Commission, June 11, 2015, No M.7484

Dear Sir/Madam,

Subject: Case M.7484 - Plastipak/ APPE

Commission decision pursuant to Article 6(1)(b) of Council Regulation No 139/20041

(1) On 4 May 2015, the European Commission received notification of a proposed concentration pursuant to Article 4 of the Merger Regulation by which undertaking Plastipak Holdings, Inc. ("Plastipak", USA), ultimately controlled by the Goldman Sachs Group, Inc. (USA) and Mr. William C. Young (USA), acquires within the meaning of Article 3(1)(b) of the Merger Regulation control of APPE by way of purchase of shares and assets.2  Plastipak and APPE are collectively referred to as "Parties".

1.  THE PARTIES

(2)   Plastipak is a manufacturer of rigid plastic packaging containers and PET preforms active throughout the entire value chain: it recycles plastic for the production of raw material (recycled PET resin or "rPET") and produces virgin polyethylene terephthalate ("PET") resin, manufactures PET preforms and plastic bottles. It is headquartered in the US and has production facilities in Brazil, US, Czech Republic, Luxembourg, France, Romania and Italy.

(3)  APPE manufactures PET packaging products, principally in the form of PET preforms and PET bottles. APPE also recycles PET material for the production of recycled PET. APPE has production facilities in the United Kingdom, Belgium, Germany, France, Spain, Turkey, Poland and Morocco. APPE is currently controlled by La Seda de Barcelona ("LSB", Spain), which has been in insolvency proceedings since July 2013.

2. THE OPERATION AND THE CONCENTRATION

(4)  Following APPE's parent company's financial difficulties and ensuing insolvency proceedings, APPE was subject to an auction process organized by an insolvency administrator appointed by the Spanish court. On 5 November 2014, Plastipak was announced as the winning bidder.

(5)  Under a Master Share and Asset Purchase Agreement, Plastipak will acquire from LSB the business units of APPE Iberia, APPE France, APPE UK, APPE Deutschland and APPE Benelux together with the related companies APPE Polska and APPE Nordic, a controlling interest in APPE-CCE and, subject to the confirmation of certain third party financial waivers and merger clearance, APPE Morocco and APPE Turkey. Plastipak will by the entirety of APPE related shares and assets currently held by LSB.

(6)  As a result, Plastipak will acquire full ownership and sole control over APPE.

(7) The proposed Transaction constitutes therefore a concentration within the meaning of Article 3(1)(b).

3. EU DIMENSION

(8) The undertakings concerned have a combined aggregate world-wide turnover of more than EUR 5 000 million3 (Plastipak: EUR […] million; APPE: EUR […] million). Each of them has an EU-wide turnover in excess of EUR 250 million (Plastipak: EUR […] million; APPE: EUR […] million), but they do not achieve more than two-thirds of their aggregate EU-wide turnover within one and the same Member State.4

(9) The notified operation therefore has an EU dimension within the meaning  of Article 1(2) of the Merger Regulation.

4.  ASSESSMENT

(10)  There are three distinct stages of the production process in relation to the manufacturing of PET packaging:

i)  manufacturing of PET resin, a malleable, gas permeable polyester resin which can be produced by combining ethylene glycol and terephthalic acid, or by recycling PET products;

ii)  manufacturing of PET preforms from PET resin: PET preforms are tube- shaped containers made from PET resins using a preform injection moulding machine; and

iii)  manufacturing ("blowing") of empty PET preforms into containers/bottles: this is the final stage of the PET packaging value chain which consists in heating the preforms, stretching and shaping them by using specific moulds in stretch blow moulding machines. The resulting transparent containers are used for drinks and liquid foods, as well as for non-food purposes, such as cosmetics, laundry and homecare products.5

(11) Plastipak and APPE are both active at all three stages of the production process.  The affected markets arise only in relation to the market for PET preforms.

4.1. Product market definitions

4.1.1. PET resin

(12)  In the past, the Commission considered that polyester resins (such as PET) could  be categorised into (i) packaging resin (bottle grade)6; (ii) polyester textile chips (fibre grade); and (iii) film polymer (film grade), but has consistently left open the exact product market definition for polyester resins.7 On a national level, the UK Competition authority held in Amcor/La Seda that PET resin forms a distinct product market which does not ought to be segmented further.

(13)  PET resin can be either "virgin", i.e. directly produced from raw material, or recycled produced from the collection of used PET bottles. The Notifying  Party argues that virgin PET and rPET are to a large extent substitutable in the manufacturing process of preforms, while the extent to which one or the other is used is driven largely by customer preferences such as bottle design and cost. For instance, rPET tends to be less clear than virgin PET, but can be used for promotion purposes as a sign of commitment to sustainable development. As concerns the price, according to the Notifying Party, the ratio between the two tends to vary depending on the cost of raw material for virgin PET resin. Customers tend to arbitrage between the two depending on which price is more favourable within the limits of the amount of rPET that can be used without altering the product quality and appearance.

(14)  The market investigation provided indications that, although rPET can replace to some extent virgin PET in a number of applications, technical and regulatory  measures limit8  the use of recycled PET in preforms.9  Customers generally confirmed that there is a limit to which rPET can be used and that typically preforms for food applications cannot be produced with 100% of rPET10  as rPET tends to be less clear.

(15)  In any event, the exact product market definition of PET resin can be left open since, under any plausible product market definition, no serious doubts arise as  regards the compatibility of the transaction with the internal market in relation to PET resin.

4.1.2.  PET preforms

(16)  There are generally two broad types of PET preforms. Most of the PET preforms are so called standard preforms simply produced from PET resin. Standard preforms can be used for the packaging of most products. However, some products may require additional protection from external factors such as for instance from oxygen ingress or light in order to extend their shelf life and preserve their quality and appearance. To this end, there needs to be an additional protection added to the preform either in a form of an additive or an extra layer. Preforms with such a protection are referred to  as barrier enhanced preforms.

(17)  There are broadly three types of enhancement to a standard preform: i) a barrier which will prevent carbon dioxide from leaking out of the package and degrading the carbonation of beers, sparkling water and carbonated soft drinks (CO2  egress barrier),ii) a barrier preventing oxygen from entering into the packaging and degrading its contents, mainly used for juices and sauces (O2 or oxygen ingress barrier), and iii) a barrier protecting from exposure to light of sensitive beverages, such as dairy products or soya drinks (UV or light barrier protection).

(18)  In previous decisions, the Commission distinguished barrier enhanced preforms used for oxygen and light-sensitive products from standard preforms used for products not requiring this kind of protection.11 Standard and barrier enhanced PET preforms were seen as not substitutable both from a demand side perspective (since a standard preform cannot be used to fill an oxygen sensitive product) and from a supply side perspective (standard preforms were seen as commodities whereas barrier enhanced preforms needed specific technology not available to all suppliers).12

(19)  The Notifying Party submits that since the last case the market has evolved and that while from demand-side the distinction between standard and barrier enhanced preforms is still relevant, from the supply side there is no reason to make such distinction. In this context, the Notifying Party submits that there are three ways to achieve barrier properties. First, a barrier additive can be blended with the resin at the preform manufacturing stage to produce monolayer barrier preforms. Second, the preform manufacturer may insert a barrier layer (usually nylon) between the PET layers and produce a multilayer barrier preform; this requires special machinery capable of producing multilayer preforms with significantly more investment. The third option involves coating a standard preform with a barrier material during the bottle blowing stage. The Notifying Party argues that whilst at the time barriers were achieved using multilayer production technology which was not available to everyone, nowadays any producer of standard preforms has the ability to produce monolayer barrier preforms as this only requires sourcing an appropriate  resin/additive technology to increase barrier performance which is widely available in the market.

(20)  The Commission investigated possible segmentations of PET preforms into standard and barrier enhanced in the market investigation. Generally, many customers indicated they use standard performs even for oxygen ingress, CO2 egress or light sensitive products.13 While there is a portion of customers who may prefer using dedicated preforms, the market investigation indeed confirmed that the technology to achieve the desired properties is available to all suppliers. Indeed, a significant majority of respondents indicate that most PET preform suppliers are able to produce all types of preforms with any type of barrier, mentioning inter alia ALPLA, Resilux, PET Verpackungen, APPE, Plastipak, Retal, Logoplaste, Esterform and Gardaplast as possible suppliers of all types of preforms.14 Customers also confirm that barrier enhancement can be achieved at the bottle blowing stage of PET bottle production by coating a standard PET preform with appropriate barrier layer, albeit it is perceived as rather costly.15

(21)  As concerns the supply-side, the market investigation indicated that there are certain barrier technologies that have been developed or acquired by preform manufacturers like Plastipak or APPE, and are not necessarily easily accessible to competitors.16 This concerns in particular oxygen ingress barriers where the Parties seem to own two competing technologies, but also CO2 egress barriers. APPE owns a barrier material (MonoBlox) which is patent-protected, while Plastipak's technology used in Europe (Monoxbar) is not patent-protected anymore, but is not sold by Plastipak to any third party.

(22)  Nevertheless, the market investigation also pointed to the existence of competing technologies: Amosorb which appears to be widely available and having comparable characteristics and performance and other oxygen protection technologies available on the market, namely Valspar Valor or Invista Polyshield.17

(23)  Furthermore, the market investigation confirmed that as regards carbonated drinks, applying a barrier on a preform to prevent CO2 egress is not necessary and the same objective may be achieved simply by producing a preform with thicker walls (and an appropriate neck design)18 – this is the reason why Plastipak does not even distinguish internally CO2  preforms from the standard ones.

(24)  It appears that if there are any types of PET preforms that are provided only by a limited number of producers, these would be multilayer preforms with oxygen or CO2 egress barrier, as indicated by both customers and competitors.19 However, there is no overlap in relation to multilayer preforms, as only APPE is active in this segment.

(25)  Based on the above, there are strong indications confirming supply-side substitutability between standard and barrier enhanced PET preforms. In any event the exact product market definition of PET preforms can be left open since, under any plausible product market definition, no serious doubts arise as regards the compatibility of the transaction with the internal market in relation to PET preforms.

4.1.3.  PET bottles

(26) In the past cases dealing with the production machinery used to produce PET packaging, the Commission identified blowing of empty bottles as one of the main stages of a typical PET packaging line, necessitating specific equipment, but did not analyse this market in detail as it was not the focus of that case.20

(27)  The Notifying Party notes that PET is just one of a number of different packaging materials used for food, drink or cosmetics, other common materials being high density polyethylene (HDPE), aluminium cans or laminated carton. The Notifying Party accepts however that PET bottles form a distinct product market from other packaging techniques.

(28) The market investigation confirmed that the users of PET bottles also use other packaging applications to sell their products. However, although majority of  customers indicated they use PET for the same applications as the alternative packaging, such as glass, metal cans, HDPE, laminated carton or aluminium, none of these materials is considered as fully substitutable with PET containers in terms of cost and performance. PET packaging is significantly cheaper, therefore even 10% increase in price of PET packaging would not incite the customers to switch to another form of packaging. More importantly, switching would require serious modifications in the whole filling chain in order to allow for the usage of other packaging materials. Furthermore, some customers indicate that only PET allowed them to achieve  a certain shape of bottle and other packaging types were considered as suitable for different target customers.21

(29) In any event, the exact product market definition of PET bottles can be left open since, under any plausible product market definition, no serious doubts arise as  regards the compatibility of the transaction with the internal market in relation to PET bottles.

4.2. Geographic market definitions

4.2.1. PET resin

(30)  In the past the Commission left open whether the relevant product market for PET resin is EEA-wide or wider.22 The UK Office of Fair Trading considered the EU to be the relevant geographic market as regards the PET resin, although  in its decision it also discusses factors pointing to a wider market definition such as imports from South Korea, Taiwan and China representing at the time around 10-15% of PET resin in Europe.23

(31)  The Notifying Party submits that the relevant geographic market for PET resin is at least EU-wide. As concerns virgin PET resin the Notifying Party refers to significant imports of virgin PET into the EEA from South Korea, the Middle East and other  parts of Asia which were confirmed during the market investigation. As concerns rPET resin, the situation is less clear in terms of the level of imports into the EEA, however, a significant proportion of customers do purchase rPET resin outside the EEA.

(32)  In any event, the exact geographic market definition regarding the PET resin can be left open since under any plausible geographic market definition, no serious doubts arise as regards the compatibility of the transaction with the internal market in relation to PET resin.

4.2.2.     PET preforms

(33)  In its decision in Case COMP/M.2416 Tetra Laval/Sidel, the Commission concluded on the basis of its investigation that the market for PET preforms is EEA- wide. According to this decision, all suppliers active throughout the EEA were  capable of providing and provided PET preforms on a cross-border basis.24

(34)  The Notifying Party submits that this market definition is still valid, arguing that there is significant cross-border trade and that both Parties sell to customers situated relatively far away from their respective manufacturing sites.

(35)  The market investigation in this case provided indications that the market can be narrower, comprising circles of 300 to 400 km25 around the production plants for standard performs, while barrier enhanced preforms tend to be transported over longer distance.26  Transport costs can represent, on average, between 0% to 10% of the  final price, depending on the distance between the preforms production facility and the customers' bottling plant. This is why suppliers tend to sell their preforms as close to their facilities as possible. While transport costs do not vary between standard and enhanced preforms, enhanced preforms may economically travel over longer distances, due to their higher value.

(36)  Some respondents clarified that in fact the transport distance as such is of little relevance, and that what matters is the ultimate price. For example, in this context, one customer explained that "there would be no geographic limit if the pricing was right and it makes sense to do so from a business perspective".27 Indeed in tenders or  request for quotes preform suppliers are requested to indicate prices for preforms at  the customer’s premises. Preform producers note that there can be lots  of arrangements in this respect – prices depend on volumes, distances and the purchasing power of customers. Finally, a significant portion of sales is done on spot, outside contractual arrangements, when customers pick up their preforms at the producers plants.28

(37)  In addition, given that besides the large EEA wide suppliers the substantial portion of the market is served by smaller local suppliers, the supply demand dynamics tend  to be different across the countries/country clusters in the EEA. This seems to be also confirmed by varying price levels. While large customers with filling plants in two or more Member States indicate that they negotiate standard PET preform prices at EEA level, the majority of customers and competitors submitted that price differences for standard preforms exist between Member States.

(38)  As regards barrier enhanced preforms, prices tend to be more homogenous across the EEA according to the market investigation.

(39)  In view of the fact that the geographic market may be narrower than EEA wide and in the absence of data at the level of regions/parts of Europe, the Notifying Party submitted market share data on a national basis, emphasising however that it did not consider this to be the adequate level for the assessment of the preform market.

(40)  It should be noted that in terms of plant location the Parties’ preform production do not overlap in any country: in the EEA Plastipak has plants only in Czech Republic, Italy, Luxembourg and Romania, while APPE's plants are located in Belgium, Germany, France, Poland, Spain, and the United Kingdom. Thus, by definition, for the Parties’ activities to overlap they must be selling their preforms on a cross border basis, which is actually the case, and which indicates that the market is rather broader than national. This is particularly the case as regards the Benelux countries for which considering national markets would not be meaningful.

(41)  The Notifying Party contends in particular that considering Benelux or each of the Benelux countries as a relevant geographic market is too narrow since these countries are highly integrated among one another and to a large extent also with their neighbours. For example as regards the Plastipak’s plant in Luxembourg, [40-50]% of the sales to  top  ten customers  (in terms  of value) are  shipped  to  companies outside Benelux, namely France, Germany and even Greece, and only the remaining [40-50]% is sold to Belgium and the Netherlands. The Notifying Party also notes that the average margins earned with preforms in the Benelux overall are not higher than in other Member States, where the Parties have lower markets shares.

(42)  The results of the market investigation support the Notifying Party’s submission that as regards the Benelux area the geographic market should not be considered at the country level but it should at least encompass all the Benelux countries. According to the replies, in the Benelux the average transport distance for the supply of preforms varies between 300 and 400 km. Moreover, a significant number of customers located in Germany purchase preforms from Luxembourg, and for some of them the average transport distance is 500-600 km, partly even more.

(43)  Consequently, the Commission will analyse the competitive impact of that transaction at the EEA and at the national or cluster of countries level (in particular for Benelux) at the national levels.

(44)  In any event, the exact geographic market definition of PET preforms can be left open since under any plausible geographic market definition, no serious doubts arise  as regards the compatibility of the transaction with the internal market in relation to PET preforms.

4.2.3.  PET bottles

(45)  In a previous decision, the Commission considered that the scope of the relevant geographic market for the supply of PET bottles was narrower than the EEA, but left the exact geographic market definition open.

(46)  The Notifying Party argues that the relevant geographic market for the supply of PET bottles should be defined at most as national in scope given the typical maximum distance of transport which is 100 kilometres.

(47)  The market investigation confirmed that PET bottles are typically transported over much shorter distances than PET preforms, given that they are very voluminous and cumbersome to transport, and therefore transport cost are higher.

(48)  In any event, the exact geographic market definition in relation to PET bottles can be left open since under any plausible geographic market definition (be it narrower than national, national or even EEA-wide), no serious doubts arise as regards the compatibility of the transaction with the internal market in relation to PET bottles.

4.3. Competitive assessment

4.3.1.  PET resin

(49)  The Parties’ activities would overlap only if virgin PET and rPET were considered to be part of the same relevant product market. This is because APPE only produces and sells rPET, whereas Plastipak only sells virgin PET on the merchant market, all its rPET production being used internally.

(50)  On a combined market including both rPET and virgin PET, the Parties' combined EEA market share is below [0-5]%.29 Furthermore, APPE’s rPET production is performed in a joint venture with Coca-Cola at its Beaune plant in France. [60-70]% of that rPET production is used internally by APPE, while the remaining [30-40]% is largely dedicated to one customer – Coca-Cola and Coca-Cola’s tolled converters.

(51)  The market investigation did not reveal any concerns in relation to PET resin market.

(52)  Therefore, it is concluded for the purpose of this decision that no serious doubts as regards the compatibility of the transaction within the internal market in relation to PET resin market irrespective of the product and geographic market definition.

4.3.2.  PET preforms

4.3.2.1.      Background on the functioning of the market

Through-the-wall manufacturing and self-manufacturing

(53)  There are three basic scenarios of how PET preforms are manufactured and sold to the market. Most of the market is represented by sales to bottlers/drink producers on a contractual or spot basis. However, some preform manufacturers integrate part of their production into downstream operations whereby preforms are being blown into bottles by the preform producers themselves at, or nearby, customers premises (referred to as "vertical integration" or "through-the-wall"). Typically these operations are dedicated to specific customers and thus the preforms used in these operations are not available on the merchant PET preform market.30 The economics of such through-the-wall arrangements are different from normal contracts as the duration of these contracts is typically much longer and the investments associated with such arrangement are shared by the parties to that arrangement. While some of the PET preforms suppliers such as Logoplaste make through-the-wall operations their business model, only around [5-10]% of the Parties’ preform production is used for such through-the-wall arrangements, thus used captively for the production of bottles.

(54)  Similarly, a number of big producers of beverages, such as Nestlé, Coca-Cola, Danone and Crystaline have internalised part of their preform demand and integrated upstream into the preform production through "self-manufacturing".

(55)  Neither the preforms produced in vertical integration nor the preforms produced by self-manufacturers are available on the merchant market, since they are produced exclusively for specific customers and as such can be considered captive. However, the Notifying Party argues that vertically integrated preform production should be taken into account in the competitive assessment, in particular because some large preform suppliers have a commercial policy of gaining preform customers by offering them through-the-wall contracts. The Notifying Party argues that this phenomenon should be perceived as a feature of competition in the preform market rather than a factor which limits competition.

(56)  Consequently, to cover the different possible scenarios, all the EEA data regarding market shares is provided for (i) total production, (ii) total production excluding self- manufacturing, and (iii) total production excluding self-manufacturing and vertical integration (see table 1 with regard to all preforms, standard as well as barrier enhanced preforms). Since the Parties tend to be less vertically integrated than their competitors (on average [5-10]% as opposed to [10-20]% for the other preform suppliers, according to the Parties), the market excluding self-manufacturing and the vertically integrated production is the narrowest plausible market and the worst case scenario for analysing the competitive position of the Parties post transaction.

Customer switching

(57)  As concerns customer switching, generally speaking the market investigation provided indications that switching is easy, which is also evidenced by the Parties’ data. While there may be some obstacles which one may encounter when switching to an  unknown supplier related to the necessity of testing in order to approve the product and there may be customer inertia in particular as concerns customers who have bespoke shapes of preforms and bottles, a big majority of customers, in particular those who buy significant volumes of preforms actually multisource and regularly switch between suppliers according to the price conditions offered. They tend to have at least one major supplier and often smaller back-up suppliers among others to mitigate the risk of supply. When a customer has several suppliers approved, switching is significantly easier, as it does not require additional testing and there is  no delay related to the approval procedure while they hire additional suppliers.

Buyer power

(58)  Buyer power seems to be a feature of this market, in particular in view of  the existence of large customers. Indeed, Parties' largest customers such as […] (representing [10-20]% of Plastipak's sales in 2014) or […] (representing around [10- 20] % of APPE's sales in the last five years) have ability to exercise pressure on their suppliers which is reflected in their relative bargaining strength.

(59)  The responses to the market investigation confirmed that it is a common practice for large customers to "threaten" the preform suppliers with the possibility of moving to self-manufacturing, switching large volumes to other suppliers or investing in "through-the-wall" operations.

(60)  As concerns the ability of large customers to self-manufacture, the Notifying Party explained that an injection machine with an annual capacity of 150-300 million preforms cost approximately EUR 2-3 million. Since large customers’ needs vary  from 200 to 800 million preforms p.a. it may be cost effective for them to internalise the preform production. Replies to the market investigation support this submission: smaller customers explain that the volume of their needs would not justify investment required for equipment necessary to internalize preform production. Medium-sized  and large customers, however, indicate that either they are already partially active in the self-manufacturing of preforms or they are regularly considering the competitiveness of such a solution. Finally, competitors of the Parties also observe  this tendency, one of the major international preform producers explained in the context  of  the  market  investigation:  "We  are  permanently  under  the  threat   that medium to big size customers make the decision to change to self-manufacturing. For sure this puts pressure on prices, as no preform supplier can afford to lose such big contracts"31.

(61)  As regards the “through-the-wall” arrangements, the market investigation confirms the Notifying Party’ submission that certain preform suppliers have developed a business strategy based on strong links with customers and set up through-the-wall arrangements close to customers’ premises to which they deliver PET preforms. Customers confirmed that both the Parties as well as other preform producers could potentially be a partner for setting up "through-the-wall" facilities. In this respect neither of the Parties would have any particular advantage as compared to other preform suppliers.32

(62)  It therefore appears that for larger customers all the options are available including "through-the-wall" arrangements and self-manufacturing of PET preforms, since they have sufficient scale to justify the associated investment. Conversely, smaller customers, whose scale would not allow for such investment, are more dependent on merchant sales. However, they have a number of alternative suppliers as they can and do turn to smaller scale local suppliers who offer competitive prices.

(63)  The fact that customers have the ability to exercise pressure on suppliers is also evidenced in some Plastipak's internal documents. For example, a Plastipak’s sales manager explains the relative bargaining power of customers as follows: “PET preform and PET bottle is a market, where the power is in the hands of buyers; Potential customer could choose producer of preforms very quickly; Every producer  of preforms is quickly replaceable; Some of potential customers want to start own production of preforms.”33

Tolling arrangements

(64)  While typically a PET preform customer pays one price for the PET preform, some customers prefer to procure PET resin independently from the PET preform manufacturer to improve their cost structure. This is commonly referred to in the market as tolling. There are two types of tolling, hard tolling and soft tolling. In case  of hard tolling the customer sources PET resin directly from a resin manufacturer and delivers it to the PET preform manufacturer. In such situation the customer pays the preform supplier only for the conversion of the resin into a PET preform. Conversely, soft tolling relates to an arrangement between a customer and a PET preform manufacturer whereby the price for PET resin is negotiated separately from the price of the conversion of the resin into a preform. 

(65)  From a commercial point of view, APPE and Plastipak do not have the same practice as regards tolling arrangements. APPE at present accepts this kind of arrangements, while Plastipak does not. While the situation post-merger is unclear, the market investigation did not provide any indications that the removal of tolling option would be  problematic  for  customers.  This  is  mainly  because  most  tolling  arrangements concern large customers who have a degree of buyer power and are ready to switch to alternative solutions (other suppliers, self-manufacturing, etc.) should they not reach a satisfactory outcome in relation to a PET preform price.

4.3.2.2. Market structure

EEA market

(66)  Table 1 below presents the position of the Parties and their competitors for all types of PET preforms, i.e. standard preforms and any types of barrier enhanced preforms combined, at the EEA level.

Table 1 - Estimated EEA shares of all PET preforms (2014)

Source: Form CO.

(67)  According to the data submitted by the Notifying Party, the combined market shares  of the Parties at the EEA level reach [20-30]% (Plastipak [5-10]%, APPE [10-20]%) for total preforms (including self-manufacturing, vertically integrated sales and merchant sales), [20-30]% (Plastipak [5-10]%, APPE [10-20]%) for the production of preforms excluding self-manufacturing and [30-40]% (Plastipak [10-20]%, APPE [20- 30]%)  for  merchant  sales  only  (i.e.  excluding  self-manufactured  preforms      and preforms that are vertically integrated i.e. turned into bottles by the preform  producers, and sold as such to the end customers).

(68)  Should the market be further subdivided into standard and barrier enhanced preforms, and assuming that also in these two categories there exists the same share of self- manufacturing and vertical integration as in the overall market, the market shares would be as follows:

Table 2 - Estimated EEA shares of PET preforms, distinguishing standard preforms as well as barrier enhanced preforms (2014):

Source: Form CO, Commission's calculations.

(69)  Following the transaction the merged entity will become the number one PET preform supplier in Europe with Alpla being the second supplier with a market share of 20%. Other suppliers such as Resilux ([…]%) and Retal ([…]%) will continue to be active on the market. In addition, a numerous small local producers compete in the market. Generally, the market investigation indicated that customers do not seem to perceive Plastipak and APPE as particularly close competitors; other international players such as Alpla, Logoplaste, Resilux and Retal, seem to be perceived as similar in terms of offering, price and service. While often APPE or Plastipak are named among the most important preform suppliers in the EEA, neither of them is consistently considered as particularly strong compared to the other main market players. Alpla is more often indicated as the strongest supplier in Europe, while other big international suppliers such as Retal, Resilux, Esterform, Pet Verpackungen, are having similar competitive position to the Parties.

(70)  Similarly, the internal documents by the Parties show that they have not perceived each other as the closest  competitors. For example, APPE estimated  that as    regards dairy products their closest competitors offering PET packaging for dairy applications are: Alpla, Logoplaste, Resilux and SGT.36

(71)  The analysis of overlapping customers provided by the Parties also shows that the number of customers present in Plastipak's and APPE's respective customer portfolios is relatively limited and concerns a few large customers: […] is the only customer present in the top 10 customers of both Parties, only […] other customers ([…]) were customers of both Parties in the last 5 years.37

(72)  Similarly, tender data analysis provided by the Parties at the EEA-wide level shows that they meet in a relatively limited number of tenders (in the last 3 years, APPE bid in [10-20]% of the tenders where Plastipak was present, and Plastipak bid in [10-20]% of the tenders where APPE was present). Further analysis, taking into account the volumes tendered, indicate that very big customers tend to ask both of the Parties for offers, thus the Parties meet much more often in large tenders. This is actually confirmed by the submissions of big customers in the market investigation, who systematically organise tenders or request for quotes addressing all the big European preform suppliers, including Plastipak and APPE, but also Alpla, Retal or Resilux.

(73)  Finally, the analysis of the non-transaction internal documents submitted by the Parties (on business and commercial policy issues), demonstrates that neither Party considered the other to be its closest competitor in any of the product of geographic markets considered.

National analysis

(74)  As a result the following countries/clusters would give rise to affected market, namely Benelux, France, Germany and Spain.

Benelux (including Belgium, the Netherlands and Luxembourg)

Table 3 - Estimated market shares of PET preforms, for all types of preforms and standard performs (2014) in the Benelux countries

Source: Parties' submission.

(75)  As shown in Table 3 above, the Parties have a very strong presence in Benelux, where the combined market share is estimated at [40-50]% as regards all types of preforms and [60-70]% as regards standard preforms

(76)  These high market shares appear to be driven by the fact that Plastipak has one of its major plants in Luxembourg and APPE has a plant in Belgium. At the same time, the Commission notes that also Resilux has a preform plant in Belgium, Alpla – one plant in Belgium and one in the Netherlands and Logoplaste a plant in the Netherlands. Customers could therefore switch to other preform producers easily.

(77)  In general it appears that the Parties’ market shares in the Benelux countries overstate their actual importance in this area, since it is unlikely that the other players mentioned in the preceding paragraph, similarly important and equally present across the whole EEA (as shown by the results of the market investigation and analysed in the other parts of this decision), would not be able to sell preforms to Benelux customers. It shows in addition that many preform suppliers consider the Benelux area as attractive and they supply from this area not only the customers located nearby, but to a large extent customers located in the neighbouring countries, namely Germany and France.

(78)  The analysis of the Parties’ data reporting switching events, which indicate the percentage of contracts lost or won to competitors for each of the Parties’ production site between 2011 and 201438 confirms that other suppliers are equally present in the area. Considering APPE's Belgian site, no loss or win was identified as gained from or lost to Plastipak. Indeed, […]% of APPE's Belgian site losses benefited the local supplier Plasco, […]% Retal and […]% Resilux (the remaining 23% could not be attributed to specific suppliers). Similarly, […]% of its wins were taken from Resilux, […]% from Constar and […]% from Petainer.

(79)  Considering Plastipak's site in Luxembourg, a portion of its wins was taken from APPE (15%). Nevertheless, this site also won large volumes from other competitors,  in particular Alpla ([…]%) and Retal ([…]%), as well as Pet Verpackungen ([…]%) and Resilux ([…]%); the remaining volumes could not be attributed to specific suppliers. In addition, only a small portion of its losses benefited APPE (only […]%), while Retal won […]% of these losses, Alpla […]% and Putoksnis […]% ([…]% of these lost volumes were transferred to customers who opted for self-manufacturing and the remaining […]% could not be attributed to specific suppliers).

(80)  This indicates that customers in the Benelux area have much more options that the market shares may suggest.

(81)  Therefore, it is concluded for the purpose of this decision that no serious doubts as regards the compatibility of the transaction with internal market arise in relation to PET preforms in Benelux.

France, Spain, Germany

(82)  Table 4 provides market share data regarding national markets other than Benelux, where affected markets arise.

Table 4 - Estimated national market shares of PET preforms, for all types of preforms (2014), for countries where there is an overlap and the Parties' combined market share exceeds 20%:

Source: Parties' submission

(83)  As shown in the table above, the Parties' market shares in these countries are below or only slightly above [20-30]% while the market share increments are low, in particular in Spain and France. Moreover, the Parties' main competitors namely Alpla, Retal, Logoplaste and Resilux are present in all three countries or in neighbouring countries and thus will continue to exert competitive pressure on the merged entity post-Transaction.

(84)  It is therefore concluded for the purpose of this decision that no serious doubts as regards the compatibility of the transaction with the internal market arise in relation to PET preforms in France, Spain and Germany.

Conclusion on PET preforms

(85)  Therefore, it is concluded for the purpose of this decision that no serious doubts as regards the compatibility of the transaction with the internal market arise in relation to PET preforms market irrespective of the product and geographic definition.

4.3.3.  PET bottles

(86)  If the relevant geographic market for bottles were to be national or narrower in scope, the Parties' bottle manufacturing activities do not overlap, given that APPE produces bottles in the UK, Belgium, Germany and Spain, while Plastipak produces bottles in Italy, the Czech Republic and Romania.

(87)  Only if the relevant market for PET bottles were to be wider than national in scope, the Parties’ activities would overlap. It is however noted that in general the Parties are less present in the bottling activities than their competitors. As explained above, they use approximately [5-10]% of their preform production internally to supply downstream PET bottle production process within the framework of "through-the- wall" arrangements, while for example ALPLA (the largest PET preform producer) or Logoplaste have much larger part of their preform production vertically integrated  into the downstream production of bottles. Consequently, other preform producers are more important players in the production of bottles than the Parties.

(88)  Furthermore, the market investigation did not reveal any concerns in relation to PET bottles.

(89)  Therefore, it is concluded for the purpose of this decision that no serious doubts as regards the compatibility of the transaction with the internal market arise in relation to PET bottles market irrespective of the product and geographic definition.

4.3.4.  Vertical links

(90)  The Notifying Party also submits that there are two vertical relationships between the Parties' activities: i) between the manufacturing and supply of PET resin and the manufacturing and supply of PET preforms, ii) between the manufacturing and supply of PET preforms and the manufacturing and supply of PET bottles, both Parties being present at all these production stages.

4.3.4.1. PET resin and PET preforms

(91)  As regards the vertical relationship between PET resin and PET preforms, the upstream products concerned are more particularly virgin PET resin for Plastipak (as its production of rPET resin is used internally) and rPET resin for APPE, and the downstream market is the market for PET preform in which both Parties are active.

(92)  Since the Parties’ position as regard the supply of PET resin is insignificant the transaction does not pose risk for input foreclosure in this respect.

(93)  Customer foreclosure could potentially arise, should Plastipak or APPE be important customers of PET resin and stop buying it from other suppliers. Given that APPE has a virgin PET resin consumption of circa […]kt, and Plastipak's total PET resin capacity available for sale on the open market is only about […]kt, Plastipak will be able to supply at most less than […]% of APPE requirements. Therefore, post- merger APPE will still need to purchase the vast majority of its virgin PET resin requirements from third party suppliers. Conversely, Plastipak produces rPET in its Luxembourg plant and vertically integrates […] the rPET produced. APPE could only sell rPET to Plastipak from its joint venture with Coca-Cola. However, as […]% of  the production is used internally by APPE and the remainder is sold to Coca-Cola and Coca-Cola's tolled resin converters, Plastipak will have to continue to buy the rPET from other suppliers. Therefore, customer foreclosure is unlikely to arise as a result of the transaction.

4.3.4.2. PET preforms and PET bottles

(94)  As regards the vertical relationship between PET preforms and PET bottles (both Parties being present on both markets), both Parties sell [above 90]% of their respective PET production to third party customers and vertically integrate the remaining [0-10]% into their PET bottle productions.

(95)  Input foreclosure could arise, should Plastipak or APPE stop supplying their customers with PET preforms and have the ability to substantially restrict access to these inputs. However, in light of the Parties' combined market share on the market  for PET preforms, the merged entity would not have the ability to engage in input foreclosure behaviour.

(96)  As a result, no serious doubts arise due to vertical relationships between the Parties' activities as regards the compatibility of the transaction within the internal market.

5. CONCLUSION

(97)  For the above reasons, the European Commission has decided not to oppose the notified operation and to declare it compatible with the internal market and with the EEA Agreement. This decision is adopted in application of Article 6(1)(b) of the Merger Regulation.

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

2                     Publication in the Official Journal of the European Union No C 156, 12.05.2015, p. 8.

3                     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).

4                     The  turnover  figures  for  Plastipak  include  Goldman  Sachs'  world-wide  and  EU  wide turnovers

respectively.

5            Case No COMP/M.2843 – Amcor/Schmalbach-Lubeca, Commission decision of 28 June 2002.

6            Both Plastipak and APPE only produce bottle grade PET resin.

7            See Case No COMP/M.1538 - Dupont/Sabanci, Commission decision of 24 November 1999; Case COMP/M.2628 - Koch/Kosa, Commission decision of 12 November 2001; Case COMP/M.5760 - Lotte Group/Artenius UK Ltd, Commission decision of 19 March 2010, paragraph 9 and Case No COMP/M.6184 – Indorama/Sinterama/Trevira, Commission decision of 9 June 2011, paragraph 11.

8            In some EU Member States rPET use appears to be capped (to 25%, Spain and 50% (Italy). Also  there are apprehensions on the ability to control the presence of certain indesirable substances or contaminants, often varying from country to country.

9            Replies to question 62, Questionnaire 1- customers.

10         Replies to question 62, Questionnaire 1- customers.

11         Case No COMP/M.2416 - Tetra Laval/Sidel, Commission decision of 30 October 2001, paragraphs 205-206.

12         Case No. COMP/M.2416 - Tetra Laval/Sidel, Commission decision of 30 October 2001, paragraph 45.

13         Replies to question 10, Questionnaire 1- customers.

14         Replies to question 12, Questionnaire 1 - customers and Questionnaire 2 – competitors.

15         Replies to question, 16 Questionnaire 1 - customers and replies to question 11, Questionnaire 2 – competitors.

16         Replies to question 15, Questionnaire 2- competitors.

17         Replies to question 10, Questionnaire 2- competitors.

18         Agreed minutes of a telephone conference on 3 March 2015 with a large international drinks  producer, paragraph 12.

19         Replies to question 13, Questionnaire 1- customers and replies to question 13, Questionnaire 2 – competitors.

20         Case No. COMP/M.2416 - Tetra Laval/Sidel , Commission decision of 13 January 2003, paragraph 22.

21                  Replies to questions 6, 7 and 8, Questionnaire 1 - customers.

22         Case COMP/M.5760 - Lotte Group/Artenius UK Ltd, Commission decision of 19 March 2010, paragraph 8.

23         OFT decision of 4 October 2007 - Anticipated acquisition by La Seda de Barcelona S.A of Amcor PET Packaging Europe, paragraph 15.

24         Case COMP/M.2416 decision of 30 October 2001, paragraph 211; confirmed in Case COMP/M.2416

– Tetra Laval/Sidel, Commission decision of 13 January 2003, paragraph 54.

25         It has to be noted however that shorter and/or significantly longer distances are also recorded for certain customer relationships.

26         Replies to question 19, Questionnaire 1- customers.

27         Reply to question 19, Questionnaire 1 – customers.

28         Replies to question 23-24, Questionnaire 2 – competitors, replies to question 24, Questionnaire 1 – customers.

29    The Parties' combined share of production including the Parties' and their competitors' captive supplies of virgin and recycled PET would amount to approximately [5-10]% in the EEA.

30    Each Plastipak and APPE use approximately [5-10]% of their preform production internally to supply its own downstream PET bottle production process, while some of their competitors, for example Logoplaste (to the largest extent) or ALPLA, choose to have much larger part of their preform production vertically integrated into the downstream production of bottles.

31         Replies to question 39, Questionnaire 2 – competitors.

32         Replies to question 55, Questionnaire 1 – customers.

33         Internal document of Plastipak: "Competitive Analysis (the impact on the Czech Republic and on the Slovakia)", 28.07.2014, page 8.

34         The data is provided on a volume basis (number of preforms produced). The Notifying Party argues

that comparing revenue data of various PET producers would not provide a meaningful picture. It explains that some producers accept so called hard-tolling arrangements, where the preform  customers purchase the PET resin on their own and pay the preform producer only for the conversion of the PET resin into a preform. Since the resin is the most valuable part of the preform (approximately 70% of the value), comparing revenue data of various PET producers would overestimate the market share of those who do not have tolling arrangements and underestimate the share of those who have hard tolling contracts.

35                  Calculation by the case team, based on the share of vertical integration of other players (excluding self-manufacturers) with regard to all types of PET preforms, as submitted by the Notifying Party (table 1).

36                  Internal documents of APPE: Dairy sector market overview (Dairy sector D1 290415).

37                  The analysis covers 80% of the top customers of each Party (in volume).

38                  The methodology consisted in identifying annual fluctuations in unit sales larger than 5 million units, primarily sales conducted via contracts as the Parties could not identify with certainty most switches that occured in theit spot business.