EUROPEAN COMMISSION DG Competition

(Only the English text is authentic)

MERGER PROCEDURE

REGULATION (EC) 139/2004

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

Date: 15/01/2016

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.

EN EN

EUROPEAN COMMISSION

Brussels, 15.1.2016 C(2016) 103 final

Public Version

COMMISSION DECISION

of 15.1.2016

(Only the English text is authentic)

TABLE OF CONTENTS

1. The Parties.................................................................................................................... 1

2. The operation and the concentration ............................................................................ 2

3. Union Dimension ......................................................................................................... 2

4. Procedure...................................................................................................................... 2

5. Introduction to the Parties' activities ............................................................................ 4

5.1. Introduction to the beverage can sector ....................................................................... 4

5.1.1. Description of the beverage can value chain and production process ......................... 4

5.1.2. Demand for beverage cans in the EEA ........................................................................ 6

5.1.3. Beverage can manufacturers in the EEA ..................................................................... 8

5.1.3.1. Location of beverage can plants serving the EEA ..................................................... 10

5.1.3.2. Location of can end plants ......................................................................................... 10

5.1.3.3. Location of research and development activities of the Parties ................................. 10

5.1.4. The pre-Transaction expansion plans of the Parties .................................................. 11

5.1.4.1. Ball in […] ................................................................................................................. 11

5.1.4.2. Rexam in […]............................................................................................................. 11

5.1.4.3. Rexam in […]............................................................................................................. 14

5.1.5. Changes in the beverage cans sector in the EEA since 1999 ..................................... 14

5.1.5.1. Changes relating to Rexam ........................................................................................ 14

5.1.5.2. Changes relating to Ball ............................................................................................. 15

5.1.5.3. Changes relating to Can-Pack .................................................................................... 16

5.1.5.4. Changes relating to Crown ......................................................................................... 16

5.1.5.5. Changes relating to other beverage can manufacturers.............................................. 16

5.2. Introduction to the aluminium bottles sector ............................................................. 16

6. Relevant product market ............................................................................................ 17

6.1. Beverage cans............................................................................................................. 17

6.1.1. Beverage cans and other forms of beverage packaging solutions.............................. 17

6.1.1.1. Past decisional practice .............................................................................................. 17

6.1.1.2. The Notifying Party's views ....................................................................................... 17

6.1.1.3. The Commission's assessment ................................................................................... 17

6.1.2. Various sizes and types of cans .................................................................................. 19

6.1.2.1. Past decisional practice .............................................................................................. 19

6.1.2.2. The Notifying Party's views ....................................................................................... 19

6.1.2.3. The Commission's assessment ................................................................................... 19

6.1.3. Can ends and can bodies for beverage cans ............................................................... 21

6.1.3.1. Past decisional practice .............................................................................................. 21

6.1.3.2. The Notifying Party's views ....................................................................................... 21

6.1.3.3. The Commission's assessment ................................................................................... 21

6.1.4. Aluminium and steel cans .......................................................................................... 22

6.1.4.1. Past decisional practice .............................................................................................. 22

6.1.4.2. The Notifying Party's views ....................................................................................... 22

6.1.4.3. The Commission's assessment ................................................................................... 22

6.2. Aluminium bottles ...................................................................................................... 24

6.2.1. Past decisional practice .............................................................................................. 24

6.2.2. The Notifying Party's views ....................................................................................... 24

6.2.3. The Commission's assessment ................................................................................... 25

7. Relevant geographic market ....................................................................................... 26

7.1. Beverage cans............................................................................................................. 26

7.1.1. Past decisional practice .............................................................................................. 26

7.1.2. The Notifying Party's views ....................................................................................... 27

7.1.3. The Commission's assessment ................................................................................... 29

7.1.3.1. Lack of national barriers ............................................................................................ 29

7.1.3.2. Pricing occurs on a customer site-by-site basis.......................................................... 30

7.1.3.3. Security of supply, rapid and just-in-time deliveries, good service levels................. 32

7.1.3.4. Transport costs ........................................................................................................... 33

7.1.3.5. Regional clusters of 700 km catchment areas around individual customer filling locations ..................................................................................................................... 34

7.2. Aluminium bottles ...................................................................................................... 37

7.2.1. Past decisional practice .............................................................................................. 37

7.2.2. The Notifying Party's views ....................................................................................... 37

7.2.3. The Commission's assessment ................................................................................... 37

8. Competitive assessment: Principles ........................................................................... 38

8.1. Non-coordinated effects ............................................................................................. 38

8.2. Coordinated effects .................................................................................................... 41

9. Competitive Assessment Beverage Cans: Non-Coordinated Effects......................... 43

9.1. Overview of the market dynamics and effects of the Transaction across the regional clusters........................................................................................................................ 43

9.1.1. The position of the Parties across the EEA ................................................................ 44

9.1.1.1. The Parties are the two largest beverage can manufacturers in the EEA................... 44

9.1.1.2. The merged entity would be the largest player in terms of sales volumes in the EEA .................................................................................................................................... 44

9.1.1.3. The merged entity would be the largest player in terms of total capacity in the EEA .................................................................................................................................... 45

9.1.1.4. The Transaction would lead to a further increase in concentration levels across the EEA ............................................................................................................................ 45

9.1.2. Intensity of competition pre-Transaction ................................................................... 46

9.1.2.1. The Notifying Party's views ....................................................................................... 46

9.1.2.2. The Commission's assessment ................................................................................... 46

9.1.3. The Transaction reduces the number of players from 4 to 3 in the EEA ................... 49

9.1.3.1. The Notifying Party's views ....................................................................................... 49

9.1.3.2. The Commission's assessment ................................................................................... 49

9.1.4. The Parties are close competitors ............................................................................... 50

9.1.4.1. The Notifying Party's views ....................................................................................... 50

9.1.4.2. The Commission's assessment ................................................................................... 50

9.1.5. The ability of the four beverage can manufacturers to compete on an equal footing 53

9.1.5.1. The Notifying Party's views ....................................................................................... 53

9.1.5.2. The Commission's assessment ................................................................................... 53

9.1.6. Large customers versus other customers.................................................................... 59

9.1.6.1. The Notifying Party's views ....................................................................................... 60

9.1.6.2. The Commission's assessment ................................................................................... 60

9.1.7. Importance of having a network of plants.................................................................. 63

9.1.7.1. The Notifying Party's views ....................................................................................... 63

9.1.7.2. The Commission's assessment ................................................................................... 64

9.1.8. The Parties are the main innovators ........................................................................... 67

9.1.8.1. The Notifying Party's views ....................................................................................... 67

9.1.8.2. The Commission's assessment ................................................................................... 67

9.1.9. Capacity is tight in the EEA ....................................................................................... 68

9.1.9.1. The Notifying Party's views ....................................................................................... 68

9.1.9.2. The Commission's' assessment................................................................................... 69

9.1.10. Switching possibilities for customers in the EEA are limited.................................... 74

9.1.10.1. The Notifying Party's views ....................................................................................... 74

9.1.10.2. The Commission's assessment ................................................................................... 74

9.1.11. Multi-sourcing ............................................................................................................ 76

9.1.11.1. The Notifying Party's views ....................................................................................... 76

9.1.11.2. The Commission's assessment ................................................................................... 76

9.1.12. Sales to wall-to-wall customers ................................................................................. 77

9.1.12.1. The Notifying Party's views ....................................................................................... 77

9.1.12.2. The Commission's assessment ................................................................................... 77

9.1.13. Sales to Red Bull ........................................................................................................ 78

9.1.13.1. The Notifying Party's views ....................................................................................... 78

9.1.13.2. The Commission's assessment ................................................................................... 79

9.1.14. There is limited competitive pressure from other forms of packaging ...................... 80

9.1.14.1. The Notifying Party's views ....................................................................................... 80

9.1.14.2. The Commission's assessment ................................................................................... 80

9.1.15. Barriers to entry and expansion are high.................................................................... 83

9.1.15.1. The Notifying Party's views ....................................................................................... 83

9.1.15.2. The Commission's assessment ................................................................................... 84

9.1.16. Insufficient countervailing buyer power .................................................................... 97

9.1.16.1. The Notifying Party's views ....................................................................................... 97

9.1.16.2. The Commission's assessment ................................................................................... 98

9.2. Assessment of the Transaction in the relevant markets

9.2.1. Market shares and concentration indicators

9.2.1.1. Description of capacity and volume shares for each catchment area around individual customer filling locations

9.2.1.2. Average capacity and volume shares and concentration indicators at regional cluster level

9.2.1.3. Average pre-Transaction HHIs, post-Transaction HHIs and HHI deltas, at regional cluster level

9.2.1.4. Differential concentration effects within regional clusters

9.2.1.5. Assessment of the Notifying Party's arguments on concentration measures in the Reply to the SO

9.2.2. Competitive assessment for each cluster of catchment areas around customer filling locations

9.2.2.1. Cluster of catchment areas around individual customer filling locations in Central Europe: Germany and Austria

9.2.2.2. Cluster of catchment areas around individual customer filling locations in the Benelux

9.2.2.3. Cluster of catchment areas around individual customer filling locations in France

9.2.2.4. Cluster of catchment areas around individual customer filling locations in Italy

9.2.2.5. Cluster of catchment areas around individual customer filling locations in Iberia

9.2.2.6. Cluster of catchment areas around individual customer filling locations in North-East Europe: Poland, Czech Republic, Slovakia, Lithuania, Estonia and Latvia

9.2.2.7. Cluster of catchment areas around individual customer filling locations in South-East Europe: Hungary, Slovenia, Croatia, Romania, Bulgaria, Greece and Cyprus

9.2.2.8. Cluster of catchment areas around individual customer filling locations in the Nordics: Denmark, Norway, Sweden, Finland, and Iceland

9.2.2.9. Cluster of catchment areas around individual customer filling locations in the UK and Ireland

10. Competitive Assessment Beverage Cans: Coordinated Effects

10.1. The Notifying Party’s arguments

10.2. SABMiller's submission

10.3. The Commission's assessment

10.3.1. Assessment of a potential strengthening of coordination

10.3.2. Assessment of a potential creation of coordination

11. Competitive Assessment: Aluminium bottles

11.1. Structure of the market

11.2. The Notifying Party's views

11.3. The Commission's assessment

12. Commitments

12.1. Legal framework for the assessment of commitments

12.2. Commitments of 18 November 2015

12.2.1. Description

12.2.2. The Notifying Party's views

12.2.3. Responses to the market investigation

12.2.3.1. Scope of the Divestment Business as included in the Commitments of 18 November 2015

12.2.3.2. Viability of the Divestment Business as included in the Commitments of 18 November 2015 and suitable purchaser

12.2.4. The Commission's assessment

12.2.4.1. Suitability to remove the significant impediment to effective competition to which the Transaction would give rise

12.2.4.2. Viability, competitiveness and suitable purchaser

12.2.5. Conclusion

12.3. Final Commitments

12.3.1. Description

12.3.2. The Commission's assessment

12.3.2.1. Suitability to remove the significant impediment to effective competition to which the Transaction would give rise

12.3.2.2. Viability, competitiveness and suitable purchaser

12.3.3. Conclusion

13. Conditions and obligations

COMMISSION DECISION

of 15.1.2016

(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's decision of 20 July 2015 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) On 15 June 2015 the European Commission received a notification of a proposed concentration pursuant to Article 4 of Regulation (EC) No 139/2004 ('the Merger Regulation') by which the undertaking Ball Corporation ('Ball') would acquire the entire issued share capital of Rexam PLC ('Rexam') (the 'Transaction'). Ball is referred to as the 'Notifying Party'. Ball and Rexam are jointly referred to as the 'Parties'.

1. THE PARTIES

(2) Ball is a company based in the United States of America ('US'). It is active worldwide in the production and supply of metal packaging for beverage, food and household products. It has production facilities in North America, Brazil, Europe and the Asia Pacific region. Ball is also present in the design, development and manufacture of aerospace systems. Ball is the largest beverage can manufacturer worldwide and the second largest beverage can manufacturer in the EEA.

(3) Rexam is company based in the United Kingdom ('UK'). It is active worldwide in beverage can manufacturing, with production facilities in North America, South America, Europe, Africa, the Middle East and Asia. It is the second largest beverage can manufacturer worldwide and the largest beverage can manufacturer in the EEA.

2. THE OPERATION AND THE CONCENTRATION

(4) On 19 February 2015, Ball announced a recommended offer to acquire sole control of Rexam by way of purchase of shares for the total consideration of GBP 4.3 billion. The Transaction is a scheme under English law and needs to be approved by Rexam shareholders and sanctioned by a court.

(5) The Transaction 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 [Ball: EUR 6.5 billion; Rexam: EUR 4.8 billion]. Each of them has a Union-wide turnover in excess of EUR 250 million [Ball: EUR […] billion; Rexam: EUR […] billion], but they do not achieve more than two-thirds of their aggregate EU-wide turnover within one and the same Member State.

(7) The Transaction therefore has a Union dimension.

4. PROCEDURE

(8) On 20 July 2015, the Commission found that the Transaction raised serious doubts as to its compatibility with the internal market and the EEA Agreement and adopted a decision to initiate proceedings pursuant to Article 6(1)(c) of the Merger Regulation (the 'Article 6(1)(c) Decision'). The serious doubts were raised in relation to beverage cans, as well as aluminium bottles.

(9) On 22 July 2015, the Commission provided non-confidential versions of certain key submissions of third parties collected during the Phase I investigation to the Notifying Party.

(10) On 3 August 2015, the Notifying Party submitted its written comments to the Article 6(1)(c) Decision ('Reply to the Article 6(1)(c) Decision').

(11) On 10 August 2015, a formal State of Play meeting took place between the Commission and the Parties.

(12) During the Phase II investigation, the Commission sent several requests for information to the Parties, in particular on 24 July; 3, 7, 10, 21 and 28 August; 1, 7 and 11 September 2015; 5, 16 and 28 October 2015; and 17, 20, 23 and 25 November 2015.

(13) The Commission also sent requests for information to competitors and customers of the Parties, in particular on 27 July 2015, on 10, 11, 18, 28 and 31 August 2015, on 15 September 2015, on 12, 16 and 19 October 2015, and on 13 November 2015. Over 100 customers responded to those requests.

5 Turnover calculated in accordance with Article 5 of the Merger Regulation and the Commission Consolidated Jurisdictional Notice (OJ C 95, 16.4.2008, p. 1).

6 A supplementary document was sent on 29 July 2015 upon consideration of a request from the Notifying Party.

(18) Pursuant to Article 18(4) of the Merger Regulation read in conjunction with Article 5 of Decision 2011/695/EU, the Hearing Officer recognised as interested third parties the competitors Crown Holdings, Inc. ('Crown') and CAN-PACK S.A. ('Can-Pack'), as well as customers SABMiller, Carlsberg and a customer who wished to remain anonymous.

(19) On 6 November 2015, the Commission sent a first Letter of Facts to the Notifying Party ('Letter of Facts of 6 November'). The Notifying Party replied to the Letter of Facts of 6 November 2015 on 12 November 2015 ('Reply to Letter of Facts of 6 November').

(20) On 12 November 2015, the Commission sent a second Letter of Facts to the Notifying Party ('Letter of Facts of 12 November'). The Notifying Party replied to the Letter of Facts of 12 November 2015 on 16 November 2015 ('Reply to Letter of Facts of 12 November').

(21) On 27 November 2015, the Commission sent a third Letter of Facts to the Notifying Party ('Letter of Facts of 27 November'). The Notifying Party replied to the Letter of Facts of 27 November 2015 on 30 November 2015 ('Reply to Letter of Facts of 27 November').

(22) On 4 December 2015, the Commission sent a fourth Letter of Facts to the Notifying Party ('Letter of Facts of 4 December'). The Notifying Party replied to the Letter of Facts of 4 December 2015 on 7 December 2015 ('Reply to Letter of Facts of 4 December').

(23) On 18 November 2015, the Notifying Party submitted commitments in order to address the competition concerns identified in the SO (the 'Commitments of 18 November 2015'). Consequently, the period for the adoption of a final Decision was extended by 15 working days pursuant to Article 10(3) of the Merger Regulation.

Questionnaire to Customers (Q1), ID 1609; Questionnaire to Customers (Q2), ID 2664; Questionnaire to Customers (Q3), ID 2665; Questionnaire to Customers (Q4), ID 2786. Q1 was sent to all can customers. Q2 was sent to a smaller group of can customers: AB InBev, Bavaria, Brasserie Champigneulles, Britvic, Carlsberg, Coca-Cola Enterprises, Coca-Cola Erfrischungsgetränke AG, Coca-Cola Hellenic, Coca-Cola Iberian Partners, Molson Coors, Diageo, Heineken, Karlsberg Brauerei, Lidl, Mahou San Miguel, Nestlé, Oettinger Brauerei, Pepsico, Red Bull, Refresco, Royal Unibrew, S.A. DAMM, SABMiller and Schweppes. The demand for cans of all these customers, except Diageo, exceeds [….] million cans a year (based either on replies to Q1, question 4 or on the information provided by the Notifying Party in its Reply to the Article 6(1)(c) Decision, Annex 5. […]. Q3 was sent to aluminium bottles customers. Q4 was sent to all can customers that were not addresses of Q2.

Decision 2011/695/EU of the President of the European Commission of 13 October 2011 on the function and terms of reference of the hearing officer in certain competition proceedings (OJ L 275, 20.10.2011, page 29).

(24) On 20 November 2015, the Commission launched a market test of the Commitments of 18 November 2015.

(25) On 3 December 2015, the Notifying Party submitted a final set of commitments ('the Final Commitments').

(26) The meeting of the Advisory Committee took place on 5 January 2016.

(27) The Hearing Officer issued his final report on 6 January 2016.

5. INTRODUCTION TO THE PARTIES' ACTIVITIES

(28) The main products involved in the Transaction are beverage cans. The Parties are also active to a lesser extent in aluminium bottles.

5.1. Introduction to the beverage can sector

5.1.1. Description of the beverage can value chain and production process

(29) Beverages are packed in a range of products made from different substrates, including polyethylene terephthalate ('PET', a form of polyester) bottles, glass bottles, aluminium and steel cans, cartons, pouches, sachets, and kegs.

(30) As can be seen in Figure 1, in Europe PET accounts for around 50 % of beverage packaging, glass bottles for 15 %, beverage cans for 9 % and other types of material for 26 %. In terms of pricing, PET is the cheapest of the main packaging substrates while glass is the most expensive one.

Figure 1: European beverage packaging shares in 2014

Source: Form CO based on Canadean data

(31) The main packaging types come in a range of shapes and sizes and it is not infrequent for the same beverage to come not only in various packaging materials but also in various shapes and sizes in each material.

(32) A beverage can is a metal container designed to hold a fixed portion of liquid such as carbonated soft drinks, alcoholic beverages (for instance beer), fruit juices, energy drinks, etc.

(33) A beverage can is made up of a can body and a lid (can end). The technique used to manufacture the can body is called 'draw and wall ironing' ('DWI'). It generally involves the following steps: blanking and deep-drawing; wall-ironing and end forming; trimming; washing; outside coating; printing; drying; internal coating; necking; flanging; end coating; testing for holes and flange cracks; testing for internal defects; and palletising.

Figure 2: Production process of beverage cans

Source: Ball's website

(34) Can ends are typically manufactured in different plants to can bodies and involve other industrial production techniques. Can ends are then shipped separately to the customer and attached to the can at the customer's filling location once the can has been filled. Contrary to other parts of the world like the US or Canada, can bodies and can ends are always sold as a package in the EEA.

(35) Beverage cans are made from either aluminium or steel. Aluminium is used more widely and accounts for around 80 % of European can demand.

(36) Beverage cans come in different diameters and sizes. While the standard in Europe is 33 centilitres ('cl') and 50 cl with a standard diameter (66.2 or 66.3 millimetres ('mm')), other can types with different height and diameters are also produced. In particular, slim and sleek cans that have smaller diameters are sometimes referred to as specialty cans. Can sizes generally range from 150 millilitres ('ml') to 568 ml. The Parties offer most, if not all possible can sizes and types.

(37) According to the Notifying Party, the variables distinguishing the can size categories (standard, tall, slim and sleek) are diameter and height (which together determine the volume of the can). In terms of diameter, standard cans and tall cans both typically have a diameter of 66 mm, sleek cans have a diameter of 58 mm and slim cans typically have a diameter of 53 mm. In terms of height, a standard can with a diameter of 66 mm with a volume larger than 37.5 cl is categorised as tall. Similarly, sleek cans tend to be slightly taller than slim cans although this is not always the case. Table 1 provides examples of volumes which tend to fall within each category size.

Table 1: Typical beverage can size volumes according to the Notifying Party

Can size category Volume (cl)

Slim 15, 20, 25

Sleek 25, 27.5, 30, 33, 35.5, 40, 45, 50

Standard 25, 29.6, 30, 33, 34, 37.5

Tall 44, 45, 47, 50, 53, 55, 56.8, 71, 75, 94.6, 100

Source: Form CO

5.1.2. Demand for beverage cans in the EEA

(38) The annual sales volume of beverage cans in the EEA amounts to more than 50 billion cans (2014). Growth in demand is modest but steady. Market participants state for instance that:

'While the US is a declining market (with 95 billion cans per annum currently, versus 105 billion cans few years ago, but still a high annual can per capita ration of approximately 300), Europe is slightly growing (but from a more modest starting point of approximately 100 cans per capita annually).'

'[T]he demand for cans is increasing as it's easier for customers to transport and to drin[k] "on the GO".'

'The beverage can drink market in the EU is rather stable, growing about 5% each year.'

(39) The Parties, in their internal documents and publicly on their website, also expect a stable beverage can demand growth in Europe, highlighting facts such as recent growth in the take home beer sector resulting from a drop in disposable income levels. Figure 3 illustrates growth forecasts of the Parties.

Figure 3: The Parties' forecasts of the European beverage can sector

[…]

Source: Ball's internal document, 'Support demand based analysis of production footprint', ID 2006-60718 (left). Rexam's internal document, 'Pre-Budget Meeting', ID 2154-9036 (right)

(40) Customers purchasing beverage cans in Europe include both bottlers and beverage manufacturers (for instance soft drinks or beer manufacturers). The largest customers of beverage cans in Europe account for a large majority of demand, as can be seen in Figure 4. Demand is also fragmented, with more than 200 smaller customers in Europe.

Figure 4: Overview of largest European beverage can customers (all can types)

[…]

Source: Internal document of Rexam

(41) The packaging choice, and thus the preference for beverage cans, is mainly driven by consumer needs and trends. These include demographic trends and health and wellness trends. Consumption patterns are therefore evolving and beverage cans are increasingly being used for off-premise consumption (e.g. consumption at home).

(42) In addition, in the EEA, the Parties also expect a growing 'specialty' segment due to differentiation and segmentation. This growing demand for 'specialty' cans is also acknowledged by respondents to the market investigation, which highlight that end-consumers (and their preferences) are driving this growth.

(43) In the EEA, there is also a strong shift to aluminium cans. As explained by one aluminium supplier, aluminium has a number of advantages over steel, including sustainability, marketing and light-weighting. Whilst 20 years ago the proportion between aluminium and steel was roughly 50/50, it is currently around 80/20. Out of the 38 beverage can plants in the EEA operating a total of 93 lines, only 23 lines in 9 plants are steel lines, with a strong trend of steel plants being converted to aluminium and new investments only being done in aluminium.

16 Form CO, Attachment G, Reply to the Commission request for information ('RFI') 1, Exhibit H10, 'BCE Strategic Plan 2015-2017', slide 29.

17 Questionnaire to Customers (Q1), question 13, ID 1609.

18 Questionnaire to Customers (Q1), question 15.1, ID 1609.

19 Ball's internal document, 'B102 BPE Strategic Plan 2015-2017'.

20 'Moreover, [customer] perceives a strong increase in the demand for sleek cans. Standard cans are around 20-30% of [customer's] total production of cans. Sufficient volumes of sleek cans to meet the growing demand are difficult to source on the market.' Courtesy translation from the Italian original: 'Inoltre [cliente] ha riscontrato un forte aumento della domanda di lattine "sleek". Le lattine standard sono usate per il 20-30% della produzione totale di lattine di [cliente]. Le lattine sleek sono difficili da reperire sul mercato in quantità sufficiente per soddisfare la domanda crescente di tale tipo di lattina.' Agreed non-confidential minutes of a call with a customer, 07.08.2015, ID 2524. See also Questionnaire to Customers (Q1), question 29. ID 1609, agreed non-confidential minutes of a call with a customer, 19.05.2015, ID 1212, agreed non-confidential minutes of a call with a customer, 22.05.2015, ID 226.

21 Agreed non-confidential minutes of a call with a supplier, 13.05.2015, ID 1163.

22 Agreed non-confidential minutes of a call with a supplier, 13.05.2015, ID 1163, agreed non-confidential minutes of a meeting with a competitor, 05.05.2015, ID 261.

23 Plants not located in the EEA such as Ball's Belgrade and Rexam's Widnau as well as Helvetia's Saarlouis and […] plants, are not included in this number. All of these plants operate/will operate exclusively aluminium production lines.

24 Agreed non-confidential minutes of a call with a customer, 19.05.2015, ID 1212.

Market participants overall expect that eventually steel beverage cans will cease to be used and all steel plants will be closed or converted to aluminium. The Notifying Party's […]: (i) […]; (ii) […]; and (iii) […].

(44) Capital intensiveness, the disruption in production during the period of conversion and the learning curve of the personnel seem to be the key reasons why not all steel plants and lines have not yet been converted to aluminium.

(45) Regarding the capital intensive nature of the conversion from steel to aluminium, this is evidenced by the Parties' recent projects. For instance, for the conversion of […], Ball estimated in an internal document the cost to be around EUR […] million. In another internal document, Ball estimated the cost for the […] to be EUR […] million. For Ball's […] plant the investment for the conversion of […] was estimated at around EUR […] million (financial interest not included). In addition, as noted by Rexam, […]. For example, for the conversion of its […] plants, Rexam estimated that the conversion would lead […] totalling […] million cans and an additional cost of delivering cans from further away (as well as other transport inefficiencies) at around EUR […] million.

Figure 5: Ball's footprint project timeline

[…]

33 Source: Ball's internal document

(46) Regarding the disruption in production during the period of conversion, the process of converting lines from steel to aluminium is time-consuming although that the time it takes to complete a particular process varies depending on a wide range of factors, with a minimum time of […] weeks per line. For example, Ball estimated the time required to convert its […] plant to be up to […] months. For […], Ball estimated the time for the conversion of the plant was up to […] months with a downtime of […] months plus the learning curve. Similarly, Rexam estimated the conversion of […] steel lines in […] to take up to […] to […] months. Although the conversion can be done in stages by keeping one steel line operating while the other one is converted, the process involves a temporary reduction in capacity.

5.1.3. Beverage can manufacturers in the EEA

(47) At global level, Ball, Rexam and Crown are the main players, as highlighted in Figure 6. According to Rexam's website, '[t]he three biggest global can makers are Rexam, Ball and Crown who together account for more than 60% of the global beverage can volumes, and close to 90% in Europe, North and South America.' Rexam's CEO in a public statement describes can making as 'a consolidated industry, […] the three largest players, including ourselves, account for about 60% of global supply.'

Figure 6: Main players in beverage cans globally*

Source: Publicly available document of Rexam (Investors presentation 2013)

* MCC is a subsidiary of AB InBev

(48) In the EEA, four main players are present: Rexam, Ball, Crown and Can-Pack.

(49) Crown is a global packaging company headquartered in the US and listed on the New York Stock Exchange. It produces metal packaging for the (i) beverage; (ii) food; (iii) health and beauty; and (iv) household and industrial products sectors. Apart from Europe, Crown is present in both North and South America, Africa, the Middle East, as well as in South-East Asia.

(50) Can-Pack is a privately-owned company headquartered in Poland. It started supplying beverage cans in 1994 through its Brzesko plant in Poland. Can-Pack also produces glass bottles. Can-Pack has plants only in some parts of the EEA (Eastern Europe, Finland and the UK), as well as in India, Morocco, Russia, Ukraine and the United Arab Emirates.

(51) In the EEA, are also active: (i) BAGPAK Polska Sp. z o.o., 'BagPak'), a beverage can manufacturer based in Poland and owned by the Polish beer manufacturer, Van Pur; and (ii) Helvetia Packaging AG ('Helvetia'), which has built a new beverage can plant at Saarlouis in Germany that is expected to be operational in the course of 2015.

38 Rexam website: http://www.rexam.com/index.asp?pageid=790 Last accessed on 28.09.2015, ID 3003.

39 Rexam. Edited Transcript, REX L – Rexam PLC Investor. https://www.rexam.com/files/presentations/investorseminar11investor seminar nov11 transcript.pdf. Last accessed on 28.09.2015, ID 3000.

40 Crown website: http://www.crowncork.com/about-crown/global-locations Last accessed on 23.10.2015, ID 3413.

41 Form CO, Paragraphs 6.70-6.73.

42 Can-Pack website: http://www.canpack.eu/?pageid=45 Last accessed on 23.10.2015, ID 3414.

5.1.3.1. Location of beverage can plants serving the EEA

(52) The locations of the beverage can manufacturing plants of Ball, Rexam, Crown and Can-Pack are presented in Figure 7.

(53) Rexam has fifteen beverage can plants in the EEA, while Ball has nine. Figure 8 also includes plants of the Parties not located in the EEA but serving EEA customers (Rexam's Russian plants in Naro-Fominsk and Vsevolozhsk and Ball's Serbian plant in Belgrade). In addition, Figure 8 also shows Rexam's new plant in Widnau, Switzerland, which started in 2015 and […].

(54) Crown has eight beverage can plants in the EEA: two in each of Spain, Greece and the UK; as well as one in each of France and Slovakia. Crown also has a plant in Tunisia.

(55) Can-Pack has five beverage can plants in the EEA: two plants in Poland, and one in Romania, Finland and the UK. Its four other plants are located in Russia (two), and one in Ukraine and Morocco.

Figure 7: Beverage can plants serving the EEA – […]

[…]

Source: Ball's internal document 'EuropeMap.February.11.2014'

5.1.3.2. Location of can end plants

(56) Rexam has two can end plants in the EEA located in France and Ireland. Rexam is currently planning […]. In addition, one of its German beverage can facilities (Recklinghausen) produces both can bodies and can ends.

(57) Ball also currently has two can end facilities in the EEA located in Germany and the UK. In addition, in June 2015, it opened a new can end facility in Poland.

(58) Crown has three can ends plants in the EEA located in the UK, Greece and Poland.

(59) Can-Pack's two can plants located in Poland also produce can ends.

5.1.3.3. Location of research and development activities of the Parties

(60) Rexam's research and development ('R&D') activities for its beverage cans sold in the EEA are carried out at its locations in Tongwell, Milton Keynes (the UK), in Mont (France) and at

43These plants sell limited volumes to the EEA (Sales into the EEA represented in 2014 […] % of Naro-Fominsk's total volume and […] % of Vsevolozhsk's total volume, data from Reply to RFI 10 Exhibit B).

44[…] % of this plant's volumes are shipped to the EEA (data from Reply to RFI 10 Exhibit B).

45Rexam also has a plant in Turkey (Izmir). In the years 2009-2014 […] from Izmir to the EEA. Reply to the SO, footnote 181 and Reply to RFI 16 of 16.10.2015.

46See recital (105). By email of 9 December 2015, Rexam confirmed that 'operations at Rexam’s Widnau plant have now commenced with one production line. […].', ID 4588.

47Rexam's internal document, 'Implementation of […], Project update 23 March 2015'.

48Form CO, Paragraph 6.64.

49The plant is currently going through the '[…]'. Reply to RFI 13 of 07.09.2015.

50Form CO, Paragraph 6.67.

51Form CO, Paragraphs 6.71.

Innovation (new product development and improvement of products); developing new processes to improve plant efficiency and performance and engineering and technical support (providing technical support in relation to, inter alia, all tooling and container performance issues). Reply to RFI 5 of 13.07.2015, Exhibit U.

RTC, Elk Grove Village (the US). In addition, some small scale R&D activities are undertaken at other locations (for instance the Graphic Centre at the Luton HQ). Rexam also has a minority shareholding in Magna Parva, a UK-based R&D company.

Ball has R&D activities for its EEA operations at its Shared Service Centre in Bonn (Germany) and at the Ball Technical and Innovation Centre, in Broomfield, Colorado (the US).

5.1.4. The pre-Transaction expansion plans of the Parties

62Pre-Transaction, both Ball and Rexam had expansion plans.

Figure 8: Beverage can plants serving the EEA – […]

[…]

Source: Form CO

5.1.4.1. Ball in […]

63[…]

58 59 60

(64) […][…][…]

61 62

(65) […][…][…]

63 64 65 66 67

(66) […][…][…][…][…]

5.1.4.2. Rexam in […]

67Although […], prior to the public announcement of the Transaction Rexam had […]. The chronology of events is set out in this section.

68 69

(68) […][…]

70 71 72

(69) […][…][…]

53 Supporting sales and marketing development team; developing and improving manufacturing end processes and Innovation (product development, testing and improvement of can ends and coatings). Reply to RFI 5 of 13.07.2015, Exhibit U.

54 Innovation (new can body and can end development) and analytical testing and beverage testing. Reply to RFI 5 of 13.07.2015, Exhibit U.

http://magnaparva.com/markets/metal-packaging/ Reply to RFI 5 of 13.07.2015, Exhibit U.

56 Innovation (developing and testing new and existing ideas & products); engineering (supporting manufacturing plants on improving & harmonizing products & processes) and laboratory - (Testing of cans, ends and materials). Reply to RFI 5 of 13.07.2015, Exhibit T.

57 Innovation and laboratory. Reply to RFI 5 of 13.07.2015, Exhibit T.

58 Form CO, Attachment G, RFI 1, Exhibit H.190, '[…].

59 Ball's internal document, 'Project […]'.

60 Form CO, Attachment G, RFI 1, Exhibit H10, 'BCE Strategic Plan 2015-2017', slide 34.

61 Update reply to question 28 of RFI 2 of 30.06.2015, '[…].

62 Update reply to question 28 of RFI 2 of 30.06.2015, '[…].

63 In its Reply to RFI 10 of 21.08.2015, Exhibit A, the Notifying Party submits that […] while in its Reply to RFI 15 of 9.10.2015, Exhibit A, the Notifying Party submits that only […]

64 Reply to RFI 15 of 9.10.2015, Exhibit A.

65 Reply to RFI 10 of 21.08.2015, Exhibit A.

66 Reply to the SO, paragraph 4.23.

67 Ball's internal document, 'BPE Deep Dive Meeting, […]', slide 68, 71 and 74.

68 '[…]'. Rexam's internal email of […] from […], subject: 'Re: AW: […]'.

69 Rexam's internal email of […] from […], subject: 'Re: AW: […]'.

(70) […]

73 74

(71) […][…]

(72) […]

(73) […]

(74) […][…]

(75) […] .

(76) […]

(77) […]

(78) […][…][…][…][…][…]

Figure 9: Rexam's expansion plan in […]

[…]

Source: […], Rexam, ID 2154-16550

86 87

(79) […][…]

(80) […]

(81) […][…]

(82) […]

Rexam's internal document, '[…]', slides 17 and 61.

Form CO, Attachment G, RFI 1, Exhibit H10, 'BCE Strategic Plan 2015-2017', slide 74 and 75.

Form CO, Attachment G, RFI 1, Exhibit H10, 'BCE Strategic Plan 2015-2017', slide 74.

Rexam's internal document, '[…]', slide 61.

Rexam's internal document, '[…]', slide 5.

Rexam's internal document, 'Strategic Planning Board Actions […]', ID 3679.

Rexam's internal document, '[…] 01/2014/098 – December 2014'.

See Rexam's internal email, '[…]', ID 4023-1073.

Rexam's internal email of 9 December 2014 from […], subject 'Board meeting output'.

Rexam's internal email, '[…] – Board response paper', ID 4141-7. According to information provided by Rexam during the course of the administrative procedure, Rexam estimated initial supply contract volumes for the […] plant to be over […] billion cans per year. Rexam also estimated that the production costs of the […] to be as low as EUR […] for […] cl cans and EUR […] for […] cl cans.

Rexam's internal document, '[…] (CER Commentary - Supplementary questions) Final', in Rexam's internal email of 15 December 2014 from […], subject: '[…] – Board response papers'.

91Rexam's internal email, '[…] – Board response paper', ID 4141-7.

92Rexam's internal email of 18 December 2014 from […], 'Board feedback', ID 4027-861.

93Rexam's internal email, 'Re: […] - Board response paper', ID 4023-196.

94Rexam's internal email, 'Re: […] - Board response paper', ID 4023-196.

95See for instance Rexam's internal emails, 'RE: […]', ID 4023-296, 'Re […]', ID 4023-195; 'Outstanding issues to resolve today', ID 4027-1625. And correspondence between Rexam and […], see for instance Rexam email to […], subject: '[…]', ID 2042-20360.

96Rexam's internal email, 'RE: […] – Board response papers', ID 4023-1096.

97Rexam's email to […], subject: '[…]', ID 2042-20360.

98Rexam's internal document (email attachment), 'SCAF Executive Summary […]', ID 4141-14.

99See Rexam's internal emails, '[…]', ID 4027-1400, 'RE: […]', ID 4027-1395 and 'Re: […]', ID 4023-280.

100See for instance Rexam's internal emails, 'WG: […]', ID 4023-943 and Rexam email to […]', ID 4023-488.

101Rexam's internal email of 7 January 2015 of […] to […], subject: 'RE: […]'.

102Rexam's internal document, '[…] content – Feb19 media statement', in Rexam's internal email of 30 January 2015 from […], Subject: 'Pre-read to SteerCom on Monday'.

103Rexam's internal document, […] content - Investor Relations', in Rexam's internal email of 30 January 2015 from […], Subject: 'Pre-read to SteerCom on Monday'.

104Rexam's internal document, '[…] content - Investor Relations', in Rexam's internal email of 30 January 2015 from […], Subject: 'Pre-read to SteerCom on Monday'.

105Rexam's internal document, '[…] content – Storyline ND', in Rexam's internal email of 30 January 2015 from […], Subject: 'Pre-read to SteerCom on Monday'.

106Reply to RFI 18 of 17.11.2015, Annex 1.

107Rexam's internal document, 'Project Alliance, Communication to ELT', ID 419-167.

108See for instance Rexam's internal emails, subject: 'RE: […]', ID 4023-776, 'Re: […]', ID 4023-445, 'Re: […]', ID 4023-772, 'Re: Annual Report 2014 - for review and comment', ID 4027-1375, 'FW: […]', ID 4141-159, 'RE: […]', ID 4023-175, 'Re: […]: video script and Q&A', ID 4027-1669, 'RE: […] slides - YE presentation', ID 4027-1828, and Rexam email to […], '[…]', ID 4141-89.

109Rexam's internal email of 31 January 2015, subject: 'RE: […]', ID 4023-333.

110Rexam's internal email of 12 February 2015 from […], subject: 'exec meeting 12th feb', ID 4027-1610.

111Rexam's internal emails, 'Re: Feb Board agenda', ID 4023-614. See also Rexam's internal emails, 'Re: Tuesday […]', ID 4023-49.

EN 13 EN

(98) […]

(99) […]

(100) […]

(101) […]

(102) […]

(103) […]

(104) […]

(105) […]

5.1.4.3. Rexam in Switzerland

106Rexam has built a new plant in Widnau, Switzerland, to […]. Absent the Transaction, Rexam's plan ([…]) was to use the plant […].

107The first line started operating in 2015. The second line is scheduled to start in […] and the third line in […].

108The start of operations at Widnau will […] Rexam's plants at […] and […] whilst […]. The Widnau plant is expected to reach a capacity of approximately […] billion cans by […].

5.1.5. Changes in the beverage cans sector in the EEA since 1999

5.1.5.1. Changes relating to Rexam

109In 1999, Rexam started to supply beverage cans in the EEA by acquiring Swedish beverage can maker PLM.

110In 2000, Rexam further expanded through the acquisition of American National Can. The acquisition reduced the number of main players in the EEA from four to three: Rexam, German

Rexam's internal document, '2015 02 17 – Minutes – Rexam Board (redacted)'.

Rexam's email of 19 February 2015 from […] to […], subject: '[…]'.

Rexam's internal email of 11 March 2015 from […] to […], Subject: 'Re: […]'.

Rexam's internal email of 13 March 2015 from […] to […] to […], subject: 'RE: […]'.

Rexam's internal email of 14 April 2015 from […] to […] and […], subject: 'RE: […]'. See also Rexam's internal email of 25 March 2015 from […] to […] and […], Subject: 'RE: […]'.

Rexam's internal email of 13 March 2015 from […] to […], subject: 'Re:[…]'.

SO, paragraph 4.217.

Rexam's internal email of 27 February 2015 from […], subject: 'RE: SteerCom meeting […] 2nd Mar.pptx'.

Rexam's internal email of 17 February 2015, subject: 'AW: WG: […]', ID 4027-1083.

[…] email to Rexam of 10 April 2015, subject: 'RE: […]', ID 2788-120.

Form CO, Paragraph 6.283.

Reply to RFI 11 of 28.08.2015.

Reply to RFI 11 of 28.08.2015.

Form CO, Attachment G, RFI 1, Exhibit H.173.

EN 14 EN

based packaging company Schmalbach-Lubeca, a subsidiary of Continental Can and Crown (Crown Cork/CarnaudMetalbox). The only other player in the EEA at the time was the small independent manufacturer Italian Tubettificio Lecco (mainly active in Italy).

In order to remove the competition concerns in the EEA raised by the acquisition of American National Can, Rexam undertook to divest three beverage can plants: La Ciotat in France, Gelsenkirchen in Germany and Runcorn in the UK. Schmalbach-Lubeca acquired the La Ciotat and Runcorn plants. As for the Gelsenkirchen plant, it was sold to a private investor group. Following the introduction of a can deposit scheme in Germany which prompted a collapse in the demand for beverage cans, the Gelsenkirchen plant experienced serious financial difficulties leading to a reacquisition of the plant by Rexam in 2003.

In 2001, Rexam took over the Ejpovice beverage can plant in the Czech Republic from Schmalbach-Lubeca in a separate transaction.

In 2007, Rexam opened a new plant in Ludesch (Austria). In 2009, it opened a new plant in Fredericia (Denmark). Rexam also invested in Finland, where it opened its Mäntsälä plant in 2013. Finally, Rexam's Widnau plant in Switzerland started operating in 2015 (see section 5.1.4.3).

Apart from plant openings, Rexam has also conducted several improvements to its plants, including […].

As regards plant closures, in 2009, Rexam closed its Dunkirk plant in France, which mainly produced steel cans for Coca-Cola. The Dmitrov plant near Moscow was also closed by Rexam in the same year.

Rexam publicly announced the closure of its Berlin plant in 2015. Rexam has […].

5.1.5.2. Changes relating to Ball

117In 2002, Ball started supplying beverage cans in the EEA through the acquisition of Schmalbach-Lubeca. Following the acquisition, there were two other main players, Rexam and Crown, and two smaller players, Tubettificio Lecco and Can-Pack.

118In 2003, Ball closed the UK Runcorn plant (previously divested by Rexam to Schmalbach-Lubeca).

119In 2005, a new plant in Belgrade (Serbia) was opened by Ball.

120Apart from plant openings, Ball has conducted several improvements to the plants, including […].

127Case M.1939 – Rexam/American National Can (2000).

128Can-Pack, who at the time had one plant in Poland, was delivering some volumes into the EEA.

129Case M.2542 – Schmalbach-Lubeca / Rexam (2001).

Reply to RFI 5 of 13.07.2015, Exhibit Q.

Moreover, the […] has been envisaged in a 5 to 10 years horizon subject to future market development in the EEA. See Rexam's internal document, '[…] – Board Presentation', slide 15.

Crown entered the European beverage can market in the1990s by acquiring Carnaud-Metalbox.

Two lines out of three were for steel cans, whose demand dropped. Moreover, the new German deposit scheme decreased the German demand. Aluminium volumes were subsequently consolidated in the nearby Wrexham plant. See Form CO, Annex 20.

Reply to RFI 10 of 21.08.2015, Exhibit A.

Reply to RFI 5 of 13.07.2015, Exhibit P.

EN 15 EN

128Outside the EEA, a larger number of suppliers exist, including Ball (DWI bottles), Rexam (DWI bottles), Exal (DWI and IE bottles), CCL Container (IE bottles), Toyo Seikan (DWI bottles), Daiwa Can (DWI bottles) and Universal Can Corporation (which manufactures DWI bottles in China and licenses its DWI bottle technology to other manufacturers around the world).

129Ball produces aluminium bottles mainly in the US. It is a large player worldwide for DWI bottles (Alumi-Tek brand). In Europe, however, Ball only supplies IE bottles. In 2011 Ball bought an aerosol can factory in the Czech Republic with the capabilities to produce IE bottles. Ball started sales of IE bottles to EEA customers in 2014.

130Rexam is the only producer and supplier of DWI bottles (Fusion brand) in Europe. These bottles are produced at Rexam's beverage can plant at Ejpovice in the Czech Republic using the DWI process. Rexam does not produce or supply IE bottles in the EEA or anywhere in the world.

136Reply to the SO, Annex 1.

137See Ball's internal document 'Tubettificio Project Overview - 2nd July 2012', and Ball's internal email of 13 July 2012 from Roy Brooks to Dominique Mercier, Subject "RE: Tubetificio - Stolle Europe".

138Reply to the Article 6(1)(c) Decision, Paragraphs 6.10 - 6.13, Reply to RFI 5, Exhibit O.10.

139Form CO, paragraphs 6.150 and 6.151.

140Form CO, paragraphs 6.150 and 6.152.

EN 16 EN

6. RELEVANT PRODUCT MARKET

6.1. Beverage cans

6.1.1. Beverage cans and other forms of beverage packaging solutions

6.1.1.1. Past decisional practice

141In Rexam/American National Can, the Commission found that beverage cans constitute a separate market from other forms of beverage packaging solutions. This finding was based on the fact that bottlers and fillers would not switch their demand to glass or polyethylene terephthalate ('PET') bottles in the event of a small but significant and non-transitory increase in the price for beverage cans. Furthermore, the Commission noted that such shift would not be motivated by a change in the relative price of various packaging solutions, but would have to consider other non-price factors, such as marketing and product image and changes to consumer preferences. Therefore, the Commission concluded that beverage cans constitute a separate product market from other forms of beverage packaging solutions.

6.1.1.2. The Notifying Party's views

132In the Form CO and its Reply to the Article 6(1)(c) Decision, the Notifying Party submits that glass, PET, and other forms of packaging impose a strong competitive constraint on beverage cans.

133The Notifying Party considers that this view is supported by examples of beverage producers switching or threatening to switch between packaging types and reflected in the Parties' internal documents. Furthermore, the Notifying Party points out that beverage producers have a significant degree of influence over the packaging mix.

134In the Reply to the Article 6(1)(c) Decision and the Reply the SO, the Notifying Party put forward examples of internal documents which show […] (see section 9.1.14).

6.1.1.3. The Commission's assessment

135For the reasons set out in this section, the Commission has reached the conclusion that beverage cans constitute a separate market from other forms of beverage packaging solutions.

First, the Parties' customers generally use a mix of packaging types. These include not only cans, glass and PET bottles, but also kegs, cartons, and pouches. Customers consider that there are several distinguishing factors between cans and other types of packaging and substantiated their views as follows:

144'Each packaging material has different product characteristics such as whether or not it is re-sealable, its shape, its environmental performance and its suitability for decoration and artwork. […] cans differ from other beverage solutions in terms of marketing, product image and customer and consumer needs and preferences'.

EN 17 EN

(142) Likewise, a large majority of customers that responded to the market investigation indicated that they would not switch to cans from glass or PET in case of a permanent 5-10 % price increase of the latter two.

6.1.2. Various sizes and types of cans

6.1.2.1. Past decisional practice

143In Rexam/American National Can, the Commission found that all can size categories (27.5 cl, 33 cl, 44 cl, 45 cl, and 50 cl) belong to the same product market, with the exception of slim cans (15 cl and 25 cl).

144From a demand-side perspective, the Commission found that customers' filling lines could easily be changed to accommodate different-sized cans provided these are of standard-body diameter, which was not the case for slim cans. From a supply-side perspective, the Commission found that can manufacturers could switch within a reasonable time and at a reasonable cost, because their manufacturing lines are designed as 'swing lines', that is to say lines that allow beverage can manufacturers to produce different can sizes on the same line.

6.1.2.2. The Notifying Party's views

145The Notifying Party considers that there is a single market for all can size categories, due to the high degree of demand- and supply-side substitutability between sizes, but that there is limited differentiation between standard and specialty cans as there is sufficient demand and supply-side substitutability between standard and speciality cans.

146On the demand-side, the Notifying Party submits that: (i) beverage producers tend to use a mix of can sizes and the same beverage is often packaged in cans of different height, diameter or volume; (ii) beverage producers can and do switch can size categories and also change the mix of cans supplied within each category; and (iii) beverage producers are generally able to accommodate different can size categories on the same filling lines. The Notifying Party also argues that there is a high degree of supply-side substitutability between can sizes, including between can size categories.

6.1.2.3. The Commission's assessment

147Due to end-consumer preferences, beverage can manufacturers generally offer products in cans of different sizes and diameters and it is not uncommon for the same beverage to be sold in cans of different formats. As explained by a competitor, different can sizes serve different consumer needs or distribution channels: 15 cl cans are predominantly used to contain beverages for airplane-consumption; 20 cl for promotional purposes and wines and 25 cl for energy-drinks (especially Red Bull). As for slim cans, they are mostly used for energy drinks, with the 33 cl size mostly used for soft-drinks and the 50 cl size mostly used for beer.

EN 19 EN

(148) For the reasons set out in this section, the Commission has reached the conclusion that various can sizes and can types belong to the same market even though they constitute differentiated products within this market. This is the case for both: (i) different can sizes of standard diameter; and (ii) all cans of standard diameter and narrow body cans (slim and sleek).

(149) As regards (i) different can sizes of standard diameter, the Commission's conclusion is based on the considerations set out in recitals (150)-(152).

(150) First, from a demand-side perspective (that is to say, from the perspective of the beverage manufacturers and fillers), filling lines can easily be changed to accommodate different can sizes provided they are of the same diameter.

151Second, from a supply-side perspective (that is to say, from the perspective of the can manufacturers) converting existing production lines to different can sizes or can types is possible with a capital investment of EUR [0-5] million per line, depending on the complexity of the retooling processes.

152Third, the four main beverage can manufacturers that supply beverage cans in the EEA are able to offer a wide range of sizes and types of cans, although for efficiency reasons each plant does not necessarily produce a complete product range.

153As regards (ii) narrow body cans (that is to say, slim and sleek), the Commission's conclusion is based on the considerations set out in recitals (154)-(158).

154First, while slim and sleek cans were traditionally considered as a niche product (specialty cans) used by certain beverage producers, in particular energy drink producers, as brand differentiators, they are now ordered in significant quantities for packaging a range of beverages even though they are generally more expensive than standard cans. As such, even though slim and sleek cans might still be considered as premium products and a brand differentiator, they no longer represent a novelty to the end-consumer.

155This is supported by a majority of customers that responded to the market investigation. Customers explained that '[h]istorically, “specialty cans” were defined as anything other than the standard 330ml can. However, as a result of evolving consumer preferences and demand, the variety of can sizes has been increasing' and that '[s]leek cans are more and more prevalent in the market, however, the overall price of sleek and slim are much higher than standard cans. Another customer pointed out that '15 years ago Slim/Sleek-cans had a insignificant volume, nowadays they do have very high volumes and therefore are to be considered normal Can sizes'.

This was also supported by a competitor that considered that '[n]on-standard in this sense means different sizes in diameter or height from the traditional 330 and 500 ml. and therefore containing different quantities of fluids (sometimes, but misleadingly, referred to as

EN 20 EN

157Second, from a demand-side perspective (that is to say, from the perspective of the beverage manufacturers and fillers), switching between different diameters is also possible, even though switching filling lines from standard diameter to smaller diameter to fill sleek and slim cans appears more costly and time consuming than switching lines to accommodate different can sizes of standard diameter. Some customers use contract fillers for narrow body cans when quantities for those cans are small and therefore do not justify the necessary investments for the switchover. Customers that regularly order large volumes of narrow body cans tend to have filling lines dedicated to filling these types of cans.

158Third, from a supply-side perspective (that is to say, from the perspective of the can manufacturers), swing lines allow for switching even between different diameters, although for reasons of efficiency, beverage can manufacturers do not switch back and forth between different sizes and diameters but tend to optimise their production lines to produce a limited amount of different sizes. As a competitor indicated, '[s]wing lines are more time consuming and less economical in terms of output, but more flexible. Overall switching height is less costly than changing the diameter of the production line. For newer lines switching sizes is easier'.

6.1.3. Can ends and can bodies for beverage cans

6.1.3.1. Past decisional practice

159In Crown/Mivisa, the Commission found that can bodies and can ends for food cans are part of a single relevant market. A majority of customers would not switch to separate purchases of can ends and can bodies in the event of a small but permanent increase in price of whole cans and would not purchase can bodies from a can manufacturer if that manufacturer could not also provide can ends.

6.1.3.2. The Notifying Party's views

160The Notifying Party submits that can ends and can bodies for beverage cans belong to the same product market. This is because typically neither of the Parties sells beverage can ends separately from beverage can bodies - all of the Parties' sales of can ends are sold and priced as part of a contract package with can bodies. This ensures that the liability for can quality rests clearly with the beverage can manufacturer for warranty purposes.

6.1.3.3. The Commission's assessment

161For the reasons set out in this section, the Commission has reached the conclusion that can ends and can bodies form part of the same product market.

First, beverage can manufacturers manufacture both can ends and can bodies and at least in Europe, sell them in a bundle to customers.

EN 21 EN

162Second, the majority of customers that responded to the market investigation indicated that they buy can bodies and can ends together, also due to both quality reasons, such as better quality assurances, and cost reasons, including better commercial conditions. As a customer pointed out '[t]he decision to buy cans and lids together is a matter of clear supplier responsibility and in case of problems the resulting warranty for the fit of the combined packaging'.

6.1.4. Aluminium and steel cans

As already set out in recital (35), beverage cans are made from either aluminium or steel. Aluminium is used widely and currently accounts for 90 % of global and 80 % of European beverage can production.

Moreover, most customers only source aluminium cans whereas only a few customers exclusively source steel cans. Customers that source both aluminium and steel cans mainly source aluminium cans.

6.1.4.1. Past decisional practice

166In Rexam/American National Can, the Commission found that in the geographic areas where both steel and aluminium were used to manufacture cans aluminium and steel cans formed a single product market. This was based, among others things, on identical prices, almost identical end-users and consumers' perceptions, and a cost difference of producing aluminium and steel cans which amounted to less than 2 %.

6.1.4.2. The Notifying Party's views

167The Notifying Party submits that there is a single market for aluminium and steel beverage cans due to the significant degree of demand- and supply-side substitutability between these metal types. While the Notifying Party recognises that there is a degree of differentiation between aluminium and steel cans, these differences should not be overstated.

6.1.4.3. The Commission's assessment

168The Commission has reached the conclusion that aluminium and steel cans belong to the same product market in the geographic areas where both steel and aluminium is used to manufacture cans.

On the one hand, two features suggest that aluminium and steel cans could potentially belong to different product markets.

First, aluminium cans are better matched with consumer preferences as they provide an overall better quality than steel cans. In particular, aluminium cans are shinier (therefore look better on designs where large parts of the can are not printed) and lighter (reducing costs of production),

EN 22 EN

are easier to recycle and have a more limited risk of corrosion than steel cans. Embossing and other decorations are also easier on aluminium cans.

For example, customers indicated that '[A]luminium does not corrode and provides an excellent barrier for oxygen, light and moisture, and hence does not contaminate the contents of the container and may be used to pack sensitive products, such as beer. […] [A]luminium cans are more easily shaped than steel cans, allowing for more product differentiation downstream'.

Second, as noted by large customers, steel does not offer the same hedging options as aluminium which protects them against swings in raw material prices and allows for better risk management. Some customers only purchase aluminium cans for this specific reason: '[t]he general direction of the market is toward aluminium for all cans partly because it is easier to hedge the cost of aluminium (LME), and because embossing and other decorations are easier on aluminium cans than on steel cans'.

On the other hand, a number of other features indicate that aluminium and steel cans belong to the same product market.

First, most customers consider that there are currently no major differences in terms of legal requirements, effect on the beverages' taste, and price. For example, one bottle filler indicated that 'we also fill products where steel is the preferred metal', while a customer stated that 'the preference of either aluminium or steel cans depends on the country in which the cans are sold'.

Second, depending on the technical characteristics of the individual filling lines of a customer, a switch from aluminium to steel and vice versa may take a few hours or several days and require no or a significant investment. A majority of customers indicated that they could switch their filling lines from aluminium to steel and vice versa with no or limited adjustments.

Third, in case of a permanent 5-10 % price increase for steel cans, customers currently using steel cans would switch in whole or in part to aluminium.

Fourth, customers in the EEA are expected to gradually move from steel to aluminium cans. This move is triggered by several factors.

In the first place, a majority of the plants of the four main beverage can manufacturers in the EEA produce only aluminium cans:

out of Ball's nine plants in the EEA, only […] steel can production lines. […].[…]

Questionnaire to Customers (Q1), question 24, ID 1609. See also agreed non-confidential minutes of a call with a customer, 22.05.2015, ID 226.

Questionnaire to Customers (Q1), question 22, ID 1609.

Questionnaire to Customers (Q1), questions 21 and 22, ID 1609. See also agreed non-confidential minutes of a call with a customer, 28.04.2015, ID 164; agreed non-confidential minutes of a call with a customer, 12.05.2015, ID 468; agreed non-confidential minutes of a meeting with a competitor, 05.05.2015, ID 261.

Agreed non-confidential minutes of a call with a customer, 28.04.2015, ID 164.

Agreed non-confidential minutes of a call with a customer, 25.08.2015, ID 2691.

Questionnaire to Customers (Q1), question 24, ID 1609.

Questionnaire to Customers (Q1), question 25, ID 1609.

Agreed non confidential minutes of a call with a customer, 22.05.2015, ID 226.

Questionnaire to Customers (Q1), question 23.1, ID 1609.

Questionnaire to Customers (Q1), questions 23 and 23.1, ID 1609.

Questionnaire to Customers (Q1), question 27, ID 1609.

EN 23 EN

out of Rexam's fifteen plants in the EEA, […] has steel can production lines and […];

out of Crown's eight plants in the EEA, only three have steel can production lines.

The Parties themselves consider that […]'and […]and therefore […].This is a reflection of the demand focusing on aluminium cans: '[…].'Even when capacity is tight, aluminium continues to be the preferred choice '[…].

In the second place, respondents to the market investigation indicated that with the development of extra thin cans, can manufacturers need to use aluminium rather than steel in the production of cans as this raw material is thinner and lighter than steel. As explained by a customer, 'aluminium cans are processed at a lower temperature than steel cans and the production of aluminium cans requires less ovens and energy [...] Consequently, the same amount of beverage cans can be produced much faster, with less energy and labour support, resulting in efficiencies of at least 5 to 15%, in addition to the lower cost for the raw material'.

6.2. Aluminium bottles

6.2.1. Past decisional practice

181The Commission has not analysed the market for aluminium bottles in past decisions.

6.2.2. The Notifying Party's views

182Both Ball and Rexam produce aluminium bottles, but the Notifying Party claims that their respective products and technologies are very different with Ball offering IE bottles whilst Rexam bottles are manufactured through the DWI process.

183According to the Notifying Party, the IE aluminium bottles are manufactured from a metal slug using the same process and equipment used to manufacture aerosol cans. The DWI aluminium bottles are manufactured through a technique similar to can production.

Therefore, while both products fall within the broader beverage packaging market, differentiation between the products limits the degree of competition between the products. Accordingly, the Parties consider that their bottles only compete to the extent that they fall within the broader beverage packaging market.

The Notifying Party submits that aluminium bottles are subject to competitive constraint from other forms of beverage packaging such as glass, PET or cans. Aluminium bottles are a relatively small and newly-offered product, therefore they have to compete with more established forms of beverage packaging materials.

Reply to RFI 5 of 13.07.2015, Exhibit O.18, '[…], November / December 2014'.

Ball's internal document, 'BPE Footprint Optimization 2015 – 2020 […]', slide 5.

Form CO, Table 6.8.

Ball's internal document, 'BPE Footprint Optimization 2015 – 2020 […]', slides 2 and 4.

Ball's internal document, 'BPE Deep Dive Meeting, September 22 2014', slide 34.

186At the same time, the Notifying Party nevertheless submits that there is no need to decide on the exact product market definition, as no competition concerns would arise under any potential product market definition.

6.2.3. The Commission's assessment

187Aluminium bottles are a relatively new and niche product. For the purposes of the present Decision, the Commission assesses aluminium bottles as a separate market from other forms of beverage packaging solutions. Moreover, the Commission considers that for the purposes of the this Decision the exact product market definition for possible sub-segments of the aluminium bottles market can be left open since the proposed Transaction would not lead to a significant impediment of effective competition under any plausible product market definitions.

The market investigation has revealed the following information about the market for aluminium bottles.

189First, aluminium bottles belong to a niche market. For instance, 'aluminium bottles are used principally for the sale or sampling of the Company’s products at prestige and promotional and sponsored events, such as the Olympic Games, and in certain niche channel, such as discos,' 'aluminium bottles are used in certain venues such as night clubs, bars and during concert or festivals', '[…] because of the high price, aluminium bottles […]are linked to very specific and limited consumer occasions and preferences, and distributed in very narrow trade channels with low volume turnover (as compared to mainstream volumes filled in cans)'.

190Second, aluminium bottles are linked to special or promotional events, such as 'for instance the football World Cup. Aluminium bottles are also used for small volume promotions or marketing events', 'Aluminium bottles are positioned for high end, special occasion consumption, and to date have been used as promotional limited edition packs'.

191Third, aluminium bottles are relatively expensive compared to other beverage packaging materials: 'aluminium bottles are considerably more expensive than cans' 'High price for the consumer compared to other formats', 'promotional packaging type with premium product positioning and at a high selling price point'.

As for a possible further segmentation of aluminium bottles between IE bottles and DWI bottles, there are arguments both in favour and against.

One the one hand, respondents to the market investigation indicated that while IE bottles and DWI bottles use different technologies, they serve the same customer requirements. 'The IE and D&I [DWI] aluminium bottles both can be reclosed and are essentially the same from the demand side perspective. The main difference between them is in the production techniques and the fact that the production costs of D&I [DWI] aluminium bottles are lower over time but requiring a higher initial investment. For D&I [DWI] aluminium bottles less metal is required which makes them lighter and therefore easier to transport. Production lines of IE aluminium bottles are limited in terms of speed with an output of approximately 250 bottles per minute in contrast to a D&I [DWI] aluminium production line with an output of approximately 600 aluminium bottles'.

Questionnaire to Customers (Q1), questions 36 and 37, ID 1609.

Questionnaire to Customers (Q1), question 36, ID 1609.

Questionnaire to Customers (Q1), question 37, ID 1609.

194One the other hand, according to internal documents of the Parties, IE bottles provide a lower output than DWI bottles, namely less than […] million bottles per year compared to DWI bottles (minimum […] million). In addition, in terms of design, IE bottles have thicker walls ([…] grams ('g') of metal versus […] g for DWI bottles), are more rigid and they use different types of closure. Moreover, as a result of their thickness, IE bottles have a higher price than DWI bottles.

Table 2: Differences between IE and DWI aluminium bottles

IE bottles DWI bottles

Technology Produced on technology similar to aerosol bottles

Could be filled on existing lines

Require investment into modification lines for glass bottles

Design Different types of closure

Uses only screw cap closure

Thicker walls

Thinner walls

More rigid

Less rigid

Volumes Higher output because of faster production

Lower output (< […] million bottles/year) at the same cost and timeframe as DWI

Price Higher than DWI

Lower than IE

(IE bottles use more metal, (approximately […] % less metal costs as namely […] g versus […] g and they use less material and are therefore heavier)

Source: Questionnaire to Customers (Q1), ID 1609, question 38; Questionnaire to Customers (Q3), ID 2665; Ball's internal documents, 'Alumi-Tek Bottle', 31 January 2013, 'AlumiTek RPM Biz Case BPE e2' and 'AlumiTek BPE Overview Internal Update'

7. RELEVANT GEOGRAPHIC MARKET

7.1. Beverage cans

7.1.1. Past decisional practice

195In Rexam/American National Can and Schmalbach-Lubeca/Rexam, the Commission found the scope of the relevant geographic market for beverage cans to be regional. In reaching this conclusion, the Commission notably took into account different factors including the location

of customers relative to plants, transport costs, national legislation, and the demand-side structure.

7.1.2. The Notifying Party's views

196In the Form CO and the Reply to the Article 6(1)(c) Decision, the Notifying Party made the following arguments.

197It submitted that the market for beverage cans is regional at most (with some local differences within the regions) due to: (i) transportation costs, (ii) actual transport patterns; (iii) the mutual benefit for can manufacturers and customers of being located close to each other to maintain good service levels, security of supplies, rapid reaction times and just-in-time deliveries; (iv) low volumes of inter-region trade flows; (v) pricing occurs on a site-by-site basis and there are differences in prices among regions; and (vi) differences in the shares held by the various can manufacturers in the different regions.

(198) First, transport costs are the main variable in customer-specific costs and increase in parallel with the distance between the can manufacturing plant and the customer's filling location. Therefore, a can manufacturer with plants located more closely to a customer's plant has a cost advantage over rivals when competing for the customer's business at that location. Due to the significance of transport costs, there is a strong locational dimension to competition. For example, the supply options for a customer filling location in Spain are entirely different to those for a customer filling location in Poland or in the UK.

(199) Second, the Notifying Party argued that a catchment area analysis shows that significant amounts of cans are transported within a […] kilometres ('km') radius (more than […] % of the volume of the Parties' EEA plants' output) or within a […] km radius (more than […] % of the output volume of the Parties' plants in the EEA) around a can plant. On average, the Parties' cans are transported less than […] km.

(200) Third, customers will generally prefer suppliers close to their filling locations as this is more environmentally-friendly, allows for just-in-time delivery and security of supply.

(201) Fourth, on the basis of its trade flow analysis, the different regions can be distinguished as: (i) the UK and Ireland, (ii) North-West Europe, (iii) South-West Europe, (iv) North-East Europe, (v) South-East Europe, and (vi) Nordic countries, because trade flows mainly occur within these regions.

(202) However, these regions are permeable at the edges and, in particular within the relatively large regions of North-West Europe and South-West Europe, significant differences in the sourcing patterns occur by country. This suggests a more localised market. Also, for each region, except the UK and Nordic countries, higher volume trade flows from adjacent regions should be taken into account when assessing the impact of the Transaction.

(203) Fifth, pricing occurs on a customer site-by-site basis and that even large customers present in several regions do not request (and suppliers do not provide) a single EEA-wide price. To support this conclusion the Notifying Party submitted Figure 10 and Figure 11 below which show that Ball's and Rexam's average prices for cans […], which, according to the Notifying Party, indicates that the scope of the geographic market is not wider than regional.

Figure 10: Ball's and Rexam's average prices by Region (2014)

[…]

Source: Reply to the Article 6(1)(c) Decision, Annex 8

Figure 11: Ball's and Rexam's variable margins by region (2014)

[…]

Source: Reply to the Article 6(1)(c) Decision, Annex 8

(204) Sixth, the Notifying Party is of the view that the local nature of competition is evidenced by the fact that market shares vary dramatically across the EEA, reflecting the fact that each beverage can manufacturer's competitive strength in a region is in large part determined by the location of its plants.

(205) In its Reply to the SO, the Notifying Party made the following additional arguments.

(206) First, it accepts the utility of a catchment area and cluster approach in defining the relevant range of geographic competition in the EEA beverage can industry.

(207) Second, the radius of the catchment areas should, however, be defined on the basis of the maximum distance at which Ball and Rexam ship 70 % of their beverage can production. It further argued that such distance should be calculated taking into account driving distances. On this basis, the Notifying Party submitted that the appropriate radius for a catchment area assessment corresponds to a 500 km area around each customer. The Notifying Party considered that this conclusion is supported by evidence in the Commission's file.

(208) Third, the relevant geographic market should not be defined on the basis of catchment areas around individual customer filling locations but rather on the basis of regional clusters grouping together catchment areas around individual customer filling locations. It supported this claim by indicating that (i) within the regional clusters competitive conditions are sufficiently homogeneous and (ii) the scope for price discrimination within a regional cluster is limited, in particular because in each regional cluster competitive choices are unlikely to vary significantly and, even if they vary, any difference would not be sufficiently obvious to enable price discrimination.

(209) Fourth, according to the Notifying Party, in some instances, the SO fails to recognise specific features of local supply patterns. Therefore, even where the regional cluster is used to define the scope of the geographic market, it is important that the competitive assessment takes full account of such features in order to ensure that a rigid application of the local cluster approach does not result in an inaccurate assessment of the competitive conditions in a specific area.

7.1.3. The Commission's assessment

(210) The Commission has reached the conclusion that, for the purpose of assessing the Transaction, the relevant geographic markets consist of regional clusters of 700 km catchment areas around individual customer filling locations (hereinafter 'regional clusters') where the competitive conditions are sufficiently homogeneous. Those regional clusters comprise individual customer filling locations located in each of the following regions: (i) the UK and Ireland; (ii) the Benelux (Belgium, Luxembourg and the Netherlands); (iii) Central Europe (Austria and Germany); (iv) France; (v) Italy; (vi) Iberia (Spain and Portugal); (vii) North-East Europe (Czech Republic, Slovakia, Poland, Lithuania, Estonia and Latvia); (viii) South-East Europe (Hungary, Slovenia, Croatia, Romania, Bulgaria, Greece and Cyprus); and (ix) the Nordics (Denmark, Norway, Sweden, Finland and Iceland). To the extent that some degree of differentiation exists between the catchment areas around individual customer filling locations in a given region, such differentiation is taken into account in the competitive assessment.

(211) As discussed in this section, the supply of beverage cans in the EEA is characterised by the fact that: (i) national barriers do not materially affect trade patterns; (ii) prices are generally set on location-by-location basis in individual negotiations with customers; (iii) customers attach importance to security of supply, rapid and just-in-time deliveries, good service levels; and (iv) transport distances and costs are an important factor influencing competitive dynamics.

(212) The methodology for the definition of the relevant geographic markets is described in section 7.1.3.5 below.

7.1.3.1. Lack of national barriers

(213) The Commission has reached the conclusion that it is not appropriate to segment the relevant geographic markets along national lines.

(214) First, the sales of the Parties and of their competitors indicate that national barriers have a limited influence on the sourcing of beverage cans. For instance, over 30 % of beverage can demand in Italy is imported. Similarly, over 60 % of can demand in the Netherlands is satisfied with cans produced in Germany and in the UK.

(215) The same is true regarding imports from non-EEA countries. For instance close to 50 % of cans sold in Slovenia are produced in Serbia and over 50 % of cans sold in Lithuania are produced in Russia.

Agreed non-confidential minutes of a call with a customer, 18.08.2015, ID 2866. Also, 'the better the footprint of a can manufacturer matches with [customer's] locations, the better the DDP price' and further: 'There is no fixed conversion price across Europe. […] The conversion price depends on several underlying factors, but competition in a given region is the most important one. For instance in Scandinavia (meaning Norway, Denmark and Sweden), where the competition is marginal, conversion costs are the highest. In Finland, where Can-Pack challenges Rexam, conversion costs are much lower'. Agreed non-confidential minutes of a call with a customer, 25.08.2015, ID 2724. See also agreed non-confidential minutes of a call with a customer, 18.08.2015, ID 2866: '[The conversion cost] differs mainly by the country where the beverage can plant is located and might differ by filling location. […] We see cases where can suppliers inflate their conversion price because they know that the alternative suppliers are farther away from their filling locations. As such, they can mark-up their conversion price due to low logistic costs and still remain below the price offered by competitors (competitors potentially having lower conversion price, but higher logistic costs)'.

The Commission compared total (can body plus end) prices across the regions by estimating a fixed effects regression of total prices on regional dummies. The fixed effects were defined as all possible combinations of can size, can size category (sleek, slim, standard etc.), metal type, product enhancement, year, manufacturer and freight cost type (delivered or non-delivered).

(220) Third, customers that have only one filling location or filling locations close to one another indicated that they pay a 'local' price which can be different from that prevalent in other parts of the EEA. As a customer indicates, 'different prices in different regions for the same supplier, expressly Rexam and Ball depending the number of competitors they have in a region. The more they have the low is the price they get into the market'. Or as another customer puts it, '[t]erms are negotiated based on competitive situation on a given market. e.g. [customer] gets better pricing in market where there is competition than where there are monopolies or duopolies'.

(221) Fourth, beverage can manufacturers can set higher prices in catchment areas where there is less competition. According to a customer, '[there are] cases where can suppliers inflate their conversion price because they know that the alternative suppliers are farther away from their filling locations. As such, they can mark-up their conversion price due to low logistic costs and still remain below the price offered by competitors (competitors potentially having lower conversion price, but higher logistic costs).'

(222) While Can-Pack's prices are at the lower end of the spectrum and Ball's at the higher end, Rexam's prices are somewhere in the middle.

(223) Fifth, the pricing behaviour of both Ball and Rexam takes into account the degree of competition in a given catchment area.

(224) Rexam has confirmed that it takes a […] approach. Market pricing in the relevant area and locational advantages of competitors around customer filling locations are key drivers of competition.

(225) Rexam is also developing […].

(226) Ball uses […] ([…] ) […].[…]

(227) Ball has also developed the […]

(228) […]

(229) […]

(230) On the basis of an understanding of local market conditions, the […] has clustered customer filling locations in and around the EEA into the following 9 regions:

-– […];

(231) Security of supply, rapid and just-in-time deliveries and good service levels have been identified by customers as key concerns in their relationship with beverage can manufacturers. Furthermore, several customers have pointed out that by shortening the distance between customers and suppliers it is easier for can manufacturers to ensure that their customers receive their required amounts of cans on time. Short distances between suppliers and customers have also been identified as a factor facilitating customer support.

(232) Beverage manufacturers generally hold limited can stock at their filling stations and customers, in particular the larger ones, may have to receive several can shipments per day and expect cans to be delivered shortly before they are needed in the production lines. As one large customer explained '[j]ust-in-time delivery is crucial. Cans can only be delivered 3-4 hours in advance before being placed in the filling line. This is usually easier to plan if a supplying can plant is within close proximity'.

(233) However, certain other customers indicated that, while closeness to a can manufacturer is desirable, cans may travel longer distances provided that inventories are managed correctly and that effective logistics are deployed. Can manufacturers could also invest in additional warehousing near a customer filling location to overcome concerns related to just-in-time deliveries. Furthermore, some customers have indicated that the actual origin of the cans they receive is not important as long as the delivered price is competitive. Finally, in peak season, it is not uncommon for beverage can manufacturers to make deliveries from plants that are located further away from customers than those from which they normally supply them.

(234) On balance, it appears that while closeness to the filling location is a desirable attribute in a beverage can manufacturer as it facilitates logistics and support, it is not a pre-condition to entering into a customer/supplier relationship. This conclusion is confirmed by the fact that, as explained in section 7.1.3.4 below, on average approximately 80 % of can sales in the EEA are made within a maximum radius of approximately 700 km of the production facility.

7.1.3.4. Transport costs

(235) Transport costs confirm that competitive conditions are largely driven by the number of alternative beverage can manufacturers that are sufficiently close to be able to competitively supply to each customer filling location the bulk of its can requirements.

(236) While customers have in principle no objection to sourcing cans over long distances this does not happen in practice, at least for the bulk of their can requirements. This is because cans are relatively voluminous goods and transport costs increase as the distance between the can plant and customer filling location increases. For example, the cost of transporting cans 1 000 to 1 500 km amounts on average to approximately […]-[…] % (for Ball) and […] % (for

'[Customer] does not give any importance to the proximity of its suppliers' plants, the only thing that really matters is the price of the cans, [customer] even ignores from which plant exactly its cans come from.' Courtesy translation from the French original: '[Client] n'accorde aucune importance à la proximité des usines de ses fournisseurs, ce qui importe est seulement le prix d'achat des canettes. Pour certains de ses fournisseurs, [client] ignore même de quelles usines exactement proviennent ses canettes.' Agreed non-confidential minutes of the call with a customer, 12.08.2015, ID 1800.

'Ball for instance has recently had huge problems with capacity; it used its network to balance it out. In […], Ball delivered to [customer] not from its […] plant, but also from […] due to capacity constraints.' Agreed non-confidential minutes of a call with a customer, 25.08.2015, ID 2724. '[Customer] was simply not able to receive cans on time. As a result, [Customer] could not produce certain types of products as much as it planned to. Rexam even had to import marginal volumes from […] in order to face the shortage', agreed non-confidential minutes of a call with a customer, 25.08.2015, ID 2691.

'There is no fixed conversion price across Europe. […] The conversion price depends on several underlying factors, but competition in a given region is the most important one. For instance in Scandinavia (meaning Norway, Denmark and Sweden), where the competition is marginal, conversion costs are the highest. In Finland, where Can-Pack challenges Rexam, conversion costs are much lower', agreed non-confidential minutes of a call with a customer, 25.08.2015, ID 2724; and '[The conversion cost] differs mainly by the country where the beverage can plant is located and might differ by filling location. […] We see cases where can suppliers inflate their conversion price because they know that the alternative suppliers are farther away from their filling locations. As such, they can mark-up their conversion price due to low logistic costs and still remain below the price offered by competitors (competitors potentially having lower conversion price, but higher logistic costs)', agreed non-confidential minutes of a call with a customer, 18.08.2015, ID 2866; and 'different prices in different regions for the same supplier, expressly Rexam and Ball depending the number of competitors they have in a region. The more they have the low is the price they get into the market', Questionnaire to Customers (Q1), question 96.2.1, ID 1609.

'Supplying from further than 900km would only be possible provided that the security of supply remains constant and that the price does not increase too much. In practice this is unlikely to happen', agreed non-confidential minutes of a call with a customer, 25.08.2015, ID 2691. 'In general, feasibility of deliveries depends rather on the DDP price than distance, which is not an issue per se', agreed non-confidential minutes of a call with a customer, 25.08.2015, ID 2724. '[Customer] does not impose a specific distance between the supplier's plants and its own filling locations', agreed non-confidential minutes of a call with a customer, 02.09.2015, ID 2888. Nonetheless, this customer received a quote from a player not producing in the EEA and considered this as an 'opportunistic quotation that was not viable (it was ex-works prices)', agreed non-confidential minutes of a call with a customer, 02.09.2015, ID 2888.

Rexam) of the total can price depending on the can size. In comparison, the cost of transporting cans less than 300 km amounts on average to […]-[…] % (for both Ball and Rexam).

Figure 12: Weighted average transport cost of Ball and Rexam in 2014

[…]

Source: Reply to the Article 6(1)(c) Decision

(237) Customers and can manufacturers therefore actively try to minimise transport distances and thus transport costs. Minimising travel distances allows manufacturers to offer competitive prices without affecting profit margins and customers to benefit from attractive price and non-price conditions (such as shorter delivery times, facilitated just-in time deliveries, and easier interactions with the can manufacturers e.g. in relation to quality issues or to the development of new products).

(238) Exceptions are slim and sleek cans, which can travel further than the standard 33 cl and 50 cl cans. There are fewer plants producing sleek or slim cans than standard cans and these plants often produce sleek or slim cans only for short periods of time each year.

(239) Another exception is when beverage can manufacturers provide 'freight cost contributions' – special discounts on total delivered price – in order to boost their price competitiveness in areas located remote from their can manufacturing sites. These are often strategic decisions to grow the beverage can manufacturer's footprint with a customer in geographical areas where the beverage can manufacturer is not present with a can plant and which offer high can volume growth potential. For instance, Rexam has often been prepared to supply cans over greater than usual distances.

7.1.3.5. Regional clusters of 700 km catchment areas around individual customer filling locations

(240) As discussed in sections 7.1.3.1-7.1.3.4, the trade of beverage cans in the EEA is characterised by the following features: (i) supply and demand conditions differ from one location to the next; (ii) customers tend to make most of their purchases within a maximum distance from their filling location plants; (iii) suppliers normally deliver the cans to customers’ locations; and (iv) prices are generally negotiated at the level of filling plants.

(241) These features allow beverage can manufacturers to price discriminate among customers based on their location and customers, in particular small customers, have limited possibilities to source cans at better conditions than those prevailing in the area in which they are based.

(242) The Commission, therefore, considers appropriate to take a customer-centric approach to the definition of the geographic market and define geographic markets on the basis of clusters of

catchment areas around individual customer filling locations where the competitive conditions are sufficiently homogeneous.

(243) Catchment areas around individual customer filling locations reflect the geographic differentiation between customers that is a characteristic of a market where proximity and the ability to price discriminate across customers play an important role in the competitive landscape. To the extent that some degree of differentiation exists between the catchment areas around individual customer filling locations in a given region, such differentiation is taken into account in the competitive assessment.

- The radius of catchment areas around individual customer filling locations

(244) The radius of these catchment areas should be sufficiently wide to capture all suppliers that can competitively supply customers with the bulk of their requirements.

(245) The Commission, therefore, analysed the travel distance at which sales are made by the four main beverage can manufacturers (Ball, Rexam, Crown and Can-Pack) to customers in the EEA on the basis of the actual sales data of the Parties, Crown and Can-Pack and not, contrary to what the Notifying Party argues, only the sales of the Parties.

(246) These delivery distances represent the maximum distance at which on average Ball, Rexam, Crown and Can-Pack make 80 % or 90 % of their sales in the EEA.

Table 4: Maximum can transport distance in the EEA and non-EEA countries […]in 2014 (km)

Maximum transport distance

Company

70 % Volume 80 % Volume 90 % Volume

Ball […] […] […]

Rexam […] […] […]

Crown […] […] […]

Can-Pack […] […] […]

EEA 486 626 922

Source: Parties' actual sales data and the Parties' estimated sales data for competitors

(247) When actual 2012-2014 sales data from the Parties and their competitors is taken into account, in the EEA, 80 % of all can sales take place within a maximum travel distance of approximately 700 km from a customer's filling location.

(248) The Commission considers that catchment areas with a radius that represents the maximum travel distance at which 80 % of beverage can sales take place adequately capture the prevailing competitive dynamics around each customer. This is for the reasons set out in recitals (249)-(254).

(249) First, while certain respondents to the market investigation indicated that they source cans only within a radius of 500 km or less, other respondents indicated that they source cans at distances significantly greater than 500 km. A majority of customers that responded to the market investigation also indicated that they could even source cans from 900 km or more if delivered prices (including transport cost) were 10 % lower than those of their current supplier(s), subject to all other trading conditions (such as quality, reliability, capacity) being equal.

(250) Second, the definition of 700 km catchment areas around customer filling locations that capture 80 % of current sales may actually understate the size of the appropriate catchment areas for the purposes of defining the relevant geographic market.

(251) In the first place, given the high variable margins of the Parties (see Figure 11 and recital (320)), manufacturers are likely to be able to supply beverage cans competitively at a maximum distance of more than 700 km.

(252) In the second place, if faced with a price increase, customers are likely to be willing to source beverage cans further away than they do today. As set out in recital (249), the majority of customers that responded to the market investigation indicated that they would be willing to source cans from 900 km or more if delivered prices (including transport costs) were 10 % lower than those of their current supplier(s), subject to certain conditions.

(253) In the third place, as indicated in Table 4, Can-Pack, which is the beverage can manufacturer that has the longest delivery distances, makes 80 % of its sales over a distance of approximately […] km.

(254) In the fourth place, whilst the 80 % of sales exclude the 20 % of sales that travel the furthest, they include sales that travel short distances (such as wall-to-wall sales). These short distance sales are not representative of the ability of customers to procure competitively over relatively long distances.

(255) Based on the data, the Commission therefore concludes that the radius of the catchment areas around individual customer filling locations should be 700 km.

- The clustering of the catchment areas around individual customers filling locations

(256) For the purpose of assessing the Transaction, catchment areas around individual customer filling locations that are subject to sufficiently homogeneous competitive conditions can be clustered together in broader geographic areas.

(257) For the reasons set out in this section, the Commission considers that competitive conditions are sufficiently homogenous for customer filling locations located in the following regions: Central Europe (Austria and Germany), Benelux, France, Italy, Iberia (Spain and Portugal),

Questionnaire to Customers (Q1), questions 42 and 43, ID 1609. Questionnaire to Customers (Q2), question 25, ID 2664 and Questionnaire to Customers (Q4), question 11, ID 2786. In particular, 63 % of respondents to Q2 (non-confidential replies) and 51 % of respondents to Q4 (non-confidential replies) would be willing to source from further away than 900 km.

North-East Europe (Poland, Czech Republic, Slovakia, Lithuania, Estonia and Latvia), South East Europe (Hungary, Slovenia, Croatia, Romania, Bulgaria, Greece and Cyprus), the Nordics (Denmark, Norway, Sweden, Finland, and Iceland) and the UK and Ireland.

(258) First, the […] groups together customers in this manner, on the basis of Ball's understanding of local market conditions.

(259) Given that the […] aims at […], it is a logical starting point for the Commission’s assessment of the homogeneity of individual catchment areas.

(260) Second, the sales data of the Parties and of their competitors indicates that plants located within a given regional cluster make the vast majority of their sales within the cluster (between 71 % and 97 %).

(261) Third, this clustering is supported by certain qualitative elements from the market investigation. For example, France has introduced a ban on Bisphenol A ('BPA'). Imports of beverage cans from other EEA countries that use BPA do not, therefore, represent a competitive constraint on beverage cans produced in France.

(262) Fourth, steel cans are produced and used in some geographic areas but not in others, for instance there is no steel beverage can production facility in the Nordics region.

(263) At the same time, due to the local nature of competition and the important role that distance plays in the competitive conditions faced by customers, certain regional clusters include customer catchment areas that are more affected by the Transaction than others based in the same cluster. The Commission has thus taken into account geographical differentiation in its competitive assessment (thus also taking into account geographical closeness of competition between the merging parties), together with the specific features of each regional cluster.

7.2. Aluminium bottles

7.2.1. Past decisional practice

(264) The Commission has not previously investigated the scope of the relevant geographic market for aluminium bottles in past decisions.

7.2.2. The Notifying Party's views

(265) The Notifying Party submits that, since aluminium bottles fall within a broader market for beverage packaging, the scope of the geographical market is the same as for cans and therefore regional with some inner-regional differences and inter-regional trade flows at the edges of some of the regional borders.

7.2.3. The Commission's assessment

(266) There are a number of reasons why the relevant market for aluminium bottles may be broader than EEA-wide.

(267) First, customers source aluminium bottles both at EEA and global level.

(268) Second, aluminium bottles are only produced at a limited number of manufacturing plants in the EEA. IE aluminium bottles are produced by Alucan at its plant near Barcelona (Spain), by Ardagh at its Beaurepaire (France) plant and by Ball at its Velim (Czech Republic) plant. In the EEA, DWI aluminium bottles are solely produced by Rexam at its Ejpovice (Czech Republic) plant.

(269) Third, because aluminium bottles are a niche product and significantly more expensive than beverage cans, longer transport distances appear more economical. This view finds support, for example, by a customer looking into the possibility to source aluminium bottles for the EEA from a manufacturer in China or other non-EEA countries.

(270) The Commission considers, however, that for the purposes of the present Decision the exact geographic market definition for aluminium bottles can be left open since the Transaction would not lead to a significant impediment of effective competition under any plausible geographic market definition.

8. COMPETITIVE ASSESSMENT: PRINCIPLES

(271) According to the Commission's Guidelines on the assessment of horizontal mergers under the Council Regulation on the control of concentrations between undertakings (the 'Horizontal Merger Guidelines'), there are two main ways in which horizontal mergers may significantly impede effective competition, in particular by creating or strengthening a dominant position:

– by eliminating important competitive constraints on one or more firms, which consequently would have increased market power, without resorting to coordinated behaviour (non-coordinated effects); and/or

– by changing the nature of competition in such a way that firms that previously were not coordinating their behaviour are now significantly more likely to coordinate and raise prices or otherwise harm effective competition. A merger may also make coordination easier, more stable or more effective for firms which were coordinating prior to the merger (coordinated effects).

8.1. Non-coordinated effects

(272) Generally, a merger giving rise to non-coordinated effects would significantly impede effective competition by creating or strengthening the dominant position of a single firm, one which, typically, would have an appreciably larger market share than the next competitor post-merger.

(273) According to well-established case law, very large market shares – 50 % or more – may in themselves be evidence of the existence of a dominant market position.

(274) The overall concentration level in a market may also provide useful information about the competitive situation. In order to measure concentration levels, the Commission often applies the Herfindahl-Hirschman Index ('HHI'). The HHI is calculated by summing the squares of the individual market shares of all the firms in the market. While the absolute level of the HHI can give an initial indication of the competitive pressure in a market post-merger, the change in the HHI (known as the 'delta') is a useful proxy for the change in concentration directly brought about by a merger. The Commission is unlikely to identify horizontal competition concerns in a market with a post-merger HHI below 1 000. Such markets normally do not require extensive analysis. The Commission is also unlikely to identify horizontal competition concerns in a merger with a post-merger HHI between 1 000 and 2 000 and a delta below 250, or a merger with a post-merger HHI above 2 000 and a delta below 150, except where special circumstances are present.

(275) Furthermore, non-merging firms in a given market can benefit from the reduction of competitive pressure that can result from a merger, since any price increase by merging firms may switch some demand to rival firms, which, in turn, may find it profitable to increase their prices.

(276) Mergers in oligopolistic markets involving the elimination of important competitive constraints that the parties previously exerted upon each other together with a reduction of competitive pressure on the remaining competitors may, even where there is little likelihood of coordination between the members of the oligopoly, also result in a significant impediment to competition. The Merger Regulation clarifies that all mergers giving rise to such non-coordinated effects shall also be declared incompatible with the internal market.

(277) The larger the market share, the more likely a firm is to possess market power. Moreover, the larger the addition of market share, the more likely it is that a merger will lead to a significant increase in market power. The larger the increase in the sales base on which to enjoy higher margins after a price increase, the more likely it is that the merging firms will find such a price increase profitable despite the accompanying reduction in output.

(278) In evaluating the likelihood of non-coordinated effects potentially caused by a merger, it is important to assess to which extent the products of one merging party are close substitutes to the products sold by the other merging party. The merging firms' incentive to raise prices is more likely to be constrained when rival firms produce close substitutes to the products of the merging firms than when they offer less close substitutes.

(279) Customers may have difficulties in switching to other suppliers where only few alternative suppliers exist or where the customers would face substantial switching costs. Such customers are particularly vulnerable to price increases. In particular, this may be the case for customers that have used dual sourcing from the two merging firms as a means of obtaining competitive prices.

(280) The Commission is unlikely to find that the merger will create or strengthen a dominant position or otherwise significantly impede effective competition when rival firms have available capacity and find it profitable to expand output sufficiently. In other words, the extent to which competitors to the merged entity constrain the merged entity from raising prices not only depends on the level of their spare capacity but also on whether these firms have the incentive to react aggressively to a post-merger price increase.

(281) Some proposed mergers would, if allowed to proceed, significantly impede effective competition by leaving the merged firm in a position where it would have the ability and incentive to make the expansion of smaller firms and potential competitors more difficult or otherwise restrict the ability of rival firms to compete.

(282) In markets where innovation is an important competitive force, a merger may increase the firms' ability and incentive to bring new innovations to the market and, thereby, the competitive pressure on rivals to innovate in that market. Alternatively, effective competition may be significantly impeded by a merger between two important innovators, for instance between two companies with 'pipeline' products related to a specific product market.

(283) Potential competition must also be taken into account in any competitive assessment. Potential competition may lead to negative competition effects where (i) the potential competitor already exerts a significant constraining influence or where there is a significant likelihood that it would grow into an effective competitive force (for example plans to enter a market in a significant way) and (ii) there is an insufficient number of other potential competitors, which could maintain sufficient competitive pressure after the merger.

(284) Countervailing buyer power should be understood as the bargaining strength that the buyer has vis-à-vis the seller in commercial negotiations due to its size, its commercial significance to the seller and its ability to switch to alternative suppliers. One source of countervailing buyer power would be if a customer could credibly threaten to resort, within a reasonable timeframe, to alternative sources of supply should the supplier decide to increase prices or to otherwise deteriorate quality or the conditions of delivery. This would be the case if the buyer could immediately switch to other suppliers, credibly threaten to vertically integrate into the upstream market or to sponsor upstream expansion or entry for instance by persuading a potential entrant to enter by committing to placing large orders with this company.

Countervailing buyer power is, however, only possessed by a customer if it is of a certain size, it is of great commercial significance to the supplier and it has the ability to switch to alternative suppliers. Moreover, even if some customers were to hold such countervailing buyer power, this is not sufficient. In addition, it is not enough that sufficient countervailing buyer

Horizontal Merger Guidelines, paragraph 31.

Horizontal Merger Guidelines, paragraph 33.

Horizontal Merger Guidelines, paragraph 36.

Horizontal Merger Guidelines, paragraph 38.

Horizontal Merger Guidelines, paragraph 60.

Horizontal Merger Guidelines, paragraph 64.

Horizontal Merger Guidelines, paragraph 65.

EN 40 EN

power may exist pre-merger, it must continue to exist post-merger (removing a credible alternative supplier may reduce buyer power).

For entry to be considered a sufficient competitive constraint on the merging parties, it must be shown to be likely, timely and sufficient to deter or defeat any potential anti-competitive effects of the merger.

8.2. Coordinated effects

In some markets the structure may be such that firms would consider it possible, economically rational, and hence preferable, to adopt on a sustainable basis a course of action on the market aimed at selling at increased prices through a coordination of their behaviour. A merger may increase the likelihood that firms are able to coordinate their behaviour and raise prices, even without entering into an agreement or resorting to a concerted practice within the meaning of Article 101 TFEU. A merger may also make coordination easier, more stable or more effective for firms that were already coordinating before the merger, either by making the coordination more robust or by permitting firms to coordinate on even higher prices.

Coordination may take various forms, such as setting prices above the competitive level, limiting production or capacity, or dividing the market, for instance by geographic area or other customer characteristics, or allocating contracts in bidding markets.

Coordination is more likely to emerge in markets where it is relatively simple to reach a common understanding on the terms of coordination. In addition, three conditions are necessary for coordination to be sustainable. First, the coordinating firms must be able to monitor to a sufficient degree whether the terms of coordination are being adhered to. Second, discipline requires that there is some form of credible deterrent mechanism that can be activated if deviation is detected. Third, the reactions of outsiders, such as current and future competitors not participating in the coordination, as well as customers, should not be able to jeopardise the results expected from the coordination. In applying those criteria, it is necessary to avoid a mechanical approach involving the separate verification of each of those criteria taken in isolation, while taking no account of the overall economic mechanism of a hypothetical tacit coordination.

For competitors to reach a common perception as to how the coordination should work coordinating firms should have similar views regarding which actions would be considered to be in accordance with the aligned behaviour and which actions would not. In particular, coordination by way of market division will be easier if customers have simple characteristics that allow the coordinating firms to readily allocate them. Such characteristics may be based on geography; on customer type or simply on the existence of customers who typically buy from one specific firm. Coordination by way of market division may be relatively straightforward if it is easy to identify each customer's supplier and the coordination device is the allocation of existing customers to their incumbent supplier.

Publicly available key information, exchange of information through trade associations, or information received through cross-shareholdings or participation in joint ventures may also help firms reach terms of coordination.

Firms may find it easier to reach a common understanding on the terms of coordination if they are relatively symmetrical, especially in terms of cost structures, market shares, capacity levels and levels of vertical integration. Structural links such as cross-shareholdings or participation in joint ventures may also help in aligning incentives among the coordinating firms.

Coordinating firms are often tempted to increase their share of the market by deviating from the terms of coordination, for instance by lowering prices, offering secret discounts, increasing product quality or capacity or trying to win new customers. Only the credible threat of timely and sufficient retaliation keeps firms from deviating. Markets therefore require sufficient transparency to allow the coordinating firms to monitor to a sufficient degree whether other firms are deviating, and thus know when to retaliate. Transparency in the market is often higher when the number of active participants in the market is lower. Further, the degree of transparency often depends on how market transactions take place in a particular market.

In some markets where the general conditions may seem to make monitoring of deviations from coordination difficult, firms may nevertheless engage in practices which have the effect of easing the monitoring task, even when these practices are not necessarily entered into for such purposes. These practices, such as meeting-competition or most-favoured-customer clauses, voluntary publication of information, announcements, or exchange of information through trade associations, may increase transparency or help competitors interpret the choices made. Cross-directorships, participation in joint ventures and similar arrangements may also make monitoring easier.

The possibility of retaliation is important for the stability of coordination. Coordination is not sustainable unless the consequences of deviation are sufficiently severe to convince coordinating firms that it is in their best interest to adhere to the terms of coordination. It is thus the threat of future retaliation that can keep the coordination sustainable. However the threat is only credible if, where deviation by one of the firms is detected, there is sufficient certainty that some deterrent mechanism will be activated.

Retaliation need not necessarily take place in the same market as the deviation. If the coordinating firms have commercial interaction in other markets, these may offer various methods of retaliation.

As regards the specific effects of a merger, a merger may increase the likelihood that firms are able to coordinate their behaviour and raise prices, even without entering into an agreement or resorting to a concerted practice within the meaning of Article 101 TFEU. A merger may also make coordination easier, more stable or more effective for firms that were already coordinating before the merger, either by making the coordination more robust or by permitting firms to coordinate on even higher prices, for example by facilitating the detection of deviation, limiting the ability and incentives of some market players to deviate and allowing more efficient retaliation.

9. COMPETITIVE ASSESSMENT BEVERAGE CANS: NON-COORDINATED EFFECTS

9.1. Overview of the market dynamics and effects of the Transaction across the regional clusters

Despite the local/regional nature of the geographic markets, competition has a strong EEA-wide element.

Certain features affecting the competitive assessment in each relevant market are similar, if not identical, across all relevant markets within the EEA. These features include: (i) the lack of intensity of competition between beverage can manufacturers; (ii) the fact that the Transaction reduces the number of players from 4 to 3; (iii) the closeness of competition between the Parties; (iv) the inability of the four beverage can manufacturers to compete on an equal footing; (v) differences between large customers and other customers; (vi) the importance of having a network of plants; (vii) the Parties' role as main innovators; (viii) the tightness of capacity in the relevant geographic markets; (ix) the inability of customers to switch; (x) the need for customers to multi-source; (xi) the irrelevance of preferred supply-agreements and wall-to-wall supplies; (xii) the limited competitive pressure from other forms of packaging; (xiii) high barriers to entry and expansion; and (xiv) insufficient countervailing buyer power.

Given the similarity in the way competition takes place in each regional cluster, the Commission will first describe the common features that currently affect competitive conditions across the EEA (sections 9.1.1 to 9.1.16).

The Commission will then assess the impact on competition of the Transaction in each regional cluster in section 9.2 below.

Judgment in Bertelsmann and Sony Corporation of America v Impala, C-413/06 P, EU:C:2008:392, paragraph 122; Horizontal Merger Guidelines, paragraph 39.

Horizontal Merger Guidelines, paragraph 42.

EN 43 EN

9.1.1. The position of the Parties across the EEA

9.1.1.1. The Parties are the two largest beverage can manufacturers in the EEA

As set out in Table 5, the Parties are the two largest beverage can manufacturers in the EEA, their sales volumes representing respectively [30-40] % (Rexam) and [30-40] % (Ball) of the total EEA-volume in 2014.

Table 5: Sales and capacity shares of beverage can manufacturers (in volume) in 2014 (all sizes, aluminium and steel cans) as presented by the Notifying Party

EEA Sales Capacity

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

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

Combined [60-70] % [60-70] %

Crown [10-20] % [20-30] %

Can-Pack [10-20] % [10-20] %

Other [0-5] % [0-5] %

Source: Form CO

Rexam is the largest player in the EEA in terms of both sales volumes and capacity. Ball is the second largest, followed by Crown and Can-Pack.

Rexam currently sells more than […] the volumes of the third largest player Crown, and about […] the volumes of the fourth player, Can-Pack. Rexam itself publicly states on its website that 'Europe is the largest of our regional markets and we are the leading can maker with more than 40% market share.'

Ball currently sells close to […] % more volumes than Crown and sells more than […] the volumes of Can-Pack.

The merged entity would be more than […] times larger in terms of sales volumes than the current third largest player, Crown. As regards Can-Pack, the merged entity would be more than […] times larger in terms of sales volumes.

The merged entity would be the largest player in terms of sales volumes in the EEA

The merged entity would be by far the largest player in the EEA, accounting for [60-70] % of sales volumes, and the increment the Transaction would bring about is significant ([30-40] %).

Furthermore, the market position of the Parties, Crown and Can-Pack in the EEA, despite a slight increase of Can-Pack's market position since 2012, has remained stable over the last years, as evidenced by Figure 13. This is also confirmed by the Parties' internal documents and by the responses to the market investigation.

Figure 13: Shares of supply of beverage cans EEA 2010-2014

[…]

Source: Form CO, Annex 21

The Parties expect the market to remain stable in the future, as illustrated in Figure 14.

Figure 14: Rexam's projections for European shares of supply of beverage cans 2015-2017

Source: Rexam's internal document, 'Can-Pack view in BCE Strategic Plan 2014-16', slide 5

The merged entity would be the largest player in terms of total capacity in the EEA

310The merged entity would control [60-70] % of overall capacity in the EEA (roughly […] out of […] billion cans), and the increment the Transaction would bring about is [20-30] %. This has a number of consequences.

First, Crown and Can-Pack, which have a joint capacity of around […] billion cans, could serve only a fraction of the […] billion demand for beverage cans in the EEA. By contrast, the merged entity would be able to serve about […] % of the demand.

Second, the Parties' combined capacity underestimates their competitive position in the EEA for several reasons.

In the first place, it does not include Ball's Belgrade plant, which, albeit located outside the EEA, serves […] the EEA (see recital (53) above).

In the second place, it also does not take into account future expansion plans (see section 5.1.4). For example, Rexam's new Widnau plant, which started production in October 2015, will […] ([…] billion cans) in the EEA despite the fact that it is not located in the EEA, as Rexam plans to […].

The Transaction would lead to a further increase in concentration levels across the EEA

315The supplier base in the EEA for beverage cans is already concentrated. Only four competitors are active, apart from the vertically integrated Polish player BagPak which cannot be considered as being active in the merchant market (see recital (50)).

316The Transaction would give rise to a high post-Transaction HHI level and HHI delta, at EEA level, as shown in Table 6. These HHI level and delta are well above the levels at which the Commission would be unlikely to identify horizontal competition concerns.

317This analysis is useful inasmuch as it helps characterise competitive conditions and the choices left to customers post-Transaction in the EEA. This analysis will also be carried out for each relevant market.

Table 6: HHI levels based on volume figures 2014 as presented by the Notifying Party

EEA

Pre-transaction HHI [2500-3000]

Post-transaction HHI [5000-5500]

HHI Delta [2000-2500]

Source: Reply to RFI 2

9.1.2. Intensity of competition pre-Transaction

9.1.2.1The Notifying Party's views

The Notifying Party claims that competition is strong, including among the members of the Beverage Can Makers Europe ('BCME') trade association (Ball, Rexam and Crown), as evidenced by the material fluctuation of the beverage can manufacturers' volume shares in the various regions over the past few years.

9.1.2.2The Commission's assessment

The Commission has reached the conclusion that pre-Transaction competition for the supply of beverage cans in the EEA cannot be characterised as strong. This is for several reasons.

First, the variable margins of the Parties (approximately between […] % and […] %) are high. These margins are even higher if they are calculated based on the price excluding the LME-hedged aluminium cost.

Second, the Parties' pricing behaviour indicates the existence of localised market power.

As detailed in section 7.1.3.2, both Parties, when deciding what price to offer for each individual customer filling location […]. In particular both Parties are […] to […].

Ball uses its […] as key elements of its pricing strategy for each customer.

Rexam is […], which will allow it to assess its […].

Third, customers describe the market as one with limited free capacity available in each production plant and where potential switches of large volumes will be avoided as much as possible. One customer pointed out that beverage can manufacturers 'know that stability is in the interest of their market'. Some respondents to the market investigation described the beverage can sector in the EEA as with a somewhat limited intensity of competition: 'We believe there is very limited competition between the beverage can manufacturers'; 'the intensity of competition is already very low amongst the four suppliers'.

Fourth, the aggregated country level information about historical beverage can fillings by Ball, Rexam and Crown that the BCME trade association collects and shares might allow them to have good information about the identity of each other’s customers and allow them to observe each other’s actions. This was emphasised by certain customers that responded to the market investigation:

'The quarterly BCME report states available capacities'; 'Beverage Can Makers Europe (BCME) publish data concerning can making capacities'; 'Information von BCME'; 'The BCME sends a report on an annual basis where they give an indication of the total capacities per plant. Obviously, they do not share the overcapacity or capacity shortage per plant, but they do give an estimate of the total demand per country and the development over the last five years divided between beer and carbonated soft drinks.'

'The suppliers are openly telling about sales, capacity, capex etc.' 'You can [access] this Information via Internet. Can supplier give Information about this.' 'We get good information t[h]rough Meetings With different Suppliers.'

'[T]he can manufacturers know the market very well and can perfectly judge which production plant of which manufacturer has capacity left open and which ones do not.'

Fifth, certain customers that responded to the market investigation indicated a general lack of interest, in particular of Ball, Rexam and Crown, in acquiring new customers. A large beverage producer described that '[d]uring [its] recent tender Rexam did not try to compete with Ball and Crown in France, Switzerland and Greece, as Ball and Crown have plants better situated to serve these markets. The same for Ball and Crown, who did not try to compete with Rexam on the Scandinavian markets.'

'Crown's perceived positioning is explained by a customer as being due to the fact 'that Crown is more focused on markets outside the EEA, like Asia and America.' Overall, as another customer pointed out no beverage can manufacturer 'is very ag[g]ressive to get more volumes.'

These statements are in line with indications stemming from the Parties' internal documents. For example, […], while Can-Pack has been a more […] player. In analysing their options to […] with a plant, Ball notes that […].

The perception that certain customers have of Can-Pack is somewhat different. They view Can-Pack as a player that 'broke [the] ranks' and made efforts to gain new customers and new volumes even though it may not always have been ideally located to do so. Can-Pack's more aggressive pricing strategy brought some movement in the market and motivated the other players, in particular Rexam, to compete more fiercely. As one major customer notes, '[h]istorically, prior to 2011, all members of the BCME (Ball, Rexam and Crown) had been pricing cans within Europe based on the highest cost country, being Germany. BCME members th[e]n began to come under pressure in Eastern Europe when Can-Pack invested in Poland and other markets like Romania. The BCME came under further pressure when Can-Pack invested in UK. In [customer's] view, Rexam then decided to break out from the BCME model and started to compete head to head with Can-Pack (and the BCME members). Ball has during this period (2011-2015) remained the least competitive supplier overall for [customer] in EEA.'

329Other customers described Can-Pack as follows: 'Big 3 seem to have similar approach, where [Can-Pack] opens the market'. 'Before Can Pack started the business and started to be ag[g]ressive on the market I never got an offer from Crown or Ball. Just have a look at the history and then you can see that these 3 players did split the market/countries between themselves.' 'The three BCME-members have known each other for a long time. They all aim for stability in the market and a good alignment between demand and production capacity. Can-Pack has entered the market much later. They aim for volume growth, which obviously goes against the objectives of the other three, but they also wish to benefit from healthy margins and conditions, since their global growth strategy requires significant investments.' 'Can-Pack growth into the European can supply market: Pricing pre-2010 was significantly higher prior to the growth of Can-Pack in the region. Can-Pack had a positive effect on driving the overall can price level down, with other suppliers such as Rexam eventually following these more competitive prices.' 'The market changed when Can Pack entered the market. My prices would be 10 - 20% higher than today if they didn't exist on the market.' 'Can Pack entered

the market and started to be ag[g]ressive on our market year 2009'. 'Can Pack is always known as the supplier which is most eager to offer competitive conditions and thus "rock the boat". […] Can Pack is not the member of the can industry association (while the other three players are) and therefore the challenger of the remaining three.'

The Transaction reduces the number of players from 4 to 3 in the EEA

9.1.3.1The Notifying Party's views

The Notifying Party submits that Crown and Can-Pack will continue to be able to compete effectively against the merged entity post-Transaction.

9.1.3.2The Commission's assessment

For the reasons set out in this section, the Commission has reached the conclusion that most customers of the Parties will, post-Transaction, face a choice between only two closely located competitors: the merged entity and either Crown or Can-Pack (see section 9.2 below). In certain regional clusters, the Transaction would even lead to a reduction of number of players from two to one, namely the merged entity.

This conclusion is confirmed by an analysis of the actual plant-level transaction data of the Parties and their estimates of Crown's and Can-Pack's transactions.

333First, Crown or Can-Pack owns the closest competing plant (for at least […] % of volumes each) for Ball customers supplied by only […] of Ball's 10 plants ([…]). For the other […] plants, Crown or Can-Pack accounts for only a small share of sales as the closest competing plant (less than […] % for all the […] plants, and less than […] % for […] of these plants).

Second, Crown or Can-Pack owns the closest competing plant (for at least […] % of customers each) for only […] of Rexam's 17 plants ([…]). For the other […] plants, Crown or Can-Pack accounts for less than […] % (and often […] %).

Figure 15: Share of Rexam and Ball sales by plant for which Crown or Can-Pack will own the closest competing plant post-Transaction

Source: Parties' transaction data

Third, in the case of Crown plants selling into the EEA, Ball owns the closest competing plant for […] % of sales, whilst Rexam owns the closest competing plant for […] % of sales (see Table 5 in Annex A to this Decision). In particular, for […] of the eight Crown plants selling into the EEA, Ball and Rexam each account for […] % of sales or more as the closest competing plant (these […] plants are […].

Fourth, in the case of Can-Pack, Ball owns the closest competing plant for […] % of sales, whilst Rexam owns the closest competing plant for […] % of sales (see Table 6 in Annex A to this Decision). In particular, for […] of the five Can-Pack plants selling into the EEA, Ball and Rexam each account for […] % of sales or more as the closest competing plant (these […] plants are […].

Customers that responded to the market investigation highlighted the fact that the Transaction would create a leading player in the EEA through the combination of Ball and Rexam, which are considered the 'two biggest suppliers of cans within the European region.' In particular customers consider that the remaining competitors would have no incentive to undercut the merged entity but would simply raise their prices. In this context it is worth quoting a customer's response to the market investigation in full:

'[Customer] anticipates that after the merger Ball-Rexam will become a powerful dominant market player which will gain a lot of bargaining power in the selling process. [Customer] estimates that it will be likely that the selling prices will be increased by Ball-Rexam which will probably also result in price increases of the other can suppliers in the market.' And '[t]he transaction might lead to higher prices, also for end-consumers. The reason is that Ball Rexam will become a large player driving the market prices and that other can manufacturers might be still competitive but will increase their prices as well. The power of the new organisation might be so strong that it will be difficult for market competitiveness to remain in place for small customers (can fillers) who do not have multiple filling locations worldwide. The new Ball Rexam organisation is likely to experience less to no competition from other can suppliers'.

9.1.4The Parties are close competitors

9.1.4.1The Notifying Party's views

In its Reply to the SO, the Notifying Party claims that the Commission overstates the Parties' closeness of competition. In particular, the Commission adopts an unduly rigid approach to assessing closeness of competition which is at odds with market reality. In particular, by simply assessing whether the Parties' plants are closest or second closest alternatives, the Commission fails to recognise the true nature of competition, as demonstrated by current patterns of supply. The Notifying Party argues that […] % of the Parties' EEA volumes were supplied from a plant which was not the closest to the customer.

9.1.4.2The Commission's assessment

For the reasons set out in this section, the Commission has reached the conclusion that the Parties are close competitors.

First, customers that responded to the market investigation consider the Parties as each other’s closest competitors in terms of customer service and reliability, rating them consistently as the best among the four beverage can manufacturers.

'Ball and Rexam are very [c]lose to each other in all [t]erms.'

'We see [Ball and Rexam] as the best [o]ption[s] in [t]erms of [q]uality, [p]rices, [flexibility] and reliability.'

Second, as explained in detail in section 9.1.8, customers that responded to the market investigation also regard the Parties as the main innovators:

'I perceive Rexam and Ball as quite close competitors in terms of service, prices, innovation. The others are followers or more specialised, for our demand I don't see them on the same level.'

Third, as explained in section 9.1.5, neither Can-Pack nor Crown benefits from a network of plants similar to that of Ball or Rexam.

Fourth, the Parties' internal documents indicate that Ball and Rexam have the largest product portfolios, while Crown is mostly focused on normal /standard and Can-Pack on large sizes. Ball and Rexam also have a higher share of slim and sleek cans than Crown and Can-Pack and the Transaction will reinforce this position, making them an important supplier for key customers that have higher demands for specialty cans.

Figure 16: Ball internal document on competitors' portfolio

Source: Ball's internal document, 'BevMod 2.0 From the “Ente” to the “Porsche”'

Customers also rate Ball and Rexam as best as regards portfolio of can types and sizes offered to customers: 'Ball and Rexam are the biggest Producer with most types and sizes from cans.'

Fifth, Ball and Rexam are perceived by the majority of respondents to the market investigation as unavoidable suppliers and challengers to each other. For instance, market participants 'regard Ball as a key competitor for Rexam'. Market participants state, among others, that there is a need for both Ball and Rexam to 'keep up competition' and as 'a challenger of the prices, of a comparable basis'.

357'Ball and Rexam are close competitors regarding quality and services.' Questionnaire to Customers (Q1), question 62.1, ID 1609.

358Questionnaire to Customers (Q1), question 64, ID 1609. 'Ball and Rexam are the biggest [p]roducers with most types and sizes [of] cans. The[ir] customer [s]ervice has big experi[e]nce.' Questionnaire to Customers (Q1), question 60.7, ID 1609.

359'Ball and Rexam […] are Close competitors in Terms of capacity, Innovation and quality' Questionnaire to Customers (Q1), question 62.1, ID 1609.

360'Ball and Rexam […] provide most broad range of sizes.'; 'Ball and Rexam are the biggest Producer with most types and sizes from cans. The[i]r customer Service has big [experience].' Questionnaire to Customers (Q1), question 60.7, ID 1609.

361Questionnaire to Customers (Q1), question 60.7, ID 1609. 'Regarding their market position (degree of innovation, service, quality, capacity) we consider Ball and Rexam to be comparable.' Courtesy translation from the German original 'Wir halten Ball und Rexam von der Marktposition (Innovationsgrad, Service, Qualität, Kapazität) vergleichbar.' Questionnaire to Customers (Q1), question 62.1, ID 1609.

362As regards Ball as a necessary supplier, market participants state: 'Ball's global production capacities are necessary from a capacity requirements point of view'; '[Ball] dominant in terms of capacity'; 'Rexam needs an equal competitor'. Questionnaire to Customers (Q1), question 63.1, ID 1609.

363As regards Rexam as a necessary supplier, market participants state: 'Rexam's global production capacities are necessary from a capacity requirements point of view'; 'Ball needs a competitor'. Questionnaire to Customers (Q1), question 64.1, ID 1609.

364Questionnaire to Customers (Q1), question 63.1, ID 1609.

(346) Sixth, the Commission has conducted a quantitative analysis of the identity of the closest and second closest competing plant for each customer (Figure 17 and Figure 18 as well as Annex A to this Decision). The Notifying Party's observation (see recital (338)) about current patterns of supply reflects the capacity-constrained nature of competition in the beverage can market. In a capacity-constrained industry like the beverage can industry, such an analysis provides a measure, albeit not the only relevant one, of the geographical closeness of the Parties, based on the location of their plant. The analysis also provides useful complementary evidence on the likelihood and strength of non-coordinated effects.

(347) The analysis of geographic closeness shows that Rexam is the closest competing plant to Ball's customers for […] % of Ball's sales in the EEA and Switzerland. This is two times larger than the average share for Crown, and close to […] times larger than the average share for Can-Pack.

(348) Moreover, out of the 10 Ball plants that sell into the EEA, there is […] Ball […] ([…]) for which Rexam does not account for at least the relative majority of sales as either the first or second closest competing plant:

– Rexam owns […] of the plants that are closest to the absolute majority of Ball's customers: […] (above […] % of their respective sales), […] (above […] % of sales) and […] (above […] % of sales). These […] Ball plants are located in the […] regions (selling volumes to customers located in these regions, and also to customers located in […] and […], especially in the case of […]);

– For an […] plant ([…]), Rexam is the closest competing plant for the largest share of volumes (over […] %). This plant largely sells to customers located in the […] and […] regions;

– Rexam is Ball's closest competitor even when one considers the position of the second closest competing plant from the perspective of Ball's customers. Rexam owns the second closest competing plant for […] % of Ball's sales, well above the corresponding shares for Crown ([…] %) and Can-Pack ([…] %). If one looks at the individual Ball plants, Rexam owns the second closest competing plant for the majority of sales from […] plants of its 10 plants selling into the EEA ([…] (above […] % of sales), […] (above […] %), […] (above […] %), and […] ([…] %)). For another plant ([…]), Rexam accounts for a […] % share of volume as the second closest competing plant.

(349) Second, Ball is the closest competing plant to Rexam's customers for […] % of Rexam's sales to customers located in the EEA and Switzerland. This is roughly […] times larger than the share for Crown and more than […] times larger than the share for Can-Pack.

(350) In particular, out of the 17 Rexam plants that sell into the EEA, there are only […] plants (out of 17 plants) for which Ball is neither the closest or second closest competitor for the majority of sales. […] of these plants are located in […], […]in […], […]in […], and […]in […]:

– Ball owns […]of the 17 plants that are the closest to the customers of Rexam. This is the case for […] (above […] % of their sales), […] (above […] % of sales) and […](above […] %). The plants are located in the […] regions (selling volumes to customers located in these regions, and also to customers located in […] regions);

– Ball is also the closest competitor for over […] % of sales from Rexam's plant in […];

– Ball owns the second closest competing plant to Rexam for a sizeable share of sales ([…] % of volumes). At plant level, Ball is the second closest competitor for the absolute majority of sales for […] (over […] % of sales), and […] (more than […] % of sales).

Ball is also the second closest competing plant to Rexam for over […] % of sales from the […] and […] plants.

(351) Third, post-Transaction, the distance from the closest competing plants would increase significantly for customers served by Ball and Rexam:

– the distance to the closest competitor would on average […] for Ball customers for whom Rexam is the closest competing plant, from […] km to […] km (see Table 3 and 4 of Annex A to this Decision). For all but […] of the Ball plants, the increase in distance would exceed […] km. Similarly, if one considers Ball customers for whom Rexam is the second closest competing plant, the distance would more than […], from […] km to […] km;

– the distance to the closest competitor would increase from 353 km to just over […] km for Rexam customers for whom Ball is the closest competing plant. For all but […] of the Rexam plants, the increase in distance would exceed […] km. In terms of the second closest competing plant, the equivalent increase would be from […] km to over […] km.

Figure 17: Share of Ball's volumes for which Rexam is the closest (in blue) or the second closest (in red) plant

Source: Ball and Rexam transaction data

Figure 18: Share of Rexam's volumes for which Ball is the closest (in blue) or the second closest (in red) plant

Source: Ball and Rexam transaction data

(352) Seventh, as described in detail in recital (229), the Parties take into account the location of the closest competing plant in their pricing.

9.1.59.1.5. The ability of the four beverage can manufacturers to compete on an equal footing

9.1.5.1The Notifying Party's views

In its Reply to the Article 6(1)(c) Decision, the Notifying Party made the following arguments.

354First, post-Transaction, both Crown and Can-Pack will continue to compete aggressively against the merged entity for a significant proportion of the overall volume of large customers.

355Second, regarding Crown, which has one plant less than Ball's EEA plant portfolio, it can be considered equivalent to Ball in terms of geographic coverage and financial strength.

356Third, regarding Can-Pack, despite a relatively lower number of plants compared to the Parties, it can supply in locations further away due to its production cost advantages.

In its Reply to the SO, the Notifying Party made the additional argument that both Can-Pack and Crown are sophisticated, global suppliers, both of whom can and do compete effectively with Ball and Rexam.

9.1.5.2The Commission's assessment

The Commission has reached the conclusion that post-Transaction neither Crown nor Can-Pack will be able to compete on an equal footing with the Parties. This is for a number of reasons.

EN 53 EN

First, it is important for a supplier to have a sufficiently wide network of beverage can manufacturing plants within the EEA . Neither Crown nor Can-Pack benefits, however, from a network of plants similar to that of Ball or Rexam, let alone the merged entity.

This is confirmed by the responses of both customers and competitors to the market investigation, as well as internal documents and public statements by the Parties.

Regarding customers that responded to the market investigation, Carlsberg explained that only Ball and Rexam 'are able to supply to multiple filling locations in Europe and they are willing to transport their cans longer distances, and if needed, subsidize the freight cost to win the business. As to Crown, Carlsberg considers Crown a small player on the European market, with limited production facilities, unlike Ball and Rexam that can benefit from their large production network, in case of capacity issues in one or the other European region.' 'Can Pack is perceived as a relatively small supplier, which is primarily active in Poland, Finland, Russia, India and the UK.'

SABMiller also stated the following: 'One has to consider that unlike Can-Pack and Crown, it is only Ball's and Rexam's plant networks that provide comprehensive footprint match with SABMiller brewery network, which has so far stimulated the competitive “network” environment between Ball and Rexam. If Ball and Rexam become one, “network” competition will disappear, as Can-Pack and Crown will not be able to compete from a “network” pricing perspective, which we fear will result in less supplier competition and significantly higher can prices to SABMiller.'

Regarding competitors that responded to the market investigation, a competitor explained that it applies a strategy that allows it 'to leverage/ trade off certain regions in order to receive better prices in other regions. […] As a result being absent from an important geographic area is a serious competitive disadvantage and the better geographic coverage one has, the more opportunities appear'.

Regarding the internal documents of the Parties, one of Ball's internal documents states that '[…].'

367'[…] it's important that they have a network of plants.' Questionnaire to Customers (Q1), question 59, ID 1609; 'to leverage technical, financial and supply contingency solutions' Questionnaire to Customers (Q1), question 59, ID 1609; 'to be flexible' Questionnaire to Customers (Q1), question 59, ID 1609; 'When more plant are available to the supplier, there is less chance to experience shortages.' Questionnaire to Customers (Q1), question 59, ID 1609; 'better for security of supply' Questionnaire to Customers (Q1), question 59, ID 1609; 'This ensures us we can be supplied by alternative locations (in case of supply issues).' Questionnaire to Customers (Q1), question 59, ID 1609.

368'[U]nlike Can-Pack and Crown, it is only Ball's and Rexam's plant networks that provide comprehensive footprint match with [our] network' Questionnaire to Customers (Q1), questions 59, ID 1609; 'Can Pack is perceived as a relatively small supplier, which is primarily active in Poland, Finland, Russia, India and the UK. Ball and Rexam are clearly the leaders: they are able to supply to multiple filling locations in Europe and they are willing to transport their cans longer distances, and if needed, subsidize the freight cost to win the business. As to Crown, [customer] considers Crown a small player on the European market, with limited production facilities, unlike Ball and Rexam that can benefit from their large production network, in case of capacity issues in one or the other European region.' See agreed non-confidential minutes of a call with a customer, 04.05.2015. ID 334. 'Ball and Rexam, as they are the hugest Players in the market. Only these two Players are covering most countries in Europe, in particu[la]r Germany' Questionnaire to Customers (Q1), question 61.2, ID 1609.

369Agreed non-confidential minutes of a call with a customer, 04.05.2015, ID 334.

370Questionnaire to Customers (Q1), question 59, ID 1609.

371Agreed non-confidential minutes of a meeting with a competitor, 13.05.2015, ID 1065.

372Ball's internal document, '[…]' , slide 4.

(365) Regarding public statements by the Parties, in November 2011, Rexam’s CEO told investors during a conference call that the 'location of the manufacturing footprint is absolutely key compared to our customer's filling plants to optimize both freight and logistic costs.'

(366) Second, the merged entity would control approximately two-thirds of all EEA-plants and have a geographical coverage in terms of number of plants three times larger than its next competitor. The merged entity would have a total of 24 plants in the EEA post-Transaction, plus one plant close to the borders of the EEA, Belgrade, which also serves […] the EEA. This does not take into account the Parties' expansion plans either (see section 5.1.4 above).

(367) Third, Crown and Can-Pack are perceived by customers to a certain extent as being capacity-constrained and, on certain occasions, unable to quote or refusing to quote for some volumes: 'Rexam and Ball are therefore the only two companies who can supply us with the volumes we require'. 'We do not have any sourcing contract with Crown, because they are at full capacity'. 'Crown is no alternative because they have not enough capacity'; 'Crown not having capacity'. 'At the last tender, Crown did not make an offer, arguing to lack available capacities (however, they indicated they aimed at offering at a later stage). Also [Can-Pack] only offered a small, limited available capacity. Hence, only the remaining Ball and Rexam are able to make an offer in the dimension needed by us'. '[Can-Pack] refused [to provide a quotation] due to lack of spare capacity'.

(368) Fourth, many customers view the Parties as the best beverage can manufacturers in terms of customer service, innovation, flexibility in meeting additional demand, reliability or portfolio of can types and sizes offered. As a customer puts it: 'Crown is not a supplier because of inferior quality of their products and their less competitive pricing. Can-Pack only has a low share of [customer's] volumes because of severe quality issues, comparable to those of Crown. Transport security and materials used by Can-Pack are also problematic. Volume-wise Can-Pack is considered to be a minor player on the market not suitable for delivering higher volumes. This has not improved over the last 15 years. Moreover, Can-Pack has no backup can-plant from where it could supply cans, if another plant cannot deliver the requested volumes. The other competitors have at least 4 alternative plants. Ball and Rexam are also more active in innovation.' 'Even with Can-Pack as a potential competitive threat, a significant concern remains as Can-Pack is less flexible in terms of volume variability, legal terms and service than Rexam.'

373Rexam. Edited Transcript, REX L – Rexam PLC Investor. https://www.rexam.com/files/presentations/investorseminar11investor_seminar_nov11_transcript.pdf. Last accessed on 28.09.2015, ID 3000.

374See recital (53). Two of Rexam's Russian plants also deliver to the EEA, […].

375Questionnaire to Customers (Q1), question 6.1, ID 1609.

376Courtesy translation from the Spanish original: '[con] Crown no tenemos contrato de suministro, por que tienen su capacidad completa' Questionnaire to Customers (Q1), question 60.7, ID 1609.

377Questionnaire to Customers (Q1), question 61, ID 1609.

378Questionnaire to Customers (Q1), question 65.1, ID 1609.

379Courtesy translation from the German original 'Crown hat bei der letzten Ausschreibung kein Preisangebot abgegeben, mit der Begründung, keine hinreichende verfügbare Kapazität zu haben (man hat aber signalisiert, langfristig in weiteren Gesprächen anbieten zu wollen). Auch Can-Pack hat nur eine geringe, limitierte verfügbare Kapazität angeboten. Somit verbleiben nur Ball und Rexam die in der Größenordnung anbieten, die unsererseits benötigt wird.' Questionnaire to Customers (Q1), question 65.1, ID 1609.

380Questionnaire to Customers (Q1), question 65.1, ID 1609.

381Agreed non-confidential minutes of call with a customer, 13.05.2015. ID 248.

382Questionnaire to Customers (Q1), question 97, ID 1609.

(369) Or as another customer states: 'Ball and Rexam are the biggest Producer with most types and sizes from cans. The[ir] customer [s]ervice has big experi[e]nce'. 'I feel Ball and Rexam are market leaders and truly focussed on beverage cans and provide most broad range of sizes. Crown is also active in a lot more metal packaging types. Can-Pack is maybe low in price against good quality but have in general longer lead times/transport time and offer less innovation'. 'Rexam and Ball have been key competitors for several years in terms of innovation, service, quality, price and capacity. Other suppliers Crown and [Can-Pack] have not focused on soft drinks can supply to the same extent'. '[Can-Pack] offers a limited number of different can sizes.'

(370) Fifth, Crown and Can-Pack would be unable to compete effectively against the merged entity due to the lack of geographical coverage. This was highlighted by a number of customers that responded to the market investigation:

'Regional can-makers are too small on a European scale if they do not have at least an EU-wide footprint.'

'[P]ost-merger the merged entity with a very large market share and large geographical coverage will have more leverage against its customers. In the long run this can lead to reduced competition in the can market.'

'Post-merger the big players will have enough power to crowd out the smaller players, such as Can-Pack (which is the only supplier in Europe without a presence in the Americas or Asia which would give it room to manoeuvre in this space).'

'The merged entity will pool volumes for procurement and R&D purposes. This will not necessarily make other players (like Can Pack) less competitive but it will be more difficult for these players to keep up with the merged entity.'

'Ball and Rexam already each have a more complete geographical coverage than other EEA can makers. If combined, they would have the majority of installed capacity in each region in the EEA with exception of Eastern Europe. In addition, the combined entity would have an almost monopolistic position in the European heartland region of the EEA which is the fastest growing region in the EEA. Ball and Rexam combined would have a significantly wider range of can sizes than any competitor in the EEA. In light of these advantages it would be extremely difficult for other can makers to compete with the combined entity.'

'The proposed transaction will create a dominant and near-monopolistic beverage can company in the EEA. We are concerned that the two remaining smaller players will not be able to compete with this entity in the future. Post transaction the merged entity will be able to leverage its advantages in terms of size, product scope and geographic coverage to the detriment of the two remaining competitors in the EEA and, ultimately, of its customers. […] We are concerned that the combined entity will be able to exclude the remaining competitors in the EEA as a result of this quasi-monopolistic market share and will thus remain uncontested by any type of competition in the future.'

(371) Sixth, there are a number of additional reasons why Can-Pack would be unable to compete against the merged entity.

(372) In the first place, Can-Pack is perceived by certain customers as not being innovative or service-oriented enough (on innovation, see section 9.1.8). In relation to Can-Pack's service level, customers have pointed out that:

'[N]ot much experience with Can-[P]ack and there we also had some Quality issues'.

'Can Pack offers good prices, but it does not offer assistance for technical services or special services for printed lids etc.'

'[Can-Pack] normally does just what is needed; No above and beyond, restrictive contracts and relatively stable performance but very limited innovation.'

'I have the impression that the service provided by [Can-Pack] is not as good as Ball's.' 'Can-Pack [cannot] meet our quality requirements currently, so we see not a close competition relation to Ball and Rexam.'

'Both Ball and Rexam are best can suppliers in terms of service level and innovation. Can Pack appears as having the lowest spend on innovation and capacity to innovate. As a result post-merger it is hard to say whether [Can-Pack] will live up to the challenge of competing with the merged entity.'

'Ball and Rexam tend to be more focused on value added solutions and technical service while Can-Pack tends to be more focused on price. Crown is somewhat in the middle'.

(373) Certain customers also consider Can-Pack's logistics capacity and flexibility as being limited.

(374) In the second place, Can-Pack's product range is somewhat limited in terms of the number of different sizes available. This was confirmed by Can-Pack, Crown and customers that responded to the market investigation. Can-Pack stated that its most common beverage can is the 50 cl can that is mainly used to contain beer. Can-Pack also explained that although like other beverage can manufacturers, it offers a full range of can sizes, given that each of its

392A competitor's non-confidential reply to RFI of 15.06.2015, ID 1129.

393Questionnaire to Customers (Q1), question 60.7, ID 1609.

394Questionnaire to Customers (Q1), question 60.7, ID 1609. Courtesy translation from the Spanish original: 'Can Pack tiene buen precio, pero no tiene asistencia de servicio técnico ni servicios especiales de impresión tapas etc.'.

395Questionnaire to Customers (Q1), question 60.7, ID 1609.

396Questionnaire to Customers (Q1), question 62.1, ID 1609.

397Agreed non-confidential minutes of a call with a customer, 04.05.2015, ID 334.

398A competitor's non-confidential reply to RFI of 15.06.2015, ID 1129.

399Questionnaire to Customers (Q1), question 86, ID 1609.

400Agreed non-confidential minutes of a meeting with a competitor, 13.05.2015, ID 1065. See also: 'Can-Pack is not a valid alternative for Ball because it does not produce the can types [customer] mainly needs (Can-Pack only produces "standard" cans.)' Courtesy translation from the Italian original: 'Can-Pack non rappresenta una valida alternativa a Ball perché non produce i tipi di lattine di cui [cliente] ha principalmente bisogno (Can-Pack produce solamente lattine "standard").' Agreed non-confidential minutes of a call with a customer, 07.08.2015, ID 2524.

401Agreed non-confidential minutes of a meeting with a competitor, 05.05.2015, ID 261.

production lines has the tooling only for a number of shapes and sizes, it cannot manufacture a complete product range in all of its plants.

(375) Crown stated that it does not see Can-Pack's strategy as one of trying to fulfil all different customer needs but rather one of focusing on investments in operations and offering simple standard can sizes.

(376) Customers and competitors that responded to the market investigation indicated that, although it offers specialty cans to a limited extent, Can-Pack's main focus is on producing standard and bigger cans for breweries.

(377) In the third place, due to its lack of extensive geographical coverage, Can-Pack is not a suitable main beverage can manufacturer for large customers in the EEA. This was confirmed by Can-Pack, customers that responded to the market investigation and the Parties’ volume data.

(378) Can-Pack itself indicated that it typically focuses on customers 'closer to home'.

(379) A number of customers that responded to the market investigation stated as follows.

'Can-Pack offered to supply sleek cans from Kiev (Ukraine), but this supply chain appeared to be unreliable and too long. Currently only Ball and Rexam can be seen as viable suppliers of larger can volumes in the Benelux.'

'Can-Pack's share has grown a bit, but it is still considered a small competitor. It was awarded volumes by [customer] in the Polish market because it has a strong presence there with two can-plants, but also in the UK and Romania. In many other EEA markets, however, where [customer] is active in, Can-Pack is not present.'

'Volume-wise Can-Pack is considered to be a minor player on the market not suitable for delivering higher volumes. This has not improved over the last 15 years. Moreover, Can-Pack has no backup can-plant from where it could supply cans, if another plant cannot deliver the requested volumes.'

'[Can-Pack] has to deliver the Cans over a [l]ong distance. This is the Problem for [Can-Pack] to be the first supplier for us.'

'[Can-Pack]'s footprint has limited overlap with Carlsberg.'

'[I]n bidding procedures Can-Pack cannot compete against suppliers with sufficient plant network that offer competitive conditions on an EEA-wide basis, such as Rexam. For this reason Can-Pack's lowest priced bids are in some regions not accepted.'

(380) As for the Parties' volume data, it shows that contrary to the Notifying Party's claim, Can-Pack never supplies the majority of the volumes to any large customer.

402A competitor's non-confidential reply to RFI of 15.06.2015, ID 2572.

403Agreed non-confidential minutes of a meeting with a competitor, 13.05.2015, ID 1065.

404Agreed non-confidential minutes of a call with a customer, 07.05.2015, ID 326; Agreed non-confidential minutes of a meeting with a competitor, 13.05.2015, ID 1065.

405Can-Pack's economic submission, ID 3440.

406Agreed non-confidential minutes of a call with a customer, 22.05.2015, ID 331.

407Agreed non-confidential minutes of a call with a customer, 12.05.2015, ID 468.

408Agreed non-confidential minutes of a call with a customer, 13.05.2015, ID 248.

409Questionnaire to Customers (Q2), question 35.1, ID 2664.

410Questionnaire to Customers (Q2), question 35.1, ID 2664.

411Questionnaire to Customers (Q2), question 35.1, ID 2664.

(381) Indeed, Can-Pack indicated that it does not compete as main supplier in Northern France, the Benelux and Germany because it is unable to meet customer requirements (for example, just-in-time delivery and flexibility) when supplying from Poland. Can-Pack estimates that its share of volume supplied as main supplier in Northern France, Benelux and Germany is less than […] %.

(382) In the fourth place, only the Parties are able to compete on an equal footing for small and large customers, while Crown and Can-Pack focus on either small or large customers. This is confirmed by Table 7, which shows that while the Parties hold similar shares for large (that is to say customers purchasing more than 500 million cans per year) and small customers (that is to say customers purchasing less than 500 million cans per year), Can-Pack's share with large customers is less than […] of its share of smaller customers and Crown's share with smaller customers is nearly […] of its share of large customers.

Table 7: Market shares of large volume customers in the EEA and non-EEA countries in […]

Market shares (%) based on customers purchasing

above 500 million cans / year

below 500 million cans / year

Ball [30-40] [30-40] [30-40]

Can-Pack [20-30] [10-20] [10-20]

Crown [10-20] [10-20] [10-20]

Rexam [30-40] [40-50] [40-50]

Source: Parties' transaction data and their estimates of competitor's sales

9.1.6. Large customers versus other customers

(383) Customers include manufacturers of many different types of beverages and of many different sizes (in terms of the volumes of cans procured). Broadly speaking, customers can be grouped according to their scale, namely (i) large customers that are usually well-known brand owners with sophisticated purchasing requirements, high quality expectations and complex logistic requirements which use global or pan-European tendering processes and (ii) other smaller customers that are more local (with generally one or two filling locations) and have less

flexibility in negotiating commercial terms with beverage can manufacturers due to their more limited scale.

9.1.6.1. The Notifying Party's views

(384) According to the Notifying Party, the size of customers varies, with a continual spectrum in terms of total EEA volume demand. Any distinction between customers is for internal analysis purposes and it is not based on volume or value of sales to those customers, but on the way in which the Notifying Party needs to structure its sales team in order to serve the customer's requirements most effectively.

(385) Large customers have different purchasing requirements in the sense that they require more complex services due to sophisticated logistics. Negotiations with key accounts in the EEA are conducted by a key account manager, a key account director and in some cases, the vice president of sales. Tenders from these customers are often organised on a global basis and therefore involve group-level coordination. For all other customers, negotiations are conducted by a sales manager, a general manager and in some cases, a sales director.

(386) In its Reply to the SO, the Notifying Party claims that the relevant evidence clearly shows that large customers typically either tender on a global basis or, more often, tender on an intra-EEA (regional) basis. Of the recent tenders organised by large customers, none took place on a pan-European basis.

(387) In Annex 1 of its Reply to the SO, the Notifying Party also acknowledges that certain cost savings can result from serving large customers. However, according to the Notifying Party, these are overstated by the Commission. According to the Notifying Party, the number of graphics changes requested by a customer (which bears no relationship to customer size) is a far more important driver of operating costs than customer size.

9.1.6.2. The Commission's assessment

(388) For the reasons set out in this section, the Commission has reached the conclusion that: (i) there are differences in the needs and demands of large customers and smaller local or regional customers; and (ii) beverage can manufacturers treat large customers differently to smaller local or regional customers.

(389) First, large customers accounted for more than 75 % of the EEA demand for beverage cans and for about […] % of the Parties' EEA sales between 2012 and 2014.

(390) Second, large customers have a large network of filling plants and prefer to source from beverage can manufacturers that match their own geographic footprint and have a spread of stock keeping units (SKUs). This has been confirmed by both competitors and customers during the market investigation. According to a competitor 'large multi-national customers prefer to deal with large suppliers who have a presence in a number of countries / geographic regions'; '[i]t would be more difficult for a small player to deal with the proliferation of can sizes' and 'regional can-makers are too small on a European scale if they do not have at least an EU-wide footprint.'

392In the first place, large customers tend to get better terms and conditions when they pool their volumes across different filling locations with the same supplier. A large customer also noted that 'all buyers with annual contract volumes that exceed 500 million cans are considered as strategic customers by the can suppliers. Thus, in these cases can suppliers are willing to lower their pricing terms and provide additional services as well as allocate additional resources to the buyer'or 'because Coca Cola and Pepsi have best price clauses in their contracts which are known to all three BCME-members there is a minimum price for regular cans which is more or less known to the bigger buyers in the market.'

393In particular, large customers negotiate volume based rebates. Rebates can be linked to overall volume targets between the beverage can manufacturer and customer or to specific volume targets per filling plant or a combination of both. For example, a volume rebate is offered by beverage can manufacturers for annual contract volumes that exceed […] million cans. These contracts are usually coupled with long-term supply agreements as this allows beverage can manufacturers to optimise asset utilisation rates. Rebates are offered by geographic regions (global or pan-European) or by bottler. The scale of the rebates will, however, depend on 'the allocation of the volume over the various countries, the product mix and the number of alternative suppliers in such markets'.

394Generally, the customer must meet an overall European-wide (or even global) volume target with specific regional or filling location-by-filling location volume targets built within this overall target: '[…]'.

395Moreover, Ball's […] automatically attaches a certain […].

396Agreed non-confidential minutes of a meeting with a competitor, 05.05.2015, ID 261.

397A competitor's non-confidential reply to RFI of 15.06.2015, ID 2572.

398Questionnaire to Customers (Q2), question 10.1, ID 2664.

399Questionnaire to Customers (Q1), question 96.2.1, ID 1609.

400'better pricing terms at a local level are generally offered by can suppliers when the annual contract volumes exceed […].'

401In such case can suppliers offer buyers volume rebates amounting to approximately EUR […] per 1000 cans. In the event that the annual contract volume reaches […] cans, can suppliers normally offer an additional volume rebate of EUR […] per 1000 cans. All buyers with annual contract volumes that exceed […] million cans are considered as strategic customers by the can suppliers. Thus, in these cases can suppliers are willing to lower their pricing terms and provide additional services as well as allocate additional resources to the buyer.

In the second place, beverage can manufacturers normally have limited storage space for their can supplies, and large customers may receive several can shipments a day. Therefore, security of supplies is a key concern for these customers as insufficient or delayed can deliveries may lead to production stoppages. As explained by one large customer: 'In case of supply issues at a plant, the can supplier can use their production network across EEA to source cans. This increases supply security and is a big advantage compared to smaller can makers who have fewer plants.' 'Suppliers with a sufficient geographic footprint can react to production issues by temporarily or permanently sourcing cans from other locations. This increases security of supply, and often enables access to a wider product range.'

In the third place, beverage can manufacturers prioritise demand of large customers in particular in peak season when supply shortages can occur. As a customer pointed out: 'in high season, when cans are requested, there are issues with can availability. Since [customer] does not need such big amounts of cans as for instance Coca-Cola, it does happen that its orders are satisfied in the second and further ranks of importance'.

In the fourth place, large customers prefer to choose beverage can manufacturers with a wide footprint to ensure the overall cost related to the sourcing of cans across all filling plants in the EEA is as low as possible. As a customer pointed out: 'the isolated supply cost for an individual plant is far less relevant than the overall cost of supply to [customer]'s entire operations in the EEA'.

Beverage can manufacturers also gain certain benefits from doing business with large customers: (i) efficiency gains when producing different can sizes on one line (the fewer customers for a line, the fewer changeovers if producing different sizes), (ii) productivity gains owing to a higher volume of cans produced (can lines are set up to run high volumes at high efficiency), or (iii) optimisation of freight routes over a larger network.

For example, some large customers stated that 'scale […] allows can makers to have larger production runs and therefore higher efficiency gains' and '[…] a greater volume of beverage cans enables the supplier to optimise his assets utilisation rates.'

The internal documents of the Parties also acknowledge the importance of footprint alignment with key customers: […]as well as the portfolio of the beverage can manufacturer […].

Third, large customers tend to issue tenders at pan-European or global level. This was confirmed by both large customers and competitors that responded to the market investigation.

Large customers that responded to the market investigation indicated that '[e]ntire global volumes of [customer] are pooled in such a tender and suppliers can bid for the entire pool or a proportion of it' and have '[c]entralized purchasing at [customer] for the global needs'.

One major customer explained that the term 'regional tenders' is defined in its internal business organisation as entire Europe. Another major customer also defines Europe as a 'global region' and runs pan-European tenders: 'In Europe, [customer] has a pan-European approach and the four suppliers are encouraged to submit an offer for all locations.'

Two major customers run global tenders: 'Last year, [customer] conducted a global tender for beverage cans for 3 to 5 years for all their regions'. '[Large customer] invited 28 can suppliers from all around the world to tender for all volumes globally.'

Competitors that responded to the market investigation stated that sourcing has evolved to become more global or European-wide for large customers and acknowledged that '[p]an-European or global customers typically negotiate contracts and prices through a central or corporate procurement function in order to leverage their scale and achieve better prices than local entities would be able to achieve on their own.' And '[n]ow procurement of cans by global beverage producers often follows a more complex global tender process and the local subsidiaries/filling plants are losing influence over who their can supplier(s) is/are and which the terms of purchase are'. Moreover, 'the way in which sales contracts are negotiated with [Large Global Customers ('LGCs')] has undergone a significant change in the recent past […] This means that while historically (and still today, for smaller customers) contracts were negotiated at local level – the purchasing department of the individual filling plant being responsible for the sourcing decisions, for the LGCs this function has moved up to group level (and sometimes even dedicated purchasing entities). Contracts are negotiated and purchasing decisions made at this level.' '[A] large majority of Crown's cans are supplied to large customers who organise international tenders. The majority of negotiations are held on either a European-wide or global level (although some national tenders are held for local customers). The larger customers organise global or pan-European tenders in order to leverage their scale. For example there was a global tender process in which more than 55 billion cans have been awarded (about 15% of global demand). Generally contracts are awarded for 1-2 years, a 3-5 years contract can be considered a long one.'

Fourth, large customers tend to have longer contracts than smaller local or regional customers. Local customers tend to conclude one-year contracts, regional customers 1-3 year contracts and large customers 3-5 years contracts.

9.1.7. Importance of having a network of plants

The Notifying Party's views

According to the Notifying Party, the size of customers varies, with a continual spectrum in terms of total EEA volume demand. Any distinction between customers is for internal analysis purposes and it is not based on volume or value of sales to those customers, but on the way in which the Notifying Party needs to structure its sales team in order to serve the customer's requirements most effectively.

Large customers have different purchasing requirements in the sense that they require more complex services due to sophisticated logistics. Negotiations with key accounts in the EEA are conducted by a key account manager, a key account director and in some cases, the vice president of sales. Tenders from these customers are often organised on a global basis and therefore involve group-level coordination. For all other customers, negotiations are conducted by a sales manager, a general manager and in some cases, a sales director.

In its Reply to the SO, the Notifying Party claims that the relevant evidence clearly shows that large customers typically either tender on a global basis or, more often, tender on an intra-EEA (regional) basis. Of the recent tenders organised by large customers, none took place on a pan-European basis.

In Annex 1 of its Reply to the SO, the Notifying Party also acknowledges that certain cost savings can result from serving large customers. However, according to the Notifying Party, these are overstated by the Commission. According to the Notifying Party, the number of graphics changes requested by a customer (which bears no relationship to customer size) is a far more important driver of operating costs than customer size.

The Commission's assessment

For the reasons set out in this section, the Commission has reached the conclusion that: (i) there are differences in the needs and demands of large customers and smaller local or regional customers; and (ii) beverage can manufacturers treat large customers differently to smaller local or regional customers.

First, large customers accounted for more than 75 % of the EEA demand for beverage cans and for about […] % of the Parties' EEA sales between 2012 and 2014.

Second, large customers have a large network of filling plants and prefer to source from beverage can manufacturers that match their own geographic footprint and have a spread of stock keeping units (SKUs). This has been confirmed by both competitors and customers during the market investigation. According to a competitor 'large multi-national customers prefer to deal with large suppliers who have a presence in a number of countries / geographic regions'; '[i]t would be more difficult for a small player to deal with the proliferation of can sizes' and 'regional can-makers are too small on a European scale if they do not have at least an EU-wide footprint.'

412First, a beverage can manufacturer that has a broad network of plants can better guarantee its customers that in case of issues with one plant, another plant can step-in to supply them. Although the Notifying Party argues that it is unnecessary to have a broad network to ensure security of supply as outages are fairly rare, a number of large customers that responded to the market investigation highlighted the importance to them that a beverage can manufacturer has a large network of plants.

413As one major customer points out: 'it is […] footprint that enables a company like Rexam to compete effectively everywhere in the EU. It can supply cans quickly and securely anywhere and then optimise supply through (sponsored) investments, optimised logistics, wall-to-wall plants etc. at a later stage without taking much risk. A smaller player cannot implement any such strategy because it will often need to make investments to be able to supply before being certain to secure the relevant contract.'

414This was also confirmed by a competitor that responded to the market investigation and which underlined that 'the benefits a large network entails during peak seasons or if capacity problems arise: Customers value the availability of back up plants with spare capacity that can be utilised in cases where the primary manufacturing plant experiences outages or capacity shortages. Given the importance of having a sufficient supply of beverage cans available at the required time, the threat of non-delivery of required volumes is a serious one. Having a large network allows the supplier to shift volumes and still meet customer demand. Even if this comes at an additional cost, it distinguishes a supplier with a large network from its smaller competitors who do not dispose of such backup facilities.'

415Second, having a network of plants allows a beverage can manufacturer to offer a wider product portfolio. Although according to the Notifying Party it is possible to manufacture a

416range of sizes on a single line, having a network of plants allows a beverage can manufacturer to manufacture a wide range of cans efficiently. This is because a network of plants allows a beverage can manufacturer to concentrate the production of formats for which demand is lower in a limited number of plants, which in turn reduces production costs as production lines do not need to be stopped in order to adjust them to the new formats.

417This is confirmed by a comparison across beverage can manufacturers between the number of lines in a plant and the actually produced can sizes in that plant. While […] produced on average around one can size per line in their EEA plants in 2014, […], produced more than two can sizes.

418This is also confirmed by customers and competitors that responded to the market investigation and by internal Rexam documents.

419A customer explained as follows: '[it] uses several sizes of cans and the demand in the market (consumption/volume and type of drinks) changes all the time (goes up and down from one moment to the next), which means that the flexibility at the supplier is very important to supply according to [the customer's] demands.'

420A competitor explained that 'It would also be more difficult for a small player to deal with the proliferation of can sizes ([…]). A big player can produce test sizes in one plant and ship them, a smaller one does not have so many plants where such test sizes could be manufactured so its risk is higher.'

421Another competitor pointed out the 'connection between the size of a can manufacturer's network and the number of production sites it has on the one hand and the ability to offer a comprehensive range of can types and sizes on the other hand. While it is technically correct that all manufacturers could produce all types and sizes of beverage cans (subject to the possible exclusionary use of IP rights by some suppliers with regard to certain can sizes), it is not commercially effective to do so in a single plant or a small number of plants.'

This competitor also underlines that this is mainly for two reasons:

422– On the one hand, less switching reduces per unit costs as '[f]ocussing production on any line on a small number of can sizes therefore reduces the relative tooling costs per unit produced.'

– On the other hand, less switching also reduces production downtimes.

423Overall, 'the ability to effectively offer complete ranges of can sizes very much depends on the number of plants a manufacturer has amongst which the production of the portfolio can be spread.'

424As for Rexam, an internal email dated 14 April 2014 states as follows: '[…]'

425Third, having a network of plants allows a beverage can manufacturer to optimise its costs by using its cheaper capacity before that of plants that have higher production costs. For example, [...], see section 7.1.3.2 above) which is […]'

426Fourth, having a network of plants allows a beverage can manufacturer to try and leverage its position in geographic areas in which there is less competition to obtain volumes in geographic areas in which there is more competition. This was confirmed by both customers and competitors that responded to the market investigation.

427'Contracts with can suppliers are concluded having a broad European perspective in mind. As a result a can supplier can offer substantive discount in one region provided that Molson Coors agrees to grant volumes in another location at a less interesting price. Thus prices differ per filling location and Molson Coors aims at choosing the best mix for its purposes.'

428'[S]uppliers [often] offer a “package deal” with the intention to gain a certain level of volume in order to achieve lowest DDP prices (incl. additional discounts). Rexam and Ball have used this strategy for EEA. With Ball - to gain discount on final DDP for […] and […] ([…] million pcs), we had to compromise on the price for […] in […] ([…] million pcs) With Rexam – to gain better prices for […], we had to maintain our current volume in […], and in general increase our volume share with Rexam.'

429'ability to bid for multiple filling locations gives a supplier the ability to tailor a bid to take the customer's focus on the overall price per can for all of its filling locations into account. This means that smaller suppliers may not be awarded contracts even if they offer the best price for a specific location.'

430Fifth, by having a network of plants, a beverage can manufacturer that is able to supply customers in different EEA regions can win more sales by offering volume rebates to customers on the basis of their total purchases which cannot be matched by beverage can manufacturers with a smaller footprint (see section 9.1.7). This was explained by several customers that responded to the market investigation:

431'Utilizing global can volumes to attract lowest DDP price per filling site Single sourcing in markets to obtain most competitive price Bundling volumes across multiple markets to attract overall lower prices from suppliers and generating a volume in which [customer] become a prioritized customer due to size of revenue share leading to strategic partnerships. Supplier footprint matching that of [customer] generates better prices than sourcing from suppliers with plants that are further away.'

432'[w]ithin the last tender for the financial years 2015-2017 in the EEA, Rexam and Can-Pack have demonstrated willingness to reduce and harmonise the conversion cost component of offered can pricing across the entire EEA area, in order to compete for larger total volume of business in the EEA.'

433Sixth, […].

9.1.8The Parties are the main innovators

9.1.8.1The Notifying Party's views

434The Notifying Party submits that although innovation is not a key parameter of competition in the beverage can industry, Crown and Can-Pack each have very strong innovation capabilities, which is confirmed by the Parties' internal documents.

435Regarding Crown, the Notifying Party submits that Crown is well-positioned to develop new innovations from its research centre in the UK and a number of 'excellence centres' globally. For example, Crown was able to use its experience in food can manufacturing in its development of the '360 degree' can end.

436As for Can-Pack, Ball considers it to be particularly strong in […], while Rexam describes Can-Pack's plants as […].

9.1.8.2The Commission's assessment

437The Commission has reached the conclusion that the Parties are the main innovators with respect to the supply of beverage cans in the EEA. This is for the following reasons.

438First, the innovations of Crown and Can-Pack are limited in terms of number and breadth as compared to the long list of innovations brought to the market by the Parties.

Figure 19: The Parties' and their competitors' innovations in the beverage can industry

[…]

Source: Ball's internal document, “Product Innovation” projectMaster file

439Second, customers that responded to the market investigation considered that Rexam and Ball are the main innovators with respect to the supply of beverage cans in the EEA (with Rexam as number one, followed by Ball): 'Both Ball and Rexam are best can suppliers in terms of service level and innovation. Can Pack appears as having the lowest spend on innovation and capacity to innovate. As a result post-merger it is hard to say whether Can Pack will live up to the challenge of competing with the merged entity.' 'By and large, Rexam and Ball are the leading innovators in the market.'

471Questionnaire to Customers (Q1), question 73.1, ID 1609.

472Reply to RFI 17 of 28.10.2015, […], ID 3448: '[…]'

473Reply to the Article 6(1)(c) Decision, paragraph 4.68, ID 1448.

474Reply to the Article 6(1)(c) Decision, paragraph 4.71, ID 1448.

475Reply to the Article 6(1)(c) Decision, paragraph 4.72, ID 1448.

476'Ball and R[e]xam are the most innovative and with highest capacity to produce sleek cans and aluminium cans'; 'In addition, Ball and Rexam are the can manufacturers characterised by innovative solutions whereas Can-Pack for instance lacks this innovation support and is less willing to take customary contractual risks.' Questionnaire to Customers (Q1), question 62.1, ID 1609. See also Questionnaire to Customers (Q1), question 60, ID 1609.

477Agreed non-confidential minutes of a call with a customer, 04.05.2015, ID 334.

478Agreed non-confidential minutes of a call with a customer, 18.08.2015, ID 2866.

This was also indicated by one supplier of the Parties:

479'In Europe, the beverage can manufacturers can be divided into two groups regarding their ability to innovate. Ball and Rexam stand out from Can-Pack and Crown in this respect. While each beverage can manufacturer tries to improve their product, Ball and Rexam today account for nearly all innovation projects currently conducted with Novelis. Also these projects follow a more long-term approach. In general, Europe is more active regarding innovation than North-America, driven in part by the absence of soft tolling in Europe.'

480Third, the innovation capabilities of the Parties are demonstrated by the fact that they are the only two beverage can players that offer aluminium bottles in the EEA. These products are considered by customers to be 'a product of the future.' This is also confirmed by an internal document of Rexam (see Figure 20), where it is underlined that […].

Figure 20: Rexam's internal document on aluminium bottles

[…]

Source: Rexam's internal document, 'Beverage Cans strategy 2014-16', slide 36

Fourth, the Transaction may reduce the current levels of innovation of the Parties. This was stated by a number of customers that responded to the market investigation:

481'Currently both Rexam and Ball compete vividly on innovation (for instance aluminium bottle) but this can change after the merger.'

482'It is to be expected that there will be less innovation projects post-merger because the innovation is partly driven by the fact that the merging parties are currently competing against each other.'

483'The proposed transaction will rather lead to less innovation as the merger entity will focus on limiting costs and the two R&D departments are likely to be reduced to one smaller department. The merged entity can use its scale to lower the costs. However, prices for customers will likely increase and not decrease.'

484A competitor that responded to the market investigation also noted that the 'notified transaction will only increase this problem. It artificially creates a behemoth that smaller players cannot hope to compete against.'

9.1.9Capacity is tight in the EEA

9.1.9.1The Notifying Party's views

The Notifying Party argues that both Crown and Can-Pack would have spare capacity to supply any customers that wish to switch suppliers post-Transaction.

486Regarding Can-Pack, it has spare capacity at its plants in the UK and Poland. Can-Pack's plant in Morocco is also able to supply into some of the southern regions of the EEA.

487As for Crown, it has capacity available at its plants in Spain, France, the UK and Slovakia.

9.1.9.2The Commission's assessment

The Commission has reached the conclusion that there is currently limited spare capacity in relation to the supply of beverage cans in the EEA. As a result, if the merged entity were to increase prices post-Transaction, competitors would have insufficient spare capacity to counteract such an increase. Moreover, Post-Transaction, the merged entity would have the incentive to do so.

First, as the Notifying Party notes in the Form CO, capacity utilisation is high in the industry (above 90 %). This was also confirmed both by competitors ('[competitor] believes that the industry in general currently is operating very close to capacity') and customers that responded to the market investigation:

492'there is currently no additional capacity in the market.'

493'[t]he industry has traditionally been very tight (capacity wise).'

494'Ball and Rexam typically try to limit capacity.'

495In particular, contrary to the Notifying Party's claim, Can-Pack confirmed that it 'does not have spare capacity in its Polish and UK plants that would allow it to increase its output in a way that would effectively restrain the merged entity in an appreciable way.' Can-Pack also confirmed that the capacity utilisation at all of its EEA plants is above 90 %.

Second, the capacities of the Parties are important from the point of view of total capacity in the EEA as Crown and Can-Pack are unable to serve large volumes, due to their limited capacity.

The overall spare capacity in 2014 of Crown and Can-Pack for plants located in the EEA was equivalent to only between 5 to 10 % of the combined sales of the Parties to EEA customers in 2014. Crown and Can-Pack would, therefore, not be able to supply sufficient additional volumes to customers in case of a price increase by the merged entity.

The levels of spare capacity estimated by third parties is even lower in 2015, implying that the 2014 data overstate the ability of third parties to increase output to offset a price increase.

Moreover, the Commission also notes that in the presence of binding capacity constraints, even if competitors hold relatively large spare capacity, this spare capacity may not be sufficient to prevent significant non-coordinated effects arising from a large increase in concentration and a significant consolidation of capacity (see M.6471-Outokumpu/Inoxum, decision of 7.11.2012 and M.6905- Ineos/SOLVAY/JV, decision of 8.05.2014).

500Questionnaire to Customers (Q1), question 64.1, ID 1609.

501Questionnaire to Customers (Q1), question 64.1, ID 1609. Courtesy translation from the Spanish original: 'Es el único proveedor que nos suministra las cantidades que nuestra empresa necesita.'

502Courtesy translation from the German original 'Andere Lieferanten wie Crown und Canpack konnten unsere Bedarfe seit 2013 nicht vollständig anbieten – Grund hierfür waren laut Crown und Canpack unzureichende Kapazitäten.' Questionnaire to Customers (Q1), question 6.1, ID 1609.

503Courtesy translation from the German original 'Dosen kaufen wir bei den Firmen Ball und Rexam zu. Andere Hersteller wie Crown und Canpack konnten die benötigte Menge nie zu 100 % anbieten.' Questionnaire to Customers (Q1), question 14.1, ID 1609.

504'[Customer] would not be able to switch from all can-manufacturers to just Can-Pack and Crown due to capacity constraints.'

Third, if any spare capacity exists, it tends to be for steel cans rather than for aluminium cans: '[…] suppliers […] have limited spare capacity. Capacity utilisation rates vary by supplier, region, can type and metal type. For instance, capacity utilisation is generally tighter on aluminium cans than on steel cans. Indeed, most suppliers of aluminium cans have a capacity utilisation rate of around 85-90%. The 10-15% of spare capacity is generally used to cover

The levels of spare capacity estimated by third parties is even lower in 2015, implying that the 2014 data overstate the ability of third parties to increase output to offset a price increase.

Moreover, the Commission also notes that in the presence of binding capacity constraints, even if competitors hold relatively large spare capacity, this spare capacity may not be sufficient to prevent significant non-coordinated effects arising from a large increase in concentration and a significant consolidation of capacity (see M.6471-Outokumpu/Inoxum, decision of 7.11.2012 and M.6905- Ineos/SOLVAY/JV, decision of 8.05.2014).

'[d]emand for beverage cans depends on the demand for beverages by end consumers which in turn varies with the season: Consumption of beverages is low in the winter time, grows during spring and peaks in summer to gradually sink in autumn to reach its winter low. For this reason, demand for beverage cans in Q1 of any given year tends to be very low, it picks up and reaches its peak in the summer months, ebbing down to a very low volume in November (with a larger demand in December (due to Christmas and the New

451Capacity is particularly tight in the 'peak' (that is to say in the summer) season, when the weather is hot and there is more demand for beverages: 'Due to seasonal peaks and lows, there are times - mostly in summer - where capacity is not sufficient to satisfy market demand.'

452'Especially in summer time the production capacity is not enough'. One of the competitors explains that while '[t]heoretically installed capacity per annum exceeds production in the market in Europe; however, in practice there is no overcapacity due to the seasonality of demand.'

453Ball in particular had difficulties in meeting demand, in particular during recent peak seasons: '[t]here has been a shortage on […] cans from Ball last year (2014) and this year (2015)'. 'Ball for instance has recently had huge problems with capacity; it used its network to balance it out. In […], Ball delivered to [customer] not from its […] plant, but also from […] due to capacity constraints. 'Especially Ball had problems in the past to deliver cans in time.' Ball also pointed out […] in internal documents: '[…]'.

454Manufacturing cans in the low season (the first quarter of a given year) and storing them until peak season in summer is also possible only to a limited extent. It is for instance not possible to produce cans in December for the following summer given that customer demand changes (in particular in terms of the printing on the cans): '[a]s some design[s] are changed in early summer there is a[l]lways a battle how to get the volume on time.'

As a competitor puts it, '[d]esign considerations influence demand for beverage cans. Because labels are actually printed onto the can during the process of producing the can (not, as in the case of bottles, during the filling process) customers must specify the design at the time the can

505Questionnaire to Customers (Q1), question 57, ID 1609.

506A competitor's non-confidential reply to RFI of 15.06.2015, ID 2572.

9.1.9Capacity is tight in the EEA

9.1.9.1The Notifying Party's views

The Notifying Party argues that both Crown and Can-Pack would have spare capacity to supply any customers that wish to switch suppliers post-Transaction.

486Regarding Can-Pack, it has spare capacity at its plants in the UK and Poland. Can-Pack's plant in Morocco is also able to supply into some of the southern regions of the EEA.

487As for Crown, it has capacity available at its plants in Spain, France, the UK and Slovakia.

9.1.9.2The Commission's assessment

The Commission has reached the conclusion that there is currently limited spare capacity in relation to the supply of beverage cans in the EEA. As a result, if the merged entity were to increase prices post-Transaction, competitors would have insufficient spare capacity to counteract such an increase. Moreover, Post-Transaction, the merged entity would have the incentive to do so.

First, as the Notifying Party notes in the Form CO, capacity utilisation is high in the industry (above 90 %). This was also confirmed both by competitors ('[competitor] believes that the industry in general currently is operating very close to capacity') and customers that responded to the market investigation:

492'there is currently no additional capacity in the market.'

493'[t]he industry has traditionally been very tight (capacity wise).'

494'Ball and Rexam typically try to limit capacity.'

495In particular, contrary to the Notifying Party's claim, Can-Pack confirmed that it 'does not have spare capacity in its Polish and UK plants that would allow it to increase its output in a way that would effectively restrain the merged entity in an appreciable way.' Can-Pack also confirmed that the capacity utilisation at all of its EEA plants is above 90 %.

Second, the capacities of the Parties are important from the point of view of total capacity in the EEA as Crown and Can-Pack are unable to serve large volumes, due to their limited capacity.

The overall spare capacity in 2014 of Crown and Can-Pack for plants located in the EEA was equivalent to only between 5 to 10 % of the combined sales of the Parties to EEA customers in 2014. Crown and Can-Pack would, therefore, not be able to supply sufficient additional volumes to customers in case of a price increase by the merged entity.

The levels of spare capacity estimated by third parties is even lower in 2015, implying that the 2014 data overstate the ability of third parties to increase output to offset a price increase.

Moreover, the Commission also notes that in the presence of binding capacity constraints, even if competitors hold relatively large spare capacity, this spare capacity may not be sufficient to prevent significant non-coordinated effects arising from a large increase in concentration and a significant consolidation of capacity (see M.6471-Outokumpu/Inoxum, decision of 7.11.2012 and M.6905- Ineos/SOLVAY/JV, decision of 8.05.2014).

500Questionnaire to Customers (Q1), question 64.1, ID 1609.

501Questionnaire to Customers (Q1), question 64.1, ID 1609. Courtesy translation from the Spanish original: 'Es el único proveedor que nos suministra las cantidades que nuestra empresa necesita.'

502Courtesy translation from the German original 'Andere Lieferanten wie Crown und Canpack konnten unsere Bedarfe seit 2013 nicht vollständig anbieten – Grund hierfür waren laut Crown und Canpack unzureichende Kapazitäten.' Questionnaire to Customers (Q1), question 6.1, ID 1609.

503Courtesy translation from the German original 'Dosen kaufen wir bei den Firmen Ball und Rexam zu. Andere Hersteller wie Crown und Canpack konnten die benötigte Menge nie zu 100 % anbieten.' Questionnaire to Customers (Q1), question 14.1, ID 1609.

504'[Customer] would not be able to switch from all can-manufacturers to just Can-Pack and Crown due to capacity constraints.'

Third, if any spare capacity exists, it tends to be for steel cans rather than for aluminium cans: '[…] suppliers […] have limited spare capacity. Capacity utilisation rates vary by supplier, region, can type and metal type. For instance, capacity utilisation is generally tighter on aluminium cans than on steel cans. Indeed, most suppliers of aluminium cans have a capacity utilisation rate of around 85-90%. The 10-15% of spare capacity is generally used to cover

The levels of spare capacity estimated by third parties is even lower in 2015, implying that the 2014 data overstate the ability of third parties to increase output to offset a price increase.

Moreover, the Commission also notes that in the presence of binding capacity constraints, even if competitors hold relatively large spare capacity, this spare capacity may not be sufficient to prevent significant non-coordinated effects arising from a large increase in concentration and a significant consolidation of capacity (see M.6471-Outokumpu/Inoxum, decision of 7.11.2012 and M.6905- Ineos/SOLVAY/JV, decision of 8.05.2014).

(465) First, switching is hindered by the fact that capacity is tight, as explained in section 9.1.9. Large volumes are reserved for large customers and '[g]iven the significant can volumes that [large customer] sources in the EEA, the specific can portfolio per filling location and the tight production capacity and limited product portfolio of can suppliers, it is very challenging if not impossible for [large customer] to move significant can volumes from one supplier to another for most if not all its filling plants'. 'We believe there is very limited competition between the beverage can manufacturers, because there are only four producers and they all fill large parts of their production capacity with large volumes from long term customers, like Coca Cola, Pepsi and the large beer producers. As a consequence, there is limited free capacity available in each production plant and potential switches of volumes will be avoided as much as possible, because the beverage can manufacturers know that stability is in the interest of their market.'

525Form CO, paragraph 8.42.

526Reply to the Article 6(1)(c) Decision, paragraph 4.11, ID 1448.

527Form CO, paragraph 6.341.

528Form CO, paragraph 6.219.

529Form CO, paragraph 6.434.

530Form CO, paragraph 6.357.

531Form CO, paragraph 6.362.

533Agreed non-confidential minutes of a call with a customer, 18.08.2015, ID 2866.

534Questionnaire to Customers (Q1), question 97, ID 1609.

(466) Second, as shown in section 9.1.9, the level of spare capacity owned by Crown and Can-Pack is limited, both at the EEA level and in individual regional clusters.

(467) Third, even in cases where customers have switched in the past, such switching was limited to a small percentage of the overall demand of customers as the alternative beverage can manufacturers lacked the capacities necessary to cover all of their volumes.

(468) As explained by a customer, '[customer] would not be able to switch from all can-manufacturers to just Can-Pack and Crown due to capacity constraints. Especially for [customer]'s filling locations in the Netherlands no competitor of Ball/Rexam would be able to provide the total amount of cans needed by [customer]. Even without consideration of transportation costs, shifting all volumes away from Ball and Rexam is not an option.'

(469) Similarly, Molson Coors explained that the reason why their switched was 'the introduction of an alternative/backup supplier for our filling plant that was dependant on one supplier prior to that'. This volume covers, however, only […] % of Molson Coors' demand.

535Questionnaire to Customers (Q1), question 55, ID 1609. See also agreed non-confidential minutes of a call with a customer, 25.08.2015, ID 2938: '[Customer's] cans volumes are small and hence this could be potentially be overcome by trying to switch volumes to another supplier. Maybe with bigger volumes this would not be feasible'.

536Agreed non-confidential minutes of a call with a customer, 12.05.2015, ID 468.

537Questionnaire to Customers (Q1), question 55, ID 1609.

538Questionnaire to Customers (Q1), question 55, ID 1609.

539Questionnaire to Customers (Q2), question 52, ID 2664.

540Questionnaire to Customers (Q1), question 55, ID 1609.

541Questionnaire to Customers (Q1), question 57, ID 1609.

542Questionnaire to Customers (Q1), question 56, ID 1609.

543Questionnaire to Customers (Q1), question 56, ID 1609.

[C]hanging supplier means extra costs for new artworks and printing plates - so it is always involving cost', 'Developing the artworks with a different supplier is a significant cost.' 'We use many different designs an[d] languages on our cans. If we would switch suppliers, for all those variations a new good for print would need to be developed to fit the manufacturing process of the new supplier. Such procedure would be very costly. As well it is not guaranteed that our filling stations can fill cans from another supplier without modifications in their filling line.' Questionnaire to Customers (Q1), question 56.1, ID 1609.

(474) Sixth, as explained further in section 9.1.13 below, the threat of switching from cans to other materials, such as PET is not credible.

9.1.11. Multi-sourcing

9.1.11.1. The Notifying Party's views

(475) According to the Notifying Party, several customers multi-source by having more than one supplier at a specific filling location. However, in most cases a customer filling location is supplied by a single supplier.

9.1.11.2. The Commission's assessment

(476) The Commission has reached the conclusion that multi-sourcing is an important factor in the beverage can industry, even on a filling plant level. As a result, the reduction in the number of competitive alternative suppliers post-Transaction is likely to reinforce the anticompetitive effects of the Transaction.

(477) First, many customers in the EEA, in particular larger ones, multi-source: 'Usually volumes are split between the can manufacturers in a global tender process. Nearly all top European customers multi-source, with the exception of Red Bull […].' 'Our group strategy is to source from minimum 2 suppliers.'

548Agreed non-confidential minutes of a meeting with a competitor, 13.05.2015, ID 1065.

549Questionnaire to Customers (Q1), question 6.1, ID 1609.

(478) The reason why multi-sourcing for smaller customers is less frequent is due to their relatively lower volumes: 'Our volume is not that much that we can split it in two suppliers because of getting the best price'; 'Volume was not that much to close contracts with 2 or more suppliers'; 'Our volume does not allow (at least it does not seem reco[m]mendable) to share sources for the same can type.'

550Questionnaire to Customers (Q1), question 53, ID 1609.

551Questionnaire to Customers (Q1), question 53, ID 1609.

552Questionnaire to Customers (Q1), question 53, ID 1609.

(479) Second, multi-sourcing is important for customers because of security of supply considerations and in order to obtain better commercial terms: '[customer] keeps two can suppliers to diversify risk and ensure it gets best conditions.' '[I]t is important for [customer] to have several actual or potential sources of supply for each of its plants, as business continuity. The multi-sourcing approach has also allowed [customer] to secure fair can pricing through competition between these can making plants.' '[O]ptimize costs and minimize risks of supply'.

554Questionnaire to Customers (Q1), question 53, ID 1609.

555Questionnaire to Customers (Q1), question 53, ID 1609.

(480) Third, multi-sourcing is a common practice both across multiple filling locations and at filling location level. The Parties' transaction data shows that approximately […] of the filling locations in the EEA source their can requirements from two or more beverage can manufacturers for any given can size. This ratio is […] ([…] %) for Rexam, […] % of whose customers single-source; and […](between […] % and […] %) for Ball, Crown and Can-Pack.

545Questionnaire to Customers (Q2), question 37, ID 2664.

546Form CO, paragraph 6.30.

(481) Multi-sourcing by large customers is often due to footprint considerations: 'The production network footprints of Ball and Rexam in the EEA are especially well overlapping with each other and with [major customer's] footprint. [Major customer] has further established a global sourcing strategy which is based upon having a minimum of two strategic global suppliers of which Rexam has fulfilled one of these positions, and Ball often has the second one. Multi-sourcing from Ball and Rexam at the same time is for [major customer] often a convenient solution because of the can manufacturers' overlapping network and same level of quality and services.'

9.1.12. Sales to wall-to-wall customers

(482) The Parties have six plants that are located wall-to-wall or in the immediate proximity of a customer. These are the Ball plants in Bierne (France) and Rugby (UK) and the Rexam plants in Ludesch (Austria), Milton Keynes (UK) and Wakefield (UK). In addition, Rexam has built a plant in Widnau (Switzerland) which started operations in 2015.

(483) Wall-to-wall sales from Ludesch and Widnau plants are further assessed in section 9.1.13 as these plants mainly serve, or are expected to mainly serve […].

9.1.12.1. The Notifying Party's views

(484) The Notifying Party argues that the Parties' sales to wall-to-wall and co-located customers 'cannot be considered truly merchant market' because of the competitive advantage which results from the co-location next to the customer. For instance the Notifying Party explains that it does not compete to supply […] filling locations in the […] region because of its freight cost disadvantage compared to Rexam's […].

9.1.12.2. The Commission's assessment

(486) The Commission has reached the conclusion that while both can beverage manufacturers and customers actively try to minimise transport costs by reducing travel distances from can plants to filling stations, sales to filling locations that are in the immediate vicinity, or indeed adjacent, to a can plant should be included in the merchant market. This is for the following reasons.

(487) First, when beverage can manufacturers are interested in developing a new customer relationship or strengthening an existing one in a geographic area where they do not have a plant in close vicinity, they accept to incur higher transport costs.

556Commission analysis based on actual transaction data supplied by the Parties and estimates of the Notifying Party for third parties.

557Questionnaire to Customers (Q1), question 62.1, ID 1609.

(488) Second, customers generally multi-source their can requirements and therefore do not necessarily source all of their can requirements from the closest plant.

(489) Third, customer filling locations in the immediate vicinity to a can manufacturer's plant still benefit from competition in the market because they are able to switch at least part of their demand to other beverage can manufacturers and can leverage competing offers to obtain better terms from the co-located or wall-to-wall supplier.

(490) Fourth, transport costs are low over short distances. The freight cost for transporting 1 000 cans for 200 km or less is approximately EUR […] and about […] % for 300 km. In this respect, it must be noted that both Crown and Can-Pack have plants that are located within a radius of approximately 200 km from the Parties' wall-to-wall and co-located plants in the UK. Therefore, the transport costs are comparably low. For example, regarding Ball's Rugby plant Crown owns the closest competing plant, which is on average […] km away from the customers served by Rugby. At Milton Keynes, Crown owns the second closest competing plant, which is […] km away from the customers served by Rexam's plant. At Wakefield both Crown and Can-Pack own plants that are the closest competing plants for a significant number of customers within 200 km from those customers on average.

(491) Fifth, as discussed in section 7.1.3.3 above, while closeness to the filling location is a desirable attribute in a can supplier as it facilitates logistics and support, it is not a pre-condition to entering into a customer/supplier relationship.

(492) Sixth, all contracts with wall-to-wall and co-located customers will be contestable in the short- to mid-term as they are set to expire at the latest in […]. The sole exception is […] that is set to expire in […] and is discussed separately in section 9.1.13.

9.1.13. Sales to Red Bull

(493) […]. Rexam's Ludesch plant is located wall-to-wall with a Red Bull filling location operated by Rauch Fruchtsäfte, Red Bull's main contract filler.

(494) As mentioned in section 5.1.4.3, Red Bull […].

9.1.13.1. The Notifying Party's views

(495) The Notifying Party argues that Rexam's sales to Red Bull should be excluded from the merchant market as they are not contestable in the short- to medium-term.

(496) Pursuant to […] and, […]. The Notifying Party is of the view that […].

(497) The Notifying Party submits that sales to Red Bull are not contestable due to […].

(498) First, […]. This is for the reasons set out in recitals (499)-(500).

(499) In the first place, […].

561Agreed non-confidential minutes of a call with a customer, 04.05.2015, ID 334: '[A]supplier [can] actively subsidizes the transportation cost in favour of winning the business, as we have seen done by Rexam to win [customer] Poland business, or Ball to win [customer] Portugal business'.

562Questionnaire to Customers (Q1), question 5, ID 1609.

563Agreed non-confidential minutes of a meeting with a competitor, 13.05.2015, ID 1065.

564Reply to the Article 6(1)(c) Decision, paragraph 3.8 and Figure 3.3.

565Form CO, Attachment G, RFI 1, Exhibit K.

566The […] is set to expire on […]. Form CO, Attachment G, RFI 1, Exhibit H.182, clause 10.1.

567Form CO, Attachment G, RFI 1, Exhibit H.182, clause 2.1.

568Reply to the SO, paragraphs 4.5-4.14.

(500) In the second place, it is highly unlikely that […]. Accordingly, this mechanism is only likely to be relevant where […].

(501) Second, […] of Rexam's plants in […] will be […]. The beverage cans manufactured in these plants will, therefore, […].

(502) The Notifying Party further submits that on the basis of the Commission including wall-to-wall plants in the merchant market because their contracts will expire at the latest in […], Red Bull volumes ought to be excluded from the market share calculations because they are not contestable for at least […] more years (until at least […]).

9.1.13.2. The Commission's assessment

(503) For the reasons set out in this section, the Commission has reached the conclusion that Rexam's sales to Red Bull should be included in the merchant market […].

(504) First, Rexam's sales to Red Bull provide an indication of its competitive strength. The fact that Rexam has won significant business from a high demand customer such as Red Bull, which accounts for nearly […] % of total beverage can demand in the EEA, is a confirmation of Rexam's success and ability to compete.

(505) Moreover, as explained in section 9.1.7, various advantages arise from having a broad network of plants and a greater scale than other beverage can manufacturers. It follows, that Red Bull's demand and the Rexam plants that serve it materially influence Rexam's overall competitiveness.

(506) Second, contrary to what the Notifying Party argues, […]. This is for the reasons set out in (507)-(511).

(507) In the first place, the Red Bull agreement provides that Red Bull (i) […], and (ii) that […]. It follows that, contrary to what the Notifying Party argues, Red Bull is […].

(508) In the second place, Red Bull can […]. Red Bull can therefore […].

(509) In the third place, a Rexam internal document (see Figure 21) […].

Figure 21: Internal Rexam document on 25 cl cans

[…]

Source: Rexam's internal document, 'Group strategy 2015-17 – Executive strategy', dated October 2014, page 12

(510) In the fourth place, pre-Transaction Ball was […].

(511) In the fifth place, Red Bull in the short-term will be mainly […].

(512) Third, as to the Notifying Party's argument that certain Rexam plants will mainly serve Red Bull and that therefore the beverage cans manufactured in these plants will not be available to other customers, the […] (i) […]; and (ii) […].

(513) Furthermore, in particular in a capacity-constrained industry such as the beverage can manufacturing, it is not uncommon that production lines are dedicated to the production of cans for one or more customers and that, once their capacity is fully utilised, they cannot serve other customers. Therefore, even assuming that Rexam's […] plants were to mainly serve Red Bull, this would not be a unique arrangement in the industry or one that would justify excluding the capacity of these plants from the merchant market.

9.1.14. There is limited competitive pressure from other forms of packaging

9.1.14.1. The Notifying Party's views

(514) The Notifying Party argues that customers can and regularly do 'flex' their packaging mix; this demonstrates that they have the ability to switch (at the very least) marginal volumes between materials such as PET or glass. The Parties monitor and react to competition from other materials on an ongoing basis. As such, even if other packaging materials are not considered as part of the same product market as beverage cans, they do impose a competitive constraint on can manufacturers.

(515) In its Reply to the SO, the Parties put forward a list of examples from internal documents to illustrate the fact that cans are constrained by other forms of packaging. Figure 22 is one of those examples.

Figure 22: Yearly market trends in packaging mix

[...]

Source: Ball's internal document 'Global Beverage Can Market based on 2012 Global Market Review'

9.1.14.2. The Commission's assessment

(516) The Commission has reached the conclusion that there is a lack of competitive pressure from other forms of packaging (such as PET or glass) on beverage can manufacturers. This is for the following reasons.

(517) First, competitive conditions are different for each type of beverage packaging as cost drivers are different (aluminium costs compared to oil costs). For example, prices for PET do not move in parallel with prices for cans. According to one competitor, this indicates that 'these products are not linked in a way that would suggest that changes in the competitive conditions of one product automatically result in parallel changes to the other product'.

(518) Second, as set out in detail in section 6.1 above, a customer's choice of packaging mix is determined by end-consumer needs and is not primarily price driven. For instance, a competitor stated that the '[p]ackaging mix […] is not influenced by changes in prices of the different forms of packaging. Rather final customers' preferences, distribution channels, recycling and environmental issues as well as having the relevant filling lines impact the packaging choice.' Customers confirm that the 'packaging mix is decided by the end consumer, beverage manufacturers will chose such packaging as is required by their clients.' Moreover, '[i]n all markets, cans are unique containers not substitutable with PET/plastic bottles or glass bottles. Each of these types of packaging serves a different consumption occasion/need and consumer perception.'

573Form CO, paragraphs 6.90-6.100.

574Questionnaire to Customers (Q1), question 13, ID 1609: 'Price of the packaging does count but for movements between cans, pet, glass, cartons takes years to change because of pricing' and Questionnaire to Customers (Q1), question 14.1, ID 1609: '[C]ost driver and suppliers are in most cases very different and sometimes even different buyer are responsible for the different packaging types'; 'Different cost drivers within the packaging types, e.g. Glass – Soda Ash; Can – Aluminum, PET – Oil'; 'Many of the packaging categories have different substrates (metal, polymers, glass, paper, etc.), most of which are linked to different cost drivers'; 'Separate materials have different dynamics and most of the times different suppliers and therefore require separate tenders'.

(519) Third, each type of packaging is purchased separately and through different tender processes, 'given that each packaging type has a totally different and independent supplier base.'

(520) Fourth, more than two-thirds of all customers who responded to the market investigation indicated that in the event of a permanent 5-10 % increase in the price of cans, they would not replace any of their purchases of cans by either PET or glass bottles as they lack sufficient flexibility in terms of filling lines.

(521) Fifth, of the respondents that indicated they would replace some purchases of cans, a vast majority said they would do so only with regard to less than 30 % of their can demand. This is because material changes in the packaging mix would require significant changes in bottling equipment because filling lines are specific to a certain form of beverage packaging (cans, PET, or glass) and different equipment is needed in the manufacturing process of different beverage packaging, namely de-palletising, filling, closing, labelling, and packaging. Switching to other types of packaging would require beverage manufacturers or contract fillers to realign their business, invest in new production lines, and source different raw materials (aluminium, polymer). This was confirmed by a competitor which considers that 'a filling line can't, for technical reasons, be "adjusted" to fill bottles instead of cans or vice versa. If a beverage producer wishes to make such a change, it will have to design, purchase, and install a new filling line. Such a step will take significant amount of time and incur significant costs'. As customers explained '[w]e have fixed volume capacities for each packaging material, and need to manage the mix within our production capabilities. A significant increase in any one packaging type would require investment in equipment, require the removal of existing equipment and reduce existing production capability' and '[a]ny large change does require planning and investment and as such usually takes 9-12 months'. But even customers vertically integrated, for example in the in-house production of PET bottles, that may have a greater ability to move some can volumes to PET bottles, are unlikely to switch because, as set

577Questionnaire to Customers (Q1), question 13, ID 1609.

578Agreed non-confidential minutes of a call with a customer, 23.04.2015, ID 129.

579Questionnaire to Customers (Q1), question 14 and 14.1, ID 1609. For example: 'Separate tenders are necessary because these are different supply markets with different market participants''; 'We consider different types of packaging as different and non-substitutable materials and industries with separate suppliers''; 'We see different supplier markets and our procurement function is also split in different packag[ing] area's''; 'Each package type is handled by different category teams and they are not jointly negotiated''; 'For each of the materials we conduct separate tenders beca[use] they have different suppliers. Only a few suppliers can supply more than one packaging type, but even then we do separate tenders per material''; 'Supplier base within the different types of packaging are different, no "global/regional" suppliers sell both glass bottles, PET and Cans. No volume synergies through tendering total volume of all packaging types. […] Different contract durations due to different competitive dynamics'.

580Questionnaire to Customers (Q1), questions 17 and 19, ID 1609.

581A competitor's non-confidential reply to RFI of 15.06.2015, ID 2572.

(522) Sixth, the Parties' internal documents submitted in the Reply to the SO do not support a finding that cans are materially constrained by glass, PET, and other forms of packaging.

(523) Figure 23 analyses market trends regarding the packaging mix used by customers, rather than an analysis of substitution between cans and other forms of packaging. […].

(524) Slide 30 of Ball's internal document 'BPE Presentation – Metal' and slide 16 of Ball's internal document 'Ball Corporation – IBP Demand Review' relate more to the […].

(525) Slide 37 of Ball's internal document 'Region Strategic Plan 2014-2015', slide 21 of Ball's internal document 'Global Beverage Can Market based on 2012 Global Market Review', slide 23 of Ball's internal document 'BPE Presentation Metal' and Rexam's internal email from Marc Stacey to Ann Healy on 8 May 2013 indicate that regional variations in trends across the EEA with respect to the growth of one packaging type over another are the result of the specific economic situation in the country or a change in consumer behaviour: […] ; […] ; or […].

(526) Internal Rexam's emails that […] switched to other types of packaging are not indicative of cans being materially constrained by other types of packaging materials as the volumes of cans that […] switched to other types of packaging is unknown and could be insignificant compared to […] overall demand for cans.

(527) As regards Rexam’s strategy plan for 2015-2017 which states […] , it relates to the supply of beverage cans in North America not the EEA.

(528) Seventh, contrary to the Notifying Party's claims, the majority of customers have had a stable packaging mix over the past 3 years. As explained by customers: '[t]he packaging portfolio is very stable and the packaging producers know very well that we are not able to switch from one to the other'. 'Our overall volume mix between packaging formats has been relatively stable over the last 3 years.' 'We have no flexibility in changing our packaging material.' 'We have not switched significant volumes from one packaging type to another. The packaging mix is fairly stable over time.' 'No flexibility to adjust volumes of each packaging material in your overall mix.'

600Questionnaire to Customers (Q1), question 11, ID 1609.

601Questionnaire to Customers (Q1), question 11, ID 1609.

(529) Eighth, most customers indicated that they have not used the threat of switching away from cans to other packaging type in their negotiations with suppliers as '[t]his argument doesn't make sense. The market (=consumers) will decide, what they buy'. In the words of one customer, 'it is not [the customer] but the consumer that determines the packaging mix. Therefore it would not make sense to use the switching argument with a packaging supplier'. As another customer explained '[t]his argument would not be considered very reliable. Can makers know very well that we [cannot] easily switch to PET bottles. Energy drinks are basically always packed in alu cans. CSD's can theoretically be filled in cans and in PET, but we do not observe any movements between the two. Family sizes are normally sold in PET and portion sizes often in cans. Still drinks, juices and juice drinks are hardly sold in cans and are therefore not relevant for the switching argument'.

602Questionnaire to Customers (Q1), questions 12 and 12.1, ID 1609.

603Questionnaire to Customers (Q1), question 12.1, ID 1609.

(530) Ninth, even the minority of the customers that have used the threat of switching away from cans to other packaging types consider this argument of limited effectiveness as it 'poses a credible threat only if price increase is significant' and '[t]he consumer ultimately dictates the choice of packaging'.

9.1.15. Barriers to entry and expansion are high

9.1.15.1. The Notifying Party's views

(531) The Notifying Party argues that barriers to entry are low.

(532) First, any intellectual property ('IP') and know-how required to manufacture beverage cans is readily available from third parties.

(533) Second, the equipment required to produce beverage cans is readily available from third party suppliers such as Stolle Machinery and Belvac Production Machinery.

(534) Third, the aluminium or steel required to produce beverage cans are commodities which are readily accessible to new entrants.

(535) Fourth, in the past 10 years, Ball, Rexam, Crown and Can-Pack have each constructed plants which initially had only one line (the plants include Belgrade, Scunthorpe, Kechnec and Mäntsälä). Moreover, Helvetia has recently entered the market with a plant which initially has only one line, demonstrating that it remains economically feasible for a newly-built production facility to have only one line.

(536) Fifth, the mere threat of entry also constrains prices. For example, Can-Pack has confirmed to the Commission that the threat of its entry into Spain resulted in the incumbent beverage can manufacturers substantially reducing prices.

600Questionnaire to Customers (Q1), question 11.1, ID 1609.

601Questionnaire to Customers (Q1), question 11.1, ID 1609.

606Reply to the SO, Annex 1.

607Form CO, paragraph 8.58.

608Form CO, paragraph 8.58.

(537) Sixth, players from outside the EEA (for instance South African Nampak Ltd., 'Nampak') could enter the EEA. Such entry would not be prevented by certain technical agreements that prevent the use of certain technology in the EEA by these non-EEA players as these players could use another technology.

9.1.15.2. The Commission's assessment

(538) The Commission has reached the conclusion that barriers to entry and expansion are high. This is for the following reasons.

(539) First, the existence of these high barriers to entry and expansion is generally confirmed by Rexam, customers and competitors alike.

(540) As Rexam itself puts it on its website, '[the beverage can industry] as a whole is characterised by high barriers to entry with the scale economies afforded by numerous production plants favouring the major players. Freight costs are substantial and in large regions a key driver of competitive advantage is an optimised network of plants in close proximity to customer filling locations.'[…].

(541) As a competitor explains in detail, '[t]here exists a number of significant barriers to entry into the European beverage can market: cost of investment – to obtain 10 % of EAA market would require a new investment of approximately 290 million EUR; entering this particular business would require specific know-how and access to the relevant technology; there exist significant economies of scale with respect to sourcing raw materials which it could be difficult to match with a single plant; the long-lasting process of purchasing machinery; to successfully enter the European market, a number of customer volum[e] commitments would be helpful to make the investment cost effective; the process of establishing a new can manufacturing plants is drawn out and [time consuming]; to be considered as a supplier by large customers new suppliers have to go through lengthy process of accreditation with each filler/bottler they would like to sell to; Quality issues.'

(542) According to customers, 'entry barriers are high and come in form of: a) large investments needed to establish a can manufacturing plant (the investment for a greenfield plant app. EUR 150 million), b) significant start-up production scale (more than 700 million cans per year), c) technological and know-how barriers linked to capability to produce cans efficiently to be able to compete effectively, d) understanding of and compliance with local legislation and regulations, e) requirement for long-term fixed volume customer agreements for can makers to justify the investment, f) the structure of the demand which fluctuates throughout the year due to its seasonability.' 'Very high barriers to entry e.g. high capital investment, required expertise etc.' 'Entry barriers are high because of existing big players and required scale and investments and knowledge;' 'Yes, the entry barriers are relatively high, because having only one plant won't be sufficient and within each plant one should have at least two efficient production lines, requiring significant investments'; 'Yes the minimum needed scale is the most important to be

610Form CO, paragraph 6.437.

611'Entry', in this Decision refers to entry of a new player into the EEA while 'expansion' refers to expansion by existing players within the EEA.

612Questionnaire to Customers (Q2), question 46.1, ID 2664.

613See Rexam's internal document 'FY12 results announcement, Encouraging performance underpinned by Beverage Cans volume and profit growth', slide 9.

614A competitor's non-confidential reply to RFI of 15.06.2015, ID 2572.

(543) Second, setting up a new, efficient beverage can manufacturing operation is costly and requires time, expertise and know-how as well as large volume long-term commitments from customers. This is demonstrated in recitals (544) to (559).

(544) In the first place, investment costs are between approximately EUR 50-100 million for a new plant with one production line. Such a one-line operation would have to have a minimum scale which is considered by market participants to be around at least 700 million cans per year. A new player would also have to establish a relationship with raw material suppliers.

(545) In the second place, setting up a new plant requires a lead time of approximately a year and a half and getting a new plant to operate efficiently takes approximately another year and a half. It may therefore, take 3-4 years before a plant is built and running efficiently. This was emphasised by several customers that responded to the market investigation 'Setting up a new plant takes roughly 1.5 years of construction and another 1.5 years from the start of the production to operating at full efficiency'; 'A new entrant would have to pass a number of hurdles to successfully enter the European beverage can market [including] […] time to build a new can plant – 18-24 months reaching high production performance – 12-18 months with support from experienced staff; time to obtain the approval from customers – an estimated 6 - 12 months from starting stable production, establishing relationships with customers 1 – 3 years, […]'.

(546) In the third place, it is unlikely that a new entrant would make the investment required to build a new plant without having first secured a sufficient volume from customers for at least a five-year period. Such a volume commitment has to be both sufficiently large and for a longer time span.

(547) As regards volumes, it seems that a commitment would have to represent at least 50 % of the capacity of the beverage can production plant to allow a beverage can manufacturer to make the decision to invest; even that may not be enough. For example, one of the Parties […] A minimum of 300-600 million cans is generally considered by market participants to be the minimum volume commitment: 'Efficiency is key, and therefore high volumes are required.

615Questionnaire to Customers (Q1), question 88, ID 1609.

616Agreed non-confidential minutes of a call with a customer, 07.05.2015, ID 326. Agreed non-confidential minutes of a call with a customer, 04.05.2015, ID 157. 'Entry barriers are high in beverage can manufacturing. Apart from the scale and technical [know-how] needed […], also considerable capital is required. The costs to build one can plant with two high speed lines and can end manufacturing is estimated at Euro 80 - 120 million.' Questionnaire to Customers (Q1), question 88, ID 1609.

617Agreed non-confidential minutes of a meeting with a competitor, 13.05.2015, ID 1065.

618Agreed non-confidential minutes of a meeting with a competitor, 13.05.2015, ID 1065.

(548) As a competitor put it, '[a]s beverage can making industry is very capital intensive, building new production capacity or considerably extending existing production lines is very risky. Thus, to some extent customers commitments are a necessary – but not a sufficient – factor in reaching a positive investment decision. It should be remembered that investing in new capacity always requires a positive business case for such an investment.' The same competitor also confirmed the necessity of both a minimum capacity commitment as well as a minimum duration of the contract: '[p]rovided that the other requirements are met, CP would require a volume commitment of at least […] of the total capacity of a new investment contracted by single customer for a duration of […] years or a commitment of at least […] of total capacity from several customers to initiate the investment.'

(549) As regards duration, a short-term commitment would not be sufficient for such an investment. As a competitor explained, '[it] does not make investments in exchange for short-term commitment (1 year contracts) because it does not provide the necessary certainty that the investments will pay off.'

(550) One major customer explained that '[f]or greenfield operations encouraged or initiated by [it], [it] usually commit[s] to purchase volumes for 5 years or longer.' Respondents to the market investigation explained that a minimum of 5-7 years is necessary for 'a supplier to gain a return on investment'.

(551) The need for a new entrant to obtain a sufficiently large and long volume commitment before deciding to build a new plant is confirmed by the recent instances of expansion by the current players:

– Rexam's Fredericia plant in Denmark was sponsored by […], whereby it contracted […] % of its demand driven volume (approx. […] billion cans per year) for a period of […] years.

– Before building its Finnish plant in Mäntsälä, Rexam obtained […].[…] .

– […] also sponsored Rexam's […] plant in […] in […] and the addition of a new line at […] in […].

– Can-Pack's Scunthorpe plant, which initially had one line, was sponsored by Carlsberg via a long-term supply agreement. According to Can-Pack, the contract which facilitated Scunthorpe was 'signed for a [1-5-year] period, with an annual volume of [100-500 million cans].' The building of the plant was also facilitated by contracts with

625Questionnaire to Customers (Q1), question 88, ID 1609.

626A competitor's non-confidential reply to RFI of 18.08.2015, ID 2874.

627A competitor's non-confidential reply to RFI of 18.08.2015, ID 2874.

628Agreed non-confidential minutes of a meeting with a competitor, 05.05.2015, ID 261.

629Agreed non-confidential minutes of a call with a customer, 04.05.2015, ID 334.

630Questionnaire to Customers (Q2), question 49, ID 2664.

– Rexam's internal document 'FY12 results announcement, Encouraging performance underpinned by Beverage Cans volume and profit growth', slide 9.

– The installation of a second line at Can-Pack's Scunthorpe plant in 2014 was sponsored by contracts with several customers: 'A [1-5-year] contract with […] for [10 % - 50 %] of their volume requirement (estimated at 120 million cans annually) was concluded. Additionally CP entered into [1-5-year] contract with […] for [100 – 500 million cans] annually.

– The extension at Can-Pack's Brzesko plant was facilitated by several customer contracts. According to Can-Pack these included a 'new [1-5-year] agreement with […] as well as the extension of existing contracts with […] for another [1-5 years].

– Can-Pack's Finnish Hämeenlinna production facility was sponsored through a long-term supply agreement.

– Crown's Slovak plant in Kechnec was built/expanded following volume commitments by Molson Coors and by SAB Miller. According to Crown, at least three customers contracted volumes in relation to the building of Kechnec's first line. The building of the second line also came about due to a contract with a customer.

– Crown converted one of 4 lines in its Botcherby plant from running 44cl/50 cl cans to run 25 cl cans. According to Crown, this was necessary because of a new contract with a major customer due to its growing demand.

625Questionnaire to Customers (Q1), question 88, ID 1609.

626A competitor's non-confidential reply to RFI of 18.08.2015, ID 2874.

627A competitor's non-confidential reply to RFI of 18.08.2015, ID 2874.

628Agreed non-confidential minutes of a meeting with a competitor, 05.05.2015, ID 261.

629Agreed non-confidential minutes of a call with a customer, 04.05.2015, ID 334.

630Questionnaire to Customers (Q2), question 49, ID 2664.

The need to have high long-term volume commitments from customers is in line with the fact that capacity utilisation needs to be high in order for a beverage can plant to be viable. This has been confirmed not only by competitors: 'Since manufacturing cans is a very volume sensitive business, the production line(s) need(s) to be as full as possible. It is necessary to run a can- plant 7 days a week 24 hours a day (and not just 5 days a week.''[C]apacity utilisation is a key element of a successful business strategy in this sector' but also by […].

636A competitor's revised non-confidential reply to certain questions of the RFI of 18.08.2015, ID 3497.

637Form CO, Paragraph 6.442.

638A competitor's revised non-confidential reply to certain questions of the RFI of 18.08.2015, ID 3497.

639A competitor's revised non-confidential reply to certain questions of the RFI of 18.08.2015, ID 3497.

640Form CO, Paragraph 6.415(iii).

641Questionnaire to Customers (Q1), question 89, ID 1609.

642Form CO, Paragraph 6.42.

643'The building of the first line in Kechnec in 2010 was associated with a [CONFIDENTIAL] year supply agreement between Crown and [CONFIDENTIAL] for approximately [CONFIDENTIAL]% of the output. We also agreed contracts with [CONFIDENTIAL] and [CONFIDENTIAL] which brought the contracted volume to approximately [CONFIDENTIAL]%.' A competitor's non-confidential reply to RFI of 18.08.2015, ID 2537.

644'The second line in Kechnec came about due to a contract with [CONFIDENTIAL]. This contract had a term of [CONFIDENTIAL] years and was for approximately [CONFIDENTIAL]% of the output of the line.' A competitor's non-confidential reply to RFI of 18.08.2015, ID 2537.

Contrary to the Notifying Party's claim (Reply to the SO, paragraph 5.46.), Crown did not, however, build a second line at Carlisle in 2011: 'We did not build a second line in Carlisle in 2011. However, we did convert one of 4 lines in our Botcherby plant from running 44cl/50cl cans to run 25cl cans. This was necessary because we have a Global contract with [CONFIDENTIAL] for 100% of their can requirements, and their growing demand in UK for this size had exhausted our capacity.' A competitor's non-confidential reply to RFI of 18.08.2015, ID 2537.

646Agreed non-confidential minutes of a meeting with a competitor, 13.05.2015, ID 1065.

647Third party comments to the SO, ID 3451.

(553) In the fourth place, smaller customers would be unable to 'sponsor' entry due to the fact that their low volumes would be insufficient to motivate an investment decision by a new entrant.

(554) In the fifth place, even large customers will have difficulties in sponsoring entry if their volumes (or indeed the estimated demand in the relevant geographic area) are insufficient: '[major customer] has had discussions with Ball, [Can-Pack] and Crown concerning entry into Scandinavia but they did not lead to any concrete results – the three Scandinavian countries are very different with only Denmark having large enough market (from the perspective of can suppliers) to merit establishing a can plant. Sweden and Norway do not have sufficiently big markets and the commitment of [major customer] volumes would not have been sufficient to incite entry, a new entrant would need to find other customers as well and this could have been difficult.'

(555) In the sixth place, large customers are unlikely to offer other financial incentives to new suppliers in order to sponsor entry. According to one customer, while it would 'consider working with a new entrant or existing supplier that is expanding[, they] would not provide financial investment.' According to another customer, 'the investment costs are very high and the suppliers should have the financial strength themselves.' Competitors also noted that '[s]ponsored entries do not exist. […] the can manufacturers still have the risks of the investment'.

(556) In the seventh place, customers are also hesitant about sponsoring a fresh entrant in the EEA. Customers explained that '[they] would need to be assured of that quality and service standards can be maintained and we would be conservative with regards to our exposure' and 'there would be still significant challenges in identifying a credible supplier'.

(557) In the eighth place, a large majority of customers indicated that they do not use the argument of sponsoring expansion or entry of new players in the commercial negotiations with beverage can manufacturers.

(558) In the ninth place, a certain degree of expertise and know-how is required to manufacture beverage cans efficiently. A competitor noted that '[b]everage can-making is a highly engineered process with bespoke complex equipment and detailed operating procedures. Whilst any new entrant would be able to draw upon the expertise of equipment manufacturers during the initial install phase (at a cost), the running of a successful and efficient beverage can making facility requires a highly skilled workforce with significant ongoing technical support and input.' It has been also added by the same competitor that can manufacturing needs '[…]

648Reply to RFI 17 of 28.10.2015, […], ID 3448: […].

649Questionnaire to Customers (Q1), question 89, ID 1609.

(560) Third, it is questionable whether a new entrant with one plant and with one production line would have the necessary scale to compete efficiently. This is demonstrated in recitals (561) to (570).

(561) In the first place, a majority of customers that responded to the market investigation considered that the minimum scale of a plant should include two production lines: 'A standard aluminium line produces approximately 750 million cans per year and generally can factories have two such lines.' As competitors also put it, '[s]tarting a new plant with one line is possible, but two production lines are required for economic operation' and '[i]n general having more than one line is more efficient because it allows for a better spread of fixed costs.'

(562) This applies equally to EEA and non-EEA players. Respondents to the market investigation that expressed a view indicated that it is unlikely that a player currently not present would enter the EEA market within the next 3-5 years because of the need to enter with more than one plant. For instance a customer explained the difficulties to enter as follows: 'Entry from a supplier from Africa or Asia into the European market is highly unlikely. At least 3-4 plants would be needed to operate economically, as a certain scale and network are needed. To have a full coverage of Europe, at least 8-12 can-plants would be needed. Moreover, Asian can manufacturers are considered to produce under a lower standard. Also, the Asian and African markets have growth-potential, which is why suppliers not active in Europe would rather invest in those regions.' '[A]n original entry of a new beverage can supplier is not to be expected in the next 3-5 years due to considerable barriers to entry.'

660Agreed non-confidential minutes of a meeting with a competitor, 13.05.2015, ID 1065.

661Agreed non-confidential minutes of a meeting with a competitor, 05.05.2015, ID 261.

662Agreed non-confidential minutes of a call with a supplier, 14.07.2015, ID 1403.

(563) Scale also matters in obtaining good prices for the raw materials and spreading investments in R&D.

(564) As regards raw materials, a competitor noted that '[a] local player entering would have to make significant investments and may have to pay a higher price for metal. Some of the clients may not only be local and the large multi-national ones prefer to deal with large suppliers.', '[…] when entering this market, one needs to reach scale in order to purchase aluminium coil at competitive prices.' A customer also noted that '[i]f Rexam and Ball merge, they will be a very strong force on the market and they will be able to buy their raw materials at cheaper rates.'

(565) With regard to R&D, as stated by a competitor, '[c]ertain investments in know-how and R&D are needed. For instance, with the recent change in regulation in France, R&D capacity was needed to quickly provide cans in line with the new requirements (BPA-free).' Moreover, this same competitor notes in relation to its R&D activities that '[i]t would not be feasible for a player with only one plant to undertake such R&D activities'.

(566) This is because scale allows the fixed costs of R&D to be spread over a larger volume. As a competitor explained, '[g]iven that the costs of innovations are similar for all undertakings, the scale of a large competitor and thus the volume of products to which the innovation is applied allows for a faster and more secure return on its investment. This mechanism makes it easier for large undertakings to innovate than for smaller suppliers.'

(567) In the second place, apart from scale, one production line cannot efficiently manufacture all the different types and sizes of cans required. As a competitor puts it, '[i]t would be more difficult for a small player to deal with the proliferation of can sizes.'

(568) This is because, as indicated in an internal Ball document, such a swing line would […]. Any new plant with just one production line would therefore most likely be limited to a few sizes of cans of one particular diameter. This was also confirmed by a competitor.

(569) In the third place, entry into the market with one production line is insufficient to compete effectively. The majority of large customers who expressed a view stated that it would be not sufficient for a newly build production facility to have only one line: 'it would make no sense economically for an existing supplier to build a production facility with only one line. In view of (i) the wide range of SKUs that are required by customers in the EEA markets; and (ii) the time and cost it entails to change production of specific SKUs, plants need at least two to three production lines to have a competitive cost base. This is also a reality in the market and is

668Agreed non-confidential minutes of a call with a customer, 12.05.2015, ID 468.

669Questionnaire to Customers (Q2), question 46.1, ID 2664.

670Agreed non-confidential minutes of a meeting with a competitor, 13.05.2015, ID 1065.

671Questionnaire to Customers (Q1), question 100.1, ID 1609.

672Agreed non-confidential minutes of a meeting with a competitor, 13.05.2015, ID 1065.

673A competitor's non-confidential reply to RFI of 18.08.2015, ID 2537.

674Third party comments to the SO, ID 3451.

(570) In the fourth place, a majority of customers that responded to the market investigation indicated that it is important for a supplier to have more than one plant in Europe so as to able to produce a wide range of can sizes and ensure security of supply. This was also confirmed by a competitor ('[l]arge multi-national customers prefer to deal with large suppliers who have a presence in a number of countries / geographic regions') and by […] )

(571) Fourth, the existence of high barriers to entry is confirmed by the fact that few new plants supplying the EEA have been built in the past three years and no major new entry is expected within the next 3-5 years, as also set out in recital (562).

(572) Fifth, incentives for current can producers to build new plants are generally low:

'Generally Ball, Rexam and Crown […] have little interest in expanding capacities, because a high utilization rate is ideal to operate a can plant economically. Limiting capacity allows for higher margins.'

'Although Ball and Rexam have sufficient production capacity, they are inflexible and focus on maintaining their margins. Ball and Rexam typically try to limit capacity.'

'[Crown] has sufficient financial resources, [but] it will probably not expand, because they like the status quo with limited capacity and attractive margins.'

'The example of Can-Pack, which started 25 years ago and has a market share of an estimated 13% today, shows that market entry is difficult. Can-Pack is the only player who succeeded to enter into the market on a long term basis in the past 25 years.'

(573) […] As a competitor notes, 'Rexam […] heavily investing in the production of slim cans in Austria and Switzerland […] may lead to the closing [of] older facilities in other regions due to environment constraints and older equipment.' 'Also […] could deliver volumes now originating from the Berlin plant, which is due to close.'

678Questionnaire to Customers (Q2), question 3, ID 2664.

679Questionnaire to Customers (Q2), question 19, ID 2664; Questionnaire to Customers (Q4), question 10, ID 2786.

680A competitor's non-confidential reply to RFI of 15.06.2015, ID 1129.

681See recital (427).

682Questionnaire to Customers (Q1), question 87, ID 1609. Questionnaire to Customers (Q2), question 46, ID 2664. See also Agreed non-confidential minutes of a call with a customer, 30.04.2015, ID 182: 'adding new capacity to the market is not straightforward and requires significant investment and time. Apart from the four main players already present in the market [customer] does not see any other can supplier who could enter the market.'

683Agreed non-confidential minutes of a call with a customer, 22.05.2015, ID 331.

684Agreed non-confidential minutes of a call with a customer, 20.08.2015, ID 2496.

685Agreed non-confidential minutes of a call with a customer, 22.05.2015, ID 331.

686Agreed non-confidential minutes of a call with a customer, 19.05.2015, ID 1212.

687Agreed non-confidential minutes of a meeting with a competitor, 05.05.2015, ID 261.

688Agreed non-confidential minutes of a call with a customer, 04.05.2015, ID 334.

(574) Rexam itself acknowledges in a public statement that 'the important thing is that the industry is very disciplined about keeping supply and demand in balance in the markets in which we operate.'[…]

(575) As for Can-Pack, apart from expansion in Sweden and Finland, it cannot be regarded as an actual or potential new entrant across the EEA. Can-Pack started supplying beverage cans in the EEA in 1994 (that is, more than 20 years ago), following which it expanded its footprint over a 20 year-period to five plants. Can-Pack has opened three plants in the EEA outside Poland and is otherwise focusing on expansion outside the EEA. Even Can-Pack describes '[i]ts market entry in the UK [as] one of few positive examples of entries into an existing market.'

(576) There have also been several examples of failed attempts to enter, including Can-Pack's failed attempt in 2014 to enter Spain.

(577) Sixth, it is doubtful whether the mere threat of entry constrains prices, as the Notifying Party claims. This is for a number of reasons. In the first place, the Notifying Party merely presents one example: Can-Pack's attempt to enter Spain. In the second place, it is unclear to what extent prices have been 'constrained' at a general level and not only for some large customers. In the third place, apart from Crown and Can-Pack, no other player can credibly envisage entry or expansion plans in the EEA. In the fourth place, it is also questionable whether any possible price reductions as a result of Can-Pack's failed plans in Spain would be maintained in the long term. As a large customer puts it, 'it is not clear whether, after this failed attempt, [confidential] would be able to come back to Spain in the future, leaving the incumbent players the opportunity to again raise prices at the next round of negotiations (in 3-4 years' time).','alignment':'left','indentation':1,'prefix':'','bold':false,'italic':false},{

(578) Seventh, the new plant recently built by Helvetia highlights both the difficulties in entering the beverage can market in a manner that would enable a new entrant to exercise any meaningful competitive constraint on existing players and that the high entry barriers go beyond the mere building of the plant. This is for the reasons set out in recitals (579)-(584).

(579) In the first place, Helvetia is only a small player with an initial capacity of 1.2 billion cans, that is to say, 3 % of total EEA demand. Customers seriously question whether 'Helvetia will be able to make any meaningful change in the current competitive landscape.' Essentially, Helvetia will not be an option for most customers: '[…] the capacity will be only 1,2 Billion cans. No real competition […] to Rexam and Ball.' According to a large customer with experience in the beverage industry, 'it is likely that in three years or less, the plant will either shut down or be bought by one of the major players. According to [this customer], the chances of a commercial success are very small.'

689Rexam. Edited Transcript, REX L – Rexam PLC Investor. https://www.rexam.com/files/presentations/investorseminar11investor_seminar_nov11_transcript.pdf. Last accessed on 28.09.2015, ID 3000.

690Questionnaire to Customers (Q1), questions 86 and 87, ID 1609.

691Agreed non-confidential minutes of a meeting with a competitor, 05.05.2015, ID 261.

692'[Competitor] cites a few examples of failed entry: in 1997 Skoda with the Ejpovice plant (bankrupt after ca. 18 months since start, bought by Ball, later taken over by Rexam) and Rusal (Russian aluminium producer), the facility of which was also eventually bought by Rexam.' Agreed non-confidential minutes of a meeting with a competitor, 05.05.2015, ID 261.

693Agreed non-confidential minutes of a meeting with a competitor, 05.05.2015, ID 261.

694Agreed non-confidential minutes of a call with a customer, 02.09.2015, ID 2888.

(580) In the second place, Helvetia will produce only 33 cl and 50 cl standard cans, and not a full range of can sizes and types. For instance it will not produce sleek or slim cans.

(581) In the third place, as explained by one large customer, it is 'also unclear how Helvetia will deal with the question of can ends. Indeed, a can ends production facility is efficient at around 4 billion units a year, while Helvetia will produce only around 1 billion cans a year. Whether setting up an efficient scale of can end operation would be efficient, is questionable.'

(582) In the fourth place, Helvetia's one plant operation cannot compete on equal footing with Ball and Rexam. As a customer pointed out 'Helvetia is not a credible entrant and will not be a credible competitor. The very little progress that Helvetia appears to have made since the announcement of its plans to build a new production facility strongly reinforces SABMiller’s doubts that this potential supplier will succeed in establishing a viable business. In any event, in view of the fact that the competitive strength of a supplier will largely depend on the size of its network, it is clear that even assuming that Helvetia’s plant becomes operational, it will not appreciably alter the competitive landscape or challenge any of the other can manufacturers. The merged entity would have more than ten times Helvetia's capacity and Helvetia will face significant obstacles because of its small size. In these circumstances it is not only unlikely but impossible that Helvetia would be an appreciable constraint for the merged entity in the near future.' And 'This new manufacturer will not increase competition. We don't expect a new reliable supplier. It seems to be a fake!'

(583) In the fifth place, certain customers are sceptical as to the long-term viability and competitiveness of Helvetia, mainly due to the fact that it is a one-plant operation.

'The experience with previous can mergers and remedies shows that a stand-alone plant like Helvetia will over time fail or be acquired by one of the larger players. And as long as it does not have any significant turnover this can even be done without any merger control review.'

'[I]t is likely that in three years or less, the plant will either shut down or be bought by one of the major players. According to [customer], the chances of a commercial success are very small.'

'[O]ne entrant with just one can plant has not enough scale to significantly influence the market.'

(584) In the sixth place, customers would consider buying from Helvetia only under strict conditions; in particular with regard to required can sizes at the right quality and qualification for the

698Agreed non-confidential minutes of a call with a customer, 18.08.2015, ID 2866.

699Questionnaire to Customers (Q2), question 61.1, ID 2664. This was confirmed by Helvetia. A competitor's non-confidential reply to RFI of 28.07.2015, ID 1573.

700Agreed non-confidential minutes of a meeting with a customer, 27.08.2015, ID 2982.

701Third party comments to the SO, ID 3424.

702Questionnaire to Customers (Q1), question 90.1, ID 1609.

703Third party comments to the SO, ID 3424.

704Agreed non-confidential minutes of a call with a customer, 18.08.2015, ID 2866.

705Agreed non-confidential minutes of a call with a customer, 22.05.2015, ID 331.

(585) Eighth, the large majority of customers do not consider Helvetia as a credible new entrant. This again highlights that the barriers to entry go far beyond the mere building of a plant. This is for the following reasons outlined in recitals (586)-(592).

(586) In the first place, a majority of customers that responded to the market investigation were unaware of the existence of such a 'new entrant'.

(587) In the second place, a majority of customers are neither in discussions to become a customer of Helvetia, nor are they aware who Helvetia' clients or sponsor(s) are or could be.

(588) In the third place, despite a press article indicating that Helvetia will have two 'key customers', a major beer and a major beverage producer, the Commission has been unable to identify a single purchaser of beverage cans that has, to date, already effectively contracted volumes with Helvetia:

– regarding the large beer producer mentioned in the article, it declared to the Commission that '[i]f Rexam does […], Helvetia or somebody else might be a commercial option for [the large beer producer]. This has to be evaluated though. In any event, [customer]'s volumes are now in general tied to contracts for the next 3-5 years'(emphasis added);

– regarding the large soft drink manufacturer mentioned in the article, it denied being the customer in question.

706Questionnaire to Customers (Q2), question 60.1, ID 2664.

707Agreed non-confidential minutes of a meeting with a customer, 27.08.2015, ID 2982.

708Agreed non-confidential minutes of a call with a customer, 26.08.2015, ID 2957.

(589) In the fourth place, none of the respondents to the market investigation stated that they are in any way (that is to say ownership or sponsorship) involved with Helvetia.

(590) In the fifth place, market participants have reservations with regard to purchasing cans from Helvetia because of the lack of public knowledge about the owners of Helvetia: 'This new manufacturer will not increase competition. We don't expect a new reliable supplier. It seems to be a fake!' 'To date, only little information exists about the "Project" Saarlouis. We know about a project under construction. Currently, we cannot [assess] whether this manufacturer can be a potential supplier for us.' 'According to rumours, the possible new entrant named Helvetia would be a self-manufacturing operation by a German brewery, based not far from Saarlouis, where Helvetia's plant is being built, although these rumours are not substantiated.' 'According to rumours that plant is financed by a group of (mostly financial) investors who over time bought breweries. Apparently the investors are experiencing problems with setting-up the plant, so that production will only be able to start by the end of 2015 or the beginning of 2016. Al[t]hough Can Pack does not have any knowledge of the owners' medium to long term strategic plans, market rumours suggest that the plant will likely be self-supplying those breweries owned by the investors and not selling on the market.'

(591) The internal documents of the Parties also show that they are not sure who the real owner of Helvetia is. They have diverging opinions on who could be the owners. Ball was confident that the owner 'is […]'. On the other hand, Rexam believed that the owner 'is a […].'

(592) This confusion and secrecy about the owners of Helvetia is perceived by market participants as a valid reason against contracting with Helvetia:

'The fact that the identity of the final owners of Helvetia is kept secret is a problem for [customer]. Indeed, in the event that the owner is a competitor of [customer], [customer] runs the risks of committing a certain volume and then of not being adequately supplied by Helvetia.'

See Questionnaire to Customers (Q2), question 56, ID 2664 for sponsorship. See Questionnaire to Customers (Q2), questions 62 and 63, ID 2664 for contracting. The three large customers that have contacted Helvetia or have been contacted by Helvetia have respectively declared on the subject: '[contracting with Helvetia] has to be evaluated though. In any event, [customer]'s volumes are now in general tied to contracts for the next 3-5 years, and in Poland next 3-7 years.' Agreed non-confidential minutes of a call with a customer, 25.08.2015, ID 2724. As late as 22 May 2015, 'no one ha[d] reached out to [customer] about the possibility to supply cans from there.' Agreed non-confidential minutes of a call with a customer, 22.05.2015, ID 226. 'It is still very unclear whether Helvetia will be able to produce cans of sufficient quality at a reasonable price'. Agreed non-confidential minutes of a meeting with a customer, 27.08.2015, ID 2982.

'It is unclear who Helvetia's shareholder is, but most likely this shareholder will get most of the can volumes. Therefore [customer] does not really consider Helvetia as a competitor for its volumes.'

'For [large customer] such obscurity is not acceptable, as trust and understanding the strategy of the supplier is crucial. [large customer]'s products are consumer goods and [large customer] must be able to tell its customers and consumers from whom their materials are sourced from.'

(593) Ninth, the existence of high barriers to entry is confirmed by the fact that non-EEA beverage can manufacturers do not envisage starting to supply beverage cans in the EEA within the next 3-5 years:

– Nampak, a beverage can manufacturer from Africa, has acknowledged that it has limited knowledge of the European can market and 'no plan to enter any EEA countries within the next 24 months.'

– GZI Industries, another African beverage can manufacturer 'also has no plans to enter the EEA beverage can market.'

(594) There seem to be a number of reasons why non-EEA beverage can manufacturers do not envisage starting to supply beverage cans in the EEA within the next 3-5 years.

(595) In the first place, emerging markets are considered by market participants as having a higher growth potential: 'the European market is rather stable and therefore probably less interesting to invest in that in emerging countries where demand for beverage cans grows much faster.'

(596) In the second place, certain technical agreements limit the ability of non-EEA beverage can manufacturers to use the technology they are currently using in territories outside the EEA within the EEA: 'Besides, some players have technical license agreements with one of the three can players. These agreements prevent them from competing with each other. Such an agreement exists between [Nampak] and Crown or between […] and […].'

(597) In the third place, the technical capability of non-EEA players to meet European standards without technical support from the incumbents was also questioned: '[customer] does not consider that either Nampa[c]k or other can suppliers from outside of EU (for instance Australian players) could enter the EU. There is no more space for a new supplier in the European market. Also Nampa[]k's technical capabilities are not necessarily up to the required standard.' '[A] new player needs technical support and it is unlikely that the global players

727Agreed non-confidential minutes of a call with a customer, 27.08.2015, ID 2716.

728Agreed non-confidential minutes of a call with a customer, 18.08.2015, ID 2866.

729Questionnaire to Customers (Q2), questions 46 and 47, ID 2664. See also agreed non-confidential minutes of a call with a customer, 30.04.2015, ID 182: 'adding new capacity to the market is not straightforward and requires significant investment and time. Apart from the four main players already present in the market [customer] does not see any other can supplier who could enter the market.'

730Email from Nampak, ID 634.

731Email from GZI, ID 593.

732Questionnaire to Customers (Q1), question 88, ID 1609; 'it seems to [customer] that since Europe has mature can markets, investment/entry is more likely to happen in Asia, Africa, Latin America.', Agreed non-confidential minutes of a call with a customer, 04.05.2015, ID 157.

733Agreed non-confidential minutes of a call with a customer, 12.05.2015, ID 468.

734Agreed non-confidential minutes of a call with a customer, 28.04.2015, ID 164.

(Crown, Rexam & Ball) would conclude a technical license agreements with a new player in Europe, as this player would compete with them.

(598) Tenth, there seems to be various mechanisms by which the Parties would have the ability and incentive to make the expansion of smaller firms and potential competitors more difficult. For example, given that the merged entity would be the only one in the market able to supply large volume orders to multiple major customers, it could threaten to increase the price of volumes or stop delivering the volumes the customer cannot get from another supplier if the customer were to consider committing volumes to support a new entrant. In a capacity-constrained industry where spare capacity is limited and old capacity is closed down when new capacity is built, this type of behaviour may well discourage customers from committing volumes to support entry or expansion.

9.1.16. Insufficient countervailing buyer power

9.1.16.1. The Notifying Party's views

(599) According to the Notifying Party, there are a number of powerful and sophisticated customers, who can, and already do, exert countervailing buyer power.

(600) First, those customers regularly use tenders and could switch volumes not only between competing beverage can manufacturers but also for other forms of packaging such as glass bottles, PET bottles and cartons. Losing a proportion of volumes from a large customer to a competitor can have a serious impact on plant economics since, for example, the top two customers at each of Ball's plants which make sales into the EEA on average account for over […] % of the plant's annual volume sales. For Rexam, the equivalent figure is over […] %.

(601) Second, customers can choose to self-supply a proportion of their packaging requirements. The Notifying Party has provided one example of self-supply in the EEA: the Polish brewer Van Pur produces its own cans from its affiliated BagPak plant in Stalowa Wola in Poland. The Notifying Party also points to the fact that as some of the Parties' customers have in-house production for other types of packaging, they could also do the same for beverage cans.

(602) Third, customers can sponsor entry and expansion. The Notifying Party has provided a number of examples which allegedly show sponsoring entry by beverage producers. These include: (i) Carlsberg sponsoring Can-Pack to build plants in the UK and Finland; (ii) Red Bull sponsoring Rexam to build a plant at […] in […] in […] and the addition of new line at […] in 2013; and (iii) Helvetia's entry at Saarlouis, Germany.

(603) Moreover, even if only large customers were able to exert some of these threats (namely sponsored entry and self-supply), the Notifying Party asserts that other customers also benefit indirectly since any new entry or expansion increases the available capacity in the market for all customers.

(604) The Notifying Party also submits that both Ball and Rexam are heavily dependent on orders from large customers and therefore customers exert buyer power by threating to switch. For

735Questionnaire to Customers (Q1), question 87.1, ID 1609.

736Form CO, paragraph 6.33 and Reply to the SO, paragraph 5.58.

737Form CO, paragraph 6.48.

738Form CO, paragraph 6.48.

example, the top two customer groups at each of Ball’s EEA plants on average account for over […] % of the plant’s annual volume sales. For Rexam the equivalent figure is over […] %.

(605) The Notifying Party further argues that the beneficial commercial terms that customers negotiate with the Parties are indicative of buyer power. For example as regards […].

(606) As for small customers, although in general they have smaller volume requirements and shorter contract terms, the Notifying Party submits that they are often able to buy opportunistically where suppliers need to fill capacity at a specific plant in their region. This flexibility of small customers can be used to impose a degree of buyer power on their suppliers.

9.1.16.2. The Commission's assessment

(607) The Commission has reached the conclusion that even large customers are currently unable to exert sufficient countervailing buyer power and that this would not change post-Transaction.

(608) Moreover, even if certain large customers were able to exert a degree of buyer power, this is not the case for the many smaller customers that represent 22 % of demand. As a customer put it, 'the can manufacturers deal with many smaller customers who have very limited countervailing power towards the can manufacturers.'

(609) There are a number of reasons why large customers are currently unable to exert countervailing buyer power.

(610) First, the scale and geographic footprint of the Parties gives them a degree of market power over large customers. As one internal Ball document explains: '[…].'

740Reply to the SO, paragraph 5.58.

741Reply to the SO, paragraph 5.61.

742Reply to the SO, paragraph 5.68.

(611) Second, tight capacity makes switching or the threat of switching difficult or even impossible. As a large customer points out 'Capacity is tight in the industry, especially during the high season, and therefore suppliers exercise a high power over their customers.' 'In general, it can be said that customers compete with each other for large capacities from the suppliers.' 'In a way, customers have to compete for the capacity of the can manufacturers.' A smaller customer also underlines that '[c]urrently [customer] must "fight" to obtain its can supplies due to the strong demand in the market. The main suppliers do not have available capacity.'

740Reply to the SO, paragraph 5.58.

741Reply to the SO, paragraph 5.61.

742Reply to the SO, paragraph 5.68.

(612) Third, while many large customers would like to source can bodies and can ends separately in the EEA (like they do in other parts of the world), the beverage can manufacturers do not allow this: 'Already today both Ball and Rexam have significant market power. This is shown, among others, by the fact that they both are able to bundle can ends and can bodies.' 'Can ends travel much more easily than can bodies and are all based on the same standard. Can ends from one producer are compatible with can bodies with another producer. Can ends are essentially a commodity product. For this reason, [large customer] constantly asks can manufacturers to quote separately for can bodies and can ends, but the can manufacturers have always refused to do so. The invoice price billed by the manufacturers for can ends is fictitious; they are intentionally not transparent on this price in order to increase their bargaining power. If beverage manufacturers were able to tender for can ends and can bodies separately they would be able to receive quotations from other can end manufacturers (given that can ends can travel further), and get lower prices overall.' Another large customer explains that '[c]an suppliers refuse to treat can ends separately from can bodies in Europe. For instance [one can manufacturer] has can ends production in France but did not quote for all volumes in France, because can bodies are sold only together with ends. [Large customer] tenders can ends and bodies separately, but suppliers –– refuse to negotiate bodies and ends separately.'

753Agreed non-confidential minutes of a call with a customer, 04.05.2015, ID 334.