CCE, 5 septembre 2002, n° M.2816

THE COMMISSION OF THE EUROPEAN COMMUNITIES,

Dear Sir/Madam,

1. On 1 July 2002, the Commission received a notification of a proposed concentration pursuant to Article 4 of Council Regulation (EEC) No 4064-89 [1] by which the undertaking Ernst & Young ("Ernst & Young"), via its French entities Ernst & Young Audit SA and HSD Ernst & Young (SELAFA) (collectively "Ernst & Young France"), enter into a full merger within the meaning of Article 3 (1)(a) of the Council Regulation with parts of the French entities of the Andersen network comprising Barbier Frinault et Associés S.A., Barbier Frinault & Autres SAS, Barbier Frinault & Cie SAS, PGA SARL, Arthur Andersen International (SELAFA), SG Archibald, Archibald (SELAFA) and JDP (SELAFA) (collectively "Andersen France"). The notification was declared incomplete on 24 July 2002 and, on 5 August 2002, the Parties completed the information required by the Form CO.

2. After examining the notification, the Commission has concluded that the notified operation falls within the scope of the Council Regulation No 4064-89 and that it does not raise serious doubts as to its compatibility with the common market and with the EEA agreement.

I. THE PARTIES

3. Ernst & Young France is a member of the global Ernst & Young network of accounting and professional services firms, which employs over 83,000 people in 125 countries. Andersen France was active as member firm of the Andersen Worldwide international network of accounting and professional services firms ("Andersen Worldwide"). Until recently the Andersen Worldwide member firms collectively employed approximately 85,000 people. Ernst & Young France as well as Andersen France are therefore the French branches of two of the so-called Big Five worldwide active accounting and professional services organisations. Apart from Andersen Worldwide and Ernst & Young, the other Big Five firms are PricewaterhouseCoopers ("PWC"), Deloitte & Touche Tohmatsu ("DTT"), and KPMG. In this decision, Ernst & Young, DTT, PWC and KPMG will be called the Big Four.

4. Ernst & Young France and Andersen France are both active in the fields of the following professional services: audit and accounting services; tax advisory services; reorganisation services; and legal services through integrated law firms. Ernst & Young is no longer active in the field of business consultancy and the business consulting business carried out by Andersen France does not form part of the transaction.

II. THE OPERATION

5. On 17 April 2002 Ernst & Young France and Andersen France entered into two principal agreements (the "Principal Agreement") concerning (1) the audit and accounting business and (2) the legal and tax advisory business of both undertakings. Furthermore, five letter agreements have been concluded among the Parties providing for temporary measures.

6. The letter agreements cover the following measures: [...]

7. The two Principal Agreements, one each concerning the audit and the legal part of Andersen France follow a similar structure and involve two separate stages: a first stage of "co-operation" between Ernst & Young France and Andersen France and a second stage in which Ernst & Young France and Andersen France will technically merge their activities.

8. For the co-operation stage, the Principal Agreements, in addition to the measures covered by the letter agreements, foresee [...]. The legal structure of the two groups will remain distinct during the co-operation stage.

9. The second stage will be a technical merger between the French entities upon Ernst & Young France's request, which must occur by [...] at the latest. The Principal Agreements already foresee the essential terms and conditions of the merger.

III. CONCENTRATION

10. The operation leads to an amalgamation of Ernst & Young, which forms worldwide one single economic entity, with Andersen France, which after the implosion of Andersen Worldwide's network forms an economic entity independent from all other Andersen Worldwide entities.

Ernst & Young constitutes one single economic entity for the purpose of the application of the Merger Regulation

11. First, it has to be decided whether Ernst & Young's international network constitutes a single economic entity for the purposes of the Merger Regulation and is, therefore, to be considered as an undertaking for the purposes of Article 3 (1)(a) of the Merger Regulation. For this purpose, it is necessary to examine whether the firms belonging to the Ernst & Young network share the same, permanent economic management and financial interests conferring to the network the character of a single economic entity for the purposes of the Merger Regulation.

12. The assessment of whether or not Ernst & Young is to be regarded as a single economic entity has to be made in the special and specific context of the audit and accounting market where there is an inherent economic incentive for member firms of the networks to act as a single economic entity. The key element for the economic success of a Big Five firm is the holding up of an international network, operating under a common brand name, and observing common professional rules and service standards which are centrally imposed on the member firms and centrally controlled. Furthermore, as discussed in a previous decision, [2] each of the networks is de facto considered as one single economic entity showing a single behaviour on the market. Compared to franchise networks in which member firms active in different geographic markets hardly depend on each other, the Big Five member firms vitally depend on the centrally organised network the possibility of working together with their international partner firms and, also, on the reputation of each individual member firm of the network

13. In addition, even where no central distribution of revenues takes place between the individual member firms, strong common financial interest is established by the systematic referral of clients across the network. As the disintegration of the Andersen Worldwide network shows, the failure of one member firm of the network to comply with professional standards puts all other member firms at risk. This leads to the result that risks created by one member firm are shared among the network. This risk sharing is underlined by the existence of one single captive insurer for the business risks for all entities of the entire network.

14. These features are all present for the Ernst & Young network, whose member firms are linked to Ernst & Young Global - the central Ernst & Young entity - through a series of contractual relationships. [...]

15. Members of the Ernst & Young network are moreover obliged [...]

16. The described relationship with Ernst & Young Global applies to Ernst & Young France [...]. The remaining Ernst & Young member firms are integrated into the network via Ernst & Young International, the older and less integrated central governance entity of the Ernst & Young network. [...]

17. Ernst & Young, irrespective whether assessed according to the Ernst & Young International or the Ernst & Young Global structure, can be considered to be a single economic entity. In the Ernst & Young Global structure, a clear permanent economic centralised management is established, supplemented by centrally formulated policies and centrally provided services. In the case of the Ernst & Young International structure, the member firms rely on the common brand name and its reputation, the worldwide network and the centrally developed and monitored professional standards and common client relationships. The strong economic links by sharing of risks [regarding insurance and reputation] and [by sharing of] revenues on the basis of referrals also apply to the Ernst & Young International structure. These links are reinforced by the central co-ordination and facilitation of standards, strategies and initiatives and the provision of common services. As set out in previous cases for DTT and Ernst & Young [3], these elements indicate a decisive degree of common economic management and common financial interest. They lead to the conclusion that Ernst & Young is to be considered as a single economic entity for the purposes of the Merger Regulation, irrespective in which structure this is assessed, and that Ernst & Young as a whole is one of the parties to the present transaction.

Andersen France does not form part of the Andersen Worldwide network for the purpose of the application of the Merger Regulation

18. Second, it has to be decided whether Andersen France (or only parts of it) forms the relevant economic undertaking for the purposes of the Merger Regulation or whether Andersen Worldwide still exists as a single economic entity. The latter conclusion would lead to the result that, in accordance with Article 5 (2) of the Merger Regulation, all transactions between Andersen Worldwide and Ernst & Young would have to be considered as one and the same concentration.

19. The Commission's investigation has shown that Andersen Worldwide is no longer able to discharge its core contractual obligations of co-ordinating the global development of the member firms' practices worldwide, their commercial strategies and their management, as laid down in the contractual relationship between Andersen Worldwide and the national Andersen firms. This is notably demonstrated by the fact that the former member firms of the global Andersen network, each individually and without awaiting Andersen Worldwide's consent, have already joined or seek to join different networks of the remaining Big Four firms on a country-by-country basis. As a result of this disintegration, no central decision making process is any longer in place and the member firms do no longer operate under a common brand name and comply with common professional and service standards. For these reasons, Andersen Worldwide is no longer considered to constitute a single economic entity for the purposes of the Merger Regulation.

20. Furthermore, following the crisis, Andersen France terminated its membership in the Andersen network, effective as of 16 April 2002. In any case, Andersen France considers that it is no longer bound by the terms of the contractual relationship with Andersen Worldwide as a result of a fundamental change of circumstances.

21. Andersen France, in contrast, constitutes a single economic entity and is therefore to be considered as an undertaking for the purposes of Article 3 (1)(a) of the Merger Regulation. Andersen France consists of the French entities of the former Andersen network and is, due to French regulatory requirements, divided into two distinct business units: (1) audit; and (2) tax and legal, acting together as part of a multidisciplinary organisation. [...] Given these elements, the business units of Andersen France share the same, economic management and financial interests conferring to them the character or a single economic entity, in the same way as already set out above for the international Ernst & Young network.

22. Given the above, Andersen France is to be regarded as the relevant economic undertaking for the purposes of the Merger Regulation.

Already the co-operation stage is to be considered as concentration

23. Already the co-operation stage, as first stage foreseen in the Principal Agreements followed by a technical merger, is to be considered as a concentration for the purpose of Article 3 (1)(a) of the Merger Regulation. Already during this stage Ernst & Young and Andersen France share the same, permanent economic management and financial interests conferring to them the character of a single economic entity. The co-operation stage will therefore lead to a de facto amalgamation of Ernst & Young France and Andersen France.

24. In the same way as set out above for the Ernst & Young network in paragraph 12, the assessment of whether or not a concentration between Ernst & Young and Andersen France already occurs in the co-operation stage has to take into account the special and specific context of the audit and accounting market.[...]

25. These legal and economic links created between Ernst & Young and Andersen France will lead to a de facto amalgamation already during the co-operation period. They are therefore sufficient to consider Ernst & Young and Andersen France as a single economic unit under a single economic management in the co-operation period and the transaction constitutes a merger within the meaning of Article 3 (1)(a) of the Merger Regulation [4].

IV. COMMUNITY DIMENSION

26. The undertakings concerned have a combined aggregate worldwide turnover of more than EUR 5 billion [5] (Ernst & Young EUR [...], Andersen France (excluding its consulting business) EUR [...]) in 2001. Each of Ernst & Young and Andersen France have a Community-wide turnover in excess of EUR 250 million (Ernst & Young EUR [...], Andersen France (excluding its consulting business) EUR [...]), but they do not achieve more than two-thirds of their aggregate Community-wide turnover within one and the same Member State. The notified operation therefore has a Community dimension.

V. COMPETITIVE ASSESSMENT

A. Relevant product markets

27. Both Ernst & Young and Andersen, before the disintegration of its network, were considered as two of the "Big Five" global audit and accounting networks. Both parties are active in the provision of a broad range of professional services to clients, which consist to a large extend of large companies, of both a national and multinational dimension, spanning a broad spectrum of business sectors, as well as to clients in the public sector. In a previous decision [6], the Commission has found that the activities of the Big Five firms can be divided into the following relevant product markets:

i) Audit and accounting services to quoted and large companies;

ii) Audit and accounting services to small and medium-sized companies;

iii) Tax advisory and compliance services;

iv) Corporate finance advisory services.

28. Management consultancy services, which were also discussed in IV/M.1016 - Price Waterhouse/Coopers & Lybrand, are not considered in the present decision, as Ernst & Young is not active in management consultancy services and as Andersen France's activity in this field is excluded from the transaction. In addition, the present transaction includes the provision of v) legal advisory services.

i) Audit and accounting services to quoted and large companies

29. Audit and accounting services consist of the performance of statutory and other audits of companies' accounts ("commissariat aux comptes") and other "audit-related" accounting services. In this context, audit-related accounting services include such services as general accounting services, systems assurance, business risks assessment, internal audit, transaction services, in particular due diligence services and preparation of reports for stock exchange listings, treasury, and advice in international accounting standards [7]. The Parties have agreed to such an approach and the market investigation confirmed the categorisation of such audit-related services in the French market.

30. As regards the definition of a separate market for audit and accounting services to quoted and large companies the Commission's market investigation focussed

- on the particularities for the French market in relation to the French regulatory framework for audit and audit related accounting services;

- on the question whether statutory audit and audit related services belong to a single market when taking into account the impact of French particularities; and

- on the question whether audit and accounting services to quoted and large companies, which are predominantly provided by the Big Five global audit and accounting networks, form a distinct product market for the purpose of this decision.

The regulatory framework for audit and accounting services to French companies

31. The French regulations on statutory audit specifically provide for conflict of interest rules. The French Commercial Code clearly specifies that statutory auditors (and any of their partner firms of the same network) are prohibited from exercising certain activities that are deemed to be incompatible with their status. In addition to these legal requirements, the French Accounting Board [8], has published ethical rules specifying, among other things, independence requirements, as well as precisely listing the activities that may be permitted to the statutory auditor and those that are definitely to be avoided. Besides French regulations, audit services to French companies could also be influenced by a new US regulation, the Sarbanes-Oxley Act of 2002, which was approved by the US Congress and signed into law by President Bush on 30 July 2002. This regulation prohibits registered public accounting firms from providing nine types of non-audit related services to clients [9]. According to the Parties, accounting firms providing audit services to French companies listed in the US financial markets are subject to the Act, and it is believed to trigger similar legislation in other countries.

32. Apart from these general features the French legislation on statutory audit services foresees the following two major particularities. First, according to the French Commercial Code companies, which are required to prepare consolidated accounts, must appoint at least two statutory auditors. The two auditors shall not belong to the same company nor even to a common network. In addition, the French stock exchange supervisory body Commission des Opérations de Bourse ("COB"), considers that the two statutory auditors must have an equally significant role. Moreover, the French Commercial Code provides that the statutory auditor is appointed for a (renewable) period of six financial years. Within this period, the statutory auditor can only be removed under very specific circumstances [10].

Statutory audit and audit-related services belong to the same product market

33. Because of the existing French restrictions and the newly created restrictions for French companies listed on the US financial markets, the Commission had to verify whether the range of services relating to audit and accounting constitute a single product market, or if statutory audit should be separated.

34. In this respect the market investigation confirmed that the prohibition to provide some audit related services to their statutory audit clients does not prevent the same firm from offering these services to other clients as similar expertise is needed for the provision of statutory audit and audit related services. Both kinds of services are therefore characterised by a clear supply-side substitutability. In addition, a large number of these services can still be provided together, also by the incumbent statutory auditor. Given these elements, statutory audit and audit related services are currently still to be considered as belonging to the same product market.

Audit and audit related services to quoted and large companies form a distinct product market

35. In previous decisions [11], the Commission identified the following main reasons for which it considered that audit and accounting services to quoted and large companies form part of a separate product market: the necessity for such companies to have audit and accounting services provided by a firm with the necessary reputation in the financial markets (in the case of quoted companies), the geographic breath to cover the companies' needs worldwide (in the case of multinationals), the depth of expertise in the particular sector (large companies in general and, in particular, regulated sectors such as banking and insurance) and significant resources (all large companies). The Commission concluded that only the worldwide active audit and accounting firms known at that time as the Big Six had both the geographic coverage and the degree of reliability with respect to financial statements required by the international capital markets, so that only they operated in the audit and accounting market for quoted and large international companies.

36. The Parties recognise that for the purpose of this decision there is a market for audit services to large companies that is distinct from the market for the provision of audit and accounting services to small and medium-sized companies. However, they also pointed out that large French companies were also purchasing a significant amount of audit and audit-related services from firms that were not Big Five firms. The Parties explained that this was, in particular, related to the French co-auditors system (CO-Commissariat Au Compte system - "COCACs"), according to which large companies and state-owned companies usually request two types of audit services:

- audit services for which Big Five firms are the exclusive providers because they require (i) a professional organisation with an international network, (ii) sufficient credibility as required by the international capital market and (iii) the ability to deploy significant resources with appropriate profile, potentially with a wide geographic coverage ("Big Five audit services"), and

- audit services which may be provided either by large national firms such as Mazars & Guérard or Salustro Reydel or by a Big Five firms ("Big Five and national firms audit services").

The Parties considered that these two types of services were not substitutable but that Non-Big Five firms, however, may have some competitive impact on the Big Five audit firms, within the framework of the COCAC system, or at the time when a client renews a mandate.

37. The market investigation underlined how the French particularities enable second tier firms and even smaller audit firms to get access to large and quoted clients. Especially French auditors such as Mazars & Guérard and Salustro Reydel, and other second tier audit firms such as Constantin, BDO or Grant Thornton International, are therefore able to provide statutory audit services to a number of large companies and large quoted companies in France. However, the market investigation showed that large companies and quoted companies mainly employ a Non-Big Five firm in addition to a Big Five firm and do not consider second tier audit firms as substitutes to their Big Five auditors. Consequently in almost all cases, Non-Big Five audit firms will actually work together with a Big Five firm, in the framework of the French COCAC system [12].

38. In practice the market investigation has shown that if a Non-Big Five Auditor collaborates with a Big Five firm, the split would often be on a geographical basis. For instance, the Big Five firm will be in charge of the international aspects or cover countries where the second tier firm would not have any network. A division by subsidiaries or by business areas was also observed, but always against the background of the Non-Big Five auditor to have the capacity and the resources for successfully providing the services required. The following statement of one large company can be seen as representative for this observation. This company currently employs one Big Five and one Non-Big Five firm as its statutory auditor and stated "we think that we have a good tandem by having as first statutory auditor a well-known French firm, which has a decent international network and above all an intimate knowledge of [our company]. [Our Big Five auditor] brings a powerful network, a strong knowledge of the US techniques (US GAAP, SEC, etc...) and a name that is internationally recognised".

39. This perception was confirmed by Andersen's current large clients addressed in the Commission's market investigation, who expressed unanimously the view that neither Andersen France on its own (without the former Andersen network), nor any of the second tier audit firms was equivalent to a Big Five firm. It was further confirmed by the replies of these companies to the question, whether they would change Andersen in favour of a Big Five firm, in case Andersen France would not be taken over by a Big Five firm. Those firms who, in addition to Andersen, had a Big Five firm as their second statutory auditor said they would not change to a Non-Big Five firm, while those who had, in addition to Andersen France, a Non-Big Five second statutory auditor said that they would have to replace Andersen France by one of the remaining Big Four firms.

40. Finally, the market investigation indicated that the vast majority of customers and competitors recognised a clear difference between the Big Five audit firms and their other competitors. According to customers, there are big differences between Big Five firms and the other auditors, in terms of their international network, their expertise about international accounting standards and their recognition from shareholders, banks and financial markets, especially outside France. Some Non-Big Five firms mentioned that a key difference between the Big Five and others resulted from their financial and commercial strength, which provides them with additional resources in terms of promotion and marketing, training, and research and which makes it possible for them to cope with the possible loss of a client.

Conclusion

41. In conclusion for the purpose of this decision it can be considered that the Big Five audit firms exclusively serve the audit and accounting market for large companies and large quoted companies. Non-Big Five firms can only exert some competitive pressure at the margin, within the scope of the COCAC system. Therefore a distinct market for audit and accounting services to quoted and large companies can be defined. This market definition is in line with the one suggested by the parties and seems to best reflect the particularities of the French market.

ii) Audit and accounting services to small and medium-sized companies

42. The market for audit and accounting services to small and medium-sized companies covers the audit and accounting services to all those companies, who do not require services that can only be provided by one of the global audit and accounting networks. This includes all audit and accounting services in France, for which the Big Five are not the exclusive providers, as set out in the above paragraphs 35-41. The parties accepted this definition and it was confirmed through the investigation that audit and accounting services to all other companies not requiring services, which can be only provided by the Big Five firms, constitute a separate relevant product market.

iii) Tax advisory and compliance services

43. The Parties consider that the provision of tax advisory services is not distinct from the larger market for legal advisory services. According to the Parties this can be documented by the fact that major reviews of the French legal market such as "La Profession Comptable" or "La radiographie 2001 des cabinets d'avocats d'affaires en France" do not make such a distinction when presenting their data on the French legal market. In line with previous decisions, the market investigation has however shown that the definition of a separate product market for tax advisory services is justified for the purpose of this decision. Although according to French legislation all qualified lawyers in France are allowed to provide tax advisory services, the provision of these services requires a strong specialisation and a constant up-date of know-how due to frequent and significant changes in tax legislation. Tax advisory services in France are predominantly provided by specialised tax experts ("conseillers juridiques") and the level of supply-side substitutability with other legal advice is therefore very limited, as it is also observed in other EEA countries.

44. The market investigation showed some features, which could justify a definition of separate markets depending on whether such services were provided predominantly with regard to national tax legislation, or to international issues. Whereas national tax advice can be provided by a firm, which is only active in one country, the provision of international tax advice services can, in principle, be provided either by an international network or by co-operation of different national firms. However, it is not necessary for the purpose of this decision to define the market precisely, as the proposed operation does not lead to the creation or strengthening of a dominant position irrespective of the precise definition retained.

iv) Corporate Finance Advisory Services

45. As regards the market for the provision of corporate finance advisory, the market investigation has confirmed the results of the case IV/M.1016 - Price Waterhouse/Coopers & Lybrand that this area can be considered as a distinct product market.

v) Legal Services

46. The market investigation has shown that there is a separate market for legal advice . It also showed some features, which could justify a definition of separate markets for services provided predominantly with regard to national legislation, and services with international dimension. In the same way as regarding tax advice, the provision of international legal advice can, in principle, be provided either by an international

network or by co-operation of different national firms. The market investigation indicated that for a number of customers the network solution is the preferred choice. However, it is not necessary for the purpose of this decision to define the market precisely, as the proposed operation does not lead to the creation or strengthening of a dominant position irrespective of the precise definition retained.

B. Relevant geographic markets

47. As regards the geographic dimension of the relevant product markets for i) audit and accounting services to quoted and large companies, ii) audit and accounting services to small and medium-sized companies, and iii) tax advisory and compliance services the Commission has stated in case IV/M.1016 - Price Waterhouse/Coopers & Lybrand that these markets are national in scope. The reasons given were in particular significant national differences for the services' main features such as the specific professional expertise required, applicable regulations, and the relevant laws for which the above mentioned advisory services are provided. The market investigation has confirmed the national scope of these product markets, as these features of the market have not changed to date. The same applies for the provision of v) legal advisory services, which can also be considered to be of national dimension.

48. With regard to the geographic markets for iv) corporate finance advisory services, the Commission has already recognised the existence of both national and international aspects for the provision of these services. In the absence of competition concerns, the Commission did not have to conclude on the geographic scope of these markets. As the market investigation has not raised new elements in this respect, and as these two markets do not lead to the creation or the strengthening of a dominant position irrespective of the precise definition of the markets' geographic scope, the Commission does not need to deviate from its previous approach for the purpose of the present case.

49. If a separate market for tax advisory services with international dimension were to be considered, the market investigation has indicated that such a market would show some national characteristics. Tax advice on international matters is linked to the national tax law of the advice seeking company and therefore presupposes knowledge of the national tax regime under which the company is operating. The company will therefore normally liase on matters with an international character with its national tax adviser, who will then usually co-operate with experts - either from the network or from independent co-operating firms - in the jurisdiction where the other end of the operation in question is located. The precise definition of the geographic scope of such a possible market, however, can be left open, as the proposed operation does not create or strengthen a dominant position irrespective of the geographic delineation of the market chosen. The same line of reasoning applies to a possible market for legal advice with international dimension. Also for this possible market the precise delineation of the market can be left open as the proposed operation does not create or strengthen a dominant position irrespective of the geographic definition of the market chosen.

C. assessment

50. The transaction will lead to horizontal overlaps in all the relevant markets mentioned above. The Parties' combined market shares will, however, only exceed 15 % in the French audit and accounting services market for quoted and large companies and the French tax advisory services market with an international dimension, assuming that such a market exists. On all other relevant markets deterioration of competition can be excluded, as the parties combined market shares will be below 15 % and the clients' choice of service providers in these areas is not limited to the Big Five audit and accounting firms.

51. On the audit and accounting services to small and medium-sized companies market, the Parties estimate their market shares to be below 15 % [13]. In any case, there are many alternative suppliers so that no competition issue arises. For corporate finance advisory services, the Parties are in competition with the large investment banks; the parties regard their combined market shares as de minimis given their low turnover in comparison to the value of the transactions achieved. The combined market shares of the Parties for the overall tax advisory and compliance services market is also below 15 % [14] and would be even lower for purely national tax advisory and compliance services, if such a market exists. Regarding the provision of legal services, be it on a national or international basis, the combined market shares of Ernst & Young and Andersen France would be less than 10 % according to the Parties' estimates. The market investigation has confirmed that customers would have no difficulty to find alternative suppliers, since the Parties compete with a large number of law firms providing legal advice, a significant number of which are integrated into international networks.

52. In view of these elements, it can be excluded that the operation could raise competition concerns in the markets for audit and accounting services to small and medium-sized companies, tax advisory and compliance services with no international dimension; corporate finance advisory services and legal services. The competition assessment of the transaction therefore focused on the markets for 1) audit and accounting services to quoted and large companies, and the possible market for 2) tax advisory and compliance services with an international dimension.

1. Market for audit and accounting services for quoted and large companies

a) Market shares in the relevant market

53. During its investigation, the Commission was faced with some difficulties regarding the calculation of market shares for the relevant market for audit services for quoted and large companies. There is no public data available on the French market, except turnover figures for the largest multi-disciplinary audit networks published by the specialised review "La Profession Comptable" on the basis of a survey among these firms. But these statistics do not match the relevant markets defined above, which made it necessary to rely on proxies and on data provided by the Parties and their competitors regarding their turnover for the relevant market. The calculation of market shares for the relevant market has required in a first step the definition of the sample, which gives the best approximation for the French market for audit and accounting services for quoted and large companies. In a second step the best estimates of market shares for this sample has to be established.

54. In order to define a sample of companies that could correspond to quoted and large companies requiring Big Five audit services the Parties and the Commission worked on some common approach. The Parties proposed and the Commission agreed that companies comprising the SBF 120 (the 120 largest French groups listed on the Paris Stock Exchange) plus the 12 largest state-owned companies constituted a suitable proxy, among other things because they represent large companies, with a turnover of at least EUR 1,500 million in 2001. On that basis, two types of calculations were considered.

55. The first method was to calculate market shares on the basis of the number of statutory audit mandates. This method has the advantage to be based on public information, since the name of the statutory auditors of big companies is published in their annual report. However, it only gives a partial picture, since it covers statutory audit only and not other audit-related services, and since it does not take into account the differences linked to the various types of clients, in terms of scope of services, and of level of fees. To address these shortcomings, the Parties have also calculated market shares based on the turnover of Big Five statutory audit clients of the relevant sample and based on the market capitalisation of these clients. However, the clients' turnover or market capitalisation are not necessarily proportionate to auditors' fees [15]; hence a second additional calculation method was used.

56. The second method, developed by the Commission in complement to the first one, was to collect turnover figures from the Parties and from competitors for the relevant market. This method has the advantage that it includes the fees linked to all services provided to the identified sample of companies -not statutory audit only- and that it is based on the actual turnover of the audit firms achieved for the relevant sample. This method has however two disadvantages. First, the data provided by the parties are confidential, so that only ranges can be disclosed for the analysis below. Secondly, and as the parties have pointed out, it may not be sufficiently reliable, because of a lack of precise analytical accounting among the Big Five audit firms for the relevant sample defined and because of the Big Five firms having slightly different ways to classify the various audit-related services among their business units. The following table summarises the market shares resulting from these calculation methods.

Table 1: Calculation of market shares

<EMPLACEMENT TABLEAU>

Sources: Parties' best estimate based on published Annual Reports and publicly data available (method 1); confidential data provided from parties and from competitors (method 2)

57. As can be seen from the table, the second method results in the parties having higher combined market shares than with the first method: more than 50 % against less than 45 %.One explanation for this difference relates to the fact that Andersen France achieves a large turnover outside statutory audit. According to the Parties audit-related services represent more than 50 % of Andersen France's total fees from SBF 120 and large state-owned companies, against only 26 % for Ernst & Young. Non statutory audit work is not reflected in the first calculation method, which only looks at statutory audit clients. Moreover, it is worth pointing out that the resulting market shares do not correspond with the position of the Big Five firms regarding the overall turnover achieved for audit services in France. The magazine La Profession Comptable in March 2002 has published a survey of the 49 largest audit firms in France, with a turnover of each at least EUR 7.6 million. If only the Big Five are considered, and taking out the book-keeping services, which large firms do not require anyway, the comparative size of the Big Five offers the following different picture of the market structure.

Table 2: Comparative turnovers of the Big Five audit firms

<EMPLACEMENT TABLEAU>

Source: La Profession Comptable, March 2002, p.24 column 'Audit and conseil'

58. As it can be seen from Table 2 the merging Parties with regard to the overall audit turnover have a shorter lead over their competitors. The market investigation has shown that one possible reason for the discrepancy between the turnover figures presented in Table 1 and Table 2 was related to the amount of referred work, which means the audit work done in France that originates from audit contracts gained through the Big Five network in other countries. It appears that PwC and KPMG, in particular, make a high turnover from referred work in proportion of their total turnover in France. This is reflected in Table 2 but not in the market shares based on the enlarged SBF 120 sample in Table 1.

b) General characteristics of the market

59. Before assessing the impact of the transaction on competition in the market for audit and accounting services for quoted and large companies, it is useful to outline the conditions of competition in this market. For quoted and large companies, the competition for audit and accounting services in France operates at the moment of competitive tenders, whereby companies invite several firms to submit a proposal. Between 1997 and 2002, the parties found that 24 tenders were launched for SBF 120 and large state owned companies and that for 41 incumbent auditors under review, 25 new auditors were appointed. As set out by a previous Commission decision [16] and confirmed by the investigation, the length of time for which audit appointments are made tend to be long-term -not only because of the 6-year legal appointment- so that the auditor-client relationship can last many years. Even if an audit appointment is very often renewed in favour of the incumbent auditor, launching an invitation to tenders imposes a competitive constraint on the incumbent auditor, often leading to a re-negotiation of the fees. The market investigation has confirmed the existence of this competitive constraint.

60. Moreover, for a bidding market to be competitive, the main requirement is that there exist a sufficient number of credible bidders that are willing to compete. The Commission's investigation showed that the respective market shares among the Big Five in this case did not fully reflect the ability to win a tender. The market investigation has shown that any of the Big Five could possibly win or lose a competitive tender. This is illustrated in the following table, which shows all the bids identified by the Parties and where newly appointed auditors are presented in bold. As can be seen, each of the Big Five audit firms has an equal chance to be invited to tenders. Even though the incumbent auditors tend to be re-appointed, there are a number of changes, and each of the Big Five has won some bids against the others over the past years.

Table 3: Details of past competitive bids on the French market

<EMPLACEMENT TABLEAU>

Source: best estimates of the Parties (annex K, form C.O.)

c) Risk of single dominance

61. As can be seen from Table 1, the merged entity would have a large market share of the French audit and accounting market for large and quoted companies, with a noticeable lead over its competitors. For that reason, the Commission had to investigate further whether the merger may lead to a situation of single dominance in the audit and accounting services market for large and quoted companies. But the following elements show that competition concerns can be excluded with regard to the risk of single dominance.

The merged entity will loose a number of clients, which will inevitably reduce its market shares

62. As a starting point, it is important to notice that the market shares calculated above will be inevitably reduced after the merger. First of all, Andersen France has already lost several clients as a result of the Enron affair. The negative impact of the Enron case on the Andersen network's reputation has already resulted in clients defections that amount to around EUR [...]million, which should be deducted from the turnover calculations in Table 1. Furthermore, the concentration will automatically lead to a loss of clients, no later than end of 2002, due to conflicts of interest between Andersen France and Ernst & Young. These losses are inevitable because, on the one hand, the Parties can no longer be the two statutory auditors for one company under the COCAC system [17] and because, on the other hand, the merger, for a number of other companies of the sample, will lead to incompatibility and independence issues between audit and audit-related services provided by each party for the same company.The Parties have estimated this inevitable reduction of turnover to be EUR [...] million. When deducting these inevitable losses of clients in the turnover of the Parties, and allocating the losses among the competitors according to their overall market shares in the total market for audit, the Commission, using the second method of calculation based on turnover in Table 1, estimates that the Parties' combined market share would be reduced to 40-45 % market shares in terms of turnover, against 20-25 % for PwC, 25-30 % for DTT and 5-10 % for KPMG.

63. Besides, Andersen France will automatically lose referred work, from the reassignment of referred work from ex-Andersen offices, which have joined on a national level some of the other Big Five networks. Even though these losses do not equate to a reduction of market shares, they are estimated by the parties as [...] of Andersen France's audit turnover to large companies, which will diminish their resources and their list of references and therefore potentially weaken their competitive abilities and to the same extent strengthen the other Big Five firms.

64. In addition, the Parties point out that some of Andersen France's existing contracts with large and quoted firms are under threat, as some clients are considered to be unwilling retain the former Andersen partners for their audit and accounting work even after the new entity has been created. This perception was also confirmed by some of the Parties' competitors during the Commission's market investigation, who expressed the view that they may be able to gain additional clients after the merger.

A bidding market with remaining competitive bidders

65. As was already explained in previous decisions [18], the audit market is a bidding market. The Commission tried therefore to clarify whether the other market participants will have the incentives and ability to gain some of the customers that the merged entity will retain, by competing against the merged entity. The Parties provided some elements that show that DTT, KPMG and PwC all meet the level required by clients in order to be considered as credible bidders. For large companies and for quoted companies these firms are considered to have the resources (the Parties estimate that each firm has more than 2000 employees in France), the international credentials and the strong global networks required to be capable to provide audit services for large and quoted companies. This perception is confirmed by the significant amount of audit work referred into France by other members of the international network, which emphasises their global credentials as effective and trusted auditors, and thus reinforces their ability to successfully participate in tenders. It appeared from the market investigation that PwC and KPMG, in particular, achieved high turnovers from referred-in work. Finally, as can be seen from Table 3, each of the Big Five has been systematically invited to tenders for new mandates and has been able to win bids over other Big Five firms in the past years.

66. This perception is also true for KPMG France, which appears as the firm among the Big Five, which has the least number of statutory audit mandates among SBF 120 firms and large state-owned companies. In this respect the Commission investigated whether this lower number of statutory mandates among SBF 120 companies for KPMG constituted a competitive disadvantage in its ability to bid successfully for new audit mandates against other Big Five firms. The investigation, however, showed that KPMG was considered a viable bidder by customers, as for instance illustrated from the fact that it is systematically invited to bid for tenders (see Table 3). The Commission was provided with examples of bids that KPMG had submitted recently to large French companies, which revealed the following main elements First, KPMG can show that it is a credible bidder on the French market through a number of references of large French clients (e.g. TotalFinaElf, Carrefour, EADS and AGF), through its presence in the French professional authorities, and through its overall size (second largest auditor in France). Second, KPMG benefits from the resources and expertise of its international networkand can show it has clients in portfolio, which are large firms in the same sector but with headquarters in different countries. Finally, the Commission was provided with evidence showing that KPMG could possibly use its international resources, and especially global partners, methods and sector expertise in addition to its French resources.

67. All these elements show that the merger will not result in single dominance of the merged entity. Even though Ernst & Young combined with Andersen France may have more mandates in the first place, the gap, in particular after the expected losses, will not be large enough to prevent the other remaining Big Four firms from being credible bidders.

Competition between appointed COCACs

68. Finally, the position of the second tier auditors, and their ability to exert some competitive pressure at the margins, is another element against the possibility of a dominant position to be held by the merged entity. Admittedly, the inevitable reduction from five to four players, in addition to issues of conflicts of interest, will reduce the choice of the large customers. Consequently, it is not clear whether any buyer power could be exerted in the market after the operation, especially in what regards statutory audit mandates.

69. But it appeared from the investigation that the COCAC system, and the presence of two statutory auditors, might provide clients with the means to react to anti-competitive behaviour from the merged entity.Non-statutory audit work represents a large part of the fees earned by the auditors, in addition to their statutory work. Clients could threat to switch part of this work to the second auditor, if the merged entity would try to abuse its position. This is all the more the case as second tier auditors are very often present within the COCAC system, so that they audit large and quoted firms in co-operation with one Big Five firm. It was also noticed that Mazars & Guérard and Salustro-Reydel are sometimes invited to the same tenders as the Big Five firms. Customers could therefore threaten to give more work to second tier auditors. Finally, clients could also threaten to give non-statutory audit work to a competitor. In conclusion, it seems that there may remain alternatives for clients, especially from competition between the two COCACs during their mandates, in the case they would be confronted with any attempted rise of fees or other unilateral behaviour from the merged entity.

Conclusion for the risk of single dominance

70. It appears therefore that the relatively high market shares calculated above do no reflect the actual market power of the merged entity, which will still be faced with competitors that are credible bidders and with customers that may be able to use alternative suppliers. Consequently, it can be concluded that the proposed concentration does not raise single dominance issues.

d) Risk of collective dominance

71. In case IV/M.1016 - Price Waterhouse/Coopers & Lybrand, the Commission considered that the market characteristics of the audit and accounting market for quoted and large companies could, in principle, lead to collective dominance. On the demand side, the Commission considered, in particular, that there was moderate growth in the market and a low price elasticity of demand. On the supply side, the Commission considered the existence of a high concentration, high market transparency for a rather homogeneous product, mature production technology and structural links between competitors. Finally the Commission considered that high barriers to entry characterise this market, in particular, due to the audit firm's required broad geographical network, extensive human resources required, and a well established reputation which has been built up over a significant period of time.

72. However, the Commission found no conclusive proof that such dominance resulted from the relevant operation in the (then) Big Six market. The Commission concluded that 'in view of the continued post-merger existence of no fewer than five suppliers, of the likely continued participation of these five suppliers in the tender offers which constitute the competitive process in the relevant markets, and of the non-emergence of any two clear leading firms following the merger; the Commission has found no conclusive proof that the merger would create or strengthen a position of oligopolistic or duopolistic dominance within any of the national Big Six markets for audit and accounting services to large companies within the Community.' [19]

73. The Commission in previous decisions [20] and the Court of First Instance [21] have identified three conditions which are necessary for a finding of collective dominance:

- each member of the oligopoly must have the ability to know how the other members are behaving (transparency)

- the tacit co-ordination must be sustainable over time; there must be an incentive not to depart from the common policy on the market;

- actual and potential competitors and clients have no possibility to jeopardise the oligopolistic behaviour.

74. In the context of the present proposed concentration, it could be considered whether the current market characteristics in combination with the further reduction of market players from five to four players could lead to sufficient market transparency, to effective retaliation mechanisms and to a lack of threatening response from actual and

potential competitors and clients, these elements resulting in the creation of collective dominance in the audit and accounting market for quoted and large companies. The market investigation has shown that the markets characteristics described in paragraph 72-73 above have not changed in such a way as to exclude the possibility that a reduction from five to four market players could lead to the creation of collective dominance. In particular, the COCAC system, and the intimate links it creates between the auditors across a range of clients may increase transparency and facilitate co-ordination among market players.

75. However, in the context of this decision, it is not necessary to examine in greater detail whether or not the proposed transaction may lead to a situation of collective dominance, as there is no causal link between the proposed operation and the possible situation of collective dominance. As the Commission already found in the previous case COMP/M.2810 - Deloitte Touche Tohmatsu/Andersen UK, the reasons for excluding this causal link are the following:

* the reduction from five to four global accounting networks was inevitable;

* the proposed merger is not more harmful for competition than other possible scenarios as regards the risk for collective dominance on the market for audit and accounting services to large and quoted companies.

Inevitable reduction from Big Five to Big Four

76. In assessing the competitive impact of the operation, the Commission had to take into account the peculiar situation of the global Andersen network, which has already beenoutlined in the Commission's decision COMP/M.2810 - Deloitte Touche Tohmatsu/Andersen UK.

77. The rapid disintegration of Andersen's world-wide network has made it impossiblefor another organisation to use the individual national Andersen entities to recreate a fifth force for the provision of audit and accounting services for large and quoted companies. Even if Andersen France could continue as an independent audit and accounting services firm, the market investigation has shown that Andersen France could no longer exist as a viable competitor in the market for audit and accounting services to quoted and large companies. The Commission's investigation clearly indicated that Andersen France, on its own, was not a Big Five audit firm and that the loss of its network would result -without a merger with another Big Five- in Andersen France losing its large and quoted clients.

78. Moreover, the investigation showed that a take-over by one of the second tier firms, like Salustro-Reydel or Mazars & Guérard, would not result in a firm capable of substituting any of the remaining Big Four, because this would require that the whole, or at least the majority of the national Andersen entities were taken over by such a new entrant. It is excluded that the take-over of Andersen France by any of the second tier would give them the automatic status of a Big Five firm. And the take-over of the entire Andersen network by a new entrant is no longer possible, since, world-wide, the different former Andersen member firms, each individually, have already joined or are about to join the different remaining Big Four networks. Even though the investigation showed that in France, the gap between the Big Five and the second tier auditors may not be as large as for instance in the UK, and that the Enron affair may further bridge the gap, it also came clearly that there was still a big difference, which could not be resolved in the short-run from the take-over of just one national entity.

79. Therefore, Andersen France's business of providing auditing and accounting services to large and quoted companies would inevitably accrue to the existing remaining Big Four.

The proposed merger is not more harmful for competition than other possible scenarios with regard to the risk of collective dominance

80. As outlined in the Commission's decision on Case COMP/M.2810 - Deloitte Touche Tohmatsu/Andersen UK, if the transaction proposed did not take place for any conceivable reasons (such as withdrawal of the notification or regulatory prohibition), only two possible alternative scenarios to the proposed transaction can be established. These two scenarios are:

- the take-over of Andersen France by one of the other remaining Big Four audit and accounting firms;

- no take-over takes place and the existing clients would be dispersed between the remaining Big Four firms (with two sub-scenarios for the attribution of shares).

81. The different scenarios, which could lead to a different repartition of market shares, are set out in Table 4 below. The ranges used in Table 4 make reference to the different approximations set out in Table 1. Scenario I refers to the hypothesis of an addition of market shares following a supposed concentration between one of the other three remaining Big Four firms. This scenario does not take into account of any possible post merger losses of Andersen France in the wake of the Enron affair. In Scenario II, relating to the hypothesis of dispersion of shares, two different alternatives have been considered. The first could be that Andersen France's market share is equally distributed to each of the remaining Big Four firms ("Equal share"). The second is a distribution on an equi-proportional basis, where Andersen France's market share is transferred to the remaining Big Four in proportion to their respective current market shares ("Equi-proportional share"). Given the difficulties explained above regarding the calculation of market shares, the market shares used are those on the basis of the number of mandates, which are based on public data. But in addition, as a matter of completeness, the total possible range [22] is presented in brackets.

Table 4. Hypothetical market shares for the different scenarios

<EMPLACEMENT TABLEAU>

Source: Calculation on the basis of market shares as calculated in Table 1

82. As indicated above, the reduction of market players from five to four appears to be inevitable. When examining all different possible situations described in scenarios I and II, the only relevant difference between them and the proposed operation for the competition analysis is the different repartition of market shares. However, as already stated above, market shares are of less importance in the audit and accounting services market for large and quoted companies, as almost all acquisition of new clients takes place via bidding procedures launched by the clients.

83. It can be seen from Table 3 that, apart from a hypothetical merger between Andersen France and KPMG and the hypothetical dispersion of Andersen France, in all other different situations analysed in scenario I and II, that the differences in the repartition of market shares regarding all possible alternatives to the operation are rather marginal compared to the market structure resulting from the proposed merger,. A hypothetical merger between Andersen France and KPMG and the hypothetical dispersion of Andersen France, however, would actually result in a more symmetric market structure than the proposed operation and therefore a potentially more problematic one with regard to collective dominance issues.

84. To exclude that any of the other alternatives could create a market structure, which could be less harmful for competition, it must be assessed whether the differences in the repartition of market shares (which would be created as a result of the proposed operation) could more likely create a situation of collective dominance than any of the possible alternatives. If this is not the case, it can be excluded that there would be a causal link between the merger proposed and a possible deterioration of the competitive structure in the market resulting from it.

85. In this respect, the market investigation has shown that, compared to the two scenarios, the operation proposed will not

- increase the likelihood that each member of the oligopoly will have the ability to know how the other members are behaving (transparency) -first condition-

- increase the likelihood that the tacit co-ordination would be sustainable over time and that there would be an incentive not to depart from the common policy on the market -second condition-;

- increase the likelihood that competitors and clients would have no possibility to jeopardise the oligopolistic behaviour -third condition-.

First condition

86. The proposed transaction does not have a different impact on the ability to monitor the competitors' behaviour from any of the other possible scenarios. The possible increase in market transparency would result from the inevitable reduction of the market players from five to four. It could not result from small differences in market shares between the competitors.. In these circumstances the degree of market transparency will not be different as a result of the proposed transaction or of any other scenario.

87. In addition, if according to scenario II, Andersen France's staff was dispersed and any further and key persons were equally spread among the Big Four firms, tacit collusion would be even more facilitated by this spread of common experience and by the creation of a network of personal links.

Second condition

88. Second, it can be excluded that the marginal differences between the repartition of market shares between any alternative and the proposed transaction could change the incentives of the Big Four firms to depart from a possible common policy on the market. In fact, the market structure resulting from the proposed operation is not particularly symmetric since the merged entity would become a clear leader. Scenario II would also create a more symmetric market structure, which could in theory increase the risk of collusive behaviour. In addition in all alternative situations presented above, the four remaining competitors will have similar cost structures, since they will all have reached a critical mass in a market where the main resources are human resources and where there are limited economies of scale, once a firm passed a minimum size. The proposed operation is not more harmful in this respect than any of the other scenarios.

Third condition

89. Finally, a different repartition of market shares between the remaining Big Four companies does not have any effect on the barriers to enter the now Big Four audit and accounting market and does therefore not affect the possibility of competitors to jeopardise the oligopolistic behaviour. The French COCAC system already provides potential entrant in this market with the means to exert some competition at the margins. Neither would a difference in the repartition of market shares change the client's choice whether to invite one or another Big Four firm to a tender. In any case, the choice is reduced from five to four. Hence the clients' ability to challenge a possible oligopoly is not different in any possible scenario compared to the proposed operation.

Conclusion for the risk of collective dominance

90. In conclusion, it can be excluded that neither scenario I nor scenario II, which represent the only other possible alternatives to the proposed transaction, could create a market structure which could, compared with the market structure resulting from the proposed transaction, decrease the likelihood for collective dominance on the market for audit and accounting services for quoted and large companies. Therefore, there is no causal link between the proposed operation and a risk of collective dominance that would result from it.

2. Market for Tax Advisory services

91. The market investigation has shown that it could be considered that the market for tax advisory services might be further divided between international tax advisory services, and national tax advisory services.

92. In this case, the market for international tax advisory services would be similar to the market for the provision of audit and accounting services to large and quoted companies, in that the service requires an international network and there are few providers on the market. The market investigation has shown, however, that the Big Five are not the only service providers on this possible market for international tax advisory services in France. To a considerable extent big law firms, which either have an international network comparable to the one of the Big Five or usually co-operate with law firms in other countries are or at least could be active on this market.

93. Given the fact that a considerable number of big international firms are active on the same market, no competition concerns can arise in the possible market for tax advisory services with an international dimension. Furthermore, if such a market exists, this market would largely coincide with the market for audit and accounting services to international firms. Thus, the same reasoning and assessment can be carried out as above. Therefore, the transaction will not lead to the creation or strengthening of a dominant position on such a market.

VI. CONCLUSION

94. For the above reasons, the Commission has decided not to oppose the notified operation and to declare it compatible with the common market and with the EEA Agreement. This decision is adopted in application of Article 6(1)(b) of Council Regulation (EEC) No 4064-89.

[1] OJ L 395, 30.12.1989 p. 1; corrigendum OJ L 257 of 21.9.1990, p. 13; Regulation as last amended by Regulation (EC) No 1310/97 (OJ L 180, 9. 7. 1997, p. 1, corrigendum OJ L 40, 13.2.1998, p. 17).

[2] Case IV/M.1016 - Price Waterhouse/Coopers & Lybrand, paragraph 118, OJ - L 050, 26/02/1999 p.27

[3] Case COMP/M.2810 - Deloitte Touche Tohmatsu/Andersen UK; case COMP/M.2824 Ernst & Young / Andersen Germany

[4] see Commission Notice on the concept of concentration under the Merger Regulation, paragraph 7.

[5] Turnover calculated in accordance with Article 5(1) of the Merger Regulation and the Commission Notice on the calculation of turnover (OJ C66, 2.3.1998, p25). To the extent that figures include turnover for the period before 1.1.1999, they are calculated on the basis of average ECU exchange rates and translated into EUR on a one-for-one basis.

[6] IV/M.1016 - Price Waterhouse/Coopers & Lybrand quoted above, paragraphs 20,22

[7] IV/M.1016 - Price Waterhouse/Coopers & Lybrand; COMP/M.2810 - Deloitte & Touche/Andersen UK

[8] la Compagnie Nationale des Commissaires aux Comptes', CNCC

[9] (i) bookkeeping services relating to accounting records or financial statements of the client; (ii) financial information systems design and implementation; (iii) appraisal or valuation of services and fairness opinions; (iv) actuarial services; (v) internal audit outsourcing services; (vi) management functions or human resources; (vii) broker-dealer, investment adviser or investment banking services; (viii) legal or expert services unrelated to the audit; and (ix) any other service that the SEC Board determines to be impermissible.

[10] The functions of the statutory auditor can only be terminated as a result of his resignation, the non-renewal of his function, the objection with good reasons requested by either one or several shareholders representing at least 5 % of the share capital, the Public Prosecutor or the COB (for listed companies).

[11] Cases COMP/M.2810 - Deloitte & Touche/Andersen UK; IV/M.1016 - Price Waterhouse/Coopers & Lybrand,

[12] The limited number of large companies and large quoted companies that only have Non-Big Five statutory auditors have mostly very limited activities outside France. They therefore do not need the specific elements that only Big Five audit firms are able to provide, in particular, a large international network, international reputation and expertise about international accounting standards.

[13] on the basis of some figures taken from La Profession Comptable, March 2002

[14] on the basis of a 2002 study "Le guide des cabinets d'avocats d'affaires", which presents the top 150 business law firms with turnover data

[15] The Parties, however, have argued that such turnover data, in particular, were not accurate to assess the market position of the Big Five firms on the affected French market for audit and accounting services to large companies.

[16] IV/M.1016 - Price Waterhouse/Coopers & Lybrand, quoted above, paragraphs 20,22

[17] e.g. for Société Générale, which is currently audited by Ernst & Young and Andersen France

[18] COMP/M.2810 - Deloitte & Touche/Andersen UK; COMP/M.2824 - Ernst&Young Germany / Andersen France; IV/M.1016 - Price Waterhouse/Coopers & Lybrand

[19] Price Waterhouse/Coopers & Lybrand, quoted above, paragraph 119

[20] Case IV/M.1741 MCI Worldcom/Sprint paragraph 259

[21] Case T-342/99 of 6 June 2002 Airtours v Commission, paragraph 62

[22] to calculate the total range, addition of market shares were made line by line in Table 1, in order to be consistent with the methods used in the calculations