Contents
Judgment of the General Court (Fifth Chamber, Extended Composition)26 March 2025note(Competition — Agreements, decisions and concerted practices — European Government Bonds sector — Decision finding an infringement of Article 101 TFEU and Article 53 of the EEA Agreement — Coordination of prices and bond-trading activities — Exchange of commercially sensitive information — Single and continuous infringement — Restriction of competition by object — Legitimate interest in finding infringements — Calculation of the amount of the fine — Basic amount — Proxy for value of sales — Unlimited jurisdiction)In Cases T‑441/21, T‑449/21, T‑453/21, T‑455/21, T‑456/21 and T‑462/21,UBS Group AG, established in Zurich (Switzerland),UBS AG, established in Zurich,represented by I. Ioannidis and C. Riis-Madsen, lawyers,Natixis, established in Paris (France), represented by J. Stratford and E. Neill, Barristers-at-Law, J.-J. Lemonnier and L. Ghebali, lawyers, and M. García, Solicitor,UniCredit SpA, established in Milan (Italy),UniCredit Bank AG, established in Munich (Germany),represented by I. Vandenborre, S. Dionnet, M. Siragusa, G. Rizza, B. Massella Ducci Teri, M. Lesaffre, T. Selwyn Sharpe and I. Antoniou, lawyers,Nomura International plc, established in London (United Kingdom),Nomura Holdings, Inc., established in Tokyo (Japan),represented by W. Howard, lawyer, M. Demetriou and C. Thomas, Barristers-at-Law, and N. Seay and S. Whitfield, Solicitors,Bank of America N.A., established in Charlotte, North Carolina (United States),Bank of America Corporation, established in Wilmington, Delaware (United States),represented by D. Bailey and D. Gregory, Barristers-at-Law, J. Turner KC, D. Liddell, Solicitor, and D. Slater, lawyer,Portigon AG, established in Düsseldorf (Germany), represented by H.‑J. Niemeyer, M. Röhrig and C. Dankerl, lawyers,vEuropean Commission, represented, in Case T‑441/21, by M. Domecq, S. Baches Opi, M. Farley and T. Franchoo, acting as Agents; in Case T‑449/21, by M. Farley, T. Franchoo and I. Söderlund, acting as Agents; in Case T‑453/21, by M. Farley, T. Franchoo, S. Baches Opi and M. Domecq; in Case T‑455/21, by S. Baches Opi, M. Domecq, M. Farley and T. Franchoo; in Case T‑456/21, by S. Baches Opi, M. Domecq, M. Farley and T. Franchoo; and, in Case T‑462/21, by M. Farley, A. Keidel, T. Franchoo and L. Wildpanner, acting as Agents,THE GENERAL COURT (Fifth Chamber, Extended Composition),composed of S. Papasavvas, President, J. Svenningsen (Rapporteur), C. Mac Eochaidh, J. Martín y Pérez de Nanclares and M. Stancu, Judges,Registrar: M. Zwozdziak-Carbonne, Administrator,having regard to the written part of the procedure,further to the hearings on 6, 7, 8, 9, 12, 13 and 14 June 2023,gives the following
JudgmentOnly the paragraphs of the present judgment which the Court considers it appropriate to publish are reproduced here.
1By their action in Case T‑441/21, brought under Article 263 TFEU, the applicants UBS Group AG and UBS AG (together, ‘UBS’) seek (i) annulment of Commission Decision C(2021) 3489 final of 20 May 2021 relating to a proceeding under Article 101 TFEU and Article 53 of the EEA Agreement (Case AT.40324 – European Government Bonds) (‘the contested decision’), and (ii) a reduction of the amount of the fine imposed on them in that decision.
2By its action in Case T‑449/21, brought under Article 263 TFEU, Natixis seeks annulment of the contested decision in so far as that decision concerns it.
3By their action in Case T‑453/21, brought under Article 263 TFEU, UniCredit SpA and UniCredit Bank AG (together, ‘UniCredit’), seek (i) annulment of the contested decision in so far as it concerns them, and (ii) a reduction of the amount of the fine which has been imposed on them in that decision.
4By their action in Case T‑455/21, brought under Article 263 TFEU, Nomura International plc and Nomura Holdings, Inc. (together, ‘Nomura’), seek (i) annulment of the contested decision in so far as it concerns them, and (ii) a reduction of the amount of the fine imposed on them in that decision.
5By their action in Case T‑456/21, brought under Article 263 TFEU, Bank of America N.A. and Bank of America Corporation (together, ‘BofA’) seek annulment of the contested decision in so far as it concerns them.
6By its action in Case T‑462/21, brought under Article 263 TFEU, Portigon AG, formerly WestLB AG, seeks annulment of the contested decision in so far as that decision concerns it.
I.Background to the dispute
A.The EGB sector
7The conduct at issue relates to European Government Bonds (‘EGBs’), that is to say, sovereign bonds denominated in euros and issued by the central governments of the Eurozone Member States. EGBs are debt securities allowing European governments to raise cash to fund certain expenditures or certain investments, and in particular to refinance existing debt.
8A Member State (the issuer) may thus issue a bond to borrow money (the notional amount) from an investor (the holder) for a fixed term which may be short (for example, 2 years) or long (for example, 10 or 30 years) and at a predefined fixed or floating rate of interest. The bondholder receives the interest (the coupon) from the issuer at regular intervals, as well as the repayment of the notional amount at the agreed maturity.
9EGBs are offered for sale for the first time by, or on behalf of, their issuer on the primary market, and are subsequently traded on the secondary market.
10The functioning of those markets, described in recitals 3 to 51 of the contested decision, is briefly set out below for the purposes of the present judgment.
1.Issuance on the primary market
11EGBs are issued by central governments on the primary market. That issue is often delegated to a debt management office, which defines the procedure for issuing those bonds. The latter procedure generally takes the form either of an auction, which is a tendering process, or of a syndication, which is a private placement process involving a more limited group of operators.
12The bonds may be obtained, in either of the aforementioned forms, only by financial institutions with the status of primary dealers. Those dealers compete on the primary market to acquire EGBs. After acquiring the EGBs, the dealers may keep them or resell them on the secondary market to other financial institutions or investors.
13In an auction, primary dealers submit a bid.
14Primary dealers are subject to various obligations. First, they are obliged, to a certain extent, to take part in the auctions and to acquire and place a significant volume of EGBs on the secondary market during a given reference period. Second, they are generally expected to take on a role as market maker on the secondary market by quoting two-way prices – namely bid prices and ask prices – and by trading at those prices generally and continuously rather than transaction by transaction. The debt management offices exert some pressure on the primary dealers by ranking them, in particular on the basis of the volumes of EGBs that they place and of their participation in auctions. A high ranking may bring with it, inter alia, privileged access to syndications and associated derivatives orders.
15The price at which primary dealers decide to bid at auctions depends on the benchmark price; this is the price at which similar bonds are quoted on the secondary market. If the auction concerns a new tranche (tap) of an existing bond – namely a new supply of EGBs which have previously been issued, with the same original maturity, the same notional amount and the same coupon but sold at the current price – the price information is readily available, since the EGBs concerned are already quoted on the secondary market.
16Within that framework, the price of the bonds is generally set as a percentage of the notional amount, and that percentage is generally expressed by reference to its first two decimal places only.
17In that context, bids made on the primary market by primary dealers in respect of those bonds are often made above the prevailing mid-price – or ‘mid’ – of that bond on the secondary market. The mid-price is the mid-point between the prevailing bid and ask prices on the secondary market and is expressed in cents (namely, to two decimal places). The latter two types of prices are those at which traders on the secondary market are ready respectively to buy or sell the relevant EGBs. The bid made by primary dealers therefore involves a spread around the mid-price.
18When the bidding window is closed, the debt management office of the issuing Member State publishes the auction results and allocates volumes of EGBs to the primary dealers, starting with the highest bidders. The latter bidders generally obtain EGBs for the entirety of the volume that they offered to buy.
19In practice, the level of the bid made by primary dealers will depend on their assessment of the market conditions and the need to win a significant share of the EGB issue in question. Thus, a given primary dealer could place a bid that is higher than the mid-price – known as overbidding – or place a bid that is closer to the mid-price, which is sometimes called a flat bid. Overbidding could, initially, lead to a financial loss on the EGBs acquired when these were issued (since the primary dealer would not be able to re-sell them at the same price on the secondary market); however, the primary dealer can then expect to be able to offset that loss subsequently, on account of its status (which gives access to syndications and other advantages) and its later trading activity on the secondary market.
20In a syndication, the debt management office instructs a group of primary dealers (the lead managers) to assist it in placing a new EGB on the market or in tapping an existing EGB. Those lead managers commit to purchasing the bulk of the EGBs issued and to sell them to investors on the secondary market. Syndication is less common than an auction, but it is generally larger in volume and longer in duration. Syndication is often used by the Member States in order to introduce large volumes of new EGBs.
21Involvement in syndications offers, for primary dealers selected as lead managers, advantages in terms of reputation but also because those lead managers are remunerated by means of underwriting fees. That selection is made on the basis, inter alia, of their ranking, which is based on their results in the auctions in which they have participated (those results being related to the volumes of EGBs acquired and the prices offered to the issuer). Thus, the prospect of being able to participate in syndication is one of the elements driving the competitive bids of those primary dealers at the auctions (see paragraph 19 above).
22The lead managers assist the debt management offices in pricing EGBs for issuance on the basis of a benchmark bond on the secondary market. To that end, they may receive information from their internal EGB trading desks.
23The lead managers must, however, maintain strict confidentiality vis-à-vis their internal desks on account of the involvement of the latter in secondary market trading. Thus syndications are, as a matter of principle, the responsibility of a team that is separate from the internal EGB trading desk of the lead manager bank.
2.Trading on the secondary market
24Unless they decide to hold them until maturity, primary dealers offer EGBs, acquired on the primary market, for sale on the secondary market, in order to trade them either with other banks that may or may not have primary dealer status (known as ‘D2D’ trades), or with other investors (known as ‘D2C’ trades). The EGBs are used for investment or hedging purposes or as a benchmark for determining the price of other assets. Trading takes place ‘over the counter’, namely without a central exchange, electronically or via brokers.
(a)Price formation of EGBs on the secondary market
25The value of an EGB on the secondary market depends on its yield to maturity – which is expressed as an annual percentage of its notional amount – on the fixed coupon and on the gains or losses from trading the bond. The price at which the EGB is traded varies according to a range of factors, such as changes in the economic environment (for example, the general interest rate evolution in the markets or inflation expectations) or issuer-specific factors (for example, the perceived risk, liquidity, or the availability of newer issues). Thus, there is an inverse relationship between the yield and the price of an EGB: a lower yield is translated in a price increase and a higher yield in a price decrease.
26In that context, a ‘yield curve’ is often used to represent the expected yield of EGBs depending on their time to maturity. Such a curve thus indirectly reflects the price of EGBs on the financial markets, taking into account investor expectations in terms of inflation developments, interest rates and general risk (economic environment and financial health of the issuer).
(b)Participants on the secondary market
27As is apparent from recitals 11 and 40 of the contested decision and footnote 10 thereto, primary dealers are generally expected to take on the role of market maker on the secondary market, quoting two-way prices for bonds as described in paragraph 14 above.
28When banks act on the secondary market, they seek to generate revenue by capturing the difference between the bid price and the ask price. The difference between those two prices is known as the ‘spread’ and constitutes the bank’s revenue on a combined buy and sell transaction.
29In managing their portfolios, traders may adopt long or short trading positions. Where traders take a long position on a bond, they hold or purchase bonds before reselling them, thereby speculating on foreseen gains if the value of the bond rises; in other words, the traders expect the price of bonds to increase, which will increase the value of the bonds that they hold. Where traders take a short position, they sell bonds that they do not own. They will have to purchase them at a later date and are speculating that the price of the bonds will fall, so that they can be bought at a lower price than that at which the traders have sold them.
30In the same context of portfolio management, traders are also led to engage in hedging, which is intended, in particular, to offset a long position by selling bonds, and a short position by buying bonds.
B.The administrative procedure which gave rise to the contested decision
31Further to an application of 29 July 2015, on 27 January 2016 the European Commission granted The Royal Bank of Scotland Group plc and The Royal Bank of Scotland plc (together, ‘RBS’) – which subsequently became NatWest Group plc and NatWest Markets Plc, respectively (together, ‘NatWest’) – conditional immunity from fines pursuant to point 8(a) of the Commission Notice on Immunity from fines and reduction of fines in cartel cases (OJ 2006 C 298, p. 17; ‘the Leniency Notice’).
32On 29 June and 30 September 2016, UBS and Natixis respectively applied for a reduction of the fine, pursuant to point 23 of the Leniency Notice.
33On various occasions, in accordance with Article 18(2) of Council Regulation (EC) No 1/2003 of 16 December 2002 on the implementation of the rules on competition laid down in Articles [101 and 102 TFEU] (OJ 2003 L 1, p. 1), or with point 12 of the Leniency Notice, the Commission sent requests for information to the addressees of the contested decision and to other banks.
34On 31 January 2019, the Commission (i) initiated proceedings with a view to adopting a decision pursuant to Chapter III of Regulation No 1/2003, in accordance with Article 2(1) of Commission Regulation (EC) No 773/2004 of 7 April 2004 relating to the conduct of proceedings by the Commission pursuant to Articles [101 and 102 TFEU] (OJ 2004 L 123, p. 18), and (ii) issued a Statement of Objections to, inter alia, UBS, Natixis, UniCredit, Nomura, BofA, Portigon – which had succeeded WestLB – and NatWest (‘the Statement of Objections’). On that occasion, the Commission also informed UBS and Natixis that they were not eligible for immunity from fines, but that they were the second and third banks, respectively, to produce evidence of the suspected infringement and that, a priori, they could be eligible for a reduction of the fines that might be imposed on them, should that event arise.
35All of the banks to which the Statement of Objections was addressed were granted access to the Commission’s investigation file. In accordance with Article 10(2) of Regulation No 773/2004, all submitted written observations on that statement. Pursuant to Article 12(1) of that regulation, they also presented their arguments at a hearing which took place from 22 to 24 October 2019.
36On 6 November 2020, the Commission sent each of the banks to which the Statement of Objections had been addressed, and on which fines could potentially be imposed, a letter (the ‘Letter on Fines’) providing further clarification of the fining methodology proposed by that institution in order to determine the proxy for the value of sales (‘the proxy’). That methodology was accompanied by underlying data and the intended amount of the proxy, for each of the addressees concerned, resulting from the application of that methodology. UBS submitted observations in response on 18 December 2020, UniCredit on 24 November 2020 and 8 January 2021, Nomura on 24 November 2020 and 8 January 2021, and Portigon on 15 January 2021.
37On 12 November 2020, the Commission sent the addressees of the Statement of Objections a letter containing a statement of facts, inviting them to give their views on the factual additions and corrections set out in that statement and relating to elements contained in the Statement of Objections (‘the Letter of Facts’). To that end, those addressees were once again granted access to the Commission’s investigation file. Most of those addressees submitted observations between 8 December 2020 and 8 January 2021.
38On 20 May 2021, the Commission adopted the contested decision on the basis of Articles 7 and 23 of Regulation No 1/2003.
C.The contested decision
1.The finding of the infringement at issue
39In the contested decision, the Commission referred to over 380 discussions between the traders of the banks concerned, via persistent chatrooms – namely, essentially, two chatrooms named ‘DBAC’ and ‘CODS & CHIPS’ – and via bilateral communications, by telephone or instant messaging, over a period between 3 January 2007 and 28 March 2012.
40According to the Commission, those discussions involved exchanges of commercially sensitive information which allowed the banks concerned to be informed about each other’s conduct and strategies, and, in that way, may have allowed them to adjust or otherwise coordinate their conduct and gain competitive advantages when EGBs were issued, placed on the market and traded (see recitals 94 and 95 of the contested decision).
41In that context, the Commission considered that the overall aim of the collaboration between the traders was to help each other in their operation on the market, by reducing uncertainties regarding the issuing and/or trading of EGBs, with the general purpose of increasing the revenues earned on both the primary and secondary markets and resulting from the participation of the banks concerned in EGB issues and subsequent trading thereof. In doing so, those banks knowingly substituted practical cooperation between them for the risks of the market, to the detriment of other market participants, their customers or debt management offices.
42According to the Commission, the discussions at issue can, for analytical purposes, be grouped into four categories of conduct which are intertwined and partially overlap (see recitals 93 and 95, together with Section 5.1.3.2 of the contested decision).
43Those four categories of conduct consist, respectively, in:– first, attempts to influence the prevailing market price on the secondary market in function of the conduct on the primary market (‘Category 1’);– second, attempts to coordinate the bidding on the primary market (‘Category 2’);– third, attempts to coordinate the level of overbidding on the primary market (‘Category 3’);– fourth, other exchanges of sensitive information, including exchanges relating, in the first place, to pricing elements, positions and/or volumes and strategies for specific counterparties related to individual trades of EGBs on the secondary market (‘Category 4(i)’); in the second place, individual recommendations given to a debt management office (‘Category 4(ii)’) and, in the third place, the timing of pricing of syndicates (‘Category 4(iii)’) (together, ‘Category 4’).
44As regards the legal classification of the conduct at issue, the Commission found that this formed part of a single and continuous infringement consisting in agreements and/or concerted practices which had as their object the restriction and/or distortion of competition in the EGB sector within the European Economic Area (EEA).
45In order to arrive at that finding, in the first place, the Commission considered that the conduct at issue constituted a single and continuous infringement having regard, in particular, to the content of the discussions, the methods employed, the systematic nature of some of that conduct and the participants involved. In particular, that conduct formed part of a common plan, since a stable group of individuals was involved in the cartel, the means of communication used were generally the same for the different practices, and the discussions were frequent, took place over the same period and covered the same or overlapping topics (see recitals 412 to 421 of the contested decision).
46In the second place, the Commission considered that the object of the single and continuous infringement at issue was to restrict and/or distort competition through the sharing of commercially sensitive information between members of a circle of competitors with the aim of coordinating their strategies for acquiring EGBs on the primary market and/or trading those bonds on the secondary market. That conduct was capable of impacting the normal course of pricing components for EGBs (see recital 495 of the contested decision).
47In the third place, the Commission found that, having regard, in particular, to the fact that the conduct at issue had occurred from trading desks situated in the European Union and EEA, that conduct had been implemented in the EEA and had substantial, immediate and foreseeable effects on competition in the EEA. Accordingly, that conduct was capable of having an appreciable effect on trade between Member States and Contracting Parties to the EEA Agreement, with the result that the Commission had jurisdiction to apply both Article 101 TFEU and Article 53 of the EEA Agreement (see recitals 725 and 726 of the contested decision).
48In the fourth place, the Commission considered, given, in particular, that the conduct at issue constituted a restriction by object, that none of the conditions under Article 101(3) TFEU and Article 53(3) of the EEA Agreement was satisfied in the present case (see recitals 728 and 731 of the contested decision).
2.The imposition of fines
49As regards the imposition of a fine on the banks concerned in respect of their participation in the infringement at issue, in the first place, the Commission stated that, pursuant to Article 25 of Regulation No 1/2003, its power to impose a fine was time-barred in so far as concerns BofA and Natixis, on account of the fact that their respective involvement had ended more than five years before the start of the investigation (see recital 778 of the contested decision).
50Pursuant to the final sentence of Article 7(1) of Regulation No 1/2003, the Commission nevertheless found the infringement at issue against those two banks (see recitals 780 to 784 of the contested decision).
51In the second place, as regards the other banks concerned – that is, UBS, UniCredit, Nomura, Portigon and NatWest – the Commission stated that it had determined the amount of the fines in accordance with the Guidelines on the method of setting fines imposed pursuant to Article 23(2)(a) of Regulation No 1/2003 (OJ 2006 C 210, p. 2; ‘the Guidelines’) (see recital 807 of the contested decision). However, taking into account the particularities of the financial sector and, in particular, the fact that financial products such as EGBs did not generate sales in the usual sense, the Commission, rather than using a ‘value of sales’ linked to the issue and trading of EGBs, deemed it appropriate to determine a proxy. The latter was based on, first, the notional amount of EGBs traded on the secondary market by the banks concerned over the course of their respective periods of participation in the infringement at issue and, second, the bid-ask spread of 32 representative categories of EGB (see recitals 813 to 832 of the contested decision).
52As to the remainder, the Commission calculated the amount of the fine for each of the banks referred to in the preceding paragraph by establishing, as is provided in the Guidelines, a basic amount, determined having regard, in particular, to the gravity and duration of the infringement, and by adjusting the basic amount in the light of considerations such as any aggravating or mitigating circumstances, the maximum legal threshold for the fine or rules stemming from the Leniency Notice.
53In that connection, first, the amount of the fine imposed on Portigon was capped at EUR 0, pursuant to the maximum threshold of 10% of total turnover during the preceding business year (as provided under Article 23(2) of Regulation No 1/2003 and point 32 of the Guidelines), having regard to the fact that that bank was gradually ceasing its activities and that Portigon’s net turnover in 2020 was negative (see recital 893 of the contested decision). Second, pursuant to the Leniency Notice, the Commission granted immunity from fines to NatWest and a reduction of 45% of the amount of the fine imposed on UBS, in respect of their respective cooperation in the investigation (see recitals 895 to 903 of the contested decision).
3.The operative part
54The operative part of the contested decision reads as follows:‘Article 1The following undertakings have infringed Article 101 of the Treaty and Article 53 of the EEA Agreement by participating, during the periods indicated, in a single and continuous infringement regarding EGB covering the entire EEA, which consisted of agreements and/or concerted practices that had the object of restricting and/or distorting competition in the EGB sector:– [BofA] participated from 29 January 2007 until 6 November 2008, with respect to the CODS & CHIPS chatroom;– [Natixis] participated from 26 February 2008 until 6 August 2009;– [NatWest] participated from 4 January 2007 until 28 November 2011 and NatWest Markets NV participated from 17 October 2007 until 28 November 2011;– [Nomura] participated from 18 January 2011 until 28 November 2011;– [Portigon] participated from 19 October 2009 until 3 June 2011;– [UBS] participated from 4 January 2007 until 28 November 2011;– [UniCredit] participated from 9 September 2011 until 28 November 2011;Article 2For the infringement(s) referred to in Article 1, the following fines are imposed:– [NatWest] and NatWest Markets NV jointly and severally liable: EUR 0– [Nomura]: EUR 129573000– [Portigon]: EUR 0– [UBS]: EUR 172378000– [UniCredit]: EUR 69442000…Article 3The undertakings listed in Article 1 shall immediately bring to an end the infringements referred to in that Article insofar as they have not already done so.They shall refrain from repeating any act or conduct described in Article 1, and from any act or conduct having the same or similar object or effect.…’
II.Forms of order sought
55In Case T‑441/21, UBS claims that the Court should:– principally, annul the contested decision;– in the alternative, reduce the amount of the fine imposed on it to EUR 51.3 million in accordance with the methodology based on UBS’ ‘net value traded’ methodology; or reduce that amount to EUR 60.6 million in accordance with UBS’ ‘adjusted net value traded’ methodology; or reduce that amount by at least 65% as a consequence of the errors and inaccuracies identified in the Commission’s methodology; and– order the Commission to pay the costs.
56In Case T‑449/21, Natixis claims that the Court should:– annul the contested decision in so far as it concerns Natixis;– order the Commission to pay the costs.
57In Case T‑453/21, UniCredit claims that the Court should:– annul Articles 1 and/or 2 of the contested decision in so far as they concern UniCredit;– in the alternative, reduce substantially the amount of the fine imposed on it;– order the Commission to pay the costs.
58In Case T‑455/21, Nomura claims that the Court should:– annul Article 1, fourth indent, of the contested decision;– in the alternative, annul Article 2, second indent, of the contested decision;– in the further alternative, reduce substantially the amount of the fine imposed on Nomura;– order the Commission to pay the costs.
59In Case T‑456/21, BofA claims that the Court should:– annul the contested decision in so far as it concerns BofA;– order the Commission to pay the costs.
60In Case T‑462/21, Portigon claims that the Court should:– annul the contested decision in so far as it concerns Portigon;– order the Commission to pay the costs.
61In Cases T‑441/21, T‑449/21, T‑453/21, T‑455/21, T‑456/21 and T‑462/21, the Commission contends that the Court should:– dismiss the action in its entirety;– order the applicants to pay the costs.
III.Law
62After hearing the parties, the Court has decided to join Cases T‑441/21, T‑449/21, T‑453/21, T‑455/21, T‑456/21 and T‑462/21 for the purposes of the present judgment, in accordance with Article 68(1) of its Rules of Procedure.
A.The applications for the omission of certain data vis-à-vis the public
63In the course of the proceedings before the Court, UniCredit, Nomura, BofA and the Commission requested, in essence, that certain data be omitted vis-à-vis the public in the present judgment, pursuant to Articles 66 and 66a of the Rules of Procedure.
64In that regard, UniCredit stated that no information contained in the procedural documents which it filed in connection with its action is secret or confidential and thus should be omitted in the text of the judgment to be delivered, with the exception of the name of its trader and several literal citations from the leniency statements of UBS and RBS.
65Nomura stated that the names of, and information that might make it possible to identify, the traders involved in the infringement at issue, the experts who have assisted it in connection with its action and in the administrative procedure which led to the contested decision as well as the companies employing those experts, and certain bond issuers should be omitted from the text of the judgment to be delivered.
66Lastly, BofA stated that (i) the names of, and information that might make it possible to identify, its employees or former employees; (ii) the content of the discussions that took place in the chat rooms, of the leniency statements and of the replies to the Commission’s requests for information; (iii) the source references for the elements referred to in (ii); (iv) quotations from the leniency applicants’ oral statements; (v) strong language or derogatory comments by its traders in relation to third parties; and (vi) references to conduct that took place outside of the infringement period should be omitted from the text of the judgment to be delivered.
67The Commission also proposed that the names of the traders involved in the conduct complained of be anonymised.
68In that regard, it should be recalled that, in reconciling the need to make judicial decisions public, on the one hand, and the right to protection of personal data and of business secrets, on the other, the court must seek, in the circumstances of each case, to find a fair balance, having regard also to the public’s right of access, in accordance with the principles set out in Article 15 TFEU, to judicial decisions (see judgment of 27 April 2022, Sieć Badawcza Łukasiewicz – Port Polski Ośrodek Rozwoju Technologii v Commission, T‑4/20, EU:T:2022:242, paragraph 29 and the case-law cited).
69Furthermore, the confidential treatment of an item of information is not justified in the case, for example, of information which is already public or to which the general public or certain specialist circles have access, information featuring also in other passages or documents in the case file in respect of which the party seeking to preserve the confidential nature of the information in question neglected to make a request to that effect, information which is not sufficiently specific or precise to reveal confidential data, or information which is largely apparent or may be deduced from other information which is legitimately available to the interested parties (see order of 14 March 2022, Bulgarian Energy Holding and Others v Commission, T‑136/19, EU:T:2022:149, paragraph 9 and the case-law cited).
70In addition, information which was secret or confidential, but which is at least five years old, must, as a rule, on account of the passage of time be considered historical and therefore as having lost its secret or confidential nature unless, exceptionally, the party relying on that nature shows that, despite its age, that information still constitutes an essential element of its commercial position or of that of interested third parties (see judgment of 19 June 2018, Baumeister, C‑15/16, EU:C:2018:464, paragraph 54 and the case-law cited).
71In the light of the foregoing, the Court has decided, in the present case, as regards the public version of the present judgment, to anonymise only the names of the natural persons referred to in the contested decision and the names of the experts used by the banks concerned as well as of the companies which employ them.
72First, there is no need to treat as confidential the figures used by the Commission in calculating the amount of the fines imposed on UBS, UniCredit and Nomura, in view of (i) the fact that no application to that effect has been made by those banks; (ii) the fact that all of those figures will be more than 13 years old as at the date of delivery of the present judgment; (iii) the fact that those figures are mainly relative or estimated figures and have already been sent to all the banks concerned, as evidenced by the versions of the contested decision notified to those banks; and (iv) the need to ensure that the reasoning of the present judgment is comprehensible, especially where some banks are claiming that the Commission made errors of calculation, the examination of which naturally requires reference to the figures used by the Commission.
73Second, there is no reason to redact the content of the discussions at issue, even if they might disclose strong language or derogatory comments by the traders of the banks concerned, particularly in relation to third parties.
74First of all, as is apparent from the contested decision, those discussions constitute almost all of the evidence on which that decision is based and the messages exchanged in the context of those discussions reveal, according to the Commission, the anticompetitive nature of the conduct of the banks concerned.
75Next, almost all of the messages contained in the discussions at issue appear in the public version of the contested decision and do not, therefore, warrant any protection. The same applies to the names of the bond issuers which Nomura seeks to have omitted.
76Lastly, as regards the excerpts from discussions not referred to in the contested decision but which the Court will make use of in the present judgment, their reference is justified by the requirement to provide an intelligible response to the arguments raised by the applicants and, in particular, by BofA.
77It is also that requirement which substantiates the Court’s finding that there is no need to treat as confidential the excerpts from documents – including the leniency documents – contained in its case file and which it considers necessary to mention when examining the arguments raised by BofA.
78Third, as regards the references to the sources of the evidence used by the Commission and referred to in the contested decision and the conduct of BofA that is unrelated to the infringement at issue, suffice it to state that they will not be used in the present judgment.
B.The object of the infringement found in the contested decision
79The pleas raised by several applicants alleging errors in the characterisation of the conduct at issue as a ‘single and continuous infringement’ and an ‘infringement by object’ as well as the Commission’s defences reveal a difference in understanding of the object of the infringement found in the contested decision.
80On the one hand, certain applicants understand, in essence, that, by Article 1 of the contested decision, the Commission found a single and continuous infringement composed of multiple separate infringements.
81In that regard, UniCredit suggests, inter alia in its fourth plea, that the Commission found a single and continuous infringement composed of two single and continuous infringements, the first relating to the primary market for EGBs and the second relating to the secondary market for EGBs.
82Similarly, in maintaining, in the case of Nomura, that the Commission erred in placing certain of the discussions at issue in one or more of the categories referred to in recitals 93, 382 and 496 of the contested decision (second plea) or, in the case of Portigon, that the discussions referred to in Category 4 could not be regarded as having an anticompetitive object, unlike those in Categories 1 to 3 (first part of its fifth plea), those banks suggest that the Commission found a single and continuous infringement itself composed of four single and continuous infringements, which correspond to each of those Categories 1 to 4 and which should therefore be examined separately.
83On the other hand, the Commission disputes those readings of the contested decision and maintains that the banks concerned participated in one single and continuous infringement; according to the Commission, whether that infringement reveals a sufficient degree of harm to competition should be assessed ‘overall’ in order to determine whether it can be regarded as having an anticompetitive object, and not category by category or even discussion by discussion.
84It should be noted that the exact determination of the scope and number of infringements found in the contested decision is essential in order to assess the legality of that decision. It is also essential in order to determine the effectiveness of certain arguments raised by the applicants, in particular those alleging errors in the classification of certain forms of conduct in one or other of the categories referred to in recitals 93, 382 and 496 of the contested decision, errors in the characterisation of a single and continuous infringement and in the participation of the banks concerned in the infringement at issue, and errors in the characterisation of the conduct at issue as a ‘restriction by object’.
85In that regard, the Court of Justice has held that, while a complex of practices may be characterised as a ‘single and continuous infringement’, it cannot be inferred from that that each of those forms of conduct must, in itself and taken in isolation, necessarily be characterised as a separate infringement. That can be the case only if, in the decision concerned, the Commission has chosen to identify and characterise as such each of those forms of conduct and then to provide evidence of the involvement of the undertaking concerned to which they are attributed (see, to that effect, judgment of 16 June 2022, Sony Corporation and Sony Electronics v Commission, C‑697/19 P, EU:C:2022:478, paragraph 67).
86Moreover, in interpreting the decision concerned, the General Court must take care not to confuse the concept of ‘conduct’ with that of ‘infringement’, otherwise it would err in law (see, to that effect, judgment of 16 June 2022, Sony Corporation and Sony Electronics v Commission, C‑697/19 P, EU:C:2022:478, paragraph 78).
87In the present case, it is therefore necessary to determine whether, in the contested decision, the Commission found that there was only one single and continuous infringement with an anticompetitive object or, on the contrary and as UniCredit, Nomura and Portigon suggest, that there was a single and continuous infringement, composed of two or four additional, separate and independent single and continuous infringements, each with an anticompetitive object.
88In that regard, there is no doubt that, in the contested decision, the Commission criticises the banks concerned for their participation in only one single and continuous infringement, and that any other interpretation results from a misreading of the contested decision.
89First, Article 1 of the contested decision merely finds that the banks concerned ‘have infringed Article 101 of the Treaty and Article 53 of the EEA Agreement by participating, during the periods indicated, in a single and continuous infringement regarding EGB covering the entire EEA, which consisted of agreements and/or concerted practices that had the object of restricting and/or distorting competition in the EGB sector’.
90Thus, unlike Article 1 of the decision at issue in the case giving rise to the judgment of 16 June 2022, Sony Corporation and Sony Electronics v Commission (C‑697/19 P, EU:C:2022:478), which referred to a single and continuous infringement composed of several separate infringements, Article 1 of the contested decision makes no reference to any other infringement which may correspond to the forms of conduct identified in the four categories of conduct referred to in recitals 93, 382 and 496 of that decision and mentioned in paragraphs 42 and 43 above.
91Second, that reading of the wording of Article 1 of the contested decision is confirmed by the statement of reasons for that decision, in the light of which its operative part must be interpreted, should the need arise (judgment of 15 May 1997, TWD v Commission, C‑355/95 P, EU:C:1997:241, paragraph 21).
92First of all, in recitals 424, 469, 484, 737, 781 and 841 of the contested decision, the Commission repeatedly uses the singular to refer to the infringement in which it considers that the banks concerned participated.
93Next, the structure of the contested decision highlights the fact that the Commission first explained the reasons why approximately 380 discussions formed a single and continuous infringement (Section 5.1.2) and then stated the reasons why that single infringement had an anticompetitive object (Section 5.1.3).
94Lastly, in recitals 93, 382, 496, 516, 549 and 550 of the contested decision, the Commission expressly stated that the four categories of conduct which it had identified were described for analytical purposes and that they were intertwined and partially overlapping, which precludes conduct coming within one or other of those four categories from also being classified, separately and independently, as infringements of Article 101(1) TFEU.
95It follows from the foregoing that the complaints based on errors of classification of the conduct at issue in one or other of the four categories of conduct referred to in recitals 93, 382 and 496 of the contested decision, and their consequences for the characterisation of the conduct at issue as a ‘single and continuous infringement’, and as a ‘restriction by object’, or based on errors in the characterisation of one or other of those four categories of conduct as a ‘restriction by object’, are ineffective and must therefore be rejected.
96However, the arguments put forward by UniCredit, Nomura and Portigon against specific discussions will be taken into consideration subsequently in order to determine whether the Commission was entitled, without making any error, to characterise the discussions at issue, taken as a whole, as a single and continuous infringement with an anticompetitive object.
97Furthermore, it must be pointed out that, in the contested decision, the Commission criticised the banks concerned for their participation in the discussions analysed in the body of that decision and in the table annexed thereto. That table refers, for certain discussions that took place on the same day, to several timestamped excerpts.
98However, contrary to what Portigon implicitly submits, such a presentation does not imply that the Commission intended to assess the anticompetitive nature of the discussions at issue on an intraday excerpt basis.
99It is apparent from both the body of the contested decision and that table that the Commission’s assessment relates to all excerpts from a given day. It is also in that regard that the Commission stated, in response to a question from the Court, that the aim of specifying time slots in the contested decision was to structure the evidence and make the decision more intelligible.
100Accordingly, when examining Portigon’s claims that the Commission erred in characterising the discussions at issue as a ‘single and continuous infringement’ and as a ‘restriction by object’, the Court will examine the discussions relied on against that bank on a daily basis.
C.The applicants’ claims for annulment
…
105The application for annulment of the contested decision made by UBS must therefore be understood as referring solely to the sixth indent of Article 1 and the fourth indent of Article 2 of that decision.
106In support of their claims for annulment of the contested decision in so far as it concerns them, the applicants put forward various pleas, which overlap to a large extent, and which it is appropriate to examine in the following order:– the pleas alleging infringement of the rights of the defence and an inadequate statement of reasons relating to the identification of the conduct at issue (the first limb of Nomura’s fifth plea; BofA’s third plea; Portigon’s first plea);– the pleas alleging errors on the part of the Commission for having held the banks concerned liable for the acts of their employees (the first part of UniCredit’s third plea; the second limb of Nomura’s fourth plea; the second limb of Portigon’s fifth plea);– the pleas alleging errors in the characterisation of the conduct at issue as a ‘single and continuous infringement’ and as regards the attributability of that infringement to the banks concerned for a certain period (UniCredit’s second plea, the second part of its third plea, part of its fifth plea and its sixth plea; part of Nomura’s first plea, its second and third pleas, the first and third limbs of its fourth plea, the third argument of the first limb and the second, third and fourth limbs of its fifth plea; BofA’s first plea; Portigon’s fourth plea and part of the first limb of its fifth plea);– the pleas alleging errors in the characterisation of that single and continuous infringement as a ‘restriction by object’, possibly accompanied by a failure to state reasons (UniCredit’s first and third to fifth pleas; Nomura’s first plea and part of its second plea; part of the first limb of Portigon’s fifth plea);– the pleas based on errors relating to the existence of a legitimate interest on the part of the Commission in making a finding of infringement against the banks which were not fined, possibly accompanied by an infringement of the rights of the defence and an inadequate statement of reasons in that regard (Natixis’s first to third pleas; BofA’s second plea; Portigon’s sixth plea);– the pleas alleging errors in the amount of the fines imposed on them, as regards the determination of the basic amount of those fines and the individual adjustments made to that basic amount (UBS’s first to fifth pleas; UniCredit’s seventh to tenth pleas; Nomura’s sixth to tenth pleas);– the plea alleging a failure to observe the principles of unity of the legal order and of consistency (Portigon’s second plea);– the plea alleging an unlawful failure by the Commission to exercise its discretion (Portigon’s third plea);– the plea alleging that the Commission exceeded its powers, reversed the burden of proof and breached the principle of the presumption of innocence as a consequence of the wording of Article 3 of the contested decision (Natixis’s fourth plea).
1.The objections alleging infringements of the rights of the defence and an inadequate statement of reasons relating to the identification or presentation of the conduct at issue
107As a preliminary point, it should be noted that Nomura, in the first branche of its fifth plea, BofA, in its third plea, and Portigon, in the first branche of its first plea, object to the presentation of the conduct relied on against them as set out in the Statement of Objections and in the contested decision, claiming infringements of their rights of defence and of the Commission’s obligation to state reasons.
108While those two objections are raised jointly, they are separate, the first alleging, in essence, deficiencies in the Statement of Objections which deprived those banks of the opportunity to exercise their rights of defence effectively during the administrative procedure which led to the contested decision, and the second alleging a failure to state reasons specific to that contested decision, which deprived the banks concerned of the opportunity to ascertain the reasons for it and the EU judicature of the opportunity to exercise its power of review.
109Those two objections must therefore be examined in turn.
(a)The objections alleging infringements of Nomura’s, BofA’s and Portigon’s rights of defence
(1)BofA’s objection alleging infringement of its rights of defence owing to the substantial difference between the evidence relied on against it in the Statement of Objections and in the contested decision
110In the first limb of its third plea, BofA claims that the Commission ‘altered the intrinsic nature’ of the single and continuous infringement alleged against it between the Statement of Objections and the contested decision.
111In the Statement of Objections, the Commission initially alleged that BofA had participated in the entire single and continuous infringement at issue, relying, principally, on discussions which had taken place in the DBAC chatroom and, to a lesser extent, on those which had taken place in the CODS & CHIPS chatroom. To that end, in recitals 596 and 618 to 622 of the Statement of Objections, the Commission relied on the fact that that bank was aware of, or could have reasonably foreseen, the conduct of the other banks participating in the infringement at issue put into effect in the DBAC chatroom or through instant messaging.
112Then, in the contested decision, it alleged only that BofA had participated in the single and continuous infringement at issue ‘with respect to the CODS & CHIPS chatroom’, after finding, in recital 441 of that decision, that ‘it cannot be concluded with sufficient certainty that [BofA] was aware or ought reasonably to have been aware and was willing to take the risk of the existence and functioning of the DBAC chatroom or other anticompetitive communications’.
113In order to enable BofA to defend itself, the Commission should have indicated in the Statement of Objections that it intended to find that that bank had participated in the infringement at issue solely by virtue of its participation in the discussions that had taken place in the CODS & CHIPS chatroom, that ‘all the various types of conduct that constituted the cartel took place in the CODS & CHIPS chatroom during the period of BofA’s involvement’ or that the discussions which took place in the DBAC and CODS & CHIPS chatrooms pursued the same objective.
114Consequently, the Statement of Objections addressed to BofA did not enable it to understand that the Commission could, in the contested decision, find that it had participated in a single and continuous infringement with respect only to the CODS & CHIPS chatroom, which, moreover, it indicated during the administrative procedure after having been informed on 16 March 2021 that the Commission was no longer claiming that it was aware of the DBAC chatroom.
115That infringement of its rights of defence is confirmed by the fact that numerous discussions which took place in the CODS & CHIPS chatroom and which were referenced only in Annex 1 to the Statement of Objections were added to the body of the contested decision, which in itself also constitutes a second infringement of BofA’s rights of defence. Those additions show that, at the stage of the contested decision, the Commission relied on a materially different presentation and analysis of the evidence in order to conclude that the two chatrooms in question pursued the same objective.
116In that regard, it should be recalled that observance of the rights of the defence requires, in particular, that the statement of objections which the Commission sends to an undertaking on which it envisages imposing a penalty for an infringement of the competition rules contain the essential elements used against it, such as the facts, the characterisation of those facts and the evidence on which the Commission relies, so that the undertaking may submit its arguments effectively in the administrative procedure brought against it (see judgment of 16 June 2022, Sony Corporation and Sony Electronics v Commission, C‑697/19 P, EU:C:2022:478, paragraph 70 and the case-law cited).
117However, the essential elements on which the Commission relies in the statement of objections may be set out summarily and the final decision is not necessarily required to replicate the statement of objections, since that statement is a preparatory document containing assessments of fact and of law which are purely provisional in nature. Thus, it is permissible to supplement the statement of objections in the light of the parties’ response, whose arguments show that they have actually been able to exercise their rights of defence. The Commission may also, in the light of the administrative procedure, revise or supplement its arguments of fact or law in support of its objections. Consequently, until a final decision has been adopted, the Commission may, in view, in particular, of the written or oral observations of the parties, abandon some or even all of the objections initially made against them and thus alter its position in their favour or decide to add new complaints, provided that it affords the undertakings concerned the opportunity of making known their views in that respect (see judgment of 18 June 2013, ICF v Commission, T‑406/08, EU:T:2013:322, paragraph 118 and the case-law cited).
118In the present case, the Commission found, in the first indent of Article 1 of the contested decision, that BofA had participated in the infringement at issue, with the material scope of that participation being narrower than that envisaged in the Statement of Objections.
119The Commission criticised BofA for having participated in that infringement no longer in respect of all the anticompetitive discussions which had taken place in the DBAC and CODS & CHIPS chatrooms and by means of bilateral communications, but only as regards those discussions which had taken place in the CODS & CHIPS chatroom.
120That narrowing of the scope of BofA’s participation in the infringement at issue found in the Statement of Objections compared with that found in the contested decision is favourable to BofA and cannot therefore, in principle, harm its interests (see, to that effect, judgment of 18 June 2013, ICF v Commission, T‑406/08, EU:T:2013:322, paragraph 123 and the case-law cited).
121However, in the light of BofA’s arguments, it is necessary to examine whether that narrowing of the material scope of BofA’s participation in the infringement at issue led the Commission to find, in the contested decision, that it had participated in an infringement so different from that set out in the Statement of Objections that it should be treated as a new objection on which that bank should have been able to submit its written and oral observations, in order to ensure that its rights of defence were observed.
122In the first place, it must be pointed out that, in its reply, BofA acknowledges that, by the contested decision, the Commission did not criticise it for participating in ‘a standalone single and continuous infringement based solely on [the discussions that took place in the] CODS & CHIPS [chatroom]’.
123In the second place, BofA does not dispute that, subject to a marginal reduction in its total duration, the material, geographic and temporal scope of the infringement alleged against all of the banks to which the contested decision was addressed does not differ between the Statement of Objections and that decision.
124Accordingly, only the material scope of BofA’s participation in the infringement at issue was altered between the Statement of Objections and the contested decision, and not the scope of the infringement as such.
125In the third place, it must be noted that, both in the Statement of Objections and in the contested decision, the Commission alleged that BofA had participated in a single and continuous infringement.
126First, as the Commission recalled, in essence, in recital 590 of the Statement of Objections, it is settled case-law that an undertaking which has participated in a single and continuous infringement through its own conduct, which meets the definition of ‘agreements’ or ‘concerted practices’ for the purposes of Article 101(1) TFEU and was intended to help bring about the infringement as a whole, may also be liable in respect of the conduct of other undertakings in the context of the same infringement throughout the period of its participation in the infringement. That is the position where it is shown that the undertaking intended, through its own conduct, to contribute to the common objectives pursued by all the participants and that it was aware of the offending conduct planned or put into effect by other undertakings in pursuit of the same objectives or that it could reasonably have foreseen it and was prepared to take the risk (see judgment of 6 December 2012, Commission v Verhuizingen Coppens, C‑441/11 P, EU:C:2012:778, paragraph 42 and the case-law cited; see also, to that effect, judgment of 16 June 2011, Team Relocations and Others v Commission, T‑204/08 and T‑212/08, EU:T:2011:286, paragraphs 36 and 37).
127Second, as is apparent, first of all, from the Statement of Objections, the Commission indicated that the ‘representative’ discussions referred to in the body of that statement and all of those referenced in Annex 1 thereto had the same general characteristics, set out in recitals 111 to 120 of the Statement of Objections, irrespective of the chatroom in which the banks concerned had participated.
128Next, the table annexed to the Statement of Objections listing all of the discussions in question showed that the Commission considered that BofA’s trader had participated in discussions that could be classed in all four categories of conduct identified in recital 529 of the Statement of Objections, as the Commission indeed noted in recital 614 of that statement.
129Lastly, in recitals 114 and 615 of the Statement of Objections, the Commission expressly stated that the traders of the banks concerned had participated ‘in some or all of these activities’ by taking part in the discussions that took place in the CODS & CHIPS ‘and/or’ DBAC chatrooms.
130It follows from the three preceding paragraphs that the single and continuous infringement set out in the Statement of Objections and alleged against, inter alia, BofA, was essentially linked to the subject matter and content of the discussions between the banks concerned, viewed independently of the chatroom in which the banks concerned participated.
131Thus, contrary to BofA’s contention, the fact that the banks concerned were connected to the two chatrooms was not a decisive factor in the single and continuous infringement detailed by the Commission in the Statement of Objections.
132That finding is confirmed by footnote 129 to the Statement of Objections, from which it is clear that, at that stage of the procedure, the Commission considered that only a limited proportion of the 3583 messages exchanged in the chatrooms in question were anticompetitive.
133Consequently, by finding in the contested decision that BofA had participated in the infringement at issue only ‘with respect to the CODS & CHIPS chatroom’, the Commission merely reduced the scope of the objections addressed to that bank and set out in the Statement of Objections.
134Therefore, in the contested decision, the Commission did not find that BofA had participated in an infringement so different from that set out in the Statement of Objections that it should be treated as a new objection on which BofA should have been able to submit its written and oral observations in order to ensure that its rights of defence were observed.
135For the same reasons, BofA cannot meaningfully claim that, following the Statement of Objections, it was not able to foresee that its participation limited solely to the discussions which took place in the CODS & CHIPS chatroom might lead the Commission to find that it participated in the alleged single and continuous infringement.
136For the same reasons again, BofA’s line of argument set out in paragraph 113 above must also be rejected.
137Those conclusions are not called into question by BofA’s line of argument, referred to in paragraph 115 above, that numerous discussions which took place in the CODS & CHIPS chatroom and which were referenced only in Annex 1 to the Statement of Objections were added to the body of the contested decision, which, according to BofA, demonstrates the ‘material’ difference in the Commission’s analysis between the Statement of Objections and the contested decision.
138As pointed out in paragraph 117 above, the Statement of Objections is a preparatory document, which the Commission may supplement in the final decision, provided that its addressee was able to reasonably deduce from it the conclusions which the Commission intended to draw.
139In addition, Annex 1 to the Statement of Objections formed an integral part of that statement, as the Commission indeed expressly indicated in recital 125, which reads, first, ‘the contacts mentioned in the chronological overview of this Statement of Objections are non-exhaustive but considered representative for the description of the pattern of conduct in the relevant period’ and, second, ‘a full overview of the selected contacts between competitors that support the preliminary conclusions set out in this Statement of Objections can be found in Annex 1’.
140BofA does not indicate to what extent the explanations relating to each of the discussions contained in the body of the contested decision and referenced solely in Annex 1 to the Statement of Objections differ materially from the preliminary analysis deriving from a combined reading of sections 4 and 5 of the Statement of Objections, the table annexed thereto setting out the categories within which each of the timestamped discussions allegedly came and, where necessary, the Letter of Facts.
141In the light of the foregoing, the first limb of BofA’s third plea must be rejected as unfounded.
142Moreover, in so far as BofA relies on an additional infringement of its rights of defence owing to the use in the contested decision of discussions referenced only in Annex 1 to the Statement of Objections, as stated in paragraph 115 above, that objection is also raised in the second limb of its third plea and, on that basis, will therefore be examined in paragraphs 143 to 189 below.
(2)The objections raised by Nomura, BofA and Portigon alleging infringement of their rights of defence due to the presentation of the conduct at issue in the Statement of Objections and in the Letter of Facts
143Nomura, in the first limb of its fifth plea, BofA, in the second limb of its third plea, and Portigon, in its first plea, claim that the Commission infringed their rights of defence due to the presentation of the conduct at issue in the Statement of Objections and in the Letter of Facts.
144In essence, first of all, that infringement is said to stem from the fact that, in the body of the Statement of Objections, the Commission detailed only a small number of discussions which it considered representative and, as to the remainder, referred to Annex 1 to that statement. BofA thus states that a very significant number of discussions which concerned that bank and were referenced only in Annex 1 to the Statement of Objections – namely 29 excerpts from 18 discussions – were explained, for the first time, in the contested decision.
145Next, that infringement of Nomura’s, BofA’s and Portigon’s rights of defence stems from the fact that, in Annex 1 to the Statement of Objections, namely a table headed ‘Selection of relevant contacts’, the Commission merely used the letter ‘Y’ for each discussion to identify the category or categories to which it belonged; those categories were, moreover, insufficiently defined. Furthermore, the difficulty in understanding the forms of conduct referred to in the table in Annex 1 to the Statement of Objections was exacerbated where, within a given day, several periods were referred to. By way of example, Portigon refers to the discussion of 5 November 2009 and to ‘six other contacts’. However, since those ‘six other contacts’ have not been clearly identified by Portigon, they cannot be examined by the Court.
146Lastly, BofA argues that not only should the Commission have indicated the market and the bonds concerned, but should also have provided a detailed explanation of those discussions. Those explanations were made all the more necessary because the language used by the traders was extremely technical and those discussions could be legitimate.
147Those omissions, taken as a whole, denied Nomura, BofA and Portigon the possibility of readily deducing the conclusions which the Commission intended to draw from all the discussions referred to in the Statement of Objections and, therefore, of effectively making their positions known. In that regard, BofA adds that, had they been able to do so, the outcome of the administrative procedure might have been different.
148As a preliminary point, first, it should be noted that, in the first part of its fifth plea, Nomura claims, inter alia, that its rights of defence were infringed due to the way in which the Statement of Objections was presented.
149However, that claim is supported only by paragraphs 112 to 115 of its application and will therefore be examined only in the light of the considerations referred to in those paragraphs.
150Second, in the second part of its third plea, BofA relies not only on an infringement of its rights of defence but also, and on the same grounds, on a failure to comply with the obligation to state reasons in the Statement of Objections. Since those two objections overlap, they will be examined together.
151In that regard, it should be noted that, in recital 112 of the Statement of Objections, the Commission stated that ‘the alleged cartel consisted of various actions and/or inactions that were interrelated and largely overlapped’ and that it ‘could potentially be achieved by means such as attempting to drive the price of an EGB bond down in the run up to an auction of further bonds (with the aim of paying less for the issue), attempting to coordinate on bidding levels and thus reduce the chances of overbidding in an auction, and exchanging information on the timing of the pricing of syndicates for new bond issues’.
152In recitals 113 and 529 of the Statement of Objections, the Commission stated that ‘these interrelated activities [could] be presented in different categories or stages’, namely, first, coordination and attempts to coordinate trading strategies, second, coordination to facilitate the alignment of bidding strategies, third, coordination and attempts to coordinate the level of overbidding, and, fourth, other exchange of forward-looking sensitive information.
153The Commission indicated the types of conduct that each of those categories covered.
154As regards, in particular, Category 4, it stated that that category included ‘the sharing of other sensitive information that actually or potentially allowed the [banks concerned] to directly or indirectly adjust or align their EGB trading activities and, in particular, the sharing of information on (i) individual recommendations given to a debt management office before an upcoming auction on the type, maturity or size of the EGB to be issued; (ii) prices and/or volumes of individual trades of EGB on the secondary market; (iii) strategies for specific counterparties; and (iv) the timing of pricing of syndications’.
155In Section 4.2 of the Statement of Objections, the Commission presented more than 130 discussions in the form either of tables containing relevant and timestamped excerpts of the discussion concerned or, in a more concise manner, quotations, which are often timestamped and are identified by inverted commas and italics.
156Moreover, in recital 81 of the Statement of Objections, the Commission stated that ‘the selection presented in Section 4.2, however, [wa]s not exhaustive, but rather constitute[d] a representative sample of the whole body of evidence on which the Commission relie[d], which [wa]s included in Annex 1’ and that ‘Section 4.2 below [wa]s complemented, therefore, by Annex 1, which identifie[d] all of the selected relevant contacts with which the Commission t[ook] issue for the purpose of th[e] … Statement of Objections’.
157Annex 1 included a table identifying approximately 480 discussions that had taken place between all or some of the banks concerned, including the 130 discussions presented in the body of the Statement of Objections.
158For each of those 480 discussions, Annex 1 indicated its date, the chatroom in which it had taken place, the time of the relevant ‘non-exhaustive extracts’, the banks which participated in it, its classification in one or more of the categories referred to in recital 113 of the Statement of Objections, the location of any reference to it in the body of that statement, the bank which provided the Commission with the transcript of the discussion and its/their index number in the Commission’s file.
159That presentation took the following form in English and German:
160Furthermore, in the Letter of Facts, the Commission rectified certain material errors and, after that letter was communicated, the banks concerned were again able to access the evidence in the Commission’s file and submit their observations in writing.
161It follows from the foregoing that, as regards the more than 130 discussions set out in the body of the Statement of Objections, a combined reading of Section 4.2 of that statement and the table in Annex 1, which identified the participants in each discussion, the excerpts referred to by the Commission and the nature of the conduct at issue by reference to the categories referred to in recital 113 of that statement, did enable the banks concerned properly to identify that evidence and the use that the Commission was likely to make of it in the contested decision.
162As regards the approximately 350 discussions referenced only in Annex 1 to the Statement of Objections, it should be borne in mind, as already stated in paragraphs 116 and 117 above, that the Statement of Objections is a preparatory document intended to enable its addressees properly to identify the conduct complained of by the Commission and the evidence available to it.
163That information may be given summarily, provided that the Commission sets forth clearly all the essential facts upon which the Commission is relying at that stage of the procedure (judgment of 7 June 1983, Musique Diffusion française and Others v Commission, 100/80 to 103/80, EU:C:1983:158, paragraph 14).
164In that regard, it must be noted that, in the present case, it is clear from recital 125 of the Statement of Objections that Annex 1 to that statement constitutes an integral part of it and is to be read in the light of all the recitals set out in the body of the statement.
165It follows that, in the table in Annex 1 to the Statement of Objections, the use of the letter ‘Y’ to indicate the category of conduct into which the Commission intended a discussion should come cannot be regarded as lacking clarity or precision.
166Since the Commission had identified and detailed, in recitals 113 to 120 of the Statement of Objections, the types of anticompetitive conduct covered by each of the four categories of conduct in question by means of a general definition and examples, the banks concerned were able to understand the type of conduct to which the Commission was referring for each discussion, without the Commission having to provide further explanations as to the unlawful nature of the discussion in question.
167In that regard, the fact that, in relation to the same discussion on the same day, the Commission was able to identify several excerpts without, however, indicating, for each of those excerpts, the category or categories of conduct to which it belonged, does not call that conclusion into question.
168In addition to the fact that organising each excerpt into a single category was not necessarily practicable owing to the rapid succession and interweaving of messages that might have different subject matters and therefore come within different categories, the methodology used by the Commission in any event enabled the banks concerned to identify with certainty the excerpts which the Commission considered, at that stage, to be anticompetitive.
169Moreover, as is apparent, implicitly but necessarily, from the Statement of Objections and in particular from recitals 577 and 578 thereof, the categories of conduct used by the Commission had been defined for analytical purposes and not to identify as many infringements of Article 101 TFEU as there were categories referred to in recital 113 of the Statement of Objections.
170Therefore, even if the banks concerned themselves had to identify the category referred to by the Commission to which each excerpt from a particular discussion belonged, they remained fully informed that the Commission was likely to use that discussion to demonstrate the existence of a single and continuous infringement with an anticompetitive object comprising all of the discussions placed in those categories.
171In that regard, in so far as concerns the discussion of 11 June 2007, objected to by BofA, Annex 1 to the Statement of Objections referred to the discussion of that day, dividing it into five excerpts, consisting, as is apparent from that annex read in conjunction with the transcript of that discussion, of approximately 1 minute and 5 messages, less than 4 minutes and 12 messages, approximately 1 minute and 5 messages, approximately 31 minutes and 21 messages and, finally, 9 seconds and 2 messages, respectively, and placing it in Categories 1, 2 and 4. Such a presentation cannot have led BofA to have been unable to identify the messages which, in the Commission’s view, were capable of being anticompetitive. That is the case, for example, with the third excerpt which, in two of its five messages, shows that RBS’s trader wrote ‘where u got mid on bono 37 vs bund?’ and that UBS’s trader replied ‘6.15’.
172It must also be found that BofA’s objection, referred to in paragraph 144 above and based on the fact that 29 excerpts from 18 discussions were explained, for the first time, in the contested decision, and that of Portigon relating to the discussion of 5 November 2009, referred to in paragraph 145 above, cannot succeed.
173First, two of the discussions objected to by BofA were analysed in the body of the Statement of Objections by reference to messages exchanged on the same day and included in other excerpts referred to in the table contained in Annex 1 to the Statement of Objections (discussion of 25 July 2007 detailed in recitals 144 and 145 of the Statement of Objections; discussion of 27 March 2008, detailed in recital 200 of that statement).
174Second, the discussion of 5 November 2009 objected to by Portigon and five of the discussions objected to by BofA (discussion of 5 June 2007, consisting of four excerpts; discussions of 25 and 30 July 2007, consisting of two and three excerpts, respectively; discussion of 11 October 2007, consisting of five excerpts; discussion of 27 March 2008, which took place in the CODS & CHIPS chatroom and consists of two excerpts) gave rise, at the time of the Letter of Facts, to detailed explanations, at least with regard to certain categorisations.
175By way of example, as regards the discussion of 5 November 2009 which took place in the CODS & CHIPS chatroom, referred to in recital 277 of the contested decision and not examined in the body of the Statement of Objections, the ‘Explanations’ column in the Letter of Facts indicates that classifying that discussion in Categories 1 and 3 is justified by the fact that ‘in accordance with [recital] 532 of the [Statement of Objections], it includes an attempt to move the curve ahead of an auction (“why does it seem when we want it flatter all these […]?”’ and by the fact that, ‘in accordance with [recital] 542 of the [Statement of Objections], it also touches upon an attempt to coordinate overbidding levels (“what we thinking of overpaying?”)’.
176Third, eight of the discussions objected to by BofA were challenged only in relation to certain excerpts and not the entirety of them, even though BofA’s trader was connected to the relevant chatroom when the messages referred to in the excerpts which are not contested were exchanged (discussion of 6 March 2007 in the CODS & CHIPS chatroom, only one of the two excerpts from which was contested; discussion of 3 July 2007, only one of the four excerpts from which was contested; discussion of 4 October 2007, only one of the two excerpts from which was contested; discussion of 11 October 2007, only three out of the five excerpts from which were contested; discussion of 8 November 2007, only three out of the five excerpts from which were contested; discussion of 23 November 2007, only one of the two excerpts from which was contested; discussion of 9 January 2008, only one of the six excerpts from which was contested; discussion of 11 January 2008 in the CODS & CHIPS chatroom, only one of the two excerpts from which was contested).
177Fourth, 10 of the excerpts or discussions objected to (excerpt from 3 July 2007, comprising 9 messages; excerpt from 25 July 2007, comprising 26 messages; 3 excerpts from 11 October 2007, comprising a total of 33 messages; discussion of 10 January 2008 in the CODS & CHIPS chatroom comprising 16 messages; excerpt from 11 January 2008 in the CODS & CHIPS chatroom, comprising 7 messages; discussion of 13 February 2008 in the CODS & CHIPS chatroom, comprising 4 messages; discussion of 4 March 2008 in the CODS & CHIPS chatroom, comprising 47 messages; discussion of 6 March 2008 in the CODS & CHIPS chatroom, comprising 23 messages) consisted of a sufficiently small number of messages to make it easy for BofA to identify them in the light of the information contained in Annex 1 to the Statement of Objections and, in particular, the Commission’s classification.
178Furthermore, contrary to BofA’s submission, the Commission cannot reasonably be criticised for having failed to clarify its intentions sufficiently as regards each of the discussions referenced only in Annex 1 to the Statement of Objections, failing to take account of the complexity or technical nature of the language used in those discussions and failing to provide detailed explanations relating in particular to the relevant market, the bonds concerned or the alleged unlawfulness of each discussion concerned.
179The complex or technical language used by the traders of the banks concerned in their discussions could not require the Commission to explain those discussions, in order to make up for BofA’s alleged misunderstanding of them.
180Since the discussions in question took place in a context known to that bank, it could be expected that it would be able to understand them or, at the very least, succeed, by its own means, in obtaining the clarifications necessary for a proper understanding of them.
181For the same reasons, it was not for the Commission to indicate to BofA which EGBs were at issue in each discussion relied on or in relation to which market – primary or secondary – it had taken place.
182Such elements could be deduced from the transcript of the relevant discussion by the banks concerned themselves, which are informed and experienced operators in the EGB sector. Moreover, those elements were not genuinely decisive in assessing whether the discussions at issue were anticompetitive, since what mattered was much more the nature of the information exchanged by the traders than the specific EGB to which that information related.…
188Accordingly, there is no need to examine whether, without the failures alleged by BofA, the outcome of the administrative procedure might have been different.
189Consequently, the first limb of Nomura’s fifth plea, the second limb of BofA’s third plea and Portigon’s first plea must be rejected as unfounded.
190Furthermore, it should be noted that Nomura, in the first limb of its fifth plea, and BofA, in the second limb of its third plea, also criticise the Commission for having referred, in the contested decision, to discussions or excerpts from discussions that were not presented either in the Statement of Objections or in the Letter of Facts, and for having criticised them for a number of discussions or excerpts from discussions referenced only in Annex 1 to that decision.
191Since the first of those objections, alleging that there was no reference to certain discussions in the Statement of Objections or in the Letter of Facts, overlaps with that made by BofA in the third part of its third plea, it will be dealt with together with the latter in paragraphs 218 to 226 below.
192Since the second of those objections, alleging that certain discussions were referenced only in Annex 1 to the contested decision and not in the body of that decision, is linked not to observance of the rights of defence of the banks concerned but to the statement of reasons for the contested decision, it will be examined in paragraphs 262 to 280 below.
(3)Portigon’s objection alleging infringement of its rights of defence due to the fact that the Commission sent a Letter of Facts instead of a supplementary Statement of Objections
193In the second part of its first plea, Portigon submits that the amendments made to the Statement of Objections by the Letter of Facts, sent by the Commission on 12 November 2020, should have taken the form not of such a letter, but of a supplementary Statement of Objections, following which the Commission should have allowed that bank to be heard at an oral hearing, in accordance with the Commission notice on best practices for the conduct of proceedings concerning Articles 101 and 102 TFEU (OJ 2011 C 308, p. 6; ‘the notice on best practices’) and, therefore, to better defend itself.
194Portigon contends that by the Letter of Facts, the Commission added and substantially altered evidence and provided a fundamentally new legal assessment of the conduct at issue, as compared with the Statement of Objections.
195In that regard, first, Portigon argues that certain discussions, such as that of 6 November 2009, were listed and categorised for the first time at the stage of the Letter of Facts. Second, the Commission altered the categorisation of certain discussions at that stage, such as those of 11 October 2007, 22 January, 13 February and 16 April 2009, and 25 January 2010, initially classified in Categories 1, 2 or 3 and then also in Category 4; those of 18 April 2007, 4 January 2008, 29 January 2009 and 24 June 2011, initially classified in Categories 1, 2 or 3 and then only in Category 4; and that of 11 January 2012 placed in Category 4 only in that letter. According to Portigon, the conduct classified in Category 4 constitutes, at most, a restriction by effect and not a restriction by object. Third, in the reply, Portigon argues that the discussion of 11 January 2012, of which it had not previously been aware, makes clear that it ‘did not, for the most part, raise any competition concerns’.
196In that regard, it should be borne in mind that, in order to ensure observance of the rights of defence of the undertakings concerned and, more particularly, their right to be heard, Articles 10 to 12 of Regulation No 773/2004 provide that the Commission may adopt, in its decisions, only objections on which the undertakings concerned have had the opportunity to comment, which they may do in the context of written submissions in response to the statement of objections notified to them by the Commission and at a hearing, if they so request in their written submissions.
197It must also be recalled that, where, during the administrative procedure, the Commission decides to add new objections to those initially raised against the undertakings, it must afford those undertakings the opportunity of making their views known in that respect (see, to that effect, judgment of 27 June 2012, Microsoft v Commission, T‑167/08, EU:T:2012:323, paragraphs 184 and 191 and the case-law cited), by sending them a supplementary Statement of Objections, following which they may submit written observations and request that an oral hearing be held in order to have the opportunity to develop their arguments.
198That obligation to allow the undertakings concerned to make known their views both in writing and at an oral hearing, in accordance with Articles 10 and 12 of Regulation No 773/2004, also applies where the Commission intends to alter materially the evidence for the infringements in respect of which proceedings are brought (see judgment of 7 January 2004, Aalborg Portland and Others v Commission, C‑204/00 P, C‑205/00 P, C‑211/00 P, C‑213/00 P, C‑217/00 P and C‑219/00 P, EU:C:2004:6, paragraph 192 and the case-law cited).
199On the other hand, in line with paragraph 111 of the notice on best practices, which can be relied on against the Commission and the validity of which is not contested by Portigon, where the Commission wishes to rely only on new evidence corroborating the objections already substantiated in the Statement of Objections, it may simply bring this to the attention of the undertakings concerned by a letter of facts, in response to which those undertakings may submit written comments within a fixed time limit.
200In the present case, it is apparent from the contested decision that, on 31 January 2019, the Commission issued a Statement of Objections to Portigon. On 12 November 2020, the Commission sent that bank a Letter of Facts in which, following the replies of the banks concerned to that Statement of Objections referring to various factual errors, it made ‘factual additions or corrections concerning certain communications reported in the Statement of Objections and its Annex in support of the objections raised’.
201In addition to certain factual corrections, the Commission added ‘a number of communications not mentioned in the Statement of Objections, including its Annex, on which [it] may rely for contextual purposes or to support the objections raised’ and which ‘support, for instance, the date of accession of a trader to a persistent chatroom and/or his ability to trade[,] the existence of social contacts between the traders or the existence of exchanges of sensitive information in 2012’.
202The Commission then invited Portigon to submit written comments, in accordance with paragraph 111 of the notice on best practices, which Portigon sent on 8 January 2021.
203It is therefore necessary to determine whether, in order to make the additions and alterations referred to in paragraphs 200 and 201 above, the Commission should have adopted a supplementary Statement of Objections, on which it should have allowed Portigon to submit its written and oral observations, or, on the contrary, was entitled simply to send that bank a letter of facts on which it allowed it to submit only written observations.
204In that regard, first, it must be held that corrections intended to remedy errors in the ID number of the chat rooms concerned or in the date, time or participants in the discussions in question – regrettable as they may be – cannot constitute material alterations to the evidence for the infringement in respect of which proceedings are brought, within the meaning of the case-law referred to in paragraph 198 above.
205That is particularly true of amendments made to discussions that were presented in the body of the Statement of Objections and referenced in the table in Annex 1 to which corrections were made.
206Moreover, Portigon does not dispute that, during the administrative procedure and even before the Letter of Facts was sent, it had access to all the discussions relied on against it, as is apparent from recital 74 of the contested decision.
207Second, the amendments made by the Commission to the classification of one or other of the discussions, such as those of 18 April and 11 October 2007, 4 January 2008, 22 and 29 January, 13 February, 16 April and 6 November 2009, 25 January 2010, 24 June 2011 and 11 January 2012, referred to by Portigon, did not result in the Commission extending the material, temporal or geographic scope of the infringement alleged against, inter alia, that bank in the Statement of Objections.
208As is clear from paragraph 6 of the Letter of Facts, that letter ‘relates solely to the objections set out in the Statement of Objections of 30 January 2019’, namely that bank’s alleged participation from 16 October 2009 to 31 August 2011 in the infringement by object at issue, which began on 4 January 2007 and ended on 1 February 2012.
209Moreover, it appears that all the discussions objected to by Portigon were referred to in Annex 1 to the Statement of Objections, with the exception of that of 11 January 2012, which, however, could not be relied on against Portigon, since it took place after the date on which that bank’s participation in the infringement at issue ended, as found in the Statement of Objections, namely 31 August 2011.
210Furthermore, with the exception of the discussions of 29 January 2009 and 24 June 2011, the discussions objected to by Portigon had been set out in detail in the body of the Statement of Objections (see recitals 141, 150, 151, 162, 163, 302, 310 to 313, 317, 318, 389, 393 and 394), thus enabling that bank, at the stage of the Statement of Objections, to understand the reasons why the Commission considered that those discussions were incriminating, and ruling out the possibility that amendments made at the time of the Letter of Facts could be regarded as material alterations to the evidence.
211As regards the discussion of 29 January 2009, it should be noted that, while it spanned a relatively long period of almost 10 hours, it was classified only in Category 1 and could therefore only have concerned an exchange of trading strategies, which should have enabled Portigon to identify the type of messages complained of.
212As regards the discussion of 24 June 2011, suffice it to note that it is in any event subsequent to the date on which Portigon’s participation in the infringement at issue ended, as ultimately found by the Commission in the contested decision, namely 3 June 2011.…
(4)BofA’s objection alleging breach of its rights of defence due to the failure to mention in the Statement of Objections five discussions referred to in the contested decision
…
239Moreover, that statement is merely a response to BofA’s argument, set out in recital 617 of that decision, that the Commission could not rely on the discussion of 25 February 2008 against that bank on the ground that it was not active on the primary market for Belgian EGBs and remained passive during that discussion.
240Moreover, in so far as BofA relies on recital 620 of the contested decision, it is clear that that recital relates to a separate discussion, namely that of 27 March 2008.
241As regards, in the third place, the discussion of 25 July 2007, BofA also criticises the Commission for having extended, in the Letter of Facts, the duration of the excerpt initially referred to in the Statement of Objections with a view to incorporating a new objection referred to in recital 121 of the contested decision.
242In this instance the Commission did indicate, in the Statement of Objections, that the messages complained of had been exchanged ‘from 06:15:58’, then stated, in the Letter of Facts, that the messages complained of had been exchanged in the time periods ‘06:15:43 – 06:29:16’ and ‘06:58:56 – 07:08:43’. However, that amendment is limited to reducing and clarifying, at the time of the Letter of Facts, the time frame of the discussion complained of in the interests of the addressee of that letter. It cannot therefore be regarded as the addition of a new objection at time of the Letter of Facts.
243In the light of the foregoing, the third part of BofA’s third plea must be rejected as unfounded.
(b)The objections raised by Nomura, BofA and Portigon alleging infringements of the obligation to state reasons due to the presentation of the conduct at issue in the contested decision
244Nomura, in the first limb of its fifth plea, BofA, in the second limb of its third plea, and Portigon, in the first limb of its first plea, complain that the Commission infringed its obligation to state reasons due to the presentation of the conduct at issue in the contested decision.
245First of all, in the table annexed to the contested decision, headed ‘Overview of anticompetitive communications in Decision’, the Commission merely used the letter ‘Y’ in relation to each discussion in order to identify the category or categories within which it came; moreover, those categories were insufficiently defined. Furthermore, the difficulty in understanding the forms of conduct referred to in the table in Annex 1 to the Statement of Objections was exacerbated where, within a given day, several periods were referred to. By way of example, Nomura refers to the discussion of 18 January 2011 and Portigon the discussion of 5 November 2009 and ‘six other contacts’. However, since those ‘six other contacts’ have not been clearly identified by Portigon, they cannot be examined by the Court.
246Next, Nomura and BofA argue that a number of discussions or excerpts from discussions are referenced only in Annex 1 to the contested decision and hence are not accompanied by any explanation as to the EGB or market concerned. In that regard, BofA states that these ‘include 38 excerpts’ from 18 discussions.
247Lastly, Nomura criticises the Commission for having provided a more limited level of detail for the discussions relating to its period of participation in the infringement at issue as compared with that relied on for prior discussions involving the other banks which participated in the infringement at issue.
248In that regard, it should be borne in mind that the obligation to state reasons provided for in the second paragraph of Article 296 TFEU requires that Commission decisions set out the facts and legal considerations of fundamental importance in the context of the decision (see judgment of 1 July 2008, Chronopost and La Poste v UFEX and Others, C‑341/06 P and C‑342/06 P, EU:C:2008:375, paragraph 96 and the case-law cited).
249In the present case, it is clear that, in recitals 376 to 383, 411 to 484, 495 to 544, 723 to 726, 737 to 752 and 760 to 771 of the contested decision, the Commission indicated the reasons why the discussions at issue, referred to in recitals 97 to 356 of that decision and in Annex 1 thereto – which forms an integral part of that decision – had to be characterised as agreements and/or concerted practices constituting a single and continuous infringement with an anticompetitive object, in which the banks concerned participated during the periods indicated in Article 1 of that decision.
250Moreover, that statement of reasons is to be read in the light of the Commission’s replies to the arguments raised by the banks concerned during the administrative procedure and referred to in recitals 384 to 402, 545 to 719 and 753 to 759 of the contested decision.
251Consequently, Nomura, BofA and Portigon cannot reasonably maintain that, in the contested decision, the Commission did not set out the facts and legal considerations of fundamental importance in the context of that decision.
252That conclusion cannot be called into question by the arguments put forward by Nomura, BofA and Portigon that, in essence, the Commission did not sufficiently explain the reasons why, taken individually, each of the approximately 380 discussions referred to in the table annexed to the contested decision, or each of the 783 excerpts from those discussions referred to in the fourth column of that table, corroborate the existence of a single and continuous infringement with an anticompetitive object.
253In that regard, first, it should be noted that, in the contested decision, the Commission did not criticise the banks concerned of as many infringements of Article 101 TFEU as there were discussions or excerpts from discussions listed in the table annexed to that decision, but of a single and continuous infringement with an anticompetitive object consisting of all those discussions (see paragraph 88 above).
254Accordingly, it was not for the Commission to demonstrate that each discussion or each excerpt from a discussion had to be characterised as an agreement, decision or concerted practice within the meaning of Article 101(1) TFEU (see, to that effect, judgment of 16 June 2022, Sony Corporation and Sony Electronics v Commission, C‑697/19 P, EU:C:2022:478, paragraphs 83 and 84).…
284First, in recitals 307 to 350 and 641 to 719 of the contested decision, the Commission set out, in sufficient detail and by means of excerpts, in most cases timestamped, the discussions which took place during the period of Nomura’s participation in the infringement at issue. It also rejected the arguments concerning those discussions raised by that bank during the administrative procedure.
285In addition, further explanatory material was provided by the table annexed to the contested decision, which, as already pointed out in paragraph 249 above, forms an integral part of that decision and the presentation of which is identical for all of the discussions, irrespective of the period during which they took place.
286Second, and contrary to Nomura’s contention, it was not for the Commission to describe ‘in more detail’ the anticompetitive nature of the first instance of each type of discussion that took place from 18 January 2011, the date on which the Commission considered that that bank began to participate in the infringement at issue.
287In the contested decision, the Commission criticised Nomura for its participation in a single and continuous infringement which had begun more than four years prior to that bank taking part in it, and not for its participation in an infringement which began at the time when that bank took part in it.
288In the light of the foregoing, the Commission did not fail to comply with its obligation to state reasons in finding, in Article 1 of the contested decision, that the banks concerned had participated in a single and continuous infringement regarding EGBs covering the entire EEA, which consisted of agreements and/or concerted practices that had the object of restricting and/or distorting competition in the EGB sector.
289Consequently, the first part of Nomura’s fifth plea, the second part of BofA’s third plea and the first part of Portigon’s first plea must be rejected as unfounded.
2.The pleas alleging that the Commission erred in holding UniCredit, Nomura and Portigon liable for the conduct of their traders
290UniCredit, in the first part of its third plea, Nomura, in the second limb of its fourth plea and Portigon, in the second limb of its fifth plea, claim that the Commission erred in attributing to them the conduct of their respective employees.
291In the first place, UniCredit, Nomura and Portigon submit that the Commission could not consider that those banks were automatically aware of all the information acquired by their respective traders during their previous employment in the service of other banks involved at that time in the infringement at issue.
292In the second place, UniCredit, Nomura and Portigon submit that the Commission could not accuse them of the conduct of their respective traders during a period or on a market where they were not authorised to trade.
293In that regard, first, UniCredit and Nomura argue that the Commission found them liable for the conduct of their respective traders carried out before those traders had obtained regulatory authorisation to trade on behalf of their employers. While the UniCredit and Nomura traders had obtained that authorisation on 4 March and 16 September 2011, respectively, the Commission took 9 September and 18 January 2011, respectively, as the dates on which UniCredit’s and Nomura’s participation in the infringement at issue began.
294Second, UniCredit submits that the Commission criticises it for positions taken by its trader related to the primary market, even though that trader had no mandate from UniCredit to trade on that market and was therefore not authorised to trade on its behalf. Similarly, Portigon argues that its trader acted without its authority and outside of Portigon’s area of activity, which was limited to the primary market in German EGBs. Thus, Portigon cannot be held liable for the conduct of its trader which went beyond the scope of that bank’s commercial activities.
295In the third place, UniCredit and Portigon submit that the Commission attributed to them the conduct not of a manager but of a rogue employee, who, moreover, in the case of UniCredit, was on a probation period.
296In the fourth place, Portigon submits that, in the absence of any negligence on its part, the Commission was wrong to attribute to Portigon the conduct of its trader.
297In that regard, first, Portigon submits that it could not have been aware of the conduct of its trader, since it could not easily have accessed or viewed the chatrooms used by him without infringing data protection law or employment law. In addition, it put various measures in place intended to ensure that its employees complied with due diligence requirements, in particular by making compliance with such requirements a term of their contracts of employment, having its trader sign an undertaking to comply with the bank’s compliance guidelines, organising compulsory compliance training and carrying out annual checks.
298Second, Portigon submits that it would be appropriate and proportionate to exclude its liability, given the particular circumstances of the case, namely that (i) its trader did not occupy a managerial position; (ii) he was the only trader from that bank involved in the infringement at issue; (iii) its involvement in that infringement began when that trader joined that bank and ended with his departure; (iv) during that period, most of the positions taken by that trader were neither beneficial nor successful; and (v) the trader’s conduct had begun before he was recruited.
299At the outset, it should be noted that Portigon’s claim, which is made only in the title of its fifth plea, alleging a failure to state reasons in the contested decision as regards attributing liability to that bank for the conduct in which its trader took part during the period in which he was employed by that bank, is manifestly unfounded.
300In recitals 755 to 759 of the contested decision, the Commission explained the reasons why the banks concerned could be held liable for the conduct of the traders they employed.
301In that regard, it stated, in essence, that the employees formed an economic unit with the companies employing them and that the anticompetitive conduct of the former was attributable to the latter, which could, as a matter of principle, be held liable, irrespective of whether the management of those companies had authorised such conduct, was aware of such conduct or whether such conduct ran contrary to instructions provided. The Commission also indicated that the trading activity was characterised by a high level of recording and supervision of the activities of the employees, which enabled the banks concerned to be aware of the essential characteristics of the collusive scheme and the involvement of their employees, and which prevented them from being able to invoke their ignorance.
302By providing such reasons, the Commission complied with its obligation under the second paragraph of Article 296 TFEU.
303As to the substance, as the Commission correctly recalled in recitals 733 to 735 of the contested decision, it is settled case-law, first, that, under EU competition law, an undertaking must be understood as designating an economic unit even if in law that economic unit consists of several persons, natural or legal, and, second, that an employee performs his or her duties for, and under the direction of, the undertaking for which the employee works and is thus regarded as forming part of the economic unit constituted by that undertaking (see, to that effect, judgment of 21 July 2016, VM Remonts and Others, C‑542/14, EU:C:2016:578, paragraphs 22 and 23 and the case-law cited).
304It follows that, for the purposes of a finding of infringement of EU competition law, any anticompetitive conduct on the part of an employee is attributable to the undertaking to which he or she belongs and that undertaking is, as a matter of principle, held liable for that conduct (judgment of 21 July 2016, VM Remonts and Others, C‑542/14, EU:C:2016:578, paragraph 24).
305Therefore, the Commission could legitimately hold Portigon, for the period from 19 October 2009 to 3 June 2011 (given its trader’s period of employment, which began on 16 October 2009 and ended on 31 August 2011), UniCredit, for the period from 9 September to 28 November 2011 (given its trader’s period of employment, which began on 1 September 2011), and Nomura, for the period from 18 January to 28 November 2011 (given its trader’s period of employment, which began on 17 January 2011), liable for the conduct of their respective traders.
306None of the objections raised by UniCredit, Nomura or Portigon can justify excluding such liability in the present case.
307In the first place, UniCredit, Nomura and Portigon cannot properly criticise the Commission for having taken into account in their regard, in particular in recitals 418, 426, 427, 764, 767 and 883 of the contested decision, the knowledge of the conduct at issue acquired by their employees prior to joining those banks, in order to characterise that conduct as a single and continuous infringement with an anticompetitive object.
308First, as has already been pointed out in paragraph 303 above, an employee performing his or her duties for, and under the direction of, the undertaking for which the employee works is regarded as forming part of the economic unit constituted by that undertaking.
309Knowledge which is acquired by an employee prior to joining a new undertaking and which the employee in fact makes available to the new employer may therefore be regarded as knowledge shared by his or her new employer.
310Second, in accordance with settled case-law, the Commission may rely on contacts prior to, or subsequent to, the period of the infringement in order to construct an overall picture and to show the preparatory stages of the cartel as well as to corroborate the interpretation of certain evidence (see, to that effect, judgments of 8 July 2008, Lafarge v Commission, T‑54/03, not published, EU:T:2008:255, paragraph 428; of 2 February 2012, Denki Kagaku Kogyo and Denka Chemicals v Commission, T‑83/08, not published, EU:T:2012:48, paragraph 188; and of 27 June 2012, Coats Holdings v Commission, T‑439/07, EU:T:2012:320, paragraph 60).
311In the present case, and in the light of the other evidence relied on by the Commission – namely, in essence, several hundred transcripts of discussions that took place between the traders of the banks concerned – the Commission was therefore entitled to use the knowledge acquired by their respective traders prior to those traders taking up employment and which those traders in fact made available to those banks.
312More specifically, the Commission was entitled to take account of the fact that, prior to joining Portigon and then UniCredit, the trader from those banks had, as a trader at RBS, and then at Natixis, directly participated in the conduct which took place within the chatrooms at issue and during bilateral discussions, and the fact that that trader had been the founder of one of those two chatrooms.
313The same finding applies to Nomura’s trader who, prior to joining that bank, had participated in the conduct at issue as a trader at RBS.
314In the second place, UniCredit, Nomura and Portigon cannot meaningfully rely on the fact that their respective traders acted on their own initiative and without their authority or on the fact that those traders did not occupy management positions within those banks.
315The attribution to an undertaking of an infringement of Article 101 TFEU does not require there to have been action by, or even knowledge on the part of, the partners or principal managers of the undertaking concerned by that infringement; action by a person who is authorised to act on behalf of the undertaking suffices (see judgments of 16 February 2017, Tudapetrol Mineralölerzeugnisse Nils Hansen v Commission, C‑94/15 P, not published, EU:C:2017:124, paragraph 28 and the case-law cited, and of 11 July 2019, Huhtamäki and Huhtamaki Flexible Packaging Germany v Commission, T‑530/15, not published, EU:T:2019:498, paragraph 72 and the case-law cited). Similarly, formal authorisation of the participation of an undertaking’s employee in a cartel is not a prerequisite of that participation being attributable to the undertaking (see judgment of 29 February 2016, Schenker v Commission, T‑265/12, EU:T:2016:111, paragraph 167 and the case-law cited).
316Although UniCredit, Nomura and Portigon dispute that they authorised their respective traders to take part in the infringement at issue, they do not dispute that, throughout the entire duration of their participation in that infringement, they did in fact employ those traders.
317Therefore, even without regulatory authorisation to trade issued by the Financial Conduct Authority, or authorisation given by UniCredit to its trader to trade not only on the secondary market but also on the primary market, the traders from those banks, by virtue of their employment with their respective banks, were put in a position, by their employers and by the means made available to them by their employers, to obtain and disclose information which they were subsequently able to pass on to other traders.
318Concerning UniCredit more specifically, it is true that the Commission acknowledged, in recital 861 of the contested decision, that it had not been able to demonstrate that UniCredit’s trader had passed on the information he had obtained in relation to the primary market, on which he was not active, to other traders from that same bank who were active on the primary market. However, the Commission also noted in recital 861 that UniCredit’s trader was able to use the information obtained in the chatrooms at issue relating to the primary market in his activity on the secondary market.
319For the same reason, Portigon cannot meaningfully rely on the fact that its activity was limited to German EGBs on the primary market, since the information exchanged in the chatrooms could be useful to it for its activity on the secondary market, which was not limited to trading German EGBs, as the Commission noted in footnote 1452 to the contested decision without being contradicted by Portigon on that point.
320Accordingly, the Commission was entitled to find, in particular in recitals 412, 426, 434, 436 or 438 of the contested decision, that UniCredit, Nomura and Portigon had been aware or were presumed to have been aware of the actions of their respective employees (see, to that effect and by analogy, judgment of 30 September 2009, Hoechst v Commission, T‑161/05, EU:T:2009:366, paragraph 55).
321In the third place, Portigon cannot meaningfully claim that it adopted various measures to prevent its employees from engaging in conduct contrary to ‘due diligence requirements’.
322In that regard, suffice it to recall that an undertaking cannot reasonably rely on any malfunctioning of its internal organisation (see, to that effect, judgment of 30 September 2009, Hoechst v Commission, T‑161/05, EU:T:2009:366, paragraph 55 and the case-law cited) and that measures adopted by an undertaking in order to comply with competition law cannot affect the reality of an infringement committed (see, to that effect, judgment of 13 July 2011, Shell Petroleum and Others v Commission, T‑38/07, EU:T:2011:355, paragraph 96 and the case-law cited).
323In addition, the fact that the various measures which Portigon claims to have adopted did not enable it to prevent the conduct at issue on the part of its employee ultimately demonstrates that those measures were inadequate in ensuring full compliance with EU competition rules within that bank.
324In the fourth place, Portigon is likewise not justified in claiming that it would have been appropriate and proportionate for the Commission to exclude its liability on the ground that the anticompetitive conduct penalised was the act of a single employee who pursued a personal course of conduct initiated before he was recruited and that that course of conduct was only marginally successful.
325The number of employees involved in anticompetitive conduct is irrelevant for the purpose of attributing that conduct to the undertaking employing them, provided that the constituent elements required by Article 101(1) TFEU are present, which will be assessed in the context of the pleas relating to the existence of a single and continuous infringement with an anticompetitive object.
326The same is true of the reasons which might have prompted the employee concerned to engage in the conduct in question or even of the fact that that employee pursued a course of conduct that was initiated previously within another undertaking, provided that the acts attributed to the undertaking employing him or her are for a period during which that same employee was actually working within that undertaking and therefore acted on its behalf.
327Such a conclusion is all the more justified in the present case since Portigon has not indicated that it has lodged a complaint or taken any steps against its former employee, despite the fact that that bank submits that it was involved in the conduct at issue without its knowledge.
328Lastly, the fact that, according to Portigon, its trader’s actions were only marginally successful and that that bank could not easily have accessed the chatrooms at issue is irrelevant for the purpose of assessing whether that trader’s conduct was attributable to that bank.
329Consequently, the Commission was entitled to find that UniCredit, Nomura and Portigon were liable for the conduct of their respective employees and to take into account the knowledge acquired by them through previous employment relationships with other banks that had engaged in the same conduct.
330Moreover, in so far as Portigon claims that, in so doing, the Commission failed to observe the principle of the presumption of innocence, it must be held that that claim is not supported by any arguments and is therefore inadmissible pursuant to Article 76(d) of the Rules of Procedure. In any event, it is unfounded for the same reasons as those set out in paragraphs 303 to 329 above.
331In the light of the foregoing, the first part of UniCredit’s third plea, in so far as it criticises the Commission for finding UniCredit liable for the conduct of its trader, the second limb of Nomura’s fourth plea and the second limb of Portigon’s fifth plea must be rejected as unfounded.
3.The pleas alleging errors as to the classification of the conduct at issue as a single and continuous infringement and as to whether that infringement can be attributed to the banks concerned over a certain period of time
(a)Preliminary observations
332Where there is a dispute as to the existence of an infringement of the competition rules, it is incumbent on the Commission to prove the infringements found by it and to demonstrate to the requisite legal standard, by means of a firm, precise and consistent body of evidence, the existence of the circumstances constituting an infringement (see, to that effect, judgments of 31 March 1993, Ahlström Osakeyhtiö and Others v Commission, C‑89/85, C‑104/85, C‑114/85, C‑116/85, C‑117/85 and C‑125/85 to C‑129/85, EU:C:1993:120, paragraphs 126 and 127; of 17 December 1998, Baustahlgewebe v Commission, C‑185/95 P, EU:C:1998:608, paragraph 58; and of 6 January 2004, BAI and Commission v Bayer, C‑2/01 P and C‑3/01 P, EU:C:2004:2, paragraph 62).
333According to settled case-law, an infringement of Article 101(1) TFEU can result not only from an isolated act, but also from a series of acts or from continuous conduct, even if one or more aspects of that series of acts or continuous conduct could also, in themselves and taken in isolation, constitute an infringement of that provision. Accordingly, if the different actions form part of an ‘overall plan’ because their identical object distorts competition within the internal market, the Commission is entitled to attribute liability for those actions on the basis of participation in the infringement considered as a whole (see judgment of 6 December 2012, Commission v Verhuizingen Coppens, C‑441/11 P, EU:C:2012:778, paragraph 41 and the case-law cited).…
340Part of the first plea in law, the second plea, the first and third limbs of the fourth plea, and the third argument in the first limb together with the second and third limbs of the fifth plea, on all of which Nomura relies, allege errors of law or of assessment relating to the finding of a single and continuous infringement, and to its participation therein. By its third plea in law and the fourth limb of its fifth plea, Nomura also relies on errors relating, in particular, to the duration of its participation in the infringement at issue.
341The first plea in law on which BofA relies alleges an error of law or of assessment relating to the finding of a single and continuous infringement. At the hearing, BofA stated that it disputed only the existence of an overall plan pursuing a single anticompetitive objective.
342The fourth plea in law relied upon by Portigon alleges incorrect findings linked to the existence of an overall plan pursuing a single anticompetitive objective, the continuous nature of the infringement at issue and the date on which its participation in the infringement at issue came to an end. Furthermore, by way of part of the first limb of its fifth plea, Portigon disputes, in particular, the anticompetitive nature of several discussions relied upon against it by the Commission.
343In the light of the considerations set out in paragraphs 332 to 338 above, it is appropriate to examine the arguments put forward by UniCredit, Nomura, BofA and Portigon by distinguishing between, first, those directed at the finding of the existence of a single and continuous infringement and, second, those directed against the finding that liability for that infringement can be attributed to those applicants for a certain period of time.
344However, prior to the examination of the arguments referred to in paragraphs 339 to 342 above, it is necessary to analyse certain findings of fact and certain assessments made by the Commission and linked to the existence of anticompetitive agreements and/or concerted practices, which findings and assessments are set out in recitals 376 to 403 and then recitals 495 to 531 of the contested decision. Those findings and assessments may have an incidence on the merits of the Commission’s findings as to (i) the existence of a single and continuous infringement and (ii) whether liability for that infringement can be attributed to the banks concerned.
(b)The criticisms put forward by UniCredit, Nomura and Portigon relating to the existence of anticompetitive agreements and/or concerted practices
345In the body of the contested decision and in Annex 1 thereto, the Commission found approximately 380 discussions, which took place between early January 2007 and late November 2011, to be anticompetitive.
346In Section 4.2 of the contested decision, the Commission set out the content, together with its interpretation, of over 100 discussions.
347In the context of its third plea in law, UniCredit expressly and specifically criticises the anticompetitive nature of the eight discussions in which the Commission found that UniCredit had participated, namely the discussions of 26 and 28 September, of 4, 12 and 19 October, and of 2, 3 and 28 November 2011.
348As regards Nomura, the Commission relied on the participation of that bank in 34 discussions that took place on 33 different days over a period running from 18 January to 28 November 2011.
349In the context of its second plea, Nomura expressly and specifically disputes the anticompetitive nature of 15 discussions that took place over the course of 14 different days, namely 2 discussions on 18 January along with the discussions of 26 January, 3 and 25 February, 5 and 12 April, 3, 5 and 18 May, 22 June, 7 July, 28 September, 4 October and 2 November 2011.
350As regards Portigon, the Commission relied on the participation of that bank in 63 discussions that took place on 60 different days over a period running from 19 October 2009 to 3 June 2011.
351In the context of its fifth plea in law, Portigon expressly and specifically disputes the anticompetitive nature of discussions or messages that arose over the course of 13 different days, namely certain excerpts from the discussions of 20 and 21 October 2009; the discussion of 6 November 2009; certain excerpts from the discussions of 14 January and 15 February 2010; two messages sent on 22 February 2010; an excerpt from the discussion of 9 September 2010; a sequence of messages from the discussion of 17 November 2010; the two discussions of 18 January 2011, the discussion of 6 April 2011; a sequence of messages from the discussion of 11 April 2011; the discussion of 12 April 2011; and a sequence of messages from the discussion of 18 May 2011.
352Lastly, it must be observed that UniCredit and Nomura have put forward horizontal arguments, namely arguments seeking to call into question the anticompetitive nature of all or a large part of the discussions referred to in the contested decision, such as the alleged excessive reliance on the leniency statements and the alleged public nature or the alleged lack of value of one or more kinds of information exchanged in those discussions.
353In that connection, first, for the purposes of the present proceedings, it must be held that the interpretation of the content of the discussions not specifically disputed by UniCredit, Nomura and Portigon is accepted (see, by analogy, judgment of 7 November 2019, Campine and Campine Recycling v Commission, T‑240/17, not published, EU:T:2019:778, paragraph 105).…
(1)The content of the discussions not specifically disputed by UniCredit, Nomura and Portigon
…
376It is also apparent from the contested decision that the traders discussed the outcome of an auction that had just ended, in particular in the following terms:‘paid 59 in end’ ‘“66” “55 64 ave”’ ‘where u have mid going in?’ ‘i used 50’ ‘50 mid?’ ‘but think it was 2 high’ (discussion of 2 December 2010 in the DBAC chatroom, referred to in recital 304 of the contested decision), ‘“got 100mm” “missed the other by a thick”’ (discussion of 3 March 2011 in the DBAC chatroom, referred to in recital 319 of the contested decision), ‘anyone get anything?’ ‘“I never moved my bids down” “got 100” “26s”’ (discussion of 7 April 2011 in the CODS & CHIPS chatroom, referred to in recital 322 of the contested decision), ‘i bid 92 for 50 for me’ (discussion of 13 April 2011 in the CODS & CHIPS chatroom, referred to in recital 325 of the contested decision), ‘i went 7-9’ ‘“kept 08” “think thats ok” “both of us”’ ‘“yes” “what mid you have 98” “or 99”’ ‘98’ ‘yup same’ ‘we had 98’ ‘04…07 ave’ (discussion of 1 June 2011 in the CODS & CHIPS chatroom, referred to in recital 330 of the contested decision), ‘was that a good auction?’ ‘“non event” “I think” “had 55 mid” “low 57” “ave 65”’ (discussion of 20 July 2011 in the CODS & CHIPS chatroom, referred to in recital 337 of the contested decision).
377In the contested decision, the Commission found that the participants had discussed their trading activity on the secondary market, as is apparent, for example, from the following excerpts:‘“04 81zx” “what mid we got july 28s”’ ‘“98” “GFI [a broker] got 96/02 x81”’ ‘“yup it s my offer” “[sold them] at 97 Six”’ ‘yeah thats about mid’; ‘“bidding 10m of the frtr 10/32 with GFI at 05 x 79” “dealt at 10”’ ‘ta’; ‘bid 40mm july 28s’ ‘ta’ ‘where u got the curve? 55.25 sound ok’ ‘55.5’ (discussion of 14 February 2011 in the DBAC chatroom, referred to in recitals 314 and 315 of the contested decision), ‘“my bid jul39s in [broker at Tullett Prebon]” “im looking for 5m only” “[broker at Tullett Prebon,] hit me at 85 at the same time someone 1ifted 89 on btec [BrokerTec, a broker] in 39s” “doh”’ ‘move curve to 52 fuk knows’ ‘“got 52 also” “feels 1ike it should be steeper” “but i aint gonna short 30yrs ahead of a 2bl tap” “think thats why 42/40s also stuck at 3.5” “my bid jan 30s and jul 39s in [broker at Tullett Prebon]”’ ‘wat u got 5-10 on day?’ ‘-0.5bp’ ‘“steeper” “10-30 1 bp steeper”’ ‘“0.8 i got” “but im high in the cash”’ ‘2-10 2 flatter last question’ ‘1.5’ (discussion of 25 January 2011 in the DBAC chatroom, referred to in recital 308 of the contested decision), ‘Where u got your mid on the jul 28s dbr?’ ‘113.08’ ‘21#’ ‘great thanks, prices crashed, offering 12m’ (discussion of 17 February 2011 in the DBAC chatroom, referred to in recital 316 of the contested decision), ‘what u pay?’ ‘“mid” “75”’ (discussion of 13 April 2011 in the CODS & CHIPS chatroom, referred to in recital 325 of the contested decision), ‘where bunds at 4.15?’ ‘127.25’ ‘“iam with ya” “lets sell oats?”’ ‘I am short already’ ‘same here staying that way want to sell more’ ‘iam long them’ (discussion of 24 June 2011 in the CODS & CHIPS chatroom, referred to in recital 334 of the contested decision).
378In recital 521 of the contested decision, the Commission referred to several discussions in the course of which the participants spoke of their trading activities with specific counterparties on the secondary market in the following terms:‘“not budging” “he’s asking” “stay the same”’ (discussion of 4 January 2007 in a non-persistent chatroom, referred to in recital 101 of the contested decision), ‘“asked us twice as well” “we’re flat”’ (discussion of 8 November 2007 in the CODS & CHIPS chatroom, referred to in recital 131 of the contested decision), ‘if he comes here I am going to tell him to fuk off’ ‘shall we fuk him’ (discussion of 12 December 2007 in the DBAC chatroom, referred to in recital 133 of the contested decision), ‘frog insurer asks me if I can offer 50m sp40s’ (discussion of 13 December 2007 in the DBAC chatroom, referred to in recital 135 of the contested decision), ‘same here i would not tell him what i had’ (discussion of 4 November 2008 in the DBAC chatroom, referred to in recital 212 of the contested decision), ‘“btp 23s” “oil asking” “wat mid”’ (discussion of 10 November 2008 in the DBAC chatroom, referred to in recital 215 of the contested decision), ‘just offered 60m dsl28 to same guy’ (discussion of 13 August 2010 in the DBAC chatroom, referred to in recital 292 of the contested decision).
379Other discussions to which the contested decision refers demonstrate that the traders discussed their trading activities with a specific counterparty on the secondary market, and in particular the discussion of 13 April 2011 in the CODS & CHIPS chatroom, referred to in recital 325 of the contested decision (‘u front run him then?’ ‘“not front run, back run” “trade already don”’ ‘A[?] [identity of the counterparty]’ ‘“yeah.” “you got them”’ ‘yee’ ‘oh i will pull offer’).
380Consequently, in the contested decision, the Commission brought to light the existence of discussions that related, in particular, to (i) the exact timing of the pricing of a syndication (see paragraphs 364 and 365 above); (ii) the recommendations made by primary dealers to debt management offices (see paragraphs 366 and 367 above); (iii) the conduct envisaged on the secondary market in yield curve matters in the run-up to an auction or subsequent thereto (see paragraphs 368 and 369 above); (iv) the levels of bidding and overbidding, mid-prices, yield spreads, bid volumes, trading positions in the run-up to an auction (see paragraphs 370 to 375 above); (v) the results of an auction (see paragraph 376 above); and (vi) the prices, mid-prices, yield spreads, volumes exchanged, trading positions and transactions carried out on the secondary market, including in respect of specific counterparties (see paragraphs 377 to 379 above).
(2)The anticompetitive nature of the discussions not specifically disputed by UniCredit, Nomura and Portigon
381It should be recalled that each economic operator must determine independently the policy which it intends to adopt on the internal market (see judgment of 23 November 2006, Asnef-Equifax and Administración del Estado, C‑238/05, EU:C:2006:734, paragraph 52 and the case-law cited).
382While that requirement of independence does not deprive economic operators of the right to adapt themselves intelligently to the existing or anticipated conduct of their competitors, it does, nonetheless, strictly preclude any direct or indirect contact between such operators by which an undertaking may influence the conduct on the market of its actual or potential competitors or disclose to them its decisions or intentions concerning its own conduct on the market where the object or effect of such contact is to create conditions of competition which do not correspond to the normal conditions of the market in question, regard being had to the nature of the products or services offered, the size and number of the undertakings involved and the volume of that market (see judgment of 19 March 2015, Dole Food and Dole Fresh Fruit Europe v Commission, C‑286/13 P, EU:C:2015:184, paragraph 120 and the case-law cited).
383In order to assess whether, in the present case, the discussions not specifically disputed gave rise to anticompetitive exchanges of information, it is necessary to examine, in the first place, whether the different types of information referred to in paragraph 380 above were commercially sensitive and, in the second place, whether that information had value for the recipients thereof and conferred an advantage on those recipients. In the third place, it will be necessary to respond to the horizontal arguments put forward by UniCredit and Nomura in that regard.
(i)The existence of exchanges of commercially sensitive information
384As is apparent, in particular, from recitals 94 and 377 of the contested decision, the Commission found that the traders of the banks concerned had exchanged various pieces of commercially sensitive information which allowed them to be informed of each other’s conduct and strategies.
385In recital 577 of the contested decision, the Commission pointed out that the exchanges of information that were the subject of that decision were not exchanges of genuinely public information which was generally and equally accessible to all competitors and customers.
386In recital 97 of the contested decision, the Commission emphasised that most of the discussions that it would examine in Section 4.2 of that decision demonstrated that commercially sensitive information had been exchanged between competitors that went beyond simple observations that any market observer could make. According to the Commission, other communications were relevant because they provided context and contained information on the organisation and functioning of the cartel. In recital 379 of the contested decision, the Commission stated that, as was clear, in its view, from Section 4 of that decision, there had been numerous instances where the traders had exchanged views with each other on their current and future bidding or trading strategy for EGBs.
387In order to determine whether the Commission rightly found that the information in question was sensitive, it is necessary, first, to assess whether the exchanges of information found to exist related to non-public information which the traders could not already have known and, second, to examine the subject, accuracy and timeliness of the information exchanged.
–The exchange of non-public information which the traders could not already have known
388First, as regards the exact timing of the pricing of syndication, the Commission essentially explained in recitals 33, 527, 528 and 586 of the contested decision that the lead manager banks assist the debt management office in pricing an EGB by reference to a benchmark EGB on the secondary market. The Commission stated that, to that end, the syndications desk of a lead manager bank can receive information from its own EGB trading desk. However, a lead manager bank’s syndications desk has to maintain strict confidentiality vis-à-vis that bank’s EGB trading desk. The Commission added that the information on the exact timing of the pricing of syndication is not made public until it is announced to the wider market by the lead managers.
389That assessment is supported by a number of the discussions referred to in paragraph 365 above, the content of which demonstrates that certain traders of the banks concerned sought to find out, before any public announcement, the exact timing of the pricing of syndication (‘“priced” “?”’; ‘is it priced?’; ‘has it pxd yet?’).
390That search for information took the form of questions put to a trader of one of the banks concerned when that trader’s bank was one of the lead managers of that syndication. Moreover, the exchanges between the traders demonstrate that the information on the exact timing of the pricing of a syndication was all the more difficult to obtain when none of the traders of the banks concerned worked for a bank that was lead manager, as is illustrated by the discussions of 19, 22 and 23 June 2009, referred to in recitals 253, 255 and 257 of the contested decision (‘trick is to know when its gonna price tho’; ‘said he can try from [another bank] but he’s positive they wont leak anyting to him’; ‘he said he doesnt really use them ever so assumes they wont give him the info though’).
391Second, in recital 525 of the contested decision, the Commission essentially pointed out that the advice given by primary dealers to debt management offices prior to auctions is an individual recommendation that is confidential in principle.
392That assessment is supported by the fact that, as is apparent from the excerpts referred to in paragraph 367 above, the traders of the banks concerned consulted one another before making a recommendation to a debt management office (‘what u gonna ask for from france tomorrow?’; ‘“what we gonna ask for?” “38s?”’).
393Furthermore, recommendations to debt management offices can be made solely by the primary dealers designated by each debt management office. Thus, banks that are not primary dealers with the issuer’s debt management office are not meant to know what recommendations are made by those dealers.
394Third, as regards the level of overbidding of one of the banks concerned in the run-up to or during an ongoing auction, it is admittedly apparent from certain discussions that the traders mentioned the levels attained in previous auctions.
395However, the discussions analysed by the Commission in Section 4.2 of the contested decision, and particularly those referred to in paragraphs 370 and 371 above, demonstrate that (i) the level of overbidding envisaged by a bank in the context of a particular auction was specific to each bank, and (ii) for that reason, the traders of the banks concerned consulted each other repeatedly regarding their respective levels of overbidding (‘wat we over bidding[?]’; ‘wat we have to pay up[?]’; ‘what u over bidding[?]’; ‘what we thinking overbidding[?]’).
396Those discussions therefore establish that the level of overbidding that had been attained in previous auctions served as a starting point only, and that the level of overbidding specific to each bank for a given auction was not known to the other banks.
397Fourth, in recitals 505 and 520 of the contested decision, the Commission explained that the overall market mid-price – that is, the average of mid-price across the market – is available on screen on either D2D or D2C platforms, but that is not the case for individual mid-prices – that is, those specific to a bank – which are not public.
398Moreover, the Commission found, as is apparent in particular from recitals 579 and 581 of the contested decision, that a bank’s yield curves and spreads were also not public.
399In recitals 581 to 583 of the contested decision, the Commission explained, in essence, that each trader made his own personal assessment of the publicly available information in order to determine his mid-price and build his own yield curves and spreads.
400In that regard, those assessments are supported by the content of a very large number of discussions analysed by the Commission and, in particular, those referred to in paragraphs 372, 373 and 377 above.
401In fact, first, in the course of those discussions, the traders of the banks concerned asked the other traders involved for their mid-prices and their yield curves and spreads (‘where we got these mid’; ‘w[h]at mid [u got]’; ‘where u got [spread] ?’ ‘where u got the curve?’).
402Second, the discussions examined by the Commission in Section 4.2 of the contested decision demonstrate that (i) the traders were not sure of the assessment and (ii) they arrived at assessments of the mid-price and the yield spread that could vary and, consequently, that that information was indeed the result of a personal interpretation, by each trader, of the various market data available.
403Those discussions in fact demonstrate that the traders of the banks concerned spoke in particular in the following, particularly explicit, terms on the subject of yield curves and spreads: ‘where u got spr[ea]d?’ ‘“fantasy land” “.4 through i have it” “37s” “think its probally -.55”’; ‘i seem igh’; ‘my mid is high’; ‘im 10tiks higher’; ‘we too high tho’; ‘same thought i was high though’; ‘got 97 here but seem high’ ‘but does look high’; ‘“wat mid 26s now” “20 76/7” “OR THAT 2 HIGH”’; ‘“I got mids well high” “71 29s” “68 39s” “I am going to cheapen”’; ‘got 37s 40 89/90 might be a bit high though’; ‘had 64 but I am high to screen’; ‘“what mid we got 37s” “34/5” “I got 91.77” “is that 2 high”/“july 28” “mid”’ ‘“that’s too high I think” “91.15 37s”/“37 91.33 mid wat about 23s ?”/“fuk the same”’; ‘might be 2 high’; ‘“my mid is high” “i’ll [c]heapen it”’; ‘we a tik higher’; ‘I got em 02 middle but not sure if thats right’.
404Fifth, the bidding volumes specific to each bank were also not known to other banks before an auction. That lack of knowledge is established by numerous discussions presented in the contested decision, and in particular by the discussions, referred to in paragraph 374 above, in the course of which the traders asked each other about the volumes for which they intended to bid (‘how many we gotta buy’; ‘how many u gotta buy?’ ‘how many btps u gotta buy [RBS trader]’; ‘how much belg u gotta try and buy’).
405Sixth, each bank’s long or short trading positions were also not known to the other banks, including on the basis of public information as is apparent from the discussions examined by the Commission in the contested decision and, in particular, from the discussion of 25 January 2010 referred to in recital 280 of that decision (‘how u positioned’; ‘short and short some front ends outright’), and the discussion of 13 August 2010 referred to in recital 291 of that decision (‘how u placed [RBS trader] btps’).
406Seventh, the discussions analysed by the Commission in the contested decision, and in particular those referred to in paragraph 376 above, demonstrate that the traders of the banks concerned put questions to each other after the auctions and that the outcomes of the auctions, in particular the volumes obtained by each bank and the price at which those volumes had been obtained, were not public (‘anyone get anything?’; ‘was that a good auction?’).
407Eighth, as is apparent from the discussions examined by the Commission in the contested decision, and in particular from the discussions referred to in paragraph 377 and 397 to 405 above, the mid-prices, yield curves, yield spreads and trading positions specific to each bank were not public on the secondary market. The same applies to the volumes traded on the secondary market by each bank, the traders’ individual strategies vis-à-vis counterparties and, more generally, transactions carried out by each of the traders with a counterparty on the secondary market.
408It follows from the foregoing that the Commission rightly found that the information set out in paragraphs 388 to 407 above was neither public, known nor accessible to the other banks, be it in the context of their activities on the primary market or in the context of their activities on the secondary market.
409That conclusion is supported by the fact that none of the information set out in paragraphs 388 to 407 above and, in particular, each bank’s mid-prices, yield curves and spreads, and trading positions, as well as the volumes traded by each bank on the secondary market and the identity of a counterparty on that market, could be inferred from the information available on the secondary market.
410The Commission in fact emphasised that, on the secondary market for EGBs, trading took place over the counter, either electronically or via brokers.
411Furthermore, the Commission admittedly acknowledged, in particular in recitals 571 and 579 of the contested decision, that some information was publicly accessible to the traders, specifically on D2D and D2C platforms. Moreover, in recital 578 of that decision, the Commission conceded that brokers could sometimes disclose to one trader the individual bid-ask prices or volumes they had received from another trader.
412However, in recitals 518 and 571 of the contested decision, the Commission explained that the information on transactions executed on D2D and D2C platforms was anonymised and/or aggregated and, consequently, a trader was not usually able to identify who was behind a transaction executed on the secondary market.
413The anonymity of the source of an offer or a bid on the secondary market and the availability of aggregated data only are established in the present case by a great many discussions examined by the Commission in the course of which the traders of the banks concerned indicated to the other traders involved the transactions that they had carried out on the secondary market and, in particular, stated that it was their offer or their bid that was displayed on their respective screens.
414The assessment set out in paragraph 413 above is illustrated, for example, by the discussion of 18 April 2007, referred to in recital 112 of the contested decision (‘traded OAT 19/21 at 5.85’ ‘ta’); the discussion of 30 July 2007, referred to in recital 122 of the contested decision (‘its my px [price] on the aug17s for info’; ‘price on tv’); the discussion of 24 September 2007, referred to in recital 124 of the contested decision (‘that is me on olo31s’); the discussion of 2 January 2008, referred to in recital 136 of the contested decision (‘my offer 38s and 55s on screen’ ‘“ta” “wat mid 37s…91/92”’); the discussion of 2 September 2008, referred to in recital 191 of the contested decision (‘my bid tv in 38s’); the discussion of 11 September 2008, referred to in recital 196 of the contested decision (‘yup [I] hit a low bid’ ‘“saw that low hit” “wondered what muppet did that”’ ‘“it was me” “mupprtbell”’); the discussion of 8 January 2009, referred to in recital 224 of the contested decision (‘my b[i]d in 23s oat again so don’t hit it’ ‘“oops its me on the offer” “pulled my offer back”’); the discussion of 20 January 2009, referred to in recital 231 of the contested decision (‘ol31 that you [UBS trader]?’ ‘“yep” “sorry” “forgot 2 say”’ ‘i joined u on the olo31s’); the discussion of 12 September 2009, referred to in recital 200 of the contested decision (‘“being asked offer in 50m” “i had 51.5”’ ‘wat did u show’ ‘“so i offered 50.9” “then i said off” “offered 50.6” “u [see it] [UBS trader]?”’) and the discussion of 2 December 2010, referred to in recital 304 of the contested decision (‘lifted 25 spain on tv’).
–The subject, accuracy and timeliness of the information exchanged
415In the first place, it is apparent from the discussions set out in the contested decision and from the excerpts reproduced in paragraphs 364 to 379 above that, in the course of those discussions, the traders exchanged information that was specific to each of them, except for information relating to the exact timing of the pricing of a syndication.
416Information relating to the exact timing of the pricing of a syndication was, for its part, known solely to the lead managers, and they were required to maintain strict confidentiality vis-à-vis their respective internal EGB trading desks (see paragraph 388 above).
417In the second place, as regards the subject matter of the information exchanged, first, it is apparent from the discussions set out in the contested decision and from the excerpts reproduced in paragraphs 370 to 373, then in paragraphs 376 and 377 above, that the information exchanged by the traders directly concerned their prices or factors used by the traders in setting their prices on the primary and/or secondary markets.
418It is in fact apparent from a great many discussions analysed by the Commission, and in particular from the excerpts from discussions reproduced in paragraph 368 above, that the traders expressed the desire to ‘steepen’ the yield curve, that is to say, to lower the price of an EGB on the secondary market by opening short positions through the sale of EGBs on the secondary market. On other occasions, the traders of the banks concerned expressed the desire, as is apparent from paragraph 369 above, to ‘flatten’ the yield curve, that is to say, to attempt to increase the prevailing market price on the secondary market.
419Furthermore, it is apparent from the discussions set out in the contested decision, and from the excerpts reproduced in paragraphs 370 and 371 above, that the traders of the banks concerned exchanged information concerning their levels of bidding and overbidding, that is to say, the price that they intended to offer on the primary market in the context of an auction. As is apparent from the discussions analysed by the Commission and from recital 28 of the contested decision, that level of bidding or overbidding has an impact on the chances of success of the bids and, in the event of a successful bid, on the price paid for the volume of bonds obtained.
420Moreover, as the Commission essentially noted in recitals 506 and 507 of the contested decision, as is apparent in the discussions examined by the Commission, each bank’s mid-prices and yield spreads were information that was used by the traders in determining their bid levels or in setting their price on the secondary market, particularly vis-à-vis a specific counterparty.
421In so far as the mid-price is the average between the prevailing bid-ask prices on the secondary market, a trader’s mid-price gives a precise indication of the prices at which that trader is prepared to buy or sell the relevant EGB on the secondary market.
422The important role played by the mid-price in the context of an auction is established by numerous discussions in the course of which the traders discussed the level of overbidding by taking the mid-price as a starting point (‘from mid’). In other words, the traders expressed their levels of overbidding by announcing the hundredths or number of cents that they would be applying to that mid-price.
423As regards each bank’s yield spreads, these are used by a bank in order to set the price of a bond on the basis of a comparison between, on the one hand, that bond and, on the other hand, a bond issued by the same issuer but with a different tenor or by a different issuer but with a similar tenor. The yield spreads therefore give a clear indication of the way in which that bank will set the price of a bond compared to the benchmark EGB. Moreover, it is apparent from the discussions examined by the Commission, and in particular from the discussion that took place on 3 August 2009, referred to in recital 261 of the contested decision, that the traders in question used the yield spreads to determine their level of overbidding in an ongoing auction (‘just easier w spread so dont have to watch futures references’).
424Second, it is apparent from the discussions set out in the contested decision and from the excerpts reproduced in paragraphs 374 and 376 to 378 above that the information exchanged by the traders concerned, first, the volumes to be issued that they would recommend to the debt management offices; next, the volumes that they intended to buy on the primary market in an upcoming auction; further, the volumes that they had actually obtained in that auction; and, finally, the volumes that they had traded or intended to trade on the secondary market, in particular vis-à-vis a specific counterparty.
425Third, it is apparent from the discussions set out in the contested decision and from the excerpts reproduced in paragraph 375 above that the traders also exchanged information on their respective trading positions. In that connection, the Commission recalled in recital 43 of the contested decision that, when a trader is or goes long on a bond, he holds or buys bonds before reselling them, and expects the price of the bonds to go up. When a trader is or goes short, he sells bonds that he does not have and expects the price of those bonds to go down in such a way as to be able to buy them at a price that is lower than that at which he sold them.
426Thus, information on trading positions itself provides information on a trader’s strategy and his current or future conduct on the primary or secondary markets. That assessment is confirmed, for instance, by the discussion of 4 November 2010, referred to in recital 302 of the contested decision, in the course of which one of the traders provided information from which it was apparent that he was long and needed to sell (‘stupid fuking trade took me long 600 lots from 76 just got risk in now’).
427In the third place, the information exchanged related to designated EGBs and was particularly detailed and precise. The bid and trade volumes were expressed in millions or, as regards bid volumes, as a percentage of the volume of bonds issued (see paragraphs 374 and 377 above). Moreover, the long or short positions related to volumes expressed in thousands or millions (see paragraph 375 above). As regards prices, overbidding was expressed in cents or base points and could be split depending on the volumes in the cases of multiple bids (see paragraphs 370 and 371 above). The mid-prices, together with the yield curves and spreads, were also quantified (see paragraphs 372, 373 and 377 above).
428In the fourth place, information was exchanged in real time, and some of that information was updated in the course of the same discussion as is demonstrated, for instance, by a number of discussions such as those of 3 February and 1 June 2011, in the course of which the traders updated the information concerning their respective prices (‘56 at the moment’ ‘“ok I am going +8 +10 on 23s” “50mm a piece” “at the moment”’). Moreover, that information concerned transactions that were soon to take place, were under way or had just taken place.
429Thus the discussions, which were examined by the Commission in the contested decision and have not been specifically disputed by the banks concerned, demonstrate that the traders involved disclosed to other traders, spontaneously or on request, precise and detailed information on their current or future strategy in auctions or syndications on the primary market or trading activities on the secondary market in the EGB sector.
430Consequently, it is clear from paragraphs 388 to 429 above that the Commission has demonstrated to the requisite legal standard that the traders of the banks concerned exchanged information with each other that was not initially known to the other traders involved and that related to their individual commercial strategy.
431The Commission therefore validly found, specifically in recitals 497 to 531 of the contested decision, that the information exchanged between the traders of the banks concerned, and referred to in paragraph 380 above, was commercially sensitive.
(ii)The value of the commercially sensitive information exchanged, and the advantage conferred on the recipients of that information
432In the contested decision, in particular in recital 389, the Commission found that, often, the discussions demonstrated that the participants took an interest in those discussions and that the information exchanged was valuable.
433Moreover, as is apparent, inter alia, from recitals 95 and 378 of the contested decision, the Commission found, in essence, that the exchanges of information conferred an advantage on the participants, on the ground that those exchanges (i) increased market transparency amongst the traders and reduced the uncertainties regarding the issuing and/or trading of EGBs, and (ii) enabled the traders to identify and pursue opportunities for coordination, to align or coordinate their conduct and to help each other when EGBs were issued, placed on the market and traded.
434In the first place, as regards the value of information concerning the exact timing of the pricing of a syndication and the advantage gained by such information, the Commission pointed out, in recital 528 of the contested decision, that knowing that information in advance, even if very shortly, enabled traders to anticipate when the benchmark bond would be at its cheapest, before the rest of the market could cover their positions. The Commission added that that information allowed the traders to raise or lower the price of the benchmark bond in the secondary market and thereby affect the success or failure of the relevant syndicated bond issue, or alternatively be aware of the timing of possible significant increased volatility and price movements in the market and adjust their trading accordingly.
435In that connection, the advantage gained by the knowledge, ahead of time, of the exact timing of the pricing of a syndication is substantiated by the interest displayed by the traders of the banks concerned in that information. It is in fact apparent from the discussions set out in paragraphs 365, 389 and 390 above that the traders sought, sometimes insistently, to obtain that non-public information from another trader whose bank was a lead manager, or else by other means.
436Moreover, the discussions of 23 June and 7 July 2009, referred to in recitals 257 and 259 of the contested decision, respectively, demonstrate that the traders discussed their strategy before and after the pricing of a syndication (‘we a little flatter? im offering some 38s’; ‘iam gonna lean on 23s’).
437Furthermore, in recital 194 of the contested decision, the Commission explained that, in the discussion of 9 September 2008, the traders had discussed the cheapening of the EGB in the run up to the syndication to increase yield and try to boost the syndication by the lead managers.
438Finally, the discussion of 19 June 2009, referred to in recital 254 of the contested decision, illustrates the disappointment on the part of one of the participant that his bank was not a lead manager and the intention of taking retaliatory measures by possibly pushing up the price of the benchmark EGB on the secondary market (‘“think this deal is going to be a blowout though” “i am going to cover early next week”’; ‘you’d think wait for the pxg call and lift the fuk out of the duration [manager]’).
439In the second place, as regards the value of information concerning recommendations made to a debt management office and the advantage gained by such information, the discussions set out by the Commission demonstrate that the traders exchanged views as to the type, maturity and volume of the EGBs that they intended to ask certain debt management offices to issue, and that those exchanges allowed the traders to influence each other or even align their recommendations.
440Thus, in the course of a discussion of 27 August 2008, referred to in recital 190 of the contested decision, one of the participants replied that he was indeed going to ask for bonds maturing in 2038 (‘“what we gonna ask for?” “38s?”’ ‘38s’). Similarly, in the course of a discussion of 9 June 2009, referred to in recital 250 of the contested decision, one of the participants expressed agreement with the analysis given by another participant who stated that he would not be asking for bonds maturing in 2037. As to the third participant in that discussion, he also stated that he would not ask for those bonds (‘“im not gonna ask for sp37s” “47” “dont think [market] needs those”’ ‘“nope” “agree”’ ‘im not asking for 37s’).
441Consequently, as the Commission essentially stated in recital 525 of the contested decision, those exchanges conferred an advantage on the traders since it allowed them better to influence the decision of the debt management office to issue EGBs the volume and maturity of which best suited them.
442In the third place, the discussions examined by the Commission in Section 4.2 of the contested decision, which have not been specifically disputed by UniCredit, Nomura or Portigon, demonstrate that a considerable number of those discussions gave rise, in the run-up to or in the course of an auction, to exchanges of information concerning the strategy intended to steepen or flatten the yield curve on the secondary market, mid-prices, yield spreads, levels of bidding or overbidding, the volumes sought and the long or short trading positions. As regards that information, the Commission found, in essence, that it was valuable, that it allowed the participants to reduce uncertainties and identify opportunities, and adjust, align or coordinate strategies; and, lastly, to maximise the participants’ chances of securing the EGBs they wanted at the cheapest price.
443In that connection, first, a considerable number of the discussions analysed by the Commission show that the traders expressed the desire to steepen the yield curve and that that conduct was aimed at pushing down the average price of an EGB on the secondary market.
444As is apparent from several of the discussions examined by the Commission, in particular those referred to in paragraph 368 above, the ultimate objective pursued by that type of conduct was, for the primary dealers, potentially to acquire an EGB on the primary market at a cheaper price in an auction (‘that is good cheapening preauction! ! ! !’; ‘need to push it cheaper cos it will prob trade like a dog after’). As the Commission stated in recital 16 of the contested decision, the price at which primary dealers decide to bid at auctions depends on the price at which similar bonds are priced on the secondary market (benchmark price).
445The discussions analysed by the Commission in the contested decision, and in particular those referred to in paragraph 369 above, also demonstrate that, on other occasions, the participants expressed the desire to flatten the curve and that the aim of the conduct was, in some cases, to push up the prevailing market price on the secondary market following an auction on the primary market. The ultimate objective pursued by that conduct was, in particular, to make a potential profit therefrom on that secondary market, as is apparent from some of the discussions examined in the contested decision.
446That information was valuable to the other participants in the chatrooms since (i) it concerned the conduct that one or more participants in those chatrooms intended to adopt on the secondary market in the run-up to an auction and (ii) it could facilitate coordination of the strategies of the various participants in order to push down the purchase prices of an EGB on the primary market or push up the sale prices on the secondary market.
447A number of discussions demonstrate, furthermore, that the exchange of such information did indeed result in a coordination of strategies. In fact, as is apparent from the discussions referred to in paragraphs 368 and 369 above, the participants repeatedly responded positively to the suggestion made by one amongst them that they steepen the curve. On other occasions, the participants responded positively to the suggestion made by one amongst them that they flatten the curve.
448Second, the information directly concerning the levels of bidding or overbidding, and therefore the prices envisaged by each of the participants in the run-up to an auction, was evidently valuable to the other participants who were envisaging making a bid in the same auction. The value of that information is illustrated by the numerous exchanges in the course of which the participants discussed their respective levels of bidding or overbidding and, on some occasions, sought to align those levels (‘lets just put the same bids in’; ‘i just moved and now have same as you’; ‘if we bid similar im sure we will get paper’).
449Third, the discussions set out by the Commission in Section 4.2 of the contested decision, and in particular those referred to in paragraph 403 above, demonstrate that a trader’s mid-prices were the result of an personal interpretation of publicly available information, but also that the traders were not sure whether their interpretation was correct. The traders in fact repeatedly asked each other about their respective mid-prices and disclosed those prices to each other in order to ascertain whether their assessment was accurate. The exchanges of information concerning mid-prices therefore reduced the participants’ uncertainty as to whether their assessment was indeed correct.
450Furthermore, it is apparent from some of the discussions analysed by the Commission that, when the participants found there to be differences in their mid-prices – and in particular when these were high – the participants sought to adjust those prices in the light of the information that they had received. That conduct is illustrated, for instance, by the discussion of 4 October 2007, referred to in recital 127 of the contested decision (‘“my mid is high” “i’ll [c]heapen it”’).
451On other occasions, the traders expressed the intention of aligning those mid-prices in the context of an upcoming or ongoing auction, as is illustrated by the discussion of 13 November 2008, referred to in recital 218 of the contested decision (‘lets se[t] our mids’), and the discussion of 6 October 2009, referred to in recital 269 of the contested decision (‘lets get our mid right in this shite’; ‘lets agree mids in 5 min’).
452Thus, the discussions examined in the contested decision demonstrate that the information on the mid-prices specific to each bank was valuable for the other banks in the run-up to an auction, since that information reduced uncertainties and allowed the bids submitted to be adjusted or aligned, and therefore increased the chances of obtaining the desired allocations of EGBs.
453Fourth, the assessment carried out by the Commission in recital 507 of the contested decision – according to which, in essence, yield spreads were of help in estimating the price of bonds before these were issued – is demonstrated by the numerous discussions, and in particular those referred to in paragraph 373 above, in the course of which the traders exchanged their yield spreads in order to compare, adjust or align their strategies, and in particular their prices, before an auction or in the context of transactions executed on the secondary market.
454In that connection, the advantage gained by the exchange of information on yield spreads is highlighted in particular by a discussion that took place on 3 August 2009 and is referred to in recital 261 of the contested decision. In that discussion, two of the participants in the DBAC chatroom informed the Natixis trader, who had been absent for several weeks, that they were currently talking more about yields than prices because it was easier (‘just easier w spread so dont have to watch futures references’). The Natixis trader replied that he understood, and the UBS trader added that ‘it helped to make loads of dosh’.
455Fifth, the information on the volumes that each participant intended to acquire and their respective long or short trading position, was also valuable for the other participants, as is illustrated by the numerous discussions relating to that type of information. That is the case, for instance, of the discussions referred to in paragraphs 374 and 375 above. The participants’ long or short positions could give an indication to the other participants as to their bidding strategies. Similarly, the levels of bidding and overbidding could vary depending on the volumes sought, and be split into tranches. Moreover, that information on the volumes sought enabled the participants to assess their chances of obtaining what they wanted, or to align their conduct, in the light of the size of the auction as a whole.
456The various types of information referred to in paragraphs 443 to 455 above were all the more valuable and liable to reduce uncertainties or facilitate coordination since such information was exchanged on the same day, simultaneously or at short intervals, in the run-up to or during an auction.
457In that connection, it is apparent, for example, from the discussion of 14 February 2011, described in recital 314 of the contested decision, that the exchanges between the participants concerned the levels of overbidding in an upcoming auction, then the bids actually made on and, lastly, activities relating to the secondary market.
458In the same way, it is apparent from the discussions of 3 March and 20 July 2011, described in recitals 319 and 337 of the contested decision, respectively, that the exchanges between the participants could concern the levels of overbidding in an upcoming auction as well as the mid-prices in an ongoing auction and the outcome of the auction, particularly in terms of the volumes obtained and prices for a specific maturity.
459Finally, it is apparent from the discussion of 1 June 2011, examined in recital 330 of the contested decision, that, in the run-up to an auction, the participants exchanged information, between themselves, on the volumes in respect of which they intended to bid, their levels of overbidding, their mid-prices and the prices at which they had actually bid.
460It follows from paragraphs 448 to 459 above that the information concerning each participant’s levels of bidding and overbidding, mid-prices and yield spreads, bid volumes and long or short positions, were of value for the other participants, and conferred an advantage on them in an upcoming or ongoing auction on the primary market.
461In that regard, first, the exchanges relating to that information raised uncertainties as to the strategies that the participants intended to adopt in an auction, in particular in terms of prices and bid volumes. Second, those exchanges created opportunities for coordination.
462Furthermore, the traders attempted to coordinate, or did indeed coordinate, their levels of bidding or overbidding as is apparent, for instance, from the discussion of 14 January 2009, referred to in recital 228 of the contested decision (‘“i think also that if we bid the same prices then they cant ignore it” “tho for the purposes of the tapes that’s not really allowed”’ ‘yeah maybe we get on a phone’ ‘yeah’ ‘“ok done” “lets do that”’); the discussion of 26 January 2009, referred to in recital 234 of the contested decision (‘lets just put the same bids in’; ‘i just moved and now have same as you’ ‘ok cool’); the discussion of 16 April 2009, referred to in recital 241 of the contested decision (‘ill bid about 100 i think u can lead me where ill do same as u’ ‘“def more likely to get em” “this level i think i’ll b bidding like 103.05” “.unless it cheapens a little before”’ ‘ok ill bid same as u’); or even the discussion of 6 October 2009, referred to in recital 269 of the contested decision (‘lets agree mids in 5 mins’ ‘ok’; ‘same now’).
463The traders’ reactions following the exchanges of information and upon their learning of the results of auctions also show that the information exchanged was valuable to the participants and that they benefitted from their exchanges and, where applicable, from coordinating their conduct.
464That is the case, for instance, of the views exchanged during a discussion of 4 January 2007, referred to in recital 101 of the contested decision (‘we did it’ ‘well done’); of 6 March 2007, referred to in recital 110 of the contested decision (‘well done everyone’ ‘all of below the average well done everybody’); of 30 July 2007, referred to in recital 122 of the contested decision (‘belg i dont see why we should p[ay] a premium’; ‘that is good cheapening preauction! ! ! !’ ‘this is the way auctions should go” “!”’) of 7 February 2008, referred to in recital 151 of the contested decision (‘worked well’); of 3 July 2008, referred to in recital 185 of the contested decision (‘well done guys’ ‘well done everybody’); of 3 October 2008, referred to in recital 205 of the contested decision (‘well done’ ‘“how the fuk we keep doing this” “wat a week”’); of 16 October 2008, referred to in recital 211 of the contested decision (‘“good work” “everyone”’); of 5 November 2009, referred to in recital 275 of the contested decision (‘we all played it well we are all great and deserve 3 mil each’); of 3 February 2010, referred to in recital 282 of the contested decision (‘well done everybody’); of 7 October 2010, referred to in recital 297 of the contested decision (‘well done everyone’); and of 12 October 2011, referred to in recital 345 of the contested decision (‘well done’).
465In the fourth place, as regards the participants’ activities on the secondary market, it is apparent from the discussions set out in the contested decision that the information on mid-prices, yield curves and spreads, trading positions, the volumes traded or the identity of a counterparty and, more generally, the transactions executed on that secondary market, were valuable to the other participants.
466As the Commission essentially explained in recitals 519 to 523 of the contested decision, and as is apparent from the discussions analysed in that decision, the information on pricing or the essential price components, trading positions, volumes and counterparties could be taken into consideration by the other participants in order to check whether their own pricing was competitive, determine the appropriate time to sell bonds, assess whether they were in competition or to know what the current market trends were and, more generally, to adjust their trading strategies.
467Those traders also coordinated their trading activities on the secondary market. In particular, the traders agreed, first, to withdraw an offer on one broker’s platform where such an offer would have harmed one or other of those participants and, second, not to lift an offer or not to hit a bid made by another participant. That type of conduct is apparent, in particular, from the discussion of 8 January 2008, referred to in recital 224 of the contested decision (‘my b[i]d in 23s oat again so don’t hit it’ ‘“oops its me on the offer” “pulled my offer back”’); from the discussion of 3 August 2009, referred to in recital 261 of the contested decision (‘“does that fuka nyone?” “shud I pull ?” “ok” “glad to help”’); from the discussion of 16 October 2009, referred to in recital 271 of the contested decision (‘“keep to urself” “dont touch yet” “but just bid 300m oat21s.”’ ‘[o]k’); and from the discussion of 2 November 2010, referred to in recital 301 of the contested decision (‘so[rr]y I will pull my offer’).
468The participants also coordinated their conduct in order to refuse to make an offer to a specific client as is illustrated, for instance, by the discussion of 12 December 2007, referred to in recital 133 of the contested decision (‘he can f off’ ‘if he comes here I am going to tell him to fuk off’) or by the discussion of 16 October 2008, referred to in recital 212 of the contested decision (‘same here i would not tell him what i had’ ‘yee nothing for him’).
469In the fifth place, as the Commission stated, in essence, in recitals 45 to 50 then in recital 479 of the contested decision, the consequence of the interrelationship between the primary and secondary markets was that information exchanged in the context of activities on the one market could confer an advantage in the context of activities on the other market.
470In that regard, first, the information exchanged in respect of strategies or activities on the secondary market conferred an advantage on the recipients of that information in the context of their activities on the primary market in connection with an auction or a syndication.
471In fact, on the one hand, the exchanges of information on the conduct that one or more of the participants intended to adopt on the secondary market in order to steepen or flatten the curve, and therefore push the prices of a bond on the secondary market up or down, were aimed at influencing the price at which EGBs would be acquired in an auction on the primary market or issued in a syndication on that market (see paragraphs 444 and 446 above).
472On the other hand, the information concerning mid-prices, yield spreads and trading positions was information specific to the banks concerned on the secondary market (see paragraphs 420 to 423 and 425 above).
473However, as is clear from paragraphs 449 to 455 above, information concerning mid-prices, yield spreads and trading positions was exchanged by the traders in the context of their activities on the primary market and, more specifically, in the context of upcoming or ongoing auctions in order to reduce uncertainties or even align or coordinate their conduct on that market.
474Exchanges of information concerning mid-prices, yield spreads and trading positions on the secondary market, which information was specific to the banks concerned, therefore conferred an advantage on the recipients of that information in the context of their activities on the primary market.
475It should be added that, as is apparent from the discussions set out in the contested decision by the Commission and from paragraphs 372, 373, 375, 420, 421 and 425 above, the traders exchanged the same types of information – namely information concerning mid-prices, yield spreads and trading positions – in the context of an upcoming or ongoing auction on the primary market and in that of their trading activities on the secondary market. Those types of information were not only used by the traders to reduce uncertainties or even to align or coordinate their conduct on the primary market, but they also provided those traders, in the context of their activities on that primary market, with useful indications as to their prices on the secondary market and their assessment of the evolution of prices on that secondary market. The same may be said of the information on the volumes traded on the secondary market which, in the context of the issuance of EGBs on the primary market, could furnish indications as to demand on the secondary market.
476Second, and conversely, the information exchanged by the participants in the context of their activities on the primary market was also relevant and conferred an advantage on the recipients of that information in the context of their activities on the secondary market.
477That finding is established by the discussions examined by the Commission in Section 4.2 of the contested decision.
478In that connection, first of all, the information on the conduct that one or more of the traders intended to adopt on the secondary market, in the run-up to the issuance of a bond on the primary market or thereafter, in order to steepen or flatten the curve and therefore push the prices of a bond up or down on the secondary market, allowed the evolution of the price of the bond concerned on the latter market to be anticipated as a result of the conduct of those traders, and action to be taken as a consequence on that secondary market.
479Next, it should be pointed out that, for the reasons set out in paragraphs 472 and 475 above, when information on mid-prices, yield spreads and trading positions was exchanged with an auction on the primary market in mind, that information could also be of value on the secondary market. In other words, when the traders disclosed information on their mid-prices, their yield spreads or their trading positions in the run-up to or during an auction on the primary market, they were providing valuable information on their prices or their position on the secondary market. That information was all the more likely to be of value on the secondary market inasmuch as it was exchanged by primary dealers who were generally expected or required to quote two-way prices continuously on that market (see paragraph 14 above).
480Furthermore, as regards information on the levels of bidding and overbidding, on the volumes sought and on the volumes and prices ultimately obtained in the context of an auction on the primary market, that information was of a certain value in engaging in activities on the secondary market.
481A number of discussions examined by the Commission in fact demonstrate that, on the primary market, the levels of bidding and overbidding, along with the volumes requested, were linked to the impression formed by the primary dealers as to the expected level of demand on the secondary market and the value of the bond concerned on the latter market. In other words, as the Commission stated in recital 479 of the contested decision, overbidding on the primary market generally reflects high demand on the secondary market, and vice versa.
482Finally, information on failed bids submitted by primary dealers, or on the volumes obtained and the prices thereof, exchanged following an auction on the primary market, allowed a precise overview to be formed as to the conduct of those primary dealers on the secondary market.
483By way of example, as the Commission explained in recital 29 of the contested decision, if the primary dealer’s bid is not entirely successful on the primary market, customer orders can only be filled on the secondary market, as that dealer has been unable to fill all those orders on the primary market.
484Again by way of example, in a discussion of 4 November 2010, referred to in recital 302 of the contested decision, the RBS trader explained that he was long and needed to sell (‘stupid fuking trade took me long 600 lots from 76 just got risk in now’).
485Thus, information concerning the outcome of auctions on the primary market, in particular the volumes obtained and the purchase prices for EGBs, was valuable in acting on the secondary market, inasmuch as it allowed the conduct of primary dealers on that secondary market, as well as the volumes of EGBs available and the prices thereof, to be anticipated.
486It should be added that, as is clear from paragraphs 434 to 438 above, information on the exact timing of the pricing of a syndication on the primary market allowed movements on the secondary market to be anticipated and, in that way, could influence the conduct, on the secondary market, of traders who had that information.
487In the sixth place, the value of the information exchanged and the competitive advantage conferred by way of those exchanges are also demonstrated by the context in which those exchanges took place and by the views expressed by certain traders.
488First of all, the information in question was exchanged in closed chatrooms. In that regard, as the Commission observed in recitals 80 and 417 of the contested decision and as is apparent, for instance, from the discussions of 13 December 2007 in the CODS & CHIPS chatroom (‘I am gonna invite [another trader] into this chat unless anyone has any objections?’) and in the DBAC chatroom (‘… do either of u guys mind if i invite [another trader] into the other chat?’), referred to in recital 134 of the contested decision, access to the CODS & CHIPS chatroom was ‘by invitation only’. The same applied to access to the DBAC chatroom, as is demonstrated in particular by a discussion of 16 October 2009 in that chatroom, referred to in recital 271 of the contested decision. During a discussion of 19 July 2007, referred to in recital 119 of the contested decision, one of the participants described the DBAC chatroom, then known as ‘30yr kings’ as a ‘select club’ (‘it’s a select club but there you go’).
489The exchanges of information therefore took place within a trusted group of traders, as is confirmed by the discussions of 13 December 2007 on the DBAC and CODS & CHIPS chatrooms, referred to in recital 134 of the contested decision, in the course of which the traders discussed the possibility of inviting new members into the CODS & CHIPS chatroom and, in particular, whether the candidates could be trusted (‘can’t be trusted’; ‘i dont trust him’).
490Next, on the basis of the discussions which it analysed, the Commission highlighted several items of evidence showing that access to the chatrooms was predicated on the expectation that participants would disclose commercially sensitive information, which would confer an advantage on all members of the chatroom.
491By way of example, the Commission observed in recital 294 of the contested decision that, when adding a new trader to the DBAC chatroom, which operated at that time under the name of CODS & CHIPS, one of the members presented the conduct implemented in that chatroom by stating that ‘we just try to help each other when we can’. Other discussions demonstrate that the participants asked for or offered help (‘i need help with that’; ‘“I got hit on telly earlier” “so just offering it back” “im helpful like that for you guys”’; ‘glad to help’). The discussion of 18 January 2011, referred to in recital 307 of the contested decision, is an equally accurate reflection of the participants’ expectations of each other, since it shows that one participant asked another to ‘make yourself useful and put it on the chat’ ‘you not on for free’.
492Moreover, it is apparent from the discussions of 4, 8 and 10 January 2008, to which the Commission referred in recital 378 of the contested decision, that some traders who were members of the chatrooms – and, more specifically, the RBS and UBS traders – complained of the fact that other members were not sharing information in the chatrooms in question. Such criticisms, which also feature in a discussion of 27 February 2008, referred to in recital 157 of the contested decision, would not have been made if the information exchanged were of no value and if the exchanges did not confer an advantage on the participants.
493It should be added that, as is apparent in particular from recital 445 of the contested decision, various views expressed in the discussions that took place in the chatrooms attest to the precautions taken by the traders in order to avoid their conduct coming to the knowledge of persons who were not members of those chatrooms.
494In that connection, for instance, a discussion in the CODS & CHIPS chatroom dated 29 January 2007, referred to in recital 104 of the contested decision, shows the precautions taken by the participants. In fact, before adding the BofA trader to that chatroom, the UBS and RBS traders expressed the wish that that addition not result in the existence and content of the views exchanged in that chatroom coming to the knowledge of the managers of the trader in question (‘“sure but we ll cut his balls” “if things get out” “ :-)”’; ‘fine as long as [BofA] doesn’t see it and when hes away from desk he logs out thats fair and no chance if [BofA] tries to fuk us’). Moreover, when he was added, the BofA trader reassured the other members of the chatroom in that regard (‘Don’t worry will close the chat if I am off the desk, nothing goes out’).…
(iii)The arguments put forward by UniCredit and Nomura in relation to all of the discussions referred to in the contested decision
…
–The alleged excessive reliance on leniency statements
…
533Furthermore, in the second leniency statement made by RBS on 3 November 2015, that bank did not omit the fact that certain discussions had given rise to exchanges of information that went beyond what was acceptable on the secondary market. Nor did it neglect to mention the consequences, on the primary market, of exchanges of information relating to the secondary market, and vice versa.
534Finally, RBS admittedly stated that the most obvious examples of anticompetitive conduct could be seen in the discussions that took place between January 2009 and February 2010, namely an infringement period that was not relied upon against UniCredit. However, that statement of 3 November 2015 also set out discussions that took place in 2011 and, in particular, 3 November 2011.
535More generally, the leniency statements made by RBS and UBS, produced by the Commission in Case T‑453/21, demonstrate that those leniency applicants described conduct arising in particular between 2007 and late 2011, that that conduct was also linked to the secondary market for EGBs, and that the UniCredit trader had participated in that conduct.
536As to the probative value of the leniency statements by UBS and RBS in so far as concerns the discussions of 3 and 28 November 2011, which are specifically disputed, this will be examined as part of the assessment of the criticisms specifically directed against those discussions by UniCredit.
537It follows that, subject to the examination of the probative value of the leniency statements in so far as concerns the discussions specifically disputed by UniCredit, the Commission was justified in finding that those statements confirmed the anticompetitive nature of a large number of the discussions that it had examined, even though the probative value of the statement made by UBS on 29 June 2016 was lower than that of the other leniency statements.
–The alleged public nature or the alleged lack of value of one or more types of information exchanged
538Inasmuch as, inter alia, Nomura annexed to the application a document described as an ‘expert report’, it is appropriate to examine, first, the arguments put forward and evidence produced by that bank and, second, the arguments put forward and evidence produced by UniCredit.– The arguments put forward and evidence produced by Nomura
539In its first plea in law, Nomura claims that the Commission did not properly examine, in recitals 3 to 51 of the contested decision, the characteristics of the primary and secondary EGB markets, which were such that the likelihood of collusion through exchanges of information was very low, in particular in 2011.
540Relying on an expert report written by an industry professional, Nomura maintains, first, that the markets concerned were highly complex, highly competitive, highly fragmented, highly transparent and well known to the players who knew the value of EGBs on the primary market given the prices charged on the secondary market; and, second, that the banks concerned occupied asymmetric and highly differentiated positions on those markets. It follows that only exchanges of information that is confidential, precise, forward-looking and highly proximate in time to the transactions to which it relates are likely to reduce uncertainty on the market. The information exchanged and referred to by the Commission does not have those characteristics, in particular in so far as it does not, for the most part, concern prices.
541Thus, Nomura states that, first, the ‘mids’, to which the Commission referred in Categories 2 and 4, constitute the combination of publicly available, rapidly outdated information the disclosure of which by a trader may be legitimate for the purposes of examining trading opportunities with counterparties, without necessarily being relevant to another trader. Second, the information on volumes concerning the primary and secondary markets referred to in Categories 2 and 4 is already known to traders and their exchanges regarding the primary market are not capable of affecting competition on account of the limited market share of the banks concerned. Third, the yield curves, referred to in Category 1, and the yield spreads, referred to in Categories 2 and 4, are predominantly publicly available information and the exchange of those spreads is legitimate in order to verify whether trading opportunities exist. Fourth, the positions referred to in Categories 2 and 4 do not necessarily indicate a trader’s intentions due to the vagueness of the exchanges involved and the complexity and asymmetry of the EGB markets. Moreover, there may be legitimate reasons for sharing positions. Fifth, the timing of the pricing of syndications, referred to in Category 4(iii), can largely be identified through the public announcements made by the debt management offices and market behaviour in response to events leading up to that process. Sixth, information relating to overbidding is, in the specific context of EGBs, foreseeable in the light of the past practice of the banks concerned and is of little relevance for traders. In the present case, that information is too vague and irrelevant due to the complexity and asymmetry of the EGB sector.
542In that connection, in the first place, it should be observed, as a preliminary point, that the expert report submitted by Nomura was in fact prepared by a third party and that there are no grounds for calling into question the absence of any conflicts of interest on which that third party relies. In that report, the third party in question relies, first, on his qualifications, namely the fact that he holds a Master in finance and professional certification awarded by the Autorité des marchés financiers (AMF) (Financial Markets Authority, France); and, second, on his 20 years’ professional experience as a trader or as a manager of a group of traders in the EGB sector.
543It must also be observed that the expert report in question was prepared at the request of Nomura for the purposes of the present proceedings and in the light, in particular, of that bank’s response to the Statement of Objections and the application lodged by that bank in the present case.
544As regards the content of that expert report, the author explains that he has been asked to discuss certain topics with particular reference to the primary market. In that regard, a reading of that report shows that the author did indeed focus on the primary market. First of all, Nomura’s expert describes the role, functioning and composition of an EGB trading desk within a bank, along with the role of primary dealers and their relationship with debt management offices. Next, that expert explains the circumstances in which a bank internally formulates and implements its bidding strategy for an upcoming auction. Finally, that expert describes the factors and information relevant to bidding in EGB auctions.
545More specifically, first, Nomura’s expert admittedly explains that the strategy and budget relating to a particular auction will be developed by a range of stakeholders at a bank; that the bank’s EGB trading desk will not have ultimate responsibility for developing that strategy; and that the sole expectation of the traders is that they implement that strategy.
546However, that expert also explains that the internal EGB trading desk plays a part in developing that strategy and, above all, that the traders usually have considerable autonomy as to how they conduct their day-to-day trading activities and they tend to have a degree of discretion as to how they might achieve the objectives of the bank’s auction strategy. Moreover, as to the approach taken in auctions, the report shows that it is the traders who are responsible for forming their own opinions on the basis of the information available to them.
547Thus, the expert report submitted by Nomura fails to establish that, having regard to the role of traders in a bank, exchanges of information between traders on the primary market for EGBs have no effect on the bids placed by those traders in an auction.
548Second, the circumstance, mentioned by Nomura’s expert, that banks may have highly differentiated priorities when going into a given auction does not support a finding that exchanges of information specific to other banks, of the type exchanged during the discussions examined by the Commission, are incapable of influencing the conduct of the recipients of that information.
549First, the existence of a differentiated strategy continues to be framed by the rules set by the debt management office and, in particular – as Nomura’s expert points out, moreover – by the obligation to participate in auctions. Such a framing may consequently result in a bank, which has the status of primary dealer, undertaking to submit a bid and acquire a minimum volume in auctions, at the risk of losing that status and the associated advantages. Second, the reciprocal exchange of sensitive information, which is specific to each bank, can enable the recipients of that information to become aware of the strategies of the banks whence that information comes and therefore remove uncertainty in that regard and, moreover, assess whether it is possible to align those strategies. Furthermore, the fact that the banks trade different volumes of EGBs does not deprive the exchange of such information of interest in so far as such an exchange can serve in adjusting their respective strategies in order to obtain the desired volume of EGBs at the best price in an auction.
550Third, it must be pointed out that it is admittedly apparent from the expert report submitted by Nomura that, before formulating a bid in the context of an upcoming auction, traders take into account a range of factors and rely on publicly available data. However, that report also explains that a trader’s appetite for risk is one of those factors. The report also emphasises that the bid will reflect a trader’s own view of the current and future value of the bond, market evolution, and his or her order book and inventory.
551It is therefore apparent from the expert report that, on the primary market, publicly available information is not the only factor taken into account by each bank, through its traders, in order to define its bids in an upcoming auction, and that the information available to traders does not allow the strategy that each trader will use to be known.
552Fourth, as regards competitors’ mid-prices, Nomura’s expert does not state that those prices are public. At the hearing, moreover, Nomura accepted that the mid-prices specific to each bank were not publicly available. Nomura’s expert points out only that knowledge of another trader’s mid-prices in an ongoing auction is of insignificant strategic value, on the ground that traders’ pricing will generally be kept within a tight range, close to the market-average mid-price publicly available on the trading platforms.
553That explanation therefore does not exclude the existence of differences between each trader’s mid-price and the available market-average mid-price. It also does not preclude a trader’s mid-price being outside the tight range in question. Thus, that explanation is not such as to affect the Commission’s assessment based on the content of the discussions that it analysed, namely that each bank’s mid-price constitutes commercially sensitive information, the disclosure of which to competitors reduces uncertainties and facilitates coordination.
554It follows that the expert report produced by Nomura does not serve to call into question the Commission’s findings that the information exchanged between the traders was sensitive on the ground that that information was not available to other traders, and on account of the subject, accuracy and timeliness of the information exchanged. That report also does not serve to call into question the Commission’s findings that the same information had a certain value for the recipients thereof and conferred an advantage on them on the ground that that information reduced uncertainties and created opportunities for coordination.
555In the second place, the other arguments put forward by Nomura, recalled in paragraph 541, must also be rejected in so far as they cannot call into question the findings made by the Commission on the basis of the content of the discussions between the traders, which discussions were confirmed by the leniency applicants.
556In particular, inasmuch as Nomura maintains that certain information was exchanged between the traders for legitimate purposes in order to assess the existence of trading opportunities, it should be pointed out – without prejudice to the examination of the content of the discussions specifically disputed by that bank – that such claims directed against all of the other discussions that are not disputed and were examined by the Commission are too general, and that the Commission expressly excluded from the scope of the infringement at issue those exchanges which were clearly and exclusively aimed at exploring bilateral trades, as is apparent from recitals 392, 561, 575 and 644 of the contested decision.
557In the third place, and in any event, the assessments made by the Commission and confirmed by the leniency applicants are also supported for the most part by the answers given by the European Securities and Markets Authority (ESMA) to the questions put by the Court pursuant to the second paragraph of Article 24 of the Statute of the Court of Justice of the European Union and Article 89(3)(c) of the Rules of Procedure of the General Court.
558In that connection, first, ESMA explains that the mid-prices specific to each bank, the yield curves and spreads specific to each bank, as well as the bid volumes on the primary market, are neither public information, easily accessible nor widespread among the public.
559ESMA adds that the availability of information on the volumes traded on the secondary market was different across jurisdictions, but that it is clear that that information has not been available across EU jurisdictions since 2011.
560As to the long or short trading positions of each trader or each bank, ESMA explains that these constitute sensitive information for a trading strategy, and that such information did not appear to be easily accessible in 2011 since, even in 2020, it was not easily accessible.
561Moreover, ESMA stresses that the choice to display banks’ bids anonymously on D2D platforms is individual to each platform but that, generally, fixed-income D2D platforms have traditionally been dark and anonymous in the aim of providing a high degree of protection to traders.
562Furthermore, as regards the timing of the pricing of syndications, ESMA explains that it appears that the type of information provided to the primary dealers depends on their role. On that point, ESMA’s answer is conclusive and therefore carries a certain probative value only in that it explains, on the basis of a document dated May 2015, that the lead manager banks may have additional information at their disposal compared with other banks that are not lead managers.
563In addition, ESMA states that, in so far as the primary dealers are selected by the debt management offices and this gives rise to certain privileges and certain obligations, their status should be known before the auction.
564Second, it is apparent from ESMA’s answers that the following information confers a market advantage on the trader who receives that information, compared with a trader who does not have that information: the mid-prices specific to each bank, yield curves and yield spreads specific to each bank, bid volumes on the primary market and the volumes traded on the secondary market, future trading positions or past trading positions relating to transactions that took place in the preceding minutes, and a trader’s level of overbidding.
565As regards the question whether the exchange of information relating to the timing of the pricing of syndications between primary dealers confers a market advantage on the trader who receives that information, compared with the trader who does not, the probative value of ESMA’s answer is subject to the question whether the timing of the pricing of syndications constitutes relevant information for the pricing of a financial instrument. ESMA states in response that, if the timing of the pricing of syndications constitutes information relevant to the pricing of a financial instrument, this might constitute a market advantage compared with a trader who does not have that information. ESMA states that it is considered that the timing of the syndications has a role in the pricing of the security concerned.
566Third, ESMA points out that the information relating to transactions on the primary market, similarly to overbidding and the different types of information referred to in paragraphs 558 to 560 above, confers a market advantage on the trader who has that information, be it taken in isolation or as a whole, on both the primary and secondary markets, compared with a trader who does not have that information. However, ESMA observes that the extent to which such information could be regarded as an advantage is to be assessed on a case-by-case basis.
567Fourth, on the one hand, ESMA states in response that it cannot be concluded that market makers are required by means of a market making agreement to maintain frequent direct contacts in which they exchange information. On the other hand, ESMA takes the view that banks acting as market makers do not necessarily have to have direct contact in order to exchange information amongst themselves.
568Nomura’s observations on ESMA’s answers are not capable of showing that those answers are devoid of probative value.
569In that connection, first, Nomura’s argument that ESMA’s experience in 2011 is limited, since that agency was only established in 2011, must be rejected. First of all, ESMA is the legal successor to the Committee of European Securities Regulators, which was itself established by Commission Decision 2001/527/EC of 6 June 2001 (OJ 2001 L 191, p. 43). Next, ESMA states that the responses to the questions put by the Court are provided by a policy officer and a senior policy officer of Trading Unit and endorsed by their managers in the Markets and Digital Innovation Department. The status of the authors of, and of the persons who endorsed, those replies at ESMA attest to their experience in the sector and, therefore, to their capacity to know and understand how the EGB sector operates, including in 2011. Finally, ESMA’s answers are based, in particular, on two documents prepared precisely by the Committee of European Securities Regulators at the times of the facts alleges against the banks concerned.
570ESMA’s status, and that of the authors of the replies provided by the latter agency and of the persons who endorsed those answers can confer probative value on those answers, including in 2011, contrary to Nomura’s claims.
571Second, it is true that, as Nomura essentially observes, ESMA pointed out that the tasks entrusted to it concern only the secondary market. Thus the probative value of that agency’s replies relating to the primary market is, in the absence of additional references, inferior to the probative value of its replies relating to the secondary market.
572However, first, ESMA explains that it fulfils its mission to enhance investor protection and promote stable and orderly financial markets through four activities: (i) assessing risks to investors, markets and financial stability, (ii) completing a single rulebook for EU financial markets, (iii) promoting supervisory convergence and (iv) directly supervising specific financial entities. Those activities necessarily require sound knowledge of how the primary market operates.
573Next, ESMA points out that the replies to Questions 5, 8 and 9, which concern the primary market, are provided on the basis of the research conducted. That is why, in response to Question 5, which concerned the timing of the pricing of a syndication, ESMA relies on a document prepared by the World Bank Group from May 2015 entitled ‘Domestic Syndications’. In response to Question 8 – which concerned the obligation and the necessity, for market markers, to have frequent contacts in the course of which they exchange information such as that referred to in the other questions put by the Court – ESMA analyses the obligations incumbent on market makers on the basis of the legislation in force in 2011, the legislation post-2011, and on the basis of the consultations that ESMA conducted in 2021. Furthermore, as regards Question 9, which concerns the question whether the status of primary dealer for a particular auction was known before that auction, ESMA refers to a report produced by the Financial Stability Board in October 2022, entitled ‘Liquidity in Core Government Bond Markets’. The sources used by ESMA, which consisted in the legislation applicable before and after 2011, as well as reports prepared by public bodies for purposes other than the present proceedings, were therefore particularly reliable.
574The sources used by ESMA are therefore capable of conferring a high probative value on the latter’s replies concerning the primary market and, more specifically, the timing of the pricing of syndications, the role of market maker, as well as the role of primary dealer.
575Finally, and more generally, it should be pointed out that the Court asked ESMA to substantiate its answers, where necessary, by means of a scientific or methodological tool capable of confirming the relevance of those answers. It is in that context that ESMA refers, in support of its answers, to two documents that are contemporaneous to the alleged misconduct. Moreover, the fact that the other documents used by ESMA were drafted and published at a later date than that on which the alleged practices came to an end does not, by that mere fact, strip all probative value from those documents or the replies given by ESMA on the basis of those documents. Indeed, certain of those later documents contain information that directly relates to the period concerned by the alleged misconduct. Furthermore, ESMA relies on other documents and data produced after the conduct at issue in order to support the content of those of its replies that relate specifically to the functioning of the EGB sector at the time of the practices alleged against the banks concerned. In particular, it is apparent from ESMA’s replies and the sources on which it relies that, despite developments in legislation and practice, characterised by growing transparency since the time of the conduct at issue, several types of information covered by the questions put by the Court were still not available to traders in the EGB sector. In other words, certain of the replies given by ESMA are based on the following line of reasoning: given that, subsequent to the conduct at issue, several types of information about which ESMA was questioned were still not available to the other traders, that information was, a fortiori, not available at the time of that conduct.
576Thus, a high level of probative value can be attributed to the replies given by ESMA concerning the primary market, equivalent to that attributed to the replies relating to the secondary market, where those replies are supported by reliable and relevant sources, such as reports prepared by public bodies or professional associations for purposes other than the present proceedings. It follows that, contrary to what Nomura essentially claims, the fact that ESMA relies on external sources and refers to legislation and reports post-2011 does not affect the probative value of the answers that it provided.
577Third, Nomura has no basis for relying on its own experience and the expert report that it has produced in annex to the application, for the reasons set out in paragraphs 542 to 556 above. In any event, in the light of ESMA’s status and mission, the status of the authors of those replies, together with the legislation and other documents on which it relies, Nomura’s observations based on its own experience and the expert report that it produced in annex to the application cannot have greater probative value than ESMA’s answers.
578Fourth, as regards Nomura’s argument that ESMA’s replies are based on flawed economic assumptions and, in particular, on the mere fact that receiving information of the type mentioned by the Court in those questions confers a competitive advantage on the recipients of that information, it should be borne in mind that ESMA was asked about certain specific types of information that are specific to a bank, namely mid-prices, yield curves and yield spreads, bid volumes on the primary market and volumes traded on the secondary market, future trading positions or past trading positions relating to transactions that took place in the preceding minutes, and a trader’s level of overbidding. ESMA’s answer, in so far as concerns the sensitive nature of that information, was therefore necessarily based on the fact that the information in question was sufficiently accurate, detailed and current or recent.
579Next, ESMA substantiated its reply that such information conferred an advantage on the recipient of that information compared to a trader who did not have it, in particular in the context of the first question put by the Court, to which ESMA referred, and in the context of its answers to the questions relating to yield curves and yield spreads, bid volumes on the primary market and volumes traded on the secondary market, future trading positions or past trading positions relating to transactions that took place in the preceding minutes.
580In fact, first, ESMA sets out the degree of transparency on the market for EGBs at the time of the conduct at issue, together with the developments in the legislation in that regard. On that basis, it observes, inter alia, that in 2011, bond trading was still done primarily over the counter and that the mid-prices of bonds traded over the counter were not public or easily accessible and widespread among the public. More generally, and by substantiating its answer, ESMA highlights the low degree of transparency on the market for EGBs and, therefore, the limited public access to information concerning those bonds.
581Second, ESMA explains, again by substantiating its answer, that the prices of other market participants are extremely valuable to make trading decisions, especially those from informed traders. According to ESMA, since obtaining information is costly, traders are not willing to share this information and, if the information – that is to say, their quotes or their orders – is public, traders do not want to disclose their identity to avoid that others know that this is a price from an informed trader; otherwise, the informed traders no longer have the incentive to research such information.
582Finally, as regards Nomura’s claim that ESMA’s replies fail to address the arguments that Nomura puts forward in the application, it should be observed that, admittedly, that authority was asked to answer the questions put by the Court by duly taking account, on the one hand, of any particularities of the sector concerned, namely the issuing and trading of EGBs on the primary and secondary markets; and, on the other hand, of assessments concerning that sector, made by the Commission in the recitals of the contested decision referred to in each question.
583However, ESMA did not have available to it the pleadings exchanged between the parties in the course of the present proceedings, such as the application lodged by Nomura. Moreover, the questions put to ESMA and the replies that it provided are not intended to give a final assessment of the pleas in law and arguments put forward by Nomura in the application, or on the existence or otherwise of errors made by the Commission in the contested decision. More specifically, ESMA’s replies do not address the question whether, in the light of their specific content and of the characteristics of the information exchanged, such as its level of accuracy, each of the discussions on which the Commission relies in the contested decision was anticompetitive. Those replies also do not address the question whether, in the present case, there were legitimate reasons for making the exchanges of information at issue. More generally, those replies do not relate to the question whether the conduct alleged against the participants constituted a restriction by object in the circumstances of the case.
584Jurisdiction for making such a final assessment lays with the Court alone.
585Nomura therefore has no basis for arguing that ESMA’s reply rests on flawed economic assumptions on the ground, first, that that authority makes unsupported assumptions regarding the value of information on the market for EGBs and, second, that it failed to address the evidence relied upon by that bank to demonstrate that increased transparency does not necessarily confer a competitive advantage.
586Fifth, as regards Nomura’s criticisms directed against the content of each of ESMA’s answers, it must be pointed out that those criticisms are confined to recalling the content of the application, are not substantiated or are not capable of demonstrating that those answers are devoid of probative value.
587In particular, it is true that ESMA’s reply highlights that there are different market conditions in different areas of jurisdictions relating to transparency. However, that answer and the documents to which ESMA refers also highlight the fact that the degree of transparency before and after trading on the secondary market for EGBs was low in most of the Member States examined.
588Moreover, the fact that ESMA did not cite a quotation from an academic publication in full is not capable of affecting either the meaning of that quotation or the probative value of ESMA’s answer. It is in fact apparent from that quotation in full that an individual trader does not know the fundamental value of a financial instrument on the market, that is, the indisputable value that would be given to that instrument if the information were perfect, and that only the markets – understood as an abstract group of market participants who independently exchange that financial instrument – produce, a posteriori, prices that are ‘close’ to that fundamental value. Moreover, Nomura has no basis for claiming that the quotation in question suggests that the disclosure, within a restricted circle of traders, of information that is not publicly available or not easily defined by a single trader increases ‘overall’ transparency in the market.…
602In that connection, UniCredit’s observations on ESMA’s answers are not capable of showing that those answers are devoid of probative value.…
614Consequently, without prejudice to the examination of the content of the discussions specifically disputed by UniCredit, the arguments put forward by that bank must be rejected in so far as they seek to demonstrate that the information referred to in paragraph 380 above was not sensitive in nature and was of insignificant value and, more generally, that the exchanges of information identified by the Commission were not anticompetitive.
–Nomura’s arguments alleging infringement of essential procedural requirements in respect of the aggregation of distinct categories of conduct
615In the context of its fifth plea, Nomura argues that the Commission infringed essential procedural requirements in so far as concerns its aggregation of different categories of conduct, on the ground that it failed to provide an explanation, at any point, as to how the four different categories of conduct, identified by the Commission in recital 93 of the contested decision, were interrelated, intertwined and overlapped. It adds that the Commission failed to establish that those four categories were intertwined and overlapped.…
(3)The anticompetitive nature of the discussions specifically disputed by Nomura, UniCredit and Portigon
…
(i)The discussions specifically disputed by UniCredit and Nomura
…
–The discussion of 5 May 2011
689As regards the discussion of 5 May 2011, classified in Categories 2, 3 and 4 and referred to in recitals 328 and 684 of the contested decision, Nomura claims that the discussion between the traders of Nomura and RBS concerned a legitimate bilateral trading opportunity on the secondary market.
690In that connection, the transcript of the discussion concerned confirms that the Nomura trader was indeed in the CODS & CHIPS chatroom on 5 May 2011. Moreover, Nomura does not specifically dispute that certain exchanges that day were rightly regarded by the Commission as attempts to coordinate bids on the primary market and as attempts to coordinate the level of overbidding on the primary market (Categories 2 and 3). In that regard, as the Commission observed in recital 328 of the contested decision, the traders requested and offered advice on how they should bid or overbid for different maturities and aligned their bids in an auction. Following the auction, the traders expressed their satisfaction. The RBS trader comments ‘man that auction came right where we wanted it around that .10 level’. The Nomura trader also responded by expressing his enthusiasm at the profits made (‘was perfect fuking made some money for a change’). The UBS trader, for his part, replied that it ‘was bang on’.
691Lastly, it is apparent from the transcript of the discussion of 5 May 2011 that, during the sequences of messages identified by the Commission in Annex 1 to the contested decision, the participants engaged in further exchanges of sensitive information conferring an advantage on the recipients of that information on both the primary and secondary markets (‘im long hopefully sell around 50/55ish then going to be short’; ‘where we got curve? ‘62.8’ ‘same we got 62.75 think may be steeper though’).
692In the light of the foregoing, and even if the exchange between the traders of Nomura and RBS at 10:36:54, referred to in recital 684 of the contested decision, were indeed a legitimate exchange intended to explore a bilateral trading opportunity, the Commission could validly take the view, in recitals 328 and 684 of that decision, that the discussion of 5 May 2011 concerned sensitive information relevant to the trading of EGBs on the primary or secondary markets and, accordingly, that that discussion was anticompetitive in nature.
693This is all the more the case since the Commission found, without Nomura raising any objections other than those examined and rejected in paragraphs 509 to 515 and in paragraphs 527 to 529 above, that the anticompetitive nature of the discussion that took place on 5 May 2011 was confirmed by leniency statements made by RBS and UBS, as is apparent from recital 684 of, and footnote 1220 to, the contested decision. Furthermore, the leniency statement made by UBS on 29 June 2016, produced by the Commission pursuant to a measure of inquiry taken by the Court, does in fact refer to the passages reproduced in paragraph 690 above.
–The discussions of 18 May and 22 June 2011
694As regards the discussions of 18 May and 22 June 2011, classified in Category 4 and referred to in recitals 329, 333, 686 and 690 of the contested decision, Nomura claims that those discussions concern syndications of suprasovereign, sub-sovereign and agency bonds (‘SSA bonds’), and not of EGBs.
695That error of assessment also led the Commission, in Nomura’s submission, to infringe its obligation to state reasons, the rights of the defence, the principle of sound administration, the presumption of innocence, essential procedural requirements and the Treaties by including, with no explanation, in the Statement of Objections and in the contested decision, conduct unrelated either to EGBs or to the operative part of that decision. Lastly, in the reply, Nomura states that not taking those discussions into consideration would have an impact on the amount of the fine imposed on it.
696In that connection, the Commission does not dispute that the discussions of 18 May and 22 June 2011 related to SSA bonds. Furthermore, at the hearing, the Commission explained that it was not relying on those discussions in order to establish the anticompetitive nature of the conduct of the traders in those discussions. However, it stated that it was relying on those discussions for context in addition to the other items of evidence available to it in order to show how the traders engaged in discussions about syndications and, in particular, exchanges of information on the exact timing of the pricing of a syndication.
697Moreover, as regards the claims referred to in paragraph 695 above, the Commission takes the view that these are ineffective since they cannot lead to the annulment of the contested decision or have an impact on the nature, gravity or duration of the infringement committed by Nomura. Furthermore, the Commission contends that (i) it afforded Nomura the opportunity to express its view on that point as part of the administrative procedure, (ii) it validly substantiated, in recitals 685, 686 and 690 of the contested decision, the reasons why it took those discussions into account, and (iii) it did not hold Nomura liable for an infringement in relation to SSA bonds and, therefore, it did not breach the presumption of innocence vis-à-vis that bank.
698In that connection, as has been recalled in paragraph 332 above, in the field of competition law, where there is a dispute as to the existence of an infringement, it is for the Commission to prove the infringements found by it and to adduce evidence capable of demonstrating to the requisite legal standard the existence of the circumstances constituting an infringement.
699The existence of discussions between the traders of the banks concerned as to the timing of the pricing of SSA bond syndications is not ‘capable of demonstrating’ the existence of a single and continuous infringement in the EGB sector, if the principle of the presumption of innocence is not to be disregarded. Furthermore, having regard to the fact that SSA bonds are not EGBs, and in so far as the Commission has failed to demonstrate that the syndication procedures for SSA bonds were similar to the syndication procedures for EGBs, that institution also cannot rely on the discussions of 18 May and 22 June 2011 for context.…
–The discussion of 28 November 2011
…
793Without it being necessary to assess the nature of the disclosures of offers made by UniCredit (‘offered 15m rag37s’; ‘offered 15m dbr42s also’), it must be held that, like all the participants in the chatroom, that bank became aware of the bids disclosed by the UBS trader concerning that day’s auction for Belgian EGBs.
794However, as has been stated in paragraphs 448, and 476 to 486 above, and as has been confirmed by ESMA in response to questions put by the Court (see paragraphs 564 to 566 above), overbidding is information that confers a market advantage on the trader who has that information, on both the primary and secondary markets, compared with a trader who does not have that information.…
(ii)The discussions specifically disputed by Portigon
…
–A sequence of messages from the discussion of 11 April 2011
…
903Moreover, an exchange of information on past transactions, albeit which had just taken place, is capable of providing all of the chatroom members with information on the direction of future trading.
904In addition, again in recital 673 of the contested decision, the Commission noted, without this being challenged by Portigon, that in the course of the exchanges to which that bank refers, the traders of UBS and Nomura had commented on what they saw on screen regarding the secondary market, and thereby disclosed information about individual clients, trading positions and exchanged forward-looking views on how the market would develop and price-related information, in plain sight of all of the four other traders in the CODS & CHIPS chatroom.
905Consequently, the Commission could validly take the view, in recitals 323 and 673 of the contested decision, that the discussion of 11 April 2011 concerned sensitive information relevant to the trading of EGBs on the primary or secondary markets and, accordingly, that that discussion was anticompetitive in nature.
–The discussion of 12 April 2011
906As regards the discussion of 12 April 2011, Portigon maintains that the content thereof has not the slightest specific link to specific commercial transactions.
907However, it has already been found, in paragraphs 673 to 676 above, that that discussion concerned sensitive information relevant to the trading of EGBs on the primary or secondary markets and, accordingly, was anticompetitive in nature.
–A sequence of messages from the discussion of 18 May 2011
908As regards the discussion of 18 May 2011, the Commission explained, in recitals 329 and 686 of the contested decision, that during that discussion, two UBS traders, a Nomura trader and the Portigon trader had exchanged information in the CODS & CHIPS chatroom, which related to the timing of the pricing of a syndication, of which they kept each other informed. Moreover, the Commission reproduced the views exchanged between the traders from 14:53:23 onwards in that forum.
909Furthermore, in Annex 1 to the contested decision, the Commission referred to three relevant excerpts from the discussion that took place on 18 May 2011 in the CODS & CHIPS chatroom, namely (i) between 09:03:28 and 09:11:48; (ii) between 10:00:05 and 10:44:08; and (iii) between 14:53:23 and 15:31:07.
910In that connection, Portigon claims that, first, the messages in the first excerpt concerned past transactions; second, the messages sent between 10:00:05 and 10:16:24 concerned only exchanges of general information that was known or could be observed on the market; and, third, the messages in the third excerpt had no specific link to specific commercial transactions.
911In that regard, first of all, it is true that the views expressed by the traders in the first excerpt identified by the Commission are not sufficient to support a finding that the discussion of 18 May 2011 was anticompetitive. Certain of those views are, at the very most, relevant inasmuch as they provide an understanding of the bonds that were to be syndicated that day (‘eib 5y deal to price later’ ‘yes swapped. i havent traded yet in cash’) and that gave rise to exchanges in the third excerpt identified by the Commission.
912Next, it is true that the messages sent between 10:00:05 and 10:16:24 constitute an exchange of general information on Spanish EGBs. However, Portigon’s challenge is confined to approximately 16 of the 44 minutes of the second excerpt identified by the Commission, between 10:00:05 and 10:44:08.
913Lastly, as regards the third excerpt, the transcript of the discussion of 18 May 2011 confirms that, at 14:53:23, one of the UBS traders asked ‘Has the call started yet’. The Nomura trader replied ‘“not that i know off” “hearing 4.15”’. The UBS trader replied ‘thanks’. Next, at 15:19:53, the Nomura trader asked ‘will the pricing have any effect’. The Portigon trader replied ‘i think unlikely much on this one’. At 15:20:19, the Nomura trader was of the same opinion, saying ‘s[am]e’. Finally, at 15:28:48, the Nomura trader wrote ‘doing it now’ and, at 15:29:10, the UBS trader wrote ‘everton’, that is, as the Commission explained in recital 252 of the contested decision, the code word used to alert the participants that a syndication had been priced.
914It is thus apparent from the discussion of 18 May 2011 that, as the Commission observed in recital 329 of the contested decision, the traders present at the time of that discussion obtained or attempted to obtain information on the exact timing of the pricing of a specific syndication.
915Moreover, Portigon has failed to put forward any argument capable of demonstrating that information on the exact timing of the pricing of a syndication is not sensitive information or confers no advantage.
916Consequently, Portigon has failed to demonstrate that the Commission erred in its assessment when it found, in recitals 329 and 686 of the contested decision, that the discussion of 18 May 2011 concerned sensitive information relevant to the trading of EGBs on the primary or secondary markets and, accordingly, that that discussion was anticompetitive in nature.…
(c)The objections raised by Nomura, BofA, Portigon and UniCredit as to the existence of a single infringement
…
(1)The arguments raised by Nomura and Portigon in their respective fourth pleas, alleging errors as to the existence of a common plan in pursuit of a single anticompetitive aim
…
(i)The other arguments put forward by Nomura
…
1040Moreover, before making the statements reproduced in paragraph 1037 above, the Nomura trader had replied to a question from the UniCredit trader as to why the RBS trader had to leave the chatroom, stating: ‘well u know 4 guys got booted for sharing info so thats reason i think’, which confirms that the Nomura trader was in a position to understand the anticompetitive nature of the conduct identified by the Commission.
1041Consequently, the arguments put forward by Nomura and recalled in paragraphs 1007 and 1027 above are rejected.
(iv)Conclusion
1042In the light of the foregoing, Portigon’s arguments, which are directed only against two of the seven objective elements on which the Commission relied in order to find that the conduct that it had found was part of a common plan in pursuit of a single anticompetitive aim, must be rejected.
1043As regard Nomura’s arguments, these must also be rejected.
1044Nomura and Portigon have failed to establish that the Commission erred in its assessment when it found that the conduct that it had identified was part of a common plan in pursuit of a single anticompetitive aim.
(2)The first plea in law raised by BofA, alleging, in essence, that the DBAC chatroom did not pursue the same aim as the CODS & CHIPS chatroom
1045In recital 441 of the contested decision, the Commission found that BofA was aware, or ought reasonably to have foreseen and was willing to take the risk, of the unlawful conduct planned or put into effect in the CODS & CHIPS chatroom. It stated, however, that it could not be concluded with sufficient certainty that BofA was aware or ought reasonably to have foreseen and was willing to take the risk of the existence and functioning of the DBAC chatroom or other anticompetitive communications. In the light of that assessment, the Commission, whilst maintaining its findings as to the classification of the conduct at issue as constituting a single and continuous infringement covering both chatrooms, essentially found BofA liable only for the discussions that took place in the CODS & CHIPS chatroom.
1046By its first plea in law, BofA does not seek to dispute the findings made as to its awareness of the conduct linked to the CODS & CHIPS chatroom, or its intention to contribute to that conduct. As it stated at the hearing, BofA disputes only the existence of an overall plan pursuing a single anticompetitive aim. More specifically, it takes the view that the Commission wrongly found that the conduct put into effect in the DBAC chatroom, on the one hand, and in the CODS & CHIPS chatroom, on the other hand, was part of an overall plan pursuing a single anticompetitive aim.
1047As a preliminary point, BofA claims that the Commission failed to apply the correct legal test. It is argued that the Commission failed to ascertain whether any of the various types of infringing conduct were capable of indicating that the conduct put into effect by the other banks did not have an identical anticompetitive object and was therefore not part of an overall plan.
1048In that connection, it should be recalled that the Commission defined the common plan in pursuit of an anticompetitive aim in recitals 411, 413 and 414 of the contested decision.
1049Moreover, in recitals 415 to 424 of that decision, it stated that a range of objective elements further confirmed that the chatrooms in question and the other discussions described in its decision were linked and complementary in nature and sought to achieve the aims that the Commission described in preceding recitals.
1050Furthermore, it should be noted that, in recitals 449 to 464 of the contested decision, the Commission replied to the arguments put forward, inter alia, by BofA alleging that the discussions at issue were disparate across a number of years and without a regular pattern. In that context, in recital 454 of the contested decision, the Commission replied to BofA’s argument alleging that its trader was not aware of the existence of the DBAC chatroom, and states that that fact was not relevant for determining the scope of the single and continuous infringement. Moreover, in recitals 465 and 466 of the contested decision, the Commission rejected BofA’s arguments alleging that EGBs were not homogeneous. Finally, in recitals 468 to 474 of the contested decision, the Commission replied, inter alia, to BofA’s arguments relating to its intention to contribute to the common plan and its awareness of the conduct at issue. In that connection, in recital 472 of the contested decision, the Commission rejected the arguments alleging that the exchanges that took place in the CODS & CHIPS chatroom were not anticompetitive. In recitals 473 and 474 of that decision, the Commission also took a view on BofA’s arguments that its trader had participated in the CODS & CHIPS chatroom only, that the discussions in the CODS & CHIPS chatroom were not anticompetitive, and that the DBAC chatroom had been created specifically to exclude its trader from sensitive discussions.
1051Thus the Commission examined the elements which, according to BofA, were capable of showing that the conduct materially put into effect in the two chatrooms in question did not share the same anticompetitive aim or effect. The fact that the Commission arrived at a different conclusion to that of BofA on those aspects is not capable of establishing an error of law on the part of the Commission.
(i)The fact that the BofA trader was deliberately excluded from the DBAC chatroom and was harmed by the conduct put into effect in that chatroom
1052First, BofA relies on several factors in order to show that the DBAC chatroom was created in order deliberately to exclude its trader from that chatroom. In that regard, BofA explains that the DBAC chatroom was established on 28 February 2007, after its trader had been invited into the CODS & CHIPS chatroom on 29 January 2007. BofA adds that the idea of creating a second chatroom was suggested by one of the founder members of the DBAC chatroom the very day that BofA joined the CODS & CHIPS chatroom. BofA also claims that several statements made by the founders and members of the DBAC chatroom and the contested decision show that the members of the latter chatroom formed a select group that was separate from the other traders, who were members of the CODS & CHIPS chatroom. BofA also refers to discussions in which members of the DBAC chatroom expressed the desire to use the CODS & CHIPS chatroom only for cheap talk.
1053Second, BofA maintains that the present case concerns two chatrooms with two different objectives, in that they were intended to benefit their respective members and not those of the other chatroom. BofA points to the fact that the participants in the DBAC chatroom pursued their own interests and agenda, without taking into account the interests of the participants in the other chatroom, namely the CODS & CHIPS chatroom, in which BofA participated.
1054In that connection, in the first place, it should first of all be noted that the discussions examined by the Commission in the contested decision show that the idea of creating a second chatroom was mentioned by one of the members of the CODS & CHIPS chatroom on 29 January 2007 in the presence of the BofA trader and was greeted with smiling emojis by the other two members of that group. It is also common ground that the DBAC chatroom was established on 28 February 2007, that is, approximately two months after the creation of the CODS & CHIPS chatroom, and approximately one month after the BofA trader was invited into the CODS & CHIPS chatroom.
1055Next, certain founder members of the DBAC chatroom, who had previously instigated the creation of the CODS & CHIPS chatroom, expressed their agreement, in the course of a discussion in the DBAC chatroom on 10 January 2008, with the use of the CODS & CHIPS chatroom for cheap talk. Moreover, one of the members of those two chatrooms expressed, in the course of a discussion in the DBAC chatroom on 27 February 2008, the desire to confine ‘big trades’ to the latter chatroom. That desire was a result of the fact that a trader participating in the CODS & CHIPS chatroom, who was not the BofA trader, shared no information in the latter chatroom.
1056Finally, during a discussion of 9 January 2009, namely subsequent to the period of BofA’s participation in the infringement, negative views were expressed by members of the DBAC chatroom in respect of traders of BofA, including in respect of the trader who had participated in the exchanges that had taken place in the CODS & CHIPS chatroom.
1057In the second place, in the contested decision, the Commission considered that the conduct that it had found was based on mutual assistance, and that the object of the single and continuous infringement at issue was to restrict and/or distort competition through the sharing of commercially sensitive information between members of a circle of competitors with the aim of coordinating their strategies for acquiring EGBs on the primary market and/or trading those bonds on the secondary market to the detriment of other market participants.
1058However, first, it should be observed that BofA does not dispute the fact that the exchanges that took place in the CODS & CHIPS chatroom pursued an anticompetitive aim and, more specifically, the anticompetitive aim defined in the contested decision.
1059In that connection, it is apparent, moreover, from the discussions collated in Annex 1 to the contested decision and from the discussions examined by the Commission in that decision, that the members of the CODS & CHIPS chatroom put into effect all of the types of conduct referred to in paragraph 43 above. Furthermore, that conduct was anticompetitive because it reduced uncertainties and created opportunities for the participants in that chatroom to coordinate on the primary market and on the secondary market.
1060As is clear from paragraphs 345 to 620 above, similar anticompetitive conduct was also put into effect in the DBAC chatroom.
1061Thus, similar anticompetitive behaviour was put into effect in both of the chatrooms in question, including after the creation of the DBAC chatroom and despite the facts recalled in paragraphs 1054 to 1056 above.
1062Second, in recital 473 of the contested decision, the Commission referred to six discussions in the course of which identical or similar sensitive information was exchanged in both of the chatrooms in question, namely the discussions of 24 September 2007; of 7 February, 27 March, 4 September and 6 November 2008; and of 14 January 2009, mentioned in recitals 125, 150, 165, 166, 192, 193, 214 and 225 of that decision, respectively.
1063Moreover, at the hearing, the Commission made mention of three other discussions in which parallel or intertwined exchanges took place in both of the chatrooms in question, namely the discussions of 11 January, 6 March and 30 April 2008, mentioned in recitals 146, 163 and 164, as well as in recitals 173 to 176 of the contested decision, respectively.
1064In that connection, in recital 125 of the contested decision, the Commission explained that, in the discussions of 24 September 2007, the traders used both of the chatrooms in question interchangeably, and simultaneously discussed a Belgian auction in both of those chatrooms.
1065It is also apparent from the discussions of 7 February 2008, referred to in recitals 150 and 151 of the contested decision, that those discussions gave rise to simultaneous exchanges, in both of the chatrooms in question, relating to a French auction, in particular at the initiative of the BofA trader in the CODS & CHIPS chatroom (‘What you thinking for the 15yr’s today’). In that regards, the participants in the CODS & CHIPS chatroom agreed, inter alia, not to ‘pay up large’, and the participants in the DBAC chatroom exchanged information on their levels of overbidding.
1066In recital 165 and 166 of the contested decision, the Commission stated that, in the discussions of 27 March 2008, the traders discussed an upcoming Spanish auction in both of the chatrooms in question and, in particular, the possibility of buying Spanish EGBs at a low price because the Spanish debt management office was more interested in volume than price.
1067In the course of discussions of 6 November 2008, referred to in recitals 213 and 214 of the contested decision, the participants in the DBAC chatroom and the participants in the CODS & CHIPS chatroom discussed a French auction taking place that day and exchanged price-related information. In that connection, the BofA trader asked the members of the CODS & CHIPS chatroom what they thought of that day’s auction and revealed his price for 10-year EGBs (‘+4 area’). In the same chatroom, the Natixis trader disclosed his strategy (‘no reason to pay over mid I can see’).
1068In the discussions of 14 January 2009, referred to in recital 227 of the contested decision, the participants in the two chatrooms in question discussed, inter alia, the timing of the pricing of a Belgian syndication. In the CODS & CHIPS chatroom, the BofA trader asked ‘is it priced??’ and received an answer in the affirmative from the Natixis trader.
1069The discussions to which the Commission referred at the hearing and cited in paragraph 1063 above also demonstrate that the same topics were discussed in the CODS & CHIPS and DBAC chatrooms and, in particular, the fact that, in the course of the discussions that took place in parallel, the participants in those two chatrooms exchanged information on an upcoming auction with regard to specific EGBs. The discussions that took place in the CODS & CHIPS chatroom also show that, on those occasions, the BofA trader disclosed or received sensitive information such as bidding levels, mid-prices or, more generally, bidding strategies, and that that information conferred an advantage on the recipients thereof.
1070It is apparent from other recitals of the contested decision that the members of the two chatrooms in question engaged in exchanges of anticompetitive information on the same bonds in parallel discussions in those chatrooms on 6 March 2007, and on 18 January and 11 February 2011.
1071Consequently, several discussions relating to the same bonds and giving rise to the same types of anticompetitive conduct took place concomitantly or successively in the two chatrooms in question, throughout the existence of those chatrooms, including after the facts set out in paragraphs 1054 and 1056 above.
1072In that connection, BofA does not dispute the specific content of the discussions referred to in paragraphs 1062 to 1071 above, with the exception of the discussions of 6 November 2008 and of 14 January 2009 which, in BofA’s submission, show that the two chatrooms in question were completely different and did not share a common objective.
1073However, BofA’s challenge to the discussions of 6 November 2008 and of 14 January 2009 is not relevant, because it does not concern the links between the two chatrooms, and in particular the fact that the same bonds were discussed in both of those chatrooms. That challenge in fact relates to the lack of contribution, on the part of BofA’s trader, to the anticompetitive exchanges. Moreover, on the one hand, the BofA trader did participate in the discussion of 6 November 2008, in particular by providing and receiving sensitive price-related information (see paragraph 1067 above). On the other hand, BofA’s participation in the discussion of 14 January 2009 was not relied upon against that bank.
1074Third, it is true that the members of the two chatrooms were not exactly the same, and that the CODS & CHIPS chatroom had a higher number of traders, namely a maximum of nine traders over certain time periods.
1075However, on the one hand, the four founder members of the DBAC chatroom were also members of the CODS & CHIPS chatroom. In particular, the Commission described the four traders who initially participated in the DBAC chatroom as ‘key players’ in the cartel precisely because they instigated the creation of the CODS & CHIPS and DBAC chatrooms, had access to those chatrooms from the outset, continued to have access to those two chatrooms despite changes of employer, and were involved in mutual discussions in those chatrooms throughout the existence thereof. Moreover, when, on two occasions, a member of the DBAC chatroom was not a member of the CODS & CHIPS chatroom, that member worked for the same bank as one of the four founder members of the two chatrooms in question.
1076Thus the conduct put into effect in the CODS & CHIPS chatroom could confer an advantage on the members of the two chatrooms in question, including the members of the DBAC chatroom. Moreover, the mutual expectations of the members of the DBAC chatroom could be met by the conduct of the members of the CODS & CHIPS chatroom. Thus, for example, over the course of the discussions of 4 September 2008, the participants in the DBAC chatroom discussed, inter alia, a French auction and checked the prices at which bonds maturing in 2038 had been cleared during a previous auction, including by means of a request put to the BofA trader in the CODS & CHIPS chatroom, to which he replied.
1077In other words, having regard to the presence of the members of the DBAC chatroom in the CODS & CHIPS chatroom, the anticompetitive aim pursued in the CODS & CHIPS chatroom coincided with the aim pursued in the DBAC chatroom.
1078On the other hand, it is apparent from several discussions, not specifically disputed by BofA, that some of the information exchanged in the DBAC chatroom was also exchanged in the CODS & CHIPS chatroom, and that the conduct put into effect in the DBAC chatroom could therefore have an impact on the CODS & CHIPS chatroom. In other words, several discussions show that members of the CODS & CHIPS chatroom received information exchanged in the DBAC chatroom, and that the receipt of that information contributed to reducing uncertainties and creating opportunities for coordination for the members of the CODS & CHIPS chatroom.
1079In that connection, it should be recalled that, as is clear from paragraphs 1064 and 1070 above, several discussions found in the two chatrooms in question related repeatedly to the same EGBs and those discussions gave rise to the same types of anticompetitive conduct. Furthermore, other discussions examined by the Commission, such as the discussions of 25 September 2007 and 18 January 2011, show that those two chatrooms were interchangeable.
1080Above all, several discussions not specifically disputed by BofA clearly show that the members of the CODS & CHIPS chatroom received information that had been exchanged beforehand in the DBAC chatroom.
1081In a discussion of 6 March 2007, referred to in recital 110 of the contested decision, and which took place in the DBAC chatroom, the participants in the latter chatroom discussed their bidding strategy and their levels of overbidding before an Austrian auction. In that chatroom, the UBS trader asked the other participants about the level of overbidding and stated that he was considering a level of overbidding between +14 and +18. At the same time, in the CODS & CHIPS chatroom, the same traders and the BofA trader also discussed the Austrian auction. That discussion took place at the initiative of the BofA trader. The latter in fact asked the other participants about their levels of overbidding and gave the level of overbidding that he was considering along with the volume he intended to ask for (‘What are we thinking for the austria? We are +16 for about 70m’). In reply, the UBS trader once again stated the level of overbidding that he was considering (‘+14 +18’) as well as the volume that he needed. The participants then updated and clarified their strategy. Thus, for example, the BofA trader disclosed his bids (‘+15 +17 +19… 20m each’). After the auction, the RBS trader and the BofA trader congratulated the group (‘well done everyone’ ‘all of below the average well done everybody’).
1082In recitals 174 and 175 of the contested decision, the Commission observed that, on 30 April 2008, the members of the DBAC chatroom shared, in the CODS & CHIPS chatroom, the strategy that they had just discussed in the DBAC chatroom regarding a French auction.
1083As regards the discussion of 14 January 2009, referred to in paragraph 1068 above, the BofA trader asked whether a syndication had been priced (‘is it priced??’) and received an answer in the affirmative from the Natixis trader. That information had been given to the Natixis trader by the UBS trader less than four minutes earlier in the DBAC chatroom.
1084Moreover, the Commission stated, in recital 307 of the contested decision, that the discussions of 18 January 2011 in the CODS & CHIPS and DBAC chatrooms also demonstrated the interchangeability between the two chatrooms. During the discussion that took place that day in the CODS & CHIPS chatroom, a trader asked ‘dont we have a separate [chatroom] form the septics [the Americans]? at least we see some flow :-)’.
1085Furthermore, the Commission observed, in recital 313 of the contested decision, that in a discussion of 11 February 2011 in the CODS & CHIPS chatroom, the RBS trader copied from the DBAC chatroom a conversation that he had had there with the Portigon trader.
1086Thus the discussions referred to in paragraphs 1081 to 1085 above, which discussions are not specifically disputed by BofA, clearly demonstrate that the information exchanged and the conduct put into effect in the DBAC chatroom was not necessarily reserved exclusively for the members of that chatroom, and that some of the same information was also exchanged in the CODS & CHIPS chatroom.
1087It must therefore be held that those exchanges of information and that conduct in the DBAC chatroom led to a reduction in uncertainties or the creation of opportunities for coordination in the CODS & CHIPS chatroom as well, to the advantage of the members of that chatroom and in particular of BofA during the period of its participation in the infringement at issue.
1088As to the circumstance, alleged by BofA, that certain discussions referred to in paragraphs 1081 to 1085 above took place after the period of that bank’s participation in the infringement, this cannot call the assessment of the Court into question. It is, in fact, not necessary that, for the purposes of demonstrating the existence of a common plan and of a single anticompetitive aim, all of the objective elements identified by the Commission be present during the period of the participation of each of the banks concerned.
1089The identical nature of the aim pursued by the discussions that took place in the two chatrooms in question is borne out by the fact, referred to in recital 88 of the contested decision, that the two chatrooms temporarily bore the same name – that is, CODS & CHIPS – from 24 September 2009 to 9 September 2010.
1090Fourth, and in any event, it should be pointed out, first of all, that the fact that an undertaking had no interest in participating in certain aspects of the infringement identified by the Commission is unrelated to the objective criterion of whether the various anticompetitive discussions were part of the same overall plan, but rather that undertaking’s specific subjective role in the infringement (see, to that effect, judgment of 7 November 2019, Campine and Campine Recycling v Commission, T‑240/17, not published, EU:T:2019:778, paragraph 252).
1091So far as the existence of the infringement is concerned, it does not matter whether or not the commission of that infringement was in the commercial interest of the undertaking concerned (see, to that effect, judgment of 7 November 2019, Campine and Campine Recycling v Commission, T‑240/17, not published, EU:T:2019:778, paragraph 252 and the case-law cited.
1092Thus, the fact that the BofA trader’s participation solely in the CODS & CHIPS chatroom was not in the interest of that bank does not serve to call into question the Commission’s assessment that the conduct put into effect in that chatroom and in the DBAC chatroom formed part of the same overall plan.
1093Next, the fact that different undertakings have played different roles in the pursuit of the single anticompetitive aim does not mean that there was no identity of anticompetitive object and, therefore, of infringement, provided that each undertaking has contributed, at its own level, to the pursuit of that common aim (see, to that effect, judgments of 15 March 2000, Cimenteries CBR and Others v Commission, T‑25/95, T‑26/95, T‑30/95 to T‑32/95, T‑34/95 to T‑39/95, T‑42/95 to T‑46/95, T‑48/95, T‑50/95 to T‑65/95, T‑68/95 to T‑71/95, T‑87/95, T‑88/95, T‑103/95 and T‑104/95, EU:T:2000:77, paragraph 4123, and of 8 July 2004, JFE Engineering v Commission, T‑67/00, T‑68/00, T‑71/00 and T‑78/00, EU:T:2004:221, paragraph 370).
1094Thus, the fact that BofA contributed to the overall plan solely by way of its participation in the CODS & CHIPS chatroom, unlike the traders who were members of both of the chatrooms in question, did not prevent the Commission from finding there to be an overall plan common to the CODS & CHIPS and DBAC chatrooms.
1095Finally, the fact of not complying with a cartel does not alter the existence of the cartel. Even on the assumption that it is proved that certain participants in the cartel succeeded in misleading other participants and in using the cartel to their advantage, the infringement committed is not eliminated by that simple fact (see, to that effect, judgments of 12 December 2014, Hansen & Rosenthal and H&R Wax Company Vertrieb v Commission, T‑544/08, not published, EU:T:2014:1075, paragraph 249, and of 11 July 2019, Huhtamäki and Huhtamaki Flexible Packaging Germany v Commission, T‑530/15, not published, EU:T:2019:498, paragraph 118).
1096Thus, the circumstance that the participants in the DBAC chatroom gained greater advantage from the conduct put into effect than those participating solely in the CODS & CHIPS chatroom – such as BofA – cannot call into question the existence of a single anticompetitive aim common to both chatrooms. In other words, the circumstance that the traders in the DBAC chatroom used the CODS & CHIPS chatroom to their own advantage cannot call into question the fact that both of those chatrooms formed part of a common plan in pursuit of a single anticompetitive aim.
1097It follows that the Commission was justified in finding that the items of evidence at its disposal were not called into question by the fact that the BofA trader was deliberately excluded from the DBAC chatroom or by the fact that the conduct in the DBAC chatroom was liable to cause detriment to excluded participants. This is all the more the case since the Commission found there to be several other objective elements which, in its view, also confirmed that the two chatrooms formed part of the same common plan.
(ii)The objective elements to which the contested decision refers
1098BofA disputes the seven objective elements to which the Commission referred in recitals 416 to 423 of the contested decision in order to confirm that the CODS & CHIPS and DBAC chatrooms were linked and complementary in nature and sought to achieve the aims pursued by the common plan found by the Commission (see paragraph 946 above).
1099As a preliminary point, first, it is true that, as BofA essentially maintains, the fact that collusive practices relate to the same product and the same activity, the fact that the same means of communication are used, and the fact that the same individuals are involved are not, taken in isolation, sufficient to demonstrate the existence of a common plan in pursuit of a single anticompetitive aim and, therefore, of a single infringement.
1100However, assuming that they are established, those objective elements are relevant to assessing the existence of a single infringement and, taken as a whole, are capable of confirming that the CODS & CHIPS and DBAC chatrooms were linked and complementary in nature and sought to achieve the aims pursued by the common plan found by the Commission (see paragraph 925 above).
1101Second, inasmuch as, by way of its arguments, BofA maintains that certain objective elements referred to in the contested decision are contradicted by the deliberate exclusion of its trader, those arguments must be rejected for the reasons set out in paragraphs 1058 to 1097 above.
1102Finally, the fact that certain events taken into consideration by the Commission in finding that there was a single infringement took place after the period of a particular bank’s participation in the infringement cannot call that finding into question (see paragraph 1088 above).
1103First, BofA argues that the EGBs covered by the conduct alleged against the banks concerned are not a homogeneous product since they have many different characteristics. Moreover, in BofA’s submission, the Commission has in no way proved that the exchange of information relating to one EGB may be relevant in the context of the acquisition or trading of another EGB, also taking into account the different positions and different investment portfolios of the various traders.
1104However, on the one hand, BofA does not dispute that various types of EGB could have been the subject of exchanges of information between the traders in the CODS & CHIPS and DBAC chatrooms throughout the existence thereof, be they taken together or in isolation. BofA also does not dispute that those traders – and, therefore, those banks – were in a position to acquire or trade various types of EGB, without any particular distinction made between them relating, in particular, to the issuer or the maturity of a given bond (see recitals 416 and 466 of the contested decision).
1105On the other hand, in so far as BofA argues that the Commission failed to prove that the exchange of information on one EGB could be relevant in the context of the acquisition or trading of another EGB, it should be noted that that bank does not specifically dispute the explanation given in that connection by the Commission in recitals 45 to 50 of the contested decision and, in particular, in recital 47 of that decision.
1106Moreover, on that point, several discussions examined in the contested decision show that the traders aimed to influence the price of certain EGBs in order to secure favourable conditions when acquiring other EGBs. For example, in recital 254 of the contested decision, the Commission explained that, in the discussion of 19 June 2009 in the DBAC chatroom, the traders agreed to buy a French benchmark EGB maturing in 2038 in order to attempt to push up the price on the secondary market and thus drive up the issue price of a bond issued by the same Member State, but maturing in 2041. Similar conduct may be seen, for instance, over the course of the discussion of 7 July 2009 in the DBAC chatroom, referred to in recital 259 of the contested decision, regarding an Italian EGB. Furthermore, other discussions show that exchanges of information on one EGB could be relevant in the context of the acquisition or trading of another EGB (see paragraphs 967 to 969 above).
1107Second, BofA confines itself to disputing the relevance of the identical nature of the means of communication used to demonstrate the existence of a single infringement. BofA therefore does not dispute the claim that the CODS & CHIPS and DBAC chatrooms constituted means of communication that were identical in their operation, or that the traders used those same means of communication, except in so far as the BofA trader did not use the DBAC chatroom or bilateral modes of communication.
1108Moreover, BofA does not direct any specific argument at the Commission’s finding that participation in the CODS & CHIPS chatroom was by invitation only. That finding is supported, moreover, in particular by the discussion of 29 January 2007, referred to in recital 104 of the contested decision, and by the discussion of 13 December 2007, referred to in recital 134 of that decision, in so far as concerns invitations into the CODS & CHIPS chatroom, and by the discussion of 9 September 2010, referred to in recital 294 of the contested decision, in so far as concerns invitations into the DBAC chatroom.
1109In that connection, the discussions of 13 December 2007, which took place in the DBAC and CODS & CHIPS chatrooms and concerned the possible invitation of two new traders into the CODS & CHIPS chatroom, show that even after the creation of the DBAC chatroom, invitation into the CODS & CHIPS chatroom was based on trust (‘can’t be trusted’; ‘i dont trust him’).
1110Third, as regards the individuals involved, it must be pointed out that, in so far as BofA’s line of argument is based on the fact that its trader was deliberately excluded from the circle of trust established by the DBAC chatroom, that line of argument must be rejected on the grounds set out in paragraphs 1058 to 1097 above.
1111As to the remainder, it should be borne in mind that the members of the CODS & CHIPS and DBAC chatrooms were not exactly the same. However, the four founder members of the DBAC chatroom were also members of the CODS & CHIPS chatroom.
1112Moreover, as the Commission observes, some traders changed banks but remained members of those chatrooms. Such a change of employer was swiftly followed by the granting of fresh access to those chatrooms. As is apparent from the discussions examined by the Commission in the contested decision, that phenomenon was observed in respect of two traders, who were invited back into the two chatrooms in question and, in respect of another trader, who was again invited into the CODS & CHIPS chatroom. In that regard, it should be noted that one of the founder members of the DBAC chatroom was twice invited back, not only into the latter chatroom, but also into the CODS & CHIPS chatroom, on 3 March 2008 then on 19 October 2009. Another founder member of the DBAC chatroom was invited back into both chatrooms on 18 January 2011.
1113Thus the Commission rightly found that the individuals involved were usually the same, showing a high degree of continuity and overlap. Moreover, it is clear from paragraph 1112 above that the founder members of the DBAC chatroom still wanted to be members of the CODS & CHIPS chatroom, including after the creation of the DBAC chatroom and after the facts set out in paragraphs 1054 to 1056.
1114Fourth, BofA acknowledges that the fact that the anticompetitive discussions at issue all took place during the same period may be relevant to assessing the existence of a single and continuous infringement. However, BofA relies on the fact that its trader was deliberately excluded from the DBAC chatroom and the fact that the discussions in the DBAC chatroom ended several months before the end of the infringement at issue in the CODS & CHIPS chatroom.
1115In that connection, the fact, claimed by BofA, that the discussions in the DBAC chatroom ended on 8 July 2011 – that is, several months before the end of the infringement at issue in the CODS & CHIPS chatroom – is not sufficient to call into question the existence of a temporal overlap in the existence of those two chatrooms and in the anticompetitive conduct put into effect in those two chatrooms.
1116In fact, as is apparent from recitals 82 to 86 of the contested decision, the CODS & CHIPS then the DBAC chatrooms were created in the space of a few weeks by the same four traders in early 2007. Furthermore, despite the facts set out in paragraphs 1054 and 1056 above, anticompetitive exchanges took place in those two chatrooms over a period of more than four years, from 2007 to 2011. Those exchanges repeatedly concerned the same topic during that period (see paragraphs 1064 to 1071 and paragraphs 1080 to 1086 above).
1117Fifth, BofA takes the view that the collusive practices did not follow the ‘same pattern’. First, the discussions in the DBAC chatroom were every day or every other day. By contrast, there were no frequent discussions in the CODS & CHIPS chatroom. Second, unlike the discussions observed in the DBAC chatroom, there was no consistent pattern in the different discussions that took place in the CODS & CHIPS chatroom, and the Commission does not identify any daily discussions or discussions on consecutive days in that chatroom in advance of an auction. Finally, the fact that the practices did not follow the same pattern in the two chatrooms at issue is supported, in BofA’s submission, by a leniency statement made by Natixis.
1118As regards the argument relating to the frequency of the discussions in the two chatrooms or to the lack of frequency of the discussions in the CODS & CHIPS chatroom, it should be observed that that frequency is not the essential aspect of the objective element in connection with the pattern of collusive practices, set out in recital 420 of the contested decision. In fact, the relevant aspect, as set out in that recital, is essentially the fact, in particular, that the traders had discussions before bonds were issued on the primary market, that they attempted to influence bond prices on the secondary market, and that they gave each other feedback subsequent to their conduct, in respect of various types of EGB. The Commission also observed that, throughout, sensitive information was exchanged allowing the traders concerned to look for potential ways to make gains on the secondary market.
1119It is indeed apparent from the discussions examined by the Commission in the contested decision that, as the latter states in recital 420 of that decision, the discussions in the CODS & CHIPS chatroom gave rise, just as the discussions in the DBAC chatroom did, to exchanges of information in advance and an alignment of strategies in the run-up to an auction, as well as feedback thereafter, particularly in the form of congratulations, including after the facts set out in paragraphs 1054 to 1056 above. Similarly, the discussions in the CODS & CHIPS chatroom show that, just as in the DBAC chatroom, sensitive information was exchanged therein throughout, which allowed the traders to look for potential ways to make gains on the secondary market.
1120Furthermore, BofA’s claim that the Commission does not identify a single instance of discussions in the CODS & CHIPS chatroom on consecutive days in advance of an upcoming auction cannot call into question the Commission’s assessment that the practices at issue followed the same pattern. In any event, as the Commission points out, the discussions that took place in the CODS & CHIPS chatroom on 9, 10 and 11 January 2008, referred to in recitals 143 to 146 of the contested decision, attest to the existence, in that chatroom, of discussions on consecutive days in advance of an upcoming auction.
1121Moreover, the statements made by Natixis, as a leniency applicant, do not support BofA’s claim that the 11 auctions that took place in 2008 were preceded by intense exchanges in the DBAC chatroom and not in the CODS & CHIPS chatroom. That claim is in fact based on the taking into account of numerous bilateral exchanges that took place outside the chatrooms and, therefore, also outside the DBAC chatroom.
1122Sixth, BofA maintains that, in the CODS & CHIPS chatroom, the discussions during the period of its participation in the infringement at issue were relatively infrequent, and the great majority of those discussions took place in the DBAC chatroom.
1123In that connection, it is true that there are sometimes gaps of several weeks or even several months between the discussions in the CODS & CHIPS chatroom.
1124However, the discussions produced by the Commission demonstrate that, over several other periods, concomitant or otherwise to the infringement period relied upon against BofA, the frequency of the discussions in the CODS & CHIPS chatroom was also high, for instance in January and March 2008, October 2009, February 2010 and April 2011, that is, including after the facts set out in paragraphs 1054 to 1056 above. Moreover, despite a transitory drop in the frequency of discussions after March 2008, numerous discussions took place in the CODS & CHIPS chatroom after that date.
1125In any event, the issue of the high frequency of the discussions is only one of the objective elements put forward by the Commission in order to find there to be a single anticompetitive aim.
1126Seventh, in so far as BofA disputes the assessment, set out in recital 423 of the contested decision, that the anticompetitive exchanges were often similar in type, that they were intertwined and overlapped, and that they took place in both chatrooms, reference should be made to paragraphs 1064 to 1071 and paragraphs 1080 to 1086 above.
1127It follows that, despite the fact that the BofA trader was not a member of the DBAC chatroom, and despite the views expressed by some of the founder members of the DBAC chatroom referred to in paragraphs 1054 to 1056 above, the existence of an overall plan common to the two chatrooms in question is confirmed by the fact that (i) the anticompetitive conduct put into effect in those two chatrooms related to all types of EGB; (ii) that conduct was put into effect by way of the same means of communication; (iii) the individuals involved were generally the same; (iv) the conduct put into effect in the two chatrooms in question took place over period that broadly overlapped; (v) that conduct followed the same pattern; (vi) that conduct was, depending on the time period, frequent; and (vii) the anticompetitive exchanges were often similar in type, intertwined and overlapped, and took place in both chatrooms.
1128BofA has therefore failed to show that the Commission made an error of assessment when it took the view that the objective elements that it had found also confirmed that the chatrooms in question were linked and complementary in achieving the aims pursued.
1129That finding is not called into question by the fact that, in its leniency statement of 29 June 2016, UBS explains that the most relevant discussion took place primarily between the four founder members of the CODS & CHIPS chatroom, who then founded the DBAC chatroom, namely the UBS trader, a trader from RBS and two traders who were initially working for ABN-AMRO (now RBS).
1130In fact, first of all, the four traders in question were members of both the CODS & CHIPS and the DBAC chatrooms. Next, UBS does not specify which chatroom that ‘most relevant’ discussion took place in, and that bank provides detailed information on the exchanges that took place in the two chatrooms in question. Finally, the fact that a ‘most relevant’ discussion took place between the four traders in question does not support the suggestion that the exchanges with the other traders were irrelevant. Furthermore, the BofA trader is amongst the participants in the discussions that took place in the CODS & CHIPS chatroom and referred to by UBS in its leniency statement.
(iii)The lack of awareness of the exchanges that took place outside the CODS & CHIPS chatroom
1131As regards the lack of awareness, on the part of the BofA trader, of the exchanges that took place outside the CODS & CHIPS chatroom, and the lack of evidence of discussions between that trader and other traders outside that chatroom, it should be recalled that those arguments are based on subjective elements linked to BofA’s liability.
1132It is clear from the case-law referred to in paragraphs 332 to 338 above that those elements are not relevant in assessing the existence of an overall plan pursuing a common objective.
1133The fact that an undertaking is not aware of an exchange of information between other participants in an infringement and does not know about it is not such as to alter the finding of a single and continuous infringement (see, to that effect, judgment of 24 June 2015, Fresh Del Monte Produce v Commission and Commission v Fresh Del Monte Produce, C‑293/13 P and C‑294/13 P, EU:C:2015:416, paragraph 160).
1134The absence of an overall plan in pursuit of a single anticompetitive aim common to the two chatrooms in question therefore cannot be inferred from the lack of awareness, on the part of BofA, of the existence of the DBAC chatroom or from the lack of evidence of discussions between its trader and other traders outside that chatroom.
1135However, that lack of awareness led the Commission to consider that BofA was liable only for conduct found in the CODS & CHIPS chatroom.
1136Even if, as BofA claims, a particular undertaking’s lack of awareness and lack of intention to contribute to the infringement could be taken into account in assessing the existence of a single anticompetitive aim, that taking into account would not be capable of calling into question the finding that there existed an anticompetitive aim common to the two chatrooms in question.
1137It is apparent from the contested decision that, through its trader, BofA actively and intentionally contributed, in the CODS & CHIPS chatroom, to the four categories of conduct referred to in paragraph 43 above and, thus, to the anticompetitive aim common to the two chatrooms in question. Furthermore, that bank has failed to present material aimed at disputing those aspects of the contested decision that establish that the bank’s conduct was intended to contribute to the infringement at issue as a whole, or its intention to contribute, at the very least, to the common objective, even if that were reduced to covering only the anticompetitive conduct of the participants in the CODS & CHIPS chatroom.…
(3)The second plea relied upon by UniCredit alleging, in essence, that the single infringement was focused on the primary market
…
1163Moreover, UniCredit’s line of argument cannot call into question the Commission’s assessments that the overall plan that it found covered both the primary and the secondary markets.
1164The second plea raised by UniCredit must therefore be rejected in so far as it concerns the existence of a single infringement.
(5)Conclusion on the existence of a single infringement
1165In the light of the foregoing, the fourth plea raised by Nomura and the fourth plea raised by Portigon must be rejected in so far as they concern the finding of the existence of a single infringement.
1166Moreover, the first plea raised by BofA must be rejected in its entirety.
1167Furthermore, the second plea raised by UniCredit must be rejected in so far as it concerns the existence of a single infringement.
(d)The objections raised by Nomura and Portigon to the continuous nature of the infringement
1168In recital 420 of the contested decision, the Commission set out the reasons why it regarded the anticompetitive discussions as following the same pattern. In doing so, it stated that the traders were in frequent contact. Moreover, in recital 422 of that decision, it pointed out that the frequency of the discussions was high. Finally, in recitals 449 to 464 of that decision, the Commission responded to certain arguments put forward by Nomura and BofA. In its reply, it recalls, in particular, that (i) there were numerous regular contacts in various chatrooms; (ii) all of the discussions referred to in Section 4 of the contested decision, and in Annex 1 thereto, were part of the body of evidence that supported its assessment that the infringement was continuous; and (iii) in the context of an overall plan extending over several years, a gap of a couple of months between the manifestations of the cartel was immaterial.
1169As a preliminary point, it should be emphasised that even if the manifestations of the cartel are separated by more or less long periods of time, the Commission may, in the context of the assessment of the continuous nature of an infringement extending over a number of years, assume that the infringement has not been interrupted, first, where the various actions which form part of the infringement pursue a single purpose and are capable of falling within the framework of a single and continuous infringement; and, second, where no indicia or evidence establishes that the infringement has not been pursued during those periods (see, to that effect, judgments of 18 March 2021, Pometon v Commission, C‑440/19 P, EU:C:2021:214, paragraphs 112 and 114, and of 17 May 2013, Trelleborg Industrie and Trelleborg v Commission, T‑147/09 and T‑148/09, EU:T:2013:259, paragraphs 61 and 63).
1170The principle of legal certainty, however, requires that the Commission should adduce at least evidence of facts sufficiently proximate in time to be reasonable to accept that the infringement continued uninterruptedly between two specific dates (see, to that effect, judgment of 29 September 2021, Nichicon Corporation v Commission, T‑342/18, EU:T:2021:635, paragraph 365 (not published) and the case-law cited).
1171Although the period separating two manifestations of infringing conduct is a relevant criterion in order to establish the continuous nature of an infringement, the fact remains that the question whether or not that period is long enough to constitute an interruption of the infringement cannot be examined in the abstract. On the contrary, it needs to be assessed in the context of the functioning of the cartel in question (see judgment of 29 September 2021, Nichicon Corporation v Commission, T‑342/18, EU:T:2021:635, paragraph 366 (not published) and the case-law cited).
1172It is in the light of these considerations that it is necessary to examine, in turn, the arguments put forward by Nomura and Portigon in their respective fourth pleas alleging, in essence, that the Commission did not properly take account of circumstances that constituted interruptions undermining the classification as a continuous infringement, at the very least during the period in the course of which, in the Commission’s view, Nomura and Portigon participated in the infringement at issue.
(1)Nomura’s arguments
1173Nomura claims that the infringement at issue is not a continuous infringement. In its view, it may only be a repeat infringement.
1174First of all, according to Nomura, economic studies suggest that the price effects of an auction are limited to a period of five trading days at most and that the effects of an anticompetitive discussion, admitting that such effects are possible, are even shorter given the volatility observed in prices.
1175Second, given the volatile and dynamic nature of the EGB sector, the functioning of the infringement at issue, assuming that it exists, depended on special ‘positive measures’. Thus, the traders needed to have regularly communicated to each other updated and current sensitive information concerning, for example, their proposed bids in a given auction, since the relevance of that information would expire rapidly. The Commission has failed to put forward any evidence of any form of agreement whereby the traders were required to exchange information as soon as they obtained it and were punished if they failed to do so.
1176Third, the Commission cannot therefore simply presume that the infringement at issue was continuous in periods in which there is an absence of evidence of anticompetitive conduct. In the present case, the Commission has improperly reversed the burden of proof. The Commission must prove not only the existence of an infringement but also its duration, including whether it is continuous or repeated in nature. There are several periods in respect of which the Commission has adduced no evidence of positive measures.
1177Fourth, Nomura relies on four gaps separating the manifestations of the infringement found during the infringement period relied upon in respect of Nomura. Nomura relies, in essence, on a gap of 45 days between the discussions of 18 January and 3 March 2011; a gap of 32 days between the discussions of 3 March and 5 April 2011; a gap of 20 days between the discussions of 1 and 22 June 2011; and a gap of 24 days between 3 and 28 November 2011. Again as regards the gaps between the discussions of 1 and 22 June 2011 and between the discussions of 3 and 28 November 2011, Nomura adds that there were no anticompetitive discussions despite auctions organised in particular by the same debt management office between those dates.
1178In that connection, in the first place, it is clear in particular from paragraphs 933 to 1041 above that, over the course of the infringement period as a whole, the conduct of the banks concerned was part of an overall plan in pursuit of a single anticompetitive aim. Thus, in so far as it had at its disposal sufficient evidence to find there to be an overall plan in pursuit of a single anticompetitive aim between 4 January 2007 and 28 November 2011, it was not necessary for the Commission, contrary to what Nomura suggests, to establish that the participants had explicitly and in advance subscribed into an overall agreement that defined their actions on the market or, more generally, that they had formalised in advance their intention to engage in anticompetitive conduct.
1179It follows that the Commission could consider that the infringement at issue had been continuous throughout the entire infringement period set out in the contested decision, unless it were to be found that the infringement had been interrupted (see paragraphs 1169 to 1171 above).
1180In the second place, Nomura’s claim that the price effects of an auction are limited to a window of at most five trading days, or narrower, is based on a reference to that bank’s reply to the Statement of Objections, which in turn refers to a list of publications on the subject without further elaboration as to the content of those publications. Moreover, Nomura’s reply to the Statement of Objections addresses only the effects of an auction on EGB prices on the secondary market. Thus, as the Commission observes, Nomura’s claim does not take account of the effect of the cartel in an initial auction on subsequent auctions of the same EGBs, or the effect of the cartel on so-called reference bonds on other EGBs that have similar characteristics (see paragraphs 964 to 970 above).
1181Furthermore, Nomura’s argument that the chatroom participants had many other opportunities to collude with each other is unconnected to any specific item of evidence.
1182Irrespective of the merits of Nomura’s claims referred to in paragraphs 1180 and 1181 above, first, it should be borne in mind that, as is apparent from recital 414 of the contested decision, the anticompetitive aim pursued in the present case was to secure the desired allocations of EGBs from the auctions and maintain the primary dealer status of the banks concerned at the lowest price. The traders aimed to support or at least respect each other’s positions and strategies when trading EGBs on the secondary market and, if possible, align with each other. The banks’ conduct on the primary and secondary market ultimately aimed to assist each other in generating more EGB trading business and increasing revenues.
1183As the Commission essentially explained in recitals 377, 378, 414, 420, 510, 537 and 540 of the contested decision, and as is apparent from the discussions examined by the Commission in that decision, the discussions between the traders concerned aimed to identify, create and take advantage of opportunities to coordinate their conduct in so far as concerned specific auctions, syndications and trading activities in the EGB sector.
1184The traders of the banks concerned pursued the anticompetitive aim identified by the Commission in the contested decision by freely exchanging information on their activities linked to upcoming, ongoing or recently completed issues of EGBs, or on their activities linked to EGB trading that was ongoing or had just occurred. As the Commission states in recital 764 of the contested decision, the discussions in question took place in chatrooms that, inter alia, allowed participating traders to follow the exchanges in real time. The use of those chatrooms thus afforded the participating traders the possibility of exchanging commercially sensitive information and identifying opportunities to coordinate with each other when the occasion arose and, in particular, when they expressed a shared interest in one or more specific EGBs on the primary or secondary markets. That is how those traders sought to achieve that anticompetitive aim in the context of anticompetitive exchanges on specific bond issues or trades or even on information that remained relevant for specific periods.
1185In those circumstances, daily interaction and very frequent participation in anticompetitive discussions was not necessary to the functioning of the infringement at issue, since the issuing and the trading of EGBs were not necessarily connected. Even though anticompetitive conduct pertaining to an EGB could have an effect on another EGB, and even though conduct on the primary market could have an effect on the secondary market and vice versa, any success of an anticompetitive discussion on a specific syndication, auction or trade was not contingent on participation in a series of discussions relating to other syndications, auctions or trades.
1186Consequently, even if, as Nomura claims, the participants in the chatrooms in question did have numerous other opportunities to collude that they did not take, that finding is not capable of calling into question the Commission’s assessment that the single infringement was continuous.
1187Second, as the Commission essentially explained in recitals 391, 417, 443, 596, 764 and 798 of the contested decision, the exchanges of information in the chatrooms in question implied an expectation that all participants disclose sensitive information, allowing all participants mutually to benefit from those exchanges. That is why, as certain discussions analysed by the Commission show, when a trader did not meet the generally held expectation of mutual exchange, he ran the risk of being excluded from those chatrooms (see paragraphs 490 to 492 above, on the analysis of the discussions not specifically disputed). As the Commission emphasised in recital 799 of the contested decision, the exchanges in those chatrooms were based on mutual trust and on the expectation of receiving similar sensitive information over time. This is why, in particular, the Commission essentially stated in recitals 482 and 798 of the contested decision that the interaction of one trader – the BofA trader or the UniCredit trader – had reinforced the mutual interdependence between the traders in sustaining their collusive behaviour.
1188Consequently, the anticompetitive aim pursued by the traders was based on their shared understanding that they helped each other in order to benefit mutually over time. On that basis, a trader exchanged commercially sensitive information with his competitors because he expected that the other traders would do the same in future. In the context of the functioning of the cartel at issue, the conduct identified by the Commission and, more generally, the overall plan of which that conduct formed part, were therefore based on reciprocal expectations that were fuelled by each anticompetitive discussion; the aim thereof was to gain a lasting advantage vis-à-vis the debt management offices of the Eurozone Member States, competing traders and their clients and, ultimately, to increase the revenues made by the participants when EGBs were issued or traded on the primary and secondary markets.
1189In those circumstances, the brevity of the time windows in which the information exchanged became obsolete, on which Nomura relies, did not require that the Commission explain how the gaps between two manifestations of the infringement had been filled. Moreover, that brevity cannot call into question the Commission’s finding that the single infringement at issue was continuous.
1190In the third place, the Commission relied on evidence of facts sufficiently proximate in time, and it was justified in considering that the gaps between two manifestations of the infringement did not support a finding that that single infringement had been interrupted.
1191First, it should be pointed out that the single infringement at issue spanned several years, namely from early January 2007 to the end of November 2011. Moreover, the Commission was justified in finding, in recitals 420 and 422 of the contested decision, that the traders in question had frequent discussions over the course of the infringement period, and that the frequency of those discussions was high. It should be borne in mind that the Commission identified approximately 380 anticompetitive discussions over the course of the infringement period between 2007 and 2011. Furthermore, the frequency of the anticompetitive conduct admittedly varied over the course of the infringement period as a whole. However, that frequency remained high in each of the years in question. In addition, that variation in the frequency of the anticompetitive conduct is explained by the context of the functioning of the single infringement, which was linked to the existence of opportunities for coordination, and by the context of the market for EGBs, which was marked by more difficult periods, as the discussion of 4 October 2011 shows (see paragraphs 746 to 749 above).
1192Second, it is apparent from an examination of, on the one hand, the discussions that preceded the gaps on which Nomura relies and, on the other hand, the discussions that followed those gaps, that those discussions contain nothing to show that the traders of the banks concerned had expressed the willingness, first, to bring the single infringement to an end or, second, to put it into effect once again. On the contrary, the tone used by the traders is particularly free and easy, and the discussions are spontaneous. In particular, the anticompetitive discussions that follow the gaps on which Nomura relies do not contain any introductory statement that might attest to the need to re-establish contact or even trust between the participants which may previously have been broken off. Those discussions thus establish an enduring and constant willingness, on the part of the traders in question, to engage in conduct that contributed to the overall plan in pursuit of a single anticompetitive aim, as defined by the Commission in the contested decision.
1193More specifically, as regards the alleged gap of 45 days separating, on the one hand, the discussions of 18 January 2011 in the CODS & CHIPS and DBAC chatrooms and, on the other hand, the discussion of 3 March 2011 in the DBAC chatroom, that gap is not capable of establishing that the single infringement was interrupted. It should be observed, in fact, that the Commission found there to be anticompetitive discussions between those two dates, on 25 and 26 January 2011, and on 3, 14, 17, 25 and 28 February 2011. The 45-day gap on which Nomura relies therefore does not correspond to a period in which there was no manifestation of the single infringement. Moreover, two other discussions, which took place on 2 and 11 February 2011, are also mentioned in the contested decision. On 2 February 2011, the participants repeatedly renamed the CODS & CHIPS chatroom and, on 11 February 2011, an RBS trader copied into the CODS & CHIPS chatroom a conversation that he had had in the DBAC chatroom.
1194In reality, the gap on which Nomura relies is based on the fact that its trader was not authorised to act on behalf of that bank during that period and is therefore, at the very most, capable of establishing that that bank’s individual participation in the infringement at issue began at a later date than that relied upon in the contested decision.
1195As regards the gap of 32 days separating the discussion of 3 March 2011 in the DBAC chatroom and the discussion of 5 April 2011 in the CODS & CHIPS chatroom, it should be observed that, while those discussions concerned different bonds, the content thereof was similar.
1196The traders in question exchanged commercially sensitive information on upcoming auctions. Moreover, as is clear from paragraphs 665 to 670 above, during the discussion of 5 April 2011, the traders also exchanged information concerning their trading activities on the secondary market.
1197The transcript of the discussion of 5 April 2011 shows an atmosphere of frank camaraderie between the traders and does not reveal anything to suggest that the willingness of the traders to put the infringement at issue into effect ceased between 3 March and 5 April 2011.
1198As regards the gap of 20 days separating the discussion of 1 June 2011 in the CODS & CHIPS chatroom and the discussion of 22 June 2011 in the same chatroom, it should be observed that, in so far as the discussion of 22 June 2011 could not be relied upon against Nomura (see paragraphs 694 to 704 above), the discussion of 24 June 2011 is the first relevant discussion to follow the discussion of 1 June 2011.
1199In that connection, over the course of the discussion of 1 June 2011, the participants discussed their bidding strategy with regard to a French auction, including overbidding and mid-prices. Furthermore, they also discussed the result of that auction. The same day, a second Nomura trader was invited back into the CODS & CHIPS chatroom and, on that occasion, that trader said ‘NOW MY LIFE [IS] COMPLETE AGAIN’ and promised that ‘anything said on this chat stays with me’. At the end of that discussion, the UBS trader stated that he would be back the following Tuesday.
1200The discussion of 24 June 2011 began with the usual greetings, which were immediately followed by a question put by the Nomura trader, who had been away the day before, in order to find out whether the previous day’s activities had gone well. The participants then exchanged sensitive information on their trading activities on the secondary market, checked their mid-prices, and the Nomura trader suggested that they sell French EGBs so as to move the market.
1201The discussions of 1 and 24 June 2011 therefore demonstrate a shared and enduring willingness to continue to put into effect the overall plan defined by the Commission in the contested decision. This is all the more the case since, between those two discussions, namely on 10 June 2011, another RBS trader was invited into the CODS & CHIPS chatroom, in addition to the access that he already had to the DBAC chatroom.
1202As regards the gap of 24 days separating the discussion of 3 November 2011 in the CODS & CHIPS chatroom and the discussion of 28 November 2011 in the same chatroom, it should be borne in mind that the content of the discussion of 3 November 2011 attests to anticompetitive conduct that took the form of the disclosure of sensitive information on the level of overbidding in an upcoming auction (see paragraphs 781 to 790 above). It is true that, in that discussion, the Nomura trader commented that ‘i don’t think these things should be discussed’. However, that discussion followed the normal course and, at the end of the discussion, the UBS trader announced his plans for that evening.
1203As to the discussion of 28 November 2011, the beginning of that discussion does not reveal anything to suggest that the infringement at issue had previously been interrupted. The beginning of that discussion shows that the traders very quickly referred to the situation of the Kingdom of Belgium which, according to the UniCredit trader, would make that Member State’s upcoming auction interesting. Thereafter, the UBS trader disclosed sensitive information on his level of overbidding in that auction (see paragraphs 791 to 797 above). It is later in the day that the RBS trader announced that he had to leave the chatroom on compliance grounds, and it is in that context that the Commission found that the discussion of 28 November 2011 marked the end of the single infringement at issue.
1204It should be added that, despite the fact that the discussion of 18 May 2011 cannot be relied upon against Nomura (see paragraphs 694 to 704 above), the gap between, on the one hand, the discussion of 5 May 2011 in the CODS & CHIPS chatroom and, on the other hand, the discussion of 1 June 2011 in the same chatroom, does not demonstrate that the single infringement found by the Commission was interrupted.
1205In fact, during those discussions, the traders in question discussed, inter alia, their levels of overbidding and their mid-prices in the context of upcoming auctions. In that regard, the discussion of 5 May 2011 contains no indication that might demonstrate that the traders envisaged putting an end to their respective conduct in future, that is, after that discussion. During that discussion, furthermore, the participants expressed their satisfaction, following their anticompetitive conduct, when they saw the results of the auction in question. As regards the discussion of 1 June 2011, this began, immediately following the usual greetings, with a question put by the UBS trader to the RBS trader in order to find out whether the latter would be taking part in that day’s French auction. Anticompetitive conduct and, in particular, exchanges of sensitive information linked to that auction then ensued (see paragraph 1199 above and paragraphs 1383 and 1384 below).
1206Furthermore, as regards the finding that the Commission has failed to show that the discussion of 4 October 2011 was anticompetitive, it should be pointed out that such a finding cannot demonstrate that the single infringement was interrupted between 28 September and 12 October 2011.
1207In fact, first, it should be recalled that the discussion of 4 October 2011 provides relevant context because it confirms the spirit of cooperation between the traders in a difficult market and demonstrates that those traders stated that they could not help each other due to the difficult market (see paragraphs 746 to 749 above).
1208Second, an examination of the content of the discussions of 28 September 2011 and of 12 October 2011 shows that these do not mention any element suggesting that the willingness on the part of the traders in question to contribute to the infringement at issue ceased between those two manifestations of that infringement.…
(e)Whether liability for the single and continuous infringement can be attributed to UniCredit, Nomura and Portigon
…
1230In its second plea and part of its third plea, UniCredit maintains that the Commission made errors in so far as concerns its participation in the single and continuous infringement at issue. That bank also relies on a sixth plea in law, alleging errors in the determination of the date on which its participation in that infringement began.
1231In its fourth plea, Nomura relies on errors made by the Commission in so far as concerns its participation in the single and continuous infringement at issue. By its third plea in law and the fourth limb of its fifth plea, Nomura also relies on errors relating, in particular, to the duration of its participation in that infringement, and breach of the principle of equal treatment.
1232In its fourth plea, Portigon submits that the Commission made incorrect findings as to the continuous nature of its participation in the infringement at issue and as to the end date of its participation in that infringement. That bank also puts forward arguments suggesting that it disputes its participation in that infringement and its liability more generally.
1233As a preliminary point, it should be stressed that, in so far as UniCredit and Nomura maintain that the Commission was not entitled to set the start date of their participation in the infringement at issue at an earlier date than that on which their trader obtained regulatory authorisation to trade on behalf of those banks, that line of argument must be rejected for the reasons set out in paragraphs 303 to 331 above. The same applies to the line of argument raised by Portigon, alleging that it cannot be held liable for its trader’s conduct.
1234In the light of the foregoing, it is necessary to examine, in the first place, the other arguments put forward by UniCredit in order to dispute the setting of the starting point of its participation in the infringement at issue at 9 September 2011; in the second place, the arguments put forward by UniCredit, Nomura and Portigon in so far as concerns their participation in that infringement and, where relevant, the continuous nature of that participation; in the third place, Portigon’s arguments in connection with the end date of its participation in that infringement; and, in the fourth place, Nomura’s arguments alleging breach of the principle of equal treatment.
(1)The setting of the starting point of UniCredit’s participation in the infringement at issue at 9 September 2011
1235In recital 761 of the contested decision, the Commission stated that, in order to determine the start date of the participation of each of the banks concerned in the infringement at issue, it used as its criterion the date when its trader received first time access to at least one of the two persistent chatrooms, DBAC or CODS & CHIPS, or renewed access where that trader had already been a member of those chatrooms when he worked for another bank. However, in the same recital, the Commission stated that a bank’s individual participation began earlier where if there was evidence that a bank’s trader already participated in other relevant anticompetitive exchanges, in non-persistent chatrooms or by other means, before receiving access to the two persistent chatrooms in question.
1236In recital 762 and Article 1 of the contested decision, the Commission took the view that UniCredit participated in the infringement from 9 September until 28 November 2011. In order to justify setting the starting point of UniCredit’s participation in the infringement at issue, the Commission stated, in recital 769 of that decision, that not only did the UniCredit trader gain renewed access to the CODS & CHIPS chatroom on 9 September 2011, but he accessed that chatroom on 6 September 2011, under the account of his former employer, namely Portigon. The Commission found that such conduct demonstrated the UniCredit trader’s continued interest and participation in the cartel. It adds that that trader used such access to provide informed advice within his bank and/or invested in the relationship with the other competing traders in preparation of his imminent ability to resume trading.
1237By its sixth plea, UniCredit maintains that the Commission erred when it set the starting point of its participation in the infringement at issue at 9 September 2011. First, the period of its participation in the infringement could not have started before it actually participated in an anticompetitive discussion, namely before 26 September 2011. Second, the fact that its trader merely had access to the CODS & CHIPS chatroom on 9 September 2011, without any discussion taking place in that chatroom, cannot be regarded as participation by that trader in an anticompetitive exchange. Third, the fact that its trader was previously an employee of Portigon and as such had access to the CODS & CHIPS chatroom cannot be relied on by the Commission to argue that UniCredit was aware of the overall plan from 9 September 2011. Its trader’s intentions are irrelevant, in UniCredit’s submission, as long as he did not participate in any anticompetitive exchange.
1238First, the Commission contends that the UniCredit trader was amongst the founders of the CODS & CHIPS and DBAC chatrooms in 2007, and that he was fully aware of the purpose of the CODS & CHIPS chatroom, since he was a member of that chatroom when he worked for other banks. Second, given that the UniCredit trader re-joined, in full knowledge of the facts, an agreement the object of which was to restrict and/or distort competition, the Commission submits that it did not need to prove that that trader had participated in a specific anticompetitive exchange on 9 September 2011 in order to establish that bank’s liability as from that date. Third, it is argued that account must be taken of the fact that the trader in question logged into the CODS & CHIPS chatroom on 6 September 2011 under the account of his former employer, which shows his ongoing interest in participating in the cartel at issue. Fourth, the Commission submits that UniCredit’s liability is based not on whether or not the trader had passed on information to his employer, but on the fact that that trader had rejoined the CODS & CHIPS chatroom on 9 September 2011 in full knowledge of the aim of that cartel, which attests to his intention to promote the objectives thereof.
1239In the first place, it should be observed that, in recital 339 of the contested decision, the Commission pointed out that, on 31 August 2011, the Portigon trader left that bank and, on 1 September 2011, he started working for UniCredit. In recital 340 of that decision, the Commission found that, in a discussion of 6 September 2011 in the CODS & CHIPS chatroom, the traders of RBS, UBS and Nomura exchanged information on mid-prices and overbidding ahead of the auction and ascertained what each participant’s bidding strategy was. In recital 341 of that decision, the Commission observed that the Portigon trader now worked for UniCredit, that he entered the CODS & CHIPS chatroom on 6 September 2011, still using the account of his former employer – that is, Portigon – but did not actively participate in that day’s discussion. In the same recital, the Commission stated that, on 9 September 2011, the trader in question was given renewed access to the CODS & CHIPS chatroom under an account linked to his new employer, namely UniCredit. In recital 342 of the contested decision, the Commission found, in essence, that on 26 September 2011, the traders of Nomura, RBS, UBS and UniCredit participated in an anticompetitive discussion.
1240Thus, it is clear from the summary in paragraph 1239 above that the Commission considered that UniCredit had incurred liability on 9 September 2011, that is, the point in time when its trader logged into the CODS & CHIPS chatroom under the account linked to that bank, even though no anticompetitive discussions took place in that chatroom between 6 and 26 September 2011.
1241In the second place, it is apparent from several passages in the contested decision that the creation of, and participation in, chatrooms are not regarded, per se, as proof of the existence of an anticompetitive agreement or concerted practice.
1242In fact, in recitals 392 and 644 of the contested decision, the Commission explained, in essence, that it did not rely, against the banks concerned, on discussions used for social purposes, for exploring bilateral trades or for exchanging market colour that is already in the public domain, even though those discussions had taken place during the infringement period and in both of the chatrooms in question.
1243Thus the anticompetitive nature of the discussions identified by the Commission is linked solely to the subject matter of those discussions and not to the environment in which they took place.
1244In the third place, it is true that a party which tacitly approves of an unlawful initiative, without publicly distancing itself from its content or reporting it to the administrative authorities, effectively encourages the continuation of the infringement and compromises its discovery. That complicity constitutes a passive mode of participation in the infringement which is therefore capable of rendering the undertaking liable in the context of a single infringement (see, to that effect, judgment of 7 January 2004, Aalborg Portland and Others v Commission, C‑204/00 P, C‑205/00 P, C‑211/00 P, C‑213/00 P, C‑217/00 P and C‑219/00 P, EU:C:2004:6, paragraph 84).
1245It is also true that an undertaking may have participated directly in only some of the forms of anticompetitive conduct comprising the single and continuous infringement, but have been aware of all the other unlawful conduct planned or put into effect by the other participants in the cartel in pursuit of the same objectives, or could reasonably have foreseen that conduct and have been prepared to take the risk (see paragraph 335 above).
1246However, first, it is common ground that no anticompetitive discussions took place between 6 and 26 September 2011, which precludes any intention on the part of the traders to conduct themselves in a specific way or to disclose their conduct on the market during that time period.
1247In the absence of any manifestation of the infringement serving to establish the intention on the part of an undertaking to participate in a single and continuous infringement, mere awareness of the infringing conduct envisaged or put into effect by other undertaking in pursuit of the single anticompetitive aim identified, or the mere possibility that an undertaking could reasonably have foreseen such conduct and have been prepared to take the risk, are not sufficient to find that an undertaking started to participate in a single and continuous infringement and, on that basis, to set the starting point of that undertaking’s participation in that infringement.
1248Second, on the one hand, in the absence of any anticompetitive discussions in the CODS & CHIPS chatroom which UniCredit had just joined through its trader on 9 September 2011, that bank cannot be regarded as having tacitly approved the content of an anticompetitive discussion on that date and encouraged the continuation of the infringement as a result.
1249On the other hand, in such a context, UniCredit – which had only just joined the CODS & CHIPS chatroom – could not be required publicly to distance itself from the content of anticompetitive discussion which had not yet taken place, or to report such discussions to the competent administrative authorities.
1250If the Commission’s line of reasoning were to be followed, it would result in allowing it to find that UniCredit participated in a single and continuous infringement even if, thereafter, that undertaking’s trader had not participated in any anticompetitive discussions or the infringement had ceased before any fresh anticompetitive discussions took place.
1251It follows that, in the absence of anticompetitive exchanges that took place on the day on which the access of the trader concerned to the CODS & CHIPS chatroom was renewed on behalf of his new employer, namely UniCredit, that renewed access is not, per se, sufficient to set the starting point of that bank’s participation in the infringement at issue at 9 September 2011, even though, owing to his previous employment relationships with RBS, Natixis and Portigon, that trader had knowledge of the fact that the exchanges that took place in that chatroom could be anticompetitive in nature.
1252That assessment is not called into question by the fact that, on 6 September 2011, the UniCredit trader logged into the CODS & CHIPS chatroom using the still active account of his former employer and was present during an anticompetitive exchange of information without actively participating therein.
1253First of all, it is apparent in particular from recital 769 of the contested decision that the Commission did not rely on the anticompetitive exchange of 6 September 2011 to determine the starting point of the period of UniCredit’s participation in the infringement at issue. Instead, it used the login that took place on that date, and above that of 9 September 2011, to demonstrate the UniCredit trader’s interest in continuing to be part of the cartel, and to set the start date of UniCredit’s participation at 9 September 2011.…
(2)The participation of UniCredit, Nomura and Portigon in the infringement at issue and the continuous nature of that participation
…
(i)UniCredit’s participation in the infringement
…
1269In the second place, in its third plea, UniCredit maintains that the Commission erred or failed to demonstrate to the requisite legal standard that the conduct of the banks concerned was such as to constitute a single and continuous infringement.
1270In particular, the Commission erred in law by basing its demonstration of the existence of a single and continuous infringement on evidence that was not contemporaneous, that is to say prior to the period of UniCredit’s participation in the infringement at issue.
1271Furthermore, the Commission failed to carry out any analysis of conduct limited to UniCredit’s own period of participation and relies on two improper imputations. First, the Commission holds that bank liable for the conduct of a trader who was not authorised to trade and, secondly, it holds it liable for that trader’s conduct and knowledge prior to his arrival at the bank.
1272The summary analysis of the eight discussions involving that bank does not meet the standard of proof required to demonstrate the classification of a single and continuous infringement since the Commission did not seek to examine the pattern of allegedly anticompetitive conduct during UniCredit’s participation in the infringement at issue, namely between 9 September and 28 November 2011. Thus, the duration of that participation corresponds to 3.5% of the total duration of that infringement, and 1.6% of the discussions identified in Annex 1 to the contested decision fall within that period.
1273Further, the period of UniCredit’s participation is a period characterised by a marked decrease in the frequency of discussions between the banks concerned, which the Commission has failed to note.
1274As a preliminary point, first, in so far as UniCredit maintains that the Commission incorrectly holds that bank liable for the conduct of a trader who was not authorised to trade, that argument must be rejected for the reasons set out in paragraphs 303 to 331 above.
1275Second, in so far as UniCredit claims that the Commission found that it became aware of the full scope of the cartel as early as 9 September 2011, it should be borne in mind that, as is clear from the examination of the sixth plea on which UniCredit relies, the Commission wrongly found that the starting point of that bank’s participation in the infringement at issue should be set at 9 September 2011 (see paragraphs 1235 to 1258 above). That starting point could, in fact, only be set at the date of the first anticompetitive discussion in which that bank’s trader participated, that is, 26 September 2011.
1276Third, it should be pointed out that, with the exception of the arguments examined in paragraphs 1143 to 1164 and paragraphs 1235 to 1258 above, alleging that the single infringement was focused on the primary market, UniCredit does not dispute the findings actually made by the Commission in recitals 413 to 424 of the contested decision in relation to the overall plan and the single anticompetitive aim. UniCredit also does not dispute the objective elements set out by the Commission in that regard in recitals 416 to 424 of that decision. Moreover, it should be recalled that, at the hearing, UniCredit stated that it did not dispute the existence of a single and continuous infringement, only its participation in that infringement.
1277Fourth, in so far as UniCredit disputes its participation in the single and continuous infringement on the ground that that infringement was focused on the primary market and that there is no relationship between the primary and secondary markets, reference must be made to paragraphs 1143 to 1164 above.
1278That said, in the first place, UniCredit disputes the fact that potentially sensitive information on the primary market may have served in its trader’s trading activities on the secondary market, or even the sensitive nature of the mid-prices shared by the traders.
1279In that connection, first, it follows from paragraphs 798 to 803 above that, with the exception of the discussion of 4 October 2011, the seven anticompetitive discussions during which the UniCredit trader was present, and in which he participated, gave rise to exchanges of sensitive information that related to both the primary and the secondary markets.
1280More specifically, first, it is apparent from the discussions in which the UniCredit trader participated that, in the course of those discussions, the banks concerned exchanged sensitive information in the context of an upcoming or ongoing auction on the primary market, such as the levels of bidding or overbidding. That was the case during the discussion of 26 September 2011, referred to in recital 342 of the contested decision (‘how much you thinking about overbidding in the Belgium longs?’ ‘about 15 cents ten years’ ‘41s im thinking 25-30 and 28s 22ish’); the discussion of 12 October 2011, referred to in recital 345 of the contested decision (‘what overpaying?’ ‘im just bidding small at avg’ ‘same here I think’ ‘plus 25 then’ ‘“I am paying +25 for part and + 15” “think it comes cheaper”’); the discussion of 2 November 2011, referred to in recital 347 of the contested decision (‘no more overbidding in Germany’ ‘or france tomorrow’ ‘till next time’ ‘agree’); the discussion of 3 November 2011, referred to in recital 348 of the contested decision (‘“what we are overbidding” “15 yrs” “today”’ ‘we re thinking like 8-12 cents’); and the discussion of 28 November 2011, referred to in recital 349 of the contested decision (‘“what u think belg” […] “gonna bid +22 +20 +18” “for 10 yrs” […] “11”’).
1281Second, it is apparent from the discussions in which the UniCredit trader participated that, in the course of those discussions, the participants exchanged sensitive information on the secondary market, such as information on mid-prices. That was the case during the discussion of 26 September 2011, referred to in recital 342 of the contested decision (‘where mids on longs’ ‘i got 99.15’), and the discussion of 2 November 2011, referred to in recital 347 of the contested decision (‘“what we got the curve to start,” “99?”’ ‘99.2’ ‘yeah 99.5 curve tho guess 99 with sell off’).
1282Lastly, it is apparent from the discussions in which the UniCredit trader participated that, in the course of those discussions, the banks concerned exchanged other sensitive information on the secondary market. That was the case during the discussion of 26 September 2011, referred to in recital 342 of the contested decision (‘where do u have rag37 to dbr37 at the moment? 59.75?’ ‘59.5’ ‘ta’); the discussion of 28 September 2011, referred to in recital 343 (‘my bloke said 05?’ ‘“makes sense I think” “I had 00 ish”’ ‘“w should go higher” “no?”’ ‘“yeah I think”’); the discussion of 19 October 2011, referred to in recital 346 (‘my bid dsl 42s for 5m in ICAP’ ‘ok its my offer on tv dont lift it!’); and the discussion of 2 November 2011, referred to in recital 347 of the contested decision (‘buying both btp spain’ ‘“im scrappy long 34s and it’s a royal dog” […] “bid 15m 42s”’).
1283It should be recalled that, as is clear from paragraphs 469 to 486 above, the Commission was entitled to find that (i) information on the primary market was relevant and could confer a competitive advantage on traders on the secondary market, and (ii) information pertaining to the secondary market was relevant and could confer an advantage in the context of activities on the primary market. Moreover, information pertaining to the secondary market, such as mid-prices, could confer a competitive advantage on traders on that market.
1284Thus the UniCredit trader participated in discussions the content and anticompetitive nature of which are proven by contemporaneous evidence of the infringement at issue and confirmed, for the most part, by the leniency applicants. In the course of those discussions, the traders exchanged sensitive information in the context of transactions on the primary market. That information could confer an advantage on UniCredit on the secondary market. Similarly, in the course of those discussions, the traders exchanged sensitive information in the context of activities on the secondary market which gave them an advantage on that market or conferred an advantage on those participants who were also active on the primary market.
1285Second, the UniCredit trader did not remain idle during those discussions, and did not only receive sensitive information when he was in the CODS & CHIPS chatroom. During those discussions, the UniCredit trader in fact actively contributed to the exchanges of information.
1286As is illustrated by the discussions of 26 and 28 September and 2 November 2011, referred to in recitals 342, 343 and 347 of the contested decision and analysed in paragraphs 711 to 737 and paragraphs 762 to 780 above, the UniCredit trader put questions to the other traders that were capable of leading, or indeed led to, the disclosure of sensitive information on the primary and secondary markets (‘what do we think for Belgium today?’ ‘“where do u have rag37 to dbr37 at the moment?” “59.75?”’ ‘you see an opening call on bunds yet?’ ‘“what we got the curve to start?” “99?”’).
1287Moreover, as is apparent from the discussions of 26 and 28 September, 12 and 19 October, and 2 and 28 November 2011, referred to in recitals 342, 343, 345, 346, 347 and 349 of the contested decision, respectively, and analysed in paragraphs 711 to 737, paragraphs 750 to 780, and paragraphs 791 to 797 above, the UniCredit trader shared sensitive information in his possession (‘i got 99.15’ ‘“makes sense I think” “I had 00 ish”’ ‘im just bidding small at avg’ ‘mid 108.58 I had’ ‘my bid dsl 42s for 5m in ICAP’ ‘92.8’ ‘“im scrappy long 34s and it’s a royal dog” “bid 15m 42s”’).
1288Furthermore, it is apparent from the discussions of 28 September, 19 October and 2 November 2011, referred to in recitals 343, 346 and 347 of the contested decision, respectively, that the UniCredit trader responded to requests from other participants, inter alia, to engage in specific conduct (‘yeah I think’ ‘ok’ ‘was tempted to lift u just then -)’ ‘til next time…’).
1289It follows that the Commission has demonstrated to the requisite legal standard that the UniCredit trader intended to contribute, and indeed did contribute, actively and intentionally, to the single anticompetitive aim of the infringement at issue on both the primary market and the secondary market.
1290It should be added that, given that the single and continuous infringement at issue covered both the primary and secondary markets and that these two markets are interrelated, UniCredit cannot be regarded as an undertaking which is not in competition with the other participants on the ground that it was active solely on the secondary market.
1291Although UniCredit was active solely on the secondary market, the Commission was therefore entitled to find that (i) that bank had subscribed to, and participated in, the overall plan in pursuit of a single anticompetitive aim put into effect by all of the banks concerned, merely on account of its presence during the anticompetitive exchanges examined in paragraphs 711 to 797 above; (ii) its conduct had distorted competition and (iii) it had taken into account the information exchanged during those discussions.
1292In the second place, in so far as UniCredit disputes its awareness of the infringing conduct beyond 26 September 2011 – which is not clear from the application – it should be recalled that, with the exception of the discussion of 4 October 2011, UniCredit has failed to demonstrate that the Commission erred in its assessment when it found that that bank had participated in anticompetitive discussions that came under the four categories of conduct that the Commission had identified for analytical purposes, namely categories that covered anticompetitive conduct on the primary and secondary markets.
1293The Commission could therefore conclude that UniCredit was aware of all of the infringing conduct envisaged or put into effect by the other banks concerned, or could reasonably have foreseen such conduct and have been prepared to take the risk.
1294This is all the more the case since, according to settled case-law, the Commission may rely on contacts prior to, or subsequent to, the period of the infringement in order to construct an overall picture and to show the preparatory stages of the cartel as well as to corroborate the interpretation of certain evidence (see, to that effect, judgments of 8 July 2008, Lafarge v Commission, T‑54/03, not published, EU:T:2008:255, paragraph 428, of 2 February 2012, Denki Kagaku Kogyo and Denka Chemicals v Commission, T‑83/08, not published, EU:T:2012:48, paragraph 188, and of 27 June 2012, Coats Holdings v Commission, T‑439/07, EU:T:2012:320, paragraph 60).
1295The Commission was therefore entitled to use the knowledge acquired by an employee prior to his or her arrival in the service of a new company and which he or she in fact makes available to that new employer when that knowledge corroborates other elements available to the Commission.
1296In the present case, before taking up his duties with UniCredit, that bank’s trader had, as a trader with RBS, Natixis then Portigon, directly participated in the anticompetitive conduct that took place in the chatrooms in question and over the course of bilateral discussions. Furthermore, the conduct in which the UniCredit trader engaged in the context of his previous duties came under all of the categories identified by the Commission.
1297In the third place, UniCredit raises no specific objection to the assessments made by the Commission in recitals 443 to 445 of the contested decision, which establish the intentional nature of the contribution of the banks involved to the single anticompetitive aim pursued by the overall plan found in that decision. Furthermore, the anticompetitive conduct in which the UniCredit trader participated took the form of exchanges of sensitive information relevant to the primary and secondary markets, or of coordination of trading activities on the secondary market in a closed chatroom. Moreover, the discussions in which the UniCredit trader participated show that he had the assurance that the information that he shared with the other traders would not be used to his detriment, and that the exchanges of information in which he participated would confer an advantage on him.
1298In the fourth place, in so far as UniCredit relies on an alleged decrease in the frequency of contacts between the banks concerned during its period of participation in the single infringement at issue, that argument must be rejected. In fact, inasmuch as that bank claims that that decrease in frequency equates to one or more interruptions in the continuity of that infringement or in the continuity of its participation in the infringement, suffice it to observe that that bank has failed to substantiate that complaint.
1299Moreover, the finding that the Commission has failed to demonstrate to the requisite legal standard that the discussion of 4 October 2011 was anticompetitive (see paragraphs 738 to 749 above) cannot call into question the continuity of UniCredit’s participation in the single and continuous infringement at issue between 26 September and 28 November 2011. It is in fact clear from paragraphs 716 to 737 and paragraphs 750 to 755 above that UniCredit’s trader participated in anticompetitive exchanges that formed part of the same overall plan and pursued the same anticompetitive aim on 28 September 2011 and then on 12 October 2011, namely just before and just after the exchange of 4 October 2011. Further, the discussion of 4 October 2011 provides context to show that the market for EGBs was marked by more difficult periods (see paragraphs 746 to 749 above). Moreover, there is no evidence or indicia to suggest that the infringement was interrupted in so far as concerns UniCredit.
1300In the fifth place, in so far as UniCredit may seek to diminish the importance of its role in the infringement at issue or the competitive advantage from which it benefitted on the ground that its trader was active only on the secondary market, it should be borne in mind that the overall plan defined by the Commission related to both the primary and secondary markets, and that that bank’s trader had an interest in receiving information on both markets in order to engage in his activities on the secondary market.
1301It should be added that the eventuality that that bank was unable to gain an advantage from the overall plan is, where applicable, to be taken into consideration not when a finding is made as to the liability of an undertaking for its participation in a single and continuous infringement, but only when the gravity of the infringement or the gravity of the participation therein on the part of the undertaking in question is assessed (see, to that effect, judgment of 16 June 2022, Sony Corporation and Sony Electronics v Commission, C‑697/19 P, EU:C:2022:478, paragraph 113 and the case-law cited). In other words, assuming that such would be the argument put forward by UniCredit, it would not be such as to call into question the Commission’s finding as to whether liability for the infringement at issue may be attributed to that bank for the period of its participation.
1302Consequently, the Commission has demonstrated to the requisite legal standard, on the basis of evidence essentially contemporaneous to the facts, that UniCredit participated, from 26 September to 28 November 2011, in the single and continuous infringement found in the contested decision, and that that bank could be held liable for that infringement during that period of time.…
(ii)Portigon’s participation in the infringement at issue and the continuous nature of that participation
…
–The continuous nature of Portigon’s participation in the infringement at issue
…
1369In so far as concerns the gap of 34 days between a discussion of 13 April 2011 and a discussion of 18 May 2011, it should be pointed out that that gap is the result of the fact that the Portigon trader was not present during discussions that took place between the participants in the CODS& CHIPS chatroom on 3 and 5 May 2011. The Portigon trader’s first words in the discussion of 18 May 2011 clearly do not tally with those of a member who previously put an end to his participation in the infringement (‘[market] got to bounce today… German auction’). The remainder of the discussion of 18 May 2011 contains nothing to show that the Portigon trader interrupted his participation in the infringement before that discussion.
1370As regards the other gaps on which Portigon relies, none of the evidence produced before the Court shows that the Portigon trader expressed (i) a willingness to end his participation in the single and continuous infringement at issue, and (ii) a willingness to participate in that infringement once again. In particular, the tone and spontaneous nature of the discussions produced before the Court thus establish a persistent and constant willingness, on the part of the Portigon trader, to continue to engage in conduct that contributed to the overall plan in pursuit of a single anticompetitive aim, as defined by the Commission in the contested decision.
1371Third, Portigon has not put forward any evidence such as to show that the Commission wrongly failed to find that Portigon’s participation in the infringement at issue had been interrupted. First of all, Portigon’s line of argument is exclusively based on the gaps separating two manifestations of the infringement at issue in which it participated. Next, Portigon has failed to produce any evidence to show that it distanced itself from the infringing conduct at issue. Lastly, Portigon has not relied on the existence of any proof or indicia capable of calling into question the continuous nature of its participation in that infringement between two manifestations thereof in which its trader participated.
1372Consequently, having regard, first, to the evidence that the Commission had in its possession on the context of the functioning of the cartel and the gaps that separated two manifestations of the participation of Portigon’s trader in the infringement at issue; second, the absence of public distancing on the part of Portigon; and, third, the absence of other circumstances raised by that bank in order to establish an interruption in its participation in that infringement, the Commission rightly found that Portigon’s participation had not been interrupted between 19 October 2009 and 1 June 2011.
1373As a consequence, the arguments put forward by Portigon alleging limitation must also be rejected.
(3)The end date of Portigon’s participation in the infringement at issue
1374In recital 761 of the contested decision, the Commission considered that the individual participation of each bank concerned ended when its trader was removed from the relevant persistent chatrooms. It adds that that individual participation ended earlier if there was evidence that the trader ceased trading for the bank concerned before that date or otherwise ended its involvement in the anticompetitive discussions.
1375In recital 762 and Article 1 of the contested decision, the Commission took the view that Portigon participated in the infringement at issue from 19 October 2009 until 3 June 2011.
1376In its fourth plea, Portigon essentially disputes the end date of its participation in the infringement at issue, set by the Commission at 3 June 2011.
1377First, that bank maintains that it cannot incur liability beyond the final anticompetitive act by its trader, namely 18 May 2011, even though the employment contract between Portigon and its trader continued beyond that date. In that connection, the discussion that took place in the CODS & CHIPS chatroom on 1 June 2011 does not, in Portigon’s submission, demonstrate its participation in an anticompetitive practice. In the course of that discussion, the statements made by its trader were not anticompetitive, while those made by the other traders did not concern Portigon’s area of activity.
1378Second, contrary to what the Commission states in its defence, the end date of an undertaking’s participation in an infringement does not depend on the duration of the employment contract of the employee who takes part in anticompetitive meetings, but rather from his or her conduct. It is argued that the Commission cannot extend the period of Portigon’s participation beyond the final anticompetitive act by its trader, solely because the employment contract between them ended at a later date.
1379In that connection, first, it should be observed that, in recital 330 of the contested decision, the Commission found that, during a discussion of 1 June 2011 in the CODS & CHIPS chatroom, the traders of RBS, UBS, Nomura and Portigon discussed their bidding strategy, including overbidding and mid-prices for an upcoming French auction. It adds that, thereafter, the traders discussed the result of that auction.
1380In recital 331 of that decision, the Commission emphasised that, in the discussion of 1 June 2011, the Portigon trader announced that he would no longer be active in the chatroom from 3 June 2011 onwards. In that regard, it observed that that trader’s employment contract was due to end on 31 August, but that he would be on leave from 3 June 2011. It adds that, after 3 June 2011, the trader in question continued to have access to the CODS & CHIPS chatroom, but it was unlikely that he traded for Portigon after that date. In recital 334 of the contested decision, the Commission noted that the Portigon trader was still present during a discussion of 24 June 2011 in the CODS & CHIPS chatroom, despite the fact that he was on leave.
1381It is in that context that the Commission found that Portigon had participated in the infringement at issue until 3 June 2011.
1382Second, Portigon does not call into question the anticompetitive nature of the discussion of 1 June 2011, only its participation therein, on the ground that that discussion concerned bonds in which it did not trade, and that the statements made by its trader were not anticompetitive.
1383In that connection, the transcript of the discussion concerned confirms that the Portigon trader was indeed in the CODS & CHIPS chatroom on 1 June 2011.
1384Moreover, the discussion that took place that day was indeed anticompetitive. The participants in the CODS & CHIPS chatroom in fact exchanged sensitive information on their bidding strategy, including overbidding and mid-prices for an upcoming French auction.
1385The Commission was therefore justified in finding that the Portigon trader participated in an anticompetitive discussion on 1 June 2011.…
4.The pleas alleging errors in the characterisation of the infringement at issue as a restriction by object
…
(a)The concept of restriction by object
…
1406It follows that that provision, as interpreted by the Court of Justice, makes a clear distinction between the concept of ‘restriction by object’ and the concept of ‘restriction by effect’, with each of them being subject to different legal and evidentiary rules (see, to that effect, judgments of 20 November 2008, Beef Industry Development Society and Barry Brothers, C‑209/07, EU:C:2008:643, paragraphs 17 and 33; of 11 September 2014, CB v Commission, C‑67/13 P, EU:C:2014:2204, paragraph 51; of 16 July 2015, ING Pensii, C‑172/14, EU:C:2015:484, paragraph 32; and of 21 December 2023, European Superleague Company, C‑333/21, EU:C:2023:1011, paragraph 160).
1407Accordingly, as regards practices characterised as ‘restrictions by object’, there is no need to investigate their effects nor a fortiori to demonstrate their effects on competition in order to characterise them as ‘restrictions of competition’, within the meaning of Article 101(1) TFEU, in so far as experience shows that such behaviour leads to falls in production and price increases, resulting in poor allocation of resources to the detriment, in particular, of consumers. Concerning such practices, all that is required is the demonstration that they can in fact be characterised as ‘restrictions by object’, though mere unsubstantiated allegations are not however sufficient (judgment of 30 January 2020, Generics (UK) and Others, C‑307/18, EU:C:2020:52, paragraphs 64 and 65).
1408On the other hand, where the anticompetitive object of an agreement, a decision by an association of undertakings or a concerted practice is not established, it is necessary to examine its effects in order to prove that that conduct has as its actual or potential effect the prevention, restriction or distortion of competition, which must be appreciable (see, to that effect, judgments of 28 May 1998, Deere v Commission, C‑7/95 P, EU:C:1998:256, paragraph 77; of 30 January 2020, Generics (UK) and Others, C‑307/18, EU:C:2020:52, paragraph 66; and of 21 December 2023, European Superleague Company, C‑333/21, EU:C:2023:1011, paragraph 169).
1409It is also apparent from the case-law of the Court of Justice, recalled in essence in recitals 486 to 488 of the contested decision, that the concept of ‘restriction by object’ must be interpreted strictly and refers exclusively to certain types of coordination between undertakings which reveal a sufficient degree of harm to competition for the view to be taken that it is not necessary to assess their effects (see, to that effect, judgments of 30 June 1966, LTM, 56/65, EU:C:1966:38, p. 249; of 23 January 2018, F. Hoffmann-La Roche and Others, C‑179/16, EU:C:2018:25, paragraph 78; of 30 January 2020, Generics (UK) and Others, C‑307/18, EU:C:2020:52, paragraph 67; and of 21 December 2023, European Superleague Company, C‑333/21, EU:C:2023:1011, paragraphs 161 and 162).
1410Furthermore, in order to determine, in a given case, whether an agreement, decision by an association of undertakings or a concerted practice reveals, by its very nature, a sufficient degree of harm to competition that it may be considered as having as its object, the prevention, restriction or distortion thereof, it is necessary to examine, first, the content of the agreement, decision or practice in question; second, the economic and legal context of which it forms a part; and, third, its objectives (see, to that effect, judgments of 11 September 2014, CB v Commission, C‑67/13 P, EU:C:2014:2204, paragraph 53; of 23 January 2018, F. Hoffmann-La Roche and Others, C‑179/16, EU:C:2018:25, paragraph 79; and of 21 December 2023, European Superleague Company, C‑333/21, EU:C:2023:1011, paragraph 165).
1411When determining the economic and legal context, examined first by the Court of Justice, it is necessary to take into account the nature of the goods or services affected, as well as the real conditions of the functioning and the structure of the market or markets in question (judgments of 11 September 2014, CB v Commission, C‑67/13 P, EU:C:2014:2204, paragraph 53; of 23 January 2018, F. Hoffmann-La Roche and Others, C‑179/16, EU:C:2018:25, paragraph 80; and of 21 December 2023, European Superleague Company, C‑333/21, EU:C:2023:1011, paragraph 166).
1412In that regard, contrary to Nomura’s claim, in the context of the first limb of its first plea, it is not for the Commission to refer to reliable and robust experience, possibly supported by independent economic studies, in order to find that a specific infringement has an anticompetitive object.…
(b)The characterisation of the single and continuous infringement at issue as a restriction by object
(1)Preliminary observations
…
1421In particular, the Commission expressly stated, in recital 496 of the contested decision, that those four categories of conduct were developed for analytical purposes, were intertwined and were partially overlapping, which is also confirmed by recital 533 of that decision.
1422Those findings have consequences not only for the effectiveness and the merits of certain arguments raised by UniCredit, Nomura or Portigon in support of their challenge to the characterisation of the infringement at issue as a ‘restriction by object’, but also for the interpretation to be given to the arguments of those banks and the answers to be given by the Court to those arguments.
(i)The criticisms alleging errors in the classification of the discussions at issue in the context of the characterisation of the infringement at issue as a restriction by object
1423As has been noted in paragraph 95 above, arguments alleging possible errors in classifying a particular discussion in one or other of the four categories referred to in recitals 93, 382 and 496 of the contested decision are ineffective for the purpose of challenging the Commission’s finding that the single and continuous infringement at issue has an anticompetitive object.
1424That is the case, in particular, with UniCredit’s claim that the Commission cannot rely on the discussion of 12 October 2011 against it, on the ground that that discussion was classified in Category 3 – which covers only discussions with regard to the primary market for EGBs – when that bank is active on the secondary market for EGBs only.
1425For the same reasons, Nomura’s criticisms, alleging error in the classification of the discussions of 28 September and 2 November 2011 in Category 1, are also ineffective.
(ii)The calling into question of the constituent elements of the single and continuous infringement at issue, in the context of the characterisation of the infringement at issue as a restriction by object
1426The fact, noted in paragraph 1396 above, that the Commission was entitled to find that all the discussions validly relied upon in particular against UniCredit, Nomura or Portigon, taken together, took the form of a single and continuous infringement means that, in the context of the assessment of the anticompetitive object of that infringement, the banks concerned cannot call into question the constituent elements of the single and continuous infringement previously found.
1427The same applies to the claims that the discussions in a specific category or specific discussions were not anticompetitive or those alleging that there was no ‘common plan’ or ‘overall plan’.
1428At the stage of assessing the anticompetitive object of a single and continuous infringement, not only the anticompetitive nature of that infringement and the elements of which it is composed, but also the existence of a ‘common plan’ or an ‘overall plan’ have now been established.
1429At this stage, therefore, all that matters is whether that infringement previously found reveals a sufficient degree of harm to competition, justifying the Commission’s not having to investigate their effects nor a fortiori to demonstrate their effects on competition (judgment of 30 January 2020, Generics (UK) and Others, C‑307/18, EU:C:2020:52, paragraph 64).
1430Consequently, Nomura’s objections raised in the context of its second plea, relating to the characterisation of the infringement at issue as a ‘restriction by object’, by way of which that bank claims that ‘[t]he Commission erred in its assessment of fact and/or law by classifying such [discussions] as anticompetitive’, must be rejected.
1431Similarly, UniCredit’s argument, raised in its fourth plea, that the Commission failed to establish that ‘the putative “common plan” led to an infringement by object with respect to secondary market trading during [its] period of involvement’ must also be rejected.
(iii)The Commission’s overall assessment of the single and continuous infringement at issue, in the context of the characterisation of that infringement as a restriction by object
1432The fact that the Commission found ‘a single and continuous infringement regarding EGB covering the entire EEA, which consisted of agreements and/or concerted practices that had the object of restricting and/or distorting competition in the EGB sector’ means that the anticompetitive object of the infringement must be assessed on the level of the single and continuous infringement as a whole.
1433Thus, in the first place, UniCredit and Portigon cannot validly rely, in order to call into question the anticompetitive object of the single and continuous infringement at issue, on the fact that, taken individually, certain discussions or categories of discussions included in that infringement are not sufficiently harmful to competition to be regarded as having an anticompetitive object.
1434There is nothing to prevent the Commission from finding the anticompetitive object of a single and continuous infringement consisting of various categories of conduct which, were they to be characterised individually, could not all have been characterised as a ‘restriction by object’. All that matters is whether, taken as a whole, that single and continuous infringement achieves a sufficient degree of harm to competition to be characterised as such.
1435Furthermore, in the present case, the Commission ‘has consistently held that the categories are intertwined and partially overlapping and have been presented in the current classification for analytical purposes’, as it notes in recital 549 of the contested decision.
1436UniCredit’s claims that the categories identified by the Commission, with the exception of certain discussions covered by Category 4, relate only to the primary market for EGBs and are therefore irrelevant for the purpose of establishing a restriction by object covering the secondary market must therefore be rejected.
1437Those criticisms must be rejected all the more since the Commission correctly found that the information exchanged in the chatrooms at issue and in connection with the primary market was of interest to the participants on the secondary market and vice versa (see paragraphs 476 to 486 above).
1438For the same reasons, Portigon’s claims that the discussions referred to in Category 4 cannot be characterised as a ‘restriction by object’, even though 14 of the 63 discussions relied on against that bank fell solely within that Category, must be rejected.
1439In the second place, UniCredit, Nomura and Portigon cannot validly rely, in order to call into question the anticompetitive object of the single and continuous infringement at issue, on the fact that, during their respective periods of participation in that infringement, the discussions in which they participated did not have an anticompetitive object.
1440The sufficient degree of harm to competition caused by a single and continuous infringement allowing it to be characterised as a ‘restriction by object’ cannot be assessed individually and separately for each undertaking participating in that conduct.
1441On the contrary, that assessment, which determines the rules of evidence applicable to the conduct at issue as a whole (see, to that effect, judgment of 30 January 2020, Generics (UK) and Others, C‑307/18, EU:C:2020:52, paragraph 63), must be carried out in the light of the objective characteristics of that conduct, taken as a whole, and without regard to the particular situation of each undertaking which participated in it.
1442That conclusion follows from the settled case-law according to which, first, an undertaking which has participated in a single and continuous infringement through conduct of its own which is intended to help bring about the infringement as a whole may also be liable for conduct put into effect by other undertakings in the context of the same infringement throughout the period of its participation in the infringement (see, to that effect, judgments of 8 July 1999, Commission v Anic Partecipazioni, C‑49/92 P, EU:C:1999:356, paragraph 83, and of 9 September 2015, Panasonic and MT Picture Display v Commission, T‑82/13, EU:T:2015:612, paragraph 81 (not published)) and, second, the mere fact that each undertaking takes part in the infringement in ways particular to it does not suffice to exclude its responsibility for the entire infringement, including conduct put into effect by other participating undertakings but sharing the same anticompetitive object or effect (judgment of 8 July 1999, Commission v Anic Partecipazioni, C‑49/92 P, EU:C:1999:356, paragraph 80).
1443The criticisms put forward by UniCredit, Nomura and Portigon, which are predicated on the assumption that only those discussions that took place during their respective periods of participation in the infringement at issue had to be taken into consideration in order to establish the anticompetitive object of those periods, must therefore be rejected.
1444Similarly, UniCredit’s claim, raised in the context of its third plea, that that bank’s participation in a ‘single and continuous infringement that amounted to a restriction of competition by object is not supported by contemporaneous evidence’ and, therefore, constitutes an error of law, must be rejected.
1445In that regard, it is admittedly true that, in full compliance with the case-law referred to in paragraphs 1441 and 1442 above, the Commission assessed whether the single and continuous infringement at issue was sufficiently harmful to competition by taking into consideration the discussions that took place prior to the period of UniCredit’s participation.
1446However, first, the Commission also relied on the discussions of 26 and 28 September, 12 and 19 October and 2, 3 and 28 November 2011 (see recital 514 and footnotes 985 and 986; recital 521 and footnote 994; by reference, in footnote 1011, to recital 420 and footnote 795; by reference, in footnote 1014, to recital 381 and footnote 732; by reference, in footnote 1015, to recitals 342, 345 to 349; by reference, in footnote 1016, to recital 342; by reference, in footnote 1017, to recitals 345 and 348; by reference, in footnote 1018, to recitals 381 and 514; recitals 702 to 705 and 708 to 717 of the contested decision). In so doing, the Commission referred to all the discussions validly relied on against UniCredit, thereby satisfying its obligation to state reasons.
1447Second, the taking into consideration of discussions that took place prior to the period of UniCredit’s participation in the single and continuous infringement at issue, for the purposes of characterising that infringement as a ‘restriction by object’, does not mean that the Commission used evidence that was not contemporaneous with the facts alleged in order to find that that bank participated in the infringement at issue.
1448Before the characterisation of that infringement as a ‘restriction by object’, which is carried out for the purposes of evidence (see, to that effect, judgment of 30 January 2020, Generics (UK) and Others, C‑307/18, EU:C:2020:52, paragraph 63), the Commission found, solely in the light of the discussions in which the UniCredit trader had actually taken part, that that bank had participated in a single and continuous infringement.
1449In the third place, in order to call into question the anticompetitive object of the single and continuous infringement at issue, UniCredit cannot validly rely on the fact that the Commission did not demonstrate that ‘the [discussions] that occurred in the CODS & CHIPS chatroom were appropriately characterised as a by object infringement in relation to each of the two markets’, namely the primary and secondary markets for EGBs.
1450Such a separate examination of the sufficient harm to competition caused by the discussions which took place in connection with (i) the primary market and (ii) the secondary market for EGBs would run counter to the very wording of the contested decision and, in particular, Article 1 thereof, which finds that the infringement in which the banks concerned participated was a single infringement.
1451The fact that UniCredit claims that the primary and secondary markets for EGBs are separate relevant markets in no way alters the scope of the infringement at issue.
1452Similarly, paragraphs 91 and 92 of the judgment of 16 September 2013, Keramag Keramische Werke and Others v Commission (T‑379/10 and T‑381/10, not published, EU:T:2013:457), relied on by UniCredit, do not support that bank’s arguments.
1453Unlike the situation in the case that gave rise to the judgment of 16 September 2013, Keramag Keramische Werke and Others v Commission (T‑379/10 and T‑381/10, not published, EU:T:2013:457), UniCredit was in direct competition with the other banks concerned on the secondary market for EGBs and the discussions relating to the primary market in which it participated conferred on it competitive advantages on the secondary market (see paragraphs 476 to 486 above).…
(2)The objections raised by UniCredit and Nomura relating to the assessment of the economic context of the infringement at issue
…
(i)The criticisms of UniCredit and Nomura relating to the incorrect assessment of the economic context of the infringement at issue
…
1492First, UniCredit and Nomura have failed to provide any evidence, in particular, evidence supported by figures, capable of demonstrating that the Commission wrongly found that the banks concerned had a ‘sizeable share of the market’, as the Commission found in recital 532 of the contested decision.
1493Second, those banks have also failed to provide evidence capable of calling into question the Commission’s finding, made in response to an argument raised by Nomura during the administrative procedure and based on that bank’s relatively small market share, that ‘this ability to check the accuracy of market information via additionally acquired knowledge clearly creates a competitive advantage for those involved’.
1494In that regard, UniCredit and Nomura have failed to demonstrate the extent to which the analysis of the size of the banks concerned would have been relevant in relation to the conduct alleged against them, namely price coordination, market sharing or customer allocation or even exchanges of commercially sensitive information.
1495Consequently, the arguments raised by UniCredit and Nomura alleging an incorrect assessment of the market for EGBs and, therefore, of the economic context of the infringement at issue, must be rejected as unfounded.
1496However, that conclusion does not predetermine whether, for each type of conduct at issue, or even for each discussion complained of, the Commission might have carried out an incorrect assessment of whether it was sufficiently harmful to competition to contribute to the characterisation of the conduct at issue, taken as a whole, as a ‘restriction by object’, which it is necessary to examine below.
(3)The claims of UniCredit, Nomura and Portigon alleging that the conduct at issue was insufficiently harmful to competition
1497UniCredit, in the context of its third plea, Nomura, in the context of its first and second pleas, and Portigon, in the context of the first limb of its fifth plea, submit, in essence, that the single and continuous infringement at issue does not reveal a sufficient degree of harm to competition to be characterised as a ‘restriction by object’.
1498In support of their criticisms, UniCredit, Nomura and Portigon essentially argue that all or part of the discussions relied on against them do not support the conclusion reached by the Commission.
1499As a preliminary point, on the one hand, it must be recalled that the assessment of the harmfulness of a single and continuous infringement to competition is carried out in the light of all the forms of conduct which constitute it, and not solely in the light of the discussions which took place during periods of participation in that infringement by undertakings challenging that classification (see paragraph 1441 above).
1500Therefore, in the present case, the characterisation of the single and continuous infringement at issue as a ‘restriction by object’ will be assessed, with regard to UniCredit, Nomura and Portigon, in the light of the harmfulness of the approximately 380 discussions that took place between 4 January 2007 and 28 November 2011, taken as a whole, taking into account (i) the discussions that are not challenged or not validly challenged by those banks – and which, therefore, must be regarded as definitively established – and (ii) the objections directed by those banks in respect of a smaller number of discussions.
1501In this respect, it should be noted that Portigon considers that ‘an exchange of confidential information before and during auctions may constitute a restriction by object’. Accordingly, of the 63 discussions relied on against it, that bank does not dispute any discussion in Categories 1 to 3 and disputes only the 14 discussions classified exclusively in Category 4, namely those of 20 and 21 October and 6 November 2009, 14 January, 15 and 22 February, 9 September and 17 November 2010, along with the 2 discussions of 18 January 2011 and those of 6, 11 and 12 April, and 18 May 2011.
1502For its part, of the 34 discussions relied on against it, Nomura lawfully challenges 12 discussions, namely those of 18 and 26 January, 3 and 25 February, 5 and 12 April, 3 and 5 May, 7 July, 28 September and 2 November 2011, it being understood that the Court has already held that the Commission cannot rely on those of 18 May, 22 June and 4 October 2011 against that bank (see paragraph 798 above).
1503In addition, UniCredit challenges the seven discussions validly relied on against it, namely those of 26 and 28 September, 12 and 19 October and 2, 3 and 28 November 2011, given that the Court has already held that the Commission cannot rely on the discussions of 4 October 2011 against that bank (see paragraph 798 above).
1504Second, as the Commission correctly observed in recitals 532 to 544 of the contested decision (see paragraphs 432 to 499 above), the single and continuous infringement at issue gave rise to practices of various kinds, namely price fixing arrangements, practices asymmetrically increasing transparency on the market for EGBs, strategies aligning trading and those exchanging sensitive information.
1505Consequently, an assessment of the merits of the characterisation of the single and continuous infringement at issue as a ‘restriction by object’ relied on by the Commission will not be carried out, as UniCredit, Nomura and Portigon seem to suggest, solely in the light of the harmfulness to competition of the exchanges of sensitive information which took place during the discussions at issue, but will also take account of the other types of conduct revealed by those discussions and referred to in the preceding paragraph.
1506In that regard, in the first place, it is settled case-law that price-fixing practices, such as those referred to in recital 535 of the contested decision, are particularly harmful to competition in so far as they are liable to lead to price increases or falls in production and, therefore, more limited supply, resulting in poor allocation of resources to the detriment of user undertakings and consumers (see, to that effect, judgments of 20 November 2008, Beef Industry Development Society and Barry Brothers, C‑209/07, EU:C:2008:643, paragraphs 17 and 33; of 11 September 2014, CB v Commission, C‑67/13 P, EU:C:2014:2204, paragraph 51; of 16 July 2015, ING Pensii, C‑172/14, EU:C:2015:484, paragraph 32).
1507Accordingly, in the present case, the Commission was entitled to find, without error, that the discussions referred to in footnote 1015 to the contested decision and which had given rise, between the traders of the banks concerned, in particular, to (i) agreements aimed at flattening or steepening the yield curve in order to attempt to alter the price of EGBs on the primary or secondary markets to their advantage (see recitals 498 to 502 of the contested decision and paragraph 418 above) and (ii) exchanges of bid levels or bidding volumes or levels of overbidding of the traders of the banks concerned, revealed such harm (see recitals 506, 508, 512 to 514 of the contested decision and paragraph 419 above).
1508Moreover, as the Commission rightly observed in recitals 538 and 542 of the contested decision, such practices proved to be all the more harmful because, on the primary market, ‘the collusion was actually or potentially detrimental to the [debt management offices] organising EGB auctions for the central governments, and therefore indirectly to the economies of the countries affected’, leading the banks concerned to ‘increase their prospects of securing their desired allocations of bonds from the auctions, at a lower price than they may otherwise have needed to pay’.
1509That is particularly the case, first, as regards specifically the period of Portigon’s participation in the infringement at issue, for the discussions of 5 November 2009 (see recital 275 of the contested decision), of 30 November 2009 (see recital 277), of 25 January 2010 (see recital 280), of 3 February 2010 (see recital 282), of 16 February 2010 (see recital 285), of 21 April 2010 (see recital 289), of 28 April, 18 May, 28 June, 1, 21 and 28 July 2010 (see recital 290), of 13 August 2010 (see recitals 291 and 292), of 10 September 2010 (see recital 295), of 27 September and 6 October 2010 (see recital 296), of 7 October 2010 (see recital 297), of 13 October 2010 (see recital 298), of 14 October 2010 (see recital 299), of 2 November 2010 (see recital 301), of 4 November 2010 (see recital 302), of 2 December 2010 (see recital 304), of 6 January 2011 (see recital 305), of 18 January 2011 (see recital 307), of 25 January 2011 (see recital 308), of 26 January 2011 (see recital 309), of 3 February 2011 (see recital 312), of 14 February 2011 in the afternoon (see recital 315), of 17 February 2011 (see recital 316), of 25 February 2011 (see recital 317), of 28 February 2011 (see recital 318), of 7 April 2011 (see recital 322), of 13 April 2011 (see recital 325), of 1 June 2011 (see recital 330). The same applies, secondly, specifically to Nomura’s period of participation in that infringement, to the discussions of 18 January 2011 (see recital 307), of 25 January 2011 (see recital 308), of 26 January 2011 (see recital 309), of 3 February 2011 (see recital 312), of 14 February 2011 in the afternoon (see recital 315), of 17 February 2011 (see recital 316), of 25 February 2011 (see recital 317), of 28 February 2011 (see recital 318), of 7 April 2011 (see recital 322), of 13 April 2011 (see recital 325), of 3 May 2011 (see recital 327), of 5 May 2011 (see recital 328), of 1 June 2011 (see recital 330), of 7 July 2011 (see recital 335), of 20 July 2011 (see recital 337), of 22 August 2011 (see recital 338), of 6 September 2011 (see recital 340), of 26 September 2011 (see recital 342), of 12 October 2011 (see recital 345), of 19 October 2011 (see recital 346), of 2 November 2011 (see recital 347), of 3 November 2011 (see recital 348), of 28 November 2011 (see recital 349). That also applies, thirdly, specifically to the period of UniCredit’s participation in that infringement, to the discussions of 26 September 2011 (see recital 342), of 12 October 2011 (see recital 345), of 19 October 2011 (see recital 346), of 2 November 2011 (see recital 347), of 3 November 2011 (see recital 348), of 28 November 2011 (see recital 349).
1510In the second place, the customer allocation practices, to which the Commission also referred in recital 535 of, and in footnote 1017 to, the contested decision also reveal a particularly high degree of harm to competition (see, to that effect, judgments of 20 November 2008, Beef Industry Development Society and Barry Brothers, C‑209/07, EU:C:2008:643, paragraphs 17 and 33; of 11 September 2014, CB v Commission, C‑67/13 P, EU:C:2014:2204, paragraph 51; of 16 July 2015, ING Pensii, C‑172/14, EU:C:2015:484, paragraph 32; and of 21 December 2023, European Superleague Company, C‑333/21, EU:C:2023:1011, paragraph 163 and the case-law cited).
1511That is particularly the case, first, as specifically regards the period of Portigon’s participation in the infringement at issue, for the discussions of 3 February 2010 (see recital 282 of the contested decision), of 7 October 2010 (see recital 297), of 2 December 2010 (see recital 304), of 17 February 2011 (see recital 316), of 7 April 2011 (see recital 322), of 3 May 2011 (see recital 327) and of 1 June 2011 (see recital 330); second, as specifically regards the period of Nomura’s participation in that infringement, for the discussions of 17 February 2011 (see recital 316), of 7 April 2011 (see recital 322), of 3 May 2011 (see recital 327), of 1 June 2011 (see recital 330), of 12 October 2011 (see recital 345) and of 3 November 2011 (see recital 348); and, third, as specifically regards the period of UniCredit’s participation in that infringement, the discussions of 12 October 2011 (see recital 345) and of 3 November 2011 (see recital 348).
1512In the third place, it is settled case-law that exchanges of information which are capable of reducing or removing uncertainty between participants as regards the timing, extent and details of the modifications to be adopted by the undertakings concerned in their conduct on the market must be regarded as presenting a degree of harmfulness to competition sufficient to render unnecessary the assessment of their effects (see, to that effect, judgment of 12 January 2023, HSBC Holdings and Others v Commission, C‑883/19 P, EU:C:2023:11, paragraphs 115 and 116 and the case-law cited).
1513As has previously been found in paragraph 1486 above, the Commission was entitled to consider that the mid-prices, the information on volumes exchanged, the yield curves, the trading positions, the timing of the pricing of syndications and the information relating to overbidding constituted non-public information, was sensitive and was capable of conferring on their holders an advantage in the market – both primary and secondary – for EGBs, in particular by reducing the uncertainty inherent in that market.
1514Consequently, the Commission rightly pointed out, in recitals 536 to 539 of the contested decision, that the regular exchanges of that information, within the small circle of the banks concerned, increased the transparency on the market for EGBs to the benefit of those seven banks alone and to the detriment of all the other banks. As a result, those exchanges generated an informational asymmetry for their exclusive benefit, enabling them to ‘[reduce] some of the normal market uncertainties and competitive tension which would otherwise have existed between the banks absent the disclosures’ and, ultimately, gave them a competitive advantage over their competitors (see, to that effect, judgment of 12 January 2023, HSBC Holdings and Others v Commission, C‑883/19 P, EU:C:2023:11, paragraph 203).
1515It follows that, in view of the requirement of autonomy inherent in Article 101 TFEU (judgment of 12 January 2023, HSBC Holdings and Others v Commission, C‑883/19 P, EU:C:2023:11, paragraph 202), the exchange of information referred to in paragraph 1513 above, during the discussions relied on against the banks concerned, reveals a sufficiently high degree of harm to competition to contribute to the characterisation of the single and continuous infringement at issue, as a whole, as a ‘restriction by object’.
1516In that regard, UniCredit, Nomura and Portigon cannot reasonably claim that the Commission failed to comply with the findings in paragraphs 186 to 193 of the judgment of 24 September 2019, HSBC Holdings and Others v Commission (T‑105/17, EU:T:2019:675).
1517First, in those paragraphs, the Court in no way found that the exchanged information in question – namely information on trading positions in the Euro interest rate derivatives sector – could not reduce or remove the degree of uncertainty on the market in question. The Court merely found that the Commission had not demonstrated to the requisite legal standard that, in the present case, that was the case (judgment of 24 September 2019, HSBC Holdings and Others v Commission, T‑105/17, EU:T:2019:675, paragraphs 188 and 191).
1518Second, it is clear, on the contrary, from paragraphs 1513 and 1514 above that the Commission highlighted the fact that the information exchanged in the present case conferred on traders an informational advantage, which could allow them to adjust their trading strategies accordingly and, therefore, reduced or removed the degree of uncertainty on the market for EGBs.
1519Lastly, it must be noted that the practices put into effect by the banks concerned arose between banks which, with the exception of Portigon, have all claimed or, at the very least, have not disputed their status as market makers on the secondary market for EGBs, which led them to carry out a particularly large volume of transactions in order to source liquidity (see recital 559 of the contested decision).
1520The Court has already had occasion to point out that that fact implied, from the point of view of competition on the market, that it was particularly fundamental that prices be determined independently (judgment of 24 September 2019, HSBC Holdings and Others v Commission, T‑105/17, EU:T:2019:675, paragraph 145), which the Commission moreover found, in essence, in recital 562 of the contested decision.
1521Accordingly, the Commission did not err in finding that the single and continuous infringement at issue, taken as a whole, revealed harm capable of leading to it being characterised as a ‘restriction by object’.
1522That finding cannot be called into question by the other arguments put forward by UniCredit, Nomura and Portigon.
1523As regards, in the first place, the arguments put forward by those banks alleging that certain discussions were not anticompetitive or even insufficiently harmful to competition, it has already been found in paragraphs 1509 and 1511 above that, in order to contribute to the characterisation of the single and continuous infringement at issue as a ‘restriction by object’, the Commission was able validly to rely on the discussion of 18 January 2011 against Portigon, on the discussions of 18 and 26 January 2011, that of 3 February 2011, those of 3 and 5 May 2011, that of 7 July 2011 and that of 2 November 2011 against Nomura, and on that of 26 September 2011, those of 12 and 19 October 2011 and those of 2, 3 and 28 November 2011 against UniCredit.
1524As regards the other discussions correctly challenged by those banks, the following should be noted.
1525With the exception of the discussion of 9 September 2010, which was taken into account by the Commission solely for the purpose of giving context for the infringement at issue (see recital 294 of the contested decision and paragraph 875 above), it is clear from paragraphs 663 and 664 (discussion of 25 February 2011 disputed by Portigon), paragraphs 669 and 670 (discussion of 5 April 2011 disputed by Nomura), paragraph 676 (discussion of 12 April 2011 disputed by Nomura and Portigon), paragraphs 721 and 736 (discussion of 28 September 2011 disputed by UniCredit and Nomura), paragraph 818 (discussion of 20 October 2009 disputed by Portigon), paragraph 827 (discussion of 21 October 2009 disputed by Portigon), paragraph 832 (discussion of 6 November 2009 disputed by Portigon), paragraph 852 (discussion of 14 January 2010 disputed by Portigon), paragraph 871 (discussion of 22 February 2010 disputed by Portigon), paragraph 883 (discussion of 17 November 2010 disputed by Portigon), paragraph 899 (discussion of 6 April 2011 disputed by Portigon) and paragraph 905 (discussion of 11 April 2011 disputed by Portigon), that the Commission rightly found that the discussions disputed by UniCredit, Nomura and Portigon were anticompetitive.
1526Moreover, as is clear from the paragraphs of the present judgment referred to in the preceding paragraph, the discussions of 20 October and 6 November 2009, of 14 January, 22 February and 17 November 2010, of 25 February, 5, 11 and 12 April and 28 September 2011, at least in part, gave rise to exchanges of information relating to either the mid-prices used by the banks concerned, the volumes exchanged by them, their yield curves, their trading positions or their levels of bidding or overbidding.
1527It has been found, in paragraph 1515 above, that the exchange of such information revealed a level of harm to competition capable of contributing to the characterisation of the single and continuous infringement at issue as a ‘restriction by object’.
1528In addition, the discussions of 21 October 2009, 15 February 2010 and 6 April 2011 concerned information relating either to the timing of the pricing of a syndication or to the services provided by the lead managers at the debt management offices of the issuing Member States.
1529In the light of the confidentiality obligations incumbent on the lead managers in the context of their interactions, to which the Commission referred in recital 33 of the contested decision and which are not disputed by UniCredit, Nomura or Portigon, both the disclosure and the receipt of such information imply a breach of those obligations, which is liable significantly to alter the process of issuing the EGBs concerned, be they issued through a syndication or an auction.
1530As regards, in the second place, UniCredit’s argument that the exchanges of information between the banks concerned would have allowed more favourable prices to be offered to their customers, it is clear that that argument is based on the premiss that the infringement at issue was confined to the secondary market for EGBs, whereas it concerned the market for those bonds – primary and secondary – as a whole. That argument therefore disregards the implications of such exchanges on the primary market.
1531Moreover, that same argument does not give rise to any reasonable doubt as to the harmful nature of those exchanges with regard to competition on the market in question (judgment of 12 January 2023, HSBC Holdings and Others v Commission, C‑883/19 P, EU:C:2023:11, paragraph 205) and, in particular, with regard to the competitors of the banks concerned and their customers.
1532In the third place, Nomura cannot reasonably rely on the fact that the characterisation of the single and continuous infringement at issue as a ‘restriction by object’ should be based on ‘reliable and robust wealth of experience’ in that area or on economic studies or reports by independent authors.
1533As has already been observed in paragraph 1413 above, the Court of Justice has expressly held that it is in no way necessary that the same type of conduct as that which must be characterised by the Commission in a given case should have previously been censured by the Commission, in order for that conduct to be regarded as a restriction of competition by object.
1534Similarly, the argument based on the existence of economic studies or reports cannot succeed without making the characterisation of conduct not yet sanctioned as a ‘restriction by object’ dependent on a purely incidental fact.
1535As has been recalled in paragraph 1414 above, for the purposes of characterisation as a ‘restriction by object’, all that matters is an individual and detailed examination of the practice concerned, which must demonstrate that that practice presents a sufficient degree of harm.…
(4)The claims of UniCredit, Nomura and Portigon alleging that the conduct at issue had no effect
…
1547It follows that Portigon, in the context of the second argument in the first limb of its fifth plea, cannot validly criticise the Commission for failing to state reasons in that regard.
1548In any event, it should be noted that UniCredit’s claim, which is, moreover, unsubstantiated, that the conduct at issue affected not all customers on the secondary market for EGBs, but only certain specific buyers of EGBs, is irrelevant.
1549Indeed, Article 101(1) TFEU, like the other competition rules of the Treaty, is designed to protect not only the immediate interests of certain identified operators but also to protect the structure of the market and thus competition as such (see, to that effect, judgment of 19 March 2015, Dole Food and Dole Fresh Fruit Europe v Commission, C‑286/13 P, EU:C:2015:184, paragraph 125 and the case-law cited).
1550It should also be noted that the econometric analysis, referred to in paragraph 1543 above, on the basis of which Nomura claims that the conduct at issue had no effects, concerns only that bank’s activities on the primary market – namely its participation in EGB auctions – when, first, the infringement alleged against Nomura also concerns the secondary market on which Nomura operated in parallel and, second, the primary and secondary markets of EGBs are interdependent (see recitals 45 to 50 of the contested decision).
1551Consequently, the arguments of UniCredit, Nomura and Portigon, alleging a lack of evidence of the anticompetitive effects of the conduct at issue, must be rejected as unfounded.
(5)Conclusion on the characterisation of the single and continuous infringement at issue as a restriction by object
1552It follows from all of the foregoing that the Commission did not err, in recitals 532 to 544 of the contested decision, in characterising the single and continuous infringement at issue, taken as a whole, as a ‘restriction by object’.
1553Accordingly, UniCredit’s third, fourth and fifth pleas, Nomura’s first and second pleas, and the first limb of Portigon’s fifth plea must be rejected as unfounded.
1554It follows from the rejection of those pleas that UniCredit’s first plea, alleging a failure to state reasons on account of the lack of definition of the relevant market or markets on which the infringement at issue occurred, is also unfounded.
1555First, the obligation to define the market in a decision adopted under Article 101 TFEU is incumbent on the Commission where it is impossible, without such a definition, to determine whether the agreement or concerted practice at issue is liable to affect trade between Member States and has as its object or effect the prevention, restriction or distortion of competition (see, to that effect, judgment of 29 February 2016, EGL and Others v Commission, T‑251/12, not published, EU:T:2016:114, paragraph 34 and the case-law cited).
1556Second, where, as in the present case, the Commission finds an infringement whose object was to restrict competition throughout the European Union, it is not necessary first to define the geographical market before finding that there is such a restriction of competition (see, to that effect, judgment of 14 May 1998, Enso Española v Commission, T‑348/94, EU:T:1998:102, paragraph 232 and the case-law cited).
1557However, UniCredit has in no way maintained that it was impossible to determine whether the infringement at issue was capable of affecting trade between Member States, without defining the relevant market or markets on which it intervened.…
5.The pleas in law alleging errors in the finding that the Commission had a legitimate interest in finding the infringement at issue with regard to Natixis, BofA and Portigon
(a)The pleas in law alleging infringement of Article 7(1) of Regulation No 1/2003, in that the Commission failed to demonstrate a legitimate interest in finding the infringement at issue with regard to Portigon, Natixis and BofA
…
(1)The pleas alleging that the Commission had no legitimate interest in finding the infringement at issue with regard to Natixis and BofA
(i)The first limb of Natixis’ third plea, alleging infringement of the obligation to state reasons
…
1583First, although Article 7(1) of Regulation No 1/2003 requires proof of a legitimate interest in finding an infringement for which the Commission’s power to impose penalties is time-barred, the Commission is not, however, required to give further reasons for its decision by reference to its previous decision-making practice by specifically setting out the reasons for reaching a different conclusion than in a previous case concerning similar or identical situations or the same market participants (see, to that effect, judgment of 14 December 2005, General Electric v Commission, C‑210/01, EU:T:2005:456, paragraph 513).
1584It is in the particular context of each case that the Commission, in the exercise of its discretion, ought to decide whether it is appropriate to find an infringement with regard to one undertaking in particular and to protect the effectiveness of competition law (see, to that effect, judgment of 29 March 2012, Telefónica and Telefónica de España v Commission, T‑336/07, EU:T:2012:172, paragraph 357).
1585Moreover, in recitals 787 to 791 of the contested decision, the Commission expressly replied to the arguments raised by Natixis and BofA in relation to the Commission’s previous decision-making practice.
1586Second, the absence of guidelines for the application of Article 7(1) of Regulation No 1/2003 also does not justify requiring the Commission to comply with an enhanced obligation to state reasons. Moreover, Natixis has failed to put forward any argument in support of that claim.
1587In the light of the foregoing, the first limb of Natixis’ third plea must be rejected as unfounded.
(ii)The first plea and the second part of Natixis’ third plea and BofA’s second plea, alleging that the reasons justifying the Commission’s legitimate interest in finding the infringement at issue with regard to Natixis and BofA are unfounded
1588First, Natixis, in the context of its first plea, and BofA, in the context of the first limb of its second plea, claim, in essence, that the Commission infringed Article 7(1) of Regulation No 1/2003 by finding that they had committed an infringement. In their view, the Commission cannot demonstrate a legitimate interest in finding the infringement at issue by justifying it on generic grounds, without substantiating that interest with evidence relating to the particular circumstances of the infringement at issue and factors specific to the undertaking concerned, without rendering meaningless the need to demonstrate that interest.
1589In that connection, according to Natixis and BofA, the four reasons given in recitals 781 to 784 of the contested decision, set out in paragraphs 1575 to 1579 above, do not demonstrate a legitimate interest in finding the infringement at issue with regard to Natixis and BofA. While they do not dispute that the Commission may carry out an overall assessment of the evidence it puts forward, they nonetheless take the view that, individually or collectively, such items of evidence do not establish such an interest.
1590In the second place, in the context of the second limb of its third plea, Natixis claims, in essence, that the contested decision breaches the principle of proportionality, in that the finding of the infringement at issue with regard to Natixis was not necessary for an effective implementation of EU competition rules and that the disadvantages for that bank, in terms of the risks to its reputation and actions for damages, are disproportionate.
1591In the third place, in the context of the second limb of its second plea, BofA adds that the reasons put forward in the contested decision should have been assessed in the light of considerations relating to the principles of legal certainty and respect for its rights of defence, which the Commission wrongly failed to do.
1592More specifically, BofA observes that, although the Commission’s previous decision making practice is not binding on it in other cases, the fact remains that previous decisions show that, in the past, the Commission has refrained from finding an infringement in the absence of an undertaking’s own legitimate interest.
1593According to the last sentence of Article 7(1) of Regulation No 1/2003, if the Commission has a legitimate interest in doing so, it may find that an infringement has been committed in the past.
1594In the present case, the Commission found that it had a legitimate interest in finding the infringement with regard to Natixis and BofA for four reasons, set out in recitals 781 to 784 of the contested decision.
1595Indeed, those four reasons relate to the fact that (i) Natixis and BofA were involved in the same single and continuous infringement as the other banks concerned; (ii) their inclusion in the contested decision and the description of their conduct were necessary to explain the overall functioning and full scope of the infringement at issue; (iii) the conduct alleged against those banks was very serious; and, lastly, (iv) such a finding enables the victims of the unlawful conduct in question to bring legal actions for damages, and the Commission to classify the bank to which it is addressed as a recidivist if it were to commit a similar infringement in the future.
1596In that context, the Court considers that the first three reasons put forward by the Commission, taken together, are in themselves sufficient to establish that the Commission had a legitimate interest in finding the infringement at issue with regard to Natixis and BofA.
1597As regards the first two reasons, the Commission rightly observes, in recitals 796 to 801 of the contested decision, that the finding of the infringement at issue with regard to Natixis and BofA contributed in particular to demonstrating the frequency of the collusive discussions between the traders of each bank, the nature of their relationships and the continuous nature of that infringement.
1598More specifically, in recital 798 of the contested decision, as regards BofA, the Commission was entitled, without committing any errors, to find that, in view of the fact that the BofA trader’s active contribution to the CODS & CHIPS chatroom reinforced the interdependence of the traders in order to maintain their collusive behaviour and to derive a profit from it, the inclusion of BofA in the contested decision had contributed to demonstrating the particular nature of the relationship between the traders and the consistency of their exchanges.
1599Similarly, in recital 801 of the contested decision, as regards Natixis, the Commission was fully entitled to rely, inter alia, on the Natixis trader’s uninterrupted participation in the discussions at issue, despite his changes in employers, in order to highlight the continuous nature of the infringement at issue.
1600In that context, Natixis and BofA cannot validly maintain that their identification in the contested decision was not relevant and did not contribute substantially to establishing the infringement at issue or to explaining the scope of the unlawful conduct.
1601It is true that the Commission could have used evidence relating to the participation of Natixis and BofA, as set out above, while preserving the anonymity of those banks in the contested decision or by refraining from finding the infringement at issue with regard to Natixis and BofA.
1602However, contrary to the argument advanced by Natixis and BofA, the Commission is under no obligation to show that the finding of the infringement, in respect of which its power to impose a fine is time-barred, is necessary. Article 7(1) of Regulation No 1/2003 requires only that it be demonstrated that there is a legitimate interest in finding that infringement, without setting out any obligation to show that such a finding is necessary. Obliging the Commission to show that it is necessary to find the infringement at issue would amount to increasing the burden of proof without due cause in contradiction with the wording of Article 7(1).
1603In any event, the Court has already held that the fact that, in previous decisions, the Commission had not held undertakings liable for equivalent conduct could not give rise to a legitimate expectation that the Commission would refrain in future cases from pursuing and penalising such conduct, where that reorientation of the Commission’s decision-making practice, even if proved, was based on a correct interpretation of the scope of the relevant legal provisions (see, to that effect, judgment of 8 September 2010, Deltafina v Commission, T‑29/05, EU:T:2010:355, paragraph 428).
1604Consequently, the mere fact that it would have been possible for the Commission not to find the infringement at issue with regard to the applicants cannot, however, preclude its legitimate interest in doing so.
1605As regards the third reason relating to the serious nature of the conduct of those banks, highlighted in recital 783 to the contested decision, the Commission validly found that that conduct was particularly serious having regard to the sector concerned and the context of the financial crisis in which that conduct occurred.
1606In that regard, Natixis and BofA cannot validly rely on the lesser degree, by comparison with the other banks concerned, of their respective participation in the infringement at issue. Natixis participated in the infringement at issue for a little over a year and five months and BofA for a little over a year and nine months, which cannot require that the Commission refrain from including them in the contested decision and from finding that infringement with regard to Natixis and BofA.
1607In the light of the foregoing, Natixis and BofA are wrong to complain that the Commission relied only on generic grounds to establish its legitimate interest in finding the infringement at issue with regard to Natixis and BofA.
1608Accordingly, without it being necessary to examine the merits of the fourth reason put forward by the Commission, the latter has demonstrated to the requisite legal standard the existence of a legitimate interest in finding the infringement at issue with regard to Natixis and BofA.
1609That conclusion is not called into question by the other arguments advanced by Natixis and BofA.
1610In the first place, as regards Natixis’ arguments, that bank cannot validly rely on a breach of the principle of proportionality by reason, first, of the fact that it cooperated under the Leniency Notice, but did not derive any benefit from it, since the Commission’s power to impose a fine on it was time-barred.
1611In that regard, while it is true that the efforts to cooperate demonstrated by Natixis will not, in the absence of a reduction of the fine, be rewarded, the Commission was nevertheless right to point out, in essence, in recitals 803 and 804 of the contested decision, that the Leniency Notice does not contemplate immunity of an undertaking from liability in connection with the finding of an infringement.
1612That notice establishes a framework for rewarding undertakings which have cooperated in a Commission investigation only by granting them immunity from, or a reduction of, that fine (see, in particular, points 8 and 23 of that notice).
1613In recital 804 of the contested decision, the Commission rightly points out that the undertakings that cooperate on the basis of the Leniency Notice do so after mature reflection and with insider knowledge of their own conduct, including the duration of the latter. Accordingly, those undertakings cooperate with full knowledge of the facts and, in particular, with knowledge of point 36 of that notice, which provides that the Commission will not take a position on whether or not to reward an application if it becomes apparent that the application concerns infringements covered by the five years limitation period set out in Article 25(1)(b) of Regulation No 1/2003.
1614Nor can Natixis compare its situation with that of the other leniency applicants, namely RBS and UBS, since, unlike in the case of the two latter banks, the Commission’s power to impose a fine on Natixis was time-barred.
1615Moreover, as regards the fact that Natixis was not in a position, at submission stage of its application for leniency, to determine with certainty whether the Commission’s power to impose a fine on it for its participation in the infringement at issue was time-barred in so far as Natixis was concerned, suffice it to recall that that uncertainty is an inherent part of the risk-benefit analysis prior to any decision by an undertaking whether or not to submit an application for immunity or reduction of a fine, with the result that the circumstance relied on by Natixis is not relevant.
1616It follows that Natixis’ cooperation, even in so far as it was not otherwise rewarded, cannot call into question the existence of a legitimate interest on the part of the Commission in finding the infringement at issue with regard to that bank.
1617Second, Natixis also cannot validly rely on the Commission’s previous decision-making practice related to the application of Article 7(1) of Regulation No 1/2003 in order to criticise the disproportionate nature of the contested decision and claim that the ‘unusual’ exercise of the Commission’s powers involves a strict examination of the proportionality of the measure at issue in the light of the importance of a clear, predictable and uniform application of EU law.
1618In so far as that line of argument is based on a comparison between the situation of Natixis and those of undertakings involved in different procedures relating to infringements of EU competition law, it must be rejected, since the Commission’s practice in previous decisions does not itself serve as a legal framework for fines or findings of infringements in competition matters and is merely indicative in nature (see, to that effect, judgment of 24 September 2009, Erste Group Bank and Others v Commission, C‑125/07 P, C‑133/07 P, C‑135/07 P and C‑137/07 P, EU:C:2009:576, paragraph 233 and the case-law cited).
1619In the second place, as regards BofA’s arguments, that bank cannot validly maintain that a proper account of the principles of legal certainty and respect for its rights of defence, in view of the time which had elapsed between the end of its participation in the infringement at issue and the time when the contested decision was adopted, would have led the Commission to the conclusion that it lacked a legitimate interest in finding the mentioned infringement with regard to that bank.
1620In that regard, first, although the finding of a legitimate interest in finding the infringement at issue must be made in compliance with the general principles of EU law and, in particular, does not authorise the Commission to delay indefinitely the exercise of its powers (see, to that effect, judgments of 14 July 1972, Geigy v Commission, 52/69, EU:C:1972:73, paragraph 21, and of 24 September 2002, Falck and Acciaierie di Bolzano v Commission, C‑74/00 P and C‑75/00 P, EU:C:2002:524, paragraph 140), it cannot, however, be held, in the present case, that the passage of time made BofA’s defence more difficult.
1621First of all, although a period of more than seven years actually elapsed between the end of BofA’s participation in the infringement at issue, namely on 6 November 2008, and the receipt, on 3 March 2016, of the request for information concerning that infringement, it must nevertheless be pointed out that a period of only five months elapsed between RBS’ formal application for immunity from a fine within the meaning of recital 14 of the Leniency Notice, submitted on 11 September 2015, and the request for information sent to BofA.
1622The fact that a period of approximately six and a half years elapsed between the end of BofA’s participation in the infringement at issue and the submission of that formal application for immunity from a fine cannot in any way be equated with negligence on the part of the Commission and validly relied on against it in the case of an infringement of BofA’s rights of defence, since BofA itself intentionally contributed to the passage of such a period by deliberately keeping the conduct in question secret (see, to that effect, judgment of 6 October 2005, Sumitomo Chemical and Sumika Fine Chemicals v Commission, T‑22/02 and T‑23/02, EU:T:2005:349, paragraph 91).
1623Moreover, as the Commission correctly points out in recital 795 of the contested decision, it follows from settled case-law that, by virtue of the general duty of care attaching to any undertaking, BofA ought to have ensured the proper maintenance of information enabling details of its activities to be retrieved, in order, in particular, to make the necessary evidence available in the event of legal or administrative proceedings (see, to that effect, judgment of 12 July 2018, Prysmian and Prysmian Cavi e Sistemi v Commission, T‑475/14, EU:T:2018:448, paragraph 101). Thus, since BofA had been the subject of a request for information from the Commission, it was a fortiori incumbent on it to act with greater diligence and to take all appropriate measures in order to preserve such evidence as might reasonably be available to it (see, to that effect, judgment of 16 June 2011, Heineken Nederland and Heineken v Commission, T‑240/07, EU:T:2011:284, paragraph 301; see also, by analogy, judgment of 25 March 2021, Xellia Pharmaceuticals and Alpharma v Commission, C‑611/16 P, EU:C:2021:245, paragraph 152).
1624Next, BofA failed to indicate in detail, if not the missing specific items of evidence, at the very least, the incidents, events or circumstances which prevented it from complying with its obligation of diligence and brought about the alleged difficulty in gathering exculpatory evidence that it alluded to (see, to that effect, judgment of 8 September 2016, Xellia Pharmaceuticals and Alpharma v Commission, T‑471/13, not published, EU:T:2016:460, paragraph 358).
1625Similarly, although BofA submits that the passage of time complicated its reading and its work of interpreting the discussions between traders, which served as evidence in the present case, the fact remains that BofA, which is an informed and experienced operator in the EGB sector, did not provide any specific example of a situation in which it was unable to propose an interpretation of one or more discussions or in which it found a significant lack of information or evidence, either in its written pleadings or when it was questioned on that subject at the hearing.
1626On the contrary, BofA was put in a position to refer to all the messages that arose during a day of discussions relied on against it, as it expressly confirmed at the hearing, in order to give context for the exchanges at issue relied on by the Commission and to put them into perspective with the remainder of the traders’ discussion, to the effect that the passage of time, as such, could only have had little impact on the understanding of the discussions.
1627Similarly, in so far as BofA takes the view that its understanding of the discussions was rendered more complex by the fact that it was unable to contact its former trader who had left the bank 10 years earlier, BofA has failed to adduce any evidence or prima facie evidence to show that it had effectively attempted to contact the latter trader and that that contact yielded no results.
1628Lastly, in so far as BofA relies on arguments alleging, in essence, selective presentations by the leniency applicants of the evidence at their disposal and the limited probative value vis-à-vis BofA of their statements, suffice it to state, first, that that selectivity is not affected by the passage of time and, second, that BofA was able to submit its observations on all the evidence in the file during the administrative procedure, so that those arguments cannot establish or corroborate an infringement of its rights of defence.
1629Second, BofA’s argument that the Commission adopted the contested decision out of time in the light of the requirements of legal certainty cannot succeed since, in the present case, it is not a question of negligence on the part of the Commission in initiating or completing the administrative procedure, but of the mere and actual passage of time from the end of its participation in the infringement at issue, it not being alleged that the Commission was aware, or even could and should have been aware of it, at a time which would have enabled it to act sooner (see, to that effect, judgment of 6 October 2005, Sumitomo Chemical and Sumika Fine Chemicals v Commission, T‑22/02 and T‑23/02, EU:T:2005:349, paragraph 90).
1630After various requests for information had been sent from 29 January 2016 onwards, the Commission notified the Statement of Objections to the banks concerned on 31 January 2019, heard the addressees of that statement between 22 and 24 October 2019 and adopted the contested decision on 20 May 2021. There is nothing in that chronology to suggest that the procedure was unreasonably long (see, to that effect, judgment of 6 October 2005, Sumitomo Chemical and Sumika Fine Chemicals v Commission, T‑22/02 and T‑23/02, EU:T:2005:349, paragraph 90).
1631In the light of all the foregoing considerations, it must be found that the Commission was entitled, without committing any errors, to find the infringement at issue with respect to Natixis and BofA.
1632Accordingly, the first plea and the second limb of the third plea raised by Natixis and the second plea raised by BofA must be rejected as unfounded.
(b)The pleas alleging infringement of Natixis’ and BofA’s rights of defence vitiating the fourth reason stated in recital 784 of the contested decision
1633Natixis, in the context of its second plea, and BofA, in the context of the first limb of its second plea, claim that the Commission infringed their rights of defence on the ground that, in recital 784 of the contested decision, it referred to a fourth reason to demonstrate its legitimate interest in finding the infringement at issue, which did not appear in the Statement of Objections.
1634In that regard, it follows from paragraphs 1588 to 1631 above that the Commission’s legitimate interest in finding the infringement at issue with respect to Natixis and BofA is sufficiently substantiated by the reasons set out in recitals 781 to 783 of the contested decision. Accordingly, any irregularity vitiating recital 784 of the contested decision is not such as to call into question the existence of the Commission’s legitimate interest in finding that the infringement at issue was committed by Natixis and BofA.
1635In any event, the Commission did not infringe Natixis’ and BofA’s rights of defence by finding, in recital 784 of the contested decision, that its interest in finding the infringement at issue was also justified, since, first, such a finding would enable it to classify those banks as recidivists if they were to commit a similar infringement in the future and, secondly, that finding would facilitate actions for damages against the undertakings involved in that infringement.
1636The mere addition, in recital 784 of the contested decision, of a fourth reason does not amount to holding those banks liable for infringements different from those already referred to in the Statement of Objections or to relying on facts and legal classifications on which they did not have the opportunity to comment, since, in recital 685 of the Statement of Objections, the Commission clearly indicated its intention to find the infringement at issue against those banks despite the fact that its power to impose a fine on them was time-barred.
1637Accordingly, that difference between the Statement of Objections and the contested decision cannot lead to a finding of infringement of Natixis’ and BofA’s rights of defence, in the light of the requirements referred to in paragraphs 116 and 117 above which the Commission is required to observe in order to ensure respect for the rights of defence of the undertakings concerned.
1638That conclusion is all the more compelling in relation to the interest in facilitating actions for damages before national courts.
1639By relying on such a legitimate interest, the Commission has merely set out the consequences inherent in the implementation of the prohibition laid down in Article 101(1) TFEU. In that regard, the Court of Justice has repeatedly held that everyone has the right to claim damages for the harm caused by an anticompetitive conduct (see, to that effect, judgments of 20 September 2001, Courage and Crehan, C‑453/99, EU:C:2001:465, paragraph 26, and of 13 July 2006, Manfredi and Others, C‑295/04 to C‑298/04, EU:C:2006:461, paragraph 60), since such actions contribute to the maintenance of effective competition in the European Union by discouraging agreements or practices, frequently covert, which are liable to restrict or distort competition (see, to that effect, judgments of 6 November 2012, Otis and Others, C‑199/11, EU:C:2012:684, paragraph 42, and of 14 March 2019, Skanska Industrial Solutions and Others, C‑724/17, EU:C:2019:204, paragraph 44).
1640It follows that, on reading recital 685 of the Statement of Objections, clearly showing the Commission’s intention to find the infringement at issue against them, Natixis and BofA could reasonably have foreseen that such a finding would necessarily have the consequence of facilitating possible actions for damages brought before the national courts by the victims of the anticompetitive conduct attributable to those two banks.…
6.The pleas raised by UBS, UniCredit and Nomura alleging a failure to state reasons, infringement of the rights of the defence and errors in the determination of their respective fines
…
1652Secondly, UBS, in its first plea, the first limb of its second plea and its third to fifth pleas, UniCredit, in its seventh and eighth pleas, and Nomura, in its sixth plea, object to the methodology used by the Commission to determine the proxies applied to them and to the amounts of those proxies.
1653Thirdly, in their respective ninth pleas, UniCredit and Nomura argue that the Commission made errors in its determination of the gravity factor applied to them (points 20 to 23 of the Guidelines).
1654Fourthly, in the first limb of its tenth plea, UniCredit argues that the Commission failed to observe the principle that penalties must be specific to the offender and the offence and the principle of proportionality in determining the rate of the additional amount.
1655In the second place, in their respective tenth pleas, UniCredit and Nomura challenge the adjustments made by the Commission to the basic amount of the fines imposed on them and the failure to make use of point 37 of the Guidelines.
1656The above objections relating to the calculation of the amount of their respective fines will be examined in turn and in the order set out in paragraphs 1647 to 1655 above.
(a)The pleas alleging infringement of UBS’s and Nomura’s rights of defence in the determination of their respective fines
1657UBS, in the second limb of its second plea, and Nomura, in its eighth plea, claim that the Commission infringed their rights of defence in not affording them the opportunity to understand, from the outset of the administrative procedure, the methodology used to determine the proxy and, consequently, the opportunity to challenge it effectively.
1658In that regard, UBS argues that, in the Statement of Objections, the Commission should have explained its calculation methodology in detail, in so far as it relied on a figures-based model in which the reduction factor was to play an essential role.
1659Nomura submits that prior to receiving the Letter on Fines, and therefore during most of the administrative procedure, it had no specific details concerning the proxy calculation, save for the notional volumes disclosed in the Statement of Objections. Furthermore, after the Letter on Fines, the Commission set out its methodology exclusively via bilateral communications, without giving Nomura the opportunity to raise specific issues orally with regard to that methodology until the State of Play meeting of 22 March 2021. Material amendments to the initial methodology were disclosed by the Commission at that State of Play meeting, for which no written details were provided until the contested decision, with the exception of a vague email of 25 March 2021.
1660As a preliminary point, it should be noted that, in the second limb of its second plea, UBS claims that the Commission failed to comply with its obligation to state reasons and infringed UBS’s rights of defence, relying on case-law applicable to the obligation to state reasons – namely paragraphs 342 to 348 of the judgment of 24 September 2019, HSBC Holdings and Others v Commission (T‑105/17, EU:T:2019:675), or paragraphs 30 and 31 of the judgment of 10 July 2019, Commission v Icap and Others (C‑39/18 P, EU:C:2019:584) – in order to prove the infringement of its rights of defence.
1661However, those are two separate claims. The first concerns the content of the contested decision, whereas the second – which UBS has not sufficiently substantiated – concerns the conduct of the administrative procedure which led to the adoption of that decision and, in particular, observance of the right of the banks concerned to be heard during that procedure.
1662Therefore, only the arguments relating to the alleged infringement of Nomura’s and UBS’s rights of defence will be examined in paragraphs 1663 to 1675 below; the arguments alleging a failure to state reasons in the contested decision will be examined in paragraphs 1684 to 1695 below.
1663In that regard, it should be recalled that, as the Commission observed in essence in recital 835 of the contested decision, in order to fulfil its obligation to respect the right of undertakings to be heard, the Commission is required to indicate expressly in the statement of objections that it will consider whether it is appropriate to impose fines on the undertakings concerned and to set out the principal elements of fact and of law that may give rise to a fine, such as the gravity and the duration of the alleged infringement and whether it has been committed intentionally or negligently (see judgment of 6 July 2017, Toshiba v Commission, C‑180/16 P, EU:C:2017:520, paragraph 21 and the case-law cited).
1664However, it is also apparent from case-law that the Commission is not required, once it has indicated the elements of fact and of law on which it will base its calculation of the amount of the fines, to specify the way in which it will use each of those elements in order to determine the amount of the fine (see judgment of 6 July 2017, Toshiba v Commission, C‑180/16 P, EU:C:2017:520, paragraph 21 and the case-law cited), any more than the level of the fines envisaged and a fortiori their exact amount (see, to that effect, judgments of 24 September 2009, Erste Group Bank and Others v Commission, C‑125/07 P, C‑133/07 P, C‑135/07 P and C‑137/07 P, EU:C:2009:576, paragraph 182 and the case-law cited, and of 7 November 2019, Campine and Campine Recycling v Commission, T‑240/17, not published, EU:T:2019:778, paragraph 360).
1665Nonetheless, where the Commission intends to calculate fines in a way that departs from its established methodology, it is required to include a particular statement to that effect in the statement of objections (see, to that effect, judgment of 18 December 2008, Coop de France bétail et viande and Others v Commission, C‑101/07 P and C‑110/07 P, EU:C:2008:741, paragraph 50).
1666In the present case, it is clear that the Commission has met its requirements.
1667In the Statement of Objections addressed to Nomura and UBS, respectively, the Commission provided details of the mandatory criteria relating to the gravity and duration of the alleged infringement and the fact that the infringement had been committed intentionally (see recitals 680 to 683 and 694 to 698 of the Statement of Objections), in accordance with the case-law referred to in paragraphs 1663 and 1664 above.
1668It also stated that, when applying the methodology set out in the Guidelines, it intended to use not the value of sales of the services supplied by the banks concerned to which the alleged infringement relates, but a proxy for the value of sales, calculated on the basis of the annualised notional amounts traded by each of those banks during their respective period of participation in the infringement at issue discounted by a ‘factor that takes account of the particularities of the financial industry, and the EGB industry in particular, allowing the Commission to calculate fines that are set at an appropriate level of deterrence’ (see recitals 686 to 692 of the Statement of Objections). In addition, in a separate and confidential Annex 2 specific to each of the banks concerned, the Commission specified the exact amount of the notional amounts and the annualised notional amounts which it intended to use.
1669During the administrative procedure, the Commission also sent Nomura and UBS a Letter on Fines, explaining in detail the method for calculating the proxy that would be applied to them. It set out, first, the method for calculating the notional amounts, the total value of which had been indicated in the Statement of Objections (see paragraphs 5 and 8 to 10 of the Letter on Fines) and, second, the level of the adjustment factor which it intended to apply and the method for calculating it, referring to detailed mathematical formulae and the data to which those formulae were applied (see paragraphs 11 to 38 of the Letter on Fines). In addition, it gave the banks concerned the opportunity to submit their written observations on the content of that letter.
1670As regards UBS in particular, it should also be noted that, as early as July 2017, that bank was engaged, in its own words, in ‘constructive discussions on the best metric to measure the “value of sales” in the context of the Commission’s EGB investigation’, which led to a meeting held on 11 July 2017, and to that bank sending, on 23 November 2017, a 16-page memorandum headed ‘Case 40324 EGB – “Value of Sales” – Q&A’. In addition, that bank was given the opportunity to submit any comments on the Commission’s proposed methodology in the context of both written observations (replies to the Statement of Objections of 19 April 2019, letter of 2 October 2019, letter of 17 February 2020) and oral observations (hearing of 22 to 24 October 2019). On those occasions, which took place before the Letter on Fines was sent, UBS states that it put forward its arguments in relation to the calculation of the proxy and the fact that it should be based on its ‘net value traded’, possibly adjusted. In addition, after the Letter on Fines was sent and UBS had submitted its observations, a State of Play meeting took place on 14 March 2021.
1671Similarly, Nomura indicated in its pleadings that, even before the Letter on Fines, it had been given the opportunity on several occasions to submit its observations on the methodology for calculating the fine that the Commission intended to impose on it (replies to the Statement of Objections of 14 May 2019 and letter of 23 December 2019) and that, after that letter, it had also been able to obtain ‘further clarifications’ from the Commission in the context of bilateral communications (email of 9 December 2020), could raise ‘specific issues’ at the State of Play meeting held on 22 March 2021 and had been able to find out, by an email of 25 March 2021, the total value of the notional volumes and the discount factor which the Commission ultimately intended to apply to it.…
(b)The pleas alleging errors in the determination of the amounts of the fines imposed on UBS, UniCredit and Nomura
…
(1)The first limb of UBS’s second plea, alleging that the basic amount of the fine was not determined in accordance with the Guidelines
…
1702However, UBS’s line of argument is based on a misreading of recital 822 of the contested decision. In that recital, the Commission did not apply any ‘discount factor’ to the proxy used for that bank, but merely justified the decision to use that proxy on grounds relating to the appropriateness of the fine imposed on, inter alia, UBS and, in particular, its proportionate and deterrent nature.
1703In addition, as observed in paragraphs 1680 and 1682 above, the Commission did not depart from the general methodology set out by the Guidelines, nor did it make use of point 37 of those guidelines.
1704In any event, it is settled case-law that deterrence is an objective of the fine and a general requirement which must be a reference point for the Commission throughout the calculation of the amount of the fine and that, therefore, the objective of deterrence does not necessarily require that there be a specific step in that calculation in which an overall assessment is made of all relevant circumstances for the purpose of attaining that objective (see, to that effect, judgment of 27 March 2014, Saint-Gobain Glass France and Others v Commission, T‑56/09 and T‑73/09, EU:T:2014:160, paragraph 380 and the case-law cited).
1705Accordingly, the first limb of UBS’s second plea, in so far as it alleges that the basic amount of the fine was not determined in accordance with the Guidelines, must be rejected as unfounded.
1706In so far as UBS additionally claims, also in the first limb of its second plea, that the Commission failed to comply with the Guidelines by using notional amounts as the basis for calculating the proxy or by not using a value of sales calculated in accordance with the rules of EU law relating to the activities of financial institutions, it should be pointed out that those objections, also set out in its first plea, are directed against the methodology used by the Commission. They will therefore be examined in paragraphs 1786 to 1800 below, in the response to the pleas alleging errors in the determination of the proxy.
(2)The pleas alleging errors in the determination of the basic amounts of the fines imposed on UBS, UniCredit and Nomura
1707As recalled in paragraphs 1643 and 1644 above, the basic amounts of the fines imposed on UBS, Nomura and UniCredit were calculated by the Commission as follows:| Proxy | Gravity factor | Duration multiplier | Variable component of the basic amount | Additional amount for deterrence | Basic amountUBS | 298490616 | 18% | 4.83 | 259686836 | 53728311 (18%) | 313415000UniCredit | 350132029 | 17% | 0.17 | 9920407 | 59522445 (17%) | 69442000Nomura | 392646351 | 18% | 0.83 | 58896953 | 70676343 (18%) | 129573000
1708In their pleas relating to the calculation of the basic amount of their respective fines, UBS, UniCredit and Nomura, first of all, argue that the Commission made errors in the determination of their respective proxies.
1709Next, UniCredit and Nomura submit that the Commission made errors in the determination of their respective gravity factors.
1710Lastly, UniCredit submits that the Commission applied an incorrect rate when setting the additional amount imposed as a deterrent.
(i)The pleas alleging errors in the determination of the proxy
1711As recalled in paragraph 1643 above, the Commission considered that, for the purpose of determining the amount of the fine for each of the banks concerned, it was appropriate to calculate a proxy.
1712In view of the specific nature of EGBs, the Commission stated, in recitals 815 and 816 of the contested decision, that it was appropriate to take, as the starting point for calculating the proxy, the notional volume and value (‘the notional amounts’) of the EGBs that the banks concerned traded on the secondary market during their individual period of participation in the infringement at issue, not including those traded on the primary market to avoid the risk of those amounts being counted twice.
1713To that end, as is apparent from recitals 817 and 819 of the contested decision, the Commission used the notional amounts of the EGBs traded by each bank concerned with counterparties located in the EEA during its period of participation in the infringement at issue, as follows:‘Thus, the notional amounts traded during the period of individual involvement in the cartel are annualised by a factor in function of the duration of the individual participation by dividing the total amount by the number of months of participation and subsequently multiplying this monthly average by 12.’
1714The Commission thus decided not to use the notional amounts traded by each bank concerned in the last full year of its participation in the infringement at issue, and departed from point 13 of the Guidelines (see recitals 817 and 819 of the contested decision).
1715In recitals 821 and 822 of the contested decision, the Commission then noted that EGBs were traded on the secondary market for a price that is expressed as a percentage of their notional amount. It deduced that the revenue on those transactions was reflected in the difference between the purchase price and the sale price of each EGB acquired and then sold by the traders (bid-ask spread), and that it was appropriate to take into account the bid-ask spread for the purpose of calculating the proxy.
1716To that end, the Commission considered it appropriate to multiply the notional amounts used for each bank concerned by a factor that takes account of the bid-ask spread levels (‘the adjustment factor’), namely an average bid-ask spread specific to each bank reduced by 50% to take account of the fact that each buy or sell transaction carried out by a bank represents only half of a combined buy and sell transaction for a specific EGB (see recitals 822 to 824 of the contested decision).
1717To arrive at the adjustment factor specific to each bank concerned, the Commission, first, chose to use 32 categories of representative EGBs, issued by eight issuers and divided into four remaining maturity ranges (0-3 years, 3-7 years, 7-10 years and above 10 years), taking into account the comments made by the banks concerned following the Letter on Fines (see recitals 827 and 846 of the contested decision).
1718The Commission thus applied, to each bank concerned, a specific weight for each of those 32 categories of EGBs, which it set out in Table 2 of Annex 2 to the contested decision and which, by way of example, for UniCredit, took the following form:
1719Second, for each bank concerned and for each day of its participation in the infringement at issue, the Commission calculated a daily bid-ask spread, in accordance with the methodology set out in recital 826 of the contested decision, as follows:
1720Therefore, for each bank concerned, that daily bid-ask spread consisted in an average of the daily bid-ask spreads of each of the 32 categories of EGBs selected, weighted according to the notional amounts traded by each bank per issuer and per maturity (see recital 828 of the contested decision).
1721For the purpose of determining the bid-ask spread for each category of representative EGBs, the Commission used public data from Bloomberg, which it considered to be the best figures available (see recitals 828 and 852 of the contested decision).
1722That platform shows the average end-of-day spread for each category of representative EGBs, calculated using the Bloomberg BGN composite price source (‘the BGN data’), which is based on executable and indicative quotes from multiple dealers (see recital 828 of, and footnote 1410 to, the contested decision).
1723Third, the Commission calculated, for each bank concerned, the final bid-ask spread as the simple average of daily bid-ask spreads computed for each day of its participation in the infringement at issue (1278 days for UBS, 57 days for UniCredit and 225 days for Nomura), as set out in paragraph 1719 above, in accordance with the methodology set out in recital 825 of the contested decision, as follows:
1724In recitals 831 and 832 of the contested decision, the Commission thus applied, for UBS, a final bid-ask spread of 0.070% and therefore an adjustment factor of 0.035%, rounded down to the third decimal place, corresponding to a proxy of EUR 298490616. For UniCredit, it applied a final bid-ask spread of 0.169% and therefore an adjustment factor of 0.084%, rounded down to the third decimal place, corresponding to a proxy of EUR 350132029. For Nomura, it applied a final bid-ask spread of 0.134% and therefore an adjustment factor of 0.066%, rounded down to the third decimal place, corresponding to a proxy of EUR 392646351.
1725The details of those calculations are set out in a confidential Annex 2 to the contested decision that is specific to each bank concerned and was notified only to the relevant bank.
1726It is in that context that UBS, UniCredit and Nomura object to the proxies applied to them.
1727First, UBS, UniCredit and Nomura claim that the Commission applied a proxy which bears no correlation to their economic activity.
1728Second, UBS claims that the Commission erred in law by failing to apply Directive 86/635 when calculating the proxy applied to it.
1729Third, UBS, UniCredit and Nomura claim that the Commission made errors in the formulation of the methodology for calculating the proxy and in its general implementation.
1730Fourth, UniCredit and Nomura criticise the Commission for errors in the calculation of the proxy for reasons relating to their specific situation.
–The claims made by UBS, UniCredit and Nomura that the proxies applied to them bear no correlation to their economic activity
1731UBS, in its first two pleas, and Nomura, in the first limb of its sixth plea, argue, in essence, that the decision to use the notional amounts of the EGBs traded by the banks concerned as the starting point for calculating the proxy led the Commission to apply a proxy which does not represent their economic activity.
1732Those notional amounts, it is submitted, are not an indicator of price, but of volume. The value of the services supplied by the banks concerned, corresponding to the difference between the purchase and sale prices, should be used, rather than the underlying value of the products traded to which a very large adjustment factor is then applied.
1733Nomura also argues that the Commission was not entitled to use the notional amounts relating to the trades of EGBs which it had effected on the secondary market when the Commission had not established its participation in an infringement on that market. Together with UniCredit, which puts forward the argument in its seventh plea, Nomura also maintains that the Commission was likewise not entitled to use the notional amounts covering the entire EGB sector, given the limited number of EGBs and auctions concerned by the discussions relied on against them.
1734In that regard, it should be recalled that, for the purpose of determining the proxy, the Commission used a two-step methodology, which, for each bank concerned, resulted in the Commission, first, determining the notional amounts traded on the secondary market during their respective period of participation in the infringement at issue and, second, applying an adjustment factor to those amounts calculated taking into account the fact that, as noted in recital 812 of the contested decision, ‘financial products such as EGB do not generate sales in the usual sense, as they are both bought and sold by the dealers and revenues are derived from the difference between the purchase price and the sale price of each bond acquired and then sold by the traders.’
1735It follows that the Commission did in fact take an amount more representative of volume than of value as the starting point for calculating the proxy, which, moreover, it expressly acknowledged in its written pleadings.
1736However, the method for calculating the adjustment factor, which is set out in paragraphs 1711 to 1724 above and was applied to the notional amounts of each bank concerned, enabled the Commission to use, as the proxy, an amount which constitutes an indicator of value, even though the starting point for calculating that proxy is an indicator of volume.
1737Applying half the average spread between their purchase and sale prices (the final bid-ask spread) to the notional amounts traded by each bank concerned makes it possible to identify the amounts on which the banks concerned generate revenues or losses.
1738In that regard, UBS cannot meaningfully criticise the Commission for having used 50% of the final bid-ask spread and not a lower rate, on the ground that that rate ‘is [admittedly] theoretical and based on the best possible outcome, but is not based on any actual data’, since the Commission could not base its methodology for calculating the fine on only part of the – albeit extrapolated – activity of the banks concerned.
1739In this way, the proxy reflects the activity of the banks concerned, regardless of whether that activity is profitable or loss-making.
1740The mere fact that determining that proxy led the Commission to use notional amounts corresponding to several hundred billion euros and then to apply an adjustment factor reducing them by more than 99% cannot, in itself, mean that that proxy bore no correlation to the economic activity of the banks concerned.
1741Similarly, that same fact is likewise not such as to discriminate, as UBS submits, against financial institutions operating in the EGB sector as compared with undertakings from other economic sectors, in that the former will be exposed to higher fines than the latter.
1742The calculation methodology used by the Commission is merely the consequence of the uncontested particularities of the EGB sector, which do not generate sales in the usual sense (see recital 812 of the contested decision) and which, as UBS has stated, leads to ‘the annualised notional volume of EGB trade [by that bank being] almost 20 times greater than the average value of sales in any other sector’.
1743Therefore, UBS and Nomura are wrong to maintain that using the notional amounts traded by each bank concerned as the starting point for calculating the proxy led the Commission to use a value that bore no relation to the economic activity of those banks, in breach of the principle of equal treatment.
1744Nomura is also wrong to argue that the Commission used notional amounts for Nomura that went beyond the scope of its participation in the infringement at issue, with the result that its proxy bore no correlation to its economic activity.
1745As follows from the rejection of Nomura’s third to fifth pleas in paragraph 1400 above, the Commission was entitled to find that, for the period from 18 January to 28 November 2011, that bank had participated in all the constituent elements of the single and continuous infringement at issue, which covered the entire EGB sector – that is to say both the primary and second markets for those bonds – and the whole of the EEA.
1746As a result, the Commission was fully entitled to use, as regards Nomura, the notional amounts relating to the EGBs traded by that bank on the secondary market.
1747For the same reason, contrary to Nomura’s contention, the Commission was entitled to include, in the notional amounts used, amounts relating to EGBs which admittedly it did not actually trade or amounts relating to auctions in which it did not participate, provided that they related to EGBs that came within the scope of the infringement at issue.
1748Unless the economic significance of the infringement committed by a particular undertaking is artificially minimised and the Commission is therefore led to impose fines which bear no actual relation to the scope of the cartel in question, the proxy – like the value of sales – cannot be calculated solely on the basis of transactions which are shown to have been actually affected by that cartel (see, by analogy, judgment of 12 November 2014, Guardian Industries and Guardian Europe v Commission, C‑580/12 P, EU:C:2014:2363, paragraphs 57 and 58), but may, as in the present case, be calculated on the basis of all the transactions coming within the scope of that cartel (see, to that effect, judgment of 16 June 2011, Team Relocations and Others v Commission, T‑204/08 and T‑212/08, EU:T:2011:286, paragraph 64).
1749In the present case, such a decision is all the more justified because, in recital 841 of the contested decision, the Commission stated, without being contradicted on this point by Nomura, that ‘discussion about specific EGB [could] have an impact on the pricing and strategies for other EGB not explicitly discussed’.
1750In the light of the foregoing, the objections raised by UBS in its first and second pleas, by UniCredit in its seventh plea, and by Nomura in the first limb of its sixth plea, to the effect that the proxy applied to them bore no relation to their economic activity, must be rejected as unfounded.
–UBS’s claim that the Commission was required to apply the methodology for determining turnover laid down by Directive 86/635
1751In its first two pleas, UBS submits that, by using the notional amounts traded by the banks concerned as the starting point for calculating the proxy, the Commission infringed the lex specialis laid down for measuring the value of the economic activity of financial institutions in the European Union, namely Directive 86/635, and thereby erred in law.
1752According to UBS, that directive provides a specific methodology for calculating the turnover of financial services intermediaries.
1753In that regard, it should be pointed out that Directive 86/635 does provide for coordination of the layout of the annual and consolidated accounts of credit and financial institutions. However, that directive is intended only to ensure ‘improved comparability of the annual accounts and consolidated accounts of these institutions’ and is in no way intended to impose, particularly on the Commission, a specific methodology for the purpose of determining the value of sales of such institutions, in the context of calculating the fines that the Commission imposes on them for infringement of EU competition rules.
1754That conclusion is supported by case-law, in accordance with which, where the Commission departs in whole or in part from the Guidelines – as in the present case in relation to the determination of the value of sales for the purposes of point 13 of those guidelines – it enjoys, within the limits laid down by Article 23(2), (3) and (4) of Regulation No 1/2003, a broad discretion as regards the method for calculating fines imposed for infringements of EU competition rules (see, to that effect, judgments of 12 November 2014, Guardian Industries and Guardian Europe v Commission, C‑580/12 P, EU:C:2014:2363, paragraph 55 and the case-law cited, and of 10 July 2019, Commission v Icap and Others, C‑39/18 P, EU:C:2019:584, paragraph 25 and the case-law cited).
1755The fact that the Commission requested UBS’s accounting data drawn up in accordance with Directive 86/635 in order to examine whether the maximum fine provided for by Article 23(2) of Regulation No 1/2003 could be applied to UBS, or the fact that Article 5(3)(a) of Council Regulation (EC) No 139/2004 of 20 January 2004 on the control of concentrations between undertakings (OJ 2004 L 24, p. 1) provides that, for credit institutions and other financial institutions, the Commission is to use, in place of turnover, ‘the sum of [certain] income items as defined in [Directive 86/635]’, does not call into question the conclusion reached in the preceding paragraph.
1756The purpose of using the methodology for calculating the turnover of credit institutions and other financial institutions set out by Directive 86/635, for the purposes of the two provisions referred to in the preceding paragraph, is to determine, by the sum of a certain number of specified income items, ‘the economic significance of a financial institution’, as UBS states.
1757By contrast, for the purpose of determining the fine to be imposed on the banks concerned in the present case, the Commission seeks to determine the volume not of the entire activity of the financial institution concerned, but only of part of its activity, which it is not disputed has specific features such as that it necessitates recourse not to nominal amounts of trades in EGBs, but to a proxy.
1758Furthermore, it is not clear within which of the income items referred to by Directive 86/635 the EGB trading activity of the banks concerned might come. In that regard, UBS maintains that it is the ‘net profit on financial operations’. However, taking into account in particular the participation of all or some of the banks concerned in syndications giving rise to the payment of underwriting fees (see recital 32 of the contested decision), the ‘commissions receivable’ income item might also be relevant.
1759Lastly, UBS cannot meaningfully rely on the practice of national competition authorities in calculating fines imposed for infringements of competition rules, which is not binding on the Commission (see, to that effect and by analogy, judgment of 18 December 2007, Cementbouw Handel & Industrie v Commission, C‑202/06 P, EU:C:2007:814, paragraph 56).
1760In the light of the foregoing, UBS’s objections based on an infringement of Directive 86/635, contained in its first and second pleas, must be rejected as unfounded.
1761Furthermore, in so far as, in its first two pleas, UBS maintains that it would have been more appropriate for the proxy applied to the banks concerned to have been calculated using the methodology laid down in Directive 86/635, which presupposes the use of the ‘net value traded’ or the ‘adjusted net value traded’, that criticism is separate from that of the mandatory nature of the use of that methodology and will be examined in paragraphs 1786 to 1800 below, in the examination of the fourth plea raised by that bank.
–The pleas raised by UBS, UniCredit and Nomura alleging errors in the methodology for calculating the proxy and in its general implementation
1762As recalled in paragraphs 1717 to 1724 above, the Commission considered it appropriate, for the purpose of determining the basic amount of the fine, to have recourse to a proxy, determined by multiplying the annualised notional amounts of EGBs traded by each bank concerned during its period of participation in the infringement at issue by an adjustment factor, calculated by using 50% of the final bid-ask spread specific to each of those banks, the final bid-ask spread itself calculated as a simple average of the weighted daily bid-ask spread of 32 categories of representative EGBs traded during the period of each bank’s participation in that infringement.
1763UBS, in the second limb of its first plea, taken together with its fourth and fifth pleas, UniCredit, in the second to fourth limbs of its seventh plea, and Nomura, in the second to fourth limbs of its sixth plea, claim that the Commission failed to use the ‘best available figures’, within the meaning of point 15 of the Guidelines, for the purpose of calculating the proxy.
1764In support of their claims, UBS, UniCredit and Nomura object to both the volume of the notional amounts used and the method for calculating the adjustment factor.
1765As regards the volume of the notional amounts, UniCredit submits that the Commission was wrong to use the volume of all secondary market trades in EGBs. In so doing, it failed to take account of the market-making activities carried out by the banks concerned, which expose them to losses in the event of a fall in prices and which require them to enter into hedging trades which, in most cases, do not give rise to a spread.
1766As regards the calculation of the adjustment factor, and more specifically the selection and weighting of the 32 categories of EGBs used by the Commission, UniCredit submits that the use of such a selection resulted in the Commission excluding certain issuing countries and using the spreads of EGBs issued by France as a proxy for those of EGBs issued by the Netherlands, Austria and Finland (footnote 1413 to the contested decision).
1767According to UniCredit, the Commission was also wrong to divide the EGBs selected per issuing Member State into maturity ranges defined not by reference to their maturity at issuance, but by reference to their remaining maturity, which renders those ranges non-homogeneous.
1768In addition, UBS argues that the Commission was wrong to rely on a selection of EGBs the weighting of which was constant throughout the duration of the infringement at issue, whereas that weighting should have been redefined monthly or annually for each maturity group and each issuer in order to take account of changes in the activity of the banks concerned. The Commission had those data and could not hide behind the increased complexity of that methodology, as it did in recital 850 of the contested decision.
1769As regards the calculation of the adjustment factor still, but now as to the data used to calculate the final bid-ask spread, UBS, UniCredit and Nomura submit that the Commission could not legitimately use the BGN data to determine the daily bid-ask spread for each of the 32 categories of EGBs.
1770In addition to the fact that the generation of the BGN data is opaque, those data constitute end-of-day bid-ask spreads, which take account of indicative prices only. It is apparent from the Commission’s decisional practice, from economic literature and from two economic studies submitted by Nomura concerning transactions executed by that bank that an actual spread is commonly much narrower than the quoted spread, particularly taking into account the bilateral negotiations taking place between traders. Moreover, the BGN data reflect neither intra-day movement in quoted bid-ask spreads, as economic literature confirms, nor the impact that the short or long positions of market makers have on the spreads actually obtained. On the contrary, according to Nomura, the BGN data tend to reflect a period of the day when spreads widen, as ‘certain academic literature’ confirms.
1771As a result, UniCredit submits that, instead of the bid-ask spreads derived from the BGN data, the Commission should have taken the bid-ask spreads from data from the MTS platform (‘the MTS data’) – wrongly dismissed in footnote 1449 to the contested decision – which constitute a solid proxy for the mid-prices.
1772Consequently, UBS, UniCredit and Nomura submit that the proxy applied to them by the Commission is imprecise, arbitrary or unrelated to their real economic activity, reveals a manifest error of assessment and is therefore contrary to the Guidelines.
1773In that regard, UBS adds that the Commission’s methodology led it deliberately to disregard the features inherent in EGB trading activity (see recital 854 of the contested decision). That activity should be viewed as a whole and over the average period during which it holds open positions in EGBs (preferably one month and, alternatively, 10 days or one week), in order to take account of the opening and closing of portfolios, cross-instrument hedging strategies, the distinction between portfolio management and the provision of trading services and the competitive pressure exerted on the banks concerned. In addition, the Commission relies on assertions relating to the market for EGBs which it has not substantiated, in particular in recital 854 of the contested decision.
1774In those circumstances, UBS, UniCredit and Nomura submit that the Commission should have used a methodology based on their own transaction data, calculated on the basis of the ‘net value traded’ by each bank concerned, as defined by Directive 86/635. In the alternative, UBS submits that the Commission should have used the ‘adjusted net value traded’, that is to say, a methodology based on the ‘net value traded’ but in which negative results from its trades are converted into positive results.
1775For UBS, UniCredit and Nomura, those two alternative methodologies do not constitute a measurement of profit, as the Commission stated in recitals 814 and 839 of the contested decision. In that regard, Nomura submits that the ‘net value traded’ and the ‘adjusted net value traded’ reflect the actual gains and losses on EGB trades, which is in line with international accounting standards, ‘under [which] these netted off figures are considered revenue and not profit’. Similarly, UBS submits that ‘P&L/revenue is a measure of income from [the traders], not profit, as costs have not been deducted from this measure’.
1776In any event, UniCredit takes the view that the Commission should have requested additional data during the administrative procedure had it considered that the data submitted by the banks concerned were insufficient.
1777In that regard, first, as pointed out, in essence, in paragraph 1754 above, it is settled case-law that the Commission enjoys a broad discretion as regards the method for calculating fines for infringements of EU competition rules.
1778While it is true that the Commission limited itself in the exercise of that discretion by adopting the Guidelines, it is entitled to depart from them, provided that it gives reasons for doing so (see, to that effect, judgment of 10 July 2019, Commission v Icap and Others, C‑39/18 P, EU:C:2019:584, paragraph 29 and the case-law cited).
1779In that regard, it must be noted that the method set out by the Guidelines, which is based on taking into account the value of sales of the goods concerned to which the infringement relates in determining the basic amount of the fines to be imposed, may sometimes prove unsuited to the particular circumstances of a case and that, in such a situation, the Commission is entitled to have recourse to a calculation method other than that set out in the Guidelines (see, to that effect, judgment of 10 July 2019, Commission v Icap and Others, C‑39/18 P, EU:C:2019:584, paragraph 27).
1780That discretion allows the Commission to depart entirely from the methodology determined in the Guidelines, on the basis of point 37 thereof, but also to depart from the Guidelines only in certain respects, by adopting a different methodology from that set out in point 13 of those guidelines and, therefore, by calculating a proxy (see, to that effect, judgments of 7 September 2016, Pilkington Group and Others v Commission, C‑101/15 P, EU:C:2016:631, paragraphs 20 to 23, and of 20 December 2023, Crédit agricole and Crédit agricole Corporate and Investment Bank v Commission, T‑113/17, under appeal, EU:T:2023:847, paragraphs 474 to 476).
1781Second, it should be noted that point 15 of the Guidelines requires the Commission to use the ‘best available figures’ in order to determine the value of an undertaking’s sales, as defined in point 13 of those guidelines, and subject to point 16 of those guidelines.
1782Consequently, in the context of the present case, where the Commission has chosen to depart from point 13 of the Guidelines and has therefore refrained from determining the basic amount of the fine by taking the value of the sales of services made by the banks concerned to which the infringement directly or indirectly relates, point 15 of those guidelines cannot be directly applicable.
1783However, where, in the exercise of its broad discretion in determining the methodology for calculating a fine, the Commission has chosen to depart from the Guidelines not in their entirety – as point 37 thereof authorises it to do – but only, as in the present case, from point 13 thereof, it cannot depart from the guiding principles or underlying logic of those guidelines (see, to that effect, judgment of 24 September 2019, HSBC Holdings and Others v Commission, T‑105/17, EU:T:2019:675, paragraphs 324 and 345) and, in particular, from the principle expressed in point 15 thereof.
1784Thus, in implementing the methodology which it lays down, it is incumbent upon it, inter alia, to ensure that it takes account of the best available figures, subject to detailed review, in law and in fact, by the EU Courts (see, to that effect, judgments of 8 December 2011, Chalkor v Commission, C‑386/10 P, EU:C:2011:815, paragraph 62, and of 24 September 2019, HSBC Holdings and Others v Commission, T‑105/17, EU:T:2019:675, paragraph 348).
1785In the present case, it is therefore for the Court to determine whether the methodology adopted by the Commission, as recalled in paragraphs 1711 to 1724 above, more appropriately reflects the activity of the banks concerned than the alternative methodologies proposed by those banks (see, to that effect, judgment of 24 September 2019, HSBC Holdings and Others v Commission, T‑105/17, EU:T:2019:675, paragraph 324).– The alternative methodologies proposed by UBS, UniCredit and Nomura
1786The alternative methodology proposed primarily by UBS as well as by UniCredit and Nomura is based on the ‘net value traded’ of the banks concerned, as defined by Directive 86/635. According to those banks, that ‘net value traded’ is capable of constituting an appropriate proxy for reflecting the economic importance of the infringement at issue and the relative weight of each bank in that infringement, for the purposes of point 6 of the Guidelines.
1787In recitals 814 and 839 of the contested decision, the Commission expressly rejected that alternative methodology on the ground that (i) it was based on the revenues of the banks concerned; (ii) ‘trading revenue involves, in essence, taking into account both profitable and loss-making transactions during the infringement period, and constitutes a figure which is comparable to the profit derived from trading activities’; (iii) ‘such a limitation would run counter to the logic applied in the methodology in the Guidelines’; and (iv) ‘the proxy aims to measure the economic activity of an undertaking in the sector concerned and not how successful this undertaking was in that activity’.
1788Those findings do not demonstrate any error on the part of the Commission and cannot be called into question by the arguments put forward by UBS, UniCredit or Nomura.
1789In the first place, as the Commission rightly pointed out, the methodology based on the ‘net value traded’ is not capable of reflecting the economic activity of the banks concerned in relation to the infringement at issue.
1790First, that methodology, which is intended to capture both the losses and profits from the EGB trading activity of the banks concerned, entails negative spreads being deducted from positive spreads.
1791As a result, even though those negative spreads are the result of an activity – admittedly loss-making, but nevertheless a genuine EGB trading activity coming within the scope of the infringement at issue – they reduce the amount of the ‘net value traded’, thereby reducing the proxy that UBS, UniCredit and Nomura propose be used as a basis for calculating the basic amount of the fine to be imposed on them.
1792Second, the methodology based on the ‘net value traded’ means that trades which come within the scope of the infringement at issue but which do not generate any spread are not taken into consideration even though such trades constitute a substantial proportion of the trades effected by the banks concerned.
1793In the second place, the fact that the costs associated with EGB trading activities are not deducted from the amounts used as the ‘net value traded’ did not prevent the Commission from taking the view, in recitals 814 and 839 of the contested decision, that the methodology based on the ‘net value traded’ was ‘comparable to the profit derived from trading activities’.
1794Apart from the fact that, in their pleadings, UBS and Nomura implicitly concede that the ‘net value traded’ constitutes ‘revenues’ or ‘gains’ made by the traders (see paragraph 1775 above), it is clear that, by recitals 814 and 839, the Commission was not seeking to establish that that value constituted profits in the accounting sense, but was essentially seeking to highlight the fact that that value did not constitute an ‘appropriate proxy for the value of sales under the Guidelines’.
1795In the third place, the fact that international accounting rules, Directive 86/635, Article 5(3)(a) of Regulation No 139/2004 and national competition authorities have recourse to different methodologies for assessing the activity of financial institutions cannot call into question the Commission’s assessment, since those texts and practices were not binding on the Commission in the specific context of calculating fines imposed under Article 101 TFEU, as found in paragraphs 1753 to 1759 above.
1796The finding made by the Commission in recitals 814 and 839 of the contested decision applies equally to the methodology suggested in the alternative by UBS, which is based on the ‘adjusted net value traded’.
1797That second alternative methodology does not remedy the inherent weaknesses in a methodology which, in any event, leads to a volume of activity being determined by netting profitable transactions against loss-making transactions.
1798It may be inferred from UBS’s pleadings, which refer to an economic analysis not provided to the Court, and from the Commission’s pleadings, that the methodology based on the ‘net value traded’ results in the conversion into positive results not of all loss-making trades, but only the negative net trading income over a chosen reference period (preferably one month and, alternatively, 10 days or one week) – which, moreover, does not coincide with the average duration of UBS’s open positions, which, according to UBS, was 11.2 days during its period of participation in the infringement at issue.
1799Furthermore, that methodology – in respect of which there is no need in the present case to assess the mathematical soundness or even the appropriateness of the reference periods – likewise does not enable account to be taken of the activity of the banks concerned which comes within the scope of the infringement at issue and which gives rise to trades that do not generate any spread.
1800Therefore, UBS, UniCredit and Nomura cannot reasonably criticise the Commission for considering that the methodologies based on the ‘net value traded’ or the ‘adjusted net value traded’ were not appropriate for the purpose of implementing the Guidelines in the present case and, therefore, for refusing to use them to calculate the proxy for each of those banks.– The appropriateness of the methodology adopted by the Commission
1801It has already been found in paragraphs 1737 and 1739 above that, notwithstanding the fact that the proxy was based on the notional amounts traded by the banks concerned, the application to those amounts of an adjustment factor calculated on the basis of the bid-ask spreads of the EGBs traded by each bank concerned made it possible to identify the amounts on which those banks generated profits or losses and, therefore, reflected the activity of the banks concerned as intermediaries, regardless of whether that activity was profitable or loss-making.
1802Therefore, it is necessary to examine whether the methodology adopted by the Commission is vitiated by manifest errors of assessment such as to call into question its appropriateness for the purpose of determining the proxy used with respect to UBS, UniCredit and Nomura.The use of all notional amounts traded by the banks concerned
1803As regards the volume of the notional amounts used by the Commission, UniCredit cannot reasonably criticise the Commission for having used, as the basis for calculating the adjustment factor, the notional amounts of all EGBs traded with counterparties located in the EEA during its period of participation in the infringement at issue.
1804There was no reason for the Commission to deduct the notional amounts that had given rise to negative spreads from the notional amounts traded by UniCredit.
1805Like the value of sales, the proxy is not a measure of the net profit of the undertakings on which the Commission imposes fines under Article 101 TFEU (see, to that effect, judgment of 15 April 2021, Italmobiliare and Others v Commission, C‑694/19 P, not published, EU:C:2021:286, paragraph 99).
1806Similarly, there was no reason for the Commission to disregard trades which resulted in a spread of zero, on the ground that they constituted hedging trades executed as a market making activity.
1807First of all, UniCredit does not explain how to distinguish hedging trades from trades executed in the context of providing trading services. Next, none of the material provided by that bank establishes that only hedging trades are capable of generating spreads of zero. Lastly, in any event, and as the Commission rightly stated in recital 854 of the contested decision, hedging trades form an integral part of the economic activity of the banks concerned.The calculation of the adjustment factor
1808As regards the calculation of the adjustment factor, the objections raised by UBS, UniCredit and Nomura concern, first, the Commission’s use of a selection of 32 categories of EGBs – issued by eight issuing Member States and divided into four maturity ranges – and, second, the use of the BGN data to calculate the daily bid-ask spread for each of the more than 100 EGBs making up those 32 categories of EGBs.
1809Concerning, in the first place, the selection of the 32 categories of EGBs chosen by the Commission, UniCredit objects to, first of all, the very principle of using a selection of Member States which issue EGBs (see recital 827 and footnote 1413) and not all of them.
1810In that regard, it should be recalled, first, that the method set out by the Guidelines, which is based on taking into account the value of sales of the services concerned to which the infringement relates in determining the basic amount of the fines to be imposed, may sometimes prove unsuited to the particular circumstances of a case and that, in such a situation, the Commission is entitled to have recourse to a calculation method other than that set out in the Guidelines (see, to that effect, judgment of 10 July 2019, Commission v Icap and Others, C‑39/18 P, EU:C:2019:584, paragraph 27).
1811Second, as already pointed out in paragraphs 1754 and 1777 above, the Commission enjoys a broad discretion as regards the methodology for calculating fines for infringements of EU competition rules.
1812It follows that, provided that it not only gives reasons for its choice but also justifies it to the requisite legal standard, the Commission cannot be deprived of the possibility, for the purpose of determining the basic amount of a fine, of using a methodology which takes account of samples which are representative of the activities of the undertakings concerned.
1813The fact that the data obtained from those representative samples are not identical to the data obtained from the analysis of their own trades is inherent in the methodology used and cannot prevent the Commission from using such a calculation methodology, provided that the samples taken into consideration are indeed representative.
1814In that respect, it must be pointed out that case-law accepts that the value of sales referred to in point 13 of the Guidelines is not a perfectly accurate reflection of the scale of the infringement at issue, and the same applies to a proxy.
1815While approving the methodology laid down by the Commission in point 13 of the Guidelines, the Court of Justice has acknowledged, first, that the proportion of turnover derived from the goods or services which are the subject of the infringement, covered by point 13, is only of such a kind as to give an indication of the scale of the infringement and to reflect the economic importance of that infringement and, secondly, that it must not be given an importance that is disproportionate (see, to that effect, judgment of 19 March 2009, Archer Daniels Midland v Commission, C‑510/06 P, EU:C:2009:166, paragraphs 74 and 76 and the case-law cited).
1816Along the same lines, the General Court has held that setting a fine under Article 101 TFEU is not a precise arithmetical exercise (judgments of 5 October 2011, Romana Tabacchi v Commission, T‑11/06, EU:T:2011:560, paragraph 266, and of 15 July 2015, SLM and Ori Martin v Commission, T‑389/10 and T‑419/10, EU:T:2015:513, paragraph 436).
1817In the present case, in recitals 825 and 826 of, and footnotes 1410 and 1411 to, the contested decision, the Commission stated that ‘the final bid-ask spread is calculated for each [bank concerned] as the simple average of daily bid-ask spreads computed for each working day of [the infringement period of each bank concerned]’ and that ‘each daily bid-ask spread is calculated individually for each [bank concerned] and consists of the weighted average of the bid-ask spread level of 32 [categories of] representative bonds to which 32 specific weights are assigned’.
1818In recital 827 of, and footnotes 1412, 1414 and 1415 to, the contested decision, the Commission stated that ‘the 32 representative EGB were selected from eight issuing [Member States] spread over four maturity ranges’. In that regard, it specified that ‘the representative sample of eight issuing [Member States] included Belgium, Germany, Ireland, Greece, Spain, France, Italy and Portugal. EGB were selected with a remaining maturity of 2, 5, 10 and 30 years, for the 0-3 years, the 3-7 years, the 7-10 years and above 10 years maturity ranges respectively’, since ‘those maturities [were] very common in the bond market and [were] widely considered as the “benchmark” maturities by the investor community’ and ‘also [corresponded] to the ones published by the majority of the DMOs in their yearly issuance calendar’.
1819Applying that weighting, the Commission provided to each bank concerned, in Annex 2 to the contested decision, its own specific weighting table, such as that reproduced in paragraph 1718 above.
1820Similarly, in Annex 2, the Commission provided Tables 3 to 6, which, for each of the 32 categories of EGBs selected, and for UBS, by way of example, appear as follows:
1821It is also apparent from recital 827 of, and footnote 1413 to, the contested decision that the Commission’s selection of 8 issuing Member States corresponds – by incorporating the EGBs issued by the Netherlands, Austria and Finland with those issued by France – to 11 Member States of the euro area, which, at the material time, consisted of 17 Member States.
1822It is implicit that six Member States were excluded from that selection, in respect of which the Commission stated, without being challenged by UniCredit, that there were no EGBs traded in one of them, Luxembourg, and that in others (Estonia, Cyprus, Malta, Slovenia and Slovakia) the notional amounts traded were ‘very limited’ (footnote 1413 to the contested decision).
1823By that reasoning, which enabled the banks concerned to become fully aware of how the 32 categories of representative EGBs were chosen, the EGBs selected in each of those 32 categories and whether or not they were ‘on the run’, the Commission also established the representative nature of the selection made, which, moreover, had to be identical for all of the banks concerned in order to ensure that the fine reflected the relative weight of each bank concerned in the infringement at issue, in accordance with point 6 of the Guidelines.
1824The fact that, in October 2011, 0.5% of UniCredit’s trades related to EGBs issued by Slovakia and were not, therefore, taken into consideration, whereas, according to that bank, it did not trade EGBs issued by Ireland or Greece, which, however, were included in the Commission’s selection, cannot call into question the representative nature of that selection, which represented more than 90% of the notional amounts of EGBs traded by that bank in October 2011.
1825The selection made by the Commission was all the more appropriate since it resulted in a weighting of the 32 categories of EGBs specific to each bank concerned, as shown by Table 2 of Annex 2 to the contested decision, reproduced in paragraph 1718 above.
1826That table shows, first, that the categories of EGBs selected by the Commission but not traded by UniCredit – such as Irish and Portuguese EGBs of all maturities – were not taken into account for the purpose of calculating the adjustment factor and, secondly, that the EGBs primarily traded by that bank – namely German and Italian EGBs – represent 26% and 61.6% of the weight matrix applied to that bank, respectively.
1827Similarly, UniCredit’s argument based on the grouping together of EGBs issued by the Netherlands, Austria and Finland with those issued by France is not capable of undermining the representative nature of the Commission’s selection.
1828In addition to the fact that UniCredit’s reply of 14 May 2018 to the Commission’s request for information of 27 March 2018 shows that the notional amounts of EGBs issued by those three Member States and traded by that bank were marginal, that bank confines itself to claiming that the grouping together of the EGBs issued by the Netherlands, Austria and Finland with those issued by France stemmed from the erroneous assumption that ‘bid-ask spreads were driven solely by credit ratings or that countries with similar credit ratings could be assumed to have identical bid-ask spreads’, whereas, ‘on the contrary, among other things, liquidity of an EGB played a much greater role in determining the relevant bid-ask spread’.
1829Next, in so far as UniCredit criticises the Commission, in the choice of the four maturity ranges (0-3 years, 3-7 years, 7-10 years and above 10 years), for having used the remaining maturity of the EGB categories and not their maturity on the date of issuance, it must be stated that that bank has not adduced any data, particularly numerical, to show that the bid-ask spreads of EGBs with the same initial maturity but a different remaining maturity are narrower than the bid-ask spreads of the EGBs selected by the Commission, that is to say, the spreads of EGBs with a different initial maturity but the same remaining maturity.
1830In addition, the Commission stated in its defence that ‘EGBs with similar remaining maturity offer investors (in the secondary market) the same investment horizon and have the same sensitivity to interest rate changes, regardless of their maturity at issuance’.
1831UniCredit’s mere assertion, in its reply, that the Commission’s statement referred to in the preceding paragraph ‘is in practice completely irrelevant since, here, those EGBs are being used as a proxy for bid-ask spreads’, is insufficient to call into question the plausibility of the Commission’s statement and, accordingly, to establish that, by dividing the EGBs of the eight Member States selected according to their remaining maturities, the Commission’s methodology for calculating the proxy was vitiated by a manifest error of assessment.
1832Lastly, UBS cannot reasonably criticise the Commission for having applied a weighting of the 32 categories of EGBs to the banks concerned that was constant for the entirety of their respective periods of participation in the infringement at issue, and not reset monthly or annually.
1833While the Commission’s decision not to perform that reset led it to apply a weighting of the 32 categories of EGBs traded that was less granular than that suggested by UBS, the fact remains that that weighting remained representative of the activity of the banks concerned during their respective periods of participation in the infringement at issue, in that it was specific to each bank and had been determined on the basis of the data provided by those banks.
1834In addition, as the Commission correctly stated in recital 850 of the contested decision, such a monthly or annual reset of the weighting applicable to each bank ‘significantly complicate[s] the methodology, by adding extra layers of detailed data gathering and calculations over thousands of trades, but do[es] not enhance the reasonable methodology in any meaningful way’.
1835Therefore, the Commission’s decision to apply a weighting of the 32 categories of EGBs to the banks concerned that was constant for the entirety of their respective periods of participation in the infringement at issue remains within the limits of the discretion available to the Commission in drawing up its methodology for calculating the proxy.
1836In the light of the above, UBS and UniCredit cannot reasonably maintain that the methodology used by the Commission to calculate the proxy, in so far as it is based on a selection of 32 categories of representative EGBs, reveals one or more manifest errors of assessment.
1837Concerning, in the second place, the use of the BGN data, it is apparent from recital 828 of, and footnote 1416 to, the contested decision that, for the purpose of calculating the final bid-ask spread, the method for calculating which was recalled in paragraph 1723 above, the Commission applied the BGN data to the 32 categories of representative EGBs.
1838In that regard, the Commission stated that the ‘Bloomberg Generic Price [was] a real-time composite price for corporate and government bonds, based on executable and indicative quotes from multiple dealers [and indicated] available consensus-forming prices’. It added that ‘historical BGN prices [were] end-of-day prices’.
1839In recitals 851 and 852 of the contested decision, the Commission dismissed the objections of the banks concerned, stating that, as ‘no intra-day bid and ask prices [were] available from the addressees for the period 2007-2011, [it considered] that the historic BGN bid and ask prices retrieved from Bloomberg [were] the best figures available that, being determined at the same moment (end of day) and hence allowing for the computation of a reliable bid-ask spread, enable[d] a fair and equitable reflection of the parties’ true values and economic importance’.
1840Lastly, in footnote 1449 to the contested decision, the Commission rejected UniCredit’s line argument as follows:‘In its submission of 15.02.2021 …, UniCredit compares the bid and ask prices from a sample of its trades with mid-prices derived from intra-day data of a trading platform (MTS) and claims that the spreads used by the Commission are wider than the spread actually earned by UniCredit. However, comparing a mid-price derived from one platform (MTS) with prices quoted by UniCredit is not consistent, as it compares data from two different sources (the result of which sometimes being negative) and hence cannot give any indication of the actual bid-ask spreads from UniCredit.’
1841In that regard, it has already been pointed out in paragraphs 1782 to 1784 above that, even where the Commission departs from point 13 of the Guidelines, it must ensure that it takes account of the best available figures.
1842It is therefore necessary to determine whether the BGN data used by the Commission to determine the final bid-ask spread constituted the ‘best available figures’ in implementing the methodology defined by the Commission.
1843In that context and having regard to the reasoning developed by the Commission in the contested decision (see recitals 828, 851 and 852), and, in particular, that by which it rejected the arguments on which UniCredit relied in the course of the administrative procedure (footnote 1449), the banks concerned cannot confine themselves to arguing that the data used by the Commission suffer from one or more shortcomings.
1844On the contrary, they must show that, within the framework of the methodology which the Commission has lawfully determined, there are in fact better data than those used by the Commission and that those data are in fact available (see, to that effect and by analogy, judgment of 24 September 2019, HSBC Holdings and Others v Commission, T‑105/17, EU:T:2019:675, paragraph 324).
1845In the present case, however, neither UBS, UniCredit nor Nomura have shown that the Commission might have had better data than the BGN data, even if the latter do not provide the spreads actually obtained by the bank concerned on the occasion of a combined buy and sell transaction for a specific EGB, but only an estimate derived from data based on ‘quoted bid-ask spreads’ at a specific point in time, namely at the end of the day.
1846First, even if it must be agreed with Nomura that ‘only the effective spread – i.e. the prices at which traders actually bought and later sold EGBs – … is the proper measure of the price of the intermediation services’, Nomura has not indicated that it offered to provide such data to the Commission during the administrative procedure; nor has it submitted those data in the context of the present proceedings. In the fourth limb of its sixth plea, Nomura merely criticised the Commission for not having used figures that are in accordance with international accounting standards and, therefore, the ‘net value traded’.
1847Moreover, the methodology based on the effective spreads obtained by the banks concerned would mean requiring the Commission to identify and apply individually, for each bank concerned, a yield spread specific to each transaction identified for each day on which that bank participated in the infringement at issue.
1848This would involve identifying, for each of the banks concerned and each purchase transaction, the corresponding sale transactions so as to enable the trade’s effective spread to be calculated, in a context where the purchase and sale transactions do not necessarily take place on the same day (see paragraph 1798 above), where the amount of those transactions is not necessarily the same, where that amount has an impact on the price of the EGBs and where the ‘volatility (including intraday) of the EGB sector’, confirmed by Nomura, makes it difficult, if not impossible, to match transactions taking place at different times.
1849Such a methodology – if it did not make it excessively difficult, if not impossible, to calculate the adjustment factor – would in any event place a disproportionate administrative burden on the Commission (see, by analogy, Opinion of Advocate General Kokott in Pilkington Group and Others v Commission, C‑101/15 P, EU:C:2016:258, point 37).
1850Therefore, neither UBS, UniCredit nor Nomura, which claim that the Commission did not ask them for their effective spreads in the course of the administrative procedure, can reasonably criticise the Commission for not having used or attempted to use a methodology to determine the effective spreads of all the transactions carried out by the banks concerned during their respective period of participation in the infringement at issue, and therefore for having used data that were ‘quoted’ by a third party.
1851Second, UBS, UniCredit and Nomura do not provide any evidence to show that complete and consistent sets of quoted data, other than the set consisting of the BGN data, are available and capable of more suitably reflecting the activity of the banks concerned.
1852In that regard, Nomura’s criticism that the way in which the BGN data are compiled is partly unknown, implying that the Commission was not entitled to use it, must be rejected at the outset.
1853It is well known and confirmed, in particular by recitals 40, 41, 505 and 518 of the contested decision, that data provided by platforms such as Bloomberg constitute data to which traders give a great deal of credence, in that such online platforms collect market information and then relay it to all market participants.
1854Moreover, as is apparent from footnote 1416 to the contested decision, which refers to a document drawn up by the Bloomberg platform, the BGN data constitute a source which is based on executable and indicative prices from multiple traders, which indicates the prices available in order to reach a consensus and which, moreover, as the Commission points out, is drawn up by a third party to the procedure.
1855Concerning, next, the claim that the BGN data are liable to overestimate the bid-ask spreads due to the time of day at which they are determined, it must be stated that that claim does not appear to be unfounded, in the light of the economic reports and academic articles produced by the banks concerned.
1856However, that finding cannot, in the present case, deprive the Commission of the possibility of using the BGN data nor, as a result, invalidate the amounts of the adjustment factors calculated by using them.
1857First, the Commission cannot be criticised for having used data which did not accurately reflect the actual spreads at which the transactions were carried out by the banks concerned, when the Commission did not have or could not have had actual data or sufficiently representative data, as is apparent from paragraph 25 of the Letter on Fines, and was therefore obliged to use a methodology based on alternative data which were necessarily less accurate in order to put together a proxy.
1858Indeed it is in this respect that point 16 of the Guidelines provides that ‘where the figures made available by an undertaking are incomplete or not reliable, the Commission may determine the value of its sales on the basis of the partial figures it has obtained and/or any other information which it regards as relevant and appropriate.’
1859Second, it is clear from paragraphs 1846 to 1850 above that a methodology based on the effective spreads obtained by the banks concerned was not a possibility.
1860Similarly, the use of the MTS data, as suggested by UniCredit and rejected by the Commission in footnote 1449 to the contested decision, would not have been more appropriate, even if those data would make it possible to provide intra-day data.
1861First, as the Commission observed in that footnote, the use of the MTS data would present shortcomings which, in the context of the methodology defined by the Commission, would make them less appropriate than the BGN data.
1862The use of the MTS data would lead to the following calculation being made in respect of each trade executed by UniCredit during its period of participation:
1863As is apparent from Annex 16 to UniCredit’s application, from which the diagram set out in the preceding paragraph is extracted, such a transaction would lead to – in the words of that bank – ‘implied spreads’ being formed on the basis of an extrapolation from an estimated bid-ask spread (taken from the MTS data) applied to a purchase or sale price actually used by that bank.
1864As the Commission noted in footnote 1449 to the contested decision, such a trade would entail data of a different nature (estimated bid-ask spread and actual purchase or selling prices) and also from different sources (MTS and UniCredit) being compared, whereas the forming of the bid-ask spread used in the contested decision and based solely on the BGN data makes it possible to compare data of the same nature and from the same source.
1865Furthermore, in view of the information provided by UniCredit and the Commission, the reliability of the MTS data cannot be regarded as greater than that of the BGN data.
1866Indeed, as the Commission noted in footnote 1449 to the contested decision and as UniCredit acknowledged in footnote 86 to the application, the MTS data ‘sometimes result in negative implied spreads’.
1867The accounting for transactions which generated such ‘negative implied spreads’ would result in a reduction in that bank’s activity even though the related transactions are part of UniCredit’s actual activity. Moreover, simply excluding such transactions could not compensate for the weaknesses inherent in the use of the MTS data, since that would lead to the activities of the banks concerned not being taken fully into consideration.
1868In the same footnote of its application, UniCredit also accepted that ‘negative implied spreads’ could be explained by ‘timing lags between when a trade actually occurred and when it [was] booked in [that bank’s] system, i.e. the MTS mid-price [was] not always exactly contemporaneous to when the trade actually occurred’.
1869Given the volatility of the EGB market, in which, according to UniCredit, ‘prices change from second to second both upwards and downwards’, such timing lags do not guarantee the reliability of bid-ask spreads calculated using the MTS data, which can therefore be both undervalued and overvalued.
1870Second, it is apparent from UniCredit’s reply to a question put by the Court that, for each of the 4965 trades that took place between 9 September and 28 November 2011, the Commission should have matched those trades with the MTS data, as a result of which it would have been possible, by using coding software, to determine the mid-price of the EGB concerned at the exact time of each trade and, subsequently, the ‘implied actual spread’ of the prices of each of its 4965 trades.
1871In order to do so and in the light of the figures submitted to the Court by UniCredit, the Commission should have cross-checked two sets of raw data, the first of which related to the transactions carried out by that bank and contained 556304 entries, and the second of which contained the MTS data, which consisted of 531 files each containing sometimes more than 40 million entries and the total weight of which was approximately 1.2 TB.
1872However, the Commission cannot reasonably be required to process such a mass of data without being placed under a disproportionate administrative burden (see paragraph 1849 above). Moreover, even if the Commission was in a position to deal with those raw data, it is clear that they were not sent to it during the administrative procedure by UniCredit.
1873UniCredit merely submitted a paper, setting out the conclusions but not attaching its body of data, an omission which it cannot subsequently attempt to remedy by criticising the Commission for not having asked it to send that body of data.
1874Therefore, in the comparative exercise which the Commission had to carry out in order to determine the best available figures which the Commission was to use for the purpose of calculating the proxy, the Commission did not err in not acting on UniCredit’s invitation – which, moreover, is insufficiently documented – to use the MTS data.
1875Furthermore, in so far as UBS claims that the Commission failed to take account of the specific features of the management of EGB portfolios when defining the methodology for calculating the adjustment factor or when using the BGN data (see paragraph 1773 above), it is clear that that bank has not established which alternative methodology – other than the use of the ‘net value traded’ or the ‘adjusted net value traded’, which the Commission rightly rejected (see paragraph 1800 above) – or what available data would have been better able to take account of those specific features.
1876Similarly, that bank cannot call into question the methodology used by the Commission by confining itself to submitting that, in the part of the contested decision in which it responds to the arguments raised by the banks concerned during the administrative procedure, the Commission did not prove certain factual allegations, such as those by which it considered that ‘seeking liquidity, holding or offloading open positions in the market [was] embedded in the bid prices and ask prices that traders [would] submit to their clients’ or that, ‘if accepting a customer order would put his/her portfolio beyond his authorised risk limits, a trader [could] also refuse to quote’ (see recital 854 of the contested decision).
1877In the light of all of the foregoing, the Commission was entitled to adopt the methodology used to calculate the adjustment factor for each of the banks concerned and, to that end, use the BGN data in respect of which it was entitled to take the view that, for the period of each bank’s participation in the infringement at issue, they constituted the ‘best available figures’.
1878For the same reasons, UBS likewise cannot reasonably maintain that the methodology adopted by the Commission is imprecise and arbitrary.
1879Accordingly, the second limb of UBS’s first plea and its fourth and fifth pleas, the second to fourth limbs of UniCredit’s seventh plea and the second to fourth limbs of Nomura’s sixth plea must be rejected as unfounded.
–The objections raised by UniCredit and Nomura alleging errors in the determination of the proxy specific to certain banks
1880In their actions, UniCredit and Nomura objected to the methodology used by the Commission for the determination of the proxy and its general application, which were rejected in paragraphs 1762 to 1879 above.
1881Those banks also raised objections more specifically relating to the determination of the proxy – calculated by multiplying the annualised notional amounts specific to each bank concerned by the adjustment factor that is also specific to each of them – which was applied to those banks and will be examined in paragraphs 1885 to 1960 below.
1882Concerning the determination of the notional amounts (see recitals 816 to 819 of the contested decision), UniCredit claims that the Commission discriminated against it by taking into account, for UniCredit as for the other banks concerned, the notional amounts traded only on the secondary market.
1883UniCredit and Nomura also criticise the Commission for having annualised the notional amounts applied to them, even though they participated in the infringement at issue for less than one year.
1884As regards the determination of the adjustment factor applied to Nomura and, more specifically, the weighting of the 32 categories of representative EGBs specific to that bank and used to determine its final bid-ask spread (see recitals 825 to 830 of the contested decision), that bank criticises the Commission for not having used the exact weighting of those 32 categories of EGBs, even though Nomura had provided it to the Commission.– UniCredit’s objection alleging breach of the principle of non-discrimination owing to the use, for UniCredit as for the other banks concerned, of the annualised notional amounts relating only to the secondary market
1885In the first and fifth limbs of its seventh plea, UniCredit argues that, by basing the proxy for all the banks concerned only on ‘the value of sales on the notional amounts of EGB traded by each [bank concerned] on the secondary market during the relevant period of individual involvement, as this [was] the activity that all [the banks concerned] had in common and [because] it [avoided] the risk of the same revenue being counted twice as part of the fining calculation’ (see recital 816 of the contested decision), the Commission discriminated against it.
1886In so far as the infringement at issue centred on the primary market in EGBs, UniCredit ‘did not participate in [the infringement at issue] with respect to primary trading’ and primary market trading was not taken into account with respect to any of the other banks concerned, the Commission should have used only the amounts traded on the primary market. By failing to do so, the Commission sanctioned UniCredit for the entirety of its participation in the infringement at issue, whereas the other banks concerned were sanctioned for only part of their participation in that infringement.
1887In that regard, the Commission’s decision, in recital 816 of the contested decision, to use, as the basis for calculating the proxy for all of the banks concerned, the notional amounts traded only on the secondary market in EGBs does not, however, reveal any discriminatory treatment of UniCredit.
1888First, UniCredit is wrong to claim that it did not participate in the infringement at issue with respect to the primary market or that that infringement centred on the primary market in EGBs.
1889In recital 861 of the contested decision, the Commission did indeed state, for the purpose of determining the gravity factor for the infringement at issue applied to UniCredit, that that bank had not been active on the primary market.
1890However, it cannot be inferred from that recital – which is unrelated to the determination of the proxy – that the Commission considered that UniCredit had not participated in the entirety of the single and continuous infringement found in the contested decision.
1891On the contrary, it is clear from the operative part of the contested decision that UniCredit is found liable for the entirety of the infringement at issue and that that infringement relates to the entire EGB market, and not exclusively the primary market for EGBs.
1892In addition, in recitals 45 to 50 of the contested decision, the Commission rightly emphasised the relationship between the primary and secondary market in EGBs and the fact that information relating to one was of interest to the other and vice versa (see paragraphs 465 to 486 and 597 to 614 above).
1893Thus, even without being active on the primary market, UniCredit, through the information that it exchanged with the traders from the other banks concerned, contributed to the functioning of, and benefited from, the entirety of the infringement at issue, as the Commission observed in recital 479 of the contested decision.
1894Second, the ground relied on by the Commission in recital 816 of the contested decision fully justifies its decision not to use the notional amounts traded by the banks concerned on the primary market in EGBs, or to include them in the notional amounts relied on in respect of the banks concerned, with the exception of UniCredit.
1895Since the activity of the banks concerned operating on the primary market involves purchasing EGBs in order to sell them quickly on the secondary market (see recital 37 of the contested decision), the accounting of the notional amounts purchased from State issuers on the primary market was unnecessary – and even a source of bias – in order to determine the extent of the activity of those banks.
1896Third, UniCredit’s specific situation was taken into account by the Commission at other stages of the calculation of the fine imposed on it (see, to that effect, judgment of 13 September 2013, Total Raffinage Marketing v Commission, T‑566/08, EU:T:2013:423, paragraph 454).
1897In recitals 861 and 868 of the contested decision, the Commission thus applied to that bank a gravity factor and an additional amount both fixed at 17%, one percentage point lower than the rates applied to the other banks concerned.
1898In the light of the foregoing, the first and fifth limbs of UniCredit’s seventh plea must be rejected as unfounded.
1899Furthermore, in so far as UniCredit might argue, in the context of those first and fifth limbs, that its situation was not properly taken into consideration in the determination of the gravity factor and the additional amount imposed as a deterrent, those objections will be examined, respectively, in paragraphs 1976 to 2003 and 2004 to 2035 below.– The objections raised by UniCredit and Nomura based on the annualisation of the notional amounts traded
1900UniCredit, in its eighth plea and Nomura, in the fifth limb of its sixth plea taken together with the second limb of its seventh plea, submit that the Commission made errors of assessment or breached the principle of equal treatment and the principle that penalties must be specific to the offender and offence by annualising the notional amounts they traded with counterparties located in the EEA during their respective periods of participation in the infringement at issue, namely from 9 September to 28 November 2011 and from 18 January to 28 November 2011.
1901That annualisation of the notional amounts traded by the banks concerned did have the effect of smoothing out the notional amounts traded by the banks which had participated in the infringement at issue for longer than one year, but also had the effect of artificially inflating the notional amounts used in respect of UniCredit and Nomura and, therefore, of overestimating their relative weight in that infringement.
1902In that regard, UniCredit states that, during its period of participation in the infringement at issue, the volume of EGBs traded – particularly Italian EGBs which represented the bulk of its trades – was particularly high and the EGB market was subject to very high volatility, involving significantly higher bid-ask spreads than in normal market conditions. As a result, UniCredit received by far the worst treatment of the banks concerned, which is ‘manifestly unreasonable and unfit’. Consequently, the Commission should have compared data relating to trading activities over the same period for all the banks concerned, namely the last year of the infringement at issue, or the total duration of that infringement.
1903For similar reasons, Nomura submits that the Commission should have used the notional amounts actually traded by that bank throughout 2011.
1904At the outset, it should be noted that neither UniCredit nor Nomura criticise the Commission for not having applied, for the purpose of determining the notional amounts to be taken into consideration for the calculation of the proxy, point 13 of the Guidelines, under which the Commission normally uses the sales made by the undertaking during the last full business year of its participation in the infringement at issue.
1905That decision on the part of the Commission was fully justified in the present case for the reasons set out in recital 819 of the contested decision and confirmed by the arguments put forward by UniCredit and Nomura, namely, first, that ‘the time period of individual involvement [was] different for most [of the banks concerned], and sometimes relatively short’, and, second, that ‘the bid-ask spread over the infringement period’ was highly volatile.
1906Ultimately, UniCredit and Nomura criticise only the fact that, in order to take account of those circumstances and despite the fact that their respective periods of participation in the infringement at issue was less than a full year, the Commission annualised the notional amounts which they traded during the full calendar months of their periods of participation in the infringement at issue, namely, October 2011 and February to October 2011, respectively (Table 1 of Annex 2 to the contested decision), as the Commission did in relation to the other banks concerned.
1907However, that decision – which forms part of defining the methodology for calculating the fine in respect of which, as already pointed out in paragraphs 1754 and 1777 above, the Commission has a broad discretion – does not reveal any manifest error of assessment or error of law on the part of the Commission.
1908The circumstances referred to in paragraph 1905 above justify the Commission annualising the notional amounts traded by all the banks concerned, in accordance with the methodology set out in recital 819 of the contested decision pursuant to which it divided the total notional amounts traded by each bank concerned during the period of individual involvement in the cartel by the number of months of participation and subsequently multiplied this monthly average by 12.
1909First, annualising the notional amounts traded makes it possible to smooth out the significant annual variations that the EGB market experienced throughout the infringement at issue and to which the banks concerned that participated in that infringement were subject, and at the same time takes into account the full activity of those banks, including that of 2011, which both the Commission as well as UniCredit and Nomura consider to be a year marked by a high volatility on the EGB market linked to the sovereign debt crisis.
1910Second, in so far as the annualisation of the notional amounts traded was applied to all the banks concerned, it reflects the relative weight of each bank in the infringement at issue, as required by the case-law applicable to the value of sales (see, to that effect, judgment of 12 November 2014, Guardian Industries and Guardian Europe v Commission, C‑580/12 P, EU:C:2014:2363, paragraph 63) and which applies equally in relation to the proxy.
1911Since the annualised notional amounts of each bank concerned constitute the first of the two elements taken into consideration in calculating the proxy specific to each of those banks, they enable the relationship existing between the proxies applied to each bank concerned and, ultimately, the fines imposed on them, to be maintained.
1912Those two conclusions are not called into question by the arguments put forward by UniCredit and Nomura.
1913First, UniCredit and Nomura cannot reasonably maintain that annualising their periods of participation in the infringement at issue of one and nine full calendar months, respectively, led the Commission to inflate artificially, and therefore wrongly, the notional amounts used in respect of those banks.
1914While it is true that that annualisation entailed extrapolating to one year the activities of UniCredit and Nomura during their one and nine full calendar months of participation in the infringement at issue, it was made necessary by the principles governing the calculation of fines under the Guidelines and at the same time led the Commission to apply a value that was directly proportional to the actual activity of those banks during their respective periods of participation in the infringement at issue.
1915First, annualising the notional amounts traded by UniCredit and Nomura was made necessary by the principles governing the calculation of fines under the Guidelines, since the Guidelines require a value of sales or a proxy to be determined on an annual basis, which those banks do not dispute, particularly in view of the alternative calculation methods they propose.
1916Consequently, the Commission could not use only the notional amounts traded by those two banks during their respective periods of participation in the infringement at issue, both of which were less than one year.
1917Second, the annualisation of the notional amounts traded by UniCredit and Nomura meant that the Commission used a value strictly proportional to the actual activity of those banks during their respective periods of participation in the infringement at issue, since it is the result of a simple division of the notional amounts traded by those two banks during their respective periods of infringement by the number of full calendar months of their participation in the infringement at issue (one full calendar month for UniCredit and nine full calendar months for Nomura), subsequently multiplied by 12, to provide an annual basis.
1918As a result and due to the fact that the notional amounts were annualised for all the banks concerned, the Commission also ensured that the relative weight of each bank in the infringement at issue and the extent of their respective involvement were respected.
1919While the annualisation of the notional amounts of UniCredit and Nomura might have led the Commission to apply higher amounts to those banks than the notional amounts they actually traded in 2011, that circumstance does not call into question the fact that the annualised notional amounts used by the Commission are strictly proportional to the notional amounts actually traded by those banks during their respective periods of participation in the infringement at issue.
1920That circumstance is simply the consequence of the fact that, during their respective periods of participation in the infringement at issue, the EGB market was more volatile than in the months of 2011 when they were not participating in it.
1921Similarly, the fact that the annualised notional amounts used for UniCredit and Nomura are, according to those banks, significantly higher compared to those used for the other banks which participated in the infringement at issue for longer or had larger market shares is merely a consequence of the fact that UniCredit and Nomura participated in the infringement at issue only during 2011, namely during a period when the EGB market was most volatile and, therefore, during a period when trades in EGBs and bid-ask spreads for EGBs were proportionally highest.
1922Second, UniCredit and Nomura cannot reasonably maintain that, in order to observe the principle of equal treatment and the principle that penalties must be specific to the offender and the offence, the Commission should have applied, as regards UniCredit, the notional amounts of the last year of the infringement at issue or of the entire period of the infringement at issue and, as regards Nomura, the notional amounts traded during the whole of 2011.
1923First, it follows from the principles governing the Guidelines, applicable in the present case, and, in particular, from points 5 and 6 thereof, that, like the value of sales, the proxy is to be calculated by taking into consideration transactions to which the infringement relates.
1924This means that the Commission cannot take into account activities carried out by the undertaking in question which do not come within the scope of the alleged cartel (see, to that effect, judgments of 7 September 2016, Pilkington Group and Others v Commission, C‑101/15 P, EU:C:2016:631, paragraph 19 and the case-law cited, and of 25 March 2021, Lundbeck v Commission, C‑591/16 P, EU:C:2021:243, paragraph 187 and the case-law cited).
1925Taking into account the periods referred to by UniCredit and Nomura might certainly be more favourable to those banks, but it would mean using transactions that took place outside their respective periods of participation in the infringement at issue as a basis for calculating the proxy applied to those banks.
1926Second, it is settled case-law that, to the extent to which reliance is to be placed on the activities of the undertakings involved in the same infringement for the purpose of determining the proportions between the fines to be imposed, the period to be taken into consideration must be ascertained in such a way that the resulting turnovers are as comparable as possible and, consequently, an individual undertaking cannot compel the Commission to rely, in its case, upon a period different from that used for the other undertakings, unless it proves that, for reasons peculiar to it, the activity which it carried out during the latter period does not reflect its true size and economic power or the scale of the infringement which it committed (see, to that effect, judgment of 29 September 2021, Tokin v Commission, T‑343/18, EU:T:2021:636, paragraph 100 and the case-law cited).
1927In the present case, UniCredit and Nomura have not furnished such proof.
1928On the contrary, it has been found in paragraphs 1915 to 1918 above that the Commission used annualised notional amounts with respect to all the banks concerned and that the notional amounts used for UniCredit and Nomura were strictly proportional to the notional amounts traded during the one and nine full calendar months of their respective periods of participation in the infringement at issue and reflected the relative extent of their involvement in it.
1929In the light of the foregoing, UniCredit’s eighth plea and the fifth limb of Nomura’s sixth plea and the second limb of its seventh plea must be rejected as unfounded.– Nomura’s objection based on the weighting of the 32 categories of EGBs traded by Nomura
1930In the second argument in the fifth limb of its sixth plea, Nomura claims that the Commission made an error of assessment in refusing, for the purpose of determining the weighting of the 32 categories of EGBs used to calculate the adjustment factor, to use the exact data which Nomura provided during the administrative procedure.
1931As already recalled in paragraphs 1716 to 1724 above, the Commission determined an adjustment factor specific to each bank concerned by using half of its final bid-ask spread, calculated by using the simple average of its daily bid-ask spreads, which themselves were calculated, for each working day of the participation of each bank concerned in the infringement at issue, by using the weighted average of the daily bid-ask spread level of 32 categories of representative EGBs to which specific weights were assigned for each bank.
1932In order to determine the weighting of each of those 32 categories of representative EGBs in the activities of each bank concerned throughout its period of participation in the infringement at issue, the Commission requested, during the administrative procedure, that each bank provide it with two sets of data, namely, first, the actual volume of EGBs traded per issuing State (regardless of maturity) and, second, the actual volume of EGBs traded per maturity range (regardless of issuer).
1933By crossing those two sets of data, the Commission determined a weighting of the 32 categories of EGBs specific to each bank concerned.
1934It thus applied the following weighting with regard to Nomura (Annex 2 to the contested decision, Table 2):
1935During the administrative procedure, Nomura sent the Commission the two sets of data referred to in paragraph 1932 above so as to enable the Commission to calculate the weighting of each of the 32 categories of representative EGBs with regard to Nomura. It was also the only bank to send the Commission, unprompted, the exact volumes that it had traded throughout its period of participation in the infringement at issue for each of the 32 categories of representative EGBs, thus providing the exact weighting of those 32 categories of EGBs.
1936Those data – in respect of which the Commission does not dispute the completeness and reliability but only that they constitute the best available figures – reveal the following differences:
1937Notwithstanding the fact that Nomura sent the Commission its exact data, the Commission used, in the contested decision, the estimated data referred to in paragraph 1934 above, on the ground that the crossing, for each bank concerned, of the actual volume of EGBs traded per issuing State and the actual volume of EGBs traded per maturity had to be considered the best approach based on the data provided by the banks (see recitals 848 and 849 of the contested decision).
1938In recital 850 of the contested decision, the Commission also emphasised the need to ‘adopt an approach that was appropriate for all [banks concerned] taking account of the information in its possession’ and the fact that, ‘absent objective differences in the … respective situations [of the banks concerned], [it could not] apply different methodologies for different [banks concerned]’.
1939In its defence and in its written reply to a question put by the Court, first, the Commission argued that using the exact data provided by Nomura would have led it to breach the principle of equal treatment between the different banks penalised in the contested decision. Secondly, it maintains that, even if it could have adopted a more granular approach in relation to that bank, its calculation of the relative weights still accurately reflected the breakdown of Nomura’s trading activity and was thus a valid methodology for the period of that bank’s infringement.
1940It is therefore necessary to examine whether the Commission was entitled to refuse to take into account the exact data provided by Nomura.
1941In that regard, it has already been pointed out in paragraph 1783 above that, even when the Commission decides to depart from the methodology set out in point 13 of the Guidelines, it cannot depart from the guiding principles or underlying logic of those guidelines.
1942In particular, it must ensure that the best available figures are taken into consideration and that partial figures are used only where the figures made available by an undertaking are incomplete or not reliable.
1943Those requirements apply all the more rigorously where the methodology used by the Commission has led it to apply an adjustment factor higher than 99% to notional amounts of several hundred billion euros (see, by analogy, judgment of 24 September 2019, HSBC Holdings and Others v Commission, T‑105/17, EU:T:2019:675, paragraph 328).
1944In the present case, the Commission does not dispute that the data provided by Nomura was complete and reliable and therefore accurately represented that bank’s trading activity for the period of its participation in the infringement at issue.
1945It is also apparent from the table referred to in paragraph 1936 above that the weighting of the 32 categories of representative EGBs applied by the Commission to Nomura differs significantly, for certain of those categories, from that bank’s actual trading activity.
1946Thus, as regards German EGBs with a maturity of 3 to 7 years, there is a difference of 2.9 percentage points, the Commission having found that those EGBs represented 16.8% of all the notional amounts traded by Nomura (approximately EUR 74.9 billion), whereas they actually represented 19.7% (approximately EUR 87.9 billion).
1947Similarly, as regards Italian EGBs with a maturity of 3 to 7 years, there is a difference of 2 percentage points, the Commission having found that those EGBs represented 9.9% of all the notional amounts traded by Nomura (approximately EUR 44.2 billion), whereas they actually represented 7.9% (approximately EUR 35.2 billion).
1948It follows that, at least for certain categories of representative EGBs constituting a significant proportion of Nomura’s trading activity (namely more than 27%), the weighting applied by the Commission to that bank differed significantly from that calculated on the basis of the trading activity actually carried out by that bank.
1949Consequently, in view of the fact that Nomura provided complete and reliable figures relating to the distribution of its trading activity between the 32 categories of representative EGBs chosen by the Commission, the weighting applied by the Commission, in Table 2 of Annex 2 to the contested decision, could not constitute the ‘best available figures’ for the purpose of calculating the proxy.
1950That conclusion is confirmed by the fact that the Commission’s decision not to use the weighting of the 32 categories of representative EGBs as derived from the data provided by Nomura during the administrative procedure – and which accurately represented that bank’s trading activity for the period of its participation in the infringement at issue – means that the fine imposed on that bank differs by more than EUR 3.927 million, being approximately 3% of the amount of the fine.
1951According to the calculations made by the Commission in response to a question put by the Court, the taking into consideration of Nomura’s actual activity for the purpose of determining the weighting of the 32 categories of representative EGBs used in respect of that bank entails a reduction in the proxy applied to that bank from EUR 392646351 to EUR 380747977 and thus a reduction in the fine imposed on it from EUR 129573000 to EUR 125646000.
1952Accordingly, the Commission cannot reasonably maintain that the weighting of the 32 categories of representative EGBs used in the contested decision in relation to Nomura ‘still accurately reflect[ed] the breakdown of [that bank’s] trades’.
1953Moreover, the Commission’s use of the data provided by Nomura would not have compromised the methodology developed by the Commission to determine the proxy.
1954Unlike, for example, the method for calculating the notional amounts traded by each bank concerned, the method for calculating the daily bid-ask spreads and the final bid-ask spread, or the choice of the 32 categories of representative EGBs, which apply to all the banks penalised and form the basis of the methodology adopted by the Commission, the use of Nomura’s data relating to the weighting of those 32 categories of EGBs in its trading activity is merely one way of implementing that methodology, which does not alter its substance, but is only capable of increasing – at that bank’s reasoned request – its accuracy.
1955Moreover, it may be observed that the Commission itself acknowledged that it ‘could have adopted a more granular approach’, thus showing that the use of the exact data provided by Nomura did not compromise the methodology adopted by the Commission.
1956Lastly, the Commission’s use of the exact data provided by Nomura would not have breached the principle of equal treatment.
1957During the administrative procedure, Nomura was the only bank to provide the Commission, unprompted and voluntarily, with precise data relating to the weighting of the 32 categories of representative EGBs chosen by the Commission, distinguishing it from all the other banks.
1958For that reason, the Commission was not required, under the principle of equal treatment, to apply a weighting to Nomura that was simply estimated, when it had an exact weighting provided by Nomura.
1959Therefore, the Commission was wrong to refuse to make use of the exact data provided by Nomura during the administrative procedure.
1960In the light of the foregoing, the second argument in the fifth limb of Nomura’s sixth plea must be upheld.
(ii)The pleas alleging errors in the determination of the gravity factor applied to UniCredit and Nomura
1961After determining the proxy applied to each of the banks concerned, the Commission stated, in recitals 859 to 863 of the contested decision, that the infringement at issue required it to set a gravity factor of 18% of that proxy for the banks concerned (points 20 and 23 of the Guidelines).
1962That rate was justified by the fact that the conduct in question ‘amounted to price-fixing arrangements, collusive exchanges of information, market sharing and customer allocation’, which were ‘among the most harmful restrictions of competition’. The Commission also stated that they permeated the whole EGB industry (that is to say, the primary market as well as the secondary market), covered the entire EEA and related to products used for raising public funding at the time of a serious financial crisis, when financial institutions, including some involved in the infringement at issue, had to be rescued.
1963However, given that the UniCredit trader was not active on the primary market for EBGs, the Commission reduced the gravity factor applied to that bank to 17% (see recital 861 of the contested decision).
1964In their respective ninth pleas, UniCredit and Nomura claim that the Commission made various errors in the determination of the gravity factor applied to them.
–The first and second limbs of Nomura’s ninth plea, alleging errors of assessment in setting the gravity factor for the infringement at issue at 18%
1965In the first and second limbs of its ninth plea, Nomura argues that the Commission made errors of assessment in setting the gravity factor for the infringement at issue at 18% of the proxy.
1966According to Nomura, the Commission was wrong to take the view, in recitals 859 and 863 of the contested decision, that the conduct at issue amounted to price-fixing arrangements, market sharing and customer allocation, when that characterisation had not previously been adopted in that decision and even though that conduct amounted to information exchange only capable of affecting price.
1967Moreover, in the light of the Commission’s previous decision-making practice, the conduct at issue should be regarded as serious and not as ‘the most harmful’, therefore justifying a gravity factor of 15% of the proxy. In that regard, Nomura also argues that the conduct at issue did not involve the manipulation of a financial benchmark and was carried out by banks with a low market share, which makes it highly unlikely that the limited number of discussions at issue could have had any material impact on the EGB sector or on public funding.
1968At the outset, it should be noted that Nomura’s objection referred to in paragraph 1966 above is based on a misreading of the contested decision, which expressly states that the conduct at issue amounted to price fixing, market sharing and customer allocation, in particular in recitals 499, 532, 535 and 568.
1969As to the remainder, it is settled case-law and apparent from point 23 of the Guidelines that horizontal cartels leading to price fixing or customer allocation reveal, by their very nature, a particular degree of harm to competition (see, to that effect, judgments of 30 June 1966, LTM, 56/65, EU:C:1966:38, page 359; of 20 November 2008, Beef Industry Development Society and Barry Brothers, C‑209/07, EU:C:2008:643, paragraphs 17 and 33; of 11 September 2014, CB v Commission, C‑67/13 P, EU:C:2014:2204, paragraph 51; of 16 July 2015, ING Pensii, C‑172/14, EU:C:2015:484, paragraph 32; and of 23 January 2018, F. Hoffmann-La Roche and Others, C‑179/16, EU:C:2018:25, paragraph 78; of 30 January 2020, Generics (UK) and Others, C‑307/18, EU:C:2020:52, paragraph 67; and of 21 December 2023, European Superleague Company, C‑333/21, EU:C:2023:1011, paragraphs 162 and 163).
1970Accordingly, the finding that the infringement at issue took the form of such conduct alone enabled the Commission to consider that that infringement ‘was among the most harmful restrictions of competition’. That justified a gravity factor being applied to that conduct at the upper end of the range provided for in point 21 of the Guidelines, set at 18%.
1971The fact, also noted by the Commission, that the conduct at issue covered the entire EEA and related to products used for raising public funding at the time of a serious financial crisis justified the choice of that gravity factor all the more.
1972Accordingly, the Commission did not make an error of assessment in applying a gravity factor for the infringement at issue set at 18% of the proxy.
1973That conclusion cannot be called into question by Nomura’s argument based on the Commission’s practice in previous decisions, which, in any event and as stated in paragraph 1618 above, does not serve as a legal framework for the fines imposed in competition matters.
1974Nor can it be called into question by the argument that the conduct at issue had no ‘material impact’ due to the low market share of the banks concerned, since, as pointed out in paragraph 1970 above, the gravity factor applied by the Commission is, in any event, fully justified by the factors referred to in recital 859 of the contested decision (see, by analogy, judgment of 16 July 2020, Nexans France and Nexans v Commission, C‑606/18 P, EU:C:2020:571, paragraph 105).
1975In the light of the foregoing, the first and second limb of Nomura’s ninth plea must be rejected as unfounded in so far as they object to the gravity factor for the infringement at issue being set at 18%.
–UniCredit’s ninth plea and the first, second and third limbs of Nomura’s ninth plea, alleging errors of assessment and breaches of the principle that penalties must be specific to the offender and the offence, the principle of proportionality and the principle of equal treatment, in the setting of the gravity factor applied to them
1976UniCredit, in its ninth plea, and Nomura, in the second and third limbs of its ninth plea, claim that the Commission failed to take sufficient account of their limited participation in the infringement at issue in setting the gravity factor applied to them.
1977UniCredit submits that the Commission breached the principle that the penalty must be specific to the offender and the offence and the principle of proportionality by not reducing the gravity factor applied to it by more than one percentage point, even though it did not participate in all aspects of the infringement at issue and played a minor role in the aspects in which it did participate.
1978Nomura argues that the Commission made errors of assessment by not taking sufficient account of the fact that it participated in only 34 discussions relating to only 9 of the 17 Member States of the then euro area, it was a primary dealer for only 10 of those 17 Member States, its market share in the primary market for EGBs was low and the conduct at issue did not affect its performance during auctions. The Commission also breached the principle of equal treatment inasmuch as Nomura did not benefit from a reduction of its gravity factor by one percentage point as UniCredit did, even though its involvement in the infringement at issue was comparable to that of UniCredit.
1979As UniCredit has observed, it is settled case-law that, where an infringement has been committed by several undertakings, the relative gravity of the participation of each undertaking must be examined for the purpose of determining the amount of the fines to be imposed on them (see, to that effect, judgment of 8 July 1999, Hercules Chemicals v Commission, C‑51/92 P, EU:C:1999:357, paragraph 110 and the case-law cited).
1980It is also apparent from the Guidelines, and from the case-law, that a distinction must be drawn between, on the one hand, the assessment of the objective gravity of the infringement concerned, carried out by way of the gravity factor provided for in points 20 to 23 of the Guidelines, and, on the other hand, the assessment of the relative gravity of the participation in that infringement of the undertaking on which the penalty is to be imposed, carried out by way of the adjustments to the basic amount provided for in points 27 to 31 of those guidelines (see, to that effect, judgment of 15 July 2015, voestalpine and voestalpine Wire Rod Austria v Commission, T‑418/10, EU:T:2015:516, paragraph 414 and the case-law cited).
1981That distinction does not rule out the possibility that, as in the present case, the Commission may apply a lower gravity factor to an undertaking participating in a cartel than to other undertakings, in particular where, in the context of a single and continuous infringement, that undertaking had no role in the implementation of certain parts of that infringement (see, to that effect, judgment of 15 July 2015, voestalpine and voestalpine Wire Rod Austria v Commission, T‑418/10, EU:T:2015:516, paragraph 415).
1982In the present case, it has been found in paragraphs 1969 to 1974 above that the Commission was entitled to set the gravity factor for the infringement at issue at 18% of the proxy.
1983Accordingly, it is necessary to examine whether, in the light of the evidence put forward by UniCredit and Nomura, the Commission should have applied a lower gravity factor than that applied to them in the contested decision.
1984As regards UniCredit in the first place, it must be borne in mind that the gravity factor applied to UniCredit was reduced by one percentage point as compared to that applied to the other banks so as to take account of the fact that its trader was not active on the primary market.
1985That reduction – which led to a gravity factor not of 18% but of 17% of its proxy being applied – constitutes a fair taking into account of UniCredit’s situation.
1986While it is true that, not being a primary dealer, UniCredit was not able to benefit directly from the anticompetitive conduct concerning the primary market, the fact remains that not only was its trader involved in anticompetitive conduct concerning the secondary market coming within Category 4, but that trader was also connected to the CODS & CHIPS chatroom on which he actively participated in discussions concerning both the secondary market and the primary market.
1987As a result, UniCredit contributed to the functioning of the infringement at issue as a whole and, given the interest in the disclosure of commercially sensitive information relating to the primary market for secondary market trading (see recital 50 of the contested decision), confirmed by ESMA (see paragraphs 566 and 601 above), it was able to benefit from the information exchanged for its activity on the secondary market.
1988Moreover, during its period of participation in the infringement at issue, UniCredit participated in all of the discussions that took place between the banks concerned and were legitimately relied on by the Commission. Therefore, the limited number of discussions relied on against it is merely the direct consequence of the reduced duration of its participation in the infringement at issue, which was reflected in the duration multiplier (see recitals 864 to 866 of the contested decision). That is also the case in relation to its lack of participation in the DBAC chatroom, which ceased being used by the traders of the banks concerned prior to the beginning of UniCredit’s participation in the infringement at issue.
1989Lastly, the facts alleged by UniCredit to the effect that it was not aware of its trader’s conduct or that it promptly let him go are irrelevant, since, as found in paragraph 305 above, that bank could be held liable for the conduct of its employee and his dismissal was not, in any event, on grounds of his participation in the infringement at issue, but, as UniCredit stated, was because of the ‘material and unjustifiable losses’ made by him.
1990As regards, in the second place, Nomura, a gravity factor of 18% of its proxy was applied, a rate which also constitutes a fair taking into account of its situation.
1991First, the Commission cannot be accused of any error of assessment of Nomura’s situation in that regard.
1992During its period of participation in the infringement at issue, Nomura participated in all of the discussions that took place between the banks concerned and that were legitimately relied on by the Commission. Therefore, the number of discussions relied on against it – which Nomura regards as low, but which nevertheless amounts to 31 – is merely the direct consequence of the duration of its participation in the infringement at issue, which was reflected in the duration multiplier (see recitals 864 to 866 of the contested decision).
1993Similarly, the fact that Nomura was active only on the primary market for certain EGBs, or was a small player on that market, cannot change the gravity factor applied to it, since the Commission rightly considered that the infringement at issue covered all EGBs and that the lesser activity alleged by that bank was taken into consideration in the determination of its proxy, which reflects the economic importance of its activity to which the infringement at issue directly or indirectly relates.
1994Lastly, as regards the fact that Nomura’s participation in the infringement at issue did not, according to Nomura, affect its performance during auctions, it is clear that the information exchanged in relation to the primary market was also relevant to its trading activities on the secondary market (see recital 50 of the contested decision, confirmed by ESMA, as is apparent from paragraph 566 above), which was not taken into account in the economic analysis submitted by that bank in support of its claims.
1995Second, Nomura cannot meaningfully rely on a breach of the principle of equal treatment on account of the fact that a higher gravity factor was applied to it than to UniCredit.
1996Contrary to that bank’s submission, its involvement in the infringement at issue is not comparable to that of UniCredit and therefore did not justify a gravity factor of 17% being applied to it.
1997While, like UniCredit, it participated in all of the discussions that took place during its period of participation in the infringement at issue, Nomura was nevertheless active on both the primary market and the secondary market for EGBs and was therefore in a position to benefit directly and fully from the information obtained in the discussions in which it participated, unlike UniCredit.
1998The fact that, according to Nomura, one of its traders was not authorised to trade prior to 4 March 2011 does not call that conclusion into question.
1999First, as the Commission stated in recital 768 of the contested decision, the fact that the Nomura trader obtained renewed access to the CODS & CHIPS chatroom establishes the interest of the traders from the other banks in reinstating him in their discussions and enabled that trader to obtain access to the anticompetitive discussions which took place prior to the renewal of his authorisation to trade and to participate actively in those discussions.
2000Secondly, unlike in relation to UniCredit, the Commission has not indicated that it was unable to establish that the Nomura trader had passed on the information obtained to colleagues active on either the primary or secondary market, contrary to what that bank suggests in paragraph 229.2 of its application.
2001Furthermore, unless otherwise substantiated, Nomura’s mere assertion that ‘insofar as [it] has been able to determine, [its trader] did not in fact execute any trades of EGB until 9 March 2011’ cannot be taken into consideration.
2002Accordingly, the gravity factor applied to Nomura does not reveal any error of assessment or breach of the principle of equal treatment on the part of the Commission.
2003In the light of the foregoing, UniCredit’s ninth plea and the first and second limbs of Nomura’s ninth plea, in so far as they concern the failure to take into account the gravity of that bank’s participation in the infringement at issue, and the third limb of that plea, must be rejected as unfounded.
(iii)The first limb of UniCredit’s tenth plea, alleging breach of the principle that the penalty must be specific to the offender and the offence and of the principle of proportionality in the determination of the rate of the additional amount imposed as a deterrent
2004In the first limb of its tenth plea, put forward in the alternative, UniCredit claims that the Commission breached the principle that penalties must be specific to the offender and the offence and the principle of proportionality, by applying, for the purpose of determining the basic amount of its fine, a rate of 17% of the proxy for the purposes of the additional amount, in accordance with point 25 of the Guidelines.
2005Since the duration multiplier (point 24 of the Guidelines) does not apply in the calculation of the additional amount (point 25 of the Guidelines), that decision resulted in the Commission applying an additional amount of EUR 59522445 to UniCredit, representing approximately 600% of the variable component of the basic amount of that bank’s fine, referred to in recital 867 of the contested decision, whereas it represents approximately 120% of that of Nomura, approximately 63% of that of Portigon and approximately 20% of those of RBS and UBS. However, a significantly lower percentage should have been used for UniCredit in view of its lesser involvement in the infringement at issue.
2006In the alternative, and for the same reasons, UniCredit submits that the Commission erred in not making use of point 37 of the Guidelines in order significantly to adjust the additional amount.
2007In the further alternative and again for the same reasons, UniCredit submits that the Commission should have set the additional amount at the minimum provided for by point 25 of the Guidelines, namely 15%.
2008As regards, in the first place, the principal line of argument raised by UniCredit, it should be borne in mind, as regards its claim alleging breach of the principle that penalties must be specific to the offender and the offence, that, in the context of calculating the amount of fines imposed under Article 23(2) of Regulation No 1/2003, differentiated treatment of the undertakings concerned is inherent in the exercise of the Commission’s powers under that provision. In exercising its discretion, the Commission is required to fit the penalty to the individual conduct and specific characteristics of the undertakings concerned in order to ensure that, in each case, the EU competition rules are fully effective (see judgment of 7 June 2007, Britannia Alloys & Chemicals v Commission, C‑76/06 P, EU:C:2007:326, paragraph 44 and the case-law cited).
2009Furthermore, in order to comply with the principle that penalties must be specific to the offender and the offence, it is sufficient that the final amount of the fine reflects the differences in the situation of the various participants, without there being any need for the Commission to differentiate the treatment of the participants at each stage of the calculation of the amount of the fine (judgment of 13 September 2013, Total Raffinage Marketing v Commission, T‑566/08, EU:T:2013:423, paragraph 454).
2010It follows that the mere fact that the Commission adopts, with regard to one of the banks concerned, an additional amount which, in relative value, constitutes a significantly higher proportion of the variable amount of the basic amount of the fine in question than for the other banks concerned cannot, in itself, establish a breach of the principle that the penalty must be specific to the offender and the offence with regard to UniCredit.
2011Moreover, in the present case, such a breach is all the less proven since, first, point 25 of the Guidelines provides that the additional amount, which reflects the characteristics of the practices of all the participants and not the individual situation of each of them (see, to that effect, judgment of 11 July 2019, Silver Plastics and Johannes Reifenhäuser v Commission, T‑582/15, not published, EU:T:2019:497, paragraph 338 and the case-law cited), is calculated ‘irrespective of the duration of the undertaking’s participation in the infringement’, in order, in particular, to deter undertakings from infringing competition law, even if only for a short period (judgment of 7 November 2019, Campine and Campine Recycling v Commission, T‑240/17, not published, EU:T:2019:778, paragraph 346).
2012Consequently, and in the absence of any challenge by UniCredit to the validity of point 25 of the Guidelines which the Commission expressly applied, there was no need for the Commission to take into account, in that regard, the duration of UniCredit’s participation in the infringement at issue.
2013Second, notwithstanding the case-law referred to in paragraph 2011 above, it is clear that, in recital 868 of the contested decision, the Commission applied a specific additional amount of 17% of the proxy to UniCredit and not 18% as for all the other banks concerned, in order to take account of the fact that ‘[UniCredit’s] trader was only active on the secondary market’.
2014Consequently, UniCredit is wrong to claim that the Commission failed to take account of UniCredit’s specific situation and therefore breached the principle that penalties must be specific to the offender and the offence.
2015That bank is also wrong to rely on a breach of the principle of proportionality by reason of the specific rate of 17% of the proxy applied to it as an additional amount.
2016First, in the light of the wording of point 25 of the Guidelines, none of the factors outlined by that bank supports the conclusion that the Commission should have refrained from imposing an additional amount on UniCredit.
2017As the Commission correctly observed in recital 871 of the contested decision, the infringement at issue took the form of agreements and concerted practices between competitors and therefore of a cartel, whose gravity and harmfulness to competition justify the imposition of a penalty intended to deter other undertakings from engaging in the same or similar conduct.
2018In the latter regard, contrary to UniCredit’s contention, the Commission was entitled to take into account, in recital 872 of the contested decision, the fact that the financial sector had already been the scene of similar anticompetitive conduct, even if that bank had not taken part in it.
2019The pursuit of a deterrent effect does not solely concern the undertakings specifically named by the decision imposing fines but also seeks to encourage undertakings of similar size and resources to refrain from participating in similar infringements of the competition rules (see judgment of 5 December 2013, Caffaro v Commission, C‑447/11 P, EU:C:2013:797, paragraph 37; see also, to that effect, judgment of 17 December 2014, Pilkington Group and Others v Commission, T‑72/09, not published, EU:T:2014:1094, paragraph 302 and the case-law cited).
2020Second, UniCredit cannot reasonably claim that the additional amount of 17% of the proxy applied by the Commission exceeds what is appropriate and should have been set at the lower end of the range of 15% to 25% of the value of sales referred to in point 25 of the Guidelines.
2021An additional amount of 17% of the proxy could, correctly, be considered by the Commission to be commensurate with, first, the gravity and inherent harmfulness to competition of the conduct at issue and the need to deter its repetition in a sector where previous similar infringements had recently been found and, second, the fact that UniCredit was active only on one of the two markets concerned by the infringement at issue.
2022Similarly, the specific characteristics of UniCredit’s participation in the infringement at issue could not justify the Commission applying an additional amount set at less than 17% of the proxy.
2023First of all, a percentage of 17% of the proxy is already at the lower end of the range referred to in point 25 of the Guidelines.
2024Next, the fact that UniCredit was active only on the secondary market or participated in the infringement at issue only for a short period could not justify a reduction of more than one percentage point compared to the additional amount imposed on the other banks concerned.
2025Contrary to UniCredit’s contention, the discussions in which it participated were just as serious or harmful to competition as those which took place before the beginning of its participation in the infringement at issue and justify the Commission finding it necessary to deter the banks concerned as well as other financial market participants from repeating the infringement.
2026Accordingly, applying a percentage of less than 17% to UniCredit would have raised questions as to the Commission’s compliance with its duty to make a consistent and objectively justified assessment of the gravity of the infringement at issue in which UniCredit and the other banks concerned participated, a duty which also contributes to compliance with the principle of proportionality (see, to that effect, judgment of 29 September 2021, Tokin v Commission, T‑343/18, EU:T:2021:636, paragraph 63 and the case-law cited).
2027Lastly, the fact that the rate of 17% applied to UniCredit results in an additional amount representing approximately 600% of the variable basic amount of the fine imposed on it, or more than 85% of the final amount of that fine, is irrelevant, since that rate is the correct application of the methodology set out by the Commission in point 25 of the Guidelines, read in conjunction with Section A thereof, the validity of which UniCredit does not challenge.
2028Furthermore, if UniCredit intended to criticise the Commission not for the rate of the additional amount applied to it, but for the fact that that rate was applied to an annualised proxy even though it had participated in the infringement at issue for only two months, it was for UniCredit to state that expressly in its application and to challenge the validity of point 25 of the Guidelines, which makes provision to that effect.
2029However, it is clear that UniCredit has not substantiated its criticism other than by making comparisons with the additional amounts applied to the other banks concerned and expressed as percentages of the variable basic amount.
2030Consequently, and without prejudice to the assessment to be made by the Court in the exercise of its unlimited jurisdiction, the Commission did not breach the principle that penalties must be specific to the offender and the offence or the principle of proportionality in implementing its Guidelines by imposing on UniCredit an additional amount of 17% of the proxy as a deterrent.…
(3)The second limb of UniCredit’s tenth plea and Nomura’s tenth plea, alleging errors in the adjustment of the basic amount of the fines imposed on those banks
…
2052First of all, as is apparent from recital 382 of the contested decision, the categories identified by the Commission were drawn up for analytical purposes in order to identify one single and continuous infringement with an anticompetitive object and not as many infringements of Article 101 TFEU as there are categories referred to in that recital. Next, those categories are intertwined and partially overlapping, which means that a discussion may come within several categories, as is the case with 10 of the 31 discussions relied on against Nomura and the discussions of 28 September and 2 November 2011, wrongly classed in Category 1. Lastly, it should be noted that, in any event, the Commission criticised Nomura for 17 discussions classed in Category 4, which, moreover, it did not classify in one or other of the subcategories, contrary to Nomura’s contention.
2053Second, that the third indent of point 28 of the Guidelines provides that the fact that an undertaking plays the role of instigator enables the Commission to increase the basic amount of the fine as an aggravating circumstance does not mean that Nomura’s exclusively passive or ‘follow-my-leader’ role, assuming it to be established in the present case, constitutes a mitigating circumstance, within the meaning of point 29 of those guidelines, entitling that undertaking to a reduction in the basic amount of the fine.
2054In that regard, it should also be noted that, although the exclusively passive or ‘follow-my-leader’ role of an undertaking constituted a mitigating circumstance under the first indent of point 3 of the Guidelines on the method of setting fines imposed pursuant to Article 15(2) of Regulation No 17 and Article 65(5) [CS] (OJ 1998 C 9, p. 3), that is no longer the case under the Guidelines (judgment of 16 September 2013, Dornbracht v Commission, T‑386/10, EU:T:2013:450, paragraph 194).
2055Third, Nomura cannot reasonably argue that the Commission should have taken account of the fact that it had not been active on the EGB market prior to 9 March 2011, or before 5 April 2011, since the Commission was entitled to find that that bank’s participation in the infringement at issue had begun on 18 January 2011, as may be seen in paragraph 1351 above.
2056In the light of the foregoing, the second limb of UniCredit’s tenth plea and Nomura’s tenth plea must be rejected as unfounded.
(c)The consequences of the Commission’s error as regards the beginning of UniCredit’s participation in the infringement at issue for the legality of the fine imposed on UniCredit
2057Notwithstanding the lack of any argument raised by UniCredit as to the impact of the duration of its participation in the infringement at issue on the calculation of the basic amount of its fine, it is necessary to examine the consequences of the Commission’s error as regards the start of that bank’s participation in the infringement at issue, which was erroneously set at 9 September 2011.
2058As found in paragraph 1302 above, that start date must be set at 26 September 2011, meaning that UniCredit participated in the infringement at issue from 26 September to 28 November 2011.
2059Such a deferral, which admittedly entails a reduction of 17 calendar days in the duration of UniCredit’s participation in the infringement at issue, has no impact on the duration multiplier (point 24 of the Guidelines) of 0.17 applied to that bank.
2060It is apparent from recitals 864 and 865 of the contested decision that that duration multiplier had been calculated ‘on the basis of the number of full months of participation in the infringement’, understood as meaning that the full month is calculated ‘as from the start date, discounting any days exceeding the last full month. For example, for an infringement [which took place] between 4 January 2007 and 28 November 2011, the duration multiplier is calculated on the basis of full months from 4 January 2007’.
2061Even after deducting the 17 days taken into account by the Commission in error, the duration of UniCredit’s participation in the infringement at issue remains longer than two full months.
2062That finding justifies the duration multiplier of 0.17 applied by the Commission.…
9.Natixis’s fourth plea, alleging that Article 3 of the contested decision is unlawful
…
2099In the present case, by ordering Natixis to refrain from repeating any act or conduct described in Article 1 of the contested decision and from any act or conduct having the same or similar object or effect, the Commission did not exceed the powers conferred on it by Article 7(1) of Regulation No 1/2003 (see, by analogy, judgment of 28 April 2010, Gütermann and Zwicky v Commission, T‑456/05 and T‑457/05, EU:T:2010:168, paragraph 65).
2100The Commission rightly found that Natixis, together with the other banks concerned, infringed Article 101 TFEU and Article 53 of the EEA Agreement by participating in a single and continuous infringement regarding EGBs covering the entire EEA, which, in its action, Natixis did not dispute.
2101In those circumstances, by ordering Natixis to refrain in the future from repeating any act or conduct described in Article 1 of the contested decision and from any act or conduct having the same or similar object or effect, the Commission merely set out the consequences which flow, as regards the future conduct of that bank, from the finding in Article 1, thus ruling out any reversal of the burden of proof and any breach of the principle of the presumption of innocence.
2102Therefore, it cannot be held that the Commission exceeded the powers conferred on it by Article 7(1) of Regulation No 1/2003 or that the obligations imposed on Natixis by the second paragraph of Article 3 of the contested decision exceed the limits of what is appropriate and necessary in order to prevent the occurrence of acts forming part of the infringements found or any conduct having the same or similar object or effect.
2103In the light of the foregoing, Natixis’s fourth plea must be rejected as unfounded.
10.Conclusion on the pleas for annulment of the contested decision
2104It follows from all of the foregoing that all of the pleas raised by the banks on which the Commission did not impose a fine, namely Natixis, BofA and Portigon, are rejected.
2106As regards the banks on which the Commission imposed a fine, first, it also follows from the foregoing that all of the pleas raised by UBS are rejected and that that bank’s first head of claim must therefore be dismissed.
2107Second, it follows from the examination of UniCredit’s second plea that its participation in the single and continuous infringement found in the contested decision is established only for the period from 26 September to 28 November 2011 and that the Commission therefore erred when it found, in the seventh indent of Article 1 of that decision, that UniCredit participated in a single and continuous infringement from 9 September 2011.
2108Consequently, as is apparent from paragraph 2067 above, the Commission erred in setting, in the fifth indent of Article 2 of the contested decision, the amount of the fine imposed on UniCredit at EUR 69442000.
2109It follows that UniCredit’s first head of claim must be upheld in part and that, first, the seventh indent of Article 1 of the contested decision must be annulled in so far as, in that provision, the Commission found that UniCredit participated in a single and continuous infringement from 9 September 2011 and, second, the fifth indent of Article 2 of that decision must be annulled in so far as it sets the amount of the fine imposed on UniCredit at EUR 69442000.
2110Third, it follows from the examination of the second argument in the fifth limb of Nomura’s sixth plea that the Commission was wrong to refuse to use, for the purpose of calculating the adjustment factor, the exact weighting of the EGBs traded by Nomura during the administrative procedure and that, consequently, the Commission erred in setting, in the second indent of Article 2 of the contested decision, the amount of the fine imposed on Nomura at EUR 129573000.
2111It follows that the second indent of Article 2 of the contested decision must be annulled in so far as it sets the amount of the fine imposed on Nomura at EUR 129573000.
D.The claims brought by UBS, UniCredit and Nomura for the exercise of unlimited jurisdiction
2112Since the illegalities found in Cases T‑453/21 and T‑455/21 have an impact on the amount of the fines imposed on UniCredit and Nomura and those banks have requested that the Court exercise its unlimited jurisdiction with regard to them, it is necessary to reassess the amount of their respective fines.
2113As regards UBS, whose fine is not vitiated by any illegality, the Court must also exercise its unlimited jurisdiction, since that bank has requested a reduction in the amount of the fine imposed on it.
2114When they exercise their unlimited jurisdiction provided for in Article 261 TFEU and Article 31 of Regulation No 1/2003, the EU Courts are empowered, in addition to carrying out a mere review of the legality of the penalty, to substitute their own appraisal in relation to the determination of the amount of that penalty for that of the Commission, which adopted the measure in which that amount was initially fixed (see judgment of 21 January 2016, Galp Energía España and Others v Commission, C‑603/13 P, EU:C:2016:38, paragraph 75 and the case-law cited) and, consequently, to cancel, reduce or increase the fine imposed.
2115In order to determine the amount of the fine imposed on an undertaking in the exercise of its unlimited jurisdiction, it is for the EU judicature to assess for itself the circumstances of the case and the nature of the infringement in question. That exercise involves, in accordance with Article 23(3) of Regulation No 1/2003, taking into consideration, with respect to each undertaking penalised, the seriousness and duration of the infringement concerned, in compliance with the principles of, inter alia, adequate reasoning, proportionality, the individualisation of penalties and equal treatment, without the EU judicature being bound by the indicative rules defined by the Commission in its guidelines, even where the latter may give guidance to the EU Courts (see, to that effect, judgment of 21 January 2016, Galp Energía España and Others v Commission, C‑603/13 P, EU:C:2016:38, paragraphs 89 and 90).
1.The form of order sought by UBS
2116As is apparent from paragraph 2106 above, the General Court has found that none of the five pleas raised by UBS, all directed against the fine imposed on it by the contested decision, was well founded.
2117However, there is nothing to support a finding that the fine lawfully imposed by the Commission should be reduced by the General Court, in the exercise of its unlimited jurisdiction, by at least 65% of its initial amount, as requested by UBS.
2118In that regard, UBS cannot meaningfully rely on the fact that the fine imposed on it should be reduced by EUR 32000000 corresponding to a correction for the difference between the end-of-day spreads derived from the BGN data and the intra-day spreads.
2119The economic analysis on which UBS relies for that purpose is not capable of justifying such a correction, since that analysis does not provide data representative of the spreads existing at the time of its participation in the infringement at issue, namely during the sovereign debt crisis.
2120First, it refers to average intra-day spreads, calculated on the basis of EGBs issued by only three issuing Member States (Belgium, France and Germany). Second, those averages were calculated on the basis of data relating to years well after the period of UBS’s participation in the infringement at issue, namely 2018, 2019 and 2021.
2121UBS likewise cannot reasonably claim a reduction of EUR 24600000 on account of the need to calculate the weights of the 32 categories of EGBs on a monthly basis.
2122Even if a monthly calculation might be more favourable to UBS, it has already been found, in paragraph 1835 above, that the Commission’s methodological choice to use constant weights remained within the limits of the discretion that the Commission has in developing its methodology for calculating the proxy. Moreover, that methodology was applied to all the banks concerned. In addition, even in the exercise by the Court of its unlimited jurisdiction, it would be neither consistent with the principle of equal treatment nor reasonable to alter the methodology, legitimately adopted by the Commission and endorsed by the Court, by replacing the constant weights of the 32 categories of representative EGBs used by the Commission with 58 separate sets of monthly weights specific to UBS, resulting in not 32 but 1856 weights being applied to that bank.
2123Lastly, UBS cannot reasonably claim that a correction should be made so that the adjustment factor is set at less than 50% of its final bid-ask spread, such correction being, moreover, unspecified.
2124As is apparent from footnote 1408 to the contested decision, that rate of 50% can be explained as follows:‘Although, conceptually, the revenues made by the bond trader amount to the full bid-ask spread when considering the two transactions together, it follows that, when one considers each of the two transactions individually, the revenues amount to the notional amount multiplied by half the bid-ask spread.’
2125Using a rate lower than 50% would lead, for each purchase transaction and for each sale transaction, to part of the spread obtained by UBS not being taken into account, even though the final bid-ask spread for that bank obtained from the BGN data was considered, in paragraph 1877 above, to be based on the ‘best available figures’.
2126Accordingly, the amount of the fine imposed on UBS must be maintained at EUR 172378000.
2.The form of order sought by UniCredit
2127In paragraphs 1302 and 2107 above, the Court found that the Commission had erred in taking 9 September 2011 as the start date of UniCredit’s participation in the infringement at issue, which meant that that date was set at 26 September 2011.
2128Nevertheless, subject to the implications of that finding on the duration of that bank’s participation in the infringement at issue and on the adjustment factor and therefore on the calculation of the proxy (see paragraph 2069 above), the Court rejected as unfounded the objections raised by UniCredit against the other elements relating to the determination of the amount of the fine imposed on it, and in particular those relating to the methodology adopted by the Commission and its general application.
2129In that context, the Court again considers that, in the exercise of its unlimited jurisdiction, it is entitled to adopt the legally well-founded assessments made by the Commission in recitals 810 to 863 of the contested decision.
2130However, first, it is for the Court to determine whether, in the light of the methodology for calculating the proxy, the reduction in the duration of UniCredit’s participation in the infringement at issue requires the proxy applied to that bank to be adjusted.
2131In that regard, it has already been noted in paragraphs 2065 and 2066 above that, in response to a question put by the Court, the Commission recalculated the proxy to take account of a possible reduction in the duration of UniCredit’s participation in the infringement at issue and that that calculation resulted in the Commission calculating an adjustment factor no longer of 0.084%, but of 0.086%, and, accordingly, a proxy no longer of EUR 350132029, but of EUR 358468506.
2132Such an increase, the arithmetical accuracy of which UniCredit expressly states that it does not dispute, is explained by the fact that, despite the reduction in the duration of that bank’s participation in the infringement at issue, the new period taken into account covers the period when the sovereign debt crisis was most intense.
2133The Court therefore considers that an adjustment factor of 0.086% and, consequently, in the absence of any change in the notional amounts applicable to that bank, a proxy of EUR 358468506, should be applied to UniCredit.
2134Second, given that UniCredit’s ninth plea has been rejected, the Court considers that a gravity factor of 17% should be applied to that bank.
2135Third, in the absence of any challenge by UniCredit to the duration multiplier of 0.17 applied to it and in the light of paragraphs 2057 to 2062 above, the Court considers that that multiplier can be retained.
2136Fourth, as regards the additional amount imposed as a deterrent, set by the Commission at EUR 59522445, it should be borne in mind, first, that that amount was disputed by UniCredit only in so far as the application of a rate of 17% of the proxy or the Commission’s refusal to make use of point 37 of the Guidelines led to an additional amount being imposed on UniCredit that was proportionally higher than that applied to the other banks and, second, that that argument was rejected in paragraphs 2008 to 2035 above.
2137However, it is apparent from case-law that, although the fines imposed on undertakings for infringement of EU competition rules must not only be punitive, but must also have a deterrent effect (see, to that effect, judgments of 10 April 2014, Commission v Siemens Österreich and Others and Siemens Transmission & Distribution and Others v Commission, C‑231/11 P to C‑233/11 P, EU:C:2014:256, paragraph 59, and of 10 April 2014, Areva and Others v Commission, C‑247/11 P and C‑253/11 P, EU:C:2014:257, paragraph 132), they cannot be excessive (see, to that effect, judgment of 12 December 2012, Novácke chemické závody v Commission, T‑352/09, EU:T:2012:673, paragraph 46 and the case-law cited).
2138Annualising the proxy relating to the two-month period of UniCredit’s participation in the infringement at issue for the purpose of calculating the additional amount imposed as a deterrent in respect of that bank would lead to a final fine of EUR 71096000 being imposed, which is excessive.
2139Consequently, in view of the reduced duration of UniCredit’s participation in the infringement at issue, the gravity of that infringement, the market-making activities carried out by that bank, the context of the financial crisis in which that infringement occurred, and also the need to ensure that the penalty for that infringement has a deterrent effect, the Court considers that a fine of EUR 65000000 constitutes a fair penalty for UniCredit’s conduct (see, by analogy, judgment of 5 October 2011, Romana Tabacchi v Commission, T‑11/06, EU:T:2011:560, paragraph 285).
2140The amount of the fine imposed on UniCredit is therefore set at EUR 65000000.
3.The form of order sought by Nomura
2141In paragraphs 1959 and 1960 above, the Court found that the Commission had erred in refusing to make use of the precise data provided during the administrative procedure by Nomura for the purpose of determining the weighting, specific to that bank, of the 32 categories of EGBs and that the second argument in the fifth limb of Nomura’s sixth plea should therefore be upheld.
2142Nevertheless, subject to the implications of that finding on the calculation of the adjustment factor and therefore of the proxy, the Court rejected as unfounded that bank’s criticisms of other aspects of the determination of the amount of the fine imposed on it, and in particular those relating to the methodology adopted by the Commission and its general application.
2143In that context, the Court considers that, in the exercise of its unlimited jurisdiction, it is entitled to adopt the legally well-founded assessments made by the Commission in recitals 810 to 872 and 874 to 878 of the contested decision.…
2147That basic amount need not be adjusted to take account of the Commission’s errors as to the anticompetitive nature of the discussions of 18 May, 22 June and 4 October 2011 (see paragraph 798 above), since they have no impact on the duration of that bank’s participation in the infringement at issue.
2148In those circumstances and in so far as there are no mitigating or aggravating circumstances to be taken into account with regard to Nomura, the fine imposed on that bank is set at EUR 125646000.
IV.Costs
2149Under Article 134(1) of the Rules of Procedure, the unsuccessful party is to be ordered to pay the costs if they have been applied for in the successful party’s pleadings.
2150Since UBS, Natixis, BofA and Portigon have been unsuccessful in all of their claims, they must be ordered to pay the costs in Cases T‑441/21, T‑449/21, T‑456/21 and T‑462/21, respectively, in accordance with the form of order sought by the Commission.
2151Since UniCredit and Nomura have been unsuccessful in almost all of their claims, they must also be ordered to pay the costs in Cases T‑453/21 and T‑455/21, respectively, in accordance with the form of order sought by the Commission.
Operative part
On those grounds,THE GENERAL COURT (Fifth Chamber, Extended Composition)hereby:
1.Joins Cases T‑441/21, T‑449/21, T‑453/21, T‑455/21, T‑456/21 and T‑462/21 for the purposes of the judgment;
2.In Case T‑441/21:– Dismisses the action;– Orders UBS Group AG and UBS AG to pay the costs.
4.In Case T‑453/21:– Annuls the seventh indent of Article 1 of Commission Decision C(2021) 3489 final of 20 May 2021 relating to a proceeding under Article 101 TFEU and Article 53 of the EEA Agreement (Case AT.40324 – European Government Bonds) in so far as it finds that UniCredit SpA and UniCredit Bank AG participated in the infringement from 9 September to 28 November 2011, and not from 26 September to 28 November 2011;– Annuls the fifth indent of Article 2 of Decision C(2021) 3489 final in so far as it sets the amount of the fine for which UniCredit and UniCredit Bank are jointly and severally liable at EUR 69442000;– Sets the amount of the fine for which UniCredit and UniCredit Bank are jointly and severally liable at EUR 65000000;– Dismisses the action as to the remainder;– Orders UniCredit and UniCredit Bank to pay the costs.
5.In Case T‑455/21:– Annuls the second indent of Article 2 of Decision C(2021) 3489 final in so far as it sets the amount of the fine for which Nomura International plc and Nomura Holdings, Inc. are jointly and severally liable at EUR 129573000;– Sets the amount of the fine for which Nomura International and Nomura Holdings are jointly and severally liable at EUR 125646000;– Dismisses the action as to the remainder;– Orders Nomura International and Nomura Holdings to pay the costs.
6.In Case T‑456/21:– Dismisses the action;– Orders Bank of America N.A. and Bank of America Corporation to pay the costs.
Delivered in open court in Luxembourg on 26 March 2025.
Source: EUR-Lex, © European Union, 1998-2026. Reuse authorised under Decision 2011/833/EU.