JUDGMENT of Mr. Justice Cooke delivered the 19th day of March, 2009.
1. Introduction
1.1 By a share purchase agreement dated 12th March, 2008, the appellant agreed to acquire by purchase from Reox Finance Ltd. the entire issued capital of two wholly owned subsidiaries of the latter namely, Breeo Foods Ltd. and Breeo Brands Ltd.
1.2 The appellant is a wholly owned member of Kerry Group plc., a quoted public company with a market capitalisation of €3.5 billion: it is a leading producer of foods and food ingredients in the State and the supplier of such goods in some 140 countries around the world. (Kerry Group plc and the appellant are referred to in this judgment as “Kerry”.)
1.3 Reox Finance Ltd. is a wholly owned subsidiary of Reox Holdings plc (“Reox”) an Irish unlisted plc whose shares are held as to 26% by Dairygold Co-Operative Society Ltd. (“Dairygold”) and as to the balance by some 7,500 shareholders who are mainly members of Dairygold.
1.4 Reox is an international food company operating in the State, the United Kingdom of Great Britain and Northern Ireland, the Netherlands and the United States of America and which exports to European countries and the Middle East.
1.5 As part of a restructuring of Dairygold, the two above Breeo companies which previously comprised the consumer foods division of Dairygold were transferred into Reox. The two Breeo companies (which will be referred to hereinafter as simply “Breeo”) had a turnover of €200 million of which some €166 million was generated in the State.
1.6 Breeo has outsourced most of its manufacturing operations to third parties and at the date of the acquisition agreement was involved in manufacturing or processing at only two locations in the State, namely, the manufacture of dairy spreads at Mitchelstown, County Cork and the slicing of cooked hams at Tallaght in County Dublin.
1.7 In practical terms, the effect of the proposed acquisition would be to transfer from Dairygold to Kerry the consumer foods division of Reox together with a number of properties including the plants at Mitchelstown and Tallaght and its business and assets, including intellectual property assets and in particular, some 225 trade marks comprising a number of well known brands including Dairygold, Galtee, Shaws, Roscrea, Mitchelstown and Calvita. Kerry is the proprietor of other brands for similar food products including, Denny, Ballyfree, Clover, LowLow, Easisingles.
1.8 Because of the size and value of the operations of the Kerry Group and Dairygold respectively and because the transfer of those assets and properties constituted a “merger or acquisition” for the purposes of s. 16 of the Competition Act 2002 (“the Act”) the appellant and Reox (hereinafter referred to jointly as “the notifying parties”) notified the transaction to the Competition Authority in accordance with s. 18(1) of the Act.
2. Notification Procedure and Appeal
2.1 The proposed acquisition was notified to the Authority on 20th March, 2008 and the notification was accompanied by 1) an economic report by Compecon Ltd. and 2) econometrics reports furnished by Dr. Vincent Hogan. The steps taken by the Authority in the course of its first phase examination as provided for in ss. 20 and 21 of the Act can be summarised as follows.
a) Notice of the notification was published pursuant to s. 20(1)(a) of the Act.
b) A questionnaire was sent to 9 retailers to which 8 responded and a questionnaire was sent to 23 competitors of the notifying parties to which 12 responded;
c) Requests for further information were issued to Kerry and Breeo;
d) On 29th May, 2008, the Authority made a determination pursuant to s. 21(2) of the Act to carry out a “full investigation” pursuant to s. 22 of the Act.
2.2 The steps taken during the full investigation phase can be summarised as follows:-a) A second request for further information was issued to Kerry on 5th June, 2008;
b) Further questionnaires were sent respectively to 6 retailers and 11 competitors;
c) The Authority meet with the parties and their experts to discuss the econometrics expertise on 3rd July, 2008;
d) Further econometrics reports from Dr. Hogan on behalf of the notifying party and Dr. Walsh on behalf of the Authority were exchanged and submitted;
e) A third questionnaire was issued to 5 retailers relating to the cheese market;
f) On 15th July, 2008, a meeting with the parties and their experts took place;
g) On 25th July, 2008, the Authority issued an “Assessment” accompanied by a list of the documents on its case file to the parties.
h) On 31st July, 2008, a fourth questionnaire was issued to 5 retailers directed at the relationship between private label and branded goods in 4 product categories;
i) On 7th August, 2008, an oral hearing was held at which submissions by the parties and their experts were made to the Authority in response to the Assessment;
j) On 15th August, 2008, the notifying parties submitted further written submissions to the Authority;
k) The Authority issued questionnaires and conducted interviews with 4 competitors in the cheese sector between 19th and 21st August, 2008;
l) On 26th August, 2008, the notifying parties made further submissions to the Authority.
2.3 On 28th August, 2008, the Authority issued its determination on the notification at the conclusion of the full investigation pursuant to s. 22(3)(b) of the Act, namely, that the proposed acquisition should not be put into effect on the ground that the result of the acquisition would be to substantially lessen competition in the markets for the goods concerned. (The “Determination”.)
2.4 By an initiating notice of motion dated 26th September, 2008, the appellant commenced an appeal pursuant to s. 24 of the Act and Order 63B rule 25 of the Rules of the Superior Courts against the Determination.
2.5 In a schedule of “Grounds of Appeal against Determination of the Competition Authority” attached to the notice of motion, the appellant set forth some 24 sets of grounds alleging errors in the Authority’s determination.
2.6 On the hearing of the appellant’s motion the Court directed pursuant to O. 63B, r. 6 that written submissions be exchanged between the parties and lodged with a view to defining precisely the issues between the parties on the appeal.
2.7 Written submissions were duly lodged and exchanged by the appellant and respondent.
3. The Regulatory Framework
The Competition Act 2002
3.1 The jurisdiction of the Authority and of this Court in relation to mergers and acquisitions is contained in Part III comprising ss. 16 to 28 of the Act. So far as is relevant to the issues raised upon this appeal the relevant provisions can be summarised as follows.
3.2 A transaction which comes within the definition of the term “merger or acquisition” in s. 16 of the Act and which exceeds the thresholds defined in s. 18, is required to be notified to the Authority within one month from the conclusion of the agreement in question.
3.3 Such a merger or acquisition may not then be put into effect until one of the events specified in s. 19(1) of the Act has occurred. In very broad terms, this provision means that a notified merger cannot be implemented unless and until the Competition Authority has either determined that it may be put into effect or has failed to make any determination within four months of the date of notification without commencing a full investigation pursuant to s. 22.
3.4 Section 20 of the Act provides that when a notification of a merger or acquisition is received, the Authority must publish a notification of it within seven days and then consider all submissions made by the undertakings involved and by other individuals and undertakings.
3.5 The Authority’s examination and investigation of a notified transaction then takes place in two phases. In the first phase the Authority has, in effect, one month in accordance with s. 21 to determine whether the transaction may be put into effect without full investigation because, in its opinion, it will not result in a substantial lessening of competition in the market concerned. Alternatively, within that period the Authority may determine that it intends to carry out a “full investigation” under s. 22 in which case it must notify the notifying parties accordingly.
3.6 In the latter case there then takes place a “full investigation” under s. 22. On completion of a “full investigation” the Authority must make one or other of the following determinations as appropriate namely “that the merger or acquisition (a) may be put into effect; (b) may not be put into effect; or (c) may be put into effect subject to conditions specified by it being complied with”. In making this determination the test upon which the authorisation or prohibition of the proposed transaction turns is whether or not the result of the merger or acquisition will be “to substantially lessen competition in markets for goods or services in the State”(s.22(3))
3.7 Under s. 24 of the Act an appeal may be made to this Court against a determination made by the Authority under s. 22(3) limited to the determinations made under paragraphs (b) and (c) of the subsection, namely, determinations which either prohibit the implementation of the transaction or permit it subject to conditions which the Authority has specified. It is to be noted, therefore, that there can be no appeal against a determination which authorises the putting into effect of the proposed merger or acquisition and under s. 24(3) the appeal can only be made by one or more of the undertakings to the notification of the transaction to the Authority. A third party cannot therefore challenge a positive determination.
3.8 Upon the hearing of an appeal under s. 24 this Court has power under s. 5(7) to (a) annul the determination; (b) confirm the determination; or (c) confirm the determination subject to such modifications as it determines and specifies in its judgment.
3.9 An appeal to the Supreme Court lies against the decision of this Court but such further appeal is confined to questions of law. (The provisions of s. 24 of the Act are considered in greater detail in the later section of this judgment dealing with the nature and scope of the present appeal.)
The Merger Guidelines
3.10 As a guide to the way in which it examines and analyses a notified merger with a view to forming its opinion as to whether or not the transaction will substantially lessen competition in the relevant market, the Authority has published a “Notice in respect of Guidelines for Merger Analysis” (“the Merger Guidelines”). While it might be said that such a notice is not part of a “regulatory framework” in the sense of being a measure with binding legal force, the Merger Guidelines correspond to similar publications issued by many equivalent national authorities charged with regulation of mergers and acquisitions and to the “Horizontal Merger Guidelines” published by the European Commission on its approach to the analysis of horizontal effects of mergers under Council Regulation (EC) No. 139/2004 of 20th January, 2004, on the control of concentrations between undertakings (2004) OJ L 24/1 (“the Merger Regulation”). (The horizontal merger guidelines are published in (2004) OJ C 31/03.)
3.11 The Authority’s Merger Guidelines set out in detail the steps which it takes and the criteria and tests which it uses in analysing the apparent effects of the proposed merger with a view to assessing whether there will be a substantial lessening of competition if the merger is put into effect. In particular, the assessment as to whether a substantial lessening of competition will occur is interpreted in terms of consumer welfare. The effect of the merger on consumer welfare is measured by whether or not the price of the products concerned in the market will rise. The guidelines are concerned primarily but not exclusively with horizontal mergers, that is to say those between undertakings that produce substitute or competing products as is the case in the present proceeding.
3.12 The detailed approach to this analysis as explained in the Merger Guidelines can be very broadly summarised as comprising the following steps:-
Step 1:The relevant product and geographic markets are identified and determined in order to set the framework in which the analysis of competition will then be made. The purpose of defining the market by reference to its product and its dimension is to identify those competitors who are capable of exercising a constraint on the merged entity and thereby prevent it acting independently of competitive pressure. The relevant product market is defined by reference to the products produced or offered by the merging parties and comprises all those products which are regarded as substitutable for the products of the merging parties by consumers having regard, particularly, to the characteristics of the products, their prices and their intended use. To identify which products come within this market the “SSNIP” test is used. This test seeks to identify the smallest market in which a hypothetical monopolist of the product could impose a “small but significant and non-transitory increase in price” above the prevailing level and maintain it profitably. If in response to such a price increase there would be a reduction in sales of that product large enough to render it unprofitable for the monopolist to impose the increase, then the products that are the next best substitute for the product of the merging firms and which attract the lost sales can be included in the relevant product market. In applying the test the Authority supposes a price increase of 5 to 10% above prevailing levels which will last for one year.
The geographic market, on the other hand, for each relevant product is delineated as a region where a hypothetical monopolist of the product could profitably impose a small but significant and non-transitory increase in price without suffering such a reduction in sales as to render the increase unprofitable as a result of a sufficient number of consumers switching to suppliers located outside the region.
While the measurement of market definition is primarily conducted by reference to the above demand side substitutability, supply side substitutability may also be examined, that is to say whether, in response to the merger, products not currently supplied in the relevant market could be supplied at short notice in response to the hypothetical price increase.
Step 2: When the relevant product market has been defined, the effect of the proposed merger on market structure is then measured. For horizontal mergers this involves notably calculating the post – merger “HHI”. This refers to the Herfindahl – Hirschmann index of concentration. The market shares held by each of the undertakings are squared and added together both pre-merger and post-merger. The “delta” is the difference between the before and after change in the index calculation resulting from the merger. The level of the post-merger HHI gives an indication of the degree of market concentration. The level and delta are then used to form thresholds of market concentration which for purposes of clarity are divided into zones A, B, C. These thresholds are intended primarily to give an initial guidance and a rule of thumb indicator as to the likelihood of a need for a closer examination of competitive effects brought about by the merger. Depending on the post-merger change or delta, a merger in zone A is less likely to have anti-competitive effects. One in zone B may raise such concerns particularly if the delta goes above 100 points. Where the calculation brings a proposed merger into zone C with a delta above 100 points, it is an indication of a highly concentrated market and therefore one which will usually give rise to competitive concerns.
Step 3:An assessment is then made as to whether the proposed merger has an effect on the level of rivalry among existing competitors in the market and for this purpose both unilateral and coordinated effects are examined. Unilateral effects arise where, as a result of the merger, the merged undertaking finds it profitable to raise prices irrespective of the reactions of its competitors or customers. The market structure is first examined and the effect of the merger on the behaviour of the merging entity is looked at. If a price increase by one of the merging undertakings caused a loss of sales which would be gained by the other party to the merger then the sales in question would not be lost post-merger and this ability to internalise sales that would be lost would make it profitable for the merged entity to increase price.
Next, the possible reactions of existing competitors are examined. Are they in a position to win sales from the merged entity if prices are raised?
The reactions of customers are also analysed to see if they have the ability to switch to alternative suppliers in response to a price increase. Countervailing buyer power is also examined at this stage. Do buyers have alternative sources of supply or the power to credibly threaten to set up alternative supply arrangements?
Coordinated effects arise where a merger facilitates competitors in tacitly or explicitly colluding to raise price. Is the market characterised by factors conducive to competitors raising prices and foregoing profitable sales in the expectation that others will do so as well?
Step 4:Having considered the effect of the merger on the position of existing competitors and their reactions, the Authority then examines whether the market power of the merged entity might be constrained by new suppliers entering the market in response to the price increase. To have such a constraining effect the appearance of new entrants would require to be “timely, likely and sufficient”.
Step 5:Finally, efficiencies are considered. If a merger will cause a substantial lessening of competition, the possibility remains that this effect could be compensated for by improvements in efficiencies arising out of the merger. The Authority uses essentially a “net price test” by considering whether the price paid by consumers will rise or fall as a result of the merger. The burden of proof in identifying and demonstrating the existence and certain implementation of such efficiencies, rests with the merging parties and the Authority requires that the efficiencies be shown to be “(a) directly achieved by the merger; (b) not achievable by other less restrictive means; and (c) they will be achieved within a reasonable timeframe and with sufficient likelihood”. Given the incentives for the parties to put efficiencies in the most optimistic light, the Authority requires that the efficiency gains claimed are clearly “verifiable, quantifiable and timely”.
4. The Determination
General Background
4.1 The Authority’s determination setting out its detailed analysis of the notified acquisition comprises 9 sections extending over 150 pages. Two introductory sections summarise the Authority’s own procedure upon the notification and give background information on the parties, their operations, products and brands; and on the structure of the markets involved with the relevant market shares. They then outline some of the main concepts and tests employed in the analysis which follows.
4.2 The Determination identifies six categories of products involved in the transaction and the markets for these are examined in sections Three to Eight of the Determination as follows:-
Section Three: Breakfast meats – sausages
Section Four: Breakfast meats – puddings
Section Five: Breakfast meats – rashers
Section Six: Cooked meats
Section Seven: Spreads
Section Eight: Cheese
A final Section Nine examines the efficiency gains claimed by the notifying parties. Broadly speaking, the analysis of competition in respect of each of the above 6 markets follows the structure outlined for such analysis by the Authority in the Merger Guidelines. (In this judgment references to particular paragraphs in the Determination are designated thus: [DET. 2.11 etc.]
4.3 Some of the background information given in Section Two in relation to the parties, their operations and the structure of the markets in question may be usefully summarised as follows.
4.4 Both Kerry and Breeo supply and distribute a wide range of consumer foods in the above market categories which they supply and distribute to wholesalers and retailers. Kerry is not involved in any primary processing but purchases meats, natural cheese and oils from a variety of suppliers for use in pre-packed products. Breeo was formerly involved in manufacture and processing of its products but lately has outsourced all processing and production so that it now (see para. 1.6 above) only manufactures spreads and is involved in processing only insofar as it slices meats bought in in cooked log form [DET table 2.1].
4.5 Both Kerry and Breeo distribute their products in similar manner, each by its own distribution network in vehicles which distribute to wholesalers and retailers. In the case of some retailers the products are delivered to a central distribution location.
Brands and Private Label
4.6 At paras. 2.17 – 2.29 of the Determination the Authority describes the respective roles and significance of brands and private label products in the markets which were the subject of the investigation. Because Breeo had little manufacturing capability, the Authority’s Assessment of 25th July, 2008, expressed the view that the price of €160 million paid by Kerry for the acquisition was in large part a reflection of the value of the brands to be acquired from Breeo. This was disputed by the notifying parties but, having consulted an independent accounting adviser to examine the share purchase agreement and the documents relating to the transferred assets, the Authority considered that the evidence was consistent with its view that the price reflected in large part the value of the Breeo brands [DET 2.21].
4.7 In the markets which are the subject of the present appeal the brands held by Kerry and Breeo are as follows:
| Products |
Kerry |
Breeo |
| Cheese |
Low-Low |
Mitchelstown |
 |
Charleville |
Xtreme |
|
Easi singles |
Three Counties |
|
Golden Vale |
Calvita |
|
Cheese strings |
Galtee |
|
Attack a Snak |
|
| Breakfast Meats |
Denny |
Galtee |
|
Ballyfree |
Shaws |
|
Clover |
Roscrea |
| Cooked Meats |
Denny |
Galtee |
|
Ballyfree |
Shaws |
|
|
Roscrea |
Brands, especially strong or premiums brands, typically command a higher price for a number of reasons including a “higher perceived quality” or innovation and advertising [DET 2.22].
4.8 At Tables 2.3 and 2.4 of the Determination the Authority sets out the premium price commanded by the parties’ branded products in the markets concerned by comparing the price for the branded products with the average price for such products in the market. This is done by volume and value. Thus, in the market for rashers, for example, the Authority asserts that the Kerry and Breeo brands are priced respectively 45% and 49% above the average price of rashers. The table also demonstrates however that Kerry and Breeo brands do not command a premium in all markets. Kerry’s are not above average in non butter products while Breeo’s are not above average in butter.
Private Label
4.9 It is noted that Kerry and Breeo not only market their own branded products but also supply “private label products” for retailers. The term “private label” is used in contrast to the term “branded products” to cover both products sold by retailers under their own name or label and usually, but not always, at a price which represents a discount from the price of the branded product. The term covers also, however, corresponding products sold by retailers designated “discounters” (being principally the Lidl and Aldi chains,) where the products concerned may be branded products in their country of origin while unfamiliar to Irish consumers. These are brought in and sold at equivalent discounts rather than packaged and sold under the Lidl or Aldi names.
4.10 At para. 2.34 of the Determination the Authority recognises the importance of retailers in these markets because a small number of them account for a large percentage of grocery sales in this State. Table 2.5 of the Determination sets out bands of market share for all the main groups of retailers and if it is assumed that each retailer’s share is at the mid point of the range given, then the leading six retailers account for 65% of all grocery products sold in the State. The four leading retailer groups or chains are Tesco, Dunnes, Musgraves and Superquinn. The parties submitted to the Authority that in 2006 those four accounted for total grocery sales of 76.9% and in 2007 for 78.8%. Irrespective of which source data is used, the Authority accepts that a small number of retailers account for a very large proportion of all grocery sales [DET. 2.34].
4.11 The Determination then describes how in these markets retailers can exercise considerable countervailing power in different forms. They can threaten to de-list one or more brands offered by a supplier; they can “derange” the products of a particular supplier, that is, stock only a limited selection of packet formats or sizes within a range offered by a supplier. They can move branded products to less favourable positions on display shelving. All of these tactics can be employed in order to obtain lower prices from a supplier with a view to improving or maintaining profitability. They can do so because of the relatively high volumes which their purchases from branded suppliers represent to the supplier and because they have alternative sources of supply available to them either in the form of other branded products or their own brands or “private labels”. Whether they can credibly threaten to employ any of these tactics may depend on the importance of the brands in question to consumers, - the retailers’ customers, - and whether those brands are “must have” brands. (In emphasising the importance of this buyer power the appellants pointed out that the sale of Kerry and Breeo brands represented only 2% of the turnover of the main multiple chains but constituted approximately 80% of total sales by Kerry and Breeo.)
4.12 As already mentioned, the main part of the Determination is concerned with the effects of the merger in relation to four broad categories of products: breakfast meats; cooked meats; spreads and cheese. Within these categories a number of different or potentially different product markets are identified and analysed as follows:-
| Breakfast meats: |
sausages, puddings and rashers |
| Cooked meats: |
poultry and non poultry cooked meats |
| Spreads: |
butters and spreads |
| Cheese: |
natural cheese – processed cheese |
(Within this last category of processed cheese a further segment is identified comprising processed cheese slices. See section 7 below.)
Analysis of Competitive Effects
4.13 The analysis of the competitive effects of the merger with a view to determining whether competition will be lessened is made in respect of each of these categories of products and is structured and explained in broadly the same way in each case as follows:-
(1) The relevant product market is examined by reference to the products offered and their characteristics and in the light of the submissions of the parties. Where there is doubt or ambiguity as to the composition of the product market the Authority investigates the market taking account of the responses of retailers and competitors surveyed and; the contents of internal documents obtained from the parties. If appropriate, supply side substitution is looked at and a conclusion on product market is then given.
(2) The geographic market is then examined and a conclusion given.
(3) The analysis of the apparent effects of the merger on competition is conducted by measuring market concentration:-
(a) The market structure is examined and the HHI calculation is made in order to categorise the merger in the zones of market concentration;
(b) The closeness of competition between the products of the merging entities themselves and with other products in the relevant market is examined taking account of econometric evidence, the responses of retailers and competitors and a conclusion is then expressed.
(4) The possibility of counterbalancing factors constraining any identified anti competitive effects is then examined under the headings of new entrants and expansion by competitors, imports and the exercise of countervailing retailer buyer power.
(5) A conclusion is then given as to the Authority’s assessment of whether substantial lessening of competition in the relevant market will occur as a result of the merger.
4.14 Thus, for example, in the case of breakfast meats, the Compecon Report submitted on behalf of the notifying parties maintained that there were three separate product markets namely sausages, puddings and rashers and the Authority agreed. There was some argument as to whether further distinctions fell to be made between cooked and uncooked meat products sold in loose and pre-packed forms but the Authority concluded that in each of the three product markets the “over the counter” (OTC) and pre-packed meats were part of the same markets.
4.15 As has already been indicated, in its Determination the Authority came to the view that no substantial lessening of competition would result in the markets for sausages, puddings, and spreads. In the case of sausages, the Authority took the view that it was unnecessary to consider whether there were separate product markets for OTC/deli sausages or pre-packed sausages as its conclusions as to the competitive effects would be unaffected by a narrowing of the product market. Similarly, in the case of spreads, arguments arose as to whether there was a single product market or two separate markets for butter and non-butter products. Again, the Authority did not consider it necessary to resolve that issue as its overall conclusion on the effect of the merger would be unaffected by a determination as to the width of the product market.
The Uncontested Markets
4.16 Although the present appeal concerns only the conclusions reached by the Authority in relation to the markets for rashers, cooked meats and cheese, it may be useful by way of comparison with those conclusions and in the light of the issues dealt with later in this judgment, to mention briefly the basis upon which it reached its different conclusions in relation to the markets in spreads, sausages and puddings. In the segment of the spreads market comprising butter, it was notable that there was one clear market leader, the Irish Dairy Board’s brand “Kerrygold” with shares by value consistently in excess of 60%. Breeo’s “Dairygold” brand hovered at 10% while Kerry had no brand with a market share above 1%. (Table 7.1). In the non butter segment of spreads, Breeo was the market leader with a share of 31% (by value) of its brands followed by Unilever (Flora) at 26% and Kerry (Low-Low) with 23% (Table 7.2). These three producers accounted for 81% of that market and their market shares tended to be relatively stable.
4.17 While the HHI market concentration calculation placed the post-merger estimate for the spreads market in zone C and the delta at 1469, the Authority nevertheless came to a conclusion upon closer analysis of the effects on competition, that the merger would not bring about a substantial lessening of competition in the spreads markets in view of the following factors:-
• Breeo’s Dairygold and Kerry’s Low-Low were not close competitors in the spreads market because one was perceived by retailers as a taste-based spread while the latter was perceived as a health-based spread.
• Unilever’s Flora brand was No. 2 in the spreads market with a share of 26.5% and competed with Kerry’s Low-Low.
• Flora was considered a “must have” brand in the spreads market.
• Breeo’s closest competitor in the spreads market was Unilever.
• The competition between the undertakings would be unaffected by the transaction as would the degree of countervailing buyer power held by retailers.
4.18 By way of contrast, the market for sausages was distinguished by a far larger number of suppliers and brands and by a more substantial presence of private label. In that market Kerry’s Denny brand with 31% was the largest single selling brand while Breeo’s brands Shaws, Galtee, Roscrea and Barcastle had a combined market share of just over 6%. The second largest brand is Clonakilty at almost 9%. Eleven other smaller brands took up a 12% market share and private label accounted for 27.8%. (Table 3.1 by volume). Within the private label supplier, Tesco, Dunnes Stores, and SuperValu represented around 4-5% each.
4.19 Based on the HHI calculation, the post-merger concentration was considered moderate in zone B (delta 400.3). While this does not exclude all risk of effect on competition it indicates a risk which is lower than if the market were judged highly concentrated.
4.20 Based on its examination of the closeness of competition, the Authority judged that the evidence indicated that Clonakilty was Kerry’s closest competitor and that more retailers considered that brand a “must have” brand in the market than those who considered the Galtee brand of Breeo. The competitive position of Clonakilty in the No. 2 position was confirmed by the content of Kerry’s internal documentation. In addition, the econometric evidence suggested that the competitive constraint on Kerry’s brands came from brands other than those of Breeo [DET 3.66].
4.21 In its examination of possible counterbalancing constraints, the Authority took the view that new entrants would be unable to establish sufficiently strong presence in that market within a two year period while the evidence indicated that entry into the supply of private label sausages to retailers was comparatively easy [DET 3.75]. It also considered that Clonakilty probably had spare capacity to expand output should retailers switch away from the merged entity. Clonakilty was also an example of a new entrant which had extended its brand image from puddings to capture a significant 9% share of the market in a short time [DET. 3.69]. In consequence, the Authority formed the opinion that the proposed transaction would not result in a substantial lessening of competition in the market for sausages.
4.22 Accordingly, in the light of the issues raised as regards the analysis of the three contested markets which are the subject of the present appeal, it is to be noted that in the case of sausages, Kerry’s Denny brand was the leader with 30% but its closest competitor was a non Breeo brand namely, Clonakilty with a market share of 9%. Breeo’s brands total of 6% was comprised mainly of the Shaws brand with 2.8% and Galtee with 2.5%. The market concentration was thus moderate and in zone B and there was some possibility of counterbalancing constraint in the existence of some spare capacity which would permit expansion by competitors and entry by new suppliers faced no significant barriers.
4.23 Similarly, in the case of the spreads market, the butter segment was characterised by the dominant presence of the Kerrygold brand (belonging to the Irish Dairy Board) and the effective absence of a Kerry brand. In the non butter segment Breeo’s position as a market leader with 30% was followed closely by Unilever’s Flora – a “must have” brand, with Kerry in third place. Also, the spreads market is characterised by a corresponding minor presence of a private label.
4.24 The Authority’s corresponding conclusion as to the absence of lessening of competition in the markets for puddings is based on similar considerations. In that market, the Denny brand was in the leading position with 20.8% (Table 4.1 by volume in 2007). While Clonakilty was in second place with 20.2% and the combined Breeo brands (notably Shaws and Galtee) in third place with 13.8%. Again, the post-merger concentration was measured in zone B and there was evidence of available counterbalancing constraint because entry barriers appeared to be relatively low and expansion by non merging firms was possible – Clonakilty maintained that it had spare capacity to expand.
4.25 Accordingly, it is to be noted that in these three markets the Kerry and Breeo brands were not respectively the closest sources of competitive constraint upon one another and the Authority considered that in each market there was evidence of possible counterbalancing constraints in the form of new entrants and expansion by competitors.
The Contested Markets
4.26 The analysis of the three markets which are contested in this appeal is examined in more detail in considering the arguments of the parties below, but in order to complete the outline of the structure and contents of the Determination the Authority’s approach and conclusions on this can be summarised as follows.
The Rashers Market (Section Five)
4.27 In 2007 the structure of this market by volume and value was found to be as follows.
Kerry’s three brands (Denny, Ballyfree, Clover) had combined market shares of 15.2% (volume) and 22% (value) almost all of which was attributable to the Denny brand products (Table 5.1).
Breeo’s three brands (Galtee, Roscrea, Shaws) held 16.6 and 24.8% respectively comprised almost entirely of Galtee. Six other brands held shares of 6.8 and 6.3% with only Cookstown being over 1% in both measurements. Private label (including discounters – Aldi and Lidl) represented 45.9 and 36.4% amongst which Supervalu was the largest (19.2 and 13.2%) followed by Tesco (10.0 and 8.2%) (Table 5.1).
4.28 It is to be noted, as was pointed out in argument to the Court, that the differences in the figures for value and volume for the brands represent a premium on sales of those products while the higher figure for volume over value represents a discount in the private label sales.
4.29 The HHI calculation placed the post-merger concentration in zone C with a delta of 1092. (Table 5.4).
4.30 The Commission’s investigation of closeness of competition led it to the view that the Denny and Galtee brands of Kerry and Breeo were the closest competitors to one another in the market as was confirmed by the responses of retailers to the effect that those two brands were “must have” brands in the market. While the Authority recognised that private label products may be a growing threat to the brand, it did not consider that source to be a sufficiently close competitor to exert adequate restraint post-merger on the merged entity. It is this assessment which constitutes one of the principal grounds of challenge in the present appeal.
4.31 In assessing the possibilities of constraints on the merged entity the Authority considered that it would be difficult and expensive for a new entrant (that is, as a supplier of branded products,) to rapidly establish a significant foothold in the market sufficient to give rise to a restraint and pointed out that Clonakilty, although a brand leader in the puddings market and in second place in sausages, had struggled to increase its market share in rashers since entry in 2002 [DET. 5.89]. Further, the Authority did not accept the submission put forward by the notifying parties and Compecon that large international suppliers such as Vion and Tulip (who frequently enter the market at the invitation of retailers for the purpose of deeply discounted periodic promotions,) could constitute a source of sufficient constraint. On this basis, the Authority concluded that the transaction would result in a substantial lessening of competition having regard, in particular, to the combination of the market shares of the two leading brands - Galtee and Denny - and the absence of credible alternative brands in the market which would enable retailers to exercise counter balancing constraint. Because Denny and Galtee are considered “must have” brands it judged that private label rashers were not a sufficiently close competitor.
The Non-Poultry Cooked Meats Market (Section Six)
4.32 In this appeal the appellants dispute the Authority’s division of the cooked meats category into separate product markets for poultry cooked meats (chicken and turkey) and non-poultry cooked meats (ham and beef). Having made that finding the Authority characterised the structure of the non-poultry cooked meats (“NPCM”) market as follows. In 2007 Kerry and Breeo were in first and second places respectively with 33% and 14.7% market shares. The proposed acquisition would therefore leave the merged entity with a 48% share with remaining brands accounting for 5.8% only and private label representing 46.2% as of 2007, a decline from 50.9% in 2005. The market concentration measurement estimated the HHI’s delta at 979 placing it in the zone C classification.
4.33 As in the market for rashers, the area of controversy between the notifying parties and the Authority in the latter stages of the investigation concerned not so much the competitive constraints of the Kerry and Breeo brands inter se but the significance to be attached to private label products as a source of post-merger constraint. In this regard, the focus was placed on the econometric evidence and in particular upon the use by the appellant’s Dr. Hogan of the “Tesco EPOS” data which was contended on behalf of the appellants to demonstrate that the private label products had an influence on Kerry which was stronger than that of Breeo. This was rejected by the Authority as being inconsistent with the preponderance of other evidence [DET 6.140]. As a consequence, the Authority concludes that while the private label products are in the same product market as the Denny and Galtee brands they are “distant competitors” to the merging parties in this market. (For an explanation of the debate on the AC Nielsen & EPOS data, see below para 9.34 et seq.)
4.34 The Authority therefore formed the opinion that the transaction will result in a substantial lessening of competition in the NPCM market having regard to the resulting market concentration, to the post-merger combined market share of Kerry and Breeo and to the fact that retailers did not appear to have sufficient countervailing buyer power to exert a counterbalancing constraint which would deter a price increase [DET 6.164].
The Cheese Sector (Section Eight)
4.35 Finally, the notifying parties also disputed the Authority’s division of the cheese sector into separate product markets for natural cheese and processed cheese. The parties had argued that the market was characterised by the fact that consumers purchased cheese by reference to what is called “usage occasion” that is for particular purposes such as for use in sandwiches, for use in cooking, as a course at dinner or as a snack and so on. It was calculated that 83.4% of consumers purchased both natural and processed cheese products in any one year with only 12.4% purchasing only natural cheese and 4.2% purchasing only processed cheese [DET. 8.5]. Moreover, the existence of a single market was evidenced by the fact that the manufacturers had switched or extended from one line into the other. Thus, Kerry’s Low-Low, traditionally a natural cheese, had stretched successfully into processed cheese and Dairylee and Calvita had gone from processed cheese brands into natural cheese [DET. 8.6]. Notwithstanding these arguments the Authority concluded, based largely on survey results and the evidence in internal documents, together with the fact that a supplier of natural cheese could not easily switch within twelve months to supplying processed cheese, that there are two separate product markets [DET. 8.28].
Natural Cheese Market
4.36 So far as concerns the natural cheese market, the Authority found that Kerry was the market leader with a 28.6% share followed by Glanbia with 17.8%, Breeo with 6.8% and Carberry with 6.4%. The combined market share for private label (including discounters) came to 34.1%. (Table 8.1, 2007 by value).
4.37 While these market shares placed the market concentration measurement in the highly concentrated zone C (delta 389) the Authority nevertheless concluded that there would not be a resulting substantial lessening of competition from the transaction. This was largely because Kerry’s closest competitor was Glanbia’s Kilmeaden brand with a market share of 17.8% and a majority of retailers regarded the latter as a “must have” brand. Breeo’s Mitchelstown brand was considered a mere distant competitor to Kerry. It is to be noted that the Authority recognised that existing countervailing buyer power was exercised by retailers and found it would remain unchanged following the merger as Glanbia would remain a credible alternative brand in the natural cheese market [DET. 8.57.]
Processed Cheese Market
4.38 A different structure pertained, however, in the market for processed cheese. Here the market leaders were again the Kerry brands with a share by value in 2007 of 35.6% followed by Breeo with 19.4%. These, however, were closely followed by Kraft with 18.9% and Fromagerie Bel with 11.3%. [DET. Table 8.4]. Unlike some of the other markets examined, private label (including discounters) represented a fairly stable 10%. Kerry’s market share had increased steadily between 2005 to 2007 while Breeo appeared to have experienced a decline from 23.2% to 19.4% in the same period. This acquisition by Kerry of the brands in third position gave a HHI post-merger calculation which placed the market concentration in zone C with a delta of 1382 thus raising concerns of adverse competitive effects.
4.39 The Authority then examines the closeness of competition in the light particularly of the argument advanced by the parties that private label cheese products were in direct competition with the brands and would constitute a real competitive constraint post-merger. While emphasising that private label cheese accounted for 34% of the total cheese market and disputing the division into separate product markets, the parties contended that this was so even on the basis that with a 10% share of the processed cheese market, private label would be in a significant position [DET. 8.66].
4.40 Based on its surveys of retailers and competitors the Authority considered that Kerry’s Easisingles and Breeo’s Calvita and Galtee brands were leaders and considered “must have” brands by retailers. The econometric advice of Dr. Walsh gave no indication as to the closeness of competition, his data not distinguishing the processed cheese segment. The internal documentation was similarly inconclusive on the issue [DET. 8.83, 8.84].
4.41 Based upon the customer/retailer responses the Authority came to the conclusion that, as there were no credible alternative processed cheese slices brands available, the degree of countervailing buyer power exercisable by retailers would decline as a result of the merger [DET. 8.114.]
4.42 It thus came to the view that the proposed acquisition would result in a unilateral price increase by the merged entity [DET. 8.115].
5. Nature and Scope of the Appeal
5.1 Although extensive submissions both written and oral have been made on this appeal on the preliminary issue as to the nature and scope of an appeal under s.24 of the Act, the Court considers that, in the final result, the issue does not present any major difficulty in the circumstances of this case by reason, largely, of the fact that no primary findings of fact made by the Authority in the Determination have been put in dispute by the parties. As appears later in this judgment, the areas of dispute concern principally the adequacy of the information or evidence relied upon and its interpretation by the Authority and, in particular, the economic conclusions which it draws from its evaluation of that material as the basis for the opinions it forms in the Determination.
5.2 Nevertheless, as this is apparently the first appeal against a negative determination under s. 22(3)(b) of the Act. It is appropriate that the Court should briefly outline the approach it has taken to the construction of s. 24 and the standard it adopts in assessing the legality and validity of the Determination.
Construction of Section 24
5.3 The parties agree that the procedure envisaged is one of appeal and that it is therefore wider than a judicial review of the legality of the determination. The main difference between them has been as to i) whether and to what extent the section permits the High Court to substitute its own view of the merits of the proposed acquisition in the light of the criterion of the Act and ii) the degree to which any error made by the Authority must be material to the conclusion upon which the Determination is based if it is to be annulled.
5.4 The pertinent provisions of s. 24 read as follows:
24.-(1) An appeal may be made to the High Court against a determination of the Authority under paragraph (b) or (c) of section 22(3).
(3) An appeal under this section –
(a) may be made by any of the undertakings which made the notification in relation to the merger or acquisition concerned, and
(b) shall be made within 1 month after the date on which the undertaking is informed by the Authority of the determination concerned or, in case the determination is one in relation to a media merger, after the expiry of the period specified in section 23 (9).
(4) Any issue of fact or law concerning the determination concerned may be the subject of an appeal under this section but, with respect to an issue of fact, the High Court, on the hearing of the appeal, may not receive evidence by way of testimony of any witness and shall presume, unless it considers it unreasonable to do so, that any matters accepted or found to be fact by the Authority in exercising the relevant powers under section 22 were correctly so accepted or found.
(5) Notwithstanding subsection (4), the High Court, on the hearing of an appeal under this section, may receive evidence by way of the testimony of one or more witnesses if it considers it was unreasonable for the Authority to have accepted or found as a fact any mater concerned.
(6) Without limiting the exercise of the judicial function with respect to a particular case, it shall be the duty of the High Court, in so far as it is practicable, to hear and determine an appeal under this section within 2 months after the date on which the appeal is made to it.
(7) On the hearing of an appeal under this section, the High Court may, as it thinks fit –
(a) annual the determination concerned,
(b) confirm the determination concerned, or
(c) confirm the determination concerned subject to such modifications of it as the court determines and specifies in its decision.
(9) An appeal to the Supreme Court against a decision of the High Court under any of the foregoing provisions of this section shall lie only on a question of law.
5.5 It is well settled that where the Oireachtas creates a statutory remedy by way of review or re-examination of a decision taken by an administrative or quasi-judicial tribunal, the primary rule for determining the nature and scope of that review is that of construing the words used by the legislature. In particular, where the statute expressly employs the term “appeal” rather than limiting the remedy explicitly to “judicial review” or to an appeal on a point of law, it is of primary importance to determine whether the Oireachtas intended the court to have power to substitute its own view of the merits of the contested decision for that of the decision-maker if it judges that the decision was incorrect on its merits and not merely wrong in law. (See the judgment of Costello J. in Dunne v. Minister for Fisheries [1984] I.R. 230 and that of Finnegan P. in Glancré Teo. v. Cafferkey [2004] 3 IR 401).
5.6 In construing the words used in s. 24, it is clear that a number of points emerge:
(a) the procedural remedy thereby created is expressly one of statutory appeal. In this it contrasts, for example, with other forms of remedy against decisions of administrative and quasi-judicial tribunals where the recourse is expressly limited to that of judicial review under O. 84 of the Rules of the Superior Courts and often to a very circumscribed form of that review. (See in that regard, for example, s. 50(2) of the Planning and Development Act 2000 (as amended) and s. 5 of the Illegal Immigrants (Trafficking) Act 2000).
(b) In principle and subject to the important limitation mentioned below, an appeal may raise “any issue of law or fact” concerning the determination. The procedure before the High Court is not therefore confined to an appeal on a point of law and thus contrasts with the further limited review which is possible before the Supreme Court as provided for in ss. (9) of s. 24. It follows, accordingly, that the appeal includes, but is wider than, a review of the substantive and procedural legality of a determination. To paraphrase the words quoted from Wade’s Administrative Law by Costello J. in Dunne v. Minister for Fisheries (above) the appeal may raise both the question “is it lawful or unlawful?” and the question “is it right or wrong?”
(c) On the other hand, it is equally clear that the procedure is not expressly an appeal by way of rehearing of the original notification in which the decision of the court fully replaces that of the Authority and, in this respect, the provision can be contrasted with, for example, the procedure for review of an application for planning permission decided by a local authority when appealed to the Planning Board under s. 37 of the Planning and Development Act 2000 (see in particular ss. (1)(g) of that section).
(d) That an appeal of limited scope is envisaged is also suggested by the duty imposed on the court by ss. (6) of s. 24 (so far as practicable) to hear and determine the appeal within two months. (That such a time limit involves considerable restraint upon the scope of an appeal can be seen from the fact that the present case, which is unlikely to be untypical, has involved an investigation lasting some five months, a determination comprising more than 150 pages and covering five product markets, appeal papers comprising more than 20 lever arch files and a hearing which has lasted eight days.)
(e) Most significantly, however, ss. (4), while permitting any issue of fact to be raised on the appeal, requires the court to presume, unless it is considered unreasonable to do so, that matters of fact have been correctly accepted or found by the Authority in its determination. The court is precluded from hearing any witness evidence on any issue of fact unless it has first been satisfied that the Authority was unreasonable in its finding or acceptance of that fact.
(f) It necessarily follows, therefore that, on the one hand, the court is required not to re-open or interfere with findings of fact so long as that presumption stands, thereby limiting in a significant degree the scope of the procedure as an appeal directed at the correctness of the determination on its merits, as opposed to one confined to its substantive and procedural legality.
(g) It also follows, on the other hand, that the court can in those limited circumstances where the presumption of reasonableness is rebutted, re-open and re-decide specific issues of fact. Accordingly, in so far as the court may thus have to make a new finding on an issue of fact material to the validity of the Authority’s determination, the section clearly envisages that the court may in principle and exceptionally substitute its own new findings of primary fact for those of the Authority.
5.7 In this last regard the appeal provided for in s. 24 can be contrasted with that provided for in s. 15 of the same Act. Under s. 4 (3) of the Act, the Authority has a power equivalent to the “bloc exemption” power of the Commission under article 81, para. 3 of the EC Treaty whereby the Authority may declare that specified categories of agreements, decisions or concerted practices which otherwise come within the prohibition in s. 4(1) as anti-competitive, comply with the exempting conditions of ss. (5). An appeal against such a declaration can be made to the High Court within 28 days under s. 15 and on such an appeal, ‘the High Court may confirm, amend or annul the declaration’ in question. The factual findings upon which the declaration may be based enjoy no presumption equivalent to that accorded to the determination under s. 24. Thus, while clearly wider in scope than a judicial review as to legality, the appeal under s. 24 is clearly narrower than that available under s. 15. In addition, of course, the decision on the appeal under s. 24 is subject to the narrow, although not absolute, time constraint of ss. (6).
5.8 It must also follow, however, from the fact that a procedure by way of appeal is created in which any issue of law or fact may be raised, that the Court is entitled to examine also the correctness of material conclusions reached by the Authority involving mixed questions of law and fact as where, without any wrong finding of primary fact being made, a legal conclusion is drawn from one or more facts.
5.9 The Court therefore concludes:
Standard of Review
5.10 While it is thus possible to construe the nature and scope of the particular statutory appeal envisaged in s. 24, this does not resolve the real difference between the parties as to the approach which this Court should take in deciding the present case because no finding of primary fact in the determination has been put in issue and the necessity to re-open such findings and admit evidence has not therefore arisen.
5.11 The difference between the parties in this regard is more accurately described as directed at the standard of review or scrutiny which the court should apply to the errors alleged to have been made by the Authority in assessing the factors which govern competition in the subject markets and upon which the conclusion of substantial lessening of competition is based. If it is open to the notifying parties to seek to prove that errors have been made which render the determination wrong, how wrong must it be if it is to be annulled by the court? What test of error is to be applied having regard, in particular, to the deference which is due to the specialist expertise of the Authority inherent in the presumption in its favour on matters of fact in s. 24 (4)?
5.12 The appellant argues for a standard of “manifest error” as applied by the European Courts in reviewing equivalent decisions of the European Commission under the Competition Rules of the Treaty or under Council Regulation No. 139/2004 (or its predecessor Council Regulation No. 4064/89). The appellant expresses that standard in its written submissions in the following terms:
“The appeal should be allowed if the appellant can demonstrate that any error as to fact or law was one which, when objectively assessed, had a bearing on the decision reached by the Authority. However, the errors need not go to the root of the decision either. Rather the errors should be material in the sense that they are objectively relevant to and have a bearing on the conclusion at which the Authority arrived. In coming to this conclusion, the court should take into account the view of the Authority given its expertise on certain specialist competition matters, but ultimately can substitute its own opinion if it takes a different view in respect of these matters.”
5.13 The Authority, on the other hand, would set the bar at a higher level and argues for “an enhanced standard” consistent with what it says is the clear intent of s. 24. It argues that the appellant must establish one or more significant erroneous inferences which are critical to and which go to the root of the decision embodied in the determination. It urges the court to adopt the standard applied by Finnegan P. in Ulster Bank Investments v. Financial Services Ombudsman [2006] IEHC 323 where, in considering an appeal to the High Court under the provisions of the Central Bank Act 1942 (as amended) he said: “To succeed on this appeal the plaintiff must establish as a matter of probability that, taking the adjudicative process as a whole, the decision reached was vitiated by a serious and significant error or a series of such errors. In applying the test the court will have regard to the degree of expertise and specialist knowledge of the defendant. The deferential standard is that applied by Keane C.J. in Orange v. Director of Telecommunications Regulation and Another and not that in The State (Keegan) v. Stardust Compensation Tribunal.”
5.14 While, as indicated above, the primary consideration in all of these cases is the correct construction of the relevant statutory provisions, the Court agrees that the standard of deference indicated by Keane C.J. in the Orange case is appropriate to be applied to the present statutory appeal. Although the terms of that statutory appeal were different and the administrative context from which the appeal arose was quite distinct (an appeal against a refusal by the Director of Telecommunications Regulation to grant a third mobile telephone licence in the State,) and which the appeal was not expressly limited by a presumption equivalent to that contained in s. 24(4) in the present case, the judicial task required does not appear to be materially different to that required here. Keane C.J. introduced the formulation cited above in the Ulster Bank Investment case by distinguishing the form of statutory appeal in question in the following terms. He said that the appeal under s. 111(2)(b) of the Postal and Telecommunication Services Act 1983 was not: “intended to take the form of a re-examination from the beginning of the merits of the decision appealed…it is accepted that at the other end of the spectrum, the High Court is not solely confined to the issues which might arise at the decision as if it was being challenged by way of judicial review,…”
He then added the statement of the standard cited by Finnegan P.
5.15 Subject to giving correct effect to the precise terms in which the statutory appeal in each case is expressed, it is obviously desirable as a matter of policy to ensure that the criteria upon which an appeal is based remain consistent with other closely analogous statutory appeals under domestic legislation and also, in the area of competition law, with the Authority’s obligation to ensure that its decisions are consistent with Community law where they have a Community dimension.
5.16 In that regard it is important to bear in mind that in exercising its functions under Part II of the Act, the Authority (and, by extension, this Court) has explicit obligations under articles 3, 5, 11 and 16 of Regulation 1/2003 to ensure that the provisions of articles 81 and 82 of the Treaty are applied consistently with Community law. While the relationship of the Authority to the European Commission in the area of the control of mergers and acquisitions is distinct from its role under Regulation 1/2003, it is nevertheless desirable that no unnecessary discrepancies should be created in the criteria applied, having regard to the fact that in certain circumstances the Authority’s functions under Part III of the Act may fall to be applied to a merger remitted for adjudication by it from the European Commission under articles 4(4) and 9 of the Merger Regulation. It is also pertinent to bear in mind that the criteria for determining a relevant product market are the same for the purposes of article 81 and 82 EC and of the Merger Regulation.
5.17 Although the scope of an appeal under s. 24 is wider than that of the review procedure of the European Courts under Article 230 EC (those courts cannot substitute their views on the merits,) the Court considers that a standard of review for the purposes of s. 24 of the Act based upon the formulation expressed by Keane C.J. in the Orange case and adopted by Finnegan P. in Ulster Investment Funds Ltd. is consistent with the standard of review applied to analogous decisions of the European Commission by the European Court of Justice and the European Court of First Instance.
5.18 The concept referred to in the Common Law jurisdictions as “curial deference” is reflected in the jurisprudence of the Community Courts in the recognition of a “margin of appreciation” or “margin of discretion” accorded to the European Commission in its appraisal of the complex economic situations that arise in the application of both competition rules and the provisions of the Merger Regulation.
5.19 As the parties have recognised in their submissions in this case, this principle was applied by the Court of First Instance in the context of mergers in its judgment of 25th October, 2002, in case T5/02 Tetra Laval v. Commission and upheld on appeal by the Court of Justice in its judgment of 15th February, 2005 in case C-12/03P Commission v. Tetra Laval. A fuller restatement of the three limbs of the Courts’ approach to the review of decisions of the Commission which turn upon matters of both economic and technological complexity can be found in the more recent judgment of the Court of First Instance of 17th September, 2007 in case T201/04 Microsoft v. Commission as follows:-
“87. The Court observes that it follows from consistent case law that, although as a general rule the Community Courts undertake a comprehensive review of the question as to whether or not the conditions for the application of the competition rules are met, their review of complex economic appraisals made by the Commission is necessarily limited to checking whether the relevant rules on procedure and on stating reasons have been complied with, whether the facts have been accurately stated and whether there has been any manifest error of assessment or a misuse of powers. (Case law omitted)
88. Likewise, insofar as the Commission’s decision is the result of complex technical appraisals, those appraisals are in principle subject to only limited review by the Court, which means that the Community Courts cannot substitute their own assessment of matters of fact for the Commissions. (Case law omitted)
89. However, while the Community Courts recognise that the Commission has a margin of appreciation in economic or technical matters, that does not mean that they must decline to review the Commission’s interpretation of economic or technical data. The Community Courts must not only establish whether the evidence put forward is factually accurate, reliable and consistent but must also determine whether that evidence contains all the relevant data that must be taken into consideration in appraising a complex situation and whether it is capable of substantiating the conclusions drawn from it. (See, to that effect, concerning merger control, case C-12/03P Commission v. Tetra Laval [2005] E.C.R. 1-987, para. 39).”
5.20 Accordingly, in a case such as the present, where primary findings of fact have not been put in issue, the Court considers that a determination by the Authority that a merger or acquisition will result in a substantial lessening of competition ought not to be set aside by this Court unless:-
(a) The Authority is shown to have committed a serious error in drawing inferences or conclusions from facts, such that the inferences or conclusions become untenable or unsound by reason of the error having been made; or
(b) It is demonstrated that the Authority has failed to take into consideration or adequately to consider, relevant information or data such that an inference or conclusion material to the determination is unsupported by or is rendered inconsistent with the clear force and effect of the available evidence taken as a whole; or
(c) A significant appraisal of economic or technical factors material to the functioning of competition in the relevant market is shown to be so inconsistent with the available evidence as to be manifestly unreasonable and unsound; or
(d) The Authority’s statement of its reasons for reaching conclusions material to the basis of the determination is lacking in cogency or coherence or is contradicted by the evidence which was available to it; or
(e) The Authority has made a material error of law either in the construction and application of the Act or by otherwise infringing some applicable principle of constitutional or natural justice.
5.21 In other words, where the Authority has, without committing significant error, exercised its specialist expertise in making judgments as to the prospective consequences of the economic and commercial factors which govern or influence competition in the relevant market, this Court should not intervene even if it is demonstrated that an opposite conclusion might plausibly have been reached by placing weight on different aspects of the available evidence or data or by attributing different or greater significance to other pertinent factors in the economic assessment. Nevertheless, the Court will be entitled and obliged to intervene to set aside a material economic conclusion if it is shown to be incorrect because it is unsupported by or inconsistent with the clear effect of the evidence, information or data upon which it is based.
6. The Grounds of Appeal
6.1 The appellant’s challenge is, of course, directed at the ultimate finding of the Authority that the acquisition will result in a substantial lessening of competition in each of the three markets for rashers, non-poultry cooked meats and processed cheese. Slightly different approaches are taken by the appellants in the way in which these grounds are presented in the Schedule of Grounds attached to the originating notice of motion and in the written submissions delivered on foot of the Courts’ order for directions but the main structure of the arguments can, for the purpose of this judgment, be outlined as follows.
A. A broad and detailed challenge which is common, in effect, to all three markets is raised against the Authority’s finding that private label products marketed by the four main retail chains (and, where appropriate, by the discounters,) while competing products in the same product market, are not sufficiently close as competitors to be in a position, post-merger, to exercise constraint upon a sustained price increase in the branded products by the merged entity. This broad challenge is grounded upon a number of arguments as to the existence of serious and significant errors in the Authority’s analysis of competition in the three relevant markets which both individually and, a fortiori when taken together, go to the root of the validity of the determination and justify annulment. In particular, it is argued that the Authority was wrong to conclude that private label products were not in close competition with the branded products because the Authority erred in the following respects:-
(i) It in effect misjudged the evidence as to the true competitive relationships between brands and private label products and ignored or misinterpreted relevant evidence in that regard;
(ii) It failed correctly to assess the true extent and significance of the countervailing buyer power of the multiple retailer chains as both “gatekeepers” to the branded products and as competing suppliers of their own private label products and wrongly rejected or ignored direct evidence of the exercise of such buyer power;
(iii) The Authority wrongly assessed the role and importance of brands in that it attributed exaggerated weight to some brands as “must have” or “must carry” brands in dis-regard of the evidence of the absence of any significant brand loyalty in these sectors;
(iv) The Authority erred in its analysis of the counterbalancing effect of the entry of new competitors or expansion by existing suppliers in the relevant markets;
(v) It erred in its evaluation of the econometric evidence and in particular it wrongly rejected and failed to consider the additional analysis submitted by Dr. Hogan based on the Tesco EPOS (electronic point of sale) data which was clearly superior both in its relevance to the key issue and in its representative quality, to that relied upon by Dr. Walsh in that it took into account the entire range of Tesco products including private label whereas no private label data had been available to Dr. Walsh whose reports in this area were relied upon exclusively by the Authority.
B. In addition to challenging the analysis of competition and the evaluation of competitive constraint from private label products in the non-poultry cooked meats market, the appellant raises a discrete ground in relation to that market in alleging that the Authority committed a fundamental error in concluding that the cooked meats sector divided into two separate product markets, namely poultry cooked meats and non-poultry cooked meats. In particular the appellant alleges that the Authority erred fundamentally in making no supply side assessment of substitutability. This is alleged to contravene the Merger Guidelines and it is the only instance in the Determination where this analysis was not carried out.
C. Similarly, in relation to the analysis of markets in the cheese sector, the appellant alleges that the Authority made a two-fold error in defining the product market by first concluding that natural and processed cheeses constituted two distinct product markets and secondly, by then further sub-dividing the processed cheese market (or appearing to do so) into a distinct sub-market for sliced processed cheese.
In addition the appellant argues that the Authority further erred in its analysis of the geographic market by finding that it was “at least as wide as the State” in dis-regard of the evidence that the geographic market in question encompasses at least the entire island of Ireland.
D. Finally, the appellant alleges that the Authority erred in its analysis of the efficiency gains which the appellant had claimed would result from the acquisition and which, it said, would be implemented and would be achieved.
In particular, the appellant alleges that the Authority is guilty of serious error in rejecting the claimed efficiency gains upon the ground that they were “almost certainly biased upwards” [DET. 9.33].
6.2 In the written submissions delivered on foot of the Court’s order for directions, the ground as originally given in the Schedule of Grounds with the notice of motion as being directed at the substantive analysis and calculation of the efficiencies in question, was expanded to a more general allegation of infringement of fair procedures in the manner in which the Authority addressed those claims. Specifically, it is alleged that the concerns originally raised by the Authority at paras. 9.26 and 9.28 of the Assessment of 25th July, 2008 had been answered and rebutted but that paras. 9.34 and 9.36 of the Determination which correspond exactly to those earlier paras. in the Assessment, put forward different arguments for rejecting the particular savings there discussed, although the appellants had been given no opportunity of addressing such new arguments.
6.3 Furthermore, it is alleged that the Authority infringed fair procedures in that, after the issue of the Assessment, the Authority conducted further investigations in relation to private label evidence and to the evidence of countervailing buyer power without communicating the questions it posed and the responses it received to the appellants to enable them to be considered.
6.4 Finally, on the day prior to the commencement of the hearing on the present appeal the appellant delivered to the respondent a book of supplementary written submissions which sought to further expand the grounds raised against the conclusion reached on the question of efficiency gains by alleging manifest errors in the figures given under the column headed “Sales (2007)” in Table 9.2 of the determination. In particular, it is alleged that the Authority having, in response to the notifying parties rebuttal of the “social gain” or “industry – wide” basis for rejection of the claimed deficiencies in paras. 9.26 and 9.28 of the Assessment, withdrawn and altered its argument in paras. 9.34 and 9.36 of the Determination, nevertheless based its crucial evaluation of the efficiencies in Table 9.2 and para. 9.74 upon industry – wide sales figures and not upon the figures for combined Kerry/Breeo sales. The conclusion that the efficiencies are insufficient is thus claimed to be manifestly wrong. The advancing of this supplementary ground is strongly opposed by the Authority as too late and therefore inadmissible.
6.5 The Court proposes to examine the above grounds by considering first the arguments raised against the definitions of the products markets (B and C above) before examining the common grounds at A.
7. Cheeses: Product Market Definition
7.1 The cheese sector and the divergent conclusions reached by the Authority in respect of the natural cheese market and the market for processed cheese are dealt with in Section Eight of the determination. Although in presenting its arguments the grounds advanced by the appellant in respect of the analysis of the cheese markets were placed in the penultimate position before the arguments on the efficiencies, the Court proposes to deal first with this aspect of the case because, as will appear later in the judgment, the view it has reached in respect of Section Eight serves to simplify the assessment it must make of the grounds raised which are common to all three markets in which a lessening of competition was found to result from the acquisition.
7.2 As already indicated, in Section Eight the Authority first examined the definition of the product market in the light of the submissions and internal documents of the parties, and of its own survey of retailers and concluded at para. 8.28 that:
“… notwithstanding the arguments of the parties, taking into account the survey results, the parties’ own view of the market as evidenced in the quoted internal documents and the fact that a supplier of natural cheese could not easily switch in a twelve month period to supplying processed cheese, the Authority considers that there are two separate product markets: the market for natural cheese and the market for processed cheese.”
7.3 In paras. 8.32 to 8.59 it then examined the natural cheese market and concluded that a lessening of competition would not result from the transaction for the reasons given at para. 8.59 as follows:“ - Although the HHI results place the proposed acquisition in zone C… Breeo is a distant competitor to Kerry in the natural cheese market;
- The merger will see the first ranked undertaking in terms of market share, Kerry on 28.6% acquire the third ranked firm Breeo on 6.8%. However, in 2007 the fourth ranked firm, Carberry, had a market share of 6.4%;
- Glanbia’s Kilmeaden brand is the closest competitor to Kerry in the natural cheese market with a market share of 17.8%;
- The majority of retailers indicated that Kilmeaden is a “must-have” brand; and
- The extent to which retailers can credibly threaten to discipline the merged entity post-acquisition will be unaffected by the proposed acquisition.”
7.4 In the remainder of Section Eight the Authority examines and analyses the structure of and competition in, the market for processed cheese and, at para. 8.115 concludes that the acquisition of Breeo by Kerry will result in a unilateral price increase by the merged entity and thus a substantial lessening of competition in that market in view of the nine considerations set out in that paragraph.
7.5 It is to be noted first that the product market in respect of which market structure is examined and the calculation of market concentration made in paras. 8.60 to 8.63 of the Determination is that of processed cheese. It is also to be noted that in Table 2.8 of the Determination the Authority recorded the total size of the market for cheese products in the State by value of sales in 2007 as €193.4 million comprised of sales of natural cheeses of €134.6 million and processed cheese of €58.9million. In other words, in 2007 sales of natural cheeses represented some 70% of the market and processed cheeses 30%. Also, as already indicated (see para. 4.7 above and Table 2.2 of the Determination) the Authority identified Kerry as having seven brands in this sector and Breeo as having five.
7.6 In the light of that general background, the Court notes that the conclusion reached by the Authority in relation to the relevant market that is, the market for processed cheese products, is based upon the 9 considerations listed in para. 8.115 and which can be described as follows.
(i) First, after recalling the market concentration calculation as placing the acquisition in the highly concentrated classification of zone C, the Authority notes that the merger would see Kerry’s leading market share of 35.6% combined with Breeo’s second ranking figure of 19.4% such that the merged entity would account for 55% of the processed cheese market.
(ii) The Authority then states that the relevant market “may be segmented by (a) processed cheese slices; (b) processed cheese snacks and (c) processed cheese spreads”.
(iii) It records that evidence from retailers indicates that the Kerry and Breeo brands are each closest competitors to one another in the “slices segment” of this market which accounts for approximately 40% of total sales in the processed cheese market.
(iv) Three of the remaining four considerations which are relied upon for the conclusion are then directed at factors relating to the slices segment of the cheese market.
(v) Thus, it is stated that there are no credible alternative brands in that segment which would enable retailers to constrain price increases by the merged entity and the expansion of non-merging firms such as Hochland and Lactalis will be insufficient to have such a restraining effect. Secondly, new entrants such as Kraft and Fromagerie Bel would be unable in a timely manner to establish a sufficiently strong presence to act as a constraint. Finally, retailers will not :-
a) have the countervailing buyer power to credibly threaten the merged entity because there are not credible alternative branded processed cheese slices suppliers;
b) entry of branded processed cheese slices will not be sufficient within a two year period; and
c) private label processed cheese is not considered a close competitor in the market and, thus, cannot be used to replace the merged entities processed cheese slices offering.
(vi) The third of the last four considerations reverts to an aspect of the relevant market: “Private label processed cheese sales account for only 10.6% of the processed cheese market and is not considered by retailers to be a close competitor to Kerry or Breeo’s processed cheese products.
7.7 It is immediately apparent therefore, that there is in the conclusion expressed at para. 8.115 an ambiguity and a lack of coherence as to the manner in which the Authority is drawing its conclusion from its analysis of the relevant market as it has been defined, as opposed to the significance it appears to attach to the fact that the Kerry and Breeo brands are one another’s closest competitors in the 40% of that market represented by the slices segment. Thus it is not clear, at least from para. 8.115 itself, what significance the Authority attaches to the sub-division of the relevant market into the three segments. The section clearly contains no finding, as such, that the slices segment constitutes a distinct product market. Nor does the paragraph appear to indicate that the fact that the Kerry and Breeo brands are one another’s closest competitors within the slices segment necessarily results in their also being closest competitors to one another in the product market as a whole when the remaining 60% of that market is taken into account.
7.8 The reliability of the Authority’s definition of the product market also appears to be put in doubt by some features of its competitive analysis in the earlier part of Section Eight. At paras. 8.71 to 8.78 the Authority gives some detail of specific responses received from six retailers who were sent a follow-up questionnaire. One retailer indicated that the processed cheese market was divided into two categories: sliced processed cheese and processed cheese snacks. Another indicated that it was divided into three categories: processed cheese singles; processed cheese snacks; and spreadable processed cheese. Another retailer who indicated that there were four “must have” brands in processed cheese (Kerry’s “Easi singles” and “Cheesestrings” and Breeo’s “Calvita” and “Galtee”) does not appear to have averted expressly to the market being further segmented. Finally, one retailer who indicated that it had no concerns with the proposed acquisition considered that the market divided into two segments (a) standard processed cheese and (b) processed cheese snacks. This retailer does not appear to have mentioned sliced processed cheese as a discreet segment but at para. 8.78 the Authority concludes that the processed cheese market is segmented into “standard/sliced processed cheese, cheese snacks and processed cheese spreads”. It is not immediately clear whether the first of those segments is the same or slightly larger than the first segment identified in the fourth consideration of para. 8.115 as “(a) processed cheese slices”.
7.9 The Court also notes that in para. 7.8 when referring to the three segments of the product market the Authority states “a majority of retailers are concerned about the impact of the proposed acquisition on the standard/sliced processed cheese segment which accounts for the majority of sales in the processed cheese market”. In para. 8.115 on the other hand, it is stated that the “slices segment” of the processed cheese market “accounts for approximately 40% of total sales in the processed cheese market”.
7.10 As regards the internal documents of the notifying parties the Authority remarks in para. 8.83 that “Kerry and Breeo are the two leading brands in the processed cheese market” but that “the internal documentation does not provide much insight into the closeness of competition” in that product market. In para. 8.84 the Authority refers to its own econometric evidence provided by Dr. Walsh which presents results for the overall cheese segment incorporating both natural and processed cheese and adds “thus the econometric analysis provides no indication on the closeness of competition in the processed cheese market”. The conclusion is then given in para. 8.85 that “the evidence indicates that Kerry and Breeo are each others closest competitor in the standard/sliced processed cheese segment of the processed cheese market”. It is to be noted that the Authority does not thereby conclude that they are also each others closest competitors in the full 100% of the defined product market.
7.11 Accordingly, it is difficult to avoid the impression that while the Authority does not make any finding to the effect that the standard/sliced cheese segment or the sliced cheese segment is a distinct product market in itself, it nevertheless appears to base its conclusion as to the closeness of competition between the Kerry and Breeo brands upon the competitive effect in that segment alone as if it was the relevant product market.
7.12 The Court considers that this ambiguity or lack of coherence in the Authority’s approach to the processed cheese products as a distinct product market necessarily puts in question its approach to the sector as a whole and thus its original division of it into the distinct product markets for natural and processed cheese.
7.13 At paras. 8.2 to 8.15 of the Determination the Authority records the submissions made by the notifying parties including the following points as to the characteristics of this sector:-
• The Compecon Report treats cheese as a single market because 94% of the product is pre-packed and mostly of the cheddar variety;
• Natural cheese is a commodity and many operators including Kerry do not produce the product but buy, slice and pack cheeses;
• The notifying parties’ research report from “Amárach Consulting” show that consumers do not buy natural or processed cheeses as such but buy products for particular uses or occasions;
• The research shows that 83.4% of consumers buy both natural and processed products within a year and 12.4% purchase only natural cheese while 4.2% purchase only processed cheese thus indicating a high degree of overlap;
• Manufacturers have successfully switched or extended from one category into the other: thus Kerry’s Low-Low extended from the natural category to processed while Kraft’s Dairlylea and Breeo’s Calvita stretched in the opposite direction;
• The products in the two categories are not sold or displayed separately on supermarket shelves but are intermingled.
7.14 In rejecting the proposition that there is a single product market for cheese the Authority relies on three factors: supply side substitutability, its retailer survey and particularly on the contents of internal documents from the parties [DET 8.16 – 8.27].
7.15 A detailed questionnaire was sent to nine leading retailers but at paras. 8.17 and 8.18 only the response to one question in “the follow-up questionnaire” directed at the SSNIP Test is explained. It asked in effect, if the price of processed cheese supplied by a hypothetical monopolist were to rise permanently by 5 – 10% would they:
“(1) Maintain your spend on processed cheese (and buy less)?
(2) Purchase the same volume of processed cheese and pay more?
(3) Switch some of your spend away from processed cheese to natural cheese. If so, what proportion of spend?”
The conclusion drawn from this question is then outlined in the two following paragraphs as follows:-“8.18. Four out of the nine retailers indicated that they would opt for either option (1) or (2). Two retailers provided unclear responses as to how they would react to a permanent rise in price of processed cheese. One retailer indicated that it was not sure how it would react to this hypothetical situation. One retailer gave an “N/A” response. Finally one retailer provided no response to this question.
8.19. Thus, the survey results, although not conclusive, lean towards supporting a view that processed cheese occupies a distinct product market from natural cheese.”
7.16 The Court can agree with this last statement insofar as it accepts that the survey results are inconclusive but must disagree with the proposition that they can form a basis for the assertion that retailers consider that there are two distinct product markets for natural and processed cheese. First, it is not at all clear whether the “follow-up questionnaire” mentioned in para. 8.17 is the same document as the “detailed questionnaire” referred to in the preceding paragraph. Secondly, it is difficult to read into that response any particular view on the part of retailers as to the definition of the product market without knowing what other detailed questions were asked of them that might have a bearing upon their handling of natural and processed cheese products and how they responded. Thirdly, the response given in para. 8.18 is equally open to the construction that 50% of the retailers in question did not consider the distinction between stocking natural as opposed to processed cheese, as such, a relevant or meaningful notion. In circumstances where it is apparently not disputed that 83% of consumers buy both natural and processed cheese in any year and given that retailers usually strive to offer the range of products their customers demand, it may not be surprising that retailers might be uncertain as to how to reply to such a SSNIP test question.
7.17 In paras. 8.20 and 8.21 the Authority cites very briefly some snippets from internal documents of Kerry and Breeo respectively in support of the conclusion that “it is clear that the parties perceive themselves as competing in separate market products for natural cheese and processed cheese”. While it may be said that the sentences or part sentences quoted do refer to “the natural cheese market” or “the number one processed cheese brand” it is difficult to conclude that these documents constitute reliable evidence to support the conclusion attributed to them in para. 8.22 without knowing what the documents in question are in more detail, why and by whom they were written. A different construction may fall to be applied to them when read as a whole depending on whether they emanate from the production department or the accounts department or the marketing department.
7.18 Finally, in paras. 8.23 to 8.27 the Authority deals with supply side substitution. A questionnaire was apparently sent to two suppliers of, inter alia, natural cheese products in the State seeking views as to whether processed cheese and natural cheese could be manufactured in the same facility. At paras. 8.25 and 8.26 the views of the two suppliers are recorded indicating that different facilities are required for production of the two products and that it is unlikely that a supplier would switch production, although the second supplier questioned indicated that a manufacturer could convert from natural cheese to processed cheese by purchasing another manufacturing plant. On this basis the Authority expresses the view that “on balance it does not believe that a supplier of natural cheese would be able to switch within a twelve month period to supplying processed cheese” in response to a permanent price increase of 5 – 10% in processed cheese.
7.19 The Court considers that the basis upon which the Authority has thus reached and explained its crucial conclusion in relation to the definition of the product market in the cheese sector is inadequate and unsound. In contrast with the analysis that is conducted of product markets elsewhere in the Determination, the analysis in section Eight contains no detailed investigation of demand side substitutability which is relied upon elsewhere by the Authority as the primary basis of assessment and as envisaged in the Merger Guidelines. In particular, as summarised by reference to paras. 8.2 – 8.15 of the Determination above, the notifying parties had put forward both in their written response to the Assessment and in a slide presentation at the oral hearing, a number of cogent propositions as to characteristics of the cheese market and how both the natural and processed products are presented to consumers and apparently treated by them without distinction, but no attempt is made to address or rebut these concrete propositions in the relatively brief analysis of paras. 8.16 – 8.28 of the Determination.
7.20 For these reasons the Court finds that the Authority has erred in its definition of the relevant product market for the cheese sector with the result that its conclusion as to a resulting substantial lessening of competition in a product market comprising processed cheese is fundamentally flawed.
7.21 It must be recalled that the definition of the product market is crucial to the reliability of the assessment of any change in competition resulting from the proposed merger. The primary source of possible constraint upon the ability of the merged entity to raise prices lies in the possibility that consumers of the product will react by switching to acceptable alternative products already on offer to them. The European Commission in its notice on market definition ([1997] OJ C372/5) defines the concept in these terms:
“A relevant product market comprises all those products and/or services which are regarded as interchangeable or substitutable by the consumer, by reason of the products’ characteristics, their prices and their intended use.”
Having regard particularly to the evidence submitted by the notifying parties as to the common characteristics and uses (from the consumer point of view) of the natural and processed products and to the manner in which most are similarly marketed by packaging, brand and intermingled on shelves, this Court considers the factors relied upon by the Authority for dividing the market unsupported by the evidence as a whole and inadequate.
7.22 Accordingly, in the Court’s judgment the Authority has erred significantly in the manner in which it conducted its analysis and reached its conclusion in that:
(a) In defining the product market as divided between natural and processed cheese, it relied upon inadequate and inconclusive evidence or data at paras. 8.16 to 8.27 of the Determination;
(b) It failed to consider, examine and to explain its rejection of material characteristics of the products and how they are dealt with by retailers or apparently perceived by consumers as evidenced by the appellants in their written submissions and in the Compecon Report; and
(c) In its analysis of the processed cheese market as defined by it, its statement of reasons as to the relationship between that product market as defined and the positions of the Kerry and Breeo branded products in the 40% segment comprised of the sliced cheese products is incoherent and fails to meet the standard indicated at para. 5.19 above of this judgment.
7.23 It is unnecessary therefore to consider the arguments raised by the appellant as to error in the determination of the geographic market for cheese products.
8. Cooked Meats: Product Market Definition
8.1 In the light of the Court’s examination of the product definition in the cheese sector it is convenient to look now at the corresponding arguments raised as regards the Authority’s division of the market in cooked meats into the separate products markets for poultry cooked meats and non-poultry cooked meats (“N-PCM”) [DET. 6.1 - 6.34].
8.2 As in the case of the cheese markets, the examination of the relevant product market for cooked meats is similarly structured. The Authority first summarises the submissions made against the Authority’s position as explained in the Assessment [DET. 6.2-6.8]. It then outlines its own investigation by reference to its survey of retailers for the purpose of the SSNIP test, its examination of internal documentation and then supply side substitution. (The Authority also looks, in passing, at the question as to whether these products in pre-packed and OTC format are in the same market and concludes that, on balance, a supplier of OTC poultry cooked meats could easily switch within a twelve month period to supplying the pre-packed product in response to a price increase [DET. 6.33].
8.3 In applying the SSNIP test the Authority issued questionnaires to nine leading retailers and put a question in similar terms to that posed for cheeses at para. 8.17. While there is some confusion in paras. 6.14 – 6.16 as to whether there was an initial questionnaire and a follow-up questionnaire and whether nine or only six responded to each, the Authority concluded at para. 6.17, correctly in the view of the Court, that the responses did not constitute useful evidence for identifying the relevant markets. Of the six retailers who apparently replied to the SSNIP question, two provided contradictory answers while two each split between the single market and dual market viewpoint thereby cancelling one another out.
8.4 The Authority next interprets a number of items of internal documentation recovered from the parties as being consistent with a view that there are two separate markets applying the criteria suggested by the International Competition Network in its Guidelines for Merger Analysis quoted at paras. 6.21 and 6.22.
8.5 The Authority then draws attention in Table 6.1 to the fact that in Kerry’s case it has separate brands for poultry cooked meats (Ballyfree) (Denny) and N-PCM. Breeo has two brands (Galtee and Roscrea) used exclusively for non-poultry cooked meats and one, Shaws, which is used for both categories of product.
8.6 Finally, the Authority refers to the question of supply side substitution but responds to the criticism that this was not examined in the Assessment by quoting para. 2.12 of its Merger Guidelines which sets out the evidence required for such a study and then observes that “evidence of this nature was not supplied by the parties in either the original notification or in any subsequent submission”. Paragraph 6.27 then asserts:
“The Authority did ask competitors whether manufacturing processes for poultry cooked meats could be easily switched to non-poultry cooked meats. Evidence supplied by (redacted) suggest that poultry and non-poultry cooked meats are not run in the same plant due to health reasons.”
(In common with many other paragraphs of the version of the Determination furnished to the Court for the purpose of the hearing, the identity of the supplier there referred to is censored or, as it is called, “redacted”. In para. 186 of the Authority’s written submissions to the Court, it is claimed that this refers to the replies of two competitors.)
8.7 It is on this basis that the Authority concludes at para. 6.34 that “there are two relevant product markets for cooked meats: poultry cooked meats and non-poultry cooked meats”.
Arguments of the Parties
8.8 The appellant submits that in the absence of any usable evidence based upon the SSNIP test, the Authority’s finding is wrong on the evidence because the internal documentation which it invokes tells nothing as to how substitutability is viewed by consumers. The appellant’s main attack, however, is directed at the complete absence of examination of supply side substitution which it regards as a fundamental error which vitiates this section of the Determination. The Authority’s assertion that the evidence regarded as necessary by the Merger Guidelines had not been submitted is said to be factually wrong because some such evidence had been adduced. In any event, the Authority ought to have undertaken its own investigation. Evidence had been supplied to the effect that Greenfarm, a traditional supplier of poultry cooked meats had begun supplying ham to Tesco. The Authority was also wrong to accept the evidence of suppliers referred to at para. 6.27 because it overlooked the fact that both categories of product are produced by Breeo at its plant in Tallaght.
8.9 The Authority rejects these criticisms and insists that its approach is in conformity with that outlined in its Merger Guidelines. It draws attention to the primacy of demand side substitutability in this regards as stated at para. 2.4 of the Guidelines:
“The substitutability of products is looked at primarily from the standpoint of consumers (demand-side substitutability), but (potential) suppliers (supply side substitution) may also be examined.”
8.10 The Authority maintains that this approach is consistent with that of the European Commission which, at para. 14 of its Notice on market definition states: “The competitive constraints arising from supply side substitutability other than those described at paras. 20 – 23 and from potential competition are in general less immediate and in any case require an analysis of additional factors. As a result such constraints are taken into account at the assessment stage of competition analysis.”
8.11 The Authority accepts that the evidence in relation to Greenfarm and the production of both categories at the Tallaght plant was given but denies that this is persuasive evidence of supply side substitutability.
Finding of the Court
8.12 The dispute between the parties on this issue appears to the Court to be a good example of the type of question upon which the Court will not intervene provided the criteria outlined earlier in this judgment as to the standard of review are clearly shown to be met. As is illustrated by the summary of the submissions set out above and the arguments exchanged with the Authority on foot of the Notification and Assessment as summarised at paras. 6.2 – 6.9 of the Determination, this is an instance in which it may well be possible to make plausible judgments as to the existence of one or two product markets depending upon where the weight or significance is allocated in the areas of evidence and data relied upon by the two sides.
8.13 The Court considers that, notwithstanding the absence of any conclusive or even usable evidence derived from the application of a SSNIP test, the Authority did have available to it a sufficient extent of evidence and data upon which it could reasonably base a judgment and provide adequate reasons for the conclusion it reached as to the existence of two separate product markets.
8.14 It must be borne in mind that in deciding whether products fall into a single or into two distinct product markets, the question that is sought to be answered is to what extent consumers of such products are willing and able to switch from one to the other. While it is both appropriate and necessary that economic mechanisms such as the SSNIP test be applied and that attempts be made to measure the elasticity subsisting between two sets of products, the issue is ultimately one for informed judgment by reference to all of the relevant factors likely to influence the choice of such customers including, obviously, the basic characteristics of the products themselves, where they come from, how they are produced, the uses to which they are put, how they are presented or marketed and how, ultimately they are regarded by consumers.
8.15 In a assessing, therefore, the soundness of the judgment made by the Authority in the present case, the Court notes that it had available to it the following factors upon which to make its judgment:
(a) It was comparing two sets of foods; chicken and turkey products with ham and beef. Both are meats and may therefore be said to have a certain functional interchangeability as such;
(b) While not expressly mentioned as such in this Determination, it is obvious that the two types of product are of different origin, the one being fowl and the other largely pig meat with some beef; and must therefore arrive at food processing stage by different routes in different time scales and with different costs;
(c) Both sets of products would appear to be sold for the most part in the same departments of retail outlets when in pre-packed format and while presented in adjacent cabinets or shelving would not appear to be intermingled in the way natural and processed cheeses were seen to be treated;
(d) Significantly, perhaps, and with one exception, the brands belonging to the notifying parties appeared to be employed exclusively for one or other category. Thus the Denny brand is used by Kerry exclusively for non-poultry cooked meats and the Ballyfree brand exclusively for poultry; the Galtee and the Roscrea brands were used by Breeo exclusively for non-poultry and only its Shaws brand was deployed in both;
(e) Shaws brand had only a minimal market share of 4.2% by value in the poultry section in 2007 and a somewhat larger one in the non-poultry section (Tables 6.4 and 6.8).
(f) There appeared therefore to be evidence that consumers tended to identify particular brands with one type of meat and it was noted that for this reason Kerry’s attempts to transfer a brand name between types had been unsuccessful; [DET. 6.153 and foot-note 94].
(g) Evidence in the form of a precedent in the consideration of the question elsewhere appeared to be meagre and inconclusive [DET. 6.4 and 6.6].
8.16 The Court considers that these factors, while comparatively slim when compared with evidence employed by the Authority in similar exercises elsewhere in the Determination, are sufficiently supportive and coherent to justify its judgment in favour of two separate markets.
8.17 This assessment is not, in the Courts view, fatally undermined by the arguments advanced on behalf of the appellant.
8.18 The Court notes that, as the Authority observed, the Compecon Report was itself equivocal on the issue of the product market in this area. Its submission said:
“The EU Commission, for example, has identified separate product markets for processed pork, beef and poultry. It has further indicated that there may be narrower product markets, although it has not made a definitive decision on that point. It should also be noted that Kerry markets its poultry and non-poultry cooked meats under different brand names. It has also indicated that these brands are identified with those specific meat types and that attempts to transfer a brand name associated with one type of meat to another have not proved successful because consumers identify the brands with the particular type of meat. This might suggest that poultry and non-poultry meats constitute distinct product markets.”
8.19 First of all, it is argued that the Authority’s conclusion in favour of separate markets is simply wrong on the evidence, because the separate use of brands for the two types of product cannot be taken as decisive since it reflects a particular branding strategy and is at variance with the use of the Shaws brand. Contrary to the appellant’s submission, however, that item of information is not treated as being decisive notwithstanding the absence of conclusive results to a SSNIP test but is clearly taken by the Authority as one of a number of factors the cumulative effect of which is to incline its judgment in favour to two markets. Contrary to the appellant’s submission, the Court considers that such a factor especially when taken in conjunction with the evidence that consumers tend to identify a type of meat with a particular brand, is something that can reliably be considered by the Authority.
8.20 While the Court has already indicated the caution it feels ought to be exercised when drawing inferences and conclusions from extracts from internal documentation it is undoubtedly the case that, as indicated by the guidance from the International Competition Network [DET. 6.22], where such documentation is adequate and pre-dates a proposed merger, it can constitute a persuasive basis for judging how the parties themselves view their own place in particular markets. While there is some similarity with the documentation reviewed at paras. 8.20 – 8.21 in the cheese sector, the documents examined in this sector at paras. 6.19 appear to be marginally more supportive of the Authority’s construction as they appear to be more directly aimed at the competitive positions of the brands in question particularly as regards the brands in the “white meat sector”. Again, however, the material is not taken as decisive but as an element amongst the number of factors considered.
8.21 Accordingly, the Court considers that even if it might be said that the Authority is taking the most favourable view of the information at its disposal, the Court does not consider that it has gone beyond the margin of judgment which it is accorded in such matters and has not committed any obvious or significant error of assessment in respect of the material before it.
8.22 As regards the main attack upon this market definition, however, it is of course true that the Determination does not rely upon a supply side substitutability test and none was conducted. Nevertheless, the Court does not consider that the conclusion is fatally undermined by any of the points raised against it. It is said that para. 6.26 is factually wrong because “some evidence” was given in the form of the reference to the position of Greenfarm as an operator in both alleged markets. The Court considers that this assertion reads too much into the sentence in question. The Authority there appears to be referring to the wider extent of evidence which is required to assess feasibility of supply side substitutability as set out in the preceding paragraph by reference to para. 2.12 of the Merger Guidelines. In substance the statement is in that sense correct and even if it could be said that it overlooks the reference to the Greenfarm information, the Court does not consider that it is so serious a mistake (if such it be) as to vitiate the conclusion. In effect, the “evidence” in relation to Greenfarm appears to be confined to the proposition that the company had commenced to supply hams to Tesco with no further information as to the extent or significance of that initiative as a move designed to seek a competitive position in pre-packed non-poultry cooked meats based upon the establishment of a permanent new line of production.
8.23 In any event, the Authority is not obliged necessarily to carry out a supply side substitutability investigation as the appellant argues or to carry it out at the stage of defining the processed market. Depending upon the circumstances of the market in question and its view of the reliability of other factors and especially the evidence of demand side substitutability, it is for the Authority to decide whether it is appropriate, necessary or feasible to do so. The importance of demand side substitutability at this stage is that it seeks to identify those products which are currently available to consumers and to which consumers may turn as acceptable alternatives in response to a price increase by the merged entity. Supply side substitutability is concerned with products not currently available but which might become available if existing available production facilities were switched to producing the relevant products as a response to the hypothetical price increase. Because the possibility of such additional supplies coming into the market can operate to constrain the exercise of enhanced market power by the merged entity, it is also possible to take it into consideration at the stage of examining the competitive effects of the merger.
8.24 It is true that the Determination at para. 2.7 acknowledges that the Breeo Tallaght plant is an example of a facility which receives the products in cooked form and, with 5 lines running at 75% capacity, produces 75 tons per week of ham, corned beef and white meats. Logically, therefore, such a plant could supply increased quantities of any one of those lines by use of its spare capacity within a short time. This however does not constitute evidence of the existence of equivalent sources of increased supplies from similar plants outside the ownership of the merging companies. Moreover, the analysis and conclusion of the Authority turns ultimately not upon the constraint that might derive from additional supplies of cooked meats as such but from credible alternative brands.
8.25 The Court therefore considers that no serious error has been made in determining that the cooked meats sector is divided into separate product markets for poultry cooked meats and non-poultry cooked meats.
9. The Common Grounds of Appeal
9.1 As indicated in section 6 above, the appellant raises a broad and detailed challenge to the validity and legality of the analysis made by the Authority of the effects of the acquisition on competition in the three contested product markets. In the light of the Court’s ruling on the definition of the product market in cheese, the Court’s examination of these grounds will now be confined to the remaining two product markets for rashers and N-PCM and, in keeping with the approach adopted by the appellant in its submissions, will be primarily directed at the former of those two markets.
9.2 It will be recalled that the main conclusion upon which the Authority’s negative determination rests in relation to these two markets is that the acquisition by Kerry as owner of the leading brands in each of Breeo, as owner of the second ranked brands and thus its closest competing brands, would result in a highly concentrated market in which private label products would not provide sufficiently close competition to constitute a constraint upon the merged entity and in which the buyer power of retailers would not permit them to credibly threaten to discipline the merged entity against a sustained price increase.
9.3 Thus, in the market for rashers the Authority concluded [DET. 5.97]:
The market shares of Galtee and Denny have been stable over the period 2005 – 2007;
These two brands are each others closest competitors in the market and post-acquisition would account for 47% of the market;
There are no credible alternative brands which would enable retailers to constrain the merged entity from raising prices;
The combined market share of other brands is only 6% with only one brand being in excess of 1% (Cookstown);
New entrants would be unable to establish a sufficiently strong presence within a two year period so as to act as a constraint on the merged entity;
Notwithstanding a combined market share of 36% by value, private label products are not considered by retailers to be sufficiently close competitors to Denny and Galtee which must be stocked as “must have” brands;
Retailers will be unable to discipline the merged entity in the absence of credible alternative branded rashers; of entry of sufficient branded rashers to the market within two years; and because private label rashers would be unable to replace the merged entities’ products as they are not sufficiently close competitors.
9.4 Similarly, in the N-PCM market it found at para. 6.164 that:o The merged entity would have 48% of the market by value post- acquisition with other brands having only 5.8% and none having more than 1%;
o Kerry and Breeo’s combined market share has increased from 2005 – 2007 from 44.7% to 48%;
o The Kerry and Breeo brands are each others closest competitors in the market;
o Private label share has declined from 50.9% in 2005 to 46.2% in 2007. Retailers do not consider private label products to be sufficiently close to the Kerry and Breeo brands;
o There are no credible alternative brands in the market which could credibly act as a constraint;
o New entrants would be unable to establish a sufficiently strong presence within two years to act as a constraint;
o Retailers do no have sufficient countervailing buyer power to act as a constraint post-acquisition because of the absence of alternative branded suppliers;
o The unlikelihood of the entry of branded alternative suppliers within two years; and the fact that private label products are not considered sufficiently close competitors to act as a constraint.
9.5 The appellants submit that these conclusions were wrong for the following main reasons:(a) Private label products are not distant competitors but direct and potent competitors to the Kerry and Breeo products;
(b) The multiple retail chains representing the bulk of all sales by the notifying parties of the products in question have very significant countervailing buyer power which the Authority has misjudged;
(c) There are many other suppliers ready and able to enter the market and they will clearly do so in response to the hypothetical price increase by the merged entity.
9.6 In support of these grounds of alleged error, the appellants advance a number of detailed arguments as follows:
A – The Close Competition Issue
(i) In finding private label a “distant competitor” the Authority wrongly ignores or misinterprets the econometric evidence and in particular it wrongfully refused to accept and consider the evidence of Dr. Hogan based on his analysis of the Tesco EPOS data;
(ii) It misjudged the nature of the competitive relationship between the branded and private label products and in particular misjudged or misinterpreted the function and significance of the price differential between the two categories as the basis upon which private label competes. It failed to consider the volatility of market shares in response to price discounts and promotions and the feeble character of brand loyalty in the markets;
(iii) It erred in finding that private label’s gain in market share was at the expense of other brands rather than those of Kerry and Breeo;
(iv) The Authority was selective and inconsistent in its use and rejection of evidence and data;
(v) It wrongly analysed, misjudged and exaggerated the role and importance of brands and wrongly attributed “must have” status to the Kerry and Breeo brands;
(vi) The Authority erred in its analysis of market shares by considering only those of the branded products in disregard of the shares attributable to the private label brands which account for 45.9% by volume of the market in rashers and 46.2% in N-PCM;
(vii) The Authority ignored evidence to the effect that since 2005 shelf space allocated by retailers to private label had increased at the expense of that allocated to brands which constitutes persuasive evidence that PL products are substituted for brands;
(viii) The Authority failed to consider the significance of the rapid growth in market share by the two discount chains who stock no brands at all.
B - Countervailing Buyer Power(i) The Authority erred fundamentally in its assessment of the ability of retailers to exercise countervailing constraint upon the merged entity in both of these markets;
(ii) It ignored or failed adequately to consider the evidence that the retailers do in fact have recourse to alternative supply arrangements;
(iii) Important international suppliers can and do enter the market and attract substantial market share;
(iv) The Authority was wrong to consider the undoubted buyer power of the retail chains diminished by the “must have” character of the appellant brands;
(v) It erred in failing to consider and give due weight to the concrete evidence of the actual exercise of de-listing and other buyer power tactics on the part of the retailers;
(vi) The Authority misconstrued the effect of the evidence available to it from both retailers and competitors as to the significance of private label and on the question of brands as “must have” products.
C – Entry and/or Expansion of Competitors within Two Years:
(i) The Authority applied the wrong test in assessing the potential constraint from entry or expansion of competitors in the rashers market within two years by dismissing the prospect on the basis that there was no evidence that two large international suppliers (Vion and Tulip) had specific plans to enter the Irish market.
(ii) The conclusion that a new entrant would not be able to establish a sufficiently strong presence in the rashers market within that time was contradicted by the fact that the Vion and Tulip brands had already sold on promotion and that when they do so (in Tesco) they can quickly gain a 30% share of Tesco sales or 6% of total pre-pack rasher sales;
(iii) The conclusion also ignores available evidence to the effect that in the rashers market new brands can be established well within that period;
(iv) The conclusion given at para. 5.96 in respect of rashers and at 6.151 in respect of N-PCM to the effect that “other brands” or “third party brands” are not sufficiently close competitors but are distant competitors to the brands of the merging parties is patently wrong and supported by no evidence available to the Authority or by any analysis conducted by it. As a result the Authority conducted no investigation as to the capacity of those other brands to expand.
Finding of the Court
9.7 The Court considers that there appears to be one finding made by the Authority in each of these product markets which was correct, which is not seriously disputed by the appellant and which logically serves as a starting point for appraisal of the above arguments. It is the finding that in each market the branded products sold by Kerry and Breeo are respectively the closest competitors to one another.
9.8 Thus, in the market for rashers the Kerry market share of 15.2% by volume and 22% by value in 2007 is comprised almost entirely of its products under the Denny brand (14.1% and 20.6% respectively). The Breeo totals are 16.6% by volume and 24.8% by value and again are made up almost entirely of the Galtee brand at 13.1% and 20.1% respectively [DET. Table 5.1].
9.9 In that market the market share totals for “other brands” are 6.8% and 6.3% respectively. Private label accounts for a total of 45.9% by volume and 36.4% by value within which the largest component market share is that of SuperValu with 19.2% and 13.2% respectively followed by Tesco with 10% and 8.2%. (Table 5.1)
9.10 In the N-PCM market the Denny brand was again in first place with 33.3% by value in 2007 with Breeo’s three brands accounting for 14.7%. The total attributable to private label by value in 2007 is 46.2%. (Table 6.4)
9.11 In each case the internal documents of the parties examined by the Authority appear to confirm that the two undertakings saw each other as closest competitors in the markets [DET. 5.56 and 6.129] although the Kerry documentation also showed that it considered private label products to be a growing threat.
9.12 The Authority considered that this assessment of the respective closeness of the Breeo and Kerry brands to one another in each of these markets was confirmed by the results it received from its expert Dr. Walsh [DET. 5.57 and 6.130]. He concluded that the Breeo and Kerry brands exercised far stronger influence upon one another compared with the weak influence of other brands. Thus the Breeo brands were a far greater competitive constraint on Kerry and vice versa. Those results were, of course, disputed by Dr. Hogan and particularly upon the ground that the A.C. Nielsen data upon which they were based did not include data for private label at the level of SKU (individual stock keeping units). (See para. 9.34 below in relation to the differences between the two experts on this data.) Whatever may be thought of the validity or otherwise of the technical differences in the respective approaches of Dr. Hogan and Dr. Walsh to the estimates of the elasticities between these products, it is clear that they are mainly concerned with the problems of measurement of the comparative influence of private label products upon the Kerry and Breeo products as compared with the influence of other brands. A resolution of those differences does not alter the fact that, quite clearly, in each of these markets the Kerry and Breeo products are close competitors of one another as, indeed, the appellants acknowledge at, for example, para. 258 of the written submissions.
9.13 It follows, accordingly, from that fact and from the respective market shares of the first and second brands in each market that the acquisition must necessarily result in a significant structural change in each of those markets. Whether or not a Breeo brand is the single closest competitor to a Kerry brand in one or other market, it is sufficiently close such that the vesting of both brands in a single undertaking which brings about joint market shares of 47% and 48% respectively, must necessarily other considerations apart, lessen competition in those markets and enhance the market power of the merged entity. Apart from anything else, the ability of retailers to play Kerry off against Breeo and vice versa is obviously diminished. It is true, of course, that the brands do not disappear and a retailer could, in theory, threaten to de-list Galtee but not Denny but the force and credibility of such a threat is obviously materially altered by the structural change in the market. A retailer may not consider each to be a “must have” brand but still be reluctant to do without both.
9.14 Accordingly, the structural change leading to a marked increase in concentration of each market would, if no other factors are involved, lead to a significant lessening of competition. The Court therefore considers that the essential issue raised by the appellant’s arguments turns upon the question as to whether or not the Authority has made any significant error, as alleged, in its prospective analysis of the effects of that change upon competition. In particular, having regard to the peculiarity of the retail grocery sector in which the four main multiple chains are in the unusual position of collectively purchasing up to 80% of the relevant products sold by the merging parties while fixing the retail price for those products and at the same time competing as suppliers and retailers of substitute products, has the Authority correctly assessed the possible constraining effects of private label products as substitutes and of countervailing buyer power?
9.15 In examining whether the conclusions reached by the Authority on the question of close competition between the appellant’s brands and the private label products (that is, the conclusion reached at para. 5.79 as regards rashers and at 6.142 as regards N-PCM,) it is important to bear in mind that the Authority is there looking at the possibility that the availability of the private label products as substitutes will by itself be sufficient post-merger to deter a sustained price increase in those products by the merged entity. As will be mentioned later, this is in fact but one of a number of interconnected factors falling to be analysed which bear upon the predictable impact of the merger on competition in these markets.
9.16 In considering the appellant’s arguments in this regard the Court considers that it is important to bear in mind that, contrary to the impression that might be taken from the way in which the argument is presented by the appellant, “private label” is not a homogeneous product or a single source of competition to the brands in either market. Clearly, the private label category is comprised of full ranges of own label products in both markets offered by each of the four main multiple chains.
9.17 It is not disputed that these product ranges are devised and marketed in a way which mimics both the pricing and the presentation formats (copycat marketing) of the branded ranges. But while the retail chains claim to match the brands for quality especially at the premium end of each range, the fundamental rationale of the private label business model and the basis of its marketing position is that the private label products compete primarily on the basis of the important average price differential which is maintained between the brands and private label products. This is, of course, subject to exceptions in that some of the own label products acquire the status of brands in their own right and some particular units may be more expensive than the corresponding branded offering. Also, periodic discounted promotions may have the effect of greatly reducing or even eliminating the actual price differential.
9.18 Again, there is no doubt and it is not disputed, that from time to time but apparently for short periods, retailers mount deeply discounted promotions (“40% extra” in a package) and that consumers do respond so that brands lose sales and private label products gain accordingly. In that sense there is clear evidence that when a retailer so decides, it can use its range of own label products in a particular market as a means of close, direct and effective competition with a corresponding branded offering.
9.19 In assessing the closeness of competition between the branded products and private label as a realistic future competitive force in reaction to the acquisition , the issue is not whether private label can in principle so compete but whether it can be predicted as a matter of reasonable probability that a sustained price increase by the merged entity in either of these markets will be defeated by a sufficient switch on the part of consumers to the private label products on the assumption that they continue to be marketed as at present and according to the existing business or marketing models.
9.20 In challenging the approach of the Authority on this issue, the Court considers that the appellant has overestimated the role of private label as a category and overlooked the fact that it is not a homogeneous product or a single source of competition. It also underestimates the importance of the fundamental feature of the retailers business model that, as compared with the different business model of the discounters, the retailers seek to stock both their own private label products and the branded products including most of the range of each branded product which they know from experience will continue to be demanded by consumers.
9.21 Retailers may periodically either as a promotion or as a tactic in negotiations with suppliers, de-list, de-range or reduce purchases of or support for one or more brands, but there is no evidence of any inclination on the part of a major retailer to forgo an entire branded product line on a permanent basis in order to fully substitute its own equivalent private label range. It is important to distinguish this assessment of closeness of competition from the short-term use of delisting, etc. as a buyer power tactic to discipline suppliers.
9.22 In that regard it is clearly relevant that the competition is not between brands on the one hand and private label on the other. Within the private label category the four rival ranges of each set of products are in active competition with one another as part of the overall vigorous competition between the multiple chains themselves. Thus, attributing a single competitive role to “private label” as such, risks overlooking the important factor that so long as particular brands such as Denny and Galtee enjoy important market shares which indicate significant support from consumer demand, individual multiple chains may have little incentive to abandon that feature of the business model by dispensing with brands as such in favour of full substitution by its own label products in the important markets of rashers and cooked meats.
9.23 The Court considers that this assessment is, for example, supported by the evidence derived by the Authority from its survey of retailers. Thus at para. 5.23 it records that six of the nine retailers questioned indicated that branded products as such were essential to them. Two of the remaining three were discounters and thus do not carry brands at all and the third retailer sells only private label products. (See DET. footnote 51 – 5.23)
9.24 Thus, even if it can be said that it is possible with a particular business model such as that of the discounters to trade successfully without, say, Denny rashers or Galtee cooked ham and to gain and retain market share for non-brand or own label products, this does not render erroneous the Authority’s judgment that, in effect, in the normal course of trade over the medium and long term, the four main multiple chains can be predicted to continue to stock and to deal with branded products according to the existing pattern including the selling of the private label ranges at a significant average discount to and alongside branded products.
9.25 It is to be noted that the evidence available to the Authority included the fact that retailers do not price their private label products (at least primarily) by reference to the cost to them of the product plus a margin, but by fixing both the retail price of those products and of the branded products so as to create a differential which optimises the overall return from both lines. The Court considers that in assessing prospectively the potential constraining effect of private label sales by themselves and independently of the other mechanisms of influence available to retailers against the merged entity, the Authority was correct to assume that the current business models including the maintenance of the average price differential would be maintained.
9.26 It is undoubtedly true that, having regard to the peculiarities of the market structure and notably the “gate-keeper” role of the multiple chains, the latter could, if they chose to do so, depart from that pricing structure, abolish the differential and thereby make their private label products direct and close competitors of the branded products. It is also true that the evidence of discounted promotions demonstrates that higher levels of substitutability can be achieved between brands and private label products but the sales lost to the brands and gained by the private label are only lost and gained so long as the retailer chooses to continue the promotion. When it ceases, the evidence indicates that the pre-existing pattern resumes. This ability of the retailers to manipulate substitutability is clearly a factor to be considered in assessing the significance of countervailing buyer power but so long as the two product markets under consideration continue to be characterised by the deliberate differentiation between the branded and private label products, the Court considers that it was open to the Authority on the evidence available to it to assess the predictable effects of the merger on competition in these markets on the basis that this model would continue.
9.27 It follows that the Court does not consider that the soundness of the Authority’s conclusions on this issue namely, that the availability of the private label products in these two markets will of itself and independently of other countervailing factors deter the hypothetical price increase, is not undermined or contradicted by the specific allegations of error alleged by the appellant.
9.28 Thus, it is alleged that the Authority erred in considering that Denny and Galtee were “must have” brands. It is true, as the appellant points out, that the discounters trade successfully in these product markets without stocking either brand. It is also true that, in theory, each retail chain could decide to dispense with Denny and/or Galtee pre-packed rashers, cooked ham and beef and offer only their own label products or increased and permanent quantities of international brands that have been hitherto brought in for promotions. The question is not, however, whether one or more retailers have the capacity and the available alternatives to do so but whether it is predictable that they would do so in response to a hypothetical price increase by the merged entity which is considerably less than the average differential maintained between the two categories of product historically. Having regard to the degree to which the stocking of the two leading brands in these markets contributes to the established business model and having regard to the ability of the retailers to accommodate such a price increase by adjusting the differential in the respective retail prices, the Court considers that the Authority was entitled, consistently with its assessment of the available evidence, to conclude that the availability of private label products as competitors in the two markets would not be of sufficient influence taken by itself to act as a sufficient constraint on a post-merger price rise.
9.29 The appellant also challenges two particular points made by the Authority in its analysis of the closeness of private label as a competitor. At para. 2.46 referring to the price differential between brands and private label the Authority observed:
“The price difference reflects the consumer’s perception of the relative value of one against the other depending on differences in quality and other characteristics.”
This is said to be an important error because it is a finding for which there is no evidence whatsoever. This, however, is not as such a finding of fact on the part of the Authority but an inference which it draws from the undoubted fact that the average price differential is maintained. If, as the appellant argues, the private label products are marketed as equal in quality and so sophisticated in the way they are presented as to be fully substitutable for the branded product, then it must necessarily follow that the consumer nevertheless continues to be willing to pay the higher price because the branded product is perceived to be in some respect superior.
9.30 The appellant also attacks the statement at para. 5.60:
“If private label is a close competitor it is not clear why it should need to discount its price, on average, substantially compared to the brand.”
It is alleged that this demonstrates a failure on the part of the Authority to consider or even mention the reason why private label is so priced at a discount and that it is self- contradictory to say that because the private label brands compete on price they are not therefore close competitors of the branded product. However, as the Authority has correctly submitted, if the consumer did in fact view the private label product as fully comparable with and substitutable for the branded product, there would be no justification for such a significant discount. It is because private label is not viewed as a full substitute that private label can retain market share on a stable basis only by maintaining such a significant discount. The fact that when the deep discount promotions end, pre-existing market share levels are resumed would tend to confirm the necessity of the price differential to the maintenance of a stable private label market share.
9.31 As already indicated, the appellant strongly criticises the Authority in relation to its use of and interpretation of the econometric evidence in evaluating the closeness of competition between the private label and branded products and in assessing the competitive constraint which would be faced by the merged entity both from other brands and, more importantly, from private label. In particular, the appellant relies strongly on the proposition that Dr. Hogan’s analysis based upon the Tesco EPOS data constituted evidence that in the case of Tesco, the private label products taken with other brands, represented a closer competitive influence on the Kerry brands than those of Breeo. It alleges that the Authority erred in dismissing this evidence without consideration and giving no reason for doing so.
9.32 It is not disputed, of course, that the Authority did indeed reject Dr. Hogan’s final analysis based upon the Tesco data but, contrary to what is alleged, it gave as its reason for doing so the fact that it agreed with the objection raised by its own expert Dr. Walsh that data from one retailer, however important, could not be relied upon in order to form a view of the market as a whole [DET. 2.87 AND 5.16].
9.33 It should be explained that during the notification period a number of reports had been exchanged between Dr. Hogan and Dr. Walsh with various issues and criticisms being raised on either side as to the methodology used, the reliability of the data and the correctness of some of the formulae employed [DET. 1.38: 2.77-2.87: 5.57: 5.77: 6.130].
9.34 The original data used by Dr. Hogan for his analysis was based upon market data from A.C. Nielsen. Having regard to the significance of the issue as to private label as a close competitor or not, this data had the disadvantage that it contained only aggregated figures for private label sales without any breakdown by reference to products or stock keeping units. Nor did it include data for “loose sales” – product sold over the counter (OTC) at butchers or deli departments. The two experts debated how to overcome the absence of data on private label. Dr. Walsh considered that engaging in conjecture would be unscientific. Dr. Hogan felt it was necessary and reasonable to make some assessment of how the estimates for elasticities would increase, particularly in the areas of cooked meats and rashers where the private label category was higher. They also argued over a technical mistake. Had Dr. Hogan employed a wrong formula in his initial report and was the segmentation correct that is, had the correct judgment been made as to the role of the package size in purchasing decisions by consumers?
9.35 In an attempt to overcome the deficiencies in the A.C. Nielsen-based estimates due to the absence of private label data at SKU level, in a final report dated 13th August, 2008, Dr. Hogan estimated the own-price and cross-price elasticities for Kerry products using EPOS data from Tesco with a view to measuring the extent to which Tesco private label products were a substitute for Kerry branded products. Dr. Hogan considered the Tesco consumer data to be broadly representative of the market as a whole because it included the actual sales of all products at all check-outs in the entirety of Tesco stores in the State. His conclusion was as follows:
“This data shows unequivocally that private label was a close substitute for Kerry products in three markets (cheese, non-poultry cooked meats, rashers). In all segments bar one, Tesco private label is a closer substitute for Kerry than is Breeo. The one exception is rashers. In this case, Tesco private label is a close substitute – only not quite as close as Breeo. However in this case, other branded and Tesco private label are jointly closer substitutes for Kerry rashers than are Breeo rashers.”
9.36 It is necessary to recall that the assessment which the Authority must make in this context concerns the constraining competitive effect of private label products offered by the four main retail chains. In relation to rashers, Dr. Hogan’s final report confirms Dr. Walsh’s assessment that in the rashers market Breeo produces the closest substitutes for Kerry products. It also concludes that “both Tesco private label and other branded products provide a degree of substitution” and that if both the other brands and Tesco products are aggregated, the joint price elasticity implies that together they are equal to Breeo in closeness of substitution for Kerry rashers. It does not necessarily follow, however, that if the same exercise was conducted in respect of the products of the other multiple chains that, for example, in Dunnes Stores the private label and other brands would also rank with the Breeo products as close substitutes to Kerry.
9.37 In the final report of 13th August, 2008, Dr. Hogan acknowledged and corrected the error in the formula used in the original report. Ironically, this error had originated in a misprint in the formula as published in an article written by Dr. Walsh himself. Dr. Hogan also adopted Dr. Walsh’s approach to segmentation while pointing out that he had originally thought it inappropriate because it had the effect of elevating the consumers decision as to size to the same importance as the decision as to product. Because the price elasticities were now being estimated separately for each segment, however, he considered this issue less relevant.
9.38 At para. 2.85 of the Determination the Authority explained that Dr. Hogan’s admission of the error in the formula which had been replicated in the estimates in the reports prior to 13th August, 2008, was its reason for disregarding his results as given in those reports. The issue that remains, accordingly, is whether the Authority erred in also failing to take account of the final report of 13th August, 2008. The Court considers that this was not an error on the part of the Authority which was entitled to rely on the advice of its own expert that data based upon results from one retailer alone (however important) could not be used as a basis for forming a view of the market as a whole. Moreover, and in any event, the additional assessment furnished by Dr. Hogan to the effect that within the Tesco group, the combination of its private label products with other brands could be regarded as equal in closeness of substitution to the Breeo products does not serve to render necessarily incorrect the Authority’s judgment that in these two markets as a whole, the availability of private label products taken by itself would not act as a sufficient constraint upon a post-merger price increase.
Countervailing Buyer Power
9.39 As has already been remarked, it is one of the prominent characteristics of the grocery market and accepted by the Authority [DET. 2.34] that a small number of multiple chain retailers account for 65% of all grocery products sold in the State. Furthermore, in the case of the food products of the notifying parties, these chains may account for something in the region of 78% of their sales of the relevant products while those sales represent only 2% of the total purchases of their customers. It is a further characteristic and unusual feature of the market, of course, that the retailers perform the “gatekeeper” role in which they fix the retail prices of the appellant’s products and control their access to shelf space while at the same time being competing producers of similar products whose prices they also fix at retail level. It is also not disputed by the Authority that in these markets the main multiple chains can exercise considerable buyer power in different forms which include not only de-listing entire brands from a particular supplier but also “de-ranging” or limiting the package sizes or formats of particular products within a range or refusing to support promotions, reducing purchases or moving products to less favourable display positions [DET. 2.35]. It is, on the face of it, not surprising therefore that this asymmetric relationship of dependence of the suppliers on the main retail chains can be readily exploited to the advantage of the latter in their dealings with suppliers.
9.40 Under this heading, accordingly, the issue which arises is whether the Authority has made its assessment consistently with the available evidence that the undoubted existence of this buyer power in these markets will nevertheless be curtailed post-acquisition by the existence of other factors such that it can be predicted with reasonable confidence that such buyer power will not operate so as to deter the merged entity from imposing a profitable price increase.
9.41 In examining the arguments on these issues it is appropriate to consider also the third source of possible constraint outlined above namely, the possibility of expansion by existing competitors or the entry of new suppliers because it is obviously a relevant feature of both markets that the multiple chains appear to be in a position to gain ready access to additional supplies of private label product when required and to procure the market entry of major international suppliers of branded product for the purpose of intermittent promotions.
9.42 In this connection the Court notes that in relation to the rashers market, the Authority accepts at para. 5.94 no problem is posed for the retail chains in having quick access to the necessary added supplies of private label product from expanded capacity in existing non-merging suppliers. Similar recognition is made by the Authority with regard to the N-PCM market at para. 6.161. Thus, so far as exercising buyer power to deter a price increase post-merger by de-listing or de-ranging or cutting back on purchases from the merged entity is concerned, the non availability of replacement quantities of own label product is clearly not a hindrance.
9.43 Secondly, although the Authority took the view that there would be no sufficient constrain from new entrants, it does so on the basis that such constraint would require brands with established reputations as such: in effect, brands that could play the “must have” role of Denny and Galtee [DET. 5.93].
9.44 While the Authority’s assessment of the time and investment required to attain such a position is undoubtedly something within its judgment as a matter of economic expertise, it nevertheless remains the position, as a matter of fact and evidence, that retailers have access also to ready supplies of branded products from international suppliers such as Tulip and Vion. The latter is of some significance in this context because since June 2008 it has been the owner of the Cookstown brand which is the leading brand in rashers in Northern Ireland and has an established, albeit minimal presence in the market in the State [DET. 5.86].
9.45 It is true that these brands are not “established” in the sense used by the Authority and their presence on the market to date has been occasional and for the purpose of heavily discounted promotions. Nevertheless, to the extent that retailers say that it is essential for them to stock “brands” as such in addition to their own label products, it is clear that those supplies are available to them such that it would be feasible as a matter of commercial practicality for them to resort to buyer-power disciplinary tactics without having to abandon, even temporarily, the business model which offers a branded product in juxtaposition to private label.
9.46 The fact that Tulip or Vion products have been hitherto invited in by retailers only for promotions at discounted prices does not contradict this assessment. It must be relatively easy for a retailer, should it choose to do so, to introduce such brand supplies on a different basis given that it is the retailer who controls the retail price in any event.
9.47 The significance of these factors in the present context is that in assessing the reality of buyer power it is clear that retailers have available to them, on the evidence, potential sources of ready supply of both own label product and brand products such that supply as such is not obstacle to the exercise of disciplinary buyer-power.
9.48 The Authority’s concerns in both markets on the question of entry and expansion are based primarily on the non availability of other branded products and this is so because it takes the view that retailers felt they could not credibly de-list the two leading brands, Denny and Galtee and that the entry of new brand suppliers would not be “likely, timely or sufficient”.
9.49 In the Determination at paras. 2.35 to 2.70 the Authority gives an introductory overview of the issues that arise in relation to countervailing buyer power and refers (at para. 2.51-2.63) to the evidence produced by the notifying parties of examples of exercise of buyer power by particular retailers. It then gives its conclusion at paras. 2.69 and 2.70 as follows:
“As explained above, the retailers have many bargaining strategies and levers pre-merger. There is no reason to assume that the retailers are not using those strategies and levers to the maximum extent possible at the moment. There is no evidence that retailers are not, so to speak, keeping any in reserve. Post-merger, the retailers will have less bargaining power as they will not be able to play off Kerry against Breeo. The issue thus becomes what options remain open to the retailer to exercise countervailing power post-merger.”
More detailed consideration of the topic is then postponed to the discussion in relation to the individual product sectors in the later parts of the Determination.
9.50 In the subsequent discussion of the rashers product market in section Five, no separate analysis is made of the possible exercise of buyer power tactics of de-listing, etc. but it is implicit in the conclusions expressed at para. 5.97 that the Authority came to the view that it was because it was essential for retailers to have the two “must have” rasher brands of Denny and Galtee that the exercise of those measures to deter a price increase could not be predicted. On the other hand, in the equivalent discussion in section Six of the N-PCM market, the Authority concludes that the retailers do not have a sufficient degree of countervailing buyer power to be able to credibly threaten to discipline the merged entity because, again, the view of the majority of retailers is accepted that they could not credibly threaten to de-list or de-range the two “must have” brands.
9.51 Although the issues are interrelated, the question that falls to be answered in considering the exercise of buyer power is distinct from that examined in the preceding section as to the closeness of private label products as substitutes for the products of the merging entities. In the earlier section the question is whether private label products, as they stand, are sufficiently close as substitutes in the perception of consumers such that a 3-5% price increase post-acquisition can be predicted, independently of any other reaction by retailers, to lead to a sufficient number of consumers switching to the private label products to render the price increase unprofitable. In the present issue, the question posed is whether it can be predicted that the retailers will in any event react so as to defeat such a price increase or whether there are other sufficient factors at play (the “must have” brand factor, for example) which will render the retailers unable or unlikely to exercise enough countervailing buyer power.
9.52 As already mentioned, the Determination recognises the existence of buyer power on the part of retailers in the Irish grocery markets [DET. 2.34] and that it can take a number of different forms [DET. 2.35]. However, it points out, citing the Guidelines of the European Commission that “it is not sufficient that buyer power exists prior to the merger, it must also exist and remain effective following the merger. This is because a merger of two suppliers may reduce buyer power if it thereby removes a credible alternative” [DET. 2.31]. In that regard, as has already been remarked, it is undoubtedly the case that this acquisition will effectively remove the ability of a retailer to play Kerry off against Breeo as separate suppliers.
9.53 In examining the basis upon which the Authority concluded that the exercise of such buyer power would be insufficient to operate as a constraint, it is first necessary to look at the direct evidence of the actual exercise of buyer power submitted by the notifying parties and summarised by the Authority at paras. 2.50 – 2.70 of the Determination. This comprised a set of email correspondence over the period 2003 – 2008 which the notifying parties claim proved some 30 instances of retailers exercising buyer power tactics. The Authority rejects the evidence in these terms:
“In general, the documents are inconclusive, either in that they do not tell us whether the threat to de-list was in fact carried out, or else no evidence is provided to show that de-listing was carried out for any motive other than the normal course of business.” [DET 2.51]
9.54 The Authority then examines five particular examples from the email correspondence and it appears clear that those considered in paras. 2.52, 2.53, 2.54, and 2.55 are open to the construction placed upon them by the Authority namely, that the contents of the emails concerned do not suggest that the motivation of the retailers for those particular threats was anything other than that stated namely, poor sales of the products in question or absence of consumer demand for such new products.
9.55 The Authority, however, accepts that the threat by Tesco to de-list Breeo products made on 15th August, 2008 [DET. 2.56] “would certainly be consistent with the exercise of buyer power since Tesco is using the threat to de-list to get better financial terms”. The Authority asked Breeo whether the de-listing in question had in fact been carried out. Breeo said that it had not because they had informed Tesco of their inability to negotiate terms pending the outcome of the Kerry/Breeo merger and Tesco then took a decision to defer the de-listing until 1st September, 2008.
9.56 The Authority then considers the value of evidence related to conduct which takes place in the final stages of a merger determination process. It concludes:
“Any evidence about marketplace behaviour occurring after the merger is negotiated or announced must be subject to careful scrutiny as to whether the behaviour may be influenced or impacted by the positions taken by the players in the market about the proposed merger. Such evidence is clearly unreliable especially in considering key issues in a merger review such as market definition or competitive effects of the merger. In general little weight could be given to such evidence.” [DET. 2.63]
9.57 Insofar as this appraisal is put forward as a reason for rejecting the implications of the Tesco threat of 15th August, 2008, the Court considers that it is unwarranted. While it is true that Breeo pleaded the existence of the merger process as a basis for persuading Tesco to defer its threat, there does not appear to be any basis for supposing that Tesco’s threat to de-list 23 product units in the breakfast meats category was in any way influenced by that process or had any objective other than the one stated namely, to procure a 4.54% improvement in margin.
9.58 The significance of the example is that it is direct evidence of pre-acquisition exercise of buyer power and if such a tactic can be employed pre-acquisition in order to obtain an improvement in margin by a retailer there is no reason for supposing that it might not be equally exercised post-merger in order to deter a price increase. It is true that, post-acquisition, Tesco would be dealing with a single undertaking as owner of both leading brands, but the “must have” character attaches to the brand and not to ownership of the business.
9.59 In any event, other examples in Annex 6 of the notifying parties’ response to the Assessment appear to corroborate the degree to which the main retail chains routinely consider themselves in a position to dictate to suppliers the terms upon which a product will be purchased from them, stocked and resold. An email from Dunnes Stores to Kerry of 12th March, 2007, indicates that a new product line from Kerry would only be accepted if an existing line is first withdrawn. Similarly an email of 27th June, 2005, confirmed no new lines would be listed until others had been withdrawn. A similar policy in relation to de-listing to make way for new lines is evident in Tesco from the email of 22nd June, 2007. Given that it is Tesco that sets the retail price for a product, doing so by reference to the differential it seeks to maintain with its own products, it is notable that in an email of 3rd January, 2006, this retailer refuses to list a number of Kerry cooked meat lines because the margins are unacceptable.
9.60 A further series of emails from, amongst others, Dunnes and SuperValu illustrate blunt refusals to accept price increases sought by Kerry or Breeo. A letter of 30th July, 2007, from Musgraves to Kerry in relation to a request for price increases is not untypical of these. The Musgrave Dairy Trading Manager “insists” on the following:
• A detailed explanation and justification for the increase
- a minimum of 8 weeks notice on official company headed notepaper;(The Court notes in passing that the propriety of the demand for “receipts from the competition” must be questionable.)
9.61 Finally in a letter dated 17th December, 2005 Tesco’s Commercial Director referring to the Christmas period writes to Kerry in these terms: “In the light of our policy of Getting Cheaper for Customers and focussing on service to ensure we create value for customers, we will not be in a position to discuss any proposals for price increases of any kind for the time being.”
9.62 It may be remarked that these exchanges are perhaps typical of tough negotiations between supplier and customer with the latter seeking to strike the best possible bargain. Nevertheless, taken as a whole and having regard to the facility and short notice with which the retailers are prepared to de-list and de-range particular lines as a weapon in that negotiation, the material is direct evidence which contradicts the proposition that the retailers cannot credibly de-list brands such as Denny or Galtee because of their “must have” character. If retailers consider these to be “must have” brands post-acquisition they must be “must have” brands before it as well. The Court therefore considers that the examples extracted from this evidence and examined at paras. 2.50 – 2.63 of the Determination for the purpose of dismissing the evidence, are selective and unrepresentative of the probative force of annex 6 to the written response taken as a whole.
9.63 It may also be true, as the Authority observed, that some specific threats are not shown to have been carried out (although the evidence of the appellant is that they were,) but the significance of the evidence in the context of the present issue is that the content, tone and frequency of the threats clearly demonstrates that recourse to such buyer-power tactics is a routine and important feature of the relationship between suppliers and the multiple retail chains in the consumer foods sector of the grocery trade. It also demonstrates that the first reaction of all retailers to a supplier’s notification of a price increase is to oppose, resist and place obstacles in the way of such requests. Thus, it is clear that in this sector no supplier can ever expect to be able to “impose” a price increase and certainly not to do so without serious delay, negotiation and justification. The Court is satisfied therefore that the correspondence in Annex 6 is compelling, direct evidence of the fact that retailers as a matter of invariable policy resist, oppose and delay all attempts by suppliers to introduce price increases.
9.64 It is next necessary to examine the second factor upon which the Authority’s conclusion as to the inadequacy of countervailing buyer power post-merger is based namely, the evidence that the buyers in question do not consider that they can credibly resort to buyer power tactics to deter the hypothetical price increase and whether this negates the weight of the above evidence.
9.65 The material is, of course, “opinion evidence” as it is described by the appellant. It may be particularly important, therefore, to exercise caution as to the inherent probative value of responses to retailer surveys in this area when regard is had to the fact that the question is posed by a competition authority to operators in a sector which may have become accustomed to the critical scrutiny of regulatory authorities. (The Court notes that in 2008 alone the Competition published three “Global Monitor Reports” on its detailed examinations of the structure and operations, on price trends in and on the application of retail planning systems to the grocery sector.)
9.66 In its surveys of retailers in the rashers market, six out of nine retailers questioned indicated that it was essential to have branded products as such within their offerings of rashers and five considered the Denny and Galtee brands to be “must have”. The three remaining retailers did not carry brands at all. The Authority then posed the following somewhat leading question in its follow-up questionnaire to the six:
“Do you have the ability to credibly threaten to de-list and/or de-range the merged entity’s rasher products post-acquisition and switch to other rasher brands such as Clonakilty/Olhausen/Country Style and/or own label rashers? If not, why not?”
The Authority analyses the replies furnished by the six retailers at paras. 5.25 – 5.45 of the Determination. Two of the retailers, customers A and C expressed concern about the impact of the proposed acquisition in this market and said there were no credible alternatives to the Denny and Galtee brands and neither considered private label to be a strong competitor. At para. 5.32 customer F which did not have its own private label rashers, is recorded as having expressed concern about the impact of the acquisition but no further explanation is forthcoming. Customers A and C both express the belief that they could not credibly de-list the brands of the notifying parties. (The concealing of the identity of the customers has the disadvantage of precluding the veracity of these claims being judged against the actual conduct of the major retailers as illustrated by the email correspondence of Annex 6 referred to above.)
9.67 Significantly, however, the three other customers B, E and D all expressly stated that they had no concern about the impact of the proposed acquisition on the rashers market [DET. 5.33 – 5.45]. Customer B confirmed that there is less brand loyalty in this segment and that sales are driven by promotional activity thereby corroborating the evidence elsewhere as to the effect of periodic deep discount promotions. Customer B also frankly conceded that any difficulties in negotiating trade terms with the merged entity could be overcome by exerting discipline. The Authority records at para. 5.35:
“The retailer stated that there were a number of negotiating tactics available, the details of which the retailer was unwilling to provide.”
While the Authority expresses doubts as to the reliability of the responses from customer E [DET. 5.36 – 5.37] it stated that its ability to discipline the merged entity would depend upon whether alternative suppliers are available and it identified three particular competitors as credible alternative suppliers in the rashers market. Nevertheless it did not believe that it could credibly threaten to exercise its option to de-list or de-range the products of the merged entity. Customer D was not concerned either about the acquisition and considered that the top sellers in this segment were its own label offering together with Denny and Galtee and that its own label rasher offering was the most developed of all its own label product ranges. This customer is not recorded as having answered the question about threatening to de-list but the implication of para. 5.39 and its lack of concern must be that customer D felt it would be well able to look after itself post-acquisition in the face of any hypothetical price increase from Kerry.
9.68 The Authority then gives its conclusion at para. 5.4:
“In terms of being able to successfully resist a post-merger price increase although the evidence is somewhat mixed, on balance it shows retailers will have difficulty resisting such a price rise.”
The Court cannot agree that this is a justifiable conclusion to draw from the evidence which the Authority has just recited. Having regard to the incentives for self-serving responses to the leading question, it is more significant, surely, that 50% of those who replied had no concern as to their ability post-merger to conduct negotiations with the merged entity as owner of the two leading brands.
9.69 In the N-PCM product market, the corresponding consideration of countervailing buyer power is dealt with more briefly at paras. 6.157 – 6.163. The survey of retailers is summarised at paras. 6.98 – 6.116. Again, somewhat broadly similar responses were received. Of the nine retailers responding to the first questionnaire four expressed concern about the impact of the transaction on the N-PCM market and four had no concerns although two of these did not stock the Kerry/Breeo brands. Of the six retailers sent the follow-up questionnaire four claimed that they would be unable to credibly threaten to de-list or de-range the branded products post-acquisition. The Authority concluded at para. 6.116 that “it is clear that a majority of retailers who have their own private label N-PCM offerings are concerned about the impact of the proposed acquisition.
9.70 The Court finds that the conclusions thus drawn by the Authority as to the absence of sufficient countervailing buyer power on the part of the main retailers is flawed and unsound because, in effect, the conclusions appear to be inconsistent with the broad thrust and effect of the evidence relating to the exercise of buyer power taken as a whole and objectively assessed.
9.71 The Court has drawn attention above to the somewhat cursory and selective way in which a limited number of examples drawn from the material in Annex 6 to the written submission are used as a basis for discounting the probative value of all of that evidence. Having considered the totality of that material the Court finds that it is more supportive of the proposition that the main retail chains can and do exercise significant buyer power by resorting to a variety of tactics of the kind acknowledged by the Authority at para. 2.35 of the Determination. The specific examples proved by the appellant moreover corroborate the general implication raised by the asymmetry of the trading relationship between the suppliers and the multiples.
9.72 Furthermore, the probative force of the Annex 6 evidence must also be seen in the light of other related evidence which was available to the Authority and which is not referred to in the Determination itself but was invoked by the notifying parties at paras. 7.59 – 7.68 of the written response to the Assessment dated 15th August, 2008. Amongst other items of evidence, they drew attention to the fact that of eleven competitors (that is, other suppliers of products to the retailers,) interviewed by the Authority and asked if they had been subjected to de-listing or de-ranging by the multiples, ten confirmed that they had. Seven said it was a frequent practice. Another customer/retailer stated “On a continual basis when listing products, reviewing terms or agreeing promotional offers, it is standard practice to threaten to switch business to rival suppliers unless competitive terms are received.” Another said:
“Suppliers have to justify why they believe a price rise is merited. One option post-acquisition would be to run fewer promotions on the merged entity’s products. Another option would be to re-examine shelf space allocation with a view to changing the merged entity’s products, e.g. by placing the products in less attractive positions on the shelf such as above or below eye level, etc.”
9.73 The fact that de-listing and related tactics are routinely resorted to for a variety of reasons in the course of negotiations does not diminish the significance of the evidence for the purpose of assessing the predictable effects of countervailing buyer power. Apart from mentioning the fact that the two leading brands in the relevant markets will come into the control of a single undertaking, the Determination does not offer any explanation as to why there should be a material change or reduction in the ability of those retailers to continue to employ the same tactics post-merger in order to defeat a hypothetical price increase.
9.74 Furthermore, given the variety of tactics available to retailers, it is clear that they do not necessarily have to make the choice of abandoning “must have” brands as such if they are to defeat the hypothetical price increase. It is also clear that a retailer would not necessarily have to maintain any of the tactics over a protacted period of time in order to make it unprofitable for the merged entity to impose the hypothetical price increase. As was pointed out by the appellant, if one assumes a margin of 25% on the products supplied, a 3% price increase will be defeated by an 11% loss of sales. Furthermore, it is clear from the evidence that buyer-power tactics are not necessarily confined to targeting products to the relevant product market. Where a supplier such as Kerry is providing products across a variety of food categories, a price increase in rashers or cooked ham might be rendered unprofitable by the de-listing of one or more other lines in other sectors where the “must have” factor was of less importance to the retailer.
9.75 For these reasons the Court finds that the Authority has made a material and significant error in its appraisal of the existence of sufficient countervailing buyer power post-merger to deter a price increase by the merged entity. It has, in the Courts view, misjudged the significance and weight of all of the evidence available to it as to the reality of the buyer power exercisable by the four main multiple chains both in the form of the evidence submitted in Annex 6 to the written submission and the other evidence gathered by its case team in the form of responses received from retailers and competitors. Furthermore, in relying upon the responses of those retailers who expressed concerns as to the impact of the acquisition, the Authority has attributed undue weight as compared with the responses of retailers who expressed no such concerns.
9.76 It is precisely because the Court defers to the Authority in matters of choice or judgment within the area of its specialist expertise that the Court must be assiduous to ensure that in appraising factual evidence upon which its expert conclusions and judgments are based, the Authority has considered all the relevant information at its disposal; that it has weighed it objectively and rationally and that it has not erred by wittingly or unwittingly selecting only those parts of it which appear best to fit a particular economic solution.
10. Efficiencies
10.1 In the light of the Court’s findings as to the errors affecting Sections Five, Six and Eight of the Determination it is unnecessary for the Court to examine the further arguments raised by the appellant as to errors in section Nine: Efficiencies.
11. Conclusion
11.1 The Court accordingly finds that the determination made by the Authority is vitiated by material error in two respects. The Authority has erred in its determination of the product market for cheese in Section Eight and has erred in finding that there will be substantial lessening of competition resulting from the acquisition in the markets for rashers and for non-poultry cooked meats in that it has failed to assess correctly the post-acquisition existence of sufficient countervailing buyer power on the part of the retailers such as will deter a price increase imposed by the merged entity.
11.2 As these errors vitiate the finding of substantial lessening of competition in each of the three product markets upon which the negative determination was based, the appeal will be allowed and the Determination annulled.
