EC, December 19, 2007, No 34.579

BUSINESS SECRETS ALLEGED BY MASTER CARD - EXPLANATION OF THE REDACTIONS IN THE NON-CONFIDENTIAL VERSION OF THE DECISION:

- Business secret la: information pertaining to MasterCard's organisational and management structure, general membership rules.

- Business secret lb: extracts (e.g. from minutes/rules) pertaining to MasterCard's organisational and management structure, general membership rules.

- Business secret 2a: information pertaining to MasterCard's interchange fee rules and procedure for determining the level of the interchange fee.

- Business secret 2b: extracts (e.g. from minutes) pertaining to MasterCard's interchange fee rules and procedure for determining the level of the interchange fee.

- Business secret 3a: information pertaining to the reasons, strategy and preparation of MasterCard's IPO.

- Business secrets 3b: extracts (e.g. from minutes/presentations) pertaining to the reasons, strategy and preparation of MasterCard's IPO.

- Business secret 4: information pertaining to MasterCard sales, turnover, market shares and number of customers.

- Business secret 5a: information pertaining to MasterCard internal commercial strategy.

- Business secret 5b: extracts (e.g. from board minutes/rules/presentations/proposals) pertaining to MasterCard internal commercial strategy.

- Business secret 6: information pertaining to the MIF, tariff evaluation, underlying cost structures and other considerations taken into account for the purpose of setting the MIF, including the list of countries where the MIF applies by default to domestic transactions.

- Business secret 7: information pertaining to exchanges between counsel.

- Business secret 8; information pertaining to studies/ expert reports submitted by MasterCard to the Commission.

THE COMMISSION OF THE EUROPEAN COMMUNITIES

Having regard to Council Regulation (EC) No 1-2003 of 16 December 2002 on the implementation of the rules on competition laid down in Articles 81 and 82 of the Treaty (1), and in particular Article 7 as well as Article 24 (1) (a) thereof, Having initiated proceedings by issuing a Statement of Objections on 6 May 1999 and having expressed preliminary concerns with respect to the matters addressed in this decision in two subsequent Statement of Objections of 24 September 2003 and of 20 June 2006, and having informed MasterCard Europe Sprl, MasterCard International Incorporated and MasterCard Incorporated of new evidence by a letter of facts of 23 March 2007, Having given MasterCard Europe Sprl, MasterCard International Incorporated and MasterCard Incorporated the opportunity to make known its views on the objections raised by the Commission pursuant to Article 19 (1) of Regulation No 17-62 (2) and Commission Regulation (EC) No 2842-98 of 22 December 1998 on the hearing of parties in certain proceedings under Articles 85 and 86 of the Treaty (3), as well as pursuant to Article 27 of Regulation (EC) No 1-2003 and Article 12 of Commission Regulation (EC) No 773-2004 (4), After consulting the Advisory Committee on Restrictive Practices and Dominant Positions, Having regard to the final report of the Hearing Officer in the present case (5), Whereas:

Executive Summary

Article 81 (1) of the Treaty

The present case deals with MasterCard's network rules and decisions of its member bank delegates and the organisation's management on Intra-EEA fallback interchange fees and SEPA fallback interchange fees. These multilateral interchange fees (MIFs) are retained for each payment card transaction and apply to cross-border card payments in the European Economic Area (EEA) and domestic card payments within several Member States of the EEA with MasterCard or Maestro branded payment cards. The fees are "fallback" fees in the sense that they are only charged in the absence of specific bilateral agreements between an issuing and an acquiring bank on interchange fees. Under the organisation's network rules the acquiring banks pay the fees to the issuing banks.

2. The MIF in MasterCard's scheme restricts competition between acquiring banks by inflating the base on which acquiring banks set charges to merchants and thereby setting a floor under the merchant fee. In the absence of the multilateral interchange fee the merchant fees set by acquiring banks would be lower.

3. The European Board [and/or] the Global Board [of MasterCard] [business secret - 2a] decisions on the level and structure of Intra-EEA fallback interchange fees and the related network rules adopted by the Global Board of MasterCard International Inc are decisions of an association of undertakings within the meaning of Article 81(1) of the Treaty. They remain decisions of an association also after the Initial Public Offering (IPO) of MasterCard Incorporated on 25 May 2006 and the related changes in the governance of the payment organisation in Europe with regard to the authority for setting the level of multilateral fallback interchange fees. The banks agreed to the IPO and the ensuing changes in the organisation's governance in order to perpetuate the MIF as part of the business model in a form which they perceived to be less exposed to antitrust scrutiny. Even after the IPO of MasterCard Incorporated interchange fees in the MasterCard organisation are not "unilaterally imposed". MasterCard's member banks still share a common interest as regards the MIF because it yields guaranteed revenues for their issuing business. There is thus a continuing commonality of the banks' interests in a MIF after the IPO which is also reflected in the setting of interchange fee rates by MasterCard [business secret - 2a] In setting interchange fee rates the Global Board cannot ignore the commercial interests of the banks without whom the system would not function, because it yields guaranteed revenues for their issuing business.

4. An open payment card scheme such as MasterCard's can operate without a MIF as is evidenced by the existence of comparable open payment card schemes without a MIF.

Article 81 (3) of the Treaty

5. As to the first condition of Article 81 (3) of the Treaty MasterCard's arguments on the existence of objective efficiencies remained unproven.

6. The Commission does not dispute in general that payment systems are characterized by indirect network externalities and that in theory a revenue transfer between issuing and acquiring banks may help optimize the utility of the network to its users. However, whether a collectively fixed interchange fee should flow from acquirers to issuers or vice versa, and at which level it should be set cannot be determined in a general manner by economic theory alone, as theories always rely on assumptions that may not sufficiently reflect market reality. Rather, any claim that a MIF creates efficiencies within the first condition of Article 81 (3) of the Treaty must be founded on a detailed, robust and compelling analysis that relies in its assumptions and deductions on empirical data and facts. MasterCard in particular failed to provide empirical evidence for its central claim that the MIF maximizes the scheme's "output" and for a causal link to other objective efficiencies claimed. It was therefore not established that the restrictive effects of the MIF in the acquiring markets are duly offset by objective efficiencies. Despite repeated requests by the Commission MasterCard failed to submit any empirical evidence on the positive effects of its MIF on system output and related efficiencies.

7. The Commission does not share MasterCard's view that the competitive process and market forces automatically lead to a MIF that can safely be assumes to be efficiency enhancing. Neither the forces of inter-system competition nor acquiring banks or merchants exert sufficient constraints on the body in charge of setting the MIF in the MasterCard organisation.

8. The specific framework underlying MasterCard's MIF is a model written by William Baxter in 1983. This model is, however, severely limited by the fact that it takes consumer and merchant demand as given in that neither strategically reacts to possible actions by the other. The Baxter model also relies on the unrealistic assumption that there is no variation in the benefits that merchants perceive from accepting cards, this means that merchants are homogeneous. Baxter's result finally relies on the unrealistic assumption of a perfectly competitive banking industry.

9. Moreover, the methodologies used by MasterCard for implementing this framework in practice are unconvincing as they do not sufficiently reflect the underlying theory. The methodologies suffer from considerable shortcomings as they establish an imbalance between card issuing and merchants acquiring solely on the basis of cost considerations while omitting to consider the banks' revenues, as well. Moreover, contrary to the merchant demand analysis, cardholders and simply assumes the relative unwillingness of this customer group to pay for the convenience of using payment cards. There are also doubts as to the usefulness of MasterCard's proxy for quantifying the willingness to pay of merchants in the credit card segment.

10. As to the second condition used of Article 81(3) of the Treaty all customer groups in a payment card system, that is cardholders and merchants (as well as subsequent purchasers) must get a fair share of the benefits which result from the efficiencies of a MIF. While merchants may benefit through enhanced network effects from the issuing side, this does not necessarily offset their losses which result from paying inflated merchant fees. The Commission bas therefore reviewed MasterCard's methodology [business secret - 2a] However, MasterCard's cost based benchmarks include cost items that are neither intrinsic in the payment functionality of a card nor related to services that clearly benefit the customers that bear the expenses of this MIF. Without further evidence, which MasterCard failed to submit, it cannot therefore safely be assumed that by maximising system output MasterCard is equally benefiting its member banks' customers. MasterCard failed to demonstrate that efficiencies outweigh restrictions to merchants (as well as subsequent purchasers).

11. As to the third condition of Article 81(3) of the Treaty MasterCard has not proven to the requisite standard that its current MIF is indispensable to maximise system output.

12. Due to unrealistic assumptions underlying the conceptual underpinnings of MasterCard's MIF, due to the lack of evidence for a causal link between this MIF and objective efficiencies claimed and due to the fact that MasterCard's methodologies do not sufficiently reflect the underlying framework and operate with inflated cost benchmarks, the Commission concludes that such MIF does not fulfil the first force conditions of Article 81(3) of the Treaty.

Remedy

13. As MasterCard's MIF restricts price competition between acquiring banks without fulfilling the first three conditions of Article 81(3) of the Treaty the Commission orders MasterCard to withdraw its intra-EEA and SEPA/intra-Eurozone fallback interchange fees within six months upon adoption of this decision. This remedy excludes one aspect of MasterCard's MIF as far as commercial cards are concerned. The Commission will further research the possibility of efficiencies in this respect. The order does not prevent MasterCard [business secret - 2a] from adopting an entirely new MIF (other than the Intra-EEA fallback interchange fees and the SEPA/intra-Eurozone fallback interchange fees) that can clearly be proven to fulfil the four cumulative conditions of Article 81(3) of the Treaty based on solid empirical evidence.

14. For this order to be effective, the Commission imposes provisional periodic penalty payments on MasterCard if MasterCard fails to comply with the remedy after the six months transition period lapses. Moreover, to speed up the pass-on of the ensuing gains for acquiring banks to merchants and subsequent purchasers, MasterCard is ordered to publish specific information on its website on the abolition of its Intra-EEA fallback interchange fees until a non confidential version of the decision is publicly available. MasterCard is finally ordered to inform all Clearing Houses and member banks in Europe on the abolition of this MIF.

Glossary

Acquiring bank: a credit institution that has a contractual relation with a merchant for accepting a certain payment card at a POS.

Association's network rules: the Amended and Restated Bylaws and Rules of MasterCard incorporated, the Amended and Restated Bylaws and Rules of MasterCard International Incorporated and the Interchange and Service Fees Manual as well as the Maestro Global Rules.

Bylaws MCI: Amended and Restated Bylaws of MasterCard International Incorporated.

Bylaws MCII: Amended and Restated Bylaws of MasterCard International Inc.

Cardholder: a consumer that holds a payment card as an instrument for payment at a POS. Cardholders may be natural or legal persons. Legal persons hold cards for employees who then use such cards for payments for the account of the legal person (their employer).

Cardholder fee: typically a yearly flat paid by a cardholder to the issuing bank for the use of a payment card.

Clearing: a service to member banks that is needed for a payment transaction to be settled. It occurs after a payment card transaction has been authorized by the issuing bank.

Clearing House: refers to a bank which sends information on successful transactions (typically) in batched form (that is to say in a package of messages) to the acquiring bank for crediting on the merchant account and to the issuing bank for debiting on the cardholder account.

Clearing "at par": a payment card transaction is cleared at the face value of the claim. Neither the issuing nor the acquiring bank receive or pay an interchange fee.

COO: Chief Operating Officer. The CCO is a manager of MasterCard International Inc. [business secret - 2a]

Cross-border interchange fees: the Intra-EEA fallback interchange fees.

Cross-border (payment card) transaction: a payment card transaction that occurs between an issuing bank and an acquiring bank that are located in different countries.

Domestic (payment card) transaction: a payment card transaction that occurs between an issuing

Bank and an acquiring bank in the same country.

Domestic MIF: See at interchange fee

European Board: a board of European bank delegates whose meetings are organized by staff of MasterCard Sprl and which set intra-EEA fallback interchange fees until the IPO of MasterCard Incorporated on 25 May 2006.

Global board refers at the same time to the Board of Directors of MasterCard International Incorporated and to the Board of Directors of MasterCard Incorporated whose meetings are typically held at the same time due to the (partial) identify of the managers involved.

Honour All Cards Rule (HACR): A network rule of MasterCard that obliges merchants to accept all valid MasterCard and Maestro branded cards and transactions equally and without discrimination according to the type of card used and the bank issuing the card. For details see section 3.2.

Interchange fee refers to a fee paid by an acquiring bank to an issuing bank (or vice versa) for each POS payment card transaction. MasterCard refers to "fallback" or "default" interchange fees for those interchange fees that apply only in the absence of other agreements on interchange fees between the issuing and the acquiring bank. An interchange fee rate can be expressed as an ad valorem (a percentage) rate and/or a flat fee per payment.

Intra-EEA fallback interchange fees: interchange fees that apply in the MasterCard payment organisation as "fallback" to cross-border and/or domestic payment card transactions with MasterCard or Maestro branded payment cards between Member States of the European Economic Area or within a Member State of the European Economic Area. As "fallback" means that the Intra-EEA fallback interchange fees apply only if the payment is not subject to (i) a bilateral agreement between the acquiring batik and the issuing batik involved in the payment ; (ii) a multilateral agreement between delegates on a national forum of MasterCard member banks which determines a specific interchange fee for the payment. MasterCard called Intra-EEA fallback interchange fees initially "intra-EU fallback interchange fees" when the fees were notified in 1992.

Intra-Eurozone fallback interchange fees: See at SEPA fallback interchange fees.

IPO of MasterCard Incorporated: Initial Public Offering of MasterCard Incorporated on 25 May 2006 when a part of its shares were for the first time publicly traded at the New York Stock Exchange.

Issuing bank: a credit institution that has a contractual relation with a cardholder which allows for the provision and use of a payment card.

Legal entities representing the MasterCard payment organisation: MasterCard Incorporated and its fully consolidated subsidiaries MasterCard International Incorporated and MasterCard Europe Spr1.

MasterCard: unless further specified references to "MasterCard" in the subsequent text are to be understood as referring to the MasterCard payment organisation. The organisation is managed and represented by MasterCard Incorporated, MasterCard International Incorporated and MasterCard Europe Spr1.

Merchant: an entity that accept a payments by means of cards. Merchants can be retailers but also other undertakings such as airlines.

Merchant service charge (MSC) or merchant fee or merchant discount rate: a fee paid for each transaction by a merchant to an acquirer for accepting a card for payment. MSCs can be expressed as ad valorem (a percentage) rate and/or a flat fee per payment.

MIF stands for multilateral interchange fee. Multilateral interchange fees are based on a collective agreement between member banks of a payment association or the decision of a body [business secret - 2a] within that association which has been empowered by the member banks to determine the level and structure of interchange fees with binding force for them. A MIF is commonly distinguished from bilateral interchange fees which are agreed upon between two parties only, that is to say the issuing and the acquiring bank.

MasterCard MIF is used as reference to the organisation's network rules and the decisions of [MasterCard] [business secret - 2a] that determine the Intra-EEA fallback interchange fees and the SEPA fallback interchange fees are multilateral interchange fees. They are based on commonly agreed network rules, on collective decisions of bank delegates and (since 2006) on the decisions of MasterCard] [business secret - 2a] whose decisions member banks of the association agreed to abide by the MasterCard MIF is the subject of this decision.

Domestic MIF (or Intra-country fallback interchange fees) are interchange fees that apply in the MasterCard or Maestro branded payment cards within a Member State of the European Economic Area. They apply to such domestic transactions unless the payments are subject to a bilateral agreement between the enquiring bank and the issuing bank involved. A domestic MIF replaces the Intra-EEA fallback interchange fees as fallback rate.

Network externalities: There is no commonly agreed definition of network externalities, but this concept is generally said to refer to a situation where the benefit of a consumer from purchasing a good service increases the more other consumers purchase the same good/services. In other words, there is a positive relationship between network size and customer valuation. Indirect network externalities refer to the fact that two separate customer groups exist in two-sided industries. The customer groups do not directly benefit from the size of their own group but rather indirectly from the size of their counterparts.

Payment cards: plastic cards that offer cardholders the possibility of making payments for goods and services at a POS. Industry commonly distinguish the following card types:

i. Debit card: a card that allows the cardholder to charge purchases directly and individually to a current account at a deposit-taking institution. The debit card serves as a device to access funds stored at bank accounts.

ii. Credit card: a card that allows the cardholder to make purchases with a certain credit amount which can be settled in full by the end of a specified period (which typically is interest free) or can be settled in part, with the remaining balance taken as credit and being charged with interest. Credit cards may be linked to a current account at a deposit taking bank, or they may be linked to an account that has been set up specifically for the use of the credit card.

iii. Charge card (or delayed debit card): a card that allows the cardholder to make purchases but does not offer credit, the full amount of the debt incurred having to be settled only after a specified period (which typically is interest free).

iv. Consumer Credit and Charge cards: Credit and charge cards which banks directly issue to natural persons. In the MasterCard organisation such cards currently carry the MasterCard, MasterCard Signia or MasterCard Electronic logos.

v. Commercial Credit and Charge cards: Credit and charge cards which banks issue to undertakings or governments and that are limited in use for business expenses of employees or civil servants. In the MasterCard organisation such cards carry the MasterCard Corporate, MasterCard Electronic Corporate, MasterCard BusinessCard, MasterCard Electronic BusinessCard, MasterCard Purchasing and MasterCard Fleetcard logos.

Payment card system (or payment card scheme or payment card network): a technical and commercial infrastructure set up to serve one or more particular card brands for payments at a Point of Sale.

Pay Later MIF: stands for Intra-EEA fallback interchange fees for payment card transactions with MasterCard branded credit and charge cards.

Pay Now MIF: stands for Intra-EEA fallback interchange fees for payment card transactions with MasterCard or Maestro branded debit cards.

POS stands for Point of Sale, (typically) a merchant outlet with a terminal or (exceptionally) another facility that allows card payments via phone (m-payments) or the internet (e-payments). The POS payment card transactions are deemed to be "domestic" transactions where the cardholder and the merchant are located in the same EEA Member State while they are deemed to be "cross- border" transactions where the cardholder is located in another EEA Member State than the merchant. POS card usage is to be distinguished from ATM card usage where a card is used for withdrawing money at an Automatic Teller Machine.

On-us and off-us transactions: On-us transactions occur if the issuing batik and the acquiring batik are identical for a card payment. In a larger sense industry uses the notion of on-us payments also where the issuing and the acquiring bank belong to the same group of banks. Off-us transactions occur where the issuer and the acquirer are different institutions.

Processing: the activity of handling electronic data that are related to a payment card transaction.

Rules MCII: Rules of MasterCard International Incorporated.

SEPA fallback interchange fees: interchange fees that may apply in the future in the MasterCard payment organisation as 'Fallback" to cross-border and/or domestic payment card transactions with Maestro branded payment cards between Member States of the Eurozone or within a Member State of the Eurozone. SEPA fallback interchange fees would only apply if the payment is not subject to a bilateral or multilateral agreement between the member banks on other specific interchange fees. According to initial announcements of MasterCard, these fallback interchange fees were to apply from 1 January 2008.

However, in June 2007 MasterCard announced that the date of their implementation was delayed. On 8 November 2007 MasterCard adopted "Intra-Eurozone fallback interchange fees" which are to apply as of 15 January 2008. The rates of these new fallback interchange fees are largely identical with those of the SEPA fallback interchange fees but they apply only as fallback to cross-border payments but not also to domestic card payments.

Settlement: the activity of deducting and crediting funds on accounts of banks after clearing. Typically, payment card systems have one central settlement bank (in MasterCard's system [bank] in the United Kingdom, domestic schemes typically use a national central bank) where all members of the card association maintain an account. On these accounts all card transactions during a certain period (e.g.: one day) are netted and settled.

Store card: a payment card that is issued by non-banking institutions and can be used in specific stores for payment. Store cards can be debit, charge or credit cards. Bank cards yo the contrary are cards issued by credit institutions.

1. Chronology and Background of the Procedure

The procedure

15. On 30 March 1992 the British Retail Consortium (hereafter " BRC "), a trade association representing United Kingdom retailers, submitted a complaint to the Commission amongst others against Europay International SA (hereafter " Europay "). In that complaint the BRC alleged that Europay and Visa both restricted competition due to their arrangements on cross-border interchange fees (6). BRC withdrew its complaint on 15 November 1999 in the light of a similar complaint by EuroCommerce which was lodged on 23 May 1997. EuroCommerce is a retail, wholesale and international trade representation in the European Union. Its complaint is directed against the major payment card systems operating in the EEA, amongst them Europay (now MasterCard Europe Sprl.) and Visa. The complaint addresses certain practices of those card systems, in particular multilaterally agreed interchange fees, the so-called No Discrimination Rule, the Honour All Cards Rule, price discrimination between merchants in levels of commissions charged by acquiring batiks and anti-competitive concentrations in the field of card acquiring (7).

16. On 22 May 1992, Europay notified the network rules of the Europay payment card association with respect to edc/Maestro branded debit cards to the European Commission. This notification included network rules and agreements/decisions on so-called intra-European interchange fees for Maestro branded payment cards (8). In May 1993 Europay submitted a second notification of its network rules for Eurocard-MasterCard branded cards (9). That notification again included network rifles and agreements/decisions on intra-European interchange fees for Eurocard-MasterCard branded credit and charge cards. On 8 September 1994 Europay submitted a third notification of its rules governing cash withdrawals with payment cards (ATM services) (11). On 22 May 1995 Europay notified its Signia program, that is to say an upmarket version of the Eurocard-MasterCard credit and charge cards program in Spain (12). Finally, the Commission received a notification by Europay effective as from 1 July 1995 in respect of all its payment systems (13).

17. On 6 May 1999 the Commission sent Europay a Statement of Objections concerning specific network rules of the Europay payment card system, namely the No-Discrimination Rules ("NDR"), the principle of territorial licensing, the Europay Rules on cross-border issuing and acquiring, the notified exclusivities rules, the "No Acquiring Without Issuing" Rule and the "Honour All Cards Rule" ("HACR"). This Statement of Objections did not yet concern the Intra-EEA fallback interchange fees. Europay responded to the Statement of objections on 28 August 1999.

18. On 13 April 2002 the Commission published a notice pursuant to Article 19 (3) of Regulation N° 17. In that Notice the Commission announced its intention to adopt a favorable position, amongst others, with respect to Europay's NDR, its HACR and the cross-border licensing rules which had been modified subsequent to the Statement of Objections of 6 May 1999 (14).

19. On 22 November 2002, the Commission opened an ex-officio investigation regarding Visa's and MasterCard's respective Intra-EEA fallback interchange fees for commercial card (15).

20. On 25 July 2003 MasterCard informed the Commission of its intention to initiate proceedings under Article 232 of the Treaty unless the Commission took a formal position with respect to MasterCard's Intra-EEA fallback interchange fees.

21. On 24 September 2003 the Commission sent a Statement of Objections to MasterCard Europe Sprl, the legal successor of Europay, as well as to MasterCard International Inc and MasterCard Incorporated addressing MasterCard's network rules and decisions on Intra-EEA fallback interchangefees.

22. On 23 October 2003 the Hearing Officer granted Visa leave to intervene as third party with a sufficient interest to be heard within the meaning of Article 13 (1) of regulation (EC) N° 773-2004 (16).

23. On 21 June 2006 the Commission sent a Supplementary Statement of Objections to MasterCard Europe Sprl, to MasterCard International Inc and to MasterCard Incorporated again addressing MasterCard's network rules and decisions on Intra-EEA fallback interchangefees. That supplementary Statement of objections complemented the preceding Statement of Objections of 24 September 2003 with a more profound analysis of the restrictive effects of MasterCard's Intra-EEA fallback interchange fees on competition.

24. On 3 and 4 July 2006 MasterCard exercised its right of access to the file by copying the paper file. On 22 August 2006 MasterCard was again provided access to the paper file upon requests of its counsels who had omitted to copy some parts of the file on 3 and 4 July.

25. In the subsequent weeks the Hearing Officer admitted 14 entities as third parties with a sufficient interest to be heard within the meaning of Article 13 (1) of Regulation 773-2004: The Royal Bank of Scotland, Airplus, three Spanish payment card systems (Servired SC, Sistema 4B SA, Euro 6000 SA), the British Retail Consortium (BRC), a Belgian retail chain (Kappé International BV), British Petroleum (BP), the European Retail Roundtable (ERRT), Hotels, Restaurants and Cafés in Europe (HOTREC), an Austrian association of hotels (OHV), the International Air Transport Association (IATA), Royal Ahold and Ikea. Visa, Airplus, OHV, Hotrec, and IATA submitted written comments.

26. On 14 and 15 November 2006 MasterCard exercised its right to be heard in an oral hearing. The hearing was also attended by EuroCommerce as well as by nine third parties (17).

27. On 23 March 2007 the Commission sent a letter to MasterCard ("letter of facts") providing MasterCard with access to documents collected since MasterCard had access to the file in July and August 2006 and setting out possible conclusions the Commission intended to draw from the new facts in a decision.

28. On 16 and 22 May 2007 MasterCard replied to the letter of facts.

1.2. The Commission's investigation

29. Since 1992 the Commission collected empirical market information on the effects of the MasterCard MIF by sending requests for information to third parties as well as to MasterCard and EuroCommerce. Information collected from 218 merchants in 2004 was computed and analysed in a "merchant market survey" which MasterCard obtained as an annex to the Supplementary Statement of Objections of 20 June 2006. The Commission held a series of meetings with market participants in particular during 2004 in order to gain a deeper understanding of the payment cards indust7 and to evaluate the effects of the intra-EEA interchange fees on competition (18).

30. The Commission repeatedly met with MasterCard and with the complainant to discuss their submissions and views. The Commission met with EuroCommerce on 11 November 2003, 19 March 2003 and 5 January 2005, in Spring 2005, on 4 July 2005 and on 14 May 2007. The Commission met with MasterCard on 27 January 2004, 18 February 2004, 27 July 2004 12 November 2004, 11 January 2005, 15 September 2005 and 25 May 2007.

31. MasterCard, EuroCommerce and third parties submitted written comments as well as opinions of economic consultants and academics in favour or against the use of multilateral interchange fees (19). The Commission studied these submissions in detail.

General Background

The Visa decisions of 2001 and 2002

32. In its 2001 Visa I decision (20) the Commission granted a negative clearance to a number of rules in the Visa system (the no-discrimination rule or NDR (21), rules on cross-border issuing, on cross-border acquiring, territorial licensing, the no acquiring without issuing rule and Visa's HACR).

33. In its 2002 Visa II decision (22), the Commission exempted Visa's intra-regional MIF in the European Union from Article 81 of the Treaty for a time period of five years subject to certain conditions, of which the main one was that the MIF be linked to, and capped at, certain costs. In its analysis pursuant to Article 81 (1) of the Treaty, the Commission rejected the argument that a MIF was essential for the existence of the Visa system. The Commission characterized the MIF as "an agreement between competitors, which restricts the freedom of banks individually to decide their own pricing policies, and distorts the conditions of competition on the Visa issuing and acquiring markets" (23).

34. The exemption was based on an offer of reforms that Visa made regarding three main elements: firstly, Visa was to progressively reduce the level of its MIFs. Secondly, the MIFs were to be capped at the level of costs for three specific "services... provided by Visa issuing banks to merchants indirectly, via the acquiring bank (24)". Those "services" were transaction processing, payment guarantee and free funding period. The costs of those services were to be determined by a cost study, and the "cost cap" would apply cumulatively with reductions in the level of the MIF offered by Visa. Thirdly, Visa was to allow member banks to reveal information about the MIF levels and the relative percentage of the three cost categories (previously considered business secrets) to retailers at their request (25).

35. The 2002 Visa II exemption decision expires on 31 December 2007.

36. On 13 June 2005 the Commission decided to open a "sector inquiry" in the area of retail banking (26). The inquiry focused on payment cards and current accounts and other services, and the Final Report was published on 31 January 2007 (27).

37. Although this decision relates to an area - that of payment cards - which falls within the scope of the sector inquiry, the Commission bas kept the case file in this case and the material gathered during the sector inquiry completely separate.

13.3. The creation of a Single Euro Payments Area

38. The project to create a Single Euro Payments Areas ("SEPA") is a self-regulated initiative of the European banking industry to move from separate national payment areas to an integrated Euro payment area. The European Payments Council ("EPC") is the coordination and decision making body for all payment issues.

39. SEPA covers direct debits, credit transfers and payment cards. Unlike for direct debits and credit transfers, the EPC chose not to create a new pan-European scheme for payment cards, mostly because international schemes already exist and operate in Europe. The EPC's "SEPA Cards Framework" (SCF) represents - as the name indicates - a framework as to how banks, card schemes and processors are to operate in the Single Euro Payments Area.

2. Parties to the proceedings

2.1 The MasterCard payment organisation

40. MasterCard is a worldwide payment organisation that is represented by legal entities, two of which are registered in the US - MasterCard Incorporated and MasterCard International Inc, and one in Europe - MasterCard Europe Sprl. This payment organization groups several thousand banks that issue cards and/or acquire merchants. In the EEA there are about [business secret 4] member banks.

41. As MasterCard's structure and governance are of importance for the legal assessment of the MasterCard MIF, the subsequent sections set out both the oragnisation's structure and governance in detail. MasterCard takes the view that on 25 May 2006 - the Initial Public Offering of MasterCard Incorporated (cf. section 2.1.4) - the organisation's nature changed fundamentally. The subsequent sections will therefore distinguish between the time period prior to and after that date.

2.1.1 The Conception of Membership

The Membership concept until 25 May 2006

42. Until 25 May 2006 the MasterCard payment organization was explicitly based on a concept of membership. Any bank wishing to issue MasterCard/Maestro branched payment cards and/or endeavoring to sign up merchants for MasterCard/Maestro card acceptance had to apply for membership in the MasterCard organization (via membership in MasterCard International Incorporated). Any entity that was not eligible for such membership or that elected not to become a member, and that engaged in or proposed to engage in any activity in which only a member may engage, either had to take action necessary to meet the membership eligibility requirements or cease any activity in which only a member may engage.

43. Only members of MasterCard International Inc. were granted a license to issue MasterCard/Maestro branded payments cards and/or to sign up merchants for MasterCard/Maestro card acceptance (28). By entering into the license agreement the member banks became subject to the MasterCard network rules. These rules regulate, among others, the setting of multilateral interchange fees and contain the HACR. Membership applications in Europe were reviewed by the membership. These were three such classes of card membership: Association/Principal/Affiliate membership. Each class of membership had different rights and obligations in terms of decision making and the level of membership fees.

The membership concept after 25 May 2006

44. The IPO of MasterCard Incorporated on 25 May 2006 MasterCard did not materially affect the concept of membership. Management now refers to banks as "licensees" rather than members, but the membership concept essentially remains the same. Member banks were already "licensees" of MasterCard International Incorporated before the IPO and they remain "members" of MasterCard International Incorporated after the IPO (29). As before the IPO only "members" can obtain a license for issuing or acquiring MasterCard payment products (30). The rules regarding the membership conditions and membership classes in MasterCard's Restated and Amended Bylaws remained equally almost unchanged (31). As before the IPO, any application for a new membership from a European bank continues to be subject to the review by bank directors on the European Board (32). The license agreements concluded between MasterCard International Inc and the member banks before the IPO remain unaltered after the IPO (33). Banks, moreover, are still groups within different classes of membership (namely Association/Principal/Affiliate) after the IPO.

2.1.2 Decentralised Structure and Decision Making

45. The MasterCard payment organization has a de-centralised governance. Member banks co-ordinate themselves in various ways, ranging from bilateral and multilateral agreements to decisions which bodies of bank-appointed directors [business secret-2 a] take with binding force for all members. This co-ordination of member banks' commercial conduct takes place on several levels: nationally, regionally and globally.

The MasterCard payment organization until 25 May 2006

46. Until 25 May 2006 the MasterCard payment organization operated world-wide in a de-centralised manner. In particular, MasterCard's brand and product strategies were managed on a regional level under the direction of several regional boards composed of member banks. Member banks were groups into six regions: Asia/Pacific, Canada, Latin America, Middle East/Africa, the United States and Europe. The Europe region comprised, amongst others, all Member States of the European Economic Area, as well as other countries such as Turkey. The regional board of the Europe region was called Europe Regional Board or European Board (henceforth "European Board") (34).

47. The European Board had the same powers as any other regional board (35). The European Board had powers to set domestic fallback interchange fees if member banks did not agree on them bilaterally or multilaterally (36). Moreover, the European Board set all-Intra-EEA fallback interchange fees for MasterCard and Maestro branded payment cards until 25 May 2006. Most of these rates are still in force today. As any other Board, the European Board also had authority to manage other important decisions such as decisions on membership applications, imposing fines and disciplinary actions for infringements of network rules, policies or procedures including recommendations to terminate the membership of a bank, approving the annual expense budget, fixing, imposing and collecting of membership fees, determining the distribution of any revenues in excess of budgeted amounts, approving inter-regional and intra-regional interchange fees, the product development and setting of co-branding rules (37).

48. Delegates on the European Board had to be directors of member banks holding a position as decision maker for the entire member bank or at least for the credit, debit or charge cards business of such bank.

49. Decisions of the European Board were binding upon member banks unless MasterCard Incorporated's Global Board decided with a majority of at least 66 2/3% of the Directors present at the meeting to overrule the decision for exceptional reasons (38).

The MasterCard payment organization as of 25 May 2006

a. The European Board

50. On 25 May 2006 MasterCard's member banks abolished all regional boards except for the European Board by approving the governance changes resulting from the European Board by approving the governance changes resulting from the IPO of MasterCard Incorporated. The European Board retained major decision making powers and MasterCard's management publicly insisted on the continuing importance of local input in Europe: We "value the input we receive from regional boards and will maintain a regional board with specified authorities in Europe as well as regional advisory boards (...) (39).

51. European banks therefore still co-ordinate their behavior through the European Board. Moreover, member bank's representatives now also serve on global and regional "advisory board" in the MasterCard organization. In Europe such advisory boards were created to group member banks in Western Europe, Southern Europe and Central Europe (40).

52. As MasterCard highlights in an internal memorandum the European Board continues to decide upon "key issues in Europe" (41). Those key issues are (1) the review of applications for membership, (2) fines, (3) intraregional operating rules, (4) assessment and fees to the extent that such assessments and fees do not have an exclusionary effect, (5) intraregional products and enhancement development to the extent that the development initiatives do not relate to competitively sensitive matters, (6) annual expense budget, (7) surplus funds, and (8) affinity and co-branding rules. With respect to matters listed in points (1)-(4) the European Board excises its powers under the guidelines of the Global Board "from time to time" (42).

53. The European Board still takes decisions independently, being bound only in relation to some of these matters by the guidelines given by the Global Board given from time to time (43). As MasterCard Incorporated explained publicly: "certain aspects of our European operations will be managed by the European Board (...) which may reach different decisions than our Global Board of Directors (...) Although our board of directors may, through majority or two thirds vote authority granted to the European Board, the European Board may reach different decisions than our board of directors would have reached on the same matter" (44).

54. However, the banks decided to delegate the authority for setting Intra-EEA fallback interchange fees from the European Board to [business secret - 2a] MasterCard Incorporated. The powers of the European Board before and after 25 May 2006 are - as MasterCard summarises - "identical (...) minus interchange fee" (45). Details on the background of this change in the organisation's governance are set out in section 2.1.4.

55. As before 25 May 2006 the consent of European banks is a necessary precondition for the Global Board to assume the powers of the European Board and to terminate the Board's activity. The Global Board may do so only by a majority of 75% of its directors and needs the consent of the European Class M stock holders (that is to say, European member banks) (46) voting separately as a class.

56. Decisions of the European Board also continue to be binding upon MasterCard's member as banks unless and until the Global Board decides with a majority of at least 66 2/3% of the Directors present at the meeting to overrule the decision (47). The Global Board can only overrule decisions of the European Board in very exceptional cases (48).

57. Following the IPO, the character of the European Board as a board of European bank representatives (49) also remained unchanged. The European Board now consists of between 12 and 25 delegates. The class M. Director representing European banks within the Global Board is automatically a member of the European Board (50). The participation of this representative of European member banks in both boards - the European and the Global Board - facilitates the exchange of views between the European Board can easily become aware of the views of MasterCard's European members and take account of them in its decision making process.

b. National For a of Member Banks in Europe

58. Besides retaining "key" decision making powers within the European Board, European banks also retained significant powers to co-ordinate their businesses in sub-regional boards, which MasterCard refers to as "for a" of member banks. They exist in countries such as in the United Kingdom, Germany, Norway and Poland.

59. MasterCard management positively encourages this form of de-centralised horizontal decision making in the payment organization. In a presentation titled "future governance of MasterCard in Europe" (51), MasterCard's management declares its support for the further development of national for a in countries where they were already in the process of being launched (the Netherlands, Denmark, Hungary, Finland) and were they could potentially be launched (Italy, Belgium (52)). Similar encouragement was given to Irish banks (53).

60. Whether or not banks choose to formality set up a country "forum", MasterCard's network rules foresee that banks can always agree upon specific national network rules that apply in a given market and that then replace parts of the global network rules. According to the MasterCard and Maestro rules (54), members holding Principal and Affiliate MasterCard/Maestro) licenses for a country (including SEPA licenses) and representing, during the year preceding the agreement, at least [business secret - 1a] of each of MasterCard/Maestro issuing and acquiring intra-country volumes, have the power to agree on fallback rules intra-country volumes, have the power to agree on fallback rules to all intra-country transactions, including those acquired by members outside the country.

61. On this basis, country by country local banks have set sometimes very extensive and detailed rules covering the matters such as the level and structure of domestic fallback interchange fees, chargeback and settlement rules (55).

Importantly, MasterCard Europe Sprl, MasterCard Incorporated and MasterCard International Incorporated do not need to endorse or certify domestic rules which are adopted by local banks. The only authority that

62. (57), (58), (59) [business secret - 2a]

2.1.3. The Legal entities representing the MasterCard Payment Organisation

63. Besides the European Board and national for a of member banks which manage the members' co-operation in the payment organization on a regional (section 2.1.2.a) and national level (section 2.1.2 b) the payment organization is also managed, and represented, by MasterCard Incorporated and its consolidated subsidiaries MasterCard International Incorporated and MasterCard Europe Sprl.

2.1.3.1 MasterCard Incorporated

64. MasterCard Incorporated is a US stock corporation with its registered office in the city of Wilmington, State of Deleware. In 2002 MasterCard Incorporated was converted from a membership association into a private stock corporation whose shares were entirely held by MasterCard member banks (the Principal Members) (60). Since 25 May 2006 MasterCard Incorporated has been listed on the New York Stock Exchange (NYSE) (the "IPO". It acts as holding company for its fully consolidated subsidiaries MasterCard International Inc and MasterCard Europe Sprl.

65. On 25 May 2006 the bank shareholders of MasterCard Incorporated decided to change the company's corporate governance.

a. Corporate governance before 25 May 2006

66. Until the IPO on 25 May 2006 there were two classes of shares in MasterCard International, namely Class A stock (59% of economic rights and 100% of voting rights) and class B (41% of economic rights and no voting rights). Only Principal member banks (61) could hold Class A stock.

67. Both MasterCard Incorporated and its subsidiary MasterCard International Inc were managed by the "Global Board" (62) which consisted of 18 persons. One third of these directors were officers of member banks in the Europe Region, one third were officers of member banks residing in the other four world regions (Asia/Pacific, Canada, Latin America, Middle East/Africa) (63).

b. Corporate governance after 25 May 2006

68. Following the IPO, MasterCard Incorporated became a publicly traded company with its capital reclassified from two into three categories of shares : (64)

69. Class A stock, listed and traded on the NYSE, represents in the aggregate, a 59% economic interest and 100% of the voting stock. This stock is held as follows: (i) 49% of economic interest (representing 83.05% voting interest) by public investors (65), and (ii) 10% of economic interest (representing 16.95% voting interest) by the MasterCard Foundation.

70. Glass B stock is held by MasterCard's Principal member banks and represents a

41% economic interest, without any voting rights.

71. Glass M stock is again held by Principal member batiks (one share each). Class M stock represents the right to elect three out of twelve Directors on the Global Board and to approve certain extraordinary matters (for example, termination of the European Board). The voting rights attached to M Stock also carry minimal minority protection rights (66) (for example, right of approval for the sale of all the assets, a merger of the company, or liquidation of its business). Glass M Stock carries no economic rights and is automatically cancelled if transferred.

72. Since the IPO MasterCard Incorporated's Global Board consists of 12 voting directors and one non-voting "advisor". The voting directors comprise the following: (i) eight directors elected by Glass A stock holders (that is to say, public investors and the MasterCard Foundation) and (ii) three directors nominated by Class M Stock holders who are representatives of MasterCard's customer financial institutions from the Americas region, Europe Region and Asia/Pacific, Middle East and Africa region, and (iii) the CEO, who is also a director. Directors are appointed by the shareholders in a general meeting. The majority of directors on the Global Board are said to fulfil the NYSE criteria for independence (67).

2.1.3.2. MasterCard International mc.

73. MasterCard International Incorporated is a US based membersh4, corporation (both before and after the IPO) (68). Its office is in the city of Wilmington, State of Deleware and its management body is (also) the Global Board (see above, section 2.1.3.1).

74. There are three classes of memberships in MasterCard International Incorporated (both before and after the IPO): (i) Class A Membership without any voting rights, (ii) Class B Membership - issued to MasterCard Incorporated solely - representing exclusive voting rights on all matters and entitlement to all dividends and other distributions approved by the Board, and (iii) other Classes of Membership (Affiliate Membership) with rights, preferences and privileges determined by the Board. The Board of Directors is elected by Class B Members (that is to say by MasterCard Incorporated) at the annual meeting of members.

2.1.3.3 MasterCard Europe Sprl

75. MasterCard Europe Sprl is a fully consolidated subsidiary of MasterCard Incorporated having its registered seat in Waterloo, Belgium. The Board of Directors of MasterCard Europe Sprl is composed of three directors: the President of MasterCard Europe Sprl, one representative of MasterCard Incorporated and one representative of MasterCard Incorporated. The company is the legal successor of the undertaking that notified the fallback interchange fees on behalf of the payment organization in 1992, that is to say Europay International SA (69).

2.1.4 The IPO of MasterCard Incorporated

76. One key reason for the IPO of MasterCard Incorporated was to modify the organisation's governance to allow its member banks and the legal entities managing it to better address intensifying regulatory and legal scrutiny of the MasterCard MIF while ensuring that member banks in Europe continue to exercise an important influence on the organisation's every day business (70).

a. Banks and management considered solutions to antitrust scrutiny

77. It would appear that the possibility of changing the governance of MasterCard Incorporated was for the first time discussed on 20 November 2003 when, following the Wal-Mart settlement (71) in the USA, the Global Board contemplated ownership and governance concern issues at a meeting.

(72), (73), (74), (75) [business secret - 3a]

b. Bank appointed directors on Global Board decide in favour of modifications to governance

78. (76) [business secret - 3a]

79. (77) [business secret - 3a]

80. (78) [business secret - 3a]

81. (79), (80)

c. Structural solutions considered and close involvement of member banks in shaping new governance of MasterCard Incorporated

82. (81) [business secret - 3b]

83. [business secret - 3a]

84. (83), (84), (85), (86) [business secret - 3a]

85. (87), (88)

i, ii, iii (89), iv [business secret - 3a]

86. (90), (91)

87. (92), (93) [business secret - 3a]

88. (94), (95), (96), (97)

89. (98), (99) (100)

90. (101)

91. (102), (103) [business secret - 3a]

92. (104), (105), (106), (107) [business secret - 3a]

d. Shareholder banks approve IPO to maintain MIF in business model

93. (108), (109), (110) [business secret - 3a]

94. (111), (112) [business secret - 3a]

95. (113) [business secret - 3a]

96. (114) [business secret - 3a]

97. In the Registration Statement filed with the US Securities and Exchange Commission (henceforth "the SEC"), MasterCard Incorporated publicly explained the reasons for the transition to the new ownership and governance structure (115). It referred to several trends as the migration of consumers and businesses to electronic forms of payment, the consolidation and globalization of MasterCard's customer financial institutions, the growing role of merchants and increased competition, but also to the regulatory scrutiny of their business model; "Many of the legal and regulatory challenges we face are in part directed at our current ownership and governance structure, in which our customers - our member financial institutions - own all of our common stock and are involved in our governance by having representatives serve on our global and regional boards of directors. While we strongly dispute these challenges, we believe that a more open ownership and governance structure following the offering transaction should leave us less prone to challenges and provide us with additional defences to the challenges that we may face (...)"

98. (116) [business secret - 2a]

f. Conclusion

99. The banks were a driving force behind the IPO of MasterCard Incorporated. They agreed to it as they knew that the new management on the Global Board would continue to act in their common interest. While the European banks were aware that they would loose control over the body setting Intra-EEA fallback interchange fees, they consented to the change in the organisation's governance with the expectation that the independence of the Global Board would reduce each individual bank's exposure to regulatory scrutiny antitrust litigations, thereby safeguarding their proceeds from interchange and maintaining the MasterCard MIF as part of the business model. Considering the IPO in its overall context, the banks voting in favour of the corporate changes to MasterCard Incorporated "outsourced" the co-ordination of their competitive behavior to an independent body while maintaining control of decision making on other key matters that are not sensitive from an antitrust perspective.

The market position of the MasterCard payment organization

a. Volume and value of MasterCard/Maestro payments

MasterCard cards

100. Both the number and value of MasterCard cross-border transactions in the Community are growing fairly steadily. In the five years 2000 to 2005 the value of cross-border payments increased by (30-50%] from EUR [5 000-15 000] million (2000) to EUR [ [...] 000]-15 000] million (2005) (117). This represents an average year increase of [[...] 0-10]%.

< table >

Maestro cards

101. In the period from 2000 to 2005 the value of Maestro cross-border transactions in the Community increased by [200-400]% from EUR [500-2000] million (2000) to EUR [3 000- 5 000] Mio (2005). This is an average year to year increase of [20-40]%.

< table >

b. Revenues from Intra-EEA fallback interchange fees

102. The total value of intra-EEA interchange fees paid by acquirers to issuers steadily increased up to 2005.

MasterCard cards

103. Between 2000 and 2005 the revenues of issuing banks from Intra-EEA fallback interchange fees due to cross-border payments with MasterCard branded cards million to EUR [150-200] million (118). This is an average year to year increase in interchange fee revenues/cost of [5-20]%.

<table>

104. Between 2000 and 2005 croos-border interchange fee revenues for Maestro branded cards increased by [200-500] % from EUR [5-10] Mio (2000) to EUR [10-50] Mio (2005) (119). This is an average year to year increase in interchange fee revenues/cost of [20-50]%.

< table >

c. MasterCard's market position

ca. General considerations

105. The primary focus for analyzing whether there is an appreciable effect on competition is the position and importance of the parties on the market taking into account the market structure (120).

In assessing MasterCard's market position and therefore the impact that its decisions have on competition, it is necessary to understand the different elements of a payment card scheme's market position include, in particular, the size of its membership network, the penetration amongst different classes of customers (cardholders, merchants), the strong acceptance rate and high transaction volumes and the relative size compared to rival networks.

106. Therefore, the market position of MasterCard as a network cannot only be measured in terms of a single set of market shares. The Commission also takes into account the network economies of MasterCard's cards business (based on an assessment on the number of payment cards currently issued by MasterCard member banks, the size of MasterCard's membership and acceptance network and on transaction volumes) and the size and growth rate of MasterCard's network.

107. Like Visa, MasterCard is characterized by important network economies, which stem from its large cardholder base and its large merchant acceptance network. Together with Visa, MasterCard's issuing and acceptance networks are unique in the EEA. The economics of network industries such as payment cards imply that established incumbent networks benefit from a strong competitive advantage over any rival network. The size of MasterCard's network is matched only by one other network, Visa.

108. MasterCard's market share (in terms of total cards issued) is significant on average and MasterCard is in the process of further increasing its market share by taking over national payment schemes in the EEA. Examples involve the migration of Bankomatkarte to Maestro in Austria (1999), of EC Karte to Maestro in Switzerland (2003), of Switch to Maestro in the United Kingdom (2004/05) (121), the ongoing efforts of MasterCard and Visa to achieve a migration of EC cash card in Germany and the [business secret - 5a] to MasterCard and Visa branded charge cards in [business secret - 5a]. Details are set out in section 8.2.2.4

cb) Numbers of issued MasterCard/Maestro cards

109. An independent industry report gives the following figures on international payment cards issued in Western Europe (EU 15 (122) plus Norway, Turkey and Switzerland) for 2004 (123). In this area, MasterCard held a market share of 57.20% in 2004 which compares to a market share of 57.30% in 2002 (124). The strong market position of the MasterCard association therefore appears stable.

110. In the entire EEA, MasterCard's penetration ratio in terms of all cards issued is currently approximately 45%. This implies that 45% of all cards issued in the EEA bear either a MasterCard or a Maestro logo.

< diagram 1 >

111. The diagram 1 is also that only 8% of all cards in circulation in Europe are "pure" domestic cards, that is to say cards not yet carrying one of the international logos in addition to a national logo.

112. MasterCard had the following country specific penetration ratios in terms of cards issued in EEA Member State in 2004.

< table 5 >

113. In July MasterCard stated on its website that there are a total of 350 million MasterCard and Maestro cards in circulation in Europe (125). 258.3 million of these cards are Maestro-branded cards (126).

114. According to MasterCard its Maestro debit cards are accepted at about 85% of all merchants accepting debit cards in 29 EEA Member States (127) and the number of merchant locations accepting Maestro branded cards stood at over 6.1 million at the end of 2005 (128).

115. An independent industry report estimated that 4.9 million merchants accepted MasterCard branded credit and charge cards in 2004 and that 3.9 million merchants accepted Maestro cards in the same year. The acceptance network of Visa seems equally strong: 4.9 million merchants accepted Visa cards and 4.3 million merchants accepted Visa Electron cards in 2004. This means that most merchants accepting MasterCard credit cards also accept Visa credit cards, a phenomenon commonly referred to as "duality" (129). The acceptance networks of American Express and Dinners Club lags far behind those of MasterCard and Visa with American Express being accepted at 2.4 million merchants outlets and Diners at 1.9 million outlets. In diagram 2 the large size of MasterCard's acceptance network becomes visible (130).

< diagram 2 >

cd) Strong membership network

The MasterCard association has more than [business secret - 4] bank issuing MasterCard/Maestro payment cards in the EEA (131). In the period of 1997 to 2002 around [business secret - 4] new banks obtain issuing (or combined issuing and acquiring) licenses from MasterCard for the European Union every year (132).

116. The MasterCard cites figures on country specific transaction volumes in its submission of 5 January 2005 to suggest that its market share is insignificant in most EEA Member States. According to these figures MasterCard's market shares are below [...]% in [business secret - 4] EEA Member States and above [...]% in [business secret - 4] EEA Member States with the highest market share in Austria [...] [business secret - 4] (133).

117. The reliability of MasterCard's data is, however doubtful. It appears that MasterCard's data on volumes are based on pure estimates. MasterCard's management has its of own data on MasterCard/Maestro transaction volumes only for countries where the association processes domestic card transactions itself. With the exception of a few EEA Member States (134), MasterCard does not process its own transactions and therefore has no direct access to data. MasterCard's management does not even have a complete overview of which member banks actually issue cards in Europe (135). MasterCard's management also collects information on transactions volumes form only a few of the [business secret - 4] member banks through so-called Quarterly Member Reports. It is unclear to what extent the data on volumes provided by MasterCard are based on those reports, to what extent MasterCard management "completes" that information with estimates, and on which sampling techniques its estimates are based.

3. Subject of this decision

118. This decision addresses the MasterCard MIF. The MasterCard MIF is a decision of an association of undertakings comprising MasterCard's network rules and decisions of the organization [business secret - 2a] on Intra-EEA fallback interchange fees and SEPA/intra-Eurozone fallback interchange fees which apply to virtually all cross-border payment card transactions and to some domestic payment card transactions (136) with MasterCard and Maestro branded payment cards in the European Economic Area. The scope of this decision excludes on the one hand bilaterally agreed interchange fees and on the other hand domestic fallback interchange fees, [business secret - 2a]. The present decision moreover deals with certain aspects of MasterCard's "Honour All Cards" Rule which enhances the restrictive effects of the MasterCard MIF.

3.1 The MasterCard MIF

3.1.1 MasterCard's longstanding " Fallback " Interchange Fee Mechanism

119. The MasterCard MIF is anchored in the MCI Bylaws and Rules, the MCII Bylaws and Rules, the Interchange and Service Fees Manual as well as in the Maestro Global Rules that are all issued by MasterCard International Inc (137).

These rules (138) determine the principle that acquiring banks must pay issuing banks an interchange fee for each POS payment card transaction with a MasterCard or Maestro branded payment card except if the banks involved in the transaction bilaterally agreed to clear and settle at other conditions. MasterCard does not dispute that in theory interchange fees could as well flow from issuing banks to acquiring banks.

120. The level of interchange fees in the MasterCard organisation can be determined in several manners and the options available to member banks increased notably after the IPO of MasterCard Incorporated. As a general principle, issuing and acquiring banks can always agree upon interchange fees bilaterally. These fees then take precedence over any other fees that bodies of [MasterCard] [business secret - 2a] within the organisation may set. However, in the absence of a bilateral agreement a " fallback " (also referred to as " default ") interchange fee with apply to the transaction.

121. MasterCard traditionally distinguishes several types of " fallback " interchange fees ranging from inter-regional fees to intra-regional fees and intra-country fees with the more specific fees taking precedence over the others (139).

122. Intra-EEA fallback interchange fees in particular are multilateral interchange fees that apply as " fallback " to :

i. Cross-border MasterCard and Maestro card payments between EEA Member States incase no other interchange fees are defines through bilateral agreements ; and to

ii. Domestic MasterCard and Maestro card payments within EEA Member States in case no other interchange fees are defined through bilateral agreements or through multilateral agreements between banks delegates (" domestic MIFs " ; see right below).

123. The level and the structure of Intra-EEA fallback interchange fees, which are the subject of this decision, are set out in Annex 1.

124. In contrast, intra-country fallback interchange fees (hereafter "domestic MIFs") apply to domestic card payments within EEA Member States. Issuing and acquiring banks holding MasterCard or Maestro licenses " have the power " (140) to agree on fallback interchange fee programs applicable to all intra-country transactions. A decision to adopt such domestic fallback interchange fees requires [business secret - 2b) (141). Those fees then replace Intra-EEA fallback interchange fees as relevant " fallback " for payment transactions in the country concerned. Being a mere " fallback ", those domestic MIFs only apply in the absence of bilateral agreements.

125. It is important to understand that MasterCard's " fallback mechanism ensures that some interchange fee must always apply to a MasterCard/Maestro payment card transaction. As the level of intra-EEA interchange fees is a positive value, the fallback mechanism excludes the possibility that in the absence of an agreement between member banks on interchange fees a card transaction is cleared and settled at the face value of the claim (" at par ").

3.2.1 MasterCard's amendments to its " fallback " Mechanism post-IPO

a. new fallback fees for the Eurozone and for Eastern Europe

126. On 4 December 2006 MasterCard publicly announced new " SEPA fallback interchange fees " for the Maestro debit card (142) as an initiative to achieve compliance with the European Payment Council's SEPA Cards Framework (143). These fees form part of this decision.

127. The fees were set at EUR 0.05 + 0.20% of the transaction value for Chip & Pin card payments (144) with " large merchants " benefiting from a special reduction to EUR 0.03 + 0.12% of the transaction value (145). According to MasterCard the SEPA fees would initially apply to Maestro euro payments within or between Member States of the Eurozone. In some countries, however, domestic MIFs would initially differ based on local arrangements " (146).

128. However, events shortly after the publication of MasterCard's SEPA fallback interchange fees led to a postponement of their application on the market. On 30 March 2007 Febelfin, the Belgium financial sector federation, announced that the migration of the domestic debit card Bancontact/Mr. Cash to Maestro cards would not take place, as initially foreseen, on 1 January 2008. MasterCard subsequently announced that Maestro would nevertheless be introduced "as a local debit card scheme in Belgium" (147). MasterCard then withdrew its announcement that the SEPA fallback interchange fees would already apply to the Eurozone from 1 January 2008 and delayed the implementation of the fees by an unspecified period of time (148).

129. On 8 November 2007 MasterCard informed the Commission that it had adopted "Intra-Eurozone interchange fees" for Maestro debit cards. (149) (150) [business secret - 2a] This time, the new interchange fee rates for the Eurozone were not published and MasterCard classified the rates as confidential. (151) [business secret - 2a]

130. On 11 January 2007 MasterCard also introduced new fallback interchange fee rates for Eastern Europe (Ukraine, Belarus, Russia, the Balkans etc). Since then, MasterCard has distinguished the following intra-regional fallback interchange fees in Europe:

i. Intra-EEA fallback interchange fees in the EEA sub-region (including Turkey, Switzerland and Israel);

ii. SEPA fallback interchange fees in the Eurozone (since 8 November 2007 re-instated as "Intra-Eurozone fallback interchange fees"); and

iii. Intra-European interchange fees in Eastern Europe (Ukraine, Belarus, Russia, the Balkans etc); these new fees are not subject to this decision (152).

b. MasterCard's offer to banks in the Eurozone

131. Having realized that the SEPA project provided a unique business opportunity, MasterCard approached banks in the Eurozone to negotiate a migration of their card portfolios to the Maestro brand. MasterCard pushes such migrations, amongst others, by promoting a "SEPA debit card offer" which MasterCard describes, as follows:

[business secret - 2b]

132. Once domestic Maestro transactions would represent a "significant share" of transactions, Maestro banks would have three options to set domestic MIFs (153):

Option 1

[Emphasis added] [business secret 2 b]

Option 2

[business secret 2 b]

Option 3

[business secret 2 b]

133. (154) (155) [business secret - 2a]

134. (156) (157) (158) [business secret - 2a]

135. [business secret - 2a]

3.1.3 Who sets the Intra-EEA fallback interchange fees

136. Until 25 May 2006 the so-called "Europe Regional Board", now referred to by MasterCard as the "European Board" or sometimes also as the "[MasterCard Europe Board" (hereinafter "European Board") sets the level of Intra-EEA fallback interchange fees. This Board was and still is composed of bank delegates from banks in the EEA area.

137. (159) [business secret - 2a]

3.1.4 The use of proceeds from Intra-EEA fallback interchange fees

138. MasterCard generally does not oblige issuing to use proceeds from interchange fees in a particular way (160). This also applies to Intra-EEA fallback interchange fees. MasterCard also does not verify in a systematic manner how issuing banks use proceeds from interchange fees (161).

3.1.5 The flow of payments and the role of the Scheme Manager

139. A typical payment card transaction in the MasterCard payment organization is performed, as follows:

Diagram 3

140. If a cardholder purchases a good/service at for example EUR 100.00 the merchant will eventually receive for example EUR 98.50, the difference being the interchange fee (e-g.: EUR 1.00) and the acquirer's other costs (fees paid to MasterCard, own processing costs etc) and profit margin (for example EUR0.50) (162)

141. In this context the role of the scheme owner MasterCard Incorporated is to process cross-border payment card transactions and in doing so to verify that the correct interchange fee is deducted to the benefit of the issuing bank. The acquirer merely indicates which interchange fee rate applies to a certain transaction. This information is then routed to the issuing bank via

MasterCard's IT network and MasterCard Incorporated's processor verifies the interchange rate. If the acquirer tried to "cheat" on the issuer by claiming a preferentially low interchange rate. If the acquirer tried to "cheat" on the issuer by claiming a preferentially low interchange fee rate, MasterCard Incorporated would step in and increase the rate claimed to a higher one (163).

3.1.6 The structure and level of intra-EEA fallback interchange fees

142. As set out in Annex 1 MasterCard currently distinguishes 27 Intra-EEA fallback interchange fee rates for MasterCard branded payment cards (that is to say credit and charge cards) and five such rates for Maestro branded payment cards (that is to say debit cards). These rates apply uniformly across all EEA Member States. For instance, the rate for a cross-border card payment between Italy and Germany is exactly the same as that for a cross-border card payment between the United Kingdom and Spain.

143. The weighted average (164) of all rates applicable to the MasterCard branded payment cards varied between [business secret - 2a] in the period 2000 to 2006. In the same period Maestro branded cards yielded on average between (165) [business secret - 2a] Thus, the interchange fees applicable to MasterCard branded cards were always roughly [business secret - 2a] in terms of percentage of the transaction value as those applicable to Maestro branded debit cards.

< table 6 > (166)

[business secret - 2a]

< table 67> (167)

144. Diagram 4 graphically illustrates the evolution of the weighted average between 2000 and 2006 (168).

Diagram 4

[business secret - 2a]

145. As can be sent from Diagram 4, the weighted average in the credit and charge cards segment [business secret - 2a] In the debit cards segment the average [business secret - 2a] This development occurred during a time period when the volume and value of transactions in the MasterCard scheme [business secret - 2a] (169)

3.1.7 The purpose of interchange fees according to MasterCard

146. MasterCard's views on the purpose of its interchange fees have evolved over time.

a) Interchange fees as a price la compensation for services rendered

147. MasterCard initially described its interchange fee as a mechanism by which the acquirer "reimburses" the issuer for the issuer's costs related to the transaction which are not reimbursed elsewhere. According to these early documents the fees were "designed to compensate a member for particular expenses that it incurs as the result of interchange transactions" (170). In 2001 the methodology for determining the level of the MIF was still depicted as to "take into account specific costs". The methodology to establish the fee levels might be described as a 'Tee-for-service approach where an imbalance of costs is redressed by acquirers compensating issuers in connection with the costs that issuers incur in providing the following services that benefit the merchant". In that early period, an internal manual also explained the purpose of interchange fees to MasterCard's member banks as follows: "the interchange fee is a price paid by the acquirer to the issuer for the service provided in relation to a POS (point of sale) transaction (...)" (172) The same manual which previously described interchange fees as "price" today explains the purpose of interchange fees as follows: "The interchange fee is a fee paid by the acquirer to the issuer to correct the imbalance between the costs and revenues that they both incur in providing a payment system to the cardholder or merchant " (173).

b) Interchange fees as o balancing mechanism

148. In its response of 5 January 2004 to the Commission's Statement of Objections of 24 September 2003, MasterCard set out the following views on the purpose of interchange fees. Rather than qualifying interchange fees as compensation for issuing banks regarding particular expenses or even as a "price" the interchange fee is referred to as tool to allocate revenues between issuing and acquiring banks in order to properly balance the demands of cardholders and merchants.

149. MasterCard see acquirers and issuers as "co-operating partners of a joint venture" supplying a "joint product". MasterCard argues that together with its issuers and acquirers, it provides card payment services simultaneously to cardholders and merchants (174). The distinct services of MasterCard, of issuers and of acquirers are to be analysed as "the MasterCard payment service"

Which is defined as "co-operation enabling service" or "demand co-ordinating service" to cardholders and merchants (175).

150. MasterCard describes as erroneous the view that interchange fees are a payment for costs incurred by issuers and attributable to acquirers and merchants (176). Instead, interchange fees would be a "proxy for merchants" elasticity of demand" taking into account payment guarantees, funding for delayed payment and processing of incoming transactions (177). MasterCard does not consider that the purpose of its cost studies was to "identify issuer costs that 'should' be reimbursed by acquirers". Instead, cost studies served to estimate the"... opportunity cost to merchants of operating their own store card product, on the theory that this opportunity cost is related to the merchant's demand for MasterCard payment cards" (178). MasterCard suggests that its cost studies should be considered as a substitute for an empirical analysis of how the demands and prices on both sides of its payment network are best balanced, since "undertaking a full Baxter analysis across the MasterCard network is simply not feasible" (179).

151. Explaining the "Baxter analysis" MasterCard states that (180): "...in a four-party payment card system individual members are free to set their own prices to cardholders and merchants and the resulting prices are determined by intersystem and intra-system competition. As a result of these two independent competitive processes, the prices charged to cardholders and merchants cannot be expected automatically to reflect the optimal price structure for payment services overall...Without a correction mechanism, a four-party system would not be able to ensure that demands on both sides are properly balanced. ... The interchange fee provides a solution to this problem, remedying the pricing deficiency of the open system. From the standpoint of the system, the interchange fee provides for a transfer of revenue between issuer and acquirer in the interest of achieving the optimal delivery of services by both acquirers and issuers to merchants and cardholders..."

152. MasterCard believes that "there is an interchange fee which maximizes system output". An "optimal" interchange fee level would reflect each side's elasticity of demand, issuers' and acquirers' respective costs and the relative strength of network effects (181). MasterCard's MIF was designed to take into account the network effects of the system and to maximize the overall "output" of the system (182).

153. In its reply to the Supplementary Statement of Objections dated 15 October 2005 MasterCard again describes interchange fees as a "mechanism for balancing cardholder and merchant demand for a payment scheme's issuers and acquirers the cost of delivering the service" (183). MasterCard moreover believes that the Commission had misinterpreted some of the association's previous statements of interchange fees and that MasterCard had indeed always qualified the MIF as mechanism to balance demands of cardholders and merchants and not as "price" (184).

154. MasterCard confirms that the economic principles underlying interchange fees in its scheme were set out in the "Baxter framework" (185). William Baxter (186) argued in 1983 that a card scheme's size will be maximized when the costs incurred by issuers and acquirers at the margin are shared between any cardholder and merchant in proportion to the value that each places on the services received (187).

3.1.8 MasterCard's procedure for setting fallback interchange fees

3.1.8.1 the decision making process between 22.5.1992 and 14.9.2006

156. The procedure and methodology for determining the level of Intra-EEA fallback interchange fees between 22 May 1992 and 14 September 2006 I set out in the subsequent recitals.

157. [business secret - 2a]

158. (188) [business secret - 2a]

159. [business secret - 2a]

160. Finally, the European Board could either adopt or reject the proposals (189).

3.1.8.2 The decision making process since 14 September 2006 (190)

161. (191) (192) [business secret - 2a]

162. (193) [business secret - 2a]

163. (194) [business secret - 2a]

164. [business secret - 2a]

165. (195) [business secret - 2a]

166. [business secret - 2a]

167. According to MasterCard the new procedure was applied for the first time in January 2007 when MasterCard split the Europe region into three sub-regions. MasterCard then slightly modified the Intra-EEA fallback interchange fee rates with effect from 11 January 2007 (196).

3.1.8.3 MasterCard's Methodology for setting fallback interchange fees

168. "MasterCard Standard Interchange Methodology"

"MasterCard Global Debit Interchange Methodology" (197) - (198)

(199) [business secret - 2a]

169. (200) (201) [business secret - 2a]

a) Methodology for setting MIF for MasterCard branded cards

170. Intra-EEA fallback interchange fees for credit cards (be they consumer or commercial cards (202)) and charge cards are set according to the so-called "MasterCard Standard Interchange Methodology". That methodology measures the following cost items of issuing banks:

i. free funding costs

ii. payment guarantee costs

iii. processing costs

171. MasterCard published the cost components of its Pay Later MIF on the internet for the fisrt time only after it received the Commissions' Statement of objections. At that time the split between the cost elements of the Pay Later MIF was 50% payment guarantee costs, 25% processing costs and 25% free funding costs. On the date of adoption of this decision, MasterCard's website indicates a split of 51% payment guarantee costs, 14% free funding costs and 35% processing costs. Cost studies are executed every two years by MasterCard's economic advisers ant they are based on the cost items set out in the preceding recital which are provided to MasterCard by a sample of MasterCard's issuing banks.

172. The purpose of the cost studies is not to assess issuing costs in order to estimate cardholder demand. Indeed, MasterCard does not execute any cost studies to estimate cardholder demand. Rather, the cost studies were developed in the USA more than 30 years ago in order to assess the demand of merchants for accepting MasterCard's credit cards (203). The willingness of the merchants community to pay is assessed by issuing a proxy, which is a study on fictitious costs merchants would incur for issuing their own store card.

173. According to MasterCard, the average interchange fee is then set somewhere below the cost study results:

"(...) So once again, the idea is to make sure that, when we set an interchange fee, we do not set that fee above an amount that, according to our measurement, a merchant should be willing to pay given that a four party bank issued credit card is a reasonably good substitute for a store card. That hopefully explains the purpose of the cost study itself".

174. The cost study therefore does not serve to identify and measure specific issuing costs which are then allocated to acquirers (204). It is merely a tool for estimating the willingness of merchants to pay.

175. MasterCard recognizes that the cost-study is a "relatively simple analysis" MasterCard tries to answer the question:

"How high could interchange fees go before we would start having either serious acceptance problems, where merchants would say: we don't want this product anymore, or by merchants trying to discourage the use of the card either by surcharging or discounting for cash (...)" (205).

176. MasterCard stated that it also takes into account other considerations such as the level of competitors' fees (206).

b) Methodology for setting MIF for Maestro branded cards

177. The "MasterCard Global Debit Interchange Methodology"

(207) [business secret - 2a]

178. (208) (209) [business secret - 2a]

179. (210) :

i. (211) (212)

ii. (213)

i. (214)

ii. (215) [business secret - 2a]

180. (216) [business secret - 2b]

181. (217) (218) [business secret - 2a]

3.1.9 Findings if the Commission's investigation

3.1.9.1 Drivers of setting Intra-EEA fallback interchange fees

182. According to the minutes of meetings of MasterCard's European Board and of the BMAC between 1998 and 2002, there were several drivers influencing the Board's decision to set and adapt Intra-EEA fallback interchange fees (219).

i. The aim of staying competitive with USA (220) ; MasterCard acknowledges that it regularly takes account of actions of its competitor with regard to interchange, even though this would not mean that " it inflates the level of interchange fee beyond what is necessary in order to attract issuers to join the Scheme and/or issue cards " (221).

ii. The wish to give incentives to issuers and acquirers to adopt new technology.

iii. The aim to keep the structure and level of Intra-EEA fallback interchange fees in Europe similar to interchange fees for cross-border transactions outside Europe (222).

iv. The results of cost studies.

183. However, neither merchant demand nor network externalities are mentioned as drivers of setting MasterCard's intra-EEA fallback interchange fees. In particular in weighing the pros and cons for an increase of these interchange fees neither the BMAC nor the European Board mention merchants' potential reactions to an increase of the interchange fees (223).

184. MasterCard's conclusion in reply to the Supplementary Statement of Objections that these minutes of meetings should be disregarded because "minutes are not transcripts" and according to the MasterCard's statesment's network externalities and merchant reactions had also been "expressly discussed" in the BMAC and the European Regional Board (224), cannot be accepted.

185. MasterCard refers essentially to three examples to demonstrate that merchants and network externalities were considered by the European Board and the BMAC. However, the respective minutes do not support such as conclusion.

186. First, MasterCard refers to a reduction of a recto in the consumer credit card segment in 1998 and links this to the "fact that acquirers-merchants might consider MasterCard's interchange fee for its Consumer Cards as possibly too high" (225). However, the minutes quoted by MasterCard reveal that in fact not all of the rates were reduced and, even if some of them were, it was only a temporary solution (226). Hence, it cannot be accepted as evidence of MasterCard's serious considerations of merchant or network externalities in setting the level of Intra-EEA fallback interchange fees.

187. Second, MasterCard cites the minutes of the 32nd meeting of the BMAC in 1999 (227). These minutes do not contain a single reference to merchants. Rather, the Board considered the potential impact of the rate changes on banks and again planned a steady increase of the MIF over the subsequent years, as can be derived from the slides attached to the minutes. If merchants had indeed been considered, then the Board must have qualified their countervailing force as non-existent, because the increase of the base rate by 10 basis points to 1.30% discussed during that meeting was later on endorsed. Only five years later, in 2007, the base rate was again reduced to the previous level of 1.20%.

188. Third, MasterCard cites the minutes of the 40th BMAC meeting where it is [MasterCard staff] (228) (229) [business secret - 2a]

189. Attempts of MasterCard to read "merchant" wherever Board minutes refer to "acquirers" (230) are unconvincing. For instance, MasterCard argues that the increase of the interchange fee differential between signature based cards and PIN cards was evidence for "creating the necessary incentive for acquirer-merchants to abandon signature based transactions". If that reading were correct, that is to say that merchants would abandon signature based card payments and promote PIN cards, then the Board does not even know its own network rules. Under the HACR once a merchant accepts MasterCard branded credit cards, then it is obliged to accept all types of such credit cards whether they are chip & PIN cards or signature based cards. MasterCard's interpretation of the minutes is illogic.

3.1.9.2 Drivers of setting SEPA fallback interchange fees

190. MasterCard's management pursued several objectives in setting the new SEPA fallback interchange fees which are closely linked to the SEPA project and the EPC's Cards Framework. MasterCard also considered competitive rates of domestic debit card schemes (231) and explored with local banks their potential reaction to the introduction of its new SEPA fallback interchange fees (232).

191. (233) (234) (235) (236) [business secret - 5a]

192. The dynamics of inter-system competition in the payment cards industry are peculiar, as industry experts and regulators in other continents also observed. The concern that inter-system competition does not constrain the level of interchange fee rates led the Reserve Bank of Australia to an intervention which culminated in a pronounced reduction of Visa's and MasterCard's respective interchange fees in 2003 (237). In the United States, where interchange fees are not yet "regulated" (but subject to increasing antitrust litigation), publicly available statistics show a consistent parallel upward development of interchange fee levels in the MasterCard and Visa schemes which industry experts attributed to inter-system competition with American Express (238).

193. The MasterCard organization has a vested interest that its member banks are successful in issuing cards and acquiring merchants for card acceptance. A high interchange fee means high revenues for the issuing business of the member banks in the scheme. A high interchange fee also means high revenues for the scheme and network owner MasterCard Incorporated charges issuing banks fees per card payment. The higher a MIF, the less sensitive the banks are to the level of membership fees charged by the scheme owner. Moreover, MasterCard Incorporated's commercial success in selling network services to banks also depends on the number of banks participating in the scheme (239). The organization as a whole and of each owner (and its shareholders), of the organisation as a whole and of each individual bank are therefore intrinsically linked. The Commission therefore cannot share MasterCard's view that "MasterCard has no incentive to keep interchange high/increase interchange fees with the view of offering issuers more revenues and/or incentivizing banks to replace domestic payment cards by MasterCard cards" (240).

194. For issuing banks interchange fees are revenues. The higher the fees, the more attractive it becomes for a bank to issue a card. Typically, issuing banks are members both of Visa and MasterCard (241) and often also of a domestic card scheme. They therefore have an incentive to promote the brand that yields the highest interchange fee revenues. High interchange fees in the MasterCard system are in particular also an incentive for banks to replace domestic payment cards by MasterCard cards (242). This in turn means that each card network has an interest to increase interchange fees or at least to keep them permanently at a high level, in order to offer issuing banks revenues.

195. The nominal level of MasterCard's intra-EEA fallback interchange fees stagnated at a high level and there was even an upward development of the fee rates due to competition with Visa in Europe. The European Board repeatedly decided to align the level of Intra-EEA fallback interchange fees with similar rates in the Visa scheme.

196. MasterCard acknowledges that its interchange fees are set "taking into account the competitive situation". Board minutes show how this is to be understood. (243) [business secret - 2a] The decisive section in the minutes reads, as follows: (244) [business secret - 2b] The European Board then decided [business secret - 2a].

197. In the debit card segment (that is to say Maestro) the BMAC proposed in August 1999 [business secret - 2a] the signature verified rate from [business secret - 2a] because, amongst other things (245), (246) [business secret - 2b] The European Board endorsed the proposal. On 23 September 1999 the European Board adopted the [business secret - 2a] Maestro POS Signature verified interchange fee tier of [business secret - 2a] referring explicitly to Visa's rate for Visa electron of 0.75% (247). On year later, the Board raised the Maestro POS Signature verified interchange fee tier a second time and [business secret - 2a] with Visa electron's interchange fee of 0.75% (248) That interchange fee rate was later [business secret - 2a] (249).

198. More recently, in 2005, [business secret - 2a] the European Board to set different [business secret - 2a] interchange fees for Eastern Europe than in the EEA in order to keep up with Visa in Eastern Europe (250). This proposal was subsequently approved by the European Board in its decision of 24 February 2005 (251).

199. In reply to the Supplementary Statement of Objections MasterCard contests that inter-system competition with Visa increased its interchange fees in Europe. MasterCard argues that the "core" nominal rates of its interchange fees in Europe had not increased during the last years and that the weighted average of the fees remained relatively stable (252).

200. MasterCard did not specify what the "core rates" of its Intra-EEA fallback interchange fees are. Based on MasterCard's statistics on transaction volumes and values the most important rate in the credit card segment is by far the enhanced electronic rate (253). MasterCard [business secret - 2a] his rate in 2001 by [business secret - 2a] (254). Similar developments increased both the base corporate cards rate (from [business secret - 2a]). These increases affected [business secret - 2a] of MasterCard's commercial cards volumes in the EEA (255). Mastercard4s claim that the "core" nominal rates of its interchange fees in Europe had not increased during the last years, must therefore be rejected.

201. As to MasterCard's other statement that the weighted average of its interchange fee rates in the credit cards had segment remained stable, this reading of MasterCard's statistics seemingly starts in 2002 only. It is a fact that the average increased considerably in 2001 and only from then on remained stable (256).

202. Finally, MasterCard kept the level of its Intra-EEA fallback interchange fees unchanged for nearly five years (257) even after Visa has started to gradually reduce its fallback rates for intra-European card payments under the terms of the Commission's exemption decision in 2002. MasterCard, however, did not reduce its Intra-EEA fallback interchange fees in parallel to those of Visa.

203. In conclusion inter-system competition is no sufficient constraint for the bodies and persons in the MasterCard organization that fix Intra-EEA fallback interchange fees.

3.1.9.4 Management proposed steady increase of rates years in advance

204. It appears that MasterCard's management has planned interchange fee increases years ahead.

205. This is indicated, for example, by the Power Point slides attached to the minutes of the Business and Advisory Committee's meeting of 12 November 1998 in which MasterCard's management [business secret - 2a] In the same meeting of the BMAC the management [business secret - 2a] In the subsequent BMAC meeting of 19/20 August 1999 MasterCard's management again [business secret - 2a] (258).

206. MasterCard's management therefore pursued a basic policy of increasing MasterCard and Maestro interchange fees over a five-year period. This policy was to a large extent [business secret - 2a] by the European Board. MasterCard was able to sustain the rate increases in the credit card segment.

207. In its reply to the Supplementary Statement of Objections MasterCard argues that "there is nothing sinister in MasterCard management considering likely evolution of the interchange fee over time, as any company might indulge in forward planning" and that "advance planning is good business". Proposals of management to gradually raise interchange fees over the coming years were simply "indicative" [business secret - 2a] to endorse them (259).

208. It is difficult to see how these arguments should be of relevance. Contrary to what MasterCard claims, its management was not simply "thinking ahead" as to "the likely evolution" of the various interchange fees in the sense of observing an extraneous development on the market. MasterCard management was proposing those increases. MasterCard management's proposals to raise the rates were also typically endorsed by the BMAC as well as by the European Board and then implemented on the market.

3.2 The Honour-All-Cards Rule

209. The HACR creates an obligation both on the merchant and on the acquiring bank. According to section 9.3.1 of MasterCard International Incorporated's rules "each member must acquire all transactions properly presented to it from each of its merchants on such terms as set forth in the merchant agreement between them and under MasterCard rules and procedures". Under section9.11 of MasterCard International Incorporated's rules merchants "must honour all valid MasterCard cards without discrimination when properly presented for payment". A similar rule exists for the Maestro brand. According to section 5.5.1 of the Maestro Global Rules "each acquirer and merchant must accept Maestro cards equally and without discrimination". Each member bank of MasterCard is bound to oblige merchants it has acquired to "honour all properly presented MasterCard card without discrimination" (260).

4. The view of the complainant and third parties

4.1 EuroCommerce

210. EuroCommerce objects in principle to interchange fees in POS cards systems, as long as they are not set in a transparent manner and as long as they remain a "non-negotiable" cost to merchants. According to EuroCommerce merchants could not indirectly exert countervailing powers on MasterCard to reduce the level of a collectively fixed interchange fee by surcharging the costs of card acceptance to their customers (261).

211. While merchants might be willing to bear the acquirers' processing costs, merchants are said to object to paying for the issuer's processing costs or "all kinds of discretionary benefits" offered by issuers to cardholders. EuroCommerce therefore requests an "unbundling" of the services covered by MasterCard's Intra-EEA fallback interchange fees, that is to say a choice for merchants not to make no use of the payment guarantee, as well as more transparency of these cost components so that merchants could make a choice which services of issuing banks they remunerate via interchange fees.

212. EuroCommerce argues that most of the issuers' costs are to the benefit of the cardholders and that merchants would not choose to receive these services if they knew true cost. In particular, EuroCommerce does not believe that the interest-free period is part of the issuer's service to merchants, but rather that it is directly related to individual cardholders, and that it is paid for by all consumers, including those who do not use credit cards. The inclusion of the payment guarantee for fraud is said to take the incentive for issuers to implement measures to reduce costs related to cardholder default losses and cross-border fraud.

213. EuroCommerce requests that commercial cards be made recognizable for merchants and that MasterCard's HACR be lifted so that merchants are in a position to choose which type of MasterCard branded payment cards they want to accept. EuroCommerce also argues that interchange fees for consumer and commercial cards should be the same.

SEPA and central acquiring

214. EuroCommerce submits that merchants are "frustrated" because MasterCard's network rules obliged cross-border acquirers to pay domestic MIFs of the country where the merchant is established instead of paying the domestic MIFs of the country where the acquirer is located. EuroCommerce believes that if acquirers were to pay the fees of their home country, this would lead to a true Single European Payments Area and to lower interchange fees across Europe.

Transparency

215. EuroCommerce points out that merchants belonging to different economic sectors currently pay different (domestic) interchange fees and merchant fees for the same payment card type and brand. Retailers seldom know the exact level of interchange fees for cross-border transactions and - due to a lack of information - were not able to break down the different interchange fees (domestic and cross-border) as part of the merchant fee they pay.

EMV liability shift as of 1.1.2005

216. Lastly, EuroCommerce submits that since 1 January 2005 both Visa and MasterCard have introduced a network rule which enacts a so-called "EMV liability shift". As a consequence of this rule acquirers bear the cost of fraud and cardholder default if they failed to upgrade their IT equipment (incl. terminals) to allow the processing of the new EMV chip payment cards. Acquirers might pass on the costs of that liability shift to merchants, even though merchants "paid" for the payment guarantee through the MIF as part of the merchant fee. In other words, merchants risked paying for a payment guarantee they would no longer receive if their acquirers did not upgrade the terminals.

4.2 Visa Europe

Visa claims competitive disadvantage

217. Visa submits that MasterCard currently has free discretion in setting its interchange fees and that Visa is at a competitive disadvantage due to its commitments under the Visa II exemption decision of July 2002.

218. Visa believes issuers and acquirers jointly supply one single product. Visa does not subscribe to MasterCard's original description of a MIF as a "fee for a service". It regards the MIF as a co-ordination mechanism necessary to run and optimize a four-party payment system. Visa also considers that the relevant market should be the market for all means of payment and that the competitive influence of other means of payments should be taken into account. Visa does not think it is appropriate to apply a SSNIP test (262) to a MIF given the interdependence of cardholder and merchant demand.

Visa's views on the restriction of competition

219. Visa argues that a MIF cannot be restriction of competition by object. While price fixing agreements were deemed to harm consumers by reducing output and raising prices, the MIF could not be faulted on this criterion. Moreover, through an increase in the MIF tends to increase prices for merchants it would at the same time reduce prices for cardholders. At the oral hearing Visa's expert also argued that it was hard to imagine how a multilaterally set interchange fee could possibly restrict competition between acquiring banks. Competition among acquirers could not be stronger with "at par" clearing than with a MIF, just as it would be hard to assume that breweries would compete more keenly if one scraps excise taxes.

Visa in favour of higher MIF for commercial cards

220. Visa submits that it is necessary to set higher interchange fees commercial cards as compared to consumer cards, because issuers of commercial cards would provide additional services and have to compete effectively with the "strongest" issuers of commercial cards such as American Express. Visa also contends that higher interchange fees for commercial cards fulfill the criteria of Article 81 (3) of the Treaty, because specific benefits accrue to merchants as a result of commercial cards.

4.3 Airplus

221. Airplus Lufthansa Servicekarten GmbH issues MasterCard branded corporate credit cards to employees of corporations in Germany. Airplus submits that interchange fees are the "main contribution margin" of its business. Airplus argues that in the event of an "exercise of competition powers the corporate credit card business under open systems would become unattractive and left to closed systems like Amex". Airplus corroborates this allegation by reference to the effects of interchange regulation in Australia where American Express and Diners would have "unjustifiably" profited from a reduction of interchange fees in the MasterCard and Visa schemes. This would have lead to a general increase of merchant fees because Amex and Diners have overall higher merchant fees than MasterCard and Visa (263). Airplus submitted a "case study" to support these views (264).

4.4 British Petroleum

222. British Petroleum, the second largest oil company in the world, submitted at the oral hearing that retail fuels were a low gross margin product. The pump price of one litre fuel would allow only for a low profit margin per litre. If a payment card was used for purchasing fuel, British Petroleum made no profit on the sale at all. Payment card transaction costs were "typically 100% + of profit" which British Petroleum considered as an unsustainable position for the fuel retailing industry. British Petroleum argued that interchange fees accounted for virtually the entire merchant service charge and that it was unable to influence the level of the interchange fee. According to internal statistics the fees for accepting MasterCard commercial cards had steadily increased during the period from 2003 to 2005 (265).

4.5 European retail roundtable

223. The European retail roundtable, a group representing the interests of large retailers in Europe, stated at the oral hearing that interchange fees inflated the costs of accepting payment cards in particular for its members, large retailers, and that those price increases would ultimately be passed on to the final consumer.

4.6 Hotrec

224. Hotrec, an association of European hotels, restaurants and bars, submitted at the oral hearing that their members paid the highest merchant fees as compared to others sectors. They had to accept payment cards, because their customers preferred paying by card. As consumers did not know the real costs of card usage for the merchant, they would not understand the reluctance of hotels, restaurants or bars to refuse this payment means. Consumers would "just not come back" if cards were refused (266).

4.7 IATA

225. IATA (267) stated that issuing banks replaced existing "business use" consumer cards by corporate cards in order to reap higher interchange fees. This was done without the knowledge of the cardholder that they were receiving a different type of card which would impose higher costs on their suppliers.

226. IATA also stated at the oral hearing that payment card costs constituted a significant portion of airlines overall distribution costs. The costs of distributing airline tickets were "dramatically reduced" during recent years as agent commission levels and other sales costs were realigned with new market realities. However, the costs of accepting credit costs represented by credit cards typically amounted to 10 % of airline distribution costs, the share had now risen to at least 25%.

227. IATA submitted in writing that the rapidly increasing penetration of both MasterCard and Visa commercial card products in the airlines sector had material impact on airlines' card settlement costs. From data provided by two IATA members in the United Kingdom IATA estimates an average 19% increase in commercial cards now consistently accounted for more than a quarter of total sales values. This adds around 0.125% to 0.150% to acquirers' interchange costs which, in turn, is reflected in higher merchant service charges borne by airlines. This increase in the usage of MasterCard and Visa commercial cards has not been accompanied by any reduction in the usage of other high cost commercial cards, for example those issued by American Express (268).

4.8 Ikea

228. Ikea stated at the oral hearing that payment cards represented the second largest variable cost of their company. Whenever Ikea negotiated a contract for accepting payment cards with an acquiring bank, the interchange fees were mentioned "as the main reason why card fees cannot be lowered". Ikea complained that it had no bargaining position to lower interchange fees in negotiations. Refusing payment cards was "not an option". Meanwhile 66% of purchases at Ikea outlets were paid for using a payment card. Ikea argued that due to the MIF economies of scale were not passed on to the consumer. Normally, product innovation and economies of scale would be reflected in lower prices, as was the case with furniture. For card payments, however, this was not the case, because the MIF set a floor for the merchant fee.

4.9 Kappé International BV

229. Kappé, a retail outlet at airports, stated at the oral hearing that payment cards were an essential aspect of their business. A high number of "cross-border customers" shop at their outlets using both consumer and commercial credit cards. MasterCard card products were "significant" as credit cards accounted for 36 % of total sales in 2005. Around 40% of these sales were made using MasterCard card products. Kappé argued that they aimed to offer customers as many payment methods as possible. However, credit cards increased the price of their goods. Kappé opposed "excessively high fees" and complained that the fees were a percentage rather than a flat fee. Kappé also complained that is was "paying" for the free funding period on credit cards while this "service" to the cardholder would not generate incremental sales at its outlets.

4.10 OEHV

230. OEHV, an Austrian association of hoteliers, criticizes the viewpoint that a MIF is needed in the MasterCard and Visa card systems based on the assumption that these systems are "open" systems and that they need a MIF to compete with "closed" systems such as American Express. OEHV refers to the market situation in Austria where MasterCard would in practice operate as "closed" scheme. Local issuing banks would hold all shares of the dominant acquirer and therefore in practice one economic entity issues and acquires MasterCard cards in Austria (269). The situation in the Visa scheme would be exactly the same (270). Nevertheless, Austrian banks would use a MIF even in this "closed" setup.

4.11 Servired SC, Sistema 4B SSA, Euro 6000 SA

231. The three Spanish schemes submitted a joint oral statement at the hearing. In this statement the schemes support the use of a MIF. The schemes argue in particular that the use of payment cards a free funding period increases turnovers at merchant outlets. As a consequence, the inclusion of free funding costs in "calculating" a MIF was justified (271).

5. The four-party/open payment card system

5.1 General

232. Before defining the relevant product market (s) it is necessary to set out briefly how open payment card systems operate in order to indentify the parties involved and their role in the system, and to examine thereafter the product characteristics and the pricing.

Parties involved in MasterCard's payment card system

233. MasterCard and Maestro branded payment cards can be used for POS purchases of goods/services from merchants, that is to say in a shop via a terminal or "at distance" by transmitting the card number and authentication details via mail, telephone or internet. Such POS transactions in MasterCard's system typically (272) involve: the cardholder, the merchant, the card issuing bank, the acquiring bank and the network operator.

Open and closed payment card systems

234. MasterCard's system is an example of an open payment card system (or "four-party" payment card system) which is open to membership by different financial institutions under a common card brand (273).

235. Such systems involve three levels of interaction. As shown in diagram 5, the first type of interaction occurs between the network operator (here: MasterCard) as service provider on the one hand, and issuers and acquirers as recipients of these services on the other hand.

< diagram 5 >

236. As shown in diagram 6, the member banks enter into a second kind of interaction when they mutually exchange data (authentication, authorisation etc.) and finally transmit funds via the network's IT infrastructure.

< diagram 6 >

237. As shown in diagram 7, a third distinct kind of interactions occurs between the issuers and acquirers and their clients, that is to say cardholders and merchants.

< diagram 7 >

238. So-called "closed" payment card systems (or "three-party" payment card systems) such as American Express are vertically integrated systems, where to system operator develops its own card brand and signs up cardholders and merchants itself. As shown in diagram 8, the structure of closed systems involves only one level of interaction between consumers and the bank (274).

Diagram 8

5.2 The operation of an open payment card system

5.2.1 Relations between the scheme owner and issuers and acquirers

5.2.1.1. Scheme owner

239. The owner of a card scheme owns and promotes a card Logo and typically also co-ordinates the behaviour of member batiks through setting common standards and protocols. Scheme owners may moreover also act as "network operator" by providing an IT infrastructure for the transmission of electronic messages required to terminate a payment card transaction.

240. In the MasterCard organisation the card logos are owned by MasterCard Incorporated and the network is operated by EPSS, a daughter company of MasterCard Incorporated. EPSS routes authorisation requests between issuers and acquirers, engages in issuer and acquirer processing and even authorises some card transactions on behalf of issuers if their IT network can not be reached for technical reasons ("stand in authorisations"). Through its activities as scheme owner and (via EPSS) as network operator MasterCard Incorporated provides a "platform" which co-ordinates the interaction and commercial activities of the member banks.

5.2.1.2. Fees charged by the scheme operator

241. Scheme owners typically charge royalties and membership fees to banks for the participation in the system. Network operators in particular charge fees for the processing of card payments. (275) [business secret - 5a]

5.2.2 Cooperation between issuers and acquirers

242. Issuers transmit data on the cardholder, the card and the card transaction after verifying the card's validity and the authorisation limit. Acquirers route transactions from the merchants' POS terminal to the issuers processing centre, provide data for the authorisation and participate in the clearing and settlement of the transaction. Issuers and acquirers provide services to their customers that are complementary in nature but have distinct features and are provided to two different customer groups (cardholders and merchants). Issuers and acquirers co-operate in order to provide these distinct services to their respective customer group.

5.2.3 Issuers' relations with cardholders

5.2.3.1Issuing bank

243. Issuing banks ensure the satisfactory completion of a payment transaction (payment functionality), that is to say the transfer of funds from a deposit account to the account of a merchant. Besides the payment transaction, issuers provide cardholders also with additional services that are not necessary for the completion of a payment card transaction and may or may not be linked to the use of a payment card. The latter may be related to credit.

5.2.3.2 Cardholder fees

244. Banks typically charge cardholders fixed annual fees for using payment cards. In many EEA countries batiks issue Visa and MasterCard payment cards at identical fees (276). Issuing banks may finance payment cards (partially) through negative float (277) and/or currency conversion fees. Banks also issue credit cards to sell consumer loans. They then charge interests for the use of the credit facility and penalty fees for belated repayments.

245. Issuers also obtain revenues through the MasterCard MIF. While MasterCard sets its interchange fees with reference to cost benchmarks, MasterCard clarified towards the Commission that its MIF does not have the functionality of a "price" in the strict sense of the word. However, the MasterCard MIF has the effect of determining to a large extent the final price merchants pay for accepting cards (278).

5.2.4 Acquirers' relations with merchants

5.2.4.1 Acquirer

246. Acquirers credit merchants by transferring funds to the merchant's bank account after receiving the funds from an issuing bank as consequence of a payment card transaction. To enable this transfer of funds, acquirers equip merchants with payment terminals and provide other technical and financial services to the merchants, which typically are all sold in one single contract to merchants. Some acquirers offer to credit merchants before they even provide a service to the cardholder (279).

5.2.4.2 Merchant fees

247. Acquirers charge merchants a fee per transaction for accepting payment cards. These fees ("merchant fee", "merchant service charge", "discount rate" or "disagio") typically are a percentage of the transaction value for credit and charge card transactions and a fixed foc for debit card transactions (280).

248. In setting merchant fees, acquirers take into consideration interchange fees as common input costs. Interchange fees are normally fully included in the merchant fee and thereby passed on to merchants. Acquirers consider, for example, the split between domestic and cross-border transactions to determine to which extent cross border and domestic MIFs must be weighted as cost element in the merchant fee.

249. Acquirers often charge one single merchant fee for Visa and MasterCard transactions (281). This means that the acquirers "blend" a merchant fee for Visa and a merchant fee for MasterCard into one fee that is then identical for both card brands. 60% of 218 merchants surveyed by the Commission in 2004 paid a blended fee for accepting Visa and MasterCard (282).

6. The relevant market

6.1 MasterCard's position

250. MasterCard contends that the relevant product market is the market where different payment card systems' services compete with each other and with all other forms of payment, including cash and cheques.

The "joint" nature of supply and demand

251. MasterCard sees acquirers and issuers as co-operating partners in a joint venture supplying a "joint product". MasterCard holds that together with its issuers and acquirers, it provides card payment services simultaneously to cardholders and merchants. The distinct services of MasterCard, of issuers and of acquirers are to be analysed as "the MasterCard payment service" which is defined as "co-operation enabling service" or "demand co-ordinating service" to cardholders and merchants.

252. Cardholders and merchants should be seen as representing a "joint demand". "Joint demand" is defined as "strict complementarity" of cardholder and merchant demand. As a consequence of its market definition, MasterCard requests the Commission to conduct a SSNIP test on the sum of the two prices charged to the two demand sides, that is to say cardholder fees and merchant fees. MasterCard submitted a study (283) as evidence that a four party payment card system involves a joint service provided in response to joint demand. The study argues in essence that the relevant product market must be "the MasterCard service" and the relevant product market is defined by determining which other payment systems compete with the MasterCard service.

Cash and cheques part of relevant product market

253. MasterCard believes that the Commission also errs in defining separate issuing and acquiring markets since, at the same time that its issuers compete with each other over the delivery of "the MasterCard service" to cardholders and its acquirers compete with each other over the delivery of the MasterCard service to merchants, they continue to compete with ail other payment systems and forms of payment.

254. In reply to the first Statement of Objections MasterCard submitted a study (284) to support its views. The study contains two surveys among merchants and cardholders, however, all of them in the United Kingdom, only. The Merchant Survey (285) is said to confirm that accepting cards is not essential to merchants and that merchants can and do discourage the use of cards at the point of sale. The second study called "Discouragement Survey" (286) is said to demonstrate that cardholders would switch to alternative forms of payment if they were required to pay more for using a credit card.

Giro Transfers

255. In reply to the Supplementary Statement of Objections MasterCard argues that the Commission should include giro transfers into the relevant product market. This is because "giros are widely used in certain EU Member States". MasterCard refers to the Swedish competition authority which had confirmed that giro transfers were among of the three most widely used payment instruments in Sweden (287).

Geographic market

256. MasterCard stated that the relevant product market used to be national, but is increasingly becoming Community-wide.

6.2. The Relevant Product Market

6.2.1 Market definition in industries with two-sided demand

257. MasterCard's approach to defining markets in industries with two-sided demand cannot be accepted. Two-sided demand does not imply the existence of one single "joint produce' supplied by a "joint venture".

a) Two-sided deman4 indirect network effects and network externalities

258. As already set out in the Visa II decision (288), there are two groups of consumers in the payment cards industry: cardholders and merchants (as well as subsequent purchasers (289). This two-sided demand is a feature of the payment card industry. Another feature of the industry is the existence of indirect network effects which means a positive relationship between network size and consumer valuation across the two customer groups.

b) Commission approach

259. MasterCard's viewpoint that in industries with two-sided demand and network effects the definition of separate product markets is an inappropriate tool for the competitive assessment cannot be accepted. MasterCard's concept of defining one single market where "the MasterCard product" is sold to a joint demand is inappropriate in particular, because it ignores the existence of three levels of interaction within open payment card systems (see section 5.1) and does not provide an appropriate analytical framework for assessing the dynamic interdependence of two-sided demand, relevant product is not merely payments, but also separate acquiring and issuing services.

ba) Joint supply

260. MasterCard's analysis of the supply side relies on an expert's opinion (290). That expert proposes to focus the competitive analysis on an inter-system level where scheme owners, issuing and acquiring banks would together offer consumers and merchants "cooperation enabling service" that allow them to interact for the purpose of a payment card transaction.

261. That approach is, however, not suited for analysing restraints to competition within payment card systems. The analysis ignores the fact that end users in four party payment card systems do not directly connect to the platform but do so only through intermediaries (issuers and acquirers) who are themselves providing services which are distinct from those of the platform. The platform run by MasterCard is not a product offered jointly to cardholders and merchants. It is a vehicle for issuers and acquirers to offer distinct services to two groups of customers (291). In other words, MasterCard's approach of defining one joint supply of co-operation enabling services offered by MasterCard/issuers/acquirers as supplier to cardholders/merchants as buyer abstracts from the relevant architecture of a payment card scheme by ignoring the different levels of interaction between the platform, the bank intermediaries and the banks' respective customers. It is therefore not an adequate framework for assessing the effect of a MIF on competition between acquiring banks. Finally, MasterCard's argument (292) that the issuing, acquiring (and network) services cannot be analysed in isolation due to the interdependence of the volume and value of a card transaction on both sides of the scheme is unconvincing. The services at stake here are not simply a transfer of money but they are on each side a bundle of complex services of a technical and commercial nature which together allow for the transfer of money.

bb) Joint demand

262. As to the demand side MasterCard's expert argues that demand of cardholders and merchants is "joint" in the sense of "strict complementarity" and claims that the Commission should carry out a SSNIP test on the sum of charges to cardholders and merchants.

263. It is obvious that demand of member banks for the platform's services at the up-stream network level is dependent on demand at the down-stream issuing and acquiring level of the system. However, this interdependence of demand in a vertically structured industry is not as such an obstacle to defining distinct product markets at each level of the production chain (293). With respect to the complementarity of demand for card usage and card acceptance at the down¬ stream level it is important to note that banks decided to cooperate in large networks to increase total card transaction volumes under a common card brand. This co-operation, however, does not imply that the competitive processes on the down-stream level disappear or that they could no longer be analysed in distinct product markets (294). Moreover, an acquirer provides a wide range of services to merchants. The same applies for services of issuing banks to cardholders. MasterCard's expert reduces ail of these services to a mere transfer of money between the issuing to the acquiring bank that is then identified as "strictly complementary". That assumption abstracts from market reality.

264. Carrying out a SSNIP test on the sum of prices charged to cardholders (295) implies that the distinct demand of cardholders for payment cards and the distinct demand of merchants for acquiring services are amalgamated into one single demand. This suggestion is conceptually unconvincing and also contradicts MasterCard's reasoning regarding Article 81 (3) of the Treaty where it justifies the very existence of an interchange fee in its system by relying on the different nature of demand of cardholders due to different price elasticities, which - in MasterCard's view - creates a need to "balance" those different demands (296).

265. For all of these reasons the concept of joint demand and MasterCard's suggestion to execute that single SSNIP test be executed on the sum of charges borne by cardholders and merchants does not appear appropriate for the purpose of defining product markets in the payment cards industry. The use of such a SSNIP test is inadequate for the purpose of assessing the potential effects of a MIF on competition within one of these schemes and in particular between acquiring banks as it ignores the different levels of interaction and supply and demand within such a scheme.

bc) Commission practice

266. MasterCard's concept of market definition is also inconsistent with the Commission's long standing case practice in defining product markets in industries with two-sided demand. The very examples MasterCard cited in reply to the Statement of Objections of 24 September 2003 (297) in support of MasterCard's market definition show that its concept is contrary to consistent Community competition policy.

267. MasterCard referred to the newspaper market as a "two-sided market", because newspapers and magazines provide services to advertisers, as well as to readers. However, the Commission has always defined services to readers and services to advertisers as separate product markets despite interdependence of demand (298). As regards MasterCard's other example for "two-sided markets", software platforms, the Commission reached the conclusion in its Microsoft decision that due to its specific characteristics and the lack of realistic substitutes, the market for streaming media players (a software application) constitutes a relevant product market that is distinct from the markets for client PC operating systems or work group server operating systems (299) despite the fact that demand for these products is "two-sided", that is to say interdependent (300).The Commission's decision was upheld by the Court of First instance (301).

bd) MasterCard reply to Supplementary Statement of Objections and Commission assessment

268. In reply to the Supplementary Statement of Objections MasterCard stated that the precise market definition appeared to be only of limited value in the present case. MasterCard nevertheless raised the following arguments to support its views that the definition of separate issuing and acquiring markets was artificial.

269. First, the Commission's case practice in the newspaper and software industry was irrelevant because there was a "fundamental difference" between these cases and this case. Unlike in the payment cards industry services of a newspaper to readers and advertisers served two different kinds of demand. Services provided to advertisement agencies consisted of the provision of advertisement space hi a newspaper while services to readers consisted of the publication of articles. In the payment cards industry, however, a payment made by cardholders corresponded to a payment received by merchants.

270. Demand in the cards industry was therefore not only "two-sided" ("simple" complementarity") but moreover "joint" ("strict" complementarity). In conclusion, two sided demand (newspapers industry) allowed for the definition of two separate product markets while joint demand (payment cards industry) required the definition of one product market (302). MasterCard refers to the business of real estate agents as another example of "joint demand" and "joint offer" similar to the payment cards industry (303).

271. Second, MasterCard believes to find support for its claim that the Commission should have executed a SSNIP test in this case on the sum of prices (304) paid by cardholders and merchants in a paper written by Emch and Thompson (305).

272. Those arguments are not convincing.

273. As to MasterCard's first argument the examples on the newspaper and software industry (see above recital 256) are MasterCard's own examples in its reply to the Commission's Statement of Objections of 24 September 2003. As regards the distinction between the "simple complementarity" of demands in the newspaper advertising industry and the perceived "strict complementarity" of demands in the payment cards industry issuing banks provide cardholders with a wide range of services. Many of those services satisfy a "one-sided" demand as the services are not intrinsic in the completion of a payment transaction and could also be provided to consumers outside the context of a card payment (306).

274. Although the respective demands of cardholders and merchants for payment transaction services are inter-related, the demand behaviour of the two customer groups is significantly different. Cardholders and merchants have different responses to final prices for the payment transaction services on the two sides of the industry, as MasterCard acknowledges (307). Therefore the interrelated nature of their respective demands for payment transaction services does not constitute a sufficient reason for qualifying issuing and acquiring services as a single product for the purpose of defining a relevant market.

275. MasterCard also alleges that the services provided to merchants and cardholders are a "joint product", because they are produced in strict complementarity and, as such, the costs of two outputs cannot be separated ("joint costs"). However, even if one adhered to this view, which is not the Commission's position, MasterCard's MIF is not merely a tool to allocate costs for the provision of "strictly complementary" payment transaction services. As MasterCard's counsel emphasised at the oral hearing, MasterCard's MIF does not allocate any specific costs (308).

276. As to the paper of Emch and Thompson, the model appears too simplistic iii several respects. It essentially assesses open card schemes as if they were a closed scheme, that is to say with one entity directly providing services to cardholders and merchants. However, this partial equilibrium analysis does not take into account the exercise of market power by issuing and acquiring banks. It ignores that end users in open card schemes do not directly connect to the platform but do so only thorough intermediaries (issuers and acquirers) (309). The model also does not meet market reality as it assumes fixed proportions, that is to say fixed marginal costs and fees charged to merchants which is why the variable margin on each transaction is also fixed. Merchant reactions to a hypothetical increase of prices are ignored in the model (310).

277. Finally, MasterCard referred to the merger cases M.3547 Banco Santander/Abbey National of 15 September 2004 and M. 1172 of Fortis AG/Generale Bank of 24 June 1998 to argue that the Commission had always defined the relevant market as large as "the market for the provision of retail banking" (311). However, typically in most merger cases it is left open whether a relevant product market can be further subdivided into more narrow markets, if even the most narrow market definition would not lead to horizontal overlaps or vertical foreclosure effects. The Commission only enters into a discussion of submarkets if a merger analysis requires it. This is not an implicit acknowledgment of MasterCard's views on the necessity of a broader market definition, as demonstrated by the merger decision in case M.4316 Atos Origin of 29 September 2006. Based on the parties' notification the Commission analysed the effects of the merger between Atos and Banksys in the horizontally related product markets for the processing of payment card transactions and on the vertically related market for commercial acquiring of national debit cards in Belgium.

6.2.2 Commission's approach to market definition

278. As set out in the Visa I decision (312) and in the Visa II decision (313), two types of competition can be distinguished in the payment cards business: competition between different payment card networks (so-called 'inter-system' competition) and competition between individual financial- institutions (usually banks) for card-related activities (essentially issuing of cards to individuals and acquiring merchants for card payment acceptance) (314). Competition can take place upstream (at the level of networks) and downstream (at the level of individual financial institutions) in the value chain.

279. Accordingly, the Commission has distinguished between an upstream "system/network market" and downstream "issuing" and "acquiring" markets (315).

280. The Commission's approach towards market definition is based on the analysis of the services provided within an open payment card system as set out in sections 6.2.3 and 6.2.4. These are services between the network and its member banks, between acquirers and merchants and between issuers and cardholders.

a) Network market

281. The upstream market is the market where card scheme owners compete to persuade financial institutions to join their schemes and on which they provide services to such institutions. As set out in section 5.2.1, MasterCard Incorporated provides such services to the organisation's member banks and charges particular fees for these services. MasterCard Incorporated competes with other card scheme owners. It is therefore appropriate to identify a distinct upstream market for the provision of network services.

b) Issuing and acquiring markets

282. Banks provide services downstream of the network market in so-called issuing and acquiring markets. The Commission will subsequently use the analytical framework laid down in the notice on the definition of the relevant market for the purpose of Community competition law (316), to assess whether the issuing and acquiring markets are to be widened to include services other than those related to card issuing and acquiring.

6.2.3 The acquiring market

6.2.3.1 Supply side analysis

283. As described in section 5.2.4.1, acquirers provide a wide range of services, which are of a technical and of financial nature. Their customers are merchants wishing to accept payment cards. The product characteristics of acquiring services are fundamentally different from those of issuing services (317). Pricing of acquiring services is structurally different from the pricing of issuing services, since it is based on a fee paid for each transaction, whereas cardholders typically pay annual fees (318).

284. The differences in product characteristics and the competitive dynamics in providing acquiring services as opposed to providing issuing services (319) further illustrate that the approach of amalgamating all these distinct services into one single "MasterCard payment service" for the purpose of defining the relevant product market is inappropriate to grasp the complexity of market reality in this industry.

285. The product characteristics of card acquiring services are also wholly different from the services that suppliers provide to merchants hi respect of acceptance of cash and cheque. Cash collectors, for example, can hardly switch to acquiring merchants for card acceptance and vice-versa. Cash collection requires high security transports and involves the risk of robbery, whereas card acquiring requires sophisticated IT equipment, involves financial risks such as unrecoverable charge back losses linked to merchant default.

6.2.3.2 Demand side analysis

a) Inappropriateness of SSNIP test

286. As set out in recitals 733 - 737 the execution of a SSNIP test on the sum of prices charged to cardholders and merchants is conceptually inappropriate. Even a SSNIP test on the single price charged by acquirers to merchants may not produce results that are sufficiently reliable for measuring acquirers' collective market power vis-à-vis merchants due to a MIF. This is because a SSNIP test in the acquiring business faces the problem of the so-called cellophane fallacy (320). This problem refers to a situation where products may appear as close substitutes, because entities providing these products already exercised market power to raise prices above a competitive level. In such situation, a SSNIP test may not yield satisfactory results for the definition of a market, because switching may occur even between products that are not substitutes in a more competitive situation.

287. In this case the risk of a cellophane fallacy is significant for several reasons. Pricing in the issuing and acquiring markets is largely determined by collectively set interchange fees (321). Also, the concentration in the acquiring business is high in most EEA countries, potentially allowing acquirers to charge supra-competitive prices. Thus, a survey with merchants in these markets asking whether they would cancel a card if the fees were raised by a small but sustainable can be expected to lead to exaggerated results. Therefore, hi assessing the MIF, the Commission attributes higher value to evidence derived from product characteristics and past switching behaviour than the results of a SSNIP test (322).

b) Evidence from product characteristics

288. The pricing of cheques and card acceptance to merchants differs considerably. Typically, banks did not (323) and still do not (324) charge merchants for drawing cheques whereas merchants pay transaction related fees for accepting cards. The fact that giro transfers are often used in countries such as the Netherlands or Sweden is as such not evidence that this payment means is substitutable with card payments. Giro transfers are typically used by consumers for recurring payments such as electricity bills, installment payments, insurance premiums, etc (325). They are, however, unpractical for paying goods or services in "over the counter" transactions as they require an interaction with the bank to establish the recurring payment. Retailers therefore can hardly invite customers to substitute a card payment with a giro transfer for over the counter transactions.

c) Evidence from customer behaviour

289. First, the Commission's merchant market survey (326) showed that merchants are reluctant to cancel wide-spread payment cards once they started accepting them. 91% of the merchants surveyed by the Commission replied that they have never cancelled a payment card. Most of the merchants surveyed are large merchants with a comparatively high degree of countervailing power.

290. Second, merchants' reasons for card acceptance suggest that they are unlikely to see cash and cheques as substitutes. The two most important reasons to accept cards in the past were that a large number of merchant customers carry cards and the merchants wish to offer good customer service. Thus, the more cardholders use MasterCard the more card acceptance becomes a "must". It was a comparatively less important reason for merchants to accept cards that cards may be less expensive or less risky than handling cash/cheques (327).

d) The [...] Dot Econ-Merchant Benefits Survey

291. Evidence submitted by MasterCard does not contradict the Commission's market definition. MasterCard's [...] Dot Econ studies, which are based on data from the United Kingdom only (328), use a complex survey methodology based on "conjoint analysis" in order to conduct two distinct surveys: one for merchant perception of payment card benefits, the other for card holders' choice of retailers.

292. Conjoint analysis is used to evaluate consumer preference. If products are composed of attributes, conjoint analysis determines which combination of attribute levels are most preferred by consumers. In particular, consumers indicate their preferences by making a sequential choice (329) between a number of different combinations of attribute levels. Conjoint analysis has attracted much interest in the marketing research field recently. There are several reasons for its increasing popularity. In particular, the task of choosing a preferred concept is similar to what buyers actually do in the market place.

293. Conjoint analysis may, however, lead to biased results and respondents are faced with complex choices. Each concept typically involves all the attributes being considered in the study, and each choice set contains several concepts. Conjoint analysis is therefore at least as hypothetical as a SSNIP test. The respondent has to process a lot of information before giving a single answer for each choice set. Although this mimics what happens in the market place, the analyst ends up with far less information than would have been available had the task been to rate each alternative in the set. Furthermore, conjoint analysis should be implemented under a strict procedure (330).

294. In any event, the [...] Dot Econ merchant benefits survey draws conclusions on cards characteristics that rather confirm that cards are not substitutable with cash and cheques from the merchants' perspective. [...] Dot Econ concludes that merchants place considerable value on payment instruments that offer a payment guarantee, rapid settlement, allegedly generate incremental spending, have reasonable level of penetration and are card-based rather than paper-based. At least, these elements, which appear specific to payment cards, seem to confirm the analysis of cards as constituting distinct markets from cash and cheques from the perspective of merchants.

295. In addition, it is noticeable that the study is geared towards some attributes, while neglecting others. For instance, the penetration of cards among competing retailers has not been retained as a criterion hi survey, although the fact that a merchant's competitor accept a cards seems to be an important driver for merchants to accept cards (331). Given the methodological constraints inherent to conjoint analysis, results may be misleading if the wrong attributes or if insufficient attributes have been taken into account, as appears the case here. The analysis will then overlook possibly important dimensions in customer preferences.

296. It cannot therefore be concluded that [...] Dot Econ's merchant benefits survey provides reliable evidence that payment card acceptance is substitutable to services related to cash and cheque. The study even tends to confirm that from the merchants' perspective payment card acceptance is not substitutable with the use of other payment means.

e) Criticism relating to empirical evidence used for market definition

297. In its reply to the Supplementary Statement of Objections MasterCard highlighted perceived methodological flaws hi the Commission's market analysis.

298. MasterCard's criticism is unjustified. MasterCard's expert report on alleged methodological flaws of the Commission's merchant market survey are analysed in Annex 2.

f) Alleged inconsistency with previous case law and national decisions

299. MasterCard considers that the market definition in this case is inconsistent with the two Visa decisions (332). MasterCard points out that in the Visa I and Visa II decisions the Commission had "examined, in particular, whether other payment services (including cash and cheques) compete on the same market as card payment services". Furthermore, the analysis would have focused on the "inter-system market" rather than the "intra-system market" and the "intra-system market(s)" were also not clearly defined in the Visa decisions (333).

300. While the Commission bas indeed examined in the Visa I and Visa II decisions whether cash and cheques are substitutes to payment cards, both decisions explicitly state that these other payment means are not sufficiently substitutable with payment cards and therefore exclude them from the relevant product markets (334).

301. Moreover, both Visa decisions take inter-system and intra-system markets into account for the competition assessment (335). In assessing the Visa MIF the Commission analysed effects on competition between issuing banks (issuing markets), between acquiring banks (acquiring markets) and also between Visa and between other card schemes (inter-system market) (336).

302. The market definition in this decision is consistent with the Commission's previous case practice. The conclusion in this decision that the MasterCard MIF restricts competition in the various national acquiring markets in the EEA takes account of the situation in the network market (section 7.2.4) and the issuing markets (section 7.2.3.2) (337).

303. MasterCard sees support for its markets definition in a prohibition decision of Dutch competition authority Nederlandse Mededingingsautoriteit (NMa) of 28 April 2004 in the so-called "PIN" case. MasterCard argues that the NMa's market definition "blatantly contradicts" the Commission's market definition in this decision (338) and that this would indicate lack of analytical rigour on the Commission's side. That view cannot be accepted.

304. First, MasterCard overstates the differences between the Dutch decision and the Commission's approach. The NMa identified a product market for "PIN transaction services" consisting in the "transmission of data between retailers and banks across a secure network". The customers of network services for PIN were the "retailers" (339). This is similar to the "acquiring market" referred to in this decision. Like the Commission the NMa also excludes payment means other than cards from the relevant product market (340) and considers the SSNIP test as inappropriate for defining the relevant market due to the cellophane fallacy problem (341).

305. Second, the Dutch decision was adopted under national completion law, only. A more recent decision of the Polish competition authority prohibiting MasterCard's and Visa's domestic MIFs in Poland uses exactly the same market definition as the one applied by the Commission in this case (342).

306. Finally, MasterCard refers to the Commission's GSA Agreement decision of 9 September 1999 (343) to corroborate its market definition. In the GSA Agreement Decision the Commission analysed the substitutability of giro transfer services and other payment services from the supply side (344) and then separately for the two demand sides (345). Clearly, the Commission did not adhere to MasterCard's concept of "joint demand" in this precedent. Moreover, the Commission again excluded payment cards from the relevant product market of bank giro transfer services. MasterCard's reference to this decision therefore does not support its view that payment cards, cash and cheques are part of the same product market.

6.2.3.3 Conclusion

307. The supply and demand side analyses show that card acquiring services are neither sufficiently substitutable with cash and cheque related services, nor with bank giro, nor with direct debit services. The Commission therefore retains as relevant product market for assessing the MIF the market for acquiring payment card transactions. It can be left open whether this market can be further sub-divided into credit card acquiring and debit card acquiring or whether acquiring for MasterCard products is a product market on its own.

6.2.4 The issuing market

6.2.4.1 Supply side and analysis

308. The issuing of cash is not substitutable with issuing of cards mainly due to legal reasons. Commercial banks are not in a position to issue cash which is a legal privilege of a State's Central Park. Thus, banks can not readily switch to issuing cash if they are not in a position to issue cards/cheques. To the contrary of commercial banks the State also has means other than pricing and marketing at its disposal to convince consumers to use and force merchants to accept "its" payment means. Cash is legal tender and merchants are prohibited from refusing the official currency of a country as payments means, whereas they are legally free to refuse cheques and payment cards.

309. Pricing of cash to consumers is also different from pricing of payment cards while cash is financed through specific revenues of a Central Bank (346), commercial banks finance card through annual fees and interests charged to cardholders and - to the extent the system uses a MIF - through interchange fees.

6.2.4.2 Demand side analysis

310. For consumers, the characteristics of payment cards differ strongly from those of other payment means (347). Consumers tend to perceive cash usage as free of charge, unless they are charged for ATM withdrawals. Card usage, to the contrary, is not perceived as free of charge, because cardholders pay annual cardholder fees and possibly also interests in case the balance on their deposit account is not sufficient to cover the costs interests in case the balance on their deposit account is not sufficient to cover the text of a payment transaction and the cardholder makes use of extended credit (348). Cash can be inconvenient, dangerous to carry in large amounts and unsuitable for expensive purchases, which may make it less attractive to use than credit, charge or debit cards (349). Contrary to cheques, payment cards cannot be used for payments to private individuals (350). Thus, any possible substitutability of cheques and cards is limited from the outset to commercial purchases. Cheque books contain a limited number of cheques whereas payment cards can be used without limitation to the number of transactions. Low cheque card limits further restrict the use by consumers of cheques as an alternative to card payments (351).

311. Cheques are also used less and less frequently by consumers. According to the European Central Bank (352) cheques are in strong relative decline in the Euro area, down from 24 transactions per inhabitant in 1997 to 19 in 2001. In the same time period the average number of payments by credit/debit cards rose from 17 to 31 and the average number of direct debits rose from 29 to 34.

312. The mere fact that cash, cheques and cards share the common characteristics of allowing consumers to access funds held in bank accounts does not, in itself, provide sufficient evidence for a substitutability (353), because this functional interchangeability is only limited (354) and consumer responsiveness to relative price changes may be influenced by other considerations, as well.

313. The pricing of cheques and cards to bank customers differs considerably (355). Under the Eurocheque system, commissions for cheque usage were charged to the issuers' clearing centre which then debited the issuer for each cheque used Thus, the costs for cheques were mainly borne by the issuers and passed on to the cheque user. The opposite is true for payment cards where cardholders sometimes even get cash rebates for using their payment cards while merchants (and subsequent purchases) bear the costs.

314. Finally, bank giro services and bank transfers are designed for situations in which the creditor and the drawee are not substitutable with over the counter payment means such as payment cards (356).

315. The conclusion that payment card related services are part of a separate product market from other payment means is confirmed by past case law of the Court of Justice of the European Communities. The Court stated already in 1994 that cheques are in a distinct product market and not in the same market as payment cards. The Court stated that the market for foreign Eurocheques drawn in the trading sector in France constituted a specific, sufficiently homogeneous market which can be distinguished from that of other international means of payment used. Like MasterCard, its predecessor Europay and the French banking association Groupement de Cartes Bancaires had argued that the relevant market included "all international payment means of payment to French traders" (357). The Court, however, did not accept this view and confirmed that "the Commission was right in finding (...) that the market directly concerned by the Helsinki Agreement is that for Eurocheques drawn in the trading sector in France" (358).

6.2.4.3. Conclusion

316. The relevant product market in this case is the market for acquiring payment cards.

6.3. The Relevant Geographic Market

6.3.1. General

317. In its Visa II decision the Commission considered the relevant geographic market to be taken into account for assessing competitive issues relating to payment card systems to be national (359). MasterCard's argument that the geographic market is "moving towards an EU market" does not lead the Commission to change its views. The evidence and facts available suggest that issuing and acquiring merchants remains national in scope for the time being.

6.3.2 Issuing

a) Cross-border issuing remains limited

318. True pan-European payment card issuers, that is to say companies issuing cards without branches or local co-operations in the country where the foreign cardholder is located, are rare. A renowned industry report cites eleven banking groups to have card issuing activities in countries other than their own, be it through subsidiaries or through joint ventures with local batiks (360) The necessity to open up branches or to enter into local co-operations / merge with local batiks in order to issue cards in other Member States indicates that in the EEA issuing markets are currently still national (361).

b) Heterogeneous market conditions

319. Card issuers face diverging consumer habits across Europe. The average number of transactions with credit cards per inhabitant in the United Kingdom is, for instance, four times higher titan in Spain, six times higher than in Germany and more than ten times higher than in Sweden (362). Issuers have to take account of these different consumer habits and market conditions across the EEA. It therefore appears plausible that banks still have to tailor card strategies for each individual Member State and that developing a pan-¬European issuing business remains a "complex challenge" (363).

320. Also, market dynamics in the payment cards business still appear to diverge across the EEA. In Greece, for example, the value of debit card transactions per inhabitant increased by 150.7% from 2000 to 2001 while the value of credit card transactions dropped by 31.4%. In the United Kingdom the value of credit card transactions grew by 25% and of debit cards by 10.5% over the same period.

c) Differences in pricing

321. Finally, annual fees charged by issuers for credit cards and MasterCard cards in particular vary strongly from one EEA country to the other. According to an independent market survey, annual cardholder fees for MasterCard cards vary according to EU country between EUR 0.00 up to EUR 64.6 and for Maestro between EUR 0.00 up to EUR 23.5 (364).

6.3.3. Acquiring

a) Cross-border acquiring remains limited

322. According to MasterCard cross-border or "central" acquiring, that is to say the acquisition of payment card transactions outside the home country of an acquiring bank, is still a niche business offered to multinational merchants in some few market segments. In MasterCard's own network central acquiring did not exceed [0-[...] 10]% of total intra-EEA transaction values in 2002 and remained concentrated on some merchant sectors such as automobile/car rental, hotels and transportation. The majority of merchants in the EEA therefore still have contracts with a domestic acquirer (365).

b) Differences in pricing

323. Similarly to card issuing strong divergences exist with regard to the pricing of acquiring services. According to a well-known industry report the average merchant fees for MasterCard acceptance for example vary between 0.75% in Denmark to 4% in Greece (366). Similar divergences were confirmed in the Commission's merchant market survey (367).

324. There are concrete indications that differences in the level of domestic MIFs and MasterCard's interchange fallback mechanism contribute to sustaining these price differentials. In [EU Member State], for instance, local banks agreed with one another on bilateral domestic MIFs (368). The bilaterally agreed domestic MIFs are set out in the subsequent table 8.

< table 8 >

325. However, cross-border acquirers who are unable to conclude bilateral agreements with the local issuing banks had to pay the Intra-EEA fallback interchange fees as "fallback" rates and these rates were typically higher. This rendered it more difficult for them to compete for acquiring Maestro transactions particularly in the food retail and petrol sector in [EU Member State].

326. A German acquirer (369) told the Commission: "It is difficult for [German acquiring bank], however,, to acquire large merchants (such as gasoline companies% as [local acquiring bank] and [local acquiring bank], with the help of their banking industry shareholders, agreed on a special, non-disclosed, "on-us" .1F, so that B+S cannot compete on the discount rate, applying the normal fallback IF ..." (370).

c) Market structure remains heterogeneous

327. The market structure in the acquiring business diverges across the BEA. Broadly speaking one can distinguish monopolistic acquiring markets such as Austria (371), Denmark (372), Finland (373), Portugal (374) from markets with oligopolies such as Belgium (375), Ireland (376), Norway (377), Sweden (378) and markets with a relatively higher number of acquirers such as Germany (379), Greece (380), Italy (381) and Spain (382).

d) MasterCard

328. MasterCard did not put forward a geographic market other than the one proposed by the Commission but notes that the Commission's analysis was incomplete because it omitted conclusions on the geographic scope of the network market and ignored ongoing efforts of creating a single market in payment card systems (383). The current decision identifies the restrictive effects of MasterCard's MIF in the acquiring markets, not in the network market. Moreover , while the SEPA project may increase cross-border competition and may allow consumers to use payment cards in the entire Eurozone, at the current moment market conditions and in particular prices are still too heterogeneous to adopt a market definition going beyond the national scope.

6.4 Conclusions on the market definition

329. The relevant market for assessing the MasterCard MIF is the market for acquiring payment cards. At present, that market is national in scope.

7. Article 81(1) of the Treaty/Article 53(1) of the EEA Agreement

330. The MIF applies to cross-border and domestic payment card transactions in the whole territory of the European Economic Area. Insofar as the MIF affects trade between Member States of the European Union, Article 81 of the Treaty is applicable. The EEA Agreement, which contains provisions analogous to those of the Treaty, entered into force on i January 1994. Article 53 of the EEA Agreement is applicable insofar as the MIF affects trade between one or more EFTA States (which have ratified the EEA Agreement) and one or more Member States. In this case the Commission has jurisdiction to apply both Article 81 of the Treaty and 53 of the EEA Agreement on the basis of Article 56 of the EEA Agreement as the MIF affects trade between EU Member States. References in this Decision to Article 81 of the Treaty are also to be understood as references to the analogous provisions of Article 53 of the EEA Agreement.

7.1 Decisions of an Association of undertakings

a) MasterCard's arguments

331. MasterCard does not contest that it was an association of undertakings within the meaning of Article 81(1) of the Treaty in the period before 25 May 2006 that is until the day where the IPO of MasterCard Incorporated took place (384).

332. MasterCard, however, submits that since the listing of MasterCard Incorporated on the New York Stock Exchange on 25 May 2006 MasterCard Incorporated would not qualify as association of undertakings but that it has become an independent undertaking pursuing its own commercial interests for the benefit of its stockholders who are detached from those of its customers. MasterCard recognises that the concept of an association of undertakings is a relatively broad and flexible one (385) but points out that neither the Commission nor the Community courts had ever applied the concept to a single publicly listed company such as MasterCard Incorporated (386).

333. According to MasterCard the decisive question is not whether undertakings coordinated their economic behaviour on the market through a collective structure or a common body, but rather whether there was an agreement between the banks to act in concert. MasterCard Incorporated sets out terms of dealing in license agreements and enforced them if needed. Bank licensees would not have to review or approve these terms nor could they require changes. Any decision of MasterCard Incorporated therefore qualifies as "unilateral" act with each bank bilaterally agreeing with MasterCard to abide by them (387). In conclusion, MasterCard submits that the Commission errs iii applying Article 81 of the Treaty to the MasterCard MIF because since the IPO of MasterCard Incorporated on 25 May 2006 the MasterCard MIF neither qualifies as agreement nor as decision of an association of undertakings (388). MasterCard's views scheme is now rather a set of bilateral relationships akin to any licensor-licensee/franchisor-franchisee relationship (389).

334. MasterCard submits that it has never considered, and has therefore never indicated in any of its documents, that pre-IPO, interchange was difficult to maintain from an antitrust perspective (390).

335. MasterCard refutes the possibility that the LPO of MasterCard Incorporated was a "disguise", i.e. that there ever was an understanding amongst the previous owners of MasterCard Incorporated and MasterCard Incorporated's management that the member banks should retain "control" over the company after the IPO and in particular over the setting of interchange fees (391). The powers of the European banks to set interchange fees had been withdrawn as a part of the IPO "without the banks' agreement" and since then these powers were vested in an independent management (392). Any assertion that MasterCard Incorporated would still be operated solely by, or in the interest of, the member banks effectively suggests that the rights and legitimate expectations of the non-licensee shareholders arising out of IPO are being ignored. If this were right, the directors on the Global Board would be operating in a manner inconsistent with their duties and obligations under us laws (393). MasterCard puts forward that MasterCard Incorporated listens to its customers, talks to them and assesses their needs and whether, in its own interests, it should satisfy those needs (394). This, however, does not mean that MasterCard Incorporated grants to its various customers all that they request and it certainly does not mean that those customers gain "control" of or exercise "influence" on the decisions of MasterCard Incorporated.

336. MasterCard submits that the veto powers of European banks to block the dissolution of the European Board (395) are irrelevant. MasterCard stresses that the European Board is "plainly subservient" to the Global Board, as the Global Board is empowered to withdraw even the "few remaining powers" (396) which the European Board still enjoys after the IPO.

337. As to the MIF MasterCard argues that there is no "commonality of interests" of the member batiks in a high level of interchange, but the only common interest is that the fees be set at a "transaction-maximising rate". This lack of a "commonality of interests" in higher interchange results from the fact that some banks are "net issuers" and therefore have a short-term financial interest in higher interchange fees, whereas other batiks are "net acquirers" with a short¬ term financial interest in low interchange fees (397).

b) Commission's assessment

7.1.1. Association of undertakings

338. Article 81(1) of the Treaty and Article 53(1) of the EEA Agreement are applicable to associations of undertakings in so far as the activities of the association, or of the undertakings belonging to the association, are calculated to produce the results which Article 81(1) of the Treaty and Article 53(1) of the EEA Agreement aim to suppress (398).

339. While the concept of an "association of undertakings" is not defined in the Treaty, it is recognised that -as a general rule- an association consists of undertakings of the same general type and makes itself responsible for representing and defending their common interests vis-à-vis other economic operators, government bodies and the public in general (399). According to the jurisprudence of the Court and the administrative practice of the Commission, an association of undertakings in the sense of Article 81(1) of the Treaty is generally characterized by the existence of a standing body or other statutory arrangements for the coordination between the members (400).

340. According to Advocate General Léger the concept of an association of undertakings "seeks to prevent undertakings from being able to evade the rules on competition on account simply of the form in which they coordinate their conduct on the market. To ensure that this principle is effective, Article 85(1) [now: Article 81(1)] covers not only direct methods of coordinating conduct between undertakings (agreements and concerted practices) but also institutionalised forms of cooperation, that is to say, situations in which economic operators act through a collective structure or a common body" (401).

341. The Commission has always applied a broad meaning to the concept of an association of undertakings including, amongst others, agricultural cooperatives (402), professional regulatory bodies, associations without legal personality, non-profit making associations and associations of associations. In this vein, the UEFA Champions League (403), the Institute of Professional Representatives before the European Patent Office (404) and an economic interest grouping administering the co-reinsurance of environmental risks (405) were qualified as association of undertakings. Some of these entities had a legal personality and a corporate form, others not.

342. For there to be an association of undertakings the Treaty does not require that - in AG Léger's words - the "common structure or common body" disposes of a proper legal personality. The Court of Justice also interprets the concept broadly and qualified acts of professional bodies (406) as decision of an association of undertakings. Conversely, neither the legal personality nor the incorporation of an entity are an obstacle to qualifying it as association of undertakings.

343. The Court of Justice confirmed also that the legal framework within which agreements between undertakings are made and decisions of an association of undertakings are taken, and the classification given to that framework by the various national legal systems, are irrelevant as far as the applicability of the Community rules on competition, and in particular Article 81 of the Treaty, are concerned (407).

7.1.1.1 MasterCard until the IPO

344. Until 25 May 2006 MasterCard was an association of undertakings within the meaning of Article 81(1) of the Treaty for the following reasons.

The MasterCard payment organisation was an "association of undertakings"

345. First, the members of the MasterCard payment organisation were credit institution and therefore engaged in a commercial activity. According to settled case-law, the concept of undertaking covers any entity engaged in an economic activity regardless of its legal status and the way in which it is finances (408) and any activity consisting in offering goods and services on a given market is an economic activity (409). MasterCard's members in the considered period were therefore undertakings within the meaning of Article 81(1) of the Treaty.

346. Second, as apparent from Article I Section 4, subsection c of MasterCard International Incorporated Bylaws and Rules, decisions of the organisation's management bodies were binding upon its members. No bank could have participated in the card activities of the MasterCard organization nor could have used any of its marks unless it agreed to comply in all respects with all bylaws, rules and regulations, and published policies of the organization. Upon joining the organization all member banks therefore submitted to the decisions of the European Board and the Global Board (410).

347. Third, any bank wishing to engage in a commercial activity with regard to the MasterCard payment cards moreover had to apply for membership in MasterCard International Incorporated. No bank could have received a licence to use the MasterCard / Maestro brands for providing payment card related services without obtaining membership. Membership applications were reviewed by the European Board.

348. Fourth, until 25 May 2006 the MasterCard payment organisation operated in a de-centralised manner in Europe where decisions regarding the most important business aspects of the organisation's operation were taken by bank appointed directors on the European Board. The purpose of this Regional Board was to manage the brands and product strategies on a regional leve1 (411). It reviewed membership applications of new banks, had power and authority to fine member banks and recommend termination of membership, approved the annual budget, fixed and collected fees from their regional member banks, etc.. Moreover, decision making on a sub-regional level was in the hands of national fora of member banks (cf. section 2.1.2).

349. MasterCard does not contest that it operated as association of undertakings within the meaning of Article 81(1) of the Treaty before the WO. It is noticeable that MasterCard mc. was already a privately held stock company from 2002 onwards and nevertheless MasterCard qualified decisions of MasterCard Incorporated until the IPO as those of an association" properly speaking. (412) Thus, the fact that a "company" decisions on behalf of banks is apparently not, in MasterCard's views, an obstacle to qualifying these decisions as decisions of an association.

7.1.1.2. MasterCard after the IPO

The MasterCard payment organisation remains an association of undertakings

350. While the IPO of MasterCard Incorporated changed the governance of this company, it did not affect the decisive elements of qualifying the payment organisation as association of undertakings.

351. First, as before the IPO (see section 7.1.1.1 above) each participant in the organisation remains a credit institution and as such an undertaking within the meaning of Article 81(1) of the Treaty.

352. Second, following the IPO, Article I Section 4, subsection c of MasterCard International Incorporated Bylaws and Rules (413) remained unchanged after the IPO. (414) Thus, decisions of the organisation's management bodies are still binding upon the organisation's members and no bank can participate in the card activities of the MasterCard organisation without complying in all respects with all bylaws, rules and regulations, and published policies of the organisation.

353. Third, the concept of membership in MasterCard International Inc. equally remained materially unchanged, as set out in section 2.1.1 member banks continue to be bound by the very license agreement they concluded before the IPO and they still have to comply with ail network rules. (415)

354. Fourth, since 25 May 2006 the MasterCard payment organisation still operates in a de-centralised manner in Europe. Although the status of Regional Boards outside Europe changed after 25 May 2006 (i.e. they were converted to "advisory bodies"), the governance of the organisation in Europe remained virtually unchanged except for decision making on "antitrust sensitive" aspects of the member banks co-ordination. The European Board was maintained as decision making body. European batiks only consented to changes to the organisation's governance after the existence of the European Board was anchored in the Amended and Restated Certificate of Incorporation of MasterCard International Inc. (416) In approving the IPO the European member banks ensured that the European Board (417) kept its "key" decision making powers. (418) The European Board continues to review membership applications of new banks, has power and authority to fine member banks, to set intraregional operating rules and to approve an annual expense budget, etc. Hence, post-IPO European banks remained in charge of the business in Europe by deciding on all matters which they previously decided upon, except those issues that were perceived as sensitive from an antitrust perspective. (419)

355. Fifth, as before the IPO, MasterCard Incorporated and its subsidiaries continue co-ordinating the market behaviour of the organisation's member banks. MasterCard for instance continues to publish the results of multilateral agreements on interchange fees in the payment organisation on the MOL database so that all banks know and abide by the agreements. As before the IPO the three legal entities representing the organisation continue to enforce interchange fee agreements between the member banks by supervising the application of the correct interchange clearing message where the scheme owner MasterCard Incorporated processes the members' card transactions. (420)

356. Sixth, in order for a bank to become member in the MasterCard organisation it is not necessary for it to acquire shares in MasterCard Incorporated. Some banks are shareholders in MasterCard Incorporated, but there is no such requirement, as before the IPO on 25 May 2006. The membership concept was indeed not materially affected by the IPO (421).

357. MasterCard's viewpoint that the IPO of MasterCard Incorporated had changed the organisation's governance so fundamentally that any decision of MasterCard Incorporated's Global Board no longer qualifies as decision of an association but rather as "unilateral" act which each member bank bilaterally agrees to abide by, (422) cannot be accepted. Where undertakings entrust a third party to perform certain tasks that enable or facilitate the restrictive behaviour of the delegating undertakings does not exclude the existence of a restriction. (423)

Change of economic ownership in MasterCard Incorporated not decisive

358. To start with, the fact that MasterCard Incorporated has proper commercial interests towards its shareholders is no novelty due to the IPO. Before and after the IPO MasterCard Incorporated was a for-profit company and MasterCard does not dispute that its decisions were decisions of an association of undertakings before the IPO. This change is, however, immaterial to the qualification of an organisation as association of undertakings. The concept of an "association of undertakings" is neutral to the question whether the bodies managing the association obtain fees and whether -if they make profits- such profits are allocated to members, only. The change in the economic ownership in MasterCard Incorporated (424) is therefore immaterial for the qualification of MasterCard organisation as an association of undertakings.

Banks remain in charge of governance in Europe

359. Moreover, in assessing whether the MasterCard payment organisation still qualifies as association of undertakings after the IPO of MasterCard Incorporated, it must be borne in mind that the Global Board is only one body which forms the common structure that banks -and European banks in particular- use to co-operate and to co-ordinate their market behaviour. Key decision making powers in Europe still rest with other management bodies, as set out in section 2.1.2 that is to say the European Board and the national fora of local member banks.

360. Moreover, post-IPO MasterCard Incorporated's Global Board and its regional managers in Europe continue to support horizontal forms of co-operation in the organisation, by:

i. Encouraging the creation of new fora of local member banks with powers to co-ordinate their behaviour on a national level (cf. section 2.1.2. b);

ii. Publishing multilaterally agreed network rules of these fora (e.g: on the MOL database) which -together with those adopted by the European Board and those adopted on a global level- form the "statutes" of the MasterCard payment organisation in Europe;

iii. Enforcing multilateral agreements on domestic interchange and other national network rules adopted by local banks (to the extent that MasterCard Incorporated acts as processor of domestic or cross-border payment card transactions).

361. The horizontal nature of the member banks' co-operation therefore has not changed after the IPO and is corroborated by MasterCard's network mules (425).

362. It is only where banks fear antitrust scrutiny that they "outsource" decision making to the new Global Board. The evidence (cf. section 2.1.4.) shows that it was not the intention of European member banks to relinquish influence over the organisation's core business beyond antitrust sensitive matters.

363. MasterCard errs in its assertion that that the powers of European member banks to block a dissolution of the European Board post-IPO are irrelevant (426) and that this Board of bank appointed directors is "plainly subservient" to the Global Board.

364. On the one hand, MasterCard's depiction of the European Board's "few remaining powers" does not correspond to the role that these powers played for European member banks in approving the IPO. The European banks vigorously fought to maintain the powers of their regional board in Europe (cf. sections 2.1.2.a and 2.1.4.).

365. On the other hand, as to the "plainly subservient" character of the European Board, it derives from the past practice of the Global Board in withdrawing powers which it had -formally- delegated to the European Board that these "delegated" powers were withdrawn only in very exceptional circumstances of exposure to antitrust scrutiny (427). There are no indications that this practice should change after the IPO. In particular MasterCard's management made sure that the Global Board's powers to override decisions of the European Board were drafted in a manner that fully protects the interests of the European banks: (428)

366. Finally, for sake of completeness, it is also irrelevant whether or not an organisation has a proper legal personality. This is immaterial for qualifying it as "association of undertakings" within the meaning of Article 81(1) of the Treaty (429), as set out above in section 7.1.1. It is therefore also immaterial for qualifying the MasterCard payment organisation as association of undertakings.

367. In conclusion, the MasterCard payment organisation remains operating as "association of undertakings" in Europe after the IPO of MasterCard Incorporated on 25 May 2006.

7.1.2 Decisions by an association of undertakings

368. Where an association of undertakings is found to exist, as iii the case at hand, its decisions and recommendations which constitute the faithful expression of the association's resolve to coordinate the commercial conduct of its members fall within the scope of Article 81(1) of the Treaty. The Court recognised that even a non-binding recommendation in such case falls under the scope of Article 81(1) of the Treaty. (430)

369. In verband der Sachversicherer the Court found that a non-binding recommendation by a committee of experts may qualify as "decision of an association of undertakings" within the meaning of Article 81(1) of the Treaty on the basis of three elements: (431)

i. the common interest of members of the organisation in the subject of the recommendation;

ii. the de facto binding nature of the recommendation and

iii. the powers vested by the statutes of the organisation in the committee of experts to co-ordinate the policy of the members (432).

370. Applying AG Léger's definition and the Court jurisprudence to the setting of Intra-EEA fallback interchange fees, it is apparent that ail elements for qualifying the MasterCard MIF as decision of an association of undertakings within the meaning of Article 81(1) of the Treaty are fulfilled, both for the time period before and after the IPO of MasterCard Incorporated. This conclusion is based on the factual elements set forth in Chapter 2 and more particular on the following considerations.

7.1.2.1 Decisions on MIF before the IPO

371. As MasterCard sets out, its interchange fees transfer revenues from acquirers to issuers in order to "correct" the prices that issuers charge to cardholders and acquirers to merchants (433). The decisions of bodies in this association on the MIF before the IPO (be it a decision of the Europe Regional Board or of the Global Board) were therefore the faithful expression of the association's resolve to coordinate the commercial conduct of its members on the market. These decisions were binding upon member banks as all banks had to submit to decisions of the European Board and to the network rules set by the Global Board upon joining the association. Member banks which disregarded the network rules or decisions of the Global Board or the European Board were faced with expulsion (434) [business secret - 3a]

372. MasterCard does not contest that its MIF was a decision of an association of undertakings within the meaning of Article 81(1) of the Treaty before the IPO (435).

7.1.2.2 Continuing effects of MasterCard MIF after the IPO

373. The changes to the governance of MasterCard Incorporated on 25 May 2006 did not affect the interchange fee fallback mechanism as such. Modifications were limited to a transfer of powers for taking decisions on the level and structure of the fees. However, the principle that some multilaterally set fallback interchange fee will always apply "as fallback" to a payment card transaction in the absence of bilateral agreements remains rooted in a network rule that was adopted before the IPO. The effect of this decision of an association of undertakings therefore continues until today. As far as the nominal level of Intra-EEA fallback interchange fees is concerned, it is important to note that the fees remained entirely unchanged until 11 January 2007. Even thereafter, the changes were minor.

374. In conclusion, the ongoing effects of the MasterCard MIF -as set out hereafter in section 7.2.3- continue to be caught by Article 81(1) of the Treaty in the period post-IPO as effects of a decision of an association of undertakings.

7.1.2.3 New Decisions on level of MIF after the IPO

375. It has been set out above in section 7.1.1 that the MasterCard payment organisation continues to qualify as association of undertakings within the meaning of Article 81(1) of the Treaty. The remaining question is whether there is scope for functional distinction [business secret - 3a]

376. In concrete terms, the question is whether the decisions of MasterCard [business secret - 2a] to apply slightly modified Intra-EEA fallback interchange fees as of 1 January 2007 as well as to adopt SEPA fallback interchange fees as of 1 January 2008 also qua1ify as "decisions" of an association of undertakings.

No evasion of competition rides simply on account of form

377. As AG Léger underlined, the concept of "association of undertakings" in Article 81 of the Treaty seeks to prevent undertakings from being able to evade the rules on competition on account simply of the form in which they coordinate their conduct on the market. The Article covers not only direct methods of coordinating conduct between undertakings. The very purpose of the concept of an "association of undertakings" is to safeguard the effet utile of the prohibition set out in Article 81(1) of the Treaty. It prevents that competitors co-ordinate their commercial behaviour on the market indirectly by delegating the co-ordination to a common body and by submitting upfront with binding effect to the respective decisions of that body (436).

378. It was set out in sections 2.1.2. and 2.1.4 that MasterCard's member banks shaped and eventually approved the IPO in order to perpetuate the MIF as part of the business model in a form that they perceived to be less exposed to antitrust scrutiny. Contrary to MasterCard's argument (437) the all of avoiding exposure to antitrust risks due to the MasterCard MIF was a clear driving force behind the IPO (438). Rather than modifying the business model to bring it in line with EU competition law, the banks chose to change the governance of their co-ordination specifically for antitrust sensitive decision making. The member banks effectively "outsourced" this decision making to a new management body and made sure that their direct influence on the Global Board's directors would be limited to minority rights. However, the banks also agreed to the IPO of MasterCard Incorporated after MasterCard's management assured them that the banks' interests will continue to be preserved under a new "enhanced customer approach" (439) and via the local input of the banks in the decision making (440).

379. It cannot be doubted that in approving the IPO and thereby delegating the decision making powers for the MIF to the new independent Global Board, the member banks legitimately expected and therefore agreed that this Board would henceforth set the MIF in a manner that is in their common interest. The member banks were a driving force behind the new governance model and they approved the preceding and subsequent "outsourcing" of interchange related decision making from boards of bank directors (the European Board, the MasterCard Member Forum in the United Kingdom) to [business secret - 2a]

The Global Board whose task it would be to manage interchange in a manner that is [business secret - 3b] (441).

380. By approving the IPO the banks effectively resolved to continue using the Global Board of MasterCard Incorporated as "common structure or body" for the co-ordination off their policies regarding the pricing off MasterCard and Maestro payment card acceptance. Decisions of the Global Board remain the "faithful expression of the association's resolve to coordinate the commercial conduct off its members" as they reflect the common interest of the organisation's [business secret - 4] European member banks (442).

MIF still adopted in banks' common interest

381. The circumstance that members of the Global Board are "independent" within the meaning of the NYSE criteria (cf. section 2.1.3.1) is not a decisive question for there to be an association of undertakings. As an organisation's members entrust decision making powers to a common body with the expectation that the body's subsequent coordination off their competitive behaviour will occur in the common interest of the members, the independence of such body is no obstacle to qualifying its decisions as decision off an association of undertakings.

382. In the case at hand MasterCard's member banks could as well chose a notary public or an independent consultant who is subject to deontological rules for setting the level of interchange fees. They rather resolved to delegate these powers to a small number of bank delegates on the European Board and - after the IPO - [business secret - 2a] the Global Board. In both instances there is no guarantee that the decision maker fulfils the wishes of each individual member bank that submits to the decision or that each bank effectively controls the outcome of the decision makers. This is not the decisive point. The decisive question here is whether overall it is in the common interest of the banks that some entity or person, whom they entrust with decision making powers, establishes through the MIF a minimum price which merchants in Europe must pay for accepting MasterCard and Maestro branded payment cards. This is the case.

383. Despite the IPO there is still a clear and pronounced commonality off interests amongst the banks of the MasterCard payment organisation in Europe in having a MIF which results from the fact that all banks benefit from a common minimum price floor merchants must pay for accepting MasterCard and Maestro branded payment cards. Both issuing and acquiring banks have an interest in the perpetuation of the system with a MIF, even if some banks may prefer a different level off the MIF than others. MasterCard's argument (443) that there is no "commonality of interests" on all interchange related issues between the banks, because some banks are "net issuers", and therefore have a short-term financial interests in high interchange fees, while others are "net acquirers" and therefore have a short-term financial interest in a lower interchange fee, is legally not relevant.

384. For there to be a "decision" it is not required that the members of an association unanimously approve it and that each of them agrees on all aspects of a decision but it is sufficient that a decision is taken by the competent body within an association (444). Hence, the fact that some banks (whether issuers or acquirers) may not be satisfied with all interchange related issues to the European Board or -since the IPO- by [MasterCard] [business secret - 2a] is irrelevant.

385. MasterCard also ignores the fact that until 31 December 2004 the "No-Acquiring-Without-Issuing" rule obliged every acquirer also to carry an issuing activity. As a result, almost ail acquirers are at the same time issuers and their interest in high interchange for the issuing part of their cards business is common. Net acquirers may be less interested in a high level of MIF than acquirers with an issuing business but they, too, have an interest in maintaining a MIF as part of the organisation's business model. This is because a MIF raises the marginal cost of all acquirers alike in a collective manner which allows them to yield a rent from merchants which they could not obtain under competitive circumstances. The additional revenues are then transferred by acquiring banks to issuing banks with the expectation that some of the money is used .by issuers to subsidise card issuing. This in turn could spur an increase of sales volumes at merchant outlets thereby raising every acquirer's profits.

386. The commonality of interests in a MIF also extends to the scheme owner MasterCard Incorporated and its consolidated subsidiaries as the membership fees they charge to banks in exchange for their co-ordination and network services are transaction related (that is to say ad valorem and/or flat fees per card payment) (445).

387. Moreover, the interests of MasterCard Incorporated's shareholders - be they banks or not - are also not opposed to those of the member banks in having a MIF (446). As MasterCard Incorporated can expect higher revenues and thus higher profits the more banks join the scheme and the higher membership fees are, it is unlikely that stockholders in MasterCard Incorporated would oppose the MIF as it artificially attracts banks to the scheme through revenues.

388. Finally, it is a fact that the Global Board still takes decisions on a MIF virtually "on behalf of the banks" (447). (448) [business secret - 5a]

389. Developments after the IPO raise indicate that MasterCard's management takes into account concrete banks interests in setting the level of fallback interchange fees. Reference is made to the background of the new domestic MIFs in) [business secret - 6] as set out in section 3.1.9.2 [business secret - 2a]

Also the European banks are still represented on the Global Board and can thereby express their views on interchange relates issues. It is unrealistic to expect that European banks will not make use of this possibility, in particular as MasterCard's management acknowledges that it regularly speaks to issuers "what their wishes are".

390. The fact that bank delegates now no longer decide upon on interchange fee related matters cannot be decisive. According to the jurisprudence (450) a decision of association of undertakings does not require that all members of the association agree upfront on a non binding recommendation for that recommendation to be caught by Article 81 (EC). If even such a non-binding recommendation was found to be anti-competitive, then the fact that the banks in the present case could not formally influence the decision making on the MIF is not important as long as the member banks adhere to [MasterCard] [business secret - 2a] decisions on interchange fees and that they remained licensees and members of the organisation. This is the case.

391. The delegation to a third party of certain tasks pertaining to a restriction does not exclude the existence of a restriction. For example, a cartel is possible also when members delegate to a facilitator certain tasks, which could in theory be the setting of the exact price level, if it will serve as a tool for the coordination of the members' activities. The lack of relevance of delegation for the assessment under article 81 EC is even more obvious in an association context. It is open for the Commission to consider that undertakings delegating a task to a common body or restricting their commercial freedom in view of the decisions of such a body form an association of undertakings via that body. To this effect, sec recital 344 of the Commission decision of 10 December 2003 (Case COMP/E-2/37.857 - Organic peroxides), where the Commission concluded that "AC Treuhand infringed Article 81 of the Treaty and Article 53 of the EEA Agreement. AC Treuhand has, as spelled out in the SO, participated in the agreement as an undertaking and/or took decisions as an association of undertakings. In analogy with its approach to leave open whether an infringement was an agreement or a concerted practice, the Commission does not consider it necessary to specify in which rote (undertaking or association of undertakings) these Articles were infringed" (451).

392. The Commission finally rejects MasterCard's argument that "by stating that MasterCard is operated solely by, or in (the interests of the licensee banks, the Commission effectively suggests that the rights and legitimate expectations of the non licensee shareholders arising out of the IPO are being ignored" and that such line of argument would imply that "the directors are operating in a manner inconsistent with their duties and obligations" (452).

393. The Commission's point is not that MasterCard Incorporated is solely operated in the interests of the organisation's member banks. Obviously, the management of any incorporated company is in the first place responsible towards the shareholders. However, as set out above, the commercial interests of a scheme owner and the commercial interests of the organisation's member banks are related as far as a MIF is concerned. By maximising member banks' proceeds from interchange fees a scheme owner can also increase its own revenues and ultimately dividends to its shareholders (be they banks and/or public investors). It is therefore implausible that MasterCard Incorporated's Global Board would act unlawfully towards shareholders if it were to consider the commercial interests of issuing banks in setting the level of interchange fee rates in Europe.

Banks agreed to new governance regarding the MIF

394. Finally, MasterCard's argument that the powers of the European Board to set Intra-EEA fallback interchange fees had been withdrawn "without the banks'agreement" (453) cannot be accepted.

395. (454) (455) (456) It is rather plausible that this strategy was conceived before the IPO with full knowledge and approval of MasterCard's member banks. [business secret-2a]

396. As to the IPO itself it was set out in section 2.1.4 that the European member banks were deeply involved in the process of creating the new governance for the MIF. Their requests regarding the scope of the European Board's powers were met by leaving essentially the same powers as pre-IPO "minus interchange" (457). Responding to shareholder banks' concerns management finally announced an (458). [business secret-2a]

397. The Commission disagrees with MasterCard's argument that since the IPO interchange fees are "unilaterally" imposed on member banks in a supplied-customer like relationship. Rather, as any other decision of the organisation's managing bodies, the MIF remains to be the faithful expression of the association's resolve to coordinate the commercial conduct of its members.

398. For the above reasons, the association's network rules that form part of the MasterCard MIF as well as decisions taken by the European Board and/or by MasterCard [business secret - 2a] which implement these rules by setting concrete levels and types of fallback interchange fees for MasterCard/Maestro payment card transactions iii the EEA have been and still remain decisions of an association of undertakings within the meaning of Article 81(1) of the Treaty. (459)

399. At any rate, even if after the IPO MasterCard Incorporated's relation to banks qualified as that of a franchisor to franchisees rather than being a horizontal form of co-operation within an association of undertakings this is no reason why the MIF should fall outside the scope of Article 81(1) of the Treaty. As is apparent from Commission Regulation (EC) N° 2790-1999 (460) anti-competitive aspects of franchise agreements may restrict Article 81(1) of the Treaty, too. MasterCard has provided no reasons let alone evidence why the MIF would escape the prohibition of Article 81 (1) of the Treaty if it were a vertical restraint.

7.2 Restriction of competition

400. In its Visa decision of 24 July 2002, the Commission considered that a multilateral interchange fee restricts competition within the meaning of Article 81 (1) of the Treaty/Article 53 (1) of the EEA Agreement by restricting competition between payment card systems (§ 61-69) (461). The Commission's findings in this case confirm that a MIF distorts competition between acquiring banks and that the effects of the MIF in the network and issuing markets reinforce the restrictive effects in the acquiring markets.

7.2.1 The object of the MIF

401. MasterCard contests of a restriction by object and refers to the Visa II decision where it is stated that a MIF is not a restriction of competition by objects as its aim is to increase the stability and efficiency of a four party scheme (462).

402. However, the concept of a restriction by object does not presuppose that the parties to an agreement have the subjective intention of restricting competition. Agreements can be restrictive by object even if the parties to it are able to show that restricting competition was not their (primary) aim, or that they had other laudible motives (463). Arguments which purport to show that such agreements may also have pro-competitive aims and effects are not be considered under Article 81 (1) of the Treaty (464), but rather under Article 81(3) of the Treaty.

403. In order to fall into the category of a restriction by object, is suffices that an agreement/a decision has by its very nature the potential for restricting competition (465), for instance that it has the obvious consequence of fixing prices.

404. In this respect, MasterCard does not contest that the MIF will typically fix a floor for MSCs because -as MasterCard realizes- it is reasonable to assume that the interchange fees affect to some degree MSCs (466) and that an MSC "typically reflects the MIF" (467).

405. The fact that the MIF typically determines a floor for the price which merchants must pay for accepting payment cards id indeed an indication that MasterCard's MIF may by its very nature have the potential of fixing prices. MasterCard's Intra-EEA fallback interchange fees are conceived as "default MIF". They apply "by default/as fallback" to a POS payment card transaction, that is only if a transaction is not yet subject to a bilateral agreement between the issuer and the acquirer concerned on the level of an interchange fee (468). In practice, however, Intra-EEA fallback interchange fees have been applied to virtually all cross-border payments with MasterCard's payment cards as well as to domestic payment transactions in [business secret-6] EEA Member States (469). For these transactions, MasterCard's MIF also acts like a minimum price recommendation for transactions on a domestic level (470). By agreeing on specific interchange fees bilaterally member banks may take the Intra-EEA fallback interchange fees into account as minimum starting point (see section 7.2.3.1.2 bb).

406. MasterCard, however, argues that the Commission was not able to draw from any document on the Commission's file or any statement from MasterCard that would support a finding that the MasterCard MIF has as its object the restriction of (price) competition between issuers and/or acquirers (471). However, MasterCard has stated in writing that its MIF was a means of "correcting" prices that would otherwise be set on both sides of its scheme by "independent competitive processes" amongst issuers and amongst acquirers (472). Hence, MasterCard also seems to view its MIF as mechanism that "remedies" the fact that prices charged between issuers and acquirers would otherwise be set through competition. The concern cannot therefore be discarded that the MasterCard MIF may have as its object the restriction of price determination through competition.

407. However, given that it can be clearly established that the MasterCard MIF has the effect of appreciably restricting and distorting competition to the detriment of merchants in the acquiring markets it is not necessary to reach a definite conclusion as to whether the MasterCard is a restriction by object within the meaning of Article 81 (1) of the Treaty.

7.2.2 The effects of the MIF

408. The assessment of MasterCard's MIF as a restriction of competition is based on its restrictive effects on competition in the acquiring markets. In the absence of a bilateral agreement, the multilateral rule fixes the level of the interchange fee rate for all acquiring banks set charges to merchants. Prices set by acquiring banks would be lower in the absence of the multilateral rule and in the presence of a rule that prohibits ex post pricing.

409. In evaluating these restrictive effects under Art 81 (1) EC, the Commission also takes into account that:

- Inter-system competition between payment card networks creates an upward pressure on interchange fees (section 7.2.4).

- MasterCard's MIF is not subject to sufficient demand-side or other competitive constraints (section 7.2.5).

7.2.3 Restriction of competition in the acquiring markets and effects in the issuing markets

7.2.3.1 Restriction of competition in the acquiring markets

410. MasterCard's MIF constitutes a restriction of price competition in the acquiring markets. In the absence of a bilateral agreement, the multilateral "default" rule fixes the level of the interchange fee rate for all acquiring banks alike, thereby inflating the base on which acquiring banks set charges to merchants. Prices set by acquiring banks would be lower in the absence of this rule and in the presence of a rule that prohibits ex post pricing. The MasterCard MIF therefore creates an artificial cost base that is common for all acquirers and the merchant fee will typically reflect the costs of the MIF (473). This leads to a restriction of price competition between acquiring banks (474) to the detriment of merchants (and subsequent purchasers).

411. A further consequence of this restriction of price competition is that customers making purchases at merchants who accept payment cards are likely to have to bear some part of the cost of MasterCard's payment the customers use. This is because depending on the competitive situation merchants may increase the price for all goods sold by a small margin rather than internalizing the cost imposed on them by a MIF (475).

412. The collective decision by the MasterCard organization to set a MIF inflates prices charged by acquirers to merchants for acquiring cross-border credit and debit card transactions with MasterCard's payments cards (476). This finding is in line with the Commission's previous case practice. The Commission, found in its Visa II decision that a MIF has the effect of distorting the behavior of acquirers vis-à-vis their customers, because it creates an important cost element which is likely to constitute a de facto floor for fees charged to merchants they acquire (477).

7.2.3.1.2 Restriction of price competition with respect to the acquiring of domestic card payments

a) cross-border acquiring of domestic payments

413. "Central acquirers" are acquiring banks that acquire payment card transactions outside their home country. According to MasterCard's network rules (478) and as already set out in country in the sense that they must pay local issuing banks the applicable domestic MIF unless local issuers agree to negotiate bilaterally a diverging set of interchange fees. If no domestic MIF exists and if no bilateral agreements are concluded, then the Intra-EEA fallback interchange fees. If no domestic MIF exists and if no bilateral agreements are concluded, then the Intra-EEA fallback interchange fees automatically apply to the central acquirer for the acquisition of domestic card payments in the target country. This MIF then effectively determines a floor for the merchant fees which the central acquirer can offer to local merchants. This situation is prevalent at least in:

i. Sweden; (479)

ii. France; (480)

iii. Spain; (481)

iv. the Netherlands. (482)

414. In addition to raising the cost base of central acquirers and thereby inflating the prices such acquirers can offer to local merchants, the interplay of the Intra-EEA fallback interchange fees (or a domestic MIF) and bilaterally agreed interchange fees may have the additional effect of hindering entry of foreign acquiring banks.

415. Central acquiring banks to agree bilaterally on interchange fees that undercut a default MIF in the target country. This can be the Intra-EEA fallback interchange fees or - where adopted - a domestic MIF. Where the applicable default MIF is higher than interchange fee rates that are bilaterally agreed between local banks, a central acquirer will incur an artificial cost disadvantage for acquiring transactions in that country because it must pay the default MIF to the local issuers. The right of MasterCard member banks to agree bilaterally on interchange fees that undercut a default MIF in the target country (be it the Intra-EEA fallback interchange fees or a domestic MIF) combined with the obligation on the foreign bank to pay the local default MIF may act as deterrent for new entrants. This was, for instance, the case in [EU Member State] where local issuing banks agreed on a set of bilateral interchange fees for certain key merchant segments that effectively undercut the applicable default MIF in that country. This hampered the entry of a central acquirer to [EU Member State] (483).

b) Domestic acquiring of domestic payments

416. MasterCard's finally also errs in contending that the economic importance of its Intra-EEA fallback interchange fees was insignificant (484) because this MIF applied merely to cross-border transactions. MasterCard's Intra-EEA fallback interchange fees are factually not only a cross-border MIF but also a domestic MIF. These fees apply as such (that is to say as fallback) to domestic card payments in [business secret - 6] EEA Member State where local members neither agreed on bilateral interchange fees nor on a domestic MIF. Moreover, the fees also provide guidance to member banks for setting the rates of specific domestic interchange fees.

ba) cross-border MIF sometimes identical to domestic MIF

417. First, the MasterCard's Intra-EEA fallback interchange fees create price restrictions with respect to domestic MasterCard/Maestro transactions (485). This happens because according to MasterCard's network rules the intra-EEA interchange fees apply as "fallback" in countries where local banks do not agree on specific domestic MIFs for domestic transactions [business secret - 2a] (or where [MasterCard] [business secret - 2a] has not adopted on the banks' behalf specific domestic "SEPA" interchange fees).

418. According to the Commission's information - which remained uncontested by MasterCard up to the date of this decision - the Intra-EEA fallback interchange fees apply to domestic card transactions with MasterCard branded credit and charge cards in at least Member States, namely (486) [business secret - 6]

419. In [business secret - 6] local member banks have explicitly decided to apply the Intra-EEA fallback interchange fees for MasterCard branded credit and charge cards to domestic payment transactions (487). In the other [business secret - 6] Member States the fees apply as "fallback" in the absence of a diverging local agreement (488).

420. In the debit card segment the Maestro Intra-EEA fallback interchange fees also apply to domestic transactions with Maestro branded debit cards in EEA Member States such as, for example, in and (489) [business secret - 6].

bb) Cross-border MIF acts as benchmark for setting specific domestic IFs

421. Second, some of MasterCard's member banks view Intra-EEA fallback interchange fee rates de facto as a minimum starting point for setting the rates of domestic interchange fees. Due to MasterCard's network rules issuing banks have the certainty that in the absence of their consent to the adoption of a domestic MIF the Intra-EEA fallback interchange fees will always automatically apply as domestic MIF in their country (490). Issuing banks have no incentive to agree to domestic interchange fees below this default rate because interchange fees are revenues. Both the adoption of a domestic MIF and of a bilateral agreement requires, however, the consent of the issuing banks (see section 3.1.1). Hence, even in countries where MasterCard's (see above), the cross-border interchange fees may act as minimum benchmark for setting the level of domestic interchange fee rates.

422. This was, for instance, the case in [EU Member State]. MasterCard's [EU Member State] member banks view the Intra-EEA fallback interchange fees as a "starting point" for setting domestic MIFs.

423. [bank], the largest acquirer for MasterCard in the [EU Member State], made the following observation: (491)

"[bank] annually negotiates POS intra-country Interchange Fees with the [EU Member State] banks. Basically the negotiated agreements have in common that the fees use the Intra-European Fallback Fee Levels as starting point. As for consumer cards, [bank] has bilaterally agreed with these banks that they receive something extra, on top of the Intra European Fallback Interchange for consumer cards. However, there are differences per issuing bank, which take many factors specific to the bank into consideration. For instance, [bank] negotiations with one [EU Member State] issuer have resulted in an agreement that a higher interchange fee will be paid, in case that issuer would generate more transactions, cq volumes at merchants acquired by [bank] than a certain target level" [Emphasis added]

424. Indeed, all domestic MIFs for MasterCard consumer credit cards that [bank] bilaterally negotiated with domestic issuing in the [EU Member State] were set above the level of the default fees. The banks simply added a certain amount or percentage on top of MasterCard's Intra-EEA fallback interchange fees (492).

7.3.1.1.3 Quantitative analysis: MIF and merchants fees

425. The Commission also carried out two quantitative analyses to see whether and to what extent the Intra-EEA fallback interchange fees set a floor under the merchant fees.

426. A first analysis was carried out for [business secret - 6] Member States [business secret - 6] where the cross-border interchange fees apply as such to domestic payment card transactions. Calculations are based on weighted averages of all merchant fees for MasterCard branded cards charged by [business secret - 6] acquiring banks. The aim of this analysis was to see to what extent the MIF sets a floor for the MSC in general.

427. The second analysis is based on data from 17 acquiring banks in 15 EU Member States. In this second analysis the MIF is compared to the MSC of the top and bottom decile of the acquirers' respective client base (493). The aim of this analysis was to investigate to what extent MasterCard's MIF determines the merchant fee of all small as opposed to large merchants so to assess the relative capacity of each merchant group to extent countervailing buyer power on acquirers to absorb the interchange fee cost element.

a) First quantitative analysis

428. The results from [business secret - 6] are based on the total population of merchants [business secret - 6] of all acquirers in [business secret - 6]. The below diagram shows that the average merchant fee for MasterCard branded credit and charge cards of all acquirers in [business secret - 6] dropped from 1.55% in 2000 to 1.51% in 2002. Both the highest and the lowest Intra-EEA fallback interchange fee rate increased from 2000 to 2001 and from then onwards remained stable at 0.80% for the lowest and 1.30% for the highest. Taking into account that the weighted average of all Intra-EEA fallback interchange fees of the MasterCard branded credit and charge cards was [business secret - 6] in 2002 (494), this means that MasterCard's MIF accounted for approximately 73% of the average merchant fee for MasterCard branded credit and charge cards in [business secret - 6] in 2002 (495).

< diagram 9 >

429. Results from [business secret - 6] are based on a population of merchants [business secret - 6] of two large acquirers. Taking into account that the weighted average of all cross-border interchange fee rates of the MasterCard brand was 1.11% in 2002 and 1.10% in 2003 (496), this means that MasterCard's MIF accounted for approximately 56% (2002) and 60% (2003) of the average merchant fee for MasterCard branded payment cards in [business secret - 6] of the acquirers observed.

< diagram 10 >

[business secret - 6]

430. Results from the [business secret - 6] are based on a total population of merchants [business secret - 6]. Data stem from one acquirer. Taking into account that the weighted average of all cross- border [business secret - 6] order interchange fee rates of the MasterCard brand was [business secret - 6] in 2002 and [business secret - 6] in 2003, this means that MasterCard MIF has accounted for approximately 57% (2002) and 60% (2003) of the average merchant fee for MasterCard branded payment cards of this [business secret - 6] acquirer. These figures of 57% and 60% are similar to those in [business secret - 6].

< diagram 11 >

[business secret - 6]

431. In conclusion the first quantitative analysis shows that the Intra-EEA fallback interchange fee rates constitute a floor for the merchant fees. Data from 2002 indicate that the MIF on average determine up to 73% of the MSC.

b) Second quantitative analysis

432. In the second quantitative analysis the Commission assessed to what extent MasterCard's MIF determines the merchant fee of small as opposed to large merchants. This was done by comparing the weighted average of MasterCard's Intra-EEA fallback interchange fees to the weighted average merchant fee which each of the 17 acquirers charged to the top and the bottom decile of its merchant base.

433. The below diagram 12 indicates the share of interchange fees in the merchant fee for the top decile and the bottom decile of the acquirer's respective client base, as well as the total share of interchange fees taking the top and bottom decile of the acquirer's respective client base together (497).

434. The analysis indicates strong variances according to the size of a merchant. With the exception of one acquirer (498), all acquirers surveyed charges a considerably higher margin on top of the interchange fee to the bottom decile of their merchant base as compared to large merchants. Thus, on average, the interchange fee determined in 2002 84.27% of the merchant fee paid by the top decile of the 17 acquirers' merchant client base, while it determined "only" 45.97% of the merchant fees paid by the bottom decile of the acquirers' merchant client base.

435. The evidence indicates that the MasterCard MIF sets a floor to MSCs for both small and large merchants. Even larger merchants, whom one would expect to have the greatest degree of negotiating power vis-à-vis their acquiring bank, cannot negotiate a MSC below the level of the MIF. Smaller merchants with less market power pay a higher mark up to their acquirer and therefore the interchange fee element of the MSC is relatively lower as compared to the share of the MIF in the MSC paid by large retailers. However, as the analysis shows, for all of the smaller and for most of the larger retailers (499) the MIF constituted a floor under the MSC (IEFB stands for "Intra-EEA fallback interchange fee").

< diagram 12 >

436. In conclusion, the second quantitative analysis shows that the MIF determines a floor under the MSC typically even for large merchants.

7.3.2.1.4 Further evidence from merchants

437. The commission asked merchants how they experience competition between acquiring banks. The examples in the subsequent recital stem from a survey with more than 200 merchants in 2004 showed that particularly large merchants experienced in their negotiations with acquiring banks that MasterCard's interchange fees determine a floor for the merchant fee (500).

438. [a petroleum company] stated: "in reality there is little scope for acquiring banks to be competitive because of the high level of interchange giving only limited potential for price differences". [United Kingdom based supermarket chain] said: "the main reason why we see limited competition between acquiring banks to acquire MasterCard and Maestro" is that " the vast majority of cost involved in acquiring these transactions is interchange, set by MasterCard and passed on to [United Kingdom based supermarket chain]. The acquirer has little influence over this interchange (...). [United Kingdom based supermarket chain] inevitably arrives in a position where the negotiation centres around the margin paid to the acquiring bank and the value and quality of their service." [an airline] stated: "Due to the fact that [MasterCard] domestic MIFs are fixed by the issuing banks the pricing competition is very restricted". A German variety shop underlined: "In Germany there is very strong competition. Problem: The Domestic MIF which is the dominant component of the Merchant Service charge is not negotiable". [a furniture shop] made the following observation: "(...) Banks in Belgium reveal very little on the structure of the MSC. They claim that the interchange fee set by MasterCard and Visa leaves very little margin, but they have never stated how high interchange fee for domestic payments actually is (...) (501).

7.3.2.1.5 Commission assessment of MasterCard's arguments why its MIF would not restrict competition between acquiring banks.

a) Arguments linked to regulation (EC) N° 2560-2001

439. In reply to the Statement of Objections of 24 September 2003 MasterCard argues that given that in general acquirers must charge the same merchant fee for both domestic and cross border transactions on the basis of Regulation (EC) N° 2560-2001 of the European Parliament and of the Council of 19 December 2001 p, cross-border transactions, it is more plausible that it is the domestic MIF that effects the setting of the merchant fee, while the impact of the cross-border MIF is very limited.

440. The Commission cannot accept that argument. Given that Regulation (EC) N° 2560-2001 applies only to payments in Euro, it is obvious that any argument based on that Regulation, even if it were valid, could only be made with respect to 13 Member States of the 30 EEA Member States of the 30 EEA Member States. In all other countries acquirers may charge different merchant fees for domestic and cross-border transactions and there is evidence that they do so. For instance, in Denmark virtually all merchants participating in the Commission's merchant market survey replied that they pay two distinct merchant fees, one for domestic and one for cross-border transactions (502).

441. Even within the Eurozone it appears to be (still) common for merchants to pay specific merchant fees for cross-border transactions (503). Moreover, in certain cases, local bank associations have restructured acquiring operations in a manner that generally avoid paying the same charge with respect to cross-border and domestic transactions (504).

442. More fundamentally, in assessing the economic effect of a MIF on prices, it is not decisive how individual acquirers precisely account for the (artificial) input cost. In Member States where Regulation (EC) N° 2560-2001 does not apply, an acquirer may decide to charge a given merchant separately for domestic and for cross-border payment transactions, or it may weight the respective share of domestic and cross-border transactions generated by the merchant and charge a single merchant fee based on the weighed average. In both alternatives, merchants and subsequent customers are harmed by the inflated cost base of merchant fees. In member States where Regulation (EC) N° 2560-2001 applies, acquiring banks will charge merchants one blended fee for cross-border and domestic card payments rather than two separate ones. They will then weight the proportion of cross-border and domestic transactions and set the blended merchant fee accordingly.

443. There is evidence that this is precisely how acquirers proceed. [bank] - one of the largest Europan acquirers - stated: (505) "The MSC [merchant fee] will take into account the percentage contribution of domestic, intraregional and interregional card volume. Generally the greater the proportion of cross border transactions, the higher the MSC, to reflect the higher interchange fees that apply."

444. In conclusion, the fact that merchant fees may be blended for cross-border and domestic transactions (or, for that matter, for payments made using different card types or card brands) does not alter the restrictive and distortive effect of MasterCard cross border interchange fees in the acquiring markets, since in both situations (blending/no blending) the costs are passed on to merchants and the MIF is a significant element of the price paid by merchants for card acceptance (506).

b) Diverging Interpretation of First Quantitative Analysis

445. MasterCard notes that the first quantitative analysis shows that MSCs may decline over time despite the fact that the interchange fee remains unchanged. From this MasterCard concludes that "the interchange fee is not responsible for the level of the MSC (507)".

446. Besides the fact that this conclusion is in conflict with MasterCard's own observations on the correlation between the MIF and the MSC (see section 7.1.2), MasterCard's interpretation of the first quantitative analysis is erroneous. The fact that the total merchant fee changes over time while the interchange fee rate remains static does not disprove that MasterCard's MIF increased the acquirers' marginal cost in the three Member States observed. An acquirer's marginal costs are composed of several elements besides a MIF. These other marginal costs may evolve over time (as well as the mark up) while the price floor due to the MIF indeed remains (almost) (508) unchanged. Hence, the decrease of the final price which merchants in [business secret - 2a] paid for MasterCard branded credit cards does not imply that the MasterCard MIF was not "responsible" for the MSC. Without the MIF the total level of MSC in the three Member States would have been considerably lower, for the reasons set out below in section f).

c) Criticism of Second Quantitative Analysis

447. MasterCard moreover criticizes the second quantitative analysis as "potentially misleading". An acquirer's margin on large and small merchants might be similar, because acquirers could possibly have higher per unit costs vis-à-vis smaller merchants than with respect to larger merchants. Hence, even if larger merchants indeed paid over fees on average, this simply meant that larger merchants could squeeze acquirer margins (509).

448. However, the purpose of the second quantitative analysis was to assess the differential between merchant fees paid by larger and small merchants to assess the extent to which larger ones are in a position to negotiate a MSC below the acquirers charge to merchants at large would be lower. This is the case, because the price each individual bank could charge to merchants would be fully determined by competition rather than to a large extent by a collective decision among (or on behalf of) the banks.

d) Acquirers may sometimes absorb MIF - Spain

449. MasterCard considers that the bargaining power of merchants may sometimes be so strong that an acquirer has to absorb a part of the interchange fee (510). MasterCard claims that it was "aware of cases where acquirers grant merchants MSCs below the level of the MIF" and referred to Spain as an example (511).

450. However, this is no reason why the MIF would not be caught by Article 81 (1) of the Treaty. Even if an acquirer may choose to absorb a portion of the MIF, the MIF still inflates that acquirer's MSCs by a certain extent. The price is still not fully determined through competition but to a large extent by a collective decision among (or on behalf of) the banks.

451. The absorption of the MIF by acquirers is moreover very exceptional. It was demonstrated in the second quantitative analysis that in the very large merchants are unable to negotiate an MSC below the MIF. This result of the quantitative analysis is corroborated by evidence collected from merchants during the Commission's merchant market survey in 2004 (see section 7.3.2.1.3 b and 7.3.2.1.4).

452. The situation in Spain is finally specific for two reasons. First, MasterCard member banks in Spain operate in three local networks (512) rather than in one single network, as in other EEA Member States. Domestic MIFs in Spain therefore existed until very recently on two levels: inter-system and intra-system. Second, banks adopted specific interchange fee rates virtually for each single merchant segment, sometimes even for individual companies. Taking into consideration the existence of hundreds and even thousands of different domestic MIF rates in Spain, an acquiring bank will blend these rates into one or a few MSCs. Inevitably, a blended MSC will be an average of all interchange fee rates and therefore sometimes lie below certain interchange fee rates. Even in this specific situation banks can be expected to consider the domestic MFIs as input cost in calculating a weighted MSC.

e) MIF no restriction because it reduces charges to cardholders

453. MasterCard argues that the analysis of the Commission was "wrongly asymmetrical" because the Commission only focused on MSCs not on cardholder fees. MasterCard sets out that "(...) if the interchange fee is viewed a "a de facto" floor for MSCs, then it is just as much a "de facto floor"

453. MasterCard argues that the analysis of the Commission was " wrongly asymmetrical " because the Commission only focused on MSCs not on cardholder fees. MasterCard sets out that "(...) if the interchange fee is viewed as a "de facto floor for MSCs then it is just as much a "de facto floor" [...] expert Prof. [...] "(...) the interchange fee - whether it is positive or zero or negative - always forms a price floor for both merchant service charges and for cardholder service charges." (513)

454. That argument cannot be accepted. Under Article 81(1) of the Treaty there is legally no reason why the negative effect of the MIF on prices in the acquiring markets to the detriment of merchants (and subsequent purchasers) should not constitute a restriction of competition because of potential benefits which a MIF may create for cardholders.

f) MIF not a restriction because effect would be like "excise tax"

455. MasterCard puts forward that the interchange fee does not favour a particular acquirer or type of acquirer over other acquirers/types of acquirers. "The interchange fee is a common identical cost, borne by all acquirers, that does not influence price competition between acquirers in terms of determining the level of MSCs" (514). Visa's expert raised a similar argument at the oral hearing by comparing the MIF to an excise tax (515).

456. Factually, MasterCard neglects that a MIF does not have neutral effects on all acquirers but that it may well disadvantage certain acquirers to the benefit of others (see section 7.2.3.1.2 a).

457. Besides this fact, even if one were to qualify a MIF as a kind of "excise tax" this is no reason why the MIF should fall outside Article 81 (1) of the Treaty. The collective act of competing undertakings to raise charges for consumers is subject to the prohibition of Article 81(1) of the Treaty.

458. If the concept of a restriction of competition within the meaning of Article 81 (1) of the Treaty had to be interpreted as MasterCard suggests, then Article 81 (1) of the Treaty had to be interpreted as MasterCard suggests, then Article 81 (1) of the Treaty would be entirely deprived of its effet utile. The MasterCard MIF not only creates an (artificial) common cost for acquirers and thereby sets a floor for the fees each acquirer charges to merchants. Acquires also know precisely that all of their competitors pay the very same fees. The price floor and the transparency of it to all suppliers involved (that is to say the knowledge of each acquirer about the commonality of the MIF for all other acquirers in the MasterCard scheme) eliminate an element of uncertainty.

459. In the absence of MasterCard's MIF, the price acquirers charge to merchants would not take into account the artificial cost base of the MIF and would only be set taking into account the acquiere's individual marginal cost and his mark up.

460. Statements of retailers demonstrate that they would be in a position to exert that pressure if acquirers were not able to refer to interchange fee as the "starting point" (that is to say, as the floor) for negotiating the MSC (516). This is because without a default that fixes an interchange fee rate in the absence of a bilateral agreement, merchants could shop around to contract with the acquirer who incurs the lowest interchange costs (517). Acquirers who bilaterally agree to pay relatively high interchange fees to issuers would ultimately not remain competitive, as other acquirers could undercut their merchant fees by refusing lower interchange fees which competitors bilaterally agree to pay to issuers would exercise a constraint on acquirers. In the long run this process can be value of the payment that is without deducting any interchange fee rate in the absence of bilateral negotiations prevents this competitive process. In the absence of such a rule (and in the presence of a prohibition of ex post pricing) acquiring banks would eventually end up setting their MSCs merely by taking into account their own marginal cost plus a certain mark up.

7.2.3.2 Effects of MasterCard MIF in the issuing markets

461. For issuing banks interchange fees are revenue which they collect irrespectively of their individual cost structure. In the MasterCard scheme [business secret - 4] member banks in the EEA issue cards. Acquiring banks are typically also issuers for historic reasons (518).

462. For each issuing there is a commercial rationale to promote cards carrying a MIF above cards without a MIF and premium cards carrying relatively high MIF above normal cards carrying a relatively low MIF. Even if the scheme grew more slowly due to reduced merchant acceptance the individual issuer would continue to promote higher MIF cards above low MIF cards as long as it increases its individual profits. The structure of MasterCard's current Intra-EEA fallback interchange fee in the credit and charge cards segment displays 27 different rates. Interchange fees for commercial cards, for instance, are higher than those for consumer cards. Therefore, banks have an incentive to promote MasterCard branded corporate cards above other MasterCard branded consumer cards.

463. According to IATA airlines have indeed observed a steady shift from consumer credit cards to more expensive corporate credit cards. Issuing banks would promote corporate card usage though cash back programs (up to 1% cash portfolios from consumer to corporate cards in the United Kingdom.

464. MasterCard's view (519) that "interchange revenues only increase if merchant acceptance does not decline" does not contradict that the individual issuer will push those cards that yield the highest interchange proceeds. There is no reason to assume that the individual issuer would stop pursuing this profit maximizing strategy if its behavior had a negative impact for all banks in a scheme due to reduced payment card acceptance on the merchant side.

466. In conclusion the effects of MasterCard's MIF in the issuing markets may contribute to the restrictive effects of the MIF in countries where banks "migrate" cards carrying no or a relative low MIF to cards carrying a higher MIF, thereby increasing the cost of card acceptance in the acquiring market.

7.2.4 Inter-system competition increases anti-competitive effects in down-stream markets

7.2.4.1 Upward pressures on interchange fees

467. MasterCard believes that the competitive process and the market forces will best ensure that the average MIF is close to an "optimum" (520). However, as set out in section 3.1.10.3 the forces of inter-system competition do not sufficiently constrain the level of interchange fees in the MasterCard scheme and even exert an upward pressure. The evidence at the end of the 1990s when MasterCard raised its interchange fee rates to the level of Visa's interchange fees is consistent with this observation. The upward pressure effect of inter-system competition on interchange fee rates is due to the fact that issuing banks are attracted by revenues from a MIF and a card scheme operating with a MIF will take this into account in its competitive behaviour towards other schemes (with or without a MIF). MasterCard does not contest that its methodology for setting interchange fee rates takes account of the rates of competing schemes.

468. The upward pressures on interchange fees which are caused by inter-system competition further amplify and aggravate the distortions and restrictions of competition in the acquiring markets.

7.24.2 The effect of inter-system competition between MasterCard and Visa

469. In its Visa II decision, the Commission expressed the concern that competition between Visa and MasterCard creates upward pressure on the level of the MIF, since most banks are members of both Visa and MasterCard and therefore are likely to issue whichever of the two brands of card has the higher interchange level and brings them most revenue (521).

470. As set out in section 3.1.9.3, there is empirical evidence that MasterCard's Board has repeatedly cross-border interchange fee rates to align its rates with those of Visa.

7.2.4.3 The effect of inter-system competition between MasterCard and domestic payment card systems

471. Furthermore, the Commission is concerned that in the SEPA process (522) MasterCard's replacement of domestic card schemes with Maestro cards will lead to the automatic extension of interchange pricing to Member States where so far no (or very low) interchange fees are being used by the dominant card scheme. Competition between MasterCard and domestic debit card networks within EEA Member States leads to higher interchange fees.

472. MasterCard intends to [business secret - 5b] (523) According to MasterCard banks in several EU Member States are in the process of evaluating a migration (524).

473. MasterCard has already displaced domestic debit card schemes with its Maestro debit card in several European countries.

474. Austria was the first country in Europe to announce the migration of its domestic debit card to the Maestro scheme. As of 1999 around 600 000 domestic Bankomat debit cards were re-issued as Maestro cards (525).

475. While Switzerland is not part of the EEA, the developments in this country confirm the general trend observed in the EEA. In 2003 the local debit card ("ec Karte") migrated to Maestro and as a consequence a new pricing model was announced with price increases ranging between 100% and 400% (526).

476. In other EEA Member States banks have announced to phase out existing domestic cards and to replace them at some point in time with a MasterCard product.

477. In Germany, both MasterCard and Visa started lobbying German banks to migrate their ec-Karte debit card scheme to their networks by offering the possibility of gaining "higher revenues". Two years later, in December 2006, German savings banks announced their plan to replace the domestic debit card with the Maestro debit card (527).

478. Similarly, on 16 May 2006 newspapers announce that in Belgium all domestic debit cards (Bancontact/Mr. Cash) will be replaced by Maestro on 1 January 2008 (528). On 30 March 2007, however, the Belgian Financial Sector Association Febelfin announced that this migration was revoked due to a lack of support by stakeholders.

479. In the above mentioned cases (recitals 474 to 477) the migration from domestic debit cards to Maestro was done through a re-branding of existing payment cards. In other instances banks achieve the same effect gradually by promoting cards that carry a MIF to the detriment of cards without a MIF. Issuing banks can, for instance, start co-branding international card logos on domestic payment cards that carry no MIF (so-called "combined cards") and then offer their customers an incentive (e-g a free funding period) for choosing at the point of sales the card functionality that carries a MIF above the functionality that carries no MIF.

480. The latter approach appears to be practiced in Finland (529) where the domestic debit card Pankkikortti is increasingly losing market share to MasterCard and [four party payment card scheme]. In Finland MasterCard has set a [business secret - 2a] interchange fee (530) while a transaction with the domestic debit card yields no interchange fees at all (531). Thus, there is a monetary incentive for each issuing bank to convince their customers to swap a domestic debit card for a MasterCard card. A Finnish banker from the domestic debit card to MasterCard's and [four party payment card scheme's] cards (532). Statistics of the Finnish Banker's Association (533) also indicate that the banks are increasingly successful in promoting MasterCard's (and [four party payment card scheme's]) payment cards above the domestic debit card (534).

481. MasterCard raises essentially two other possible reasons that might explain the ongoing migration of domestic cards to Maestro that are unrelated to interchange (535). First, MasterCard contends that most domestic schemes were "proprietary schemes" owned by the domestic banks and therefore could have an interchange fee mechanism in place if they wished, provided this was allowed by national regulators. The MasterCard MIF was not in itself an incentive for banks to migrate from their domestic schemes to MasterCard products. Second, replacing domestic cards with an international logo was a manner of "outsourcing" SEPA related investments to the international schemes.

482. MasterCard's first observation that domestic banks could install an interchange fee mechanism without migrating their domestic payment card portfolio to Maestro is technically correct. However, for reasons related to competition law where banks have issued cards in domestic schemes that successfully operated without a MIF so far, they must prove that from a given moment onwards the scheme could no longer operate without this mechanism. Both in Germany and in the Netherlands, for instance, local banks have tested this avenue with local competition authorities, decades after schemes started to operate. In both instance no exemption was granted due to an apparent lack of the restrictions' indispensability (536). The possibility of "outsourcing" the antitrust sensitive setting of a MIF to some manager in the MasterCard payment organization may be perceived as opportunity for banks to rid themselves off the litigation and antitrust enforcement risks of pricing card acceptance services through a MIF (537). MasterCard management praises its capacity of fixing interchange fees "on behalf of" local banks in proposals to replace domestic debit cards with Maestro. Contrary to MasterCard's assertion (538) this is for instance clearly evidenced by MasterCard's proposals to (539) [business secret - 6]. For all these reasons the Commission therefore rejects MasterCard's first argument.

483. As to MasterCard's second reason for migrations, the Commission does not dispute that the MasterCard MIF may not be the only reason for a bank to migrate its domestic debit cards to Maestro. For instance, MasterCard' management also offers [incentives] (540). However, this does not mean that the MIF would not play an important role in the migration process (541) [business secret - 6].

484. While the SEPA project certainly acts to some extent as catalyst for the migration of domestic schemes to Maestro (542), MasterCard neglects that according to the EPC cards Framework banks have more than one possibility to render their cards "SEPA compliant". The replacement of domestic cards with a MasterCard or Visa card is not the only option. Rather, the EPC also offered banks the possibility to co-brand their cards with an international card or to join an alliance of domestic card schemes. An issuing bank's choice of how to render its domestic payment card portfolio "SEPA compliant" is influenced amongst others by the MIF, as the European Central Bank pointed out (543).

485. In conclusion, the MasterCard MIF fuels to a certain extent, in conjunction with other factors, the ongoing migration of domestic debit cards to Maestro. The Commission's competitive assessment of the MasterCard MIF and its impact on merchants fees must therefore take account years and that the economic importance of MasterCard's Intra-EEA fallback interchange fees (544) and possibly the SEPA fallback interchange fees will in particular expand to the 13 Member States in the Eurozone where local schemes applied no MIF or a lower MIF (545).

486. MasterCard's claim that the Commission's concern with respect to scheme migrations was "attacking rather than defending competitive processes (546) therefore misses the point. The Commission's concern is not that one scheme is taken over by another one. The Commission's concern is that MasterCard's MIF has become an (artificial) element of inter-system competition that risks inflating merchant fees in countries where the MasterCard scheme replaces card schemes without a MIF. MasterCard has taken no measures to prevent this detrimental side-effect of a migration and even promotes migrations with references to its MIF.

7.2.4.4 Evolution of MasterCard's Intra-EEA fallback interchange fees

487. Finally, the analysis of MasterCard's Intra-EEA fallback interchange fee rates empirically confirms that inter-system competition does not sufficiently constrain MasterCard in maintaining a high level of interchange fees.

a) Intra-EEA fallback Interchange fees for MasterCard branded cards

488. In 2000 MasterCard raised the base rate for consumer credit and charge cards from and in 2001 from. In 2001 MasterCard increased the enhanced electronic rate from MasterCard also increased in 2000 from and a second time in 2001 from. The enhanced electronic rate was raised in 2001 from [business secret - 2a]. The fee increases concerned the majority of all transaction volumes with MasterCard branded payment cards. Transactions subject to the enhanced electronic MIF amount to approximately [business secret - 2a]% of total credit card cross-border transaction volumes and the base rate amounted to 26 [business secret - 2a]% (547). Only interchange fees for business purchasing cards were lowered in 1999 and 2000. However, this category only accounts for a handful of cards, roughly [business secret - 4] cards (548) out of approximately [business secret - 4] MasterCard branded payment cards (549).

489. MasterCard acknowledges that the weighted average cross-border interchange fee of MasterCard branded cards increased from in 1996 to [business secret - 2a] in 2003 (550). This development occurred despite an increase in transaction volumes with MasterCard's payment cards in Europe which should have led to reduced costs due to scale gains in processing transactions. As set out in section 7.4 aa) total sales volumes with MasterCard branded cards in the European Union have steadily increased during the last years (551).

b) Intra-EEA fallback interchange fees for Maestro branded debit cards

490. When MasterCard created the Maestro brand in 1992, the interchange fee was set at [business secret - 2a] or on-line PIN based transactions and at [business secret - 2a] or on-line signature based transactions (552). The gradual replacement of signature based debit cards with chip and PIN cards should therefore have lead to a drop of the average interchange fee if everything else remained constant. MasterCard's European Board, however, raised the nominal rate in 1999 and 2000. At the end of 2006 the chip and PIN rate was at 0.55% (553) and the signature based rate was at 0.80% (554).

491. Contrary to MasterCard's allegation (555), the slight decrease in the weighted average interchange fee in the Maestro debit card segment between 2000 and 2006 does not inter-system competition. The evolution of the weighted average is not the most reliable measure to assess the effect of inter-system competition on MasterCard's MIF because MasterCard sets many different levels of interchange depending on the card type and the manner of authorization. The weighted average is therefore influenced by modifications in the cards portfolios of MasterCard issuers, for instance a gradual replacement of signature based cards (high MIF) with chip cards (low MIF). For assessing the upward pressures on the MIF it is rather of importance to observe why the nominal rates of a MIF are changed. In this respect the European Board's decision [business secret - 2a] was explicitly justified by an improvement of MasterCard's prospective competitive position as compared to Visa (556).

7.2.4.5 MasterCard arguments in reply to Supplementary SO

492. MasterCard argues that what the Commission described as "upward pressure" on interchange fees was "simply competition" between MasterCard and Visa. The Commission also only partially assessed competition, between payment schemes, as it focused solely "on one constituent element of such competition", that is to say interchange (557).

493. The Commission cannot see how these arguments support MasterCard's claim that the competitive process and market forces would best ensure that the average interchange fee will be close to an "optimum level" and that an interchange fee set at "sub-optimal level" would "not be sustainable" (558).

494. In the Commission's view are two forces that could in theory impact on the MIF, one on the acquiring side. On the acquiring side, it is shown blow (section 7.2.5) that the countervailing power of acquirers and merchants is insufficient to constrain the MIF. On the issuing side it is clear that the interchange fee constitutes revenues which drive issuers to promote cards carrying a MIF above those without or with a lower MIF. This effect of the MIF in the issuing markets thereby becomes an artificial parameter of inter-system competition that creates upward pressures on the level of interchange fees.

495. It is a fact that MasterCard has repeatedly aligned its Intra-EEA fallback interchange fees with cross-border interchange fees of the Visa payment association (section 3.1.10.3). Whether the MIF is only one constituent of inter-system competition besides others is not of relevance. The relevant issue is whether market forces can sufficiently constrain a scheme and its members in collectively setting the level of interchange fees and whether as a consequence there is no necessity for the Commission to intervene. Considering that (i) most of the [business secret - 4] member banks of MasterCard are also members of Visa, (ii) MasterCard repeatedly set its Intra-EEA fallback interchange fees in parallel with Visa (559), (iii) merchants often pay a "blended" MSC for accepting Visa and MasterCard card products and (IV) the hidden cost of interchange fees is not visible either for consumers or for (the large majority of) merchants, MasterCard's contention is entirely implausible.

496. The development of MasterCard's Intra-EEA fallback interchange fees after the Visa II decision corroborates this conclusion. The average level of MasterCard's fees for credit cards remained stable at around [business secret - 2a] since 2002 while Visa gradually reduced its cross-border interchange fees under the terms of the Visa II decision an average of [business secret - 2a] in 2007. Contrary to its contentions MasterCard was well in a position to maintain interchange fees without facing sanctions from the market.

7.2.5 MIF not subject to constraints from acquirers or merchants

497. MasterCard takes the view that "it is in the interest of MasterCard system, its members, its cardholders and merchants [and] competition in the payments market to set a MIF at an optimal output optimizing level". MasterCard believes that "overall, the competitive process and the market forces will best ensure that the average MIF will be close to such optimum over a relevant period of time" (560). In the same vein MasterCard's expert states that a "profit maximising IF is the volume maximizing IF" in case the pass-through on both sides is symmetric so it would be in the own interest of MasterCard to maximize system output in order to maximize profits. As maximizing system output equals maximizing competition, the "government" would not need to intervene (561).

498. However, the Commission's market research indicates that the market forces do not sufficiently constrain the MasterCard MIF.

7.2.5.1 Insufficient constraints from acquirers

499. As MasterCard acknowledges (562), the setting of interchange fee rates is not akin to a contentious process such as a price negotiation where opposing interests of buyers and sellers meet. Rather, all banks eventually share a common interest that merchants pay a higher price than they would in a fully competitive environment. A MIF allows a payment organization to raise the marginal cost of all acquirers alike in a collective manner which enables the acquirers to set a higher price for merchants. The additional revenues are then transferred to the card issuing bank. For banks without any issuing business the MIF also generates revenues. For acquirers without any issuing business the MIF also generates revenues even though indirectly. Acquirers expect that some issuers "invest" a portion of their revenues from the MIF in promoting further card usage which in turn may lead to some increase of transaction volumes ay merchant outlets. Hence, both issuers and acquirers may commercially benefit from a MIF, be it directly or indirectly.

500. Minutes of meeting of the European Board and the BMAC confirm that acquirers did not try to oppose increases of MasterCard's Intra-EEA fallback interchange fee rates in the period from 1999 to 2006. The decision to increase the rates was planned by MasterCard's management and subsequently endorsed by the banks' delegates in the BMAC and the European Board (563).

501. Contrary to MasterCard's argument (564), the fact that MasterCard management rather than a board of bank delegates has set the level of Intra-EEA fallback interchange fees since 14 September 2006 will de facto not change the situation. As the developments preceding and following the IPO of MasterCard Incorporated indicate (565), management will consider the commercial interests of member banks in setting new interchange fee rates. Both issuing and acquiring banks have a continuing commonality of interests in the MIF after the IPO.

7.2.5.2 No sufficient constraints from merchants

a) Analysis of merchant's demand elasticity

502. First, and contrary to MasterCard's assertion (566), merchants' countervailing power has not been considered by the BMAC and the European Board when they decided to raise Intra-EEA fallback interchange fees in past years. Reference is made only as far as it tries to estimate their extreme limit of endurance rather than the merchant's benefits. MasterCard assesses the merchants' maximum willingness to pay by estimating their opportunity cost of issuing store credit cards (see section 3.1.8.3 a).

503. Second, even very large merchants are typically unable to negotiate a merchant fee which is below interchange fee costs. Merchants view the costs of MasterCard's interchange fees as the non-negotiable part of the merchant fee, a price element that is "not subject to competitive market forces (567).

504. Third, paradoxically, a MIF may at the same time reduce merchants' resistance to an increase of the MSC as inflates the MSC. This is because to the extent that a MIF is used by car issuing banks for promoting further card usage merchants will displease an increasing number of cardholders if they were to cease to accept the payment card in question (568).

505. Fourth, merchants may pass the cost of the interchange fee on to their customers by raising the final price. When faced with in interchange fees and consequently an increase in merchant fees, recovering that incremental cost through a smaller fall in turnover than ceasing to accept MasterCard cards (569). This fact further reduces the likelihood that merchants will constrain an increase of the MIF.

506. Fifth, in deciding whether or not to accept a particular payment card a merchant not only takes account of "transactional befits" (such as speed of payment, security, convenience etc) but in particular also the potential impact on its business volumes. Payment cards are not like other products. Their intrinsic feature is to allow for the transfer of funds to a merchant and the merchant will balance the profits it can potentially yield on the entire sale with the costs of accepting the card, if that sale would otherwise no longer take place. Only a minority of the 218 merchants surveyed by the Commission indicated that they accept cards because they are faster and more secure than cash and cheques (570).Merchants accept cards in the first place to respond to consumer preferences and because their competitors also accept cards (571). Once merchants accept cards they are hesitant to stop accepting them again (572). There are of course merchants who do not accept payment cards due to the costs involved. This is, however, no reason why the MasterCard MIF would fall outside the scope of Article 81(1) of the Treaty (573).

b) Assessment of MIF in context with HACR

507. Finally, in assessing the capacity of merchants to constrain the MIF the Commission also considers other network rules.

508. In the MasterCard scheme the so-called Honour-All-Cards Rule (HACR) obliges merchants and acquirers to accept all valid MasterCard cards

509. The "honour-all products" functionality reinforces the restrictive effects of the MasterCard MIF on price competition between acquiring banks. MasterCard applies significantly higher interchange fees for some cards (commercial credit cards in particular) than for others (for example, consumer credit cards). By obliging merchants to accept all card products (including commercial credit cards) if they wish to accept MasterCard cards, the HACR enables MasterCard's member banks to exert collective market power through the MIF by allowing issuing banks to introduce new card products in the market while at the same time pre-determining their price through the MIF for merchants who are bound to accept those cards. Merchants cannot prevent this by specifically refusing such new card products but merely have the choice of refusing MasterCard branded cards altogether. The "honour all products functionality" of MasterCard's HACR therefore further decreases the countervailing buyer power of merchants in the presence of a MIF.

c) Assessment of MIF in context with NDR

510. MasterCard used to operate with a "No Discrimination Rule" (or "No Surcharge Rule") until 1 January 2005. That rule used to prohibit merchants from passing on the costs of accepting MasterCard and Maestro cards in the form of a surcharge fee.

511. MasterCard's decision to abolish the No Discrimination Rule is unlikely to change merchants' capacity to constrain the collective exercise of market power of MasterCard's member through the MIF for the following reasons.

512. First, there is no indication that surcharging has actually been implemented to an appreciable extent since 2005 nor that acquirers, in their contracts with merchants, meanwhile deleted the prohibition to surcharge for the MasterCard brands (574). Furthermore, Visa maintains that rule until today.

513. Second, the Commission market surveys in the past showed that surcharging does not take place to a significant extent even in countries where surcharging has been allowed in the past (575). The Commission's merchant market survey has confirmed that merchants do not surcharge for card transactions, save for a limited number of merchants in the travel industry, particularly those that use internet sales methods (576).

514. Third, while MasterCard abolished the NDR as network rule as of 1 January 2005, acquirers are not forced to change existing contracts with merchants containing the NDR as a contractual obligation of the merchants. The clause will at best only disappear gradually with the renewal of existing merchant contracts (577).

515. MasterCard submitted a "model" on surcharging and argues that "under the plausible assumption that many of the administrative costs of price differentiation are likely to be one-off", the threat of surcharging or of using other forms of discouragement exercises a strong competitive constraint on MSCs and thus the MIF" (578). This is because even if a card scheme, having triggered widespread discouragement, were subsequently to revert to its earlier pricing, merchants would in all likelihood continue to discriminate according to the method of payment used. Also, "one would expect that discouragement, once triggered, will not be confined to individual merchants, but become widespread" (579).

516. The Commission has analysed that argument in detail in Annex 4 and considers it to be unconvincing in several respects. Moreover, the theoretic considerations set out in the model ignore the market reality:

517. First, in practice merchants often pay one" blended" fee for accepting both MasterCard and Visa payment cards (580). Thus, a merchant simply does not know which scheme to "sanction" by surcharging (or other methods of discouragement) if the merchant's bank announces an increase of the blended merchant fee.

518. Second, the administrative costs of surcharging are considerable, as MasterCard seems to acknowledge (581).

519. Third, as long as Visa prohibits surcharging, it remains difficult for merchants to surcharge for MasterCard, only. They would risk displeasing customers that carry MasterCard once they realize that customers using Visa do not pay the surcharge fee. Good customer service is however, one key reason for merchants to accept cards in the first place (582).

520. Fourth, Intra-EEA fallback interchange fees apply for all 30 Member States in the EEA. Thus, if a merchant wanted to exercise countervailing power onto the MasterCard payment organization for raising these fees, its surcharging would need to gain such importance that it affected the payment organisation on pan-European scale.

521. Finally, and contrary to the views expressed by MasterCard in the proceedings, MasterCard's management seemed very well aware that surcharging was not an option for merchants at the time the rule was abolished. When MasterCard abolished the rule in 2005, the then [senior manager] of MasterCard Europe Sprl, gave an interview where he expressly stated that "the pressure on merchants" to refrain from surcharging is "simply too strong" (583). In an open letter to its member banks the MasterCard management made another similar statement with references to the fuel retail sector (584). At least at this point of time MasterCard management believed that apart from taxi drivers and travel agents, other merchants would not impose a surcharge to cover the costs of accepting payment cards to consumers even once they were allowed to do so (585). Statements of Hotrec at the oral hearing again confirmed the difficulty for merchants to sanction price increases in the travel and entertainment sector (586).

7.2.6 MasterCard members exert market power through the MIF

522. The [business secret 4] members of the MasterCard payment organization collectively have market power vis-à-vis merchants and their customers. The MIF allows them to exploit this collective market power by effectively putting a floor under the MSC charged to merchants. In the presence of a MIF the marginal cost of acquirers are inflated, thereby setting a floor under the merchant fee (587).

523. Contrary to the view of MasterCard (588) the MIF has continued to be a collective exercise of market powers to set the level of the cross border interchange fees were transferred from the European Board to the Chief Operating Officer of MasterCard Incorporated. As set out in section 2.1.4, MasterCard management sets the fees in the interest of the member banks and the members are bound by those decisions under the organisation's network rules which they approved by consenting to the IPO of MasterCard Incorporated and the related changes in the governance of the payment organisation.

7.3 The MIF does not fall outside the scope of Article 81(1) of the Treaty

7.3.1 The framework for the legal assessment

524. According to the Court of First Instance of the European Communities if it is established that a restriction is directly to related and necessary for achieving a main operation, the compatibility of that restriction with the competition rules must be examined with that of the main operation. Thus, if the main operation does not fall within the scope of the prohibition laid down in Article 81(1) of the Treaty, the same holds true for the restriction directly related to and necessary for the operation. If, on the other hand, the main operation is a restriction within the meaning of Article 81(1) of the treaty but benefits from an exemption under Article 81(3) of the Treaty, that exemption also covers those ancillary restrictions (589).

525. In Pronuptia the Court of Justice held that franchises do not in themselves restrict competition. Restrictions that are essential to avoid the risk of the franchisor's know-how and assistance from benefiting competitors fall outside Article 81(1) of the Treaty (590). That case also illustrates clearly the narrow scope of the ancillary restraint doctrine and the distinction between restrictions that are necessary for the implementation of an agreement and those which are desirable in terms of commercial success. The Court noted that restrictive clauses could infringe Article 81(1) of the Treaty even if a prospective franchise would not take the risk of becoming part of the chain, investing his own money, paying a relatively high entry fee and undertaking to pay a substantial annual royalty, unless he could hope, thanks to a degree of protection against competition on the part of the franchisor and other franchises, that business would be profitable (591). In other words, restrictive clauses desirable with a view to the commercial success have to be assessed within the context of Article 81(3) of the Treaty. If a restraint is merely needed to render the main operation profitable for (some of) the parties involved, it is not for that very reason an "ancillary restraint" that falls outside the scope of Article 81(1) of the Treaty.

526. The objective necessity test of the ancillary restraint doctrine does not involve an assessment of whether the main operation would be commercially less successful in the absence of the restraint, as set out by the Court of First Instance in Metropole (592). It is clear from that judgment that such and other efficiency considerations are to be assessed within the legal framework of Article 81(3) of the Treaty: (593) "it is only in the precise framework of that provision [Art 81(3) EC] that the pro and anti-competitive aspects of a restriction may be weighed (...). Article 85(3) of the Treaty would lose much of its effectiveness if such an examination had to be carried out already under Article 85 (1) of the Treaty (...)."

527. The Gottrup-Klim (594) and the Wouters judgments (595) further corroborate the principle that restrictive clauses desirable with a view to the commercial success of the main operation but not necessary for its viability must be assessed within the context of Article 81(3) of the Treaty.

528. In Gottrup-Klim the Court of Justice ruled that a provision in the statutes of a co-operative agricultural forbidding its members to participate in other forms of organized co-operation that are in direct competition with it, is not caught by Article 81(1) of the Treaty as long as the provision is restricted to what is necessary to ensure that the co-operative functions properly and maintains its contractual power in relation to produces (596). The reason why the prohibition of dual membership was deemed no to restrict competition was a peculiarity of the market in question: prices varied according to the volume of orders and producers were deemed to have market power (597). Since dual membership could have jeopardized the countervailing market power of the agricultural purchasing association, the association's decision was considered to fall outside Art 81(1) EC (598).

529. In Wouters (599) the Court of Justice qualified a decision of the Dutch Bar to prohibit multi-disciplinary partnerships between attorneys and accountants as not infringing Art 81(1) EC. The Bar was charged by law with adopting deontological rules that assured the correct exercise of the attorneys' profession (600). The Bar's decision to ban multidisciplinary partnership was taken in the context of this legal task and in the context with other measures of the Bar obliging lawyers to act for clients in complete independence and in their sole interest and to observe strict professional secrecy. Considering that there may indeed be a degree of incompatibility between the advisory activities of an accountant, the Court qualified the prohibition of multi-disciplinary partnerships as necessary in order to ensure the proper practice of the legal profession (601) and that the restrictions of competition did not go beyond what was necessary to achieve that aim (602).

530. In none of these judgments did the Court consider possible efficiencies of the restrictive agreements under Article 81(1) of the Treaty: the decisive question in Wouters was whether the prohibition of multidisciplinary partnerships was necessary for the proper practice of the legal profession. Whether or not, for example, Dutch attorneys could provide more efficient advice to companies if one entity could provide more efficient advice to companies if one entity could provide legal and accounting services at the same time was not an issue. Likewise, in Gottrup-Klim the agricultural association's capacity to concentrate the buying power of individual farmers was the very reason for its existence (603) and might have been endangered without the restrictive membership rule. Thus, the rule was deemed necessary for the very existence of the association, not simply for a more efficient operation.

531. Finally, in Métropole the Court (604) expressly required that the anti-competitive clause in the joint venture agreement 'exclusively) is directly related to and necessary for the "main operation" in order to fall outside Article 81(1) of the Treaty. The notion of "main operation" in turn is used by the Court to refer to the establishment, that is to say the creation, of the joint venture TPS (605). In this case, "main operation" must be understood as the operation of an open payment card system such as MasterCard's.

532. For those reasons and based on the Commission consistent case practice the following questions will be examined to see whether MasterCard's MIF falls outside the scope of Article 81(1) of the Treaty:

i. First, could banks co-operate in an open payment card system without the MIF?

ii. Second, if a MIF were indeed necessary for such co-operation, are less restrictive means available than the MasterCard MIF that would allow the banks to co-operate, as well?

533. In analyzing these questions account will be taken of the arguments put forward by MasterCard concerning the objective necessity of its MIF

7.3.2 MasterCard's arguments

534. MasterCard believes that the Commission should apply a two-stage analysis to investigate the objective necessity of its MIF. First, the Commission should ascertain whether there is an alternative fallback mechanism to a MIF that is "feasible" (606). Second, the Commission should also evaluate whether that alternative mechanism, "leads to a better outcome (in terms of level of MIF, system output, etc) than the MasterCard default MIF" (607).

535. MasterCard also argues that the objective necessity test consists in assessing whether, without the restriction, the scheme would be "less capable" of pursuing its objectives for the benefit of its members. This is a test of "relative difficulty" that is to say of comparing the "relative difficulty that would be encountered in the operation of the scheme with and without the alleged restriction" (608). The correct legal test would therefore be whether the MasterCard scheme, with an HACR, could operate "with no less difficulty" without a default interchange fee (609). An operation which is "less efficient" is also "less certain". Thus, if in the absence of the MasterCard MIF the scheme would be "less efficient", that MIF would not be a restriction of competition within the meaning of Article 81 of the Treaty (610).

536. On the substance, MasterCard argues that a MIF is necessary in its scheme due to the existence of the HACR (611). In the presence of such a rule acquiring banks cannot refuse to accept an issuer's cards what-ever the interchange fee the issuer deducts. Based on an option of its expert (612) MasterCard concludes that due to the HACR there must be some kind of arrangement which sets out the terms and conditions under which issuers and acquirers agree to provide payment services to cardholders and merchants (613). Without the MIF acquiring banks would be "at the mercy of issuers", because issuers would be in a position to deduct excessive interchange fees from the funds they transfer to acquiring banks. Thus, it would also be impossible to sustain the HACR (614).

537. According to MasterCard there would be no less restrictive alternative to a collectively set default interchange fee and bilateral negotiations between all banks in its scheme would be impossible in view of the high number of banks involved, the transactional costs of negotiations and the risk of a break down of bilateral negotiations between issuers and acquirers (615). If the default rule were to leave issuers the freedom to determine the level of interchange fees in the absence of an agreement with the acquirer, this would result in excessive (616) fees and acquirers would exit the scheme. If in contrast, acquirers were to set the fees by default, the fees would tend towards zero (617) and issuers would derive insufficient revenues to cover their "fixed costs" (618). Larger issuing banks in the scheme would leave and start alternative payment card systems either on their own or together with other banks (619). MasterCard does not contest that the default interchange fee could be replaced through arbitration (620). However, in practice the level of interchange fees as before (621).

538. MasterCard also argues that if a MIF was replaced with a rule that orders issuing banks and acquiring banks not to charge any interchange fees in the absence of a bilateral agreement, the scheme would have to be overhauled "in a number of important respects" in order to remain competitive and, even with these, "main types of default financial terms of dealing" that would have to be modified in the absence of the MIF. These are the circumstances in which payments are to be made or reversed (that is to say, chargeback rules), the amount of the payment (that is to say, settlement delays) (623).

539. If MasterCard's scheme used such a fallback rule that imposed "at par" clearing instead of a certain interchange fee rate, this would amount to just as much of a restriction of competition as the current rule (624). Moreover, the change of other network rules would "in all likelihood" lead to the "demise" of the scheme. However, this "was not to say the demise of its business", as MasterCard could operate a (smaller) three party payment card scheme" (625).

540. MasterCard finally also refers to US case law in the "NaBanco case" of 1984 to corroborate its view that the MasterCard MIF is objectively necessary (626).

7.3.3 The Commission's assessment of the burden of proof

541. The Commission agrees in principle with MasterCard on the first step of the proposed analysis, that is to say to analyse whether the MasterCard scheme would be "feasible" without a MIF and whether therefore the MIF is a precondition for the provision of services by MasterCard's acquiring banks to merchants.

542. However, the Commission refutes the second step of the proposed analysis, that is to say whether alternatives to a default interchange fee lead to a better system output. This would blur the boundaries of Article 81(1) and 81(3) of the Treaty. MasterCard essentially requests the Commission to compare the current the potential efficiencies of a MIF, namely a higher system output. This would effectively reverse the burden of proof under Article 81(3) of the Treaty. In the Commission's constant case practice (627) and under settled case law, the framework of considering possible efficiencies is not Article 81(1) of the Treaty, but Article 81(3) of the Treaty (628). This is now also clearly set out in Article 2 of Regulation (EC) No 1-2003.

543. MasterCard (629) misreads the Gottrup-Klim judgment (630) by isolating paragraph 33 from the context. Rather, paragraphs 33 to 35 must be read together. In paragraph 33 the Court acknowledges that the removal of the restriction could make the association less capable of pursuing its objectives. However, the Court goes on to say that, even though removing the restriction could make the scheme less capable of achieving its objectives, the restriction could nonetheless have adverse effects on competition. The Court therefore concludes that the restriction will only escape Art 81(1) if it is: "limited to what is necessary to ensure that the cooperative functions properly". Gottrup-Klim therefore does not introduce a test along the lines suggested by MasterCard. It is simply the objective necessity test, which is another way of asking whether the co-operation is viable in the absence of the restriction.

544. In Metropole the Court of First Instance applies an objective necessity test. It is true that the Court of First Instance says that if removal of the ancillary restriction makes the main operation "difficult" to implement, it "may" be regarded as objectively necessary. However, MasterCard (631) quotes the citation out of context. The CFI makes the comment in the course of rejecting the commercial success of the main operation". The CFI goes on to say that the restriction must not only be objectively necessary to implement the main operation but must not go beyond what is necessary in terms of scope, i.e must be proportionate.

545. The importance of the Metropole judgment for the case at hand is in particular that the Court of First Instance excluded the analysis of the commercial success of a co-operation from the ambit of the objective necessity test. Thus, contrary to the argument of MasterCard, it is not relevant whether in the absence of the MasterCard MIF the scheme would have less commercial success (632).

546. MasterCard (633) also does not cite the relevant paragraphs from the recent 02 judgment (634) (paragraphs 74-79) where the Court of First Instance criticizes the Commission for failing to consider whether the potentially restrictive agreement was "indispensable/necessary" in order for 02 to enter the 3G market. 02 is therefore another application of a pure objective necessity test, not a test of "unchanged difficulty" or even of "less efficiency".

547. MasterCard's test of "unchanged difficulty" must therefore be rejected. The applicable test is whether the MIF is objectively necessary for the implementation of the main operation, which is the co-operation of banks in an open payment card scheme.

548. As already set out in the Visa II decision with respect to the Visa MIF (635) the only provision necessary for the operation of an open payment card system, apart from the technical arrangements on message formats and the like, are the obligation on the creditor bank to accept any payment validity entered into the system by a debtor bank and a prohibition on ex post pricing by one bank to the other. A mechanism that shifts costs and revenues between issuing and acquiring banks is not necessary for the banks' co-operation as issuing and acquiring services can be remunerated directly by the respective consumer groups. Thus, the free play of market forces will then determine the revenues banks can obtain from consumers and businesses.

7.3.4 The Commission assessment of the objective necessity

549. Banks are able to co-operate in an open payment card scheme as MasterCard's if the scheme operates without a rule that sets a certain interchange fee rate in the absence of bilateral agreements between an issuing and acquiring bank. As a matter of economic principle there is no reason why this should be unfeasible.

7.3.4.1 The MasterCard scheme in the absence of a MIF

550. A payment card scheme such as MasterCard's (see section 2.2) is characterized by a large number of cardholders and merchants that accept the scheme's cards as a means of POS payments. Cardholders obtain a payment card from issuing banks and merchants are able to accept card payments thanks to the card acceptance services that the acquiring bank supplies to them. Both issuing and acquiring banks are remunerated by their clients. The scheme owner (here: MasterCard Incorporated) obtains fees from issuing and acquiring banks but has no contractual relationship with cardholders and merchants.

551. The absence of a MIF in a payment card scheme such as MasterCard's implies that on each side of the scheme prices result from profit-maximising issuing and acquiring, that is, in concrete terms:

i. Issuing banks charge for the service they provide to them. Issuing banks individually set their prices so to maximize their profits. As there is transparency in that the product (payment card) and the services provided by the issuing bank are known to the cardholder, the cardholder chooses the issuing bank with the best price/quality combination to the detriment of inefficient issuers.

ii. Acquiring banks charge merchants for the services they provide. Similar to what happens on the issuing side, prices are set according to the profit maximizing level of each individual acquirer. Banks decide about the price they charge to merchants without that price being predetermined to a large extent by a MIF.

iii. The scheme owner MasterCard Incorporated charges issuing and acquiring banks the fees (in the MasterCard scheme.

iv. Cardholders obtain from their issuing bank payment cards that are priced in a transparent manner and will compare banks on the merits of their offer.

v. Merchants are able to negotiate the merchant service charge with their acquiring bank. The price each merchant pays is not (largely) pre-determined by a MIF.

552. A MIF is not objectively necessary for the operation of an open payment card scheme such as MasterCard's. In the absence of a MIF the prices banks charge to their customers are established through competition rather than to a large extent through competition rather than to a large extent through a multilateral agreement.

7.3.4.2 Acquirers are not "at the mercy" of issuers without a MIF

553. MasterCard argues that acquiring banks in open payment card systems would be "at the mercy of issuers", because without a MIF that applies by default in the absence of bilaterally agrees interchange fees the scheme's issuing banks would be in a position to deduct unilaterally any interchange fee they wish. Acquiring banks could then not prevent issuing banks from retaining excessive interchange fees as acquirers are bound under MasterCard's HACR to process all card transactions properly presented to them. Based on an opinion of its expert (636) MasterCard concludes that due to the HACR there must be some kind of arrangement which sets out the terms and conditions under which issuers and acquirers agree to provide payment and conditions under which issuers and acquirers agree to provide payment services to cardholders and merchants (637). The MIF is such arrangement and without the MIF it would be impossible to sustain the HACR (638).

554. That argument cannot be accepted. As already set out in the Commission's Visa II decision, the possibility that some issuing banks might hold up acquirers who are bound by the HACR could be solved by a network rule that is less restrictive of competition than MasterCard's current solution that, by default, a certain level of interchange fees applies. The alternative solution would be a rule that imposes a prohibition on ex-post pricing on the banks in the absence of a bilateral agreement between them. The rule would oblige the creditor bank while prohibiting each bank from charging the other bank in the absence of a bilateral agreement on the level of such charges (639). That solution to "protect" acquirers if issuers should indeed abuse their power under an HACR is less restrictive of competition than a MIF as it does not set a minimum price level on either side of the scheme.

7.3.4.3 Other open payment card schemes without a MIF

555. In 2002 the Commission took the view that there were no examples of four party payment card schemes that function without a MIF (640). This finding in the Visa II case was based on the assessment of three payment card schemes which the complainant EuroCommerce has referred to, that is the German ec cash Karte system, the Australian EFTPOS system and the Canadian Interac system. The Commission's research after the Visa decision of 2002 went beyond these examples and established the existence of open card schemes operating without a MIF sufficiently comparable to a, open payment card scheme such as MasterCard's.

556. Several payment card schemes in the European Economic Area have indeed successfully been operating without a MIF for a long time period. These schemes have been established between 1979 and 1992. The number of issuing banks in these ranges from 10 to 344 banks, as set out in the diagram below.

< table > (641), (642), (643), (644), (645)

557. The Commission considers the existence of these open card schemes as evidence that a MIF is not objectively necessary for the co-operation of banks within an open payment card scheme such as MasterCard's. If comparable card schemes that are smaller than MasterCard's are viable without a MIF than, a fortiori, schemes such as MasterCard's with a considerably more important cardholder and merchant base (see section 2.2) must also be viable without a MIF (646). MasterCard's scheme is not in a start up situation where a newcomer has to reach a critical scale on both sides of the scheme during the start up phase.

Scheme structure comparable to MasterCard's

558. The structure of the MasterCard scheme is sufficiently similar to the structure of the five domestic card schemes set out in recital 556. The common feature between MasterCard and the five domestic card schemes is that the activity of card issuing and of merchant acquiring is legally separated (647). MasterCard errs in its assertion that those domestic card schemes were not comparable to MasterCard because some of them were "not truly open" (648). In all of these five schemes different financial institutions can join the system under the same brand (649). MasterCard's specific point seems to be that the five schemes are "not truly" open because in some of them issuing banks engage in acquiring through a common joint venture. This, however, ignores the fact that in MasterCard's own scheme the formal separation of issuing and acquiring activities does not exclude that issuing banks also act as acquiring banks. Industry refers to so-called "on us" payments where the issuer and the acquirer happen to be the same bank. Indeed, in most open card schemes issuing banks also act as acquirers, whether they acquire themselves (650) or through a joint venture that they co-own with other issuing banks. Such a joint venture could be a monopolist (651). According to MasterCard most institutions in the MasterCard scheme act both as issuers and acquirers (652).

559. This has historic reasons. In several markets MasterCard's issuing banks have together set up inter-bank organizations mainly during the 1980's to acquire MasterCard's (and Visa's) payment cards. This was for instance the case in Austria (653), Italy (654), Germany (655), Belgium (656), Finland (657), the Netherlands (658), and Portugal (659). An industry report therefore summarized the situation of payment card acquiring in Europe as follows: "In Europe, local issuers traditionally acquire merchants for card transactions" (660). As OEHV emphasized during the oral hearing, the setup of MasterCard and Visa in Austria, for example, is to a large extent similar to that of a three party scheme because one inter-bank association in each scheme acquires and issues (a large part of the) payment cards on behalf of its shareholders (661). Where issuing banks concentrated acquiring for MasterCard in the hand of a jointly owned inter-bank organization, this organization would hold -often over a long period of time- a monopoly or quasi monopoly position for acquiring one or both of the international brands on that market. These inter-bank organisations of issuing banks still exist but over time had to face competition (662). The specific setup of the MasterCard scheme in countries where domestic card schemes operate without a MIF is discussed below in sections 7.3.4.3.1 to 7.3.4.3.5, country by country.

Domestic and cross-border functionality of cards

560. Contrary to MasterCard's argument (663) the cross-border functionality of MasterCard and Maestro branded payment cards cannot be a decisive difference to assessing the objective necessity of MasterCard's MIF.

561. If the necessity of a MIF was related to the cross-border functionality of a scheme's cards one would expect to see that a MIF is set for cross-border transactions, only. This is, however, not the case. MasterCard's MIF is not limited to cross-border payments but applies also to domestic transactions (664). Moreover, there is no reason why it should be possible to operate without a MIF in domestic systems but impossible in systems that also switch cross-border and domestic payments is very similar as both cross-border and domestic payment card transactions are cleared centrally in a clearing house and claims are subsequently settled centrally at a settlement bank. Once clearing and settlement are centralized, the number of participants does not matter as inter-bank debts and claims are established electronically. It simply becomes a matter of IT capacity. MasterCard did not contest this fact. Finally, many MasterCard cross-border card products are in practice identical with cards issued in one of the above-mentioned five schemes, because banks typically co-brand the Maestro logo on top of an existing domestic payments card.

562. Thus, these five domestic card schemes (PIN, BAX, Dankort, Bankomat and Pankkikortti) are relevant for assessing the objective necessity of the MasterCard MIF. Contrary to the argument of MasterCard these schemes have operated with considerable success despite the absence of a MIF and despite the existence of MasterCard and Visa who offer issuing banks revenues from a MIF.

7.3.4.3.1 The Dankort scheme

563. In Denmark, 148 banks issue Dankort debit cards. Acquiring for Dankort was traditionally centralised in the hand of a joint venture, PBS (665) with the exception of the time period between 1999 (666) and October 2005 (667) when the acquiring contracts were transferred from PBS to the bank with which a merchant held its current account. During those six years acquiring for Dankort was done by 145 banks (668) while from Octobe 2005 onwards acquiring in the Dankort scheme has again been centralised in the hand of PBS, an inter-bank organisation which was set up in 1986. PBS is at the same time also the sole acquirer for MasterCard and Visa branded payment cards in Denmark. PBS is owned by most of the Danish issuing banks (669). The setup of the MasterCard scheme in Denmark is the same as that of Dankort (670).

564. Dankort operated entirely without (multilataral or bilateral) interchange fees between 1984 and 1999 (ecept for one three-months period) and since 1999 96.5% of Dankort card usage is still not subject to interchange fees (671). Since 1984, when the Dankort scheme was set up, acquirers have paid no interchange fees to issuers with the exception of a short time span between January and March 2005 (672). The banks even paid retailers for accepting cards at their outlets. For so-called " non physical " payment card transactions where the payment card is used for a purchase via the internet or the telephone, interchange fees were, however, introduced in 1999. These non physical transactions account for only approximately 3.5% of the total volume of Dankort payments (673).

565. MasterCard doubts that Dankort is a "valid comparator" (674) for the objective necessity analysis of its MIF. Besides the more general arguments that Dankort was a "small domestic debit scheme" and that is had "no cross-border functionality" MasterCard discards the relevance of the scheme because the absence of a MIF in Dankort was due to statutory regulation of merchant fees (675), a situation which -according to a Danish bank- was commercially untenable (676). MasterCard moreover refers to the opinion of PBS and of Finansradet who argued in the Commission's sector inquiry that Dankort could not form "the basis for assumption that domestic card system can work without the use of interchange fees" (677).

566. In this respect the Commission's market research has shown the following. The Danish debit card scheme Dankort is a considerable success (678). Danish adults hold on average 1.6 payment cards (679) and Denmark boasts the highest per capita usage of payment cards in the EU 25 (680) in 2004 (681) and the second highest per capita usage in 2005 (682). Debit card usage in Denmark represented approximately 90% by volume and 87% by value of all Danish payment card transactions in 2004 (683).

567. Between 1984 and 2004 the Danish Act on Payment Instruments prohibited banks from charging merchants a fee for Dankort card acceptance. The act was eventually modified so that from 1999 PBS was allowed to charge merchants a fee per transaction for the use of debit cards in the non physical trade (that is to say telephone and internet orders) and from March 2005 also an annual lump sum of EUR 16.6 million for the use of debit cards in the physical trade (that is to say over the counter payments) (684).

568. Whatever the underlying historic reasons for the absence of a MIF is not objectively necessary for the operation of an open payment card scheme. The opinion of PBS and its shareholders that the statutory regulation of merchant fees of the scheme "untenable" is not surprising, because it hampers the acquirer from obtaining revenues while in the Visa and MasterCard schemes the same acquirer is allowed to charge merchants up to 0.75% per transaction. It can be expected that commercial banks will consider any form of price regulation as unacceptable.

569. The decisive fact for the objective necessity analysis of the MIF is not whether the Danish banks accept the Danis Act on Payment Instruments which is a political matter for the Danish government. The situation in Denmark is, however, a natural experiment of whether an open card scheme needs a MIF to be viable. This is not the case. As to MasterCard's argument that Dankort "runs at a loss", if indeed there had been no sufficient commercial benefit for banks to issue Dankort cards, then they would not have done it over a time span of more than 20 years. After all, the Act on Payment Instruments does not oblige banks to issue cards, it merely caps the fees on one side of the industry (685).

7.3.4.3.2 The Pankkikortti scheme

570. In Finland, the debit card system Pankkikortti has operated since 1979 and in this scheme, too, issuers receive no (multilateral or bilateral) interchange fees from acquirers (686). In 2004, 344 banks issued Pankkikortti debit cards. Since 1979 the number of debit card payment transactions and POS terminals accepting debit cards in Finland increased steadily, as shown in the following table 10 (687).

< table 10 >

571. Finland had the second highest per capita usage of payment cards in the EU 25 in 2004 (688) and the highest per capita usage in 2005 (689). Pankkikortti represented approximately 71% of the total payment card transactions and 66% of the total payment card values in this country in 2003 (690).

572. MasterCard does not seem to contest that "interchange fees are not in the Pankkikortti scheme" (691). MasterCard nevertheless disputes the relevance of Pankkikortti for a range of commercial and technical reasons that would render it irrelevant as "valid comparator" for the objective necessity analysis regarding the MasterCard MIF. Besides the more general argument that Pankkikortti was a domestic debit card scheme that operates in a small country and for domestic transactions, only, MasterCard discards the relevance of the scheme for the following specific reasons.

573. First, the Pankkikortti scheme would de facto operate with a "balancing mechanism" as card issuing would be balanced through three different manners (692):

i. [...] Luttokunta, the sole acquirer for credit cards in Finland, paid a "commission" to the issuing banks for every co-issued Visa card. This would "amount to a subsidisation of the Pankkikortti scheme".

ii. "Some acquiring banks" in Pankkikortti, which also operate as issuing banks, would be financed by " bundled fees " to merchants for the acquisition of domestic debit and/or international cards and other domestic proprietary credit card schemes. Even though interchange fees did not exist as such, transfers of funds in favour of issuing banks existed which amounted to " a form of balancing mechanism " (693).

iv. Annual cardholder fees in the Pankkikortti scheme were "unusually high" amounting up to EUR 40.30, "the highest in the EU" (694). The cardholder fees in the Pankkikortti scheme were an indication that the scheme was "not properly balanced" (695).

574. Second, the Finnish Pankkikortti scheme was "little more than a rudimentary debit scheme operating with antiquated technology". In particular Pankkikortti cards would not carry an EMV (that is to say Europay/MasterCard/Visa) chip. This lack of EMV functionality was a "significant difference" between the MasterCard and the Pankkikortti scheme (696).

575. Third, Pankkikortti scheme was a "clumsy scheme" because it operated with a "floor limit" of EUR 50. Identity checks were "slow" (697) and the floor limit "ineffective" (698).

576. Fourth, in the SEPA process Finnish bankers intended to shut down Pankkikortti, because SEPA compliance would have required "heavy investments in technical solutions and large structural changes in scheme management". This in turn indicates that the Commission should not compare Pankkikortti to "cross-border SEPA compliant schemes" such a MasterCard (699).

577. In respect of these arguments the Commission's market research has shown the following.

The structure of Pankkikortti and the reasons for absence of MIF

578. In the Pankkikortti scheme individual banks act as issuers and acquirers. Issuing banks in the Pankkikortti scheme have a choice to issue only or to acquire, as well (700). De facto, all issuers in Pankkikortti chose to acquire (701). The setup of MasterCard concentrates the acquiring of payment cards in the hand of one single entity, [...] Luottokunta. This monopolist acquirer is co-owned amongst others by MasterCard's Finnish despite the fact that the centralized setup basically puts MasterCard's acquiring bank in a better position to internalize network externalities relative to the de-centralised setup of the domestic Pankkikortti scheme which functions without a MIF.

579. The introduction of a MIF in the Pankkikortti scheme was an option and indeed discussed between the banks. If a MIF had been introduced, the interchange model would most likely have operated similary to a MIF in the international schemes, that is to say the Pankkikortti acquirer would have paid the interchange fee to the Pankkikortti issuers. However, eventually no decision to introduce a multilateral interchange fee in this scheme was taken because when put to a vote the member banks could not agree. Decisions in the Pankkikortti scheme require unanimity (702).

580. MasterCard's broader arguments on why domestic card schemes cannot validly be compared with the MasterCard scheme have been analysed above. As to MasterCard's other specific issues with regard to Pankkikortti, the Commission takes the following position.

MasterCard's argument on the existence of a balancing mechanism based on cross-subsidisation

581. The Commission does not share MasterCard's view on the existence of a "balancing mechanism" in Pankkikortti that would result from cross-subsidisation of "commissions" paid by [...] Luottokunta to Finnish issuing banks. It is not implausible that the acquirer [...] Luottokunta pays such "commissions" to Finnish issuing banks for Visa transactions as Visa generally operates with a MIF that is paid by acquirers to issuers. However, MasterCard's contention that proceeds from the Visa MIF would be used to "balance demands" in the Pankkikortti scheme is absurd as it presupposes that Visa sets the domestic MIF in Finland to cross-subsidise a competitor.

582. MasterCard's second argument (703) why Pankkikortti operated with a balancing mechanism similar to a MIF is the allegation that "some acquiring banks" in Pankkikortti were "financed by bundled fees" to merchants for the acquisition of domestic debt and/or international cards and other domestic proprietary credit card schemes". The Commission has invited MasterCard to provide the evidence underlying this claim. However, MasterCard declined to submit the relevant document invoking legal professional privilege (704). MasterCard's argument is at any rate not convincing. In Finland international payment cards such as those of MasterCard and Visa are acquired by many individual banks. Thus, proceeds from acquiring are collected by different entities (705).

583. MasterCard's third argument that cardholder fees can act as "balancing mechanism" is also unconvincing. Cardholder fees are not a balancing mechanism but the charges issuers levy on their customers. Moreover, MasterCard's information on the level of cardholder fees in the Pankkikortti scheme is erroneous. According to the Finnish Association of Bankers annual fees for pure Pankkikortti cards range between EUR 4.80 and EUR 12.00. Contrary to what MasterCard argues (706), this information clearly disproves Pankkikortti scheme was EUR 40.40. This range of EUR 4.80 to 12.00 is more than reasonable if one considers also that in the Nordic countries prices are generally higher than in the rest of Europe. It is finally surprising that MasterCard views the level of annual cardholder fees in the Pankkikortti scheme as an indication that the system was "severely out of balance" (707). Finland displays the second highest per capita usage of payment cards in the EU 25 in 2004 and the highest per capita usage in 2005. This excellent relative performance casts doubts on MasterCard's understanding of a " proper scheme balance ".

Other reasons

584. MasterCard seems to assume that only an open payment card scheme that has exactly the same technical a its payment cards at this very moment (that is to say, carry a Europay-MasterCard-Visa chip and have on-line authorisation etc.) can serve as valid " comparator " for analysing its MIF. However, MasterCard seems to forget that its credit cards in Europe are operating with magnetic stripe autorisation.

585. Similary, the use of " floor limits " for autorisation purposes in the Pankkikortti scheme is not a reason to discard the relevance of that system. When MasterCard notified its MIF in 1992 its scheme also still operated with floor limits autorisation requests (708).

586. Finally, the fact that in the medium or long term Finnish bankers might possibly choose to shut the system down does not disprove ex-post that it operated successfully without a MIF more than 25 years.

7.3.4.3.3 The Bancomat scheme

587. In Luxembourg, ten banks issue the domestic debit card Bancomat. The scheme has successfully operated since 1983 without (multilateral or bilateral) interchange fees (709). The banks decided against introducing interchange fees in the Bancomat scheme in order to stimulate the acceptance of electronic payment means (710).

588. MasterCard's assertion that the Bancomat scheme is irrelevant due to an " asymmetrical " allocation of costs in that scheme remains unproven (711). The fact that the scheme owner and acquirer of Bancomat (Cetrel) is co-owned by local banks is also not a reason to discard Bancomat as example. Acquiring for MasterCard cards in Luxembourg is also centralised (Cetrel), so the capacity of banks in the two schemes to internalise externalities without a MIF is the same. Moreover, MasterCard's member banks also used to co-own the scheme owner MasterCard Incorporated until the IPO of 25 May 2006 and still hold a large portion of the shares.

589. In the Netherlands there are currently 31 acquiring and 15 issuing banks in the national debit card system PIN (712). Until March 2004 acquiring was centralised with Interpay, an inter-bank joint venture of the largest issuers. In 2004 acquiring PIN was de-centralised and since then 31 banks compete with one another in the acquiring business. The fact that an inter-bank joint venture acquired PIN cards on behalf of the shareholder issuing banks of PIN until 2004 was equivalent to the structure of MasterCard's scheme in the Netherlands until today (713). After 2004 the structure of PIN was de-centralised and PIN still operates without a MIF.

590. Despite the fact that the Dutch debit card system PIN has been operating without a MIF since 1991 and still operates without a MIF today, the number of transactions with the domestic debit card " PIN " has grown explosively since its introduction in 1991 from 20 million to more than 954 million in 2001 and it keeps growing (714). The Netherlands had the fifth highest per capita usage of payment cards in the EU 25 in 2004 (715) and 2005 (716). The domestic debit card sytem PIN represented approximately 87% of all payment card transactions in the Netherlands in 2004 (717).

591. MasterCard discards PIN as a relevant example that open payment card schemes can operate without a MIF for two reasons.

592. First, MasterCard argues that between 1991 and 2004 Interpay had operated with a MIF -like mechanism because interpay used to pay its shareholders a yearly dividend. This dividend was an " implicit interchange fee ". Since March 2004 the dividend had ended, but he banks would " have been discussing and negotiating the introduction of multilateral or bilateral interchange fees for PIN. According to MasterCard, these interchanges fees will, at some point in time, enter into force retroactively from March 2004 onwards. MasterCard concludes that the Dutch PIN had actually never operated without an interchange fee mechanism in place (718).

593. Second, MasterCard argues that the PIN scheme was at any rate not comparable to MasterCard, because PIN cards could only be used for domestic payments (no international functionality) and Interpay had " a history of being loss making ". MasterCard infers this from a study of [...] McKinsey which was conducted on behalf of the Netherlands Bankers' Association and the Netherlands Bank in 2005. According to MasterCard the report concludes that PIN operates with a loss of EUR 101 million. MasterCard emphasises that " the procedures and methods for this research were discussed and approved in advance by all stakeholders including the NMa. The content of the report on the results of PIN was never disputed (719) ".

The dividend Mechanism in PIN and the retroactive application of bilateral interchange fees

594. On the first argument, MasterCard's viewpoint that the payment of a dividend such as in the PIN scheme has comparable effects to the MasterCard MIF must be rejected.

595. As regards the time period until March 2004 the dividends which Interpay used to pay to its shareholders did not have the characteristics of a MIF, because no minimum dividend per transaction was agreed upfront (720) and the banks could not count on a specific revenue per payment card transaction in advance. During several years (1991 until 1998) Interpay did not pay any dividends to its shareholder issuing banks (721). These are important differences to a transfer mechanism such as MasterCard's MIF. Contrary to the dividend payment mechanism in PIN the MasterCard MIF upfront inflates the base on which acquirers set their charges to merchants and at the same time shares out the profits between acquiring and issuing banks with the latter obtaining a constant revenue stream.

596. As regards the time from April 2004 until today, MasterCard's argument that PIN operated with a MIF like balancing mechanism after March 2004, when acquiring in the PIN scheme was de-centralised, must be rejected. As the NMa (722) confirmed, Dutch issuing banks indeed started negotiating bilateral interchange fee agreements. Based on MasterCard's market information the Dutch banks entered into discussions on bilateral interchange fees as of Autumn 2006 in relation to (i) the settlement of interchange fees for PIN transactions between March 2004 and December 2006 and (ii) for the post-December 2006 period (723).

597. Assuming that the PIN acquirers do indeed actually retroactively pay bilaterally negotiated interchange fees with a multilateral interchange fee from 2004. It is at best evidence that the scheme operated with bilateral interchange fee arguments. It is at best evidence that the scheme operated with bilateral interchange fee agreements which are a less restrictive means of balancing demands. Moreover, the specificity of the bilateral agreements in PIN for the period between March 2004 and December 2006 is also that the acquirers only retroactively negotiated the level of the fees with the issuers. Such a process cannot ex post restrict the competition that has already taken place between the PIN acquirers between March 2004 and December 2006 (724). The fees merchants paid between March 2004 and December 2006 were then determined by the profit maximising behaviour of each individual acquiring bank rather than by a collective decision.

598. In conclusion, the fact that PIN operated with a dividend payment mechanism until 2004 and that the banks may retroactively apply bilateral interchange fees from 2004 onwards does not demonstrate that the scheme has so far operated with a MIF-like transfer mechanism. Rather, both the Dutch dividend mechanism and the bilaterally agreed interchange fees provide examples of alternatives to a MIF which are less restrictive of competition between acquiring banks.

The [...] McKinsey study

599. On the second argument, the study of [...] McKinsey of 10 July 2006 sets out that in 2005 banks in the Netherlands suffered a total loss on payment services of EUR 23 million on revenues of EUR 3996 million. [...] McKinsey moreover alleges that the banks had in particular incurred losses of EUR 101 million with the PIN debit card (725). The study does not, however, support MasterCard's conclusion that the Dutch PIN payment card scheme has "a history of being loss making".

600. First, the [...] McKinsey study only covers the period from January to December 2005 which was an exceptional year (726). Possible losses of PIN in this year are hardly representative for the past.

601. Second, there are concerns regarding the methodology use by [...] McKinsey. Amongst others, the allocation of a bank's general overhead costs to individual payment products such as the PIN debit card is a complex exercise and ultimately depends on a subjective cost allocation of a largely fixed cost base (727). [...] McKinsey's study is of a highly aggregate level and does not provide any information on how costs were actually allocated to PIN.

602. Third, MasterCard's allegation that the procedures and methods used by [...] McKinsey had been approved upfront by the Dutch competition authority NMa is simply incorrect. Rather, the opposite is true (728).

603. Finally, the allegation that PIN had "a history of being loss making" is also contradicted by the fact that the network operator Interpay paid dividends to its shareholders banks between 1999 and 2004. The fact that Interpay incurred losses in the early years of its existence until 1998 is not untypical for a company in the start-up phase. Any business must reach a certain scale to break even. For the network operator Interpay this has apparently taken a few years. This circumstance does not disprove that the PIN scheme operates successfully without a MIF.

7.3.4.3.5 The Bank Axept scheme

604. In Norway issuing and acquiring for the domestic debit card BAX is legally separated and the vast majority of commercial and savings banks issues and/or acquires. The network is operated by Bank Axept and BBS (729). Banks never considered it to be an "issue" to introduce interchange fees in this open system (730). Despite the absence of (multilateral or bilateral) interchange fees in this scheme, the Bank Axept card became a success. In Norway, debit cards represented approximately 89% of the volume and 94% of the value of all payment card transactions in 2005 (731). Nearly of all these transactions are done with the domestic debit card of the Bank Axept card transactions are done with the domestic debit card of the Bank Axept card scheme (BAX). Bank Axept cards represented approximately 87% of the volume and 77% of the value of all Norwegian payment card transactions in 2005 (732). The yearly increase of card payments in Norway is impressive. Debit card transactions grew from 129 million payments in 1996 to over 463 million in 2002 (733). Norwegian cardholders are used to paying a fee per transaction for using their BAX cards (734).

605. MasterCard does not contest that the Norwegian debit card scheme operates without a MIF. MasterCard argues, however, that the absence of a MIF in that scheme is due to the fact that Bank Axept was faced with a "threat of blunt regulation" by the government (735).

606. MasterCard's claims are invalid. As the EFTA Surveillance Authority clarified (736), the Norwegian Competition Authority even granted an individual exemption for interchange fee agreements between banks in the Bank Axept scheme. Despite that exemption the banks did not make use of the possibility to charge interchange fees in practice (737). It is therefore unlikely that the banks refrained from enacting interchange fees in Bank Axept due the threat of regulatory intervention (738). The Bank Card Regulation of 1989 referred to by MasterCard targeted exclusively international card schemes, not Bankaxept (739). The Regulation was withdrawn after the international schemes reached an agreement with the Norwegian merchant associations (740). Until today Norwegian banks do not charge interchange fees for BAX transactions but they do so MasterCard and Visa payments.

607. MasterCard moreover argued that as acquiring was not profitable in the Bank Axept scheme the Norwegian banks had set up BBS as a "de facto" acquiring monopoly" (741). However, based on the EFTA Surveillance Authority's sector inquiry in the field of payment cads in Norway, this is incorrect (742). Banks in the Bank Axept scheme operate independently of one other, as issuing and/or services levels towards their own customers in competition with the other acquiring banks in the scheme. Only the processing of card payments is entrusted to a common institution, BBS, but this does not amount to a " de facto " monopoly in acquiring.

608. In conclusion, the long existence and considerable success of open payment card schemes in the EEA that operate without a MIF demonstrates that a MIF is not objectively necessary for the co-operation of banks in an open payment card scheme such as MasterCard's.

7.3.4.4 Cardholder fees would not be set at a level where demand is insufficient

609. MasterCard further argues that "clearly, a four party-payment system could not operate at a zero MIF. At that level, the system would fail one of the basic requirements of a two-sided product, namely to get both sides on board: Without a transfer payment from acquirers, issuers would have to set cardholder fees at levels at which there would be insufficient demand from cardholders for a viable payment system" (743).

610. MasterCard provided no concrete information as to the level at which cardholder fees would be established in the absence of its current MIF and why that new price level would lead to the collapse of the scheme. It is, in any event, unlikely that in the absence of the MIF, cardholder fees would increase to a level where the viability of the MasterCard scheme would be at stake.

611. First, demand for the cards in open payment card schemes without a MIF was clearly sufficient to render these schemes viable (see section 7.3.4.3).

612. Second, issuing banks in MasterCard's scheme are unlikely to increase cardholder fees to excessive levels in the absence of a MIF because of the issuing banks' other non-MIF related commercial benefits due to credit and debit card issuing. Issuing banks obtain considerable non MIF related revenues from cardholders which they would put at stake by raising cardholder fees to excessive levels as a reaction to the absence of the MIF. Issuing banks benefit from important cost savings due to debit card issuing (744).

613. Third, competitive constraints are likely to deter MasterCard's issuing banks from increasing cardholder fees to excessive levels. Large open schemes such as MasterCard have created an important demand for their payment cards across the EEA (745) and the number of MasterCard issuers in the EEA is high. The existence of competitive pressure on the individual issuer renders it implausible that issuing banks at large would raise cardholder fees excessively or stop offering MasterCard's payment cards entirely.

614. For the above considerations it is concluded that in the absence of a MIF, issuing banks in the MasterCard scheme would not raise cardholder fees to a level where demand was insufficient for the viability of the scheme.

7.3.4.5 No instrinsic link between MIF and other network rules

615. MasterCard argues that "amendments to remedy the setting of interchange fees to zero " would not necessarily mean lower costs for acquirers and, therefore, lower MSCs for merchants (746). This is because "the scheme would have to make the necessary amendments to its existing rules in view of achieving a balance that closely reflects the allocation that used to be realized through the interchange fee. These amendments [...] would not necessarily mean lower costs for acquirers, and, therefore, lower MSCs for merchants" (747).

616. MasterCard's claim that in the absence of the MIF a collective re-allocation of costs in the scheme would have to occur and would have similar effects on merchant fees to the present situation, must be rejected. Contrary to MasterCard's assertion (748), the allocation of fraud and default costs between issuers and acquirers as well as the timing of settlements in its system are not intrinsically linked to the level of its MIF.

617. First, MasterCard's global chargeback and settlement are identical worldwide despite the variance of interchange fees across the world ranging from an average of [business secret - 2a] in Australia to [business secret - 2a] in the USA (credit cards) (749). Moreover, when the Australian Reserve Bank reduced MasterCard's interchange fees for credit cards from 0.95% to 0.55% in 2003 the organization maintained its global chargeback and settlement rules entirely unchanged (750). If there was an intrinsic link between the MIF and these rules, one would expect to see the rules modified in order to take account of that sudden decrease in issuers' interchange revenues. That was not the case.

618. Second, as to a link between the timing of payments and the interchange fee, it is difficult to see the relevance of MasterCard's comparison between Europe and Brazil (751). While it may be correct that merchants in Brazil wait in average 28 days for the funds of a credit card transactions are settled daily, a representative of the Brazilian Finance Ministry linked the 2 days settlement period in Brazil to the past hyperinflation in this country, not to interchange fees (752). Moreover, according to this representative the average of MasterCard and Visa's interchange fees in Europe of [business secret - 2a] for credit cards and [business secret - 2a] for debit cards (753). If there was an intrinsic link between interchange fees and the settlement period, one would expect to find considerably lower interchange fees in Brazil than in Europe. That is not the case.

619. In conclusion, MasterCard's allegation (754) that in the absence of the MasterCard MIF other collective measures would inevitably have to be taken that would yield similar (negative) effects on merchant fees as the MasterCard MIF, are unconvincing. At any rate, the existence of open card schemes without a MIF (section 7.3.4.3) shows that MasterCard would not be "destabilized" as MasterCard puts it (755) - if it had to revise its network rules in the absence of the MIF.

7.3.4.6 MasterCard could compete with closed schemes absent the MIF

620. MasterCard argues that closed payment card systems such as American Express have a number of distinct advantages over open payment card systems such as MasterCard or Visa. One of the advantages of closed schemes would be that they do not face intra-system competition between member banks. The three-party schemes could therefore take advantage of any degree of market power they might have, however slight or transitory, to raise prices, and can use the resulting incremental revenue to compete more effectively against open payment card systems (756). While only open payment systems have an explicit MIF, closed systems would operate with an implicit MIF in the sense that cardholder fees would not necessarily reflect costs in the issuing side and MSCs would not necessarily reflect costs on the acquiring side (757).

621. MasterCard concludes that a MIF was objectively necessary for MasterCard to compete with American Express and other closed payment card systems, in particular in the (758) commercial card business (759). MasterCard argues that "any action by the Commission that would prevent MasterCard from establishing a MIF for commercial credit cards at higher levels than those of consumer credit cards would be tantamount to delivering the entire commercial cards business into the hands of American Express" (760) and that American Express has a "price led strategy" making commercial cards for some clients free of charge and even granting corporate customers rebates for the use of commercial cards.

622. That argument cannot be accepted.

623. First, there are open payment card system card schemes in the EEA which have faced competition from American Express and Diners for decades and have nevertheless been able to succeed without a MIF (see section 7.3.4.3).

624. Second, evidence from Australia discredits the plausibility of MasterCard's claim that the absence of a MIF would lead to the demise of the scheme as member banks would be motivated to evolve towards three party systems (see below section 7.3.4.8).

625. Third, regarding MasterCard's more specific argument regarding the sub-segment of commercial credit cards, MasterCard's claim that it would lose its entire commercial cards business to American Express in the absence of a MIF must be rejected. MasterCard's current position in Europe is robust (761).In contrast to the USA the growth of American Express in the Community is also hindered by the relatively small size of its acceptance network (762). Industry analysts therefore consider American Express as hardly capable of stopping MasterCard's further growth potential in the commercial cards (763). An independent market report states that the growth potential in the commercial cards segment is stronger overall than in the consumer cards segment (764) and expects that mainly Visa and MasterCard -not American Express - will capitalize on this growth potential (765). Even if MasterCard would lose all of its commercial cards business, which is highly unlikely, this would not endanger the viability of the entire scheme. MasterCard acknowledges that "commercial cards do not represent a significant part of [its] business". In 2003 commercial cards accounted for only [business secret - 4] of the total number of MasterCard payment [business secret - 4] of the payment cards issued in the EU 25 (766). Hence, [business secret - 4] of the payment cards carrying MasterCard logos would remain in circulation if MasterCard's entire commercial cards business was lost to the competition (767).

7.3.4.7 Memorandum submitted by the [bank]

626. MasterCard asked one of its members in the United Kingdom, [bank], to consider from a banks' perspective the possible consequences for its business if four party consumer credit card schemes were to operate without interchange fees in the United Kingdom. [bank] followed up on MasterCard's request by submitting on 5 December 2006 a memorandum to the Commission which addresses two questions, namely (i) the consequences for [bank]'s, and the commercial options available to [bank], if the MasterCard and Visa consumer credit card schemes were to operate on the basis of at par clearing in the United Kingdom and (ii) the potential consequences for cardholders, merchants and the MasterCard and Visa schemes if [bank] (and other banks) were to adopt one or a number of these commercial options available to them.

627. MasterCard believes that [bank]'s memorandum contradicts "the main argument put forward by the Commission that it would be financially viable for issuing banks to continue to issue MasterCard and Maestro branded payment cards in the absence of interchange fees" (768).

b) General

628. It is important to note from the outset that MasterCard misinterprets the Commission's objective necessity analysis by assuming that the financial viability of payment card issuing in the absence of a MIF was a "main argument" of the Commission. The Commission demonstrates that a MIF is not objectively necessary for the operation of an open payment card scheme based on (amongst others) evidence that five open payment card schemes have been operating in the EEA without a MIF for a long time period (section 7.3.4.3).

629. In theory, one could argue that the lack of objective necessity in the present case can be established by assessing the financial viability of payment card issuing in the absence of a MIF. However, this would be too narrow an approach. Arguably, the "financial viability" of an entire payment card scheme depends not only the financial viability of card issuing, but also of card acquiring and of the services provided by the scheme owner and the network operator to member banks. Moreover, precisely due to a MIF the issuing banks may incur certain costs (769) that they would possibly not choose to incur in the absence of a MIF. Hence, any attempt to benchmark the financial viability of issuing by mechanically comparing the banks' revenues including a MIF with revenues excluding a MIF while leaving costs constant would not yield reliable results. Rather, one would have to use a dynamic model that is based on certain assumptions with respect to the cost items would incur if they knew in advance that they are to issue cards without a MIF (770).

630. In addition, there are considerable procedural difficulties involved in providing MasterCard with access to confidential revenue and cost data of individual issuing banks (see Annex 7).

631. Turning to the contents of [bank]'s memorandum, [bank] has insisted on classifying its main observations as confidential towards [bank].

Bank (771)

Bank [business secret - 7]

632. Bank [business secret - 7] (772)

633. The Commission therefore has analysed the confidential version of the memorandum and concludes that [bank]'s memorandum does not establish that in the absence of interchange fees Bank MasterCard's payment card scheme would collapse.

7.3.4.8 Australia

634. In 2001 MasterCard argued towards the Reserve Bank of Australia (RBA) that regulation of the scheme's interchange fees could lead to a "death spiral" of its scheme in Australia if interchange fees were reduced and set too low as MasterCard's member banks would be motivated to evolve towards three party systems. That argument is not dissimilar to key elements of MasterCard's defence in this case (773). As the RBA set out in a public document addressed to the OECD, MasterCard's death spiral argument was proven wrong by the events following the regulation of interchange fees in Australia in 2003 (774).

635. On 1 July 2003 the RBA brought into force a regulation ("standard") which obliged MasterCard and Visa to set their interchange fees for credit cards below or equal to a certain cap. According to the RBA the schemes subsequently complied with the standard and reduced their interchange fee rates as of 1 November 2003 so that, as a consequence, the weighted average interchange fee for MasterCard and Visa credit cards dropped abruptly from around 0.95% (775) to 0.55% (776). The RBA adopted another standard on credit cards which prescribed a second reduction of these interchange fee for credit cards in either scheme may not exceed 0.50% (777). [business secret - 2a]. In the debit card segment the RBA reduced the MIF of Visa and MasterCard to 12 Australian cents per transaction which corresponds approximately to 7 euro cents.

636. The decrease of interchange fees for MasterCard and Visa credit cards in Australia was allowed by a sharp decrease of the merchant fees in both schemes. The fees of the closed schemes American Express and of Diners Club were not regulated by the RBA but their merchant fees also decreased, even though in a less pronounced manner (778). As American Express capacity to offer rewards to cardholders is dependent on the level of merchant fees (779), this drop of American Express and Diners Club's merchants fees also constrained their possibility to attract cardholders at the expense of MasterCard and Visa. The combined market share of American Express and Diners Club in Australia therefore increased only slightly from 15% to 17% and then remained stable (780). The development of MasterCard's market share and the continuing growth of its sale values disprove MasterCard's argument that a reduction of interchange fees could lead to a "death spiral" as MasterCard's member banks would be motivated to evolve towards three party systems (781).

637. MasterCard's market share in Australia evolved as follows. In 2002, the year preceding interchange fee regulation in Australia, MasterCard's market share varied between [[...] 10% to 20%] and [[...] 10% to 20%] (calculated as portion of total sales values). In November 2003, when the RBA's interchange fee regulation entered into force, MasterCard's market share had risen to [[...] 15% to 25%]. In 2004 MasterCard's market share further increased to [[...] 15% to 25%] in the first quarter of the year and to [[...] 15% to 25%] in the last quarter of the year. In 2005 MasterCard ended with no less than a [[...] 20% to 30%] market share in the last quarter of the year. The latest available figures on MasterCard's market share reached [[...] 20% to 30%] (782). Diagram 13 illustrates the evolution of MasterCard's total sales volume (in million Australian dollars) and the market share of MasterCard in the credit and charge cards segment in Australia for the time period of January to March 2006.

< diagram 13 >

638. While MasterCard's own data indicate that its market share even increased subsequent to the entry into force of the RBA's regulation of interchange fees in 2003, it must also be considered that this development coincided with the ongoing decrease of the market share of another four party scheme in Australia called Bankcard. Some or all of the decrease of MasterCard's market share may therefore be due to the disappearance of that competitor.

639. However, if MasterCard's claim that a reduction of its interchange fees would lead to a "death spiral" of its scheme was correct, one could expect to see that MasterCard's market shares and sales volume in Australia plummeted after interchange fees were reduced drastically by approximately 50% within a time span of only three years (1 November 2003 to 1 November 2006). This, however, was not the case.

640. Turning to the relevance of that evidence for the assessment of MasterCard's MIF, MasterCard discards the events in Australia as irrelevant for the objective necessity analysis at stake in this case arguing that the Australian regulation did not impose "a zero interchange fee" but only "a cap" on MasterCard's MIF. Australia was moreover "a national market which numbered few cardholders, issuers, acquirers and merchants as well as very different levels of issuance and "acceptance" and competitive conditions between Australia and Europe were different, because for instance American Express would be stronger in Europe than in Australia (783).

641. First, the Commission is well aware that MasterCard's interchange fees were reduced by 50% and not by 100%. However, the fact that MasterCard kept growing after the reduction of its MIF by 50% in 2003 is a clear trend and it is this trend that the Commission takes into account for assessing the plausibility of MasterCard's claim that the hypothetical absence of its MIF would trigger the "demise" of the "scheme (784) and a migration to three party schemes, an argument similar to the "death spiral" argument used by MasterCard in the Australian case to fend off a decrease of its domestic MIF. After all, MasterCard's transaction volumes in Australia nearly doubled since its MIF was reduced by 50% in 2003.

642. Second, MasterCard's allegation that there is a decisive difference between its competitive situation in Australia and the EEA remained unsubstantiated and is, moreover, unconvincing. MasterCard refers, in particular, to the allegedly stronger position of American Express in Europe as compared to Australia. In this respect it appears that American Express and Diners Club hold a combined market share of around 17% in Australia (785). MasterCard argues (see below) even that this figure would be too low and that it should be increased by "at least 2%" (786). This market share of 17% and possibly even 19% in Australia compares to a combined market share of American Express and Diners Club in Europe of only 2% to 3% (787).

643. MasterCard moreover also argues that the publicly available statistics of the RBA are unreliable (788). MasterCard believes that the measure of market shares used is flawed, because the RBA relied on data from acquiring banks rather than from issuing banks which would have lead to an exaggeration of MasterCard's and Visa's market shares. This is, because acquirers' sales volumes did not comprise card expenditures of Australians abroad but included card expenditure of tourists in Australia. MasterCard estimates that the impact of this "flaw" was at least 2% (789).

644. It is, however difficult to see how this should make any difference for the conclusions drawn in recitals 637 to 639. Assuming with MasterCard that in the observed time period of 2002 to 2006 its combined market share with Visa was indeed 2% lower than assumed by the RBA, this does not affect the Commission's analysis of how the reduction of the MasterCard MIF by 50% in 2003 impacted on the schemes' relative market position towards the closed card schemes hardly changed after the RBA's intervention. Whether in the time period observed (2002 to 2006) MasterCard's market share was (in each single year) 2% higher or not, is irrelevant for comparing the relative change between this scheme and the closed card schemes after the position of the RBA's

• regulation in 2003 (790). Since MasterCard did not raise any other reasons why the Commission should discard the RBA's statistics, it is concluded that those statistics are relevant and a valid basis for the Commission's assessment.

7.3.4.9 US NaBanco jurisprudence

645. In contrast to MasterCard, the Commission does not consider the US NaBanco judgment as valid precedent that MasterCard's MIF falls outside the scope of Article 81(1) of the Treaty.

646. First, US case law pertains to a different legal order and is therefore not relevant for the assessment of Article 81(1) of the Treaty.

647. Second, the US court in NaBanco based its decision that Visa's MIF did not infringe the Sherman Act on the factual analysis that credit card networks could not exercise market power because they participated in an "all payment systems" market that was highly competitive. However, the competitive landscape has changed considerably since 1986, when the NaBanco decision was issued. Visa and MasterCard now dominate the credit and charge cards business in the USA and Europe. As a consequence, in 2003, a US Appellate Court determined that Visa and MasterCard exercise market power, collectively and individually, in a product market narrowly defined as credit and charge cards (791). US card experts meanwhile question whether NaBanco is still "good law" (792). This has become even more doubtful after both Visa and MasterCard had to compensate US retailers for abusing their HACR to the debit and credit cards in order to yield excessive interchange revenues (793).

7.3.5 Conclusion

648. The considerations in the preceding recitals 555 to 647 show that a MIF and its restrictive effects on price competition between acquiring banks are not objectively necessary for the co-operation of banks in the MasterCard payment organization and for the viability of the scheme.

7.4 Appreciability

649. A decision of an association of undertakings will be caught by the prohibition set out in Article 81(1) of the Treaty if it has the object or effect of appreciably restricting competition within the common market (794). To the contrary, a decision falls outside the scope of Article 81(1) of the Treaty when h has only an insignificant effect on the market taking into account the weak position of the parties concerned (795). When determining if there is appreciable effect on competition, the primary criterion is the position and importance of the parties on the market taking into account the market structure (796).

650. The Commission considers that the MasterCard MIF appreciably restricts competition in most EEA Member States because of:

i. the substantial economic impact of the MIF within the EEA;

ii. the strong position of MasterCard on the relevant markets, in particular due to network effects;

iii. the fact that the MIF is part of a network of inter-related agreements that together have a reinforced cumulative effect and

iv, the fact that the MIF is a collective price agreement that determines to a significant degree prices for cross-border and domestic transactions in the downstream acquiring markets and allows MasterCard's members to exploit their market power.

a) Economic significance of the MasterCard MIF

651. The economic impact of the MasterCard MIF is substantial MasterCard is not a small scheme in its start-up phase but a mature system that has been using a MIF for almost 30 years (797). Due to the substantial number of MasterCard and Maestro cards in circulation and the high number of merchants accepting them in the EEA, a restriction of price competition between member banks of such a scheme cannot but be considered as appreciable.

b) The strong position of MasterCard on the relevant markets

652. MasterCard has a strong position on the relevant markets, as set out in section 2.2.

c) Network of inter-related decisions and agreements

653. MasterCard's MIF forms part of a network of inter-related or similar arrangements that, taken together, have a cumulative restrictive effect on competition.

654. First, the MasterCard Intra-EEA Fallback interchange fees not only apply to cross-border card payments. It also sets a reference for interchange fees regarding domestic card payments. This is because Intra-EEA fallback interchange fees (and SEPA fallback interchange fees) apply automatically to domestic payments if local member banks do not set country specific rates (or ask MasterCard [business secret - 2a] to set such rates on their behalf). Intra-EEA fallback interchange fees should not be seen in isolation as they form part of a set of similar interchange agreements within the MasterCard scheme that, together, determine the complete amount of costs loaded onto merchants and the price distortion that ensues.

655. Second, MasterCard's MIF is mirrored by similar interchange agreements adopted under the Visa scheme (798). Several domestic payment card systems also operate with multilaterally agreed interchange fees (799). It appears, therefore, that the payment card markets are generally governed by networks that apply similar decisions and agreements. There is therefore a cumulative effect on the relevant markets.

d) Considerations regarding blending

656. In several parts of its submission to the Commission MasterCard puts forward the theory that the MIF does not appreciably restrict competition as long as it applies only to a minor part of the total domestic and cross border transaction volumes with MasterCard branded payment cards. MasterCard uses this argument in general (800) for the Intra-EEA fallback interchange fees and more particularly for the Intra-EEA fallback interchange fees as far as commercial cards are concerned (801). Essentially MasterCard argues that acquirers blend cross border and domestic or consumer and commercial cards fees into one single MSC they charge to merchants. MasterCard then goes on to contest that the MIF sets a sufficiently significant floor for the MSC and thereby does not appreciably restrict competition.

657. However, the Commission has demonstrated in section 7.2.1.1 that the blending of interchange fees in a single merchant fee is irrelevant for the purpose of assessing the restrictive effects.

658. The overall impact of MasterCard's MIF on individual merchant fees can be considerable, as illustrated in Diagram 14 on the proportion of cross-border usage of MasterCard branded credit and charge cards in the EU 25.

< diagram 14 >

[business secret - 4]

659. Diagram 14 shows that in [business secret - 4] Eurozone countries cross-border transactions amount to between [business secret - 4] of total sales volumes. The higher the proportion of cross-border sales in a given Member States, the more cross-border interchange fees are of importance as cost element for acquirers.

660. The countries with high cross-border interchange fee revenues. According to MasterCard, (EUR) [business secret - 2a] and (EUR) [business secret - 2a] banks obtained the largest part of total cross-border interchange fees in 2003, followed by issuing banks in (EUR) [business secret - 2a], (EUR) [business secret - 2a] (802).

7.5 Effect on trade

661. According to the case-law of the Court of Justice for an agreement between undertakings to affect trade between Member States it must be possible to foresee with a sufficient degree of probability on the basis of a set of objective factors of law or fact that it may have an influence, direct or indirect, actual or potential, on the pattern of trade between Member States. Whilst Article 81(1) does not require that provisions have actually affected trade between Member States, it requires that it be established that the agreements are capable of having that effect (803).

662. In the present case MasterCard's MIF affected and still affects trade between Member States in the EEA, because they caver cross-border payments and the MIF moreover applies by default to domestic payment card transactions in whole Member States of the EEA. MasterCard has not contested that the MIF affects trade between Member States.

7.6. Conclusion on Article 81(1) of the Treaty/Article 53(1) of the EEA Agreement

663. The MasterCard MIF constitutes a decision of an association of undertakings, both before and after the IPO of MasterCard Incorporated at the New York Stock Exchange on 25 May 2006.

664. That decision restricts competition between acquiring banks by inflating the base on which acquiring banks set charges to merchants and thereby sets a floor under the merchant fee. In the absence of the multilateral interchange fee the prices set by acquiring banks would be lower to the benefit of merchants and subsequent purchasers.

665. The MasterCard MIF is net objectively necessary for the operation of an open payment card scheme such as MasterCard's. Evidence from five open card schemes without a MIF shows that an open card scheme such as MasterCard's would be viable in the absence of a MIF.

8. Article 81(3) 0f the Treaty/Article 53(3) of the EEA Agreement (804)

8.1 General

666. Agreements between undertakings or decisions of associations of undertakings which are caught by Article 81(1) of the Treaty/Article 53(1) of the EEA Agreement may benefit from an exception if they satisfy the four cumulative conditions set out in Article 81(3) of the Treaty/Article 53(3) of the EEA Agreement. Thus, multilateral interchange fees that have the object or effect of restricting price competition between a scheme's member banks are net as such illegal as they may potentially fulfil the conditions of Article 81(3) of the Treaty/Article 53(3) of the EEA Agreement. The two positive conditions of Article 81(3) of the Treaty/Article 53(3) of the EEA Agreement require that the agreement (a) contributes to improving the production or distribution of goods or to promoting technical or economic progress (b) while allowing consumers a fair share of the resulting benefits, and the two negative conditions oblige (c) not to impose on the undertakings concerned restrictions that are not indispensable to the attainment of these objectives and (d) not to afford such undertakings the possibility of eliminating competition in respect of a substantial part of the products in question.

667. Where a restriction of competition exists within the meaning of Article 81(1) of the Treaty/Article 53(1) of the EEA Agreement, Article 81(3) of the Treaty /Article 53(3) of the EEA Agreement can be invoked as a defence. According to Article 2 of Regulation 1-2003 the burden of proof under Article 81(3) of the Treaty rests on the undertakings invoking the benefit of the exception rule. The four conditions are cumulative (805).

668. The starting point for the Commission's assessment is MasterCard's arguments as to the purpose and function of its MIF. As set out in detail in section 4.1.3, MasterCard's views have, however, evolved over time. While during most of the period of it existence MasterCard defined its MIF as a price or fee paid for services provided by issuers to acquirers and merchants, MasterCard later referred to the MIF as a tool to balance cardholder and merchant demand. In MasterCard's favour the Commission therefore analysed both possibilities under Article 81(3) of the Treaty in the two Statements of Objections.

669. As set out in section 3.1.8. MasterCard clarified in reply to the Supplementary Statement of Objections that its MIF was not a "fee for services" or a "price" for specific services but a balancing mechanism for cardholder and merchant demands.

8.2 First condition: Technical or economic progress

670. Agreements capable of benefiting from the exemption provision of Article 81(3) of the Treaty are those that contribute to improving production or distribution or to promoting technical or economic progress. Firstly, it follows from the case law of the Court of Justice that only objective benefits can be taken into account. This means that efficiencies are not assessed from the subjective point of view of the parties. Cost savings that arise from the mere exercise of market power by the parties cannot be taken into account. Cost reductions that do not produce any pro-competitive effects on the market and that merely allow the undertakings concerned to increase their profits, are therefore irrelevant from the point of view of Article 81(3) of the Treaty (806). Secondly, an improvement only meets the first exemption condition if the parties have demonstrated that it brings appreciable objective advantages of such a character as to compensate for the disadvantages caused by the agreement in relation to competition (807).

671. The purpose of the first condition of Article 81(3) of the Treaty is to define the types of efficiency gains that can be taken into account and be subject to the further tests of the second and third conditions of Article 81(3) of the Treaty. The aim of the analysis is to ascertain what the objective benefits created by the agreement and what the economic importance of such efficiencies are. Given that for Article 81(3) of the Treaty to apply the pro-competitive effects flowing from the agreement must outweigh anti-competitive effects, it is necessary to verify what the link between the agreement and the claimed efficiencies and what the value of these efficiencies are (808).

672. All efficiency claims must therefore be substantiated so that the following can be verified: (809)

i. The nature of the claimed efficiencies;

ii. The link between the agreement and the efficiencies;

iii. The likelihood and magnitude of each claimed efficiency; and

iv. How and when each claimed efficiency would be achieved.

8.2.1. MasterCard's arguments

673. Firstly, MasterCard provides a list of efficiencies that it claims represent the technical and economic progress "of payment card systems" (810). The list includes: increased sales for merchants, improved cash-flow for merchants, improved security and back-office operations for merchants, payment guarantee for merchants against cardholder default, payment guarantee for merchants against fraud, new sales channels, an alternative ability for cardholders to defer payment for goods and services over a convenient period, increased personal security for cardholders and fraud protection for cardholders.

674. Secondly, MasterCard claims that the technical and economic progress "of the MasterCard system" is achieved "as a result of the MasterCard default MIF" (811). MasterCard attributes such progress to the fact that the MIF renders the MasterCard payment system more efficient. "The MasterCard default MIF creates considerable benefits and efficiencies, thereby contributing to economic and technical progress, by allowing for an additional payment card system to operate and to compete against other payment card systems as well as other payment systems. This additional payment system, i.e., MasterCard's, also facilitates intra-system competition by both issuers and acquirers, thus further promoting competition and innovation in the payments market" (812).

675. In describing the way in which its interchange fee contributes to that progress, MasterCard relies on the theory that "the default MIF is a method of balancing the demand of cardholders and merchants in a market with two-sided demand" and therefore "undoubtedly" fulfils the first condition (813).

676. The argument consists of two elements. Firstly, according to MasterCard, issuers and acquirers provide a joint service to cardholders and merchants and face joint demand. "From an economic perspective" there would be common costs to that joint service which have to be allocated between the two groups of users, and, therefore, an interchange fee is "inherent and indispensable to the operation of a four-party payment card system" (814). Secondly, according to MasterCard neither issuers nor acquirers take into account the effect of the prices they set on the other side. Interchange fees rectify this "problem" because payment system operators, when setting the interchange fees, can take that effect into account, maximise the "overall value of the product on both sides" and balance demand between cardholders and merchants (815).

677. MasterCard also suggests that its MIF allocated costs between cardholders and merchants according to their relative price sensitivity and that "the costs that are measured by [MasterCard] methodology represent a legitimate proxy of the actual costs incurred" (816). The interchange fee should be characterized as balancing mechanism "for the costs incurred in the mutual exchange of services between issuers and acquirers" (817). The balancing mechanism operated by allocating between the scheme's issuers and acquirers "the cost of delivering the service" (818).

678. MasterCard argues that the Commission was wrong to request MasterCard to establish under Article 81(3) of the Treaty that the interchange fee "set at a certain level" was indispensable to achieve objective efficiencies within the meaning of Article 81(3) of the Treaty, because such requirement amounted to an "attempt to regulate the level of MasterCard's interchange fees" (819) and the Commission would lack such powers to set MasterCard's interchange fees at a certain level (820).

8.2.2 The Commission's Analysis

679. It is not disputed that payment card schemes such as MasterCard may represent, as such, economic and technical progress. The decisive question is, however, whether the MasterCard MIF specifically contributes to that progress.

8.2.2.1 Assessment of MasterCard's "joint service" argument

680. MasterCard's starting point is that payment services in a four-party (or open) payment card system, which it characterizes as a "joint venture type arrangement", are a "joint service" provided cooperatively by issuers and acquirers. The fact that certain costs are at the issuing side and others at the acquiring side would be irrelevant, because from an economic perspective they are "common costs to a joint service" which had to be allocated to the two sides of the scheme (821).

681. The Commission does not consider MasterCard's payment system as a "joint production" venture (822) and MasterCard's argument that its MIF allocates "joint costs" is misleading. In an economic sense "joint" or "common" costs are costs associated with a production process that results in two different products that are sold to two different customer categories. For example, if a mining company takes gold and silver from the same mine, it is impossible to distinguish the costs of mining each. Hence, the company will have to decide how to allocate the "joint costs" of mining on gold and silver. In an open card scheme, however, issuers incur distinct costs for offering card holders payment cards while acquiring banks incur distinct costs for enabling merchants to accept cards. The issue of common costs does not arise.

682. In a wider sense, it may be efficient for distinct companies in two-sided industries to voluntarily reallocate costs in another way than letting each company bear the costs that it incurs itself, in order to make use of network effects. However, it cannot be presumed that cost transfers are necessarily beneficial for the customers of the system as they may also be used to increase market power of the firms by restricting competition on one side, where this competition would otherwise compete profits away. Therefore, convincing evidence will have to be brought forward that demonstrates the beneficial effect of the restriction of competition. As will be set out below in section 8.2.2.2, no such evidence was produced by MasterCard.

8.2.2.2 Assessment of MasterCard's "balancing of demand" arguments a) "MasterCard MIF operates by allocating costs" - lacks clarity

683. MasterCard's claim that its MIF operates "by allocating costs" remains opaque. MasterCard refers both to "costs incurred in the mutual exchange of services between issuers and acquirers" (823) and to costs of delivering "the service", (824). MasterCard did not, however, specify what the "mutual services" of issuers to acquirers (or vice versa) are and also did not provide further specifications about the alternative concept of costs incurred in delivering "the service" (825). Whether the MasterCard MIF indeed operates "by allocating costs", as MasterCard claims, is particularly doubtful as MasterCard's counsel acknowledged that in practice the scheme does not allocate any specific costs (826).

b) "Imbalance" between issuing and acquiring

684. One of the crucial (827) assumptions underlying the MasterCard MIF is a perceived imbalance between the issuing and the acquiring business in the scheme. MasterCard derives that imbalance from the fact that the average issuer would "incur the vast majority of the Scheme costs" because in the United Kingdom market 95% of the costs are skewed towards the issuing side (828).

685. The argument that a MIF was required because issuing banks incurred 95% of the total costs in the United Kingdom is a circular reasoning because it is precisely due to the MIF that issuing banks incur certain costs that they would not incur in the absence of a MIF. To the extent that a MIF provides a stimulation to issuing banks to issue cards they may incur all kind of (marketing) cost to push card usage and these cost then determine ex post the objective necessity for MasterCard to "cover" these cost with a MIF. In economic terms, MasterCard's argument suffers from endogeneity.

686. Also, an "imbalance" between issuing and acquiring cannot be assumed on the basis of cost considerations, but has to comprise an analysis of revenues, as well. A cost imbalance is as such no sufficient evidence to explain why MasterCard's MIF is always paid by the acquirer to the issuer, irrespective of the concrete market situation (829). If receipts (interests, money exchange fees, penalty fees, etc.) or other monetary benefits (resulting from cost savings such as reduction of staff, paperwork etc.) from payment card issuing provide sufficient commercial incentives for banks to invest in incremental card issuing, a transfer from acquiring to issuing may be superfluous and even counterproductive as the revenue transfer dampens card acceptance due to the increase of costs on the merchants' side. Robust empirical evidence is therefore required to establish the necessity for and the direction of a fallback interchange fee (830).

c) The MasterCard MIF "maximizes system output"

687. MasterCard has failed to demonstrate on the basis of appropriate empirical evidence, how and to what extent the Intra-EEA fallback interchange fees are causal for appreciable objective efficiencies of such kind as to offset the resulting disadvantages for competition (831).

688. The central efficiency claim of MasterCard rests on the alleged capacity of the. MasterCard MIF to help the scheme to maximise system "output" by balancing cardholder and merchant demands. This effect is then said to contribute to other efficiencies, such as sales for merchants; improved cash-flow for merchants; improved security and back-office operations for merchants; payment guarantee for merchants against cardholder default; payment guarantee for merchants against fraud; new sales channels; an alternative, cardholders' ability to defer payment for goods and services over a convenient period; increased personal security for cardholders; fraud protection for cardholders (832), increased competition in bath issuing and acquiring, lower costs through economies of scale and increased innovation (833).

689. As set out in the Statement of Objections of 21 June 2006 to MasterCard, it cannot just be assumed, as MasterCard does, without detailed economic and empirical analysis, that a MIF maximises the overall benefits of a system to merchants and cardholders "by reducing costs, increasing services levels and contributing to overall economic welfare". The mechanism may overburden one side of the scheme with (artificial) costs while not yielding any positive effects on scheme growth and overall efficiency.

690. Hence, whether a MIF should be paid by acquirers to issuers or vice versa, and whether it should be set at a certain amount or at zero, cannot be determined in a general manner by economic theory alone (834). A claim that an interchange fee mechanism creates efficiencies within the meaning of Article 81(3) of the Treaty therefore must be founded on a detailed, robust and compelling analysis that relies in its assumptions and deductions on empirical data and facts. Apart from MasterCard's general assertion that balancing of the demand of cardholders and merchants leads to a better performance of the MasterCard system, is inherent and indispensable to the operation of a four-party payment card system, contributes to overall economic welfare and therefore "undoubtedly" fulfils the first condition of Article 81(3) of the Treaty (835), no such analysis and empirical evidence was provided to the Commission (836).

691. MasterCard's claim that increased sales volumes equal increased competition in both issuing and acquiring to the benefit of all users (837) is moreover unconvincing.

692. It has also been set out in section 7.3.4 that MasterCard would not collapse in the absence of a MIF. MasterCard's argument that an efficiency of the MasterCard MIF was to allow an additional card scheme to operate (which could be seen as increase of inter-system competition) must therefore be rejected. Contrary to MasterCard's views in reply to the Supplementary Statement of Objections, the increase of a scheme's output also does not necessarily equal an increase of effective intra-system competition. An artificially created input cost such as the MIF may reduce the number of banks on the acquiring side, because the expected higher cost level makes entry less desirable. On the issuing side, the MIF may lead to entry of banks that would be unable to offer cards without the subsidy in view of their marginal cost. However, the entry of these inefficient suppliers in the issuing market is as such no gain for competition.

693. Finally, any transfer which results from a restriction of competition must be more than compensated by benefits for somebody else other than the undertakings participating iii the restrictive arrangements. The increase of sales volumes in MasterCard's scheme is clearly to the advantage of MasterCard's member banks. An increase of system output only contributes to appreciable objective advantages if parties other than the organisation's member banks benefit from it. The analysis of the first condition is therefore linked to the second condition of Article 81(3) of the Treaty, i.e. whether there is a sufficient pass-on of benefits to cardholders as well as merchants (and subsequent purchasers). A methodology to set a MIF must consider the benefits to the final users of a card scheme.

e) Need for empirical evidence - No excessive burden of proof on MasterCard

694. MasterCard (838) errs in its assertion that the Commission imposes an excessively high burden of proof on MasterCard if it requires MasterCard to demonstrate empirically a causal link between the MIF and actual effects on system output as well as the objective efficiencies that could result from increased system output.

695. It is on the undertakings in the first place to present to the Commission the evidence intended to establish that the agreement in question fulfils the conditions laid down by Article 81(3) of the Treaty (839). In the context of the first condition it has to be ascertained that the restrictive effects are offset by efficiencies. In this context the undertakings concerned must demonstrate whether a MIF generates the positive effects which the underlying model claims to achieve, here: an increase of system output and possible related efficiencies (840). To the extent that objective efficiencies cannot be established empirically, they cannot be balanced with the restrictive effects. Some form of convincing empirical evidence on the actual effect of a MIF on the market is therefore required.

696. The Commission's request for empirical evidence for a positive impact of MasterCard's MIF on the scheme's "output" and possible related efficiencies is supported by the relatively good output of card schemes without a MIF. As statistics of the European Central Bank show, payment card usage per capita is above Community average particularly in Member States where payment card schemes operating without a MIF have been present for a long time period. Finland, Denmark, the Netherlands and Luxemburg rank among the countries with the highest card usage per capita (841).

697. Diagram 15 displays the number of card payment transactions with debit-, charge and credit yards per capita in the EU-25 for 2005 (842).

698. On average a European citizen uses his/her payment card 50 times a year. In Finland, however, 128.5 card payments occurred per capita in 2005. This is the highest card usage in the EU 25. Denmark displayed 122.75 card payments per capita, the Netherlands 86.16 and Luxembourg 81.5. The "no-MIF" schemes operating in those Member States hold between 71% and 90% of all payment card transactions in their respective territories (see section 7.3.6).

699. In view of the performance of payment card schemes without a MIF, it is justified to require empirical evidence for the actual effects of a MIF on system output if this is the central efficiency claim of a scheme. MasterCard has merely contended this (843). Moreover, MasterCard did not demonstrate how and to what extent an increase of its sales volume benefits cardholders, merchants (and subsequent purchasers).

700. None of the three Annexes, the two exhibits and the 64 appendices to MasterCard's reply to the first Statement of Objections dated 5 January 2004 (844), the subsequent submissions of MasterCard's expert (Annexes 1.1 to 1.3), or the report on the Merchant Acceptance of payment cards and other payment means, nor the Research Market Study on the characteristics of retail credit contain empirical evidence that would have enabled the Commission to assess how and to what extent the MasterCard MIF contributed and continues to contribute to an efficient balance of cardholder and merchant demand in the MasterCard scheme, what that efficient balance would be and how this could contribute to economic and technical progress. This is even more surprising as supposedly this type of evidence forms the basis which MasterCard uses to set its MIF.

701. In conclusion, the MasterCard MIF does not meet the first condition of Article 81(3) of the Treaty as MasterCard failed to demonstrate a causal link between the MIF and objective efficiencies.

f) The Baxter framework

702. Notwithstanding the lack of evidence to bolster MasterCard's efficiency claim the Commission has moreover also assessed the theoretic underpinnings of MasterCard's MIF, as well as MasterCard's methodology for setting the level of the Fees. These theoretic underpinnings arc problematic and the methodology MasterCard applies has little in common with the theory.

703. The Baxter framework, on which MasterCard relies (845), finds that a scheme's size will be maximised (as will the welfare of the scheme's participants in the Baxter framework) when the total costs incurred by issuers and acquirers at the margin are shared between any cardholder and merchant in proportion to the value that each places on the services received. This conclusion is not convincing for several reasons.

704. First, the Baxter framework is severely limited by the fact that it takes consumer and merchant demand as given in that neither strategically reacts to possible actions by others. Baxter assumes that merchants' willingness to pay for card acceptance through an interchange fee is a measure of the benefits that they enjoy from card acceptance. There is, however, consensus in the recent academic literature that the Baxter framework leads to an overestimation of the level of an efficient interchange fee for several reasons (amongst others) by ignoring the fact that merchants, as strategic players, are willing to accept cards in order to avoid a competitive disadvantage relative to other merchants (846).

705. Merchants will accept cards, in part, to attract customers from each other and this will increase the amount they will be prepared to pay to accept cards above a level determined solely from transactional benefits they obtain from accepting cards. Since some or most of the additional sales won by one will be taken from its competitors, the acceptance of cards may have little or no effect on total sales from the perspective of merchants as a whole (847).

706. Second, Baxter's result also relies on the unrealistic assumption of a perfectly competitive banking industry. Once that assumption is lifted, the result fails to hold. If issuing banks have some degree of market power in the Baxter framework, an interchange fee allows issuing banks to extract rents from merchants. This means that the interchange fee influences not only the number of card transactions, but also the sum of prices that merchants and cardholders have to pay. Therefore, the result that banks choose the interchange fee to maximise system output is invalid if the- assumption of perfect competition does net hold. Likewise, the result that the interchange fee is set to maximise participant's welfare fails to hold.

707. Third, the Baxter framework does not take into account the effect that card usage bas downstream of the acquiring markets. However, when merchants do not surcharge card users for paying with a card, the higher merchant service fee that results from the interchange fee will be passed on to all consumers through higher retail prices. Hence, card usage does not involve a direct cost on users, but drives up retail prices. The economic literature has clearly shown that this may lead to a reduction of consumer and merchant welfare. Therefore, maximisation of output is not necessarily optimal for the customers of the system. Hence, Baxter's result also fails to hold once repercussions on the markets downstream of the acquiring markets are taken into account.

708. Fourth, the Baxter framework relies on the unrealistic assumption that there is no variation in the benefits that merchants get from accepting cards, i.e. that merchants are homogenous. It is, however, well established in the literature that merchant heterogeneity makes the determination of the optimal interchange fee much more complex than it is suggested in the Baxter framework (848).

g) The methodologies used to apply the Baxter framework in practice

709. As MasterCard recognizes (849), the Baxter framework is theoretical and does not set out a practical way to calculate an interchange fee. MasterCard has devised two methodologies to apply the theory iii practice, namely the MasterCard Standard Interchange Methodology and the Global MasterCard Debit Interchange Fee Methodology. For a description of these methodologies see above in section 3.1.9.3.

710. The "MasterCard Standard Interchange Methodology" (850) in the credit and charge card segment (851) purports to apply the Baxter framework in practice by measuring the utility of cards for merchants through cost studies with banks.

711. Baxter requires knowledge of the "contours" of merchant and cardholder demand. Baxter suggests that the "contours" of those demands are revealed by the willingness of merchants and cardholders to pay for bank-card related services (852). Baxter proposes to measure this willingness of cardholders and merchants to pay by examining the "historical" charges imposed by banks on merchants and cardholders. In a next stop Baxter then aggregates the two demand curves to cross the joint demand with the "joint supply" curve of all member banks in the scheme where the latter is determined by the costs of offering card payment services for acquiring and issuing banks (853).

712. MasterCard applies the Baxter framework by assessing which costs card issuing banks incur. The data collected in cost studies are said to serve as "proxy" for measuring the willingness to pay of merchants (854). This approach is not satisfactory.

713. First, MasterCard's claim that it can measure merchants' elasticity of demand by estimating merchants' opportunity costs for issuing store cards through benchmarking some cost items incurred by its issuing banks does not guarantee that - as MasterCard claims - "at this level of cost a merchant may consider it viable to establish its own alternative payment method" (855). As not every merchant wishes to issue store credit cards, the potential cost savings some merchants may theoretically incur for not having to issue store credit cards is a bad measurement to approximate overall merchant willingness to pay for MasterCard's credit cards. Moreover, beyond the payment functionality store cards serve different purposes than a bank issued credit card. Store cards are in particular issued to enhance customer loyalty towards a specific merchant (or a specific group of shops, if they together issue and accept a certain store card). By buying again at a shop the customer typically gets rebates which are counted on the store cards. The shop in turn wins the loyalty (and often a home address of the client to mail advertisements). Bank issued credit cards do not have such loyalty enhancing effects towards specific merchants. They are a tool for banks to sell (unsecured) bans to bank customers in return for (high) interests. MasterCard's assumption that merchants save on costs for issuing their own store card products if they accept bank issued credit cards (and that these cost savings are in turn the opportunity cost of merchants for declining bank issued credit cards) is therefore unrealistic. Indeed, many merchants accepting credit cards still issue their own store cards to enhance the loyalty of their customers. Also, not all (in particular not smaller) merchants accepting bank issued credit cards are actually in a position to issue their own payment products. Furthermore, the analysis neglects merchants' strategic desire to offer MasterCard payments taking into account other merchants' behaviour. Overall, the costs of issuing banks appear to be a particularly poor indicator of merchants' willingness to pay for card services.

714. Second, while Baxter suggests relying on historic data on Fees cardholders pay for using credit cards, MasterCard does not study in an objectively verifiable manner the willingness to pay of cardholders, that is to say their aggregate demand for MasterCard payment card usage. To rely simply on "conversations and information that is provided by issuers" is not sufficient. The fact that issuing banks tell MasterCard management "what their desires are" can at best be a starting point. It is an entirely different question to which extent cardholders are indeed willing to pay for MasterCard/Maestro card usage. This cannot be assumed without a robust methodology and verifiable studies (856).

715. The empirical evidence submitted by MasterCard to support its claim that cardholders are unwilling to pay for card usage is sparse.

716. In 2003 MasterCard submitted a model on cardholder discouragement (857). The Commission has set out in paragraphs 171-175 of the Supplementary Statement of Objections of 20 June 2006 why it doubts the probative value of this study. The study was at any rate not instrumental for setting the level of MasterCard's MIF. Finally, where payment card schemes without a MIF impose card fees per transaction the magnitude of such fees has not discouraged consumers from using the schemes' cards. Reference is made to section 7.3.4.3.

717. In 2006 MasterCard submitted a second report on merchant discouragement, which is analysed more in detail in Annex 3 to this decision. The surveys attached to that report, that is to say the merchant acceptance survey (858), the "Retail Payment Mechanisms in Europe" survey and the survey on the use of retail credit (859), do not analyse cardholder willingness to pay, either. All these studies on cardholder and merchant discouragement were also produced after the initiation of the proceedings in September 2003 and did not serve to set the direction and the level of the cross-border interchange fees. They are no evidence that MasterCard quantified the willingness to pay of cardholders in an objective manner prior to determining the direction of the interchange flow and its amount.

718. MasterCard therefore failed to establish a conceptual link between the problems allegedly addressed by interchange fee according to MasterCard (860) and the methodology used in practice for determining the level of Intra-EEA fallback interchange fees.

719. The Global MasterCard Debit Interchange Fee Methodology, which MasterCard applies only since April 2005, appears conceptually closer to the Baxter framework as it not only considers costs on the issuing side but also costs on the acquiring side of the scheme.

720. However, in establishing an imbalance between issuing and acquiring MasterCard again omits to assess banks' receipts without interchange fees. The possibility cannot therefore be ruled out that banks have sufficient commercial incentives to issue MasterCard's debit cards without an interchange fee (861). Thus, the MasterCard MIF may as well rather impede than promote Maestro debit card usage as the utility of an interchange fee transfer to the issuing side was not objectively established a priori.

721. MasterCard's assumption, on the basis of experience with cardholder demand in the United Kingdom for credit cards, that demand for debit card transactions is predominantly on the merchant side is not conclusive. Even if merchants "may be viewed as 'transaction hungry" in the debit card segment, this does not mean a contrario that cardholders would be unwilling or even unable to pay a transparent price for debit card usage. MasterCard's attribution of total issuer and acquirer costs ([business secret - 6] (862)) without empirical studios on cardholders' willingness to pay for debit card usage is simply arbitrary (863).

722. A major concern with respect to MasterCard's Debit Global Interchange Fee methodology is also that the methodology does not only take account of costs directly attributable to the issuing of a debit card (864). MasterCard in particular also includes a portion of current account costs in its cost study arguing that "the nature and operation of a debit card" necessitate the inclusion of these additional costs on the issuing side (865). This is not convincing: (866) In practice banks often provide debit cards to current account holders and the two products are often linked, but it is not clear why that link should be particular and intrinsic to debit cards. Any non-cash payment instrument needs to be linked to some account on which the funds to be transferred are annotated (867). Neither is it necessary for a bank to offer debit cards together with a current account and account holders do not necessarily hold debit cards (868).

723. Finally, while MasterCard argues that it did "not pursue a policy of MIF maximisation to the benefit of issuers (869), the Commission has analysed MasterCard's past practice in setting interchange fees (section 4.1.5). From that analysis it appears that the objective to get issuing banks to issue MasterCard card products rather than those of the Visa organisation was an important if not predominant incentive of MasterCard's European Board in determining the level and the structure of Intra-EEA fallback interchange fees. MasterCard's management also planned interchange Fee increases years ahead (870). Decisions of the European Board's member bank delegates actually had little to do with internalising externalities or at best are biased towards underrating the externalities generated by merchants (871).

724. For all these reasons MasterCard's view (872) that its methodology can safely be assumed to favour the setting of interchange fees at art efficient level must be rejected.

8.2.2.3 Assessment of other efficiency arguments

"Fee for services"

725. As to MasterCard's initial views that the MIF was a 'fee for services" provided by issuing banks which benefit merchants, MasterCard explicitly abandoned this concept in its more recent submissions. It is therefore not necessary to enter into a discussion of this previous argument.

"War on cash"

726. MasterCard's management argued in meetings with the Commission that it used the MIF to fight a "war on cash". MasterCard has, however, not provided evidence that its MIF was causal for the replacement of cash and cheques in Europe (873).

727. MasterCard's further argument that four party schemes have done "better" in replacing cash and cheques than three party schemes, remains equally unsubstantiated. Moreover, neither the Pay Later nor the Pay Now cost studies have taken account of the "cost of cash" or of other paper based payment means such as cheques. In addition, these costs also do not figure among the criteria MasterCard listed as determinants for setting its Intra-EEA fallback interchange fees.

728. In conclusion, MasterCard's view that the MasterCard MIF is used to fight a "war on cash" cannot be accepted.

8.2.3 Conclusion on the first condition of Article 81(3) of the Treaty

729. In its Visa II decision, the Commission stated that an interchange fee agreement can in principle contribute to economic and technical progress within the meaning of Article 81(3) of the Treaty. The Commission does not dispute that, in principle, in a payment card system characterised by indirect network externalities, interchange fees can help optimise the utility of the network to its users (merchants and cardholders) (874). Nor does the Commission dispute that there are theoretical models such as the model developed by Baxter that - using a number of assumptions - aim at calculating "optimal" prices to cardholders and merchants for a payment card product through interchange fees (875).

730. However, whether an interchange fee should be paid by acquirers to issuers or vice versa, and at which level it should be set to enhance scheme output, cannot be determined in a general manner by economic theory alone. There is no presumption that MIFs in general enhance the efficiency of card schemes just as there is no presumption that they do not fulfil the conditions of Article 81(3) of the Treaty and are therefore illegal. A MIF may be used by banks to achieve efficiencies as well as to extract rents. The Commission's conclusion on the efficiencies of a MIF will depend on the concrete evidence brought forward by the parties.

731. Contrary to MasterCard's perception (876) the Commission's position is not that only the level of a MIF is a decisive criterion for assessing whether that MIF fulfils the first condition of Article 81(3) of the Treaty. Rather, the existence of objective appreciable efficiencies is assessed in relation to the MIF as such, the effects it produces on the market and the manner in which it is set. In particular, the Commission verifies on the basis of the evidence submitted whether the model underlying a MIF is based on realistic assumptions (which is not the case here), whether the methodology used to implement that model in practice is objective and reasonable (which is not the case for the two methodologies used by MasterCard) and whether the MIF indeed bas the positive effect on the market to the benefit of both customer groups which the model claims.

732. Any claim that a MIF creates efficiencies within the meaning of Article 81(3) of the Treaty must therefore be founded on a detailed, robust and compelling analysis that relies in its assumptions and deductions on empirical data and facts. MasterCard has not provided such analysis and empirical evidence, only a general assertion that the balancing of the demand of cardholders and merchants through a MIF leads to a better performance of the MasterCard system, is inherent and indispensable to the operation of a four-party payment card system, contributes to overall economic welfare and therefore "undoubtedly" fulfils the first condition of Article 81(3) of the Treaty.

733. In conclusion, the MasterCard MIF does not meet the first condition of Article 81(3) of the Treaty.

8.3 Second condition: Fair share of benefit to consumers

8.3.1 General

734. According to the second condition of Article 81(3) of the Treaty consumers must receive a fair share of the efficiencies generated by the restrictive decision/agreement. The condition that a fair share of the benefits is passed on to consumers implies that the claimed efficiencies must outweigh the anticompetitive effects produced by the agreement. If a restrictive agreement is likely to lead to higher prices, consumers must be compensated through increased quality or other benefits (877).

8.3.2 MasterCard's arguments

735. MasterCard argues that the key direct benefit of its MIF for consumers is to "optimize the use of a service which is characterized by collaborative joint service and joint demand and indirect network effects". Bath cardholders and merchants would benefit from the MIF "through a better functioning system". The MasterCard MIF also leads to more general efficiencies of four party systems, such as "competition among issuers and among acquirers, greater network benefits, greater coverage and overall better performance which, in each case, benefits both groups of users, that is to say merchants and cardholders" (878).

736. MasterCard also believes that "if issuers presented cardholders with the issuers costs and acquirers presented merchants with acquirers" costs, the outcome would be many fewer completed transactions and, ultimately, the demise of the MasterCard payment system" (879). Whether costs initially fall on the issuing or acquiring side would be a product of the terms laid down for the operation of the scheme. The scheme could as well be operated with a different cost allocation. For instance the average free funding period could be reduced or the settlement at the merchant bank account could be delayed (880). MasterCard believes in particular that costs related to the funding of delayed payment, the payment guarantee against cardholder default and fraud, as well as processing costs should legitimately be considered in a cost study for setting the level of interchange fees in the credit and charge card segment (881).

737. In support of its arguments MasterCard submitted a study on "Merchant Acceptance of Payment Cards" of 12 October 2006 (882) which draws from two merchant surveys (883). Moreover, MasterCard also submitted a theoretical model on surcharging credit card sales in a simple model of retail price setting (884).

Euro Commerce's arguments

738. EuroCommerce stated that the MIF makes transactions by payment cards more expensive than they should be. It considers that the MIF would be to the detriment of merchants and consumers (885). All kinds of "discretionary benefits" offered by issuers to cardholders should be excluded from a MIF (886).

8.3.3 Assessment by the Commission

739. Most arguments submitted by MasterCard with respect to the second condition of Article 81(3) of the Treaty are the same as those submitted under Article 81(1) of the Treaty on the absence of a restriction of competition. MasterCard's general arguments on the "the key direct benefit of the MasterCard MIF" (887) which is to "optimise the use of a service that is characterised by (a) collaborative joint service ands (b) by joint demand and (c) indirect network effects remained unsubstantiated. As these arguments have been dealt with in the context of the analysis under Article 81(1) of the Treaty and the first condition of Article 81(3) of the Treaty reference is made to sections 7.3 and 8.2.

740. There is no reason to assume from the outset that an interchange fee paid by acquirers to issuers increases the utility of the payment card system to bath groups of consumers alike (888). The Commission does not dispute that merchants may benefit through enhanced network effects from the issuing side, but this does not necessarily offset their losses which result from paying inflated merchant Fees. In setting a MIF the member banks of a card scheme must guarantee a fair share of the benefits to all customers, not only to those that are on the side of the scheme which receives the MIF. In a scheme where the MIF is paid by the acquirer to the issuer, the efficiencies must in particular counterbalance the restrictive effects to the detriment of merchants (and subsequent purchasers). MasterCard has not submitted evidence in this respect.

741. The Commission has therefore reviewed the methodologies which MasterCard uses as "starting point" for setting the level of the Intra-EEA fallback interchange fees. It has already been stated in section 8.2.2.2 g that these methodologies do not appear to properly assess the willingness to pay on the cardholder side. MasterCard's issuing banks are also not obliged under the network rules to use interchange proceeds from MasterCard's MIF in a particular manner and it remained unclear to which extent MasterCard's issuers indeed pass the fee proceeds on to their customers. However, it can be left open in this case whether cardholders sufficiently benefit from MasterCard's MIF. The Commission's concerns under the second condition of Article 81(3) of the Treaty in this decision relate to the customer group which bears the cost of the MIF, that is the merchants.

742. While merchants may benefit through enhanced network effects from the issuing side, this does not necessarily offset their losses which result from paying inflated merchant fees. The Commission bas therefore reviewed how MasterCard sets an upper limit to its interchange fee rates. MasterCard in [business secret - 6] (see section 3.1.8.3 a). As set out in detail in the Supplementary Statement of Objections (889) this benchmark includes cost elements that are not related to services which sufficiently benefit merchants: (890)

- Intra-EEA Fallback Interchange Fees for consumer credit and charge cards: It remains unproven that merchants benefit from bearing the financial burden of issuers for the provision by issuers to cardholders of also-called "free funding period" (891); moreover, the Commission doubts that merchants sufficiently benefit from bearing the financial costs of issuers writing off bad debts and collecting debts from cardholders.

- Intra-EEA and SEPA Fallback Interchange Fees for debit cards: It remains unproven that merchants benefit sufficiently if debit card interchange fees are inflated by including an issuing banks' costs of setting up and managing current bank accounts.

743. Hence, without further evidence -which MasterCard failed to submit- it cannot safely be assumed that by pursuing its aim of maximising system output MasterCard is creating objective efficiencies that benefit all customers, including those that bear the cost of its MIF (merchants and subsequent purchasers).

744. In reply to the Supplementary Statement of Objections MasterCard insisted on including all above mentioned cost items in MasterCard's cost studies. MasterCard's arguments in this respect are, however, not convincing.

745. MasterCard errs in considering the Commission's analysis of the distinct cost items as an attempt to "deconstruct" the MasterCard payment card (892). The features of those cards are not static. They change over time and differ from one country to another depending on the issuing strategies adopted by local banks (893). In the absence of interchange fees issuing banks retain their free choice to attach certain value added features to a payment card that go beyond its transactional function. However, those features are then priced in a manner that is finally transparent to consumers and subject to the free play of market forces.

746. Finally, a bank in the United Kingdom submitted on 22 December 2006 a study conducted by [...] [consultant] called "evidence on the benefits to merchants form the extension of credit". In the Commission's view that study does not establish that merchants sufficiently benefit from a cost benchmark which includes the cost of a free funding period, that is a period during which a cardholder can make use of free credit. For details of the [...] [consultant] study and the Commission's assessment see Annex 6.

8.3.4 Conclusion on second condition of Article 81(3) of the Treaty

747. The MasterCard MIF does not allow consumers to obtain a fair share of possible efficiencies.

8.4. Third condition : Instability

8.4.1. MasterCard's argument

748. MasterCard's claim that its MIF is indispensable to achieve efficiencies is based on arguments similar to those used to support the efficiency claim in relation to the first condition of Article 81(3) of the Treaty. MasterCard argues that its MIF setting methodology " is designed to lead to an optimal MIF level " (894) and that the elements of that methodology, that is to say, the use of MasterCard's discretion in setting the cross-border interchange fees, the use of a " proxy for the merchants' elasticity of demand " and " the competitive environment of the payment system market ", ensure that the fees were set at the " optimal " level. (895)

749. MasterCard does not demonstrate that its MIF is efficient or optimal argues " that there is no reason to suspect that the level at which the MIF is currently set is not the optimum one ". (896)

750. MasterCard also seems to argue that its MIF causally contributes to maintaining the MasterCard network at its present size, and that a reduction in the level of the MIF would " be expected to result in significant contradiction, if not the disintegration, of the system as it currently exists " (897) and would " hamper the development of e-commerce ". MasterCard does not offer any factual or empirical evidence for any of these conjectures.

8.4.2 Assessment by the Commission

751. The assessment under Article 81(3) of the Treaty of the indispensability of the restrictions refers to whether the restrictions contained in the decision are or are not indispensible to attain the efficiencies that may result from the restrictive decision. Whatever efficiencies are claimed to result from a MIF, it has to be demonstrated that the restrictive effects of the MIF are indispensable to achieve the claimed efficiencies. In the present case MasterCard has argued that its MIF tends to maximise system output and that certain objective efficiencies may result from this effect of its MIF. MasterCard has, however, not proven to the requisite standard that its current MIF is indeed indispensable to maximise system output and to achieve any related objective efficiencies. MasterCard did not even provide empirical evidence on the actual effects of its MIF in the market regarding system output. However, several payment card schemes in the European Economic Area have successfully been operating without a MIF for a long time. These schemes have been established between 1979 and 1992 and they are not merely viable but indeed successful. ECB statistics indicate that in EEA Member States where domestic payment card schemes operate without a MIF, card usage per capita is particularly pronounced on a pan-European average (see section 8.2.2.2.). There are also concerns with regard to the objectiveness of MasterCard's methodologies for setting the direction, the structure and the level of the intra-EEA fallback interchange fee rates (see section 8.2.).

752. In conclusion, the MasterCard MIF does not fulfil the third condition of Article 81(3) of the Treaty.

8.5. Conclusion on the assessment of Article 81(3) of the Treaty and Article 53(3) of the EEA Agreement

753. In overall conclusion, MasterCard has not demonstrated that its MIF fulfils the first three conditions of Article 81(3) of the Treaty/Article 53(3) of the EEA Agreement.

9. Addressees of the decision

754. This decision should be addressed to the legal entities managing and acting on behalf of the MasterCard payment organisation, that is to say MasterCard Incorporated, MasterCard International Incorporated and MasterCard Europe Sprl. The documentation submitted to the Commission suggests that MasterCard Incorporated controls MasterCard International Inc. Who in turn sets Bylaws and Rules that include the HACR and are the basis for MasterCard's entire interchange fee mechanism. MasterCard International Inc. also figures as publisher of the Interchange and Service Fees Manual, which specifies the different interchange fee programs. Finally, MasterCard International Inc. Controls the deduction of the " correct " interchange fees wherever it processes cross-border and/or domestic payment card transactions. Since September 2006 [...] [business secret - 2a] MasterCard Incorporated also sets the level of intra-EEA Fallback Interchange Fees and of SEPA Fallback Interchange Fees.

755. In addition, this decision should be addressed to MasterCard Europe Sprl because it is also the legal successor of Europay International SA who in turn is the legal successor of EuroCard International S.C. and Eurocheque International Holding SA, who had filled the initial notification in case COMP 34 324, 34 579 and 35 578. MasterCard Europe Sprl is also the addressee of this decision, because until September 2006 its management prepared decisions of the European Board on Intra-EEA fallback interchange fees.

10. Remedies and fines

10.1 Article 7(1) of Regulation 1-2003

756. Where the Commission finds that there is an infringement of Article 81 of the Treaty and Article 53 of the EEA Agreement, it may, in accordance with Article 7 (1) of regulation (EC) No 1-2003, require by decision the undertakings and associations of undertakings concerned to bring such an infringement to an end. According to Article 5 of Council Regulation (EC) No 28-94 of 28 November 1994 concerning arrangements for implementing the Agreement of the European Economic Area (898) " the Community rules giving effect to the principles set out in Article 85 and 86 [now Articles 81and 82 ] EC [...] shall apply mutatis mutandis " in respect of the EEA.

757. A decision pursuant to Article 7(1) of Regulation (EC) No 1-2003 may impose on the undertakings and associations of undertakings concerned to bring such infringement to an end. For this purpose, it may impose on them any behaviour or structural remedies which are proportionate to the infringement committed and necessary to bring the infringement effectively to an end. The Commission may moreover require the undertakings and association of undertakings concerned to submit to it proposals with a view to bringing the situation into conformity with the requirements of the Treaty (899).

758. The requirement that a remedy has to be effective (900) empowers the Commission to oblige an undertakings or an association of undertakings to refrain from adopting any measures having an equivalent effect as the conduct identified as restrictive.

10.1.1 Remedy

759. In order to remedy the restriction of competition by MasterCard Incorporated, MasterCard International Incorporated and MasterCard Europe Sprl these undertakings should be obliged to cease and desist from determining in effect a minimum price merchants must pay for accepting payment cards by way of setting Intra-EEA fallback interchange fees.

760. This obligation should exclude MasterCard's Intra-EEA fallback interchange fees as far as they affect commercial credit and charge cards, because the Commission has not yet finalised its investigation of possible efficiencies in that regard.

761. Accordingly, the legal entities representing the MasterCard payment organisation should be require to modify the organisation's network rules on interchange fees and to repeal all relevant decisions of the European Board and of MasterCard [...] [business secret - 2a] with respect to Intra-EEA fallback interchange fees (except for the fees affecting commercial credit and charge cards), SEPA fallback interchange fees and Intra-Eurozone fallback interchange fees.

Timing

762. The MasterCard payment organisation should have sufficient time to modify its network rules. The order therefore does not immediately take effect upon notification but only six months thereafter.

Information

763. For this decision to be effective it is necessary that member banks, clearing houses and settlement banks are informed in a timely and adequate manner of the changes to the organisation's network rules.

764. In the light of the foregoing, the legal entities representing the MasterCard payment organisation should be required to inform all member banks, all clearing houses and all settlement banks in European Economic Area (with currently clear and settle POS payment card transactions with MasterCard or Maestro branded payment cards) that inter-bank claims and debts resulting from POS payment card transactions with MasterCard and Maestro branded payment cards do not give rise to the payment of Intra-EEA fallback interchange fees (or SEPA fallback interchange fees or of Intra-Eurozone fallback interchange fees) as of the day the Commission's cease and desist order takes effect six months after the notification of the present decision.

765. To remedy the competitive harm in the market as fast as possible it is important that aquiring banks pass-on the cost reductions resulting from the absence of MasterCard's Intra-EEA fallback interchange fees to their customers. This presupposes in turn that merchants are adequately informed that their banks are now in a position to offer considerably lower merchant fees.

766. If on the day the Commission's cease and desist order takes effect six months after the notification of the present decision, no non-confidential version of the present decision is publicly available, the legal entities representing the MasterCard payment organisation should publish the information set out in Annex 5 on the internet. This information shall remain on MasterCard's website until a non confidential version of the decision is publicly available. During this time period MasterCard is to establish on the first page of each country specific website of MasterCard for Member States of the European Economic Area (e.g.: http://www.mastercard.com/be/ for Belgium) a clearly visible hyperlink to the information set out in Annex 5. MasterCard Incorporated and MasterCard International Incorporated should provide a print out of the first page of MasterCard's country specific websites in the EEA to the European Commission.

Remedy is sufficiently determined, adequate and proportionate

767. The requirement on MasterCard to cease and desist from setting Intra-EEA fallback interchange fees is sufficiently determined, necessary and proportionate to remove the competitive harm.

768. The obligation on MasterCard to inform member banks, clearing houses and settlement banks is necessary and proportionate, because without such information there is a real risk that mere internal change of MasterCard's network rules would not, at least not immediately, be known to the stakeholders in the market, thereby prolonging the effects of MasterCard's Intra-EEA fallback interchange fees in the market.

769. The obligation on MasterCard to publish the information referred to in Annex 5 on the internet is also the necessary and proportionate, because that information will enhance the information available to merchants until the publication of a non confidential version of decision. This will in turn speed up the pass-on of the acquirers' gains resulting from the absence of MasterCard's Intra-EEA fallback interchange fees to merchants and their customers.

770. The scope of the decision should exclude interchange fee arrangements in the MasterCard payment organisation that were not subject to the Commission's proceedings. There are bilateral agreements on interchange fees and multilateral agreements between delegates on a nation forum of MasterCard member banks that determine an interchange fee for domestic POS payment card transactions within a specific EEA Member State. Several of these multilateral agreements between delegates on a national forum of MasterCard member banks are also subject to national antitrust investigations.

771. The six months period granted to the legal entities representing the MasterCard payment organisation to implement this Decision is appropriate, because MasterCard may have to review, modify and re-publish its network rules in order to render the remedy effective. The period is proportionate as it may be necessary for MasterCard Incorporated to invoke an extraordinary meeting of its Global Board (and -if required- of its stockholders) in order to bring about the necessary rule changes. The period is also sufficient to enable MasterCard Incorporated to inform all member Banks, clearing houses and settlement banks of the rule changes.

11. Article 23 (2) Regulation N° 1-2003 - fines

772. As the MasterCard MIF has been notified to the Commission and given the specific circumstances of the present case no fine should be imposed pursuant to Article 15(5) of Regulation No 17-62 and Article 23(2) of Regulation (EC) No 1-2003.

12. Article 24 (1) Regulation No 1-2003 - periodic penalty payments

773. Under Article 24 (1) (a) of Regulation (EC) 1-2003, the Commission may, by decision, impose on undertakings or associations of undertakings periodic penalty payment not exceeding 5% of the average daily turnover in the preceding business year per day and calculated from the date appointed by the decision, in order to compel them to put an end to an infringement of Article 81 or Article 82 of the Treaty, in accordance with a decision taken pursuant to Article 7 of Regulation 1-2003.

774. The serious risk that MasterCard continues to apply Intra-EEA fallback interchange fees to payment card transactions in its scheme, that MasterCard implements the SEPA fallback interchange fees or the Intra-Eurozone fallback interchange fees or that it attempts to take measures which will effectively circumvent the remedy is sufficient ground for considering it necessary to impose periodic penalty payments on the legal entities representing the MasterCard payment organisation to ensure compliance with the remedy. The imposition of a periodic penalty payment in a decision according Article 7 of Regulation (EC) no 1-2003 is in line with the Commission's long-standing and undisputed decisional practise to impose periodic penalty payments pursuant to Article 24 (1) (a) of Regulation 1-2003 already in the decision finding an infringement of Article 81 or Article 82 EC and requiring it to be brought to an end. (901)

775. In setting the level of the periodic penalty payments, the Commission has taken into account the necessity of imposing periodic penalty payments sufficient to ensure MasterCard's compliance with the present decision. It is necessary to set periodic penalty payments at a level which reinforces the incentive to comply with a decision taken pursuant to Article 7 of Regulation No 1-2003 by rendering it economically rational for the undertaking concerned to comply with such a decision rather than to reap the benefits of non-compliance. Similarly, the Commission must take into account the necessity of setting periodic penalty payments which are proportionate and sufficient to compel compliance from undertakings such as MasterCard, with its substantial size and financial resources. It also has to be considered that MasterCard already attempted to hamper antitrust enforcement with respect of its MIF through the measures taken in the context of MasterCard's Initial Public Offering on 25 May 2006 (see section 2.1.4).

776. For all these reasons the Commission fixes the daily periodic penalty payment at 70% of the maximum amount which is in turn 5% of MasterCard's daily turnover in the preceding business year. This means that the period penalty payment is set at 3.5% of MasterCard Incorporated's daily consolidated global turnover in the business year preceding an infringement of the Commission's order.

Has adopted this decision:

Article 1

From 22 May 1992 until 19 December 2007 the MasterCard payment organisation and the legal entities representing it, that is MasterCard Incorporated, MasterCard International Incorporated and MasterCard Europe Sprl, have infringed Article 81 of the Treaty and from 1 January 1994 until 19 December 2007, Article 53 of the EEA Agreement by in effect setting a minimum price merchants must pay to their acquiring bank for accepting payment cards in the European Economic area, by means of the intra-EEA fallback interchange fees for MasterCard branded consumer credit and charge cards and for MasterCard or Maestro branded debit cards.

Article 2

The MasterCard payment organisation and the legal entities representing it shall bring at an end the infringement referred to in Article 1 in accordance with the subsequent Article 3 to 5.

The MasterCard payment organisation and the legal entities representing it shall refrain from repeating the infringement through any act or conduct as described in Article 1 having the same or equivalent object effect. They shall in particular refrain from implementing the SEPA/the Intra-Eurozone fallback interchange fees.

Article 3

Within six months after notification of this decision the legal entities representing the MasterCard payment organisation shall formally repeal the Intra-EEA fallback interchange fees, as well as the SEPA/Intra-Eurozone fallback interchange fees. They shall moreover modify the association's network rules to reflect this order and the order according to Article 2 second paragraph. They shall repeal all decisions taken by MasterCard's European Board and/or by MasterCard [...] [business secret - 2a] Intra-EEA fallback interchange fees on SEPA fallback interchange fees and on Intra-Eurozone fallback interchange fees.

Article 4

Within six months after notification of this decision the legal entities representing the MasterCard payment organisation shall communicate all changes of the association's network rules and the repeal of decisions referred to in Article 3 to all financial institution holding a license for issuing and/or acquiring in the MasterCard payment organisation in the European Economic Area and to all clearing houses and settlement banks which clear and/or settle POS payment card transactions in the MasterCard payment organisation in the European Economic Area. A copy of this correspondence shall be provided to the European Commission to the attention of the Director General of the Directorate General for Competition within seven months after notification of this decision.

Article 5

The legal entities representing the MasterCard payment organisation shall publish the information set out in Annex 5 of this decision on the internet as from the first day after the six months implementation period of the order in Article 3 has lapsed and until the first day following the publication of a non confidential version of the present decision on Commission's website and/or in the Official Journal of the European Union, whichever comes first. During this time period the legal entities representing the MasterCard payment organisation shall publish a clearly visible hyperlink to this information on the first page of MasterCard's website for each Member State in the European Economic Area. Within a week after publishing the information in Annex 5 they shall moreover provide a print out of the first page of these country specific websites as well as of the published information to the European Commission to the attention of the Director General of the Directorate General for Competition.

Article 6

The Commission may at its sole discretion and upon a reasoned and timely request by MasterCard grant an extension of the various time limits provided for in the preceding articles. The Commission may at its sole discretion and upon a reasoned and timely request by MasterCard also agree to an adequate alternative of providing merchants with the information set out in Annex 5 of this decision other than a publication on MasterCard's website or to a modification of the text in Annex 5.

Article 7

If the legal entities representing the MasterCard payment organisation fail to comply with any of the order set out in the preceding Article 2 to 5 of this decision, the Commission hereby imposes a daily penalty payment on the legal entities representing the MasterCard payment organisation of 3.5% of MasterCard Incorporated's daily consolidated global turnover in the preceding business year according to Article 24(1) a of regulation (EC) No 1-2003. This penalty will be calculated as from the first day after the infringed order takes effect.

Article 8

The Decision is addressed to MasterCard Incorporated, 2000 Purchase Street, Purchase NY 10577, USA, to MasterCard International Incorporated, 2000 Purchase Street, Purchase NY 10577, USA, and to MasterCard Europe Sprl, Chaussée de Tervuren 198A, B-1410 Waterloo.

This Decision shall be enforceable pursuant to Article 256 EC and Article 110 of the EEA Agreement.

---

Annex 1

MasterCard Intra-EEA Fallback Interchange Fees for credit and charge cards

< table >

[business secret - 2a]

Maestro intra-EEA Fallback Interchange Fees for debit cards

< table >

SEPA Fallback Interchange Fees for Maestro debit cards (announced 4 December 2006, not effective

< table >

Intra-Eurozone Fallback Interchange Fees for Maestro debit cards (announced 8 November 2007 and effective as of 15 January 2008)

< table >

[business secret - 2a]

Annex 2

Analysis of MasterCard's methodological criticism regarding the Commission's merchant market survey

1. Sampling and weighting

1.1 Sampling and number of requests sent

First, MasterCard doubts the validity of the Commissions empirical findings in the merchant market survey (904), because of an over-representation of large companies in the sample of respondents, as well as a bias towards merchants that accept cards and because of the "low number of requests" issued and the "even lower number" of replies obtained (905).

2. In terms of the respondents' size, the Commission expressly made clear that its sample over represents large companies (906). 65% of the 218 valid responses received in the Commission's merchant market survey came from companies with more than EUR 50 Mio animal turnover. These companies are qualified as "large companies". 17% of the respondents were companies with annual turnovers between EUR 10 to 50 Mio, 10% had a turnover of EUR 2-10 Mio and the remainder had a turnover of below EUR 10 Mio. As set out in the first chapter of its report on the merchant market survey, there are several reasons for the sample bias in favour of larger companies. First, the merchants operating in four of the seven merchant segments surveyed by the Commission are typically larger companies (airlines, petrol station chains, supermarkets, hotels). These segments were in turn selected on the basis of MasterCard's own information on the relative importance of merchant segments for cross-border payments with MasterCard and Maestro branded cards (907). Second, the bias in favour of larger companies is also justified because multinationals are more likely to engage in cross-border payment card transactions than small merchants. This is of importance, because the MasterCard MIF applies in particular to this type of payments.

3. MasterCard's consultants also attribute particular importance to large merchants in their survey sample based on the opinion that these merchants are "of disproportionate importance" for the questions at stake (908).

4. In conclusion, the Commission's survey bias towards larger companies does not invalidate the importance of the survey results, as MasterCard claims. The over-representation of larger merchants is the result of a careful and objective sampling technique that is in turn based on MasterCard's own market information. If the bias affects the survey results at all, it can only play in favour of MasterCard, because larger merchants have stronger market power than smaller merchants and are therefore rather in a position to exert countervailing buying power with regard to the MasterCard MIF than smaller undertakings.

5. Second, MasterCard criticises that most respondents to the market survey accept payment cards (909). This criticism is misguided. Questionnaires were sent to all kind of merchants. The Commission's aim in compiling the sample was to target in particular larger merchants as they generate the highest card volumes and are therefore representative in terms of card volumes. In pursuing this aim, the Commission asked MasterCard acquiring banks to provide addresses of large merchant customers. It is natural that such merchants accept cards, because otherwise they would not be clients of an acquiring bank. On the other hand the Commission also compiled a random sample of smaller and medium sized merchants from a public data base to balance the sample of the 400 merchants which were chosen as addressees of these 400 merchants eventually 218 replied. Further details of the Commission's sampling technique are set out in section 1 of the merchant market survey (910).

6. Third, as to the number of requests sent, the Commission sent out approximately 400 requests for information under Article 11 of Regulation 17-62 and Article 18 of Regulation 1-2003 (911). This is an exceptionally high number of formal requests for information in an antitrust procedure. The Commission chose this method to obtain data rather than "outsourcing" it to a telephone interview company, because telephone conversations with merchants did not appear to yield sufficiently reliable results. Merchants might reply without due care simply to save time and to "get rid" of the interviewer. It is also crucial to ensure that the person replying for a merchant is not simply the first one who takes the phone but a knowledgeable person such as the CEO or CFO. As merchants replied to formal requests for information, the Commission services can safely assume that the replies are accurate as addressees of a request for information are always made aware, in the cover letter of the request, that the Commission can sanction inaccuracy, even negligent one, with fines. The overall high quality of the replies received confirms this.

7. Finally, MasterCard's view (912) that the Commission should not have disclosed its identity in asking questions to merchants is absurd. Obviously, any request for information under Article 11 of Regulation 17/Articlel8 of Regulation 1 must necessarily inform the addressee of the Commission's identity as the Commission is bound to reveal the legal basis and the purpose of its request (913). MasterCard's claim implies that the Commission should not seek information under Article 81 of the Treaty with a key instrument foreseen for this purpose in the law, that is to say the request for information.

1.2 Weighting

1.2.1 Stratification

8. MasterCard alleges that Commission survey results are "not re-weighted in order to make their representative for the merchant population". The Commission would therefore have over-sampled large merchants "in a non-systematic way" (914). MasterCard's consultant used "sampling stratification techniques" in trying to correct the over-representation of large merchants in its merchant survey (915).

9. Stratification is the process of grouping members of a population (that is to say a group of persons interviewed) into relatively homogeneous subgroups before sampling. While a stratification process might improve the "representativeness" of a sample if accurately done, it also presents a number of dangers. First, it may be difficult to select relevant stratification variables. For instance, the criterion chosen by MasterCard's consultants to stratify the sample of large and small merchants according to the number of employees is inadequate for the relevance of a merchant in terms of payment card acceptance (916). Second, stratification also requires accurate information about the population (e.g.: size). Third, stratification is not useful if there are no homogeneous subgroups. Thus, once stratification criteria are badly selected the stratification process rather increases than corrects the sample bias. Alternatively to stratification, one can account for a sample bias in the analysis of a survey's results. This was the Commission's approach (917).

1.2.2 Weighting

10. Weighted averages may be more informative than simple arithmetic averages of replies received. However, the weighting of averages in turn also creates risks of distorting the survey results if the weighting process is done inadequately. Again, this is the case for the MasterCard's study as the "weight" in the merchant survey, the number of merchant's employees, is not an appropriate criterion to produce reliable country and sector "weighted" averages (918). Under these circumstances, it is unclear how the use of a weighted average, such as the one used by MasterCard would improve rather than distort the result of the market survey. The Commission chose to apply a simple arithmetic average, which is the less risky option. MasterCard also did not specify how the use of weighting would yield results that are significantly different (and more reliable) as compared to the use of a simple average in the Commission's market survey. Again, if larger merchants are over-represented in the sample, this only plays in favour of MasterCard. The countervailing market power of large merchants is greater than that of small merchants.

2. Questioning technique

11. MasterCard claims that the Commission's questionnaire was "poorly designed" and that conclusions were "not supported by evidence". MasterCard argues that robust results "necessitate easy question?' that do not leave scope for interpretation. For complex questions interviews should be undertaken "face-to-face, or at least over the phone" in order to provide assistance. The Commission, by contrast, would have "mailed an arguably complex questionnaire to merchants without providing any point of contact should respondents need clarification" (919).

12. This criticism is misplaced for the following reasons. First, MasterCard's questionnaire in its merchant market survey is at least as complex as the questionnaire used by the Commission (920). The Commission also does not consider that merchants surveyed were overburdened with the questions it asked. Second, it is incorrect that "no point of contact" was provided to merchants in the questionnaires to obtain clarifications on questions asked. If MasterCard had carefully read the Commission's questionnaires before criticising them, it would have realised that in the cover letter of each request for information the Commission services had precisely indicated such points of contact (921).

13. MasterCard moreover argues that questions 7 and 8 in the Commission's merchant market survey on reasons why merchants accept payment cards contained "certain elements" that would "taint the reliability" of the responses (922). This argument is, however, unsubstantiated and untenable.

14. First, MasterCard assumes without further explanations that merchants are simply "not able to distinguish between different types of cards" (923). The relevance of this unsubstantiated observation for analysing the validity of responses to questions 7 and 8 is unclear. Both questions in the' merchant market survey referred to "payment cards" in general and did not distinguish between credit and debit yards or between different brands of cards (924). The lack of weight of this criticism becomes apparent if one compares the Commission's questioning technique with the one used by MasterCard in the merchant market survey. In question Q12 of the questionnaire MasterCard's consultant asked merchants to indicate the reasons why they accept debit cards. In the subsequent question Q13 the question is repeated for credit cards and in question 13 for store cards. In MasterCard's own words, such approach is flawed, because it cannot reasonably be assumed that merchants are able to distinguish a debit card from a credit card. Apparently MasterCard and its expert have different perceptions on the capacity of merchants to distinguish between different types of yards.

15. Second, MasterCard criticises that the Commission had offered merchants a long list of "preformatted answers" rather than asking "open questions" on the importance of payment cards. However, if the Commission had followed this suggestion it would have been impossible to quantify in a statistically relevant manner the importance of each reason mentioned. Moreover, MasterCard seems to ignore that its own consultants chose the same approach as the Commission did in their merchant market survey where merchants had to choose among a list of "preformatted reasons" as to why they accept payment cards (925).

16. As to the number of reply options in Q 7 and 8, the Commission observes the following.

17. First, MasterCard's consultants asked merchants to choose among eight options (in Q 12 to 14 of its merchant market survey), while the Commission offered 12 and 13 options respectively in Q 7 and 8 of the Commission's survey. It is unclear why the "critical threshold" for a reasonable number of reply options should be eight rather than twelve or thirteen.

18. Second, in MasterCard's survey merchants had to decide on the spot over the phone which of the eight options they agreed with. To the contrary, the Commission gave merchants sufficient time to reply in writing to reply to its questionnaire. Merchants could therefore carefully contemplate and balance their options, which is not necessarily the case in a telephone interview.

19. Third, the difference of four more reply options in the Commission's survey (Q 7 and 8) as compared to Q 12 - 14 in the MasterCard's survey is also a plus rather than a minus. The Commission requested merchants to indicate whether they view technical differences of payment cards as opposed to other payment means (cheques and cash) as important element of their decision to accept cards. MasterCard entirely excluded this aspect from the ambit of its questionnaire. This is a blatant omission. MasterCard argues that even though a high percentage of merchants in the Commission's merchant market survey had indicated that they never cancelled a card this "does not indicate a lack of price responsiveness (926). In its reasoning leading to this conclusion MasterCard, amongst others, bases itself on the assumption that "the only reason which is specific to the acceptance of cards pertains to the transactional benefits related to the use of such cards" (927) which is opposed to the Commission's findings in the merchant market survey that transactional benefits are of limited importance for card acceptance (928). MasterCard was not in a position to provide a source reference for its assumption. This is not surprising, as MasterCard had precisely excluded reply options in Q 12 to 14 that would have allowed merchants to compare relative transactional benefits of payment means in their replies.

20. Finally, MasterCard's suggestion to randomise the order of reply options in questionnaires to avoid that merchants view the first reply option as the most important one is a suggestion that MasterCard itself did not adhere to. The instruction for the interviewers in the MasterCard's merchant market survey do not contain any instruction for the interviewing person to randomise the order of reply options 1 to 8 in Q 12 to 14. At any rate as merchants had sufficient time to reply to the Commission's questionnaire, and as typically the Commission addressed knowledgeable staff (the CEO or CFO of a merchant), it can be assumed that the responding person was able to understand that the first option on top of questions 7 and 8 is not necessarily the most important one for its business.

3. Individual points of criticism

21. MasterCard sees a contradiction in merchants' replies to Q 7 and 9 of the Commission merchant market survey that would undermine the Commission's conclusion that the more cardholders use MasterCard the more card acceptance becomes a "must". This criticism is not convincing.

22. The Commission's conclusion on merchant captivity is in particular supported by the following empirical elements of the merchant market survey. In reply to Q 4 of the survey 91% of the respondents observed that they never cancelled a payment card once they started accepting it (229). In reply to Q 7 merchants indicated that the most important reasons to accept cards are not related to transactional benefits (cheaper/faster/more secure than cash or cheques) but rather linked to the fact that most customers carry a particular card (90%) and the need to maintain an image as customer friendly company (67%) (930).

23. As to the perceived contradiction between replies to Q 7 and Q 9 it is correct that 35% of respondents replied to Q 9 that they considered themselves not unable to refuse a card once the number of cardholders had increased above a certain threshold. However, more respondents (41%) replied in the affirmative. MasterCard seems to assume that 41% of respondents are not a significant number. This is wrong. If 41% of the merchant communities are assumed to have to accept cards as opposed to 35%, then the majority of merchants consider payment card acceptance as "a must" or -in MasterCard's words "merchants are transaction hungry" (931). Thus, contrary to MasterCard to results of Q 9 do not undermine the conclusion that the more frequently particular payment cards are used, the more it becomes necessary for a merchant to accept them.

24. MasterCard's finding of a "relative low acceptance of payment cards across Europe" is no evidence for the absence of merchant captivity. It is not because some consumers do not buy a certain product/service that those who do are in a position to exercise sufficient constraints to avoid a price increase above supra-competitive levels. In particular it does not disprove the Commission's finding that the majority of merchants accept cards, in particular because they risk losing business if they did not ("business stealing effects") (932).

4. Conclusion

25. In conclusion the Commission discards MasterCard's criticism with respect to the Commission's merchant market survey as unfounded.

Annex 3

Analysis of the MasterCard's " Expert Report " and the Research on the use of retail credit

[business secret - 8]

Annex 4

Analysis of [...] prof. [...]

Export report on behalf of MasterCard

51. The report [...] prof. [...] is presented in the form of an annex (Annex i to MasterCard's response to the Supplementaty Statement of Objections) (982). This Annex i provides a "critique of certain sections of the Supplementary Statement of Objections" while sub annexes 1.1, 1.2 and 1.3 develop some of the arguments included in Annex 1.

52. Annex I addresses four questions: (i) whether the MasterCard MIF is a restriction of competition, (ii) whether a SSNIP test has to be performed on the total service charge (cardholder charge plus merchant charge), (iii) whether merchants can discourage the use of payment cards and thereby impose a competitive pressure on MasterCard to keep prices low and (iv) what would happen to prices if the Commission forced MasterCard to reduce the MIF.

53. The conclusions of annex 1 can be summarized as follows:

The interchange fee as set by MasterCard would tend to maximise output and thus would net be a restriction of competition. The report concludes that "provided that the "pass on ratio" is approximately equal on the issuing side and on the acquiring side (symmetric pass on ratios) we can characterise the volume maximising interchange fee. It is the one which also maximises the total profit from the .system (page 9)". The report also claims that "there is no valid indication that a different interchange fee, in particular a zero interchange fee would imply a lower total service charge (= cardholder charge plus merchant charge) for the system (page 1)".

- The SSNLP-test needs to be performed on the total service charge (= cardholder charge plus merchant charge).

- The Commission is wrong in denying the ability of merchants to surcharge the use of payment systems that they perceive to be too expensive. The Commission misinterprets the results of its own merchant survey. While merchants would prefer to use price differentiation by means of payment, the administrative cost implied by this price differentiation explains why "surcharging" is so infrequently used. In conclusion, the high level of price coherence is not an indication of MasterCard market power.

- If the Commission were right that merchants cannot successfully surcharge the use of the expensive payment systems, then a forced reduction in the interchange fee would raise the total service charge and would disadvantage small merchants vis-à-vis large merchants.

Commission assessment of the expert report [...] by prof. [...]

1. Regarding the claim .that the interchange fee as set by MasterCard tends to maximise output and thus is not a restriction of competition

54. The model that [...] prof. [...] presents to derive the welfare effects is very simplistic in nature, leaving aside many important features of credit card markets and wrongly specifying others. In particular, the model is designed to represent a three-party system while MasterCard operates a four party system (983). This is an important limitation as it well established that an interchange fee is a redundant instrument in three party schemes, which is not the case in four party schemes. The model therefore lacks relevance for the case at hand.

55. There are additional weaknesses of the model that invalidate the conclusions within the chosen setting. A critical assumption of the model is that there is a symmetric pass-on ratio for merchants and cardholders (i.e., that issuing banks and acquiring banks pass on changes in the interchange. fee to. their customers in exactly the same proportion). This is implausible and also not reflected by the Australian experience, where the decrease in the interchange fee that was brought about by regulation, was passed on in a very asymmetric fashion on the issuing side and on the acquiring side.

56. The theory developed by Rochet and Tirole (984), to whose research [...] by prof. [...] refers regarding other aspects of competition, features asymmetric pass-on ratios with a 100% pass-on on the acquiring side and a 0-100% pass-on on the issuing side. The authors state that on the issuer side, "there is some cost absorption" (p. 556), while on the acquirer side "they just pass through the interchange fee to the merchants" (p.557).

57. Without the unrealistic assumption of a symmetric pass-on ratio, however, the interchange fee has an impact on the "total service charge" in [...] by prof. [...]'s model. Hence, firms will not use it to maximise output (as the MIF influences both output and the total service charge); i.e., the result is invalid.

58. It is instructive to investigate which outcome [...] by prof. [...]'s model produces if an asymmetric pass-on ratio à la Rochet and Tirole is incorporated. In the model, a lower pass-on on the issuer side means that a>b (p. 41). By the third equation on p. 41, this implies that a higher interchange fee directly leads to a higher total service charge.

59. In line with the above argument, it is not surprising that [...] by prof. [...]'s conclusion that the interchange fee maximises output is strongly refuted by empirical evidence. Indeed, as the Commission has shown, in all four Member States where payment card schemes operate without a MIF, card transactions per capita are considerably higher than the average of card transactions of all Member States (see the diagram in section 8.2.2.e).

60. Finally, even if [...] by prof. [...]'s argument were not based on implausible assumptions and refuted by empirical evidence (which it is), a maximisation of output is not a goal in itself, insofar as it may not coincide with fostering competition and promoting consumer welfare, that is benefits to the customers of the system (cardholders, merchants and subsequent purchasers) (985). As is apparent from the economic literature on payment card systems, output maximisation does not necessarily go hand in hand with the maximisation of consumer welfare. To refer to a concrete example, see again the model by Rochet and Tirole, where output maximisation can lead to a reduction of both consumer welfare and total welfare (986). To the extent that a large number of card transactions is artificially subsidized by merchants and non cardholding consumers, maximisation of output will not be in line with the maximisation of consumer welfare. Therefore, even if MasterCard had convincingly shown that the MIF is set so as to maximise output (which is not the case), this would by no means imply that it has also been shown that the MIF maximises consumer welfare and that the restriction of competition on the acquiring aide should therefore be accepted.

61. In summary, the argument that "the interchange fee as set by MasterCard tends to maximise output and thus is not a restriction of competition" should be dismissed as it is unfounded both on theoretical and empirical grounds.

2. Regarding the allegations that the Commission is wrong in denying the ability of merchants to surcharge the use of payment systems which they perceive to be too expensive and that the high level of price coherence is not an indication of MasterCard market power.

62. According to [...] by prof. [...]'s report, "other things equal, two products, which have different marginal costs, in all likelihood are sold at two different prices" (page 31), and so "the Commission has not shown that the loss in demand implied by surcharging MasterCard is greater thon the gain in demand by the price reduction for users of other payment systems implied by this surcharging" (page 31). In other words, [...] by prof. [...] suggests that merchants will not be afraid to surcharge, because even though they may loose turnover from credit cardholding customers they may (more than) offset these losses due to increased sales which result from an overall decrease of the merchants' retail prices. Chapter D of the paper supports this result (987).

63. lit theory, merchants' willingness to surcharge MasterCard users will depend on a number of factors including the size of transaction costs of surcharging, the intrinsic value that consumers attach to paying by card, the degree of consumer information about which stores accept which cards, and the expected surcharging strategy of other merchants. Hence, economic models will necessarily produce differing results, which will depend on the specific circumstances at hand.

64. Therefore, economic theory tells us that whether surcharging is profitable will depend on the specifics of the situation. In other words, this is an empirical question. As things stand, the vast majority of merchants does not surcharge the use of MasterCard, suggesting that surcharging card usage generally is not a profitable strategy to counter the increase in merchant fees that the MIF generates, contrary [...] by prof. [...]'s allegations.

65. This allows MasterCard to increase merchants' fees beyond the level that would arise absent an interchange fee as abandoning acceptance of MasterCard as such is no viable option for a large part of merchants given that other merchants accept the card.

3. Regarding the claim that aims the Commission were right that merchants cannot successfully surcharge the use of expensive payment systems, then a forced reduction in the interchange fee, as envisaged by the Commission, would raise the total service charge and would disadvantage small merchants vis-à-vis large merchants.

66. The report develops a model that aims at showing that any reduction of the current level of MasterCard's MIF would lead to an increase "on the average total service charge" (i.e. the sum of cardholder fees and merchant fees) whenever merchants are not responsive to changes in the merchant discount. In other words, a hypothetical decrease in the interchange fee would lead to an increase of the cardholder fee that would be larger than the decrease of the merchant service fee and thereby increase the total service charge for both consumer groups.

67. The model is based on a substantial misunderstanding on [...] prof. [...]'s side of the Commission's argument. The fact that merchants refrain from surcharging MasterCard users of course does not mean that merchants would not pass on part of the cost of an increase in the merchant discount (which in turn was induced by an increase in the MIF). It simply means that pass on occurs through a general increase of retail prices rather than through selective surcharges for card users. As a consequence, a hypothetical reduction of a MIF in all likelihood leads to lower costs for merchants' customers.

68. Therefore, the Commission stresses that it does not consider merchants to be completely irresponsive towards changes in the merchant discount (which would imply dx/dq=0 in [...] prof. [...]'s model). Rather, we consider that the burden of higher interchange fee is shared by merchants and the entire group of consumers, including cash users.

69. As the results of an increase in the total service charge and of a discrimination of small merchants rely on the erring presumption that the Commission considers merchants to be fully non-responsive to input prices at all price levels (which, we agree [...] by prof. [...] would be absurd), they are both void.

4. Regarding market definition and the proposal to perform a SSNIP test on the total service charge

70. [...] by prof. [...] criticizes the Commission's approach to market definition and argues that the relevant market must be the same for consumers and merchants. He proposes to perform a SSNIP test on the sum of prices that cardholders and merchants face.

71. It is important to note that the SSNIP test on total prices was proposed by Emch and Thomson (to whose paper [...] by prof. [...] refers) in order to assess competition between payment card networks. The authors suggest that this test may then be uses to investigate possible restrictions of inter-system competition (which would, for instance, result from hypothetical price agreements between MasterCard and Visa). [...] by prof. [...] neglects, however, that the primary question in the present investigation is not in how far the MIF affects competition between payment systems, but in how far it affect competition within the system, i.e. the way in which the MIF allows to restrict competition on one side of the industry in order to increase profits to the detriment of consumers. What is relevant is the internal organisation of the scheme, i.e., the degree of competition between banks that issue or acquire for MasterCard.

72. If anything, the possibility of competitive pressure from other payment card schemes rather tends to reinforce the anticompetitive problems that are generated by a collectively set MIF. This is because inter-system competition may in fact reinforce the desire to increase the interchange fee as issuing banks are more likely to promote cards which offer prospective income from interchange fees.

Annex 5

"Information to be published on the website of MasterCard as set out in Article 5 of the decision"

"From 22 May 1992 until 19 December 2007 the MasterCard payment organisation and the legal entities representing it, that is MasterCard Incorporated, MasterCard International Incorporated and MasterCard Europe Sprl, have infringed Article 81 of the Treaty and, from 1 January 1994, Article 53 of the EEA Agreement by in effect setting a minimum price merchants must pay to their acquiring bank for accepting payment cards in the European Economic Area, by means of the Intra-EEA fallback interchange fees for MasterCard branded consumer credit and charge cards and for MasterCard or Maestro branded debit cards.

Accordingly, MasterCard was ordered to bring to an end the infringement and to refrain from repeating the infringement through any act or conduct having the same or equivalent object or effect. They shall in particular refrain from implementing the SEPA/the Intra-Eurozone fallback interchange fees.

Within six months after notification of this decision MasterCard must also formally repeal the Intra-EEA fallback interchange fees, as well as the SEPA/the Intra-Eurozone fallback interchange fees. MasterCard shall moreover modify the association's network rules to reflect this order. MasterCard shah repeal ail decisions taken by MasterCard's European Board and/or by MasterCard's Global Board and/or its delegate the President and Chief Executive Officer of MasterCard Incorporated and/or his designee the Chief Operating Officer or other persons in the association on Intra-EEA fallback interchange fees on SEPA fallback interchange fees and on Intra-Eurozone fallback interchange fees.

The Commissions orders become effective only six months after notification of the Commission's decision to MasterCard Incorporated, MasterCard International Incorporated and MasterCard Europe Sprl This notification occurred on [insert date].

Intra-EEA fallback interchange fees are paid by the acquiring bank to the card issuing bank. Acquiring banks regularly pass the fees on to merchants through inflated merchant discount rates. MasterCard's Intra-EEA fallback interchange fees for MasterCard and Maestro branded payment cards used to be the following:

[insert Intra-EEA fallback interchange fees except for those pertaining to commercial cards]

Annex 6

Analysis of the [...] [consultant]-study

"Evidence on the benefits to merchants from the extension of credit"

The [consultant]-study

1. On 22 December 2006 a bank submitted, in support of MasterCard, a study by [...] [consultant] (called "Evidence on the benefits to merchants form the extension of credit" (hereinafter "the study"). The study was accompanied by a memorandum ('cover memorandum').

2. The study addresses "the issue of whether merchants benefit in aggregates (i.e. as a class) from extension of credit to cardholders" (988), and consists of two parts.

i. The first part reviews a number of economic theories on the issue of whether the extension of credit has real effects on consumption. The first part of the study briefly describes the standard "permanent income hypothesis", which suggests that the proportion of income that people will choose to spend is based on their "permanent" income - that is, the income they expect to have not simply in the current period, but also in future periods. [...] [consultant] criticises the standard "permanent income hypothesis" in light of a number of economic models and concludes that the extension of credit can increase overall consumption through a relaxation of constraints on spending which are driven by a reduced desire to hold precautionary savings as buffer, or which result from removing liquidity constraints.

ii. The second part is an econometric exercise assessing whether an increase in the credit line to cardholders leads to an increase in average credit card balance. A further sensitivity analysis was also conducted using sales (i.e., cardholder purchases) instead of balances as the dependent variable of interest.

3. The econometric results can be summarized as follows:

i. There is a positive marginal propensity to consume an increase in credit. This effect was stronger for "high-utilisation" cardholders, which would indicate that these individuals are liquidity-constrained and that increase spending as credit is made available to them. The finding that, even at low levels of utilisation, extra credit results in extra spending would demonstrate that as an individual's precautionary buffer increases (in absolute terms) through more available credit, average spending can increase.

ii. While "it is no obvious that a distinction can be made between the provision of the interest-free period and the extension of credit more generally (989), the study found that "the effects of credit-line extension are strongest for those who consistently pay interest and there is also a positive effect for cardholders who made of the interest-free period" (990).

Commission assessment of the [...] [consultant] study

4. Regrettably, the econometric results as wee as data description were claimed confidential. Therefore, the Commission cannot on the econometric results obtained and data description. The Commission will nevertheless comment on the general line and findings of the study.

5. The study aims at assessing the potential impact of an increase in credit limits to cardholders on merchant sales, but does not address the relevant question, which is whether credit card usage increases sales volumes in aggregate. [...] [consultant]'s analysis of credit card balances and transactions provides no information over an increase of total spending. Cardholders whose credit limit was temporarily increased may simply increase have switched from paying by debit card to paying by credit card. Any possible increase in balances or sales with credit cards may therefore reflect a shift from one payment means to another. Unfortunately, [...] [consultant] does not control for this possibility and does not provide information on switching between payment means. The study is therefore irrelevant for assessing the benefit of a credit card's free funding period to merchants.

6. The [...] [consultant] study mainly relies on econometric specifications that use balances as a dependant variable of interest. The alternative econometric specification uses sales as a dependant variable and is considered only as a sensitivity test to the first model (991).

However, balances are not a good proxy for cardholder spending (992). Indeed, balances (net of explicit balance transfers) equal, by definition, total credit card values minus repayments. The fact that the balances held on a single credit card increase do not imply that the spending on that credit card has increased, as balances could have increased because of reduction repayments. Therefore, the balances model used in the study introduces an important bias to the extent that the estimates can be explained by other factors than the explanatory variables used in the model.

7. The report tries to circumvent the problem relates to the choice of the correct dependant variable with a further "sensitivity test", which would show that even where a measure of sales rather than balances is used as the dependant variable, the results show that there is a significant impact on spending behavior from an increase in credit" (993). However, irrespectively of the model specification chosen and the dependant variable used (i.e. balances or sales), there are other model specification problems that undermine the econometric results.

8. In particular, the study does not distinguish between the timing of consumer spending and net changes to total consumer spending. If consumers respond to an increase in the credit limit by borrowing more and spending more in the current period, they must repay their debt thereafter. Everything else equal, cardholders must reduce their spending in the future. If credit card holders cannot generate incremental income to finance their incremental purchasing, their capacity to spend will stagnate. A euro spent today cannot be spent tomorrow.

9. The study moreover does not even distinguish the provision of the interest-free period and the extension of credit more generally. Issuing banks finance the extension of credit to cardholders through interests. At no point the study considers why a MIF would be necessary for the extension of credit through credit cards. Rather, the study limits itself exclusively to analysing the effects of extending credit as such. It does not establish a causal link between the issuing bank's capacity to extend credit to cardholders and a MIF.

Conclusion

10. For the reasons set out above, the study does not address the relevant question at hand which is whether a credit card's free funding period can increase aggregate consumer spending and thereby benefit merchants. Moreover, the study suffers from serious methodological flaws that cast doubts on the alleged results obtained from the econometric exercise in the second part of the paper.

Annex 7

Procedural issues regarding a note to the file on the UK market

In order to assess whether the absence of revenues from the domestic MIFs of MasterCard and Visa in the United Kingdom would trigger a "significant" relative reduction of [bank]'s total income from credit card issuing in the United Kingdom as the bank claimed in its memorandum of 5 December 2006 (see section 7.3.4.7 of the decision) the Commission required the bank to report its total revenues and cost associated with the issuing of Visa and MasterCard branded credit cards in the United Kingdom over the period 2003 to 2005. Three other major UK issuing banks were likewise asked to provide such data hi order for the Commission to have a basis for comparison between [bank] and other credit card issuers in this market.

2. The replies were analysed in a note to the file ("the Note"). The note also contained a table setting out some of the banks' data in encrypted form and with the banks' names being redacted in the table (the "Table"). This table analysed the impact of the absence of the MIF on each bank's issuing business. The note was transmitted to MasterCard and to their lawyers together with other documents annexed to the Commission's letter of facts of 23 March 2007. The Commission services informed MasterCard of their intention to take the note into account for assessing whether and how MasterCard's issuing banks would continue issuing MasterCard and Maestro branded payment cards in the hypothetical absence of the MIF (994).

3. On 20 April 2007, while reviewing the documents to which they had access as a result of the Letter of Facts of 23 March, MasterCard's lawyers addressed a letter to the Commission services asking them detailed questions about the Note to the file and the table. In one question, the lawyers asked for "a more detailed explanation of the approach taken in order to guarantee anonymity, namely the application of a common multiplicative factor. There may be a simple explanation, but we have been unable to replicate the Commission's calculation of profit rates, with and without interchange revenues (...) using the numbers reported for incarne, interchange revenues and total costs for each issuer, year and card scheme".

4. In reply to this question the Commission services explained on 26 April 2007 to MasterCard's lawyers (but not to MasterCard) why the figures could not match in the accessible version of the table: "You informed us that you had been unable to replicate the Commission's calculation of profit rates using the numbers reported for total income, interchange revenue and total costs of each issuer, year and card scheme. We suspect that this is because we had to multiply all original data in the table by a "common factor". This also includes the original profit rates." The Commission services also provided an example further explaining the encryption process.

5. On 2 May 2007 MasterCard's lawyers addressed a letter to the Hearing Officer where they take the view that on the basis of the Commission's explanations on the encryption process they were now in a position to calculate the multiplier that was used for encryption. "The Commission seems to have applied a multiplicator of [xx] to all incomes, costs and profit rates reported in the table" and that "[u]nless the above assessment is inaccurate, we note that the Commission's file does not therefore entirely protect the information received by the Commission". The lawyers clarified that "although we have reviewed and assessed the information we have not shared it with MasterCard" and that "we believe that we have a professional duty to share information with our client, unless it is to be withdrawn". The lawyers then sought guidance on how to proceed: "We hereby request the Hearing Officer's guidance as to whether the information is to be withdrawn and, if not, how this information is to be referred to in the on-going process".

6. As a consequence of this letter the Commission services concluded that there was a risk that the information received by the Commission services may indeed not be entirely protected in the note to the file. Taking this factor into consideration, together with reasons relating to substance, the Commission services decided not to rely on the note for a finding in its decision. By letter of 10 May 2007 the Commission services requested MasterCard's lawyers to destroy any (electronic or paper) copy of the note to the file on the 11K market and to return the original CD ROM attached to the letter of facts of 23 March 2007. The Commission services also informed MasterCard's lawyers that subject to the condition of receiving the original of our letter as well as the CD-ROM, and your (and your client's) confirmation on the destruction of the above mentioned copies, we hereby confirm that we will not make use of the note on the UK market in a decision". A similar letter was sent to MasterCard equally requesting the return of the original CD Rom which contained the note on the UK market.

7. However, by letters of 14 May 2007 the lawyers refused to return the CD ROM claiming that the lawyers would otherwise act "in breach of our duties towards our client if we were to hand them back and agree for the file to be amended: the evidence is arguably exculpatory (.)". MasterCard returned the original CD ROM with the note to its lawyers and destroyed ail available copies thereof. MasterCard's lawyers refused to return the original CD ROM and the explanatory letter.

8. On 16 May 2007 MasterCard replied to the Commission's letter of facts. At paragraph 367 of its reply MasterCard requests that the Note on the UK market "should be discounted, or withdrawn". The reasons MasterCard puts forward are methodological flaws of the data analysis.

9. As set out above the Commission decided not to rely on the note on the UK market and the table contained therein for a finding in this decision among other reasons to safeguard the protection of business secrets of the banks concerned. MasterCard's justification for not returning the CD ROM that the note may also contain "arguably exculpatory" evidence is implausible, as its lawyers at the same time repeatedly requested the Commission services to withdraw reliance on the note and the underlying data from the UK banks. The Commission is considering legal action to ensure the return of the original CD and has informed the banks concerned of the incident by separate letter of 30 November 2007.

Notes

1. OJ L 1, 4.1.2003, p.1; Regulation as amended by Regulation (EC) No 411-2004 (OJ L 68, 6.3.2004, p. 1).

2. OJ P 13, 21.2.1962, p.204.

3. OJ L 354, 30.12.1998, p.l8.

4. OJ L 123, 27.4.2004, p.l8.

5. OJ C ...

6. Case IV-34.293-British Retail Consortium v. Visa & MasterCard/Eurocard/Access.

7. Case COMP-D.1- 36.518 -EuroCommerce.

8. Case COMP-D.1-34.324 - Edc/Maestro.

9. Case COMP-D.1-34.579 - Europay (Eurocard-MasterCard).

10. Case COMP-D.1-34.324 - Edc/Maestro.

11. Case COMP-D.1-35.217 -Europay ATM program.

12. Case COMP-D.1-35.517 - SigniaProgram.

13. Case CQMP-D.1-35.578 - Europay Membership and Licensing Rules.

14. OJ C89, 13.4.2002, p.9, paragraph 21.

15. Case COMP-D.1-38.580 - Commercials Cards.

16. OJ L 123, 27.4.2004, p.18.

17. Visa Europe, Servired, Euro 6000, Sistema 4B, Kappé International BV, European Retail Roundtable, Ikea, the International Air Transport Association (IATA), Deutsche BP, Hotels, Restaurants and Cafés in Europe (HOTREC) and an Austrian association of hotels (OHV).

18. Such meetings were, for example, held with: the Royal Bank of Scotland (a Principal Member and shareholder of MasterCard Inc.), Citibank (a Principal Member and shareholder of MasterCard Inc.), the Austrian Payment Systems Services (a processor), Clear2Pay (a processor), American Express (a competitor of MasterCard), B&S Kartenservice (an acquirer), Airplus International (an issuer) and merchants from the airlines industry.

19. In particular submissions in reply to the Commission's Statement of Objections of 24 September 2003 by EuroCommerce (28 December 2003), by MasterCard (5 January 2004), by Visa International (5 January 2004); submissions in reply to the Commission's Supplementary SO of 21 lune 2006 by MasterCard (15 October 2006), by Visa Europe (29 September 2006), by EuroCommerce (13 October 2006).

20. Commission Decision of 7 August 2001 case COMP-D1-29.373, OJ L293, 10.11.2001, p.24.

21. The NDR was cleared on the grounds of lake of appreciable effects. The Visa I decision confirmed that the NDR restricts competition. The only reason for the negative clearance was that the restriction of competition lacked an appreciable effect. This conclusion was based on the studies regarding the effects of the abolition of the NDR in Sweden and the Netherlands. See recitals 11-12 and 54-58 of the decision. The assessment of " appreciability " may obviously change over time as market conditions evolve.

22. Commission Decision of 24 July 2002 in case COMP-D1-29.373, OJ L318, 22.11.2002, p.17.

23. Ibid., pt. 66.

24. Ibid., pt. 90.

25. In May 2004 Visa MasterCard decided to publish their cross-border MIF rates on their websites. See Commission press release IP-04-616 of 7 May 2004.

26. The enquiries were initiated through two Commission decisions of 13 June 2005.

See IP-05 of 13 June 2005 and MEMO-05 of the same date.

27. http://ec.europa.eu/comm/competition/antitrust/others/sector_inquiries/financial_services/retail.html

28. 5934, Mastercard's reply of 21.11.2003 to the Commission's request for information, Q.2.4.

29. 33958, Mastercard's reply of 22 December 2006 to the Commission's request for information, Q.24. : [...]. [Business Secret - 1b]

30. 35390, Mastercard's reply of 19 January 2007 to the Commission's request for information, Q. 27-28.

31. 35854, pages 138-145 of the document, Article I - Membership of Amended and Restated Bylaws of MasterCard International Incorporated with limited changes and 35390, MasterCard's reply of 19 January 2007 to the Commission's request for information, Q 27-28.

32. Cf. below, section 2.1.2.a.

33. 34496, MasterCard's reply of 1 December 2006 to the Commission's RFI, Q 14, 15 and 16.

34. 26227; Bylaws MCI Article VII.

35. 26227; Rules MCII Chapter 12.1[...], Chapter 18.1and 18.5. [Business Secret - 1b]

36. 26227; Rules MCII Chapter 18.B.2.2. [...]. [Business Secret - 1b]

37. Ibid. In addition, the regional boards made of member banks exercised any powers and authorities delegated to them "by the Board of Directors of the Corporation or MasterCard International Incorporated" (26227, Bylaws MCI Article VII sections 1 and 4). This board was in turn composed of directors appointed by member banks.

38. 26227; Bylaws MCI Article VI section 2 and VII section 6. The decision of a Regional Board could have been moreover overruled only in very exceptional cases, i.e. if the Regional Board's decision (i) has effects on other regions (i.e.: the Regional Board has exceeded its powers) or if (ii) the decision in inconsistent with a published policy, practice or strategy of the entire Corporation.

39. 35876, Proxy Statement filed with the US Securities and Exchange Commission, including a series of slides of 31 August 2005.

40. 35341-35328, three such boards were established in Europe: Western Europe Customer

Advisory Forum, Southern Europe Customer Advisory Forum and Central Eastern Europe Customer Advisory Forum.

41. 35929, page 834 last paragraph at the bottom.

42. According to the Amended and Restated Certificate of Incorporation of MasterCard

Incorporated, Section 6.7(A).

43. 33858, Point 11 (B) of Amended and Restated Certificate of Incorporation; [...] [Business Secret - 1a]

44. 35856, pdf page 28.

45. 35929, MU page 1158.

46. 35854, MCI page 134, In accordance with point Eleventh (C) of Amended and Restated

Certificate of Incorporation of MasterCard International Incorporated.

47. The purpose of this high quorum was the following: [...]. [Business Secret - 1b]

[emphasis added] (35927, MCI page 1543).

48. These conditions cover situations where in Global Board's sole judgment and discretion, any action or failure to take action by the European board (1) could subject Corporation or any of its subsidiaries to risk of legal or regulatory liability, (2) would be contrary to the Corporation's global strategy, (3) would be reasonably likely to have an effect outside Europe or on U.S.commerce or (4) relates to any matter outside the authority granted to the European Board.

49. 35464-35461, Procedure on composition and nomination of a licensed bank delegates on the European Board.

50. 35458, Amended and Restated Certificate of Incorporation of MasterCard Incorporated, Section 6.7 (C).

51. 35929, MCI pages 1154-1172, " future governance of Mastercard in Europe " of 22 September 2005.

52. 35929, MCI pages 1170-1171, [...]. [Business Secret - 1b]

53. 35944 [...]. [Business Secret - 1b]

54. 35413-35411, Mastercard Bylaws and Rules, Chapter 18 and Maestro Global Rules, Chapter 19.

55. 35389-35387 and 35845-35799, MasterCard reply of 19 January 2007 to the Commission's request for information, Q.31 with Annexes. According to Mastercard's own knowledge, such country specific rules were adopted in respect with MasterCard in [...]. [Business Secret - 6] With respect to Maestro, according to Mastercard's knowledge, such rules were adopted in [...]. [Business Secret - 6]

56. 35389-35388, Mastercard's reply of 19 January 2007 to the Commission's request for more information, Q.31 : " MasterCard does not endorse or certify domestic rules. MasterCard would, however, verify compliance if a complaint were received regarding a domestic rule ".

57. 34494-34493, MasterCard's reply of 22 December 2006 to the Commission's request for information, Q.20 : [...]. [Business Secret - 2b]

58. 33376, [...]. [Business Secret - 2a]

59. 35921, page 14, Minutes of the meeting of the global Board on 18 November 2004.

60. This change in the corporate governance of the payment organisation did not yet modify the way MasterCard's intra-EEA fallback interchange fees were set.

61. For the different membership classes in the MasterCard payment organisation refer to section 2.1.1.

62. 26227, Bylaws MCI Section IV.2. As the persons sitting on the Board of MasterCard Incorporated are identical with those of the Board of Directors of Mastercard International Inc., management commonly calls both boards together " the Global Board ".

63. 26277, Bylaws MCI Section IV.2.

64. 34185-34184, MaterCard's reply to the Supplementary Statement of Objections.

65. " Public investors " means any investor from the general public be it in an individual, legal entities or the State.

66. Amended and Restated Certificate of Incorporation of MasterCard Incorporated, Article IV, section 4.3(A).

67. The NYSE criteria for independence protect investors by ensuring that effective management board of listed companies exercise " independant judgment ". Section 303 A01 of the NYSE rules foresees that majority of board directors must be independent in order to reduce the risk of " damaging conflicts of interests " (34397). " Independent " directors may not have a material relationship, either directly or as a partner, officer or director of an organisation that has a relationship with the listed company (http://www.nyse.com/pdfs/finalcorpgovrules.pdf). According to MasterCard the Global Board fulfils these criteria.

68. 34496-34495, MasterCard's reply of the 22 December 2006 to the Commission's request for information, Q 14-16 (...) MCI remains a membership corporation, essentially isolating the rights of MCI's Class A customers to provisions in MCI's rules and bylaws, and controlled wholly by its Class B member, MasterCard incorporated, which is in turn controlled by public shareholders ".

69. Europay International S.A. was in turn created on 1 September 1992 following a merger between EuroCard International and Eurocheque International. Europay was subsequently integrated in the MasterCard Europe Sprl.

70. 35924, page 236, draft Letter to Regional Boards : [...]. [Business Secret - 3b]

71. US Case No. 96-CV-5238 (JG)), also referred to as the " Wal-Mart case ".

72. 35924, page 288-292, 360-364 : [...]. [Business Secret - 3b]

73. 35924, page 360.

74. 35924, page 246 and 35927 page 1261.

75. 35924, page 246 : [...]. [Business Secret - 3b]

76. 35922, page 147. Minutes of the N&CG Committee's meeting on 26 October 2004.

77. 35921, page 11 ; [...] (35922, page 148). [Business Secret - 3a]

78. 35922, page 152. On 18 November 2004 the Global Board adopted the following resolution : [...] (35908, pt. 7 on the agenda) [Business Secret - 3b]

79. 35921, page 12, 2nd resolution.

80. 35924, MCI page 12

81. 35921, MCI page 12, Resolution of the Global Board during the meeting on 18 November 2004.

82. 35921, MCI page 21.

83. See, for instance, [...]. [Business Secret - 3a] 35927, page 1493 and 1494.

84. See, for instance, internal training material for staff presenting "[...]" [Business Secret - 3a] to European shareholders banks (35929, page 833 to 835; the fact that the training material was used to brief banks derives in particular from questions 12 and 13 and that European banks were addressed derives from emphasis of the European Board and the SEPA project uin questions 9 to 10) as well as on page 1099 to 1101.

85. See, for instance, 35922, page 210 ("[Senior MasterCard Manager] [...]. [Business Secret - 3b])

86. 35922, MCI page 163-170, Minutes of the N&CG Committee's meeting on 21 January 2005, see in particular on MCI page 168.

87. 37041-37070, [...]. [Business Secret - 3b]

88. 35921, page 22, Minutes of the meeting of the Global Board on 10 February 2005 : "[Senior MasterCard Manager] [...]. [Business Secret - 3b]

89. 35921, MCI page 23-24.

90. 37054 to 37053.

91. 37044, [...]. [Business Secret - 3b]

92. 35921, MCI page 22, Minutes of the meeting of the Global Board on 10 Feburary 2005. See also : [...] [Senior MasterCard Manager] [...] [Senior MasterCard Manager] [...] [Business Secret - 3b]

93. [Business Secret - 3a]

94. 35921, MCI page 36, Minutes of the Meeting of the Board on 17 March 2005, [...]

[Business Secret - 3b]

95. 35922, MCI page 175, Minutes of the meeting of the Nominating and Corporate Governance Committee of the Global Board on 2 March 2005, [...] [Senior MasterCard Manager] [...] 35922, MCI page 182, Minutes of the

N&CG Committee's meeting on March 2005: [...] [Senior MasterCard Manager] [...] [Senior MasterCard Manager] [...] [Senior MasterCard Manager] [...] [Senior MasterCard Manager] [...] [Senior MasterCard Manager] [...] [Senior MasterCard Manager] [...] [Senior MasterCard Manager] [...] [Business Secret - 3b]

96. see for example 35925, MCI page 1258-1259, [...] and 35925, page 1353-1354, [...] [Business Secret - 3b]

97. 35921, MCI page 21-22, Minute of the Meeting of the Global Board on 10 February 2005, in particular : [...] [Senior MasterCard Manager] [...] [Business Secret - 3b]

98. 35922, page 225, N&CG Committee's document [...] [Business Secret - 3b]

99. [...] the N&CG Committee during its meeting on 13 April 2005. [...], see 35922, MCI page 195 [...] [Senior MasterCard Manager] and MCI page 198 [...] [Senior MasterCard Manager] [...] [Senior MasterCard Manager] [...] During the following meeting on 28 April 2005, the N&CG Committee [...] [Senior MasterCard Manager] [...] [Senior MasterCard Manager] [...] [Senior MasterCard Manager] [...] [Senior MasterCard Manager] [...] [Business Secret - 3b]

100. 35922, page 202, Minutes of the Committee's meeting on 5 May 2005.

101. See 35927, MCI page 1224, [...] [Business Secret - 3a]

102. 35922, page 204, Minutes of the N&CG Committee's meeting on 5 May 2005 : [...] [Senior MasterCard Manager] [...] 35922, pages 205-206, Minutes of the N&CG Committee's meeting on 20 May 2005 : [...] [Senior MasterCard Manager] [...] [Senior MasterCard Manager] [...] [Business Secret - 3b]

103. 35921, page 145, Nominating and Corporate Governance Committee Recommendations of 10 June 2005. [...] [Business Secret - 3b] 35922, MCI page 225, Minutes of the N&CG Committee's meeting on 6 April 2005, Agenda Item 3 [...] [Business Secret - 3b]

104. 35921, page 40-42, Minutes of the meeting of the Global Board on 16 June 2005. [...] (35921, page 1094 to 1097). [Business Secret - 3b]

105. 35921, Minutes of the meeting of the Global Board on 16 June 2005. [...][Bank][...] [Business Secret - 3a]

106. 35921, page 54-59. Minutes of the meeting of the Global Board on 14 July 2005.

107. 35921, page 56-57. [...][Senior MasterCard Manager] [...] [Business Secret - 3b]

108. 35921, page 62 and pages 127-143, "MasterCard Structure".

109. 35929, page 880. [...] [Business Secret - 3b]

110. Ibid., page 881 (second paragraph from to the top) and 882 (second paragraph from the top).

111. 35929, 1099 and 1100 [...] [Business Secret - 3b]

112. 35929, page 1011; see also the power point presentation on page 857 [...] as well as page 858 [...] [Business Secret - 3b]

113. 35929, page 864. See also [...] (36064, page 799, 3rd paragraph from the top): [...] [Business Secret - 3b]

114. 35849-35847, Minutes of tise Special Meeting of Stockholders of MasterCard Incorporated of 28 November 2005.

115. 35856, page 58-59 Section "Reasons for Transition to New Ownership and Governance

Structure" of the Registration Statement failed with the SEC on 15 September 2006.

116. 35908, pt. 7 on the agenda.

117. 28099, Mastercard reply 8 May 2006, Annex 1.

118. 5923, Q 7.3; 27513, MasterCard reply 14 April 2005, Q 14.

119. 5923, Q 7.3; 27513, MasterCard reply 14 April 2005, Q 14.

120. Case 56-65, Société Technique Minière v Maschinenbau Lim, [1996] ECR 235, p. 250.

121. MasterCard's CEQ stated la February 2006: "MasterCard Europe uniquely has the experience in migrating national-use-only debit acceptance locations ta Maestro. Take, for example, the recent upgrade of over 600,000 merchant locations from Switch-only to Maestro in the UK over a period of just 2 years. (...) Over [...] 2 billion previously national-only debit branded transactions [in the United Kingdom and Switzerland] have migrated to Maestro".

122. The term 'EU 15" stands for the following Member States: Austria, Belgium, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Luxembourg, The Netherlands, Portugal, Spain, Sweden, United Kingdom.

123. Retail Banking Research Ltd., Payment Cards in Western Europe 2006, International Card Organisation page 58.

124. Retail Banking Research Ltd, Payment Cards in Europe'2002 market overview, International Payment Organisations, p.7-I5 and 68 and are based on data from the 15 EU Member States, plus Norway.

125. 26252.

126. 26248.

127. 26245.

128. 26248.

129. Retail Banking Research Ltd., Payment Cards in Western Europe, International Overview page 27.

130. The diagram shows the total number of outlets per card brand. The total results from adding the number of outlets in each Member State of the EEA (2004). The "largest national scheme" stands for outlets accepting payment cards of the largest domestic payment card scheme in each respective EEA Member State (2004).

131. 21856, MasterCard reply of 6 April 2005 Annex 3.

132. All in all, the following number of licenses per year were granted: 1997 ([200-500]), 1998 ([100-300]), 1999 ([0-100]), 2000 ([0-100]), 2001([0-100]), 2002 ([0-100]). The number of licenses are granted in 1997 and 1998 are not representative according to MasterCard. See 5929, MasterCard reply 22.11.2003 Q 3.1.

133. 6504, MasterCard submission 5.1.2004, p. 60, figure 5.

134. According to MasterCard (see 6036, Mastercard reply 3.12.2003, page 2), the association processes [0-5]% of domestic transaction in Austria, [10-50]% in Belgium, [0-5]% in Denmark, [0-5]% in Finland, [0-5]% in France, [50-100]% in Germany, [10-50]% in Greece, [50-100]% in Ireland, [50-100]% in Italy, [5-10]% in Luxembourg, [0-5]% in the Netherlands, [0-5]% in Portugal and Spain, [50-100]%, [0-5]% in Sweden, [50-100]% in the UK.

135. 21638 at Q 3.

136. [...] [Business Secret - 2a] See below at recital 418.

137. 26227, Bylaws MCI Section 18.B.2.5.

138. Section 10.4. of Mastercard International Inc. Rules [...] Section 18.B.2. of the Rules [...] See also the Interchange & Service Fees Manual (Jan 2007) section 2-2 [...] [Business Secret - 2b]

139. Section 18.B.2.6.2 of the MasterCard International Inc. Bylaws and Interchange and Service Fees Manual (section 1-3 in version Nov 2004 and section 2-1 in version Jan 2007).

140. See Interchange & Service Fees Manual (version1/2007), section 2-2.

141. See MCI Rules section 18.B.1. and Interchange & Service Fees Manual (version 1/2007), section 2-2.

142. The rates were endorsed by [MasterCard] [...] [Business Secret - 2a] 35402, Mastercard reply dated 19 January 2007 to Commisssion's request for information, Q (7), and the documents on pages 35370-35364.

143. 36125.

144. Less important rates are the secure e-m commerce rate of EUR 0.05 + 0.25 %, the signature base rate of EUR 0.05 + 0.3%.

145. For more details see Annex 1.

146. 36125.

147. MasterCard " News Release " of 30 March 2007 " MasterCard Remains Commited to Introducing Maestro in Belgium as a SEPA Solution ".

148. Newspaper article " MasterCard Europe Extends Timetable re : Maestro SEPA Fallback Interchange rates " http://www.paymentsnews.com/2007/06/mastercard_euro.html.

149. For more details see Annex 1.

150. MasterCard also introduced a new rate for PayPass transactions.

151. See MasterCard letter to Commission of 8 November 2007.

152. These new " intra-European fallback interchange fees " confusingly bear the identical name as the predessecor of MasterCard's " intra-European Economic Area fallback interchange fees ". While the scope of the present decision encompasses the " intra-European fallback interchange fees " as notified by Europay in 1992 and in force until 11 January 2007, it excludes the " intra-European fallback interchange fees " as understood since 11 January 2007.

153. 35295 (section 3.3) and 35268.

154. Interchange &Service Fees Manual Jan 2007, section 2-1 to 2-3.

155. Ibid., section 2-3[...] Ibid., section 2-2 [...] [Business Secret - 2b]

156. See the power point presentation entitled [...] of 22 November 2006 (35268) [...] 35370-35366, Memorandum of 21 November 2006 [...] and in particular 35370 [...] [Business Secret - 5b]

157. [...] 35298 and 35288 [...] [Business Secret - 5b]

158. 35398. See also section 3.1.10.2. The Italian fees are in force since April 2007. They are not subject of this decision.

159. 33376, resolution of the MasterCard Global Board adopted at the Meeting of the 14 September 2006 ; 34190, MasterCard's reply of the 15 October 2006 to the Supplementary Statement of Objections, pt. 51 : [...] [Business Secret - 3b]

160. 22307 at Q 4.

161. Mastercard's Inc.'s Associate General Counsel stated at the oral hearing of 14 November 2007 : " As to the question : How do we know that if we raise the interchange fee for example, that we are not harming the merchant without getting any benefit on the other side ? The perfectly honest answer is : We don't know with absolute certainty. This is a business to run, we are not, as Professor [...] mentioned, we are not trying to mesure some theoretical " social welfare ". We are trying to run a business. "

162. These figures are only illustrative,

163. 20874, 2004 Interchange & Service Fees Manual of Europay Int., p. 4-9 Validation of Fees.

164. For details on the weighting measurements sec 27513 MasterCard reply of 14 April 2006.

165. 27513, MasterCard reply of 14 April 2006, Q 13.

166. 27513 and 35791. The average of [...] [Business Secret - 2a] is calculated on data for the first to third quarter of 2006, only.

167. 27513 and 35791. The average of [...] [Business Secret - 2a] is calculated on data for the first to third quarter of 2006, only.

168. The diagram does not take account to the newly adopted intra-EEA fallback rates which entered into force on 11 January 2007, only.

169. See section 8.5. with further details on the development of transaction values and volumes in the MasterCard scheme.

170. 26383. Bylaws MCI May 1999, section 11.09.

171. 3442, MasterCard letter of 13.6.2001, Annex 1, "The methodology of Europay for determining the level of the MIF".

172. 24217 b; Chapter 1-1, section "Basic fee types"; 37399 Chapter 1-1, section "Basic fee types".

173. 27171, Interchange & Service Fees Manual March 2005 Chapter 1-1, section "Basic fee types".

174. 6565, pt. 22.

175. 6803, pt. 17 ("On this basis, the issuer and the the acquirer are jointly providing a service to their customers namely "enabling payment from cardholder to the merchant" and, as in the threeparty example, there is both joint production and joint demand") and pt. 69("The relevant product market must, therefore, be the MasterCard service and the relevant product market is defined by determining what other payment systems compete with the MasterCard service").

176. 6563, pt. 31.

177. 6493, pt. 368.

178. 6536., pt. 151.

179. 6536., pt. 149.

180. 6542, pts 111-114.

181. 6566, 3rd bullet from the top.

182. 6564, pt. 23 ("On the contrary, the MIF is designed to take into account the network effects of the system and to increase the overall output of the system") and 6534, pt. 154 ("As the function of the MIF is to maximise output to the benefit of the system and its users ...").

183. 34001, pt 649.

184. 34163, pt. 134-135. MasterCard does not contest the existence of previous statements according to which interchange fees were a "price" properly speaking. However, MasterCard argues that such previous statements ought to be discarded, because "the MasterCard manual did not seek or purport to explain all of the factors taken into account in establishing the fees". MasterCard alse believes that individual statements were cited out of context. The Commission does not share this view.

185. 33998.

186. 23485, Baxter, Bank interchange of transactional paper. Legal and Economic Perspectives, Journal of Law & Economics, vol. XXVI (October 1983).

187. For more details, see section 8.2.2.2.f.

188. 3442, MasterCard letter of 13 June 2001, Annex 1, [...] [Business Secret - 2a]

189. On April 2004 [...] (see at 26485). [Business Secret - 2a]

190. 34189, 35377

191. 34190 at pt. 51.

192. See the MasterCard Europe Interchange Fees Setting Process at 35377

193. For details on the cost items collected cf. section 3.1.8.3.

194. 35373

195. [...] [Business Secret - 2a]

196. In essence, EEA fallback interchange fees for MasterCard branded credit and charge cards were modified by lowering one of the 34 interchange fee rates, the so-called base rate (from 1.30% to 1.20%). The interchange fees for Maestro cards were transformed from ad valorem fees (i.e.: a percentage) into a combined ad valorem and fixed fee MasterCard also introduced as of 11 January 2007 new fallback interchange fees for the Eastern European region, which are not subject of this decision.

197. [...] [Business Secret - 2a]

198. 33997, 660.

199. 33997, 660.

200. 33997, 660.

201. 6493, pt. 369.

202. [...] [Business Secret - 2a]

203. MasterCard Inc.'s Associatc General Counsel at the oral hearing of 14 Nevember 2007:

"MasterCard has employed more or less the same type of cost study for many years. I believe it started in the US probably 30 years ago or so and that cost study measured some specific costs of issuers, namely the processing cost, the interest fee period costs, fraud loses and things related to fraud and things related to credit losses like collection charges and things like that. However, even though these are measurements of the issuers' costs, they are really directed at giving us a better understanding of merchant demand. In other words the theory is that a four party credit yard, which is, which can be seen as a replacement for a store card, and I don't just make that up, we have numerous examples around the world of merchants converting store cards to co-branded MasterCard cards so we know that our four party card is a reasonably good substitute for store cards. The theory is that if we can understand better which are the minimum costs to an issuer for issuing and processing transactions on their cards, we will have a sense of how much a merchant should be willing to pay for our cards, hence the purpose of the cost study is essentially to estimate the merchants' willingness to pay for a MasterCard card. So it really goes to that side of the equation. It is really not an attempt to measure issuer or cardholder demand, hopefully that is understood."

204. MasterCard Inc.'s Associate General Counsel at the oral hearing of 14 November 2007: "If [the MIF] is not designed to allocate specfic costs. So it is... What we're saying is... We are allocating the costs of the system. We are not specifically allocating travel insurance costs. We are not allocating the credit losses. We're not allocating specifically that. We are not specifically allocating any cost [..]. It may be useful to measure somebody's costs. But we are not allocating specific costs."

205. MasterCard Inc.'s Associate General Counsel at the oral hearing of 14 November 2007.

206. [...] 3442, MasterCard letter of 13.6.2001, Annex 1, [...], page 4 and 5. [Business Secret - 2a]

207. 35401, Q 10. For information on the SEPA fallback interchange fees see section 3.1.2.a.

208. [...] 33983, pt. 706. [...] [Business Secret - 2a] [...] [Business Secret - 2a]

209. [...] [Business Secret - 2a].

210. 24698, 2nd paragraph; 33984 at pts 700-703.

211. [...] [Business Secret - 2a].

212. [...] [Business Secret - 2a].

213. [...] [Business Secret - 2a].

214. [...] [Business Secret - 2a].

215. [...] [Business Secret - 2a].

216. 33984 and 33983 at pt. 702.

217. 24716 to24715, Annex 3, page 3-4.

218. [...] (35400, Q 10) [Business Secret - 2a].

219. 4340, minutes of meetings of BMAC and of the Europe Regional Board as submitted by MasterCard in its reply of 3.9.2003.

220. 24716, [...] [Business Secret - 2b]

221. 33997, 661.

222. [...] [Business Secret - 2b] see 4340, minutes of the 36th Europay Business and Marketing Advisory Committee 19-20.8.1999, page 2 and 5.

223. [...] [Business Secret - 2b] See 18794, Extract of Minutes of MasterCard International Board Meeting of March 2003, Annex IB, page 1 and 2.

224. 34160, section 2.2.

225. 34159, pt. 151 1st bullet.

226. [...] [Business Secret - 2a].

227. 34159, pt. 151 2nd bullet.

228. 34159, pt. 151 3rd bullet.

229. [...] [Business Secret - 2a].

230. 34157, pt. 157, pt.158, 159 etc.

231. 36008, page 47, section 9A in MasterCard's Maetro Debit Card Proposal for an Italian bank of 26 September 2006 [...] [Emphasis added] [Business Secret - 5b].

232. See hereafter in section 7.1.2.2.

233. 35298 [...] and 35288 [...] [Business Secret - 5a]

234. [...] (35368) [...] [Business Secret - 5b].

235. 35398, table 13.2. [...] [Business Secret - 2a]

236. 36008, page 12, diagram in section 3.5 [...] [Business Secret - 5b].

237. 36796 [...] [Business Secret - 2b]. See also at 36794 on the Reserve Bank Rationale for reducing the Visa and masterCard MIF by 50%. Tho RBA concludes in pt 22 that interchange fees "are not subject to the normal forces of competition. This was both in the credit yard schemes where they were multilaterally determined by the same banks in each system, and in the EFTPOS system, where they were determined bilaterally but, once negotiated, almost never changed thereafter.

238. In the USA American Express is turning from a closed into an open card scheme and tries to convince issuing banks to join its scheme rather than that of MasterCard or Visa. American Express offers issuers fees for issuing its cards. Visa and MasterCard steadily increased their interchange fees to match increases of American Express' fees. For a graphic illustration of this upward development of interchange fees in the USA see 28621, Table C "Interchange fees of Visa and MC in the USA 1983 - 2004. Data are based on on the US Nilson Report. In the February 2004 issue (no. 804, pages 1 and 8) the Nilson Report takes note of the upward trend of Visa and MasterCard interchange in the USA and reasons as follows: "By increasing the fees that merchants must pay to U.S. credit card issuers, MasterCard and Visa will begin to close the 75 basis point gap between what issuers earn from them and what issuers could earn from American Express. Higher interchange fees will help, to discourage issuers from following MBNA into an Amex partnership. MasterCard will raise interchange beginning April 2 [2004] and Visa will need to do the same shortly afterwards to remain competitive. Over time, American Express can be expected to respond by widening the gap again." [Emphasis added] In reply to the Supplementary Statement of Objections MasterCard did not contest the correctness of the underlying data.

239. Note that MasterCard charges issuing banks a membership fee per card payment and that this fee is [...] [Business Secret - 5a] as the fee charged to acquirers. See section 6.3.1.2.

240. 34155, MasterCard reply to Supplementary Statement of Objections of 15 October 2006, pt. 166.

241. See already Commission decision of 24 July 2002, Visa II, pt. 80.

242. See section 7.2.4.3. for details.

243. 4526, the 32nd Europay BOD, Waterloo, of 9 December 1999.

244. 4529, Ibid.

245. Another reason was to incentivise acquirers to migrate from signature based Maestro transactions to chip transactions.

246. 4340, 36th Europay Business and Marketing Advisory Committee 19-20 August 1999, annexed to MasterCard reply of 3.9.2003.

247. 4340, 31st Europay Board Meeting of 23.9.1999, annexed to MasterCard reply of 3.9.2003.

248. 4340, 36th Europay Board Meeting of 28.9.2000, annexed to MasterCard reply of 3.9.2003.

249. The rate remained at [...] [Business Secret - 2a] until 11 January 2007. See Annex 1 for more details on the intra-EEA fallback interchange fees of the Maestro debit product.

250. "Eastern Europe" refers to countries in Eastern Europe outside the EEA such as Russia, Ukraine and the Balkans.

251. 27194, MasterCard reply of 14 April 2006 Annex 11 and 12 [...] [Business Secret - 2b].

252. 34153 at pt. 171 and 172.

253. The Commission notes in this respect that in the "enhanced electronic" rate which accounts for approx. [...] [Business Secret - 2a] of total consumer credit card usage in 2004. Sec 10127, MasterCard letter of 26 March 2004, page 3, Q 3.

254. For a more detailed overview on the development of MasterCard's intra-EEA fallback interchange fee rates in the credit and charge cards segment see Annex 1.

255. The majority of transaction volumes of MasterCard branded payment cards is subject to the anhanced electronic rate ([50-70]%) and the base rate ([10-30]%); sec 10127, Mastercard letter of 26.3.2004, page 3, Q 3. Only interchange fees for business and purchasing cards were lowered in 1999 and 2000. However, this category only accounts for a handful of cards, e.g. +/- [3.000-10.000] cards out of approx. [10-50] Mio MasterCard branded payment cards (sec 21640, MasterCard reply 21.3.2005, Table 1). MasterCard reported to have in issuance as of 17 Fabruary 2004, the following: [Bank] (since September 2003) - to date, +/- [100-1.000] cards; [Bank] (since January 2004) - to date, +/- [50-500] cards; [Bank] (since September 2002) - to date, [20-150] cards; [Bank] (for several years - outside the EU) to date, +/- [2000-10.000] cards.

256. Section 3.1.6, table 1.

257. Only in 2007 MasterCard reduced one rate on the consumer credit card segment, i.e.: the base rate from 1.3% to 1.2%. The other rates remained unchanged. See Annex 1 for a more detailed overview on the development of MasterCard's intra-EEA fallback interchange fee rates.

258. See 4340, MasterCard reply 3.9.2003 and the attached minutes of the 32nd meeting of Europay International's Business and Advisory Committee in Zaventem on 12.11.1998 with power point slides (pages 5 and 7) and the minutes of 36th BMAC meeting in Berlin 19/20.8.1999, page 2, section 3.1.

259. 34152 and 34153.

260. 24211, MCII Bylaws 2004, Article I Section 5 c.

261. EuroCommerce made the following statement during the oral hearing: "First, how would you impose a cost on consumer when he does not know about the cost? Two, have you heard about competition? Imagine Sainsbury imposing a surcharge without Tesco doing it? This is what we mean by real competition. Three, which MIF should we apply? How would we know it? This is still a business secret in many cases, even to the largest retailers. Four, if surcharging is technically difficult or impossible for large retailers, can you imagine the situation for SMEs? Five, despite what was said, MasterCard contracts still prevent in many cases surcharging. Six, it would be most interesting to hear Visa's opinion on the NDR? Seven, and most important, why does MC ask a third party -the marchant- to pass on to all its customers charges that the bank should - and could-pass directly to the one who benefit from the specific service?".

262. The SSNIP test (a Small but Significant and Non-transitory Increase in Price) is a tool to define the relevant product market. It is an alternative to determine the relevant product market by analyzing the similarity of different products.

263. 37507.

264. 37501 and 37502.

265. 37574.

266. 32529.

267. 19674, minutes of meeting between Commission and airline representatives of 19 JuIy 2004.

268. 36136.

269. 32657.

270. 32058.

271. 32051.

272. Save for "on-us" transactions where cardholder nid merchants are clients of the same bank. This is not infrequent.

273. See alse MasterCard's definition of four party payment card schemes in MasterCard's submission of 5 January 2007, pt. 95 ("Four-party systems are open systems, in the sense that dtifferent financial institutions can join the system under the same brand").

274. In practice, the borderlines between open and closed card schemes are fluent. Closed schemes such as American Express meanwhile issue a portion of their cards also through independent banks and in "open" schemes such as MasterCard and Visa payment card transactions often occur "on-us". This means that one and the same banks acts as issuer and acquirer for a given card transaction. In concentrated markets or markets where issuers co-own the dominant acquirer such on-us payments are not uncommon.

275. In 2003 MasterCard's member banks paid a total of € [50-100] Mio for intra-EU cross border transactions to MasterCard with the payment scheme service fees € [30-80] Mio being the large bulk. Taking domestic payment card transactions into account, as well, MasterCard annually yields approx. € [100,000-200,000] Mio in fees (2003) 20509, MasterCard reply of 31.1.2005, table 1.

276. This appears to be the case at least in Belgium, Germany, Greece, Ireland, Italy, the Netherlands, Spain and the UK; see Retail Banking Research Ltd., Payment Cards in Europe 2004, Belgium, page 14; Germany, page 22 and 23; Greece, page 24 and 25; Ireland page 12; Italy page 18; the Netherlands page 15; Spain page 33; Sweden page 14; the UK page 36 and 37.

277. "Negative float" is referred to as a situation where a carholder is debited on the bank account one or more days retrospectively so that the transaction is artificially made to occur before it was executed. Banks thereby gain funds without paying interests.

278. See in section 7.3.2.1.3. and 7.3.2.1.4.

279. The merchant is paid "now" for a service he delivers to his customer only in the future. This entails an increased risk for the acquirer if the cardholder later claims that he never received the service and if the merchant has meanwhile gone bankrupt; 12411, nc reply of [Bank] of 18.5.2004, Q 15: "Advance crediting is negotiated as part of the contract with the merchant. Each Merchant's Service Charge will depend on a number of factors, including funding costs".

280. There are, however, exceptions: Merchants sometimes pay combined fees (percentage plus flat fee) for debit card transactions and in some countries (particularly Central and Eastern European countries) they pay the same MSC for credit and debit card transactions. See, for example, English translation of Polish National Association of Trade and Services letter of 25.1.2005 "Another shortcoming of the system in Poland is that the fees charged for transactions by debit card and credit card are the same, whereas in most countries a lower fee is charged for payment by debit card". A Czech bank also told the Commission that they are generally charging merchants the same fees for Maestro and MasterCard transactions; see 17686 Q 6, Table 3.

281. See for example 8641, nc reply of [Bank] of 16.1.2004, p.2, 2a.

282. 28417. [MasterCard Consultant]

283. 6807, " Econometric analysis of the MIF " by [...] by Prof. [...]

284. 6769, " Economic evidence in support of MasterCard's response to the Statement of Objections " by Dot Econ.

285. 6708.

286. 6680.

287. 34090.

288. Commission decision of 24 July 2002, Visa II, pt. 82.

289. Commission Notice on the Application of Article 81(3) of the Treaty, OJ C101, 27.4.2004, p.97 to 118 at pt. 84.

290. 6807, " Econometric analysis of the MIF " by [...] by Prof. [...]

291. The scheme owner (here: Mastercard Incorporated) acts as platform that interlinks banks, issuers and acquirers, to perform a payment transaction. For details on the concept of a platform in payment systems see in particular 24197, Rocket & Tirole, Platform competition in two-sided markets, Journal of European economic Associations, 13 December 2002, page 5 to 9.

292. 34101, MasterCard response to Supplementary statement of Objections, pt. 324.

293. Indeed, since the early days of EU competition law, it has been constant practice to distinguish separate levels in a production chain in order to analyse competitive constraints on retailers; see already Commission Decision of 23 September 1964 in Case Consten-Grundig, OJ 2545-64, [1964] 489.

294. To illustrate this, consider the typical example of an industry with two-side demand: the matching business. If singles clubs werc te co-operate by creating a matching web portal this would not imply that there is no more any downstream competition between male-single clubs for males on the one side, and between female-single clubs for females on the other.

295. See also 8369, Visa submission 5 January 2004, pt. 3.3. and 2.5. which contradicts MasterCard's views in this respect. A two-sided "market" is defined as a "market" where efficiency can be improved by changing the structure of prices without changing the overall pricing level. Hence the split of prices matters more than their sum.

296. 6524, MasterCard submission 5 Januasy 2004, pt. 202, nid 6494 at pt 362 ("The MasterCard default MIF rectifies the problem. As a result of the transfer payment between issuers and acquirers the MIF balances demand between cardholders and merchants").

297. 6548, FN 40: "Examples of other multi-sided platforms are newspapers and magazines (which offer a simultaneous service to advertisers and readers) and software operating systems (which offer a simultaneous service to users and software application develapers)".

298. Commission Decision of 20.4.1999 in Case IV-M.1455 -Gruner + Jahr-Financial Times, pt. 15 (" From a product production point of view, news papers editors operate in two broad markets : the market for a written press, in which consumers are the buyers of the newspapers as a source of information and the market for advertising space, in which the consumers are the advertisers who buy space in order to promote sales ") ; see also Commission Decision of 1st february 1999 in Case IV-M.1401 - Receletos/Unidese, pt. 16 ; Commission Decision of 29 November 1995 in Case IV-M.665 - CEP/Groupe de la Cité, pt. 9 and Commission Decision of 15 march 1994 in Case IV-M.423 - Newspaper Publishing, pts. 15 and 16.

299. See Commission Decision of 24.3.2004 in Case COMP-C3-37.792 - Microsoft, chapter 5.1.1.3, para 402-117. (http://europa.eu.int/comm/competition/antitrust/cases/decisions/37792/en.pdf)

300. 34105, pt.307-309.

301. Case T-201-04, Microsoft, of 17 September 2007, not yet reported.

302. 34095.

303. 34J75

304. 34100.

305. Eric Emch and T. Scott Thompson, Market Definition nid Market Power in Payment Card Networks, Review of Network Economics, Vol5 - March 2006.

306. They include, for example, the extension of credit to cardholders in the form of loans or delayed re-payment or other services such as a travel insurance.

307. MasterCard argues in particular that merchants are 'transaction hungry" as opposed to cardholders (24716 to 24715, Annex 3, page 3-4).

308. See section 3.1.8.3. a). [...] [Business Secret - 6]

309. See already at section 6.2.1 ba).

310. Moreover, the SSNIP test in this model of Emch and Thompson is inappropriate to analyse the exercise of market power by the scheme owner or collective action of a scheme's member banks. A SSNIP test on the sum of price charged to cardholders and merchants should in theory address the question whether a competing payment system would act as constraint on MasterCard is raising interchange fees. However, this test fails as an increase in the interchange fee increases issuing banks' willingness to promote a given scheme relative to others. Hence, even if one were to accept the Emch and Thompson methodology and a SSNIP test on the sum of prices indeed found a competing scheme to be a significant competitive constraint on MasterCard, this would not imply that this other scheme places a constraint on the way in which the MasterCard MIF may be increased to the detriment of consumer welfare (that is the consumer surplus of cardholders, merchants and subsequent purchasers). Quite to the contrary, competition between payment schemes may in fact reinforce an anticompetitive increase in the MIF.

311. 34094 and 34093 at pt. 341.

312. Commission decision of 9 August 2001, Visa I, pt. 34.

313. Commission decision of 24 July 2002, Visa II, pt. 43.

314. Commission decision of 9 August 2001, Visa I, pt. 34 and Commission Decision of 24 July 2002, Visa II, pt. 43.

315. Commission decision of 24 July 2002, Visa II, pts. 34 and 66.

316. Commission Notice on the definition of the relevant market for the purpose of Community competition law, OJ C 97 372-03, 9.12.1997, page 5.

317. See the description of product characteristics in sections 5.2.3.1. and 5.2.4.1. as well as - more in detail - in tise Commission's Supplementary Statement of Objections of 20 June 2006 on section 6.3.3.1 nid 6.3.4.1 (28780 to 28781).

318. Sec sections 5.2.3.2 and 5.2.4.2. Tise exception is tise Norwegian BAX scheme where cardholders are used to paying a fee per debit card transaction.

319. Acquiring of payment cards is typically more concentrated than card issuing.

320. See commissio guidelines on the definition of the relevant market, OJ 97-C 372-03 of 9.12.1997, page 5 section III on the use of evidence in competition cases, pt. 38.

321. MasterCard in particular uses cost studies to set its MIF just below a level where it would expect to incur serious problems with card acceptance (cf. MasterCard's Associate General Counsel at the oral hearing of 14 November 2007).

322. The Commission executed a SSNIP test in the acquiring markets, netherless, to verify some findings in the market studies submitted by MasterCard, in particular the results of [...] Dot Econ discouragement study and its conclusion that the market should be defined wider to include cash and cheques, as well. While the Commission doe not attribute decisive importance to it, it notes that the results (28393 to 28395) of its hypothetical analysis tend to confirm the findings based on product characteristics and on past behaviour of merchants.

323. Under the terms of the so-called "Package Deal Agreement", notified to the Commission in 1982, the members of Eurocheque International agreed the no commission is to be charged by the cashier at the time off encashment or by the trader when accepting the cheque; fees may only be charged when the cheque reimbursed by the clearing house of the cheque issuing bank; thus ultimately the cheque user would pay for the costs incurred by issuing and drawing cheques. See ECJ 23.2.1994, GCB nid Europay Intl. v Commission, T-39 nid 40-1992, [1994] page 49 ss at pt. 2; The Package agreement was exempted by the Commission.

324. Cheque users are charged, for example, fees for buying American Express traveller cheques. (http://www.americanexpress.com/uk/tc/faqs.shtml#4).

325. 34090, pt. 357.

326. 28440, pt. 6.3.

327. 28668, Commission Merchant Market Survey at section 4.

328. 6517 and 6770, MasterCard submission 5 Januasy 2004, page 56, Annex II.

329. The [...] Dot Econ study is rather a choice-based conjoint analysis. The main characteristic distinguishing choice-based conjoint analysis from other type of conjoint analysis like ranking-based conjoint analysis (ex : preferences based on the ranking a number of different combinations of attribute levels) is that the respondent expresses preferences by choosing comcepts from sets of concepts, rather than ranking them.

330. In particular different tasks must be undertaken to implement such experiment (i) Identification of the set of attributes ; (ii) Selecting the mesurement unit for each attribute. This can be ordinal (low, medium, high) or cardinal (absolute value, percentage,...) ; (iii) Specification of the number and magnitudes of attribute levels. As an important rule, levels have to be believable ; (iv) Statistical design. The statistical design is an important point in the survey. It can be different according to the objective of the survey. In other words, what we are going to do with the data from the survey and how we are going to use them should be known to construct a design with good properties ; (v) The statistical design has to be translated into a set of questions and for excecution of the data collection phase ; (vi) Selection of an appropriate estimation procedure.

331. 28668, Merchant Market survey, section4. The second most important reason for the merchant to accept payment cards, according to 64% of the merchants participating in the survey, is that " our competitors accept payment cards in their shops as well ".

332. 34098, section 1.3.

333. 34097, pt. 331.

334. Commission decision of 9 August 2001, Visa I, pt. 38 ("First of all (...) cash can be excluded form the relevant market, for several reasons(...)"); pt. 39 ("Secondly, cheques can be excluded from the rekvant market ... "). Commission decision of 24 July 2002, Visa II, pt. 48 ("Next, as concerns cash and cheques, neither of these can be considered as substitutable with payment card, ..").

335. Commission decision of 9 August 2001, Visa I, pt. 55 ("All these types of competition are affected") and Commission decision of 24 July 2002, Visa II, pt. 44 ("Both types of competition are affected by the Visa Rules and by the MIF in particular"). See also Commission decision of 9 August 2001, Visa I, pt. 55 ("This fact indicates a limited effect of the no-surcharge rule on competition on the inter-system market between cardsystems, since (...)"); pt. 63 ("The principle of territorial licensing in the Visa rules restricts the commercial freedom of action of the bank participating in the Visa card payment systems: banks are in principle not allowed to issue cards and to acquire merchants' transactions outside the area in which they have established a branch or subsidiary and for which they hold a licence (...)"); pt. 65 ("The No Acquiring Without Issuing Rules does not in it self create significant barriers to entry on the acquiring market (...)").

336. See for instance the Commission Decision of 24 July 2002, Visa II, pt. 106 ("The MIF does not eliminate competition between issuers, which remain free to (...). Moreover, although it sets de facto a floor in the merchant fees, it does not eliminate competition between acquirers either, (...). Nor does it eliminate competition between Visa and its competitors, (...).")

337. The one difference to the previous case practice in that the present decision leaves it open, in MasterCard's favour, whether the business of crad acquiring and issuing are brand specific depends on the situation in each relevant geographic market. From the Commission's merchant market survey in the present case it would appear, for instance, that on a pan-European basis the majority of acquirers charge identical fees for accepting Visa and Mastercard cards and in some Member States Visa and MasterCard cards acceptance is typically sold as a package. This indicates a market definition for acquiring the "Visa/MasterCard product" rather than acquiring for the MasterCard products on the other hand. However, the present decision assumes, in MasterCard's favour, the broadest possible market definitionas "acquiring of all payment cards".

338. 34093, pt. 344 to 346.

339. NMa decision in case 2910-700 of 28 April 2004 pts. 22, 31.

(http://www.nmanet.nl/Images/Interpay_tcm16-78624.pdf)

340. NMa decision in case 2910-700 of 28 April 2004 pts. 52 and 53.

341. Ibid., pt. 56.

342. Decison of the 29 December 2006 in case DDF3-580-1-01-DL-EK.

343. Commission Decision of 09.12.1999 in Case IV-34.010 - Nederlandse Vereniging van Banken (1991 GSA Agreement), OJ L 271, 21.10.1999, page 28.

344. Pt. 35 reads: "From the banks' viewpoint, however, it is considerably cheaper to process direct debit orders than acceptance giro transfers (...) ".

345. In recital 36 the Commission addresses the demand side substitutability from the viewpoint of the drawee (the originator) while in recital 38 it analysed the demand side substitutability from the viewpoint of the creditor (the beneficiary, payee).

346. The revenues of a central bank originate from "seigneuriage", which is the consequence of a central banks monopoly -granted by the State - to print chartal money. Seigneuriage revenues derive from the fact that a central bank does not have to reward its liabilities with interest in contrast with a typical commercial bank (the only negligible cost for the central bank is the printing cost). On the opposite side of the balance sheet, the central bank records standard revenue-generating assets; hence a relatively large stream of' profits results as a consequence of this.

347. See also Commission decision of 24 July 2002, Visa II, pts. 48-50 and Statement Objections of 24 September 2003, pts 72-73.

348. 6545, MasterCard submission 5 January 2004, pt. 100.

349. Commission Decision of 24 July 2002, Visa II, pt. 49.

350. Except for in very exceptional circumstances where intermediaries such as E-Bay enable users to pay by card using Pay Pal.

351. Merchants who are willing to accept cheques above this limit lose their protection from fraud or the customer's inability to pay. Also, only very few banks still provide their costumers with internationally accepted cheques that are guaranteed since the guaranteed Eurocheque was faded out by Europay at the end of 2001; see http://www.bundesbank.de/bildung/bildung_glossar_e.php ("Die Einlösung des Eurocheques wurde dem Schecknehmer durch Vorlage der Scheck-Karte bis Ende 2001 zu einem bestimmten Höchstbetrag garantiert. Die Karte dient auch der Abhebung von Bargeld am Geldautomaten und als Debitkarte (mit PIN oder nur mit unterschrift). Der kartengestützte Zahlungsverkehr hat den Eurocheque inzwischen weitgehend verdrängt.")

352. 26509, http://www.ecb.int/pub/pdf/other/ecbbluebook2001en.pdf; Table 12.

353. 6512, MasterCard submission 5 January 2004, pts. 267 to 269.

354. Even this simple functional interchangeability of payment means is only partial, because the way aconsumer accesses his/her funds at a bank deposit is different from cash to cheques or cards. Contrary to POS cards payment, a payment by cash implies that the consumer has with drawn an appropriate amount of cash at an ATM (or physically at a bank branch) in advance of the payment. Also, the time the consumer is debited diverges according to the payment means.

355. In the Eurocheque system banks decided to operate an interehange fee in the opposite direction as in the later Europay payment card system in order to promote cheque acceptance. The interchange fees in the Eurocheque system influenced pricing of cheques at the consumer level. For details see Commission Decision of 10.12.1984, OJ L 35-43, 07.02.1985, pt. 33.

356. See already Commission decision of 9 December 1999, case IV-34.010, Nederlandse Vereniging van Banken (1991 GSA Agreement), recital 32.

357. See T-39-and 40-92, CB and Europay v. Commission, [1994] ECR II-49, at pt. 93.

358. Ibid., at pt. 103.

359. Commission decision of 24 JuIy 2002, Visa II, pt. 53.

360. Retail Banking Research Ltd. Doubts the xistence of pan-European issuers. See RBR Payment Cards in Europe 2004, International Overview, page 13. The Commission tried to obtain figures on payment cards through issued cross-border from MasterCard, but MasterCard could not provide such figures; see reply of 8.9.2003, Q 10.

361. To the extent that banks offer payment cards via internet, internet applications for cards typically appear restricted to customers located in the country of the bank. Even the largest card issuer in Europe issue cards through local branches or local co-operations instead of cross-border through the internet. Neither the Royal Bank of Scotland, Barclays or BNP Paribas offer credit cards through the internet to customers with a zip code outside their country where the bank is located.

362. 21158, ECB Blue Book Addendum incorporated 2001 figures September 2003, page 15, table 7.

363. See Retail Banking Research Litd, Payment Cards in Europe 2004, International Overview, page 1.

364. See "Payment Cards in Europe 2004; International Survey and Analysis" of Retail Banking Research Ltd, July 2003, International Overview, Section 1.4 Card Pricing, page 11-12, Figure 9.

365. 6559, MasterCard submission of 5 January 2004, pt. 46 md 68.

366. See Retail Banking Research Ltd., Payment Cards in Europe 2004, International Overview, page 32.

367. 28419, Commissions merchant market survey, section 14.

368. 11989, reply of [Bank] of 13.5.2004 in the nc version of 1.4.2005.

369. 10700, Minutes of meeting with [German Acquiring Bank] of 27 April 2004, page 9; Minutes of telephone conversation with Mr. N. [German Acquiring Bank] of 5 October 2004.

370. See also 19837, minutes of telephone conversation with [German Acquiring Bank] of 5.10.2004, section 2: "Issuers und [Local Acquiring Bank] können alles vereinbaren und jeden torpedieren").

371. In [EU Member State], merchant acquisition for MasterCard use to be done mainly by [Acquiring Bank] who was also the main acquirer for PIN-based Maestro payments. Visa transactions to the contrary used to be mainly acquired by another quasi-monopolist, [Acquiring Bank]. Both entities were co-owned by [EU Member State] issuing banks. In 2007 both companies started dual branding.

372. In Denmark, Pengeinstitutternes BetalingsSysteme (PBS) is the only acquirer for MasterCard cards in Denmark; see Retail Research Ltd., Payment Cards in Europe 2004, Denmark, page 8, 19, 20.

373. Luottokunta is the only acquirer for MasterCard, Maestro, Visa and Visa electron card transaction in Finland ; See Retail Banking Research Ltd., Payment Cards in Europe 2004, Finland page 26.

374. Unicre even acquirers merchants for Visa, Diners Club and JCB. Only the local debit cards ("Multibankco") are acquired by another monopolist, SIBS.

375. In Belgium, Bank Card Company ("BBC"), Citibank and Europabank currently acquire MasterCard transactions; See Retail Banking Research Ltd., Payment Card in Europe 2004, Belgium, p. 23.

376. AIB, BoI, NIB-NAG and Ulster bank; see Retail Banking Research Ltd., Payment Cards in Europe 2004, Ireland, pages 21 to 23.

377. 21679, nc reply of SEB Kort Norge of 30 April 2004, page 1.

378. In Sxeden, servo used to be the sole acquirer for MasterCard. Acquiring contracts have however, recently been transferred from this acquiring organizations to its member banks. Today, FöreningsSparbanken has approximately 50% of the entire merchant acquiring market, the next large acquirer is SEBM; see see Retail Banking Research Ltd., Payment Cards in Europe 2004, Norway, page 24.

379. In Germany, EURO Kartensysteme (now: ConCardis) has been the monopolist acquirer for MasterCard until July 1999. Today, mainly three large acquirers (ConCardis, B+S Card Service and Citicorp Kartenservice) provide acquiring for Visa and MasterCard. Besides these three acquirers several small acquirers also took up activity in Germany, although most of them on a marginal scale; see 18851, B&S' submission of 13 JuIy 2004, p. 2.

380. In Greece, all MasterCard transactions were acquired by the National Bank of Greece until 1997. Today there appear to be at least nine acquirers for MasterCard according to a public market survey; see Retail Banking Research Ltd., Payment Cards in Europe 2004, page 34.

381. See Retail Banking Research Ltd., Payment Cards in Europe, Italy, page 27.

382. 9089, MastcrCard reply of 27 January 2004 (in its revised version as in the e-mail of MasterCard Europe Sprl of 19 February 2004).

383. 34074.

384. The status of MasterCard as an association of undertakings until May 2006 is not disputed by Mastercard. See the opinion expressed by the legal attorney of MasterCard at the oral Hearing : " But what was the situation before IPO, the historic case, Before May 2006 ? Was Article 81 then applicable ? We do not contest that, then, there was an association of undertakings. MasterCard was owned by the banks and European Board, composed of bankers, had the authority to approve the Ifs. In fact, it is precisely for those reasons, because there was an association of undertakings, that i twas notified 14 years ago. The questions for the historic case are : was the IF a restriction or rather was the IF a necessary arrangement falling outside the scope of Article 81 ? "

385. Pt. 136 of MasterCard's response to the Letter of facts of 16 May 2006.

386. Ibid., pt. 137.

387. 34070, pt. 394.

388. 34206, pt.3.

389. 34187, pt. 63.

390. Pt. 161 of MasterCard's response to the Letter of facts of 16 May 2006.

391. Ibid. pt. 144.

392. Ibid. pt. 125.

393. Ibid. pt. 128.

394. Ibid. pt. 169.

395. Ibid. pI. 174.

396. These "few remaining power", in accordance with the Amended and Restated Certificate of Incorporation cf MasterCard Incorporated, Section 6.7(A), include: (1) a review of applications for membership, (2) fines, (3) intraregional operating rules, (4) assessment and fees to the extent that such assessments and fees do not have an exclusionary effects, (5) intraregional products and enhancement development to the extent that the development initiatives do not relate to competitively sensitive matters, (6) annual expense budget, (7) surplus funds, and (8) affinity and co-branding rules.

397. Pt. 182 of MasterCard's response to the Letter of facts of 16 May 2006.

398. Case 71-74, Frubo v. Commission, [1975] ECR 563, at paragraph 30; Joined Cases 209 to 215 and 218-78, Van Landewuck and Other v. Commission, [1980] ECR 3125, at paragraph 88; and Joined Cases 96 to 102,104,105,108 and 110-82, IAZ and Others v. Commission, [1983] ECR 3369, at paragraph 20.

399. Opinion of Advocate General Léger in Case C-309-99, Wouters, [2002] ECR I-01577, paragraph 61.

400. See in particular the judgement in Wouters, recital 64, and the Commission decision of 13.03.1969 in Case Exposition européenne de la machine-outil, OJ L69, 20.03.1969, pp. 13 to 20. See also legal commentary: "Kommentar zum deutschen und Europäischen Kartellrecht, Volume 2, Europäischen Kartellrecht" by Langen and Bunte, at paragraphs 62 to 64, pp. 65 and 66.

401. Langea and Bunte, Kommentar zum deutschen und europaischen Kartellrecht, Volume 2, Europäischen Kartellrecht, 62.

402. Case T-217-03 and T-245-03, FNCBV cmd others v Commission, judgment of 13 December 2006, not yet reported, at paragraph 4 ("la Féderation nationale des syndicats d'exploitants agricoles (FNSEA), la Féderation nationale boving (FNB), la Féderation nationale des producteurs de lait (FNPL) et les Jeunes agriculteurs (JA) sont des syndicats de droit français.

403. Commission Decision of 23.07.2003 la Case COMP-C2-37.398 - Joint selling of the commercial rights of the UEFA Chanptons League, OJ L 291, 8 November 2003, pages 25 to 55 at pt. 109.

404. Commission Decision of 7 April 1999 in Case COMP-36.147 - EPI code of conduct, OJ L106, 23.04. 1999, pages 14 to 27 at pt. 25.

405. Commission decision of 14 January 1992, OJ L No 37 cf 14 February 1992, pages 16 to 28 at pts. 1 and 26.

406. See for instance Wouters, at paragraphs 56 and 64.

407. Case 123/83, BNIC v Clair. [1985] ECR 391, paragraph 17 and Case C-35-96, Commission v Italy [1998] ECR I-3851, CNSD, paragraph 40.

408. Case C-41-90, Höfner and Elser, [1991] ECR I-1979, paragraph 21; Case C-244-94, Fédération Française des Sociétés d'Assurances and Others v Ministère de l'Agriculture et de la Pêche, [1995] ECR I-4013, paragraph14; and Case C-55-96, Job Centre, [1997] ECR I-7119, paragraph 21.

409. Case 118-85, Commission v Italy, [1987] ECR 2599, paragraph 7.

410. Referrence is made to the detailed obligations on member banks set out in Chapters 3 to 9, as well as in Chapter 18 of the MasterCard International Inc. Rules. [...] [Business Secret - 2a]

411. 24220b, MCI Bylaws Article VII Section 1.

412. See section 7.1.a.

413. For details on Article I Section 4, subsection c of MCII Bylaws and Rules see section 7.1.1.1..

414. See the text of the Amended anti Restated Bylaws of MasterCard International Incorporated as submitted to the SEC on 26 October 2005, 35854, MCI pages 137-194.

415. 34489. MasterCard's reply of 22 December 2006 to the Commission's request for information: "The rules, including the HACR and the interchange mechanism are enforced through the License Agreement entered into between MCI and each of the Licensees. Section 6 of the License Agreement provides that: 'At all times Licensee shall observe all Rules adopted in connection with Authorised Marks as some may be adopted, modified, supplemented, changed or rescinded at any time and from time to time. Such Rules are incorporated herein by reference and made apart of this License Agreement'. The IPO did not impact the way in which the rules are enforced".

416. 35843, page 134; 35854, MCI page 134. Pint Eleventh (C) of Amended and Restated Certificate of Incorporation of MasterCard International Inc., according to which the European Board can be terminated only by the decision of the Global Board taken by 75% of its directors and provided that the directors of the European banks agree by the affirmative vote at least of majority of the votes cast.

417. 35929, MCI page 834, [...] [Business Secret - 3b]

418. 35929, MCI page 1158, [...] [Business Secret - 3b]

419. 35929, MCI page 1158, [...] [emphasis added] [Business Secret - 3b]

420. See recital 141.

421. See sections 2.1.1 and 7.1.1.2.

422. Pt. 169 of MasterCard's reply to the letter of facts of 16 May 2006.

423. Commission Decision 10.12.2003 in Case COMP-E2-37-857 - Organic peroxides, OJ L 110, 30.04.2005, p. 44, at recitals 8 and 9 where the Commission found that the facilitator AC Treuhand violated Article 81 of the Treaty by organising meetings, mediating conflicts, proposing market shares anti hiding incriminating evidence for the cartel members.

424. See section 2.1.3.1.b.

425. Reference is made to all network rules which delegate powers to decide on intra-country

functioning of the network to the local banks such as the rules on intra-country interchange and other fees in MasterCards Bylaws and Rules, Chapter 18 and maestro Global Rules, Chapter 19, 35413-35411.

426. Pt. 174 of MasterCard's response to the letter of facts of 16 may 2006.

427. E.g. this was the case of the delegation of authority to establish MasterCard Brand Domestic UK Multilateral Interchange Programs, see 35921, MCI page 14: [...] [A Senior Manager of MasterCard Incorporated] [...] [Business Secret - 3b]

428. 35927, MCI Page 1543, Presentation discussing pros and cons of Option 4 and 5.

429. See already Commission Decision in Case Cecimo, OJ L69-13, 20.03.1969 and in Case EMO, OJ 1979 L 11-16, 13.01.1979 as well as in Case Assurpol, OJ L37-16, 15.02.1992.

430. Case 45/85, Verband der Sachwersicherer, [1987] ECR 405 at 32 ("The recammendation

constituted the faithful expression of the applicant 's resolve to coordinate the conduct of its members on the German insurance market").

431. See Verband der Sachversicherer/Commission, p.405 at 26-31, sec footnote 429.

432. On MOL "MasterCard On-line"

433. 6542, MasterCard submission 5 January 2004, section 3.4.3.

434. 24220, Bylaws by MasterCard International Inc., Section 10 [...] [Business Secret - 3b]

435. See Section 7.1.a.

436. Loe wenheim/meesen/Riesenkampff, Kartellrecht, Bd 1, Europäisches Recht (2005), Art 81 (1) pt.62.

437. [...] [Business Secret - 3b]

438. See section 2.1.4.

439. 35929, MCI page 834, [...] [Business Secret - 3b]

440. 35929, MCI page 834, [...] [Business Secret - 3b]

441. 35925, MCI page 1258-1259, message of [A Senior Manager of MasterCard Europe SPRL], of February 2005: [...] [emphasis added]. [Business Secret - 3b]

442. See Verband der Sachversichere, 32 ("the recommendation... constituted the faithful expression of the applicant's resolve to coordinate the conduct of its members on the Germany insurance market").

443. Pt.182 of MasterCard's reply to the letter of facts.

444. Case T-39-92 and T-40-92, Groupement des cartes bancaires "CB" et Europay International SA v Commission, [1994] ECR II-49; Verband der Sachversicherer, at pts. 30 to 32.

445. The higher a MIF, the less sensitive the banks are to the level of membership fees charged by the scheme owner. Moreover, MasterCard Incorporated's commercial success in selling network services to banks also depends on the number of banks participating in the scheme. A MIF attracts issuing banks to the detriment of other card schemes operating without a MIF or with a lower MIF. Hence, MasterCard Incorporated shares its member banks' interest in having a MIF.

446. MasterCard's argument that by setting interchange fees in the member banks interests, the new Global Board would necessarily neglect the interests of public (non-bank) stockholders is unfounded.

447. Sec sections 3.1.2. b) and 3.1.9.2.

448. The notion [...] [Business Secret - 5b] is expressly used by MasterCard in correspondence with [...] [Business Secret - 5a] banks (36008, page 12).

449. See for instance 36016 on MasterCard's statements [...] [Business Secret - 5b]

450. See Verband der Sachversichere, at paragraph32; and Case 8-72, Vereniging van Cementhandelaren / Commission (Cement), [1972] ECR 977, at paragraphs 15-22.

451. See Commission decision of 10 December 2003, Organic peroxides, at recital 344.

452. Pt. 128 of MasterCard's response to the Letter of facts of 16 May 2006.

453. Ibid., pt. 125.

454. 35922, page 152. [...] [Business Secret - 2a]

455. 35921, page 33. See section 2.1.4..

456. 35908, pt. 7 on the agenda.

457. See section 2.1.2.a).

458. 35924, page 246, [...] [Business Secret - 3b]

459. This conclusion on the nature of MasterCard's intra-EEA Fallback Interchange Fees is notwithstanding the possibility that decisions of bank delegates on a national forum of MasterCard member banks regarding the level of multilateral fallback interchange fees for domestic card payments qualify as " agreements " or as " concerted practise " between the undertakings represented on such forum.

460. OJ L 336, 29 December 1999, on page 21.

461. Commission decision of 24 July 2002, Visa II.

462. 34070, pt. pt. 396 of MasterCard Response to the Supplementary Statement of Objections referring te paragraph 69 of the Commission's decision of 24 July 2002 in Visa II.

463. Sec, for example, Joined Cases 96-102, 104, 105, 108 and 110-82, NV IAZ International Belgium and others v Commission, [1983] ECR 3369.

464. Cases T-374-94 etc. European Night Services v Commission, [1998] ECR II-3141, para. 136; see also Case 19-77, Miller International Schallplatten v Commission, [1978] ECR 131.

465. Commission Guidelines on Article 81(3) EC [2004] OJ C101-97, para 21. Miller International Schallplatten v Commission judgment.

466. 34052, MasterCard's Response to the Supplementary Statement of Objections, pt. 285 ("As will be set out further in this response to the SSO, MasterCard does not dispute that interchange fees may sometimes function, and could therefore be viewed, as a floor for MSC's") and 475 ("MasterCard has not, in the past, taken issue with the Commission's contention that the interchange fee constitutes a"de facto floor" for MSCs. This is because it is reasonable to assume that interchange fees, as an input cost to acquirers, affects to some degree MSCs. However, it is worth nothing that sometimes, merchant's bargaining power is so strong that acquirers have no choice but to set the MSC at a level below the interchange fee, thereby making a loss. For instance, this is the case in Spain where it is not uncommon for the MSC to be set at a level which is lower the the interchange fee").

467. See also MasterCard's acknowledgement in reply to the first statement of Objections that "it is therefor supposed that the MSC typically reflects ths MIF (...) as well as the acquirer's other costs" (6545, pt. 100).

468. See aboce at section 3.1.1.

469. See at recital 418.

470. See Suppelmentary Statement of Objections of 20 June 2006 (28728) at pt. 241 ("It must be conclude that MasterCard's cross-border interchange fees, as default interchange fee for domestic payment transactions, produce an effect akin to a "recommendation" of an association of undertakings of a minimum price. The default fees are more than simply a "reference fee". The cross-border interchange fees apply automatically in the absence of an explicit agreement by local banks. It therefore sets a minimum level for domestic interchange fees and, in turn, for prices charged to merchants".

471. 34062, MasterCard reply to Supplementary Statements of Objections 15 October 2006, pt. 430.

472. 6542, MasterCard submission 5 January 2004, section 3.4.3.

473. 6545, MasterCard submission 5 Januasy 2004, pt. 100 ("The MSC typically reflects the costs of the MIF").

474. A MIF can of course only restrict cempetition in a market where there is competition. So, to the extent that MasterCard operates with only one acquirer in a market no restriction of price competition arises. Note, however, that a MIF may impede access of new acquirers to such a market (foreclosure), for instance in the manner set out in recitals 325 and 326. Also, the existence of a monopolistic acquirer, that is an inter-bank organisation which is co-owned by several issuing banks, may as such constitute an infringement of Article 81(1) EC. In this vein the Dutch competition authority NMa fined several banks for extending their cooperation within a joint venture (Interpay Nederland B.V.) beyond the technical aspects of the cards business to the joint sales of network services for PIN transactions, as a result of which they eliminated competition in the period from 1 January 1998 to 1 March 2004 (decision in Case 2910-700 of 28 April 2004).

475. See also Commission Decision of 24 July 2002, Visa II, pt. 80.

476. See alos section e (quantitative analysis) and f (merchants comments).

477. Courts and competition agencies in other jurisdictions concur with the Commission's analysis. See, for example, the rulling of the US Court of Appeal for the 2nd Circuit of 08.05.2003 in USA v Visa and MasterCard, Docket No. 02-6074 / 02-6078 / 02-6078, page 7:"...the price merchants pay for acceptance of payment cards (the merchant discount) is affected by the size of the interchange fee, which is set by the network."

478. See sections 18.B.2.6.2. and 18.B.3.7. of MCI Rules.

479. Swedish issuers and acquirers agreed on interchange fees bilaterally; see at 9089, MasterCard reply of 27.1.2004, Appendix G, Country Report: in Sweden the MIF applies to a small part of domestic transactions by default.

480. French issuers and acquirers pay one another interchange fees that are calculated for each pair of banks on the basis of pre-agreed formula; see at 9089, MasterCard reply of 27.1.2004, Q 11, Appendix G, Country report for France; [...] [Business Secret - 6]

481. Until very recently Spanish issuing and acquiring banks bilaterally agreed within each of three networks on interchange fees that are specific per merchant segment); for details on these sector specific fees see 9115 MasterCard reply of 27 January 2004, Appendix G.

482. 21298, the reply of [Bank] of 15.2.2005 and 25719, note to the file on telephone conversation with [Employee of A Bank] of 9.7.2004.

483. See at section 6.3.3.

484. 6504 ('... it is hard to see how the MasterCard MIF could exert any significant effect on the market or impede, distort to undermine competition. Moreover, intra-European cross-border transactions amount to only a tiny fraction of card payments (in the EU approximately [...] [Business Secret - 4] of all MasterCard transactions and [...] [Business Secret - 4] of all payament card transactions)'.

485. See also the subsequent section e (quantitative analysis) and f (merchant comments) of section 7.2.1.1.

486. MasterCard reply of 27 January 2004, see Appendic G.

487. 11046, MasterCard reply of 30 April 2004, pt. 6, Q 3.

488. 11047, MasterCard reply of 30 April 2004, pt. 5.

489. 20378, MasterCard's MOL Database, transmitted to the EU Commission by e-mail on 17 January 2005.

490. 24220b, MCI Rules Section 18.B.2.6.2.

491. 21291, [Bank] reply of 15.2.2005 in the version of 31.3.2005, additional comments after table 1.

492. 28625; 21298, table I and connnents on page 3, 3rd paragraph.

493. Rather than defining "large" and "small" merchants according to an arbitrary benchmark such as "more/less than EUR 1 Mio turnover", a relative approach was used in the model. Acquirers were asked to segregate their client pool in four sub-groups. Group one has generated the first 10% of the total card turnover. Group two consisted on the top 25 percentile of turnover. Group 3 of merchants: bottom 25 percentile of turnover and group 4: bottom decile of turnover. The top decile therefore comprises merchants whose total aggregate sales transaction volume (turnover, in local currency) generated in 2002 represented 10% of the total turnover of all the acquirer's clients.

494. 27513, MasterCard reply of 24 April 2006, Q 13; we use pan-European averages as proxy country averages as country averages are not available.

495. The acronym "MIF min" in diagram 7 stands for the lowest nominal level of the intra-EEA fallback interchange fees in the consumer credit card segment (chip rate) while MIF max stands for the highest nominal level (Base rate). Average MSC stands for the average merchants discount rate, i.e.: the average merchant fee.

496. 27513, MasterCard reply of 14 April 2006, Q 13; we use pan-European averages as proxy for country averages as country averages are not available.

497. The results of this analysis have to be qualified, as follows. MasterCards intra-EEA fallback interchange fees were used as proxy for interchange fees in all EU Member States, i.e. including countries where domestic MIF rates diverge from the fallback rates. This calculation therefore does not yield 100% reliable results for the total share of interchange fees in the MSC. However, for the purpose of benchmarking the relative difference between interchange fee costs in the MSC of small as opposed to large merchants, the use of the IEFB (the intra-EEA fallback interchange fees) as proxy for the weighted average interchange fee cost in each country is appropriate. The observations of one Danish acquirer were eliminated from the sample as merchant fees in that country are regulated by law, which renders a benchmark comparison of weighted average cross-border interchange fees and the regulated MSC purposeless for the present exercise.

498. This particular acquirer is exceptionally co-owned by retailer associations and charges identical MSCs to large and small merchants.

499. All 17 acquirers charged small merchants an MSC above the weighted average interchange fee and 12 also charged top merchants an MSC above the weighted interchange fees. As the diagram shows, for five acquirers (B, E, F, G and I) the share of the IEFB (the intra-EEA fallback interchange fees) as part of the merchant fee exceeds (slightly) 100%. There are two possible explanations for this. On explanation is that the actual weight average interchange cost in these countries may in reality be lower than the weighted average of the cross-border interchange fees which was used as proxy in the analysis. Using the "right" country average, the share of the IEFB of these five acquirers may be below the MSC. A second possible explanation is that the five acquirers have absorbed some percentage points of the interchange fee cost. The commercial rationale for doing this could be that the acquirer valued the overall business relationship with top merchants so much that it accepted making some loss on the acquiring business order to recoup these losses with revenues on other parts of its business relationship (e.g.: mortgages, loans etc.). Even in this hypothesis, however, the analysis shows that the five acquirers passed on the large part of interchange fee costs on to the top merchants.

500. For the sources of the subsequent quotations see the footnote references in the Commission's merchant market survey (cf. 28649) in section 12 with further examples.

501. 28645-28649, Commission Merchant Market Survey 2004, Chapter 12.

502. 28644, Commission Merchant Market Survey 2004, Chapter 14.

503. Ibid.

504. For instance, in Germany the [A Company] reorganised Maestro and ec-cash card acquiring licenses in such as way as to ensure that the licenses for Maestro and ec-cash are no longer held by the same entities. Since Maestro transactions are solely international transactions, Maestro acquirers are no longer obliged under Regulation 2560/2001 to apply the " same " charge as applied domestically and as a consequence Maestro transactions acquired in Germany yield a merchant fee than domestic ec-cash debit card transactions. See minutes of meeting with [Bank] of 27.4.2005.

505. 8847, [Bank] reply of 26.10.2004, Q2b; 19935 [Bank] letter of 5.11.2004.

506. See at recitals 453 and 454.

507. 34052, 474.

508. An interchange fee price floor may for instance fluctuate when interchange fee rates are modified and/or when different interchange fees are fixed per card type and the relative use of each card type evolves over time.

509. 34050.

510. 34052, 475.

511. 34044 at pt. 497, third bullet ("MasterCard is aware of cases where acquirers grant merchantsMSCs below the level of its MIF" "As indicated above, this is the case for Spain where it is not uncommon for MSCs to be set at a level which is lower than the MIF").

512. Issuing and acquiring banks in Spain must be member of one of the three local networks (Servired, Euro 6000 and Sistema 4B). These three networks process domestic credit card payments with MasterCard and Visa and set the level of domestic MIFs (inter-system and intra-system). Note that this was situation at the time of MasterCard's reply to the Supplementary Statement of Objections and that thereafter, under regulatory pressure, the Spanish banks abandoned the approach of fixing an interchange fee rate par merchant segment and also agreed with retailers to reduce the overall level of domestic MIFs.

513. 34053, pt.468.

514. 34053, pt. 470.

515. See section 4.2. Visa's expert argued that competition among acquirers could not be stronger with "at par" clearing than with a MIF, just like it would be hard to assume that if one scraps excise taxes, breweries would compete more keenly.

516. See the summary of oral statement by Ikea and the British Petroleum during the second day of the hearing in section 4.

517. Note that in the Commission's view in the absence of a default MIF banks may or may not enter into bilateral greements on interchange fees. The xistence of a such bilaterally agreed interchange fees is no pre-requisite for the viability of the MasterCard payment card scheme. In the absence of a default MIF prices are established on both sides of the MasterCard scheme as set out in section 7.3.4.1., i.e.: each bank determines its service levels and prices in a manner that maximizes its individual profits.

518. MasterCard used to oblige acquirers to issue cards, as well, until the abolition of the so-called "No Acquiring Without Issue Rule" in 2005. This is why today most of the acquirers also issues cards.

519. 34029, pt. 554; 34048, pt 481.

520. 6455, MasterCard response 5 January 2004, pt. 499.

521. Commission decision of 24 July 2002, Visa II, pt. 80.

522. For more informationon the Single Euro Payments Area projects see section 1.3.3.

523. 20771, See MasterCard's Europe's SEPA Plan for Debits Cards, section 4 on page 7.

524. 21006, speech of Mr. L. to Delegates of Fourth Annual MasterCard Debit Conference, Geneva, Switzerland "SEPA: The Future Beyond Cash" of 10 March 2005, page 6.

525. Both ATM yards (Bankomat) and POS debit cards (Bankomatkasse) were originaly managed by the industry-owned service company GABE Geldausgabeautomaten-Service GmbH. In 1993 GABE and Eurocard Austria GmbH, the Austrian licensee for the Eurocard brand established jointly by the Austrian banks in 1980, merged to form Europay Austria GmbH; see Retail Banking Research Ltd, Payment Cards in Europe 2004, Austria: Page 6.

526. 21002, Sonntagszeitung 30.5.2004, A. Flütsch "Streit mit Banken wegen Maestro": "Nach einer ersten Präsentation von Telekurs (...) rechneten Zahlungsverkehr Spezialisten hoch, dass die Gebühren pro Transaktion gegenüber heute bis zu 400% aufschlagen könnten ". .. ,,Ein nachgebessertes Gebührenmodell dürfte die heutigen Transaktionen im Schnitt nur mehr verdoppen ".

527. 36217.

528. 28104.

529. 25442. point A.

530. 20378, MasterCard MOL data base, country data Finland, since 5.4.2002 set at [...] % [Business Secret - 2]

531. 21262, nc reply of [Bank] of 23.3.2003Q 6.

532. 25444 "Some banks in Finland have in the last years actively pushed the migration from "pure" Pankkikortti and "pure" ATM cards to Visa electron or combined cards. The migration from pure Pankkikortti and pure ATM cards to combined cards and Visa electron has of course some implication with interchange fees. Banks must in the end be aware that all of their products are profitable and a Pankkikortti card does not yield interchange fee revenues as compared to a Visa/MC or Visa electron card. This development is accelerated by the EMV migration which could create costs for upgrading existing debit cards."

533. The ratio between "simple" Pankkikortti cards and Pankkikortti cards that are equipped with a MasterCard or Visa credit/charge card functionality ("combined cards") changes progressively since 1995. Between 1995 and 2004 the number of combined cards grew by 47%, while the number of "pure" Pankkikortti cards only increased by 28%. In 2002 the number of "simple" Pankkikortti cards dropped in total by 8,25% until 2004. In the same time period combined cards kept growing at an accelerated speed, in total by 21,8% (from 946.000 cards in 2002 to 1.210.000 cards in 2004); see 36479, reply of the Finnish Bankers' Asssociation dated 16.1.2007.

534. According to MasterCard the number of MasterCard cards issued in Finland Grew by [1000% to 2000%] between 1998 and 2002; 6468, MasterCard submission 5.1.2004, page 105, Figure 9.

535. MasterCard's further reason to explain the migration of domestic card scheme to Maestro remained unsubstantiated. These were the contention that the domestic card schemes in Luwemburg, in Finland and in Belgium were "loss making" and that MasterCard was at any rate "more effective" than these schemes (see MasterCard rply to letter of facts of 16 May 2007, pt. 408 and pt. 413).

536. Both the ec cash card scheme and the PIN card scheme today still operate successfully without a MIF.

537. In addition, ensuing price increases can be blamed on SEPA.

538. MasterCard reply to letter of facts of 16 May 2007, page 134 to 136.

539. 36008, page 2 [...] [emphasis added] [Business Secret - 6]

540. Mastercard Europe Sprl management offered e.g.: [INCENTIVES]

541. SeE for instance Appendix D tO MasterCard's offer to [...] [Business Secret - 5a] (35990)

542. MasterCard states (34048 pt., 484) that "it is a matter of public record that some of the migrations to whitch the Commission refers are in large part driven by SEPA".

543. ECB: The Eurosystem's Views on a SEPA for Cards, published in 2006, page 2: "The decision to close national card schemes and replace thens with an international one may be driven by the following two considerations: 1) this represents a quick and easy way to adapt to the SCF, and 2) this is an attractive solution to banks as international card schemes typically apply higher interchange fees than national schemes (and the latter tend to be partly retained by the banking system)." (http://www.ecb.int/pub/pdf/other/eurosystemsviewsepacards200611en.pdf)

544. And - if they eventually replace intra-EEA interchange fees as fallback in the Eurozone - the new SEPA fallback interchange fees.

545. To illustrate the economic impact assume that the domestic debit cards in the Netherlands (the "PIN" cards) were re-branded as Maestro cards. If the SEPA interchange fees for Maestro would then apply a "fallback" to domestic card payments in the Nettherlands, merchants and their customers would bear incremental costs of approximately EUR 120 million annually (based on the Chip & PIN rate of the SEPA fallaback interchange fees for Maestro of 0.20% +EUR 0.05 and an estimated total value of payments with debit cards in the Netherlands of EUR 60 billions).

546. 34048, pt. 482.

547. 10127, MasterCard letter of 26 March 2004, page 3 Q 3. The remaining 8% were split between other types of transactions such as those with chip cards.

548. 93539353, MasterCard e-mail 17.2.2004. MasterCard reported to have in issuance as of 17 February 2004, the following: [Bank] (since September 2003) - to date, +/- 500 cards; [Bank] (since Januasy 2004) - to date, +/- 100 cards; [Bank] / [Bank] (since September 2002) - to date, 49 cards; [Bank] (for several years - outside the EU) to date, +/- 5,000 cards.

549. 21638, MasterCard reply 21 March 2005, Table 1. This figure only accounts for cards issued by Principal License Holders in MasterCard's system. MasterCard does not dispose of precise figures on the total number of MasterCard cards issued in the EEA.

550. 6486, MasterCard submission 5 January 2004, figure 7 on page 87.

551. See also Statement of Objections of 24 September 2003, page 29, FN 104 with further references for the development of MasterCard transaction volumes before 1998.

552. COMP 34.324, 338.

553. Until it was transformed in 2007 into a combined rate [...]; [Business Secret 2a] cf. Annex 1.

554. Until it was transformed into a combined rate of [...] [Business Secret - 2a]. cf. Annex1.

555. 34049, 3rd bullet.

556. See section 3.1.9.3.

557. 34048.

558. 6542, MasterCard response of 5 January 2004, pt. 499.

559. See section 3.1.9.3.

560. 6455, MasterCard response 5 January 2004, pt. 499.

561. 32536, Annex 1 to MasterCard's response to the Supplementary Statement of Objections,

Introduction: "The economist expects from a restriction of competition that prices are higher and the output is lower than without that restriction. Yet the interchange fee as set by MasterCard tends to maximise output [..] Thus, the MasterCard interchange fee is not a restriction of competition"; [...] Prof. [...] in the oral hearing: "[...] We have derived that tire profit maximising IF is the volume maximising IF so by ils own interest if that assumption is accepted it is in the own interest of MasterCard to do what we have defined to be the implementation of that IF, which is not a restriction of competition. The system itself maximises competition and government does not have to intervene. This is the point."

562. MasterCard stated that when the European Board set the level of cross-border interchange fees the interchange fees levels were "not the result of opposition, i.e.: a fight", between banks with opposing interests (34046 at pts. 492 and 493).

563. See section 3.1.9.4.

564. 34046.

565. For example, MasterCard Incorporated's management set country specific fallback interchange fees for Maestro in [...] [Business Secret - 5a] after talks with banks showed "tensions on the issuing side". These "tensions" were solved by adopting domestic MIFs that are above the SEPA fallback interchange fees which were then planned to apply as of 1 January 2008 as fallback in the entire Eurozone (see 3.1.10.3 with further details on the adoption of the country specific fallback interchange fees for [...] [Business Secret - 5a] after the IPO).

566. 34044, pt. 497, first bullet.

567. See for example, the statement of Total UK, Sainsbury and British Petrol cited in the Commission's market merchant survey, section 12.

568. See in this context 26627, letter of Fédération des Entreprises du Commerce et de la Distribution in France of 28 January 2005 ("il est inenvisageable de revenir en arrière et de refuser dans nos commerces les cartes Visa ou MasterCard"; 20735 letter of EDEKA of 11 February 2005, page 2 (".. ist es nicht möglich Barzahlungen anstelle von Transaktionen mit Visa/MasterCard vorzunehmen "); 20664, letter of Dansk Handel & Service of 25 January 2005, page 1 ("De handelnde har hverken et reelt alternative til at modtage Visa/MasterCard eller an forhanlingsposition"); 20401, letter of Metro of 28 January 2005 ("In most of our countries Metra sales divisions have to accept credit and debit cards in order to meet the demand of our customers and not to have a competitive disadvantage"); 20668, letter of the Swedish Federation of Trade of 4.2.2005 ("Swedish retailers achieve over 40% of their turnover via Visa and MasterCard ... refusing Visa and MasterCard transactions cannot be considered an option"); 20411/20413, Portuguese Association of Distributors APED of 28 January 2005 ("The Portuguese companies represented by Aped in the food and specialised ratail sector ... cannot even choose to refuse payments with Visa and MasterCard cards, in so far as they represent close on 50% of all consumer payments").

569. See already Commission decision of 24 July 2002, Visa II, pt. 80.

570. According to 90% the main driver was that a high number of their customers wished to use a particular card and 67% stated that they had to keep a customer friendly image. 64% also indicated that they feared to lose business if they were to refuse cards but to their competitors did not ("business stealing affect"). Only between 10% and 39% mentioned transactional benefits as main reasons for accepting cards.

571. See 28668 and 28667, Commission Merchant Market Survey, sections 4 and 5.

572. Only 7% of the merchants surveyed by the Commission ever cancelled such payment cards and none of the merchants ever cancelled Visa or MasterCard, i.e. one of the two main card networks. See 28666, Commission Merchant Market Survey, Annex 1, section 6. Q 4 asked merchants whether they ever cancelled one of MasterCard, Maestro, Visa, Visa electron, Amex, Diners, JCB or the local debit card. See also section 11 on captivity. The market studies which MasterCard submitted (Merchant Acceptance Survey by [...] Dot Econ/TNS and "Retail Payment Mechanisms in Europe" research by [...] [Consultant] are conceptually unconvincing as set out in detail in Annex 3.

573. In particular because the fact that some merchants carniot afford to accept MasterCard cards may precisely be due to bbc MIF.

574. On 1 January 2005, MasterCard also abolished the "No Acquiring Without Issuing" ("NAWI") Rule within the EU so that an eligible member would be entitled te operate as an acquirer if it wished to do so.

575. http://euroea.eu.int/comm/competition/antitrust/cases/29373/studies/.; It would appear that already in the past MasterCard did not sanction breaches of the NDR; see the examples given by MasterCard itself in Exhibit B of MasterCard submission 5 January 2004; see findings of the Dutch competition authority NMa in the Interpay decision of 20 December 2004 where the NMa stated that Interpay was not applying MasterCard's NDR in practice any more (http://www.nmanet.nl/nederlands/home/besluiten/besluiten_2004/2159bbme.asp).

576. 28660, Commission Mervharit Market Survey, Annex 1, section 8.

577. 20300, letter of MasterCard International to acquirers and processors in Germany, as attached to e-mail of [German Acquirer and Processor] of 3.1.2005: "MasterCard erwartet daher nicht von den Acquirern, dass die Barzahlungsklausel aus Verträgen entfernt wird, die über dan 1. Jänner 2005 hinaus gütlig sind. Wenn diese dann mit der Zeit im Rahmen des normalen Vertragszyklus auslaufen und erneuert werden, ist die neue Regelung in den Vertrag aufzunehmen".

578. MasterCard however, does not contest that surcharging is still not widespread in the market although it was allowed since 1 January 2005. Dot Econ stated at the oral hearing "Now, of course surcharging is not widespread or rampant but neither would have expect it to be.".

579. 33265.

580. 28642. 59% of the merchants surveyed in the Commission's merchant market survey in 2004 paid a blended fee for acccpting Visa and MasterCard consumer credit cards.

581. 33228, pt. 111. MasterCard's experts even consider that the decision to surcharge ("price differentiate") is primarily based on the merchants' consideration with respect to the admisnistrative costs.

582. 28667, pt. 4.2.

583. 23040, Interview with MasterCard Europe Sprl's [Senior Manager] in Cards - Karten - cash, issus no. 4, November 2004, Page 10: "Glauben Sie, dass es überhaupt zu einer Weitergabe der Kosten von den Händlern an die Kunden kommen wird? Nein, das glaube ich nicht. Die Erfahrungen aus anderen Ländern wie Grossbritannien, in denen die no-Surcharge Rule abgeschafft wurde, zeigen, dass es hierzu nicht kam. Der Druck auf die Händler, den Kunden nicht zu verlieren, ist einfach zu gross".

584. 20300, letter of MasterCard International to acquirers and processors in Germany, as attached to e-mail of [German Acquirer and Processor] OF 3.1.2005 [...] [Business Secret - 5B]

585. 26483, MasterCard Question and Answers on the Abolition of the No Surcharge Rule in Germany, F 4. ("MasterCard geht nicht davon aus, dass der Handel im EWR rege von der Möglichkeit Gebrauch machen wird, den Kunden Aufschläge zu berechnen. In der Praxis beschränken sich hier die Aufschläge auf Bereiche wie Taxifahrer, Reisebüros sowie Sportveranstaltungen").

586. See section 4.6.

587. Commission guidelines on the application of Article 81(3) of the Treaty, OJ 2004-C 101-08 of 27 April 2004, pt. 25.

588. 34034, pt. 531 ("As already explained ... the MIF is set by MasterCard. The MIF is not set by the licensees. The Commission's assertion that licensees have market power by virtue of their power to set the MIF is therefore incorrect.").

589. Case T-112-99, Métropole Télévision and others v Commission, [2001] ECR II-2459 at paragraphs 115-116.

590. Case 161-84, Pronuptia de Paris v Schilgalis, [1986] ECR 353.

591. Ibid., paragraph 24.

592. Case T-112-99, Métropole Télévision and others v Commission, [2001] ECR II-2459 at para 109 ("Consequently, as the Commission has correctly asserted, examination of the objective necessity or a restriction in relation to the main operation cannot but be relatively abstract. It is not a question of analysing whether, in the light of the competitive situation on the relevant market, the restriction is indispensable to the commercial success of the main operation but of determinating whether i, in the specific context of the main operation, the restriction is necessary to implement that operation ").

593. See Métropole Télévision judgment at paragraph 72-77.

594. Case C250-92, Gottrup-Klim v Dansk Landbrugs AmbA, [1994] ECR-I-5641.

595. Case C-309-99, Wouters, [2002] ECR I-1577.

596. See Gottrup-Klim v Dansk Landbrugs AmbA judgement, paragraphs 35-45.

597. Ibid., paragraph 32.

598. Ibid., paragraph 34.

599. See the Wouters judgment.

600. Ibid., paragraph 100.

601. Ibid., paragraph 107.

602. Ibid., paragraph 109.

603. See tise Gottrup-Klim judgment at paragraph 3; DLG is a cooperative society the object of which is "to provide its members with farm supplies ... at the lowest price".

604. See the Metropole judgment, pt. 104

605. Ibid. pt. 122.

606. 6565, Mastercard submission 5 January 2004, pt. 20 does not specify, however, whether "feasible" relates to the operation of its system or something else. If pt. 20 is read in conjunction with the preceding pt. 19 it would appear that"feasible" means the viability of a system containing an HACR.

607. 6565, Mastercard submission 5 January 2004, pt. 20.

608. 34024, pt. 576.

609. 34021, pt. 589.

610. 34024, pt. 580.

611. For details of the Honour All Cadrs Rule (HACR) see section 3.2.

612. 6540, pt.122.

613. 6540, pt.124.

614. 6515, pt. 244.

615. 6565, pt. 19.

616. 6539, pt. 132.

617. 6538, pt. 139.

618. 6456, MasterCard submission 5 January 2004, pt. 491 ("... issuers would derive insufficient revenues fram interchange fees and from cardholder fees to cover their fixed costs. ").

619. 6456, pt. 491.

620. 6538, pt. 140 -142.

621. 6537, pt. 142.

622. 34068, pts.402 and 403.

623. 34173, pts.106-117.

624. 34068, pts.400 and 401.

625. 34005, pt. 634 and footnote 612.

626. 6515, pt. 244.

627. See Commission guidelines on the application of Article 81 (3) of the Treatry, OJ 2004-C 101-08 of 27 April 2004, pt. 11.

628. In its ruling case T-328-03 of 2 May 2006 the Court annulled a Commission decision on national roaming agreements because the Commission failed to carry out an objective assessment of the "impact of the agreement on the competitive situation" (paragraph 77). The Commission was to assess "what the competitive situation would have been in the absence of the agreement" (paragraph 73). However, the Court did not require the Commission to undertake any analysis of efficiencies accruing from the agreement in order to establish a restriction of competition.

629. 34025, 575.

630. Case C-250-92 Gottrup-Klim v Commission [1994] ECR I-5641.

631. 34024, 577.

632. See Métropole judgment at patagraph 109. The CFI stated that "the examination of the objective necessity of a restriction in relation to the main operation cannot but be relatively abstract. It is not a question of analysing whether, in the light of the competitive situation on the relevant market, the restriction is indispensable to the commercial success of the main operation but of determining whether, in the specific context of the main operation, the restriction is necessary to implement that operation".

633. 34022, 586.

634. Case T-328-03, O2 v Commission, judgment of 2 May 2006 (not yet reported).

635. Commission decision of 24 July 2002, Visa II, pt. 59 with further references.

636. 6540, pt. 122.

637. 6540, pt. 124.

638. 6515, pt. 244.

639. See Commission decision of 24 July 2002, Visa II, pt. 59.

640. Ibid, pts. 28, 69 and 100.

641. 26487, nc e-mail of Mr. K. (Finish Bankers Association) of 8.4.2005.

642. 21725, nc reply [Bank] of 13.4.2005, p.5.

643. 21702, Q 4.

644. 25382, nc reply [Bank] 7.4.2005, Bijlage 2.

645. 21713, nc reply of [Bank] of 12.4.2005, Q 4 ("all Nowegian savings and commercial banks")

646. Franken and Shampine, The Economic effects of Interchange Fees, Antitrust Lax Journal, 2006, No 3 p 655 "By its nature, a network externality is likely to become less important (and a less persuasive justification) as a network matures"; Rochet, The theory of interchange fees: a synthesis of recent contributions, Review of Network Economics, 2003, 2, 97 "Payment card network are also characterized by a more classical network externality ... this externality becomes less and less important as the network matures, when virtually all potential users have joined".

647. This is different to "closed" card schemes such as American Express, Diners or JCB where one single legal entity always acts as issuer and acquirer. On the legal separation see Retail Banking Research Ltd., Payment Cards in Europe 2004, Finland: page 13 and 23; Denmark,: Page 10 and 20; Netherlands: Page 14, 25. Belgium: Page 12 and 23; Norway, page 9, 20; for Belgium, see also nc. Reply of ING Belgium of 30 April 2004, Q 2.

648. Mastercard reply to Letter of facts of 16 May 2007, page 62, pt. 188 (2nd bullet from the bottom).

649. MasterCard response of 5 January 2004, pt. 95 ("Four-party systems are open systems, in the sense that different financial institutions can join the system under the same brand").

650. For instance, this situation occurs in the MasterCard scheme in the UK where banks such as the Royal Bank of Scotland, Barclays and HSBC are both large issuing banks and large acquiring banks.

651. Due to the existence of on-us transactions the boundaries between open ("four-party") card schemes and closed ("three party") card schemes are in practice blurred. The fact that American Express recently started licensing individual banks to distribute its cards on behalf of American Express further softens the structural difference between so-called "closed" and "open" card schemes.

652. 6540, MasterCard submission 5 January 2004, pt. 96, last bullet ("most institutions act as both issuers and acquirers.

653. The inter-bank organisation Europay Austria was founded in 1980 (http://www.paylife.at/plb/opencms/en/Home/About_Europay/Facts_Figures/Company_history/index.html).

654. 23191, Datamonitor, Marchant Acquiring and Transaction Processing 2001, 1st paragraph from the top ("CartaSi is the card issuing and acquiring arm of interbank association Servizi Interbancari, which is made up of the 16 largest Italian banks, the Associazione Bancari Italiana (ABI) and 4 central institutes, the founding partners").

655. 22510, Datamonitor, Central Marchant Acquiring 2000, page 14, 3rd bullet from the bottom setting out that the market was held by two inter-bank associations namely EuroKartenSysteme and B&S Card Services.

656. 22463, Datamonitor, Central Merchant Acquiring 2000, page 61, 2nd paragraph from the bottom. The Belgian inter-bank acquiring organisation Bank Card Company (now: Atos Origin) was founded in 1988 and co-owned by virtually all Belgian issuing banks until December 2006.

657. The inter-bank organisation Luottokuta acquires MasterCard and Visa.

658. 23181, Datamonitor, Merchant Acquiring and Transaction Processing 2001, page 136, 1st paragraph ("Interpay is the principle acquirer for Eurocard/MasterCard transactions with a 95% share of the Dutch market"). See also at 22446.

659. The inter-bank organisation UNICRE acquires MasterCard and Visa.

660. 20306, Datamonitor, Central Merchant Acquiring 2000, page 43, 2nd paragraph from the bottom.

661. See section 4.10.

662. Today, these inter-bank acquiring organisations in the MasterCard scheme still hold strong market positions in [EU Member States].

663. Note that MasterCard's arguments on the objective necessity of a MIF were initialy not geared at international payment card systems but that the alleged need for system pricing and the role of the interchange fee was identified for open card scheme sin general (cf. 6540, MasterCard submission 5 January 2004, pt. 86 and the subsequent section 3.4.1.: "In order to assess the MasterCard default MIF under Article 81 (1), it is necessary to take into account the specific characteristics of such payment card systems in general and four-party payment card system in particular.". Only after MasterCard was confronted with the Supplementary Statement of Objections on 20 June 2006 MasterCard modified its argument and since then claims that a MIF would be intrinsic in the operation of an international four party payment card scheme (cf. 34027 pt. 564 "The Commission does not put forward any evidence that an international four party payment scheme could operate without an interchange fee").

664. [...] [Business Secret - 6]

665. 10741, nc reply of [Bank] of 29.4.2004, accompanying letter, para 3.

666. Retail Banking Research Ltd., Payment Cards in Western Europe 2006.

667. 34017, 3rd bullet from the top.

668. 21697 nc reply of [Bank] of 11.4.2005, page 4, Q 2 Annex ./2, Q 2; see Retail Banking Ltd., Payment Cards in Europe 2004, Denmark Page 20.

669. 37556 ("PBS Holding is owned by most of the Danish Banks, although only three have possession of 5% or more of the capital"). PBS Holding is the mother company of PBS International A/S which "is the only Danish acquirer of MasterCard, Maestro, Visa, Visa Electron and JCB cards".

670. Except for the time period from 1999 to October 2005 when acquiring in Dankort was decentralized while acquiring MasterCard and Visa remained concentrated with PBS. Between 1986 and 1999 and from October 2005 until today the setup of MastrCard and of Dankort on the acquiring side has been exactly the same as PBS acquired transactions in both schemes.

671. 10740. In 1984 the Danish legislator prohibited banks from charging merchants for debit card acceptance. This legislation also fixed merchant service charges for domestic credit card transactions at 0.75%.

672. In these three months interchange fees for physical payment card transactions (card-present payments) were introduced and subsequently again revoked.

673. 37429, at pt. 2.

674. 34017 to 34020.

675. MasterCard reply to letter of facts of 16 May 2007, page 65, pt. 197 (1st bullet).

676. Danske Bank (10740) complained in a reply to the Commission that the prohibition of merchant fees in the Dankort scheme was an "untenable situation", because it would not allow the scheme to operate on "a commercial basis".

677. MasterCard reply to letter of facts of 16 May 2007, page 71, pt. 220.

678. Today, debit cards are used in Denmark to pay for goods on a day by day basis. Card payments account for 56% of all non cash payment instruments (incl. cheques, cards, credit transfers, direct debits and electronic purge) and the share of cards is growing annually. The Danish paymant cards market is increasingly mature as most people fulfilling the criteria for cardholding the Dankort card meanwhile hold such card (37553).

679. 37555.

680. The term "EU 25" stands for the following Member States: "Austria, Belgium, Cyprus, Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Ireland, Italy, Latvia, Lithuania, Luxembourg, Malta, the Netherlands, Poland, Portugal, Slovakia, Slovenia, Spain, Sweden, United Kingdom.

681. 37477 based on statistics of the European Central Bank.

682. 37479 based on statistics of the European Central Bank.

683. 37557.

684. 37430. This sum of 16.6. million is dedicated to cover a portion of the scheme's total operation costs and is borne by larger merchants through annual subscription fees.

685. 37002 to 37012. MasterCard's claim (in MasterCard's reply to the letter of facts of 16 May 2007, page 72, pt. 221) that Dankort was overall "loss-making" remained unsubstanciated.

686. 10676, nc reply of [Bank] of 27.4.2004, Q 1 and 2.

687. 20705, Finnish Bankers' Association, Statistical data on the banks' payment system in Finland 1994-2003, page 6.

688. 37477 based on statistics of the European Central Bank.

689. 37479 based on statistics of the European Central Bank.

690. 20700 to 20699.

691. MasterCard reply to letter of facts of 16 May 2007, pt. 229.

692. 34015 ("Issuing banks in the Pankkikortti debit scheme are subsidised tn three different manners, which operates as balancing mechanism: (...) Therefore, even though interchange fees do not exist as such, transfers of funds in favour of issuers exist (i.e. a form of balancing mechanism").

693. 34015.

694. 34015.

695. MasterCard reply to letter of facts of 16 May 2007, pt. 248, 3rd bullet from the top.

696 Ibid., pts. 232 to 236.

697. A "floor limit" is a threshold above which the cardholder's identity must be verified for the transaction to be authorized.

698. MasterCard reply to the letter of facts of 16 May 2007, page 75, pt. 248, 2nd bullet from the top.

699. Ibid., pts. 239-241.

700. 25392 at Q 1.

701. 40635, Q 1.

702. 21262, nc reply of [Bank] of 24.3.2005, Q 6 : " The process was ongoing then but was stopped last Autumn without result because all banks did not agree to have the fee. As the Pankkikoriti program mis owned by Finnish Banks' Association, the consensus decision is needed to make things accepted. "

703. MasterCard reply to letter of facts, page 77, pt. 247.

704. 36347, 1st bullet from the top.

705. It is finally also unclear how MasterCard's argument on cross-subsidisation (this time on the acquiring side) would relate to the viability of the Pankkikortti scheme and in particular why the situation in Pankkikortti would be any different from that in the international schemes. Merchants regularly pay "bundled" merchant fees for accepting Visa and MasterCard cards (28642. About 60% of the merchants surveyed by the Commission in the Cornmission's merchant market survey paid a blended fee for accepting Visa and MasterCard credit cards).

706. MasterCard reply to letter of facts of 16 May 2007, pt. 244. MasterCard confused the Finnish Banker's Association reply on the level of annual cardholder fees for Pankkikortti usage (EUR 4.80 to EUR 12.00) with the fees Finnish consumers pay for a payment card that carries a Visa or MasterCard logo (so-called "combined cards") which range from zero to EUR 75.

707. MasterCard reply to letter of facts of 16 May 2007, pt. 246.

708. Cf. for instance the rules on floor limits of Eurocard at the time Europay notified the MIF in 1992 on page 34579, 12a ("Intra-European Floor Limit: Limits quoted in ECU, by requested by an Europay International, above which an authorization must be requested by an European acquirer for all European cards used outside their country". Local Floor Limits: Limits quoted in the local currency, set by the acquirer and applicable at Merchant Establishment level. Below these limits it is not necessary for the merchant to request an authorization, Local floor limits must be equal to or lower than intra-European or Inter-Continental Floor Limits") and page 693, section E 7.01. of the Eurocard International S.A. rules.

709. 21725, nc reply [Bank] of 13.4.2005, p.5. The annual cardholder fees for Bancomat were EUR 16 in 2004.

710. 34014, pt. 616.

711. 34014, pt. 616. According to the ECB, which relies on data from 1999, the costs of the Bancomat scheme "are shared by the banks and retailers and by bank customers through the annual card fee. The banks costs are shared between the costumer's bank, which bears one-third of the cost, and the retailer's bank, which covers two-thirds of the cost". MasterCard simply qualifies this cost allocation in the Bancomat scheme as "assymetrical", without however providing a cost breakdown of its debit cards in Luxembourg. The ECB Blue Book does not provide a cost breakdown for the MasterCard and Visa schemes either. On page 323the Blue Book merely mentions that costs in the credit card schemes are "shared by the customers who pay an annual card fee, and the retailers, who pay a retailer discount fee covering the service of guaranted payment for all cards transactions." The ECB also pointed out that acquirers and issuers in this scheme directly charge their respective customers for accepting and using debit cards which is contrary to MasterCard's assumption that without a MIF an open card scheme cannot operate.

712. 25384, nc reply of [Bank] of 7.4.2005, Bijlage 1.

713. Acquiring of MasterCard cards is centralised (Paysquare/Currence).

714. For the figures 1991 - 2000 see 26592, public version of NMa decision case no. 2910-700, Interpay; see also Interpay's Annual Report 2003, page 14, with further details on the increase of PIN transactions in the Netherlands.

715. 37477 based on statistics of the European Central Bank.

716. 37479 based on statistics of the European Central Bank.

717. According to "De Nederlandsche Bank occasional studies" Vol 3, table 2.1 (37537 to 37527) the total number of transactions in 2004 was 1.423 million including debit, credit e-purse. Of this total 1.247 are due to debit card usage. The study specifies that debit card usage is PIN usage (37529). According to the publicly available annual report of Interpay, the scheme owner of PIN, the total number of payment card transactions with PIN payment cards was 1.247,1 in 2004. This confirms the statement in the first source. Thus, PIN debit card usage made up for 87% of total payment card usage in 2004.

718. 34012.

719. See MasterCard reply to the letter of facts dated 16 May 2007 page 116, pt. 259.

720. 36380, Q 3 ("There was no minimum dividend per transaction").

721. 36380 According to [Bank] "... there was obviously no arrangement on the compensation of the issuing banks (it was not possible because of the losses)"; 36377 The NMa also disagrees with MasterCard argument that the dividend mechanism in PIN until 2004 was asking at the MIF ("The NMa therefor does not agree with the claimed fact that there was a de facto interchange fee in place which compensated in issuing banks in the period from 1991 until March 2004. The NMa concluded that the dividend payments were merely and not more than the distribution of profits to Interpay's shareholders" (...). The arguments presented by MasterCard does not prove that the Dutch PIN Scheme ever operated with an interchange fee mechanism in place. They only prove that the Dutch banks from August 2003 onwards wanted some kind of reimboursement/fee for their authorization costs of PIN transactions"). The NMa investigated the dividend payment mechanism of PIN retrospectively at a time where the Dutch issuing banks planned to introduce a MIF in PIN. Dutch banks justified this move in particular by arguing that the historic dividend mechanism of Interpay had effects similar to a MIF. The NMa's investigation therefore focused in particular on the question whether the dividend mechanism has specifically allocated authorization costs as the issuing banks had argued that they needed a MIF to cover these expenses of authorization costs. The NMa concluded that the historic dividend mechanism had no such purpose but merely allocated profits and informed the banks informally of that conclusion and that it could therefore not grant an exemption, also due to a change in the legal regime.

722. 36375.

723. See MasterCard reply to the letter of facts dated 16 May 2007 page 84, pt. 270.

724. The NMa's Financial Sector Monitor gives a good overview over the competitive process since 2004. Merchant fees in PIN have been dropping since Match 2004 due to enhanced competition between Dutch acquirers. See the analysis from page 26 onwards in the Monitor (http://www.nmanet.nl/Images/Financial%20Sector%20Monitor%202005_tcm-16-83961.pdf).

725. [...] McKinsey also mentions that Dutch banks incurred a loss of EUR 779 million with cash payments which is considerably higher than the "losses" with PIN.

726. In March 2004 the Dutch banks decided to transfer the PIN acquiring contracts from their joint venture Interpay to the shareholder banks in order to start offering PIN acceptance services in a fully competitive manner. The transfer of contracts was completed on 1 September 2005. 76% of the 40 largest merchants in the Netherlands (and 12% of all merchants including small ones) immediately enterd into negociations with the banks on the contractual conditions of PIN payment cards acceptance. This lead to a general drop of prices which may explain the losses PIN banks incurred in 2005. Once the market is in equilibrium, this situation may change again. (http://www.nmanet.nl/Images/ Monitor%20Financiele%20Sector%202005_tcm-16-82572.pdf)

727. Different cost allocation methods may yield very different results. With retail banks, cost allocation is particularly difficult and largely arbitrary. Banks typically provide a portfolio of products, including current accounts, overdraft facilities, deposits, personal loans, credit cards and montgages, for both personal and business customers. These are offered through a number of channels (including branches, ATMs, telephone and internet). A significant proportion of a retail bank's operating costs, in particular its operations and call centers, its risk departments, the costs of the branch network and IT systems, will be share by the portfolio of products and customers. In particular it is unclear from the [...] McKinsey study share (and on what basis) of the common costs such as account maintenance, customer service, strategy, administration, marketing and advertising, general overheads etc. has been attributed to PIN debit cards, and which to currents accounts or other payment instruments.

728. In March 2006 the NMa submitted critical observations on the methodology to be used by [...]. The Dutch Banking Federation discarded these observations and announced that they would nevertheless proceed with the study. The NMa's observation related amongst others to the important issue of whether and how overhead costs are allocated to individual payment products. In spite of the NMa's criticism [...] McKinsey continued his methodology unchanged. According to the NMa their "silence" with regards to the study's final results is due to the fact that the NMa's initial comments on the methodology had been ignored by [...] McKinsey and that it would not make sense to give further comments on possible conclusions based on such methodology.

729. 21713, nc reply of [Bank] ef 12.4.2005, Q 4, table 1, 1st and 2nd row.

730. 21714, Q 3.

731. 31618.

732. 31618.

733. Ibid., Norway Page 16, section 4.2.1.; see also the impressive growth of total values with BAX card usage between 1996 (NOK 141,5 billion) and 2005 (NOK 391,8 billion) as set out in the Norwegian Central Banks Annual Report on Payment Systems 2005 at 37491. The total use of international payment cards was NOK 88,8 billion in 2005, i.e.: 22% of total card usage.

734. The fees ranged between NOK 2.07 [EUR 0.268] and NOK 2.50 [EUR 0.324] between 1998 and 2007, 31616, Table 24.

735. 33965-33967 pages 6 to 8 and at 34011, pt. 625.

736. 31990 to 31987.

737. 31989, section 2, last two paragraphs.

738. Note that according to the EFTA Surveillance Authority (36663) the banks' decision not to introduce a MIF in the Bankaxept scheme was a business decision.

739. The Act pre-dates the BankAxept scheme which was not launched until 1991.

740. The reason why the Bank Card Regulation of 1989 was not implemented was due to an agreement between the international card schemes MasterCard, Visa, Amex and Diners with the two merchant associations on a cap for merchant fees. MasterCard errs in alleging that this agreement (also) concerned the BankAxept scheme.

741. 33966 page 7.

742. 36665.

743. 6539, pt. 129.

744. For more details on non MIF related revenues of issuing banks and cost saving related to debit card issuing in section 8.3.3.2. of the Supplementary Statement of Objections of 20 June 2006 (28728).

745. For details on MasterCard's cardholder base see 8,5. b).

746. 34052, pt. 472.

747. 34052, pt. 472.

748. 34173, pt 107; MasterCard states on the allocation of fraud costs: "Instead of this arrangement, alternative Rules could have been adopted by MasterCard whitch allowed issuers to charge back all or wider categories of fraudulent transactions. This would initially have placed the cost of fraud losses onto acquirer, who might have passed this cost onto merchants either by charging back fraudulent transactions to merchants or by charging higher MSC's"; 34172 pt 107; MasterCard states on cardholder default "Once again, the rules could either require the issuer to pay in any event, or permit the issuer not to pay".

749. See the chargeback and settlement manuals at COMP 34.579, 35126 and 35104.

750. 35124. MasterCard's replies to the Commissions requests for information of 1 and 13 December 2006 and in particular MasterCard's reply dated 8 January 2007 on the absence of changes to its chargeback rules, settlement periods etc. despite the reduction of MasterCard's MIF in Australia in 2003; see in particular replies to Q 1-12, 17 and 28-30.

751. 34171, pt. 110.

752. 37517.

753. See section 3.1.6, tables 6 and 7. To compare: The level of MasterCard and Visa credit card interchange fees in Brazil [...] [Business Secret - 2a] and Europe [...] [Business Secret - 2a] is higher than the average in Australia (0.95%) before the Reserve Bank's intervention in 2003. See 36797. In the USA, crerdit card interchange fees of MasterCard and Visa were at an average of [...] [Business Secret - 2a] in 2003.

754. 34173 pt. 107 and 34172 pt. 107.

755. 33977, pt. 717 3rd bullet of MasterCard's Response to the Supplementary statement of Objections.

756. 6469, MasterCard submission 5 January 2004, Footnote 218 on page 119.

757. 6499, pt. 335.

758. Diners Club is the second closed card scheme in Australia.

759. Commercial cards are cards issued by banks to corporate customers whose employees use such cards to cover expenses they incur on business trips and the like. Corporate cards may also be used by staff for private expenses depending on their arrangements with the employer.

760. 6469, MasterCard submission 5 January 2004, pt. 463.

761. In the commercial cards segment MasterCard's market share is approximately [...]

[Business Secret - 4] as that of American Express and it is number [...] [Business Secret - 4] in the market after Visa. 4663, MasterCard reply 1.8.2003, Q 14. MasterCard estimated that its commercial cards account for a total transaction value in the EU of [EUR 10 to 20 billion] (2001), compared to [EUR 20 to 30 billion] (2001) of Visa and American Express [EUR 5 to 10 billion] (2001).

762. 23703, Datamonitor, Commercial Cards in Europe, 2004, p. 18.

763. Ibid, p. 26: "American Express and Diners offer Iittle danger to Visa and MasterCard".

764. Ibid, p.20 and 21.

765. Ibid., p.27. Visa, like MasterCard, sets comparatively higher interchange fees to push commercial cards. Contrary to Visa's consumer cards MIF, the Visa commercial cards MIF was not exempted in 2002.

766. 6469, MasterCard submission 5 January 2004, pt. 464.

767. Note also that according to American Express its strong position in the corporate cards segment is due to innovation ("MC responded that Amex indeed regarded itself as the best provider of corporate cards and that such position of Amex was precisely the result of its innovation efforts. Visa/MasterCard have put much less efforts in this type of cards and that is the reason why they find it difficult to compete with Amex in this field").

768. Pt. 374 of MasterCard reply to letter of facts of 16 May 2007.

769. Such as those related to offering cardholders a free travel insurance, cash rebates, rewards or a free funding period.

770. Even such a dynamic analysis would suffer from inevitable shortcomings related to the need for making assumptions on the concept of profitability and on financial viability. To start with, there exists no commonly agreed concept of "profitability" for the purpose of such exercise. Economists generally distinguish "accounting" profitabllity (a pure comparison of cost and revenues) and "economic" profitability (which enriches the cost revenue comparison by including some opportunity cost such as the rent on the secondary market). What ever approach is eventually chosen, it remains an open question (i) how many of MasterCard's issuing banks (10%, 25%, 50%, 75%, 100%) would need to remain "profitable" to draw valid conclusions for the financial viability of the entire scheme; (ii) to which extent each individual bank must remain profitable to safely assume that it would continue issuing payment cards in the MasterCard organisation in the absence off a MIF (a return on equity of 5%, 10%, 15% or more?).

771. [Bank]'s lawyers provided the Commission services with the confidential version of the memorandum on 5 December 2006 [...] [Business Secret - 7]

772. [Bank]'s lawyers convincingly explained in their letter to the Commission of 12 December 2006 that the redacted information was confidential to [Bank] as it provides an insight into the strategic options of [Bank] how to react hypothetically to the absence of interchange fee proceeds and is "highly confidential" in nature.

773. See 6537 MasterCard submission 5 January 2004 at pt. 369 ("if issuers presented cardholders with the issuers' costs and acquirers presented merchants with acquirers' costs, the outcome would be many fewer completed transactions and ultimately, the demise of the MasterCard payment system").

774. 36212. The RBA stated that: "...contrary to MasterCard's argument prior to the intervention, the scheme did not experience a death spiral. Credit card use in Visa and MasterCard schemes continues to grow".

775. MasterCard confirmed this figure upon request of the Commission. See at 35120, Q 2.

776. MasterCard confirmed this figure upon request of the Commission. See at 35119, Q 2.According to MasterCard its weighted average interchange fee grew between 1 November 2003 and 3 October 2006 to [...] [Business Secret - 2a] and then was again reduced to [...] [Business Secret - 2a] as of 1 November 2006.

777. 36214.

778. 36783. The merchant fees of Amex dropped on average by 20 basis points after the RBA reduced MasterCard's and Visa's interchange fees.

779. See alos the Reserve Bank of Australia's conclusion (cf. at 36781, pt. 35): "It is important to recognize that the main reason that American Express offer relatively high reward points has nothing to do with interchange fees, but rather stems from the relatively high price that American Express has been able to charge merchants for accepting cards".

780. More precisely, according to publicly available statistics of the RBA the market share of the closed schemes increased by 2.1% if calculated on the basis of transaction numbers and by 1.99% if calculated on the basis of transaction values (36782).

781. 36212. The RBA stated that: "...contrary to MasterCard's argument prior to the intervention, the scheme did not experience a death spiral. Credit card use in Visa and MasterCard schemes continues to grow".

782. 35106 to 35107.

783. See MasterCard reply to the letter of facts of 16 May 2007 pts. 352 and 353.

784. See 6537 MasterCard submission 5 January 2004 at pt. 369 ("If issuers presented cardholders with the issuers' costs and acquirers presented merchants with acquirers' costs, the outcome would be many fewer completed transactions and ultimately, the demise of the MasterCard payment system").

785. Ibid.

786. MasterCard reply to the letter of facts of 16 May 2007, pt. 308 and footnote reference 296.

787. See diagram 1 at recital 110.

788. MasterCard reply to letter of facts of 16 May 2007, page 98, pts. 307 and 308.

789. Ibid., page 98.

790. Adding 2% to the market share of MasterCard, the situation is the following. In 2002, the year preceding interchange regulation in Australia, MasterCards market share varied between [[]10 % to 20%] and [[]10% to 20%]. In November 2003, when the RBA's interchange regulation entered into force, MasterCard's market share had risen to [[]15% to 25%]. In 2004 MasterCard's market share further increased to [[]15% to 25%] in the first quarter of the year and to[[] 20% to 30%] in the last quarter of the year. In 2005 MasterCard ended with no less than a [[]20% to 30%] market share in the last quarter of the year. The latest available figures on MasterCard's market share in Australia are those for the third quarter of 2006 when MasterCard's market share reached [[]20% to 30%].

791. US Court of Appeals fort the 2nd Circuit, 344 F.3d 229; 2003 U.S. App. Lexis 19281; 2003-2 Trade Cas. (CCH) P74, 151.

792. 22918, Retail Banking Research Limited, Interchange: The Hidden Dynamo of the Cards Business, August 2001, Chapter 3, Page 77 states that David A. Balto, Assistant Director at the US Federal Trade Commission, published an influential paper on interchange fees in the middle of 2000 where he declared; "A critical linchpin to the Nabanco analysis is the court's finding that significant costs are born by the card issuing bank, primarily of loss and the float. But those casts have changed substantially. The risk of loss is far less substantial in credit cards than it was decades ago. (...) There are fewer regulatory barriers preventing card issuing banks from recovering these costs directly from consumers".

793. Sec 28720, Commission's Supplementary Statement of Objections of 20 June 2006, section 8.4.2 with further references to the "Walmart" settlement in footnote reference 322.

794. See Case 5-69, Volk v Vervaecke, [1969] ECR 295, at Grounds 5/7; and Case C-22-71, Béguelin Import v G.L. Import Export, [1971] ECR 949, at paras 16-18; Case T-374-94, European Night Services and Others v Commission, [1998] ECR II-3141, at paragraphs 93 to 105.

795. See the Volk v Vervaecke judgment, at Grounds 5-7 and Béguelin Import judgment, at paragraph 16.

796. Case 56/65, Société Technique Minière v Maschinenbau Lim, [1966] ECR 235, at page 250.

797. 35990, Appendix D, 1st sentence.

798. See Commission decision of 24 JuIy 2002 in Visa II.

799. Such as the MIF in the domestic debit card schemes Pagobancomat (Italy), the Groupement de Cartes Bancaire's "Commission Interbancaire de Paiement" (France) or the EC cash card scheme's "Händlergebühr" (Germany).

800. 6519, MasterCard submission 5 January 2004, pt. 225: "... the impact of the MIF in setting the MSC is very limited".

801. 6469, MasterCard submission of 5 January 2004, pt. 464 "commercial cards do not represent a significant part of MasterCard's business ... the value of intra-EU transactions made with commercial cards only accounts for [...] [Business Secret - 4] of value" and pt. 484 "the size of Mastercard's commercial cards business is but a fraction of the total. Hence, even if one were to conclude erroneously that the application of the MasterCard HACR to its commercial cards constituted a restriction, that restriction would clearly be de minimis".

802. 18613, MasterCard reply 9.7.2004, page 11.

803. See Case 56-65, Société Technique Minière, [1966] ECR 282, paragraph 7; Case 42/84, Remia and Others, [1985] ECR 2545, paragraph 22, and the Cement case, see footnote 449. See also Case C-306-96, Javico, [1998] ECR I-1983, paragraphs 16 and 17; and Case T-374-94, European Night Services, [1998] ECR II-3141, paragraph 136.

804. Where ever reference is made to Article 81(1) of the Treaty or Article 81(3) of the Treaty, this reference is to be understood as referring to Article 53 (1) of the EEA Treaty or Article 53 (3) of the EEA Treaty, as well.

805. see for instance the judgment of 8 October 2002 in Case T-185-00 and others, Métropole television SA (M6), [2002] ECR II-3805, paragraph 86; Judgment of 15 July 1994 in Case T-17-93, Matra ECR [1994] II-595, paragraph 85; and Judgment of 17 January 1984 in Joined Case43-82 and 63-82, VBVB and VBBB, [1984] ECR 19, paragraph 61.

806. Commission guidelines on the application of Article 81 (3) of the Treaty, OJ C101-08, 27.04.2004, paragraph 49 with further references to the Community court's case law.

807. See Consten and Grundig, paragraphs 347-348, see footnote 292; see also Case T-65-98, Van den Bergh Foods ltd v Commission, [2003] ECR II-04653, at paragraph 139.

808. Ibid, paragraph 50.

809. Ibid, paragraph 51

810. 6499, MasterCard submission 5 January 2004, pt. 337.

811. 6497, section 5.3.3.

812. 6497, pt. 339.

813. 6495, pt. 345.

814. 6496, pt. 345 and 6494, pt. 360.

815. 6494, pts. 360 and 362.

816. 6502, Mastercard submission 5 January 2004, pt. 323.

817. 34007.

818. 34001, pt. 649.

819. 34006, pt. 629, 2nd bullet from the top.

820. In the oral hearing MasterCard expanded this argument by referring to Commission's Visa II decision, footnote 19.

821. 6544, section 3.4.2.3., pt. 360.

822. See section 7.2.1.

823. 34007.

824. 34001, pt 649.

825. In earlier submissions MasterCard used the notion of "the service" or "the MasterCard" service very specifically. See for instance at 6502, MasterCard submission 5 January 2004, pt. 22 where MasterCard defines "the service" as "co-operation enabling service" that is delivered not only by issuers and acquirers but also by the scheme owner MasterCard. During the oral hearing of 14 to 15 November 2006 MasterCard's Associate General Counsel also referred to this broader concept of "the costs of the system" rather than merely costs of "mutually exchanged services' between issuing and acquiring banks.

826. During the oral hearing of 14 to 15 November 2006 MasterCard's Associate General Counsel stated: "We are not allocating the credit losses. We are not specifically allocating any cost. (...) It may be useful to measure somebody's costs. But we are not allocating specific costs. Is that clear?".

827. MasterCard reply to Supplementary Statement of Objections 15 October 2006, 33998, pt. 660 1st to 3rd bullet.

828. Ibid, 34057, footnote 429 ("... the division of costs which is the product of at par settlement (...) In the UK, for example, would be 95% of costs being borne by issuers and 5% of costs being borne by acquirers"). This statement was corroborated with reference to paragraph 22 of MasterCard's Notice of Appeal against a decision of the Office if fair Trading dated 6 September 2005 in which MasterCard, however, did not submit to the Commission. MasterCard also left it open whether the data derive from one single bank or whether they represent and average and how such average was calculated.

829. In the UK for instance issuing banks generated 90% of their revenues with income from cardholders (mainly interests) and only 10% from interchange fees (European card review March 2007, 18; data for the year 2000). The magnitude of issuers revenues from cardholders in the UK show that not only the costs but also the revnues of credot cards may be skewed to the issuing side in the UK market. MasterCard neglects this is assuming an imbalance. Moreover, it should be clarified in this context that in the Visa II decision the Commission accepted a cost benchmark for exempting Visa's MIF for a five year period until 2007. Visa's Board has unlimited discretion for setting interchange fee rates. This cost benchmark is unrelated to the question why Visa's MIF is payed by acquirers to issuers and not vice versa. The Visa II decision expires on 31 December 2007.

830. This also derives from the Baxter framework which MasterCard relies upon as theorrtic basis of its MIF. According to Baxter an interchange fee should "bring the receipts of each bank in equality with the marginal cost it has incurred in providing transactional services to facilitate the exchange" (23473). Baxter is therefore "neutral" on the direction of the transfer payment. The direction will depend on the perceived inequality of marginal costs and marginal revenues. Mastercard's studies are limited in scope to costs.

831. For a contribution towards improvement in production or distribution see T-168-01, GlaxoSmithKline services / Commission, [2006] II-2969, para 247 with further references to preceding jurisprudence.

832. Such as increased sales for merchants; improved cash-flow for merchants; improved security and back-office operations for merchants; payment guarantee for merchants against cardholder defau1t; payment guarantee for merchants against fraud; new sales channels; an alternative, cardholders' ability to defer payment for goods and services over a convenient period; increased personal security for cardholders; fraud protection for cardholders.

833. 34007, pt. 629, 1st bullet.

834. Note that in the Baxter framework, which MasterCard relies upon in defence of its MIF, a MIF may be optimal in either direction, that is paid by acquirer to issuers or vice versa. As will be shown below, the conclusion that a collectively set MIF will maximise system output follows only under very restrictive assumptions. Likewise, a maximisation of system output can not be presumed to be consumer welfare maximising (that is maximising the consumer surplus for cardholders, merchants and subsequent purchasers) on thearectical grounds.

835. 6542, MasterCard submission 5 January 2004, sections 3.4.3 and 5.3.3

836. MasterCard's general references to economic theory in paragraphs 116-118 of its submission of 5.1.2004 do not go beyond paraphrasing (partly inaccurately) same general conclusions that might be drawn from economic models, but do not specify the models used, the assumptions relied on and the facts and data that support the analysis. Similarly, [...] [...] Prof. [...] in his "Economic Analysis of the MIF" (paragraph b) on behalf of MasterCard refers in general terms to and "optimal interchange fee" which is "business volume maximising", but does not present a detailed and robust economic and empirical analysis.

837. 34007, pt. 629.

838. 34006, pt. 629. last bullet ("The Commission is plainly wrong to assert that MasterCard would need to establish that the interchange fee set a certain level is indispensable to achieving the identified benefits. It is the freedom to set the level of the interchange fee so to balance the Scheme which is indispensable to the achievement of the benefits. The Commission's position, on the other hand, amounts to a rather transparent attempt to regulate the level of MasterCard's interchange fees; but it does not have this competence under the Treaty").

839. Case C-552-03 P, Unilever Bestfoods v Commission, [2006] I-09091, paragraph 102; see also Case 42-84, Remia v Commission, [1985] ECR 2545, paragraph 45.

840. Again, it should be noted that an increase in system output does not constitute an objective efficiency if the benefits of increased card usage only accrue to banks, while customers and merchants are worse off due to higher retail prices and increased merchant fees. Hence, evoking the maximisation of system output also requires a convincing analysis that consumers benefit from this.

841. It is likely that Norway, too, displays a card usage above the EU-25 average. However, the ECB does not publish data on card usage per capita in Norway. Based on data from Norwegian statistics (www.ssb.no/english/subjects/02/main.html which indicate a total population in Norway of 4 681 000 inhabitants) and statistics of the Norwegian Central Bank (table 28 on 37491 which indicate total card usage of 864.4 million transactions in 2005) one can estimate that per capita usage amounted to 184 transactions per year in 2005. This would be clearly above EU average. However, the figure of 184 also includes cash withdrawals and the Norwegian statistics do net allow to isolate the over the counter (POS) card usage from cash withdrawals (ATM).

842. European Central Bank, Blue Book, "Payment and securities Settlement systems in the European Union and in the acceding countries addendum", 2005, based on diagram 7.4a on page 34 (37478). The ECB defines "card payments" as payment transactions performed through cards with a debit, credit or delayed debit function at a terminal or via other channels excluding e_money and m-payments (cf. page 712). ATM cash withdrawals and ATM cash deposits with payment cards are also excluded (cf. page 713).

843. 6540, MasterCard submission 5 January 2004, pt. 118 (" Given the nature of four-party payment systems (...), the ulitimate objective of the interchange fee is therefore to promote issuance, acceptance, and ultimately use of the system's card's. The interchange fee operates in such a way that the overall benefits if the system to merchnats and cardholders are maximised, i.e., the car dis issued ta as many consumers as possible and accepted by as many merchants as possible, thereby reducing costs, increasing services levels and contributing to overall economic welfare ").

844. In particular Annex I - The Report by [...] Prof. [...] on the necessity of a MIF due to joint demand and joint supply and considerations regarding the market definition, Annex II - " Economic Evidence " by [...] Dot Econ with a two stage nested model on merchants benefits due to card acceptance in the UK and Annex III with Responses to Linklater's Question by [...] ; - Edgar Dunn & Company - which clarified aspects of the MasterCard Pay Later Cost Studies.

845. 33998, pt. 660 ; 34007, pt. 629 3rd bullet.

846. 23060, Rochet and Tirole, "Cooperation among Competitors: some economics of payment card association", RAND Journal of Economics, 549 at p.5; Katz, 24105 - 24104 at pages 25-26. This was empirically conflrmed in the Commission's merchant market survey. See section 8.2.3.2. with further references to the Commission's merchant market survey.

847. See Katz, 24120 - 24118 (pages 10-12 of the article), on why the use of cards is unlikely to lead to a permanent increase in aggregate sales and consumption.

848. 24102; see in particular page 28 of the article.

849. 33998, pt. 660 last bullet.

850. Section 3.1.8.3.

851. Which still stems from a time where MasterCard used to consider its interchange fees predominantly as "remuneration" or even as "price".

852. 23447.

853. This is graphically illustrated in figure 4 of Baxter's article. 23474.

854. It is interesting to observe, even though not decisive for the assessment at hand, that in correspondence with member banks MasterCard management underlines one other purpose of its cost studies: "Cost studies provide a solid basis for any discussion with regulatory authorities." (36008, page 11, section 3.4, 1st bullet)

855. 33996, pt. 667; 6493, pt. 368.

856. Oral statement of MasterCard's Associate General Counsel at the oral hearing: "The meref act that we do not have a comparable study to measure or estimate or approximate issuers' demand does not mean that we don't take issuers demand into account. There are number of ways that MasterCard approximates or estimates issuers and cardholders demand. One is frankly just through conversations and information that is provided by issuers themselves. Issuers talk to us, they tell us what their desires are, what their plans are, what their needs are so that gives us a very good indication of what it is that issuers believe they need. That said, it does not mean that whatever they ask we give them and I can tell you that very often we don't ask them. We don't give them what they ask for."

857. Annex II to MasterCard's submission of 5 January 2004 at 6680. MasterCard's consultants establish on the basis of a model that if all UK merchants were to charge 2% to 5% to credit card users, total spending with credit cards would drop by between 12% and 39%.

858. 33273, "Merchants acceptance of payment cards" of October 2006 prpared by [...] Dot Econ /TNS.

859. 33094, Survey prepared by [...] Q Research Ltd.

860. "Card holders and merchants will not be willing to pay a price higher than the benefits (which on the merchant side, are not limited simply to transactional benefits) they each derive from the transaction. Similarly, issuers and acquirers will not be willing to supply the service if they cannot recover their costs." (33998 at pt. 660, second bullet)

861. See 28728, Commission Supplementary Statement of Objections of 20 June 2006 in section 8.3.3.2 with examples of cost savings incurred by banks due to debit card usage.

862. 24716 to 24715, Annex 3, page 3-4. [...] [Business Secret - 6]

863. The arbitrariness of this approach becomes even more apparent if one assesses the fees cardholders are prepared to pay in schemes that operate without interchange fees. In Finland, for instance, consumers pay between EUR 4.80 and 12.00 annually for a package for banking services including the use of the domestic debit card Pankkikortti (36478). Assuming that cardholders are "unwilling to pay" for debit card usage and that merchants are "transaction hungry" requires profound empirical evidence.

864. 23473.

865. 33984, pt.701.

866. Moreover, the approach of including non card related cost items on the issuing side while keeping the measurement strictly limited to card related activities on the acquiring side distorts the assessment of an "imbalance" between transactional costs incurred on the two sides of the scheme.

867. This applies to debit cards in the same way as to credit yards, cheques and other means of making payments.

868. It is alsi imaginable that a debit card is issued by an entity with which the cardholder dues not have a "current account relationship". Payment cards and current accounts remain separate products. "Current account infrastructure costs" (customer recruitment, account statements etc.) are not costs that are inherently incurred in payment card transactions but belong to -and depend on- the customer relationship between a bank and its account holders.

869. 6542, MasterCard submission 5 January 2004, section 3.5.2.

870. Sec section 3.1.9.1. and 28788 Supplementary Statement of Objections, section 4.1.6.3

871. See section 3.1.9.1.

872. 33999, pts. 658 to 661.

873. The only " evidence " MasterCard could offer in this respect was the convention that " four-party schemes operating with interchange fees have been more successful overall than any " at par " schemes in term of volumes of transactions " (33995 and 33996 at pt. 668). See, however, in this context the findings of the comparative study by Bolt/Humphrey/Uittenboogaard, 23841 to 23786, which indicates that a " war on cash " may rather be won by an increase of POS terminals than by subsiding the issuing side through interchange fees. Banks have also been very successful in replacing cash and cheques in payment card systems where transactions have traditionally been cleared without a MIF (section 8.3.3.3).

874. See commission decision 24 July 2002, Visa II, pts. 81 and 83.

875. See, for example, 23485, Baxter, Bank interchange of transactional paper. Legal and Economic Perspectives, Journal of law and Economics, vol. XXVI (October 1983), page 541.

876. 33981, MasterCard reply to Supplementary Statement of Objections, pt.712.

877. Commission Guidelines on Article 81(3) EC [2004] OJ C101/97, para 86.

878. 6492, MasterCard submission 5 January 2004, pt., 360-362.

879. 6492, MasterCard submission 5 January 2004, pt., 369.

880. 33993, MasterCard submission 15 October 2006, pt., 673.

881. 33994 to 33986.

882. 33096 to 33273 and 33226 to 33198.

883. 33226 to 33198.

884. 33177 to 33164.

885. 8210, EuroCommerce observations on the statement of Objection of 24 September 2004, pt.4.1.3.

886. Ibid., PT. 4.1.7.

887. 6494, MasterCard submission 5 January 2004, section 5.4.2.2., pt., 360.

888. See also at 28698, supplementary Statement of Objections of 20 June 2006 pt. 456.

889. 29068, Supplementery Statement of Objections of 20 June 2006 pts 469 to 530.

890. For commercial cards MasterCard sets the level of the MIF without any objective benchmark. It therefore is also doubtful that merchants may sufficiently benefit from services that issuing banks provide to corporate cardholders for using commercial cards.

891. The Commission accepted costs related to the free funding period as element in a cost benchmark for setting a cross-border interchange fees in its decision of 24.7.2002, Visa II, OJ L 318/17, pt. 99. However, as the Commission set out in footnote 44 of this decision, the exemptability of the inclusion of a free funding period in a MIF for domestic card payments might conceivably reach a different conclusion. In the framework of the Commissions research with respect to MasterCard's MIF (which applies both to cross-border payments and to domestic payments in eight EEA Member States) the Commission re-assessed to which extent a free funding period may lead to incremental spending at merchant outlets. Tise resuit of its research did net provide support for upholding its assumption in the Visa case of 2002 (see the Commission's merchant market survey et 28665). MasterCard in turn did not provide convincing empirical evidence that the free funding perFiod of its credit cards contribute to incremental spend at merchant outlets accepting such credit cards.

892. 33994, 671 and 672.

893. For instance, MasterCard's Global Rules do not determine the free funding period. It is in each issuers' discretion to offer the benefit to credit card holders and if so, to determine the duration of the interest free period. [...] The ruel MasterCard refers to (MCI By Laws and rules, Rule 9.3.2.) does not contain such obligation but dispose that " each member must pay its merchant for all transaction received from the merchant " whithout fixing the settlement time. [Business Secret - 1a] As MasterCard sets out (33993, pt. 673), [...] [Business Secret - 5a]

894. 6475, MasterCard submission 5 January 2004, section 5.5.3. ("MasterCard's methodology is designed to lead to an optimal MIF')

895. Ibid., section 5.5.3.3.

896. Ibid., section 5.5.3.1.

897. Ibid., section 5.5.3.2

898. OJ L 305, 30.11.1994, p.6.

899. Cases 6 and 7-73, Commercial Solvents and Others v Commission [1974] ECR 223, at pt. 45.

900. Ibid., at pt. 46.

901. See, for instance, Commission Decision 76-353-EEc of 17 December 1975 relating to a procedure under Article 86 of the EEC Treaty (IV-26699 - Chiquita) (OJ L 95, 9.4.1976, p. 1); Commission Decision 97-624-EC of 14 May 1997 relating to a proceeding pursuant to Article 86 of the Treaty (IV-34.621, 35.059-F-3 - Irish Sugar pic) (OJ L 258, 22.9.1997, p. 1); Commission Decision 98-273 - VW) (OJ L 124, 25.4.1998, p. 60); Commission ecision of 21 December 2000 relating to a proceeding under Article 81 of Treaty (Case COMP.F.1-35.918 - JCB)(OJ L 69, 12.3.2002; p.l); Commission Decision 2002-75 8-EC of 10 October 2001 relatinf to a prceeding under Article 81 of the Treaty (Case COMP-36.264 - Mercedes Benz) (OJ L 257, 25.9.2002, p. 1).

902. MasterCard Corporate and MasterCard Electronic Corporate.

903. MasterCard and MasterCard Electronic. MasterCard World Signia is not depicted in the above diagram.

904. 28642-28673.

905. 34137.

906. 28669 ("When evaluating the results of the market survey, one has to be aware, therefore, that the majority of the answers stems from larger undertakings with a relatively high degree of market power vis-à-vis acquiring banks at small and medium undertakings usually not have"). For the sampling methodology of the Commission's merchant market survey generally see chapter 1 of the survey on pages 28672 - 28669.

907. 28671 and 28672, pts. 1.3 to 1.4.

908. "In order to obtain a representative view of larger merchants, whose acceptance behavior is of disproportionate importance, the sample was stratified according to size and sector" (33223, pt. 111).

909. 33249, pt. 51 ("It is equally instructive to consider why merchants do not accept credit cards (including both those who have previously accepted cards, and those who have considered accepting cards but ultimately decided not to do so. Given the extremely poor return form such merchants in the Commission's merchant survey (only 5 out of 218 respandents did not accept at least one credit card) there is little insight the Commission could have gained from its analysis of this important point").

910. 28672, Commission Merchant Market Survey.

911. 28672.

912. 33186, pt. 162 "Where it is common practice in the conduct of market research not to reveal the identity of the sponsor in order to avoid strategic responses, the Commission has not only revealed its identity but has provided an explicit description in the questionnaire".

913. Article 18(3) of Council Regulation (EC) No 1-2003 of 16 December 2002.

914. 33223, pt. 111.

915. Ibid., pts. 111 and 112. According to [...] Dot Econ in its survey a stratified sample of merchants from each country was randomly selected using "the most appropriate source of information on the entire merchant population".

916. See Annex 3 with an in-depth analysis of [...] Dot Econ's survey and "export report".

917. See 28669, pt. 1.8.

918. Sec Annex 3. For instance, an efficient merchant might generate a large amount of card transaction with a few employees.

919. 33186, pt. 165.

920. For the MasterCard's merchant acceptance questionnaire (67 pages) see 33163 - 33096: for the Commission questionnaire (9 pages) see 28641 - 28632.

921. For instance, the French version of the standard request for information used in the merchant market survey (15594.) contained the following sentence on page 3: "Si vous souhaitez obtenir des renseignements complémentaires sur la présente demande, veuillez contacter Mme XX (telephone +32 2 29XX ; e-mail [email protected]) our M. YY (telephone +32 2 2XX; email [email protected])". [Names have been replaced to safeguard the privacy of case handlers concerned]. In practice, the Commission services were contacted by a number of merchants and spent considerable time during 2004 by explaining questions to the 400 merchants surveyed (in particular to merchants of medium and smaller size). Also, the Commission services actively clarified replies obtained if there were doubts as to the correctness of specific statements, in particular if they contradicted others in the same questionnaire, or if a reply was simply unclear or exceptional as compared to the overall picture. It is astonishing that MasterCard interprets these clarifications as attempt to influence merchants in their responding behaviour (cf. 34118, pt. 283).

922. 34125, pt. 265.

923. 34125, pt. 265, 1st bullet.

924. 28639, Q 7 and Q 8.

925. 33145 "Now I will list some reasons for accepting credit cards. Please tell us how far the following statelents are correct..." (free translation form German).

926. 34085, pt. 381, 2nd bullet.

927. 34086, FN 360.

928. 28668, sections 4 and 5 and the underlying questions Q 7 and 8 on page 28639.

929. 28665.

930. 28668, pts. 4.1. to 4.3. (reply to Q 7)

931. 24715 to 24716.

932. For more details see the examples on pages 28665 and 28656 and the analysis of past merchant behavior on pages 28653 and 28654 of the Commission's merchant market survey.

933. to 981. [...] [Business Secret - 8]

982. 34579 or 32536.

983. "Whether this payment platform is a three-party system like the American Express or a four-party system like MasterCard is not of primary interest", page 12.

984. See Jean-Charles Rochet and Jean Tirole, Cooperation among competitors: some economics of payment card associations, RAND Journal of Economics, Vol. 33, No. 4, 2002, p. 556-557.

985. As a theoretical example, consider the case of perfect price discrimination, where output is larger than under uniform pricing, but consumer welfare (that is the consumer surplus for cardholders, merchants and subsequent purchasers) is lower.

986. See Rochet and Tirole, supra note 3, p. 559.

987. Annex 1.2. "Some economics of the interchange fee" develops a theoretical model.

988. Cover memorandum, page 2.

989. [...] [Consultant] study, page iii.

990. [...] [Consultant] study, page 16.

991. This means that the core analysis is on balances.

992. The [...] [Consultant] study motivates the use of balances as dependent variable by referring to the preceding paper of Gross and Souleles "Do liquidity constraints and interest rates matter for consumer behaviour? Evidence from credit card data", Quarterly Journal of Economics, February 2002, with respect to the relationship between credit limits and levels of borrowing on individual credit card accounts in the US. Gross and Souleles find that a $1,000 increase in credit limit results in an increase in balances a year later of between $83 and $138 depending on the specification. They find that most of this increase in balances is not explained by changes in balances on other credit cards and so does not primarily represent balance shifting. However, there remain reasonable alternative explanations of this result other than those that consumers have increased spending.

993. [...] [Consultant] study, page 17.

994. Commission Letter of Facts to MasterCard of 23 March 2007 at pt. 25.