EUR-Lex Access to European Union law

Back to EUR-Lex homepage

This document is an excerpt from the EUR-Lex website

Document 31998D0602

98/602/EC: Commission Decision of 15 October 1997 declaring a concentration to be compatible with the common market and the functioning of the EEA Agreement (Case No IV/M.938 - Guinness/Grand Metropolitan) (notified under document number C(1997) 3169) (Only the English text is authentic) (Text with EEA relevance)

OJ L 288, 27.10.1998, p. 24–54 (ES, DA, DE, EL, EN, FR, IT, NL, PT, FI, SV)

In force

ELI: http://data.europa.eu/eli/dec/1998/602/oj

31998D0602

98/602/EC: Commission Decision of 15 October 1997 declaring a concentration to be compatible with the common market and the functioning of the EEA Agreement (Case No IV/M.938 - Guinness/Grand Metropolitan) (notified under document number C(1997) 3169) (Only the English text is authentic) (Text with EEA relevance)

Official Journal L 288 , 27/10/1998 P. 0024 - 0054


COMMISSION DECISION of 15 October 1997 declaring a concentration to be compatible with the common market and the functioning of the EEA Agreement (Case No IV/M.938 - Guinness/Grand Metropolitan) (notified under document number C(1997) 3169) (Only the English text is authentic) (Text with EEA relevance) (98/602/EC)

THE COMMISSION OF THE EUROPEAN COMMUNITIES,

Having regard to the Treaty establishing the European Community,

Having regard to the Agreement on the European Economic Area, and in particular Article 57(2)(a) thereof,

Having regard to Council Regulation (EEC) No 4064/89 of 21 December 1989 on the control of concentrations between undertakings (1), as amended by the Act of Accession of Austria, Finland and Sweden, and in particular Article 8(2) thereof,

Having regard to the Commission Decision of 20 June 1997 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 (2),

WHEREAS:

(1) On 16 May 1997 Guinness plc ('Guinness`) and Grand Metropolitan plc ('GrandMet`) notified to the Commission their intention to create a new company, to be called GMG Brands plc ('GMG`) in which they will merge all their business activities.

(2) By Decision of 6 June 1997 the Commission ordered the continuation of the suspension of the notified concentration, pursuant to Article 7(2) and Article 18(2) of Regulation (EEC) No 4064/89 ('the Merger Regulation`) until it takes a final decision.

(3) On 20 June 1997, after examination of the application, the Commission concluded that the operation fell within the scope of the Merger Regulation and raised serious doubts as to its compatibility with the common market and decided to initiate proceedings pursuant to Article 6(1)(c) of the Merger Regulation.

I. THE PARTIES

A. GUINNESS

(4) Guinness is the UK-registered holding company of a group whose principal business activities are the production and distribution, throughout the world, of spirits (United Distillers - 'UD`) and the brewing of beer. It also has interests in hotels and in publishing.

B. GRANDMET

(5) GrandMet is the UK-registered holding company of a group whose principal activities are the production and worldwide distribution of spirits (International Distillers and Vintners - 'IDV`), also food manufacturing (Pillsbury, Haagen-Dazs) and ownership of fast-food restaurants (Burger King).

II. THE OPERATION

(6) The operation consists of the merger of the two parties' businesses to create GMG. It will be effected by means of a Scheme of Arrangement of GrandMet under section 425 of the UK Companies Act 1985. Guinness will be renamed GMG Brands plc and its shareholders will retain their holdings through shares in the new company. The business of GrandMet will be vested in GMG; shareholders in GrandMet will receive one GMG share for each GrandMet share. Immediately after the merger, former GrandMet shareholders will hold approximately 52,7 % of GMG, and former Guinness shareholders the remainder.

III. THE CONCENTRATION AND THE COMMUNITY DIMENSION

(7) The operation is a merger of the two parties and is accordingly a concentration under Article 3(1)(a) of the Merger Regulation. The parties have combined aggregate worldwide turnover of more than ECU 5 000 million (Guinness has turnover of more than ECU 5 000 million and GrandMet more than ECU 11 000 million) and each has Community-wide turnover of more than ECU 250 million (each party having turnover of more than ECU 3 000 million). They do not achieve more than two thirds of the latter within one and the same Member State. The operation therefore has a Community dimension. It does not fulfil the criteria under Article 2 of Protocol 24 to the EEA Agreement, and therefore does not fall to be treated as a cooperation case under that Agreement.

IV. COMPATIBILITY WITH THE COMMON MARKET AND WITH THE FUNCTIONING OF THE EEA AGREEMENT

A. RELEVANT PRODUCT MARKETS

1. Horizontal aspects

(8) The activities of the parties overlap materially in the supply (manufacture and wholesale distribution) of spirits throughout the world. Both parties manufacture a range of spirits covering all the main internationally-recognised spirit types. However, they do not manufacture local spirit types such as Korn, popular in Germany, or genever, popular in Belgium and the Netherlands. Spirits are produced by the distillation - heating and condensation - of fermented fruits or cereals with water to yield a strong yet palatable alcohol. The distillate may then be matured in bulk and/or blended with other alcohols or other flavourings before bottling.

(9) A continuum of possible product market definitions was put forward in the course of the Commission's investigations. In decreasing order of breadth, they comprise: 'all spirits` (as originally proposed by the parties); various groupings - for example, 'brown` and 'white` spirits, (that is, separating whiskey (3), brandy and so forth on the one hand from gin, vodka and so forth on the other); separate markets for each spirit of the same general type, such as gin, whisky, brandy, vodka, rum, further segmentation by origin/quality, especially for whiskey, separating, for example, Scotch whisky from other types; and, finally, subdividing each (narrow) spirit type by price/quality - thus, Scotch would be subdivided into, for example, 'de luxe`, 'premium` and 'standard`. The various possibilities are examined below.

(a) All spirits

(10) The parties provided details of consumer surveys which, in their view, suggested that consumers were willing to substitute one type of spirit for another, and even to substitute other drinks, according to the occasion, availability and price. The Commission noted, however, that where those surveys (most of which were originally aimed at addressing taxation issues) employed price-change data, the overall levels of change (which mainly reflected changes in taxation) were much higher than those normally used by competition authorities as an aid to market definition. It also noted that occasion-based consumption patterns did not of themselves indicate a wider product market. The tendency to consume different spirit types on different occasions (for example, gin before a meal, brandy afterwards) implies that consumers have preferences for several specific types rather than that they are indifferent as to which type they consume (which they would have to be if the products were close substitutes). This was supported by the Commission's own investigations. The bulk of competitors and customers who responded to the Commission's question on the subject indicated that an increase in the price of their preferred spirit brand or type would not induce them to change to a different product category. Moreover there are significant supply-side differences. All spirit types involve distillation, but ingredients and processes vary considerably. On the basis of all these considerations, the Commission considered that an 'all spirits` market definition was unduly wide for the purposes of assessing this case.

(b) White and brown spirits

(11) 'White` spirits: As the parties pointed out, there appears to be some scope for employing the same basic distilled grain spirit to produce a variety of white spirit types (such as gin or vodka and their derivatives). However, other aspects of the manufacturing process differ - for example, to produce gin, juniper (and sometimes other flavours) must be added during distillation or afterwards; vodka is usually unflavoured, but may be distilled several times before bottling, white rum uses a different fermentation process (and ingredients) from those used in gin or vodka production. On the demand side, the parties themselves revised their initial view to some extent, stating that in practice consumers did not consider the different main white spirit types (gin, vodka) to be substitutes for each other; the age and social profiles of the consumers concerned were significantly different.

(12) 'Brown` spirits: For brown spirits such as whiskey, brandy, or dark rum, there is effectively no scope for supply-side substitution. They acquire their distinctive flavour from the ingredients used in the distillation and from the maturation process, whereas white spirits are seldom matured and have little flavour of their own. On the demand side, as already mentioned, consumption patterns do not support the notion of a brown spirits market - consumers may, for example, drink both brandy and whiskey, but on different occasions, not as substitutes.

(c) Segmentation by spirit type

(13) The strongest determinant of product market boundaries in spirits appears to be that of consumer demands and preferences, since they will drive the stocking and marketing policies of retailers, wholesalers and ultimately manufacturers. Third parties - both competitors and customers - consulted by the Commission generally supported a relatively narrow definition, characterising most spirit drinkers as having a degree of loyalty towards one or a few specific brands within the category (or categories) of choice, and with occasion-based consumption patterns, which were well-entrenched and unlikely to be seriously disturbed by relatively small price variations between types.

(14) The importance of branding, and its application to individual spirit types, is a key characteristic of competition in the spirits industry and is not consistent with a product market wider than that for each main spirit type. Product development, advertising and promotion expenditure are normally focused on the brand rather than the category or company, especially at consumer level. Thus, for example, advertisements are typically for Johnnie Walker' whisky rather than for Guinness's range of whiskies or for their spirit products collectively. It is also of interest in this connection that, in general, brands do not appear to be easily transferable between spirit types; for example, there is no Johnnie Walker gin or Gordon's whisky. This behaviour is consistent with product markets based on each spirit type, since it implies that manufacturers have adopted a strategy of branding for, and within, each spirit type in order to satisfy specific consumer demands in terms of taste, price and image.

(d) Segmentation within spirit type

(15) Some third parties considered that country of origin was an important defining factor for some, if not all spirit types. This point was most frequently raised in regard to whiskey, and in particular Scotch. It was pointed out that the industry markets Scotch, Irish, American, Canadian and other whiskies separately, and that for Scotch in particular, national legislation laid down certain requirements (it must be wholly distilled in Scotland and matured there for at least three years before it can be sold as Scotch).

(16) In the Community as a whole, Scotch accounts for the great majority (over 80 %) of whiskey sales, and the parties' main interests are in Scotch rather than other whiskey types. For the purposes of the present case, therefore, the issue of segmentation of whiskey by origin is only analysed further where the different possible definitions would lead to significantly different shares and overlaps (notably, Spain and Ireland).

(17) A similar approach is adopted to the possible substitutability between gin and genever, where it has been suggested that it is unnecessary to separate 'London` types from local products, and, in the Benelux countries in particular, from 'Genever`. The issue is considered further in the section on Belgium/Luxembourg below.

(e) Further subdivision by price/quality

(18) In most of the main spirit types, and especially in whiskey, there is a wide range of products available at different prices. Some third parties have suggested that there are in effect separate markets for each quality level because, for example, a consumer who habitually drank a premium brand would not regard a cheaper one as providing an adequate substitute in terms of taste, image and so forth. Price data provided by the parties show generally that in most markets there tend to be products available (both from the parties and from their competitors) at all points along the range. However, those price series data do not take account of the importance of particular brands in terms of their market share. Accordingly this point is examined further in those product and geographic markets where the distinction is likely to be relevant.

(f) Liqueurs

(19) Given that there are many different liqueurs, each with highly distinctive taste and other characteristics, it appeared reasonable to assume, as a point of departure, that in general each liqueur constituted a separate, niche product market. No information or argument was received during the Commission's investigation to suggest otherwise.

2. Vertical issues

(20) Like most of the major players in the spirits industry, the parties' distribution systems in their developed markets show some degree of vertical integration, with the use of wholly owned subsidiaries responsible for the exclusive distribution of their products at national level. Like their competitors, the parties may distribute various spirits on behalf of third party manufacturers or brand owners. In general, however, such third-party distribution accounts for only a relatively small proportion of their total distribution activity. This suggests that the primary function of these vertically integrated distribution operations is to protect the parties' brands rather than to undertake distribution on behalf of others. Integration into distribution allows the supplier to maintain control over the marketing and distribution of the brands which it owns, thereby safeguarding the all-important image of those brands in the market-place. Accordingly, their role as independent distributors, competing actively for distribution of third parties' brands, is likely to be limited (for example, to those instances where they do not possess an important brand of their own in a particular product category). In those instances where the parties appear to play a more significant role in the distribution of third-party products, this is considered in more detail below.

(21) On-trade and off-trade (4): A question arises as to whether the conditions of competition are sufficiently different to justify separate treatment for supplies to and from the on-trade and off-trade. A number of elements could be taken to suggest that those retail channels might constitute separate product markets. For example, on-trade consumers usually buy a smaller measure (a glass rather than a bottle) and the purchase also includes the provision of the related services (use of the premises and so forth). Consumers may also be less price-sensitive when purchasing from the on-trade; for example, in the United Kingdom, a 5 % change would typically give a difference of a few pence on a glass compared to some 50 pence on a bottle. However, in the present case the primary impact of the operation is in terms of supplies to wholesalers and large retailers, since small retailers, whether in the on-trade or the off-trade, acquire their supplies from those sources rather than direct from the parties. Therefore the different characteristics of each channel are taken into account where relevant in the assessment of the affected markets.

(22) Duty-free sales: The parties submitted that duty-free sales were a separate market. Duty-free purchases of spirits are only available to those travelling by air or sea across borders. They are often made as gifts (as evidenced by the wide range of special packages available in duty-free shops) and probably compete more strongly with other duty-free goods such as perfume rather than with non-duty-free spirits. No third-party evidence was received to contradict this view.

3. Conclusion

(23) In view of the above, the Commission considers that the relevant product markets in this case, at all levels of the supply chain, are in general no wider than those for each of the individual internationally-recognised main spirit types (whiskey, gin, vodka, rum and so forth) and for each liqueur. Narrower definitions may, however, be appropriate to specific product or geographic areas. Where such distinctions are likely to be relevant to the merger's effect on competition, they are considered further below. The Commission accepts, however, that duty-free sales of spirits constitute a separate market.

B. RELEVANT GEOGRAPHIC MARKET

(24) The parties contended that the relevant geographic market was Community (or EEA) wide, pointing to relatively low transport costs, absence of import restrictions, moves towards harmonisation of taxation, and the extent of parallel trade, assisted by the development of the 'under-bond` warehousing system by which spirits can be shipped across borders without payment of duty. The Commission, however, considers the relevant markets to be essentially national, for the following reasons.

(25) Consumption patterns vary widely between Member States. The total amount consumed per capita varies greatly. In several Member States, locally produced spirits are prominent which are virtually unknown elsewhere, such as Korn in Germany or ouzo in Greece. Even among the internationally recognised types, consumption varies significantly between countries, both as regards the type of spirit and in terms of brands.

(26) Distribution is also organised primarily on a national basis. Although production facilities are concentrated, with a few plants serving the whole of the Community, the parties, like their main competitors, operate wholly owned subsidiaries (or, in some cases, joint ventures) which handle primary distribution and marketing in each national territory concerned.

(27) There is varying scope for parallel trade in spirits between Member States (according to factors such as geographic location and relative tax and currency levels). But major taxation and regulation differences continue to exist between Member States, and these are reflected, to a large extent, in retail prices. If parallel trade (or cross-border purchases by consumers) were sufficient to create a single market across the Community, it would become impossible for national governments to sustain these distinctions in the long term, and retail prices would vary little between Member States - indeed, there would cease to be any incentive to engage in parallel trade.

(28) The position is, however, different for duty-free sales. These have characteristics, such as packaging and quantity (and sometimes alcoholic strength) differences compared to non-duty-free equivalents, and an international (not just Community) and transient customer base, which suggest that the market for them is likely to be at least EEA-wide.

(29) Accordingly, the Commission concludes that, with the exception of duty-free sales, the relevant geographic markets in this case are no wider than national.

C. ASSESSMENT

1. The structure of the market

(30) The operation will combine the activities of the two largest spirits suppliers in the world (and, on the basis of sales figures by value, the two largest in the Community) creating a company approximately twice the size of its nearest rival, Allied Domecq.

(31) However, as explained above, competition takes place principally within each different spirit type, and at national level. The Commission's analysis has accordingly focused on those product/national markets where the merger produces potentially significant market shares and aggregations, and/or the possibility of a 'portfolio effect` as described below. The individual country markets are considered in detail in the appropriate sections of this Decision. Nevertheless there are some general features common to the individual country assessments, and these are considered here.

(a) Manufacturing

(32) Although the parties will have the scope to rationalise their production facilities, the merger does not seem likely to have a major effect on competition at that level. Their production facilities in the Community are already quite concentrated, especially with regard to gin and vodka, with a few plants responsible for most of the output, reducing the opportunity for rationalisation of capacity or other economies of scale which might lead to increased barriers to entry at that level. In whiskey (the parties' production being effectively all Scotch) the scope for rationalisation or other economies from the merger is also limited. There is some excess capacity, but distilleries can fairly easily be closed temporarily and brought back into operation if demand recovers (the parties have several plants in this condition). In the case of malt whisky - the high-quality product derived exclusively from malted barley, which is mixed with the cheaper grain whisky (mainly made from maize and so forth) to produce the popular blends and is also sold on its own - the product of each distillery is highly distinctive in colour and flavour. In consequence, the different malts are in continual demand both for sale on their own and for blending and there is accordingly little scope for rationalisation.

(33) Grain whisky production, in which the parties will have a [40-50 %] (5) combined share of capacity - increment to Guinness (already the leader) of [15-25 %] - presents relatively few barriers to entry by comparison with malt whisky, and there remain significant competitors, namely Allied Domecq with [15-25 %] and Grants with [15-25 %]. The plant (or still) employed is effectively the same as that used for white spirits, and in contrast to the pot stills used for malt whisky can be run continuously. Moreover, flavour and colour characteristics, and hence the nature and quality of the ingredients (especially the water) used to produce them, are of much less importance.

(34) Some third parties raised the possibility that the parties would be able to foreclose supplies of, in particular, malt whiskies to competitors, many of whom use some of the parties' malts to make their blends. The parties' shares, and the increment, in malt whisky are lower than in grain (combined shares of [25-35 %] increment to Guinness, the leader, of [< 10 %]) and again there will remain some substantial competitors, such as Seagram with [10-20 %] and two others each with about [5-15 %].[ . . . ]

(35) In the light of the above, therefore, the Commission takes the view that the merger will not create or enhance a dominant position at the manufacturing level in any relevant product market.

(b) Branding

(36) As explained above, the spirits industry is to some extent vertically integrated, with most of the largest players (including the parties) active in both manufacture and distribution, although integration into retail is uncommon (the United Kingdom being a partial exception). Vertical integration into distribution is advantageous in product markets where branding is important, because it allows the brand-owner to retain full control of product development, promotion and marketing. Retailers play relatively little independent role in advertising branded spirit products - apart from their own brands, if any - although they will work closely with manufacturers where there is mutual advantage in doing so.

(c) Distribution

(37) The larger spirits manufacturers also engage in distribution on behalf of other suppliers, typically through exclusive agreements covering a particular national territory. This feature provides access to a route to market for smaller suppliers without a distribution network of their own. However it also reduces the scope for competition in distribution from independent distributors, and increases the barriers to entry for a potential new supplier of a brand which would compete directly with one owned by the vertically-integrated supplier.

(d) Portfolio effects

(38) One competitor remarked that 'In short the market power deriving from a portfolio of brands exceeds the sum of its parts`.

(39) In addition to horizontal overlaps, assessed in detail below, a key result of the merger, recognised as a major part of its rationale by the parties and by third parties, is that it combines the two parties' ranges or portfolios of products and brands.

(40) The holder of a portfolio of leading spirit brands may enjoy a number of advantages. In particular, his position in relation to his customers is stronger since he is able to provide a range of products and will account for a greater proportion of their business, he will have greater flexibility to structure his prices, promotions and discounts, he will have greater potential for tying, and he will be able to realise economies of scale and scope in his sales and marketing activities. Finally the implicit (or explicit) threat of a refusal to supply is more potent.

(41) The strength of these advantages, and their potential effect on the competitive structure of the market, depends on a number of factors, including: whether the holder of the portfolio has the brand leader or one or more leading brands in a particular market; the market shares of the various brands, particularly in relation to the shares of competitors; the relative importance of the individual markets in which the parties have significant shares and brands across the range of product markets in which the portfolio is held; and/or the number of markets in which the portfolio holder has a brand leader or leading brand.

(42) In addition the strength of a portfolio effect has to be considered in the context of the relative strength of competitors' brands and their portfolios.

(43) The portfolio effect has been recognised in two recent cases in the soft drinks sector (6).

(44) Furthermore in response to the Commission's enquiries, competitors and customers recognised the portfolio effect in practice; for example, of ten firms responding to the question 'does possession of a leading brand in all or most spirit categories help sales of spirits in general?` eight replied that it would help a lot.

(45) At the hearing, major competitors of the parties confirmed the existence of a portfolio effect and provided evidence of how their portfolio could be used.

(46) The parties have said that consumers buy brands and not portfolios. That is true. However, the parties do not sell their products to the final consumers. They sell them to intermediaries, multiple retailers, wholesalers and others. Those customers would buy a range of products from GMG, and the fact that GMG would be able to offer in many national markets a wide and deep portfolio of leading brands would give the combined entity advantages in dealing with its clients. The strength of any portfolio effect will vary from geographic market to geographic market. In the present case, the only market in which it has been considered to be a significant feature in the context of the assessment is Greece, and it is accordingly assessed further in the section on Greece below.

2. Barriers to entry

(47) Portfolio power can accordingly be seen, in part, as a barrier to entry. There are, however, a number of other barriers, which are outlined in general terms here, although their precise strength will vary from country to country and they are accordingly examined in more detail, where appropriate, in the relevant sections.

(a) Commission's initial view

(48) In its 'Statements of objections`, the Commission identified the following as potentially significant barriers.

(49) General market conditions do not favour entry, since there are barriers at all levels of the supply chain.

(50) As mentioned above, in the Community as a whole spirits demand is in decline. In general, therefore, sales can only be obtained at the expense of competitors unless a new niche can be found. Entry is further restrained by government restrictions on alcohol advertising, promotion and consumption. Retailers and wholesalers generally require a licence, in many instances their number and location is restricted to what governments consider appropriate, and their opening hours are also usually restricted. As in the case of tobacco, most countries restrict advertising (especially on television). There are also a variety of restrictions, varying from country to country, on matters such as labelling, strength and quantities that may be offered for sale.

(51) Entry into spirit manufacturing is not technically difficult, although entry into Scotch is constrained by the need to have access to supplies of matured malt whisky from Scotland or else to develop one's own maturing facilities in that country (which by definition will take several years).

(52) Distribution and marketing are potentially significant barriers. As explained above the supply chain is characterised by branding and portfolio effects which increase the costs and time required for successful entry even with a clearly distinctive and innovative product or range. It is said for example that it took 10 years for GrandMet to develop its Bailey's Irish Cream liqueur into a recognised brand carried by most retailers, and during most of that time it was effectively unopposed. Together with niche products like fruit-flavoured, lower-alcohol liqueurs (such as Archer's peach schnapps and Malibu coconut liqueur) it is almost the only significant product innovation in the market of recent years. There has been more line extension (for example by introducing a premium whisky brand with a name similar to an existing standard one) but that option is available only to those with established products. Famous Grouse Scotch (Matthew Gloag) is almost the only 'new` brand of a major spirit from the parties' competitors to have gained a material share across the range of Community markets, and again its development to that position has taken a considerable number of years. By contrast, Johnnie Walker is, in the words of its own advertising 'still going strong` after some 150 years, an illustration of the power of an established brand. Japanese whiskies, although strong on their home markets and based closely on Scotch in terms of quality and nature of ingredients and the manufacturing process, have so far failed to make significant inroads into Community markets despite considerable promotional efforts. Spanish whiskey is virtually unknown outside Spain.

(53) New entrants need to have, or quickly develop or acquire, not just one desirable product, but a range of them, in order to counter the portfolio power of competitors such as GMG. Using a third-party distributor may assist in this, since it enables the entrant's products to be sold alongside the distributor's existing range. But the vertical integration of the major players in the industry limits the effectiveness of this approach, since an entrant's product is unlikely to be accepted or marketed vigorously by the distributor if it competes directly with a brand which he owns or for which he has the agency. Entrants will need to advertise and promote their brands heavily over a long period in order to overcome consumer resistance and the barrier presented by the reputation of the established players. Retailers and wholesalers, especially large ones, will often expect a significant listing fee for carrying a new product (one respondent from the United Kingdom indicated that the fee could reach hundreds of thousands of pounds per retailer). The entrants will need to take account of the importance of on-trade consumption in shaping consumer demand for a product, since customers will be more likely to try a new product if they only have to buy a glass of it rather than a bottle. They will also need to have resources to defend themselves from attack by the parties using their incumbency advantages, for example, by targeting selective discounts or price reductions at new entrants' brands. General market conditions are not favourable to the taking of such risks. Overall, therefore, entry barriers appear generally to be very high and likely to deter entry.

(b) The parties' response

(54) In their response to the Commission's 'Statements of objections`, the parties made a number of general points seeking to rebut this view. Those points are examined immediately below while their points on specific markets are dealt with in the relevant sections.

(55) The parties referred to a number of substantial international spirits companies which had entered the Community market within the last 10 years, namely American Brands, Suntory, Bacardi Martini and Brown Forman. Although all these are undoubtedly substantial companies, their strength varies between the national markets and several are not new entrants in the sense that they were already well established in the spirits sector, albeit not necessarily in the same national or product markets.

(56) American Brands owns the Jim Beam bourbon whiskey brand and has a range of other brands which it either owns or distributes. It has acquired two UK distillers, White & McKay and Invergordon, and as a result has a share of around [5-15 %] in the UK whiskey and vodka markets. But its brands are generally secondary rather than category leaders and its share in the individual country markets identified in this Decision as causing concern is generally very small, both overall and in the main spirit categories. Furthermore, its products were established before they were acquired by American Brands.

(57) Suntory is a Japanese producer, primarily of whiskey. It has acquired two Scotch malt whisky distilleries and undoubtedly has substantial resources. But its share in any spirit category in any of the markets identified as being of concern is very small (less than [< 5 %]).

(58) Bacardi-Martini is only a new entrant in a very narrow sense, since it is the result of the merger in 1993 of two already very well-established companies, Bacardi and Martini-Rossi, the latter being domiciled in the Community. The products were already established. On the same basis, presumably GMG would itself be classed as a new entrant after the merger.

(59) Brown Forman's position is similar to that of American Brands. It owns Jack Daniels, a leading bourbon, and Southern Comfort, an orange-flavoured bourbon-based liqueur. But outside those niche areas its share is very small in the markets of concern. Both products were already well established.

(60) The parties also pointed to expansion by existing suppliers as evidence that barriers were low, asserting that such expansion showed that there was an adequate threat of or potential for entry to ensure that competition was not significantly reduced following the merger.

(61) Various incumbents have indeed diversified geographically and/or in product terms, for example Matthew Gloag has successfully developed its Famous Grouse whisky outside the United Kingdom and other Scotch producers, such as Grants, have diversified into gin and/or vodka. Nevertheless, those developments have generally occurred at a slow pace. For example, it took some four years for Famous Grouse to achieve a 5 % share of whiskey in Greece, despite considerable advertising spend and the brand's high profile in the United Kingdom.

(62) The parties sought to rebut the Commission's argument that advertising and promotion costs represented a significant barrier by suggesting, first, that regulation of advertising merely presented a level playing field to new entrants. The Commission does not accept the level playing field argument. Restrictions on advertising favour incumbents, whose products are known to customers and so require less advertising, against entrants, whose products are by definition unknown to customers. Generally, and especially for entry into the mainstream categories (that is internationally recognised spirit types or categories) where incumbents are well-established, significant sunk costs in terms of advertising and promotional costs are an essential prerequisite of entry, especially given the importance of branding.

(63) The parties suggested that listing fees charged by retailers were not a significant barrier but exemplified retailers' buyer power. Matthew Gloag, for example, had evidently found the necessary resources to fund them for Famous Grouse and retailers were free not to charge them if they wished to encourage an entrant. The Commission accepts that listing fees are not an insuperable deterrent. Nevertheless they raise the price of marketing a new brand or product in comparison with that of an established one and to that extent they constitute an entry barrier.

(64) The parties also argued that distribution was not the bottleneck in preventing and limiting entry that the Commission, in its 'Statements of objections`, had claimed it to be. They pointed out that in all Member States there were a number (in some, a large number) of distributors willing to undertake distribution of new products and brands. They provided information suggesting that agency distribution agreements were typically of short duration (if they had any fixed term) and that changes were frequent. They also indicated that part of the rationale for their (and competitors') move towards wholly-owned distribution facilities was to guarantee a route to market for their products in the face of the risks of losing existing distribution arrangements if a competitor's offer became attractive to the distributor concerned. In itself this reduces the scope for independent distributors who no longer have access to many brands, no matter what terms they might offer.

(65) Nevertheless it remains the case that in several Community markets the parties already have a large share of spirits distribution. Where this occurs there is a risk that new entrants will find themselves unable to obtain a route to market on favourable terms for any product which the parties may consider to threaten those in their own portfolio. Since the parties' portfolio will be much enhanced by the merger, the merger is likely to increase this risk. Direct distribution, by supplying direct to retailers without the use of an agent or intermediary such as a wholesaler, may be a possible alternative in some product and/or geographic markets. But in practice this is unlikely to be a viable option in most cases. In the case of supplies to supermarkets, new entrants are likely to face a requirement to pay a listing fee or to have to agree to similarly onerous terms and conditions as the price of getting their product onto the shelves. The on-trade is more fragmented, making direct distribution generally uneconomic for suppliers.

(c) Conclusions on entry barriers

(66) The Commission accepts that new entry is possible and has indeed taken place. But the scale and scope of that entry does not, in general, appear to have been significant, especially in the main spirit categories. This, together with the other factors described above, tends to reinforce the Commission's initial view that barriers to entry in this market are generally high. However, their precise importance varies from country to country and between products and is accordingly taken further into account in the sections below concerning individual countries.

3. Countervailing buyer power

(67) Where entry barriers are high and market structures concentrated, the size and strength of customers in the individual national markets is particularly important to the assessment of the effect of a merger both in terms of horizontal overlap and of any associated portfolio power. Again, the position in the present case varies considerably between countries and is discussed further in the next section of this Decision. But there are also some common elements which are examined here.

(a) Commission's initial view

(68) The Commission considered that, in summary, the requisite structure, in consumer markets, for the exercise of countervailing power as a means of preventing creation or reinforcement of a dominant supply position at a higher level in the distribution chain appeared to be as follows. The retail and wholesale customers must include several who are each responsible for a significant share of sales by the dominant supplier and by his competitors and have the necessary technical facilities and bargaining skills to put that advantage to use in the buying process. Crucially, there must be alternative suppliers capable of offering an equivalent range of products on equally favourable terms and conditions and the retailer or wholesaler must have effective power to delist brands if the terms on which he is offered them are not satisfactory. The Commission doubted whether those conditions were generally met in the spirits sector, especially as regards the ability to delist. A number of suppliers, retailers and wholesalers, claimed that it would be uneconomic to delist a leading brand.

(b) The parties' response

(69) The parties strongly disagreed with that analysis and made a number of submissions in support of the contrary view, as follows.

(70) The parties produced data showing the total share for all products of the top five retailers in the Member States identified by the Commission as of concern. Those shares ranged from some [5-15 %] in Greece to some [30-40 %] in Spain. They produced similar data showing the shares of their own sales in the countries concerned accounted for by their largest five customers in each. Those shares varied from some [15-25 %] for Greece to over [55-65 %] for Belgium (7). They also stressed the importance of own-label sales, not only by supermarkets, but also in some instances in their view, by the on-trade.

(71) The parties also quoted the views of a leading UK stockbroker and merchant bank (HSBC) on buyer power and its implications for the merger. It suggested that the supplier had more to lose than the retailer in any negotiation over the possible delisting of the supplier's brands, principally because, if delisting occurred, the retailer would still have all the competitors' brands, plus their own brand (which in the United Kingdom - the example given - accounted for as much as 60 % of the retailer's sales) whereas the supplier would have lost a major customer.

(72) The parties disputed the Commission's view that customers for spirits were ill-informed about prices because the products were strongly differentiated (branding) and were purchased relatively infrequently. The parties observed that spirits were purchased more frequently than some other household products such as electrical goods and suggested that some spirit brands were so-called known value items since they were widely advertised and consumers could, for example, compare their prices with that of a duty-free equivalent.

(73) Finally, the parties challenged any notion that any of their products were unique, must-stock items. They suggested that their brands were no more (or less) essential to retailers than those of their main competitors. Retailers and wholesalers would need a range of the main brands, but not necessarily all of them. They also provided details from various countries of instances where their products had been delisted.

(c) Commission's further assessment

(74) The Commission has not claimed that buyer power is non-existent in the spirits sector; rather its concern is that it will be insufficient to prevent the creation or reinforcement of dominance in the present case. As far as the dependency figures given by the parties are concerned, the Commission notes that dependency varies considerably between countries and that the share of the parties' turnover accounted for by the five largest customers in, for example, Greece is relatively low (some [15-25 %]). Also, since the on-trade is, in most Member States, much more fragmented than the off-trade, both at wholesale and retail level, those figures may overstate the true strength of countervailing power, since many on-trade retailers and their suppliers will not have it. These factors are taken into account in the consideration of buyer power in the individual country sections that follow.

(75) In the off-trade, the importance of own brands is acknowledged. But if retailers and consumers really found own-brands (whose margins are, according to the parties, much higher than those of the main brands) so attractive, it would seem reasonable to expect a much greater degree of commoditisation and the disappearance of brands from the shelves than has actually occurred. Retailers appear to find it necessary to stock a wide range of brands, presumably because, as several have indicated to the Commission in this case, if they do not they risk losing a customer, not only for the whisky and so forth, but for all the rest of their purchases from the store as well.

(76) It follows that permanent delisting of major brands is generally not a realistic option for retailers; accordingly, suppliers of those brands can, if necessary, get tough in the knowledge that the worst they can expect is a temporary delisting, perhaps limited to the less important sizes. That could very well be survivable, especially if, like the parties, the supplier also supplies a wide range of other brands and products. The stronger the brand and the portfolio in question, the less realistic the threat of delisting. By contrast, a less fortunate supplier could well risk a permanent loss of his main or only product.

(77) As regards consumer awareness of price and ability to compare, the Commission remains of the view that this is relatively limited in the spirits sector. Comparisons between on- and off-trade prices are very difficult, and within the on-trade, are complicated by the bundling of goods and services. In the off-trade, although purchases of spirits are, as the parties suggest, more frequent than of electrical goods for example, the frequency is still insufficient to allow effective comparison. Purchase of spirits is unlikely, for example, to be on, a weekly basis. With items such as electrical goods, by contrast, comparisons are facilitated by the fact that they are in most cases more expensive than spirits, the consumer has more to gain by shopping around, and consumer programmes and magazines provide further help. Duty-free prices of spirits may provide a base-line for comparison, but differences in pack sizes, and in some cases, product strengths, as between duty-free and other sales complicate the matter. Moreover, duty-free purchases will, for most people be even less frequent than non-duty-free ones.

(d) Conclusions on buyer power

(78) The strength of retailers and independent wholesalers obviously varies greatly between countries. The issue therefore needs to be analysed at that level. But on the basis of the foregoing arguments, it is not obvious that, throughout the markets of concern to the Commission, buyer power is likely to be sufficient to prevent the creation or reinforcement of a dominant position as a result of the merger.

V. EFFECTS IN SELECTED NATIONAL/PRODUCT MARKETS

(79) In this section those markets in which the Commission considers that dominant positions will be created or strengthened are examined in detail.

A. GREECE

1. General overview of the market

(80) Greece is the seventh largest consumer of spirits in the Community and in 1995 accounted for about 3 % of total Community sales by volume.

(81) Table 1 shows the relative importance of the different categories of spirits in Greece. An important feature is the large share of consumption accounted for by 'other spirits`, largely due to the sale of ouzo, the national drink. A further feature of the Greek market is the fact that although, in line with Europe as a whole, overall spirits consumption has fallen, the consumption of whiskey has grown. However, operators in this market believed that consumption of vodka, tequila, rum and gin, as well as some innovative spirits, such as fruit schnapps, was growing as well. Whiskey is the most consumed imported spirit, making Greece the second whiskey market in Europe on a per capita basis.

>TABLE>

2. Relevant product markets

(82) The relevant product markets under consideration in the Greek market are the internationally recognised categories of spirits, that is whiskey, vodka, gin, rum, brandy, the various liqueurs (each of which may constitute a separate niche product market), and the local ouzo aperitif.

(83) Scotch whisky accounts for 95 % of all whiskey consumed in Greece (8). However, on the basis of the market shares of the parties, the assessment of the operation would be similar no matter whether Scotch whisky or all whiskey is used. Therefore, the assessment below is made on the basis of all whiskey.

3. Market profile and position of the parties

(a) Position in the market

(84) The parties operate in Greece through wholly owned distribution subsidiaries. These are United Distillers Greece, a subsidiary of UD (Guinness) and Metaxa S & H & A AEBE, a subsidiary of IDV (GrandMet). The major brands of the parties are: Johnnie Walker Red Label, Dewar's, White Horse, Bell's, Haig, VAT 69 (Guinness) and J& B (GrandMet) in Scotch whisky; Smirnoff (GrandMet) in vodka; Metaxa (GrandMet) in brandy; Gordon's (Guinness) in gin; Ouzo 12 (GrandMet) in ouzo; and Baileys, Malibu and Archer's (GrandMet) in liqueurs and fruit schnapps. In addition, Guinness distributes Bacardi rum and Wyborowa and Finlandia vodkas, whereas GrandMet distributes tequila Cuervo, both on a brand agency basis. Table 2 shows the parties' individual and combined market shares in 1995 by value at brand owner and brand agency distributorship level, for the most important product categories in Greece.

>TABLE>

(b) Categories and brands of the parties

(85) The parties are present in all the major categories of spirits. Their various categories of spirits, as well as the relevant brands are shown in Table 3 below.

>TABLE>

(86) It may be seen from Table 3 that the combined entity will cover a broad range of spirits categories, in fact all the major and popular types of spirits. Taken separately, Guinness has currently a strong position in whiskey, gin and rum, whereas GrandMet is strong in brandy, ouzo, tequila and liqueurs. The merger thus fills the gaps in the respective portfolios of each party. The resulting combined portfolio will be by far wider and deeper than that of competitors.

(87) After the proposed operation has been completed, the combined entity will account for above [45-55 %] of the overall trade of spirits ([20-30 %] from Guinness and [10-20 %] from GrandMet), covering all the major categories of spirits marketed in Greece. The next largest competitors, Karoulias/Berry Brothers and Allied Domecq, have shares of [5-15 %] and [5-15 %] respectively.

(88) More specifically, GMG will be the driving force in the whiskey market, with a market share above [45-55 %]. In addition, it will be the largest supplier in categories such as gin with a market share above [75-85 %] (Gordon's), brandy, with a market share above [75-85 %] (Metaxa), and rum with a market share of [75-85 %] (Bacardi). Moreover, GMG will supply other categories, such as tequila (Cuervo), ouzo (Ouzo 12), cream liqueurs (Baileys and Malibu) and fruit schnapps (Archer's).

(c) Aggregation

(89) In so far as it concerns horizontal overlap in the individual categories, the proposed operation will give rise to an aggregation of market shares in the whiskey market. As a result, the combined entity's share in this market would amount to [45-55 %] (that is [40-50 %] from Guinness and [< 10 %) from GrandMet). The parties considered the accretion of [< 10 %] as a de minimis increment of shares in this market. However, the Commission considers this accretion significant, in particular in the light of the pre-existing high market share that one of the parties already had in this market, and of the fact that it is the result of the addition of one single brand, namely J& B, to the existing wide range of brands held by the other party. As can be seen in Table 4 in the top five leading whiskies in Greece in 1995, GMG would occupy positions one, three and five respectively with Johnnie Walker Red Label, Dewar's and J& B [ . . . ].

>TABLE>

(90) The rest of GMG's product range would be composed of a variety of other brands (Black & White, White Horse, VAT 69), and specialities, such as deluxe whiskies (Johnnie Walker Black Label, J& B Jet, Dimple) or malt whiskies (Lagavulin, Glenkinchie and so forth). By contrast GMG's competitors do not have such a broad range of whiskey brands. Karoulias-Berry Bros. supply Cutty Sark ([15-25 %] market share) and Allied Domecq Ballantine's ([5-15 %] market share), but have no other significant brands that would give them advantages similar to those that GMG will acquire.

(d) Portfolio effects

(91) Although there is no horizontal aggregation in other categories, the merger will bring together existing high market shares in gin, brandy and rum. Guinness' Gordon's gin accounts for over [75-85 %] of the gin market and is complemented by two premium quality brands, that is Tanqueray (Guinness) and Bombay Sapphire (GrandMet). Competing brands in this market include Allied Domecq's Beefeater [< 10 %] and Four Seasons [< 10 %], Amvyx's Nicholson's [< 10 %], and Seagram's Burnetts [< 10 %]. In brandy, Metaxa is the uncontested market leader with a market share of [70-80 %], whereas competitors offer Remy Martin [< 2 %], Martell [< 2 %], Courvoisier [< 2 %] and Hennessy with an insignificant market share. In rum, Bacardi is the leading brand with a [75-85 %] market share, whereas its only potential competitor is Seagram's Captain Morgan [5-15 %]. For the rest of the rum market, an industry report (Canadean) makes reference to 'cheap imitations of Bacardi`. Apart from the leading brands in their respective categories, GMG would also have a number of second-string brands including White Horse, Black & White, Bell's, Haig, VAT 69, Mackenzie, Crawford's, Dimple and Cardhu Scotch whiskies, Finlandia and Wyborowa vodkas, Grand Marnier and Sheridans liqueurs, Tanqueray gin and Karavaki and Kaloyannis ouzo.

(92) Accordingly, and given that market entry and countervailing buyer power are not significant constraints, as explained below, the Commission considers that the parties have existing dominant positions in the markets for gin, rum and brandy.

(93) Overall, GMG will be at least four times as big as the next largest competitors, none of whom account for more than [5-15 %] of the spirits market. It will therefore become the largest spirits importer and distributor in Greece. In 1996, Guinness achieved a turnover of GRD [ . . . ] and GrandMet of GRD [ . . . ], in a total market of GRD 118,2 billion. The next largest competitor, Karoulias-Berry Bros., which distributes the number two Cutty Sark whisky, achieved a turnover of GRD [ . . . ].

(94) The issue of portfolio power is of particular relevance in the assessment of the operation with regard to the Greek market. This is mainly due to the fact that the combined entity will be present across the major categories of spirits, that is whiskey, gin, rum and brandy, where it will be able to supply the leading brands, with the exception of vodka.

(95) To date, the Greek market has been characterised by the presence of various suppliers, none of which was strong across all the categories of spirits. As a result, customers, whether wholesalers or retailers, have obtained their spirits from a variety of suppliers, according to the latter's strength in the various categories. It is precisely in those market conditions that the combination of the most important spirits categories in one single supplier's portfolio is expected to enhance that supplier's market power in individual categories.

(96) As can be seen in Table 5, the portfolio of brands which GMG will bring together will include the top-selling brand in each of the main spirits categories, that is whiskey, gin, rum and brandy, with the exception of vodka. It will also contain the best-selling brands of tequila, ouzo and various liqueurs.

>TABLE>

(97) As stated above, whiskey is by far the largest category of spirits sold in Greece. However, the fact that GMG will include brands with very significant market shares in smaller categories is also important. For example, even if gin or rum have lower sales than whiskey, the presence of Gordon's and Bacardi is of crucial importance to a particular outlet, as these brands have been driving their respective categories for a long time and are identified with the category to which they belong. According to Canadean 1996 on Greece, 'gin continues to grow, largely due to a strong performance by Gordon's`. The same industry report refers to the rum market as consisting essentially of Bacardi.

(98) It is true that other competitors supply important brands, some of which have achieved high sales volumes. For instance, on the merits of their sales performance, Cutty Sark whisky and Stolichnaya vodka would not face particular problems in access to the trade. However, the potential power of those brands is significantly reduced by the fact that they are spread out among different suppliers. That fragmentation of the market, as contrasted to the combined portfolio of GMG, deprives such brands of their potential portfolio power.

(99) More particularly, a deep portfolio of whiskey brands, spread out across the various quality and price segments, confers considerable price flexibility and marketing opportunities. Therefore, the supplier is shielded from market pressures, as he is able to face price competition from other suppliers' brands by positioning and pricing his various brands within the category. For instance, with its secure high market performance of the best-selling whiskey brands, GMG will be able to devote as many resources as necessary in order to maintain its secondary brands in their position or to reposition the weaker brands upwards by expanding their share at the expense of competing brands, or in order to counter eventual competitive pressure coming from those brands. The parties have argued that such 'pull-through` has not occurred in the past and is accordingly unlikely to occur in the future. However, this argument ignores the substantial increase in the parties' market shares and resources that the merger will create.

(100) Moreover, a wide portfolio of categories confers major marketing advantages, giving GMG the possibility of bundling sales or increasing the sales volume of one category by tying it to the sale of another category. Both Guinness and GrandMet have made use of their portfolios of brands in bundling deals. For instance, in 1995, GrandMet rewarded customers who collected and delivered [ . . . ] bottle-caps of Smirnoff, Cuervo and J& B, by offering them a free [ . . . ]. For the same number of caps, wholesalers received a credit note of [ . . . ]. Moreover, Guinness made joint promotions of different categories and brands, whereby customers purchasing a 12-bottle pack containing Johnnie Walker Red Label (7 bottles), Gordon's gin (2 bottles) and White Horse (3 bottles) obtained a discount of [ . . . ]. It should be noted that these promotional campaigns were run with the cooperation of wholesalers who supplied the various trade channels. Finally, in 1996 GrandMet carried a similar discount campaign for wholesalers, offering discounts for purchases of [ . . . ] cases of a pre-selected mixture of the following GrandMet brands: J& B ([ . . . ]), Smirnoff ([ . . . ]), Cuervo ([ . . . ]), Baileys ([ . . . ]), Grand Marnier ([ . . . ]) and Malibu ([ . . . ]).

(101) In the on-trade, where spirits producers build a brand's strength and image, GMG, through its broad portfolio of brands, would be able to influence what products are stocked or displayed in the limited space available behind the bar, (the so-called back-bar), thus further strengthening its market power. For small outlets which have smaller back-bars, or for Greek night-clubs which concentrate on whiskey, the combined entity would be an attractive solution for one-stop-shopping considerations. In addition, larger modern outlets, which usually stock a much broader variety of brands, may also become a target of the combined entity, should it attempt to gain more back-bar space or use the image of fashionable clubs in order to launch its brands. GMG could afford to make substantial offers, discounts and credits or organise and finance promotional events, that would also accrue to the outlet itself, and use its strength in leading brands, such as Johnnie Walker Red Label, Gordon's gin and Bacardi rum, in order to induce bars to list brands in the same or another category. Given that those premises could not afford not to stock the brands set out above, the negotiating power of GMG would be significantly strengthened. Therefore, it would be much easier for GMG to induce bartenders to adopt GMG brands as pouring brands (that is, the brand offered when a customer fails to specify a brand by name), thus increasing their sales volumes and public awareness.

(102) In the off-trade, the elimination of competition between Guinness and GrandMet for in-store promotions will serve to enable GMG to plan jointly the timing of promotions, negotiate jointly the terms of promotions and coordinate any price changes. Moreover, through its variety of brands, GMG could also alternate branded products promoted over a period of time, thus occupying long promotion periods and excluding competitors from access to the promotion calendar for long periods.

(103) By comparison, competitors have weaker portfolios and fewer strong brands, the most important being Cutty Sark, accounting for [15-25 %] of whiskey sales, and Stolichnaya and Serkova vodkas, accounting for [20-30 %] and [15-25 %] of vodka sales respectively. As stated in the preceding paragraphs, although those brands may have performed well, they lack the support of a strong portfolio of brands. Indeed, in contrast to the complete GMG portfolio, the discontinuity of the competitors' portfolios would deprive them of price flexibility and make them more vulnerable to market pressures. For example, when their brands start losing sales volume, they will have to commit disproportionately stronger resources in order to avoid situations that could in the long run restrict their competitive scope.

4. Other potential competitive constraints

(a) Countervailing buyer power

(104) Such power could come from the trade and in particular the various trade intermediaries that are, in large part, the immediate customers of the spirits suppliers. In Greece, the various trade channels comprise wholesalers supplying on-trade and off-trade outlets, and retailers, including supermarkets, hypermarkets and cash and carry chains. Table 6 illustrates the split among the various trade channels, in terms of percentage of traded spirits.

>TABLE>

(105) The major wholesalers buy from importers and sell to small retailers and on-trade outlets in various areas of Greece. As a result of its investigation, the Commission concluded that wholesalers would not be able to exercise countervailing power against the combined entity for the reasons that follow.

(106) The parties said that out of a total of 700 wholesalers, 40 accounted for some 40 % of all imported spirits purchases. However, none of those were identified as being of a particularly significant size, whereas the remainder were small. The wholesale channel is therefore very fragmented. It follows that, overall, wholesalers would not be in a position to exercise countervailing power vis-a-vis the combined entity. On the contrary, in view of the incentives that GMG will be able to offer to them, wholesalers would have an obvious interest in staying on good business terms with GMG's marketing policy. That influence over wholesalers is particularly important to GMG, since wholesalers handle almost three quarters of the spirits trade in Greece. Moreover, since wholesalers are responsible for the supply of smaller retailers (29 % of their sales) and on-trade outlets (43 % of their sales), GMG will be able to take advantage of their access to those outlets in order to put into practice some of the marketing and promotional strategies, resulting from its large portfolio of brands, as described in the previous paragraphs on portfolio.

(107) Cash and Carry outlets constitute only a minimal portion of the market, 4 %, and their possible countervailing power is, therefore, not sufficient to reduce GMG's power.

(108) Retailers, such as supermarkets and smaller liquor shops handle 23 % of spirits trade in Greece. Although the big supermarket chains have characteristics enabling them to develop countervailing power, their replies to the questionnaires indicate that they have not done so far. On the contrary, their views suggested that because GMG has the largest number of best-selling brands and the broadest portfolio of spirits among the suppliers, it would be difficult to fulfil their needs, which are driven by consumer demand, without passing through GMG. Referring to the combined entity, a leading buying group in Greece noted that 'the two companies combined represent 56 % of all spirits purchases`. Referring to a number of GMG brands, in particular Johnnie Walker, Dewar's, White Horse, Haig, J& B, Gordon's, Smirnoff, Baileys and Grand Marnier, the buying group stated that those brands 'stand out on their own and impose themselves to the retail trade`. A large retailer, operating more than 30 supermarkets and hypermarkets, noted that 'there are certain brands of these companies which are essential for our stores to stock in view of their increased demand, such as Johnnie Walker, Bacardi, Metaxa, Ouzo 12, Dewar's, Dimple, etc.` The same retailer considered that the merger would have an impact at both consumer and retail level, since the reduction of competition would be prejudicial for the ultimate consumer.

(109) Moreover, the eight own-label whiskey brands which several supermarkets have developed account for a marginal proportion of total whiskey demand (estimated at around 5 %) and, taking into account the fact that in general whiskey consumers are brand-sensitive, own labels are not expected to put substantial competitive pressure on branded products.

(110) The parties provided a number of statements supplied by customers, which implied that the customers concerned did not believe the merger would have adverse impacts on their business. However, only limited reliance can be placed on those statements, in view of the fact that those customers have an important commercial relationship with the parties, which asked them to make the statement.

(111) Moreover, the parties presented a number of examples of on-trade outlets that had refused to stock some of Guinness' brands without payment of a listing fee. However, those examples were isolated cases which could not be considered as representative of the countervailing power of on-trade outlets. Indeed, they only concerned a limited number of on-trade outlets (six) and the delisting period was relatively short (one to three months).

(b) Parallel trade

(112) While the existence of cross-border sales is indicative of alternative supply sources and therefore of residual competition, in the case of Greece parallel trade does not appear to be a significant constraint. First, the volumes involved are not significant (9). Second, parallel trade seems to be motivated by specific circumstances that arise at random when a quantity of spirits becomes available in another Member State (usually Spain), or when variations in exchange rates resulting in cheaper prices favour this business activity. Moreover, the various wholesalers, retailers, buying groups and other respondents to the questionnaires were not aware of the extent of parallel trade, because their companies did not engage in such trade.

(c) Barriers to entry

(113) As stated above, the entry of new products into a highly branded and regulated market is a particularly difficult task in Greece, and as a result of the creation of GMG and for the reasons set out in preceding paragraphs on GMG's portfolio power and foreclosure effects, entry of new products is likely to become more difficult. Due to its negotiating power stemming from its leading brands, GMG could afford to negotiate with wholesalers and retailers lower listing fees and year-end bonuses and as a result, in order to maintain the profitability of the category, retailers could be compelled to impose higher listing fees or higher retail margins on new brands. GMG could also enter into exclusive distribution agreements or impose conditions on various on-trade outlets so as to make the launch and development of new brands both more difficult and more costly.

5. Conclusions on Greece

(114) On the basis of the above, GMG will account for [45-55 %] of the whiskey market. Coupled with its broad portfolio of whiskey brands, ranging across all the various quality and price sub-segments, that will confer on the combined entity considerable marketing advantages.

(115) Moreover, for the reasons set out above, GMG already has dominant positions in gin [80-90 %], brandy [70-80 %] and rum [75-85 %] and a very broad portfolio of brands, including the best-selling brands across all the spirits categories, with the exception of vodka, and will have [35-45 %] of overall spirits consumption in Greece.

(116) Furthermore, existing competitors do not have such a portfolio of brands that they would be able to constrain GMG's market power. In addition, the various trade channels are not able to exercise countervailing buyer power. Finally, barriers to entry are important, preventing thus new entrants from limiting the power of GMG.

(117) Therefore, for all these reasons, the merger will result in the creation of a dominant position in the Greek market for the supply of whiskey.

(118) Finally, through the portfolio effects set out above, the existing dominant positions in gin, rum and brandy will be reinforced.

B. SPAIN

1. General overview

(119) In terms of total spirit sales, Spain is the third biggest consumer in the Community, with 16,5 % of total Community sales by volume (1995).

(120) Table 7 shows the structure of spirits consumption in Spain.

>TABLE>

(121) Whiskey accounts for the greatest proportion of spirit sales in Spain with about 30 % market share by volume or value, about two thirds of which is accounted for by Scotch, which has been a growing market. Spain is the fourth largest consumer of whiskey in the world, after the United States, the United Kingdom and France.

(122) Brandy/cognac is the second most important spirit, with a share of about 17 %. Gin comes third with 17 % by volume and 10 % by value and has been declining. Vodka, although it accounts for a small part of the market with a share of about 3 %, has been growing over the last years.

2. Relevant product markets

(123) As explained above, the point of departure for product market analysis is the individual spirit type. In Spain, with regard to whiskey, given that the parties are not involved in the supply of Spanish whiskey, the key issue was whether separate markets should be defined for Scotch and Spanish whiskey. The parties claimed that there was no ground to differentiate between Scotch whisky and other types of whiskey, in particular, the locally produced whiskeys, namely DYC and Doble V (both produced by Allied Domecq). However, for the reasons set out in the following paragraphs, the Commission considers that Scotch whisky and Spanish whiskey are two distinct markets.

(124) Spanish whiskey does not comply with the distillation and ageing process requirements imposed on authentic Scotch whisky. Spanish whiskey is produced by combining a proportion (usually 30 %) of authentic Scotch with locally produced grain spirit. Consumers also regard the two products as different due to differences in image (for example certain Spanish brands have particular historical associations dating from the time when international brands were not readily available). Over the period 1985 to 1995, the growth in Scotch was twice that of Spanish whiskey. Furthermore, following the ending of preferential tax treatment for the local brands, their sales declined, whereas Scotch consumption generally has risen.

(125) Evidence provided by the parties on the retail prices per litre of whiskey in Spain in 1996 showed that the two Spanish whiskeys, DYC and Doble V, were at the lower end of the price range, and that only two brands of Scotch ([ . . . ]) were sold more cheaply (a difference of a few pesetas). In their response to the statement of objections the parties said that DYC had been seen in a Spanish supermarket priced considerably above two of the retailer's own-label Scotches, and a few other Scotches which were not listed on the original price series, as well as one (William Lawson) which, according to the price series, was usually considerably more expensive than either DYC or Doble V. However, when account was taken of which of the whiskies in the series were of any significance in Spain from the viewpoint of market share, it appeared that the next significant whisky in the price series ([ . . . ]) was more than [5-15 %] more expensive and the parties' biggest seller, [ . . . ], was about [25-35 %] more expensive again.

(126) The parties also adduced econometric evidence against the existence of a separate Scotch market in Spain. They provided quantitative estimates of own price elasticities to argue that Scotch whisky was not a separate market from Spanish whiskey. This was challenged by other parties. The validity of the results presented can be questioned, since the quantitative analysis suggests that some systematic influence is left out (to show up in significant residual serial correlation of unknown structure), which may contribute to the elasticity estimates. Furthermore, despite the importance of whiskey in the segment expenditure (which was also an explanatory variable), no account was taken when isolating the price elasticity of how segment expenditure varied with incremental changes in Scotch whisky price. There were therefore a number of insecure elements in the quantitative results. In the circumstances the Commission concluded that the evidence concerned could not be said either to confirm or refute the proposition which it was seeking to prove.

(127) In the light of the foregoing, the Commission considers that the relevant markets in Spain for the purposes of this assessment are Scotch and Spanish whiskey.

3. The market position of the parties

(128) Table 8 below shows the parties' shares at brand owner level by main categories of spirits in 1995.

>TABLE>

(129) Spain accounts for about [15-25 %] of Guinness's sales in the Community and about [10-20 %] of those of GrandMet. Importation and distribution are carried out by local wholly owned subsidiaries, Udie SA for Guinness, and Anglo Española Distribución for GrandMet.

(130) Table 9 shows the parties' market shares in Scotch by brand in volumes terms (1995 data).

>TABLE>

4. Aggregation

(131) The parties will have a strong position in Scotch. They have [40-50 %] combined market share by value (40-50 %] by volume) with an increment of approximately [10-20 %] by value ([10-20 %] by volume). J& B alone has over [25-35 %].

5. Competitors

(132) Overall, GMG will have more than twice the market share of its nearest competitor Allied Domecq, whose Ballantine's has a [15-25 %] market share. The third-placed competitor is Seagram's Passport, with a [< 10 %] share. Those shares, and in particular the share of Allied Domecq, have to be seen against the fact that the parties have the leading brand which on its own has a share significantly larger than Allied Domecq's total share.

6. Countervailing power of customers

(133) The parties have provided details of their estimates of the breakdown of sales by type of customer for Spain as a whole, and also in terms of their own sales. For Spain as a whole, they estimate that: direct sales to on-trade account for only [5-15 %] of the total; direct sales to off-trade retailers account for some [20-30 %], and that the balance of [60-70 %] is accounted for by sales to wholesalers including cash and carry outlets. It appears that the parties' own split of sales between these channels is [ . . . ].

(134) The parties claim that countervailing power exists because the wholesale and direct retail sectors are concentrated. They provided estimates showing that approximately [40-50 %] of their total sales are made to [ . . . ] large wholesalers/cash and carry customers and about another [15-25 %] to hypermarkets and large supermarket chains. Four hypermarket chains are said to account for approximately [55-65 %] of GrandMet's direct retail sales (equivalent figures were not available for Guinness.)

(135) It is to be noted however, that GrandMet sales through the main hypermarkets represent [5-15 %] of its total sales, suggesting that the countervailing power of those customers is limited. At the wholesale level, the majority of the parties' sales appear to be made to smaller wholesalers, who would not possess countervailing power.

(136) Moreover, it would be harder to exercise any countervailing power where the supplier has important leading brands, because any retailer or wholesaler would be taking substantially increased commercial risks in attempting to trade without being able to offer those products.

(137) In addition, one of the important elements for the exercise of countervailing power, namely a strong presence of own brand, is missing from the Spanish market, where such products account for 5 % of whiskey sales.

7. Barriers to entry

(138) There was no evidence to suggest that barriers to entry in Spain in the relevant markets were substantially different from those to be found in other European countries, and therefore the general comments in the introductory section of this Decision about barriers to entry apply. More specifically, the parties pointed to several whisky brands which had been introduced into the Spanish market since 1994. However, the majority of those brands were owned either by Guinness or by GrandMet, and the rest by existing competitors, rather than by genuine new entrants. All of them were special types of malt or premium whisky, which could be expected to capture only a small proportion of the market and which were mainly an extension of existing product lines rather than genuinely new products.

(139) It was noted that, apart from VAT 69 which has achieved approximately [< 10 %] since its recent reintroduction, in the 10 years to 1995, only two brands, Cutty Sark [< 10 %] and Passport [< 10 %], achieved a material market penetration. Passport is distributed and promoted by Seagram and has therefore benefited from that company's support, whereas Cutty Sark entered the market through independent distributorship. The relative scarcity of successful new entries over the period bears out the view that new entrants would face significant barriers.

8. Conclusions on Spain

(140) For the above reasons, the Commission considers that the concentration will lead to the creation of a dominant position in the market for Scotch whisky in Spain.

C. IRELAND

1. General overview

(141) In terms of spirit sales Ireland is the 12th largest consumer in the Community, accounting for approximately 2 % of sales.

(142) Table 10 below shows the structure of spirits consumption in Ireland in 1995.

>TABLE>

(143) Irish whiskey, with about one third of sales, is the most important spirit category, followed by vodka with about 20 %. Other important categories include brandy/cognac and gin.

(144) The situation in Ireland differs from other European countries in a number of ways. Irish whiskey accounts for 70 % of total whiskey sales whereas elsewhere Scotch whisky is predominant. All the leading brands of Irish whiskey are produced by Irish Distillers Ltd a subsidiary of Pernod Ricard. The Irish Distiller's Group distributes [85-95 %] of the Irish whiskey sold in Ireland and [45-55 %] of the total spirits.

(145) The distribution system for spirits is unusual, since two of the four major distributors are jointly owned by important international spirits producers, and only one is wholly owned. Normal practice in mature markets would be for all such subsidiaries to be wholly owned.

(146) The on-trade in Ireland is very dispersed with scarcely any chains of outlets. The largest chains of public houses have about 10 outlets. They are largely supplied by wholesalers. The Irish on-trade sector is important, and accounts for about 55 % of total spirits sales.

2. Product market

(147) The Commission considers that the markets for Scotch and Irish whiskeys can be differentiated on the basis of taste and consumption patterns which have been remarkably stable. This view is confirmed by the fact that over [> 95 %] of sales of Irish whiskey are made at prices higher than the price of Scotch whiskies, despite the fact that all Scotch whisky has to be imported. Three brands are listed by the parties in their reply to an Article 11 letter of 11 July 1997 as being cheaper than Scotch whiskies: [ . . . ]. Of those, [ . . . ] is not technically a whiskey but an Irish spirit, since its alcohol content is much lower than the standard required of whiskey. The other two brands are not listed individually in either IWSR or Canadean. Assuming that those brands account for all of the parties' sales of Irish whiskies classified as unidentified in Canadean, the maximum sales of both brands together amount to [ . . . ] cases or about [< 5 %] of 'unidentified` Irish whiskey. They do not therefore significantly affect the definition issue. A recent decision of the Irish Competition Authority (Decision No 285 - Irish Distillers Group/Cooley Distillery of 25 February 1994) also found a separate Irish whiskey market.

(148) In relation to products other than whiskey the conclusion reached in the introductory section of this Decision applies, namely that the relevant product market is no wider than the internationally recognised main spirit types.

3. Market profile

(149) In contrast to most other Member States, where the parties and their major competitors distribute their products through wholly owned subsidiaries, in Ireland many of the leading spirit manufacturers distribute their products through joint ventures with competing suppliers. In this situation it is necessary to consider the distribution level separately. Table 11 shows the main brands and their distributors in Ireland.

>TABLE>

4. Aggregation

(150) The aggregation arises from the fact that GMG would have substantial influence, albeit not de iure control, over the behaviour of three of the four major spirits distributors in Ireland. Gilbeys of Ireland (which currently distributes GrandMet spirits) would become a 100 % subsidiary of GMG. The merged entity would hold 49,6 % of Cantrell & Cochrane (the remainder of the shares being held by Allied Domecq), which wholly owns the distribution subsidiary Grants of Ireland, responsible for distributing the spirits products of Allied Domecq. Finally, Guinness holds 33 % of the shares of Edward Dillon, which principally distributes Guinness's spirits products.

(151) [The parties have argued that Guinness has no control or influence over Cantrell and Cochrane, and thus over Grants of Ireland and that their share holding is purely financial. They point out that the presence of three directors on the board of Cantrell and Cochrane is to protect Guinness' investment].

(152) However, the views of a shareholder with nearly 50 % of the equity could not realistically be ignored by the other shareholder, if Guinness, and in future GMG, chose to express them. Similarly, it is unlikely that Allied Domecq, though the majority shareholder, would be able to act in a way which Guinness might perceive as contrary to its interests without risking the dissolution of the joint venture. Consequently neither party can be said to be acting independently of the other and thus, as regards this joint venture, they should not be viewed as competitors. Moreover, it is unlikely that Guinness's current interest and that of GMG in the future would be purely financial given that Cantrell & Cochrane not only distributes spirits but is also involved in the wholesaling of alcoholic and non-alcoholic beverages and the production and distribution of cider and soft drinks, as are a number of Guinness subsidiaries.

(153) In relation to Edward Dillon, the parties point to the comfort letter recently issued by the Commission concerning this operation and the limited powers of the company's board. Two factors should be noted in this context, apart from its distribution of Bushmills Irish whiskey, an arrangement dating from the time when Irish Distillers was a shareholder in Edward Dillon, the company distributes (with very minor exceptions) only the spirit brands of its parents. Dillon would be unlikely to take on new products competing with its current range. Furthermore the situation obtaining at the time of the comfort letter will be changed by the implementation of the merger, whereby the Guinness and GrandMet interests in spirits distribution in Ireland would be combined.

>TABLE>

(154) Edward Dillon and Grants of Ireland are already associated through the significant shareholding Guinness has in each of them. The proposed operation would give rise to aggregations greater than 5 % in Scotch whisky (total [70-80 %]) and in brandy/cognac (total [90-100 %]). In addition, the three undertakings in which GMG would have an interest would also have very strong combined positions in vodka [60-70 %] and rum [85-95 %].

5. Competitors

(155) The Irish Distillers distribution operation is strong in Irish whiskey [85-95 %] and gin [70-80 %] but with the exception of vodka, where its Huzzar brand has a [20-30 %] share, it neither owns nor distributes other products with significant shares in other spirit categories. Therefore, although its parent company is dominant in the production of Irish whiskey and has control of most of the distribution of that product and of gin, it is not well placed to compete in other sectors.

(156) In the categories where distributors controlled or significantly influenced by GMG have very important market shares, Irish Distillers, the only competitor at this level of distribution, has few brands with any strength. In Scotch whisky, its Clan Campbell has only a [< 10 %] share against the [20-30 %] of Teacher's, and the [10-20 %] of Black and White. In cognac/brandy and rum, its Bisquit with [< 10 %] and Kiskadee with [< 5 %] are against Hennessy and Bacardi with [75-85 %] and [80-90 %] respectively. In vodka, its Huzzar brand has [20-30 %] compared with [60-70 %] for Smirnoff, but Huzzar has been losing volume faster than the category in general and much faster than Smirnoff itself.

(157) In Ireland, the multiple grocers have not developed own brands which might restrain the behaviour of powerful manufacturers and distributors in the off-trade. A number of retail chains do have agency brands, that is a proprietary product which they distribute exclusively, but those brands account for less than 5 % of the total off-trade. The recent acquisition by Tesco of retail grocery chains in Ireland is unlikely to have any substantial effect in the short to medium term and will have only a limited effect in the long term. This is because Tesco is maintaining the Irish identity of those stores, so that it cannot simply transfer its UK own brands to Ireland. Even if it were to do so, the Tesco brand is not established in Ireland and would, at least initially, carry little weight.

(158) The on-trade is more important in Ireland, accounting for about 55 % of spirits consumption, but the industry is very fragmented and is served by wholesalers. The countervailing power of those wholesalers would constrain the behaviour of strong distributors. However both Guinness and Cantrell & Cochrane have wholesaling operations including spirits which could be used either to marginalise other wholesalers or to bring pressure to bear on them, thereby reducing their effectiveness as a constraint.

(159) It seems, therefore, that there would not be sufficient constraining power from Irish Distillers in either the on-trade or the off-trade to counteract the strength of the distributors who would be controlled or significantly influenced by GMG in the Scotch whisky, and brandy/cognac markets.

6. Barriers to entry

(160) Ireland is not a densely populated country and has a large dispersed rural population. New entrants to the Irish market, who would almost certainly have a limited range of products, would not be able to establish their own distribution networks and would have to rely on using an existing distributor to provide a route to market for their product. As a result of the proposed operation, the number of major distribution channels will be reduced to two, one under the influence of GMG, the other under Irish Distillers.

7. Conclusions

(161) The proposed operation would have the effect of reducing to two the number of important independent spirits distributors in Ireland and would strengthen, through its impact at the distribution level, the parties' existing dominant positions in Scotch whisky and brandy/cognac.

D. BELGIUM/LUXEMBOURG

1. Product market definition

(a) All whiskey or Scotch

(162) On the basis of the market shares of the parties, the assessment of the operation would be similar irrespective of whether Scotch whisky or all whiskey is used. Therefore, the assessment below is made on the basis of all whiskey.

(b) Gin

(163) The question arose as to whether the definition of gin should include or exclude genever. The parties argued that genever should be included, pointing out that the Distillers case (10) had used a definition of 'juniper-based spirits`, relying on the classification of the market in Council Regulation (EEC) No 2658/87 (11). They also pointed to their own market research purporting to show that [30-40 %] of gin drinkers also drink genever and that [20-30 %] of genever drinkers also drink gin. They also referred to the ease of supply-side substitutability between genever and gin, and to the fact that gin and genever were normally positioned adjacent to one another on supermarket shelves.

(164) The fact that there is a 'juniper-based spirit` definition in Regulation (EEC) No 2658/87, or that it was used in an earlier case, does not imply that it is the appropriate definition to use in the present case. In relation to the market research, the fact that consumers will drink more than one type of spirit is not evidence that the spirits concerned are substitutes for one another, nor does adjacent positioning on supermarket shelves of itself say anything about whether the spirits concerned are in the same product market. Moreover, genever is not normally drunk with a mixer, and is often drunk as an accompaniment to beer or to coffee. In this respect it is totally different from the 'London` gins supplied by the parties.

(165) In addition, the Commission examined retail price data for gin and genever provided by the parties, and in particular it compared prices for leading brands of the two products. Its analysis showed that, over a recent two and a half year period, genever prices remained substantially below those of the gins (a difference of at least BEF [ . . . ] or approximately [5-15 %] of the bottle price). That supports the view that genever is not in the same product market as gin in Belgium/Luxembourg. Accordingly the relevant product market in Belgium/Luxembourg is considered to be the market for gin, excluding genever.

(c) Vodka

(166) There was no evidence to suggest that the appropriate product market was any narrower than the internationally recognised category for vodka. Accordingly the Commission considers the product market to be vodka.

2. General overview

(167) Belgium/Luxembourg accounted for about 2 % of total spirits sales by volume in the Community in 1995. Table 13 shows the structure of the market.

>TABLE>

(168) It will be noted that, in terms of its proportion of overall spirits sales, whiskey is the most important single spirit category and that some 90 % of whiskey sales are accounted for by Scotch, Gin and brandy are also important, but vodka and rum less so.

3. Position of the parties

(169) Both parties have wholly-owned subsidiaries operating in Belgium/Luxembourg for the distribution of their products there.

(170) Table 14 shows the parties' individual and combined market shares for all spirits and for certain spirit categories in Belgium/Luxembourg.

>TABLE>

4. Assessment

(a) Market shares following the merger

(171) As shown by Table 14 above, market shares of [40-50 %] would arise in an all-whiskey market, with an increment of [10-20 %]. The principal Guinness (UD) brands are Johnnie Walker Red Label ([10-20 %] share of all whiskey sales by volume at distribution level in 1995), Ainslies ([< 10 %]), Haig ([< 10 %]), and Black & White ([< 5 %]) as well as Cardhu, Dimple, 'Classic` and VAT 69. GrandMet's main brand is J& B ([5-15 %]). In an all-whiskey category, the next largest competitor has a market share of about [5-15 %]; GMG would therefore be at least five times as big as any competitor.

(172) In gin, according to the Canadean Report, 1996, Gordon's (Guinness) is the leading brand with a [35-45 %] market share, followed by Gilbey's (GrandMet) with [10-20 %] market share. By adding Tanqueray ([< 2 %]) and Bombay Sapphire ([< 2 %]), GMG would have a [50-60 %] combined market share in gin. The next largest competitor would be Booths ([< 5 %]) and the rest of the competitors would have minor shares well below [< 5 %], (Silver Top: [< 5 %]; Bosford [< 5 %]; Beefeater [< 2 %]; and Burnetts White Satin: [< 2 %]).

(173) In vodka, the post-merger combined market share would be over [55-65 %] with a significant increment arising from Guinness' distribution of third-party brands (Wyborowa and Zubrowka) added to the [40-50 %] market share of GrandMet's Smirnoff. The Commission recognises that if the distribution arrangements for Wyborowa and Zubrowka were to end, as the parties have suggested might happen, there would be no aggregation of market share arising from the operation. However, in the absence of any undertaking from the parties to end the agencies, the Commission must assume that current arrangements will continue. No other competitor has more than a [< 10 %] market share.

(174) The parties have argued that, notwithstanding those market shares, there is no case for finding dominance in any of the markets. Their three general arguments are that there is substantial buyer power, effective competition from own-label products, and substantial parallel trade. They observe that their margins on key products have fallen substantially in recent years.

(b) Whiskey

(175) The parties contended that powerful purchasers such as supermarkets were able to exert significant pressure on spirits suppliers. They contended that there was a high level of concentration in terms of their own sales. Figures provided by them showed that the top five customers of both Guinness and GrandMet accounted for over [35-45 %] of their total sales. The parties pointed out that, according to an IWSR report, over [45-55 %] of whiskey sales in Belgium were accounted for by large supermarkets.

(176) In terms of buyer power in branded products, the Commission doubts whether the power of large buyers such as supermarkets is likely to be sufficient to constrain the parties following the merger. In whisky, not only will the parties be the largest supplier with [40-50 %] combined market shares, compared with the nearest competitor's share of below [10-15 %], but the merger would also bring together the two leading brands, with a combined share from those two brands alone of [20-30 %]. With brands and shares of that importance relative to those of competing suppliers, supermarkets would be taking substantially increased commercial risks in attempting to trade without being able to offer those products.

(177) The parties also contended that own brands were a significant constraint in the direct retail sector, pointing out that own brands (including private label) accounted for some [25-35 %] to [45-55 %] of whiskey sales in key Belgian retail chains. In value terms the share of own label should be smaller, because it is generally cheaper than the branded product. The parties' figures also overstate the importance of own label in terms of its contribution to spirits sales in Belgium/Luxembourg. Canadean data suggests that own label might account for about [30-40 %] of all whiskey sales in Belgium.

(178) In the Commission's view, the fact that certain retailers have their own labels will not have more than a limited effect in constraining the prices for the parties' brands, in view of the market shares they possess. Although the possession of a range of own brands may strengthen the retailers' hands in negotiations it will not sufficiently offset the power which the parties know they possess through their control of the supply of leading brands.

(179) Own brands (including private label brands) are also handicapped in a number of ways. They are restricted to the outlets of the retailer or supplier in question and their scope for expansion is thus limited. The retailer cannot hope to replace altogether branded spirits by his own brands. Any attempt to do so would alienate those of his customers who want to buy specific brands, leading to the loss not merely of sales of the brands in question but possibly also of other products which might have been bought on the same occasion.

(c) Gin

(180) The parties make similar arguments in respect of gin. The parties observe that own label accounts for about [25-35 %] of sales in one leading Belgian retailer, suggesting similar shares in other comparable outlets. However the same basic arguments apply as for whiskey. The combined strength of own-brand gins appears to be significantly less than that of the parties' two brands. In addition the arguments set out above concerning the restrictions on the expansion and constraining ability of own brands of whiskey apply equally to gin.

(d) Vodka

(181) The parties estimate that the shares of own-label sales of vodka are in the order of [35-45 %] for two leading Belgian supermarket chains. Once again, however, that total is less than the share of Smirnoff. In addition, the arguments set out above concerning the restrictions on the expansion and constraining ability of own brands of whiskey apply equally to vodka.

5. Conclusions on Belgium/Luxembourg

(182) In the light of the above, the Commission considers that the merger will create dominance in the markets for whiskey, gin and vodka.

VI. UNDERTAKINGS SUBMITTED BY THE PARTIES

(183) In order to achieve clearance of the proposed concentration, the notifying parties proposed the following undertakings to be completed within the divestment period of 15 months or such extended period as may be approved by the Commission:

'(i) Parties will divest, within a period of 15 months from the date of the Decision or within such extended period as may be approved by the Commission (together "the divestment period") the rights in all EU/EEA/EFTA Member States, Poland, Hungary, the Czech Republic, Slovakia, Slovenia, Croatia, Bosnia, Serbia, and Macedonia ("the territory") to the Dewar's and "Ainslie's" Scotch whisky brands ("the brands") together with such confidential information, and related copyright and know-how specific to those brands as is necessary for their satisfactory production. In addition, the parties will undertake, in so far as whisky indispensable to the blending of the Brands can only be sourced from distilleries under their ownership, to continue to supply such whisky to the purchaser, if requested, on reasonable arms' length commercial terms. In the event of any dispute concerning the supply of such whisky, the parties shall refer the matter to be resolved by arbitration under the Rules of the London Court of International Arbitration applying the law of England and Wales.

(ii) Parties will bring to an end by 31 December 1998 at the latest the brand agency distribution arrangements for Belgium/Luxembourg for "Wyborowa" vodka currently held by Guinness.

(iii) Parties will within the divestment period:

either:

(a) [ . . . ].

or

(b) [ . . . ]

[Dispose of certain interests in Ireland in order to ensure continued competition in the distribution of spirits after the formation of GMG brands which otherwise would have effectively reduced the number of distributors in Ireland from four to two.]

(iv) The parties will, within the divestment period, entrust, for a period not less than nine years, the distribution of Gilbey's gin in Belgium to an independent third-party distributor on reasonable arms' length commercial terms. If such a distributor shall not be appointed within the divestment period, the parties will appoint an independent trustee (who shall be the same as that appointed in relation to the brands and the shareholding) who shall be mandated to appoint such a distributor on the best available terms and conditions and in any event by 30 June 1999.

(v) Parties will, within the divestment period, discontinue the brand agency distribution arrangement for Bacardi rum in Greece currently held by Guinness.

Mechanisms for divestment

(vi) The parties will immediately following the date of the Decision, and in any event, not later than four weeks, propose to the Commission for its approval the names of two institutions whom they consider appropriate to be appointed as trustees to act on GMG Brands' behalf, following its formation, in overseeing the divestment of the brands and [ . . . ]. The Commission shall not, without good cause, withhold its approval of any trustee proposed by the parties.

(vii) The parties will, as soon as possible after receiving the Commission's approval of the proposed trustee, appoint such trustee ("the trustee") to act from the date of formation of GMG Brands on its behalf in overseeing the divestment of the brands and [ . . . ] shareholding for full and fair market value during the divestment period. Such appointment shall be made, subject as more fully described below, on an irrevocable basis save only in the circumstances that Grand Metropolitan and/or Guinness should announce that the proposed merger has been abandoned in which case such appointment shall be deemed automatically revoked.

(viii) Pending divestment of the brands, the parties shall act, and shall instruct the trustee to act, so as to ensure the continued viability and market value of the brands and their rapid and effective divestiture from the rest of GMG Brands' activities, as more fully described below.

(ix) The parties undertake that they will give the trustee a mandate to find on behalf of GMG Brands a satisfactory purchaser or purchasers for the brands and [ . . . ] (subject to (xi)(b) below), it being understood that such purchaser or purchasers shall be a viable existing or prospective competitor independent of, and unconnected to Guinness or Grand Metropolitan and possessing the financial resources and proven expertise enabling it to maintain and develop the divested brands and/or shareholding as an active competitive force in competition to the parties' spirits business on the various markets concerned ("the purchaser standards").

(x) [ . . . ]. Should any of the divestitures not be effected within the divestment period, such divestitures must in any event be the subject of a binding agreement to sell by the end date agreed with the Commission.

(xi) The parties shall ensure that the mandate of the Trustee includes the following rights and obligations:

(a) to provide to the Commission written reports (with a copy to GMG Brands) on a monthly basis, (or, at the option of the Commission at such other reasonable time in the event of significant developments in the divestment process), concerning relevant developments in its negotiations with third parties interested in purchasing the brands [ . . . ], including the time-frame within which an agreement with interested third parties would be implemented, and, in particular, sufficient information to enable the Commission to assess whether each bidder satisfies the purchaser standards;

(b) to continue negotiations with an interested third party only if the Commission does not, within two weeks of receipt of the Trustee's report, formally indicate that it does not approve of the third party specifying its reasons;

(c) to receive remuneration from GMG Brands on a basis which will provide incentives for a prompt divestiture.

(xii) The parties undertake that GMG Brands shall provide the trustee with all reasonable assistance required in carrying out the mandate.

(xiii) If there is more than one prospective purchaser unopposed by the Commission for all or any of the brands [ . . . ], GMG Brands shall be free to select the offer of its choice.`

VII. ASSESSMENT OF THE UNDERTAKINGS

A. WHISKEY

(184) The parties have offered to divest two brands at European level. Dewar's is an important international brand and the parties' third most important one in worldwide sales terms, after Johnnie Walker and J& B. Ainslie's currently sells only in Belgium/Luxembourg but is the third most important brand in that territory.

(185) The effect of divestiture on the various whiskey markets in which the Commission has identified problems is shown in Table 15.

>TABLE>

(186) Together the sales of these brands in the Community in 1995 were [ . . . ] million nine-litre cases, or about [< 5 %] of Community consumption by volume. For comparison, sales of Famous Grouse, the leading Scotch whisky not sold by a major spirits producer, were [ . . . ] million cases in 1995.

B. VODKA

(187) The termination of the Wyborowa agency in Belgium/Luxembourg will reduce the parties' market share from [60-70 %] to [45-55 %], the remaining market share being accounted for in its entirety by GrandMet's Smirnoff, and remove the overlap created by the merger.

C. GIN

(188) The appointment of an independent third-party distributor for gin in Belgium/Luxembourg would reduce the parties' market share for gin from [50-60 %] to [35-45 %].

D. RUM

(189) Bacardi accounts for [75-85 %] of rum sales in Greece. The termination of the Bacardi agency agreement in Greece would, together with the divestment in whisky, reduce significantly the number of product categories in which the parties would have significant shares. Although GMG will continue to hold high market shares in some categories, notably brandy, its ability to exercise portfolio power will be satisfactorily restrained, since whiskey and rum are respectively the first and third largest spirit categories in Greece.

E. IRELAND

(190) The parties propose the two following alternatives. Either:

(a) [ . . . ].

(b) [ . . . ].

[Which will involve the disposal of certain interests in Ireland so as to ensure continued competition in the distribution of spirits.]

(191) Either of these options would satisfactorily address the competition issues identified in the Commission's analysis of the Irish market by substantially reducing the parties' influence over the distribution of their own products and those of others in that territory. [ . . . ]. [Either alternative will ensure that the influence of GMG over the distribution of spirits in Ireland is significantly reduced.]

F. TERMS AND CONDITIONS

(192) The terms and conditions for the various divestitures, which are consistent with the practice in previous such cases, are considered adequate.

VIII. CONCLUSION

(193) In the light of the foregoing, the Commission has concluded that the concentration notified by Guinness plc and Grand Metropolitan plc on 16 May 1997, relating to the merger of all their business activities, should be declared compatible with the common market and with the functioning of the EEA Agreement, subject to the condition of full compliance with the commitments made by the parties in their undertaking to the Commission as set out in section VI above,

HAS ADOPTED THIS DECISION:

Article 1

The concentration notified by Guinness plc and Grand Metropolitan plc, relating to the merger of all their business activities, is declared compatible with the common market and with the functioning of the EEA Agreement, subject to the condition of full compliance with the commitments made by the parties in their undertaking to the Commission as set out in section VI of this Decision.

Article 2

This Decision is addressed to:

Guinness plc

39 Portman Square

London W1H 0EE

United Kingdom

and

Grand Metropolitan plc

8 Henrietta Place

London W1M 9AG

United Kingdom.

Done at Brussels, 15 October 1997.

For the Commission

Karel VAN MIERT

Member of the Commission

(1) OJ L 395, 30. 12. 1989, p. 1; corrected version OJ L 257, 21. 9. 1990, p. 13.

(2) OJ C 329, 27. 10. 1998.

(3) The term 'whiskey` is used throughout to describe all whiskey types (Scotch, Irish, American, Spanish) both collectively and, except for Scotch, individually. For Scotch, the traditional spelling - without an 'e` is used.

(4) The terms 'on-trade` and 'off-trade` are commonly employed in the industry to distinguish the two retail channels: 'on-trade` that is, consumption on the premises where the product is purchased, comprises supplies to and from bars, restaurants, cafés and hotels (sometimes referred to as the 'horeca` (hotel/restaurant/café) sector); 'off-trade`, that is consumption away from the place of purchase: covers sales through general and specialised shops, supermarkets, retail grocers, and so forth.

(5) This version of the Decision has been edited to ensure that confidential information is not disclosed.

(6) Commission Decision 97/540/EC, Case No IV/M.794 - Coca-Cola Enterprises/Amalgamated Beverages GB (OJ L 218, 9. 8. 1997, p. 15), Case No IV/M.833 - Coca-Cola Company/Carlsberg AS.

(7) Parties' response to the 'Statement of objections`, p. 24.

(8) In 1995, Scotch whisky accounted for 95 % of an all-whiskey market, that is 2 541 thousand cases of a total of 2 654 thousand cases (Source IWSR 1995).

(9) The parties' estimates point to a figure of some [5-15 %] of their total sales between 1995 and 1997 (see 'Response to the statement of objections`).

(10) Commission Decision 85/562/EEC, Case No IV/30.570 - The Distillers Company, (OJ L 369, 31. 12. 1985, p. 19).

(11) OJ L 256, 7. 9. 1987, p. 1.

Top