increased to EUR […]* million. This equates to approximately […]*% of EMI's worldwide revenues. The EEA-wide revenues of the package amounted to EUR […]* 540 million, equating to approximately [60-70]*% of EMI's EEA-wide revenues .

(884)Out of these EEA-wide recorded music revenues, EUR […]* million was generated by owned content, equating to […]*% of EMI's EEA-wide owned content recorded music revenues (that is to say, excluding revenues received from compilations, third party distribution or licensing deals and ancillary revenues). The package includes Anglo repertoire that represented around […]*% of EMI's worldwide Anglo repertoire revenues, […]*% of its EEA-wide 541 Anglo repertoire revenues and […]*% of its UK Anglo repertoire revenues .

8.5.Commission's assessment of the Final Commitments

(885)The Commission's assessment of the Final Commitments is based on the results of the market test and further information from the Notifying Party. As concerns the latter, the Notifying Party in particular provided details on the artist contracts included in the Divestment Business. This information included the relevant artists' future delivery obligations, the length of the relevant agreements, the remaining retention period on catalogues, as well as any change of control clauses in the case of the transfer of the legal entity they are signed to and/or any other consent that may be required for the transfer of any artist included in the remedy package.

(886)As explained in Section (241) of this Decision, the proposed concentration would be likely to give rise to a significant impediment to effective competition, in

(918)To address the concern raised during the market test that the rights in the Divestment Business should not be split between the EEA and the rest of the world, the Final Commitments include worldwide rights. This will put the purchaser in a position to effectively develop the marketing and promotion of artists included in the Divestment Business. The fact that the purchaser is no longer required to grant Universal a licence-back for the exploitation of the repertoire outside of the EEA, this also ensures that the links between the merged entity and these artists are permanently severed. The ownership of worldwide rights will allow the purchaser of the Divestment Business full freedom to exploit those rights and to work with the artists and the catalogues in the Divestment Business on a global basis. The Commission considers that this further ensures the viability and competitiveness of the Divestment Business.

(919)On a separate but related note concerning the viability of the Divestment Business, the Commission also considers that, although the likely significant impediment to effective competition that it has identified in this case relates to the wholesaling of digital recorded music, the inclusion in the Divestment Business of both the relevant physical and digital exploitation rights is necessary for the viability of the Divestment Business. This reflects the fact that, in order to ensure the viability of the Divestment Business, it is important that the Purchaser of the Divested Business is able to exploit all rights attached to the divested repertoire. The Commission considers that an artificial split whereby the rights for the wholesaling of digital music would be with the Purchaser, but the rights for the wholesaling of physical music would remain with the merged entity, would prevent the Purchaser from coherently exploiting the artists and catalogues included in the Divestment Business. The Divestment Business therefore includes the right to exploit the divested repertoire both in digital and physical format.

8.5.4.Suitability and effectiveness of the "no MFN commitment"

(920)Although the market participants, in particular customers, welcomed in principle the underlying idea of behavioural commitments, the market test clearly showed that implementing and monitoring the types of behavioural commitments included in the Commitments of 27 July 2012 would have been very difficult and complex. This is because the various commitments were vaguely formulated and included a number of restrictions and exceptions, the application of which would be difficult to monitor.

(921)The Remedies Notice makes it clear that commitments relating to the future behaviour of the merged entity may be acceptable only exceptionally in very specific circumstances and that these commitments will generally not eliminate 549 competition concerns resulting from horizontal overlaps. The Remedies Notice

549Remedies Notice, paragraph 17.

also makes it clear that those types of remedies can only exceptionally be accepted if their workability is fully ensured by effective implementation and 550 monitoring and if they do not risk leading to distorting effects on competition . Where the behavioural remedies that are proposed are so extensive and complex that it is not possible for the Commission to determine with the requisite degree of certainty, at the time of the decision, that they will be fully implemented and that they are likely to maintain effective competition in the market, a clearance decision cannot be granted. The Commission may reject such remedies in particular on the grounds that the implementation of the remedies cannot be effectively monitored and that the lack of effective monitoring diminishes, or 551 even eliminates, the effect of the commitments proposed .

550 551550 Remedies Notice, paragraph 17. 551 Remedies Notice, paragraphs 14 and 69.

(922)The Remedies Notice equally makes clear that according to the case-law of the 552 Court of Justice of the European Union, the basic aim of commitments is to ensure competitive market structures. Accordingly, commitments which are structural in nature, such as the commitment to sell a business unit, are as a rule preferable from the point of the view of the Merger Regulation's objective as they prevent, durably, the competition concerns which would be raised by the merger 553 as notified .

552 553552 See Recital 8 of the Merger Regulation; Case T-102/96 Gencor v Commission [1999] ECR II-753, paragraph 316; Case C-12/03 P Commission v Tetra Laval [2005] ECR I-987, paragraph 86; and Case T-158/00 ARD v Commission [2003] ECR II-3825, paragraph 192. 553 Remedies Notice, paragraph 15.

(923)In line with these principles, following the market test of the Commitments of 27 July 2012, the Notifying Party has offered to divest additional repertoire. The Commission considers that this additional commitment contributes to decreasing the merged entity's post-merger market power on a lasting basis and therefore to removing the competition concerns raised by the proposed concentration on a durable basis.

(924)In light of these additional structural commitments and the concerns associated with the implementation and the monitoring of the Behavioural Commitments, the Final Commitments now only include a straightforward behavioural commitment by the Universal to refrain from including MFN clauses in any contract with any digital customer that is active in the EEA. This commitment was significantly strengthened as compared to that included in the Commitments of 27 July 2012, as it now applies without qualification to any digital customer that is active in the EEA, irrespective of the business model chosen by that customer. The duration of the commitment was moreover increased from five years to ten years.

(925)The removal of MFN clauses assists in fully eliminating the significant impediment to effective competition that the proposed concentration would otherwise be likely to give rise to. The effect of these MFN clauses was that Universal, due to its larger size post-merger, would not only be able to extract favourable terms and licensing conditions for its own repertoire, but would also

550 Remedies Notice, paragraph 17. 551 Remedies Notice, paragraphs 14 and 69. 552 See Recital 8 of the Merger Regulation; Case T-102/96 Gencor v Commission [1999] ECR II-753, paragraph 316; Case C-12/03 P Commission v Tetra Laval [2005] ECR I-987, paragraph 86; and Case T-158/00 ARD v Commission [2003] ECR II-3825, paragraph 192. 553 Remedies Notice, paragraph 15.

benefit from any better terms, including, but not limited to, higher payments, that its competitors, in selected cases and despite their smaller size, would obtain from digital customers. Universal now commits, from the date of this decision onwards, not to enter into agreements with digital customers that are active in the EEA that contain such MFN clauses. This commitment applies in every case that Universal enters into agreements with these digital customers, be it through the conclusion of a new agreement or the re-negotiation or renewal of an existing agreement. Moreover, the commitment to refrain from including MFN clauses in its contracts with digital customers apply to the entirety of Universal's repertoire, including the limited EMI repertoire that Universal would acquire after the proposed concentration. Contrary to the Behavioural Commitments included in the Commitments of 27 July 2012, this unequivocal commitment not to include any MFN clause in Universal's favour, which will apply to any digital customer without qualification, is simple to monitor and thus effective.

(926)Moreover, this behavioural commitment will also assist in levelling the playing field between competitors, including in favour of the purchaser(s) of the Divestment Business, and in ensuring that competitors, including the purchaser of the Divestment Business, have the ability to effectively compete with the merged entity in the future to the ultimate benefit of consumers. Indeed, the removal of MFN clauses in favour of the merged entity in the agreements between the merged entity and digital customers will allow competitors, including the purchasers of the Divestment Business, and digital service providers to negotiate more freely and not under any constraint that the existence of an MFN clause in favour of the merged entity could have posed in these negotiations.

8.5.5.Likelihood of effective implementation of the Final Commitments

(927)The Commission concluded that a number of the concerns that were raised during the market test in relation to the implementation of the Commitments of 27 July 2012 were valid. The Final Commitments include a number of safeguards to ensure effective implementation of the remedies package.

(928)First, to address the uncertainty to which the artists included in the Divestment Business may be subject during the transitional period until their transfer to a purchaser, the period during which Universal is required to sell the Divestment Business […]*. A speedy transfer is essential in this case given the importance of personal relationships in the music industry, the risk that artists renegotiate or terminate their contacts in periods of long uncertainty and that active management of these artists needs to be continuous in order to prevent the erosion of the value of the assets in the Divestment Business.

(929)Second, to minimize the risk associated with the fact that transfer of individual artist contracts may be subject to the artist's agreement, the Final Commitments provide for the divestiture of legal entities. This reflects the fact that artists' consent is generally not required when the entire legal entity, to which they are signed is being transferred. The Commission has also analysed the Notifying Party's information on the artist contracts included in the Divestment Business in order to verify that these agreements do not include so-called "change of control" clauses, that is to say clause, which would require the artist's content in case the

227

legal entity to which the artist is signed is sold. The Commission's analysis confirmed that, with some limited exceptions, the relevant artist contracts do not include any such change of control clauses.

563The correlation coefficient will range from minus one to plus one. A correlation coefficient of minus one would indicate that the two variables move in perfect opposition to each other and a correlation coefficient of one would indicate perfect co-movement, such that whenever variable A rises variable B also rises.

564All correlation coefficients are significant at the 5% level, except for Platform 2 at a yearly level.

565The data provided for the recorded music companies that were not amongst the top seven were dropped for the purposes of this regression analysis. Otherwise, the Commission would end up attaching an artificially high market share for this "Other" catch-all variable. For example, the combined sales generated by the "Other" catch-all variable in many instances are in the range of […]*%. As the margins provided for the "Other" catch-all variable correspond to a weighted average of the individual record company's margins whose market share is significantly lower than […]*.

566The star signs reflect significance level: * indicates significance at 10%, ** at 5%, and *** at 1%. For confidentiality reasons, the Commission anonymised the identity of the platforms.

236

(11)Table 7 shows that the correlation coefficient between the margins and the revenue share is, overall, larger on smaller platforms.Platforms 1, 2 and 3 are the three largest Platforms for which the Commission has collected data while Platforms 4, 5 and 6 are the three smallest. The Commission therefore notes that the size effect is likely to be more pronounced in smaller platforms, as seen by the much higher correlations for Platforms 4, 5 and 6 compared to Platforms 1, 2 and 3. This is also confirmed by the platform regression analysis.

(12)The Commission also recognises, however, that correlation coefficients are descriptive statistics that provide a first screen of the data which do not capture certain characteristics of the data such as time trends and country specificities.

(13)The Commission therefore dealt with these issues by carrying out regression analysis.

(14)In this stage of the analysis, the Commission used a linear regression methodology to investigate the relationship between the size of a recorded music company and its ability to extract rents. Regression analysis is a statistical tool for understanding the relationship between two (or more) variables.

(15)In the base regression, the Commission regresses the margin of recorded music companies against their relative downstream revenue share, controlling for country specific, platform specific and time specific fixed effects. The platform, time and country fixed effects control for otherwise unobserved characteristics present in a given platform. month and country respectively, which may affect margins (technically this is done by introducing a group-specific constant term in the regression model).

(16)The identification strategy of the Commission is based on the cross record company differences in margins and revenue shares. These differences provide the necessary variation in the data for the Commission to identify the size effect. To the contrary, […]*, and given that data exist only for a very limited time period ([…]*), it is not

567On the yearly correlations the Commission observes that on Platform 2 the relationship is not statistically significant, yet positive (and similar to the monthly correlation for Platform 2). However, the correlation coefficient for this Platform is not significant at the yearly level […]* (that is to say, even though the correlation coefficient is larger in the yearly compared to the monthly results). Whether estimates of statistical relationships are found to be significant or not is a function of the number of observations in the sample. Other things being equal, effects are harder to detect in smaller samples. Increasing sample size is often the easiest way to boost the statistical power of a test. As in the yearly data there are fewer observations that in monthly data, this effect is likely to affect the significance of yearly estimates Also, the Commission notes that for platforms 5 and 6, which are smaller platforms, the yearly correlations are substantially higher than for the monthly correlation.

(17)The Commission considers that there are two ways to capture the size effect. One is directly through the coefficient on the revenue shares in the base regressions which do not include record company fixed effects. Alternatively, in this context, whenever record company fixed effects are included in the base model, an analysis of the magnitude of such fixed effects should be undertaken in order to evaluate the existence of the relationship between size and margins.

(18)Table 8 shows, for different regression specifications, the sign of the coefficient estimate for the revenue share.Using the estimated model's coefficients it is possible to calculate the implied margin effect of the proposed concentration (hereafter called "implied margin effect"), which is also shown in the Table. The implied margin effect is the combined entity's margin level, compared to Universal's and EMI's weighted average pre-merger margin level.

(19)Table 8 - Regression analysis - Platform data

Revenue Share

Margin Effect

Model I 1 - Base yearly

+***

[…]*

Model II - Log yearly

+***

[…]*

Model III - IV Lag 1-2

+***

[…]*

Model IV –IV country_platform

+***

[…]*

Model V – Base monthly

+***

[…]*

(19)The results of the yearly base regression (Model I) show a positive (and statistically significant) coefficient between the size of the recorded music company and the margin it is able to extract. The estimated coefficient implies that the combined entity's margin, compared to Universal's and EMI's weighted average pre-merger margin, is expected to increase by around […]*%.

570The analysis presented by the Notifying Party includes these fixed effects. The Commission has assessed the Notifying Party' analysis and concluded that to a large extent the results of the findings based on the base regression still hold. The Commission has dealt with this issue in detail in Annex II (Sections 1.6.2 and 2.1).

571A more detailed summary of the regression results are provided in the Appendix. Due to confidentiality reasons, the regression outputs are displayed in a restricted way. The tables report only the sign and significance level of the estimated main coefficients, as well as their corresponding p-values. The estimated fixed effects (controls) are not reported. The estimated margin effects are reported in ranges.

572In all the regressions the observations are weighted by the revenues generated. Therefore a data point which reflects a high turnover (such as […]*) is given more weight in the regression. In case the results are not weighted by revenues then the impact of downstream revenue share on margins is much higher. This reflects the finding that the impact of the size of the recorded music company on its ability to extract better terms is greater in smaller platforms.

573Platform data for 15 EEA countries, 6 platforms. Robust p-values, * p<0.1, ** p<0.05, *** p<0.01.

238

(20)The Commission has run several robustness checks, in line with the Best Practices for submission of economic evidenceto ensure that the results obtained are not driven by the specification chosen and also to address the Notifying Party's criticisms.

(21)The different specifications demonstrate that the results are indeed robust and show expected margin increases that are both economically and statistically significant. Model I provides the yearly results of the base regression. The logarithmic ("log") specification of the yearly data (Model II) also shows a positive and statistically significant impact of the proposed concentration, albeit with a lower predicted margin increase of […]*%. Several specifications with instrumental variables have also been tested.Two instrumental variables specifications (Model III and Model IV) show results that are in line with the base regression. The monthly specification (Model V) results also point to similar effects.

(22)The regression results computed at an individual platform level also show that the effect is present and significant for all individual platforms, irrespective of the size of the platform and whether it is a download or streaming service.

(23)Table 9 and Table 10 provide confidence intervals for the estimated/implied effect on margins (on the basis of both the yearly and the monthly data) for the individual platform level analysis.

574DG Competition, "Best Practices for the submission of economic evidence and data collection"

575Such tests include i) excluding some fixed effects, ii)clustering in terms of platforms and recorded music companies iii) using instrumental variables (different instrumental variable have been used as explained in more detail in Section 2 of this Annex) iv) introducing the margins and the revenue share variables in logarithmic form,v) running the analysis on a yearly basis vi) restricting the analysis only on the Notifying Parties vii) performing the analysis on an platform by platform basis and viii) performing the analysis on a country level. To address the Notifying Party's comments to the Statement of Objections, the Commission has carried out further robustness checks as shown in Annex II.

576Instrumental variables allow consistent estimation when the explanatory variables (in this particular case the revenues share) are correlated with the error terms of a regression relationship, leading to endogeneity concerns. The choice of instrumental variables in this case is discussed in Section 2 where more sets of instrumental variable regressions are presented (see also Section 2.2 of Annex II).

577Two instrumental variable specifications are presented here; each representing the two sets of instruments discussed in Section 4.2.2. In Model III lagged revenue shares are used as instruments while in Model IV the instruments used are revenue shares of the instrumented recorded music company in all the different countries and platforms at the same time period.

239

245

2.1 Data

(46) As explained above data were collected from six major digital music platforms active in the EEA. For each of these platforms, monthly data for all their digital music sales were requested for fourteen countries in the EEA. Platforms were requested to provide sales for the top seven recorded music companies, aggregating all other recorded music companies into an "Other" category. In particular, monthly information was requested for the downstream/retail revenues that the platform generates from the repertoire of each of the recorded music companiesand the net payment ("royalties") paid back to each recorded music company. Data were provided for the three year period from 2009 to 2011, with some platforms providing data for a shorter period, notably since they were not active in the whole period in all fourteen countries.

(47) To ensure advances and discounts are properly and consistently reflected in the data provided, the Commission has requested that advance payments and discounts are distributed across months proportionally to sales and has confirmed with the platforms by way of follow up correspondence that they have computed the data provided on this basis. The Commission has also checked the consistency of the data provided by the platforms and followed up with questions on the underlying data.

2.2 Methodology and Results

1.2.2 Base regression.

(48) The Commission used a linear regression methodology to investigate the relationship between the size of a recorded music company and its ability to extract rents. In the base regression, the Commission regresses the margin against the downstream revenue share, controlling for country specific (c), platform specific (p) and time specific (t) fixed effects. These fixed effects introduce a group-specific constant term in the regression model. These platform, time and country specific fixed effects (similar to dummy variables) control for unobserved characteristics present in a given platform, year or month and country, respectively.

(49) The following equation is estimated (with r denoting a recorded music company):

margins = α + β∗revenue_share + γ∗FE + δ∗FE+ ε∗FE+ εtcpr tcpr p c t tcpr

(50) In all the regressions the observations are weighted by the revenues generated. Therefore a data point which reflects a high turnover […]* is given more weight in the regression.

(51) Also, the data provided for the recorded music companies that were not amongst the top seven were dropped for the purposes of this regression analysis. Otherwise, an artificially high market share would be attached to the "Other" catch-all variable.

(52) In the base regressions as well as in the further robustness checks, the estimated coefficients are not provided, however the sign and the significance levels are reported. Also, the implied marginal impact is provided. The model is able to give a prediction of the average margin of a record company as a function of its revenue share (in a given platform's revenues). In particular, the revenue share variable is changed for Universal and EMI to the combined revenue share of the two companies, post transaction. Then the estimated model implies new (higher) margins, both for Universal and EMI. The implied marginal effect formula (%) for the level specifications is the following:

(53) The weights, (w), are based on the recorded music companies' downstream revenues. The formulas are evaluated at the averages of the estimation sample. The statistical uncertainty of the estimated price effects is calculated using the Delta-method.

(54) The results of this yearly and monthly regression are reported in Table 8 above.

2.2.2 Excluding some fixed effects and introducing fixed effects for recorded music companies.

(55) The Commission has also checked that the results are not driven by the inclusion of fixed effects. Therefore, different specifications – only country fixed effects, only platform fixed effects, only time fixed effects – were tested. This allows additional sources of variations to affect the results. The results, however, remain largely unaffected as shown in Table A-1for monthly data and in Table A-9 for yearly data.

(56) In a second stage, fixed effects for each recorded music company were included. As shown in the results in Table A-2, the inclusion of such fixed effects affects the results. The impact on the revenue share is now mostly captured by the fixed effects. This implies that the variation across companies is likely an important parameter to identify the effect on the base regression.

(57) Indeed, the identification strategy of the Commission is based on the cross company differences in margins and revenues shares. These differences provide the necessary variation in the data for the Commission to identify such effects. This is because given the nature of the negotiations, […]*, it is unlikely that for a given record company there will be significant variation […]*. Furthermore, given that data exist for a very limited time period only ([…]*) it is not possible to exploit the across time variation to a significant extent. The Commission considers that there are two ways to capture the size effect. One is directly through the coefficient on the revenue shares in the base regressions which do not include company fixed effects. Alternatively, in this context, whenever company fixed effects are included in the base model, an analysis of the magnitude of such fixed effects should be undertaken in order to evaluate the existence of the relationship between size and margins. This is because the company fixed effects represent the relevant cross company variation in the data. The parties have presented models including company fixed effects. The Commission has assessed the parties' analysis and concluded that to a large extent the results of the findings based on the base regression are still accurate. The Commission has dealt with this issue in detail in Annex II.

3.2.2 Clustering in terms of platforms and recorded music companies.

(58) In a regression model if the assumption of independently and identically distributed (iid) error terms is violated, though the parameter estimates are still unbiased, the calculated standard errors may be biased. Consequently, the Commission used the so-called robust standard errors to assess the statistical uncertainty of the estimates.Alternatively, it is possible to calculate unbiased standard errors using a clustering (grouping) assumption. Clustering the errors allows the observations within group i to be correlated in some unknown way, inducing correlation in the error terms within the group i (while groups i and j do not have correlated errors). Clustering errors by time, platform and by recorded music companies are provided below. However, it should be noted that for a small number of clusters (or very unbalanced cluster sizes) clustering errors may provide worse results.Therefore in particular results on platform (as there are only six platforms) clustering are more likely to be unreliable.

(59) Results again are largely unaffected, with the exception of platform clustering, and are provided in Table A-3 for monthly data and in Table A-10 for yearly.

4.2.2 Instrumental variables.

(60) Instrumental variables allow consistent estimation when the explanatory variables (in this particular case the revenues share) are correlated with the error terms of a regression relationship, leading to endogeneity concerns. To dispel concerns of endogeneity on the revenue share variable in the base regression different instrumental variables (IV) have been selected. The first set of instrument used is lags of the independent variable (that is to say, of the revenue share). Different lag specifications of the revenue share variable have been used. By lagging the revenue share (and in particular selecting lag further back in time for the monthly specification) it is likely that in case of time specific shocks this set of instruments can address the endogeneity concerns. The second set of instrumental variables ("instruments") is revenue shares computed either in other countries for each recording companies, (the number of instruments thus equal the number of countries in the dataset), or in each platform across the different countries (the number of instruments is thus the number of countries times the number of platforms, if the platform is active in that country) or on average in the dataset for each recorded music companies (one instrument). In case the shocks have the characteristics of being country specific and/or platform specific, the selection of these instruments can solve the endogeneity given that an instrument based on other countries would remove the correlation between the independent variable and the error term. Last, the analysis also tested instruments that reflect the repertoire contribution of each recorded music company on a platform's repertoire. This set of instruments relies on the fact that it is likely that the repertoire size is predetermined or at least more predetermined than revenues market shares. As this instrument was only available from two platforms, data on the repertoire contribution from each of these two platforms were applied to the other platforms.

(61) The results of the instrumental variables regression suggest that the first set of instruments is more likely to be a good set of instruments. These IV specifications pass the specification tests (Rank-test, Hansen-test, weak IV test), while the other two set of instruments typically do not pass all the three tests. Further instrument selection is discussed in Section 2.2 of Annex II.

(62) The results obtained are largely in line with the results of the base regression, notably in the lag IV specifications. The direction of the potential endogeneity bias tends to magnify the size effect, if anything. Detailed results are provided in Table A-4-Table A-7 for monthly data and in Table A-11-Table A-14 for yearly data.

5.2.2 Introducing the margins and the revenue share variables in logarithmic form (log-linear regression).

(63) The base regression was repeated while introducing the dependent and the revenue share variables in logarithms (logs) as shown in the following equation:

ln(margins )= α + β∗ln(revenue_share )+ γ∗FE + δ∗FE+ ε∗FE+ εtcpr tcpr p c t tcpr

(64) In this specification, the impact on the margins is based on the following equation:

(65) Results for monthly data are provided in Table A-9 and show an implied margin increase of […]*%. Similar results are obtained for the yearly data as shown in the summary Table 8 in Section 1 above and in Table A-10.

6.2.2 Analysis on a yearly basis (aggregate the monthly data into yearly).

(66) The Commission has replicated the base analysis on a monthly level and also on a yearly basis. The reason for doing this is twofold. First, it ensures that the results are robust when aggregated and it further confirms that the results are not driven by the way the discounts and advances are allocated (even though the Commission has confirmed with the platforms that both discounts and advances have been allocated on a (revenue weighted) basis in the data provided). Furthermore, monthly results may simply provide essentially "duplicates" of the same underlying observation as contracts may cover more than a single month.This could potentially lead to lower standard errors and therefore results that are more likely to be statistically significant.

(67) The Commission considers that the regressions on yearly observations better reflect the economic reality of the market as the length of the contracts between record companies and platforms are typically […]*.

(68) Results for the base regression (including the log specification) are provided in the summary Table 8 in Section 1. Further yearly results are provided in the Appendix. Results suggest that the estimated impact is robust and very similar to the monthly level base regression.

7.2.2 Restricting the analysis only to Universal and EMI data.

(69) While not representative of the impact on all the recorded music companies the analysis was replicated only on the basis of Universal and EMI data. This addresses concerns that the inclusion of larger independent recorded music companies ("Large Indies") or the commercial policy of a given recorded music company would drive the results.

(70) Results (for both monthly and yearly data) are provided in Table A-15-Table A-21. The predicted impact on margins is slightly, however, the predicted margin effect is in line with the previous results and it is statistically and economically significant.

8.2.2 Analysis on a platform by platform basis.

(71) The analysis was also replicated for each platform. The regression results at an individual platform level shows that the effect is present and significant for all individual platforms, irrespective of the size of the platform and whether it is a download or streaming service. Very importantly, and in line with the correlation coefficient results provided above (Table 7), the results of the regression analysis suggest that the impact on smaller platforms is greater than on larger platforms.

(72) Platform by platform results for both monthly and yearly data are provided in Appendix Table A-22-Table A-33.[…]*. The implied effect on the merging parties' margins in the base regression is […]* for the smaller platforms, while more limited for larger platforms.

250

(73) This result also indicates that the overall impact computed (across the six platforms) likely underestimates the overall effect in the market as only the main platforms at the EEA level are included in the analysis.

9.2.2 Analysis on a country level.

(74) The monthly and yearly regressions are performed on the fifteen Member States for which the Commission has data.The results, in Table A-34-Table A-39, show that the identified effect is statistically significant in most Member States (in the monthly data it is not significant in Member States, largely due to very few observations for these Member States). The strength of the effect differs by country, with a great number of countries with an implied margin effect greater […]*.

3 RECORD COMPANY DATA ANALYSIS

(75) This Section explains the methodology employed in the record company data (aggregate and track-level data), explains the reasons for the different robustness checks and reports the different results obtained

3.1 Data

1.3.1 Aggregate data

(76) The Commission requested aggregated data from six recorded music companies on their digital sales. The companies include Universal and EMI, as well as two large and two smaller competitors. The time span and frequency of the data is from 2008 to 2012 on a monthly basis, with one of the smaller competitors being able to provide data only on a yearly basis. The geographic coverage of the data includes 14 countries (Austria, Belgium, the Czech Republic, Finland, France, Germany, Greece, Italy, the Netherlands, Norway, Poland, Spain, Sweden and the United Kingdom), with one of the smaller competitors being able to provide data for only four countries. For each country and month, each record company provided data for its largest customers (platforms).The customers include download and streaming platforms, as well as mobile operators. The data is broken down by transaction type (download/stream) and product type (track/album).

(77) The key variables included in the datasets are sales units, wholesale revenue and repertoire size. The revenue is the actual wholesale revenue the record company gets from the platform, with the restriction that it does not include advance payments and

(78) A comparison with the platform data reveals that a data selection is necessary. As explained above, the wholesale revenue variable of the record company data does not include all wholesale revenues, while the royalty measure of the platform data is complete. Table 3-1 shows the absolute differences between the wholesale revenue shares of the platform data and the record company data. For example, for Customer (platform) […]*, the two revenue share measures differ on average by […]* percentage points with a standard deviation of […]*, and for Customer […]* the average difference is […]* percentage points. However, for other customers the differences are more substantial and the standard deviations of the differences are also larger (for example, for Customer […]*, the difference on average is […]* percentage points with a standard error of […]* percentage points). The pattern of differences is similar when the data is restricted to the […]* large countries, which represent […]*% of the sales of the full data.

Table 3-1 Platform vs. record company data: Absolute differences in wholesale revenue shares (%point)

13 countries

4 big countries

mean

st.dev.

mean

st.dev.

Customer_1

[…] *

Customer_2

[…] *

Customer_3

[…] *

Customer_4

[…] *

Customer_5

[…] *

Customer_6

[…] *

Note: Data from platforms and record companies, 2009-2011. […]*.

(79) Therefore, for the two most important customers, Customer 1 and Customer 2 in Table 3-1,the aggregate record company data's sales figures provide an adequate measure of the total wholesale revenue on these platforms, given the relatively small discrepancy between record company and platform data for these Customers. As a consequence, the subsequent analysis has been restricted to these two download customers.

(80) The record company data have also been augmented with a repertoire size variable from the platform data. This variable is available for all record companies on a monthly basis but only from one customer (Customer 2 in the above table). Also, the geographic coverage of this variable is more limited as it is only available for Austria, France, Germany and the United Kingdom.

(81) The data have been aggregated to yearly level, using only data related to the years 2009 to 2011 and the two largest customers to ensure consistency across companies. The selection of these two customers is also motivated by the comparison of the record company data with the platform data. In particular, though the record company data generally provides a less precise measure of the rent extraction abilities of the record companies than the platform data, in the case of the two large customers even the record company data is useful as the discrepancy from the platform data is smaller.Consequently, the analysis of the record company data can be viewed as an auxiliary piece of evidence, providing further support to the results of the platform data analysis.

2.3.1 Track-level data

(82) The Commission also requested track-level data from four recorded music companies on their digital sales. The companies include Universal and EMI, as well as two large competitors. The time span and frequency of the data is from 2008 to 2012 on a monthly basis. The geographic coverage of the data includes four large countries for the Parties and two for the competitors. For each country and month, each record company provided data for its largest customers. The customers include download platforms, as well as mobile operators. Each company selected the 5000 most successful tracks of the whole period, and submitted data for these tracks.

(83) The key variables included in the datasets are sales units and wholesale revenue. The definition of these variables is similar to the same variables in the aggregate datasets.Tracks are identified by their song and track title and the name of the artist.In addition, there is information for each track regarding the date of its release or the date when it entered the underlying accounting system. Also, as in the case of the aggregate record company data, the track-level record company data have been augmented with a repertoire size variable from the platform data.

(84) Using these data, it has been possible to calculate further variables. For each country/customer/month triplet, a "digital chart" of the four companies is calculated by ranking the tracks by their units of sales. Also, the proportion of top 50 (or top 100 or top 200) songs in a given firm's repertoire can be calculated. The age of the tracks is calculated as the number of months since the track's release or its date of entry into the system. Finally, three wholesale price tiers can be identified based on the distribution of unit prices. These tiers are identified as below EUR […]*, between EUR […]* and EUR […]*, and above EUR […]*.

(85) The data have been aggregated to yearly level, keeping only the years from 2009 to 2011 to ensure consistency across companies. Only data related to Customer 1 and Customer 2 have been used subsequently for similar reasons as for the aggregate datasets, (i.e. namely that companies' data do not include advances and some discounts and that only for these customers/platforms the discrepancy in the data provided by record companies and by the customers/platforms is relatively small). In addition, some price outliers have been dropped from the sample (accounting for less than 0.5% of the observations, and less than 0.4% of the total revenue covered by the data).

3.2 Methodology

(86) The data have been analysed using linear regression methodology. Similarly to the platform data analysis, the goal of this regression analysis is to investigate the relationship between the size of the record company and its ability to get better wholesale terms from the platforms. The size of the company here is measured by the repertoire length variable, and the wholesale terms by the observed wholesale unit price.

1.3.2 Aggregate data

(87) In the base regressions of the aggregate data, the log of wholesale unit price is regressed on the log of the repertoire size, customer, country and time dummies, as well as on a constant. The regressions use the units as weights so as to give higher importance to observations corresponding to more popular music offerings. The regressions are run separately for album and track downloads.

(88) Several robustness checks have been carried out. First, the two different repertoire size variables are tried.Second, regressions for each year separately are run. Third, regressions for the four large countries representing […]*% of sales are run. Fourth, specifications with the lagged value of the repertoire size variable, as well as instrumental variables estimations of the regressions are tried. These latter two specifications attempt to control for potential endogeneity of the repertoire size variable. As the data is on a yearly level, the lags of the size variable are most probably predetermined, but they are still correlated with the current value of the variable. Hence, they have the potential for being valid instruments. Hence, in the instrumental variables estimations, the first and second lags of the repertoire size are used as instruments. Fifth, regressions of the wholesale price levels on the repertoire variable (as opposed to the above log-linear specifications) are run.

(89) The form of the log-linear regressions is the following:

(90) In this formulation, price is the wholesale price variable, repertoire is the repertoire size variable, X is the set of control variables, ε is the econometric error term, and cct denotes country/customer/period triplets.

(91) The level specifications are:

(92) Using the estimated model's coefficients it is possible to calculate the implied price effect of the Transaction. The implied price effect is the combined entity's price level, compared to Universal's and EMI's weighted average pre-merger price level. The model is able to give a prediction of the average wholesale price level of a record company as a function of the size of its repertoire. In particular, the repertoire size variable is changed for Universal and EMI to the combined repertoire size of the two companies. Then the estimated model implies new (higher) price levels, both for Universal and EMI. In particular, the implied price effect (in percentages) for the log-linear specifications is of the form:

(93) The implied price effect formula (%) for the level specifications is the following:

(94) The weights, w, are based on units sold. The formulas are evaluated at the averages of the estimation sample. The statistical uncertainty of the estimated price effects is calculated using the Delta-method.

2.3.2 Track-level data

(95) Similarly, in the base regressions of the track-level data the log of wholesale unit price is regressed on the log of the repertoire size, controlling for the log of the rank

615W. H. Greene (2005) Econometric Analysis, Prentice Hall, Upper Saddle River, New Jersey, pp. 913 and 914.

255

and age of the song, tier, customer and time dummies and a constant. The regressions are separately run for the two countries where data from all four companies is available. Again, the regressions use the units as weights so as to give higher importance to observations corresponding to more popular music offerings.

(96) The following robustness checks have been carried out. First, regressions using the two different repertoire size variables. Second, regressions by year. Third, specifications with the lagged size variable, and instrumental variables estimations (again, using the lags of the size variable as instruments). Fourth, specifications are tried where, in addition to the repertoire size variable, the log of the proportion of the top 50 songs in the firm's repertoire is added as a regressor. Fifth, regressions of the wholesale price levels on the repertoire variable (as opposed to the log-linear specifications) are run.

(97) The form of the estimation equations is similar to those of the aggregate data analysis. For the log-linear specifications the following equation is estimated:

(98) In this formulation, price is the wholesale price variable, repertoire is the repertoire size variable, X is the set of control variables, ε is the econometric error term, and ctt denotes customer/track/period triplets.

(99) For the level specifications the following equation is estimated:

(100) The implied price effect formulas and their calculations are the same as above (Points (92)-(94)).

3.3 Results

1.3.3 Aggregate data

(101) Appendix B shows the results from the regressions on the aggregate data. Table B-1 and Table B-2 show the results of the log-linear regressions on the aggregate data for tracks and albums, respectively. The estimated statistical relationship between the log wholesale price and the log repertoire size is positive and statistically

616Due to confidentiality reasons, the regression outputs are displayed in a restricted way. The tables report only the sign and significance level of the estimated main coefficients, as well as their corresponding p-values. The estimated fixed effects (controls) are not reported. The estimated price effects are reported only in ranges.

256

2.3.3 Track-level data

(102) Appendix C shows the results from the regressions on the track-level data. Table C-1 and Table C-3 report the track-level regression results for the two countries, respectively, and Table C-2 and Table C-4 report the corresponding yearly regressions. The estimated statistical relationship between the log wholesale price and the log repertoire size is positive and statistically significant, with some exceptions for the regressions of Country 2, where the main coefficient is not statistically significant. The same pattern is also true for the regressions using the platform data repertoire size variable (Table C-5-Table C-8), and the level specifications (Table C-9 and Table C-10). The fit of the regressions is strong with the exception of the IV estimations. These IV estimates, with one exception, also tend to be rejected by the instrument specification tests. It should be noted, however, that the OLS regressions using the lagged repertoire variables tend to give similar results to the simple OLS regressions. Given that endogeneity is less of a problem in the lagged specifications, the results of these regressions provide sufficient evidence for the robustness of the simple OLS results. The estimated price effects tend to be around […]*% for Country 1, and […]*% for Country 2.

257

(18) However, as explained by the Commission (Annex I, Section I, point 6), the margin variable it uses is not the economic margin described above in point (17). It is rather the ratio of the record company's total wholesale royalties to the retail revenues that were generated by its repertoire on a in a given country, given platform, in a given time period. That is, the margin variable used essentially measures the proportion of the total retail revenues generated by the repertoire which the platform has to pay to the record company. It is important to note that the production costs do not enter into this margin definition. This means that lower production costs do not lead ceteris paribus to higher margins. It follows that the positive relationship between margins and company size found by the Commission is not the result of differences in cost efficiencies across record companies.

(19) Similarly, the way the Commission defines margins and company size controls for quality differentials between the different record companies' repertoire.

635 Response to the Statement of Objections, paragraph 6.65. 636 Response to the Statement of Objections, Appendix 3, page 2. 637 Response to the Statement of Objections, paragraph 6.66 (a). 638 Response to the Statement of Objections, Appendix 2, Section 2, points 1-2, p. 3. 639 Response to the Statement of Objections, Appendix 3.

322

(20) The Commission uses the retail revenue shares as the measure of company size in a given country/platform/time period triplet. This retail share is essentially a quality adjusted measure of the record company's product line length. The quality adjustment here means that each company's repertoire size is weighted by its "popularity" as measured by the retail revenues it generates. Two repertoires with different sizes (i.e., repertoires with different number of songs) can generate the same amount of retail revenues and, hence, from the retailer's point of view, offer similar qualities if more of the smaller repertoire is sold . As such, the retail market share of a given record company can be thought of as the number of "quality units" the company offers the retailer.

640 (21) It follows that the margin variable, as defined by the Commission,measures the proportion of revenues the record company is able to extract from the retailer per quality units supplied. This means that the correct interpretation of the Commission's results is that larger companies are able to extract better terms per quality units from digital platforms. Hence, this empirically validated positive relationship is not a spurious one driven by quality or cost differentials across companies.

1.3.2. Higher margin may reflect a pure product mix effect due to differing price demand elasticity for songs

(22) The Notifying Party further argues that a higher margin may reflect a pure product mix effect and not a higher bargaining power. In particular, the Notifying Party claims that […]* the Commission's results are purely driven by a composition 641,642 effect.According to its reasoning, this is […]*. The Notifying Party argues that the size is a priori irrelevant to explain such differences in margins.

643 (23) The Commission notes […] *. […]*.

(24) Finally, it is also worth mentioning that the Commission in its track-level record company data analysis has explicitly taken into account the effect of the tier structure of the different record companies' repertoires (as well as their qualities) on the wholesale prices. The results of this analysis were found to be consistent with those of the platform data analysis. Also, the results of the empirical analysis are consistent with the results from the market investigation. In the market investigation platforms explained that they received worse terms from larger record music companies largely 644due to their larger size (market share/repertoire).

640 Annex I, Section I, point 6. 641 Response to the Statement of Objections, paragraph 6.66 (b)(i). 642 Response to the Statement of Objections, Appendix 4, Section 3.4, page 46. 643 […]*. 644 As explained in Section 7.1.5.5 Recital 527 of the Decision, the vast majority of digital customers (thirteen out of sixteen digital customers who expressed a view on the question whether the size of a recorded music company matters for its bargaining position and leverage in commercial negotiations with digital customers) indicated that the size of a recorded music company increases its bargaining power in relation to digital customers. If reference is made to the underlying repertoire of the record company this is done to indicate that such repertoire is able to generate significant revenues (and not that due to the nature of such repertoire digital customers make lower margins).

323

1.3.3. Higher margin may reflect a composition effect due to different level of margins extracted on album and songs

645 (25) […]* . […]*.

(26) […]*.

(27) Second, to assess the validity of the Notifying Party's claims, the Commission has checked these data and made the following findings.

(a)[…]*.

(b)After taking into account the effect of product mix, by introducing the share of albums in the repertoire as an additional explanatory variable into the Commission's the base regression of the platform, the effect of revenue shares on the margin is still positive, statistically significant and similar in magnitude. The effect of album shares is negative and statistically significant.

(c)The overall impact of the transaction on the margins is very similar, as shown in Section 2.5 of this Annex. This is because the magnitude of the impact on margins from the change in the product mix due to the merger is second order relative to the magnitude of the change in the revenue share. […]*As a conclusion, the overall impact of the album/track composition change is likely to be less than second order in magnitude and, hence, does not affect the conclusions of the Commission's analysis.

(d)[…]*. The Commission's results, however, show that the ability of larger record companies to extract better terms is present in all platforms (which confirms that if any such effect exists it is insignificant for the purposes of the current analysis).

1.3.4. The retail price applied by a platform may result from the strategic commercial decision of the platform.

(28) The Notifying Party further claims that by focussing on margins, the Commission is not able to decipher whether any resulting patterns in margins are driven by the record company's choice to set different wholesale prices, or the platform's decision 646,647 to set retail prices, or some combination thereof.According to this argument, the retail price applied by a platform may result from the strategic commercial decision of the platform. Further, the Notifying Party argues that it follows that a low margin may result from a platform promoting heavily certain songs, which can be said to attract customers on a website to sell them other products. This in turn would introduce an upward bias in the relationship between size and repertoires if larger, more hit driven, record companies have repertoire which is more likely to be used as "loss-leading" by the digital platforms.

645 […]*. 646 Response to the Statement of Objections, paragraph 6.66 (c). 647 Response to the Statement of Objections, Appendix 4, Section 3.4, pp. 46 and 47.

324

(29) To support these claims, the Notifying Party presents two graphs, both using data on 648 Universal's album sales […]*. The first shows the evolution of the average wholesale prices, margins and retailer mark-ups for Universal's top 10 selling albums. […]*. The second graph shows the margins and [digital platform]* mark-ups for Universal's top 10 selling albums and the other albums separately. […]*.

(30) The Commission notes, first, that the evidence provided by the Notifying Party does not support its claims. The data plotted on the two graphs […]*. It is also important 649 to note that the Notifying Party does not provide any systematic evidence […]* .

(31) Second, the Commission's results, showing a positive relationship between margins and size, hold across all platforms – […]*. Moreover, they also show that the estimated effect is stronger for smaller platforms (which are, overall, less likely to sell other products). This directly contradicts the claims of the Notifying Party.

650 (32) Third, the Commission notes that […]* .

(33) Finally, the Commission notes that in its track-level regressions it takes into account the effect of the chart position of tracks as well as the proportion of hits in each company's repertoire. These variables would explicitly control for observable repertoire characteristics that would proxy for the effects claimed by the Notifying Party. The results show a positive, significant relationship between wholesale prices and company (repertoire) size consistent with the results of the platform dataset.

1.3.5. Intrinsic characteristics of the catalogue may lead to the observed margins pattern.

(34) The Notifying Party also claims that some platforms would appear to apply a 651 different policy in which both retail and wholesale prices vary.According to this reasoning, a higher margin for a music provider (relative to others) can thus be explained either by a higher wholesale price or a lower retail price. Lower retail prices could be explained for instance by the characteristic of the catalogue (being more price elastic). The Notifying Party claims that if the size of a catalogue is correlated with such characteristics, one should explicitly take into account the effect of the characteristics of the catalogue. It argues that failing this, the Commission might wrongly conclude that margins are explained by the bargaining power that size confers when it is merely associated with intrinsic characteristics of the catalogue.

(35) The Commission notes, first, that the Notifying Party has failed to provide any evidence supporting its claims, let alone evidence showing that the alleged phenomena are such that they lead to a systematic bias in the Commission's analysis.

(36) Second, setting retail prices adjusted to price elasticities is a tool for the platform to maximize retail revenues. Consequently, the resulting retail revenues (and revenue shares) reflect these different characteristics of repertoires. Moreover, these same characteristics are also reflected by the wholesale royalties as the bargaining between

648 Response to the Statement of Objections, paragraph 6.66 (c), Figures 1-2, pp. 71 and 72. 649 […]*. 650 […]*. 651 Response to the Statement of Objections, Appendix 4, Section 3.4, p. 46.

325

the platform and the record company is about the split of the retail revenues, where the repertoires' different retail revenue generating abilities are taken into account. As explained above (Section 1.3.1), the Commission's margin variable (that is to say, the ratio of wholesale royalties and retail revenues) measures the proportion of revenues the record company is able to extract from the retailer per quality units supplied (through the effective royalties). Moreover, the Commission's size variable (that is to say, the retail revenue shares) is a quality adjusted measure of repertoire size or, in other words, a measure of the quality units that the record company's repertoire represents to the platform. It follows that the identified rent extraction (that is to say, the estimated positive relationship between margins and size) tackles the issue of different repertoire characteristics as it measures the rent extraction per quality units.

(37) Third, the Commission's results, showing a positive relationship between margins and size, hold across all platforms. […]*. The fact that there is evidence of a positive relationship between margins and size consistently across these different platforms directly contradicts the claims of the Notifying Party.

(38) Fourth, the Commission's track-level regressions explicitly control for the observable repertoire characteristics (such as chart position and age of tracks, as well as the proportion of hits in each company's repertoire). The results still show a positive, significant relationship between wholesale prices and company (repertoire) size.

1.4. The data collected by the Commission is problematic

(39) The Notifying Party claims that the platform data collected by the Commission have various defects as they do not include a number of key variables (split between albums versus tracks, ad-supported versus subscription streaming services, individual versus subscription download services). Moreover, there is insufficient transparency and control on how advances and minimum guarantees have been collected. The Notifying Party argues that advances and minimum guarantees are typically negotiated for […]* territories and time periods, and allocation of recent advances/guarantees is speculative as the recoupment period has not expired. Consequently, any allocation may be an over-allocation. The Notifying Party also claims that it has not been in a position to verify how each and every advance has been accounted for by each of the platforms. Accordingly, the Notifying Party claims that the data collected by the Commission provides an unsatisfactory empirical basis for an econometric study, whose results, as a consequence, should be disregarded as evidence.

652 Response to the Statement of Objections, paragraph 6.67-6.69.

(40) As an assessment of these claims, the Commission notes, first, that in any 653 quantitative analysis the first step is to formulate the relevant empirical question. In the present case this amounts to asking whether larger record companies are systematically able to extract better terms from digital platforms than smaller companies. This question had to be addressed for all the royalties received by the recorded music companies irrespective of whether they are related to tracks or

652 Response to the Statement of Objections, paragraph 6.67-6.69. 653 DG Competition, Best practices for the submission of the submission of economic evidence and data collection in cases concerning the application of Articles 101 and 102 TFEU and in merger cases, Section 2.1.

326

albums or are generated from ad-supported or subscription packages. The Commission requested the minimum amount of data sufficient to answer this 654 question.

654 Ibid, Section 2.2, paragraph 21: "it is necessary to identify the relevant facts to validate the theoretical assumptions and employ data which is appropriate to respond to the empirical question under investigation".

(41) Second, the Commission has explicitly requested the data providers (the platforms) to take into account minimum guarantees and advances. In particular, the Commission has indicated a methodology to allocate minimum guarantees and advances. After discussions with each of the data providers on the methodology to be implemented the Commission has relied on the knowledge of platforms to perform the allocation of such payments. In this context, the Commission considers that the platforms are likely to be the entities in the best position to carry out such an exercise of payment allocation. Indeed, unlike the Notifying Party, no platform mentioned to the Commission an inability to comply with the allocation of minimum guarantees and advances, in accordance with the methodology indicated by the Commission. Moreover, the reliability of the computation was certified by the platforms to the Commission via exchanges that are available to the Notifying Party. This is in line with the best practices of economic analysis and data work.Therefore, the Commission based its analysis on the dataset with the highest possible quality.

655 Ibid, Section 2.2, paragraph 22: "not all facts can be observed or measured with high accuracy and most datasets are incomplete or otherwise imperfect. Hence, parties and/or the Commission should become familiar with the facts and data and acknowledge its limitations explicitly. As regards quantitative data, for example, this requires (i) a thorough inspection of the data, including summary statistics and graphs, and (ii) a sufficient understanding of how the data were gathered, the sample selection process, the measurement of the variables and whether they bear a close relationship with their theoretical counterparts. Quantitative data may contain anomalies because of miscoding or other errors, which should be discussed with the data providers to decide how to best adjust the data to address these problems".

(42) Third, the Commission considers that the fact that advance and minimum guarantees are negotiated […]* does not affect the reliability of the data as the allocation methodology has been performed on the basis of the revenues generated […]* (for a given contract). Besides, in the yearly aggregated data any allocation error of the monthly data is likely to be cancelled out due to the aggregation. In this context, it is important to point out that the Commission's yearly results suggest that the identified effects are robust and support the overall conclusion of the analysis. Using these alternative methods as robustness checks is, again, consistent with the best practices of empirical work.

656 Ibid, Section 2.2, paragraph 23: "Failure to observe and validate all key assumptions or deficiencies in the data should not prevent an economic analysis to be given weight, though caution must be exercised before relying on its conclusions. Furthermore, statistical techniques have been developed to deal with measurement errors, missing observations and sample selection problems. While these techniques may not be able to improve the data, they may help to deal with some of its imperfections.".

327

(44) As a conclusion, the Commission finds that it collected a data of sufficient quality and made this transparent to the Parties. Hence, the data can be used as a basis of a reliable quantitative analysis.

1.5. Criticisms of the methodology used by the Commission in its econometric analysis

(45) The Notifying Party claims that the Commission's econometric results are erroneous as the analysis has only taken a limited number of variables into account in explaining margins, namely size and variables controlling for fixed differences between platforms, countries and over time, and has not taken into account 657 differences between recorded music companies unrelated to size.

(46) The Commission makes two observations. First, the Commission notes that in its track-level regressions it did take into account other record company specific variables. In particular, to control for repertoire quality, for each track the chart position, age (months since first sale) and price tier, as well as for each company the proportion of hits (top 50 songs) in its repertoire were included as explanatory variables. The reported results from these track-level regressions are consistent with those of the platform data analysis. Namely, they show a strong, statistically significant, positive relationship between repertoire size and wholesale terms.

(47) Second, in the platform analysis the variables capturing the wholesale terms and record company size (that is to say, the ratio of wholesale royalties to retail revenues, the "margin", and the retail revenue shares, respectively) have different definitions than those in the track-level data analysis (wholesale price and repertoire size). In particular, and as explained above in Section 1.3, the variables of the platform data capture more completely the rent extraction, and control for quality. Hence, there is less need for other variables to be included in the model. It is therefore plausible to assume that the effects captured by the platform data analysis are also valid and 658 economically meaningful.

1.6. Criticisms of the results of the Commission's quantitative analysis

1.6.1. Criticisms on the basis of regressions done at the record company level

(48) The Notifying Party claims, based on the platform data used by the Commission, that there is no positive relationship between margins and size, that is to say, the ability to extract rents from digital platforms is not correlated with the size of the record company. In particular, when restricting the analysis to separate regressions on each of the four major record companies' data, the Notifying Party claims that the results 659 show no positive relationship between margin and size.Moreover, it claims that

657 Response to the Statement of Objections, paragraph 6.70-6.71. 658 The Commission also notes the Notifying Party was not able to provide data split by genres in a consistent way. This implies that this additional variable might not be relevant for the negotiations and, hence, for the econometric analysis. 659 Response to the Statement of Objections, paragraph 6.72.

328

[…]* the estimated relationship is negative and statistically significant and 660 statistically non-significant […]*.

(49) To assess these claims, the Commission notes, first, that such company specific results are not in contradiction with the effect identified by the Commission. The specification suggested by the Notifying Party discards to a large extent the identification strategy proposed by the Commission which is primarily based on the cross-company variation in the data (see in more detail below in Sections 1.6.2 and 2.1.2), as now each regression relates to a different company. Second, the relationship identified by the Notifying Party captures a different type of variation in the company specific data in which statistically non-significant or even negative estimated coefficients are expected given the nature of the negotiations between record companies and platforms, as explained in the point (50).

(50) The monthly (major) record company specific datasets used by the Notifying Party are such that for each record company there is information on the margins and revenues shares on each country/platform/month triplets. The company specific regressions as proposed by the Notifying Party include country, platform and month fixed effects. This specification amounts to removing from the margins and revenue shares the specificities that are contingent on the countries, platforms and month (for a given record company). Introducing country/platform/month fixed effects accounts for the average different characteristics and market dynamics that are peculiar to, respectively, each country/platform/month. The variation in the margin left to be explained is thus provided by the deviations from these country/platform/month specific averages. In the company specific regressions proposed by the Notifying Party, this residual variation in the margins is to be explained by the residual variation in the revenue shares. The Commission, however, notes that […]*. This generates the negative relationship between margins and revenue shares.

(51) The Commission has tested whether the company specific results are robust under the yearly aggregation. The yearly aggregation is more representative of the 661 underlying dynamics […]*. […]*. All these results, however, do not imply that larger recorded music companies are not able to extract better terms from digital platforms than their smaller competitors.

1.6.2. Criticisms on the basis of the introduction of company fixed effects on the overall and by platform and by country regressions

(52) Further, the Notifying Party claims that had the Commission included company specific dummy variables (company fixed effects) in its base regressions it would have found that the greater size of the recorded music company does not increase its ability to extract rents from digital platforms.In addition, the Notifying Party claims that as these company fixed effects are jointly statistically significant there is 662 no basis for excluding them from the Commission’s base case model. Moreover, the Notifying Party claims that once the company fixed effects are included, there is no evidence of a positive relationship between the market share of a record company and its margins.Also, the estimated coefficients on the record company variables show that there is no relationship between average margins and average markets shares across the record companies.The Notifying Party also extends this analysis to regressions by platform, by country, and to models restricted on the four majors.

663 (53) To assess these claims, the Commission notices several aspects of the analysis. First, the Commission notes that the company fixed effects indeed provide explanatory power in terms of fitting the observed margins. This is consistent with the Wald tests 667 reported by the Notifying Party.However, the Commission observes that the inclusion of the company fixed effects into the base model (which already includes the revenues share as explanatory variable) does not result in a substantial improvement of the ability of the model to explain margins. This can be seen by comparing the adjusted R-squares of the Commission's and the Notifying Party's models.The comparison of the different specifications shows that the improvement in the fit due to the inclusion of the company fixed effects is at best very limited, as reflected by the very similar adjusted R-squares. Indeed, in the monthly specifications the adjusted R-square increases from 0.232 only to 0.249. Moreover, the improvement practically disappears when using the models on the yearly data. Here, the base model's adjusted R-square changes from 0.382 to 0.383 when introducing the company fixed effects. It should be noted that the yearly specifications better reflect the features of the market […]*.

(54) Second, when the company fixed effects are included in the model the estimated coefficient on the revenue share variable has different signs across the estimations 669 using different subsamples.This is also an indication that the company fixed effects and the revenue share variable capture the same variation of the margins and, hence, the coefficient on the revenue share is not properly identified when the company fixed effects are included.

663 Response to the Statement of Objections, Appendix 4, pp. 13, 15 and 16. 664 Ibid, pp. 13 and 15. 665 Ibid, pp. 13 and 17. 666 Ibid, pp. 13 and Section 3.1.3. 667 The Wald test is a statistical procedure to establish whether an estimated coefficient (or a set of estimated coefficients) is statistically (jointly) significantly different from a given value, e.g., zero. This is the methodology used by the economic consultants of the Notifying Party to establish the joint statistical significance of the estimated record company fixed effects. 668 The adjusted R-square is a measure of the goodness of the model to fit the data. Its possible values range from 0 to 1, with values close to 1 indicating an almost perfect fit and 0 an almost perfect lack of fit. In evaluating the goodness of the fit, the adjusted R-square also trades-off, on one hand, the improvement that the additional parameters bring to the fit of the model and, on the other hand, the resulting increased complexity derived from adding such additional parameters. This latter is particularly important in the case of the models of the Notifying Party which introduce company fixed effects. While the Commission's base model estimates 56 parameters, the Notifying Party's company fixed effects specification estimates 92 parameters, hence, almost doubling the number of parameters. This is a substantial increase in complexity which results in very similar adjusted R-squares. 669 Response to the Statement of Objections, Appendix 4: by platform (Table 4, p. 18), by country (Table 5, p. 20).

330

(55) These elements, namely the similarity in the goodness of fit between the base and extended models and the alternating signs of the revenue share variable in the 670 subsamples' estimation, indicate the existence of a multicollinearity problem. In other words, the set of company fixed effects, on the one hand, and the revenue shares, on the other hand, are competing for the same variation in the data. This means that they both are able to explain to a very similar extent the same variation in the margins. Thus, it is not surprising that the revenue share coefficient becomes insignificant once the record companies fixed effects are included in the model.

(56) Furthermore, one cannot conclude based on the statistical tests used by the Notifying Party that such lack of statistical significance implies a lack of relationship between the size of the company and the margins. Rather, the proper interpretation of this evidence is that the record company fixed effects and the revenue share are likely to be substitutes and thus able to explain the same underlying phenomena.

(57) It follows that the company fixed effects model provides an additional way to test the hypothesis that larger companies extract higher margins. This is because the estimated company fixed effects can be thought of as the average margin of each record company (once time, platform and country effects are filtered). Therefore, the test can be implemented by plotting the company fixed effects against the revenue shares. The Commission has undertaken this exercise and concluded that the estimated (average) margins show a positive relationship with the company size measure, that is, the overall revenue shares of the record companies (as explained in Section 2.1.2 of this Annex). The Commission has also tested the other specifications put forward by the Notifying Party, namely, the platform specific regressions. The results robustly show a positive relationship between company size and (average) 671 margin (as shown in Annex I).

(58)Finally, it should be noted that the Notifying Party's economic consultants have also attempted to plot the estimated company fixed effects (average margins) against the revenue shares, and based on this they have concluded that there is no positive relationship between the average shares and average margins. However, because of a methodological mistake their analysis is flawed and, hence, unreliable. In particular, the Notifying Party's economic consultants have computed the average market share of a given company by taking a simple, un-weighted average of the underlying market shares across different countries, platforms and months. This methodology biases the picture as it is essentially amounts to "adding apples and oranges" without properly taking into account their differences.

(59)The proper way to compute market shares in this context is to weight each market share in a given country/platform/time triplet according to the underlying total revenues. Hence, this calculation properly reflects the economic significance of each country/platform/time triplet and gives a correct representation of the relative size of each recording company. Unlike the erroneous average shares calculated by the Notifying Party's economic consultants, these properly calculated average shares do show a positive relationship with the estimated margins.

(60)To illustrate this methodological issue, the Commission notices that the revenue shares calculated by the Notifying Party's economic consultants substantially differ from the correct market shares. For example, for record company RC30 the market share goes from (the correct) 1.5% to 6.6%, which is a more than fourfold increase.Moreover, the Commission notes that the bias is systematic, with the major record companies' shares biased downward (for each of them the un-weighted market share is less than half of the correct figure) and those of the small companies biased upwards. This systematic bias greatly reduces the positive relationship between the estimated average margins (fixed effects) and the average shares. This is precisely because larger record companies have larger margins. The correlation of the incorrect, un-weighted shares with the average margins is just […]*, while for the proper shares it is […]*.It should be noted that this latter correlation is consistent with the results of the Commission's base model showing a positive relationship between record company size and its rent extraction ability. The Commission considers that this establishes the consistency between the Commission's base models and the company fixed effects models suggested by the Notifying Party.

1.6.3.The Commission's analysis of the record companies' data is affected by the same errors (in relation to fixed effects)

(61)The Notifying Party also claims that the Commission's analysis of the record companies' data is affected by the same errors as that of the platform data, and to an even greater extent.According to the Notifying Party, for both the aggregate and track-level regressions the estimated positive relationship between repertoire size and

(99)The Commission has used data from one major digital platform to assess whether there is a substantial difference in the composition of albums versus tracks for different record companies, and in particular for the majors, and secondly to perform the regression explicitly accounting for the different composition of albums versus tracks of each company.The results indicate a very small decrease in the implied margin difference when one accounts for this difference […]*.

The Commission only has such information on the split between albums and tracks for one platform.

342

Table 2-1 Regression results when introducing albums vs songs dummy

base

base with composition dummy

Estimates

+***

Revenues shares

-0.002

-0.009

-**

Composition dummy

0.016

Diagnostics

287

0.536

0.558

R-square

Implied margin difference

[…]*

[…]* […]*

[…]*

[…]* […]*

[…]*

[…]* […]*

[…]*

[…]* […]*

[…]*

2.6.The assumption of sub-additivity as a function of the platform's repertoire

(100)The Notifying Party also questions whether it is reasonable to assume that there are decreasing returns relative to the size of a platform's repertoire irrespective of the size of the platform's repertoire.According to this argument, it is common that platforms cannot be successful when they have access to the repertoire of only one recorded music company. If the decreasing returns assumption is relaxed the bargaining model no longer predicts that a merger will increase the merged entity's bargaining power and lead to higher prices for platforms. The Notifying Party argues that while the value of additional repertoire may fall when a platform has secured agreements with all majors, this is unlikely to be the case when the platform has access to the repertoire of only one or two of them. The Notifying Party therefore expects to find a super-additive area with respect to the repertoire size before sub-additivity arises. It further expects that the merger between two majors will result in a combined entitiy which is placed in in the super-additivity area the value function, and thus the merger will lead to lower prices.

Response to the Statement of Objections, Appendix 2, Section 3, pp. 5-7.

343

(101)First, the Commission notes that even the Notifying Party acknowledges that the value of additional repertoire may fall when a platform has reached a certain size of repertoire. Second, the Commission notes that its econometric analysis has shown the presence of the "greater extraction", or "size", effect. Therefore the only remaining issue raised is how one can make an out-of-sample prediction (that is to say, as the merged entity would enjoy higher market shares than reflected in the sample, it is possible that for these market shares, larger firms receive lower rents). The Commission acknowledges that there are limitations in any econometric analysis when making out-of-sample predictions. However, as detailed in Annex I of this Decision the Commission has shown that the size effect is currently present in the market. Furthermore, the evidence collected during the market investigation did not provide any support for the argument that if the merger were cleared the merged entity would request lower terms from the platforms. Also, the Commission notes that for EMI's repertoire this effect alleged by the Notifying Party is even less likely to arise as post-merger the EMI terms would become similar to Universal's – and, given the size effect identified by the Commission, would therefore become better from the merged entity's perspective.

2.7.The assumption of Cournot competition

(102)The Notifying Party also claims that the Commission would also overestimate the impact of the transaction as the Commission implicitly holds fixed the volumes of rival producers when undertaking its analysis. Restricting supply will be met by an expansion of supply from rivals and therefore the merging parties combined post-transaction share will likely shrink.To show this, the Notifying Party uses a baseline model, of homogeneous product quantity setting (Cournot); and claims that margin is equal to the ratio of revenue share to elasticity of demand. In the context of this base Cournot model, it also discusses the issue of endogeneity and the use of lagged dependent variables to solve this issue.

(103)First the Commission notes that the setting which is relevant is a bargaining setting and therefore the Cournot model predictions put forward by the Notifying Party are not applicable. In particular, it is well noted in economics that the reaction of rivals depends on the type of competition. While for Cournot competition rivals are expected to increase their output, in price competition (Bertrand) competitors would be expected to increase prices too (which, in the same spirit as the Notifying Party explains, would lead to the underestimation of any merger effect).In any event, the main point is that the baseline model put forward by the Notifying Party is not adapted for the case at hand.

Response to the Statement of Objections, Appendix 2, Section 4, pp. 6-11.

344

2.8.The use of revenues to measure margins

(104)The Notifying Party also argues that there is a significant risk in obtaining a "false positive" result since both the dependent (margin) and independent (revenue share) variables appear constructed from largely the same underlying platform revenue data.Moreover, any variable which affects both will introduce a correlation which may have nothing to do with market power.

(105)The Commission notes that the way the margin and revenue share variables are constructed does not lead to a "false positive", that is to say, a spurious positive correlation. The revenues of a given record company appear in the denominator of the margin and the numerator of the share variable. Purely in technical terms, this cannot give rise to a positive correlation (if anything, there is an inverse relationship stemming solely from these variable definitions). Hence, the size effect, that is to say, the positive relationship between margins and shares, as identified by the Commission, cannot be driven by the construction of these variables.

3.RECORD COMPANY DATA ANALYSIS – FURTHER POINTS

3.1.Aggregate record company data

3.1.1.Introduction of company fixed effects

(106)Similarly to the platform data analysis, the Notifying Party claims that had the Commission included company specific dummy variables (company fixed effects) in its regressions of the aggregate record company data it would have found that there is no positive relationship between the (repertoire) size and the wholesale price.The Notifying Party argues that the estimated coefficient on the repertoire size variable becomes statistically non-significant once company fixed effects are included in the model.

(107)The Commission notes that, similarly to the platform data case discussed in Section 1.6.3 of this Annex, the estimated company fixed effects can be interpreted as the average wholesale price of each record company (once time, platform and country effects are filtered). It follows that the company fixed effects model provides an additional way to test the hypothesis that larger companies set higher wholesale prices. Therefore, the test can be implemented by plotting the estimated company fixed effects against the revenue shares. The Commission has undertaken this exercise and concluded that the estimated average wholesale price shows a positive relationship with the repertoire size of the record companies (Figure 3-1 and Figure 3-2). The correlations between the average estimated wholesale price and the repertoire size are always positive, for tracks in the range of […]* and for albums in the range of […]*. This shows a strong, positive relationship between the estimated wholesale price and the repertoire size. It is also important to note that the Notifying Party has failed to carry out this analysis.

Response to the Statement of Objections, Appendix 2, Section 4, p. 12.

345

Figure 3-1 Estimated company fixed effects and repertoire size, aggregate record company data, tracks, CONFIDENTIAL

[…]*

Note: the left panels ("rec. comp.") use the record company data's repertoire size variable, the right panels ("platform") use that of the platform data.

Figure 3-2 Estimated company fixed effects and repertoire size, aggregate record company data, albums, CONFIDENTIAL

[…]*

Note: the left panels ("rec. comp.") use the record company data's repertoire size variable, the right panels ("platform") use that of the platform data.

3.1.2.Goodness of fit of regressions

(108)The Notifying Party also claims that the aggregate regressions are not capable of explaining the wholesale prices.According to this argument, there is a high degree of correlation between the wholesale prices and the regression residuals from the Commission's models (the unexplained variation in wholesale prices), and this shows that as the models are not capable of explaining the wholesale price, a significant portion of the wholesale price essentially becomes the residual. To substantiate this claim, the Notifying Party plots the wholesale prices against the residuals (replicated below in Figure 3-3 and Figure 3-4 for albums and tracks, respectively) and argues that there is a high degree of correlation between these two variables.

(109)The Commission notes that the Notifying Party's assessment of the aggregate models' goodness of fit is flawed as it fails to take into account that these models use weights to properly reflect the economic significance of each country/platform/year triplet of observations. For example, observations from a large market should be considered as more important than those from a small market whose size is just a fraction of that of the large one. Without weighting the regressions, these smaller markets would have been treated as being the same size as the large ones. Consequently, the resulting estimated relationship between the wholesale price and the repertoire size would have been driven by economically less important observations and would not have reflected the market conditions. It also follows from this that this weighting should also play a role when measuring a weighted regression's performance in terms of goodness of fit.

(110)Once the weighting is taken into account, it is clear that the models have a reasonably good explanatory power. This is illustrated by Figure 3-5 and Figure 3-6 which plot the weighted margins against the weighted regression residuals from the same models as in Figure 3-3 and Figure 3-4. As shown by these Figures, the correlations between the weighted variables are significantly weaker than those between the un-weighted variables, and there seems to be no systematic relationship or a much weaker systematic relationship between the price and the residual. This implies that the relationship between the unexplained part of the wholesale price (the residual) and the wholesale price is substantially weaker than what the Notifying Party claims. Hence, it is not true that "the wholesale price essentially becomes the residual". In other words, the models do have a reasonably good explanatory power once it is taken into account that the estimation deliberately attempts to explain certain groups of observations (those with higher weights) better than others. The reason for this discrimination is precisely the higher economic importance of the observations with higher weights.

(111)The Commission also notes that in assessing the goodness of fit of weighted regressions it reports the standard diagnostic to (R-squares) which are adapted to reflect the underlying weighting. On the other hand, analysing the simple, un-weighted correlations, as the Notifying Party does, is flawed as these diagnostic statistics were not designed to properly take into account the nature of the estimated weighted regressions (they are more suitable for assessing un-weighted regressions).

Figure 3-3 Wholesale prices and regression residuals, un-weighted, albums (correlations in parentheses)

Source: CRA calculations replicated by the Commission; excluding observations with log price below -1.

347

Figure 3-4 Wholesale prices and regression residuals, un-weighted, tracks (correlations in parentheses)

Source: CRA calculations replicated by the Commission; excluding observations with log price below -1.

Figure 3-5 Wholesale prices and regression residuals, weighted, albums (correlations in parentheses)

Source: Commission calculations; excluding observations with log price below -1.

348

Figure 3-6 Wholesale prices and regression residuals, weighted, tracks (correlations in parentheses)

Source: Commission calculations; excluding observations with log price below -1.

3.1.3.Wholesale price versus repertoire size

(112)The Notifying Party further claims that there is no relationship between the wholesale price and repertoire size at all.To substantiate this claim, the Notifying Party simply plots the wholesale price against the repertoire size variables for tracks and albums (replicated below in the left panels of Figure 3-7, Figure 3-8 and Figure 3-9, Figure 3-10 for albums and tracks, respectively) and argues that "[t]here is absolutely no pattern whatsoever between prices and the repertoire size".

(113)The Commission notes that the Notifying Party's analysis is flawed. The graphs used by the Notifying Party (reproduced as the left panels of Figure 3-7-Figure 3-10) pool together observations from different countries/platforms/years without taking into account the county/platform/year specific context of each of them. It may be the case that within each county/platform/year triplet there is a positive relationship between the wholesale price and the repertoire size, but this cannot be seen if the observations are put together. This is because the average level of prices (and to some extent even that of the repertoire size) can be different across the different triplets.

(114)The Commission has tested whether this is the case by simply substracting from the wholesale price and repertoire size variables their respective county/platform/year specific averages. These "filtered" observations are plotted at the right panels of each figure.The results show very clearly that within the county/platform/year specific triplets there is always a strong, statistically significant and positive correlation. This correlation is always stronger than in the case of the "unfiltered" observations. For the latter, in the case of albums the correlations are not even statistically significant. This is a clear indication that the pooled graphs, advocated by the Notifying Party, hide the relevant tendencies in the data.

Technically, these observations are the residuals of regressions of the wholesale price and the repertoire size on country, platform and year dummy variables.

349

triplets there is always a strong, statistically significant and positive correlation. This correlation is always stronger than in the case of the "unfiltered" observations. For the latter, in the case of albums the correlations are not even statistically significant. This is a clear indication that the pooled graphs, advocated by the Notifying Party, hide the relevant tendencies in the data.

(115)In other words, the analysis put forward by the Notifying Party fails to take into account that the relevant variation in the price data (that is to say, the price deviations from the country, platform and year specific averages) has a positive and statistically significant relationship with the repertoire size variables. It is also interesting to note that the Commission, by including country-, platform- and year specific fixed effects in its base regressions, used the "filtered" observations to identify the size effect. Hence, the right panels in the figures are also consistent with the Commission's regression results.

[…]*

Note: left panel: replication of CRA's graph; 'unfiltered': original price and repertoire variables, 'filtered': residuals of a regression of the original variables on country/customer/time fixed effects; dotted line: best linear fit of data points; p-values of correlations in parentheses.

350

Figure C-33 Base with company fixed effects and without revenue share

[...]*

Figure C-34 Base with company fixed effects and without revenue share (restricted)

[...]*

Figure C-35 Base with company fixed effects and revenues shares interacted

[...]*

Figure C-36 Base with company fixed effects and revenues shares interacted (restricted)

[...]*

PLATFORM 2 - LOGS

Figure C-37 Base in Logs with fixed effects

[...]*

Figure C-38 Base in Logs with fixed effects (restricted)

[...]*

Figure C-39 Base in Logs with fixed effects without revenue shares

[...]*

Figure C-40 Base in Logs with fixed effects without revenues shares (restricted)

[...]*

Figure C-41 Base in Logs with company fixed effects and revenues shares interacted

[...]*

372

Figure C-42 Base in Logs with company fixed effects and revenues shares interacted (restricted)

[...]*

PLATFORM 3 –LEVELS

Figure C-43 Base with company fixed effects

[...]*

Figure C-44 Base with company fixed effects (restricted)

[...]*

Figure C-45 Base with company fixed effects and without revenue share

[...]*

Figure C-46 Base with company fixed effects and without revenue share (restricted)

[...]*

Figure C-47 Base with company fixed effects and revenues shares interacted

[...]*

Figure C-48 Base with company fixed effects and revenues shares interacted (restricted)

[...]*

PLATFORM 3 –LOGS

Figure C-49 Base in Logs with fixed effects

[...]*

Figure C-50 Base in Logs with fixed effects (restricted)

[...]*

Figure C-51 Base in Logs with fixed effects without revenue shares

[...]*

Figure C-52 Base in Logs with fixed effects without revenues shares (restricted)

[...]*

Figure C-53 Base in Logs with company fixed effects and revenues shares interacted

[...]*

Figure C-54 Base in Logs with company fixed effects and revenues shares interacted (restricted)

[...]*

373

PLATFORM 4 -LEVELS

Figure C-55 Base with company fixed effects

[...]*

Figure C-56 Base with company fixed effects (restricted)

[...]*

Figure C-57 Base with company fixed effects and without revenue share

[...]*

Figure C-58 Base with company fixed effects and without revenue share (restricted)

[...]*

Figure C-59 Base with company fixed effects and revenues shares interacted

[...]*

Figure C-60 Base with company fixed effects and revenues shares interacted (restricted)

[...]*

PLATFORM 4 -LOGS

Figure C-61 Base in Logs with fixed effects

[...]*

Figure C-62 Base in Logs with fixed effects (restricted)

[...]*

Figure C-63 Base in Logs with fixed effects without revenue shares

[...]*

Figure C-64 Base in Logs with effects without revenues shares (restricted)

[...]*

Figure C-65 Base in Logs with company fixed effects and revenues shares interacted

[...]*

374

Figure C-66 Base in Logs with company fixed effects and revenues shares interacted (restricted)

[...]*

PLATFORM 5 -LEVELS

Figure C-67 Base with company fixed effects

[...]*

Figure C-68 Base with company fixed effects (restricted)

[...]*

Figure C-69 Base with company fixed effects and without revenue share

[...]*

Figure C-70 Base with company fixed effects and without revenue share (restricted)

[...]*

Figure C-71 Base with company fixed effects and revenues shares interacted

[...]*

Figure C-72 Base with company fixed effects and revenues shares interacted (restricted)

[...]*

PLATFORM 5 -LOGS

Figure C-73 Base in Logs with fixed effects

[...]*

Figure C-74 Base in Logs with fixed effects (restricted)

[...]*

Figure C-75 Base in Logs with fixed effects without revenue shares

[...]*

Figure C-76 Base in Logs with effects without revenues shares(restricted)

[…]*

Figure C-77 Base in Logs with company fixed effects and revenues shares interacted

[…]*

Figure C-78 Base in Logs with company fixed effects and revenues shares interacted (restricted)

[...]*

375

PLATFORM 6 - LEVELS

Figure C-79 Base with company fixed effects

[...]*

Figure C-80 Base with company fixed effects (restricted)

[...]*

Figure C-81 Base with company fixed effects and without revenue share

[...]*

Figure C-82 Base with company fixed effects and without revenue share (restricted)

[...]*

Figure C-83 Base with company fixed effects and revenues shares interacted

[...]*

Figure C-84 Base with company fixed effects and revenues shares interacted (restricted)

[...]*

PLATFORM 6 -LOGS

Figure C-85 Base in Logs with fixed effects

[...]*

Figure C-86 Base in Logs with fixed effects (restricted)

[...]*

Figure C-87 Base in Logs with fixed effects without revenue shares

[...]*

Figure C-88 Base in Logs with effects without revenues shares (restricted)

[…]*

Figure C-89 Base in Logs with company fixed effects and revenues shares interacted

[…]*

Figure C-90 Base in Logs with company fixed effects and revenues shares interacted (restricted)

[…]*

376

COMMITMENTS TO THE EUROPEAN COMMISSION

Pursuant to Article 8(2) of Council Regulation (EC) No 139/2004 as amended (the “Merger

Regulation”), Universal Music Group (“UMG”) hereby provides the following Commitments (each a

“Commitment”, together the “Commitments”) in order to enable the European Commission (the

“Commission”) to declare the acquisition by UMG of EMI Group Global Limited’s (“EMI”)

recorded music business, (the “Transaction”) compatible with the internal market and the EEA

Agreement by its decision pursuant to Article 8(2) of the EC Merger Regulation (the “Decision”).

The Commitments shall take effect upon the date of adoption of the Decision.

This text shall be interpreted in the light of the Decision to the extent that the Commitments are

attached as conditions and obligations, in the general framework of EU law, in particular in the light

of the Merger Regulation, and by reference to the Commission Notice on remedies acceptable under

Council Regulation (EC) No 139/2004 and under Commission Regulation (EC) No 802/2004.

Section A. Definitions

For the purpose of the Commitments the following terms shall have the following meaning:

Affiliated Undertakings: undertakings controlled by UMG and/or by the ultimate parents of UMG

(i.e. Vivendi S.A.), whereby the notion of control shall be interpreted pursuant to Article 3 of the

Merger Regulation and in light of the Commission Consolidated Jurisdictional Notice under Council

Regulation (EC) No 139/2004 on the control of concentrations between undertakings.

Closing: the transfer of the legal title of all or each of the Divestment Assets to one or more

Purchasers.

Divestment Business: each of the (i) Legal Entities; (ii) EMI’s share in NOW; and (iii) UMG's share

in Jazzland as defined in Section B and Schedules 1 to 3 of these Commitments that UMG commits to

divest (each of the Legal Entities, EMI’s share in NOW, and UMG's share in Jazzland and is referred

to as a “Divestment Asset”).

Digital Music Service Company: a company operating in the EEA and providing an online or mobile

music service offering a catalogue of music from a multitude of artists (including but not limited to à-

la-carte download stores and subscription, streaming services (free to user and paid for)). For the

avoidance of doubt, short term premium music offers associated with brands, marketing partnerships

and services using music purely as a subsidiary to the purpose of the overall service e.g. games

services incorporating music, shall not fall within the definition of Digital Music Service Company.

Divestiture Trustee: one or more natural or legal person(s), independent from UMG, who is

approved by the Commission and appointed by UMG and who has received from UMG the exclusive

Trustee Mandate to sell the Divestment Business to one or more Purchasers at no minimum price.

Effective Date: the date of the adoption of the Decision.

First Divestiture Period: the period of […]* from the Effective Date.

Hold Separate Manager: the person(s) appointed by UMG to manage the day-to-day business of the

Divestment Business under the supervision of the Monitoring Trustee.

Key Personnel: all personnel necessary to maintain the viability and competitiveness of the

Divestment Business, as listed in Schedule 1.

377

Legal Entities: the (i) following companies currently wholly owned and under the sole control of

either EMI or UMG some of which are largely operated as stand-alone local companies (namely EMI

Belgium; EMI Czech Republic; EMI Denmark; EMI France; EMI Poland; EMI Portugal; EMI Spain;

EMI Sweden; EMI Norway; and UMG Greece) and some of which are not organisationally operated

on a stand-alone basis and whose support functions are (at least to some extent) provided by EMI’s

and UMG’s respective central or regional group functions, namely: Chrysalis Records Limited; Co-

Op; EMI Records Limited (“EMIRL”); Ensign Limited; Mute Records Limited; King Island

Roxystars Recordings AB; Sanctuary Records Group Limited; and (ii) the newly formed legal entity

to which the MPS Records catalogue will be transferred, each a “Legal Entity” whose assets that are

being divested are identified and described in Schedule 1 of these Commitments, with EMI Belgium;

EMI Denmark, EMI Czech Republic; EMI France; EMI Poland; EMI Portugal; EMI Spain; EMI

Sweden; EMI Norway, Chrysalis Records Limited; EMIRL; Ensign Limited; and Mute Records

Limited being hereafter referred to as the “EMI Legal Entities” whereas as the remaining legal

entities are referred to as the “UMG Legal Entities”.

Local Opcos: each of the following companies, namely EMI Belgium; EMI Czech Republic; EMI

Denmark; EMI France; EMI Poland; EMI Portugal; EMI Spain; EMI Sweden; EMI Norway; and

UMG Greece as described and specified in Schedule 1 of these Commitments.

Main Package: EMIRL and the Local Opcos as described and specified in Schedule 1 of these

Commitments.

Monitoring Trustee: one or more natural or legal person(s), independent from UMG, who is

approved by the Commission and appointed by UMG, and who has the duty to monitor UMG’s

compliance with the conditions and obligations attached to the Decision.

Personnel: the personnel currently employed by EMI or any UMG Legal Entities, including Key

Personnel that are working for each Divestment Asset (including staff on secondment and shared

personnel) as specified in Schedule 1.

Purchaser: with regard to each Divestment Asset the entity or entities approved by the Commission

as acquirer in accordance with the criteria set out in Section E.

Residual Classics Artists: Virgin Classics and EMI Classics branded artists in the EEA which are not

currently signed to any of the Legal Entities and who will be transferred to the Legal Entities as

follows: Virgin Classics branded artists to EMI France or EMI Records Limited; EMI Classics

branded artists to EMI Records Limited.

Third Party Undertakings: undertakings which are not controlled by UMG and/or by the ultimate

parents of UMG whereby the notion of control shall be interpreted pursuant to Article 3 of the EC

Merger Regulation and in light of the Commission Consolidated Jurisdictional Notice under Council

Regulation (EC) No 139/2004 on the control of concentrations between undertakings.

Trustee(s): The Monitoring Trustee and/or the Divestiture Trustee (as the case may be).

Trustee Divestiture Period: the period of […]* from the end of the First Divestiture Period.

UMG: Universal Music Group, incorporated under the laws of Delaware, with its registered office at

1755 Broadway, New York, NY10019 and registered with the State of Delaware Division of

Corporations. under number 2961379.

Vivendi S.A.: a French Société Anonyme whose principal place of business is at 42 avenue de

Friedland, 75380 Paris, France

Section B. The Divestment Business

378

Commitment to divest

In order to restore effective competition, unless otherwise agreed with the Commission, UMG

commits by the end of the Trustee Divestiture Period to divest, or procure the divestiture of the Main Packageto one Purchaser with an option to sell some or all of the remaining Divestment Assets

either also to that Purchaser or to one or more other Purchasers on terms of sale approved by the

Commission and in accordance with the procedure and terms described in paragraph 20, i.e. with the

totality of the Divestment Business comprising:

(a)the Legal Entities (subject to the exclusions particularised in Schedule 1), including the

Residual Classics Artists (subject to paragraph 6(d) of these Commitments);

(b)EMI’s 50% share in the NOW joint venture (together with the name and associated rights

to the NOW brand/trademark within the EEA) as described in Schedule 2 (the “NOW Joint Venture”). UMG also commits for a period of ten years after the Effective Date, to

continue to license its repertoire to the NOW Joint Venture on terms equivalent to its

existing contractual arrangements with the NOW Joint Venture (the “NOW Licence Commitment”); and

(c)UMG’s share in Jazzland.

To carry out the divestiture of the Divestment Business, UMG commits to enter into one or more (as the case may be) final binding sale and purchase agreement(s) for the sale of the Divestment Assets within the First Divestiture Period. If UMG has not entered into such an agreement(s) at the end of the First Divestiture Period, UMG shall grant the Divestiture Trustee an exclusive mandate to sell the Divestment Assets to one or more purchasers in accordance with the procedure described in paragraph 29.

In order to restore effective competition, UMG also commits to terminate the exclusive sales and distribution agreement between Universal Music Operations Limited with the UK Indie record label Ministry of Sound Recordings Limited (“Ministry of Sound”) dated […]* (“MoS Agreement”) pursuant to the […]* notice period in the MoS Agreement (or such longer period as possibly requested by Ministry of Sound and agreed with the Commission) and not to enter into new distribution and/or licence agreements for any rights covered by the MoS Agreement for a period of […]*.

UMG shall be deemed to have complied with the Commitment to divest set out at paragraph 1 and 2 above if by the end of the Trustee Divestiture Period UMG or an Affiliated Undertaking:

(a)has entered into a final binding sale and purchase agreement(s), if the Commission approves the Purchaser(s) and the terms in accordance with the procedure described in paragraph 20 and if the closing of the sale of the Divestment Business takes place within a period not exceeding […]* after the approval of the purchaser (or all purchasers, as the case may be) and the terms of sale by the Commission; and

(b)has terminated the MoS Agreement.

In order to maintain the structural effect of the Commitments, UMG shall, for a period of 10 years after the Effective Date, not acquire direct or indirect influence over the whole or part of the Divestment Business, including by soliciting, re-signing, or otherwise acquiring an interest in artists or bands (as the case may be) forming part of the Divestment Business in order for them to enter into agreements with UMG for the exploitation of their recorded music rights, unless the Commission has previously found that the structure of the market has changed to such an extent that the absence of influence over the Divestment Business (or any part thereof) is no longer necessary to render the Transaction compatible with the internal market. For the avoidance of doubt, this provision shall not apply to any artists or bands merely on the basis that they have featured on any compilations (such as the NOW compilations) which form part of the Divestment Business. UMG shall also for a period of […]* not renew the MoS Agreement or enter into any new licence and/or distribution agreement for any rights covered by the MoS Agreement, unless the Commission has previously found that the structure of the market has changed to such an extent that the absence of UMG’s participation in an agreement akin to the MoS Agreement is no longer necessary to render the Transaction compatible with the internal market.

Structure and definition of the Divestment Business

The Divestment Business includes Anglo-American and domestic repertoire (both catalogue and front-line) associated with labels and held within the Legal Entities. The present legal and functional structure of the Divestment Business as operated to date is described in the Schedules 1 to 3. The Divestment Business, described in more detail in the Schedules, includes:

(a)all tangible and intangible assets (including intellectual property rights), which contribute to the current operation or are necessary to ensure the viability and competitiveness of the Legal Entities. For the avoidance of doubt, UMG shall retain all rights to the EMI and Virgin names (subject only to a transitional licence to the Purchaser(s) of the Main Package for the use in relation to certain pre-existing stock, and for […]* from Closing the EMI name) and UMG shall retain a perpetual licence back for the use of the Parlophone name in respect of the Beatles and former Beatles (as specified in Schedule 1).

(b)insofar as relevant all licences, permits and authorisations issued by any governmental organisation for the benefit of the Divestment Business;

(c)all contracts, leases, commitments and customer orders of the Divestment Business; all customer, credit and other records of the Divestment Business (items referred to under (a)-(c) hereinafter collectively referred to as “Assets”);

(d)the Key Personnel identified in the Schedule 1 to these Commitments;

(e)[…]*; and

(f)at the option of the Purchaser(s) and if determined necessary by the Trustee and required by the Commission for a transitional period of up to […]* after Closing (or such other period as agreed to allow the Purchaser(s) to establish all or any of the Divestment Assets as a viable and independent business) and on terms agreed on an arm’s length basis, the provision of support services by UMG or Affiliated Undertakings to any of the Divestment Assets. This would include continuing with the inclusion of the repertoire held in the Divestment Assets in any existing licence or distribution agreements with an appropriate allocation of revenues. The scope and nature of these services will be tailored to meet the

reasonable requirements of the Purchaser(s) and to be agreed between UMG and the Purchaser(s).

Section C. Related Commitments

Preservation of Viability, Marketability and Competitiveness

For UMG Legal Entities and UMG's shares in Jazzland from the Effective Date until Closing and for EMI Legal Entities and EMI’s share in the NOW Joint Venture from the closing of the Transaction until Closing, UMG shall preserve the economic viability, marketability and competitiveness of the Divestment Business, in accordance with good business practice, and shall minimise as far as possible any risk of loss of competitive potential of the Divestment Business. In particular UMG undertakes:

(a)not to carry out any act upon its own authority that might have a significant adverse impact on the value, management or competitiveness of the Divestment Business or that might alter the nature and scope of activity, or the industrial or commercial strategy or the investment policy of the Divestment Business;

(b)to make available sufficient resources for the development of the Divestment Business, on the basis and continuation of the existing business plans;

(c)to take all reasonable steps, including appropriate incentive schemes (based on industry practice) to encourage all Key Personnel to remain with the Divestment Business; and

(d)whereas UMG has minimised as far as possible any potential legal obstacles to the transferability of rights, to the extent that in some exceptional cases artist consent is required to transfer an artist with the Legal Entity to which that artist is primarily attached, […]* and not to re-sign such artist(s) for a period of 10 years unless the Commission has previously found that the structure of the market has changed to such an extent that it is no longer necessary to preclude UMG from signing such artist(s) to render the Transaction compatible with the internal market.

Hold-separate obligations

UMG commits, from the Effective Date until the date of closing of the Transaction, to keep EMI in its entirety separate from the UMG business and to have no involvement in any of EMI’s businesses. UMG shall assist the Monitoring Trustee in monitoring UMG's compliance with this obligation.

UMG commits, for UMG Legal Entities and UMG's shares in Jazzland from the Effective Date until Closing and for EMI Legal Entities and EMI’s share in the NOW Joint Venture from the closing of the Transaction until Closing , to keep these Divestment Assets separate from the businesses it is retaining (the "Retained Business”) and to ensure that Key Personnel of these Divestment Assets– including the Hold Separate Manager(s) – have no involvement in any business retained and vice versa (to the extent this is reasonably practicable and does not adversely affect the viability of any part of the Divestment Business or the Retained Business in which case appropriate safeguards shall be put in place to protect confidential information).

UMG shall also ensure that the Key Personnel of these Divestment Assets do not report to any individual outside these Divestment Assets. For the avoidance of doubt, this Commitment does not prohibit:

(a)under supervision from the Monitoring Trustee, communications between on the one hand these Divestment Assets (and their Personnel)and on the other hand UMG and the Retained Business (and vice versa) which are necessary in order to enable UMG or the Retained

Business to continue to provide where necessary general corporate, commercial, financial governance, and HR support to these Divestment Assets (and vice versa) during the First

Divestiture and (insofar as relevant) Trustee Divestiture Periods;

(b)under supervision from the Monitoring Trustee, the taking of any steps, including all

related communications between on the one hand these Divestment Assets and on the other

hand UMG and the Retained Business (and vice versa), which are necessary in order to

enable UMG or the Retained Business to take the benefit of and comply with contractual

obligations under existing licensing and distribution agreements; and

(c)under supervision from the Monitoring Trustee, communications between the Personnel

and UMG and/or the Retained Business which arise from and in the exercise of functions

performed by the Personnel which do not relate to these Divestment Assets (and vice versa)

For UMG Legal Entities and UMG's shares in Jazzland from the Effective Date until Closing and for

EMI Legal Entities and EMI’s share in the NOW Joint Venture from the closing of the Transaction

until Closing, UMG shall assist the Monitoring Trustee in ensuring that these Divestment Assets are

managed as a distinct and saleable entity (or entities) separate from the Retained Business. UMG shall

appoint one or more Hold Separate Managers who shall be responsible for the management of these

Divestment Assets, under the supervision of the Monitoring Trustee. The Hold Separate Manager(s)

shall manage these Divestment Assets independently and in the best interest of these businesses with a

view to ensuring their continued economic viability, marketability and competitiveness and their

independence from the Retained Business.

10 For UMG Legal Entities and UMG's shares in Jazzland from the Effective Date until Closing and for

EMI Legal Entities and EMI’s share in the NOW Joint Venture from the closing of the Transaction

until Closing, to ensure that these Divestment Assets are held and managed as a separate entity

(entities) the Monitoring Trustee shall exercise UMG’s rights as shareholder in any relevant Legal

Entities (except for its rights for dividends that are due before Closing), with the aim of acting in the

best interest of the business, determined on a stand-alone basis, as an independent financial investor,

and with a view to fulfilling UMG’s obligations under the Commitments. Furthermore, the

Monitoring Trustee shall have the power to replace members of the supervisory board or non-

executive directors of the board of directors, who have been appointed on behalf of UMG or EMI (as

the case may be). Upon request of the Monitoring Trustee, UMG or EMI (as the case may be) shall

resign as member of the boards or shall cause such members of the boards to resign.

Ring-fencing

11 Without prejudice to paragraph 8, UMG shall implement all necessary measures to ensure that it does

not after the Effective Date obtain any business secrets, know-how, commercial information, or any

other information of a confidential or proprietary nature relating to the Divestment Business. In

particular, the participation where relevant of the Divestment Business in a central information

technology network shall be severed to the extent reasonably practicable following the closing of the

Transaction, without compromising the viability of any of the Divestment Business or the Retained

Business.

12 UMG may obtain information relating to EMI which is reasonably necessary with a view to the

closing of the Transaction and the subsequent integration of the Retained Business into UMG and,

subsequently, information relating to the Divestment Business which is reasonably necessary for the

divestiture of the Divestment Business or whose disclosure to UMG is required by law. UMG may

also obtain information relating to the Divestment Business pursuant to paragraph 7(a), 7(b) and 7(c)

above as well as information required to properly implement the hold-separate, in each case under the

supervision of the Monitoring Trustee.

Non-solicitation clause

13 UMG undertakes, subject to customary limitations, not to solicit, and to procure that Affiliated

Undertakings do not solicit, the Key Personnel transferred with the Divestment Business for a period

of […]* after Closing.

Due Diligence

14 In order to enable potential purchasers to carry out a reasonable due diligence of the Divestment

Business, UMG shall, subject to customary confidentiality assurances and dependent on the stage of

the divestiture process:

(a)provide to potential purchasers sufficient information as regards the Divestment Business;

and

(b)provide to potential purchasers sufficient information relating to the Personnel and allow

them reasonable access to the Personnel.

Reporting

15 UMG shall submit written reports in English on potential purchasers of the Divestment Business

and developments in the negotiations with such potential purchasers to the Commission and the

Monitoring Trustee no later than 10 days after the end of every month following the Effective Date (or

otherwise at the Commission’s request).

16 UMG shall inform the Commission and the Monitoring Trustee on the preparation of the data room

documentation and the due diligence procedure and shall submit a copy of an information

memorandum to the Commission and the Monitoring Trustee before sending the memorandum out to

potential purchasers.

Section D. Behavioural Commitment in relation to Most Favoured Nation Clauses

17 As from the Effective Date, UMG commit as regards Digital Music Service Companies not to enter

into any agreement which insofar as it applies to the EEA (or parts thereof) provides that such

companies shall guarantee to Universal commercial terms that are better, equivalent to or at least as

good as those agreed with another recorded music company (the "MFN Commitment").

18 UMG commits to act in compliance with the MFN Commitment for a period of 10 years from the

Effective Date unless the Commission has previously found that the structure of the market has

changed to such an extent that it is no longer necessary to comply with this Commitment.

Section E. The Purchaser

19 In order to ensure the immediate restoration of effective competition, the Purchaser(s), in order

to be approved by the Commission, must:

(a)be independent of and unconnected to UMG;

(b)have the financial resources, proven expertise and incentive to maintain and develop the

Divestment Business as a viable and active competitive force in competition with UMG

and other competitors;

(c)neither be likely to create, in the light of the information available to the Commission,

prima facie competition concerns nor give rise to a risk that the implementation of the

Commitments will be delayed, and must, in particular, reasonably be expected to obtain all

necessary approvals from the relevant regulatory authorities for the acquisition of the

Divestment Business (; and

(d)be a company in a group which is either currently or was previously active in the recorded

music or music publishing business(the before-mentioned criteria under (a) to and

including (d) for the purchaser(s) hereafter the “Purchaser Requirements”).

20 Unless otherwise agreed with the Commission, UMG commits to selling at least the Main Package to

a single Purchaser with an option to selling some or all of the remaining Divestment Assets either also

to that Purchaser or to one or more other Purchasers. The final binding sale and purchase

agreement shall be conditional on the Commission’s approval. When UMG has reached an

agreement with one or more purchasers, it shall submit a fully documented and reasoned

proposal, including a copy of the final agreement(s), to the Commission and the Monitoring

Trustee. UMG must be able to demonstrate to the Commission that the purchaser(s) meets the

Purchaser Requirements and that the Divestment Business is being sold in a manner consistent

with the Commitments. For the approval, the Commission shall verify that the purchaser fulfils

the Purchaser Requirements and that the Divestment Business is being sold in a manner

consistent with the Commitments. The Commission may approve the sale of the Divestment

Assets as a whole to a single purchaser or may approve the sale of different parts of the

Divestment Business to different purchasers. The Commission may also approve the sale of the

Divestment Business without one or more Assets or with a replacement asset or parts of the

Personnel, if this does not affect the viability and competitiveness of the Divestment Business after

the sale, taking account of the proposed purchaser(s).

Section F. Trustee

I Appointment Procedure

21 UMG shall appoint a Monitoring Trustee to carry out the function specified in the Commitments for a

Monitoring Trustee. If UMG has not entered into a binding sales and purchase agreement one month

before the end of the First Divestiture Period or if the Commission has rejected a purchaser proposed

by UMG at that time or thereafter, UMG shall appoint a Divestiture Trustee to carry out the functions

specified in the Commitments for a Divestiture Trustee. The appointment of the Divestiture Trustee

shall take effect upon the commencement of the Extended Divestment Period.

22 The Trustee shall be independent of UMG, possess the necessary qualifications to carry out its

mandate, for example as an investment bank or consultant or auditor, and shall neither have

nor become exposed to a conflict of interest. The Trustee shall be remunerated by UMG in a

way that does not impede the independent and effective fulfilment of its mandate. In particular,

where the remuneration package of a Divestiture Trustee includes a success premium linked to

the final sale value of the Divestment Business, the fee shall also be linked to a divestiture within

the Trustee Divestiture Period.

Proposal by UMG

23 No later than one week after the Effective Date, UMG shall submit a list of one or more persons

whom UMG proposes to appoint as the Monitoring Trustee to the Commission for approval. No later

than one month before the end of the First Divestiture Period, UMG shall submit a list of one or more

persons whom UMG proposes to appoint as Divestiture Trustee to the Commission for approval. The

proposal shall contain sufficient information for the Commission to verify that the proposed trustee

fulfils the requirements set out in paragraph 22 and shall include:

(a)the full terms of the proposed mandate, which shall include all provisions necessary to

enable the Trustee to fulfil its duties under these Commitments;

reasonable requirements of the Purchaser(s) and to be agreed between UMG and the Purchaser(s).

(b) the outline of a work plan which describes how the Trustee intends to carry out its assigned

tasks; and

Divestiture Trustee as well as whether the proposed Trustee is to monitor compliance with the

Behavioural Commitments or whether different trustees are proposed for the three

functions.

Approval or rejection by the Commission

24 The Commission shall have the discretion to approve or reject the proposed Trustee(s) and to approve

the proposed mandate subject to any modifications which it deems necessary for the Trustee to fulfil

its obligations. If only one name is approved, UMG shall appoint or cause to be appointed, the

individual or institution concerned as Trustee, in accordance with the mandate approved by the

Commission. If more than one name is approved, UMG shall be free to choose the Trustee to be

appointed from among the names approved. The Trustee shall be appointed within one week of the

Commission’s approval, in accordance with the mandate approved by the Commission.

New proposal by UMG

25 If all the proposed Trustees are rejected, UMG shall submit the names of at least two more

individuals or institutions within one week of being informed of the rejection, in accordance

with the requirements and the procedure set out in paragraphs 21 and 24.

Trustee nominated by the Commission

26 If all further proposed Trustees are rejected by the Commission, the Commission shall

nominate a Trustee, whom UMG shall appoint, or cause to be appointed, in accordance with a

trustee mandate approved by the Commission.

II Functions of the Trustee

27 The Trustee shall assume its specified duties as described herein in order to ensure compliance

with the Commitments. The Commission may, on its own initiative or at the request of the

Trustee or UMG, give any orders or instructions to the Trustee in order to ensure compliance

with the conditions and obligations attached to the Decision.

Duties and obligations of the Monitoring Trustee

28 The Monitoring Trustee shall:

(a)propose in its first report to the Commission a detailed work plan describing how it intends

to monitor compliance with the obligations and conditions attached to the Decision;

(b)oversee the on-going management of the Divestment Business with a view to ensuring its

continued economic viability, marketability and competitiveness and monitor compliance

by UMG with the conditions and obligations attached to the Decision. To that end the

Monitoring Trustee shall:

(i)monitor the preservation of the economic viability, marketability and

competitiveness of the Divestment Business, and the keeping separate of the

Divestment Business from the business retained by the Parties, in accordance

with paragraphs 6 and 7 of the Commitments;

(ii)supervise the management of the Divestment Business as a distinct and saleable

entity, in accordance with paragraph 9 of the Commitments;

(iii)subject to 5(f) above (1) in consultation with UMG, determine all necessary

measures to ensure that UMG does not after the effective date obtain any

business secrets, know-how, commercial information, or any other information

of a confidential or proprietary nature relating to the Divestment Business, in

particular strive for the severing of the Divestment Business’ participation in a

central information technology network to the extent possible, without

compromising the viability of the Divestment Business, and (2) decide whether

such information may be disclosed to UMG as the disclosure is reasonably

necessary to allow UMG to carry out the divestiture or as the disclosure is

required by law;

(iv)monitor the splitting of assets and the allocation of Personnel between the

Divestment Business and UMG or Affiliated Undertakings;

(c)subject to 5(f) above, assume the other functions assigned to the Monitoring Trustee under

the conditions and obligations attached to the Decision propose to UMG such measures as

the Monitoring Trustee considers necessary to ensure UMG’s compliance with the

conditions and obligations attached to the Decision, in particular the maintenance of the full

economic viability, marketability or competitiveness of the Divestment Business, the

holding separate of the Divestment Business, and the non-disclosure of competitively

sensitive information;

(d)review and assess potential purchasers as well as the progress of the divestiture process and

verify that, dependent on the stage of the divestiture process, potential purchasers receive

sufficient information relating to the Divestment Business and the Personnel in particular

by reviewing, if available, the data room documentation, the information memorandum and

the due diligence process, and potential purchasers are granted reasonable access to the

Personnel;

(e)provide to the Commission, sending UMG a non-confidential copy at the same time, a

written report within 15 days after the end of every month. The report shall cover the

operation and management of the Divestment Business so that the Commission can assess

whether the business is held in a manner consistent with the Commitments and the progress

of the divestiture process as well as potential purchasers. In addition to these reports, the

Monitoring Trustee shall promptly report in writing to the Commission, sending UMG a

non-confidential copy at the same time, if it concludes on reasonable grounds that UMG is

failing to comply with these Commitments;

(f)within one week after the receipt of the documented proposal referred to in paragraph 20,

submit to the Commission a reasoned opinion as to the suitability and independence of the

proposed purchaser(s) and the viability of the Divestment Business after the Sale and as to

whether the Divestment Business is sold in a manner consistent with the conditions and

obligations attached to the Decision, in particular, if relevant, whether the Sale of the

Divestment Business without one or more Assets or not all of the Personnel affects the

viability of the Divestment Business after the sale, taking account of the proposed

purchaser(s).

Duties and obligations of the Divestiture Trustee

29 Within the Trustee Divestiture Period, the Divestiture Trustee shall sell at no minimum price the

Divestment Business to a purchaser, provided that the Commission has approved both the purchaser

and the final binding sale and purchase agreement in accordance with the procedure laid down in

paragraph 20. The Divestiture Trustee shall include in the sale and purchase agreement such terms and

conditions as it considers appropriate for an expedient sale in the Trustee Divestiture Period. In

particular, the Divestiture Trustee may include in the sale and purchase agreement such customary

representations and warranties and indemnities as are reasonably required to effect the sale. The

Divestiture Trustee shall protect the legitimate financial interests of UMG, subject to UMG’s

unconditional obligation to divest at no minimum price in the Trustee Divestiture Period.

30 In the Trustee Divestiture Period (or otherwise at the Commission’s request), the Divestiture Trustee

shall provide the Commission with a comprehensive monthly report written in English on the progress

of the divestiture process. Such reports shall be submitted within 15 days after the end of every month

with a simultaneous copy to the Monitoring Trustee and a non-confidential copy to the Parties.

III Duties and obligations of UMG

31 UMG shall provide and shall cause its advisors to provide the Trustee with all such co-operation,

assistance and information as the Trustee may reasonably require to perform its tasks. The Trustee

shall have full and complete access to any of UMG or the Divestment Business’ books, records,

documents, management or other personnel, facilities, sites and technical information necessary for

fulfilling its duties under the Commitments and UMG and the Divestment Business shall provide the

Trustee upon request with copies of any document. UMG and the Divestment Business shall make

available to the Trustee one or more offices on their premises and shall be available for meetings in

order to provide the Trustee with all information necessary for the performance of its tasks. UMG

shall provide the Monitoring Trustee with all managerial and administrative support that it may

reasonably request on behalf of the management of the Divestment Business. This shall include all

administrative support functions relating to the Divestment Business which are currently carried out at

headquarters level. UMG shall provide and shall cause its advisors to provide the Monitoring Trustee,

on request, with the information submitted to potential purchasers, in particular give the Monitoring

Trustee access to the data room documentation and all other information granted to potential

purchasers in the due diligence procedure. UMG shall inform the Monitoring Trustee on possible

purchasers, submit a list of potential purchasers, and keep the Monitoring Trustee informed of all

developments in the divestiture process. UMG shall grant or procure Affiliated Undertakings to grant

comprehensive powers of attorney, duly executed, to the Divestiture Trustee to effect the sale, the

Closing and all actions and declarations which the Divestiture Trustee considers necessary or

appropriate to achieve the sale and the Closing, including the appointment of advisors to assist with

the sale process. Upon request of the Divestiture Trustee, UMG shall cause the documents required

for effecting the sale and the Closing to be duly executed.

32 UMG shall indemnify the Trustee and its employees and agents (each an “Indemnified Party”) and

hold each Indemnified Party harmless against, and hereby agrees that an Indemnified Party shall have

no liability to UMG for any liabilities arising out of the performance of the Trustee’s duties under the

Commitments, except to the extent that such liabilities result from the wilful default, recklessness,

gross negligence or bad faith of the Trustee, its employees, agents or advisers. At the expense of

UMG, the Trustee may appoint advisors (in particular for legal advice), subject to UMG’s

approval (this approval not to be unreasonably withheld or delayed) if the Trustee considers the

appointment of such advisors necessary for the performance of its duties and obligations under

the Mandate, provided that any fees and other expenses incurred by the Trustee are reasonable.

Should UMG refuse to approve the advisors proposed by the Trustee the Commission may

approve the appointment of such advisors instead, after having heard UMG. Only the Trustee

shall be entitled to issue instructions to the advisors. Paragraph 32 shall apply mutatis mutandis.

In the Trustee Divestiture Period, the Divestiture Trustee may use advisors who served UMG during

the Divestiture Period if the Divestiture Trustee considers this in the best interest of an

expedient sale.

IV Replacement, discharge and reappointment of the Trustee

33 If the trustee ceases to perform its functions under the Commitments or for any other good

cause, including the exposure of the Trustee to a conflict of interest:

(a)the Commission may, after hearing the Trustee, require UMG to replace the Trustee; or

(b)UMG, with the prior approval of the Commission, may replace the Trustee.

34 If the Trustee is removed according to paragraph 33, the Trustee may be required to continue in its

function until a new Trustee is in place to whom the Trustee has effected a full hand over of all

relevant information. The new Trustee shall be appointed in accordance with the procedure referred to

in paragraphs 21-26 above.

35 Beside the removal according to paragraph 33, the Trustee shall cease to act as Trustee only

after the Commission has discharged it from its duties after all the Commitments with which the

Trustee has been entrusted have been implemented. However, the Commission may at any time

require the reappointment of the Monitoring Trustee if it subsequently appears that the relevant

remedies might not have been fully and properly implemented.

Section H. Fast Track Dispute Resolution

36 In the event that a third party claims that UMG or an Affiliated Undertaking is failing to comply with

the Commitments vis-à-vis that third party, the fast track dispute resolution procedure as described

herein shall apply.

37 Any third party who wishes to avail itself of the fast track dispute resolution procedure (a

“Requesting Party”) shall send a written request to UMG (with a copy to the Monitoring Trustee)

setting out in detail the reasons leading that party to believe that UMG is failing to comply with the

requirements of the Commitment (the “Request”). The Requesting Party and UMG will use their best

efforts to resolve all differences of opinion and to settle all disputes that may arise through co-

operation and consultation within a reasonable period of time not exceeding fifteen (15) working days

after receipt of the Request.

38 The Monitoring Trustee shall present its own proposal (the “Trustee Proposal”) for resolving the

dispute within eight (8) working days, specifying in writing the action, if any, to be taken by UMG or

an Affiliated Undertaking in order to ensure compliance with the Commitment(s) vis-à-vis the

Requesting Party, and be prepared, if requested, to facilitate the settlement of the dispute.

39 Should the Requesting Party and UMG (together the “Parties to the Arbitration”) fail to resolve

their differences of opinion in the consultation phase, the Requesting Party shall serve a notice (the

“Notice”), in the sense of a request for arbitration, to the ICC (hereinafter the “Arbitral Institution”),

with a copy of such Notice and request for arbitration to UMG.

40 The Notice shall set out in detail the dispute, difference or claim (the “Dispute”) and shall contain,

inter alia, all issues of both fact and law, including any suggestions as to the procedure, and all

documents relied upon shall be attached, e.g. documents, agreements, expert reports, and witness

statements. The Notice shall also contain a detailed description of the action to be undertaken by

UMG (including, if appropriate, a draft contract comprising all relevant terms and conditions) and the

Trustee Proposal, including a comment as to its appropriateness.

41 UMG shall, within 10 working days from receipt of the Notice, submit its answer (the “Answer”),

which shall provide detailed reasons for its conduct and set out, inter alia, all issues of both fact and

law, including any suggestions as to the procedure, and all documents relied upon, e.g. documents,

agreements, expert reports, and witness statements. The Answer shall, if appropriate, contain a

detailed description of the action which UMG proposes to undertake vis-à-vis the Requesting Party

(including, if appropriate, a draft contract comprising all relevant terms and conditions) and the

Trustee Proposal (if not already submitted), including a comment as to its appropriateness.

Appointment of the Arbitrators

42 The Arbitral Tribunal shall consist of three persons. The Requesting Party shall nominate its arbitrator

in the Notice; UMG shall nominate its arbitrator in the Answer. The arbitrator nominated by the Requesting Party and by UMG shall, within five working days of the nomination of the latter, nominate the chairman, making such nomination known to the parties and the Arbitral Institution which shall forthwith confirm the appointment of all three arbitrators.

389

43Should the Requesting Party wish to have the Dispute decided by a sole arbitrator it shall indicate this in the Notice. In this case, the Requesting Party and UMG shall agree on the nomination of a sole arbitrator within five working days from the communication of the Answer, communicating this to the Arbitral Institution.

44Should UMG fail to nominate an arbitrator, or if the two arbitrators fail to agree on the chairman, or should the Parties to the Arbitration fail to agree on a sole arbitrator, the default appointment(s) shall be made by the Arbitral Institution.

45The three-person arbitral tribunal or, as the case may be, the sole arbitrator, are herein referred to as the “Arbitral Tribunal”.

Arbitration Procedure

46The Dispute shall be finally resolved by arbitration under the ICC rules, with such modifications or adaptations as foreseen herein or necessary under the circumstances (the “Rules”). The arbitration shall be conducted in London in the English language.

47The procedure shall be a fast-track procedure. For this purpose, the Arbitral Tribunal shall shorten all applicable procedural time-limits under the Rules as far as admissible and appropriate in the circumstances. The Parties to the Arbitration shall consent to the use of e-mail for the exchange of documents.

48The Arbitral Tribunal shall, as soon as practical after the confirmation of the Arbitral Tribunal, hold an organisational conference to discuss any procedural issues with the Parties to the Arbitration. Terms of Reference shall be drawn up and signed by the Parties to the Arbitration and the Arbitration Tribunal at the organisational meeting or thereafter and a procedural time-table shall be established by the Arbitral Tribunal. An oral hearing shall, as a rule, be established within two months of the confirmation of the Arbitral Tribunal.

49In order to enable the Arbitral Tribunal to reach a decision, it shall be entitled to request any relevant information from the Parties to the Arbitration, to appoint experts and to examine them at the hearing, and to establish the facts by all appropriate means. The Arbitral Tribunal is also entitled to ask for assistance by the Trustee in all stages of the procedure if the Parties to the Arbitration agree.

50The Arbitral Tribunal shall not disclose confidential information and apply the standards attributable to confidential information under the Merger Regulation. The Arbitral Tribunal may take the measures necessary for protecting confidential information in particular by restricting access to confidential information to the Arbitral Tribunal, the Trustee, and outside counsel and experts of the opposing party.

51The burden of proof in any dispute under these Rules shall be borne as follows: (i) the Requesting Party must produce evidence of a prima facie case and (ii) if the Requesting Party produces evidence of a prima facie case, the Arbitral Tribunal must find in favour of the Requesting Party unless UMG can produce evidence to the contrary.

Involvement of the Commission

52The Commission shall be allowed and enabled to participate in all stages of the procedure by (a) Receiving all written submissions (including documents and reports, etc.) made by the Parties to the Arbitration;

(b)Receiving all orders, interim and final awards and other documents exchanged by the Arbitral Tribunal with the Parties to the Arbitration (including Terms of Reference and procedural time-table);

390

(c)Giving the Commission the opportunity to file amicus curiae briefs; and

(d)Being present at the hearing(s) and being allowed to ask questions to parties, witnesses and experts.

53The Arbitral Tribunal shall forward, or shall order the Parties to the Arbitration to forward, the documents mentioned to the Commission without delay.

54In the event of disagreement between the Parties to the Arbitration regarding the interpretation of the Commitment, the Arbitral Tribunal may seek the Commission’s interpretation of the Commitment before finding in favour of any Party to the Arbitration and shall be bound by the interpretation.

Decisions of the Arbitral Tribunal

55The Arbitral Tribunal shall decide the dispute on the basis of the Commitment and the Decision.

56Issues not covered by the Commitment and the Decision shall be decided (in the order as stated) by reference to the Merger Regulation, EU law and general principles of law common to the legal orders of the Member States without a requirement to apply a particular national system. The Arbitral Tribunal shall take all decisions by majority vote.

5756 Upon request of the Requesting Party, the Arbitral Tribunal may make a preliminary ruling on the Dispute. The preliminary ruling shall be rendered within one month after the confirmation of the Arbitral Tribunal, shall be applicable immediately and, as a rule, remain in force until a final decision is rendered.

5857 The Arbitral Tribunal shall, in the preliminary ruling as well as in the final award, specify the action, if any, to be taken by UMG or an Affiliated Undertaking in order to comply with the commitments vis-à-vis the Requesting Party (e.g. specify a contract including all relevant terms and conditions). The final award shall be final and binding on the Parties to the Arbitration and shall resolve the Dispute and determine any and all claims, motions or requests submitted to the Arbitral Tribunal. The arbitral award shall also determine the reimbursement of the costs of the successful party and the allocation of the arbitration costs. In case of granting a preliminary ruling or if otherwise appropriate, the Arbitral Tribunal shall specify that terms and conditions determined in the final award apply retroactively.

5958 The final award shall, as a rule, be rendered within six months after the confirmation of the Arbitral Tribunal. The time-frame shall, in any case, be extended by the time the Commission takes to submit an interpretation of the Commitment if asked by the Arbitral Tribunal.

6059 The Parties to the Arbitration shall prepare a non-confidential version of the final award, without business secrets. The Commission may publish the non-confidential version of the award.

6160 Nothing in the arbitration procedure shall affect the power to the Commission to take decisions in relation to the Commitment in accordance with its powers under the Merger Regulation.

Section I. The Review Clause

61The Commission may, where appropriate, in response to a request from UMG showing good cause and accompanied by a report from the Monitoring Trustee (unless the Commission waives the need for such a report):

(a)Grant an extension of the time periods foreseen in the Commitments, or

(b)Waive, modify or substitute, in exceptional circumstances, one or more of the undertakings in these Commitments.

391

Where UMG seek an extension of a time period, it shall submit a request to the Commission no later than one month before the expiry of that period, showing good cause. Only in exceptional circumstances shall UMG be entitled to request an extension within the last month of any period.

………………………………………

Duly authorised for and on behalf of UMG

Richard Constant, General Counsel, Universal Music Group International Limited

24 August 2012

392

SCHEDULE

The Divestment Business is described in the Schedules to these Commitments as follows:

(a)each Legal Entity and the assets comprised within them, together with the revenues

generated by each Legal Entity, is described in Schedule 1;

(b)the NOW JV is described in Schedule 2 together with the revenues generated by that asset;

and

(c)the MoS Agreement is described in Schedule 3 together with the revenues generated by

this

asset.

393

M.6458 -- UMG / EMI Music

Schedule to Divestiture Commitments

Table of Contents

Schedule 1 – Legal Entities

A - EMI Records Limited

B - Chrysalis Records Limited

C - Ensign Records Limited

D - Mute Records Limited

E - EMI Belgium

F - EMI Czech Republic

G - EMI Poland

394

H - EMI Portugal

I - EMI France

J - EMI Sweden

K - EMI Norway

L – EMI Denmark

M - EMI Spain

N - Sanctuary Records Group Limited

O - Co-Op Music Limited

P - King Island Roxystar Recordings AB

Q - UMG Greece

R - UMG’s Shareholding in Jazzland

395

S - MPS

Schedule 2 - NOW! compilation business

Schedule 3 – The MoS Agreement

Future Deliverables (as at 24 August 2012) Appendix

396

SCHEDULE 1

LEGAL ENTITIES

A.EMI Records Limited

Legal and functional structure

1.1EMI Records Limited (“EMIRL”), has its registered office and principal place of business at 27 Wrights Lane, London, W8 5SW. EMIRL is a subsidiary of EMI Limited (with one share legally held by EMI Nominees Limited) and currently the main trading entity of EMI. Please see the EMI UK

723Structure Chart at Confidential Annex 1;

1.2EMIRL will be transferred with the following key assets:

(i)"Parlophone" branded artists signed to EMIRL including Coldplay (see further

section 5 below);

(ii)Pink Floyd;

(iii)The rights to David Guetta’s future recordings pursuant to his current recording

724agreement with EMIRL;

(iv)“EMI Classics” branded artists signed to EMIRL, which in turn constitute the

substantial majority of "EMI Classics” branded artists in the whole EMI Group;

(v)“Virgin Classics” branded artists signed to EMIRL, which in turn constitute the

substantial majority of "Virgin Classics” branded artists in the whole EMI

725Group;

(vi)“Virgin Classics” branded Residual Classics Artists which are not transferred to

EMI France and “EMI Classics” branded Residual Classics Artists (see further

section 6 below); and

723Please note that all EMI structure charts reflect the pre-separation EMI Group, i.e. before the music publishing business was extracted by way of corporate reorganisation. Revised structure charts are in the process of being produced and, when complete, can be provided at the Commission’s request.

724[…]*.

725There are also some “Virgin Classics” branded artists who are signed to EMI France, a separate divestment asset. Also, as UMG retains the Virgin Records asset, the Virgin brand will be retained by UMG.

(vii)all other non-“Parlophone” and non-“Virgin Records” branded artists signed to

726EMIRL (see Confidential Annex A.1),

727(together the “EMIRL Divestment”).

Description of the EMIRL Divestment

2.1The EMIRL Divestment includes, but is not limited to the tangible and intangible assets set out below.

At Confidential Annex 2 are details of the principal artist agreements signed to EMIRL and which will be transferred with EMIRL. At Confidential Annex 3 are details of the turnover generated by the EMIRL Divestment’s principal artists.

2.2The EMIRL Divestment includes current pop artists with future hit potential such as Coldplay, David Guetta, Tinie Tempah, Gorillaz, Lily Allen, Eliza Doolittle, Blur and Conor Maynard, as well as top-selling catalogue artists such as Pink Floyd, Iron Maiden, David Bowie, Duran Duran, Gorillaz, Radiohead, Kylie Minogue, Cliff Richard, Kraftwerk Tina Turner, Kate Bush, Sigur Rós, Blur, Pet Shop Boys, Whitesnake, Marillion, David Gilmour, Morrissey and Joe Cocker. It also includes current classical/cross-over artists with significant future revenue generating potential (as demonstrated by their past success) including Sir Simon Rattle, Daniel Barenboim, Alison Balsom, Thomas Adès, Angela Gheorghiu, Philippe Jaroussky and the Berlin-based Artemis Quartet, as well as catalogue recordings by Herbert von Karajan, Maria Callas, Vanessa-Mae and Maksim Mrvica.

2.3The recorded music revenues generated in relation to EMIRL Divestment are set out as follows:

728(a)UK “Parlophone" branded artists (other than the Beatles) (please see Confidential Annex A.2 for details of these artists) signed to EMIRL generated […]* of recorded music revenues across the EEA and […]* in 2011/12. This asset generated recorded music revenues in the UK alone of […]* in 2011/12. These revenues only account for the recorded music revenues and do not include ancillary revenues, such as merchandising, synchronisation and neighbouring rights (public performance). Please see Confidential Annex 4 for further details;

(b)Pink Floyd generated […]* of recorded music revenues across the EEA and […]* globally in 2011/12, of which […]* was in the UK in 2011/12. Please see Confidential Annex 4 for further details.

(c)“EMI Classics” branded artists, including those signed to EMIRL (which constitute the majority of "EMI Classics” branded artists in the whole EMI Group), and including those

729“EMI Classics” branded Residual Classics Artists(see further section 6 below) generated […]* across the EEA and […]* globally in 2011/12. This asset generated recorded music revenues in the UK alone of […]* in 2011/12. These revenues only account for the recorded music revenues of these assets, and do not include ancillary revenues, such as merchandising, synchronisation and neighbouring rights (public performance). Please see Confidential Annex 4 for further details;

(d)“Virgin Classics” branded artists, including those signed to EMIRL (which constitute the

730majority of "Virgin Classics” branded artists in the whole EMI Group),and those signed to EMI France, and including those “Virgin Classics” branded Residual Classics Artists which are not transferred to EMI France or EMIRL generated […]* across the EEA and […]* globally in 2011/12. This asset generated recorded music revenues in the UK alone of […]* in 2011/12. Further, these revenues only account for the recorded music revenues of these assets, and do not include ancillary revenues, such as merchandising, synchronisation and neighbouring rights (public performance). Please see Confidential Annex 4 for further details;

731(e) […]*; and

(f)All other non-Parlophone and non-Virgin Records branded artists signed to EMIRL, excluding Pink Floyd, (please see Confidential Annex A.1 for details of these artists)

732generated […]* across the EEA and […]* globally in 2011/12.This asset generated recorded music revenues in the UK alone of […]* in 2011/12. Further, these revenues only account for the recorded music revenues of these assets, and do not include ancillary

729Excluding those EMI Classics artists signed to EMI Belgium and EMI Portugal, which will be divested in any event as part of the EMI Belgium and EMI Portugal divestments described further at sections E and H of this Schedule.

730Although there are some “Virgin Classics” branded artists that are signed to EMI France, a separate divestment asset.

731See footnote 2 above.

732These artists include David Bowie, Cliff Richard, Iron Maiden, Whitesnake, Kraftwerk and Kate Bush. The commercial and artistic value of these artists is very significant. All these artists are recognised for career longevity, not to mention influence on subsequent generations of musicians. With global recording sales estimated at 140 million, David Bowie is one of the rock era’s most revered performers; an anniversary reissue in 2012 of his “Ziggy Stardust” returned the album to the UK charts for the fourth time since 1972. Cliff Richard, with total recording sales exceeding 250 million, is the only singer to have had UK No. 1 singles in five different decades. Iron Maiden are estimated to have sold 85 million recordings since their EMI debut. Their 2012 “Final Frontier” album topped charts in 28 countries; the band played for 2 million+ fans in 36 countries during the accompanying concert tour. Whitesnake also sustains their popularity through global touring, while one of their trademark songs was featured in the recent London/Broadway smash hit musical, “Rock of Ages.” Germany’s Kraftwerk (admired by U2) and Britain’s Kate Bush are further examples of musicians with enduring influence and sales. Bush’s work has sold consistently throughout Europe, including two new albums issued last year. Her music – and Bowie’s – was featured in the closing ceremony of the 2012 Olympics. Duran Duran have been delivering major hit singles and albums worldwide since they emerged with “Planet Earth” in 1981. One hundred million record sales and 14 UK Top 10 singles later, they have 13 studio albums in their catalogue and a deserved reputation for bespoke visuals to accompany their music.

399

revenues, such as merchandising, synchronisation and neighbouring rights (public performance). Please see Confidential Annex 4 for further details.

2.4EMI (IP) Limited holds the “Parlophone” trademark registrations as set out in Annex A.5. EMI (IP) Limited. UMG commits to procure the assignment of the Parlophone trademarks as part of the EMIRL Divestment, subject to a perpetual licence back for their use in respect of the Beatles.

Personnel

Group 1: Key Personnel

Name

Position

Employing entity

[…] *

3.1[…]*.

Group 2: Other Personnel

7333.2[…]*.

3.3[…]*.

3.4[…]*.

Beatles personnel:

[…]*.

Virgin Records personnel:

[…]*.

733[…]*.

400

Transitional Arrangements

4.1It is not anticipated that it would be necessary for arrangements to be put in place for the supply of any of any corporate support services for a transitional period, but UMG or their Affiliated Undertakings commit to agree to supply any such services on terms to be agreed on an arm’s length basis, if determined necessary by the Trustee and required by the Commission at the request of the Purchaser, for a transitional period of […]*.

4.2The Purchaser will be granted licences to use the EMI trademark in connection with physical product manufactured prior to the date of completion of the sale of the EMIRL Divestment to the Purchaser to avoid recall or scrapping of existing product and will only be required to change artwork which includes the EMI trademark on future manufacturing runs. The Purchaser will also be licensed to use the EMI trademark as part of the corporate name of the relevant legal entity and for associated corporate purposes such as office signage and stationary during a transitional period of […]* following the date of completion of the sale of the EMIRL Divestment to the Purchaser before effecting a corporate change of name.

Exclusions

5.1Prior to the divestment, UMG will transfer by way of a reverse carve-out to a retained EMI Group entity contracts with The Beatles (including the individual repertoire of the Beatles, i.e. John Lennon, George Harrison, Paul McCartney, Wings and Ringo Starr) and all “Virgin Records" branded artists other than David Guetta. Please see Confidential Annex A.7 for details of these “Virgin Records” branded artist contracts. For the avoidance of doubt, it is noted that the Abbey Road premises and management team are excluded from the EMIRL Divestment (consistent with the exclusion of the Beatles).

5.2UMG has the option to transfer the EMI classics business, comprising the artists released on the EMI Classics and Virgin Classics labels as well as the Residual Classics Artists (the “EMI Classics Business”), to a separate purchaser (other than the purchaser of the EMIRL Divestment), such as a company with particular interest or expertise in the classical music business, subject only to any delivery issues of such a EMI Classics Business package, in which case the EMI Classics Business will be transferred together as part of the Main Package. In case of a separate divestment of the EMI Classics Business, UMG would use its best commercial endeavours to transfer any EMIRL signed EMI Classics and/or Virgin Classics out of EMIRL.

7345.3[…]*, […]*.

Additions

7356.1[…]* .

734[…]*.

735[…]*.

401

B.Chrysalis Records Limited

Legal and functional structure

7.1Chrysalis Records Limited (“Chrysalis”), has the following legal and functional structure:

(a)Chrysalis’s registered address and place of business is 27 Wrights Lane, London, W8 5SW, United Kingdom.

(b)Chrysalis is a wholly owned subsidiary of the EMI Group (see the EMI UK Structure Chart at Confidential Annex 1).

Description of the asset

8.1Chrysalis includes, but is not limited to the tangible and intangible assets set out below.

8.2Rights to artist recordings, for which see:

(a)Confidential Annex 2 for details of the principal artist agreements signed to Chrysalis; and

(b)Confidential Annex 3 for details of the turnover generated by Chrysalis’s principal artists.

8.3Chrysalis Holdings Limited holds the “Chrysalis” trade mark as set out in Annex B.1. While Chrysalis Holdings Limited is not an EMI Group company (and hence not part of the Divestment Business), the terms of the trademark licence from Chrysalis Holdings Limited to Chrysalis require Chrysalis Holdings Limited to grant a licence on similar terms to a purchaser of Chrysalis. By divesting Chrysalis, UMG can therefore procure a license of the Chrysalis trade mark to the purchaser as part of the Chrysalis Divestment.

8.4Chrysalis’ roster includes top-selling catalogue artists in genres including the Ramones, Jethro Tull, UFO, Spandau Ballet, the Specials, Generation X, Split Enz, Runrig and the Proclaimers.

8.5The recorded music revenues generated by this asset were […]* across the EEA and […]* globally in

7362011/12. This asset generated recorded music revenues in the UK alone of […]* in 2011/12.

Further, these revenues only account for the recorded music revenues of these assets, and do not include ancillary revenues, such as merchandising, synchronisation and neighbouring rights (public

performance). Please see Confidential Annex 4 for further details.

736The figures in this paragraph are consolidated with those of Ensign Records Limited.

737For the avoidance of doubt, the divestment of Chrysalis Records Limited will not include any shareholdings in any other legal entities. Any such shareholding will be transferred to another retain group company before the divestment.

402

Personnel

Group 1: Key Personnel

9.1There are no Key Personnel.

Group 2: Other Personnel

9.2Chrysalis does not itself employ any personnel, but relies on A&R, marketing and sales support by personnel some of whom are employed by EMI Records Limited (another asset in the Divestment Business).

9.3[…]*.

10 Transitional Arrangements

10.1As noted above, Chrysalis currently benefits from general corporate support functions provided by the wider EMI Group. It is not anticipated that it would be necessary for arrangements to be put in place for the supply of any of these support services for a transitional period, but UMG or their Affiliated Undertakings commit to agree to supply any such services on terms to be agreed on an arm’s length basis, if determined necessary by the Trustee and required by the Commission at the request of the Purchaser, for a transitional period of […]*.

10.2The Purchaser will be granted licences to use the EMI trademark in connection with physical product manufactured prior to the date of completion of the sale of Chrysalis to the Purchaser to avoid recall or scrapping of existing product and will only be required to change artwork which includes the EMI trademark on future manufacturing runs. The Purchaser will also be licensed to use the EMI trademark as part of the corporate name of the relevant legal entity and for associated corporate purposes such as office signage and stationary during a transitional period of […]* following the date of completion of the sale of Chrysalis to the Purchaser before effecting a corporate change of name.

11 Exclusions

11.1UMG will retain by way of a reverse carve-out the catalogue of Robbie Williams which will be transferred by way of an intra-group assignment to a retained EMI Group company.

403

C. Ensign Records Limited

12Legal and functional structure

12.1Ensign Records Limited (“Ensign”), has the following legal and functional structure:

(a)Ensign’s registered address and place of business is 27 Wrights Lane, London, W8 5SW, United Kingdom.

(b)Ensign is a wholly owned subsidiary of the EMI Group (see the EMI UK Structure Chart at Confidential Annex 1).

13Description of the asset

13.1Ensign includes, but is not limited to the tangible and intangible assets set out below:

13.2Rights to artist recordings, for which see:

(a)Confidential Annex 2 for details of the principal artist agreements signed to Ensign; and

(b)Confidential Annex 3 for details of the turnover generated by Ensign’s principal artists.

13.3EMI (IP) Limited holds the “Ensign” trademark registrations as set out in Annex C.1, with the exception of the Mexican registration, which is owned by Ensign. EMI (IP) Limited will transfer the rights to these trademarks to the Purchaser.

13.4Ensign’s roster includes artists ranging from critically-acclaimed bands such as World Party and the Waterboys to Phil Fearon and Galaxy.

13.5The recorded music revenues generated by this asset are consolidated with those of Chrysalis Records Limited. Please see Confidential Annex 4 for further details.

14Group 1: Key Personnel

14.1There are no Key Personnel.

Group 2: Other Personnel

14.2Ensign does not itself employ any personnel, but relies on A&R and marketing and sales support by personnel some of whom are employed by EMI Records Limited (another asset in the Divestment Business).

14.3[…]*.

738For the avoidance of doubt, the divestment of Ensign will not include any shareholdings in any other legal entities. Any such shareholding will be transferred to another retain group company before the divestment.

404

15 Transitional Arrangements

15.1As noted above, Ensign currently benefits from general corporate support functions provided by the wider EMI Group. It is not anticipated that it would be necessary for arrangements to be put in place for the supply of any of these support services for a transitional period, but UMG or their Affiliated Undertakings commit to agree to supply any such services on terms to be agreed on an arm’s length basis, if determined necessary by the Trustee and required by the Commission at the request of the Purchaser, for a transitional period of […]*.

15.2The Purchaser will be granted licences to use the EMI trademark in connection with physical product manufactured prior to the date of completion of the sale of Ensign to the Purchaser to avoid recall or scrapping of existing product and will only be required to change artwork which includes the EMI trademark on future manufacturing runs. The Purchaser will also be licensed to use the EMI trademark as part of the corporate name of the relevant legal entity and for associated corporate purposes such as office signage and stationary during a transitional period of […]* following the date of completion of the sale of Ensign to the Purchaser before effecting a corporate change of name.

405

(b) Confidential Annex 3 for details of the turnover generated by EMI Portugal’s principal artists.

artists.

33.3EMI Portugal has a lease of the following property: Alameda Dos Oceanos, Lote, 2.11.01 E-Piso 2, Parque Das Nações, 1998-020, Lisboa, Portugal.

33.4EMI Portugal’s roster includes current artists with future hit potential such as Camané and Carminho (both Fado singers that are highly respected in Portugal and abroad), singer-songwriter Jorge Palma, the Rock group Clã and pop artists Deolinda, Vincente Palma, Mendes, Ricardo Ribeiro, as well as top-selling catalogue artists such as Rui Veloso, who holds the title for the highest selling Portuguese album of all time with “Mingos & Os Samurais” in 1990.

33.5In addition UMG undertakes to use reasonable endeavours to include the rights to any local

744compilations signed to EMI Portugal within the divestment package.

33.6The recorded music revenues generated by EMI Portugal overall were […]* across the EEA and […]*

745globally in 2011/12.This asset generated recorded music revenues in the Portugal alone of […]* in

744For the avoidance of doubt, the EMI Portugal divestment will not include any shareholdings in any other EMI Group companies (if any).

415

2011/12. Further, these revenues only account for the recorded music revenues of these assets, and do

not include ancillary revenues, such as merchandising, synchronisation and neighbouring rights

(public performance). Please see Confidential Annex 4 for further details.

34Personnel

Group 1: Key Personnel

Name

Position

Employing entity

[…]*.

Group 2: Other Personnel

34.1EMI Portugal team comprises the following personnel:

[…]*.

34.2[…]*. See further the Portugal Personnel Structure Chart at Confidential Annex H.2.

34.3[…]*.

34.4[…]*.

35Transitional Arrangements

35.1As noted above, EMI Portugal acts as a “territory cluster” with EMI Spain. […]*.

35.2[…]*.

35.3The Purchaser will be granted licences to use the EMI trademark in connection with physical product

manufactured prior to the date of completion of the sale of EMI Portugal to the Purchaser to avoid

recall or scrapping of existing product and will only be required to change artwork which includes the

EMI trademark on future manufacturing runs. The Purchaser will also be licensed to use the EMI

trademark as part of the corporate name of the relevant legal entity and for associated corporate

purposes such as office signage and stationary during a transitional period of […]* following the date

of completion of the sale of EMI Portugal to the Purchaser before effecting a corporate change of

name.

745To avoid double counting, the figures in this paragraph do not include revenues generated by artists signed to EMI Portugal that have already been accounted for elsewhere in the Divestment Business, i.e. Virgin Classics artists signed to EMI Portugal.

416

I.EMI France

36Legal and functional structure

36.1EMI Music France SAS (“EMI France”), has the following legal and functional structure:

(a)EMI France’s registered address and place of business is 118-126 Rue de Mont Cenis, 75018, Paris.

(b)EMI France is a wholly owned subsidiary of the EMI Group (see the EMI France Structure Chart at Annex I.1).

(c)Most of EMI France's essential functions are performed within the business to be divested, including A&R and marketing. However, certain functions of EMI France are provided centrally by the EMI Group, such as manufacturing and distribution.

36.2The EMI France divestment will also include the […]* shareholding in PLAY ON S.A.S. (“Play On”) held by EMI France. In addition, EMI France will also include those “Virgin Classics” branded Residual Classics Artists not transferring to EMIRL.

37Description of the asset

37.1EMI France includes, but is not limited to the tangible and intangible assets set out below:

37.2Rights to artist recordings, for which see:

(a)Confidential Annex 2 for details of the principal artist agreements signed to EMI France;

and

(b)Confidential Annex 3 for details of the turnover generated by EMI France’s principal artists (including Play On).

37.3EMI France has a leasehold interest of the following properties:

(a)118-126 Rue de Mont Cenis, 75018, Paris, France; and

(b)22 rue de l'Equerre - 95310 Saint Ouen L'Aumône, France (over which EMI France has exercised its option to acquire the freehold interest for EUR 1).

37.4EMI France’s roster includes current artists with future hit potential such as Diam, Matt Pokora, Daft Punk and Rohff, as well as top-selling catalogue artists such as Edith Piaf, Alain Souchon and

746Julien Clerc.In addition, EMI France owns the rights to David Guetta’s hugely successful

747catalogue.Play On’s roster includes current artists such Leslie, Matt Houston, Gérard Lenorman,

417

Colonel Reyel, Merlot, Volo, Vadel, Elisa Tovati, Iselym, Junior Caldera, Zaz, Tom Dice, Veronique Jannot, Vincent Liben, Klem, and Benjamin Boehm.

37.5In addition UMG undertakes to use reasonable endeavours to include the rights to any local

748compilations signed to EMI France within the divestment package.

37.6The recorded music revenues generated by EMI France overall (excluding those related to EMI Classics and Virgin Classics artists, which were reported above, but including revenues attributable to

749David Guetta) were […]* across the EEA and […]* globally in 2011/12. This asset generated

recorded music revenues in France alone of […]* in 2011/12. Further, these revenues only account

for the recorded music revenues of these assets, and do not include ancillary revenues, such as

merchandising, synchronisation and neighbouring rights (public performance). This makes EMI France the largest EMI local operating company across the EEA after the UK. Please see

Confidential Annex 4 for further details.

38Personnel

Group 1: Key Personnel

Name

Position

Employing entity

[…]*

75038.1 […]*. […]*.

Group 2: Other Personnel

38.2Most of EMI France’s essential functions are performed by personnel within the business to be divested. EMI France has approximately […]* employees. Please see the EMI France organisation chart at Confidential Annex I.2. The management personnel of EMI France are (including again the Key Personnel to show their relationship to the Other Personnel):

[…]*.

38.3[…]*.

748For the avoidance of doubt, the EMI France divestment will not include any shareholdings in any other EMI Group companies (if any).

749This excludes revenues attributable to EMI Classics, and Virgin Classics.

750[…]*.

418

39Transitional Arrangements

39.1As noted above, EMI France currently benefits from general corporate support functions provided by the wider EMI Group. It is not anticipated that it would be necessary for arrangements to be put in place for the supply of any of these support services for a transitional period, but UMG or their Affiliated Undertakings commit to agree to supply any such services on terms to be agreed on an arm’s length basis, if determined necessary by the Trustee and required by the Commission at the request of the Purchaser, for a transitional period of […]*.

39.2The Purchaser will be granted licences to use the EMI trademark in connection with physical product manufactured prior to the date of completion of the sale of EMI France to the Purchaser to avoid recall or scrapping of existing product and will only be required to change artwork which includes the EMI trademark on future manufacturing runs. The Purchaser will also be licensed to use the EMI trademark as part of the corporate name of the relevant legal entity and for associated corporate purposes such as office signage and stationary during a transitional period of […]* following the date of completion of the sale of EMI France to the Purchaser before effecting a corporate change of name.

40Exclusions

40.1[…]*.

419

J.EMI Sweden

41Legal and functional structure

41.1EMI Music Sweden AB (“EMI Sweden”), has the following legal and functional structure:

(a)EMI Sweden’s registered address and place of business is Linnegatan 89B, 10450 Stockholm.

(b)EMI Sweden is a wholly owned subsidiary of the EMI Group (see the EMI Sweden Structure Chart at Annex J.1).

(c)Most of EMI Sweden’s essential functions are performed within the business to be divested, although certain functions of EMI Sweden are provided centrally by the EMI Group, such as manufacturing and distribution. In addition, EMI has in recent years rationalised its local operations across the Nordic region (see further details below).

(d)EMI Sweden acts as a “territory cluster” with the other Nordic countries (see further details below).

42Description of the asset

42.1EMI Sweden includes, but is not limited to the tangible and intangible assets set out below:

42.2Rights to artist recordings, for which see:

(a)Confidential Annex 2 for details of the principal artist agreements signed to EMI Sweden;

and

(b)Confidential Annex 3 for details of the turnover generated by EMI Sweden’s principal artists.

42.3EMI Sweden has a leasehold of the following properties: Linnegatan 89B, 10450 Stockholm, Elite Avenue Hotell, van3, Kungportsavenyn 36, 41136 Goteborg, and Karlavagen 100, 115 26, Stockholm (archives).

42.4EMI Sweden’s roster includes current artists with future hit potential such as Roxette, Eldkvarn, Timbuktu, Lasse Stefanz, Eric Amarillo and Magnus Uggla.

42.5In addition UMG undertakes to use reasonable endeavours to include the rights to any local

751compilations signed to EMI Sweden within the divestment package.

751For the avoidance of doubt, the EMI Sweden divestment will not include any shareholdings in any other EMI Group companies outside of Sweden.

420

42.6The recorded music revenues generated by EMI Sweden overall were […]* across the EEA and […]*

752globally in 2011/12.This asset generated recorded music revenues in Sweden alone of […]* in

2011/12. Further, these revenues only account for the recorded music revenues of these assets, and do

not include ancillary revenues, such as merchandising, synchronisation and neighbouring rights (public performance). Please see Confidential Annex 4 for further details.

43Personnel

Group 1: Key Personnel

Name

Position

Employing entity

[…]*

Group 2 Other Personnel

43.1As noted above, EMI has in recent years rationalised its local operations across the Nordic region such that managers have both local and regional responsibilities. Set out below are the details of personnel and their connection with Sweden:

[…]*.

43.2[…]*.

44Transitional Arrangements

44.1As noted above, EMI Sweden currently benefits from general corporate support functions provided by the wider EMI Group. It is not anticipated that it would be necessary for arrangements to be put in place for the supply of any of these support services for a transitional period, but UMG or their Affiliated Undertakings commit to agree to supply any such services on terms to be agreed on an arm’s length basis, if determined necessary by the Trustee and required by the Commission at the request of the Purchaser, for a transitional period of […]*.

44.2The Purchaser will be granted licences to use the EMI trademark in connection with physical product manufactured prior to the date of completion of the sale of EMI Sweden to the Purchaser to avoid recall or scrapping of existing product and will only be required to change artwork which includes the EMI trademark on future manufacturing runs. The Purchaser will also be licensed to use the EMI trademark as part of the corporate name of the relevant legal entity and for associated corporate purposes such as office signage and stationary during a transitional period of […]* following the date

of completion of the sale of EMI Sweden to the Purchaser before effecting a corporate change of

name.

752To avoid double counting, the figures in this paragraph do not include revenues generated by artists signed to EMI Sweden that have already been accounted for elsewhere in the Divestment Business, i.e. EMI Classics and Virgin Classics artists signed to EMI Portugal.

421

of completion of the sale of EMI Sweden to the Purchaser before effecting a corporate change of

name.

422

K.EMI Norway

45Legal and functional structure

45.1EMI Music Norway AS, has the following legal and functional structure:

(a)EMI Norway’s registered address and place of business is Karl Johans Gt. 12J, 0105 Oslo, Norway.

(b)EMI Norway is a wholly owned subsidiary of the EMI Group (see the EMI Norway Structure Chart at Annex K.1).

(c)Most of EMI Norway’s essential functions are performed within the business to be divested, although certain functions of EMI Norway are provided centrally by the EMI Group, such as manufacturing and distribution. In addition, EMI has in recent years rationalised its local operations across the Nordic region (see further details below).

(d)EMI Norway acts as a “territory cluster” with other Nordic countries (see further details below).

46Description of the asset

46.1EMI Norway includes, but is not limited to the tangible and intangible assets set out below:

46.2Rights to artist recordings, for which see:

(a)Confidential Annex 2 for details of the principal artist agreements signed to EMI Norway;

and

(b)Confidential Annex 3 for details of the turnover generated by EMI Norway’s principal artists.

46.3EMI Norway has a leasehold of the following property: Karl Johans Gt. 12J, 0105 Oslo, Norway and Archive, Grensen 3, Oslo, Norway (archives).

46.4EMI Norway’s roster includes current artists with future hit potential such as Sivert Høyem, Big Bang and Marit Larsen, as well as top-selling catalogue artists such as Lene Marlin and Morten Abel, one of Norway’s biggest pop artists, who has released many commercially successful studio albums.

46.5In addition UMG undertakes to use reasonable endeavours to include the rights to any local

753compilations signed to EMI Norway within the divestment package.

753For the avoidance of doubt, the EMI Norway divestment will not include any shareholdings in any other EMI Group companies outside Norway.

423

46.6The recorded music revenues generated by EMI Norway overall were […]* across the EEA in

2011/12. This asset generated recorded music revenues in Norway alone of […]* in 2011/12. Further,

these revenues only account for the recorded music revenues of these assets, and do not include

ancillary revenues, such as merchandising, synchronisation and neighbouring rights (public

performance). Please see Confidential Annex 4 for further details.

47Personnel

Group 1: Key Personnel

Name

Position

Employing entity

[…]*

Group 2: Other Personnel

47.1As noted above, EMI has in recent years rationalised its local operations across the Nordic region such that managers have both local and regional responsibilities. Set out below are the details of personnel and their connection with Norway:

[…]*.

47.2[…]*.

48Transitional Arrangements

48.1As noted above, EMI Norway currently benefits from general corporate support functions provided by the wider EMI Group. It is not anticipated that it would be necessary for arrangements to be put in place for the supply of any of these support services for a transitional period, but UMG or their Affiliated Undertakings commit to agree to supply any such services on terms to be agreed on an arm’s length basis, if determined necessary by the Trustee and required by the Commission at the request of the Purchaser, for a transitional period of […]*.

48.2The Purchaser will be granted licences to use the EMI trademark in connection with physical product manufactured prior to the date of completion of the sale of EMI Norway to the Purchaser to avoid recall or scrapping of existing product and will only be required to change artwork which includes the EMI trademark on future manufacturing runs. The Purchaser will also be licensed to use the EMI trademark as part of the corporate name of the relevant legal entity and for associated corporate purposes such as office signage and stationary during a transitional period of […]* following the date

of completion of the sale of EMI Norway to the Purchaser before effecting a corporate change of

name.

424

L.EMI Denmark

49Legal and functional structure

49.1EMI Music Denmark A/S, has the following legal and functional structure:

(a)EMI Denmark’s registered address and place of business is Falkoner Alle 7, DK-2000 Fredericksberg, Denmark.

(b)EMI Denmark is a wholly owned subsidiary of the EMI Group (see the EMI Denmark Structure Chart at Annex L.1).

(c)Most of EMI Denmark’s essential functions are performed within the business to be divested, although certain functions of EMI Denmark are provided centrally by the EMI Group, such as manufacturing and distribution. In addition, EMI has in recent years rationalised its local operations across the Nordic region (see further details below).

(d)EMI Denmark acts as a “territory cluster” with the other Nordic countries (see further details below).

50Description of the asset

50.1EMI Denmark includes, but is not limited to the tangible and intangible assets set out below:

50.2Rights to artist recordings, for which see:

(a)Confidential Annex 2 for details of the principal artist agreements signed to EMI Denmark; and

(b)Confidential Annex 3 for details of the turnover generated by EMI Denmark’s principal artists.

50.3EMI Denmark has a leasehold of the following property: Falkoner Allé 7, 3 sal. 2000 Frederiksberg.

50.4EMI Denmark’s roster includes current artists with future hit potential such as Christophe, Bjornskov, Turboweekend, When Saints Go Machine, Specktors and Freja Loeb, as well as top-selling catalogue artists such as Kim Larsen, possibly the most successful artist in Danish popular music history and for whom a new release is expected, and Soes Fengers.

50.5In addition UMG undertakes to use reasonable endeavours to include the rights to any local

754compilations signed to EMI Denmark within the divestment package.

754For the avoidance of doubt, the EMI Denmark divestment will not include any shareholdings in any other EMI Group companies outside Denmark.

425

50.6The recorded music revenues generated by EMI Denmark overall were […]* across the EEA in

2011/12. This asset generated recorded music revenues in Denmark alone of […]* in 2011/12.

Further, these revenues only account for the recorded music revenues of these assets, and do not

include ancillary revenues, such as merchandising, synchronisation and neighbouring rights (public

performance). Please see Confidential Annex 4 for further details.

51Personnel

Group 1: Key Personnel

Name

Position

Employing entity

[…]*

Group 2: Other Personnel

51.1As noted above, EMI has in recent years rationalised its local operations across the Nordic region such that managers have both local and regional responsibilities. Set out below are the details of personnel and their connection with Denmark […]*:

[…]*.

51.2See further the Denmark Personnel Structure Chart at Confidential Annex L.2.

51.3[…]*.

52Transitional Arrangements

52.1As noted above, EMI Denmark currently benefits from general corporate support functions provided by the wider EMI Group. It is not anticipated that it would be necessary for arrangements to be put in place for the supply of any of these support services for a transitional period, but UMG or their Affiliated Undertakings commit to agree to supply any such services on terms to be agreed on an arm’s length basis, if determined necessary by the Trustee and required by the Commission at the request of the Purchaser, for a transitional period of […]*.

52.2The Purchaser will be granted licences to use the EMI trademark in connection with physical product manufactured prior to the date of completion of the sale of EMI Denmark to the Purchaser to avoid recall or scrapping of existing product and will only be required to change artwork which includes the EMI trademark on future manufacturing runs. The Purchaser will also be licensed to use the EMI trademark as part of the corporate name of the relevant legal entity and for associated corporate purposes such as office signage and stationary during a transitional period of […]* following the date

of completion of the sale of EMI Denmark to the Purchaser before effecting a corporate change of

name.

426

M.EMI Spain

53Legal and functional structure

53.1EMI Music Spain SL (“EMI Spain”), has the following legal and functional structure:

(a)EMI Spain’s registered address and place of business is Alcalá 44, 3a planta, Madrid, 2814, Spain.

(b)EMI Spain is a wholly owned subsidiary of the EMI Group (see the EMI Spain Structure Chart at Annex M.1).

(c)Most of EMI Spain’s essential functions are performed within the business to be divested, although certain functions of EMI Spain are provided centrally by the EMI Group, such as manufacturing and distribution.

(d)EMI Spain acts as a “territory cluster” with EMI Portugal (see further details below).

54Description of the asset

54.1EMI Spain includes, but is not limited to the tangible and intangible assets set out below:

54.2Rights to artist recordings, for which see:

(a)Confidential Annex 2 for details of the principal artist agreements signed to EMI Spain;

and

(b)Confidential Annex 3 for details of the turnover generated by EMI Spain’s principal artists.

54.3EMI Spain has a lease of the following property: Alcalá 44, 3a planta, Madrid, 2814, Spain.

54.4EMI Spain’s roster includes current artists with future hit potential such as Macaco, Bebe, Luz Casal, Pablo Alborán, Jose Merce, Amaral, Estrella Morente, Luz Casal, Alvaro Laguna, Camela, Pablo Alboran, Wally Lopez and Abel the Kid, as well as top-selling catalogue artists such as

Héroes del silencio, El ultimo de la fila and Amaral.

54.5In addition UMG undertakes to use reasonable endeavours to include the rights to any local

755compilations signed to EMI Spain within the divestment package.

54.6The recorded music revenues generated by EMI Spain overall were […]* across the EEA and […]*

756globally in 2011/12.This asset generated recorded music revenues in Spain alone of […]* in

755For the avoidance of doubt, the EMI Spain divestment will not include any shareholdings in any other EMI Group companies (if any).

427

2011/12. Further, these revenues only account for the recorded music revenues of these assets, and do

not include ancillary revenues, such as merchandising, synchronisation and neighbouring rights

(public performance). Please see Confidential Annex 4 for further details.

55Personnel

Group 1: Key Personnel

Name

Position

Employing entity

[…]*

Group 2: Other Personnel

55.1EMI Spain team includes the following personnel:

[…]*.

55.2See further the Spain Personnel Structure Chart at Confidential Annex M.2.

55.3[…]*.

56Transitional Arrangements

56.1As noted above, EMI Spain acts as a “territory cluster” with EMI Portugal. […]*.

56.2[…]*.

56.3The Purchaser will be granted licences to use the EMI trademark in connection with physical product

manufactured prior to the date of completion of the sale of EMI Spain to the Purchaser to avoid recall

or scrapping of existing product and will only be required to change artwork which includes the EMI

trademark on future manufacturing runs. The Purchaser will also be licensed to use the EMI

trademark as part of the corporate name of the relevant legal entity and for associated corporate

purposes such as office signage and stationary during a transitional period of […]* following the date

of completion of the sale of EMI Spain to the Purchaser before effecting a corporate change of name.

756To avoid double counting, the figures in this paragraph do not include revenues generated by artists signed to EMI Spain that have already been accounted for elsewhere in the Divestment Business, i.e. Virgin Classics artists signed to EMI Spain.

428

EUC

AI-Powered Case Law Search

Query in any language with multilingual search

Access EUR-Lex and EU Commission case law

See relevant paragraphs highlighted instantly

Get Instant Answers to Your Legal Questions

Cancel your subscription anytime, no questions asked.Start 14-Day Free Trial

At Modern Legal, we’re building the world’s best search engine for legal professionals. Access EU and global case law with AI-powered precision, saving you time and delivering relevant insights instantly.

Links

Features Pricing Blog Terms of Use Privacy Policy Sitemap

Contact Us

Tivolska cesta 48, 1000 Ljubljana, Slovenia

+386 64 117 549

support@modernlegal.si

Similar Documents

We Found Similar Cases for You

UNIVERSAL MUSIC GROUP / EMI MUSIC

M.6458

Case No COMP/M.6458 - Universal Music Group/ EMI Music

REGULATION (EC) No 139/2004 MERGER PROCEDURE

Article 8 (2) Date: 21/09/2012

Brussels, 21/09/2012 C(2012) 6459 final

PUBLIC VERSION

COMMISSION DECISION

of 21/09/2012

TABLE OF CONTENTS

COMMISSION DECISION

of 21/09/2012

1. THE NOTIFICATION

2. THE PARTIES

3. THE PROPOSED CONCENTRATION

4. UNION DIMENSION

5. PROCEDURE

6. MARKET DEFINITION

6.1. Introduction to the recorded music industry

6.1.1. Background

6.1.2. The recorded music sector

6.1.2.1. Major record companies and independent record companies

6.1.2.2. The recorded music value chain

6.1.3. Other activities related to recorded music

6.1.3.1. Ancillary services

6.1.3.2. Music publishing

6.2. Relevant product markets

6.2.1. A&R

6.2.1.1. The Notifying Party's view

6.2.1.2. The Commission's assessment

6.2.2. Wholesale of recorded music

6.2.2.1. The Notifying Party's view

6.2.2.2. The Commission's assessment

6.2.2.2. The Commission's assessment

6.2.5.2. The Commission's assessment

6.3. RELEVANT GEOGRAPHIC MARKETS

6.3.1. A&R

6.3.1.1. The Notifying Party's view

6.3.1.2. The Commission's assessment

6.3.2. Wholesale of recorded music

6.3.2.1. The Notifying Party's view

6.3.2.2. The Commission's assessment

6.3.3. Retail of recorded music

6.3.3.1. The Notifying Party's view

6.3.3.2. The Commission's assessment

6.3.4. Music-related activities outside the recorded music sector

6.3.4.1. Ancillary services

6.3.4.2. Music publishing

7. THE COMPETITIVE ASSESSMENT IN THE RELEVANT MARKETS

7.1. HORIZONTAL NON-COORDINATED EFFECTS

7.1.1. Analytical framework

7.1.2. Market shares

Available market share sources in the recorded music sector

7.1.2.1. IFPI data

7.1.2.2. GfK data

7.1.2.3. OCC data

7.1.2.4. Platform data

7.1.2.5. Chart and radio airplay data

7.1.2.6. Market shares according to the different data sources

7.1.2.7. The Notifying Party's view

IFPI data

GfK and OCC data

Platform data

Chart and radio airplay data

7.1.2.8. The Commission's assessment

Introduction

The reference to market shares throughout the case

The relevance of the various data sources

7.1.3.2. The Commission's assessment

EMI is a major record company

7.1.4.Customers' limited possibilities of switching supplier

7.1.5.Likely effects of the proposed concentration

7.1.5.1.The Notifying Party's view

7.1.5.2.The Commission's assessment

7.1.5.3.Comparison of Universal and EMI's agreements with online platforms

7.1.5.4.Formal economic analysis of commercial conditions for digital customers

7.1.5.5.Results of the market investigation

7.1.5.6.Evidence from the merging parties