EUROPEAN COMMISSION DG Competition

(Only the English text is authentic)

MERGER PROCEDURE

REGULATION (EC) 139/2004

Article 8(2) Regulation (EC) 139/2004

Date: 19/05/2015

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.

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.

EUROPEAN COMMISSION

Brussels, 19.5.2015 C(2015) 3370 final

Public Version

COMMISSION DECISION

of 19.5.2015

(Only the English text is authentic)

TABLE OF CONTENTS

1. INTRODUCTION........................................................................................................ 7

2. THE OPERATION AND THE CONCENTRATION ................................................. 8

3. UNION DIMENSION ................................................................................................. 8

4. THE PROCEDURE ..................................................................................................... 8

5. REFERRAL REQUEST .............................................................................................. 9

6. RELEVANT MARKETS........................................................................................... 10

6.1. Retail supply of fixed voice services ......................................................................... 12

6.1.1. Product market definition ........................................................................................... 12

6.1.2. Geographic market definition .................................................................................... 13

6.2. Retail supply of fixed Internet access services .......................................................... 13

6.2.1. Product market definition ........................................................................................... 13

6.2.1.1. Internet access services with speeds below and above 30Mb/s ................................. 15

6.2.1.2. Segmentation by distribution technology................................................................... 16

6.2.1.3. Mobile compared to fixed broadband Internet access services .................................. 16

6.2.1.4. Residential and small business customers compared to Large business customers ... 17

6.2.2. Geographic market definition .................................................................................... 18

6.2.3. Conclusion.................................................................................................................. 18

6.3. Retail supply of mobile telecommunication services ................................................. 19

6.3.1. Product market definition ........................................................................................... 19

6.3.2. Geographic market definition .................................................................................... 19

6.4. Multiple play services ................................................................................................ 20

6.4.1. Product market definition ........................................................................................... 20

6.4.2. Geographic market definition .................................................................................... 22

6.4.3. Conclusion.................................................................................................................. 23

6.5. Wholesale market for call termination services on fixed networks ........................... 23

6.5.1. Product market definition ........................................................................................... 23

6.5.2. Geographic market definition .................................................................................... 24

6.6. Wholesale access and call origination on mobile networks ....................................... 24

6.6.1. Product market definition ........................................................................................... 24

6.6.2. Geographic market definition .................................................................................... 25

6.7. Wholesale mobile call termination market ................................................................ 25

6.7.1. Product market definition ........................................................................................... 25

6.7.2. Geographic market definition .................................................................................... 25

2

6.8. Wholesale broadband access services ........................................................................ 25

6.8.1. Product market definition ........................................................................................... 25

6.8.2. Geographic market definition .................................................................................... 26

6.9. Affected markets ........................................................................................................ 27

6.9.1. Horizontally affected markets .................................................................................... 27

6.9.2. Vertically affected markets ........................................................................................ 28

7. COMPETITIVE ASSESSMENT............................................................................... 29

7.1. The Spanish Telecommunication Markets ................................................................. 29

7.1.1. Providers of telecommunication services in Spain .................................................... 29

7.1.2. Next-Generation Access Networks and Very High-Broadband Services .................. 30

7.1.3. Convergent offers ....................................................................................................... 32

7.1.4. Upcoming changes in the regulation of the fixed telecommunication sector in Spain .................................................................................................................................... 34

7.1.5. The acquisition of ONO by Vodafone in summer 2014 ............................................ 36

7.1.6. Alternative Jazztel/Yoigo merger scenario ................................................................ 37

7.2. Retail markets including fixed Internet access services ............................................. 37

7.2.1. Overview of possible retail markets including fixed Internet access services ........... 38

7.2.2. Retail market for fixed Internet access services ......................................................... 40

7.2.2.1. Horizontal non-coordinated effects ............................................................................ 40

7.2.2.2. Market structure ......................................................................................................... 42

7.2.2.3. Removal of two important competitive forces ........................................................... 52

7.2.2.4. Change of the incentives of the merged entity to compete ........................................ 65

7.2.2.5. Likely reaction of competitors ................................................................................... 73

7.2.2.6. Quantitative analysis .................................................................................................. 78

7.2.2.7. Limited likelihood of sufficient entry ........................................................................ 80

7.2.2.8. Conclusion on horizontal non-coordinated effects on the market for fixed Internet access services ............................................................................................................ 80

7.2.3. Possible retail market for dual-play services ............................................................. 80

7.2.3.1. Removal of two important competitive forces ........................................................... 80

7.2.3.2. Change in the merged entity's incentives to compete ................................................ 81

7.2.3.3. Likely reaction of competitors ................................................................................... 81

7.2.3.4. Quantitative analysis of horizontal non-coordinated effects ...................................... 81

7.2.3.5. Limited likelihood of sufficient entry ........................................................................ 82

7.2.3.6. Conclusion horizontal non-coordinated effects on the possible market for dual-play services ....................................................................................................................... 82

7.2.4. Possible retail market for triple- and quadruple-play services ................................... 82

7.2.4.1. Removal of two important competitive forces ........................................................... 83

7.2.4.2. Change in the merged entity's incentives to compete ................................................ 84

7.2.4.3. Likely reaction of competitors ................................................................................... 85

7.2.4.4. Quantitative analysis of horizontal non-coordinated effects ...................................... 85

7.2.4.5. Limited likelihood of sufficient entry ........................................................................ 86

7.2.4.6. Future competitive strength of the Parties in light of the growing share of quadruple-play bundles ............................................................................................................... 86

7.2.4.7. Commission's assessment of the Notifying Party's efficiencies claims ..................... 93

7.2.4.8. Conclusion on horizontal non-coordinated effects on the possible market for triple- and quadruple-play offers .......................................................................................... 93

7.2.5. Possible retail market for triple-play services ............................................................ 93

7.2.6. Possible retail market for multiple play services ....................................................... 94

7.2.7. Quantitative analysis of horizontal non-coordinated effects ...................................... 94

7.2.7.1. Measurement of diversion ratios ................................................................................ 95

7.2.7.2. Measurement of margins and volumes ...................................................................... 99

7.2.7.3. The results of the calibrated merger simulation presented in the Statement of Objections ................................................................................................................ 100

7.2.7.4. The Notifying Party's main arguments in the response to the Statement of Objections ................................................................................................................ 101

7.2.7.5. The Commission's assessment of the main arguments by the Notifying Party ....... 101

7.2.7.6. The Commission's assessment of the expected price increases ............................... 105

7.2.7.7. Conclusions from the quantitative analysis.............................................................. 108

7.2.8. Limited likelihood of sufficient entry in the market ................................................ 108

7.2.8.1. Likelihood of sufficient entry into markets involving a fixed Internet access services component ................................................................................................................ 109

7.2.8.2. Likelihood of sufficient entry into multiple play markets involving a mobile component ................................................................................................................ 118

7.2.9. Impact of the proposed transaction on the deployment of NGA networks .............. 122

7.2.9.1. Plans for deployment of NGA networks in Spain .................................................... 122

7.2.9.2. Notifying Party's view .............................................................................................. 123

7.2.9.3. Commission's assessment ........................................................................................ 125

7.2.10. Commission's assessment of the Notifying Party's efficiencies claims ................... 137

7.2.10.1. Efficiencies stemming from allegedly increased fibre roll-out post-merger ........... 138

Verifiability ............................................................................................................................ 138

Merger specificity .................................................................................................................. 139

Benefit to consumers .............................................................................................................. 140

Conclusion on Efficiencies stemming from allegedly increased fibre roll-out post-merger . 140

7.2.10.2. Efficiencies from an improved offer of quadruple-play bundles ............................. 141

Notifying Party's view ............................................................................................................ 141

Commission's assessment ....................................................................................................... 141

7.2.10.3. Improved ability to serve consumers based on fibre network and to save xDSL access charges...................................................................................................................... 142

7.2.10.4. Efficiencies stemming from the elimination of double marginalisation of mobile services provided by Orange to Jazztel .................................................................... 142

7.2.11. Conclusion on markets including fixed Internet access services ............................. 146

7.3. Retail market for fixed voice services ...................................................................... 146

7.3.1. Notifying Party's view .............................................................................................. 147

7.3.2. Commission's assessment ........................................................................................ 147

7.3.3. Conclusion................................................................................................................ 151

7.4. Retail market for mobile telecommunication services ............................................. 151

7.4.1. Notifying Party's view .............................................................................................. 151

7.4.2. Results of the Market Investigation ......................................................................... 152

7.4.3. Commission's assessment ........................................................................................ 152,

7.4.4. Conclusion................................................................................................................ 153

7.5. Wholesale market for broadband access services

7.6. Vertical assessment

7.6.1. Wholesale market for fixed call termination services – Retail supply of fixed voice services and retail supply of mobile telecommunications services

7.6.1.1. The Notifying Party's view

7.6.1.2. The Commission's assessment

7.6.2. Wholesale market for mobile call termination services – Retail supply of fixed voice services and retail supply of mobile telecommunications services

7.6.2.1. Notifying Party's view

7.6.2.2. Commission's assessment

7.6.3. Wholesale market for access and call origination services on mobile networks – Retail supply of mobile telecommunications services

7.6.3.1. Notifying Party's view

7.6.3.2. Commission's assessment

8. COMMITMENTS

8.1. Procedure

8.2. Analytical framework

8.3. Assessment of the Commitments

8.3.1. First Commitments

8.3.1.1. Description

8.3.1.2. Results of the First Market Test

8.3.1.3. Commission's assessment of the First Commitments

8.3.2. Modified Commitments

8.3.2.1. Description

5

8.3.2.2. Results of the Second Market Test

8.3.2.3. Commission's assessment of the Modified Commitments

8.3.3. Final Commitments

8.3.3.1. Description

8.3.3.2. Commission's assessment

8.3.4. Conclusion

9. CONDITIONS AND OBLIGATIONS

6

COMMISSION DECISION

of 19.5.2015

(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 concentrations between undertakings, and in particular Article 8(2) thereof,

Having regard to the Commission Decision of 4 December 2014 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,

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

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

Whereas:

1. INTRODUCTION

(1) On 16 October 2014, the European Commission received a notification of a proposed concentration by which Orange SA ("Orange" or the "Notifying Party"), based in France, intends to acquire sole control over Jazztel p.l.c. ("Jazztel"), based in the United Kingdom, by way of a public bid (the "Proposed Transaction"). Orange and Jazztel are jointly referred to as the "Parties".

(2) Orange, through its fully owned subsidiary France Telecom España S.A.U which operates under its trade name Orange España, offers mobile telecommunication, fixed telephony and Internet access services to customers in Spain. Orange is the third largest Mobile Network Operator ("MNO") in Spain. For its provision of fixed Internet access and fixed voice services, Orange mainly relies on access to the local copper network of the incumbent telecom operator Telefónica de España, SAU

1 OJ L 24, 29.1.2004, p. 1 ("the Merger Regulation"). 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 407, 8.12 2015. p.14.

3 OJ C 407, 8.12 2015, p.16.

7

4 ("Telefónica")through local loop unbundling ("LLU") (via its proprietary backbone network, as well as technical equipment installed in a number of local LLU exchanges operated by the incumbent telecom operator Telefónica) and partially on its own Fibre to the Home ("FTTH") network, covering 800 000 building units ("BUs") as of end 2014. On the retail market for fixed Internet access services, Orange is the fourth largest player based on revenues and the third largest player based on subscribers.

(3) Jazztel offers fixed telephony, Internet access and mobile telecommunication services in Spain. Jazztel offers fixed line services through its proprietary network consisting of a backbone, equipment in local LLU exchanges of the incumbent telecom operator Telefónica and a FTTH network that covers 3 million households in Spain. Jazztel offers mobile telecommunication services as a mobile virtual network operator ("MVNO") on Orange's network. In the retail markets for fixed telephony and for fixed Internet access services, Jazztel is the third largest player based on revenues and the fourth largest player based on subscribers.

2. THE OPERATION AND THE CONCENTRATION

(4) Orange intends to acquire control over Jazztel by way of a voluntary cash tender offer over the entire share capital of Jazztel. Orange publicly announced that offer on 16 September 2014. On 16 October 2014, Orange requested authorization from the Spanish stock exchange authority Comisión Nacional del Mercado de Valores ("CNMV") to launch a takeover bid over Jazztel.

(5) The proposed concentration constitutes a concentration within the meaning of Article 3(1)(b) of the Merger Regulation.

3. UNION DIMENSION

(6) The undertakings concerned have a combined aggregate world-wide turnover of more than EUR 5 000 million (Orange: EUR 40 981 million; Jazztel: EUR 1 044 million). Each of them has a Union-wide turnover in excess of EUR 250 million (Orange: EUR 39 251 million; Jazztel: EUR 1 044 million), but they do not achieve more than two-thirds of their aggregate Union-wide turnover within one and the same Member State. The notified operation therefore has an EU dimension within the meaning of Article 1(2) of the Merger Regulation.

4. THE PROCEDURE

(7) Based on the results of the first phase market investigation ("Phase I Market Investigation"), the Commission raised serious doubts as to the compatibility of the transaction with the internal market and, on 4 December 2014, adopted a decision to initiate proceedings pursuant to Article 6(1)(c) of the Merger Regulation (the "Article 6(1)(c) Decision").

(8) The Notifying Party submitted its written comments to the Article 6(1)(c) Decision on 15 December 2014. On 14 January 2015 the Notifying Party submitted a report by its external consultant Analysys Mason (the "AM Report") on the subject of the

Parties' fibre roll-out plans before and after the merger which was followed on 16 January 2015 by a report of its economic consultants on the use of upward pricing pressure/ illustrative price rise ("UPP/IPR") analysis in the Article 6(1)(c) Decision (the "First UPP/IPR Report"). A supplementary report on UPP (the "Second UPP Report") was submitted on 4 February 2015.

(9) Based on the results of the second phase investigation ("Phase II Market Investigation") which supplemented the findings of the Phase I Market Investigation (jointly referred to as "Market Investigation"), the Commission addressed a Statement of Objections (the "Statement of Objections") dated 25 February 2015 to the Notifying Party pursuant to Article 18 of the Merger Regulation in which the Commission further substantiated its competition concerns.

(10) On 6 March 2015, before replying to the Statement of Objections, the Notifying Party submitted a first set of commitments in order to address the competition concerns identified in the Statement of Objections. The Commission carried out a market test of those commitments on 13 March 2015.

(11) The Commission issued a letter of facts to the Notifying Party on 10 March 2015 (the "Letter of Facts"). The Notifying Party submitted its comments on the Letter of Facts on 13 March 2015.

(12) The Notifying Party replied to the Statement of Objections (the "Reply to the Statement of Objections") on 11 March 2015. Masmovil, an interested third party, submitted comments on the Statement of Objections on the same date. Other interested third parties, namely Vodafone and Yoigo made submissions on 12 March 2015 and R Cable on 30 March 2015.

(13) An oral hearing took place on 16 March 2015 during which the Commission, the Parties and certain interested third parties, namely Vodafone, Yoigo and Masmovil presented their views on the competition concerns addressed in the Statement of Objections. The competition authorities of several Member States including the Spanish competition authority Comisión Nacional de los Mercados y la Competencia ("CNMC") attended the oral hearing and had the opportunity to ask questions.

(14) On 18 March 2015, the Commission adopted a decision pursuant to Article 11(3) of the Merger Regulation, following the Notifying Party's failure to respond to several requests for information pursuant to Article 11(2) of the Merger Regulation from the Commission. That decision had the effect of suspending the time limits for the Commission’s procedure as of 4 March 2015. The Notifying Party responded to the Request for Information on 27 March 2015 and the procedure was resumed on 28 March 2015.

(15) On 29 March 2015 the Notifying Party submitted a second set of commitments. A market test of those commitments was sent to market participants on 8 April 2015. The Notifying Party submitted another set of commitments on 6 April 2015 and a final set of commitments on 20 April 2015.

(16) The Advisory Committee discussed the draft of this Decision on 7 May 2015 and issued a favourable opinion.

5. REFERRAL REQUEST

(17) On 5 November 2014, the CNMC, submitted a request pursuant to Article 9(2)(a) of the Merger Regulation for full referral of the Proposed Transaction from the Commission to the Kingdom of Spain (the "Referral Request").

9

(18) In the Referral Request, the Kingdom of Spain considered that the Proposed Transaction threatens to significantly affect competition in the Spanish markets for: (i) fixed voice services; (ii) fixed Internet access services; (iii) mobile telecommunication services; and (iv) the possible market for multiple play services. Furthermore, the Kingdom of Spain considered that those markets are national in scope and present all the characteristics of distinct markets within Spain. Moreover, the Kingdom of Spain considered that it is best placed to deal with the Proposed Transaction.

(19) Given the Commission's decision of 4 December 2014 to initiate proceedings in this case, the Commission did not have to take a position on the Referral Request in Phase I pursuant to Article 9(4)(a) of the Merger Regulation.

(20) After the initiation of proceedings, the Kingdom of Spain submitted a reminder of its Referral Request on 19 December 2014, pursuant to Article 9(5) of the Merger Regulation.

(21) The Notifying Party was informed of Spain's Referral Request on 10 November 2014 and submitted its comments on 14 November 2014. Furthermore, the Notifying Party was informed of the Kingdom of Spain's reminder on 5 January 2015.

(22) The Notifying Party essentially submitted that the Proposed Transaction does not threaten to affect significantly competition and, therefore, that the legal requirements for a referral are not met. In any event, the Notifying Party submitted that the Commission is better placed to examine the case. In addition, a referral of the Proposed Transaction to Spain would cause an unnecessary administrative burden for the Notifying Party. Finally, it would generate additional legal uncertainties, which would be particularly disruptive given that Orange proposes to acquire Jazztel by way of a public bid.

(23) With a letter dated 9 January 2015 the Kingdom of Spain was informed of the Commission’s intention to refuse the Referral Request and given the opportunity to make known its views. In response, the Kingdom of Spain informed the Commission that it had no further comments.

(24) On 26 January 2015 the Commission adopted a decision rejecting the Referral Request of the Kingdom of Spain. The Commission considered that, at least as regards some of the markets discussed in the Referral Request, the Proposed Transaction threatens to significantly affect competition. Moreover, each of those markets presents all the characteristics of a distinct market within Spain. In any event, even if the conditions for a referral provided for in Article 9(2)(a) of the Merger Regulation are fulfilled with regard to the Proposed Transaction, in exercising its discretion the Commission took into account that it is well placed and is the most appropriate authority to investigate the Proposed Transaction for a number of reasons, including the need to ensure a coherent and consistent approach when addressing mergers in the converging fixed and mobile telecommunication sectors in different Member States falling into the Commission's competence and the fact that the Commission has developed significant expertise in the European tele-communication markets over the last years.

6. RELEVANT MARKETS

(25) The Proposed Transaction gives rise to certain horizontal overlaps between the Parties' activities in a number of relevant telecommunication services markets in Spain.

(26) In particular, at retail level, the Parties' activities overlap in:

(a) the supply of fixed voice services;

(b) the supply of fixed Internet access services; and

(c) the supply of mobile telecommunication services to end consumers.

(27) The Parties offer the services referred to in recital (26) as stand-alone products, as well as bundles of those services (so-called "multiple play" services).

(28) At wholesale level, Orange and Jazztel both operate in:

(a) the wholesale market for call termination on fixed networks;

(b) the wholesale market for mobile call termination;

(c) the wholesale provision of domestic call transit services on fixed networks;

(d) the wholesale broadband access services;

(e) the wholesale Internet connectivity market;

(f) the global telecommunication services ("GTS") market; and

(g) the wholesale international carrier services market.

(29) In addition, Orange is active in:

(a) the wholesale access and call origination on mobile networks;

(b) the wholesale provision of international roaming on mobile networks;

(c) the wholesale market for end-to-end calls; and

(d) the retail supply of Pay-TV services.

6.1. Retail supply of fixed voice services

(30) Fixed voice services include the provision of connection services or access at a fixed location or access to the public telephone network for the purpose of making and receiving calls and related services.

6.1.1. Product market definition

(31) The Notifying Party does not take a view on the exact product market definition.

(32) In previous decisions, the Commission considered drawing a distinction between residential and non-residential customers as well as between local or national and international calls but ultimately left the product market definition open. In the more recent Vodafone/ONO decision, the Commission considered the market for retail

the market for wholesale international carrier services is not affected and will not be further discussed in this Decision.

10 International roaming is a service which allows mobile subscribers to use their mobile handsets and SIM cards to make and receive calls, to send and receive text messages and to use other data services when abroad. There is no horizontal overlap between the Parties' activities as Jazztel does not operate a mobile network in Spain and does not provide international roaming services. Orange provides such international roaming services over its mobile network in Spain and has a [20-30]*% market share by revenue in 2013. As a consequence, the wholesale market for international roaming is not horizontally affected. The Notifying Party was not able to provide market shares for Orange in other European countries where Orange is active as an MNO, however given Orange's overall presence on the mobile telecommunications markets in France, Poland and Romania, the Notifying Party cannot exclude the possibility that Orange's market share in the wholesale provision of international roaming in these three countries could be above 30%. Therefore it cannot be entirely excluded that the wholesale markets for international roaming outside of Spain (in France, Poland and Romania) are vertically affected. The Commission discusses such potential vertical links in recital (124) below. For market definition, see Commission decision of 1 March 2010 in Case M.5650 – T-Mobile/Orange paragraphs 32-34.

11 Fixed telephone network operators package origination and termination services and provide communications providers without their own network a package which they can use to offer retail fixed voice services to consumers without the need to invest in infrastructure. The Notifying Party estimates Orange's market share in terms of revenue in 2013 at below 1%. Jazztel does not provide such services. As a consequence, the wholesale market for end-to-end calls is not affected and will not be discussed further in this Decision. For market definition, see Commission decision of 3 July 2012 in Case M.6584 – Vodafone/Cable & Wireless, paragraphs 19-21.

12 In previous decisions (Commission decision of 25 June 2008 in Case M.5121 – News Corp/Premiere, paragraphs 17-20 and Commission decision of 21 December 2010 in M.5932 – News Corp/BSkyB, paragraphs 92-99), the Commission distinguished between Free-to-Air TV (FTA) and Pay-TV services. Jazztel does not provide its own retail Pay-TV services since 2010 and its activities on the market are limited to being an agent of Distribuidora de Television Digital, S.A. ("DTS"). Orange has a market share on the Spanish Pay-TV market of [0-5]*% in terms of revenue and [0-5]*% in terms of subscribers in 2013. As a result, the market for retail supply of Pay-TV services is not affected and will therefore be further discussed in this Decision only in the context of multiple-play bundles involving a Pay-TV component.

13 Commission decision of 29 June 2009 in Case M.5532 – Carphone Warehouse/Tiscali UK, paragraphs 35 and 39; Commission decision of 29 January 2010 in Case M.5730 – Telefónica/Hansenet Telekommunikation, paragraphs 16-17; Commission decision of 20 September 2013 in Case M.6990 – Vodafone/Kabel Deutschland, paragraphs 126-131.

14 Commission decision of 2 July 2014 in Case M.7231 – Vodafone/ONO, paragraphs 26-28.

12

(37) Therefore the Commission considers that the relevant market for the retail provision of fixed voice services is national in scope.

6.2. Retail supply of fixed Internet access services

(38) Fixed Internet access services consist of the provision of a fixed telecommunications link enabling customers to access the Internet.

6.2.1. Product market definition

(39) A relevant product market comprises all those products and/or services which are regarded as interchangeable or substitutable by the consumers, by reason of the products' characteristics, their prices and their intended use. In determining the relevant market, the Commission assesses demand substitution by determining the range of products which are viewed as substitutes by the consumers.

(40) The Notifying Party submits that broadband Internet services with download speed of up to 30 Mb/s and very high broadband ("VHBB") services with download speed above 30 Mb/s in Spain are largely substitutable and therefore belong to the same

15 Commission decision of 29 June 2009 in Case M.5532 – Carphone Warehouse/Tiscali UK, paragraph 56; Commission decision of 20 September 2013 in Case M.6990 – Vodafone/Kabel Deutschland, paragraph 137.

16 See replies to Commission Questionnaire to competitors Q1 of 17 October 2014, question 11, and Questionnaire to customers Q2 of 20 October 2014, question 9.

17 Commission Notice on the definition of relevant market for the purposes of Community competition law ("Notice on Relevant Market"), paragraph 7 (OJ C372, 9.12.1997, p.5).

18 Notice on Relevant Market, paragraph 15.

19 Notice on Relevant Market, paragraph 20.

20 Notice on Relevant Market, paragraph 20.

13

(41) In previous decisions, the Commission considered, but ultimately left open, a subdivision of the market for the retail provision of fixed Internet access services by product type, distinguishing narrowband, broadband and dedicated access. The Commission also considered a distinction by broadband distribution technology, namely digital subscriber line ("DSL") and cable but ultimately found that the relevant product market included services offered both through DSL technology and cable. In that respect, in Hutchison 3G Austria/Orange Austria, the Commission found that fixed broadband services are not a substitute for mobile data services and therefore do not form part of the same product market.

(42) Finally, the Commission distinguished between residential and small business customers on the one hand and large business customers on the other, based on their different needs for Internet services. Large business customers require Internet access services based on higher performance in terms of both security and bandwidth. In addition, those customers could also require services such as dedicated access (leased lines) and virtual private network ("VPN") technology. In light of those differences, the Commission considered that retail broadband Internet access services for large business customers belong to a distinct product market. In more recent Commission decisions, that separate product market for large business customers is referred to as retail business connectivity market. The Commission considers that in line with previous Commission practice, for the purpose of this Decision, there is a separate market for retail business connectivity catering for large enterprise customers.

21 Such chain of substitution occurs where it can be demonstrated that although products A and C are not directly substitutable, product B is a substitute for both product A and product C and therefore products A and C may be in the same product market since their pricing might be constrained by the substitutability of product B.

22 Commission decision of 29 January 2010 in Case M.5730 – Telefónica/Hansenet Telekommunikation, paragraph 11 and Commission decision of 22 July 2009 in case M.5532 - Carphone Warehouse/Tiscali UK, paragraph 14.

23 Commission decision 29 June 2009 in Case M.5532 - Carphone Warehouse/Tiscali UK, paragraphs 17-21.

24 Commission decision of 12 December 2012 in Case M.6497 - Hutchison 3G Austria/Orange Austria, paragraph 57.

25 Commission decision of 29 June 2009 in Case M.5532 – Carphone Warehouse/Tiscali UK, paragraphs 24-27.

26 Commission decision of 29 January 2010 in Case M.5730 – Telefónica/Hansenet Telekommunikation, recital 6 and subsequent.

27 For business customers the estimated combined market share of Orange and Jazztel amounts to [10-20]*% for 2013 and [10-20]*% as of third quarter of 2014 in terms of number of subscribers. However this share covers all types of business customers: big enterprises, small and medium enterprises ("SMEs") and small offices/ home offices ("SoHo") as the Notifying Party was not able to provide an estimate of the market and relevant market shares for large business customers only. In any event, given the limited combined market share for all business customers and absent any indication that the combined market share in the market for business connectivity services, the Commission considers that the retail business connectivity market is not an affected market and will not be discussed further in this Decision.

14

6.2.1.1. Internet access services with speeds below and above 30Mb/s

(43) The Market Investigation did not yield clear results as to whether a distinction should be made between retail provision of Internet access services of up to 30 Mb/s and services above 30 Mb/s.

28

(44) Some respondents consider that speed is an important differentiating factor while other respondents state that Internet access services offered on a stand-alone basis (without being packaged with another service such as fixed voice telephony or television("TV")) compete only on price. The majority of respondents however consider that Internet access services below and above 30 Mb/s are substitutable. A respondent explains that the speed of what is considered an appropriate connection varies depending on the consumer's needs and usage patterns: for browsing, checking emails, publishing photos and videos (including video calls) speeds of up to 30 Mbps seem sufficient while for consumers with more intensive usage (multiple Internet users per household for example), including online gaming and watching High Definition TV ("HDTV") over the Internet, speeds above 30 Mb/s are necessary.

(45) Market participants also confirm that there is a price gap between Internet access services of different speeds with higher prices for higher speeds. The majority of the respondents that consider that substitutability is possible between Internet access services of speed of up to 30 Mb/s and Internet access services of speed above 30 Mb/s in case of a permanent 5% to 10% increase in the price of Internet access services of speed of up to 30 Mb/s, expect rather strong rates of substitution (ranging between 20% and 50%) in case both services are available to the consumers. Respondents explain that a potential price increase of the Internet access services of speed of up to 30Mb/s would reduce the current price gap between Internet access services of speed of up to 30 Mb/s and VHBB access services.

(46) The Commission considers that on the one hand, even if Internet access services below and above 30Mb/s may not be perfect substitutes for all consumers, there is possibly a chain of substitution between the different fixed Internet access services that is strongly influenced by the specific usage patterns of the individual consumer. The price points of retail broadband Internet offers also vary significantly depending on the corresponding speed of connection. That makes it difficult to determine the specific point in the chain of substitution where a potential rupture could occur. On the other hand, while Telefónica has an obligation to provide indirect wholesale access to its network, including its fibre network, the National Regulatory Authority ("NRA") limited the access remedies to a maximum bandwidth of 30Mb/s.

28 Respondents to the Phase II Market Investigation include competitors of the Parties in the retail provision of fixed Internet access products, MVNOs, resellers (such as supermarkets reselling telecommunication products) and business connectivity providers. In the Phase I Market Investigation, also business connectivity customers have been addressed.

29 See replies to Commission Phase II Questionnaire to Competitors of 23 December 2014, question 3.

30 See replies to Commission Phase II Questionnaire to Competitors of 23 December 2014, questions 9 and 10.

31 See replies to Commission Phase II Questionnaire to Competitors of 23 December 2014, question 12.

15

regime, the distinction between speeds above and below 30 Mb/s is not artificial, since different regulatory obligations apply and potentially impact the competitive situation.

(47) However, for the purpose of this Decision, the question whether fixed Internet access services of speeds below 30 Mb/s and above 30 Mb/s belong to the same product market can be left open, as the transaction would significantly impede effective competition irrespective of the precise market definition.

6.2.1.2. Segmentation by distribution technology

(48) The large majority of respondents to the Market Investigation confirm that the retail market for provision of fixed Internet access services comprises all the different distribution infrastructures: copper, hybrid fibre coaxial ("HFC") cable and FTTH. Consumers can select the service that corresponds to their preferences in terms of price, speed and availability. The geographic coverage of the different technologies varies and some technologies, such as FTTH and HFC cable, have more limited coverage compared to copper so users are able to choose only between the technologies that are present in their respective geographic area. The Market Investigation did not result in a clear picture as to whether consumers would switch to copper-based Internet access services in case of a 5% to 10% permanent price increase of Internet access services delivered over HFC cable or FTTH. Some respondents point out that currently there is a 20% to 30% price difference between the less expensive services delivered over copper and higher-priced services over another infrastructure.

(49) The Commission considers that the fact that all the different technologies currently compete on the market and given that consumers tend to base their choice of Internet access services on price and speed rather than on the exact delivery technology, do not justify a distinction between the different technologies available on the market. Therefore the Commission considers that for the purposes of this Decision there is no reason to divide the relevant market according to different infrastructures such as copper, HFC cable and FTTH.

6.2.1.3. Mobile compared to fixed broadband Internet access services

(50) As regards mobile broadband Internet access services the vast majority of respondents consider that fixed and mobile broadband services are not substitutable for a number of reasons, such as different uses (mobile broadband permits mobility but also access to mobile applications), different capacity and pricing models (mobile broadband unlike fixed broadband is almost always capped at a certain amount of data transferred and is also invoiced per amount of data traffic) as well as different speeds (only 4G/LTE mobile technology can provide comparable speeds to fixed broadband). In addition, respondents explain that customers demand both mobile and fixed broadband which demonstrates the complementary character of the two services. Some respondents consider that mobile broadband can in some cases

33 See conclusion in recital (415).

34 See replies to Commission Phase II Questionnaire to Competitors of 23 December 2014, question 4.

35 See replies to Commission Phase II Questionnaire to Competitors of 23 December 2014, question 5.

36 4G stands for "fourth generation of mobile telecommunication technology" and LTE stands for "Long Term Evolution". 4G/LTE is a mobile technology which increases the speed and capacity of the network and is adapted for improved voice quality and high speed data transmission from wireless devices (for example, to stream video, Internet TV and to use broadband Internet).

16

substitute fixed Internet access services but not the other way around. A small majority of respondents also expect that there will be very little to no substitution between fixed and mobile Internet access services in case of a permanent 10% price increase for fixed Internet services of speeds of up to 30 Mb/s while price of mobile broadband remains unchanged .

(51) Currently all four mobile network operators in Spain (Telefónica, Vodafone, Orange and Yoigo) are deploying 4G and overall 4G coverage extended to more than 50% of Spanish population in 2014. That large-scale 4G coverage however does not impair the existing differences in use, capacity, pricing and speed between mobile and fixed Internet access services that have been pointed out by the market respondents.

(52) The Commission has previously found that there is one overall market for retail mobile communication services which comprises voice calls, short messaging services ("SMS") and data services. The Commission has not previously found that there is a separate market for data-only services. In the more recent Vodafone /ONO case the Commission concluded that mobile broadband services are not yet substitutable to fixed Internet services in Spain but stated that there is a possibility of future convergence of the two services .

(53) In light of previous Commission's decisions and the differences between mobile and fixed Internet access services indicated by the respondents to the Phase II Market Investigation the Commission considers that a distinction between mobile and fixed Internet access services is justified at the present point in time and that for the purpose of this Decision the relevant product market is the retail market for provision of fixed Internet access services.

6.2.1.4. Residential and small business customers compared to Large business customers

(54) The large majority of respondents to the Market Investigation consider that the retail provision of Internet access services to residential and small business customers on the one hand and to large business customers on the other hand constitute two separate markets. In particular, most competitors confirm that operators provide similar services at similar prices both to residential customers and to Small and Medium Enterprises ("SME") and Small Offices or Home Offices ("SoHo"), while there are several differences between large enterprises and residential and small business customers. Large enterprises regularly require value added services (that is to say, VPN, hosting, etc.) and customer and technical support which has an impact on the final price paid for the provision of Internet access services .

(55) As mentioned in recital (42) the Commission previously considered that retail broadband Internet access services for large business customers belong to a distinct product market. In more recent Commission decisions, this separate product market for large business customers is referred to as retail business connectivity market.

46 47 product market. In more recent Commission decisions, this separate product market for large business customers is referred to as retail business connectivity market.

(56) In light of the differences in the needs of large business customers compared to the services required by residential and small businesses, the Commission considers that for the purpose of this Decision the relevant product market is the retail market for the provision of Internet access services to residential and small business customers, which is separate from the market serving large business customers with Internet access services.

6.2.2. Geographic market definition

(57) The relevant geographic market comprises the area in which the undertakings concerned are involved in the supply of and demand for products or services, in which the conditions of competition are sufficiently homogeneous and which can be distinguished from neighbouring areas because the conditions of competition are appreciably different in those areas .

(58) In line with the Commission's conclusions in the Vodafone/Kabel Deutschland case, the Notifying Party submits that the geographic scope of the relevant market in the present case is national. Correspondingly, nearly all respondents to the Phase I Market Investigation, both competitors and customers, consider that the geographic scope of the relevant market is national as competition takes place on a national basis and there are no geographically differentiated retail offers for fixed Internet access services in Spain .

(59) Aside from the three regional cable operators (Euskaltel, S.A., Telecable de Asturias, S.A.U. and R Cable, S.A.) all major retail telecommunication services providers in Spain operate networks with national reach and compete on the national market.

(60) For the reasons stated in recitals (58) and (59), for the purpose of this Decision, the Commission considers that the relevant geographic market is national in scope.

6.2.3. Conclusion

(61) The Commission concludes that, for the purposes of this Decision, fixed Internet access services to residential and small business customers, regardless of whether their speeds is less or more than 30Mb/s and irrespective of the distribution technology used for the delivery of those services to the end user, belong to the same relevant market for the retail supply of fixed Internet access services in Spain.

46 Commission decision of 29 June 2009 in Case M.5532 – Carphone Warehouse/Tiscali UK, paragraphs 24 -27.

47 Commission decision of 29 January 2010 in Case M.5730 – Telefónica/Hansenet Telekommunikation, recital 6 and subsequent.

48 Notice on Relevant Market f, paragraph 7.

49 Commission decision of 20 September 2013 in Case M.6990 - Vodafone/Kabel Deutschland, paragraph 197.

50 See replies to Commission Questionnaire to competitors Q1 of 17 October 2014, question 11, and Questionnaire to customers Q2 of 20 October 2014, question 9.

18

6.3. Retail supply of mobile telecommunication services

(62) Mobile telecommunication services to end customers include services for national and international voice calls, SMS (including MMS and other messages), mobile Internet data services and retail international roaming services.

6.3.1. Product market definition

(63) The Notifying Party submits that there is a single market for the retail provision of mobile telecommunication services.

(64) In previous decisions, the Commission considered that there is one single market for retail supply of mobile telecommunication services without segmenting the market according to the type of customer, whether business or private, the type of tariff (pre-paid or post-paid contracts), by network technology (2G/GSM, 3G/UMTS and 4G/LTE), or by type of service (Internet data services, voice and text services).

(65) In Hutchison 3G Austria/Orange Austria the Commission considered the segmentations referred to in recital (64), but eventually concluded that there is a single market for the provision of mobile telecommunications services to end customers.

(66) The large majority of respondents to the Commission's Market Investigation consider that the provision of mobile telecommunication services to end customers constitutes a single market. Several respondents note the current trend toward convergence of different services (fixed voice telephony, Internet, mobile telecommunication services and Pay-TV) taking place on the Spanish market .

(67) For the purpose of this Decision, the Commission considers that the exact definition of the product market can be left open as the Proposed Transaction does not raise competition concerns under any alternative product market definition considered.

6.3.2. Geographic market definition

(68) In line with previous Commission decisions, the Notifying Party submits that the market for mobile telecommunications services is national in scope.

(69) The large majority of respondents to the Market Investigation in the present case confirmed that the relevant geographic market is national and corresponds to the territory of Spain. A respondent noted that in Spain there are no geographically differentiated retail offers .

(70) The Commission notes also that customers are only able to obtain mobile communication services from licensed network operators and licenses for the operation of a mobile network in Spain are granted on a national basis.

(71) The Commission considers the relevant market to be national in scope.

6.4. Multiple play services

(72) Multiple play services comprise a bundle of two or more services to end-consumers: among the fixed telephony services, fixed Internet access services, mobile telecommunication services and TV services. Such packaged services can consist of so called dual-play, triple-play or even quadruple-play packages comprising some or all of those services.

(73) Table 1 summarises the relative size of the different product types offering fixed broadband, either as stand-alone or as part of a bundle .

(74) The most common product types including a fixed broadband component are (i) the double play product type, consisting in fixed voice and fixed broadband ("2p"), and (ii) the triple and quadruple-play product types, consisting in double play plus mobile and possibly TV ("3p/4p"). Other product types included in the market for fixed internet access, such as fixed broadband only and fixed broadband plus mobile products only represent approximately 1% of the retail market for fixed internet access services.

Table 1: Shares of subscribers of product types

Source: Commission calculation based on data obtained from operators

6.4.1. Product market definition

(75) The Notifying Party submits that multiple play services are not a relevant market but rather a trend that affects different relevant markets, such as retail mobile telecommunications services, retail fixed voice services, retail fixed Internet access services and TV services. The Notifying Party argues that operators are free to decide whether to offer their services either bundled or not. Furthermore, the Notifying Party points out that consumers would also have the choice of cancelling an individual component or individual components of a bundled service and procuring it or them from an alternative operator.

(76) In previous decisions, the Commission ultimately left open the question as to whether there is a market for multiple play services that is separate from the markets for each of the components of the bundles. If multiple play services would not form a separate market, they would consist of two or more separate products.

(77) A majority of respondents to the Phase I Market Investigation stated that multiple play services should not be considered a distinct market from the retail supply of the

57

1p means Fixed Broadband, 2p means Fixed Broadband and Fixed Voice, 3p/4p means Fixed Broadband, Fixed Voice, Mobile and possibly TV.

It was not possible to split 3p from 4p because certain operators could not provide data at such a level of disaggregation.

Commission decision of 16 June 2011 in Case M.5900 – LGI/KBW, paragraphs 183-186; Commission decision of 25 January 2010 in Case M.5734 – Liberty Global Europe/Unitymedia, paragraphs 43-48; Commission decision of 3 July 2012 in Case M.6584 – Vodafone/Cable&Wireless, paragraphs 102-104; Commission decision of 20 September 2013 in Case M.6990 – Vodafone/Kabel Deutschland, paragraph 261.

20

respective unbundled components (that is to say, Internet access services only, fixed telephony services only, mobile communication services and TV only). However, most competitors supported the view that bundled services play a significant role in the residential segment in Spain. They explained that bundles allow end-consumers to obtain better prices in a price-driven market and to simplify the customer's purchasing decision in exchange for less flexibility in selecting one or part of the services in the bundle.

(78) The majority of the respondents that provided a detailed answer to the relevant question in the Phase II Market Investigation point to the price of the bundle as indeed being the main driver of consumer choice for packaged services over a combination of stand-alone services. Convenience and additional services, such as TV for example, are other important factors influencing the consumer choice. However, price is designated as the most important driver of convergence by far.

(79) From a supply side's perspective, bundled services are a way of increasing customer loyalty (reducing the customer churn) as it is more onerous for a customer to change several services at once. Some respondents to the Phase II Market Investigation consider that there are certain cost savings in the provision of multiple play services, given that the different services are provided over the same infrastructure, such as single installation cost, lower commercial cost and some reduction in the administrative costs. At the same time, the emergence of bundled services makes it more difficult for retailers that are not capable of offering such services to compete successfully in the market .

(80) Currently, bundles that contain a TV component, have a relatively minor share in Spain as the most popular bundles, according to CNMC data, are dual-play (fixed Internet access services plus fixed voice services) and triple-play convergent bundles (fixed Internet access services, fixed voice services plus mobile telecommunications services). Market participants explain that the relatively lower penetration of Pay TV services is mainly due to the availability of attractive Free to Air ("FTA") TV content and the relatively expensive price of Pay-TV. Few respondents consider that bundles comprising a Pay-TV component may gain higher uptake in the coming two or three years facilitated by the increased availability of high-speed Next Generation Access ("NGA") networks and possibly by availability of content, including premium content, at reasonable price.

(81) The Commission notes that due to the different services included in different multiple play bundles, instead of one possible market for multiple play products, there could also be several markets for different multiple play products.

(82) Whereas dual-play services are provided over the same, fixed infrastructure, triple- or quadruple-play services are provided over more than one infrastructure. Triple-play services combine fixed and mobile infrastructures and quadruple-play services

60

See replies to Commission questionnaire to competitors Q1 of 17 October 2014, question 10 and Questionnaire to customers Q2 of 20 October 2014, question 8.

61

See replies to Commission questionnaire to competitors Q1 of 17 October 2014, question 5 and Questionnaire to customers Q2 of 20 October 2014, question 5.

62

Table 2:

Dual-Play Triple-Play Quadruple-Play Fixed    Mobile   TV 

(83) If multiple play services were not to form a separate market from their respective components, the market for fixed Internet access services would include all fixed Internet access services irrespective of whether they are sold as stand-alone services or as part of a bundle with other services. In Table 2, such assessment would include the entire horizontal line "Fixed" in addition to a small number of fixed Internet access services that are provided on a stand-alone basis without being bundled in at least a dual-play product.

(84) If multiple play services were to form separate product market(s), the market for fixed Internet access services would include only stand-alone services sold or (for the supply-side considerations discussed in recital 21) as part of a dual-play bundle but not Internet access services sold as part of triple- or quadruple-play bundles.

(85) A possible multiple play market could be a general market for multiple play products or separate markets for dual-, triple- and quadruple-play products. In the table above, such definition would require separate assessments of the vertical columns for "dual-play" and for "triple-play" and "quadruple-play". One could also regard dual-play services as not belonging to a separate market while grouping triple-play services and quadruple-play services together. Therefore, there are in total five possible markets for multiple play services: (i) a general market for all multiple play services, (ii) separate markets for dual-play services, (iii) triple-play services, (iv) quadruple-play services, and (v) a market combining triple- and quadruple-play services .

(86) In any event, the Commission considers that, for the purpose of this Decision, the question as to whether multiple play services constitute a separate product market (including all dual, triple and quadruple-play combinations) or several separate product markets (including selected combinations of bundled components, for example dual-play only, triple-play only, triple and quadruple-play only, quadruple-play only) distinct from each of the underlying telecommunications services or whether such services combine several separate products can be left open, as the Commission concludes in recital (415) that the transaction would significantly impede effective competition irrespective of the conclusion regarding this point.

6.4.2. Geographic market definition

(87) The Notifying Party does not take a view as to the geographic definition of the possible market for multiple play services.

64

As regards the involved infrastructures, both triple- and quadruple-play offers combine services provided over fixed and mobile infrastructure.

65

Other market that could be thought of would comprise of the combination of fixed voice with pay-TV services or of fixed Internet access with pay-TV services. Furthermore, there could be a market for the combination of fixed voice, fixed Internet access and pay-TV services. However, as Jazztel offers no TV services of its own and Orange has only a very limited market share in such services, the Decision will not further discuss these possible markets.

22

(88) Previously, the Commission considered that a possible market for triple-play services comprising fixed voice services, fixed Internet access services and Pay-TV services would be national in scope. Equally, in its Vodafone / Kabel Deutschland decision, the Commission considered the possible triple-play market comprising fixed voice services, mobile telecommunications services and fixed Internet access services as national in scope. However, ultimately, in both cases the exact geographic market definition was left open.

(89) The large majority of respondents to the first phase market investigation considered that, in the event that the provision of multiple play services constitutes a separate market, the relevant geographic scope of such market would be national. This would be because, first, the main operators offering those services are active at national level, second, the offers do not differ by territory (with small exceptions for Pay-TV, where local TV channels might be included depending on the region) and, third, each of the components of the multiple play offers is offered individually at national level so that packaging the services in a bundle would not change the geographic scope of its components.

(90) In any event, the Commission considers that, for the purpose of this Decision, the question as to whether the geographic scope of the possible market for multiple play services – and in particular, dual-play or triple-play services – is national or regional can be left open, as the Commission concludes in recital (415) that the transaction would significantly impede effective competition irrespective of the conclusion on this point.

6.4.3. Conclusion

(91) The Commission concludes that, for the purposes of this Decision, the question whether multiple play services constitute a separate market or separate markets distinct from each of the underlying telecommunications services can be left open, as the Commission considers that the transaction would significantly impede effective competition irrespective of the conclusion on this point. For the same reason, the question of whether such market or markets are national or regional in scope can be left open.

6.5. Wholesale market for call termination services on fixed networks

(92) Call termination is the service provided by a network operator on the supply side to other network operators on the demand side, whereby a call originating in a demand side operator's network is delivered to a user in the supply side operator's network allowing users of different networks to communicate with each other. Call termination is a wholesale service that is resold or used as an input for the provision of downstream retail telephony services.

6.5.1. Product market definition

(93) In line with the Commission's previous decisions, the Notifying Party submits that wholesale termination on each individual fixed network constitutes a separate

66

Commission decision of 16 June 2011 in Case M.5900 – LGI/KBW, paragraphs 183-186.

67

Commission decision of 20 September 2013 in Case M.6990 – Vodafone/Kabel Deutschland, paragraphs 263-265.

23

The majority of the respondents to the market investigation in the present case agreed with this product market definition .

(94) The Commission considers that there is no substitute for call termination on each individual network since the operator transmitting the outgoing call can reach the intended recipient only through the operator of the network to which the recipient is connected.

(95) The Commission considers that, as regards wholesale call termination services, termination on each individual fixed network constitutes a separate product market.

6

6.5.2. Geographic market definition

(96) In the Notifying Party's view the wholesale market for call termination services on fixed networks is national in scope. In its previous decisions, the Commission concluded that the geographic scope of the market is national .

(97) All respondents to the market investigation in the present case take the view that the geographic market is national in scope.

(98) Moreover, given that the geographic dimension of fixed networks is usually limited to national borders owing to regulatory barriers, for the purpose of this Decision, the Commission considers that the relevant geographic market is national.

6.6. Wholesale access and call origination on mobile networks

(99) MNOs provide wholesale access and call origination services which enable operators without their own network, that is to say MVNOs, to provide their own retail mobile telecommunication services to end customers. There are different types of MVNOs, ranging from so-called full MVNOs that typically own some of the core infrastructure but do not own radio network access or spectrum, to service providers and pure resellers of MNOs.

71

6.6.1. Product market definition

(100) The Notifying Party did not express its view on the exact product market definition.

(101) In the past, the Commission considered that wholesale network access and call origination services belong to the same relevant product market. The majority of respondents to the market investigation in the present case agreed with that product market definition .

73

(102) For the purpose of this decision, the Commission considers that the exact definition of the product market can be left open.

68 Commission decision of 1 March 2010 in Case NM.5650 – T-Mobile/Orange UK, paragraph 37; Commission decision of 3 July 2012 in Case M.6584 – Vodafone/Cable & Wireless, paragraphs 22- 23 and Commission decision of 20 September 2013 in Case M.6990 – Vodafone/Kabel Deutschland, paragraphs 117.

69 See replies to Commission questionnaire to competitors Q1 of 17 October 2014, question 7.

70 Commission decision of 1 March 2010 in Case M.5650 – T-Mobile/Orange, paragraph 38; Commission decision of 3 July 2012 in Case M.6584 – Vodafone/Cable & Wireless, paragraph 24; Commission decision of 20 September 2013 in Case M.6990 – Vodafone/Kabel Deutschland, paragraph 119.

71 Both service providers and resellers rely on the same network elements and services provided by MNOs (including SIM cards) with the main difference being that resellers rely on the MNO also for the provision of equipment (handsets, etc.).

72 Commission decision of 1 March 2010 in Case M.5650 – T-Mobile/Orange, paragraph 27; Commission decision of 2 July 2014 in Case M.7231 – Vodafone/ONO, paragraph 56.

73 See replies to Commission Questionnaire to competitors Q1 of 17 October 2014, question 7.

6.6.2. Geographic market definition

(103) In line with previous Commission's decisions, the Notifying Party submits that the relevant geographic scope of the market for wholesale access and call origination on mobile networks is national .

(104) All respondents to the market investigation in the present case take the view that the geographic market is national in scope.

(105) In light of the above, for the purpose of this decision, the Commission considers that the relevant geographic market is national.

6.7. Wholesale mobile call termination market

(106) When someone calls a mobile phone connected to a different network, that call is terminated on the network of the receiving mobile phone. In order for a retail mobile services provider to be able to provide calls to a different network, it must purchase wholesale terminations services on other networks by means of interconnection agreements between the various network operators.

6.7.1. Product market definition

(107) The Notifying Party submits that, in line with previous Commission's decisions, call termination on each individual mobile network constitutes a separate relevant product market. Nothing in the course of the investigation in this case justifies a deviation from previous Commission findings. Therefore, the Commission considers, in line with its previous decisions, that termination on each individual mobile network constitutes a separate product market.

6.7.2. Geographic market definition

(108) The Notifying Party submits that the definition of the market is national in scope. The Commission has consistently found that the markets for wholesale mobile call termination are national in scope, given that licenses to operate a mobile telecommunication network are granted for national territories. Nothing in the course of the investigation in the present case justifies a deviation from previous findings. Therefore, the Commission considers that the wholesale markets for mobile call termination are national.

6.8. Wholesale broadband access services

(109) Wholesale broadband access includes different types of access that allow Internet service providers to provide services to end consumers. It comprises physical access at a fixed location such as LLU, of non-physical or virtual network access such as bitstream access and of the resale of fixed broadband offerings of a third party, most often the incumbent operator.

6.8.1. Product market definition

(110) The Notifying Party did not express its view on the exact product market definition.

74 Commission decision of 1 March 2010 in Case M.5650 – T-Mobile/Orange, paragraph 31; Commission decision of 2 July 2014 in Case M.7231 – Vodafone/ONO, paragraph 59.

75 Commission decision of 3 July 2012 in Case M.6584 - Vodafone/Cable & Wireless, paragraphs 47- 48; Commission decision of 20 September 2013 in Case M.6990- Vodafone/Kabel Deutschland, paragraph 111-122.

76 Commission decision of 1 March 2010 in Case M.5650 – T-Mobile/Orange, paragraph 38; Commission decision of 2 July 2014 in Case M.7231 – Vodafone/ONO, paragraph 76.

(111) In the market investigation for the Carphone Warehouse/Tiscali UK decision, the Commission found significant differences in the characteristics, price, performance and service between the following different types of wholesale broadband access:

(a) LLU: unbundled (shared) access to metallic loops of the local access network in a number of local telecommunications exchanges (in particular in urban areas), as it is the most cost-efficient way for alternative operators to provide differentiated retail broadband services;

(b) Bitstream: indirect or virtual wholesale access in the form of "bitstream", where a wholesale capacity of bidirectional data transmission to and from the end-customer is provided;

(c) Resale of the fixed incumbent's broadband offering: reselling the line of the owner of the telecommunication network. That third form of obtaining wholesale Internet access does not enable the alternative provider to differentiate its service from that of the wholesale provider, but allows it to offer a broadband product with no investment and at mostly variable costs.

(112) In Carphone Warehouse/Tiscali UK, as well as in the more recent Vodafone/Kabel Deutschland, the Commission, however, ultimately left open the exact product definition of the market for wholesale broadband access services.

(113) The majority of the respondents to the market investigation in the present case confirmed that the market definition, considered by the Commission in the Carphone Warehouse/Tiscali UK case, is still valid. A number of competitors consider that each of the different type of broadband access services constitutes a separate market. A respondent states that LLU, bitstream and resale of the fixed incumbent's broadband offering should be treated as separate markets in line with the concept of "investment ladder", used by national regulators and telecom operators referring to the gradual nature of investments by alternative operators in telecommunications. The different types of wholesale broadband access services constitute different steps of the ladder where LLU requires more investments by the alternative provider but grants it more autonomy when creating a fixed retail offer. Reselling the fixed incumbent's broadband offering does not allow the reseller to differentiate the product but allows the operator to offer a fixed broadband product without significant investments .

(114) In any event, for the purpose of this Decision, the Commission considers that the exact product market definition can be left open as the Proposed Transaction does not raise competition concerns under any alternative product market definition considered.

77 Commission decision of 29 June 2009 in Case M.5532 - Carphone Warehouse/Tiscali UK, paragraphs 28- 34.

78 Commission decision of 20 September 2013 in Case M.6990 – Vodafone/Kabel Deutschland, paragraph 161.

6.8.2. Geographic market definition

(115) The Notifying Party did not express its view on the exact geographic market definition. In Carphone Warehouse/Tiscali UK, while there were indications supporting a national scope of the market, the Commission ultimately left open the exact geographic market definition .

(116) The Spanish regulator NRA had considered segmenting geographically the national wholesale broadband market based on the number of active players in each area covered by each of Telefónica's local switches differentiating between zones considered competitive areas and zones considered non-competitive. The final decision of the regulator, however, declared the geographic market to be national in scope with no further geographic segmentation .

(117) In any event, for the purpose of this Decision, the Commission considers that the exact geographic market definition can be left open.

6.9. Affected markets

(118) The Proposed Transaction gives rise to the affected markets referred to in recitals (119) and (121).

6.9.1. Horizontally affected markets

(119) The Spanish retail market for fixed Internet access services, possible separate Spanish retail markets for multiple play services, the Spanish retail market for fixed voice services, the Spanish retail market for mobile telecommunication services and the Spanish wholesale market for broadband access services are horizontally affected since the Parties’ combined shares on those markets or on a potential definition of those markets exceed 20%.

(120) As regards the affected retail market for fixed voice services, as well as the retail market for mobile telecommunications services, the Commission considers that competition concerns are not likely to arise. Post-transaction the combined entity will continue to face strong competition on the retail market for fixed voice services from both Telefónica and Vodafone, who have market shares in terms of number of subscribers of 48% and 22% respectively. Furthermore as the same underlying infrastructure necessary for the distribution of fixed Internet access services is used also for the delivery of fixed voice services to consumers any commitments aimed at addressing competition concerns identified on the retail market for fixed Internet access services would remedy also any potential competition concern on the retail market for fixed voice services. On the retail market for fixed mobile services Jazztel operates as an MVNO with a rather limited market share so the Proposed Transaction will have no impact on Orange's position as MNO on the market. In addition the merged entity will continue to face competition on the market from the other three MNOs in Spain: Telefónica, Vodafone and Yoigo.

80 Commission decision of 29 June 2009 in Case M.5532 - Carphone Warehouse/Tiscali UK, paragraphs 48-53.

81 CMT Decision of 22 January 2009. The draft CNMC proposal on the wholesale broadband regulation currently under public consultation (see section 7.1.4 for more details) foresees a geographical segmentation in competitive and non-competitive areas. See CNMC "Consulta Pública Relativa A La Definición Y Análisis Del Mercado De Acceso Local Al Por Mayor Facilitado En Una Ubicación Fija Y Los Mercados De Acceso De Banda Ancha Al Por Mayor, La Designación De Operadores Con Poder Significativo De Mercado Y La Imposición De Obligaciones Específicas (EXPEDIENTE ANME/DTSA/2154/14/Mercados 3a 3b 4)", published on 19 December 2014, available at http://www.cnmc.es/Portals/0/Ficheros/Telecomunicaciones/Consultas_Publicas/Consulta_cnmc/20141219_ProyectoMedida.pdf.

6.9.2. Vertically affected markets

(121) The Spanish wholesale market for fixed call termination services and the Spanish markets for retail supply of fixed voice services and retail supply of mobile telecommunications services, the Spanish wholesale market for mobile call termination services and the Spanish markets for retail supply of fixed voice services and retail supply of mobile telecommunications services as well as the Spanish wholesale market for access and call origination services on mobile networks and the Spanish market for retail supply of mobile telecommunications services are vertically affected.

(122) As regards the affected wholesale markets for fixed call termination services, for mobile call termination services, and for wholesale broadband services, the Commission considers that competition concerns are not likely to arise from the Proposed Transaction as both the markets for wholesale fixed call termination and wholesale mobile call termination are subject to an ex ante regulation by the national telecom regulator CNMC and on the wholesale broadband services market the Parties have only a marginal market share of less than 1%. In addition, the respondents to the Commission's Market Investigation have not raised any specific concerns regarding those markets.

(123) The Commission considers that competition concerns are unlikely to arise from the Proposed Transaction also on the wholesale market for access and call origination services on mobile networks. Orange's position on that market will not change as a result of the Proposed Transaction as there is no reduction in the number of MNOs and a sufficient number of providers of wholesale access and call origination services on mobile networks will remain post-transaction. Furthermore Orange is subject to a regulatory obligation imposed by the CNMC to meet all reasonable requests for access.

(124) As regards the wholesale market for international roaming Orange is active on that market in other European countries where it operates as an MNO but the Notifying Party was not able to provide the corresponding market shares of Orange in those countries. Given Orange's overall presence on the mobile telecommunications markets in France, Poland and Romania, the Notifying Party submitted that it cannot exclude the possibility that Orange's market share in the wholesale provision of international roaming in those three countries could be above 30%. As there is a vertical link between the Spanish retail market for mobile telecommunications services and the wholesale international roaming services in France, Poland and Romania, where Orange operates as an MNO it cannot be excluded that the wholesale markets for international roaming services in those three countries would be affected by the Proposed Transaction if Orange's market share in any of those wholesale markets is indeed above 30%. However the Commission considers that even if there were any such affected wholesale international roaming markets competition concerns are unlikely to arise. The Regulation (EC) No 717/2007 on roaming imposes a price cap on the wholesale prices that MNOs can charge their roaming customers. At the wholesale level, that Regulation caps prices for operators from Member States for voice roaming charges, SMS and data. In addition, MNOs must meet all reasonable requests for wholesale roaming access. MNOs are,

82 Regulation (EC) No 717/2007 of the European Parliament and of the Council of 27 June 2007 on roaming on public mobile telephone networks within the Community and amending Directive 2002/21/EC.

28

therefore, prevented from refusing access to their network and from charging excessive termination fees. Therefore the wholesale markets for international roaming are not discussed any further in this Decision.

7. COMPETITIVE ASSESSMENT

(125) The Commission considers that the Proposed Transaction would lead to significant impediment to effective competition in the market for the retail supply of fixed Internet access services and the possible markets for dual-play services, triple-play services, triple- and quadruple-play services or for multiple play services. The reasons for such conclusion are presented in the sections 7.1 and 7.2 of this Decision.

7.1. The Spanish Telecommunication Markets

(126) The Proposed Transaction combines two national providers of telecommunication services in Spain. Its effects on competition have to be assessed in the light of the structure and developments of the telecommunication markets in Spain.

7.1.1. Providers of telecommunication services in Spain

(127) In Spain, there are currently four providers of fixed telecommunication services at a national level as well as four mobile network operators that offer mobile telecommunication services of which Telefónica, Vodafone and Orange operate a fixed as well as a mobile network.

(128) Telefónica is the incumbent provider of telecommunication services in Spain and still by far the largest operator. Telefónica operates a nation-wide copper network for fixed voice and xDSL services, a large FTTH network as well as a mobile network. As the incumbent provider, Telefónica is subject to mandatory access regulation and has to offer certain wholesale access to its copper and FTTH network.

(129) Despite the access regulation, Telefónica is still the largest provider of fixed voice, fixed Internet access and mobile telecommunication services. Telefónica's market share in the market for fixed voice services in Spain amounted to more than 49% in terms of subscribers and more than 71% in terms of revenues in 2014. At the same time, the market shares of Telefónica for fixed Internet access services have been almost 45% in terms of subscribers and more than 42% in terms of revenues. For mobile telecommunication services, Telefónica in 2013 achieved market shares of 34% in terms of subscribers and of 36% in terms of revenues.

(130) After the acquisition of the cable operator Ono in Summer 2014, Vodafone became the second largest provider of telecommunication services in Spain. The integration of the two businesses of Vodafone and Ono is still ongoing. Vodafone operates a mobile network and offers fixed voice and fixed Internet access services through its own backhaul network combined with wholesale access to the "last mile" of Telefónica's copper network. Under a co-deployment agreement with Orange, Vodafone also started to roll-out an own FTTH network. After the acquisition of Ono, Vodafone also owns Ono's cable network.

(131) In the market for fixed voice services, the combined market shares of Vodafone and Ono in 2014 amounted to 21% in terms of subscribers and to 14% in terms of revenues. In the market for fixed Internet access services, Vodafone and Ono had, in 2014, combined market shares of approximately 21% in terms of subscribers and of 18% in terms of revenues. In the market for mobile telecommunication services, Vodafone achieved market shares of 22.5% in terms of subscribers and 27% in terms of revenues in 2014.

(132) Orange is the third integrated provider of telecommunication services. It operates a mobile network and offers fixed voice and fixed Internet access services through its own backhaul network and wholesale access to the "last mile" of Telefónica's copper network. Recently, Orange started to roll-out an own fibre network under a co-investment agreement with Vodafone.

(133) In the market for fixed voice services, Orange's market shares in 2014 amounted to 13% in terms of subscribers and 2.4% in terms of revenues. Orange provided Internet access services in 2014 to 15% of all subscribers and earned a share of 16% of the total revenues in that market. Orange's market share for mobile services has been 23% in terms of subscribers and 20% in terms of revenue.

(134) The fourth nationwide provider of fixed voice and fixed Internet access services is Jazztel. Jazztel offers xDSL services through its own backhaul network and wholesale access to Telefónica's copper network as well as through its FTTH network. The largest part of that FTTH network was built under a co-deployment agreement with Telefónica. Jazztel offers mobile services under a wholesale agreement with Orange bundled with its fixed products.

(135) In 2014 Jazztel's market share in the market for fixed voice services amounted to 11% in terms of subscribers and 5% in terms of revenue in 2014. Its shares in the market for fixed Internet access services have been 12% in terms of subscribers and 14% in terms of revenue. Jazztel's position in the market for mobile telecommunication services is very modest. Jazztel achieved market shares of 2% in subscribers and of 1% in terms of revenue in 2013.

(136) Yoigo is the smallest of the four mobile network operators in Spain with 2014 market shares for mobile telecommunication services of approximately 7% in terms of subscribers and 5% in terms of revenue. Yoigo does not offer fixed voice or Internet access services. Instead, Yoigo resells Telefónica's fixed products since October 2013 .

(137) In addition to the nationwide providers of telecommunication services referred to in recitals (128) to (136), the three cable companies Euskaltel, Telecable de Asturias and R Cable have each sizeable market shares in a certain region of Spain. Euskaltel is active in the Basque country, Telecable serves customers in Asturias and R Cable conducts its business in Galicia. All three regional cable companies offer fixed voice services and fixed Internet access services through their own fixed networks. In addition, each of them has a wholesale contract for mobile services and offers bundles of fixed and mobile services to consumers.

7.1.2. Next-Generation Access Networks and Very High-Broadband Services

(138) As of the third quarter of 2014, there were 12.8 million active lines in Spain for the provision of fixed Internet access services. Currently, the main technology (70%)

85 Yoigo launches a fixed+mobile offering which is almost identical to that of Telefónica "Yoigo lanza una oferta fijo+móvil casi igual a la de Telefónica", in Expansion dated 14 October 2013, http://www.expansion.com/2013/10/14/empresas/tmt/1381761241.html, Doc ID2651.

30

price. In its most recent analysis, the CNMC reports that in 2013 44.6% of all households currently demanding fixed and mobile telecommunications services procure such services from a single operator. That figure was 37.2% in 2012. Moreover, the possible market for triple-play services is broader, given that almost all the households which have subscribed to a fixed Internet access service are also customers of mobile telecommunications services. According to an internal document of Jazztel of 2013, the market for convergent fixed-mobile services could attain 60% of the total number of subscribers in 2015, a result which was achieved already by the end of the third quarter of 2014, showing the success for this type of products.

7.1.4. Upcoming changes in the regulation of the fixed telecommunication sector in Spain

(154) The CNMC is currently reviewing the regulatory framework for wholesale access to Telefónica's fixed networks (markets 3a, 3b and 4 of the Commission Recommendation on Relevant Markets of 2014). On 19 December 2014, the regulatory division of the CNMC, the Directorate for Telecommunications and Audiovisual Sector ("DTySA") published a draft regulation according to which Telefónica would be obliged to, inter alia, grant wholesale access to its FTTH network in all of Spain except for nine municipalities which are covering approximately 16% of the Spanish population. The terms of such wholesale access are not yet proposed by the CNMC.

93 (155) Market participants were able to comment on that draft until 23 February 2015 . After the public consultation, the CNMC Council will have to approve the proposal for a regulation. Once the proposed regulation is formally notified to the Commission

90 Presentation by Jazztel: "2013 - 2017 Business Plan", dated 15 March 2013, Doc ID 39, slide 13.

91 Commission Recommendation of 9 October 2014 on relevant product and service markets within the electronic communication sector susceptible to ex ante regulation in accordance with Directive 2002/21/EC of the European Parliament and of the Council on a common regulatory framework for electronic communication networks and services (Text with EEA relevance) – Brussels, 9.10.2014 – C(2014) final.

92 CNMC Public Consultation concerning the definition and market analysis of the market for wholesale local access provided at a fixed location and of the markets for wholesale broadband access, of the designation of operators with significant market power and of the imposition of specific obligations (RECORD ANME / DTSA / 2154/14 / Markets 3a 3b 4) "Consulta Pública Relativa A La Definición Y Análisis Del Mercado De Acceso Local Al Por Mayor Facilitado En Una Ubicación Fija Y Los Mercados De Acceso De Banda Ancha Al Por Mayor, La Designación De Operadores Con Poder Significativo De Mercado Y La Imposición De Obligaciones Específicas (EXPEDIENTE ANME/DTSA/2154/14/Mercados 3a 3b 4)", published on 19 December 2014, available at http://www.cnmc.es/Portals/0/Ficheros/Telecomunicaciones/Consultas Publicas/Consulta cnmc/20141219_ProyectoMedida.pdf.

34

94 in accordance with the procedure pursuant to Article 7a of Directive 2002/21/EC , the Commission will need to agree to the proposed regulation. A decision is due one month after notification, extendable by two months if a Phase 2 consultation is opened.

(156) Changes in the regulatory environment are a relevant element to be taken into consideration in the Commission's assessment, if they can be reasonably predicted . The draft regulatory proposal by the CNMC indicates that the proposed changes follow the guidelines of the Recommendations on Relevant Markets and Costing and Non-discrimination methodologies. Notwithstanding the relevance of potential regulatory changes the Commission considers however that the outcome of the CNMC's market review and the content and effect of future regulatory provisions cannot be reasonably predicted at this stage. Firstly, the draft proposal might be amended in response to the public consultation or the proceedings with the Commission. Secondly, the proposed regulation will be adopted only after the Commission would have decided on the Proposed Transaction. Thirdly, important elements, such as wholesale prices, are not part of the current proposal. For those reasons, the CNMC's regulatory proposals cannot be relied upon for the purpose of the competitive assessment.

(157) Independently of the proposed review of the regulatory framework for wholesale access to Telefónica's fixed networks referred to recital (154) above, on 30 March 2015, the CNMC published a public consultation concerning a potential reduction of the current capacity prices related to the regulated indirect access product called Nuevo servicio Ethernet de Banda Ancha ("NEBA") . The consultation proposes a reduction by approximately 45% the NEBA capacity prices. Other elements of the regulated NEBA offer such as the monthly recurrent prices of the service and prices related to the Points of Interconnection are not considered by the draft proposal. As to the timing, the consultation was published on 30 March and is open for 20 days. It is unknown if and when the revised regulation will enter into force. Moreover, the outcome of the consultation and the effect of a potential overhaul of the NEBA regulation on the market conditions of a potentially revised NEBA regulation are uncertain at this stage. For those reasons, the CNMC's proposal to modify the NEBA regulation cannot be relied upon for the purpose of the competitive assessment.

94 Directive 2002/21/EC of the European Parliament and of the Council of 7 March 2002 on a common regulatory framework for electronic communications networks and services (Framework Directive), OJ L 108 of 24.4.2002, p.33-50.

95 According to paragraph 9 of the Commission's Guidelines on the assessment of horizontal mergers under the Council Regulation on the control of concentrations between undertakings, (2004/ C 31/03), generally, the conditions existing at the time of the merger constitute the relevant comparison for evaluation the effects of a merger. Only in some circumstances, the Commission may take into account future changes to the market that can reasonably be predicted.

96 Commission recommendation 2014/710/EU of 9 October 2014 on relevant product and service markets within the electronic communications sector susceptible to ex ante regulation in accordance with Directive 2002/21/EC of the European Parliament and of the Council on a common regulatory framework for electronic communications networks and services (Recommendation on Relevant Markets), OJ L 295, 11.10.2014, p.79.

97 Commission Recommendation 2013/466/EU of 11 September 2013 on consistent non-discrimination obligations and costing methodologies to promote competition and enhance the broadband investment environment (Recommendation on non-discrimination and costing methodologies), OJ L 251, 21.9.2013, p.13.

98 CNMC, Consulta pública sobre la revisión del precio de la capacidad en PAI del servicio de banda ancha mayorista NEBA, MTZ 2014/1840. NEBA stands for Nuevo servicio Ethernet de Banda Ancha.

7.1.5. The acquisition of ONO by Vodafone in summer 2014

(158) The Spanish telecommunication markets are further affected by the recent acquisition of ONO by Vodafone in summer 2014. After reviewing that merger, the Commission concluded that the transaction would not raise competition concerns, because the Parties' activities are largely complementary: ONO's main field of business have been Pay-TV and fixed telecommunication services, whereas Vodafone is mainly active in mobile telecommunication services. The Commission further found that the impact of the transaction on the markets for fixed and mobile telecommunication services is likely limited because the combined entity would continue to face significant competition from other market players, such as Telefónica, Orange and Jazztel.

99 (159) The reasons for clearing the transaction between Vodafone and ONO cannot be applied to this case. First of all, the combined market shares of Orange and Jazztel are higher than the ones of Vodafone and ONO. The combined market shares of Vodafone and ONO in the market for fixed Internet access have been less than 25%, the threshold under which, according to the Commission's Guidelines on the assessment of horizontal mergers (the "Horizontal Merger Guidelines") , a concentration is generally presumed to be compatible with the internal market. Based on subscribers and according to data provided by the Notifying Party, the combined market shares of Vodafone and ONO amounted to 21.4% in terms of subscribers and to 18.4% in terms of revenues in the period between the first and the third quarter in 2014. The respective combined market shares of Orange and Jazztel amount to 26.9% and 30.0% in the same period.

(160) Next, the complementary of the two businesses was reflected in the market shares of Vodafone and ONO as the position of Vodafone in the market for fixed Internet access services was rather weak compared to ONO whereas the share of Orange or Jazztel in that market are closer to each other. While Vodafone achieved only a market share in terms of subscribers of 8.9% and in terms of revenues of only 4.8% in the period between the first and the third quarter in 2014, the respective market shares of Orange and Jazztel were 15.0% or 11.9% in terms of subscribers and 15.7% or 14.3% in terms of revenues in the same period.

(161) Furthermore, the Commission took into account the continued competition by other market players which included in particular competition from the other three remaining nationwide operators Telefónica, Orange and Jazztel. After the Proposed Transaction, there would be only three nationwide operators left.

(162) Finally, the Commission also notes that the current investigation showed that the combined gross add shares of Vodafone and ONO have been and still are much lower than the ones of Orange and Jazztel (see recitals (310) et seq.).

99 As stated in the respective press release of the Commission dated 2 July 2014.

100 Guidelines on the assessment of horizontal mergers under the Council Regulation on the control of concentrations between undertakings ("Horizontal Merger Guidelines"), paragraph 18 (OJ C31, 5.2.2004, p.5).

101 Commission decision of 2 July 2014 in Case M.7231 – Vodafone/ONO.

7.1.6. Alternative Jazztel/Yoigo merger scenario

(164) Before Orange announced its intent to acquire Jazztel, Jazztel was in the process of negotiating a potential acquisition of Yoigo, the smallest MNO in Spain, with Yoigo’s shareholder, TeliaSonera. According to the Horizontal Merger Guidelines, the effects of the proposed merger have to be compared with the conditions that would have prevailed without the merger and such alternative scenario might include future changes in the market that can be reasonably predicted. Therefore, the competitive effects of the Proposed Transaction might have to be compared to the potential Jazztel/Yoigo transaction (the "Jazztel/Yoigo Scenario").

(165) The Jazztel/Yoigo Scenario would have combined the only operator of a mobile network that does not own a fixed network and the only operator of a nationwide fixed network that does not own a mobile network. The Spanish market would have featured four integrated fixed-mobile players (Telefónica, Vodafone, Orange and Jazztel/Yoigo). According to Orange's own assessment in its internal documents,

103 [reference to the Notifying Party's internal business documents]* .

(166) The Notifying Party submits that the acquisition of Yoigo by Jazztel cannot be reasonably predicted and therefore it is not an appropriate alternative scenario for the current investigation.

(167) Jazztel submitted a document setting out […]*. In that document, Jazztel describes the negotiations and further procedural steps that took place and those that were still outstanding with regard to a potential acquisition of Yoigo by Jazztel. [conclusion relating to a strategic decision of Jazztel]*.

(168) The information currently available to the Commission indicates that the negotiations between Jazztel and Yoigo had not reached a stage where the transaction appears to be reasonably predicted in the absence of the present transaction. Therefore, the Commission bases its competitive assessment of the Proposed Transaction in this Decision on the current competitive conditions and does not take into account the situation that would have resulted from the Jazztel/Yoigo Scenario.

7.2. Retail markets including fixed Internet access services

(169) Orange and Jazztel are two of the four nationwide operators offering retail fixed Internet access services. Both Parties provide fixed-line telecommunication services mainly through wholesale access to Telefónica's unbundled local copper loops (xDSL network). Furthermore, each of the two operators has started deploying its own NGA network.

(170) In Spain, fixed Internet access services are being sold most often as part of multiple play bundles. As discussed in section 6.4, multiple play offers either constitute a market separate from the ones of their individual components or can be seen as consisting of several separate products. After providing an overview over possible retail markets including fixed Internet access services in section 7.2.1, the Decision analysis the different relevant markets in sections 7.2.2 to 7.2.6.

(171) Section 7.2.7 sets out the results of a calibrated merger simulation which quantifies the extent to which the elimination of competition between the Parties will generate,

(172) Section 7.2.8 shows that entry into the market involving fixed Internet access services by potential new competitors is unlikely due to high barriers to entry. The Commission considers that there could be a risk if the current regulatory framework does not allow for a timely entry of competing players in the current market conditions. Wholesale (physical) access to the local loop of Telefónica's copper network would require considerable investments which could not be recouped in time before the likely transition of the market to NGA networks. Likewise, the conditions for regulated bitstream access to Telefónica's network, in particular as regards its FTTH network, do not allow an alternative player to compete with attractive commercial terms on the downstream markets. The Commission acknowledges that the regulatory framework is in the process of being amended, but deems that without the adoption of a final text of such new regulation, the future changes are not sufficiently certain to change the findings, namely that there currently are high barriers to entry (see section 7.1.4).

(173) Section 7.2.9 discusses the impact of the Proposed Transaction on the future roll-out of NGA networks by the Parties. The Commission concludes that contrary to the Notifying Party's claim, the merger is unlikely to lead to any significant pro-competitive effects arising from reaching additional households that Orange or Jazztel or both would not have covered in a standalone scenario.

(174) Finally, section 7.2.10 assesses efficiency claims brought forward by the Notifying Party. As a consequence of the Commission's analysis of the likely future roll-out of NGA networks, any efficiency claims based on additional roll-out are not sufficiently verifiable. As to the loss of competition in areas where the NGA networks of the Parties would have overlapped in the future, the Commission notes that such a potential anti-competitive effect cannot be established with the required degree of certainty. The Commission acknowledges in section 7.2.10.4 that the Proposed Transaction will likely create efficiencies based on elimination of double marginalisation of mobile services provided by Orange to Jazztel.

7.2.1. Overview of possible retail markets including fixed Internet access services

(175) If multiple play services were not to form a separate market from their respective components, the market for fixed Internet access services would include all fixed Internet access services irrespective of whether they are sold as standalone services or as part of a bundle with other services. By reference to Table 2, such assessment would include the entire horizontal line "Fixed".

(176) Narrower markets have to be assessed if multiple play products were to be regarded as markets separate from their respective components. Such market could be a general market for multiple play products or, even narrower, separate markets for dual-, triple- and quadruple-play products. Finally, one could also consider that dual-play services do not belong to a separate market while grouping triple-play services and quadruple-play services together. Therefore, there are in total five possible markets for multiple play services: (i) a general market for all multiple play services,

104As regards the involved infrastructures, both triple- and quadruple-play offers combine services provided over fixed and mobile infrastructure.

market for triple-play services are Telefónica and Vodafone. Post-transaction, Telefónica, Vodafone and the merged entity will account for a substantial part of the triple-play market. For the reasons set out in recital (182), the assessment of that possible market is not further detailed in this section.

(182) On a possible market combining triple and quadruple-play services, the Parties' combined market share would account for [30-40]*% (Orange: [10-20]*%; Jazztel: [10-20]*%) at the end of 2014. Section 7.2.4 assesses the impact of the Proposed Transaction of the possible market combining triple- and quadruple-play services. It concludes that the Proposed Transaction gives rise to competitive concerns on the possible market for triple- and for quadruple-play services. Consequently, the Proposed Transaction would give rise to even more serious concerns on a possible market for triple-play offers, as both Parties mainly offer triple-play services. Therefore, the assessment of the possible market for triple-plays services is not carried out separately in more detail.

(183) The Commission will first assess, in section 7.2.2, a market that includes all Internet access services, whether provided as part of multiple play bundles or not. For the assessment of competitive concerns related to fixed Internet access services, this is the broadest possible market definition as it includes products that would otherwise be part of separate markets. This Decision then discusses whether the outcome of the competitive assessment related to fixed Internet access products would change in case multiple play services were to form one or several separate markets in section 7.2.3 to section 7.2.6.

7.2.2. Retail market for fixed Internet access services

(184) The retail market for fixed Internet access services includes all fixed Internet access services irrespective of whether such services are sold standalone or as part of a bundle with other services.

7.2.2.1. Horizontal non-coordinated effects

(185) The Commission considers that the proposed merger would lead to a significant loss of competition as it reduces the number of fixed telecommunication operators active at national level from four to three. This loss of competition concerns mainly competition in fixed Internet access services based on wholesale access to the xDSL network of Telefónica, which currently represents the large majority of all fixed Internet access services, as in particular Orange has currently only a very limited own FTTH network.

(186) As regards future competition in fixed Internet access services based on NGA networks in those geographic areas which in the absence of the merger would have been covered by the separate NGA networks of both Jazztel and Orange, the loss of competition caused by the Proposed Transaction cannot be established with the required degree of certainty. According to the Commission's estimates, the Proposed Transaction will lead to a reduction of choice for NGA customers in an area comprising approximately 20% to 30% of all BUs in Spain .

(187) Section 7.2.2 shows in detail that the Proposed Transaction would cause a higher degree of concentration in the market for fixed Internet access services. The market

structure following the merger would consist of the incumbent operator, Telefónica, with a market share of 42% in revenues and 45% in number of subscribers, followed by Orange with market shares of 30% in revenues and 27% in subscribers and by Vodafone with market shares of 18% in revenues, and 21% in subscribers. The combined market share of the three regional cable operators in terms of subscribers amounts to 4.75% in 2014.

(188) While the Proposed Transaction would not lead to the creation or strengthening of a (single) dominant position of the merged entity, the Merger Regulation explicitly recognises that mergers in oligopolistic markets involving the elimination of important competitive constraints that the merging parties previously exerted upon each other together with a reduction of competitive pressure on the remaining competitors, could, even if there is little likelihood of coordination between the members of the oligopoly, also result in a significant impediment to effective competition .

(189) Section 7.2.2.3 shows that Orange and Jazztel are currently important competitive forces in Spain in terms of prices and innovation. The two companies are perceived by the majority of market players (competitors and customers) as innovative and price aggressive players. This is further confirmed by the internal documents of Orange and Jazztel, as well as the percentage of total new contracts in these markets gained by each of Orange and Jazztel in 2013 and 2014 which is well above the percentage to be expected based on their market shares. This is a clear indication of the attractiveness of the two companies and their importance for the competitive dynamics in the market.

(190) If the merger takes place, Orange would gain a significant customer base. Section 7.2.2.4 sets out that this is likely to reduce the Parties' current incentives to compete aggressively, likely leading to price increases by the merged entity. In addition, the review of internal documents of Orange has shown that the merged entity would no longer have the incentive to compete as aggressively as each of the two companies before the merger. In fact, in its internal assessment of the impact of the merger, Orange expects retail prices to be higher following a merger with Jazztel compared to a scenario without such merger.

(191) Section 7.2.2.5 shows that the loss of competition between Orange and Jazztel would also reduce the competitive pressure on the remaining operators who would not have the incentive to compensate for such loss. In its internal documents, Orange expects Telefónica and Vodafone to behave rather rationally and not to assume an aggressive stance. This is further supported by public statements of representatives of Telefónica and Vodafone: both companies expressed the view that, in the future, they will likely accommodate price increases rather than oppose them. Furthermore, smaller operators without significant own network infrastructure or with a limited regional presence do not have the ability and/or incentive to counter possible price increases of the merged entity.

(192) Section 7.2.7 sets out the results of a calibrated merger simulation presented in section 7.2.2.6 for the retail market for fixed Internet access services.

(193) Finally, section 7.2.9 discusses the impact of the Proposed Transaction on the future roll-out of NGA networks by the Parties. The Commission concludes that contrary to the Notifying Party's claim, the merger is unlikely to lead to any significant pro-

competitive effects arising from reaching additional households that Orange and/or Jazztel would not have covered in a standalone scenario. Consequently, any efficiency claims based on additional roll-out are not sufficiently verifiable. As to the loss of competition in areas where the NGA networks of the Parties would have overlapped in the future, the Commission notes that such a potential anti-competitive effect cannot be established with the required degree of certainty.

7.2.2.2. Market structure

Overall market

(194) The Notifying Party submits that the Proposed Transaction does not give rise to any competition concerns because of the limited combined market share of the Parties post-transaction, which will not exceed 30%. Moreover, the merged entity will continue to face strong competition post-transaction from several other players active at national level such as the incumbent operator Telefónica as well as Vodafone, and smaller players. Other players' combined share is 8.6% by revenue and includes (i) the three regional cable operators operating in 3 out of the 17 Spanish regions and (ii) largely service-based fringe competitors (mainly relying on bitstream or resale of fixed telecommunication products) such as Masmovil or Pepephone. In addition, the Notifying Party submits that the businesses of Orange and Jazztel are largely complementary as Orange is mainly active in mobile telecommunication businesses whereas Jazztel is mainly active in fixed Internet access services.

(195) Orange and Jazztel offer retail fixed Internet access services in Spain mainly by relying on regulated direct wholesale access to Telefónica's copper network through LLU (necessitating own infrastructure in local exchanges as well as backhaul and core networks). Furthermore, the Parties compete to a limited extend through their own infrastructure (deployment of own NGA networks) and through regulated indirect wholesale access to Telefónica's copper and fibre networks through bitstream or bitstream-like access.

(196) With regard to the market for the provision of retail fixed Internet access services in Spain, the Notifying Party only provided market share data for an overall market for fixed Internet access that also includes large enterprise customers. However, as set out in recital (55) above, large enterprise customers are part of a separate product market. Orange and Jazztel have smaller market shares in the "business segment" – that includes, as defined by the Notifying Party, Small and Medium Enterprises ("SMEs") and Small offices/Home offices ("SoHo") as well as large enterprise customers (see Table 10 below at recital (208)). Therefore, the Parties' market shares in the fixed Internet access market defined as including only residential and smaller enterprise customers are likely higher than those that are set out in recital (197) and in Table 8, and closer to the market shares in the residential segment (see Table 10 below).

(197) In an overall market for fixed Internet access services comprising also large business customers, in 2013, Orange had a market share of 13.7% in terms of revenue and

of revenues, the Parties increased their market shares by 7.9% (an increase of 99%) and 6.2% (an increase of 76%) respectively.

(200) In contrast, Telefónica has experienced a strong decline in the same period, losing 12.5% of market share by revenues (a decrease of 23%) and 7.8% by subscribers (a decrease of 15%). The other national operators, Vodafone and ONO, had a stable performance. Vodafone has slightly increased its market shares by revenues and subscribers, but only by 0.2% and 1.8% respectively. ONO has been losing market share by subscribers, but slightly strengthened its position in terms of its market share by revenues. The other operators (mainly the three regional cable operators), did not experience any significant change in the last years regarding their combined position in the overall market.

(201) As regards the concentration levels, the post-transaction Herfindahl-Hirschman Index ("HHI") as regards the overall retail market based on revenue market shares would be 3157, with a delta of 461. The corresponding HHI based on subscriber market shares would be 3273, with a delta of 370. These concentrations levels significantly exceed the safe harbour provided for by paragraphs 19 and 20 of the Horizontal Merger Guidelines, and provide an indication of the oligopolistic structure of the retail market for the provision of fixed Internet access services in Spain.

(202) As regards the argument of the Notifying Party, that the businesses of the Parties are largely complementary, the Commission notes that while that holds true in the market for mobile telecommunication services where Orange operates an own mobile network and Jazztel utilises Orange's network to offer mobile telecommunication services as an MVNO, both Parties are significant and successful competitors on the market for fixed Internet access services. On that market, instead of complementing each other, the activities of the Parties overlap.

(203) The Commission considers that the Proposed Transaction will create a market with three nationwide fixed Internet access providers and lead to a market with all the characteristics of an oligopolistic market. An oligopolistic market structure is characterised by a limited number of sizeable firms. Because the behaviour of one firm has an appreciable impact on the overall market conditions, and thus indirectly on the situation of each of the other firms, oligopolistic firms are interdependent .

(204) The market share increments brought about by the Proposed Transaction in the retail market for fixed Internet access services and in particular in its residential segment are analysed in recitals (208) to (231).

(205) Moreover, on this national market for retail provision of fixed Internet access services, the activities and therefore the competitive constraint of regional cable

117The Notifying Party did not provide market shares of small competitors other than Telefónica, Orange, Jazztel and Vodafone ("Other players") on this market. Therefore, the HHI was calculated assuming that Other players constitute a single entity, which may slightly overstate the HHI. However, even under a more conservative approach (i.e., not including such players at all in the calculation of the overall HHI), the proposed transaction would still significantly exceed the safe harbour provided for by the Horizontal Merger Guidelines (post-transaction HHI level of 3082 in revenue and 3236 in number of subscribers – the delta figures would not change).

(211) Table 11 shows how, both Telefónica and ONO are losing segment shares in term of subscribers, while Orange and Jazztel are the two operators experiencing the highest growth with an increment of 3.6% and 2.6% respectively, as compared to 2011, followed by Vodafone with an increment of 2.5%, as compared to 2011 as well.

Internet access services with speeds below and above 30 Mb/s

(212) At the end of 2014, there were 12.8 million active lines in Spain for the provision of fixed Internet access services, and by the end of 2013 there were eight hundred thousands more. recitals (146) to (148) set out that the main technology (73%) used by Spanish customers is a Digital Subscriber Line ("DSL"). Only a minority of consumers (around 26% of active lines) use NGA network lines. Moreover, a third of HFC cable or FTTH customers have, or chose to pay only for speeds below 30 Mb/s. Therefore, across all technologies, only about 20% of all Spanish broadband customers currently have Internet access services that reach speeds higher than 30 Mb/s.

Segment/Market with speeds below 30 Mb/s

(213) In the segment or market of fixed Internet access services with download speeds of up to 30 Mb/s (representing currently 80% of the overall market in terms of active lines), Orange had a 17.2% share in terms of number of active broadband lines for the first half of 2014. Jazztel's share for the same period amounted to 13.3%, resulting in a combined market share of 30.5% post-merger.

(214) In its response to the Article 6(1)(c) Decision, the Notifying Party submits that the share mentioned in recital (213) has to be put into perspective since it is quickly diminishing as services provided over copper networks are being replaced with services provided over fibre networks and customers are rapidly migrating from Internet access services with download speeds of up to 30 Mb/s to VHBB services. Furthermore, Telefónica España CEO's Luis Gilpérez Miguel has already announced on 20 November 2014 the company's intention to dismantle, in about five years, a significant part of the copper network over which the Parties provide their service .

(215) The Commission noted in its Article 6(1)(c) Decision that fibre take-up in Spain so far appears to be relatively modest as the active FTTH lines represent only around 10% of all active broadband lines at the end of the third quarter of 2014. The Notifying Party submits that the fibre uptake in Spain for the first half of 2014 stands close to 20% and expects a significant increase in the fibre uptake in the next 3 years ([30-40]*% in 2015, [40-50]*% in 2016 and [50-60]*% in 2017) .

(216) The Commission notes that the Notifying Party's figures are calculated on the basis of fibre subscribers out of the number of existing fixed broadband subscribers that fall into the areas covered by FTTH. That is different from calculating FTTH uptake by number of fibre subscribers out of the total number of addressable households (that have been passed with FTTH). Under the latter method, the current uptake is indeed around 10%.

(221) As regards Telefónica's announcement to dismantle its copper access network in numerous other densely populated areas by 2020, the Commission notes that in 2009, the former Spanish telecommunication regulator CMT had imposed conditions on the dismantling of Telefónica's copper network. Most importantly, if any alternative operator is co-located at a local exchange, Telefónica needs to provide wholesale services for 5 years to protect those operators' investments. Therefore, over the next five years, there will be no dismantling of Telefónica's copper network without the consent of all operators co-located at the local exchange so that alternative operators can continue to compete on the basis of their LLU infrastructure.

(222) The Commission considers that while its importance will likely decrease, the segment for speeds below 30 Mb/s – currently 80% of the market – is likely to still be relevant for the competitive situation in the overall market in the years to come. Similarly, there will probably be co-existence of the copper and NGA networks in the coming years. Therefore, contrary to the view expressed by the Notifying Party in the Reply to the Statement of Objections, the Commission considers that the merged entity's position on the market or segment for Internet access services with speeds below 30Mb/s is of relevance for the competitive assessment of the Proposed Transaction.

VHBB market/segment (speeds above 30 Mb/s)

(223) As to the market or segment of fixed Internet access services with download speeds exceeding 30 Mb/s ("the VHBB segment"), the Notifying Party submits that the combined share of Orange and Jazztel has been 4.0% (respectively, 0.7% and 3.3%) in the first half of 2014. The Notifying Party further submits that since the wholesale access obligation imposed on the incumbent Telefónica is capped at 30Mb/s, other operators, including the Parties, would have to invest in own infrastructure to be able to offer high-speed Internet access services. In its response to the Article 6(1)(c) Decision and to the Statement of Objection, the Notifying Party argues that the Commission's approach of relying on BUs instead of active lines to describe the Parties' future position in VHBB is unjustified, as it ignores Telefónica's first mover advantage in deploying fibre. Instead, the Commission should rely on current shares of the Parties for the purposes of its competitive assessment.

(224) As regards the VHBB segment, there are no data publically available for each operator. A good proxy for shares in the VHBB market or segment can be obtained by analysing the market or segment by technology since HFC cable and FTTH (together "the NGA segment") account for 86% of all VHBB lines, as illustrated in

covering 13m BUs. Under the more recent Telefónica roll-out scenario of 20m BUs, figures would be [60-70]*%, [40-50]*% and [30-40]*%, respectively.

Speaking at the Smart City Expo World Congress in Barcelona on 20 November 2012, Telefónica España CEO Luis Gilpérez Miguel said according to press reports that Telefónica would close down practically the entire copper network through which the operator provides ADSL services. See http://www.telecompaper.com/news/telefonica-to-replace-spanish-adsl-network-with-ftth-by-2020--1050512, Doc ID 1567.

With this CNMC decision, Telefónica has received approval to dismantle its first 2 local exchanges (out of the more than 8800 that it operates), Doc ID 1566.

(230) Table 14 demonstrates that while each of the Parties currently lags behind as regards the roll-out of NGA networks, and hence in the VHBB segment, each of them would considerably increase its footprint and become a stronger player also in that market or segment over time, when this market/segment becomes more important in terms of consumer take up of VHBB lines .

Document by Telefónica: "Results January – December 2014", Doc ID 2652, p. 36: http://www.telefonica.com/en/shareholders-investors/pdf/rdos14t4-pres.pdf ]

Note, however, that Telefónica recently announced it would slow down FTTH roll-out in 2015 by 35% over its previous internal plan (3.6 m instead of 5.5m BUs) in reaction to the CNMC's draft regulation published on 19 December 2014, Doc ID 1568. http://economia.elpais.com/economia/2015/02/01/actualidad/1422824643 028880.html. According to the CNMC's draft proposal, Telefónica would be obliged to grant wholesale access to its FTTH network in all of Spain except for nine cities covering approximately 16% of the Spanish population.

[Reference to the Notifying Party's internal business documents]*.

The term building units ("BUs") comprises both residential buildings and office. The total number of BUs in Spain is around 25 million, with the total number of residential BUs being around 18 million.

As discussed above, currently, the segment with speeds below 30 Mb/s represents 80% of the market.

(231) The Commission therefore considers that the Proposed Transaction has less impact in the VHBB segment in the short term where both the Parties have a limited presence. However, the dynamics as regards NGA deployment are such that the Parties are expected to play a more significant role in that segment in the coming years. The Commission considers that absent the Proposed Transaction, it is likely that the future VHBB market or segment will be characterised by four operators active at national level. The impact that the Proposed Transaction will have in the NGA market or segment in the future will be further discussed in section7.2.9.

Conclusion on market structure

(232) The Commission concludes that the Proposed Transaction will reduce the number of nationwide players in the overall market for fixed Internet access services from four to three by merging the two most successful operators, in terms of market share growth, in the last years.

7.2.2.3. Removal of two important competitive forces

Notifying Party's view

(233) The Notifying Party submits in the Form CO that the key question is not whether Orange or Jazztel are important competitors but rather whether the merger of both entities is likely to impede effective competition in the respective market. In that regard, the Notifying Party submits that the combined market share of the merged entity will be 27.45% in terms of revenues in 2013. That figure would be only slightly above the 25% threshold mentioned in the Commission's Horizontal Merger Guidelines, and in any case below 30%. Moreover, the merged entity would face strong competition from the incumbent Telefónica and Vodafone/ONO. As a result, the Proposed Transaction is not likely to raise competition concerns.

(234) In its reply to the Commission's Article 6(1)(c) Decision, the Notifying Party submits that the Commission should not ignore the existence of MVNOs also selling fixed Internet access services as well as the regional cable operators, the market for fixed Internet access services is moving towards VHBB, Telefónica, not the Parties, is the real competitive force on that market, and the use of shares of new customers, referred to as gross adds, to estimate the Parties' attractiveness is inadequate.

(235) The Notifying Party's view, when assessing the impact of the Proposed Transaction on the fixed Internet access services market in Spain, the Commission should not underestimate the role played by MVNOs and regional cable operators. The Notifying Party submits that MVNOs such as Pepephone, Masmovil, Eroski Movil, BT and Oceans are currently providing multiple play offers combining mobile and fixed telecommunications services. Likewise, regional cable operators such as Euskatel, R Cable and Telecable would hold strong market positions in their respective regions (Euskatel between 30% and 40% in the Basque Country; R Cable between 40% and 50% in Galicia; Telecable around 32% in Asturias).

(236) Moreover, as the market for fixed telecommunications services in Spain is evolving towards VHBB, the Notifying Party submits that the Commission should not focus on competition on services based on copper, but rather take a more long-term approach. As far as VHBB is concerned, the Parties are currently lagging behind much larger players such as Telefónica and Vodafone/ONO as regards both deployment of NGA networks and customer take-up for VHBB services. In that

Horizontal Merger Guidelines, paragraph 18.

(245) The Commission considers that some firms have more of an influence on the competitive process than their market share would suggest. A merger involving such a firm would therefore also have a higher impact on competition than suggested by the mere market shares of the involved undertakings. Contrary to the argument of the Notifying Party, for an undertaking to have a higher impact on competition than suggested by its market share it is not necessary that the said undertaking should be identified as the "most aggressive" competitor. Whether or not an undertaking is an important competitive force within the meaning of the Horizontal Merger Guidelines does not depend on a ranking of all competitive forces in the market but rather on a comparison of an undertaking’s actual influence on competition with the influence suggested by its market share.

(246) Based on such measurement, the results of the Commission's in-depth investigation have shown that both Orange and – particularly – Jazztel have played an important role in exerting competitive constraints upon each other and on remaining competitors in the Spanish market for fixed Internet access in the recent years.

(247) Jazztel and Orange have been competing, although differently, in proposing a number of price-attractive and innovative offers for fixed Internet access services and convergent offers containing fixed Internet access services in Spain. In particular, the Commission notes that the Parties have been important competitive forces on the market not only thanks to their price-aggressiveness but also and especially for their perceived value-for-money (that is, offering the best product for the cheapest price), which is generally also associated to brand reputation and customer satisfaction).

(248) The Notifying Party submits that the proposed merger could not significantly impede effective competition because of the limited combined market share of the Parties and the presence of alternative strong operators. However that a concentration can significantly impede effective competition not only by creating or strengthening a dominant position, but also by removing important competitive constraints that the merging parties were previously exerting upon each other, together with a reduction of competitive pressure on the remaining competitors .

(249) In this section the Commission will therefore analyse Jazztel and Orange's role in the retail market for fixed Internet access services.

(250) The Commission has gathered evidence submitted by the Notifying Parties, by third parties and general data publicly available, to substantiate its findings.

Jazztel's role

(251) Overall, Jazztel's market share growth provides a first indication of the success of its activities in the fixed Internet access services market in Spain. In that respect, the CNMC published, in October 2012, a blog post indicating how the market had changed since 2007. In that post, the CNMC affirms that "Also noteworthy is Jazztel's growth, which has gone from a 3% to a 11% market share". That growth

Horizontal Merger Guidelines, paragraph 37.

Horizontal Merger Guidelines, paragraph 25.

"También destaca el crecimiento de Jazztel, que ha pasado de 3% de cuota, al 11%."

is visualised by a graph published alongside the post, which is reproduced in Graph 1 .

Graph 1:

Source: CNMC

(252) The Commission's in-depth investigation shows the importance of Jazztel as an aggressive competitor that is characterised by a reputation of "value-for-money".

(253) The assessment is based, among others, on the Market Investigation. The Commission deems it important to assess which players are regarded as price aggressive, capable of innovation, have a good customer perception and offer value-for-money. Given the diversified nature of products offered in the Spanish retail market for fixed Internet access services, a price comparison between the different players on the Spanish market seems rather inconclusive.

(254) According to the competitors of the Parties responding to the Market Investigation the brand "Jazztel" is most often associated with value-for-money, that is, offering the best product for the cheapest price. The brand "Jazztel" was also regarded by most of the competitors responding to the Market Investigation as the brand under which the second most-innovative tariffs in Spain have been offered during the last three years (through the packs "Pack Ahorro" and "Pack Sin Límite fibra 200"), following Telefónica's "Fusión" tariffs. Furthermore, most respondents to the Market Investigation indicated the brand "Jazztel" more often than any other competing brand as the most price aggressive. The company Jazztel was regarded as having the second most price aggressive tariffs in the past three years by the respondents, following Orange, considered by the highest number of respondents as being the most price aggressive company .

(255) The Commission notes that the Market Investigation results provide a first indication of Jazztel's importance as a competitive force on the market. Regarding price aggressiveness, innovative tariffs and value-for-money, Jazztel is ranked first or second according to the Market Investigation. The Commission finds that those

"La evolución del mercado de la banda ancha en España", CNMC blog, 5 October 2012, available at http://cnmcblog.es/2012/10/05/la-evolucion-del-mercado-de-la-banda-ancha-en-espana/, Doc ID 1496.

See replies to Commission Phase II Questionnaire to Competitors of 23 December 2014, question 33.

See replies to Commission Phase II Questionnaire to Competitors of 23 December 2014, question 32.

See replies to Commission Phase II Questionnaire to Competitors of 23 December 2014, question 30.

See replies to Commission Phase II Questionnaire to Competitors of 23 December 2014, question 31.

results provide a strong indication that Jazztel is an important competitive force on the market even though Jazztel was only the second most often mentioned by the respondents regarding some of the mentioned parameters.

(256) Jazztel's important role as a competitive force in the Spanish market is further supported by evidence contained in the internal documents submitted by the Parties. Orange's internal documents describe Jazztel as an innovative and aggressive player with high customer satisfaction and a perception of value-for-money.

(257) As regards its commercial strategy, internal documents of Orange express the view that Jazztel has established its growth in recent years on two main strategic products on offer, namely convergence – with a strong emphasis on cheap mobile telecommunications offers – and a commitment to develop its fibre offer.

(258) Internal documents of Orange suggest that Jazztel has been one of the first players on the Spanish market to have understood the importance of convergent offers in Spain. In a strategic document produced by Orange in January 2013, Orange provides a background to the convergent market in Spain, before its own evaluation of the possibility to introduce a convergent offer. That document reads:

[Reference to the Notifying Party's internal business documents]*

(259) The Notifying Party submits that Jazztel's decision to launch the first convergent pack in August 2012 was only meant as a promotion. The Commission notes, nonetheless, that Jazztel's move, despite being a promotion, was the first one of its kind in the Spanish telecommunications market. This would further support the impression from the Market Investigation, the respondents to which point at Jazztel as one of the most tariff-innovative brands in Spain.

(260) Jazztel's entry into the emergent convergent market in Spain was made possible thanks to the MVNO contract that Jazztel has with Orange. That is further recognized in another internal document of Orange:

[Reference to the Notifying Party's internal business documents]*

(261) Therefore, due to its presence in the fixed Internet access market, Jazztel has been able to leverage its position in fixed Internet access services (mainly DSL services) to make an aggressive convergent offer comprising attractive commercial terms as regards the mobile component. Another of Orange's internal documents recognizes Jazztel's strategic move as follows:

[Reference to the Notifying Party's internal business documents]*

(262) While improving its convergent offers comprising a fixed xDSL Internet access component and a mobile telecommunications component, it appears from the internal documents of Orange, that Jazztel is investing also in FTTH technology in order to improve speed and quality of its commercial offer. In internal documents, Orange describes Jazztel as follows:

[Reference to the Notifying Party's internal business documents]*

[Reference to the Notifying Party's internal business documents]*

[Reference to the Notifying Party's internal business documents]*

[Reference to the Notifying Party's internal business documents]*

(263) Internal documents of Orange explain how Jazztel is perceived as an aggressive competitor, focussing on the segment of the Spanish telecommunications market for low-cost convergent offers, and having a significant impact on the overall decline of the average revenue per user ("ARPU").

(264) In a marketing plan for 2012, Orange benchmarks its customer acquisition strategy on Jazztel. That document recognizes that Jazztel has one of the best and most aggressive [a specific customer acquisition scheme]* for the following reasons:

[Reference to the Notifying Party's internal business documents]*

(265) Orange's internal documents dated October 2013 and July 2014 consistently recognize that:

[Reference to the Notifying Party's internal business documents]*

[Reference to the Notifying Party's internal business documents]*

[Reference to the Notifying Party's internal business documents]*

(266) In one of the documents referred to in recital (265), Orange evaluates the aggressiveness of each competitor by assigning them […]*. Jazztel is attributed […]* for its aggressiveness, the highest value. […]* .

(267) Finally, another presentation prepared for Orange by investment banks dated September 2014 stresses that in the recent years, Jazztel has been one of the main contributors to the pressure on prices. In more detail, the document reads:

[Reference to the Notifying Party's internal business documents]*

(268) Another aspect of Jazztel's success reflected in Orange's internal documents relates to Jazztel's perceived leadership in terms of customer satisfaction.

(269) Orange's internal documents usually report Jazztel as the company which is highly recommended by its clients and the company that […]*

(270) Orange's marketing team analysed Jazztel's strengths and weaknesses. A document prepared in March 2013 reads:

[Reference to the Notifying Party's internal business documents]* .

(271) In the same presentation, Jazztel was ranked first in customer satisfaction and customer recommendation for 2012, an element which is constant over time.

(272) A customer survey conducted by an external consultant for Orange in late 2014 shows that Jazztel's customer perception is high and the reasons why a customer usually prefers Jazztel are the following:

[Reference to the Notifying Party's internal business documents]*

[Reference to the Notifying Party's internal business documents]*

[Reference to the Notifying Party's internal business documents]*

[Reference to the Notifying Party's internal business documents]*

(273) As a result of its aggressiveness, the perception that it offers value-for-money and its optimal customer reputation, Jazztel's success was significant in the Spanish market. In internal documents, Orange acknowledges that this is reflected in the high proportion of new customers which have decided over time to subscribe with Jazztel for fixed Internet access services or convergent offers. The Commission notes that this finding is in line with the assessment of Jazztel's gross add shares as set-out in recitals (311) et seq.

(274) Orange's internal documents closely review on a monthly or quarterly basis the number of new customers of the company and of its rivals. According to Orange Spain's financial plan for 2012 to 2016, Jazztel was the leading company as regards net adds of customers of fixed Internet access services for 2010, 2011 and 2012 .

Net adds being the difference between gross adds and the customers lost to the competitors.

[Reference to the Notifying Party's internal business documents]*.

[Reference to the Notifying Party's internal business documents]*.

[Reference to the Notifying Party's internal business documents]*.

[Reference to the Notifying Party's internal business documents]*.

(275) A document of Orange called [Reference to the Notifying Party's internal business documents], describes Jazztel as the leader in terms of gross adds of fixed Internet access services […]* and the company having the highest share of gross adds, along with ONO […]* as regards cross-selling convergent packs to mobile customers .

(276) Orange's strategic plan for 2015 to 2017, dated 2 June 2014, describes Jazztel as the leader in acquiring customers for its convergent offers […]* and the second best in cross-selling convergent offers to mobile customers .

[Reference to the Notifying Party's internal business documents]*.

(277) In an internal document of Orange, […]*, provided some insights on the performance of Jazztel over the last years. That supporting document reads:

[Reference to the Notifying Party's internal business documents]*

(278) Likewise, a presentation by the investment banks […]* in support of the Proposed Transaction consistently explains that:

[Reference to the Notifying Party's internal business documents]*.

(279) An internal document of Orange discussing the possibility of an acquisition of Jazztel and dated […]* reads:

[Reference to the Notifying Party's internal business documents]*

(280) The Notifying Party submits that Jazztel's (and Orange's) aggressiveness in the past should not be projected in the future. In particular, the Notifying Party submits that the Parties' aggressiveness will come to a halt once each of the Parties will have cross-sold fixed or mobile telecommunications products to its respective existing customer base.

(281) The Commission disagrees with the Notifying Party's argument. The Commission notes that, as already explained in recitals (274) to (275), the Parties have been successful players not only by cross-selling fixed or mobile customers to their existing customer bases, but also and mostly by attracting new convergent clients. In particular Jazztel's high gross adds represented by customers which were not previously Jazztel's subscribers and therefore, which are new clients for both fixed and mobile telecommunications services – cannot be explained by such effects.

(282) The Commission concludes that Jazztel has been and is an important competitive force in the fixed Internet access market in Spain.

Orange's role

(283) Orange also is an important competitive force in the Spanish telecommunications market. Orange's aggressive growth in recent years has been achieved not only by competing aggressively on prices, but also by means of its innovative products and its capacity to react to successful competitors' offers, like Jazztel's.

(284) This assessment is based on the Market Investigation addressed to the Parties' competitors. On average, respondents to the Market Investigation pointed at Orange most often as the company with the most price aggressive tariffs in the past three years, either under its own brand Orange (for example, "Canguro 35") or under the low-cost brand "Amena" (for example, "Amena.com en casa").

(285) The competitors of the Parties that took part in the Market Investigation also associated Orange's low-cost brand "Amena" and the brand "Orange" itself most often with price-aggressiveness in Spain, after Jazztel.

(286) The Commission has further analysed the internal documents of both Parties and according to these documents, Orange is perceived as a threat by Jazztel. The documents also reflect that Orange has become one of the most successful companies in the Spanish market, along with Jazztel.

(287) In an internal document of Jazztel dated April 2013, Jazztel analyses Orange's changed position in the market:

[Reference to the Notifying Party's internal business documents]*

(288) The perceived aggressiveness of the Canguro offer is shown in a presentation by Jazztel of June 2013 where, comparing the situation before and after the launch of Canguro on the market, Jazztel noticed that: […]* .

(289) Likewise, in June 2013, Jazztel elaborates internally a commercial campaign aimed at […]* and […]*. An internal e-mail of Jazztel dated 25 June 2013 explains the strategic consideration of Jazztel for such a campaign.

(290) In particular, a presentation attached to the e-mail referred to in recital (289) reads as follows:

[Reference to the Notifying Party's internal business documents]* .

(291) That e-mail contains the details of the following […]* commercial campaign:

[Reference to the Notifying Party's internal business documents]*

(292) As a result of its aggressiveness, Orange has become a successful company in the Spanish market. This is further reflected in the high proportion of new customers which have decided over time to subscribe with Orange for fixed Internet access services or convergent offers. The Commission notes that that finding is in line with the assessment of Orange's gross adds shares as explained in recitals (311) et seq.

(293) In that regard, Orange's internal documents closely review on a monthly or quarterly basis the amount the company's new clients and that of its rivals. Those documents reflect Orange's success as explained in recitals (294) to (297).

(294) Firstly, Orange Spain Financial plan for 2012-2016 clearly shows that Orange was the second leading company as regards net adds of fixed Internet service clients for 2010, 2011 and 2012 (Jazztel being the first).

[Reference to the Notifying Party's internal business documents]*.

(295) Secondly, an internal presentation of Orange dated August 2013 intended to analyse the market for residential offers in Spain shows that, between December 2012 and June 2013, Orange is the operator that grows the most. […]* .

[Reference to the Notifying Party's internal business documents]*.

(296) Thirdly, a document of Orange called […]* describes Orange as the leader in terms of gross adds of newly-attracted convergent clients […]* and of cross-selling convergent packs to fixed Internet access customers […]* .

[Reference to the Notifying Party's internal business documents]*.

(297) Finally, an internal document of Orange entitled "Orange Spain Country Business Review" and dated December 2014 portrays Orange as the operator which has the highest net adds since the third quarter of 2013 and strong net adds in the mobile market […]* .

(298) As a result of its aggressiveness and its commercial success, Orange has seen its customer base and financial results increase over time. The internal documents indicating Orange and Jazztel's performance are described in recitals (277) to (279) above.

(299) The Commission concludes that Orange has been an important competitive force in the fixed Internet access market in Spain, at least in the last two to three years.

Telefónica's role

(300) The Notifying Party submits that Jazztel and Orange are not to be regarded as important competitive forces because Telefónica would be by far the most aggressive player on the fixed Internet access market in Spain. The Commission disagrees with that argument for the reasons explained in recitals (301) to (308).

(301) For a concentration to result in a significant impediment to effective competition – and contrary to the Notifying Party's argument, the Commission is not required to show that the merging parties have been the most important competitive forces, as long as the merging parties have exerted significant competitive constraints upon each other and on the remaining competitors for a period of time before the proposed transaction.

(302) In this case, even with the assumption that Telefónica would be the most aggressive player on the Spanish market or just as aggressive as the Parties, the removal of important competitive forces on the market, such as Orange and, particularly, Jazztel, could still result in a significant impediment to effective competition in view of the likely influence on the ability and incentives of the merged entity to compete in the resulting new market structure.

(303) Moreover, internal documents of Orange […]*. For example, an internal document of Orange detailing the financial plan of the company for the period 2015 to 2018 and dated December 2014 provides useful insights of the role played by Telefónica.

(304) In a slide describing the competitive positioning of each of the Spanish telecommunications players along the lines of "Low-cost/Mainstream" and "Mobile-only/Convergence", Orange describes Telefónica […]* .

(305) The Notifying Party argues that aggressiveness should be measured as the development of the ratio between revenue and subscriber market shares. Since the share of Telefónica in terms of customer number has fallen "much less dramatically" than its revenue share, the Notifying Party concludes that "the price of Telefónica has fallen relative to the average price of its competitors". That would make Telefónica, in the view of the Notifying Party, an aggressive competitor. Regarding Jazztel and Orange, the Notifying Party finds that their shares in terms of revenue have increased faster than their shares in terms of subscribers and concludes that the reduction in prices of both companies has been less steep than that of their main competitors.

(306) The Commission finds that the arguments put forward by the Notifying Party are unconvincing. The Notifying Party implies that the absolute ratio between market shares in terms of revenues and market shares in terms of subscribers mirrors the price. However, this seems to be a gross over-simplification in particular in markets with differentiated products and product types (such as fixed Internet access services ("1p"), dual-play ("2p"), triple-play ("3p") and quadruple-play ("4p") products).

(307) Differences in the ratio between revenue shares and subscriber shares could be caused in the market for fixed Internet access services by factors such as speeds offered, product mix, brand perception or customer service, to name just a few potential factors. Indeed, to take the example of Jazztel which, according to the Notifying Party, must have become less aggressive because the revenue or subscriber ratio increased, was able to gain in particular new subscribers to its triple-play offers in 2014 as illustrated in Table 16 and Table 19 while the overall subscriber share for Internet access services grew less.

(308) Next, looking at the development of such ratio over time does neither have to be indicative for the developments of the prices of the respective operator. The same factors mentioned in recital (306) that could distort a potential relation between the revenue or subscriber ratio and prices, could also distort any potential link between the development of such ratio and the development of prices. Those effects

[Reference to the Notifying Party's internal business documents]*.

(309) For the reasons outlined in recitals (301) to (308), the Commission discards the arguments of the Notifying Party derived from the alleged significance of the ratio between subscriber and revenue market shares or the developments thereof.

(310) The Commission will analyse Telefónica's incentives to compete in the market post-transaction in section 7.2.2.5.

Gross add shares

(311) The finding that the Parties are two important competitive forces is further supported by their gross add shares.

(312) Existing subscriber market shares do not fully depict the nature of competition in that market. That is because many customers may be bound by longer-term contracts, which means that, at any given time, only a fraction of the total customer base is contestable. At any given moment, competition occurs only in respect of contestable customers and entirely new customers (that is to say those who are not yet fixed subscribers at all), whose numbers have been declining significantly in recent years as the market has matured. Consequently, it would take a number of years before trends in winning new business were reflected in overall market share.

(313) In this respect, the Market Investigation has provided indications that on average the minimum contractual commitment period (that is to say the duration during which a customer cannot freely terminate its contract) of fixed Internet access services contracts is around 12 months. The minimum contractual commitment period of multiple play offers comprising fixed Internet access services, along with fixed voice services and mobile telecommunications services goes from 12 to 18/24 months, depending on whether a mobile handset is included in the offer.

(314) Gross add shares provide a more dynamic picture of the market, since they show how successful different providers are in attracting new customers. In contrast, market shares indicate the "stock" of clients that each provider is currently serving. As such, gross add shares that are higher than the respective (static) market shares provide an indication that a certain provider is more important for competition than suggested by its market share (and the likely increase of its market share in the future).

To provide an example: If operator A has many customers for its product A+ that pay on average EUR 50 and also customers for its product A- that costs EUR 35, whereas operator B has mostly customers that pay on average EUR 30 for the product B-, there could be several reasons to explain such difference based, for example, on the products offered, quality of customer service, brand value or subsidies. If then operator B successfully starts targeting customers using product A+ with a product B+ that is priced at EUR 45, the ratio between operator B's revenue and subscriber shares will increase whereas the one of operator A will decrease. The increase of the ratio between revenue and subscriber share of B does, however, not imply that B has become less aggressive – in fact, quite the contrary would be true –, neither is the decrease of the same ratio of operator A a sign of a particularly aggressive conduct as, in this example, for operator A changes nothing except that he is losing customers of his product A+.

For the purpose of this Statement of Objections, operators A's quarterly gross adds relating to product type X have been defined as the number of new subscribers acquiring product type X with operator A in each quarter. New subscribers have been defined as subscribers of the operator who were not subscribers of a product type containing a fixed broadband component at the beginning of the relevant quarter and were subscribers of a product type containing a fixed broadband component at the beginning of the following quarter.

See replies to Commission Phase II Questionnaire to Competitors of 23 December 2014, question 25.

(315) The figures relative to the total active fixed subscribers and gross adds of Table 16 show how both Orange and Jazztel attract more customers than their market share would suggest. In 2014, Orange attracted […]*% of the total new customers or customers switching their provider, compared to a subscriber market share of […]*%. Likewise, in the same year Jazztel attracted.[…]*% of new or switching customers, as opposed to a subscriber market share of […]*%. Moreover, both Orange and Jazztel grew in terms of number of active subscribers and in terms of market share. Therefore, Orange and Jazztel are to be considered as important competitive forces on the Spanish telecommunications market.

(316) Furthermore, the Commission points out that Telefónica provided the number of fixed lines and not the number of fixed subscribers, as Telefónica records the number of lines in its systems. To the extent subscribers have on average more than one fixed line, both the subscriber figures as well as the number of gross adds of Telefónica in Table 16 could be overstated.

(317) The Notifying Party submits that, according to the Commission's calculations, Telefónica has the highest gross add share. That would further support the Notifying Party's claim that Telefónica, not the Parties, is the most aggressive player on the market.

(318) The Commission disagrees with this argument. As explained in paragraph 37 of the Horizontal Merger Guidelines, a merger involving undertakings that have "more of an influence on the competitive process than their market shares or similar measures would suggest" could change the competitive dynamics in a significant, anti-competitive way. Thus, the Horizontal Merger Guidelines use the term "important competitive force" to describe an undertaking that is more important for competition than its market shares would indicate. Therefore, it is very relevant to note that the gross add shares of both Parties are significantly higher than their market shares.

(319) While the Commission does acknowledge that Telefónica is a very significant competitor, its outstanding role in the Spanish telecommunication markets in general and in the retail market for fixed Internet access services in particular is already reflected in Telefónica’s very high shares in those markets. The argument of the Notifying Party that Telefónica has the highest gross adds – even though less than its market shares – thus misses the point.

(320) The Notifying Party further submits that the Parties' gross add shares have been […]*. As a result, in the Notifying Party's view, the Parties cannot be considered as the most aggressive players on the market.

(321) However, the argument in recital (320) misses the point of the use of gross adds to assess whether an undertaking is an important competitive force as well. As explained in recital (318), gross adds can provide an indication of whether undertakings have more of an influence on the competitive dynamics in a specific market than suggested by their market shares. The fact that the Parties’ gross add shares have consistently been significantly higher than their market shares provides a particularly strong indication of such importance. For the purposes of assessing whether an undertaking has been an important competitive force, the comparison between gross adds and market shares is relevant to assess whether such undertaking punches "above its weight", whereas a comparison of gross adds in any given year with gross adds in the previous year does not affect the finding that the respective undertaking is an important competitive force as long as the gross add shares are significantly higher than the market shares.

(322) The Commission concludes that both Orange and Jazztel have played an important role in exerting competitive constraints upon each other and on remaining competitors in the recent years.

Change of the incentives of the merged entity to compete

(323) The Notifying Party argues that the transaction will allow Orange to better compete with Telefónica and Vodafone particularly because of the higher fibre coverage of the merged entity.

(324) The Notifying Party submits that the evidence from the Market Investigation is inconclusive and Orange's internal documents do not demonstrate an intention to modify its competitive strategy. Moreover, Orange points out that a large number of respondents to the Market Investigation were not concerned about potential price increases. Orange further argues that its internal documents demonstrate that "the main aim of the acquisition is to increase the ability of Orange to compete in the Spanish market by investing in the development of a fibre network". In addition, Orange points out that its strategic plan expects ARPUs to continue decreasing after the transaction.

(325) The Notifying Party draws attention to the fact that the quotes that were used in the Article 6(1)(c) Decision to conclude that the merged entity would have less incentives to compete refer to the general evolution of the market "from a passive perspective" and do not mention active changes in the competitive strategy of the merged entity like new pricing policies or less aggressive marketing strategies.

(326) Moreover, the Commission would ignore the structural changes in the market – notably the transition to NGA networks, that prevent the merged entity from increasing prices or reducing aggressiveness. According to Orange, Telefónica and Vodafone/ONO enjoy a first mover advantage to acquire clients interested in VHBB. Price increases by the merged entity would only accelerate customer migration to VHBB services to the detriment of the Parties that are lagging behind in NGA roll-out. Therefore, low prices for broadband over copper provided by the merged entity would be important to disincentivise customers from migrating to competitors' VHBB services.

(327) The Notifying Party further submits in its Reply to the Statement of Objections that the internal documents of Orange on which the Commission relies in its assessment in the Statement of Objections do not mention a direct impact of the Proposed Transaction on retail prices. The Notifying Party argues that […]*[…]*.

(328) In that regard, the Notifying Party points out that the Commission acknowledges in the Statement of Objections that most internal documents […]*. The Notifying Party further argues that all internal documents of Jazztel have no evidentiary value when it comes to evaluating the incentives of the merged entity. Moreover, the Notifying Party objects to the use of statements by investments banks as those could not identify the reasons behind the acquisition decision of Orange.

(329) In its reply to the Letter of Facts, the Notifying Party submits that […]*.

(330) Finally, the Notifying Party claims that it would be unable to increase retail prices for fixed Internet access services offered both via DSL and via FTTH post-transaction due to the competitive pressure of the larger NGA networks of Telefónica and Vodafone. If the merged entity increased prices for DSL-based fixed Internet access services, it would create an additional incentive for its customers to switch to NGA networks. Given the current coverage of NGA networks, those customers would likely switch to either Telefónica or Vodafone. As regards the retail prices of fixed Internet access services offered via its FTTH network, the Notifying Party argues that it will have an "empty" NGA network and thus the incentive to gain customers quickly rather than to increase prices.

(331) The Commission finds that the merged entity would have lower incentives to compete on the retail market for fixed Internet access in comparison to Orange's and Jazztel's incentives on a standalone basis. That finding is based on the market structure and the expected loss of competition as a consequence of the Transaction as well as on indications of a decrease in competition contained in internal documents of Orange.

(332) The Horizontal Merger Guidelines state in paragraph 24 that "[a] merger may significantly impede effective competition in a market by removing important competitive constraints on one or more sellers, who consequently have increased market power. The most direct effect of the merger will be the loss of competition between the merging firms. For example, if prior to the merger one of the merging firms had raised its price, it would have lost some sales to the other merging firm.

(333) As established in section 7.2.2.3, the Proposed Transaction will combine two important competitive forces on the fixed Internet access market in Spain.

(334) For the reasons developed in this section, the Commission considers that the loss of competition between the Parties provides the merged entity with the incentive to raise prices compared to prices absent the merger. In a market characterised by decreasing prices, that could also mean an incentive to a slower decrease of prices following the Proposed Transaction. Loss of competition as a result of a merger can arise because at least some of the sales that would have been lost to competitors as a result of a price increase by one of the merging parties pre-merger will be diverted to the other merging party post-merger. A merger thereby changes pricing incentives (since some sales that would be lost after a price increase pre-merger will be diverted to the products of the other merging party post-merger) and may lead to upward pricing pressure. Section 7.2.7 sets out the quantitative assessment of the consequences of the Proposed Transaction and suggests also that the merged entity is likely to have substantial incentives to significantly increase its prices as a consequence of the loss of competition between the Parties.

(335) In particular the transaction will lead to a loss of competitive pressure by the low-cost, convergent offers of Jazztel. Internal documents of Orange describe Jazztel as the only low-cost, convergent operator. According to internal documents of Orange, the low-cost, convergent offers of Jazztel […]* . Telefónica expects a "more rational behaviour in the Spanish markets" and thus less competitive pressure from Orange after the merger .

(336) Those considerations further are supported by a statement in an internal document of Orange according to which […]* .

(337) The main reasons for the Commission to conclude that the incentives of the merged entity to compete will decrease after the merger, emerge from internal documents of Orange showing that Orange assumes that the Proposed Transaction will lead to less competition in the Spanish markets. That is expressed in different ways – usually without mentioning a direct impact on (retail) prices […]* . Despite the argument of the Notifying Party that that document was not drafted during the preparation of the acquisition of Jazztel approximately half a year later, the Commission finds that the document is in sufficiently close context to the transaction to provide insights on the effects of the acquisition of Jazztel as expected by Orange.

(338) Other documents refer to […]* as a result of the transaction or […]* […]* .

(339) That internal due diligence assessment of Orange was featured prominently in documents summarizing the transaction for the purpose of a decision on the transaction by Orange's internal hierarchy. […]* .

(340) The most important synergies which Orange assumed would be the result of the transaction are […]*. Those synergies imply, according to this document that the consolidation […]* . […]*. On the same slide in this document, Orange points out that […]* .

(341) The Commission assumes that the synergies expected from the transaction by Orange have been an important factor in deciding whether to pursue the acquisition of Jazztel or not. They have even been taken into account in calculating the price Orange offers to Jazztel's shareholders. According to a press release of Orange, Orange's offer values Jazztel "at Enterprise Value / EBITDA 2015-multiple of 8.6x after accounting for the synergies generated by the integration of the two entities". The press release continues to state that the transaction "will generate estimated global synergies for the combined entity of up to 1.3 billion euros" .

(342) Those synergies of "up to 1.3 billion euros" referred to in recital (341), that have been used to calculate the price offered to Jazztel's shareholders […]* (340). Therefore, the argument of the Notifying Party, that the assessment of the incentives of the merged party to compete is based only on a limited number of documents misses the point. The document [Reference to the Notifying Party's internal business documents]* summarizes the key expectations and calculations of Orange that formed the basis for the internal decision to acquire Jazztel as well as for calculating the consideration offered to shareholders of Jazztel. Its evidentiary value for the reasons and expectations of Orange regarding the transaction is thus of high significance.

(343) […]* .

(344) […]* .

[Reference to the Notifying Party's internal business documents]*.

[Reference to the Notifying Party's internal business documents]*.

[Reference to the Notifying Party's internal business documents]*.

[Reference to the Notifying Party's internal business documents]*.

[Reference to the Notifying Party's internal business documents]*.

[Reference to the Notifying Party's internal business documents]*.

(345) The presentation, referred to in recital (341), provides a clear explanation of how such synergies have been derived. […]* . […]*. […]* .

(346) Furthermore, based on the presentation, the argument of the Notifying Party that […]* seems implausible. […]*. […]*. […]* .

(347) It also seems unreasonable to assume that […]*. All those factors are general market developments that evolve irrespective of the Proposed Transaction. […]*

(348) […]*. For those reasons, the Commission concludes that it is evident, even from the presentation itself, that […]* .

(349) The Commission made substantial efforts to retrace in internal documents […]* and requested further information several times and finally in a decision pursuant to Article 11(3) of the Merger Regulation. In its response to the Article 11(3) decision, the Notifying Party provided some additional documents that contain the same figures as in the [Reference to the Notifying Party's internal business documents]* presentation and show that those figures have been provided by Orange to its investment bank. However, the respective values are always imported from another excel file .

(350) In its [Reference to the Notifying Party's internal business documents]*, Orange assumes that in a stand-alone case, the ARPU will amount to, by the end of […]*, EUR […]* for convergent offers, EUR […] for mobile offers. In the same documents, Orange also forecasts the development of ARPUs in case of a merger with Jazztel. Under such assumption, Orange envisages the average ARPUs of its customers to amount at the end of […]* to EUR […]* for converged offers, EUR […]* for fixed broadband offers and EUR […]* for mobile offers .

(351) […]* whereas the main opportunity mentioned is […]* .

(352) A further document, an Excel file named [Reference to the Notifying Party's internal business documents]*, contains forecasts by Orange of several figures related to the mobile and fixed broadband ("FBB") business of all major telecommunication operators in Spain. Regarding the ARPUs, the document states within the slide entitled [Reference to the Notifying Party's internal business documents]* the estimated ARPUs of Orange for mobile and FBB customers from 2012 to 2018 as well as the resulting compound annual growth rate ("CAGR") between 2013 and 2018. That information is complemented by a similar set of figures a few lines further down within the document that are described as […]* .

(353) The respective figures are set out in the following table:

Table 17:

ARPUs of ORANGE '12 '13 '14e '15e '16e '17e '18e CAGR 13/18 […]* […]* […]* […]* […]* […]* […]* […]* […]* […]* […]* […]* […]* […]*

[…] * […]* […]* […]* […]* […]* […]* […]* […]*

[…] * […]* […]* […]* […]* […]* […]* […]* […]*

Source: Notifying Party

(354) According to the figures in Table 17, the […]*. That becomes obvious if the CAGR is calculated based on the figures in Table 17 for the period between 2015 and 2018, as follows:

Table 18:

ARPUs of ORANGE '15e '16e '17e '18e CAGR 15/18 […]* […]* […]* […]* […]* […]* […]* […]* […]*

(355) The change in the CAGR between 2015 and 2018 in case of […]* amounts to […]*% in the mobile market and […]*% in the fixed market. […]* […]* […]* .

(356) The Commission finds that those numbers do not appear to be "calculations, trash and other byproducts" that could be found in a working file at some point in time as argued by the Notifying Party. While the Notifying Party is correct in pointing out that the respective table in the excel file available to the Commission is not used in other parts of the excel file, the Commission notes that the values in the excel file for the […]* scenario are derived from another excel file – a certain cell in a slide called […]* of an excel file named […]* which was located in a subfolder called […]* .

(357) The Commission requested from Orange that other excel file and further related documents by a decision pursuant to Article 11(3) of the Merger Regulation. While Orange was not able to provide the exact files linked in the excel file, the additional documents provided in response to that decision show that those calculations have neither been byproducts nor have they been made by investment banks. Several other excel files contain those or similar values which are all not calculated within the respective excel file but included by referencing to another excel file.

(358) Neither the excel file […]* nor any other document that actually calculates the ARPUs following the Proposed Transaction has been provided to the Commission. […]* .

(359) Even though the Commission was thus unable to fully retrace and understand how the changes in the ARPU following the merger have been calculated, the Commission finds that for the reasons set-out, the respective internal documents provide sufficient evidence that the […]* assumed by Orange will be caused by a loss of competition from the merged entity following the transaction compared to the competitive pressure of Orange and Jazztel as separate entities going forward.

(360) Orange's estimate is also shared by others. An internal document of Jazztel […]* . In the same document, […]*'s analysis of the proposed merger is quoted […]* .

(361) Contrary to the Notifying Party’s argument, the higher revenues are not just "passive" developments. They are, first of all, triggered by the acquisition of Jazztel by Orange. Furthermore, the different assumptions for the average revenues per customer show that Orange does not just assume to profit from a larger customer base but rather to generate more revenues per customer. The Commission deems it highly unlikely that such different averages can be solely explained by factors other than higher prices. Such potential explanation would also conflict with the risk mentioned in the document as the highest risk in the stand-alone scenario that clearly links "ARPU erosion" to "competitive aggressiveness". A slower decrease of ARPUs therefore seems to be caused by less competitive aggressiveness.

(362) The Commission acknowledges that [Reference to the Notifying Party's internal business documents]* also shows that Orange plans to continue increasing its market shares after the merger with Jazztel and thus aims at retaining its competitiveness and aggressiveness. However, taking into account that Orange also estimates that its ARPUs and, as already explained, the prices for its products […]*, the Commission finds that the merged entity will likely compete less aggressively than the two Parties absent the merger.

(363) The structural changes in the market which, according to the Notifying Party, would prevent the merged entity from reducing aggressiveness because of the first mover advantage of Telefónica and Vodafone/ONO in NGA networks will, in the Commission's assessment, not provide sufficient incentives for the merged entity to compete as aggressively as the two Parties on a stand-alone basis. The Notifying Party argues that price increases by the merged entity would accelerate customer migration to VHBB services and thus benefit Telefónica and Vodafone because of their larger NGA networks.

(364) The argument in recital (363) assumes that only planned or likely price increases of the merged entity would indicate lesser incentives to compete. However, in particular in a market in which prices decreased consistently over the last few years, also the likelihood of prices decreasing slower than without the merger is an indicator for fewer incentives to compete. Furthermore, the trend towards VHBB services might have even increased competition between Orange and Jazztel absent the Proposed Transaction. Given that Telefónica and Vodafone can offer VHBB services to a much larger amount of customers, Orange and Jazztel would have been both forced to compete for customers with attractive xDSL offers in a large area of Spain where both would have had no NGA network.

(365) The responses to the Market Investigation did not provide a clear opinion of the market participants about the incentives of the merged entity to compete. Respondents to the Phase II questionnaires had no majority view on whether the incentives of the merged entity to compete will increase, decrease or remain the same. Five respondents answered that the incentive to compete will decrease, four answered that it will stay the same and three respondents replied that it will increase.

[Reference to Jazztel's internal business documents]*.

[Reference to the Notifying Party's internal business documents]*.

(366) Respondents replying that the incentives will increase or stay the same provided only few reasons for their opinion. One respondent stated that the market is competitive and is likely to remain that way. The Commission also notes that there is a discrepancy between Telefónica's response to the market questionnaire and other public statements by Telefónica. In response to the market questionnaire Telefónica stated that the merger will increase the incentives of the merged entity to compete and explained that the Proposed Transaction "will create a stronger and more complete operator" and "enable Orange to compete even more fiercely against Telefónica and other infrastructure operators" . In a statement addressed to investors, however, Telefónica expects "a more rational pricing scenario" in which Vodafone and Orange "need to defend higher ARPUs in order to make sure that the value that they acquired is not destroyed" and "being more focused on defending their core customer segments". While it is not obvious whether only the answer to the questionnaire or only the statement addressed to investors or both of those statements have been guided by strategic considerations, the Commission finds that the strong discrepancy between those statements render Telefónica's response to the question for the incentives of the merged entity to the Market Investigation non-credible. Instead, the detailed arguments, addressed to Telefónica's investors, about why Telefónica expects Orange to behave more "rationally" provide another indication that the incentives of the merged entity to compete are likely to decrease.

[Reference to the Notifying Party's internal business documents]*.

[Reference to the Notifying Party's internal business documents]*.

(367) Respondents taking the view that the merged entity will have fewer incentives to compete provided a number of arguments for their opinion. Competitors point to the disappearance of one of the major players in the market for fixed Internet access services and to the fact that only three nation-wide players will remain active on the market. Another respondent replied that the high market shares of the remaining nation-wide operators will decrease their incentive to compete and that the merger would remove Jazztel, a particularly challenging and innovative competitor. However, the Commission also takes into account that those, as well as all other, answers might have been guided by individual interests of the respective parties in the outcome of the Commission's assessment.

(368) As regards the Notifying Party's argument that it will not be able to increase its prices for DSL Internet access services to prevent its subscribers from switching to NGA networks and thus likely to either Telefónica or Vodafone, the Commission underlines, first of all, that in markets with falling prices, a slow-down of the price decrease would also give rise to competitive concerns and that, based on the assessment in section 7.2.2.4, is precisely what Orange is assuming to be the result of the Proposed Transaction.

(369) Moreover, absent the Proposed Transaction, Jazztel would also compete in most areas for customers that are not willing to pay a premium for fixed Internet access services via a NGA network. Therefore, regardless of the pricing pressure exercised by existing NGA networks, competition for the remaining DSL customers will be reduced. Absent such additional competitive pressure from Jazztel, Orange could likely decrease its prices less without triggering higher churn towards NGA offers of Telefónica or Vodafone.

(370) Orange also claims that the merged entity will have an incentive to fill its FTTH network with new customers and thus to compete aggressively in areas in which the merged entity rolls-out an own fibre network and thus will have no incentive to raise prices. However, again, it is more likely that, absent the Proposed Transaction, the situation would have been even more competitive. If either Jazztel or Orange would have rolled-out fibre in the same area absent the transaction, they would have had the same incentives to "fill-up" their respective networks. Therefore, according to Orange's argument, competition would have increased and thus, prices would have likely decreased more, in a stand-alone scenario than if the Transaction was concluded.

Conclusion on the change of the incentives of the merged entity to compete

(371) The Commission concludes that the merged entity's incentives to compete aggressively would be significantly weaker than those of Orange and Jazztel without the transaction.

7.2.2.5. Likely reaction of competitors

Notifying Party's view

(372) The Notifying Party argues that the remaining competitors, and in particular Telefónica, are already "price challengers". Telefónica had added services to its multiple play bundles without increasing prices (for example, TV channels) and will continue to compete aggressively by rolling out its fibre network. In the Notifying Party's view, the Transaction will rather incentivise Telefónica to accelerate its fibre roll-out in order to continue benefiting from a first mover advantage.

(373) As regards the Commission's interpretation of internal documents of Orange, the Notifying party submits that the Commission misinterprets the statement that Vodafone/ ONO will limit itself to consolidate its position. Instead, the merger between Vodafone and ONO would rather allow Vodafone to cross-sell its services.

(374) Furthermore, the Notifying Party argues that Telefónica is a dominant firm and very aggressive in getting additional clients and revenues. According to Orange, Telefónica's statement quoted by the Commission whereby Telefónica does "try to drive the market" and that it provoked a "painful" process towards convergence shows the importance of Telefónica in the Spanish telecommunication markets.

(375) With regard to the price increases recently announced by Telefónica for dual-play services, the Notifying Party argues that it is just another example of Telefónica's market power by which Telefónica tries to further drive the market towards quadruple-play offers.

Commission's assessment

(376) According to the Horizontal Merger Guidelines, "Non-merging firms in the same market can also benefit from the reduction of competitive pressure that results from the merger, since the merging firms' price increase may switch some demand to the rival firms, which, in turn, may find it profitable to increase their prices. The reduction in these competitive constraints could lead to significant price increases in the relevant market" .

(377) In line with the Horizontal Merger Guidelines, the Commission decision concerning Hutchison 3G Austria/Orange Austria noted that just as the merged entity, "competitors would also face the same trade-off between attracting additional new customers by practising lower prices and cannibalising the flow of customers who would anyway have switched to them" and found that "other competitors are unlikely to increase supply or reduce prices in response to a price increase by the merged entity. Even assuming competitors are not capacity constrained, it is unlikely that they would increase supplies in response to a price increase of the Parties. Since the products are endogenously differentiated in terms of their market positioning, generally accepted and robust economic theory demonstrates that the profit-maximising response of competitors to a price increase would be to increase prices themselves" . A similar conclusion was drawn in Hutchison 3G UK/Orange Austria . Although the cases cited concerned primarily the retail mobile telecommunications services market, the Commission considers that the market characteristics of the retail market for fixed Internet access services are very similar and the Spanish nationwide operators are active in both markets.

(378) In the Commission's view, the reasoning in recitals (376) and (377) applies to this case. Both merging Parties exert an important competitive pressure on all other competitors, including Telefónica and Vodafone. The change of incentives of the merged entity following the merger and the likely price increase by the merged entity would significantly reduce that pressure on those operators' customer bases and induce some of the merged entity's customers that would have remained with the Parties in the absence of the transaction to switch to Telefónica and Vodafone. Accordingly, those operators would find it easier to retain their customers and could even attract new customers who would switch away from the merged entity. The resulting increase in the demand faced by those operators provided them with an incentive to raise its prices in turn.

(379) The finding that competing firms have incentives to raise prices as a response to a price increase by another firm is a general characteristic in standard models of oligopolistic competition. That characteristic is also reflected in the Commission's quantitative analysis which shows that, in response to a price increase by the Parties, it is likely that the Parties' rivals, including Telefónica and Vodafone, would, all else being equal, also raise their prices so that the transaction is likely to lead to general price increases on the market compared to what would be the case in the absence of the merger.

(380) In the present case, the Commission found several indications that operators indeed tend to follow price changes of their competitors. Therefore, it must be anticipated

Horizontal Merger Guidelines, paragraph 24.

Commission Decision of 12 December 2012 in Case No COMP/M.6497 – Hutchison 3G Austria / Orange Austria, recital 374.

Commission Decision of 12 December 2012 in Case No COMP/M.6497 – Hutchison 3G Austria / Orange Austria, recital 367.

Commission Decision of 28 May 2015 in COMP/M.6992 Hutchison 3G UK / Telefónica Ireland, recital 582.

See also Commission Decision of 12 December 2012 in Case No COMP/M.6497 – Hutchison 3G Austria / Orange Austria, recital 369. See also BULOW, J. I., J. D. GEANAKOPLOS, AND P. D. KLEMPERER (1985): “Multimarket Oligopoly: Strategic Substitutes and Complements,” The Journal of Political Economy, 93(3), 488–511.

that competing operators will respond to price increases of the merging parties with raising their own prices, implying a more pronounced market-wide price increase.

(381) Internal documents of Orange show prices of its products are intentionally set close to its competitor's prices. […]* […]* .

(382) Likewise, the document discussed in recital (286) documents that Jazztel reacts to the introduction of the attractive […]* products by Orange by offering more attractive conditions to its own customers. That is another instance that illustrates that competitors tend to follow price changes of their competitors.

(383) Regarding the competitive behaviour absent the Proposed Transaction by Vodafone, an internal document of Orange setting out Orange’s strategy for the 2015-2018 period explains that […]* The Commission considers that those quotes distinguish cross-selling opportunities from other, presumably price-related aggressive moves. The latter are clearly not expected from Vodafone by Orange.

(384) In the same document, […]*’s future behaviour is described as […]*. Customer retention, as opposed to customer acquisition, cannot be described as a particularly aggressive focus that Orange expects from […]* .

(385) Another internal document of Orange describes the current priorities and main risks for the main market players. For Telefónica, Vodafone and Orange, one of the main risks is described as […]*. It follows from that document that Orange perceives those players as unlikely to compete aggressively by themselves.

(386) The expectations of Orange for the development of the Spanish market as of 2015 are set forth in another internal document of Orange. According to that document, Orange assumes that […]* .

(387) In general, it appears from those documents that […]*

(388) That is also confirmed by certain public statements of Telefónica executives according to which Telefónica expects consolidation to lead to a "more rational" market with stabilising prices and that Telefónica will be driving the market toward a rational price policy, described as a market where players would mainly try to consolidate their existing customers in fixed and mobile services. Thus, the Chief Operating Officer, CEO and Chairman of Telefónica Europe Division, Mr. Álvarez-Pallete López, stated the following at an earnings call with investors on 12 November 2014:

"So overall, if you ask me to give you an overall picture of the Spanish market, we see a much more rational behaviour of the market. It has been painful because we have been provoking this convergent process and now it looks like we are reaching a point with the different players, namely the consolidated or convergent players of the market -ourselves, Vodafone and Orange-, would have similar ARPUs to defend on the existing customer base both on the wireline and on the wireless side. So, we forecast a more rational Spanish market and this is starting to flow through the

[Reference to the Notifying Party's internal business documents]*.

stabilisation that you saw in the ARPU of Fusión. So we think this is the way to go. This is the way we, as the leaders, are going to try to drive the market and I think that the consolidation processes are going to be one more step into that direction."

(389) Telefónica's Europe's CEO's statement supports, in the Commission's view, that the other nation-wide operators do not have an incentive to make-up for the loss of competition by Jazztel in several ways. First, it shows that Telefónica expects consolidation to lead to a "more rational market". Secondly, Telefónica does not only assume that to be a mere result of the transaction but will also aim to "drive the market" that direction. It is clear that Telefónica will not compete aggressively if the Proposed Transaction is implemented.

(390) Furthermore, the opinion supports the assessment of the Commission concerning Orange's reduced incentives to compete after the merger. It also expresses that the same reasoning that let the Commission conclude that the merged entity will have lesser incentives to compete also applies to Telefónica and Vodafone. Both face the same trade-off when deciding whether to gain new customers or to protect the margins of their current customers.

(391) In the view of the Commission, this statement also shows that Telefónica does not expect either Orange or Vodafone to compete aggressively after the merger between Orange and Jazztel. More precisely, Mr. Álvarez-Pallete López argues that Orange and Vodafone will have "higher ARPUs" to defend as well as "being more focused on defending their core customer segments".

(392) The finding of a less competitive behaviour on behalf of Telefónica is supported by the fact that on 5 February 2015, Telefónica reportedly announced a price increase for its fixed broadband products that are not bundled with a mobile or TV component, applicable as from 7 April 2015. Prices for its ADSL products will increase by EUR 3, an increase of 10-12%. Telefónica will also raise prices for optical fibre capped at 100 Mbps by EUR 2 or 4%, from EUR 48 to EUR 50. That increase demonstrates Telefónica's intention to drive the retail market of fixed Internet access services toward higher prices. Given the price increases for broadband connections that are not bundled with mobile or TV services, it can be expected that Telefónica would not counter any price increases for fixed Internet access products by the merged entity.

(393) That the pricing decision was not taken to facilitate the migration of customers to quadruple-play products, as argued by the Notifying Party, is demonstrated by the subsequent price increase of all quadruple-play services. Effective as of early May 2015, Telefónica increases the price of all Fusion products by EUR 5 per month, a

[Reference to the Notifying Party's internal business documents]*.

price increase of 12% from EUR 42 to EUR 47 for the least expensive Fusion product. That additional price increase by a roughly equal percentage than the price increase for dual-play products demonstrates that Telefónica does not intend to facilitate the migration of customers to quadruple-play products through those price increases but rather increases the prices for all products including a fixed Internet access service.

(394) As to Vodafone/ ONO, both the CEO and the European CEO of Vodafone have stated that consolidation in Spain would permit Vodafone/ ONO to raise prices and that today the current situation is hindering such price increases. Philipp Humm (Regional CEO, Europe of Vodafone) said: "[W]e are still faced with a very aggressive price in bundle proposition in Spain. As long as this is the case, it's very hard for us to raise prices as TV has been right now given in for free in the big bundles from Telefonica. As this things [sic] maybe change over time, we should be able to increase prices; and then, we are then in a very good position again here with Ono.".

(395) Philipp Humm added that Vodafone had already "carried out quite a lot of price increases in the last times, be it in KD [Kabel Deutschland in Germany], on fixed line in Spain…". Responding to a question enquiring about possible price increases by Telefónica in Spain, Vittorio Colao (Group CEO Vodafone) noted that "[W]ith more consolidation in Spain, with more investment in 4G, with more homes reached by fiber from us and Orange, in theory, longer term, even Spain could follow a more kind of healthy path." .

(396) On 2 March 2015, Vodafone's international CEO Vittorio Colao said at the Mobile World Congress in Barcelona that he would like the Spanish market to be more profitable to support investments. In this regard, he considers it necessary to increase the prices of these telecommunications services if the strong pace of investments that companies are making is to be maintained, especially in the deployment of the 4G mobile network and fibre optical network.

(397) In the view of the Commission, that statement of Vodafone's international CEO further shows that Vodafone has no incentive to compete aggressively. It complements the statements of Philipp Humm (Regional CEO, Europe of Vodafone) and the earlier statement of Vittorio Colao quoted in recital (395).

[Reference to the Notifying Party's internal business documents]*.

(398) As regards the internal documents of Orange, the statements […]*, […]*, already imply that also in the view of Orange, the loss of competition by Jazztel is not completely offset by the other remaining operators.

(399) Specifically regarding low-cost, converged offers, internal documents of Orange suggest that […]* .

(400) The same document sets out that other players have difficulties in capturing relevant market shares with low-cost, convergent offers. With regard to […]* .

(401) The documents go on explaining that it is also difficult for "ethnic players", that is to say, in particular mobile operators that focus on specific ethnic minorities, to address the low-cost, convergent part of the market .

(402) As regards the argument of the Notifying Party that Telefónica is a strong player that has considerable market power, the Commission notes that the existence of strong players before and after the transaction is not sufficient to assume that the merger would not significantly impede effective competition in the respective markets. In the view of the Commission, the statements by Telefónica quoted in this section and the announced price increases by Telefónica of up to 12% for all products including a fixed Internet access service demonstrate that Telefónica will not make-up for the loss of competition caused by the Transaction.

(403) While thus Telefónica and Vodafone lack the incentives to make-up for the loss of competition caused by the Proposed Transaction, other market participants and in particular the three regional cable operators lack the ability to do so. The three regional cable operators are each only active in one region of Spain. Together, they account for less 5% of all subscribers to fixed Internet access services in Spain. The Commission deems their limited regional presence to be too small in order to make-up for the loss of competition by a nation-wide competitor since customers are only able to procure fixed Internet access services from an operator that is present where the services are to be provided.

Conclusion on the likely reaction of competitors

(404) The Commission concludes that competitors of the merged entity, namely Telefónica and Vodafone/ ONO, are unlikely to counter potential price increases of the merged entity after the Proposed Transaction.

7.2.2.6. Quantitative analysis

(405) The quantitative analysis presented in section 7.2.7 contains an assessment of the likely effect of the transaction on the average price of products including a fixed Internet broadband component . Even though that assessment predicts price increases for product bundles that include not only fixed Internet access services, the Commission deems that the results of the analysis in section 7.2.7 apply to fixed Internet access for several reasons.

(406) As regards the data that is used for the quantitative analysis, bundles consisting of several products that are sold for one joint price constitute the vast majority of all

[Reference to the Notifying Party's internal business documents]*.

fixed Internet access services. Looking at new contracts for fixed Internet access services, they even represent more than 99% of all contracts. Excluding such bundles would disregard the vast majority of fixed Internet access services.

(407) Furthermore, taking into account the prices for bundles is the only way to properly reflect the actual market conditions in Spain. Pricing decisions of the market participants are made on the basis of the price and the margin of the bundle, not on hypothetical prices of the included components. The overall price cannot reasonably be split into prices for individual components as any allocation of the overall price to the individual components would be arbitrary.

(408) As regards the results of the quantitative analysis, the Commission notes that the likely price increases predicted in section 7.2.7 are expressed in terms of percentages and not in terms of absolute amounts. In light of this, and considering the difficulties of disaggregating the price of the bundle into the prices of the included components, for the purpose of this Decision the Commission considers it reasonable to assume that the same relative price increases apply to each of the components included in the relevant bundles (including the fixed Internet access component).

(409) The Commission further notes that the Notifying Party itself assessed potential price effects on the market for fixed Internet access services by taking account of all bundles that include such services and by including the entire revenues and costs of bundles into the analysis.

(410) According to the baseline scenarios, the average price increase across all operators and across all products including a fixed broadband component is in the range of 3 to 7% based on 2013 data and 3 to 6% for 2014 data. The sensitivity scenario using diversion ratios that are exclusively based on purely horizontal switches (that is, customer's switching across operators but within a given product type) indicates that the predicted price increases could also reach up to 10% according to 2013 data and 7% for 2014 data.

(411) Those average price increases are a combination of a significant expected price increase for triple and quadruple-play products and a more limited expected price increase for dual-play products. As explained in section 7.2.7, no price increases have been computed for fixed broadband only products (possibly including mobile as well), as those products only represent 1% of the new contracts including a fixed broadband component.

(412) The lower bound of the estimated price increases has been obtained by including in the analysis an assumption of strong diversion to the outside good (that is, the hypothesis whereby, following a price increase, customers may choose not to subscribe to any fixed broadband product at all, instead of simply switching provider) equivalent to the one put forward by the Notifying Party. The Commission considers that such an assumption is rather extreme, mainly because the elasticity and the modelling assumed by the Notifying Party lead to rather implausible results . In incorporating this rather extreme assumption reduces significantly the estimated price effects but not to an extent that allows dispelling any concern of potential price increases post-merger. The price increases from the analysed transaction are expected

The estimated price increases from the baseline scenarios presented in the Statement of Objections were 5-10%.

(413) The details of the robustness checks of the Commission's quantitative analysis and the Commission's responses to the arguments raised by the Notifying Party are set out in in section 7.2.7 and Annex A. In particular, the sensitivity analysis conducted on the effect of the assumption of no cross bundle diversions (section 6.2 of Annex A) shows that while considering that cross bundle diversions slightly increase the price effects for 2p and slightly decreases the price effects for 3p, it does not materially affect the overall price increase expected on average across all products including a fixed broadband component.

7.2.2.7. Limited likelihood of sufficient entry

(414) In section 7.2.8.1, the Commission sets out why it deems that sufficient entry to markets including fixed Internet access services is unlikely. Based on the analysis in that section, the Commission concludes that the likely effects of the Proposed Transaction on the retail market for fixed Internet access services are not offset by low barriers to enter this market.

7.2.2.8. Conclusion on horizontal non-coordinated effects on the market for fixed Internet access services

(415) In light of recitals (185) to (414), the Commission considers that the Proposed Transaction would give rise to non-coordinated anti-competitive effects because it involves, in an already highly concentrated market, the elimination of important competitive constraints that the merging parties previously exerted upon each other together with a reduction of competitive pressure on the remaining competitors. Moreover, as discussed below in section 7.2.8.2 and 7.2.4.7, those non-coordinated anti-competitive effects are not offset by countervailing factors such as efficiencies. Therefore, the Commission considers that the Proposed Transaction would significantly impede effective competition on the market for fixed Internet access services in Spain.

7.2.3. Possible retail market for dual-play services

(416) In a possible market for dual-play services, Orange holds a market share in terms of subscribers of [10-20]*% and Jazztel of [10-20]*%. The Parties' combined market share would thus amount to [20-30]*%. As such, the combined market share of the Parties does not exceed the market share threshold indicated in paragraph 18 of the Horizontal Merger Guidelines, under which there is a presumption that the transaction is unlikely to impede effective competition.

7.2.3.1. Removal of two important competitive forces

(417) As to whether the two Parties can be regarded as important competitive forces in a possible market for dual-play services, the Commission notes that certain elements in the respective assessment for fixed Internet access services relate specifically to triple-play products while others also apply to dual-play products.

(418) As regards Jazztel, the Commission notes that the significant growth of Jazztel from a share in the overall market for fixed Internet access services of 3% in 2007 to 11% in 2012, as noted in a publication of the CNMC in October 2012, was mainly due to

[Reference to the Notifying Party's internal business documents]*.

its success with dual-play offers as those offers were the standard proposition until 2012 .

(419) Furthermore, the questions in the Market Investigation enquiring about the most aggressive or the most innovative market participant did not distinguish between different multiple play offers. Therefore, the respective conclusions from the Market Investigation as set out in section 7.2.2.3 similarly apply to the possible market for dual-play products.

(420) Other elements in the assessment of the importance of Jazztel for the competition in the market for fixed Internet access products, however, do not apply to dual-play products. For example, Jazztel's strategy to provide one or even two mobile lines for free (as part of a triple-play offer) to win customers and to decrease customer churn does only apply to triple-play products. Also the internal Orange documents relating to Jazztel's convergent offers do not relate to Jazztel's competitive position in a possible market for dual-play offers.

(421) As regards the role of Orange in a possible market for dual-play offers, the Commission notes that while Orange's success in recent years is most likely mainly based on dual-play offers up until at least 2012, the current strategy of Orange is centred around its converged offers.

(422) Overall, the Commission regards Jazztel and Orange as important competitive forces in the possible market for dual-play services. However, that importance is lower than their importance in the market for fixed Internet access services as not all elements of the assessment in section 7.2.2.3 also apply to dual-play products.

7.2.3.2. Change in the merged entity's incentives to compete

(423) The elements of the assessment presented in section 7.2.2.4 do also apply to the merged entity's incentives to compete on a possible market for dual-play offers. In particular, Orange's internal documents cited in that section indicate that Orange expects higher average revenues per user following the transaction. Such internal documents of Orange do not distinguish between fixed Internet access services or dual-play services. They only separate effects related to fixed products and to mobile products. Therefore, the conclusion of the Commission in section 7.2.2.4 also applies to dual-play products.

7.2.3.3. Likely reaction of competitors

(424) Similarly, the likely reactions of competitors also stay the same if assessed in the context of a possible market for dual-play offers. Neither the internal documents of Orange nor the quotes from Telefónica and Vodafone distinguish between specific telecommunication products.

7.2.3.4. Quantitative analysis of horizontal non-coordinated effects

(425) The quantitative analysis presented in section 7.2.7 contains an assessment of the likely effect of the transaction on the average price of dual-play products.

(426) According to the baseline scenarios, the average price increase for dual-play products across all operators is approximately 1.2% to 4.5% and 0.6 to 1.9% for 2013 and 2014 respectively. The price increases for the Parties are in the range of 2% to 9% and 1% to 6%, according to 2013 and 2014 data, respectively).

(427) Those price increases are significantly lower than the predicted prices increases for triple- and quadruple-play offers and are thus also lower than the overall price increases predicted for all products containing a fixed Internet access component.

(428) The lower bound of the estimated price increases has been obtained by including in the analysis an assumption of strong diversion to the outside good equivalent to the one put forward by the Notifying Party . Incorporating that rather extreme assumption reduces significantly the estimated price effects but not to an extent that allows dispelling any concern of potential price increases post-merger. The price increases from the analysed transaction are expected to be closer to the upper bound of the Commission's estimates than to that purely illustrative lower bound.

(429) The details of the robustness checks and the Commission's responses to the arguments raised by the Notifying Party are set out in in section 7.2.7 and Annex A. In particular, in light of the sensitivity analysis conducted on the effect of the assumption of no cross bundle diversions (section 6.2 of Annex A), the Commission considers that price effects for 2p products could be slightly higher than those suggested by the baseline scenarios and in the range of 3% to 6% and 2% to 3% for 2013 and 2014 respectively.

(430) Therefore, the light of recitals (425) to (429) and in particular the results of the baseline scenario, it cannot be shown with sufficient likelihood that the Proposed Transaction will lead to a significant price increase in the possible market for dual-play services.

7.2.3.5. Limited likelihood of sufficient entry

(431) In section 7.2.8.1, the Commission sets out why it deems that sufficient entry to markets including fixed Internet access services is unlikely. Based on the analysis in that section, the Commission concludes that the likely effects of the Proposed Transaction on the possible market for dual-play services are not offset by low barriers to enter that market.

7.2.3.6. Conclusion horizontal non-coordinated effects on the possible market for dual-play services

(432) In conclusion, the Commission finds that the competitive concerns raised with regard to a possible separate retail market for dual-play services are less strong but still present. Indeed, in such a market for dual-play services, the Parties are overall less aggressive but still important competitive forces.

(433) In particular, the Commission concludes that the Proposed Transaction will remove two important competitive forces, reduce the merged entity's incentives to compete without offsetting such loss of competition by additional competitive pressure from either existing or new competitors. Therefore, the Commission concludes that the Transaction raises significant competitive concerns also in a possible market for dual-play services.

7.2.4. Possible retail market for triple- and quadruple-play services

(434) In a possible market for triple- and quadruple-play services, Orange holds a market share in terms of subscribers of [10-20]*% and Jazztel of [10-20]*% calculated on the basis of the Notifying Party's estimates and CNMC data. The Parties' combined market share would thus amount to [30-40]*%. As such, the combined market share in terms of subscribers of the merged entity in a possible market for triple- and multiple play offers is slightly higher than in the market for fixed Internet access services discussed in section 7.2.

(435) The Commission considers that the Proposed Transaction would significantly impede effective competition on the possible market for triple- and quadruple-play services in Spain. The Proposed Transaction would result in the removal of the Parties as two important competitive forces on that possible market, and in the reduction from four to three of the number of convergent players active at national level.

(436) In that respect, the Commission considers that (i) each of the Parties is an important competitive force in the market, (ii) the merged entity would have less incentives to compete post-transaction, (iii) the Parties' competitors will have reduced incentives to counter the likely reduction of competitive pressure of the merged entity and, (iv) entry of new players in response to price increases into the possible retail market for triple-play services is unlikely. Generally, the Commission's assessment of the Spanish retail market for fixed Internet access services applies to the possible market comprising triple- and quadruple-play services. To the extent the assessment of the possible market comprising triple- and quadruple-play services is based on different or additional considerations than the competitive assessment of fixed Internet access services, the Commission will explain its findings in recitals (437) to (491).

7.2.4.1. Removal of two important competitive forces

(437) Most of the Commission's assessment and conclusions listed in section7.2.2 relating to the retail market for fixed Internet access services in Spain apply to the possible retail market for triple- and quadruple-play services. The competitive concerns expressed in the market for fixed Internet access services are even further compounded as regards the possible market for triple- and quadruple-play services for the reasons set out in recitals (438) to (441).

(438) Both Parties are described in [Reference to the Notifying Party's internal business documents]*, as the most successful companies in acquiring new customers for convergent services: Jazztel is mentioned as most successful with acquiring […]*% new customers for triple-play services followed by Orange with the acquisition of […]*% of all new customers at the end of 2013 .

(439) That finding is in line with the Parties' gross add shares, as calculated by the Commission. Gross add shares for a possible market comprising triple- and quadruple-play services indicate that Orange and Jazztel are even more important competitive forces in triple- and quadruple-play services than in fixed Internet access services.

as presented in Table 6. According to data the Commission collected from all operators, the market share in a market for triple- and quadruple-play services of the Parties amount to [10-20]*% and [10-20]*% and the combined market share to [30-40]*%.

(440) According to data the Commission gathered from all providers of triple- and quadruple-play services, the Parties acquired more than half of all customers who concluded a new contract for triple- or quadruple-play services in 2013 and 2014. In 2013, the combined gross add share of the Parties amounts to […]*%, in 2014 they acquired […]*%. Those gross add shares are significantly higher than their combined market shares of […]*% in 2013 and […]*% in 2014.

(441) In conclusion, the Commission considers that the Parties are particularly important competitive forces in the possible Spanish market for triple- and quadruple-play offers, even more significantly than in the market for fixed Internet access services discussed in section 7.2.

7.2.4.2. Change in the merged entity's incentives to compete

(442) Regarding the incentives of the merged entity to compete, the Commission notes that the assessment set-out in section 7.2.2.4 for fixed Internet access services also applies to a possible market for triple- and quadruple-play services.

(443) Moreover, the internal forecasts of the Notifying Party indicate that the effect of the transaction on ARPUs for mobile component is even stronger than on ARPUs for fixed components. Therefore, according to those estimates, the overall impact on multiple play services is stronger than on fixed Internet access services only.

(444) In conclusion, the Commission finds that the assessment set-out in section 7.2.2.4 also evidences that the merged entity will have less incentives to compete in a possible market for triple- and quadruple-play services in Spain after the Proposed Transaction.

7.2.4.3. Likely reaction of competitors

(445) The Notifying Party submits that other strong players, like Telefónica and Vodafone, will remain active at national level. The Notifying Party describes Telefónica as an aggressive player, which introduced its "Fusión" offer in October 2012. Vodafone would also be a strong competitor, in particular as a result of its access to Pay-TV services after the acquisition of ONO. In addition, other competitors in the triple- and quadruple-play market are the regional cable operators, as regards the respective regional markets where they are active.

(446) Moreover, the Notifying Party submits that the Proposed Transaction will not have any anticompetitive effects on the possible market for multiple play services because:

(a) there would be a large number of alternative operators, including Telefónica, Vodafone/ONO, MVNOs and the three regional cable operators;

(b) Telefónica would be the undisputed leader; and

(c) the merged entity will be in a better position to compete with the leading operator, Telefónica.

(447) The Commission finds, however, that the assessment of the likely reactions of competitors conducted in section 7.2.2.5 also applies to a possible market for triple- and quadruple-play products. The quotes of Telefónica and Vodafone cited in that section expressly refer to convergent products. Furthermore, the Commission notes that […]* predicted by Orange as a result of the transaction already factor in the reactions of the Parties' competitors. […]*.

(448) Therefore, the Commission concludes that competitors of the Parties will not make-up for the loss of competition caused by the transaction in the possible retail market for triple- and quadruple-play products in Spain.

7.2.4.4. Quantitative analysis of horizontal non-coordinated effects

(449) The quantitative analysis presented in section 7.2.7 contains an assessment of the likely effect of the transaction on the average price of triple- and quadruple-play products.

(450) According to the baseline scenarios, the average price increase for triple and quadruple-play products across all operators is approximately 5% to 11% and 4% to 10% for 2013 and 2014, respectively. The price increases for the Parties are in the range of 7% to 18% and 5% to 14%, according to 2013 and 2014 data, respectively.

(451) The details of the robustness checks and the Commission's responses to the arguments raised by the Notifying Party are set out in Annex A. In particular, in light of the sensitivity analysis conducted on the effect of the assumption of no cross bundle diversions (see section 6.2 of Annex A), the Commission considers that the price effects for triple- and quadruple-play products could be slightly lower than those suggested by the baseline scenarios.

(452) The lower bound of the estimated price increases has been obtained by including in the analysis an assumption of strong diversion to the outside good equivalent to the one put forward by the Notifying Party . Incorporating that rather extreme assumption reduces significantly the estimated price effects but not to an extent that allows dispelling any concern of potential price increases post-merger. The price

This is discussed in more detail in section 5.7 of Annex A.

increases from the analysed transaction are expected to be closer to the upper bound of the Commission's estimates than to the purely illustrative lower bound.

(453) Overall, the Commission's quantitative assessment indicates that the merger will likely lead to significant price increases in the possible Spanish retail market comprising triple- and quadruple-play services. Those effects are even more pronounced than the price effects found in the retail market for residential fixed internet access services.

(454) The larger price effects in the possible Spanish retail market comprising triple- and quadruple-play services compared to the overall market for residential fixed internet access services (and in particular compared to a possible market for dual-play services) are due to a number of factors. In particular, the Parties' diversion ratios and benchmark diversion ratios based on gross add shares are systematically significantly higher for the triple- and quadruple-play product type compared to those for the dual-play product type. Those high diversion ratios for triple- and quadruple-play, coupled with the higher margins earned by the Parties in triple- and quadruple-play services compared to dual-play services, determine price effects for triple- and quadruple-play services which are more pronounced compared to dual-play services.

(455) It is also noted that the more pronounced price effects for triple- and quadruple-play services are consistent with Orange's own internal documents which […]* .

(456) The Commission concludes that the Proposed Transaction will likely lead to significant price increases in the possible market for triple- and quadruple-play services.

7.2.4.5. Limited likelihood of sufficient entry

(457) In section 7.2.8.1, the Commission sets out why it deems that sufficient entry to markets including fixed Internet access services is unlikely. In section 7.2.8.2, it is explained why entering multiple play markets that also include a mobile component is confronted with even higher barriers to entry.

(458) Based on the analysis in the sections mentioned in the previous recital, the Commission concludes that the likely effects of the Proposed Transaction on the retail market for triple- and quadruple-play services are not offset by low barriers to entry.

7.2.4.6. Future competitive strength of the Parties in light of the growing share of quadruple-play bundles

View of the Notifying Party

(459) Given that the Parties' position in providing triple-play services is much stronger than their combined position in the provision of quadruple-play services ([30-40]*% of all triple-play bundles are sold by either Orange or Jazztel compared to only [0-5]*% of all quadruple-play bundles (see Table 7), in the Reply to the Statement of Objections the Notifying Party argues that the growing number of quadruple-play offers will limit the competitive strength of Orange and Jazztel in the future.

(460) The Notifying Party submits that Telefónica would already have (and would be further developing) a strong position in the Pay-TV market because (i) it would acquire directly premium content and (ii) it would be currently acquiring the leading Pay-TV provider in Spain, Distribuidora de Television Digital, S.A, ("DTS").

(461) Furthermore, in the Notifying Party's opinion, Vodafone/ONO would have a strong position in Pay-TV as well, thanks to the acquisition of the leading cable operator ONO. In that respect, ONO would have a solid customer base for the provision of Pay-TV services.

(462) The Notifying Party also submits that all regional cable operators have a customer base of Pay-TV services larger than the one of Orange.

(463) The Notifying Party submits that the merged entity would be in a weaker position if compared to Telefónica and Vodafone/ONO as regards VHBB and Pay-TV services and, thus, would have no incentives to raise prices post-transaction. As a result the merged entity would face increasing difficulties to attract customers for its triple-play services.

Results of the Market investigation

(464) When asked about the future take-up of quadruple-play services in Spain, the respondents to the Market Investigation indicated that the importance of quadruple-play services is likely to increase in the future. Half of the respondents to the Market Investigation indicated that they expect quadruple-play services to attain the same level of popularity as triple-play services in 2 to 3 years. A minority of respondents expects quadruple-play services to attain a higher uptake than triple-play services, and an equal minority to attain a lower uptake than triple-play services. A majority of respondents also indicated that they expect quadruple-play services to achieve higher take-up then dual-play services (comprising fixed Internet access services and fixed telephony services) or at least the same level of take-up .

(465) The reported reasons for such an increased take-up of quadruple-play services are two-fold. First, telecommunications operators use Pay-TV as a differentiator in their service offers. Second, the demand for premium content would drive the take-up .

(466) Furthermore, respondents to the Market Investigation took the view that the six TV-channels offered by Telefónica in its entry-level Fusión quadruple-play offer (namely, Fox, AXN, Calle 13, Comedy, Cosmo, Fox Life) are generally not crucial to successfully sell multiple play offers with a TV component in Spain. In that respect, respondents generally indicated that the success of quadruple-play services in Spain rather depends on the availability of premium or exclusive content and especially content related to sports events and in particular football. Moreover, all respondents that currently offer Pay-TV services indicated that they currently offer all or some of the channels mentioned in this recital .

Commission’s assessment

(467) As regards the role of Pay-TV services in Spain, the Commission considers that the future relevance of a Pay-TV offer for competing successfully in a possible market for triple- and quadruple offers is uncertain. It will depend on many factors like the attractiveness of bundled Pay-TV services, the price for such services, the discount for bundling Pay-TV services with triple-play services, and whether there will be any popular content that could be accessible exclusively as part of bundled offers.

(476) In this respect, the Commission notes that on 22 April 2015 the CNMC approved the proposed acquisition of DTS by Telefónica with conditions and obligations. In its decision, the CNMC conditioned its approval of the proposed acquisition of DTS by Telefónica to three major sets of commitments, relating to commercial conditions for Telefónica's Pay-TV customers, wholesale access to Telefónica's exclusive television content and access to Telefónica's Pay-TV clients to providers of Over-the-top services.

(477) As regards access to its exclusive television content, the combined Telefónica/DTS entity committed to provide a wholesale access to its premium channels at cost. In the framework of those commitments, a television channel is considered "premium" if it contains non-sports-related content produced by any of the Hollywood "Majors" (Sony, Warner, Disney, Universal, Paramount and Fox) or if it contains live sports-related content, on which Telefónica enjoys exclusive broadcasting rights for Spain. The totality of the "premium" channels of the combined Teléfonica/DTS will be made available through this wholesale offer, although each acquirer will be able to purchase a maximum of half of them. Prices for wholesale access to such television content will be generally at cost, and will be supervised by the CNMC.

(478) The Commission notes that access to television content in the Spanish market will be facilitated by the commitments imposed by the CNMC over the combined Telefónica/DTS entity. However, the Commission considers that, despite the formal set of commitments entered into by Telefónica, it is too early to take a view on the effectiveness of such commitments to allow alternative providers of Pay-TV services in Spain to compete effectively on the market.

(479) Furthermore, the Commission notes that, in its internal documents, […]*, as explained in recitals (480) and (481) here below.

(480) By means of example, an internal document of Orange Spain intended to respond to Orange Group's questions as regards the group’s strategic review of the Spanish business plan for 2015 to 2018 reveals that the acquisition of DTS by Telefónica […]*. Orange foresees two different strategic options in reaction to Telefónica's proposed acquisition of DTS. […]*

(481) Moreover, and consistently with the Commission's Market Investigation findings described in recital (466) above, […]*

(482) As an additional element, the Commission notes that, with the exception of Telefónica, currently almost all the Spanish players offering multiple play services comprising Pay-TV (Vodafone/ONO, Orange, Euskaltel, R Cable, Telecable) offer almost the same TV channels, despite having completely different market shares in terms of Pay-TV subscribers.

(483) Therefore, the Commission concludes that access to TV content is not an obstacle to provide quadruple-play services, given that Orange itself does not consider it an issue.

(484) As regards the necessity to offer fibre products in order to be able to offer Pay-TV within quadruple-play services on competitive terms in Spain, the Notifying Party submits that the Proposed Transaction will allow the merged entity to become a stronger player against Telefónica, in particular as regards the deployment of its own FTTH network.

(485) The Commission disagrees with the Notifying Party's argument for two reasons.

(486) Firstly, as regards the combination of Orange and Jazztel's fibre networks, the considerations explained in section 7.2.9 apply. In that regard, the merged entity will not deploy a FTTH network with a significantly larger footprint than the combined footprint of Orange and Jazztel absent the merger. In that respect, the Commission has concluded in section 7.2.9 that, contrary to the Notifying Party's claim, it is unlikely that the Proposed Transaction leads to any significant increase in the FTTH coverage by the merged entity, as compared to the standalone scenario. On the other hand, the Proposed Transaction removes a competitive player on fibre in the areas where the footprints of the fibre networks of Orange and Jazztel would have overlapped.

(487) Secondly, as regards the importance of fibre for a fixed network operator to be able to provide Pay-TV services, the Commission notes that currently each of the Parties are able to provide Pay-TV services over their existing xDSL network.

(488) According to data submitted by the Notifying Party, from a technical point of view, Table 23 sets out the bandwidths needed in order to stream IPTV channels.

Table 23: Bandwidth associated with IPTV channels

Quality of channel Bandwidth

1 channel 3/3.5 Mbps

2 channels simultaneously 6/7 Mbps

High quality Approx. 4.5 Mbps

7 Mbps

Source: Notifying Party

(489) At the moment, both Orange and Jazztel offer fixed Internet access services with maximum speeds above 10 Mbps. In particular, Orange offers xDSL Internet with a maximum download speed of 20 Mbps and Jazztel of 30 Mbps. That means that currently, even if the actually observed speeds are lower than the advertised maximum speeds, the Parties are able to offer simultaneously high quality Pay-TV services and fixed voice services with enough bandwidth left to allow a user to access the Internet at broadband speed.

(490) To further support that finding, Orange and Jazztel currently offer (or in the case of Jazztel, distributes) IPTV services over their xDSL networks. Therefore, the current download speed offered by each of Orange and Jazztel would be already sufficient to provide quadruple-play services.

(491) The Commission concludes, first, that the Proposed Transaction will not have any impact on the quadruple-play market. In second place, the Commission considers that, contrary to the Notifying Party's argument, the findings set out in sections 7.2.4.2, 7.2.4.3 and 7.2.2.5 above as regards incentives which the merged entity and its competitors are likely to have post-transaction apply in the same way to the possible market for triple-play services.

7.2.4.7. Commission's assessment of the Notifying Party's efficiencies claims

(492) The reasons provided in recital (703) also apply to triple- and quadruple-play offers. The elimination of double marginalisation of mobile services provided by Orange to Jazztel applies in particular to triple- and quadruple-play offers as they include a mobile component (see section 7.2.10.4). As set out in section 7 of Annex A, the predicted price effects of 3p/4p products can be expected to decrease by approximately 0.8 percentage points, but the remaining expected price increases are still significant.

(493) Therefore, the Commission concludes that the claimed efficiencies are not such to balance identified anti-competitive effects.

7.2.4.8. Conclusion on horizontal non-coordinated effects on the possible market for triple- and quadruple-play offers

(494) The Commission considers that the Proposed Transaction would give rise to non-coordinated anti-competitive effects on the possible retail market for fixed-mobile triple- and quadruple-play services in Spain because it involves, in an already highly concentrated market, the elimination of important competitive constraints that the merging parties previously exerted upon each other together with a reduction of competitive pressure on the remaining competitors. Moreover, those non-coordinated anti-competitive effects are not offset by countervailing factors such as efficiencies. Therefore, the Commission considers that the Proposed Transaction will significantly impede effective competition on the possible market comprising triple- and quadruple-play services in Spain.

7.2.5. Possible retail market for triple-play services

(495) On the possible market for triple-play services, the Parties' combined market share would account for almost [40-50]*% of the total subscribers (Orange: [20-30]*%; Jazztel: [10-20]*%) at the end of 2014. Other main players active on the possible market for triple-play services are Telefónica and Vodafone. Post-transaction,

third quarter of 2014, available at http://www.minetur.gob.es/telecomunicaciones/es-ES/Servicios/CalidadServicio/informes/Documents/Seguimiento_SAI_T3_14.pdf.

(496) The position of the Parties on a possible market for triple-play services is thus much stronger than their combined market of [30-40]*% (Orange: [10-20]*%; Jazztel: [10-20]*%) at the end of 2014 in a possible market for triple- and quadruple-play services assessed in section 7.2.4. Consequently, the Proposed Transaction would give rise to even more serious concerns on a possible market for triple-play offers, as both Parties mainly offer triple-play services. Therefore, the assessment of the possible market for triple-plays services is not presented separately in more detail.

7.2.6. Possible retail market for multiple play services

(497) The position of the Parties in a general market for all multiple play services would be almost identical to their position in the market for fixed Internet access services as the share of fixed Internet access services provided as a stand-alone service outside of a bundle is negligible. Therefore, an assessment of a general market for multiple play services reaches the same result as the competitive assessment in section 7.2.2 for fixed Internet access services.

7.2.7. Quantitative analysis of horizontal non-coordinated effects

(498) The Commission has also carried out an assessment of the extent to which the elimination of competition between the Parties will generate an incentive to increase price for the merged entity post-concentration. The Commission applied a calibrated merger simulation approach that uses observed diversion ratios between competitors and observed margins and shares of new customers to calibrate demand and predict the likely price effects of the concentration. In the context of differentiated products markets, similar approaches were used in previous cases . The Commission's quantitative analysis is described in more detail in Annex A to this decision.

(499) The incentive for the Parties to raise prices arises because the concentration eliminates competition between them. More precisely, since the Parties offer substitute products, following a unilateral price increase by one merging Party, a number of subscribers would switch to products offered by the other merging Party. Pre-merger, from the viewpoint of one merging Party, customers that switch to products of the other merging Party are lost after a price increase. Post-merger, customers that switch to products that were previously offered by the other merging Party (and after the merger are offered by the merged entity) increase the profits of those products. Put differently, the threat of losing consumers to competitors is reduced post-merger since some of the alternative products that pre-merger were owned by a competitor are offered post-merger by the merged entity. Since some of the demand that would be lost following a price increase pre-merger will thus be recaptured by the merged entity post-merger, it will be more profitable to increase price post-merger relative to the non-merger scenario.

(500) The analysis of the Commission aims at quantifying how strong the Parties' incentives to raise price are as a consequence of the loss of competition. In that respect, the calibrated merger simulation is used to quantify the "most direct effect" of a merger and the additional effect that, in reaction to price increases by the merged entity, the Parties' rivals could also increase price, which are described in paragraph 24 of the Horizontal Merger Guidelines. The calibrated merger simulation abstracts from other factors (such as, for example, barriers to entry, countervailing buyer power and repositioning by rivals) which could affect the merged entity's ability and incentives to raise prices.

(501) The Commission's analysis focusses on two product types, namely (i) the double play product type consisting in fixed voice and fixed broadband and (ii) an aggregation of triple and quadruple-play services consisting in double play plus mobile and possibly TV . All of those product types offer fixed broadband as part of the bundle . Other product types included in the market for fixed Internet access, such as fixed broadband only products and fixed broadband plus mobile products have not been explicitly modelled in the analysis, as they represent approximately 1% of the retail market for fixed Internet access services, as detailed in Table 1. The analysis is therefore directly relevant to the market for fixed Internet access, since that product is practically always sold as part of the bundles considered in the quantitative assessment.

7.2.7.1. Measurement of diversion ratios

(502) Diversion ratios provide a measure of the extent to which one operator is able to attract the sales that another operator would lose as a result of a price increase. The diversion ratios thus capture the competitive constraint that the Parties impose on one another pre-concentration and which is lost as a result of the concentration.

(503) The main source used for the measurement of diversion ratios is the Fixed Number Portability ("FNP") data of the Asociación de Operadores para la Portabilidad

(504) The Commission has addressed that issue by complementing the number portability data with information on the number of migrations relating to WLR, which has been provided by Telefónica. The details of that adjustment are presented in section 3.1.2 of Annex A.

(505) The Commission has also analysed more disaggregated portability information (at the product-type level) collected directly from the operators. That information complements the FNP because it allows separating the business segment from the residential segment and allows identifying the origin and destination product-type of the port (that is to say the product type the customer ports from and to).

(506) Each of Orange, Jazztel, Telefónica and Vodafone has been asked to provide quarterly data on subscribers that ported their number to and from each operator from and to each of its rivals. For each of port-in, the Commission asked to provide information on the operator of origin and the new product type that the customer subscribed to. For each port-out, the Commission requested information on the operator of destination and the product type to which the customer was subscribed before leaving .

(507) The Commission considers that data covering actual switching events from each operator to the other operators is informative on the extent of substitution patterns between different operators. For instance, if one observes a large share of lost customers of Jazztel porting their number to Orange, that provides a good indication that Orange is an important alternative for Jazztel customers, even if some of the observed switches are not driven by price changes.

(508) There was a significant shift from 2p to 3p/4p products in 2013 and 2014 (both internally within each operator and externally, across operators). In contrast to observed horizontal switching, vertical switching across product types may be less likely to reflect consumer reactions to marginal price changes compared to observed switching across operators (especially within the same product type). Because the observed vertical diversion ratios do not necessarily reflect consumer reactions to marginal price changes, in its calibrated merger simulation the Commission assumes that no price based switching from one product type to another occurs. The analysis is thus focused on the horizontal choice between operators .

Report "Comments on the use of UPP / IPRs in the 6(1)(c) decision" of 16 January prepared by the Notifying Party.

Wholesale Line rental is known in Spain as "Acceso Mayorista a la Línea Telefónica (AMLT)". The terms WLR and AMLT will therefore be used interchangeably in the remainder of this document.

Relating to the first month of each quarter only (for the period Q1 2013-Q4 2014).

The Commission has applied to this disaggregated product type-level portability data a similar correction to the one applied to the FNP data from the AOP.

That is, customer's switching across operators but within a given product type.

For example switches from a double play product to a convergent product.

This assumption is further motivated by the fact that many subscribers who switched from 2p to 3p/4p products would have presumably switched to 3p/4p products even if the prices of 3p/4p products had been marginally higher (that is, for these switchers the original 2p products were unlikely to be the second best choice anymore).

In order to isolate switches which are purely horizontal, the Commission has matched the ported-out numbers (and the corresponding operator of destination) of all operators with the numbers ported-in to all operators (and corresponding operator of origin). That provides for each switcher a comprehensive view of (i) the operator of origin, (ii) the product type of origin, (iii) the operator of destination and (iv) the product type of destination. The Commission then retained only those ports whose origin and the destination product type were identical.

Based on that information, the Commission has calculated the diversion ratios for switching which occurs within the same product type, that is to say from 2p to 2p and from 3p/4p to 3p/4p, which allows isolating those switching occurrences that are least likely to be driven by shifts in preferences over time, notably from lower value product types such as 2p to higher value product types such as 3p/4p .

The diversion ratios based on number ports across operators but within the same product type are summarised in Table 24.

Table 24: Diversion ratios matrix for 2p and for 3p/4p product types (based on switching within the same product type, adjusted for WLR)

2013 2014

2p 3p/4p 2p 3p/4p

Diversion from Orange to Jazztel [...]* [...]* [...]* [...]*

Diversion from Jazztel to Orange [...]* [...]* [...]* [...]*

Source: Commission calculation based on data obtained from operators

As Table 24 illustrates, the diversion ratios for 3p/4p product types between the merging parties are especially significant. The diversion ratios from Jazztel to Orange were […]*% and […]*% in 2013 and 2014, respectively. The diversion ratios from Orange to Jazztel were […]*% and […]*% in 2013 and 2014, respectively. This suggests that Orange and Jazztel particularly constrain each other's pricing for their 3p/4p offers.

The Commission has also calculated the diversion ratios by product type of destination (irrespective of the product type of origin).

Calculating the diversion ratios based on an aggregation of the migrations across all product types of origin, and conducting the merger simulation analysis by product type of destination (2p or 3p/4p), reflects the implicit assumption that consumers who switched vertically, would have only considered subscribing to the destination product type, but not the product type they were currently subscribed to. In the context of the present case, in light of the market developments driving the trend to 3p/4p products, it seems reasonable to assume that many subscribers who switched from 2p to 3p/4p products did not consider, at the time of their new choice, subscribing again to a 2p product.

Compared to the choice of matching the product type-level portability data, aggregating over the origin product types has the further advantage of increasing the number of switchers on which diversion ratios are based, thereby improving the reliability of the resulting estimated diversion ratios.

The resulting diversion ratios for the 2p product type and for the 3p/4p product type are summarised in Table 25.

Table 25: Diversion ratios matrix for 2p and for 3p/4p product type, based on operator data on product type-level portability (adjusted for WLR), aggregated across all product types of origin

2013 2014

2p 3p/4p 2p 3p/4p

Diversion from Orange to Jazztel [...]* [...]* [...]* [...]*

Diversion from Jazztel to Orange [...]* [...]* [...]* [...]*

Source: Commission calculation based on data obtained from operators

As regards the 3p/4p product type, the diversion ratios from Jazztel to Orange were […]*% and […]*% in 2013 and 2014, respectively. The diversion ratios from Orange to Jazztel were […]*% and […]*% in 2013 and 2014, respectively. That level of diversion again suggests that Orange and Jazztel significantly constrain each other's pricing for their offers in the 3p/4p product type.

The Commission has also derived benchmark diversion ratios based on the assumption that each operator attracts from the others a share of subscribers proportional to its gross add share. The diversion ratios between the merging parties are roughly of the same level as set out in Table 25 and are presented in Table 26 and below.

Table 26: Benchmark diversion ratios matrix for 2p and for 3p/4p product type (based on gross add shares)

2013 2014

2p 3p/4p 2p 3p/4p

Diversion from Orange to Jazztel [...]* [...]* [...]* [...]*

Diversion from Jazztel to Orange [...]* [...]* [...]* [...]*

Source: Commission calculation based on data obtained from operators

Overall, the level of diversion between the merging parties indicates that the two operators impose a significant competitive constraint on each other. As regards 3p/4p products, the observed diversion ratios suggest that the competitive constraint imposed by the merging parties on each other is comparable or larger than the constraint imposed by Vodafone and Telefónica, respectively. The loss of that constraint as a consequence of the notified transaction is likely to induce incentives to raise the price.

Measurement of margins and volumes

The Commission requested from the Parties detailed quarterly information on revenues relating to Orange's and Jazztel's main product types (in particular, 2p and 3p/4p). The Commission used the information collected on revenues to construct a measure of the ARPU for the 2p and for the 3p/4p product types. That was obtained by dividing the total revenues allocated to the product type by the number of active fixed subscribers to the product type.

The Commission also requested from the Parties detailed quarterly information on costs relating to Orange's and Jazztel's main product types (analogously to what was done for revenues).

For products including further components along with fixed broadband (such as fixed voice or mobile services), the Commission included also the revenues and costs relative to those components, and the analysis has been done on the basis of the entire bundle, in line with the analysis performed by the Notifying Party. Indeed, pricing decisions of the market participants are made on the basis of the price and the margin of the bundle, not on hypothetical prices of the included components.

The incremental costs for each product type have been defined as the merging parties' estimates on costs that the operator could avoid (that is to say no longer incur) in the hypothetical case of a substantial reduction of the subscriber base. Those include all direct costs (that is to say those costs that can be directly attributed to a subscriber and that therefore vary in direct proportion with the number of subscribers) and a portion of costs and overheads which could be reduced in case of a substantial reduction in the subscriber base (for example costs relating to outsourced call centre services).

Based on the computed ARPU and incremental cost per user, the Commission has calculated the incremental margins per user (a conservative measure of margins, compared to other definitions which would uniquely account for the direct costs).

Incremental margins are considered an appropriate measure of profitability for the purposes of the calibrated merger simulation in that context. Incremental margins reflect an operator's profitability from the new subscribers acquired and this is the profitability measure that governs the decisions of a profit maximising operator when setting the price of its products.

Finally, the Commission also based its analysis on estimates of the number of new subscribers to the relevant fixed Internet products (that is to say the gross adds). The Commission considers that the share of the operators's new fixed subscribers (the so called gross add share) is a good indicator of the share of contestable subscribers attracted by each operator and therefore a good indicator of the current competitive strength of each market participant .

Section 3.2 of Annex A contains further details on the construction of margins and section 3.3 of Annex A contains further details on the use of gross add shares as a measure of demand.

The results of the calibrated merger simulation presented in the Statement of Objections

In support of the Statement of Objections, the Commission has calculated the post-merger expected price increase for each of the main operators, namely the Parties, Telefónica and Vodafone.

In the first baseline scenario the market wide expected price increase, that is, the weighted average price increase across all operators and across all product types including a fixed broadband component was 10% for 2013 and 7% for 2014 .

In the second baseline scenario, the market wide expected price increase was 7% for 2013 and 5% for 2014.

(531) The Statement of Objections contained a variety of further sensitivity scenarios. For instance, the Commission considered a scenario based on benchmark diversion ratios calculated from gross add shares and a scenario based on diversion ratios calculated from the FNP data from the AOP. The results of all those further scenarios also

313See also section 3.3 of Annex A.

314Where the Commission used incremental margins computed as discussed in section 7.2.7.2 and diversion ratios based on purely horizontal number ports as presented in Table 24.

The average price increase for each product type has been calculated as the weighted average of the price increase expected from each operator for that product type (where the weights used are the gross add shares of each operator for that product type and no price increase is assumed for operators other than the main four). The market wide price increase has been calculated as the weighted average of the (average) price increases expected for each product type (where the weights used are the gross add share of each product type and no price increase has been assumed on average for the fixed broadband only product type and for the fixed broadband plus mobile product type).

316Where the Commission used incremental margins computed as discussed in section 7.2.7.2 and diversion ratios based on the product type-level portability data collected from the operators for residential customers, aggregated across all product types of origin and disaggregated by product type of destination as reported in Table 25.

(100)

pointed to significant price increases as a consequence of the notified transaction (the details are presented in section 4.3 and 4.4 of Annex A).

7.2.7.4. The Notifying Party's main arguments in the response to the Statement of Objections

(532) In its reply to the Statement of Objections, the Notifying Party strongly criticises the calibrated merger simulation carried out by the Commission, claiming that the switching data used to calculate price effects is tainted by the rapid shift to multiple play products. The Notifying Party argues that the switching data is confounded by the rapid and significant shift of consumers' choice from dual-play ("2p") to triple-play ("3p") and quadruple-play ("4p") products (in particular internal switching from 2p to 3p).

(533) According to the Notifying Party, using the available data on internal switching invalidates the merger simulation model because the level of internal diversions by far exceeds the levels compatible with profit maximisation (that is, only a positive own price elasticity could reconcile the two). The Notifying Party considers that the issue cannot be solved by using the available data and that the solution proposed by the Commission in the Statement of Objections, by ignoring switching across bundles and estimating the expected price effects separately for 2p and 3p products, not only fails to address the issue but also results in an upward bias in the diversion ratios.

(534) In addition to the overarching critiques to the use of a merger simulation in this case, the Notifying Party submitted a number of more technical comments pointing to problems of data inconsistency in the Commission's quantitative analysis.

(535) The Notifying Party also noted that the Commission did not allow for any diversion of subscribers to an outside good, that is, it did not allow for the possibility that some subscribers could choose to no longer subscribe to any fixed broadband product following to a price increase.

(536) The main comments by the Notifying Party are summarised and addressed in section 7.2.7.5 and the remainder is addressed section 5 of Annex A.

7.2.7.5. The Commission's assessment of the main arguments by the Notifying Party

(537) This section presents the Commission's assessment of the main arguments raised by the Notifying Party in response to the Statement of Objections.

(538) As a general principle, the Commission notes that the precision of a quantitative analysis where the effects are estimated based on observed parameters in a constantly evolving environment is inevitably lower than in stable environments.

(539) The Commission agrees that observed switching between operators from the portability data is unlikely to be exclusively based on consumers' reactions to small price changes. Therefore, caution has to be applied when calibrating demand based on observed switching patterns. In particular, in light of the observed on-going shift of the product mix in the retail market for fixed Internet access services, it cannot be simply assumed that as a consequence of a small price change consumers would switch from 2p to 3p (or 4p) products in the same proportion as observed in the past.

(540) However, in line with its practice in recent past cases, the Commission considers that diversion ratios based on observed switching in the present case are a reasonably good indicator to calibrate horizontal switching between competitors in case of a unilateral price increase of one of the merging parties. In particular, as stressed in recital (507), the Commission considers that the portability data is informative on the extent of substitution between different operators.

(541) The Commission considers that focusing its analysis on the choice between competing operators in the present case can be expected to produce results which provide relevant insights to the assessment of the case for several reasons.

(542) Firstly, in the applied differentiated goods framework, the magnitude of predicted anti-competitive effects from a horizontal merger is mainly driven by diversions ratios between the merging parties and not by diversion ratios between product types within each merging party (that is to say the internal diversions). Indeed, the Notifying Party simulated how sensitive the estimated results are to modifications of internal diversion ratios. As long as the internal diversions between 2p and 3p products do not exceed 20% the estimated price effects change mostly by less than one percentage point (for internal switching beyond 20% the profit maximisation assumptions underlying the merger simulation no longer hold in this case). Moreover, the Notifying Party's simulations indicate that the predicted price effects are lowest when assuming zero internal diversions. That suggests that assuming no competitive interaction across product types is a conservative assumption in this context.

(543) Secondly, the Commission disagrees with the Notifying Party's claim that focusing on the competition between operators is not a reasonable approximation because the calculations only lead to consistent results for internal diversion rates that are much smaller than those actually observed. As regards observed internal switching from 2p to 3p products, it has to be taken into account that during the past years the prices of 3p products have significantly decreased relative to 2p products. Many subscribers who switched from 2p to 3p/4p products would have presumably switched to 3p/4p products even if the prices of 3p/4p products had been marginally higher (that is, for those switchers the original 2p products are unlikely to represent their second best choice). That implies that the internal diversion ratios between 2p and 3p products in case of small price changes are likely to be much lower than the Notifying Party's estimation of the past switching that occurred between those product types. The past internal diversion ratios do therefore not point to inconsistencies that would in themselves invalidate the results.

(544) Thirdly, the arguments of the Notifying Party as to why the Commission's estimates of the diversion ratios between the Parties would be biased upwards, as opposed to being less precise, are not convincing. The Notifying Party suggests that high switching between the merging Parties might be due to the fact that in 2013 and 2014 Orange mobile customers who did not have a preference for a particular mobile offer, but had an affinity to Jazztel as a fixed provider would have switched to the Jazztel 3p bundle. Likewise, customers who were not particularly committed to the Jazztel fixed offer, but had a strong preference for Orange’s mobile service would appear as switchers from Jazztel 2p to Orange 3p.

(545) However, Jazztel has been an MVNO on Orange's mobile network, so that there is presumably no significant quality difference in the mobile services offered at the retail level by both companies. Thus, the argument as put forward by the Notifying Party is more convincing when applied to the remaining nationwide competitors, that is to say, Jazztel customers who were particularly committed to the Jazztel fixed offer could have switched to Telefónica or Vodafone because they had a strong preference for the latter's mobile services. Similarly, previous Orange mobile customers who did not have strong preferences on the mobile provider may have switched to a 3p product of Telefónica or Vodafone. Therefore, to the extent the introduction of 3p products has triggered additional switching, the argument of the Notifying Party does not imply that diversion ratios between the Parties would be overestimated.

(546) Moreover the observed diversion ratios from Jazztel to Orange do not support that argument of the Notifying Party. When focusing on purely horizontal switches between 3p/4p products, the diversion ratios from Jazztel to Orange tend to be significantly higher than the diversion ratios from Jazztel to Orange based on ports from 2p to 3p/4p products.

(547) Fourthly, the Commission notes that the high diversions between the Parties are unlikely to be due to temporary effects which are unrelated to the strength of the Parties. Indeed, the high observed gross add shares, especially for 3p products, and

325The observed internal diversions do not necessarily reflect preference shifts as claimed by the Notifying Party but are likely caused also by changes in the relative prices of the relevant product types.

(548) The Commission also notes that when deriving a full diversion matrix that contains both flows within operators and across operators it is important to distinguish flows of number porters from flows of switching customers, as not all switching customers request that their number be ported. Moreover, depending on the products between which subscribers switch, the telephone number may be automatically preserved so that no number request is necessary.

327Notifying Party's response to the Statement of Objections, p. 31.

328See also section 5.3. of Annex A.

the observation that especially Orange grew significantly in 2014 are in line with the high observed diversion ratios for the Parties. Moreover, Jazztel would have likely remained an MVNO on the Orange mobile network in the absence of the merger, which would have implied a persistent similarity between the mobile services of both Parties.

(548) Therefore, whereas it is true that the introduction of 3p offers could have resulted in switching that is not exclusively driven by price, the Notifying Party has failed to provide arguments as to why the observed past diversions between the merging parties would overestimate their competitive constraint on each other. Moreover, 3p products were mainly introduced in the third quarter of 2012 or in the beginning of 2013, so that the observed diversion ratios are presumably less affected by that product introduction in 2014. That is, for example, reflected in more stable diversion ratios over the quarters of 2014 as compared to 2013.

(549) As regards the elimination of data that the Notifying Party criticises, the Commission notes that the elimination of switching across product types is done in order to focus the analysis on the horizontal switching (2p to 2p and 3p/4p to 3p/4p), which are the instances in which observed switching is most likely to be driven by consumers' reactions to small price changes.

(550) The observed vertical switching across product types in the portability data (for example switches from a double play product to a convergent product) is less likely to reflect consumers' reactions to marginal price changes compared to observed switching across operators within the same product type, for the reasons already discussed. Since the observed vertical diversion ratios do not necessarily reflect consumer reactions to marginal price changes, in its calibrated merger simulation the Commission assumes that no price based switching occurs from one product type to another. The analysis is thus focused on the horizontal choice between operators and does not explicitly model the possibility of switching to other product types as a consequence of small price increases.

(551) That assumption is implemented in several ways. In one scenario, the assumption is implemented by eliminating the cross bundle diversions across operators and retaining only the diversions relating to switching across operators but within the same product type. In another scenario, the data is aggregated by product type of origin, which implies assuming that all consumers acquiring a 2p product had only 2p products in their choice set and all consumers who acquitted a 3p/4p product only had a 3p/4p product in their choice set. In a third scenario, diversions between competitors are derived based on the gross add shares of the operators.

(552) The Notifying Party claims that the first methodology, by focussing only on the purely horizontal switching instances and discarding the diagonal instances produces biased diversion ratios due to a number of data issues. The Commission carefully considered those issues (see sections 5.2 and 5.3 of Annex A) and decided to attribute less weight to the scenario based on a matching of the product type-level FNP data and no longer retain it amongst the baseline scenarios of its updated analysis.

(553) The Notifying Party further claims that the all scenarios result in biased IPRs, due to implicit assumption of no competitive interaction between product types. To analyse

(554) that claim, the Commission compared the expected price increases in case no cross-bundle switching is assumed and the expected price increases in case a certain degree of cross-bundle switching is assumed. The results of that sensitivity (see section 5.3.2 of Annex A) suggest that including a certain degree of cross bundle switching has three implications: (i) it slightly increases the IPRs for 2p products, (ii) it slightly decreases the IPRs for 3p/4p products and (iii) it does not significantly affect the market wide IPRs.

(555) Overall, the Commission considers that in the present case a calibrated merger simulation based on observed portability patterns between competitors is a useful indicator to assess the likely anti-competitive effects of the loss of competition between the merging parties. Although vertical switching as a consequence of price increases is difficult to predict within this framework, the estimated magnitude of likely price increases for the relevant product types primarily depends on switching between competitors. Therefore the quantitative assessment performed by the Commission can still provide additional insights to assess the present case.

(556) In any event, the Commission does not exclusively rely on predicted effects based on observed portability data. Alongside the scenarios based on portability data, it has carried out a scenario where price effects are estimated based on benchmark diversion ratios derived from gross-adds. That approach assumes that diversion between competitors is proportional to their share of gross adds, which effectively implies that all operators are equally close . In its updated analysis, the Commission has decided to give more prominence to the results obtained from that scenario. As set out in section 7.2.7.6, the results point to significant price increases.

(557) The Commission has also revised the range of its estimated price effects, specifically by considering the impact of introducing a strong assumption that a significant proportion of subscribers would choose to no longer subscribe to any fixed broadband product after a price increase.

(558) Following to the Statement of Objections and the Notifying Party's reply, the Commission has refined its predictions of the post-merger price increases.

(559) While a number of sensitivity scenarios are contained in Annex A, the Commission presents, in recitals (563) to (575), the results of its baseline scenarios. These scenarios differ in the diversion ratios that are used to calibrate the customers' switching behaviour.

(560) For each of the baseline scenarios considered, the upper bound of the estimated price effects corresponds to the figures presented in the Statement of Objections, which are based on an assumption of no diversion to the outside good. An extreme lower bound to the price effects has been computed instead by reducing the diversions between the operators so as to attain a market elasticity of minus 1 (as suggested by the Notifying Party). That should be seen as an extremely conservative sensitivity assumption (see section 5.7 of Annex A).

(561) Adding that extreme diversion to an outside good reduces the estimated price effects but not to an extent that allows dispelling any concern of potential price increases post-merger. In addition, for both scenarios, the Commission expects the diversions to outside goods to be significantly lower than those assumed for that lower bound. Therefore, the reported lower bound to the expected price increases should be interpreted as an illustrative and extremely conservative reference. The price increases from the analysed transaction are more likely to be closer to the upper bound of the Commission's estimates than to the purely illustrative lower bound.

(562) The Commission has not explicitly modelled the changes in the incentives of the remaining market participants (in particular the three regional cable operators active in Spain). Those changes are present in the Commission's calculations in terms of market shares (see Table 16) but their pricing behaviour has been assumed to be unchanged post-merger compared to the situation prior to the transaction. That assumption is expected to be conservative, because under the assumed horizontally differentiated demand model with Bertrand competition modelling the pricing behaviour of those additional players would likely lead to further increases in price for all market participants .

(563) In one baseline scenario, the Commission used the incremental margins computed as discussed in section 7.2.7.2 and diversion ratios based on the product type-level portability data, aggregated across all product types of origin and disaggregated by product type of destination as reported in Table 25.

(564) For the 2p product type, the resulting average price increase across all operators is approximately 2.2% to 4.5% in 2013 and 1.0% to 1.9% in 2014. The corresponding price increases for Orange and Jazztel are in the range of 4% to 9% in 2013 and 2% to 6% to […]* in 2014.

(565) For the 3p/4p product type, the resulting average price increase across all operators is approximately 5% to 11% in 2013 and 4% to 7% in 2014. The corresponding price increases for Orange and Jazztel are in the range of 7% to 17% in 2013 and 5% to 10% in 2014.

(566) The market wide expected price increase, that is, the weighted average price increase across all operators and across all product types including a fixed broadband component is 4% to 7% for 2013 and 3% to 5% for 2014.

(567) Another baseline scenario considered by the Commission assumes the incremental margins computed as discussed in section 7.2.7.2 and benchmark diversion ratios based on gross add shares as reported in Table 26.

(568) For the 2p product type, the resulting average price increase across all operators is approximately 1.2% to 2.4% in 2013 and 0.6% to 1.1% in 2014. The corresponding price increases for Orange and Jazztel are in the range of 2% to 5% in 2013 and 1% to 3% in 2014.

(569) For the 3p/4p product type, the resulting average price increase across all operators is approximately 6% to 11% in 2013 and 5% to 10% in 2014. The corresponding price increases for Orange and Jazztel are in the range of 6% to 18% in 2013 and 6% to 14% in 2014.

(570) The market wide expected price increase, that is, the weighted average price increase across all operators and across all product types including a fixed broadband component is 3% to 6% for 2013 and 2014.

(571) Across the two scenarios, the Commission notes that for both 2p and for 3p/4p product types the magnitude of the predicted price increases is lower for 2014 data compared to 2013 data. However, the Commission also notes that the relative share of 3p/4p subscribers increased significantly from 2013 to 2014 as reported in Table 1. In its assessment of results of the quantitative analysis the Commission has taken into consideration both the fact that overall price effects are less pronounced based on more recent data but also the fact that a forward looking analysis which takes into consideration the direction in which the market is evolving must inevitably place significant weight on the higher price increases expected for the 3p/4p product types.

(572) As explained in section 7.2.2.4, an internal document of Orangesuggests that the ARPUs by Orange in a […]* scenario for 2018 would be […]* for the mobile segment and […]* for the fixed broadband segment than the respective ARPUs absent […]*. Another internal document of Orangesuggests that the ARPU […]* for the mobile segment and […]*for the fixed broadband segment (compared to the respective ARPUs absent […]*.

(573) Overall, the results of the Commission's calculations are aligned with Orange's expectations, as set out in Orange's internal documents. For the 2p product type, the price increases estimated by the Commission for each of Orange and Jazztel are in the range of 2% to 9% in 2013 and 1% to 6% in 2014. For the 3p/4p product type, the expected price increases for each of Orange and Jazztel are in the range of 6% to 18% in 2013 and 5% to 14% in 2014.

(574) Since in some of the relevant internal documents also other synergies are analysed, the expected price increases indicated in these internal documents presumably already account for the countervailing effect of efficiencies. As explained in section 7.2.10.4, the elimination of double marginalisation for mobile services included in Jazztel's 3p products can be expected to reduce the estimated average price increase for 3p/4p products (across all operators) by approximately 0.8 percentage points. Therefore, once the elimination of double marginalisation is taken into account, the predicted price increases for 3p/4p products are more in line with the estimates of internal documents of Orange.

(575) The Commission has estimated the price effects for a variety of further sensitivity scenarios. For instance, the Commission has conducted sensitivities on the margins assumed for the non-merging parties (see section 4.6 of Annex A). Another sensitivity scenario included diversion ratios based on the operator level FNP data from the AOP (see section 4.5 of Annex A). The results of all sensitivity scenarios also pointed to significant price increases as a consequence of the notified transaction.

337 [Reference to the Notifying Party's internal business documents]*.

338 [Reference to the Notifying Party's internal business documents]*.

7.2.7.7. Conclusions from the quantitative analysis

(576) The quantitative analysis conducted indicates that the merging parties impose a significant competitive constraint on each other, in particular as regards 3p/4p products. This follows from the significant size of the Parties' gross add shares and from the fact that the diversion ratios between the merging parties are even higher than suggested by their gross add shares. In deriving its price predictions, the Commission took into account a number of arguments raised by the Notifying Party in its Reply to the Article 6(1)(c) Decision and throughout Phase II. These arguments (which are summarised and addressed in section 7.2.7.5 and in more detail in Annex A) have been given thorough consideration.

(577) As explained in the detailed assessment of the Parties' arguments, as a general principle the Commission notes that the precision of a quantitative analysis with regard to a constantly evolving market (as is it occurs in this case) is inevitably lower than with regard to stable markets. Nevertheless, the Commission considers that in this case the quantitative analyses performed provide a valuable additional insight in addition to the other elements presented.

(578) The Commission has revised downwards the range of its predicted price effects compared to the Statement of Objections, also to reflect a degree of precision loss in this case. Overall, the final estimated market-wide price increases in the two baseline scenarios likely are in the range of 3% to 7% based on 2013 data and 3% to 6% for 2014 data. The Commission considers that the estimates based on 2014 data are more relevant than the 2013 ones for the likely price effects of the merger.

(579) The Commission expects the diversions to outside goods to be significantly lower than those assumed for this lower bound. Therefore, the reported lower bound to the expected price increases should be interpreted as an illustrative and extremely conservative reference.

(580) The revised predicted price increases do not materially alter the concerns presented in the Statement of Objections, that is, the elimination of competition for customers between the merging parties is likely to lead to significant price increases .

7.2.8. Limited likelihood of sufficient entry in the market

(581) A merger is unlikely to pose any significant anti-competitive risk if entering a market is sufficiently easy. For entry to be considered a sufficient competitive constraint on the merging parties, it must be likely, timely and sufficient to deter or defeat any potential anti-competitive effects of the merger .

(582) In the present case, entry of a viable operator able to compete at national level could dispel the concerns in relation to the affected retail market for the provision of fixed Internet access services as well as in relation to the possible retail markets for the provision of multiple play services. The Commission has assessed the likelihood of entry by a new operator and whether such entry would be sufficient to deter or defeat the potential anti-competitive effects on the merger.

(583) The following assessment discusses, first, the likelihood of a sufficient entry into markets based on products provided only over fixed infrastructure (e.g., fixed Internet access services and a possible market for dual-play products). Second, the additional barriers to entry are assessed when looking at convergent markets that combine fixed and mobile services (e.g., triple- and quadruple-play services), including regarding potential entry of fixed infrastructure-only operators into such convergent markets.

7.2.8.1. Likelihood of sufficient entry into markets involving a fixed Internet access services component

(584) In the Form CO, the Notifying Party submits that it is relatively easy to enter the retail fixed telecommunication markets. While barriers to entry to most telecommunications markets are very relevant, as in all network industries, the EU telecommunications regulatory regime implemented by Spain imposes on operators with significant market power the obligation to provide access to the fixed network. Therefore, the most relevant barrier to entry, the deployment of infrastructure, is significantly reduced. Moreover, wholesale access services provided by Telefónica have improved over the last years.

(585) In its replies to the Article 6(1)(c) Decision and to the Statement of Objection, the Notifying Party points to recital 183 of the Commission's decision in the case Vodafone/ ONO where it stated that "the merged entity will not have the ability to foreclose its competitors in multiple play offers by denying access to fixed wholesale services, as competitors have a regulated alternative option." In Orange's view, this statement would be proof that the Commission had, in that decision, concluded that the regulated access to Telefónica's network eliminates the barriers to entry.

(586) On the other hand, the Notifying Party acknowledges that "the existing barriers to entry are not new and definitely have not been exacerbated over the last years", but are the same faced both by Orange and Jazztel to develop and sustain their current position . The Notifying Party states: "The deployment of a full parallel network, either a copper network or a fibre network, requires a significant investment. Orange is fully aware as it is actually undertaking such a deployment. However, such investment is not lower for Orange than for any other possible market entrant." .

(587) Finally, the Notifying Party submits in its reply to the Statement of Objection, that at this point in time, there is "no doubt" that alternative carriers will have at least two wholesale access products to Telefónica's FTTH network: (i) access to FTTH at the local level (through Virtual Unbundled Local Access ("VULA")-type services) in a significant part of Spain; and (ii) access at a regional level through bitstream services (including for speeds above 30 Mb/s) in a significant part of the territory. However, the Notifying Party acknowledges that "details" of access services such as the access conditions and the geographic segmentations still have to be determined.

Results of the Market Investigation

(588) The vast majority of respondents to the Phase I Market Investigation questionnaire to competitors consider that barriers to entry are high for fixed Internet access services . First, significant investments are required to develop own fixed infrastructure connecting to end-customers (such as a FTTH network or a HFC cable network). Vodafone points out that there has not been any significant entry in this market in the last 5 years.

(589) Second, investments into a backbone network connecting to local exchanges to enable physical access to Telefónica’s regulated unbundled copper network are significant, considering in particular the long return cycle for such investment. Several respondents point out the market is mature, requiring also significant resources to build up a sufficient customer base to amortize the significant investment .

(590) Third, several respondents to the Phase 1 and the Phase II Market Investigations pointed out that they had considered to enter the market but they would have needed access to fixed broadband at reasonable price to compete on the market. According to them, prices set by the existing wholesale regulation for indirect access to Telefónica's network through bitstream do not enable a new entrant to compete effectively in the market. While Telefónica has an obligation to provide indirect access to its network for speeds below 30Mb/s (regardless of the underlying broadband technology), the regulated terms and conditions are currently such that new entrants cannot compete effectively on this basis at least in the residential segment of the market (where customers do not demand value-added products such as security or data centre services) and in particular with respect to convergent products (whose retail prices tend to be significantly discounted as opposed to the standalone fixed and mobile components). Moreover, some respondents submitted that the nationwide fixed Internet service providers other than Telefónica do not have the incentive to offer wholesale access at reasonable cost. One respondent states that it is "extremely difficult to provide fixed Internet access services because […] there are a limited number of players, not willing to provide a competitive wholesale offer and current regulated offers are not feasible." .

Commission's assessment

(591) Before possibly deploying its own network, a new entrant is most likely to enter markets involving fixed Internet access services by climbing up the ladder of investment including by benefitting from regulated access. Such regulated access in Spain is currently limited to the segment with speeds below 30 Mb/s. In the following, the Commission considers in more detail three ways for a new entrant to enter the retail markets involving fixed Internet access services without rolling out its own last mile access network. These ways are currently limited to the segment with speeds below 30 Mb/s and concern, notably, (1) the resale of another operator's product, (2) regulated indirect (or virtual) access to the incumbent's last mile network, and (3) regulated direct (or physical) access to the incumbent's last mile copper network. Yoigo is an example for reselling a fixed Internet access product of another operator (Telefónica). Orange and Jazztel rely on regulated direct access for the vast majority of their xDSL customer base (approximately […]*% and […]*% respectively). Their other xDSL customers (located in less densely populated areas) are served through regulated indirect access. (4) The remainder of this section discusses potential entry into the VHBB segment by deploying its own NGA network.

Resale

(592) To the extent that non-fixed operators currently succeed in agreeing commercial terms with a fixed Internet access provider to resell the latter’s services – for instance, so that mobile-only players can offer a convergent product – the Commission observes that these products do not seem to be successful in the market. Internal documents of Orange show that […]*[…]* .

(593) In any case, the Commission considers that agency contracts to resell a product do not allow a reseller to act as a competitive force in the market as the reseller generally does not set the price. Moreover, customer invoicing for the fixed component may be handled by the fixed operator, which means that the fixed operator has direct access to the customer and can therefore more easily entice the customer to switch from the reseller to the fixed operator for the services it does not yet supply to the customer. This would help explain why a reseller would hesitate to compete offensively in the market with a resale product for fixed Internet access.

(594) These findings are confirmed by Orange's internal documents where, assessing the ability to compete of low-cost mobile only players in the convergent segment, Orange states that it is […]* .

Regulated indirect wholesale access to Telefónica's network

(595) Regulated indirect wholesale access to Telefónica's network (so-called "bitstream access") is considered to be the second step on the ladder of investment for an alternative operator. However, as shown below, the Commission does not consider that a new entrant can, currently, compete effectively in the markets including fixed Internet access services on the basis of bitstream access.

(596) Telefónica offers two legacy regulated products for indirect access to its network. First, GigADSL is a service for the indirect access to the local loop with an ATM hand-over and 109 points of access to reach national coverage. Second, ADSL-IP is a service for the indirect access to the local loop with an IP hand-over and either 50 points of access (regional service) or one point of access (national) to reach national coverage.

(597) The Commission had expressed serious doubts based on a draft of the current regulation that "by setting price levels […] for bitstream legacy products (GigADSL and ADSL-IP) up to 50% above cost-efficient levels, CMT is safeguarding and promoting competition on the Spanish broadband markets." . For the final regulation, the CNMC revised the relevant access prices criticised by the Commission.

(598) In any event, a new Telefónica product called NEBA was formally approved by the Spanish NRA on 10 November 2011 to replace the bitstream legacy products. On 1 April 2014, the CNMC confirmed the effective availability of NEBA services .

(599) The Commission considers that the NEBA product as currently regulated, and in particular NEBA delivered over Telefónica's FTTH network, is unlikely to allow a new entrant to replicate current retail prices, let alone undercut those prices in order to induce customers to switch provider and to replicate the competitive pressure on prices that is lost with the merger. First, according to the Notifying Party's own submission in the Form CO, NEBA contains several important weak points that have limited the appeal of the offer to alternative operators. For one, the service is limited to a maximum of 30Mb/s bandwidth. While in theory, NEBA is technology neutral and includes different access technologies such as ADSL2+, VDSL2 and FTTH, the bandwidth restriction excludes the effective provision of indirect access to fibre. Moreover, the geographic coverage of NEBA is more limited than the coverage of the bitstream legacy products such as GigADSL as not all DSL access multiplexers support NEBA. This is particularly the case in low-density areas, which are the areas where NEBA is of more relevance for operators that already have direct access to the local loop or fibre deployment in high density areas. Furthermore, the bandwidth limitations and quality of service parameters in NEBA create problems for the provision of IPTV services. The Notifying Party notes that NEBA was not specifically developed to replicate the TV services provided by Telefónica. Thus, Orange submitted to the CNMC that it is impossible to replicate Telefónica's convergent FTTH products (Fusión) that all incorporate, as from April 2014, TV services. Finally, Orange submits that […]*. The current traffic charges are as follows:

(600) Second, almost after one year after being operational, actual usage of the regulated NEBA product has been minimal. On 31 January 2015, there were only 1801 NEBA xDSL lines and 29 248 NEBA FTTH lines. In its public consultation published on 30 March 2015 concerning a potential reduction of the current NEBA capacity prices, the CNMC therefore states that the current regulated prices (approved more than a year ago based on the model results of 2013) should be reviewed without awaiting the outcome of the review of markets 3a, 3b, and 4 as otherwise they would lead, given the evident non-use of NEBA, unjustified additional cost for operators that use NEBA .

(601) Third, Orange's internal documents further corroborate that NEBA does not allow a new entrant to compete effectively in the retail markets involving fixed Internet access services. The current regulated NEBA FTTH offer has, in Orange's own words, […]* . The price payable to Telefónica for fixed voice and broadband (double play) would be EUR […]* (excluding VAT of 21%, assuming the projected 2015 average traffic for double play of […]* kbps). According to Orange, NEBA fees payable to Telefónica for a convergent offer including TV amount to EUR […]* (excluding VAT of 21%, assuming the projected 2015 average traffic for fixed-TV triple-play of […]* kbps). Another Orange document describes NEBA as […]* .

(602) Fourth, based on the figures in recital (601) above, a new entrant cannot compete effectively in the markets involving fixed Internet access services. A dual-play fixed package would cost a new entrant a monthly fee of EUR […]* per customer (including VAT of 21%, based on Orange's capacity assumption). On top of this monthly cost calculated by Orange, one-off costs for setting up a line and certain other charges may need to be added. Furthermore, this cost excludes customer acquisition cost (sales commission, logistics or promotional costs) and any additional monthly recurrent costs of billing, collection, and customer care etc .

(603) For the above reasons, it seems unlikely that regulated indirect wholesale access to Telefónica's network allows operators to enter or geographically expand in the market. The regulated wholesale prices and conditions do not allow replication of current retail prices in the market, let alone undercut those prices to induce customers to switch provider and to replicate the competitive pressure on prices that would be lost with the Proposed Transaction. With respect to the consultation published on 30 March 2015 in which the CNMC proposes certain changes to the regulation the Commission considers that the outcome, as well as the timing of the possible entry into force of a potentially revised NEBA regulation are not sufficiently certain to be taken into account in the present assessment. While the consultation proposes to reduce by 45% the NEBA capacity prices shown in Table 27, other elements of the regulated NEBA offer such as the monthly recurrent prices of the service and prices related to the Points of Interconnection are not considered as part of the proposed draft measure. Therefore, the effect of a potential overhaul of the NEBA regulation on the market conditions and the barriers to entry cannot be determined.

(604) As to the Notifying Party's argument that the Vodafone/ ONO decision is proof that, in that decision, the Commission had concluded that the regulated access to Telefónica's network eliminates barriers to entry, the Commission notes that this reading misinterprets the said decision. The statement invoked by the Notifying Party refers to a foreclosure scenario which is subject to the legal test of whether the merged entity has the ability to foreclose a competitor from the market by not offering wholesale services for fixed Internet access services. The Commission answered this question by stating that a competitor seeking access to fixed Internet access services has the option to rely on Telefónica's wholesale access products, such as bitstream products. The question discussed in the context of the present case is different as it concerns the question as to whether a new entrant would be able to enter the market and replicate the competitive role Jazztel is currently exercising in the market. The conclusion reached in this sub-section is that this is not the case due to the specific economics of the regulated bitstream products (as compared to the more favourable economics of services based on LLU access, which Jazztel is mainly relying on). As shown in the next sub-section, a new entrant is, moreover, unlikely to roll-out its own infrastructure in order to be able to rely on LLU access to Telefónica's copper network.

Local loop unbundling

(605) Orange and Jazztel have undertaken significant investments in order to rely on Telefónica's regulated LLU offer delivered to end-customers over Telefónica's copper network. They have rolled-out their own aggregation and backbone networks in Spain. For a certain percentage of local exchanges that each covers, they have deployed their own optical fibre backhauling networks. They have installed xDSL equipment in, respectively, […]* and […]* "local exchanges" (Telefónica's central offices containing the so-called Main Distribution Frames ("MDFs").

These prices are EUR 15,02 for NEBA copper, and EUR 19,93 for NEBA FTTH per month and per line (excl. VAT).

Jazztel has connected […]* of the local exchanges it uses with owned optical fibre backhaul networks. For Orange, the percentage is approximately […]*%.

According to the Notifying Party, Jazztel has deployed, or will deploy shortly, LLU equipment in […]* local exchanges. Orange is present (or will be shortly) in […]* local exchanges. According to the Notifying Party, the geographic overlap concerns at least […]* local exchanges, corresponding to […]* million lines or at least […]*% of all Spanish copper lines. Orange's network overlaps with Jazztel's xDSL network for […]*%. Jazztel overlaps with Orange's xDSL network at […]*%.

the former Spanish telecommunication regulator CMT had imposed conditions on the dismantling of Telefónica's copper network. Most importantly, it imposed that if any alternative operator is co-located at a local exchange, Telefónica needs to provide wholesale services for 5 years to protect those operators' investments. Moreover, Telefónica can only close a Local Exchange if more than 25% of customers covered by it are connected by alternative means (such as FTTH).

(610) This limited time horizon of about five years after which large parts of Telefónica's copper network are likely to be dismantled (and the xDSL network becomes obsolete), is unlikely to allow a new entrant to amortize the significant cost for investing into the roll-out of a xDSL network given significant upfront infrastructure cost (as established in recital (608) above) and high customer acquisition cost to induce subscribers to switch provider and to gain a sufficient customer base in a market with relatively high broadband penetration.

(611) This is also the main reason why the Notifying Party is incorrect in stating that the barriers to entry are currently the same as they used to be when Orange and Jazztel entered the market. Jazztel, for instance, started to operate in 1997 and therefore has had a long period to recoup its investments into direct access via LLU. Today, the recoupment period for a new entrant would be significantly shorter, limited by the regulated period of 5 years that Telefonica is obliged to operate a local exchange that meets the regulated criteria for closure.

(612) As regard the Notifying Party's argument that there is "no doubt" that alternative carriers will have at least two wholesale access products to Telefónica's FTTH network: (i) access to FTTH at the local level (VULA type service) in a significant part of the territory; and (ii) access at a regional level through bitstream services (including for speeds above 30 Mb/s) in a significant part of the territory, the Commission considers, for the reasons set out in section 7.1.4 above, that the outcome, terms and conditions including wholesale prices and geographic coverage, as well as the timing for such potential regulated access options are not sufficiently certain and cannot, therefore, be relied upon for the purposes of this Decision.

Deployment of a NGA network

(613) Access to VHBB Internet services is currently not regulated in Spain. An operator who wishes to compete in the VHBB segments currently needs to deploy its own NGA network to the homes of end-customers. Operators deploying NGA networks in Spain may rely on the civil infrastructure sharing imposed by the CNMC in 2009 via the so-called MARCo reference offer. This sharing service features cost oriented prices and the availability of ducts in the outside plant up to the building entrance, avoiding for operators using it the need for trench digging. Moreover, there is an IT-based graphical information system for access to the complete duct database of Telefónica, made available to alternative operators, to enable them to construct business cases for fibre deployment. This helps explain that costs for civil

engineering appear to be lower in the fibre deployment in Spain as compared to other European countries and that all four nationwide operators – Telefonica, Vodafone, Jazztel and Orange – have been deploying in recent years FTTH networks.

(614) Despite the MARCo regulatory regime granting access to ducts, the Commission considers that the deployment of an own NGA network is a high barrier to enter the retail market involving Internet access services. This is because a new entrant would likely need to build up first its own customer base in fixed Internet access services to be able to profitably amortise the significant investments of deploying its own access network.

(615) First, as discussed in recital (588) above, respondents to the Market Investigation stated that significant investments are required to develop own fixed infrastructure connecting to end-customers (such as a FTTH network or a HFC cable network).

(616) Second, the Notifying Party submits itself (see recital (586) above) that rolling-out an NGA network such as an FTTH network or a HFC cable network to the building/homes of end-customers requires significant investments. Jazztel invested EUR […]* million to roll out a FTTH network to 3 million BUs. Orange publicly announced to invest EUR 300 million to roll out a FTTH network to almost 1.5 million BUs. Given that both companies have network sharing agreements for the mentioned footprint (Jazztel with Telefonica, and Orange with Vodafone), the total investments to reach this footprint likely are likely to be higher.

(617) Third, as discussed in more detail in recital (665) below, Orange’s internal documents as well as the AM report submitted by the Notifying Party state that one of the main criteria for making the roll-out profitable for a company is that it already has an existing customer base in fixed Internet access services in the roll-out areas that can be converted to the new NGA network. Therefore, deployment of an own (NGA) network would only make business sense if there is a minimum customer base in fixed Internet access services. By definition, a new entrant does not have such customer base and can therefore not profitably enter the market by deploying its own network.

Conclusion on the limited likelihood of sufficient entry in retail markets involving fixed Internet access services

(618) For the above reasons, the Commission considers that the barriers to enter the retail markets involving fixed Internet access services are high. This applies both to the VHBB segment for which the market is not regulated in Spain as well as to the segment for speeds below 30 Mb/s that is subject to regulation. The regulated wholesale prices and conditions for indirect access to Telefónica's copper network do not currently allow replication of the retail prices in the market, let alone undercut those prices, in particular with respect to convergent product offers. As regards direct access to Telefónica's copper network, a new entrant today is unlikely to amortize the significant cost for investing into xDSL equipment and backhauling, aggregation and backbone networks given the limited time horizon after which large parts of Telefónica's copper network are likely to be dismantled.

As discussed in recitals (700) to (701). Access to Telefónica’s civil infrastructure is complemented with a symmetrical obligation regarding in-house wiring also imposed by the CNMC in 2009 to all operators. The first operator that installs in-house fibre wiring in a building is obliged to grant access at reasonable prices to any requesting operator.

(619) As already explained in section 7.2.8.1, because of the structure of the market and because of the current regulatory framework, the Commission considers that access to the fixed network infrastructure of Telefónica will likely not allow a timely or sufficient entry of alternative providers of fixed Internet access services, in the event that the merged entity and its competitors decided to raise prices post-transaction.

(620) This finding is even further exacerbated if considering that providers of triple- and quadruple-play services need to have competitive access to mobile networks as well. In fact, providers of triple-play services often cross-subsidize either the fixed or the mobile component of the bundle by offering it at competitive commercial terms (sometimes, as it is the case of Jazztel, "for free", that is, at no additional charge to subscribers of their triple-play services).

(621) Whereas considerations about entry into retail markets involving fixed Internet access services have been explored in section 7.2.8 above, the current section will discuss the additional obstacles for successfully entering into multiple play markets involving a mobile component.

The Notifying Party's view

(622) The Notifying Party explains that effective regulation at wholesale level, as regards both the mobile and the fixed telecommunications markets, ensures that other competitors can easily enter the possible multiple play markets involving a mobile component.

(623) The Notifying Party submits that, as regards the mobile component of the possible market for multiple play services, entry is facilitated by the existing regulation of the Spanish national wholesale access and call origination on mobile networks market ("wholesale mobile regulation"). Such wholesale mobile regulation imposes on MNOs, including Orange, the obligation to provide such service to MVNOs.

(624) In its reply to the Commission Article 6(1)(c) Decision, the Notifying Party submits that regulation on access to the mobile network is scarce if compared with access on the fixed network of Telefónica. In fact, in the Notifying Party's view, there are no regulated reference offers or regulated prices, but a simple obligation to provide "reasonable prices". This would be explained by the fact that in mobile communications, there is competition in the provision of wholesale access services between MNOs.

(625) The Notifying Party submits that, even if the wholesale mobile regulatory regime is light compared to the obligations imposed in other telecommunication wholesale markets, the number of MVNOs is high, since the market works effectively, and MNOs are in competition against each other to host MVNOs in order to recoup network investments. This would be explained by the fact that MVNOs generate significant revenues for their hosting MNO.

(626) As regards access to 4G, the Notifying Party submits that access to this technology is being negotiated with all MVNOs. The Notifying Party further submits that an agreement was reached with Jazztel in July 2014 and with the MVNO HP in 2014.

(627) The CNMC issued a resolution on 30 October 2014 clarifying that access to 4G technology is part of the obligations contained in the wholesale mobile regulation. The Notifying Party indicates that the above resolution is not binding for any specific operator since it was only an answer to a consultation by an association of MVNOs.

(628) According to data of the CNMC, 4G coverage was around 60% of the population in Spain in 2013. The Notifying Party submits that, as of April 2014, Orange had more than 1 million customers for 4G services, out of a total of around 8.2 million mobile customers for the first half of 2014. It is expected that 4G penetration will progressively grow as customers change to 4G-enabled smartphones and coverage expands.

Commission's assessment

(629) In 2006, the Spanish NRA issued a decision on wholesale access and call origination on mobile networks (at that time, corresponding to market 15 of the Commission's recommendation on relevant products and service markets susceptible to ex-ante regulation). In this wholesale mobile regulation, the Spanish NRA demonstrated that the three MNOs (Telefónica, Vodafone and Orange) jointly held a position of collective dominance on the wholesale market for access and call origination on public mobile telephone networks in Spain and imposed on MNOs the following obligations:

1)1) meet all reasonable requests for access;

2)2) provide services at reasonable prices;

3)3) intervention of the Spanish regulator in case no interconnection agreement can be reached with another operator.

(630) Several MVNOs entered the retail market for mobile telecommunications services and gained sizeable market shares in the last years. As reported by the CNMC in its annual report for 2013, MVNOs have registered the largest increase in mobile customers in 2013 (+1.7 million) as compared to MNOs and have reached a combined market share of around 13% in terms of revenues. The Commission notes

The CNMC recently held that the three leading MNOs are obliged to give access to their 4G networks, Doc ID 1581 (see: http://www.cnmc.es/Portals/0/Ficheros/Telecomunicaciones/Resoluciones/2014/Noviembre/141030 Acuerdo CNS-DTSA-1730-14-Consulta%20AENOM%20Servicios%204G.pdf).

Commission Recommendation of 11 February 2003 on relevant product and service markets within the electronic communications sector susceptible to ex ante regulation in accordance with Directive 2002/21/EC of the European Parliament and of the Council on a common regulatory framework for electronic communication networks and services (OJ L114, 8.5.2003, p.45). That recommendation has now been replaced by Commission Recommendation of 9 October 2014 on relevant product and service markets within the electronic communications sector susceptible to ex ante regulation in accordance with Directive 2002/21/EC of the European Parliament and of the Council on a common regulatory framework for electronic communications networks and services (OJ L295, 11.10.2014, p.79).

Informe Económico de las Telecomunicaciones y del Sector Audiovisual 2014, available at http://informetelecom.cnmc.es/docs/Informe%20economico%20sectorial/Informe%20Telecomunicaciones%20CNMC%202014.pdf

(631) Respondents to the Market Investigation indicated that, as of today, the regulation does not oblige MNOs to make wholesale offers that would allow an MVNO to replicate retail prices for multiple play bundles. In the respondents' view, MNOs would not have the incentive to offer adequate conditions to MVNOs. In this respect, respondents to the Market Investigation indicate that the fact that MNOs have been, and still are, reluctant to offer wholesale 4G mobile services to MVNOs is telling.

(632) Moreover, a number of respondents to the Market Investigation have raised concerns in relation to access to 4G technology at national level in the framework of their agreements on wholesale access and call origination on mobile networks.

(633) For example, one respondent explains that "The mobile service is provided in precarious conditions, with contracts imposed by the provider of the network, with uncompetitive wholesale prices, short durations and the refusal to provide 4G services, which is of fundamental importance in order to be able to compete".

(634) Likewise, another respondent highlighted that to provide multiple play services including a mobile component at national level, access to 4G at reasonable price and conditions would also be needed. The same respondent argues that 4G is essential to compete on the market. In its view, the inability to offer such services would have a strong impact on reputation.

(635) The same respondent submits that the CNMC resolution of 30 October 2014, which was adopted following a complaint by an association of MVNOs and which was aimed at receiving clarification by the NRA as regards the obligation for MNOs to offer wholesale access to 4G as part of their regulated wholesale mobile offers, despite confirming that obligation, did not determine the prices at which access should be granted. The resolution simply confirmed that prices must be "reasonable" and allow potential entrants to compete in the market. According to the respondent, this resolution leaves a high degree of uncertainty in the market and MNOs can further delay granting any effective access to 4G to MVNOs.

(636) The Commission notes that, as regards the possible multiple play markets involving a mobile component in Spain, operators must be able to offer mobile and fixed telecommunications services in a bundle at prices that, while at least replicating the retail prices observed in the market, allow for a positive margin for the operator. Therefore, reasonable wholesale prices for mobile telecommunications services – including 4G technology – are key.

(637) The Commission considers that access to the possible multiple play markets involving a mobile component on the basis of the current wholesale mobile regulation is difficult for players that do not have their own fixed telecommunications infrastructure based on ULL or NGA technology.

(638) As revealed by the Market Investigation, it appears that access to mobile telecommunications services on competitive terms (including 4G services) is key for convergent operators to offer credible multiple play services involving a mobile component in Spain. The Market Investigation has shown that the current wholesale mobile regulation implies substantial legal uncertainties, first of all as regards the price of access – which is not regulated - and, second, as regards access to 4G technology.

(639) The Commission notes that, in its resolution dated 30 October 2014, the CNMC considered that 4G technology is not excluded from the market product of the 2006 resolution and therefore subject to the current obligations under market 15. In the same resolution, the CNMC pointed out to the various MNOs the non-discrimination obligation and the obligation to provide such regulated services at reasonable prices.

(640) Nonetheless, the Commission notes that, despite the resolution dated 30 October 2014, there is currently legal uncertainty if and to what extent access to 4G services can be obtained by MVNOs. In this regard, the Commission notes that the views of MNOs such as the Notifying Party and of MVNOs such as the respondents to the Market Investigation diverge as regards the legal force of the CNMC opinion of 30 October 2014. Even if it is assumed that the resolution of the CNMC is legally binding on MNOs, the Commission stresses that it would be still time-consuming and generally difficult for MVNOs to obtain access to 4G services. Because of the MNOs' reticence in providing such access, MVNOs would need to clarify the legal situation before the CNMC or even before a court at the end of each negotiation with an MNO.

(641) The Commission has explained that, to be able to offer aggressive multiple play offers involving a mobile component on the market, telecommunications operators need to have access to the fixed and mobile components of the bundle. Whereas Telefónica, Vodafone/ONO and Orange are both MNOs and operate a fixed network, Jazztel (and previously, ONO) have been the only convergent players who have been able to compete aggressively on the convergent market by cross-subsidizing the terms of their retail mobile offers through their fixed offers (where they operated owned fixed networks based on ULL – Jazztel – or HFC technology – ONO).

(642) After the acquisition of ONO by Vodafone and following the Proposed Transaction, no other player will be able to timely and sufficiently enter the market to outweigh the anticompetitive effects of the Proposed Transaction. Access to the fixed component of the bundle is difficult for the reasons explained in section 7.2.8. At the same time, no other player operating its own fixed network would be able to enter the possible multiple play markets involving a mobile component by relying on wholesale access to the mobile component, since there is no other fixed network operator active at national level other than Telefónica, Vodafone/ONO and the merged entity. Finally, because of the currently unattractive commercial terms attached to the regulated wholesale broadband access in Spain and to the regulated wholesale access to mobile networks in Spain, no other new player would be able to enter the markets. In fact, the Commission notes that no player has successfully

(643) entered the possible multiple play markets involving a mobile component in the past by relying entirely on regulated wholesale broadband access and wholesale access and call origination on mobile networks.

(644) In light of the above and given the additional barriers to entry for providing the mobile telecommunication services, the Commission considers that the barriers to entry into multiple play markets involving a mobile component (e.g., the possible retail market for triple- and quadruple-play services) are even higher than the barriers to entry into the retail markets based on products provided only over fixed infrastructure (e.g., the market for fixed Internet access services and for dual-play products).

7.2.9. Impact of the proposed transaction on the deployment of NGA networks

In parallel to the current evolution of the Spanish telecommunication markets towards convergence (see sections 7.1.3, 7.2.3 and 7.2.4), the other major trend is the rapid deployment of NGA networks, in particular FTTH. This section discusses the likely impact of the proposed transaction on the deployment of NGA networks. A distinction is made between the pure FTTH roll-out which comprises only the operator's own physical fibre infrastructure and its NGA deployment/coverage which can include also other types of infrastructure different from fibre (like HFC cable) as well as bitstream access to the NGA infrastructure of another operator under a wholesale or access agreement between the two operators. In this section the Commission concludes that, contrary to the Notifying Party's claim, the proposed transaction is unlikely to lead to any significant increase in the FTTH roll-out by the merged entity, as compared to a standalone scenario of the combined FTTH footprint of Orange and Jazztel. The proposed transaction could also lead to a loss of future competition in those geographic areas which, absent the merger, would have been covered by the separate FTTH networks of both Jazztel and Orange. However, in the Commission's view, this loss of future competition cannot be established with the required degree of certainty. While the future FTTH footprint of the Parties may give a certain indication as to the Parties' future competitive position, it is nevertheless an imperfect proxy for measuring likely market power in the future (see recital (226) above).

7.2.9.1. Plans for deployment of NGA networks in Spain

(645) The incumbent operator Telefónica was the first to start rolling out FTTH on a commercial scale in 2011, reporting a significant national footprint by the end of 2014.

(646) The other major national players (Jazztel, Vodafone and Orange) reacted by each starting the roll-out of their own FTTH network: Jazztel entered in a co-deployment agreement with Telefónica for the roll-out of fibre to 3 million BUs while Orange joined forces with Vodafone in order to put in place a FTTH network expected to cover 2 million BUs.

Under the initial co-deployment agreement between Orange and Vodafone the FTTH target was 3 million BUs but after the acquisition of local cable operator ONO by Vodafone in July 2014 the agreement was amended and the target changed to 2 million BUs with the possibility of Orange to request bitstream access on the HFC cable network of ONO for up to 1 million BUs.

(647) The table hereafter shows the coverage of NGA networks at the end of 2014 as well as the expected coverage by the end of 2017. The total number of BUs in Spain is around 25 million. BUs comprise both residential buildings and offices, with the total number of residential homes/households being around 18 million.

Table 29: NGA networks (HFC cable and FTTH) by operator per building unit in Spain (given in millions of BUs)

End of: Telefónica Vodafone/ Ono Orange Jazztel

2014 10 mln BUs 8 mln BUs 0.8 mln BUs 3 mln BUs

2017 20 mln BUs 10 mln BUs […]* mln BUs […]*mln BUs

Source: Commission based on information provided by the Parties

(648) Telefónica and Vodafone are at present much more advanced in the deployment of NGA networks than the Parties. The figures for 2017 show, however, that each of the Parties will significantly expand its NGA network in the coming years.

7.2.9.2. Notifying Party's view

(649) The Notifying Party submits in the Form CO that one of the main rationales of the transaction is to allow the merged entity to significantly increase its FTTH roll-out in order to compete more effectively with the two leading operators (in terms of NGA coverage) Telefónica and Vodafone. The Notifying Party submits that combining the existing customer base of the Parties will allow the merged entity to deploy fibre in areas where investment into FTTH would not have been economically justified on a standalone basis for neither Orange nor Jazztel.

(650) In its reply to the Article 6(1)(c) Decision, the Notifying Party claims that the geographic footprint of the FTTH network of the merged entity will be wider than the geographic footprint of the combined networks of each Jazztel and Orange pre-merger. The Notifying Party argues that, in a standalone scenario, Jazztel's plans are limited to the deployment of a fibre network to cover 5 million BUs, while Orange's objective was to reach […]* million BUs. In its reply to the Article 6(1)(c) Decision, Orange assumes that the total footprint of the Orange and Jazztel combined networks would reach a […]* million BUs with an overlap of […]* million BUs between the two networks. Thus, in a standalone scenario, the combined fibre footprint of the independent entities Jazztel and Orange would be limited mainly to high density areas of large cities and their metropolitan areas but would not include medium density areas.

(651) Orange further submits in its reply to the Article 6(1)(c) Decision that the coverage of the merged entity could reach […]* million BUs covering all medium density areas in Spain.

(652) On 15 January 2015, Orange submitted the AM report which compares the combined future FTTH roll-out of Orange and Jazztel on a standalone basis with the expected

See recitals (670) to (683) for more detailed explanation on the Parties' figures for 2017. As for Telefónica and Vodafone / ONO, refer to recitals (225) to (228).

(653) According to the AM report, the total combined fibre footprint of Orange and Jazztel in a standalone scenario (excluding any overlap between their two fibre networks) will amount to […]* million BUs at the end of 2017, (with the overlap amounting to […]* million BUs).

Table 30: Orange and Jazztel standalone fibre roll-out according to the AM report

Footprint - BUs End of 2014 End of 2017

Orange standalone (A) […]* […]*

Jazztel standalone (B) […]* […]*

Overlap under standalone scenario (C) ( […]*% of Jazztel's coverage and […]*% of Orange's coverage)

Orange and Jazztel taken together (standalone scenario) (D=A+B-C)

Source: AM report submitted by the Notifying Party

(654) Based on the theoretical profitability model developed in the AM report, the merged entity would extend its FTTH coverage to a total of 11.1 million BUs by the end of 2017. Therefore, the transaction would result in […]* million more BUs covered compared to the standalone scenario.

(655) In addition, Orange submits that the projected […]* million BUs increase in the fibre coverage to be achieved post-merger should be assessed as an efficiency, as it complies with the three cumulative conditions for being (i) beneficial for consumers; (ii) merger-specific; and (iii) verifiable. With respect to the merger specificity, Orange submits that such scale effects can be achieved only through a merger, as opposed to co-deployment.

(656) The Notifying Party has also submitted the paper "Competitive landscape of NGA networks in Spain", which illustrates in a more detailed manner the current NGA footprint of the Parties' main competitors Telefónica and Vodafone, and aims at substantiating the Notifying Party's claim that the future fibre roll-out of both the merged entity and both Orange and Jazztel on a standalone basis would occur in areas where […]*.

(657) In response to the Commission's concerns as set out in the Statement of Objections, the Parties argue that the fibre roll-out figures that the Commission considers in its Statement of Objections as the likely standalone scenario for Jazztel (that is, […]* million BUs) are unrealistic […]*. In addition, the Notifying Party considers that the […]* BUs due to bitstream access on ONO's network ("the ONO bitstream BUs") that the Commission counts towards the likely NGA coverage of the merged entity should be excluded from any fibre roll-out as it does not represent infrastructure deployment.

(658) As regards the NGA coverage of the merged entity considered by the Commission in its Statement of Objections as likely (that is, […]* million BUs including ONO bitstream BUs, or […]* million BUs without those) the Notifying Party claims that the […]* million coverage, although stated repeatedly in Orange's internal documents and announced in public statements, should no longer be considered relevant in light of the theoretical model on investment incentives developed in the AM report (which predicts a fibre roll-out to […]* million BUs, excluding the ONO bitstream BUs). According to the Notifying Party, only the theoretical model should be taken into account as it represents the most relevant and likely scenario.

(659) In particular, the Notifying Party submits that the Commission has based its assessment in the Statement of Objections on the most extreme figures available for both the standalone roll-out of Orange and Jazztel and for the merged entity's fibre roll-out. In this regard the Notifying Party further submits that other plausible roll-out figures should be analysed. Such alternative roll-out figures, presented by the Notifying Party, aim at demonstrating that in all other scenarios, except the most extreme ones used in the Statement of Objections, the merger would result in a neutral or wider footprint (up to an extra […]* BUs).

(660) Moreover the Notifying Party argues (i) that any loss of competition in the overlapping footprint of the standalone Orange and Jazztel (amounting to […]* million BUs according to the theoretical investment incentives model) is not significant, and (ii) that the negative price effects in the overlapping area where the merger would result in a reduction of infrastructure operators would be largely outweighed by the positive price effects in the areas where the merger would result in an increase of infrastructure operators due to an enlarged fibre roll-out.

7.2.9.3. Commission's assessment

Likely NGA and FTTH footprint of the merged entity

(661) As regards the future fibre roll-out of the merged entity, the theoretical profitability model established in the AM report submitted by the Notifying Party calculates a maximum roll-out area of […]* BUs (excluding the BUs covered by the ONO bitstream) where the merged entity could deploy its FTTH network post-transaction. The Commission's review and analysis of Orange's internal documents revealed, however, that there is a discrepancy between this theoretical maximum coverage and the strategic NGA deployment plans of Orange post-transaction. Orange's […]* targets an NGA footprint in a merger scenario to a total of […]* consisting of: […]*.

[Reference to the Notifying Party's internal business documents]*.

(662) The NGA coverage target of […]* million BUs is confirmed by additional internal documents where Orange assesses […]* and considers two different scenarios to reach this objective: with or without bitstream access to ONO's HFC cable network. The NGA coverage target of […]* BUs is mentioned in a number of other strategic internal documents including the synergies documents prepared by Orange for the purposes of evaluating the Jazztel transaction (e.g. a presentation of the Jazztel acquisition project to […]*) and a document called […]* .

(663) On 17 March 2015, Orange has publically announced its plans to accelerate NGA deployment in Spain with the aim of reaching coverage of 10 million BUs by the end of 2016 instead of 2017 as previously planned. […]*[…]* . However, according to the same document, there is no additional deployment planned for the period between the end 2016 and the end of 2017. Therefore, the number of BUs covered by the merged entity with NGA networks at the end of 2017 would remain unchanged at 10 million BUs. The Commission therefore notes that the announced acceleration in the merged entity's NGA deployment will not alter the final number of BUs covered in 2017.

(664) In its assessment, the Commission needs to decide whether to accord higher evidentiary value to a theoretical model established by external consultants to the Notifying Party for the specific purpose of the merger review process than to relevant internal documents that are drawn up in the ordinary course of business, including internal documents evaluating the synergies brought about by the transaction.

(665) First, the Commission observes that there is a difference between the AM report and Orange’s business practices in the criteria based on which certain geographic areas are identified and designated as areas of interest where FTTH could be rolled out. The main criterion used in the AM report is the number of existing xDSL customers of the Parties out of the total number of BUs per city, while internal documents of Orange indicate that the Notifying Party determines the business potential for FTTH deployment of each city by taking into account […]*, […]* .

(666) Second, some of the assumptions on which the theoretical profitability model is based do not take into consideration the behaviour of the Parties in the market. For example, the model assumes that the percentage of gross adds for both Orange and Jazztel is equal to their respective market shares for retail provision of fixed Internet access services (Orange has a 24% market share in 2014, hence the assumption is that its gross adds also amount to 24%). It follows from this assumption that the merged entity, by combining Orange's and Jazztel's individual shares, will have a higher combined market share and a higher gross adds percentage than each of Jazztel and Orange would have in a standalone scenario. The higher the gross adds percentage, the higher the incentive for the operator to deploy more FTTH and acquire more new subscribers. The Commission considers that this assumption

ignores the fact that gross adds are in fact related to the aggressiveness of a given operator on the market, and not a direct result of its market share. An operator with a larger market share would not necessarily be more aggressive (resulting in higher gross adds) compared to another competitor that has a smaller market share.

(667) Third, the Commission verified the plausibility of the minimum accesses per BU threshold as set out in the AM report, which also have an impact on the incentive to deploy fibre. The model distinguishes different thresholds of existing xDSL customers depending on the geographic type of the area ([…]*). These thresholds seem to match (at least for the densely populated areas) the minimum share required in the AM report as previously indicated by Orange. However, the Commission notes that for cities where Orange's and Jazztel's deployments are already planned on a standalone basis, the assumed thresholds in the AM report are not always met (that is, Orange rolls out in regions with lower shares), which puts into question the reliability of this assumption and, by extension, of the model as a whole.

(668) Lastly, the Commission considers the model's result of a fibre roll-out of […]* million BUs to be unrealistic as it assumes that the merged entity will be able to deploy FTTH in all BUs located a given city. Such assumption fails to reflect the fact that, as explained by both Orange and Jazztel, due to various technical and practical constraints (such as civil work permits, difficulty to get licenses, etc.), the final number of BUs covered with FTTH infrastructure per city is always smaller than the initial gross amount of BUs targeted. In its internal FTTH planning, Orange uses an efficiency rate equal to […]*% of the initial gross BU deployment target for a given area. Jazztel's predicted efficiency rate for its deployment up to 2015 is equal to […]*%. Applying Orange's efficiency rate to the FTTH roll-out of the AM model would result in a maximum fibre roll-out by the merged entity of […]* million BUs. It would therefore be wrong to consider the […]* million BUs as the merged entity's likely roll-out.

(669) In light of the above, the Commission considers that for the assessment of the future potential NGA network deployment of the merged entity, more weight should be given to the strategic plans as discussed by Orange in its internal documents and its public announcements, as they reflect with higher reliability the future plans of the merged entity, than to the theoretical profitability model established in the AM report.

Likely NGA coverage and likely FTTH footprint of Orange and Jazztel in a standalone scenario

(670) As mentioned in recital (661) above, regarding the fibre roll-out of Orange and Jazztel absent the transaction on a standalone basis, the Commission considers, that a distinction should be made between NGA deployment and strict FTTH roll-out or footprint. When analysing FTTH roll-out targets and plans the […]* BUs on ONO's HFC cable network to which Orange will obtain bitstream access under its amended agreement with Vodafone should not be taken into account as bitstream

access does not represent infrastructure deployment. The bitstream access can be counted towards target figures relating to total NGA deployment. This approach is in line with the Notifying Party’s submission. In a standalone scenario, Orange would have rolled out its FTTH network to […]* million BUs: […]* million BUs under the co-deployment agreement with Vodafone and […]* million BU as standalone deployment. Adding the […]* million BU bitstream to this FTTH roll-out would result in a NGA coverage of […]* million BU.

(671) These figures submitted by the Notifying Party are further confirmed by internal documents of the Notifying Party that establish Orange's objective to roll-out a fibre network to […]* households by […]* in a standalone scenario and to […]* BUs in case of acquisition of Jazztel .

(672) Regarding Jazztel's FTTH network in the standalone scenario, the AM report considers a roll-out of […]* million BUs, assuming that Jazztel will stop investing in fibre deployment after 2015. The Notifying Party submits that, […]*, this roll-out could not be considered as likely.

(673) The Commission considers that although the possible fibre roll-out of Jazztel […]* million BU by 2017 seems at the present moment to some degree uncertain, there are a number of indications that it is a possibility that, absent the merger, Jazztel's footprint could have reached […]* million by 2017.

(674) On 24 July 2014, Jazztel publicly announced that it is working on an additional plan which envisages the roll-out of up to a total of 7 million BUs by the end of 2017. An internal document of Jazztel reveals that this additional plan has been discussed internally as Jazztel […]*. In a presentation to the Board dated 29 October 2014, which analyses the internal standalone roll-out plan, this target has been divided in three parts: (i) co-deployment with Telefónica reaching […]* BUs (achieved), (ii) Jazztel standalone Phase I for […]* (ongoing); and (iii) Jazztel standalone Phase II for an additional […]* BUs. […]*. However Jazztel seems confident in its ability to "pursue the fibre project while maintaining its strong financial position". Some preparatory steps of a technical character for the Phase II standalone deployment have been undertaken as well: conducting geographic planning and preliminary engineering works (including reserving space in Telefónica's OLT rooms) .

(675) In internal discussions, Orange also considered the Jazztel target of […]* million BUs when analysing the FTTH roll-out plans of Jazztel in internal documents discussed both within Orange España and at Board level within Orange France .

In addition, in an internal analysis of its competitors business plan, Orange considers that […]* .

(676) Jazztel however claims that its standalone Phase II deployment consists […]* .

(677) […]*, […]*. […]*.

(678) In light of this information, the Commission considers that […]* FTTH deployment covers […]*.

(679) However, the fact that […]* should not constrain the Commission to extend its analysis further into medium-term NGA deployment in the next two to three years. For this reason, the Commission's view is that Jazztel's Phase II standalone plan should also be considered. It emerges from the analysis of Jazztel's internal documents that […]*; […]*; […]* .

(680) Moreover, Orange's internal due diligence […]* .

(681) In conclusion, it appears that a number of preparatory steps for additional deployment […]* have taken place. On this basis, […]* it is nevertheless a possibility that, absent the merger, Jazztel's footprint could have reached […]*. Therefore, the Commission’s analysis will consider both scenarios, a Jazztel footprint of 5.5 million BUs and of 7 million BUs.

(682) Furthermore, the Commission notes that, contrary to the AM report, it is coherent in its analysis when considering both the additional […]* BUs for Jazztel […]* and the […]* BUs to be deployed on a standalone basis by Orange, […]*. As shown in recital (670), the AM report also considers the standalone deployment of additional […]* BUs by Orange.

(683) In the light of the above, the Commission estimates a possible NGA coverage of the networks of Orange and Jazztel under a standalone scenario and of the merged entity by the end of 2017 as follows:

Table 31: Likely NGA footprint of Orange and Jazztel by the end of 2017

Footprint AM report Commission estimate

Orange standalone (A) […]* BUs […]* BUs

(including BUs via ONO bitstream access)

Jazztel standalone (B) […]* BUs […]* BUs

(taking into account additional roll-out plans of Jazztel)

434[…] * BUs […]* BUs

Overlap under standalone scenario (C)

Orange and Jazztel taken […]* BUs […]* BUs together (standalone scenario) (D=A+B-C)

Merged entity […]* BUs […]* BUs

Source: Commission based on Notifying Party

(684) This possible NGA coverage in Table 31 is calculated using the same assumption used in the AM model for a 100% overlap between the Parties' networks at city level (scenario C) which as explained in recitals (685) and (686) below the Commission considers to be unrealistic.

(685) It should be recalled that, as pointed out in the AM report; "the model calculates the maximum overlap, therefore the number of BUs will be significantly lower" . Consequently, the combined coverage should be significantly higher. The reason is that, for lack of more granular data, in the AM report and in the Commission's analysis the future overlap analysis was undertaken at city level, which overestimates the total overlapping area compared to a potential analysis made closer to the building unit (i.e. at local exchange level) or even at building level. In the AM report, the analysis of the overlap as of October 2014 was made both at city and at local exchange level. While at city level the calculated overlap stands at […]* BUs, performing the same analysis at local exchange level resulted in a decrease in the overlap by one quarter to […]* BUs. The reason for this difference is that in each city there are different local exchanges and each operator may choose to deploy in different ones based on their own internal strategy and customer footprint. Analysing the data at city level suggests the presence of an overlap between the parties that is, however, likely more limited (or might not exist at all) because the operators chose to deploy in (partly) different local exchanges.

434 An overlap of […]* BUs is assumed in case of a footprint of Jazztel which is […]* BUs. Assuming the total size of Jazztel’s network to equal […]* BUs at the end of 2017, the overlap would likely amount to […]* BUs.

(686) In order to make a best estimate of the likely FTTH roll-out in a standalone scenario, the Commission has first calculated the maximum possible overlap between Orange and Jazztel. Using the methodology adopted in the AM report for identifying the overlap in a standalone scenario at city level, adjusted by the Jazztel and Orange standalone roll-out targets that the Commission considers the most likely ([…]* BUs and […]* BUs, respectively), the maximum overlap would be […]* BUs. As explained in recital (684) above, this calculation overestimates the likely overlap and should therefore be viewed as the upper bound for the overlap. Consequently, the […]* BUs set out in Table 31 as the combined footprint of Orange and Jazztel in a standalone scenario are to be understood as the absolute minimum footprint under the (highly unrealistic) assumption that the Parties' FTTH networks perfectly overlap at city level.

(687) In order to estimate a more realistic overlap of the Parties in a standalone scenario, the Commission finds it useful to rely on Orange's own estimates as expressed in recent internal documents (see also the extract in graph 1) and on most likely results obtained when comparing the model developed in the AM report with two different scenarios: one based on Parties' internal documents and the second comparing the results of the AM reports obtained at city level with results obtained at local exchange level).

(688) In Orange's Strategic Plan 2015-2018 of 14 November 2014, Orange assumes: (i) an overlap of almost […]* BUs between Orange and Jazztel upon implementation of the Proposed Transaction; and (ii) an additional deployment of […]* BUs by the merged entity, after deducting an overlap estimated at […]* FTTH-covered BUs between Orange and Jazztel standalone plans. This means that, when taking into account the […]* BUs of Jazztel to be rolled out until end of 2015 and the internal Orange deployment plan of […]* BUs, Orange expects a total overlap of […]* BUs. This contrasts with the overlap of […]* BUs calculated in the AM report using the same roll-out figures as input.

Graph 2: Likely NGA footprint of Orange by the end of 2017 in a standalone and merger scenario

[Reference to the Notifying Party's internal business documents]*

Source: Internal document of Orange

(689) In the Statement of Objections, the Commission considered that AM report's methodology has a […]*% efficiency rate in predicting the maximum overlap compared to Orange's internal estimations. This percentage had been calculated dividing the best estimate overlap obtained internally by Orange with the maximum overlap calculated in the AM report for an equivalent scenario (i.e. […]* BUs divided by […]* BUs = […]*%). In its response to the Statement of Objections, the Notifying Party submits, first, that the Orange internal estimation of an overlap of […]* million BUs is methodologically flawed as the two Orange estimations reproduced in graph 1 have been produced independently from each other. Second, the Notifying Party argues that this ratio is artificial as it compares the estimation done by two different models – Orange's internal estimations compared to the AM report - using different methodologies and different sources.

(690) The Commission considers, first, that while the two documents were produced in two different periods – one from October 2014, the other from November 2014 – Orange itself links them. In the Strategic Plan 2015-18 of November 2014, […]*. Therefore, the figure of […]* can well be considered as Orange's estimate for the overlap between the two FTTH networks. With regard to the second argument, the Commission notes that Orange does not substantiate the methodology used internally to estimate the NGA coverage, as compared to the AM report. In any event, the Statement of Objections had also used a second efficiency rate that does not compare different models but compares different scenarios established with the methodology of the AM report. The Notifying Party has not objected to using this efficiency rate. Therefore, in the present Decision, the Commission will use this efficiency rate for the purposes of its analysis.

(691) This efficiency rate compares the estimate of the overlap by the end of 2014 at city level, as established by the AM report, (i) with the overlap estimated at local exchange level in the AM report and (ii) with the overlap at BUs level calculated using data provided by the Parties. In this case, the efficiency rate amounts to […]*% at local exchange level (which is similar to the […]*% obtained with the methodology based on Orange's internal documents) and to […]*% at BUs level.

(692) Using these efficiency rates, the Commission estimates, for the end of 2017, ranges for different scenarios concerning the NGA footprint and FTTH rollout of Orange and Jazztel in a standalone scenario and of the merged entity. Given the uncertainty related to the NGA deployment plans of the Parties, different scenarios are considered that take into account the two uncertain aspects related to future NGA deployment of the Parties, namely (i) the […]* BU bitstream access to Vodafone/ONO's HFC network and (ii) the additional Jazztel's FTTH rollout to […]* BU via its Phase II standalone deployment plan.

(693) Table 32 below summarises the results for a number of scenarios that the Commission has run using the Parties' detailed geographic roll-out plans by city, in order to cover all plausible variations both for fibre rollout and for total NGA deployment .

(695) Therefore, according to the most likely scenario (the "median" scenario in Table 32 above), the combined NGA footprint of Jazztel and Orange on a standalone basis would be similar to, or even greater than, the estimated footprint of the merged entity ([…]* BUs), thereby at least substantially reducing or more likely eliminating any residual potential benefit of additional fibre roll-out in a merger scenario as claimed by the Notifying Party.

Overlap of Orange's and Jazztel's NGA networks

(697) Absent the Proposed Transaction, as indicated in Table 31 above, according to the Commission's estimate, Orange's and Jazztel's standalone networks would have a […]* to […]* BU overlap by the end of 2017. Given the extensive current coverage of Telefónica's fibre network and its ambitious roll-out plans, consumers located in the geographic areas of the Orange-Jazztel overlap would have been able to choose their provider of retail Internet access services among at least three operators each having its own proprietary NGA infrastructure: Telefónica, Orange and Jazztel. Taking into account the fact that the HFC cable network of ONO, now operated by Vodafone, has a national presence of nearly 8 million BUs (expected to expand to 10 million BUs by the end of 2017), it can be assumed that in a portion of the Orange-Jazztel overlap, consumer's choice for a NGA provider would have been possible between four operators: Telefónica, Orange, Jazztel and Vodafone/ONO.

(698) The Proposed Transaction could therefore result in a reduction of the number of operators with NGA network within the areas of future overlap between the standalone Orange and Jazztel fibre networks, between […]* and […]* BUs by the end of 2017 – a significant part of the combined entity's footprint (between […]*% and […]*% of the merged entity's footprint). This reduction would bring the number of NGA operators from four to three in the areas where both Telefónica and Vodafone have already or would have deployed NGA infrastructure and from three to two in those areas covered with a NGA only by Telefónica. Therefore, the Proposed Transaction could lead to a loss of competition concerning not only existing competition (mainly based on wholesale access to the xDSL network of Telefónica) but also future competition in those geographic areas which absent the merger would have been covered by the separate FTTH networks of both Jazztel and Orange.

(699) The Notifying Party submits, in essence, that post-merger, there will be three remaining NGA networks nationwide, which compares favourably to other European countries where fewer NGA networks are being rolled-out .

(700) In this respect, the Commission notes that Spain is an example of intense NGA infrastructure roll-out in Europe. According to internal documents of Jazztel, this is due in particular to several important factors that facilitate FTTH deployment:

 favourable regulated access to the existing passive infrastructure(the ducts installed and operated by the incumbent Telefónica), which eliminates a significant part of the civil works costs (up to 80% of the overall roll-out cost);

 high population density in the urban areas (Spain has a high-rise apartment stock of 30% while in France are 16% and only 6% in Germany);

 regulated shared access to vertical elements of the network (last portion of the fibre network reaching the individual households within the building): whichever operator is the first one to install vertical fibre elements in a certain building unit is obliged to provide access to those vertical elements to any other operator that has deployed a fibre line to the said building unit .

(701) Jazztel estimates the average FTTH roll-out cost per home passed in densely populated areas to be significantly lower compared to Germany and the UK .

(702) The Commission considers therefore that the assessment of the competitive situation needs to be specific to the geographic market concerned. Comparisons of the number of NGA networks being rolled-out across different countries, as suggested by the Notifying Party, are not helpful as the conditions for NGA roll-out vary significantly between countries, as established in recital (700) above.

Conclusion on the impact of the Proposed Transaction on the roll-out of NGA networks

(703) The Commission concludes that, contrary to the Notifying Party's claim, the Proposed Transaction is unlikely to lead to any significant increase in the FTTH coverage by the merged entity, as compared to the standalone scenario. As regards the potential loss of competition in areas where the NGA networks of the Parties would have overlapped in the future, the Commission considers that such a finding of loss of future competition cannot be established with the required degree of certainty for the reasons discussed in recital (226) above.

7.2.10. Commission's assessment of the Notifying Party's efficiencies claims

(704) The Commission considers any pro-competitive effects of efficiencies that benefit consumers as part of its overall assessment of the merger. For the Commission to take account of efficiency claims in its assessment of the merger and to be in a position to reach the conclusion that, as a consequence of efficiencies, there are no grounds for declaring the merger to be incompatible with the internal market, the efficiencies must be substantiated and satisfy the three cumulative criteria defined in the Horizontal Merger Guidelines :

(a) Verifiability: efficiencies have to be verifiable such that the Commission can be reasonably certain that the efficiencies are likely to materialise and be substantial enough to counteract a merger's potential harm to consumers ;

(b) Merger specificity: efficiencies have to be a direct consequence of the merger and cannot be achieved to a similar extent by less anticompetitive alternatives ;

(c) Benefit to consumers: efficiencies have to benefit consumers in the sense that they should be substantial and timely and should, in principle, benefit consumers in those relevant markets where it is otherwise likely that competition concerns would occur .

(705) The Horizontal Merger Guidelines further explain that the burden of proof for showing that efficiencies fulfil the above criteria lies with the Notifying Party as most of the information is solely in the possession of the Parties. It is therefore incumbent upon the Notifying Party to provide in due time all the relevant information necessary to demonstrate that the claimed efficiencies are merger-specific and likely to be realised. Similarly, it is for the Notifying Party to show to what extent the efficiencies are likely to counteract any adverse effects on competition that might otherwise result from the merger and therefore benefit consumers. Furthermore, evidence relevant to the assessment of efficiency claims should include, in particular, internal documents that were used by the management to decide on the merger, statements from the management to the owners and financial markets about the expected efficiencies, historical examples of efficiencies and consumer benefit, and pre-merger external experts' studies on the type and size of efficiency gains, and on the extent to which consumers are likely to benefit .

(706) In the following recitals, the Commission will assess whether the submitted efficiency claims fulfil the three criteria defined in the Horizontal Guidelines.

7.2.10.1. Efficiencies stemming from allegedly increased fibre roll-out post-merger

(707) As already mentioned in recital (654) the Notifying Party submits that according to the theoretical profitability model developed in the AM report the Proposed Transaction would result in a […]* million BUs increase in the fibre footprint of the merged entity, which would be a clear benefit to the consumers located in the area of these additional […]* million BUs covered by FTTH.

Verifiability

(708) The Commission's assessment of the future fibre coverage under both the standalone and the merger scenarios, carried out in recitals (661) to (695) above, already showed that the Proposed Transaction is unlikely to lead to any significant increase in the FTTH coverage by the merged entity, as compared to the standalone scenario. In the more likely scenario, the merged entity's footprint would even be […]*% smaller. Under the (highly unrealistic) assumption of perfect overlap at city level in the Parties' NGA networks, the merged entity's coverage would increase by around […]*% as compared to the standalone scenario.

(709) Moreover, none of the internal synergies documents prepared by Orange for the purposes of assessing the acquisition of Jazztel envisage an increase in the coverage achieved by the merged entity (relative to the joint coverage of the two standalone networks). There is no reference in these documents to a higher value (in the form of greater profits for the merged entity) generated by the merger and deriving from more effective competition against rival operators. The fibre-related synergies contained in the Orange documents […]*. However, these synergies do not relate to higher value created by the merged entity via an increased fibre roll-out. Therefore, the Notifying Party has not been able to produce any of the documentary evidence indicated in paragraph 88 of the Horizontal Merger Guidelines, in support of its claim on greater fibre roll-out post-merger.

(710) The Commission therefore considers that the claimed efficiency in terms of increase in the NGA network coverage is not sufficiently verified.

Merger specificity

(711) The Notifying Party in its efficiency claims submitted via the AM report argues that the increased fibre footprint of the merged entity could be achieved only through the merger. Orange submits that such a large increase in the fibre footprint could not result from a possible co-deployment agreement between the Parties for a number of reasons. First, a possible co-deployment scheme would impose certain limitations in the open exchange of confidential and granular information between the Parties which inevitably would affect the scope of the co-deployment areas. The Notifying Party considers that full transparency can be obtained only in the framework of the Proposed Transaction. Second, under a co-deployment scheme there would be a certain number of areas where co-deployment would still not make the business case for one or both of the Parties as their respective thresholds of existing customers in the said areas would be too low. And third, Orange submits that the level of uncertainty in co-deployment is much higher compared to a merger scenario, one practical example being the revised scope of the joint fibre roll-out of Orange and Vodafone following the acquisition of ONO by Vodafone in 2014.

(712) The existence of co-deployment agreements in Spain (between Jazztel and Telefónica and between Orange and Vodafone) shows that the possible limitation in the exchange of confidential information between the co-deployment parties does not seem to represent an unsurmountable hurdle for the efficient co-deployment of FTTH. Although the level of uncertainty as regards joint roll-out may be higher compared to a merger scenario it should be noted that the Parties are free to negotiate contractual provisions in a potential co-deployment agreement to safeguard against major dramatic changes in the deployment strategy of one of the partners. The Commission notes that although the […]* BUs target under the Orange/Vodafone agreement has been altered following the Vodafone/ONO transaction to […]* BUs, in fact the amended agreement envisages Orange to obtain bitstream on Vodafone/ONO's HFC cable network in order to offset this reduction in the fibre coverage target. In case such bitstream access is not technically feasible the parties would increase their roll-out commitment to […]* BUs.

(713) As to the Notifying Party's argument that, under a co-deployment scheme, there will be a number of areas where fibre roll-out would still not make the business case for one or both of the partners due to their asymmetric shares of existing DSL customers, the Commission notes that this argument relies on the assumption that the split of the deployment costs for each area would always be equal (50:50) between the co-deployment partners. However, it is likely that co-deployment partners could design a way that would overcome this issue, for example by sharing roll-out costs

[Reference to the Notifying Party's internal business documents]*.

according to the partners' shares of subscribers in a given area where fibre is rolled out. Alternatively it could be agreed that fibre is rolled out in some areas where one partner has more subscribers and in other areas where the other partner has more subscribers so that, overall, both partners achieve an acceptable balance between benefits and costs.

(714) The Commission considers that the Notifying Party has failed to prove the merger specificity given that a number of co-deployments have happened in the market in the past and joint roll-out is still an option for the future. Orange, in a presentation discussing the potential coverage of its standalone plan and ranking the Spanish territory for its business potential, states that there is […]*.. In the same document, […]* .

Benefit to consumers

(715) The Notifying Party argues that the increased Orange/Jazztel footprint post-merger would generate entry in areas that would otherwise be served by Telefónica and possibly Vodafone only. This effect tends to lower the prices of Telefónica and Vodafone, which in the context of national pricing policies could offset the upward pricing pressure due to the first effect.

(716) The Commission considers that whereas the claim that a larger fibre footprint of the merged entity may indeed intensify competition with Telefónica and Vodafone, which due to national pricing policies could have an effect on the overall price level of the relevant products, the entire argument relies on the assumption that the notified transaction entails merger specific additional fibre rollout. However, a significant increase in the merged entity's fibre footprint due to the transaction is unlikely to materialize.

Conclusion on Efficiencies stemming from allegedly increased fibre roll-out post-merger

(717) In light of the above, the Commission concludes that the AM report significantly overstates the potential increase of fibre coverage resulting from the Proposed Transaction by underestimating the standalone coverage of Orange and Jazztel absent the merger and by overestimating the merged entity's future roll-out. According to the Commission's estimate the Proposed Transaction is unlikely to lead to any significant increase in the FTTH coverage by the merged entity, as compared to the standalone scenario. In the more likely scenario, the merged entity's footprint would actually be […]*% smaller. In a more unrealistic scenario, the merged entity's coverage would increase by up to […]*% as compared to Orange and Jazztel taken together absent the transaction.

(718) Furthermore, the Parties failed to prove that any additional FTTH rollout would be merger specific. From the analysis of both past co-deployment agreements and internal documents of the Parties on future rollout, the Commission showed that similar objectives may be achieved through partnership with other market players.

(719) The Commission therefore concludes that the merger will not lead to any significant pro-competitive effects arising from reaching additional households that Orange and/or Jazztel would not have covered in a standalone scenario. Consequently, any efficiency claims based on additional roll-out appear to be not sufficiently verifiable or merger-specific.

7.2.10.2. Efficiencies from an improved offer of quadruple-play bundles

Notifying Party's view

(720) The Notifying Party argues that the notified transaction will accelerate the move of the merged entity into the top tier of offerings compared to a non-merger counterfactual, since it will enjoy better fibre coverage. To the extent Orange enters due to the merger into the quadruple-play product type, intensified competition will lead to lower prices for that product type. Lower prices in the quadruple-play product type constrain the prices of triple or dual-play products. As a consequence, potential anticompetitive effects from the loss of competition must therefore be counterbalanced with the pro-competitive effect of price reductions by Telefónica in its quadruple-play offers.

(721) In its reply to the Statement of Objections, the Notifying Party further argues that greater scale in terms of footprint and customers means greater ability to obtain access to attractive content. By increasing the customer base this allows the Parties to compete more credibly for content and as a result also for 4p customers.

Commission's assessment

(722) The underlying assumption that the merged entity will be in a better position to offer quadruple-play products as a consequence of the Proposed Transaction appears not to be supported by the facts. The Notifying Party raised two arguments for being in a better position to offer quadruple-play products. First, the Notifying Party has clarified that the higher transmission speeds with greater reliability on the basis of fibre connections is particularly important for the delivery of audio-visual content. However, as set out in recital (718) of this Decision, the Commission considers that the notified transaction does not lead to significant merger specific additional fibre roll-out. This in turn implies that the Proposed Transaction is unlikely to improve the ability of the merged entity to compete effectively to offer quadruple-play products.

(723) Although each of the merging parties will in the absence of the merger be likely to have a smaller fibre footprint than the merged entity, each party will be able to offer a TV component in particular wherever it will have fibre access. Moreover, audio-visual content can be also transmitted on copper to the extent the transmission speed is high enough.

(724) As to the second argument of economies of scale in the acquisition of content, the Notifying Party has not submitted any evidence (other than a claim that this basic principle is recognised by regulatory authorities such as OFCOM) to support or quantify its claim. The Commission is therefore not in a position to validate this claim.

(725) Moreover, the Commission notes that Orange is already successfully selling products that include a TV component now. For instance, Orange had […]* residential subscribers of products which include a TV component at the beginning of Q4 2014. Of these, […]* subscribers had quadruple-play products .

(726) Therefore, the Commission considers that increased competitive pressure based on a merger specific improved ability by the merged entity to offer quadruple-play products has not been demonstrated.

7.2.10.3. Improved ability to serve consumers based on fibre network and to save xDSL access charges

Notifying Party's view

(727) According to the Notifying Party, the merger-related build-out of FTTH also makes it possible to shift consumers onto fibre. This allows the operators to avoid access charges to copper to deliver xDSL services, thus lowering the effective marginal cost of their services – even when offering exactly the same bundle (by capping the speed and service at the existing xDSL level).

Commission's assessment

(728) In Annex A to the Statement of Objections, the Commission pointed out that the claims relating to xDSL access charges need to be assessed by applying the three cumulative criteria for efficiencies. These efficiency claims must be verifiable, merger specific and passed on to consumers, for the Commission to be able to take them into account.

(729) In particular, the Commission pointed out that the Notifying Party could quantify the claimed effects reasonably well by way of additional submissions. This would allow the Commission to balance consumer benefit from the claimed cost savings against competitive harm within the Commission's quantitative approaches .

(730) However, the Notifying Party has failed to produce the necessary information to verify and possibly quantify this claim.

(731) The Commission stresses that it is incumbent upon the Notifying Party to provide in due time all the relevant information necessary to demonstrate that the claimed efficiencies are merger-specific and likely to be realised. Given that the Notifying Party has not provided the information required to assess the above-mentioned claims, the claimed efficiencies cannot be acknowledged.

7.2.10.4. Efficiencies stemming from the elimination of double marginalisation of mobile services provided by Orange to Jazztel

(732) The Notifying Party also claimed that the notified transaction will reduce marginal costs in the provision of mobile services, as it eliminates the wholesale margin that Jazztel currently pays to Orange for hosting its MVNO activity. This efficiency claim is based on the elimination of so-called "double marginalisation", and it can be the result of mergers that bring together complementary assets .

(733) In Annex A to the Statement of Objections, the Commission noted that the Notifying Party had not submitted the information required to evaluate this efficiency claim, and therefore did not assess the impact of the claimed elimination of double marginalisation . Following the Statement of Objections, the merging parties provided additional information in support of this claim.

Verifiability

(734) Based on the additional submissions by the Parties, the Commission considers that the efficiencies stemming from the elimination of double marginalisation of mobile services are sufficiently verified. Based on Orange's average cost per MVNO mobile line, the difference between Orange's costs and the charged average wholesale price per 3p product to Jazztel amounts to EUR […]* in 2013.

(735) The Commission considers that after the merger Orange will likely base its price for hosting Jazztel as an MVNO on the actual costs of providing the service. This would reduce the (transfer) price that Jazztel pays to Orange in relation to the provision of mobile services. It is likely that Jazztel currently bases its pricing decisions inter alia on the wholesale costs for the mobile services that it pays to Orange. To the extent that the actual costs for providing the mobile services are below the wholesale cost faced by Jazztel, the costs that are relevant for pricing decisions concerning the former Jazztel products would thus decrease after the merger.

(736) Jazztel submitted that it paid on average EUR […]* and EUR […]* MVNO fees to Orange per fixed triple-play (3p) products subscriber per month in 2013 and 2014, respectively. The Commission further notes that the MVNO contracts between Jazztel and Orange […]*. Furthermore, Jazztel submitted that there were on average […]* and […]* active SIM cards per fixed 3p subscriber in 2013 and 2014, respectively. Jazztel also submitted an average usage per active SIM card of […]* outgoing voice minutes, […]* incoming voice minutes, […]* outgoing SMS and […]* MB of mobile data usage per month in 2013 .

(737) Orange submitted the mobile costs it incurred in 2013 for mobile service provided to full MVNOs in accordance with regulatory cost accounting data submitted to the

firm individually maximises profits and does not take into account that a higher markup reduces the demand and thus the profits on the mobile services of the other firm. This may result in inefficiently high prices.

Annex A to Statement of Objections, paragraph 28.

To the extent Orange's actual costs of providing the mobile services are below the wholesale prices charged to Jazztel.

See file "Template 01 - aggregate.xls" submitted on 26/01/2015. The average cost per fixed subscriber is derived by dividing the total wholesale costs paid by Jazztel to Orange for 3p products by the number of fixed subscribers. 3p products offered by Jazztel contain mostly fixed voice, fixed broadband and mobile services.

See LTE Agreement signed Jazztel 17 June 2014, "Acuerdo firmado LTE Jazztel 17-6-14.pdf", "Acuerdo renovación OMV Jazztel_Vdef.doc" and "Versión firmada renovación OMV Jazztel 26-9-13_vdef.pdf".

There are more active SIM cards for 3p products than fixed subscribers because sometimes 3p packages include more than one SIM card per package for additional mobile users.

See Jazztel reply of 20 April 2015 to RFI of 17 April 2015. The average usage per active SIM card of […]* outgoing voice minutes, […]* incoming voice minutes, […]* outgoing SMS and […]* of mobile data per month in 2014.

(739) Based on the submitted figures, the estimated cost savings are derived as follows. First, the total CACR for MVNOs is divided by the number of hosted active MVNO SIM cards. This yields costs of EUR […]* per month per active SIM card. Second, taking into account Jazztel's ratio of active mobile SIM cards per 3p subscriber, this results in mobile costs of EUR […]* per 3p fixed subscriber. The 2013 estimate of the costs per active MVNO SIM card (EUR […]* per month) is also used for 2014 as a proxy, which implies mobile costs of EUR […]* per 3p subscriber. Based on these figures, the resulting indicative cost saving for 2014 is of EUR […]* per 3p fixed subscriber.

Merger Specificity

(740) The Commission considers that the cost reduction in the incremental costs for mobile services compared to the wholesale price paid by Jazztel cannot be realistically achieved by other, less anti-competitive means.

(741) One theoretical possibility to reduce the incremental costs would be to commit to pay a fixed wholesale fee (independent of the number of hosted subscribers or the traffic generated by the hosted subscribers) that is determined for a period of several years in advance in return for a low variable wholesale price.

(742) The Notifying Party submits that its MVNO contracts in Spain have a similar pricing structure and include typically fees per active subscriber and variable price per service usage (mobile voice, SMS and mobile data). On top of this general pricing structure that can be described as a pay-as-you-go model, there are ad hoc temporary promotional agreements based on growth objectives to help and promote growth of the MVNOs.

See Orange reply of 17 April 2015 to RFI 18 and Orange reply of 20 April 2015 to RFI of 17 April 2015.

See section 7 of Annex A to this Decision for more details.

See Orange reply of 17 April 2015 to RFI 18 and Orange reply of 20 April 2015 to RFI of 17 April 2015. Alternatively, the CACR per active MVNO SIM card could be estimated on the basis of the average usage per active SIM card of Jazztel in 2013 yields a monthly cost of EUR […]* per month per active SIM card. This results in mobile costs of EUR […]* per 3p product and hence in a slightly lower difference between Orange's cost and the wholesale charges to Jazztel. Based on this alternative estimate the same conclusion of paragraph (750) would apply.

(743) The Notifying Party further submits that to its knowledge there are no MVNO contracts with a price structure whereby the MVNO would pay a fixed wholesale price (independent of the number of hosted subscribers or the traffic generated by the hosted subscribers) that is determined for a period of several years in advance and a low variable wholesale price. According to the Notifying Party, it is practically impossible to negotiate a fixed wholesale price (independent of the number of hosted subscribers or the traffic generated by the hosted subscribers) which would cover the non-incremental costs of rendering the service. The Notifying Party further claims that, as the market has demonstrated, it is more straightforward to negotiate pay-as-you-go wholesale contracts with a unitary price per service (that is, Euros per minute of voice or per GB of data) covering average costs of delivering the service to the MNO.

(744) The Commission accepts that contracts as described in recital (741) are very uncommon (outside the context of recent merger commitments) and probably inexistent in Spain. The Commission also considers that agreeing on such contracts may be particularly difficult if both parties do not share common information as to the number of subscribers that the MVNO can realistically attract in the contracted period. Therefore, the Commission considers that the cost reduction in the incremental costs for mobile services compared to the wholesale price paid by Jazztel cannot be realistically achieved by these types of contracts.

(745) For the above-mentioned reasons the Commission considers that Jazztel's mobile cost savings are merger specific.

Benefit to consumers

(746) In Annex A to the Statement of Objections, the Commission pointed out that if the variable mobile cost savings of Jazztel were to be quantified, the Commission could balance consumer benefit from the claimed cost savings against competitive harm within the quantitative analysis conducted. The information provided by the Notifying Party shows that the mobile cost savings which Jazztel would benefit from as a result of the merger are of a variable cost nature, and as such could benefit consumers.

(747) Within the Commission's quantitative framework, the pass through of variable cost savings to subscribers is treated similarly to the anti-competitive effects. The Commission considers that this similar treatment is justified by the similarity of these concepts from an economic point of view.

(748) The Commission has thus considered the effect of accounting for the verified expected cost savings in the calibrated merger simulation. As set out in detail in section 7 of Annex A to this Decision, for both baseline scenarios considered in the quantitative analysis, the predicted anti-competitive effects of the merger are partially offset by Jazztel's reduced incentives to raise prices due to Jazztel's (perceived) cost savings. Compared to the results reported in section 7.2.7, the expected cost savings reduce the predicted market wide price increases only by 0.4 percentage points. This means that in both scenarios significant market-wide price increases remain after accounting for efficiencies.

(749) The Commission reiterates that the (perceived) cost savings applicable to former Jazztel products in 2014 are imprecise as no accurate estimates for 2014 were submitted by Orange. The Commission therefore bases its assessment on the benefit to consumers due to the elimination of double marginalisation using the 2013 data. The estimated effects of this efficiency for 2014 are therefore only indicative.

Conclusion on the elimination of double marginalisation of mobile services provided by Orange to Jazztel

(750) For the above mentioned reasons, the Commission accepts that the elimination of double marginalisation relating to mobile services provided by Orange to Jazztel will likely reduce the incentives to raise prices of former Jazztel products post-merger. However, these efficiencies are not expected to entirely offset anti-competitive effects of the merger. The anti-competitive effects net of this efficiency remain significant.

7.2.11. Conclusion on markets including fixed Internet access services

(751) In light of the above, the Commission concludes that the Proposed Transaction will significantly impede effective competition on the market for fixed Internet access services, possible markets for multiple play services, on the possible market for dual-play services, on the possible market for triple-play services, and on the possible market comprising triple- and quadruple-play services in Spain. Given that Jazztel is not active on the possible market for quadruple-play services, the Proposed Transaction will have no impact on this latter market.

(752) The Commission considers that the impediment of effective competition resulting from the Proposed Transaction concerns mainly competition based on wholesale access to the xDSL network of Telefónica. As regards future competition based on NGA networks, the Commission considers that the loss of competition caused by the Proposed Transaction cannot be established with the required degree of certainty.

7.3. Retail market for fixed voice services

(753) Both Orange and Jazztel provide retail fixed voice services in Spain and had a combined market share on the overall market for fixed voice services of 22.84% by volume (Orange 12.38% and Jazztel 10.46%) and 7.56% by value (Orange 2.53% and Jazztel 5.03%) for the first half of 2014. This overall market comprises all fixed voice services regardless of whether these are sold on a standalone basis or as part of a bundle.

(754) On the narrower segment of provision of fixed voice services to residential customers the combined market share of the Parties is 24.11% by number of subscribers (Orange 12.63% and Jazztel 11.48%). The Notifying Party was unable to provide market shares based on revenue.

(755) The Notifying Party submits that the combined market share of the Parties in 2013 both in terms of volume, measured in number of minutes, (18.98%) and in terms of value (6.87%) is below the 20% threshold, and only if the number of subscribers is taken into consideration the combined market share of the Parties is slightly above the 20% threshold: 21.96%.

(756) The Notifying Party argues that the market shares in terms of revenues and in minutes of traffic are more relevant than the market share by subscribers as the market share in revenues illustrates more correctly the real economic impact of the operator for the retail provision of fixed voice services, while the market share in minutes shows the real volume of traffic generated by the customer base of each operator. The Notifying Party submits that subscribers can have very different consumption patterns (some subscribers generating significant number of minutes and consequently revenue while others make or receive a very limited number of calls thus generating very little traffic and revenues for the operator) and market shares by subscribers does not provide insight on the profitability of the subscriber base of the operator.

(757) Therefore the Notifying Party further submits that it does not consider the market for retail supply of fixed voice services in Spain to be an affected market.

(758) In any event the Notifying Party argues that in the retail fixed voice market the merged entity will be the third supplier in the market as both Telefónica and Vodafone/ONO will have larger market shares.

7.3.2. Commission's assessment

(759) The market share of the merged entity based on subscribers amounts to 23% in the retail market for supply of fixed voice services in Spain. This share is slightly higher for residential customers: 24.11% (12.63% for Orange and 11.48% for Jazztel) and much lower for business customers: 16.14% (11.08% for Orange and 5.06% for Jazztel).

147

would remedy also any such potential competition concern on the retail market for fixed voice services.

7.3.3. Conclusion

(767) In light of the above, the Commission concludes that the Proposed Transaction would not significantly impede effective competition on the Spanish retail market for fixed voice services, including its possible segments.

7.4. Retail market for mobile telecommunication services

(768) Both Parties provide mobile telecommunication services to end customers in Spain. While Orange acts as an MNO, Jazztel is an MVNO hosted on Orange's network. On the Spanish market, there are four active MNOs: Telefónica, Vodafone, Orange and Yoigo.

(769) In the first half of 2014, the Parties had a combined market share of 22.5% by revenues (Orange 20.9%, Jazztel 1.6%) and 25.7% by subscribers (Orange 22.6%, Jazztel 3.1%). On the market segment of mobile post-paid telecommunications services, the Parties had a combined market share of 24.1% by revenues (Orange 22.2%, Jazztel 1.9%) and 27.6% in terms of subscribers (Orange 23.3%, Jazztel 4.3%).

7.4.1. Notifying Party's view

(770) The Notifying Party submits that the Proposed Transaction will not lead to competition concerns on the retail mobile telecommunications market in Spain as Orange’s competitive position in the Spanish retail market for mobile telecommunications services will remain virtually the same after the acquisition of Jazztel.

(771) First, the removal of Jazztel as a provider of retail mobile telecommunication services is not likely to impede effective competition in the market due to the limited market share of Jazztel. The Proposed Transaction, as a consequence, will not have an important impact on Orange’s position in the market. Orange will remain the third largest provider in the market by revenue (22.5% combined market share) behind Telefónica (34.1%) and Vodafone (27.6%).

(772) Second, the Notifying Party submits that Orange and Jazztel are not close competitors. Jazztel is focused on the low-cost segment of the market and its mobile telecommunications services are bundled with its fixed telecommunications services in order to defend its customer base in fixed services by attempting to reduce churn in those services. Therefore, direct competition between Orange and Jazztel in the retail market for mobile telecommunication services would only be relevant if bundled offers are considered as a relevant market segment, which Orange considers not appropriate.

(773) Third, the Notifying Party notes that post-merger the combined entity will continue to face competition from three MNOs: Telefónica, Vodafone and Yoigo, with a combined market share of 67.5% in terms of revenue and 61.8% in terms of subscribers. Furthermore, the Notifying Party submits that after the concentration, the merged entity will continue to face competition from a high number of MVNOs (31 active, out of which 9 are full MVNOs and 22 are service providers). According to Orange, MVNOs are aggressive players on the Spanish market for retail mobile telecommunications and represent more than 10% of the market in terms of subscribers. Their price aggressiveness has not decreased and their entry has been successful in the last years.

(774) Fourth, the Proposed Transaction will not have any impact on spectrum rights due to the spectrum cap regime in Spain.

7.4.2. Results of the Market Investigation

(775) Based on the results of the Market Investigation, the Commission did not identify competition concerns in relation to the retail market for the supply of mobile telecommunications to end customers in Spain on a standalone basis.

(776) The majority of respondents recognize that mobile-only services are still important today, especially for certain types of customers (such as young people) who do not require a fixed component. In the next 2-3 years, this situation may change, as almost half of the respondents expect mobile-only services to become less important due to the increasing popularity of fixed-mobile triple-play bundles. A higher proportion of customers are expected to purchase mobile telecommunications services within a bundle. In this context, one competitor notes that access to fixed services will be required to avoid the erosion of the mobile customer and revenue base.

(777) The Market Investigation was inconclusive as regards the closeness of competition between Orange and Jazztel on this market. On the one hand, a majority of respondents to the Market Investigation consider Jazztel and Orange to be close competitors in the provision of mobile telecommunication services. Respondents argue that Orange and Jazztel's offers target the same customers in the residential segment and they are both rather aggressive competitors on this market. On the other hand, respondents point out that Jazztel mainly provides mobile telecommunications services together with its multiple services bundles (which include fixed services) and that Orange is a much larger competitor in mobile services.

(778) Respondents to the Market Investigation consider that a sufficient number of providers of retail mobile telecommunications services will remain active on the market. However, entry on this market is considered to be difficult. One competitor perceives a possible positive impact on prices which would decrease.

7.4.3. Commission's assessment

(779) The key question is whether the removal of Jazztel is likely to significantly impede effective competition on the retail mobile market. The Commission notes, first, that Jazztel acts as an MVNO and its presence on the market is rather limited (1.6% market share by revenues). The increment brought about by the Proposed Transaction is not higher than 3% (in terms of subscribers) in the overall market and marginal share of slightly more than 1%. Moreover, competing providers of wholesale access to broadband services could easily increase their resale offerings, if the merged entity were to raise prices for the resale of broadband services. For this reason, it is unlikely that the Parties will be able to significantly raise prices for the resale of broadband services post-transaction.

(787) In conclusion, the Commission finds the Proposed Transaction would not significantly impede effective competition on the overall wholesale market for broadband access services. As regards a possible wholesale market for the resale of broadband offerings, which would be of marginal size, the Commission notes the high combined market share of the Parties, the significant constraints exercised by wholesale bitstream offers as well the possibility of providers of bitstream services to also offer the resale of broadband offerings.

7.6. Vertical assessment

(788) The Proposed Transaction gives rise to a number of vertically affected markets.

7.6.1. Wholesale market for fixed call termination services – Retail supply of fixed voice services and retail supply of mobile telecommunications services

(789) Both Orange and Jazztel provide call termination services in fixed networks in Spain.

(790) The wholesale market for fixed call termination services where the Parties have a 100% market share in their respective networks ("one net-one market" principle) is upstream of the markets for the retail supply of fixed voice services (Orange 2.5%, Jazztel 5.0%) and of the retail supply of mobile telecommunication services (Orange 20.8%, Jazztel [0-5]*%) as of the first half of 2014.

7.6.1.1. The Notifying Party's view

(791) The Notifying Party submits that any possible competition concerns are excluded from the outset, as the wholesale market for fixed call termination in Spain is subject to regulatory obligations. According to a CNMC regulation of September 2014, all operators including Orange and Jazztel have the following obligations: (1) provide fixed call termination to all operators; (2) provide the service at cost-oriented prices and (3) on a non-discriminatory basis; (4) intervention of the Spanish regulator in case no interconnection agreement can be reached with another operator. The CNMC defined cost-oriented prices following the Bottom-Up Long Run Incremental Cost (BULRIC) standard indicated in the Commission Recommendation on the Regulatory Treatment of Fixed and Mobile Termination Rates in the EU. A maximum termination price applying to all operators has been set at 0.0817 €cent/min which is charged by the second starting from the first second.

7.6.1.2. The Commission's assessment

(792) The Commission notes, first, that no respondent to the Market Investigation raised any issues related to vertical competition concerns arising from the Proposed Transaction on the market for fixed call termination services on the one hand, and the markets for retail supply of fixed voice services and of mobile telecommunications services on the other hand.

(793) Second, the Commission notes that there are detailed regulatory obligations applying to the wholesale fixed call termination market, including an obligation to provide call termination to all operators on a non-discriminatory basis and with a regulated maximum price, subject to intervention by the Spanish regulatory authority.

(794) Third, in any case, the Parties' combined market share on the downstream market for retail supply of fixed voice services remains under 15% (by revenues). In relation to the downstream market for retail supply of mobile telecommunication services, the Commission notes that the Proposed Transaction only brings an insignificant increment (less than 2% by revenues)

(795) In light of the above, the Commission concludes that the Proposed Transaction would not significantly impede effective competition as regards the relation between the market for wholesale fixed call termination services on the one hand, and the markets for retail supply of fixed voice services and retail supply of mobile telecommunication services on the other hand.

7.6.2. Wholesale market for mobile call termination services – Retail supply of fixed voice services and retail supply of mobile telecommunications services

(796) Orange is active on the market for wholesale mobile call termination services on its own network. Jazztel is also active on this market and provides mobile termination services for its own network.

(797) The wholesale market for mobile call termination services where the Parties have a 100% market share in their respective networks ("one net-one market" principle) is upstream of the markets for the retail supply of fixed voice services (Orange 2.5%, Jazztel [5-10]*%) and of the retail supply of mobile telecommunication services (Orange [20-30]*%, Jazztel [0-5]*%) as of the first half of 2014 .

7.6.2.1. Notifying Party's view

(798) The Notifying Party submits that any possible competition concerns are excluded from the outset, as the wholesale market for mobile call termination in Spain is subject to detailed regulatory obligations. According to a regulation adopted in May 2012 , all operators have the following obligations: (1) provide fixed call termination to all operators; (2) provide the service at cost-oriented prices and (3) on a non-discriminatory basis; (4) transparency obligations; and (5) intervention of the Spanish regulator in case no interconnection agreement can be reached with another operator. Moreover, a cost accounting separation obligation is imposed on Telefónica, Vodafone and Orange. As from 1 July 2013, the regulated mobile termination maximum price is set at EUR 0.00109 per minute, applying to all operators and charged by the second starting from the first second.

507 Source: data provided by the Notifying Party on the basis of the CNMC data (http://data.cnmc.es/datagraph/).

508 CMT Decision dated 10.5.2012 in case MTZ 2011/2503 corresponding to market 7 in Commission Recommendation of 17 December 2007 on relevant product and service markets within the electronic communications sector susceptible to ex ante regulation in accordance with Directive 2002/21/EC of the European Parliament and of the Council on a common regulatory framework for electronic communications networks and services.

156

7.6.2.2. Commission's assessment

(799) The Commission notes, first, that none of the respondents to the Market Investigation raised any concerns related to vertical issues arising from the transaction on the market for wholesale mobile call termination services on the one hand, and the retail supply of fixed voice services and retail mobile telecommunication services on the other hand.

(800) Second, the Commission notes that there are regulatory obligations, including price caps, applying to the wholesale mobile call termination market

(801) Third, the Parties' combined market share on the downstream market for retail supply of fixed voice services remains under 10% (by revenues). In relation to the downstream market for retail supply of mobile telecommunication services, the Commission notes that the Proposed Transaction only brings an insignificant increment (less than 2% by revenues).

(802) In light of the analysis above, the Commission concludes that the Proposed Transaction does not significantly impede effective competition on the markets for wholesale mobile call termination services on the one hand, and retail supply of fixed voice services and retail mobile telecommunication services on the other hand.

7.6.3. Wholesale market for access and call origination services on mobile networks – Retail supply of mobile telecommunications services

(803) MNOs supply wholesale access and call origination services which enable MVNOs to provide their own retail mobile services. In Spain there are four MNOs, out of which three are currently providing such services (Vodafone, Telefónica and Orange). Yoigo is not active at present on this market due to the fact that it depends on roaming services provided to it by Telefonica. The market for wholesale access and call origination services where only Orange, and not Jazztel, is active ([20-30]*% in terms of revenues) is upstream of the market for retail supply of mobile telecommunications services where both Parties are active (Orange 20.8%, Jazztel [0-5]*% in revenues).

7.6.3.1. Notifying Party's view

(804) The Notifying Party submits that the Proposed Transaction will not lead to any input or customer foreclosure concerns.

(805) Firstly, the concentration does not create any overlap, as Jazztel is not active in this market. Only Orange, an MNO, provides wholesale access and call origination on mobile networks services.

(806) Secondly, the market share of the only party active in the market, Orange, is below the 25% threshold specified in paragraph 18 of the Commission’s Horizontal Merger Guidelines.

(807) Thirdly, Orange is the smallest provider of the service in the Spanish market. Vodafone and Telefónica have larger market shares in this market.

(808) Fourthly, the size of the market, EUR 433 million, is limited if compared with the retail market for the provision of mobile telecommunication services.

(809) Finally, this is a regulated market and Orange, as well as Vodafone and Telefónica, have the regulatory obligation to provide the service, which is supervised by the Spanish NRA.

157

7.6.3.2. Commission's assessment

(810) The Market Investigation confirmed that Spanish legislation imposes an obligation on Vodafone, Orange and Telefónica to provide wholesale access and call origination services on reasonable terms. The majority of respondents indicated that while entry on the retail mobile market may be difficult, there would be a sufficient number of providers of wholesale access and call origination on mobile networks in Spain post-transaction .

(811) In order to be competitive in the market, most respondents to the Market Investigation consider it essential to be able to offer 4G services. Data usage has an increasing importance in the mobile sector and customers require higher speeds. Respondents submit that if MVNOs do not have access to 4G on reasonable terms they will not be able to compete and may be forced to leave the market. Moreover, half of respondents consider that the incentive of Orange to provide 4G access to MVNOs will decrease after the transaction. Post-transaction, the merged entity would give priority to its own MVNO brands Jazztel and Simyo. One respondent submits that Orange would countenance a profit loss for not providing 4G services to its MVNOs as it would gain profits from expanding sales through Jazztel, by raising prices to consumers and by expelling MVNOs from the market (who no longer can compete without 4G services).

(812) The Market Investigation thus showed that MVNOs currently experience difficulty in contracting 4G services at reasonable prices from MNOs. The Commission notes that the relevant question as regards the assessment of possible input foreclosure by the merged entity is whether Orange’s ability and incentives to provide this service change after the transaction.

(813) With respect to ability, the Commission notes, first, that Orange’s position on the wholesale market for access and call origination service on mobile networks does not change. The transaction does not reduce the number of MNOs providing wholesale access to their networks as Jazztel is not present on the upstream market.

(814) Second, Orange’s market share in the upstream market is below 25%, which typically does not confer the necessary market power to engage in input foreclosure. Two alternative providers, Telefónica and Vodafone, are present on the upstream market. The Commission notes that at least one MVNO, Pepephone, was able to conclude an agreement with an MNO (notably with Telefonica) for the provision of 4G services. Therefore, Orange on its own does not have the ability to foreclose MVNOs from the market.

(815) Third, in 2006, the Spanish telecommunication regulator imposed on Telefonica, Vodafone and Orange obligations to meet all reasonable requests for access to their network and to provide services at reasonable prices. According to MVNOs responding to the Market Investigation, it appears unclear to what extent these provisions are enforceable with respect to 4G services.

509 See replies to Commission Phase I Questionnaire to Competitors Q1 of 17 October 2014, question 15.3.

510 See replies to Commission Phase I Questionnaire to Competitors Q1 of 17 October 2014, question 28.

511 See replies to Commission Phase I Questionnaire to Competitors Q1 of 17 October 2014, question 17.

512 See replies to Commission Phase II Questionnaire to Competitors of 23 December 2014, question 78.

513 Resolución de la CMT de 2 de febrero de 2006, por la que se aprueba la definición y análisis del mercado de acceso y originación de llamadas en las redes públicas de telefonía móvil, la designación de los operadores con poder significativo de mercado y la imposición de obligaciones específicas, y se acuerda su notificación a la Comisión Europea, expe. AEM 2005/933.

158

(816) One competitor submits that without the initiated merger, Jazztel would have initiated a regulatory conflict against Orange before the CNMC. According to that respondent, without the merger, Jazztel would have continued to exercise pressure on Orange to make it comply with its legal and regulatory obligation to provide mobile wholesale access in 4G and would have supported the demands for 4G access of the rest of the MVNOs. The Commission notes that if this is correct, other MVNOs than Jazztel would have the possibility to initiate a regulatory conflict against Orange to oblige it to grant access to 4G services.

(817) With respect to incentives, the Commission considers, first, that Orange’s incentive to provide wholesale 4G services to MVNOs does not significantly change post-transaction. Orange is already present on the downstream market for retail mobile telecommunications services and the increment in the downstream market for retail mobile telecommunication services brought about by the acquisition of Jazztel is small ([0-5]*% in terms of revenues).

(818) Second, the Commission notes that pre-transaction, Orange had not concluded an agreement to provide 4G services with any of the MVNOs it hosts except with Jazztel. Therefore, Orange does not appear to have had the incentive to conclude such agreements with MVNOs. Therefore, if Orange were to continue post-transaction to refuse to provide 4G wholesale services, its incentives would not have changed post-merger and therefore would not be merger-specific.

(819) In light of the analysis carried in recitals (813) to (818), the Commission concludes that the Proposed Transaction would not significantly impede effective competition on the markets for wholesale access and call origination on the one hand, and retail supply of mobile telecommunications on the other hand.

8. COMMITMENTS

8.1. Procedure

(820) In order to address the Commission's serious doubts raised during the Phase I investigation, the Notifying Party submitted commitments on 13 November 2014 and revised commitments on 19 November 2014 pursuant to Article 6(2) of the Merger Regulation consisting of several different wholesale access offers.

(821) The Commission considered that these commitment packages did not address in full and in a clear-cut fashion the competition concerns raised during the Phase I investigation and that they therefore did not meet the standard for an acceptable remedy in Phase I. As a result, the Commission decided not to market test the proposed set of commitments.

(822) Upon completion of the Phase II Market Investigation, on 25 February 2015 the Commission addressed the Statement of Objections to the Notifying Party. The Commission further substantiated its competition concerns in the Letter of Facts.

(823) To address the Commission's concerns raised in the Statement of Objections, the Notifying Party submitted a commitments package on 6 March 2015 (the "First Commitments").

(824) The Commission launched a market test of the First Commitments on 13 March 2015 (the "First Market Test"). Questionnaires were sent to competitors of the Parties in Spain and outside Spain. The CNMC also submitted its comments on the First Commitments on 19 March 2015.

(825) Following the First Market Test, the Notifying Party submitted a new set of commitments on 29 March 2015 ("the Second Commitments") and a revised set of commitments on 6 April 2015 ("the Third Commitments", together with the Second Commitments referred to as the "Modified Commitments").

159

(829) Under the Merger Regulation, it is the responsibility of the Commission to show that a concentration would significantly impede effective competition. The Commission then communicates its competition concerns to the parties to allow them to formulate appropriate and corresponding remedies proposals. It is then for the parties to the concentration to propose commitments that would be suitable to address entirely such competition concerns. The Commission only has 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 of 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. The commitments must also be proportionate to the competition concerns identified .

(830) In assessing whether the proposed commitments will likely eliminate the competition concerns identified, the Commission considers all relevant factors including inter alia the type, scale and scope of the proposed commitments, judged by reference to the structure and particular characteristics of the market in which the competition concerns arise, including the position of the parties and other participants on the market .

(831) In order for the commitments to comply with these principles, commitments must be capable of being implemented effectively within a short period of time. Where, however, the parties submit remedies proposals that are so extensive and complex

514 Commission notice on remedies acceptable under Council Regulation (EC) No 139/2004 and under Commission Regulation (EC) No 802/2004 (the "Remedies Notice"), OJ 2008/C 267/01, paragraph 5.

515 See Case T-209/01 Honeywell International, Inc. v. Commission [2005] II-5527, paragraph 99.

516 Remedies Notice, paragraph 6.

517 Remedies Notice, paragraph 9.

518 See also Case C-202/06 P Cementbouw Handel & Industrie v Commission [2007] ECR 2007 I-12129, paragraph 54.

519 Remedies Notice, paragraphs 9 and 61.

520 recital 30 of the Merger Regulation. The General Court set out the requirements of proportionality as follows: "the principle of proportionality requires measures adopted by Community institutions not to exceed the limits of what is appropriate and necessary in order to attain the objectives pursued; when there is a choice between several appropriate measures recourse must be had to the least onerous, and the disadvantages caused must not be disproportionate to the aims pursued" (T-177/04 easyJet v Commission [2006] ECR II-1931, Paragraph 133).

521 Remedies Notice, paragraph 12.

522 Remedies Notice, paragraph 9.

(832) As concerns the form of acceptable commitments, the Merger Regulation leaves discretion to the Commission as long as the commitments meet the requisite standard. Structural commitments will meet the conditions set out above only in so far as the Commission is able to conclude with the requisite degree of certainty that it will be possible to implement them and that it will be likely that the new commercial structures resulting from them will be sufficiently workable and lasting to ensure that the significant impediment to effective competition will not materialise .

(833) Divestiture commitments are generally the best way to eliminate competition concerns resulting from horizontal overlaps, although other structural commitments, such as access remedies, may be suitable to resolve concerns if those remedies are equivalent to divestitures in their effects .

(834) It is against this background that the Commission analysed the proposed Commitments in this case.

8.3. Assessment of the Commitments

8.3.1. First Commitments

(835) To address the Commission's competition concerns raised in the Statement of Objections, the Notifying Party submitted the First Commitments on 6 March 2015.

8.3.1.1. Description

(836) The First Commitments comprise two elements: (i) the divestment of a FTTH network located in some of the cities where Jazztel's and Orange's networks currently overlap (the "Divested FTTH Network"); and (ii) the wholesale access to Jazztel's DSL network ("Wholesale ADSL Bitstream Access").

Divested FTTH network

(837) Orange commits to divest an FTTH network that covers [500-600 000]* BUs. These BUs are located on the current fibre network of Jazztel at 11 local exchanges located in the four cities of Barcelona, Madrid, Sevilla and Valladolid.

(838) According to Orange, the Divested FTTH Network constitutes an independent, coherent network at the level of the cables (which bundle many fibre lines). It would also cover BUs located on parts of Jazztel's non-overlapping fibre network in exchange for BUs located in the other […]* local exchange areas, where Orange's and Jazztel's networks currently overlap. Therefore, the divested network infrastructure would be independent from Orange.

(839) As the Divested FTTH Network includes parts of Jazztel's fibre network that are non-overlapping, the First Commitments provide for an Indefeasible Right of Use ("IRU") to the benefit of Orange for 35 years (renewable for another 35 years) for a "limited capacity" on the Divested FTTH Network. In return, Orange commits to pay

523 Remedies Notice, paragraphs 13, 14 and 61 et seq.

524 Case T-177/04 easyJet v Commission [2006] ECR II-1913, paragraph 197.

525 Remedies Notice, paragraph 10.

526 Remedies Notice, paragraph 19.

(840) To allow the purchaser to build up the necessary capability to manage the Divested FTTH Network, during a minimum of 12 months, renewable at the purchaser's request (the "Transition Period"), Orange will operate the Divested FTTH Network on behalf of the purchaser.

Wholesale ADSL Bitstream Access

(841) The structural divestment of the Divested FTTH Network is complemented by granting the purchaser wholesale access to the merged entity's ADSL network. The Notifying Party commits to provide a national bitstream service to its LLU network, with interconnection in one single point of presence and for a duration of seven years. This wholesale access would be constructed using as an input Telefónica's regulated physical LLU offer, thus providing access to either Orange or Jazztel's ADSL network ([…]* local exchanges, covering 78% of the Spanish territory). The purchaser could choose among three different pricing structures.

(842) The first pricing structure comprises two elements: (i) a fixed fee of EUR […]* which would pay for a block of […]* lines usable over seven years, and (ii) an additional monthly fee of EUR […]* per line. In addition, a number of non-recurrent fees are payable, such as a one-off set up fee per line of EUR […]*.

(843) The second option does not have any fixed fee but a higher monthly fee of EUR […]* per line. The set up cost for a line is EUR […]*. There is a minimum of […]* lines to be activated during the first 12 months of the contract.

(844) The third option is modeled on the current prices of the regulated Telefónica offer called NEBA (copper), but […]*.

(845) For both the first and the second option, the average peak hour subscriber throughput is limited to 375 kb/s per customer. In the event of traffic in excess of the threshold of 375 kb/s per customer, an additional EUR […]* per Mb/s per customer per month would be payable.

(846) For all options, other non-recurrent fees apply for disconnecting a line, incident management, on-site support, equipment, etc.

8.3.1.2. Results of the First Market Test

Divested FTTH network

(847) Respondents to the First Market Test provided mixed results as regards the scope of the Divested FTTH Network. Half of the respondents to the First Market Test indicated that the geographic scope of the Divested FTTH Network would be enough to compete on the national Spanish market for fixed Internet access services. A small minority indicated that such geographic scope would be sufficient only to be active on part of the Spanish retail markets involving fixed Internet access services, whereas the remaining respondents to the market test indicated that the geographic scope of the Divested FTTH Network would not be of a sufficient size to compete on the Spanish market .

527 See replies to the Commission's First Market Test, question 30.

(848) Respondents to the First Market Test indicated that the cities where the Divested FTTH Network is located are generally all commercially attractive areas .

528 See replies to the Commission's First Market Test, question 31.

(849) A number of respondents indicated that the scope of the Divested FTTH Network would only partially cover the four cities included in the First Commitments, and that the purchaser would need to invest further to acquire sufficient coverage. As regards the parts of the four cities not covered by the Divested FTTH Network, the majority of the respondents to the market test indicated that it would be difficult for the purchaser to offer fixed Internet access services with NGA technology to those parts. In particular, a respondent stressed that, in order to achieve significant coverage in those parts, access to the merged entity’s FTTH network would be needed .

529 See replies of Pepephone and Euskaltel to the Commission's First Market Test, question 30.

(850) When asked whether the purchaser will have the ability and the incentive to roll out additional fibre, respondents provided mixed replies. Some respondents to the market test affirmed that the purchaser would need time to acquire a sufficient customer base for it to be able to extend its FTTH network. As a consequence, wholesale access to the merged entity’s FTTH network would be needed. A majority of the respondents to the market test stated that, for Orange to remove completely the competition concerns identified in the Statement of Objections, the commitments package should include a wholesale access to the merged entity’s FTTH network .

530 See replies to the Commission's First Market Test, question 33.

(851) As explained in recital (839), Orange will be granted an IRU over a limited capacity of the Divested FTTH Network, in exchange for a one-time fee and a recurrent fee. A large majority of the respondents to the First Market Test affirmed that, despite the existence of such IRU, the Divested FTTH Network can be considered a viable and independent business. Respondents generally explained that IRUs are industry standard practice. However, as regard the exact definition of the IRU, the large majority of the respondents to the market test stressed that the formulation of the "limited capacity" needed clarification .

531 See replies to the Commission's First Market Test, question 48.

(852) Regarding the possibility for the purchaser to require Orange to provide maintenance services for the operation of the Divested FTTH Network, the large majority of the respondents to the First Market Test generally acknowledged the need of such maintenance services for the purchaser to be able to acquire the necessary knowledge during the Transition Period. The majority of the respondents indicated that a sufficient duration for such maintenance services would be 12 months .

532 See replies to the Commission's First Market Test, questions 51 and 52.

(853) Likewise, the large majority of the respondents to the market test affirmed that it may be necessary for Orange to operate the Divested FTTH Network on behalf of the purchaser during the Transition Period. Respondents generally explained that the Transition Period would ensure continuity of the service, and it would allow the purchaser to acquire the necessary knowledge to operate the network by itself. Moreover, all respondents expressed the view that a duration of 12 months for the Transition Period is an adequate timeframe .

533 See replies to the Commission's First Market Test, question 53.

(854) A majority of the respondents to the market test affirmed that the Divested FTTH Network does not need to include other network layers, such as the Metropolitan Area Network for it to be considered a viable business .

(855) In its comments submitted on 19 March, the CNMC comments that the proposed remedies would be insufficient to address the competition concerns identified in the Statement of Objections and to allow the effective entry of a fourth operator. As regards the FTTH network, while such a divestment has positive aspects, its scope is too small and geographically dispersed. Given the short time that a purchaser has to acquire clients on the DSL network and considering the terms of the Wholesale ADSL Bitstream Access, the new entrant may be unlikely to acquire a critical mass of clients, to be able to profitably expand its FTTH footprint.

Wholesale ADSL Bitstream Access

(856) The Wholesale ADSL Bitstream Access will be provided to the purchaser in one national interconnection point. As regards the underlying technology for such interconnection point, the majority of the respondents affirmed that an IP-based interface would be technically performing and future-proof . A respondent indicated, however, that Ethernet technology would represent a newer alternative .

(857) The large majority of the respondents to the market test also explained that a web-based management service is industry standard .

(858) As regards the technical capabilities of the Wholesale ADSL Bitstream Access, the majority of the respondents affirmed that it is technically possible for the purchaser to offer fixed voice services via an ADSL bitstream offer, both as a standalone service or in bundle with fixed Internet access services, including by relying on Voice over IP ("VoIP") technology. Respondents generally proposed to include Quality of Service ("QoS”) levels similar to those offered by Telefónica for its regulated wholesale access offer NEBA. Another respondent explained that QoS should include agreed service levels on latency, packet loss and availability .

(859) The large majority of the respondents to the market test stressed that the Wholesale ADSL Bitstream Access should contain agreed terms as regards QoS and guaranteed minimum download/upload speed .

(860) Respondents' views were inconclusive as regards the technical possibility for the purchaser to offer Pay-TV services with the Wholesale ADSL Bitstream Access offer, both as a standalone service or bundled with fixed Internet access services and fixed voice services. The possibility to offer Pay-TV services depends mainly on

539 See replies to the Commission's First Market Test, question 54.

540 See replies to the Commission's First Market Test, question 45.

541 See replies to the Commission's First Market Test, question 6.

542 See reply of Telefónica to the Commission’s First Market Test, question 6.

543 See replies to the Commission's First Market Test, question 7.

544 See replies to the Commission's First Market Test, questions 10.1 and 10.3.

545 See reply of BT to the Commission’s First Market Test, question 12.1.

546 See replies to the Commission's First Market Test, question 12.1 and 12.2.

547 See replies to the Commission's First Market Test, questions 10.2 and 10.4, three respondents consider it possible to offer Pay-TV services, three disagree and four do not reply. Masmovil considers it possible to offer standard definition services but not HDTV.

(861) In its Wholesale ADSL Bitstream Access offer, Orange considers that, once an average peak hour throughput of 375 kb/s per subscriber is exceeded, an additional traffic charge will be charged to the purchaser. The majority of the respondents to the market test explained that a throughput of 375 kb/s is not generally sufficient to provide fixed Internet access services, in particular in the coming years when Internet traffic is expected to grow. Moreover, already today, the provision of IPTV services in standard- or high-quality or a combination of fixed Internet access services and IPTV could not be provided without exceeding the traffic cap .

(862) With regard to the duration of the Wholesale ADSL Bitstream Access offer, the majority of the respondents to the market test affirmed that 6 months after the adoption of the Commission's decision is an adequate timeframe for the purchaser to start providing fixed Internet access services via the Wholesale ADSL Bitstream Access offer, although a large minority of the respondents indicated that a shorter timeframe would be sufficient. In particular, the majority of the respondents to the market test indicated that a period shorter than 6 months would be enough to allow the purchaser to make all the necessary preparations for it to meaningfully benefit from the Wholesale ADSL Bitstream Access. Likewise, a shorter period would also be sufficient for Orange to provide the purchaser with the Wholesale ADSL Bitstream Access, in particular because both Parties would be already today providing similar services to third parties .

553 See replies to the Commission's First Market Test, question 16.

(863) Finally, the large majority of the respondents to the market test indicated that a duration of 7 years for the Wholesale ADSL Bitstream Access would be sufficient for the purchaser to establish itself as a fixed network operator in Spain .

(864) As to the pricing of the first and the second options, respondents to the First Market Test expressing a view generally indicated that monthly recurrent per line prices should be lowered to a level close to the recurrent monthly fee of OBA regulated access, with which both Orange and Jazztel provide today most of their fixed Internet access services .

554 See replies to the Commission's First Market Test, questions 20 and 24.

(865) The Wholesale ADSL Bitstream Access is also offered under a third pricing option. According to the third pricing option, the purchaser will be offered the same prices as Telefónica’s regulated wholesale broadband offer NEBA, but […]*. According to the majority of the respondents to the market test, […]* would not be enough for the purchaser to compete effectively on the market. In particular, some respondents indicated that such a pricing option, which provides for a monthly fee and a recurrent fee based on the effective traffic consumption, would be insufficient to allow the purchaser to compete against retail offers that do not provide for traffic limitations .

556 See in particular the replies of Yoigo and Aire Network to the Commission’s First Market Test, question 26.

(866) The CNMC considers, first, that DSL technology will be competitive in the short term (probably the next two years). However, in the mid and long term, DSL will be insufficient to compete with the other three competitors who likely have NGA networks in place in the areas that are covered by the DSL bitstream access offer. Second, the minimum capacity the purchaser is required to purchase is very limited. Third, variable cost per subscriber is significant. Fourth, the average peak hour throughput per customer of 375 kbps will effectively hamper the purchasers' ability to compete in the quadruple-play segment (to the extent that the ADSL line allows offering TV).

Lack of access to mobile services

(867) When asked about the criteria necessary for the purchaser to be able to preserve competition in the retail market involving fixed Internet access services, respondents to the market test provided mixed views. Generally, respondents indicated that a sufficiently large client base, either in fixed telecommunications or in mobile telecommunications services, would be needed. Other respondents stressed the importance for the purchaser to be able to offer mobile telecommunications services . Several respondents indicated that wholesale access to the mobile network of Orange would need to be included in the commitments package .

(868) In its comments of 19 March, the CNMC also considers that the remedies lack a wholesale mobile offer including 4G services. Therefore, a new entrant would not be able to compete in the possible retail market for multiple play services except if the purchaser has its own mobile network with a considerable coverage in Spain.

Respondents’ overall assessment of the commitments package

(869) The respondents to the First Market Test provided mixed replies as to the ability of the First Commitments to remove the competition concerns set out in the Statement of Objections. In particular, respondents pointed to the lack of a wholesale access to Orange’s mobile network and the lack of access to the merged entity’s FTTH network as factors limiting the effectiveness of the remedy package.

8.3.1.3. Commission's assessment of the First Commitments

Divested FTTH Network

(870) The Commission notes that the Divested FTTH Network includes the current fibre network […]* at 11 local exchanges in the four cities of Barcelona, Madrid, Sevilla and Valladolid and covers a total amount of [500-600 000]* BUs. While this is a coherent network – involving the divestment of parts of Jazztel's non-overlapping network – the Divested FTTH Network does not cover the entire existing overlap between Orange and Jazztel. According to Orange, the overlap between the FTTH networks of Orange and Jazztel amounted to [500-600 000]* BUs by 12 February 2015. Moreover, these BUs are located in […]* different local exchanges that are located (in addition to the four aforementioned cities) in particular in the major Spanish cities of Valencia and Malaga.

562 See replies to the Commission's First Market Test, question 26.

(871) Taking into account that the size of the Divested FTTH Network does not reach the size of the current overlap of the Parties’ FTTH networks (measured in BUs) and that it does not reflect its geographic distribution, the Commission considers that it falls short of what is required to address the identified competition concerns.

Wholesale ADSL Bitstream Access

(872) As regards the Wholesale ADSL Bitstream Access, the Commission considers that, in the present case, in order to be an acceptable component of the remedies and for the reasons explained in recitals (929) and (929) below, the wholesale agreement needs to replicate as much as possible the current pricing structure of the Parties so as to create an incentive for the purchaser to behave in a similar manner in the market as the Parties. This means that the Notifying Party would need to commit to give access to a fixed capacity for a price that has the highest possible fixed component. The fixed capacity should be sufficiently large to achieve a customer base of a size similar to that of Jazztel's customer base, and may be unlimited. In other words, the variable parts of the price stipulated in the Wholesale ADSL Bitstream Access agreement should match the Parties incremental costs. The actual price of the "capacity" acquired by the purchaser (which corresponds to the fixed component of the price) would be a matter of commercial negotiations between Orange and the purchaser.

(873) The three proposed pricing options fall short of allowing the purchaser to replicate the incentives and ability of Jazztel to compete with its LLU-based network. First, options 2 and 3 are essentially pay-as-you-go pricing options, as for each additional subscriber a relatively high recurrent price has to be paid, with no or minimal upfront capacity commitment. Therefore, the purchaser would not have incentives similar to Jazztel or Orange who have invested upfront into their LLU-networks and price aggressively in order to recuperate their cost. While option 1 includes an upfront payment for […]* subscribers, such commitment is too small to ensure that the purchaser comes close to replicating Jazztel's current role (more than […]* gross adds in 2014) that is lost as a result of the merger in the retail markets involving fixed Internet access services .

(874) Second, the monthly payable fees per subscriber combined with charges for Internet traffic under the three options are above Orange's variable cost of providing the service. The cost structure of Jazztel and Orange based on the regulated offer to access subscriber loop "Oferta de acceso al Bucle de Abonado" ("OBA") implies incremental costs of EUR 8.60 per subscriber (the regulated OBA price payable to Telefonica), plus a minor cost payable to Telefonica for energy and cabling in the local exchange, as well as other cost such as the traffic transport through different network layers (the backhauling network, the aggregation network and the backbone network). For the purchaser to be able to compete as aggressively as Jazztel or Orange currently do, its variable (recurrent) cost should be aligned with Orange’s incremental cost for providing the service. Any incremental wholesale price

562 Moreover, the upfront fee payable by the purchaser (EUR 1.54 million for a 20K subscriber package) should be a matter of negotiations with the purchaser and should not be fixed in the Commitments text.

(865) The Wholesale ADSL Bitstream Access is also offered under a third pricing option. According to the third pricing option, the purchaser will be offered the same prices as Telefónica’s regulated wholesale broadband offer NEBA, but […]*. According to the majority of the respondents to the market test, […]* would not be enough for the purchaser to compete effectively on the market. In particular, some respondents indicated that such a pricing option, which provides for a monthly fee and a recurrent fee based on the effective traffic consumption, would be insufficient to allow the purchaser to compete against retail offers that do not provide for traffic limitations .

556 See in particular the replies of Yoigo and Aire Network to the Commission’s First Market Test, question 26.

(866) The CNMC considers, first, that DSL technology will be competitive in the short term (probably the next two years). However, in the mid and long term, DSL will be insufficient to compete with the other three competitors who likely have NGA networks in place in the areas that are covered by the DSL bitstream access offer. Second, the minimum capacity the purchaser is required to purchase is very limited. Third, variable cost per subscriber is significant. Fourth, the average peak hour throughput per customer of 375 kbps will effectively hamper the purchasers' ability to compete in the quadruple-play segment (to the extent that the ADSL line allows offering TV).

Lack of access to mobile services

(867) When asked about the criteria necessary for the purchaser to be able to preserve competition in the retail market involving fixed Internet access services, respondents to the market test provided mixed views. Generally, respondents indicated that a sufficiently large client base, either in fixed telecommunications or in mobile telecommunications services, would be needed. Other respondents stressed the importance for the purchaser to be able to offer mobile telecommunications services . Several respondents indicated that wholesale access to the mobile network of Orange would need to be included in the commitments package .

(868) In its comments of 19 March, the CNMC also considers that the remedies lack a wholesale mobile offer including 4G services. Therefore, a new entrant would not be able to compete in the possible retail market for multiple play services except if the purchaser has its own mobile network with a considerable coverage in Spain.

Respondents’ overall assessment of the commitments package

(869) The respondents to the First Market Test provided mixed replies as to the ability of the First Commitments to remove the competition concerns set out in the Statement of Objections. In particular, respondents pointed to the lack of a wholesale access to Orange’s mobile network and the lack of access to the merged entity’s FTTH network as factors limiting the effectiveness of the remedy package.

8.3.1.3. Commission's assessment of the First Commitments

Divested FTTH Network

(870) The Commission notes that the Divested FTTH Network includes the current fibre network […]* at 11 local exchanges in the four cities of Barcelona, Madrid, Sevilla and Valladolid and covers a total amount of [500-600 000]* BUs. While this is a coherent network – involving the divestment of parts of Jazztel's non-overlapping network – the Divested FTTH Network does not cover the entire existing overlap between Orange and Jazztel. According to Orange, the overlap between the FTTH networks of Orange and Jazztel amounted to [500-600 000]* BUs by 12 February 2015. Moreover, these BUs are located in […]* different local exchanges that are located (in addition to the four aforementioned cities) in particular in the major Spanish cities of Valencia and Malaga.

562 See replies to the Commission's First Market Test, question 26.

(871) Taking into account that the size of the Divested FTTH Network does not reach the size of the current overlap of the Parties’ FTTH networks (measured in BUs) and that it does not reflect its geographic distribution, the Commission considers that it falls short of what is required to address the identified competition concerns.

Wholesale ADSL Bitstream Access

(872) As regards the Wholesale ADSL Bitstream Access, the Commission considers that, in the present case, in order to be an acceptable component of the remedies and for the reasons explained in recitals (929) and (929) below, the wholesale agreement needs to replicate as much as possible the current pricing structure of the Parties so as to create an incentive for the purchaser to behave in a similar manner in the market as the Parties. This means that the Notifying Party would need to commit to give access to a fixed capacity for a price that has the highest possible fixed component. The fixed capacity should be sufficiently large to achieve a customer base of a size similar to that of Jazztel's customer base, and may be unlimited. In other words, the variable parts of the price stipulated in the Wholesale ADSL Bitstream Access agreement should match the Parties incremental costs. The actual price of the "capacity" acquired by the purchaser (which corresponds to the fixed component of the price) would be a matter of commercial negotiations between Orange and the purchaser.

(873) The three proposed pricing options fall short of allowing the purchaser to replicate the incentives and ability of Jazztel to compete with its LLU-based network. First, options 2 and 3 are essentially pay-as-you-go pricing options, as for each additional subscriber a relatively high recurrent price has to be paid, with no or minimal upfront capacity commitment. Therefore, the purchaser would not have incentives similar to Jazztel or Orange who have invested upfront into their LLU-networks and price aggressively in order to recuperate their cost. While option 1 includes an upfront payment for […]* subscribers, such commitment is too small to ensure that the purchaser comes close to replicating Jazztel's current role (more than […]* gross adds in 2014) that is lost as a result of the merger in the retail markets involving fixed Internet access services .

(896) As regards specifically the excess traffic fee, the majority of the respondents indicated that the current level of EUR […]* per kb/s would affect the future viability and/or competitiveness of the purchaser's ADSL offer. In this regard, a respondent expressed the concern that the purchaser will not be able to provide IPTV services without incurring the additional traffic fee, and that, in the next two years, even the simple provision of fixed Internet access services would likely exceed the cap of 375 kb/s, triggering significant additional monthly charges per subscriber .

(897) As regards the Notifying Party's proposed indexation mechanism of the excess traffic fee, respondents provided mixed views. While some respondents welcomed the indexation, others indicated that such a mechanism, based on the review of Telefónica's regulated prices for wholesale leased lines, would not be effective since the revision of Telefónica's regulated prices is undertaken rather infrequently (approximately every three years). The majority of respondents indicated that the cost of maintaining, upgrading and extending the backhaul, the backbone and the aggregation network, which is the infrastructure whose maintenance justifies the excess traffic fee, would decrease or in any case not increase in the future .

(898) As regards the technical possibility for the purchaser to offer VoIP services by relying on the Wholesale ADSL Bitstream Access offer, the majority of the respondents stressed the importance of including agreed Quality of Service levels in the commitments package. In particular, a respondent indicated that agreed QoS level should encompass jitter, latency and packet loss .

572 See replies to the Commission's Second Market Test, question 1.

573 See replies of Yoigo, R cable and Aire Network to the Commission’s Second Market Test, question 1.

574 See replies to the Commission's Second Market Test, question 3.

575 See reply of Yoigo to the Commission’s Second Market Test, question 3.

576 See replies to the Commission's Second Market Test, question 3.1.2.

577 See in particular replies of R Cable and Yoigo to the Commission’s Second Market Test, question 3.1.2.

578 See replies to the Commission's Second Market Test, question 2.

579 See replies to the Commission's Second Market Test, question 4.

580 See reply of Yoigo to the Commission’s Second Market Test, question 5.

(899) Finally, a respondent expressed the opinion that a second point of interconnection would be needed as a back-up .