EC, February 17, 1992, No 92-157

THE COMMISSION OF THE EUROPEAN COMMUNITIES,

Having regard to the Treaty establishing the European Economic Community, Having regard to Council Regulation No 17 of 6 February 1962, First Regulation implementing Articles 85 and 86 of the Treaty (1), as last amended by the Act of Accession of Spain and Portugal, and in particular Article 3 thereof, Having regard to the application for negative clearance and notification made on 4 January 1988 by the Agricultural Engineers Association Ltd of an information exchange agreement called the UK Agricultural Tractor Registration Exchange concluded with and between eight undertakings active in the sector of agricultural tractors, Having regard to the decision taken by the Commission on 11 November 1988 to initiate proceedings in this case, Having given the undertakings concerned the opportunity of being heard on the matters to which the Commission has taken objection, in accordance with the provisions of Article 19 (1) of Regulation No 17 and with Commission Regulation No 99-63- EEC of 25 July 1963 on the hearings provided for in Article 19 (1) and (2) of Council Regulation No 17 (2), Having consulted the Advisory Committee on Restrictive Practices and Dominant Positions,

Whereas:

I. FACTS

Introduction

(1) On 4 January 1988, the Agricultural Engineers Association Ltd (AEA), the United Kingdom trade association of manufacturers and importers of agricultural machinery, notified an information exchange agreement called the UK Agricultural Tractor Registration Exchange ('the Exchange'). That agreement concerns an exchange of information identifying the volume of retail sales and market shares of eight manufacturers and importers of agricultural tractors on the United Kingdom market. The Exchange is managed by the AEA with the service of the computer bureau Systematics International Group of Companies Ltd ('SIL').

(2) This information exchange has existed at least since November 1975. The Commission started its investigation of the Exchange in 1984. During investigations about complaints of interference with parallel trade by individual manufacturers, the Commission discovered the existence of the Exchange in the course of inspections at the offices of some of the members of the Exchange and at the offices of the AEA and SIL. After this investigation, the AEA notified the Exchange with an application for a negative clearance or an exemption under Article 85 (3) of the EEC Treaty. The Commission objected to this information exchange by a Statement of Objections of 11 November 1988 which was followed by an oral hearing between 25 and 28 April 1989.

(3) The members of the Exchange are the following manufacturers or sole importers of agricultural tractors in the United Kingdom:

- Ford (Ford New Holland Ltd which was recently taken over by Fiat),

- Massey-Ferguson (Massey-Ferguson United Kingdom Ltd),

- Case (J. I. Case Europe Ltd),

- John Deere (John Deere Ltd),

- Renault (Renault Agriculture Ltd),

- Watveare Ltd (importer of Deutz),

- Fiat (Fiatagri UK Ltd),

- Same-Lamborghini (Same-Lamborghini UK Ltd).

Structure of the market

(4) The volume of the United Kingdom tractor market amounts to approximately 20 000 new tractor units per year. Agricultural tractors sold in the United Kingdom are largely produced in other Member States, in particular in Germany, Belgium, France and Italy. The imports of tractors from Community countries into the United Kingdom account for over 50 % of the total United Kingdom market.

(5) The United Kingdom market is dominated by four suppliers which together held some 76 to 77 % of the market. These suppliers are Ford, Case, Massey-Ferguson and John Deere. Since the takeover of Ford New Holland by Fiat, these four suppliers now hold an approximate total market share of 80 % of the United Kingdom market. These four major suppliers, which each hold market shares of between 15 and 25 %, are followed by suppliers which hold between 2 and 3 % of the United Kingdom market. This is the case for Renault, Deutz and Same-Lamborghini. The eight, now seven, members of the Exchange together hold some 87 to 88 % of the United Kingdom tractor market while the remaining 12 % of the market is shared by several small manufacturers who are not members.

(6) This high concentration of the United Kingdom tractor market is further increased by the fact that in smaller geographic areas, in particular at dealer territory level, not all the suppliers are actively selling or even represented. The smaller suppliers on the United Kingdom market, i.e. Renault, Deutz and Same-Lamborghini, do not have the same extensive sales network as the four major suppliers. Their small volume of sales does not support the creation of a country-wide sales network with 120 sales outlets which is said to be the necessary number of retailers in the United Kingdom tractor market.

The Exchange provides for a common system of organization of dealer territories for all the participating companies based on aggregations of the 8 250 postcode areas of the United Kingdom.

The already high concentration (four firms which ± 80 %) is also reinforced by the fact that, in the upper horsepower categories, the number of suppliers is sometimes reduced to only three or four, which in combination with the unequal representation at dealer level can lead to situations where in particular territories and in particular horsepower categories only two or three suppliers compete in particular geographic and/or product areas.

(7) The suppliers of the United Kingdom market are also major suppliers in the other Community markets with the result that imports from these other markets into the United Kingdom are controlled to a large extent by the same suppliers.

(8) The market of agricultural tractors is further characterized by high barriers to entry for the following reasons:

- the sale of tractors requires a dense distribution and service network: dealers must be located in a 25 to 35 km radius of the farmer as it is vital to a farmer that his tractor can be repaired almost immediately; to cover the United Kingdom territory adequately a manufacturer needs about 120 sales outlets; the smaller the sales volume, the more difficult it is to finance an optimal distribution network. Distribution networks require a considerable investment and involve a high commercial risk,

- the tractor market is a low volume and stagnant / declining market with general overcapacity: in such a market, new investment from either manufacturers or dealers is unlikely. In particular, dealers do not want to invest in dealerships where the prospects are limited,

- in the tractor market brand loyalty is an important factor in the final choice of consumers: farmers normally give preference to a well-established brand which is a guarantee of quality and increases the second-hand value in the case of resale. The members of the Exchange are all producers enjoying a high brand reputation; the acquisition of an equivalent reputation by a newcomer would take a long time and require substantial investment.

This difficulty is increased by the fact that the eight, now seven, members of the Exchange hold nearly nine-tenths of that market. This market share has been almost constant since 1976 with the result that nearly nine out of every 10 tractors in use in the United Kingdom were bought from members of the Exchange; this established high market penetration has an important influence on a market which is mainly a 'replacement' market (i.e. one in which customers are not first-time buyers but are replacing existing equipment),

- finally, the information exchange itself considerably increases the barriers to entry because it permits the established suppliers to recognize immediately any new penetration into the United Kingdom market or any increase in sales by non-members and then to react and defend their market positions. It is indeed the case that the market share of non-members has shown no substantial change over the entire period of the existence of the Exchange.

(9) The tractor market in the United Kingdom and in the Community generally is further characterized by the fact that, unlike the car market, it does not face any significant price or quality competition from imports from outside the Community. Imports of agricultural tractors from outside the Community are limited to the lower horsepower groupings (less than 30 hp) which are mainly used for non-farming purposes or to Eastern European brands which do not compete on equal terms with the high quality brands of Community manufacturers. This applies particularly to the United Kingdom which is mainly a market for high-powered tractors in which Eastern European brands are even more disadvantaged.

Exchanges of information in the United Kingdom tractor market

(10) The eight, now seven, main suppliers of the United Kingdom market have set up, with the help of the AEA and SIL, an information exchange identifying the retail sales and market shares of each member of the United Kingdom market with detailed breakdowns by product, territory and time periods.

(11) Since its notification on 4 January 1988, the Exchange is under the direct control of the AEA. Before that period, the AEA considered itself only as a secretariat to the 'Tractor Registration Exchange Committee' created by the eight members of the Exchange within the AEA. This Tractor Registration Exchange Committee, whose existence is confirmed by the notified exchange agreement, is composed of a representative of each member and is responsible for the administration of the Exchange. This committee holds at least one annual meeting under the chairmanship of one of its members.

(12) The information exchanged between the members of the Exchange through the AEA and SIL is mainly taken from the V55 forms which are used for the registration of tractors with the United Kingdom Department of Transport.

For tractors with need not be registered (e. g. those not to be used on a public road), the dealer must undertake to supply a copy of the V55 form directly to SIL to be included in the database together with the data on the tractors which must be registered.

(13) The Department of Transport has agreed to make its registration documents available to the AEA for industry analysis. However, no document suggests that the Department of Transport would have agreed to make these registration documents available to individual companies where such disclosure would identify the sales and market shares of competitors in the detailed form in which they have been made available by the AEA to the members of the Exchange. The contrary is suggested by an internal note of the AEA dated 31 August 1979 which states: 'it is doubtful that the vehicle licencing information would be made available by the Department of Transport to a group of individual companies without the unifying influence of a trade association which is meant to represent broader commercial interests . . .'.

(14) The V55 registration forms used by SIL for its analysis contain the following information:

- make (manufacturer),

- model number,

- serial number / chassis number,

- original dealer (code number, name, address and postcode),

- selling dealer (code number, name, address and postcode),

- postcode of the registered keeper (to ther fifth digit),

- name and address of the registered keeper (this information is provided on a voluntary statistical section).

(15) This information is made available by AEA and SIL to the members of the Exchange in the form of reports and analyses described hereafter or in form of direct computer access by each member according to their specific requirements.

Aggregate industry data

(16) Each member may obtain information on aggregate industry sales with or without a breakdown by horsepower groupings or by driveline. This aggregate information can in addition be broken down by geographic areas, i.e. the United Kingdom, MAFF (Ministry of Agriculture, Fisheries and Food) regions (10 regions), land use (arable / horticultural - livestock - mixed arable - mixed livestock - rough grazing - urban), counties, dealer territories and postcode sectors. This information is or can be made available for yearly, quarterly, monthly and weekly time periods.

The Commission has not in principle objected to the availability of these aggregate industry data because they do not identify the retail sales of the individual members of the Exchange. However, the Commission objected to the exchange of aggregate industry data to the extent that in respect of specific geographic areas product breakdowns or time periods, the reports supplied by SIL contain less than 10 tractor units sold for any such specific breakdown by territory, product or time period. Below this minimum number of total sales there is a high risk that even aggregate data will allow, directly or indirectly, the identification of the exact sales volume of individual competitors.

Data identifying sales of individual competitors

(17) In addition to the aggregate industry sales data described above, each member may obtain information identifying the volume of retail sales and market shares of each individual member of the Exchange with detailed breakdowns by model, by product groups (horsepower grouping and driveline), by geographic areas (including counties, dealer territories and postcode sectors), and by yearly, quarterly, monthly and daily time periods.

(18) The data exchanged between the members via the AEA and SIL identifies in respect of each member the following information:

- the exact volume of the retail sales and market shares of each member-competitor on the United Kingdom market at national, regional, county, dealer territory and postcode sector level: this makes it possible to compare the market penetration and performance of each supplier down to the smallest geographical level; it is possible to compare the performances not only of the manufacturers but also of their dealers in each dealer territory,

- the exact volume of the retail sales and the exact market shares of every specific model sold by each member: this permits comparison of the performance of specific models of each member-competitor; the comparison is equally done at the level of geographic zones which can be determined by each member,

- the exact volume of the retail sales and the market shares of specific horsepower groups for each member: this permits comparison of the performance and market penetration of each member-competitor in specific horsepower bands which are usually categories of 10 horsepower grouping; there are also breakdowns by driveline (two-wheel four-wheel / articulated),

- the daily and monthly retail sales and market shares at United Kingdom level for each member: this permits comparison of the latest most up-to-date sales performance of each participating supplier on the United Kingdom market and their evaluation during the current month.

(19) The above information is made available for time periods broken down by year (calendar year, fiscal year, rolling year), year to date, quarter and month, and in respect of total United Kingdom retail sales the figures are even available on a daily basis. This Therefore permits each member to follow the sales performance and market penetration of each participating competititor on a yearly, quarterly, monthly and daily basis in respect of all the products, specific products and within the smallest geographic areas.

(20) In addition, the members release to their dealers all this individualized market information as regards each dealer's own territory and information on the units sold and market share of each member at national level. Thus commercially sensitive information is exchanged not only between the manufacturers of tractors but is also made available to the retailers in the market. This creates full transparency between competitors at manufacturer level and at the dealer level within each dealer territory.

(21) Finally, the Exchange provides to each member aggregate data on the retail sales and market shares of non-members (called 'others') with approximately 12 % of the United Kingdom market. Sales by the individual non-members are not identified but members of the Exchange nevertheless have access by this means to up-to-date information about changes in the market shares of the 'others' - the data covers the same periods as the identifying information (yearly, quarterly, monthly and daily).

(22) As is shown in various documents quoted below, the members of the Exchange classify the information exchanged between them via the AEA and SIL as sensitive confidential information, to the extent that it identifies the exact retail sales of each member, and consider that this information is exchanged only on a mutual and reciprocal basis:

- policy outline drawn up by the AEA on 12 January 1979:

'As it is the intention of the Association to respect the interests of companies no analysis allowing identification of a particular company's product will be made without direct permission of that company . . . In addition, companies have the option to arrange, through the Association, for further analysis of their own information on a group basis. The information relating to the members of such a group would be analysed according to particular requirements on a mutual and reciprocal basis. At no time may members of such a group obtain access to detailed information (make or model), either directly or by deduction, on products of companies not within the group.'

- Notes on the UK Agricultural Tractor Registration Exchange drawn up by the AEA on 12 January 1979:

'The computer bureau then analyses the forms of those companies that are within the scheme according to the instructions of the companies. Information which is considered confidential (i.e. make and model) is therefore only available on those machines which were supplied by companies which participate in the exchange . . . They (the members) do not receive information of a confidential nature on companies not within the group.'

- Minutes of a meeting of the Exchange dated 8 November 1979 where the members classify the information exchanged between them into two categories: sensitive and non-sensitive information. Total registrations by horsepower groups without identification of manufacturers or models are considered to be non-sensitive information if shown at United Kingdom level or MAFF-region level on a yearly or quarterly basis. On the other hand, information identifying manufacturers or models is considered to be sensitive information which is only exchanged on a confidential and reciprocal basis.

- Minutes of a meeting of the Exchange dated 24 June 1980 where it is stated in respect of the release of information to dealers: 'The exchange accepted the necessity of stressing to dealers that detailed registration information was available only as a result of close cooperation between manufacturers. Any misuse could result in arrangements being jeopardized and the information should therefore be treated confidentially.'

- Minutes of a meeting of the Exchange dated 28 October 1980:

'Deutz were also informed that the Exchange existed totally on the principle of reciprocal exchange of information . . .'

- Letter of Massey-Ferguson to all its UK dealers dated 30 April 1981:

'We obtain the information through a mutual and confidential exchange of data between participating manufacturers. A participant is entitled to withdraw if for example confidentiality is not maintained. The data is expensive to obtain because of the cost of processing required and circulation is restricted to the manufacturers who pay for it . . . You are asked therefore to continue to treat your own territory data and any total UK data which is made available to you with the strictest confidence.'

- Letter by Ford to all members of the Exchange:

'I need hardly remind you of the value to us of such statistical information. It is strictly confidential and is only obtained by mutual trust of all participating manufacturers respecting that confidence.'

- Letter by International Harvester to its dealers dated 22 April 1981:

'. . . these figures are released on a strictly confidential basis and are for use only within the International Harvester organization. Any forms showing unit and market shares must be regarded as security documents and treated accordingly. Release to the press or any other source could lead to the collapse of the system which is beneficial to you and us.'

- Minutes of a meeting of the UK Agricultural Tractor Registration Exchange held on 20 April 1983: it is recorded that the chairman of the meeting objects to the release of make data inter alia for the following reasons:

- 'Recently there has been a statistical war in the car trade, to publish the market shares is to increase the likelihood of this and the exchange will only survive as long as amicable relationships between participants exist.'

- 'The exchange is based upon and still exists upon the idea that, for instance, MF (Massey- Ferguson) and IH (International Harvester taken over by Case in 1985) will swap their registrations/figures. The exchange can then be viewed as a collection of swapping decisions.'

- Letter of the AEA to the National Farmers' Union dated 27 May 1983:

'In so far as more detailed information is concerned we have a difficulty which may not be surmountable. Trade statistics are provided to us on a confidential basis and we are not at liberty to divulge them to anyone without the consent of all participants.'

(23) The members have discussed the possibility of releasing market data to the press but until April 1987 only aggregate industry data relating to United Kingdom and MAFF regions were published. From April 1987 to October 1988, when the members decided to suspend the operation of the Exchange following the opening of proceedings by the Commission, they published data identifying market share of individual members in respect of total United Kingdom figures on a yearly and quarterly basis with a delay of three months. All other information made available to the members remains confidential to the participating members.

Data on dealer sales of own company

(24) In addition to the data identifying sales of other competitors, the AEA and SIL provide, to each manufacturer individually, information on sales made by its own dealers.

(25) This information on own dealer sales is also extracted from the V55 registration system. The V55 forms allow the destination of each tractor to be traced by a simple delivery / registration comparison. The V55 forms contain for each registered tractor information on the serial number / chassis number, original dealer, selling dealer (if different from the original dealer) and the five-digit postcode of the final purchaser.

(26) There are basically two versions of V55 forms: the precompleted V55/1-4 forms which must be used by the dealers for the registration of all new tractors delivered to them by their United Kingdom suppliers and the V55/5 form which must be filled in by independent importers for the registration of new tractors imported into the United Kingdom.

The V55/1-4 registration forms are analysed by SIL to provide, upon demand, to each individual manufacturer the following data on sales of its own dealer network:

- trend analysis: comparison of own company's sales with total industry sales at national, MAFF region, county, dealer territory and postcode sector levels with product breakdowns (horsepower groups) on a monthly basis; at dealer territory and postcode sector level these monthly trend analyses can entail such small numbers of sales for specific product groups that identification of sales by competitors in the these small geographic areas is not excluded,

- dealer import analysis: this analysis identifies the number of tractors imported by other dealers into a given dealer territory and compares these imports with the sales made by the home dealer within his own dealer territory,

- dealer export analysis: this analysis identifies the number of tractors exported by a given dealer into other dealer territories and compares these exports with the dealer's home sales within his own dealer territory,

- selling dealer analysis: this analysis permits manufacturers to identify the selling dealers in a particular postcode sector and to compare their respective sales with the industry sales in that given postcode sector.

(27) Finally, until 1 September 1988, the V55/5 registration forms used for the importation of new tractors into the United Kingdom were passed on by SIL to the individual manufacturers. These Documents permit the manufacturers to identify the importer, the serial number of the imported tractor, the model and with this information to trace the origin of the imported tractor, i.e. the selling dealer in other Community countries if the imports come from within the Community. The V55/5 documents thus permit the manufacturers to identify the source of parallel imported tractors on the United Kingdom market. SIL stopped sending the V55/5 forms to the individual manufacturers on 1 September 1988.

(28) SIL described the system in a fax addressed to John Deere in 1984: 'our database picks up all tractors registered for use on UK roads, i.e. 98 % over 30 hp irrespective of source. For those companies in our Exchange we also store data relating to which dealer actually retailed the unit and we return to each company the base registration documents. So if a JD (John Deere) unit was imported by an EEC dealer, JD UK would have sight of the documentation referring to the tractor - serial No etc.'

(29) Massey-Ferguson described the system in a memorandum dated 4 May 1981: 'From this system we are able to track and verify registration data on all MF (Massey-Ferguson) tractors and some other products dispatched by MF (UK) Ltd to UK distributors. We also have clear visibility of registration data on units shipped some time ago and of first time registration of units shipped more recently. Most important in the light of recent Grey Import activity we will have visibility of any registration on V55/5 forms of MF products and the serial numbers.'

(30) There are a number of documents discovered during investigations which evidence the fact that some of the members have actually used the V55/5 forms received from SIL to trace the source of parallel imports and to stop these imports coming from other Community countries. The documents in question are those contained in Annex 3 to the Commission's Statement of Objections and those supplied to the parties after the oral hearing by a Commission letter of 22 May 1989 (documents Nos 1, 5 to 8, 16 and 28 to 48).

Suspension of the Exchange

(31) On 24 November 1988 the parties decided to suspend the operation of the Exchange to the extent that the Commission had objected to it in its Statement of Objections of 11 November 1988. The importer of Deutz tractors, Watveare, left the Exchange in December 1988.

(32) After the oral hearing, the Commission informed the parties by letter of 25 October 1989 of its position. This position as refined on 28 November 1989 was accepted as a matter of principle by two parties - Watveare and Case.

(33) On 12 March 1990, five members, i.e. Ford, Fiat, John Deere, Renault and Massey- Ferguson notified an amended Exchange, which they called 'UK Tractor Registration Data System'. On 9 July 1990 Massey-Ferguson informed the Commission that it too accepted the Commission position of 28 November 1989 and consequently withdrew its application for negative clearance and/or exemption made on 12 March 1990.

The four, now three, remaining parties have notified an information exchange system which they have not implemented. This new notification reduces to some extent the information made available to the members but continues in particular to provide information identifying the sales and the market shares of the members for periods below one year, i.e. monthly information at United Kingdom, MAFF region, land use, county, dealer territory and postcode sector levels including breakdowns by models. The amended Exchange also provides to each party details on the chassis number and date of registration of that party's tractors sold. The details of the amended Exchange are set out in Appendix 2 to the new notification.

II. LEGAL ASSESSMENT

A. Legal assessment under Article 85 (1) of the exchange as practised before notification and as notified on 4 January 1988

(34) The AEA and the eight, now seven, members of the Exchange are (3) parties to an agreement within the meaning of Article 85 (1). The parties have agreed on a common system of organization of dealer territories (see recital (6) above) and the AEA are only allowed to release the information identifying the sales of each member on the basis of a mutual and reciprocal acceptance by each member of the release. This agreement was formalized with the notification of an information exchange agreement by the AEA in January 1988. As regards the release of information on retail sales of own dealers to each individual member, this release is also based on an agreement between each member and the AEA and SIL. All these data are provided by the AEA and SIL against the payment of fees.

The role of SIL is that of a computer bureau which provides its services to the AEA and the eight members of the Exchange against payment of a fee. The responsibility for the Exchanges however mainly with the AEA and the members of the Exchange. Therefore, the Commission does not consider it necessary to continue to include SIL in these proceedings.

Data identifying sales of individual competitors

(35) For the assessment of that part of the Exchange which identifies the sales of the individual members, the Commission takes account of:

- the market structure, i.e.:

- high concentration: four firms with approximately 80 % of the market, seven firms with 87 to 88 % of the market, a still higher degree of concentration for smaller geographic areas or product breakdowns, the members are also major suppliers in the other Community markets, and the gap and dispersion of the market shares of the non-members who together hold only about 12 % of the market,

- high barriers to entry: the need for an extensive distribution and service network, the low volume and stagnant / declining market, the advantage of brand loyalty for established suppliers and the effects of the information Exchange,

and

- the absence of significant imports from outside the Community for agricultural tractors above 30 hp;

- the nature of the information exchanged: i.e. the exact quantities of the retail sales and the exact market shares which are trade secrets between genuine competitors in a highly concentrated market,

- the detail of the information exchanged: i.e. detailed product breakdowns, small geographic breakdowns and time periods below one year, i.e. daily, monthly and quarterly market sensitive information on competitors, and,

- the fact that the members meet regularly within the AEA Committee which gives them a forum for contacts.

(36) The Exchange leads to restrictions of competition for two reasons:

(a) Prevention of hidden competition in a highly concentrated market

(37) The Exchange restricts competition because it creates a degree of market transparency between the suppliers in a highly concentrated market which is likely to destroy what hidden competition there remains between the suppliers in that market on account of the risk and ease of exposure of independent competitive action. In this highly concentrated market, 'hidden competition' is essentially that element of uncertainty and secrecy between the main suppliers regarding market conditions without which none of them has the necessary scope of action to compete efficiently.

Uncertainly and secrecy between suppliers is a vital element of competition in this kind of market. Indeed active competition in these market conditions becomes possible only if each competitor can keep its actions secret or even succeeds in misleading its rivals.

This reasoning, however, in no way undermines the positive competitive benefits of transparency in a competitive market characterized by many buyers and sellers. Where there is a low degree of concentration, market transparency can increase competition in so far as consumers benefit from choices made in full knowledge of what is on offer. It is emphasized that the United Kingdom tractor market is neither a low concentration market nor is the transparency in question in any way directed towards, or of benefit to, consumers.

On the contrary, the high market transparency between suppliers on the United Kingdom tractor market which is created by the Exchange takes the surprise effect out of a competitor's action thus resulting in a shorter space of time for reactions with the effect that temporary advantages are greatly reduced. Because all competitive actions can immediately be noticed by an increase in sales, the consequences are that in the case of a price reduction or any other marketing incentives by one company the other can react immediately, thus eliminating any advantage of the initiator. This effect of neutralizing and thus stabilizing the market positions of the oligopolists is in this case likely to occur because there are no external competitive pressures on the members of the Exchange except parallel imports which are however also monitored as has been explained above.

(38) The United Kingdom tractor market is clearly a highly concentrated market where competition is already weakened by the fact that:

- four firms dominate the market with a combined market share of approximately 80 %,

- these four firms have created an information exchange with four, now three, other well- known suppliers capable of challenging their market position with the effect that the conditions of a narrow oligopoly have been artificially created between the eight, now seven, best established competitors on this market by giving them information on any change in volumes and market shares at manufacturer and dealer level,

- these eight, now seven, suppliers are active in all other Community markets and know the pattern of trade and products supplied on the United Kingdom market for a long period of time; in particular, there is no difficulty knowing the prices charged by each member because in this industry there are list prices and a simple telephone enquiry will readily reveal the general level of discounts applied by each manufacturer's dealer network; this information is also available from customers who inform the dealers of competitors' prices in a given territory,

- the market is protected from competition from outside the United Kingdom by the fact that there are high barriers to entry and that there are only insignificant imports from outside the Community.

(39) In addition, demand is very dispersed. There are numerous buyers, the majority of whom do not have the possibility of purchasing tractors in other Member States (transport difficulties, import formalities, registration, service, trade-ins, etc.). Thus, demand transferability in this market is very low which weakens the competitive pressure from the demand side on the limited number of suppliers established within the United Kingdom and therefore reinforces their economic strength on that market.

(40) On the United Kingdom tractor market, therefore, the only difficult, but very important, market data to obtain is the exact volume of sales of each manufacturer / dealer so as to be able to notice instantly changes in sales volumes and market shares of each member of the oligopoly and of each dealer at the level of dealer territories. This market knowledge allows each member and dealer to react immediately and thus to neutralize whatever initiative any one of the members / dealers of the oligopoly would take to increase its sales (see recital (37)). However, the result in practice is that few such initiatives will be taken precisely because every supplier knows very well that the position of each of the others is and that, thanks to the transparency created by the system, any initiative on his part can be detected at once by the others.

(41) The very detailed product and geographic market information on retail sales (see recitals (17) to (23)) gives each manufacturer and dealer fully reliable market knowledge (100 % coverage and 100 % accuracy) which is accessible on an instant or very short time basis (daily, monthly or quarterly). It allows them:

- to establish with accuracy the market positions and performances of their rivals and to follow constantly any changes of these market positions,

- to see at once whether there has been any increase in the retail sales of a rival, to see the territory in which such an increase takes place, to detect the models which contribute such an increase and finally to follow whether and to what extent any price or other marketing strategies of rivals are successful,

- to limit price competition as far as possible by allowing suppliers and dealers to react to any price-cutting or other market strategies selectively by limiting their response to the absolute minimum degree necessary in terms of product and territory and by being sure to hit the right target,

- to react more quickly if the market positions start changing.

(42) The Exchange further helps the established firms and dealers with considerable market shares (i.e. the four biggest firms holding approximately 80 % of the United Kingdom market) to defend their market positions more efficiently than they could do without the detailed information on retail sales of their rivals and any changes thereof. By their wider market coverage and sales volume, these firms and dealers already have the advantage of better market knowledge which is further strengthened by the fact that they can react more effectively to any increase in sales by the smaller competitors on the market. This is confirmed by the fact that, during their participation in the Exchange, the main four suppliers have essentially maintained their combined market share vis-à-vis the other members of the Exchange on the United Kingdom market.

(43) In the absence of the Exchange, firms would have to compete in a market with some measure of uncertainty as to the exact place, degree and means of attack by rivals. This uncertainty is a normal competitive risk bringing about stronger competition because reaction and reduction of prices cannot be limited to the absolute minimum degree necessary to defend an established position. Uncertainty would lead the firms to compete more strongly than if they knew exactly how much of a response was necessary to meet competition. They would have to exceed a minimum response, for instance by offering more favourable discounts to move their stock or by offering discounts for more products and in more territories. The Exchange reduces uncertainty by revealing the actions and reactions of all participating competitors who represent 87 to 88 % of the United Kingdom market. There is thus a prevention of hidden competition which results necessarily from the Exchange.

(b) Increase of barriers to entry for non-members

(44) The Exchange not only lessens competition between members of the Exchange and between their dealers, it also restricts competition between members and non members of the Exchange even if the Exchange in principle admits any manufacturer or importer to the information Exchange.

(45) If a supplier chooses not to become a member of the Exchange, he is disadvantaged by the fact that he does not have available the detailed and accurate market information about other suppliers which is available to members of the Exchange. Detailed knowledge of the sales pattern for tractors on the United Kingdom market improves the members' ability to defend their positions vis-à-vis non-members.

(46) If a supplier chooses to become a member of the Exchange, he must reveal his exact retail sales by product and by every small geographic territory with the result that the Exchange then permits the established suppliers with considerable market shares and extensive dealer networks to become aware of the existence of new entry and to instantly detect the market penetration by any such new member. This market information on any new member will permit the established suppliers to defend their acquired positions by placing selective actions designed to contain the new member.

(47) As a result, for a small supplier it is neither advantageous to become a member of the Exchange nor to stay outside the Exchange. In both cases the Exchange advantages the big suppliers who already belong to it. The presence of the smaller suppliers in the Exchange shows indeed that these suppliers have not been able to contest the position of the four biggest suppliers, i.e. Ford, Case, Massey-Ferguson and John Deere, and that market expansion is only possible through acquisition - such as the recent acquisition of Ford by Fiat. The presence of the four smaller suppliers can therefore only be explained by the fact that the UK Exchange is part of a network of similar Exchanges in other Community countries and that the eight suppliers participate in various exchanges on a reciprocal basis. These other exchanges are presently under investigation.

(48) In consequence, the Exchange permits its members to effectively prevent new entry and to contain the expansion of other suppliers on the United Kingdom market who are not members of the Exchange. This constitutes a serious restriction on the development of new competition in a highly concentrated market which is already characterized by high barriers to entry with the result that the members of the Exchange can jointly maximize profits to the detriment of farmers. This restriction affects in particular the smaller suppliers from other Community countries who will not be able to penetrate the United Kingdom market if they wanted to sell or increase their sales on that market.

(c) Main arguments put forward by the parties

(49) The parties have submitted that the Exchange is based on registration data made available to the industry by the United Kingdom Department of Transport and that these data do not constitute trade secrets. This argument cannot be accepted.

First, the Exchange does not deal in mere estimates, but in wholly accurate information on the quantities sold and the market shares of each participating member and of its dealers. Such information is normally considered to be a trade secret which confers legitimate competitive advantages and is, for that reason, not to be revealed or shared with competitors (see Court of Justice, Case 236-81, Celanese Chemical Company Inc. v. Council and Commission (4), Case T-30-89, Hilti v. Commission (5), Seventh Report on Competition Policy, Chapter I, paragraph 2, points 5-8; Commission Decisions 87-1-EEC (6), 77-592-EEC (7) and 79-90- EEC (8)). Unlike prices or other trading conditions which can be found out by telephone enquiries or customer feedback, exact quantities sold and market shares are the most difficult market information to be obtained and the only accurate indicator of the trade pattern in a market and changes thereof.

Secondly, the parties have themselves described information identifying the exact quantities sold and market shares of individual competitors as sensitive and confidential which can only be shared on a mutual and reciprocal basis and with the express agreement of the firms whose figures are being revealed (see recital (22)).

Thirdly, even if the registration data of the United Kingdom Department of Transport were accessible to each manufacturer on an individual basis, they could not be exploited to the same extent as they are through the Exchange. The Exchange provides detailed product, geographic and periodic break downs in respect of the retail sales and the market shares of all members which allows manufacturers instantly and constantly to follow the evolution of each member's market position.

In addition, the members have all agreed to organize their respective dealer territories on the basis of the five digit postcode areas which is a prerequisite for any comparison of the full and detailed information on the individual members and their dealers. Furthermore, for tractors sold but not registered, the dealers have undertaken to return the completed V55 forms to the manufacturer who sends them to SIL for processing. Thus, even if the registrations were accessible on an individual basis, they are not exploitable without the Exchange's agreement.

Lastly, the fact that a Government department makes available to the industry registration data identifying the sales of individual competitors in a given market as opposed to aggregate data not identifying individual companies does not prevent the application of Article 85 to the conduct of the undertakings in question. On the contrary, it only means that the public authority may, in certain circumstances, also be laying itself open to the allegations of an infringement, in this case of Article 5, for it follows from the combined provisions of Article 85, Article 3 (f) and the second paragraph of Article 5 that provisions of national law or national administrative practices may not adversely affect the application in full of the Community competition rules (see the judgment of the Court of Justice in Case 45-85, Verband der Sachversicherer v. Commission [1987] ECR, p. 405 (9)).

(50) The parties have further submitted that the information relates only to past transactions as opposed to future planned actions. Therefore, in their view the exchanged information is historical information which has no impact on competition in the market concerned. This argument cannot be accepted either.

In a market where demand is stable or declining as is the case for the agricultural tractor market in the United Kingdom, forecasts of competitors' future actions can largely be determined on the basis of past transactions. All the restrictive effects of the Exchange described above depend on the observation of competitors' actions and performance in the past. The more accurate and the more recent the information on quantities sold and on market shares, the more impact this information has on future behaviour of the firms in the market (see the judgment of the Court of Justice in Case 172-80, Zuechner v. Bayerische Vereinsbank AG [1981] ECR, p. 2021 (10)).

However, the Commission accepts that from a certain point in time, market information relating to past transactions becomes truly historic and no longer has any real impact on future behaviour. In the present case, the Commission considers that an annual exchange of one- year-old sales figures of individual competitors at the United Kingdom, MAFF region and land use level and with a breakdown by model can be accepted as commercial data with no appreciable distorting effect on competition between the manufacturers or between the dealers operating on the United Kingdom market.

(51) The parties have also submitted that the Commission has no proof of actual restrictions of competition resulting from the Exchange and that on the contrary an analysis of the past performance of companies on the United Kingdom market shows changes in market shares which are not reconcilable with the anticompetitive effects of the Exchange alleged by the Commission in the Statement of Objections. This argument cannot be accepted either.

The nature of the information exchanged and the detail of that information is such that it creates a degree of transparency in a highly concentrated market shielded from outside competition (high entry barriers, virtually no extra-Community imports) which necessarily leads to a reduction of competition between the competitors operating on that market.

In addition, account must be taken not only of the immediate visible effects on an agreement but also of its potential effects and of the fact that an agreement can create a structure likely to be used for anticompetitive purposes.

Article 85 (1) must be interpreted as including potential anticompetitive effects because the objective of that provision is the maintenance of an effective competitive structure within the meaning of Article 3 (f). This objective is particularly material in a highly concentrated market where an information exchange creates a structure of transparency which prevents hidden competition and increases barriers to entry for non-members.

As to the changes in the market shares, it first needs to be pointed out that the overall United Kingdom market shares for the participants in the Exchange have remained relatively stable during the existence of the Exchange. However, fluctuations in market shares have taken place at national and local levels. These variations in themselves do not exclude a lessening of competition on the United Kingdom tractor market as a result of the Exchange. In order to be restrictive of competition, the Exchange does not need to create an absolute stability of market shares or exclude even fierce competition in certain areas. In the present case, the anticompetitive effect of the Exchange lies in the fact that, for the reasons explained above, it can be said with sufficient certainty that the market shares and market positions of the members would have been different in the absence of the Exchange and are likely to be different in the future if the Exchange does not continue. However, to provide proof of actual differences is impossible in practice. In this case, it is the economic context, the structure of the market and the nature of the information exchanged which all lead to the conclusion that a lessening of competition as a result of the Exchange is inevitable.

(52) Finally, the parties submit that the heterogeneous nature of the products concerned creates strong non-price competition (quality competition) in the tractor market as each manufacturer tries to differentiate its products from those of its competitors in order to gain the consumers' preference.

This argument neglects the fact that for farmers, the purchase of a tractor represents a very high financial investment. Therefore, in the farmer's final choice, the price of the tractor remains an important element, in particular since all tractors of the members of the Exchange, however differentiated thay may be, fulfil the same functions and are compatible with all other agricultural machines to be attached to a tractor. Reliability and brand loyalty are therefore not the only elements taken into account by a farmer. Price competition still plays an important role in this market. This means that there is scope for the manufacturers and dealers to lessen the intensity of competition between them by avoiding price fights which would erode their margins. By increasing the transparency in a highly concentrated market and by strengthening the cohesion between the major suppliers in that market through regular and secret contacts, it is possible to maintain a general high price level in the market concerned in spite of price differences between the various products offered on the market.

Data on dealer sales of own company

(53) The Exchange also reveals to each individual manufacturer detailed information about the retail sales of its own dealers. This in itself is not objectionable provided that such information does not permit the identification of sales of competitors and provided that such information does not permit interference with the retail activity of dealers or parallel importers.

(54) As to the possibility of identifying sales of individual competitors, this risk exists where, in respect of geographic areas, product breakdowns or time periods, the number of industry sales amounts to less than 10 tractor units sold for any such specific breakdown by territory, product or time period. In such cases, a comparison between the industry sales and own company sales can facilitate, either directly or by deduction, identification of the exact sales made by individual competitors or dealers of competitors.

(55) As to the possibility of interfering with the retail activity of dealers or parallel importers, the Exchange entails the following potential anticompetitive effects:

- the dealer import and dealer export analyses show for each dealer territory the exact number of sales made to customers within that territory both by the local dealer and by other dealers of the same network. These analyses further show in which dealer territories outside his own a dealer makes sales. The Exchange thus reveals the exact destination of each specific tractor and reveals the export activities of each dealer into the territories of other dealers in the United Kingdom. This monitoring of dealers' sales outside their allocated territory allows manufacturers to exercise pressure on dealers engaging in any perceived excessive export activity. It is the very nature of such reports to give rise to misuse which in most cases cannot be controlled a posteriori (a manufacturer could, for instance, reduce supplies to a dealer who sells too many tractors in other dealers' territories). The import/export analyses are thus highly likely to reduce intrabrand competition with the result that dealers can maintain higher profit margins for both themselves and the manufacturers,

- until September 1988, the members of the Exchange also received from SIL a copy of each V55 registration form which enabled manufacturers to identify parallel imported tractors and to trace the origin of these tractors (see recitals (27) to (30)). The documents referred to in recital (30) show that the information made available to each member has been used to stop parallel imports of tractors from other Community countries into the United Kingdom. The Exchange was thus a facilitating mechanism for interference with parallel imports which brought about competitive pressure on the United Kingdom dealerships.

(56) As regards the restriction of parallel imports, the present decision only addresses the Exchange in so far as it constitutes a facilitating mechanism for interference with parallel imports. The use made by certain members of the Exchange actually to stop parallel imports through this mechanism will be dealt with in separate proceedings.

Effect on trade between Member States

(57) An exchange of information identifying in great detail the exact retail sales volume and the market shares of 88 % of the suppliers of a national market, which includes all major importers from other Community countries, which prevents hidden competition in a highly concentrated market and which increases the barriers to entry for non-members, is liable to substantially affect trade between Member States because the lessening of competition resulting from this exchange necessarily influences the volume of imports to the United Kingdom.

The pattern of trade between the United Kingdom and the rest of the common market will be different from what it would be in the absence of this information exchange. In addition, the Exchange has been and could in future be used to hinder parallel imports to the United Kingdom.

(58) The appreciable effect on trade between Member States is all the more likely to exist as the information exchange is not one between two or three market participants but a quasi industry-wide exchange arrangement covering nearly nine-tenths of a national market which itself is a substantial market within the Community. In the same way, the information system facilitates interference with parallel trade not just for one or two undertakings but for virtually all suppliers of the United Kingdom market. Finally, the possibility of interference with the retail activity of the dealers within the United Kingdom by misusing the dealer import/export analyses reduces intrabrand competition on the United Kingdom market which is a substantial part of the common market. All these effects are likely to lead to the result that the prices charged in the United Kingdom for agricultural tractors can be maintained at a supracompetitive level while tractor imports by non-members to the United Kingdom market are rendered more difficult as a result of the increased market knowledge in the hands of the major United Kingdom suppliers.

B. Legal assessment under Article 85 (3) of the exchange as practised before notification and as notified on 4 January 1988

(59) The UK Agricultural Tractor Registration Exchange has been in existence since 1975 and was only notified on 4 January 1988 after several investigations had revealed the existence and nature of that industry-wide information exchange. The parties have thus operated the Exchange for over 12 years without notification to the Commission. The parties cannot invoke the benefit of Article 4 (2) (1) of Regulation No 17 because their Exchange relates also to imports of tractors into the United Kingdom.

Since the Exchange was not notified in accordance with Article 4 (1) of Regulation No 17 until 4 January 1988, it is not eligible for an exemption under Article 85 (3) for the period preceding that notification.

(60) In their notification of 4 January 1988 and in their later submissions, the parties have argued that they use the detailed information identifying the sales of each member for quicker product development, better production planning and better monitoring of dealers. The Commission takes the view that the disadvantages resulting from the above described restrictions of competition outweigh these alleged benefits.

In any event, the parties have not proved the indispensability of information which identifies the performance of individual manufacturers and individual dealers to achieve the alleged benefits. The Commission takes the view that these benefits can be derived from own company and aggregate industry information.

(61) The Commission does not object to the exchange of aggregate industry information with geographic breakdowns by United Kingdom, MAFF region, land use, county, dealer territory and even postcode sector and with product breakdowns by horsepower groupings or driveline provided that the reports contain at any time a minimum of 10 tractor units sold for any such report. The Commission does not object either to the release of own company information except for certain specific reports with clear anticompetitive effects.

Therefore, the parties may have access to detailed and up-to-date industry and own company figures. In addition, in this particular market, the Commission does not object to the annual exchange of one-year-old figures identifying the sales volume and the market shares of individual competitors at United Kingdom, MAFF region and land use levels with a breakdown by models.

All this information permits each member to identify any overall market trends and changes in demand in the industry and also to forecast production needs, to fix sales targets for their dealers and to maintain adequate stocks of tractors and spare parts. It enables each member to judge the performance of its products and its dealers in relation to the industry as a whole.

(62) Since the oral hearing in this case, three members of the Exchange, i. e. Watveare (for Deutz), Case and Massey-Ferguson have informed the Commission that they can accept and operate with an Exchange which is reduced to a purely statistical exchange not identifying units sold and market shares of competing manufacturers for periods below one year. This can only confirm the conclusion of the Commission that own company data and aggregate industry data are sufficient to operate in the agricultural tractor market.

(63) As regards data on sales by own company dealers which enable manufacturers to interfere with the retail activity of dealers or parallel importers (dealer import and dealer export analysis / transmission of a copy of the V55/5 registration forms to the manufacturers), neither arrangements for the exchange of these data nor such practices are eligible for an exemption. It is difficult to see how the sharing of such data or such practices could be said to be indispensable for the achievement of better production or distribution of tractors. The performance of a dealer can be assessed by the number of total sales achieved by the dealer without identifying the exact destination of each specific tractor sold.

(64) In conclusion, without having to examine whether all four conditions of Article 85 (3) are fulfilled, the condition of indispensability is not in any event fulfilled in respect of either the data identifying sales of individual competitors or the date on dealer sales of own company dealers which enables interference with the retail activity of dealers or parallel importers. Therefore, the notified Exchange cannot benefit from an exemption pursuant to Article 85 (3).

C. Legal assessment under Article 85 (1) and (3) in respect of the amended notification of 12 March 1990

(65) The reasoning in Article 85 (1) and (3) developed hereabove applies mutatis mutandis to the amended notification of 12 March 1990. This second notification is made only by four parties, i.e. Ford, Fiat, John Deere and Renault, although the Commission understands that the other four parties to the original Exchange would certainly agree to further collaboration with them if the amended notification was acceptable to the Commission vis-à-vis Article 85. The amended notification, however, goes far beyond what the Commission considers to be acceptable for the market concerned because it continues, in particular, to provide information identifying sales volumes and market shares of the members and dealers for monthly periods and to give details on the chassis number and date of registration of each tractor sold. This latter information, like the V55/5 forms, permits the identification of the origin and destination of all tractors. If this information was essential for the checking of warranty or bonus claims, as is suggested by the parties, it could be limited to specific cases and not be made available for all tractors sold,

HAS ADOPTED THIS DECISION:

Article 1

The UK Agricultural Tractor Registration Exchange as practiced since November 1975 and notified on 4 January 1988, including its amendment of 12 March 1990, infringes Article 85 (1) of the EEC Treaty in so far as it gives rise to an exchange of information identifying sales of individual competitors, as well as information on dealer sales and imports of own products.

Article 2

The UK Agricultural Tractor Registration Exchange as notified on 4 January 1988, including its amendment of 12 March 1990, does not fulfil the conditions for an exemption under Article 85 (3) of the EEC Treaty and the application for an exemption is accordingly refused.

Article 3

The AEA and the eight members of the Exchange shall immediately put an end to the infringement established in Article 1, in so far as they have not already ended the infringement, and shall in future refrain from entering into any agreement or concerted practice that may have an identical or similar object or effect.

Article 4

This Decision is addressed to:

1. The Agricultural Engineers Association,

Samuelson House,

Paxton Road,

Orton Centre,

UK-Peterborough,

Cambridgeshire PE2 OLT;

2. Ford New Holland Ltd,

Cranes Farm Road,

UK-Basildon,

Essex SS14 3AD;

3. Massey-Ferguson (United Kingdom) Ltd,

Stareton,

UK-Kenilworth,

Warwickshire CV8 2LJ;

4. J. I. Case Europe Ltd,

PO Box 121,

Wheatley Hall Road,

UK-South Yorkshire DN2 4PN;

5. John Deere Ltd,

Langar,

UK-Nottingham NG13 8HT;

6. Renault Agriculture Ltd,

Shipston House,

Darlingscote Road,

UK-Shipston-on-Stour,

Warwickshire CV36 4D7;

7. Watveare Ltd,

Headquarters Road,

West Wilts Trading Estate,

UK-Westbury,

Wiltshire BA13 4JY;

8. Fiatagri UK Ltd,

Newmarket Road,

UK-Bury St Edmunds,

Suffolk IP33 3YA;

9. Same-Lamborghini (UK) Ltd,

Barby,

UK-Rugby

Warwickshire CV23 8FA.

(1) OJ No 13, 21. 2. 1962, p. 204/62.

(2) OJ No 127, 20. 8. 1963, p. 2268/63.

(3) The present tense is used here because the notification of 4 January 1988 has not been withdrawn in the light of the amended notification of 12 March 1990.

(4) [1982] ECR, p. 1183.

(5) [1990] ECR II, p. 163.

(6) OJ No L 3, 6. 1. 1987, p. 17.

(7) OJ No L 242, 21. 9. 1977, p. 10.

(8) OJ No L 21, 30. 1. 1979, p. 16.

(9) Paragraph 20, p. 453.

(10) Paragraph 21, p. 2033.