Evans (Respondent) v Barclays Bank Plc and others (Appellants)

Case summary

Case ID

UKSC/2023/0177

Parties

Appellant(s)

(1) Citibank N.A. (2) Citigroup Inc.

Respondent(s)

Phillip Gywn James Evans

Judgment appealed

Judgment details

Judgment date

18 December 2025

Neutral citation

[2025] UKSC 48

Hearing dates

Full hearing

Start date

1 April 2025

End date

2 April 2025

Justices

Lord Sales

Lord Leggatt

Lord Burrows

Lady Rose

Lord Richards

Judgment details

Michaelmas Term

[2025] UKSC 48

LORD SALES, LORD LEGGATT AND LADY ROSE (with whom Lord Burrows and Lord Richards agree):

1. Introduction

1. Amendments made to the Competition Act 1998 (“the 1998 Act”) in 2015 introduced a new and important procedure in cases where breaches of competition law are alleged. Under section 47B of the 1998 Act, collective proceedings, a form of class action, may be brought in the Competition Appeal Tribunal (“the Tribunal”). Anti-competitive behaviour can affect large numbers of people but the loss suffered by any one individual may be too small or hard to establish to justify bringing a separate claim. The new procedure allows proceedings to be brought on behalf of a class of persons who are affected and allows damages to be awarded for the aggregate loss suffered by the class as a whole, without the need to show what loss each individual member of the class has suffered.

2. The statutory regime gives the Tribunal an important gatekeeper function to ensure that the new procedure is used in appropriate circumstances. Collective proceedings may be pursued only if the Tribunal makes a collective proceedings order. In Merricks v Mastercard Inc [2020] UKSC 51; [2021] Bus LR 25; [2021] 3 All ER 285 (“Merricks”) this court examined the circumstances in which the Tribunal should make such an order. The present case is concerned with another aspect of the regime: when should the Tribunal direct that the collective proceedings are to be “opt-out” rather than “opt-in” proceedings?

3. Where collective proceedings are brought on an “opt-in” basis, the class representative may pursue claims on behalf of every class member who decides to opt in to join the proceedings. Where collective proceedings are brought on an “opt-out” basis, the class representative may pursue them on behalf of the whole class affected by the alleged breach of competition law except for any member who opts out by notifying the class representative that their claim should not be included. Opt-out collective proceedings are an especially powerful vehicle for seeking redress from alleged wrongdoers because they enable damages to be claimed on behalf of what may be thousands or millions of people who have not made any positive choice to sue and may not even be aware of the litigation. But they also carry the risk that defendants may be driven to settle unmeritorious claims by their sheer size and the heavy costs of defending them.

4. Any collective proceedings order made by the Tribunal must specify whether the proceedings are to be opt-in or opt-out collective proceedings. That decision is governed by rule 79(3) of the Competition Appeal Tribunal Rules 2015 (SI 2015/1648) (“the Tribunal Rules”), which gives the Tribunal a broad discretion to take into account all matters it thinks fit in making this decision but expressly stipulates that two matters are relevant. They are: (a) the strength of the claim; and (b) whether it is practicable for the proceedings to be brought as opt-in collective proceedings having regard to all the circumstances, including the estimated amount of damages that individual class members may recover.

5. This appeal is principally concerned with the significance of these two matters. The Tribunal (Sir Marcus Smith and Professor Anthony Neuberger, with Mr Paul Lomas dissenting on this issue) [2022] CAT 16; [2022] Bus LR 1334 decided that the collective proceedings which the respondent to this appeal, Mr Phillip Evans, was seeking authorisation to bring as class representative should not be brought on an opt-out basis. In reaching that conclusion, the Tribunal attached weight to its assessment of the strength of the claim. The Tribunal considered the claim pleaded to be so weak that it was liable to be struck out. It did not at that stage strike out the claim because it thought that, as a matter of procedural fairness, Mr Evans should be given an opportunity to address the Tribunal’s concerns. But the Tribunal treated the weakness of the claim as a powerful factor against allowing the proceedings to be brought as opt-out collective proceedings. The Tribunal also considered that it was practicable for the proceedings to be brought as opt-in collective proceedings and that, although the evidence showed that proceedings would not in fact be brought on that basis, this appeared to be because potential class members who had most to gain from the proceedings if successful did not want to participate in opt-in proceedings and not because opt-in proceedings are not practicable.

6. The Court of Appeal (Green LJ, with Sir Julian Flaux C and Snowden LJ agreeing) overturned the Tribunal’s decision. They considered that, having allowed an opportunity to reformulate the claim, the Tribunal had been inconsistent and unfair in treating its “provisional” view of the merits as “final” when deciding that the proceedings should not proceed on an opt-out basis, knowing that this would bring the claims to an end. More generally, the Court of Appeal considered that the strength of a claim would generally be a neutral factor in deciding whether proceedings should be opt-in or opt-out and criticised the Tribunal for failing to show how its assessment of the merits made opt-in preferable to opt-out. In relation to practicability, the Court of Appeal disagreed with the inferences drawn by the Tribunal from the evidence about the number and size of potential individual claims and regarded this evidence as showing that opt-in proceedings were impracticable. The Court of Appeal remitted the proceedings to the Tribunal to reconsider the merits and set out at some length its own more positive view of the strength of the claim. In expressing that view, the court drew heavily on a decision of the European Commission, published after the Tribunal’s judgment was handed down, finding that a third party had committed infringements of competition law similar to those admitted by the present appellants.

7. On this appeal the appellants argue that the Court of Appeal was wrong to overturn the Tribunal’s decision. They submit that, in doing so, the Court of Appeal improperly interfered with the discretionary case management role of the Tribunal and adopted an approach to the criteria of strength and practicability which, if generally applied, would permit all or virtually all collective proceedings, unless struck out or summarily dismissed as having no real prospect of success, to be pursued on an opt-out basis. These arguments raise questions of general importance in the developing field of collective proceedings. The appeal also raises an issue about the admissibility and relevance in proceedings before the Tribunal of a decision of the European Commission addressed to a third party.

8. Before discussing these issues, we must explain in more detail the factual background and legal context.

2. The infringements

(a) The Commission decisions

9. The claims which the respondent, Mr Evans, is seeking to combine in collective proceedings are what are known as “follow-on” claims for damages pursuant to section 47A of the 1998 Act in respect of infringements established by two decisions of the European Commission. The decisions were both issued on 16 May 2019. They are:

• (i) Commission decision in Case AT.40135-FOREX (Three Way Banana Split) C(2019) 3631 final, addressed to UBS, RBS, Barclays, Citigroup and JPMorgan; and

• (ii) Commission decision in Case AT.40135-FOREX (Essex Express) C(2019) 3621 final, addressed to UBS, RBS, Barclays and Bank of Tokyo Mitsubishi (“BOTM”).

10. The names of the two cases are taken from the names of online chatrooms in which certain individual traders employed by the banks to whom the decisions were addressed exchanged information relating to spot trading of currencies. The Three Way Banana Split decision concerned communications that took place within three private chatrooms collectively called the “Three Way Banana Split chatrooms” between 18 December 2007 and 31 January 2013. The Essex Express decision concerned communications that took place within two private chatrooms collectively called the “Essex Express chatrooms” between 14 December 2009 and 31 July 2012.

11. The Commission found that in each case the banks in question had infringed article 101 of the Treaty on the Functioning of the European Union (“TFEU”) and article 53 of the Agreement on the European Economic Area (“EEA”), which prohibit agreements between undertakings and concerted practices that “have as their object or effect the prevention, restriction or distortion of competition”. Both decisions found infringements by “object”, meaning that the undertakings colluded with the object of restricting or distorting competition. There was therefore no assessment in the decisions of the effect that their conduct had on competition in the relevant market.

12. Both decisions were settlement decisions whereby the undertakings admit their participation in the anti-competitive behaviour described in the decision in return for immunity from fine or a reduced fine. Settlement decisions are typically much shorter and less detailed than ordinary decisions of the Commission issued where allegations of anticompetitive conduct have been disputed.

(b) The infringements found

13. What follows is a much simplified description, sufficient for present purposes, of the infringements established by the two settlement decisions.

14. The infringements related to foreign exchange (“FX”) spot trading in eleven different currencies, referred to as the “G10”. An FX “spot trade” involves an agreement to exchange an amount of one currency for an amount of a different currency at a rate fixed at the moment of the agreement, and where the actual exchange (settlement) usually takes place two days later.

15. For any pair of currencies, a trader dealing in FX typically offers the market a “bid” and an “ask” price. The “bid price” is the price at which the dealer is ready to buy a currency against another, and the “ask price” is the price at which the dealer is ready to sell a currency against another currency. The difference between the bid and ask prices is known as the “bid-ask spread”. Generally, a dealer aims to buy currency at a low price (ie the bid price) and sell it at a higher price (ie the ask price), thereby generating a return on its activities. As such, FX dealers will prefer wider bid-ask spreads, as this increases the potential profits that they can make from their trading.

16. As described in the Three Way Banana Split decision, the gravamen of the infringement in that case was an underlying understanding reached among the participating traders to use the chatroom to exchange commercially sensitive information about certain of their trading activities: para 1. The traders were in near daily communication and agreed to exchange information on an extensive and recurrent basis. This exchange of information, the Commission found, enabled the traders at times to predict more accurately each other’s market conduct, and potentially informed their subsequent decisions, “allowing for occasional opportunistic coordinated behaviour relating to trading activities”: para 49. The Commission described the information they exchanged and its value. It comprised (para 52):

“certain current or forward-looking commercially sensitive information about their trading of either immediate commercial value, or of commercial value lasting for a period of minutes or at most hours after it had been shared, depending on the type of information or until it had been superseded by new updated information that overrode it …”

17. In its legal assessment, the Commission described the parties as having “knowingly substituted practical cooperation between them for the risks of competition”: para 80. The Commission noted that, as this was an infringement by object, there was no need to take into account the actual effects of the conduct: para 86.

18. Applying well-established legal principles to the facts of the case, the Commission found that the exchange of information:

• (i) enabled participating traders trading on behalf of competing undertakings to engage in occasional coordination of their trading activities in specified ways (para 88);

• (ii) may have facilitated occasional tacit coordination of those traders’ spreads, thereby tightening or widening the spread quote in that specific situation (para 89); and

• (iii) led to the traders occasionally coordinating their actions or refraining from action in order to help a member of the chatroom “with higher stakes at play” (para 90).

19. All this activity, the Commission decided, was capable of altering the terms on which the participating traders competed in the market compared to how they would have competed in the absence of that activity. The conduct was therefore capable of affecting the basic parameters on which the participating undertakings should be competing autonomously and was an infringement of article 101 of the TFEU.

20. The Commission found that the anticompetitive conduct had the characteristics of a single, continuous infringement. The significance of this is clear from the case law of the Court of Justice of the European Union: see eg European Commission v Verhuizingen Coppens NV (Case C-441/11P) EU:C:2012:778; [2013] 4 CMLR 8, paras 43 and 44. That finding meant that the five undertakings could be treated as liable for the infringement as a whole, including acts committed by other participants, during their respective periods of participation: see paras 97 to 109 of the decision. Despite this, although the infringement continued from 18 December 2007 until 31 January 2013, the Commission found that only one of the undertakings participated for the whole period when the chatrooms were operating. Others joined and left at different times, with the participation usually ending when the relevant employee left the bank: see paras 37-44, 65 and 148.

21. As regards remedy, the Commission found that the infringement had been committed intentionally rather than negligently: para 152. The approach of the Commission in setting fines is to take as a starting point the value of each undertaking’s sales of the goods or services to which the infringement relates in the relevant geographic area in the EEA. As the aim is to obtain a broad measure of the economic importance of the infringement for the purpose of punishment rather than to estimate harm caused, the value is not limited to sales which are shown actually to have been affected by the cartel: see UBS Group AG v European Commission (Case T-84/22) EU:T:2025:752; [2025] 5 CMLR 12, para 403. Identifying the value of sales was complicated by the fact that, as the Commission noted, G10 FX transactions do not generate any value of sales that is directly traceable in the accounts of the parties: paras 159-161. The Commission therefore adopted a proxy measure based on the (annualised) revenues made by the undertakings during their respective periods of participation in the infringement. This was multiplied by an adjustment factor designed to reflect the applicable bid-ask spreads in G10 FX spot transactions.

22. UBS was granted total immunity from fines because it had been the first to disclose the existence of the cartel. The other parties also received substantial reductions ranging from 10% to 50% for having provided evidence not previously in the Commission’s possession. There was a further reduction of 10% across the board to reflect the fact that the parties had settled rather than fought the case. The fines ultimately imposed ranged from nil for UBS to €310.8 million for Citigroup.

23. The Essex Express decision concerned similar conduct and was drafted on similar lines to the Three Way Banana Split decision. The Commission again found there had been a single, continuous infringement which had as its object the restriction of competition but also found that the duration of the undertakings’ involvement was not the same and was in some cases shorter than the overall period covered by the infringement: see para 148. The fines imposed were nil for UBS as the party which had first disclosed the existence of the agreement to the Commission, €94.2 million for Barclays, €93.7 million for RBS and €69.8 million for BOTM.

3. The legal context

(a) The development of the collective proceedings regime

24. The first competition law collective proceedings regime was introduced by section 19 of the Enterprise Act 2002, which inserted a new section 47B into the 1998 Act to allow collective proceedings to be brought by a “specified body” on behalf of a group of named individual consumers. The consumers had to consent to the proceedings (original section 47B(3)), so the proceedings were opt-in in nature. The only body which was specified under this regime was the Consumers’ Association. The regime was not a success. Only one action was ever brought, for which there was hardly any take-up: see Lloyd v Google LLC [2021] UKSC 50; [2022] AC 1217, para 26.

25. There followed a number of reviews and consultations in the UK and across Europe to examine whether collective action regimes could be made more effective. Consideration was given to the possibility of introducing an opt-out procedure and to the safeguards to be attached to such a regime.

26. On 19 December 2005 the Commission published a Green Paper on damages actions for breach of European antitrust rules (SEC(2005) 1732) COM/2005/672 final, which was subsequently adopted by the European Parliament by resolution dated 25 April 2007. Recital F stated that there was a need for measures to facilitate the bringing of actions for damages, which should increase compliance with European competition law, but at the same time that such changes “should not lead to a situation in which undertakings engaging in lawful economic behaviour are placed at undue risk of having to pay unjustified claims, or to change their behaviour, in order to avoid costly litigation”.

27. The tension between these objectives and the need to strike a balance between them underlies all the reform proposals and is at the heart of the issues arising on this appeal.

28. In April 2007 the Office of Fair Trading (“the OFT”) commenced a consultation based on a discussion paper entitled “Private actions in competition law: effective redress for consumers and business” (OFT916). This noted that “consumers and small and medium-sized businesses (in particular) face a number of practical barriers which have to date made them reluctant to take action to enforce their rights”: para 1.2. The OFT proposed reform of representative actions to help overcome these barriers, whilst seeking to avoid concerns raised by class actions in other jurisdictions (such as in the USA), including that “they are sometimes run in the interests of the law firm (the agent), rather than the claimants (the principals)”: para 4.34. In November 2007 the OFT published its recommendations (OFT916resp) to provide greater access to redress by means of what it called “a balanced system of private actions”. This would allow representative bodies to bring standalone and follow-on representative actions on behalf of consumers and businesses whilst at the same time providing safeguards “to minimise the risk of ill-founded litigation”, which would involve “potential social costs” of which the OFT was “acutely conscious”: paras 4.3, 7.21. Opt-out and opt-in procedures could be provided for, with strong judicial supervision.

29. In 2008 the Commission published a White Paper on damages actions for breach of European antitrust rules. It proposed the introduction of representative actions and opt-in collective actions, but not opt-out collective actions (see para 2.1 of the White Paper COM(2008) 165 final). Ultimately, however, the proposal was not adopted and the Damages Directive (Directive 2014/104/EU OJ L 349/1 of 26 November 2014) did not require Member States to introduce any collective redress mechanism into their domestic procedures: see para 13 of the recitals.

30. In November 2008 the Civil Justice Council in the UK (“the CJC”) published a report with recommendations for reform, entitled “Improving Access to Justice through Collective Actions: developing a more efficient and effective procedure for collective actions”. The recommendations included an opt-out form of collective procedure subject to “positive certification” to ensure that this was the most appropriate procedural mechanism “in terms of efficiency, economy, fairness and access to justice”: para 28. The procedure should involve “a robust assessment of the preliminary merits” to protect defendant interests and avoid “blackmail suits, ie generally unmeritorious actions commenced purely as a means to extract a settlement from a defendant”: para 29. The CJC set up a working group to draft court rules, which reported in 2010. It recommended that the main criterion for certification should be that “the collective proceedings are the most appropriate means for the fair and efficient resolution of the common issues”. The court should be left to choose whether to certify opt-in or opt-out proceedings.

31. In April 2012 the Department for Business, Innovation and Skills (“the Department”) published a consultation paper on options for reform of private actions in competition law. This noted that it was rare for consumers and small and medium-sized enterprises (“SMEs”) to obtain redress and it could be difficult and expensive for them to go to court. Views were sought on forms of opt-out procedure to be applied in the Tribunal to improve access to redress. At the same time, the consultation paper stated that the Government recognised “the importance of ensuring that any changes to the regime do not create a disproportionate risk of exposing business to vexatious or spurious claims, or unwittingly foster a compensation culture. Potential defendants should not be burdened with time-consuming processes and legal costs unless it is clear that there is a case for them to answer”: para 3.19. The paper proposed that appropriate safeguards should be put in place, including a preliminary process of certification, which might involve a preliminary merits test, to ensure that a collective action is the most suitable procedure and to exclude unsuitable cases. The proposed regime should be designed “to prevent vexatious or unmeritorious claims”: para 5.32. While providing access to justice for the claimant as well as increasing deterrence of anti-competitive behaviour, it was “equally important to ensure that the defendant’s right to justice is maintained: the Government does not wish to bring about a regime in which the correct move for a defendant with a strong and winnable case is nevertheless to settle to avoid the risk of damages or legal costs”: para 5.32.

32. The Department published its response to the consultation in January 2013. It decided to introduce a limited opt-out collective actions regime for competition law, limited to proceedings in the Tribunal, with “a set of strong safeguards”, including “[s]trict judicial certification of cases so that only meritorious cases are taken forward”: para 2. There should be “a preliminary merits test, an assessment of the adequacy of the representative and a requirement that a collective action must be the best way of bringing the case”: para 5.55. The regime was intended to “avoid the pitfalls of a US-style class action system”, with “sensible controls on the number and cost of cases”: para 227 of the Department’s final impact assessment, also published in January 2013.

33. In June 2013 the Department published its draft Consumer Rights Bill to carry through these proposals as well as making many other significant reforms to consumer law. It was introduced in Parliament in January 2014 and, in relation to collective proceedings, led to the enactment of revised sections 47A, 47B and 47C of the 1998 Act. Before the introduction of the Bill the Department asked the Tribunal to develop procedural rules to implement the new regime, to inform the legislative scrutiny of the Bill. This step was recommended in the report on the Bill published by the Business, Innovation and Skills Committee of the House of Commons on 23 December 2013 (HC 697-I), in view of the “key role” the rules would have in setting out how the intended reforms would work in practice and because they would “be vital to their effectiveness and clarity”: para 268.

34. The President of the Tribunal established a working party comprising experts in competition law and collective actions to draft the rules. They were published on 10 March 2014, while the Bill was in the Committee stage in the House of Commons. The draft rules were placed in the libraries of the House of Commons and the House of Lords to inform the debate on the Bill. Rule 7 of the draft rules was substantially identical to what became rule 79 (see paras 41-43 below). The significance of the Tribunal Rules for the new regime is confirmed by the express reference to them in the new sections 47A(1) and 47B(1).

35. The Bill became the Consumer Rights Act 2015 (“the 2015 Act”) when it received Royal Assent on 26 March 2015. The legislative history that we have outlined shows that the new collective proceedings regime was intended to strike a balance between facilitating claims for redress, particularly those brought on behalf of consumers and SMEs which it would otherwise be impracticable to bring, and protecting businesses from unmeritorious but burdensome litigation.

(b) The primary legislation

36. The new section 47A of the 1998 Act, introduced by para 4 of Schedule 8 to the 2015 Act, provides that a person may make a claim in the Tribunal for damages for loss arising from relevant infringements, “subject to the provisions of this Act and Tribunal rules” (subsection (1)). Section 47B makes provision for collective proceedings to be brought in the Tribunal to vindicate claims of this kind. Section 47C deals with damages and costs in collective proceedings.

37. Section 47B, introduced by para 5 of Schedule 8 to the 2015 Act, provides in material part as follows:

“47B Collective proceedings before the Tribunal

(1) Subject to the provisions of this Act and Tribunal rules, proceedings may be brought before the Tribunal combining two or more claims to which section 47A applies (‘collective proceedings’).

…

(4) Collective proceedings may be continued only if the Tribunal makes a collective proceedings order.

…

(7) A collective proceedings order must include the following matters—
…

(c) specification of the proceedings as opt-in collective proceedings or opt-out collective proceedings (see subsections (10) and (11)).

…

(10) ‘Opt-in collective proceedings’ are collective proceedings which are brought on behalf of each class member who opts in by notifying the representative, in a manner and by a time specified, that the claim should be included in the collective proceedings.

(11) ‘Opt-out collective proceedings’ are collective proceedings which are brought on behalf of each class member except—

(a) any class member who opts out by notifying the representative, in a manner and by a time specified, that the claim should not be included in the collective proceedings, and

(b) any class member who—
(i) is not domiciled in the United Kingdom at a time specified, and
(ii) does not, in a manner and by a time specified, opt in by notifying the representative that the claim should be included in the collective proceedings.”

(c) The Tribunal Rules

38. The Tribunal Rules came into force at the same time as sections 47A, 47B and 47C of the 1998 Act. Rule 2(2) provides that the rules “are to be applied by the Tribunal and interpreted in accordance with the governing principles set out in rule 4”. In this respect, the Tribunal Rules follow the pattern of the Civil Procedure Rules 1998 in England and Wales, in which rule 1.1 sets out the overriding objective in civil proceedings in similar terms to rule 4 of the Tribunal Rules and rule 1.2 requires the rules to be applied and interpreted so as to give effect to the overriding objective.

39. Rule 4 of the Tribunal Rules provides, in relevant part:

“(1) The Tribunal shall seek to ensure that each case is dealt with justly and at proportionate cost.

(2) Dealing with a case justly and at proportionate cost includes, so far as is practicable—
(a) ensuring that the parties are on an equal footing;
(b) saving expense;
(c) dealing with the case in ways which are proportionate—
(i) to the amount of money involved;
(ii) to the importance of the case;
(iii) to the complexity of the issues; and
(iv) to the financial position of each party;
(d) ensuring that it is dealt with expeditiously and fairly;

(e) allotting to it an appropriate share of the Tribunal’s resources, while taking into account the need to allot resources to other cases; and

(f) enforcing compliance with these Rules, any practice direction issued under rule 115, and any order or direction of the Tribunal.

...

(4) The Tribunal shall actively manage cases.”

40. Rule 41(1) gives the Tribunal power “of its own initiative or on the application of a party” to strike out a claim, including if “it considers that there are no reasonable grounds for making the claim”. Rule 43(1) sets out a power for the Tribunal “of its own initiative or on the application of a party” to give summary judgment against a claimant or defendant, including if it considers that the claimant has no real prospect of succeeding on the claim or issue.

41. Rule 79(2) provides that, in determining whether the claims are suitable to be brought in collective proceedings, the Tribunal shall take into account all matters it thinks fit, including:

“(a) whether collective proceedings are an appropriate means for the fair and efficient resolution of the common issues;

(b) the costs and the benefits of continuing the collective proceedings;

(c) whether any separate proceedings making claims of the same or a similar nature have already been commenced by members of the class;

(d) the size and the nature of the class;

(e) whether it is possible to determine in respect of any person whether that person is or is not a member of the class;

(f) whether the claims are suitable for an aggregate award of damages; and

(g) the availability of alternative dispute resolution and any other means of resolving the dispute, including the availability of redress through voluntary schemes whether approved by the CMA under section 49C of the 1998 Act or otherwise.”

42. Rule 79(3) is central to this appeal. It provides:

“In determining whether collective proceedings should be opt-in or opt-out proceedings, the Tribunal may take into account all matters it thinks fit, including the following matters additional to those set out in paragraph (2)—

(a) the strength of the claims; and

(b) whether it is practicable for the proceedings to be brought as opt-in collective proceedings, having regard to all the circumstances, including the estimated amount of damages that individual class members may recover.”

43. Rule 79(4) provides that, at the hearing of an application for a collective proceedings order, the Tribunal may hear any application by the defendant under rule 41(1) to strike out claims or under rule 43(1) for summary judgment.

(d) The Guide to Proceedings

44. In 2015 the Tribunal issued a Guide to Proceedings. This has the force of a practice direction issued under rule 115 of the Tribunal Rules (which gives the Tribunal power, subject to the Tribunal Rules, to regulate its own procedure). Section 6 of the Guide deals with collective proceedings under section 47B. The question whether proceedings should be opt-in or opt-out is addressed at paras 6.38 and 6.39. Para 6.39 states:

“The Tribunal will consider all matters it thinks fit in determining whether proceedings should be opt-in or opt-out. Rule 79(3) lists two specific factors the Tribunal will consider:

- Strength of the claims (Rule 79(3)(a))

Given the greater complexity, cost and risks of opt-out proceedings, the Tribunal will usually expect the strength of the claims to be more immediately perceptible in an opt-out than an opt-in case, since in the latter case, the class members have chosen to be part of the proceedings and may be presumed to have conducted their own assessment of the strength of their claim. However, the reference to the ‘strength of the claims’ does not require the Tribunal to conduct a full merits assessment, and the Tribunal does not expect the parties to make detailed submissions as if that were the case. Rather, the Tribunal will form a high level view of the strength of the claims based on the collective proceedings claim form. For example, where the claims seek damages for the consequence of an infringement which is covered by a decision of a competition authority (follow-on claims), they will generally be of sufficient strength for the purpose of this criterion.

- Whether it is practicable for the proceedings to be brought as opt-in proceedings (Rule 79(3)(b))

The Tribunal will consider all the circumstances, including the estimated amount of damages that individual class members may recover in determining whether it is practicable for the proceedings to be certified as opt-in. There is a general preference for proceedings to be opt-in where practicable. Indicators that an opt-in approach could be both workable and in the interests of justice might include the fact that the class is small but the loss suffered by each class member is high, or the fact that it is straightforward to identify and contact the class members.”

4. The certification process and the decisions of the courts below

(a) Mr Evans’ pleaded case

45. Mr Evans applied to commence collective proceedings under section 47B of the 1998 Act by filing a claim form on 10 December 2019 seeking damages for losses caused by the infringements found in the two Commission decisions. A similar application had already been brought by Mr Michael O’Higgins. Both applications sought approval to act as class representative for collective proceedings to be brought on an opt-out basis. At this point, therefore, Mr Evans and Mr O’Higgins were competing with each other to secure the certification of the Tribunal for their respective claims, since where more than one applicant seeks approval to act as class representative in respect of the same claims the Tribunal must decide who would be the most suitable: see rule 78(2)(c). And in practice such a choice must also be made where, as here, the proposed claims, though not identical, substantially overlap. In the event the Tribunal decided this issue - referred to as the “carriage issue” - in favour of Mr Evans and the Court of Appeal affirmed that decision. Mr O’Higgins has since discontinued his application for a collective proceedings order and has not participated in this appeal. In our description of the proceedings, we therefore need refer only to Mr Evans’ application and proposed claim.

46. Mr Evans’ amended collective proceedings claim form runs to over 100 pages. Filed with the claim form were expert reports from three different experts, a witness statement from Mr Evans explaining his suitability to act as the class representative, a litigation plan, a litigation funding agreement and an after-the-event (“ATE”) insurance policy to cover any award of costs made against the claimants if the claim fails.

47. For present purposes it is relevant to note the following features of Mr Evans’ claim.

48. Mr Evans sought permission to continue opt-out collective proceedings on behalf of two classes of claimants, designated as Class A and Class B. Class A encompassed persons who entered into transactions directly with the proposed defendants in the EEA during the periods when, according to the Commission decisions, they participated in the infringements. Class B encompassed persons who entered into transactions with other financial institutions in the EEA during the period between 18 December 2007 and 31 January 2013 when one or both infringements were taking place or who entered into transactions with any of the proposed defendants during that period at a time when that defendant was not participating in either of the infringements.

49. The primary effect of the infringements alleged by Mr Evans is that, by unlawfully coordinating their behaviour, the proposed defendants were able to widen the bid-ask spreads that they applied to FX spot trades (and certain forward transactions based on the spot price) of G10 currencies and thereby to offer to sell currency at a higher price and offer to buy currency at a lower price than would otherwise have been the case. It should, however, be noted that, although the claim form describes Class A as encompassing the “direct harm” caused by the infringements, the definition of this class assumes that the discussions in the chatrooms affected many more transactions than those directly affected by the commercially sensitive information exchanged.

50. The Commission found that the exchange of commercially sensitive information between the traders participating in the chatrooms enabled them “occasionally” to coordinate their trading in the ways summarised at para 18 above. The commercial value of the information was said to last for “a period of minutes or at most hours” (see para 16 above). The Commission inferred that this may have enabled the participating traders to quote wider bid-ask spreads for specific currency pairs for certain trade sizes to particular clients during these periods. Class A is not, however, limited to customers who entered into such transactions. It assumes that anyone who entered into a spot transaction with the bank that employed a participating trader to exchange any G10 currency pair at any time during the period when an employee of the bank was participating in any of the chatrooms was quoted a wider bid-ask spread than would have been the case in the absence of the infringing conduct and lost money as a result. A further assumption is that this occurred “irrespective of the method by which FX trading took place”: in particular, that it applied to electronic trading that operated automatically, even though the Commission decisions were confined to “voice trading” conducted by telephone.

51. The definition of Class B assumes that the effects of the infringing conduct extended even further. This class comprises persons who allegedly suffered loss because of a so-called “umbrella effect” (as to which see eg Niels, Jenkins and Kavanagh, Economics for Competition Lawyers, 3rd ed (2023), para 10.2.2 and Kone AG v ÖBB-Infrastruktur (Case C-557/12) EU:C:2014:1317, [2014] 5 CMLR 5, paras 28-30). The theory advanced in the claim form is that unlawful widening of the bid-ask spread by traders participating in the chatrooms “reduced the competitive pressures” on other banks, “enabling them, in turn, to charge widened spreads to their customers”. Further, the infringements were alleged to have increased the costs of buying and selling currency in the inter-dealer market.

52. Mr Evans estimated that Class A would consist of between 14,201 and 42,015 class members and Class B of between 27,814 and 42,015 members. It was recognised that the same person may well be in both classes because of different transactions.

53. The claim form provided “indicative figures” for the quantum of the claims. The figures were calculated by estimating the value of commerce (in terms of the amount of currency traded) falling within the class definitions and applying an estimated overcharge attributed to half of the spread paid by members of the proposed class. An expert instructed by Mr Evans had estimated an overcharge of 18% derived as “a median overcharge drawn from a survey of cartel overcharges”. The value of commerce was estimated at some £41 billion for Class A and £74 billion for Class B. The total indicative estimate of damages claimed by both classes combined (before adding interest) was £2.155 billion.

(b) The Tribunal hearing, later submissions and the judgment

54. The hearing of Mr Evans’ and Mr O’Higgins’ applications for certification took place before the Tribunal over five days from 12 to 16 July 2021. Before the hearing, there was a “teach in” regarding the economics which underlie the proceedings: see para 26 of the Tribunal judgment. The appearances recorded on the front-page of the judgment show that 21 counsel attended, including 11 Queen’s Counsel. The Tribunal noted that the material adduced before them “was vast”. As listed in Annex 2 to the judgment, it included 28 factual witness statements and 15 expert witness reports from five different experts, all instructed by either Mr Evans or Mr O’Higgins. At the hearing, five expert economists, all instructed by the applicants, gave live evidence and were cross-examined.

55. During the hearing, the Tribunal made it clear to the parties that the question of causation “was a matter that was troubling the Tribunal”: see para 133 of the judgment. After the hearing ended, the Tribunal wrote to the applicants asking them to crystallise their current positions on the “theory of harm” they were asserting “in a pleading-style manner” but with the support of their experts. In response to this request, the applicants submitted documents further articulating their respective cases on causation. The proposed defendants provided a joint response to these documents, to which in turn the applicants each replied. The Tribunal therefore had that material before them when writing the judgment.

56. The judgment of the Tribunal was handed down on 31 March 2022: [2022] CAT 16.

57. In light of its concerns relating to the question of causation and the prospect that the litigation of these issues “is going to involve all parties in enormous expenditure of cost and time, and considerable court time”, the Tribunal decided that it should consider of its own initiative whether the applicants had reasonable grounds for bringing the claims: para 148. The Tribunal concluded that both claims, as currently pleaded, were liable to be struck out: para 240. However, the Tribunal also concluded, at para 241, that it would be inappropriate to strike out either claim at that stage, in particular because:

“It is right that the strike-out jurisdiction not be exercised in an area of law that is subject to some uncertainty and is in a state of on-going development and not without the Applicants having the opportunity to address the concerns we have articulated much more clearly in this Judgment than we did during the hearing.”

58. The Tribunal unanimously concluded that each application, if it were the only application in issue, could and should be certified as collective proceedings: para 364. As this court held in Merricks, para 60, when making that determination - in contrast to when choosing between opt-in and opt-out proceedings - the merits of the claims are not a relevant factor.

59. On the issue of whether the proceedings should be opt-in or opt-out, the majority (Sir Marcus Smith and Professor Neuberger) concluded that the claims should not be certified on an opt-out basis. In particular, the majority found that - for reasons that we will consider later in this judgment - the “strength” and “practicability” factors both pointed “clearly and strongly” away from certifying on an opt-out basis: para 384.

60. The majority accepted that proceedings would not go ahead on an opt-in basis. They considered that, “[o]rdinarily, that would be a significant factor. But in this case we have concluded that the claims, as presently framed, are so weak that they are deserving of strike out”: para 385(1); and that “[t]he weight of this factor is also attenuated by the ‘practicability’ factor”: para 385(2).

61. Mr Lomas dissented on the opt-in/opt-out issue. He placed great weight on the need for access to justice and on the likelihood that opt-in proceedings would not proceed at all or, even if they did, would result in very few of the 40,000 proposed class members (conservatively estimated) opting in: see paras 415(2), 451, 456.

62. In the result, the Tribunal stayed the applications and gave permission to each applicant (if so advised) to submit a revised application for certification on an opt-in basis within three months of the date of the judgment: see para 411.

(c) The Sterling Lads decisions

63. Both Mr Evans and Mr O’Higgins appealed to the Court of Appeal against the Tribunal’s decision, with permission granted by the Tribunal.

64. On 5 July 2022, after the Tribunal had handed down its judgment but before the hearing of the appeals, the Commission published two further decisions:

• (i) Commission settlement decision in Case AT.40135-FOREX (Sterling Lads) C(2021) 8613 final, addressed to Barclays, RBS, UBS and HSBC (“the Sterling Lads settlement decision”); and

• (ii) Commission ordinary decision Case AT.40135-FOREX (Sterling Lads) C(2021) 8612 final, addressed to Credit Suisse (“the Sterling Lads ordinary decision”).

65. The Sterling Lads settlement decision was in similar form and made similar findings to the two earlier settlement decisions. The decision concerned communications that took place between certain individual traders employed by the four undertakings to whom it was addressed and by Credit Suisse within a private chatroom called the “Sterling Lads chatroom” between 25 May 2011 and 12 July 2012. As in the Three Way Banana Split and Essex Express decisions, the Commission found that the conduct described amounted to a single, continuous infringement which had as its object the restriction of competition and for which the parties were liable during their respective periods of participation. Again, the parties did not all participate for the full period. The fines imposed ranged from nil for UBS to €174.3 million for HSBC.

66. Credit Suisse did not enter into a settlement with the Commission or admit liability, but was found to have participated in the infringements to the extent and for reasons set out in the Sterling Lads ordinary decision. That decision contains similar findings and legal analysis to the Sterling Lads settlement decision but describes in far more detail the evidence on which the findings were based and records the Commission’s assessment of arguments advanced by Credit Suisse in its defence. The Commission found Credit Suisse liable for the full period of its participation in the chatroom, which was from 7 February 2012 until 12 July 2012, and imposed a fine of €83.3 million.

67. Credit Suisse challenged this decision before the General Court. On 23 July 2025 the General Court upheld the finding of infringement but substantially reduced the fine, concluding that the Commission had used incorrect data to calculate the proxy value of sales leading to a miscalculation of the basic amount of the fine: see Case T-84/22 UBS Group AG v European Commission.

68. After the Sterling Lads decisions were published, Mr Evans amended his claim form to add a claim based on those decisions and joined both HSBC and Credit Suisse as additional proposed defendants in the proceedings.

(d) The Court of Appeal’s judgment

69. The Court of Appeal’s judgment was handed down in final form on 9 November 2023. For reasons that we will consider in more detail when discussing the issues raised on this appeal, the Court of Appeal concluded that the Tribunal had erred in its assessment of the two key factors of strength of claim and practicability: paras 134 and 138. As summarised by Green LJ at para 134:

“First, in relation to “strength”, having concluded that it would form no final view on the merits pending the submission of reformulated cases by the applicants, it was wrong to treat that necessarily provisional view as definitive and accord it more or less decisive weight in the scales against opt-out, knowing and intending that this would bring the claim to an end. Further, it was necessary, in any event, to link any conclusion the CAT did have on strength to the choice it had to make. … it is wrong to treat strength as a sliding scale with a weaker case going to opt-in and a stronger case to opt-out. Secondly, in relation to practicability, the statistical evidence, which in broad sweep is unchallenged, explains why opt-in is impracticable. I respectfully disagree with the inferences drawn from this data and from the evidence by the [Tribunal] majority. That being so there was no reason why the proceedings should not proceed upon an opt-out basis.”

70. The Court of Appeal’s order set aside the Tribunal’s decision on the opt-in/opt-out issue and remitted Mr Evans’ application for an opt-out collective proceedings order to the Tribunal for a further decision and case management in accordance with the Court of Appeal’s judgment.

5. The issues in the appeal

71. There are four issues raised by this appeal:

• (i) Was the Court of Appeal wrong to find that the Tribunal erred in relying on its view that the claims were weak as a factor weighing against opt-out proceedings? (Issue 1)

• (ii) Was the Court of Appeal wrong to find that the Tribunal erred in its assessment of the practicability of bringing opt-in proceedings? (Issue 2)

• (iii) Was the Court of Appeal wrong to hold that the principles of “facilitating the vindication of rights” and deterring future wrongdoers are factors which weigh in favour of opt-out proceedings in general, or in the circumstances of this case? (Issue 3)

• (iv) Was the Court of Appeal wrong to treat the Sterling Lads ordinary decision against Credit Suisse as admissible on a strike-out and to rely on that decision in its own judgment? (Issue 4)

72. We address these issues in turn.

6. Issue 1: the relevance of the strength of the claims to the opt-in/opt-out issue

73. It is very clear from the Tribunal’s judgment, and had been made very clear to the parties in the course of the hearing, what the Tribunal saw as a fundamental difficulty facing the claims. It was not apparent how Mr Evans proposed to prove a causal link between, on the one hand, findings by the Commission of infringements involving four or five named individual traders exchanging information which was commercially valuable for a few hours at most and, on the other hand, loss from unrelated transactions allegedly suffered not only by customers of the banks for whom those traders worked, but also by customers of many other banks across a range of FX deals affecting trillions of pounds worth of commerce.

74. This concern played an important part in two aspects of the Tribunal’s decision. First, it prompted the Tribunal to raise with the parties the possibility that the claim should be struck out under rule 41(1)(b) because there were no reasonable grounds for making the claim. Second, although the Tribunal decided that it would not at that stage strike out the claim, the majority treated its assessment that Mr Evans’ claim was weak as a powerful factor against certifying the claim on an opt-out basis.

75. The Court of Appeal held that the Tribunal had erred in two respects:

• (i) Having decided to allow Mr Evans to submit a reformulated case before reaching a final decision on whether the claim should be struck out, it was illogical and unfair for the Tribunal to treat its necessarily provisional view of the merits as definitive in the context of whether to make an opt-out order (paras 91 and 134).

• (ii) The Tribunal also failed to explain how the merits were relevant to the choice between opt-in and opt-out, given that in the view of the Court of Appeal the strength of the claim is generally a neutral factor (paras 92-93 and 134).

76. On the footing that the Tribunal had not yet decided whether the claim should be struck out and would need to reconsider this question, the Court of Appeal went on to express its view that: (a) the Sterling Lads ordinary decision, issued after the judgment of the Tribunal was handed down, was admissible and had substantial “probative value” in relation to these claims (paras 96-108); (b) disclosure in the proceedings was “intrinsically likely to generate relevant material” (paras 109-112); (c) the Tribunal had extensive powers at its disposal to overcome evidential obstacles and had shown a “more confident and robust approach” in other more recent cases, “which is generally to be welcomed” (paras 113-114); and (d) there was much to be said for the approach to the strength of the case taken in the judgment of the Tribunal member (Mr Lomas) who dissented on this issue (paras 115-116).

77. We have concluded that the Court of Appeal had no proper basis for interfering with the Tribunal’s assessment of the strength of the claim and with the weight that the Tribunal gave to that assessment in choosing between opt-in and opt-out, as we explain in the following paragraphs.

(a) The relationship between strike out and the choice between opt-in and opt-out

78. In our opinion, the Court of Appeal was wrong to identify an inconsistency between the Tribunal’s decision to postpone further consideration of whether to strike out the claim and the view of the majority that the weakness of the claim was a powerful factor against certifying on an opt-out basis.

79. There is no doubt that the Tribunal has jurisdiction to consider of its own motion whether to strike out a pleaded claim even where, for whatever reason, the proposed defendants do not apply for such an order. Collective action claims absorb a very considerable amount of court time and resources. The Tribunal has the power to bring claims to a halt at an early stage if they deserve to be struck out, even if the parties themselves are prepared to spend the money required to pursue or defend the action. Rule 41(1) (referred to at para 40 above) expressly states that the Tribunal may on its own initiative, and after giving the parties an opportunity to be heard, strike out the claim at any stage of the proceedings if it considers that there are no reasonable grounds for making the claim. The same is true in relation to the power to give summary judgment under rule 43(1) (see para 40 above).

80. The Tribunal explained in robust terms what prompted it to raise the striking out issue of its own motion. There has been no suggestion that this led to any procedural unfairness. In deciding whether to strike out the claim, the Tribunal made clear that it was considering all the additional material provided by the applicants, and not just the pleadings. The panel assessing that material included Professor Neuberger, a former director of the Financial Mathematics MSc course at Warwick Business School and currently Professor Emeritus of Finance at Bayes Business School. The Tribunal’s conclusion was that, even assuming that Mr Evans and his team would gather together and present to the Tribunal evidence in support of the pleaded allegations, those allegations themselves did not add up to a plausible theory of harm. As the Tribunal explained in para 140:

“The claims as they have been articulated before us are founded on economic theory. That theory is not articulated in the pleadings, but can be derived from a careful consideration of the expert reports referenced in those pleadings (which we have listed and described in Annex 2 hereto) by a person well-versed in the relevant economics. The Tribunal is fortunate that it has the expertise to consider the implications of the Annex 2 materials.”

81. On this key question of causation, the Tribunal noted at para 239, as regards the Evans application, that even the loss claimed by Class A was not limited to direct harm in the sense of comprising loss suffered only by those persons whose trades were actually the subject of an unlawful information exchange identified in the Commission decisions: the class included anyone dealing with a defendant during the period of its participation in the infringement. The pleaded case did not explain how it was alleged that transactions that were not the subject of an unlawful information exchange were affected. As the Tribunal said at para 239(1):

“The mere fact that a person A trades with a counterparty who – in other, unrelated trades, has behaved unlawfully – is of itself no basis for impugning A’s trades with that counterparty. The causal route by which A is harmed is unclear and not clearly stated in the Evans claim form.”

82. As to Class B, the Tribunal noted that the claim was not put on the basis of market-wide harm but was specific to a segment – albeit a large segment – of the market. But that introduced the prospect of intra-market competition particularly where, as here, the products were fungible and readily substitutable, so that intra-market elasticity ought to be very high. The assertion in the Evans claim form that the infringements enabled the defendants unlawfully to provide a wide bid-ask spread was insufficiently justified.

83. The majority summarised its conclusion at para 237:

“we are satisfied that the facts and matters necessary to support a proper pleading have not been articulated in the pleadings as they stand …. We are satisfied that this is not because of a failure to translate specific details that are contained in the expert reports into a legally framed document. Although the expert reports are detailed, these details amount to no more than a detailed expansion of a theoretical position. Our conclusion is that they do not contain material sufficient to support a proper plea of causation, loss and damage.”

84. In considering how provisional this assessment of the merits was, one must bear in mind that these are follow-on actions relying solely on the infringements found by the Commission as having caused the loss claimed. It took three years of investigation into the foreign exchange markets before the Commission formally opened proceedings. The findings set out in the decisions are the totality of the infringing conduct that the Commission determined had been committed by the addressees of the decisions. Mr Evans is advised by an expert and experienced team of lawyers and economists who, having had sight of the unredacted versions of the Commission’s decisions, should have constructed the strongest case that can plausibly be put forward out of the material they have before they brought the proceedings. The Tribunal was therefore entitled to expect that, by the time of the hearing of the application for a collective proceedings order, the case would be far enough advanced for the nature of the causal link between the infringements relied on and the billions of pounds of loss claimed to be explained. But what was still missing, in the Tribunal’s view, was any properly pleaded case as to how the infringements were alleged to have caused the loss. That was despite the fact that the collective nature of the action itself reduces the burden on the claimants since there is no need to prove that individual members of the class have suffered loss: see Lloyd v Google LLC [2022] AC 1217, paras 30-31.

85. The Tribunal explained at para 241 why, although it was satisfied that the pleadings did not disclose reasonable grounds for making a claim, it ultimately decided not to strike out the claims at that stage. It took account of the fact that the claims raised novel and difficult questions in what was an untested area, citing Hughes v Richards (t/a Colin Richards & Co) [2004] EWCA Civ 266; [2004] PNLR 35 as authority for that being a reason not to strike out a claim. The Tribunal also recognised that the concerns articulated in its judgment were the result of a development of the Tribunal’s thinking during its private deliberations following receipt of the post-hearing submissions. The applicants had not had an opportunity to address the Tribunal’s “final thinking” on the question of pleading.

86. The Tribunal made it very clear that the applicants were, as the Tribunal put it, “in the ‘last chance saloon’” (para 147(3)) and that, absent significant amendment and revision, a future strike out application “may very well be on the cards” (para 241(3)). The Tribunal nonetheless considered, perhaps generously to the applicants, that it should not exercise its power to strike out the claims without giving them a yet further opportunity to attempt to plead a plausible case on causation.

87. It was neither illogical nor unfair to Mr Evans for the Tribunal, while affording him this latitude, to take into account its assessment of the claim as weak in deciding that he should not be allowed to proceed on an opt-out basis. The Tribunal had been able to analyse the claims and the difficulties which they faced on the issue of causation in considerable depth. Mr Evans was not asking the Tribunal to postpone its decision on opt-in or opt-out. The Tribunal was entitled to take the view that, even if (which they clearly thought unlikely) Mr Evans could plead a case which did not deserve to be struck out, the prospects of the claim succeeding were very low. It was entirely legitimate for the Tribunal to attach weight to that view and to treat it as a powerful reason against certifying on an opt-out basis.

(b) The relevance of merits to the choice between opt-in/opt-out

88. The other error which the Court of Appeal attributed to the Tribunal was a failure to explain how its negative view of the merits of the claim made opt-in collective proceedings preferable to opt-out. There is, however, no mystery about the significance of this factor. If a claim is weak, that militates against affording claimants the advantages of an opt-out process, with the concomitant disadvantage that a defendant may feel commercial pressure to settle such a claim even though it would be likely to fail at trial. This is apparent from the history of the legislation discussed at paras 24-35 above and is made clear in the Guide at para 6.39 (see para 44 above).

89. In Merricks, two of us (Lord Sales and Lord Leggatt) noted, by reference to the legislative history and experience in other jurisdictions, that collective proceedings - particularly if they are opt-out - confer substantial legal and commercial advantages on claimants and burdens on defendants which are capable of being exploited opportunistically. As stated in Merricks at para 98:

“A class action procedure which has these features [the potential to opt out and to recover aggregate damages] provides a potent means of achieving access to justice for consumers. But it is also capable of being misused. The ability to bring proceedings on behalf of what may be a very large class of persons without obtaining their active consent and to recover damages without the need to show individual loss presents risks of the kind already mentioned, as well as giving rise to substantial administrative burdens and litigation costs. The risk that the enormous leveraging effect which such a class action device creates may be used oppressively or unfairly is exacerbated by the opportunities that it provides for profit.”

Although this was said in a minority judgment, the majority judgment in Merricks given by Lord Briggs said nothing inconsistent with these observations and we do not consider them to be controversial.

90. In the present case the Tribunal in its judgment spelt out in detail the “leveraging” effect of opt-out collective proceedings, explaining the incentives which the opt-out procedure creates to settle even weak claims for more than merely “nuisance value” and hence to bring weak claims as opt-out claims: see para 88(3). This “leverage” includes, in particular, the fact that, if claims proceed to judgment, damages are awarded in respect of the whole class even though it is well understood that not all of the class members will come forward to claim their entitlement.

91. If the merits of a claim are very weak, as the Tribunal assessed them to be in this case, it is unlikely that the Tribunal will be justified in conferring on the proposed class of claimants the significant leveraging advantage associated with opt-out proceedings. To allow such a claim to go forward on that basis would be unlikely to be fair, as between claimants and defendants, and would tend to be contrary to the obligation on the Tribunal under rule 4(1) to deal with the case “justly”, both as a general matter and by reference to the individual elements of that overriding objective specifically identified in rule 4(2) (para 39 above).

92. As regards rule 4(2), if an opt-out order were made in such circumstances the parties would not be likely to be “on an equal footing”, because the claimants would be given assistance to extract value from unmeritorious claims against defendants who would thereby be impeded in defending themselves and in vindicating their own rights. There would tend not to be a “saving [of] expense”, but rather the exposure of the defendants to having to incur expense without that being justified by the need to secure access to justice for claimants to vindicate meritorious claims. The method of dealing with the case would be unlikely to be “proportionate … to the amount of money involved” and “the importance of the case”, since if the claims are unmeritorious then ultimately no money is properly or substantially in issue and the case is not important. The Tribunal would be unlikely to ensure that the case “is dealt with … fairly”, because if the claims are unmeritorious it is usually not fair to expose the defendants to the leveraging advantage which an opt-out order would give the claimants. The Tribunal would also be likely to fail to allot “an appropriate share of the Tribunal’s resources” to the claim, “while taking into account the need to allot resources to other cases”, because the Tribunal’s time and attention would have to be diverted to managing complex proceedings which did not reflect meritorious claims, and diverted away from managing and trying other cases.

93. To allow collective proceedings to be pursued on an opt-out basis in such circumstances would also tend to be contrary to the Tribunal’s obligation under rule 4(1) and (2) to deal with a case “at proportionate cost”, for similar reasons. The aim under this limb of rule 4 should be for the cost of proceedings to be proportionate to the contribution they are capable of making to a just disposal of the underlying dispute, involving the vindication of rights of claimants and defendants. But if it appears that the claims do not have merit, the proceedings will not contribute to the just resolution of the dispute or to the genuine vindication of rights, and the cost involved will tend to be disproportionate to those aims. Similarly, if the claims are of little or questionable merit, the advantages for claimants associated with opt-out proceedings, and the concomitant risk that this tactical advantage will result in the defendant settling the case on terms which do not reflect the true rights of the parties, will be out of proportion to those aims.

94. To sum up, if the claim is very weak it is likely to be more difficult to justify resort to the opt-out procedure as striking a fair procedural balance between the claimants and the defendant(s). It is also likely to be difficult to justify resort to the opt-out procedure as a proportionate way of providing access to the Tribunal for the vindication of such a claim.

95. The assessment involved is not all or nothing. The detriments and potential unfairness to a defendant of an opt-out procedure are real, and do not drop out of the picture just because the claimants have claims which might have some merit. The less the Tribunal has confidence about that, the harder it is to conclude that imposing those additional detriments on the defendant is fair and a price worth paying in the overall interests of justice. Contrary to the conclusion of the Court of Appeal (see para 69 above), it is indeed appropriate, in our view, to view the strength of the claim in terms of a sliding scale.

96. The Tribunal adopted this approach, stating (para 374(3)) that “as a general rule it seems to us that the weaker a case, the less justification there is for certifying on an opt-out basis. Of course, this is one factor amongst many”. We agree. The Tribunal analysed the pleaded causes of action with care, in the light of Mr Evans’ proposed economic theory of harm. Having done so, it concluded that the pleaded case on causation was both intrinsically weak and that it lacked the necessary particularity: para 375. On the basis that the pleaded claims suffered from such fundamental weakness, we consider that the Tribunal majority were right to regard this as “a powerful reason against certifying on an opt-out basis”: ibid (emphasis in original).

(c) Reliance on Le Patourel

97. The Court of Appeal referred to its earlier decision in Le Patourel v BT Group plc [2022] EWCA Civ 593; [2022] Bus LR 660; [2023] 1 All ER (Comm) 667 (“Le Patourel”) as authority for the proposition that in most cases the strength of a claim will be a neutral factor in the choice between opt-in and opt-out. In Le Patourel the Court of Appeal upheld an opt-out collective proceedings order granted by the Tribunal in respect of a claim by customers of BT for aggregate damages for alleged abuse of a dominant position. BT had applied for summary judgment to dismiss the claim, but that application had failed. The Tribunal had considered the strength of the claim again at the stage of deciding between opt-in and opt-out and said (as quoted at para 102 of the Court of Appeal’s judgment):

“But that could only assist BT here, in our view, if it could persuade us that this is a very weak claim even if it could surmount the summary judgment/strike-out threshold. For all the reasons given, however, we are not so persuaded.”

98. BT argued that the Tribunal had misdirected itself by requiring a defendant to establish that a claim is “very weak” before opt-in becomes appropriate. BT contended that the Tribunal should have adopted a quite different approach and required the applicant to demonstrate that the claim had strong merits in order “to displace the general preference under the legislative regime for proceedings to be certified as opt-in”: para 103. The Court of Appeal rejected that contention.

99. In doing so, the Court of Appeal started from the position that, where a collective proceedings order is to be made, section 47B and rule 79 do not establish any legislative presumption in favour of either opt-in or opt-out; rather, the choice is to be made by the Tribunal on the basis of all the circumstances of the case: para 68. We agree that there is no general legislative presumption one way or the other as regards the choice between opt-in and opt-out proceedings and that the Tribunal has a wide discretion to decide between them. At the same time, however, rule 79(3) specifically identifies the strength of the claim as a factor which, depending on the circumstances, may have weight in exercising that discretion. The legislation would not have singled out this factor (and that of the practicability of opt-in proceedings) from the entire range of factors potentially relevant to the choice between the opt-in and opt-out procedures if it was intended that it should generally be regarded as a neutral consideration. The object of these procedures is to promote access to justice for claimants while not imposing a disproportionate burden on defendants. Therefore, the natural explanation of the significance attached to this factor is that the stronger the claim, the greater the confidence the Tribunal can reasonably feel that certifying an opt-out claim will achieve a just outcome (eg by way of settlement) notwithstanding the additional burden imposed on the defendant.

100. The Court of Appeal in Le Patourel further observed, at para 105, that:

“in many cases the [Tribunal] might be hard pressed at the certification stage to form a sufficiently clear view of the merits to enable it, with confidence, then to take merits any further in the overall balancing exercise. This might be because of the complexity of the legal and economic issues arising and the absence at the certification stage of expert evidence and disclosure.”

It is certainly true that in cases where the Tribunal concludes that it cannot form a view of the merits at the certification stage, it may be unable to place any weight on this factor when choosing between opt-in and opt-out.

101. Equally, however, there will be cases in which the Tribunal does feel able to form a view, albeit necessarily provisional, of the strength or weakness of the claim to which it considers that weight can properly be attached in deciding between opt-in and opt-out proceedings. If the Tribunal is able to form the view not only that the claim has a realistic prospect of success but that it appears to be a strong case, that will provide support for a decision to make the proceedings opt-out. The same point applies in reverse. There can equally be cases - of which the present case is one - where, even if the claim surmounts the summary judgment/strike-out threshold, the Tribunal is able to form the view that the claim is so weak that this points against allowing collective proceedings to be brought on an opt-out basis.

102. The Court of Appeal in Le Patourel had difficulty in making sense of the Tribunal’s reasoning on this point. This difficulty stemmed from the Court of Appeal’s view that “finding that a claim has a realistic prospect of success and is not fanciful and is more than arguable, logically precludes a conclusion that it is simultaneously ‘very weak’”: para 109. We disagree. Surviving strike-out or summary judgment is a low bar and the merits of the claim may not be the only factor at play when summary disposal is rejected. There is nothing illogical in finding that a claim is not susceptible of being struck out or summarily dismissed and yet finding also that the claim appears to be very weak. Nor is there any illogicality in treating the weakness of the claim as pointing against opt-out proceedings even if it is not so weak as to be struck out. Indeed, that is precisely what rule 79(3) contemplates in specifically identifying the strength of the claim as a factor relevant to the choice between opt-in and opt-out proceedings. That choice only arises where the claim has not been struck out or summary judgment granted.

103. The same misconception that, if a claim is not capable of being struck out or summarily dismissed, its perceived lack of merit cannot militate against certifying on an opt-out basis is reflected in the Court of Appeal’s judgment in the present case.

104. Here the Court of Appeal has gone further than in Le Patourel in asserting that “[g]enerally, the strength of a claim will be neutral” (para 93) and “in most cases the merits will be a neutral factor” (para 134). In our view, these assertions are unwarranted. We would expect that by the time of the certification hearing the Tribunal will be able in many, if not most, cases to form a view (even if provisional) of the merits of the claim and, where it is able to do so, the merits should not be a neutral factor for the reasons we have given.

105. Counsel for Mr Evans sought to rely upon the statement in para 6.39 of the Guide (para 44 above) that in proceedings which follow on from a decision of a competition authority (such as the Commission) the claims “will generally be of sufficient strength for the purpose of [the] criterion” in rule 79(3)(a). But in the present case the claimants are only able to bring follow-on claims in respect of decisions which found that the object of the practices was to restrict competition. The Commission’s decisions did not determine what the effects of the practices were and hence did not cover a necessary element in the damages claims being brought, being to establish causation of loss. The Tribunal considered that it was precisely this element which the claimants had failed to explain, despite being given more than one opportunity to do so. On a proper reading, the statement in para 6.39 must be interpreted to mean that the relevant decision will satisfy the strength of claim criterion to the extent that the decision covers the elements which have to be satisfied in order for the claim to succeed.

(d) Other points made on the strength of the claim

106. We would add that the further opinions expressed by the Court of Appeal about the strength of the claim were also, in our view, misplaced. In summary, they were that the exercise of disclosure would require the defendants to disclose vast quantities of data and information and was intrinsically likely to generate relevant material. Mr Evans said in his claim form that his expert witness would use multiple regression analysis using data from the proposed defendants, multi-bank platforms, CLS Bank International (which settles FX transactions) and other inter-dealer trading platforms. The Court of Appeal suggested that such material might provide a basis, for example, for comparing the profitability of the defendants’ trading during periods when they did or did not participate in infringements and when the infringements were or were not taking place: paras 109-112. Further, the Court of Appeal noted that in decisions reached more recently than its decision in this case, the Tribunal took a more confident and robust approach to evidence and that was generally to be welcomed: paras 113-114.

107. These comments do not give sufficient credit to the rigorous analysis of Mr Evans’ case on causation which the Tribunal majority carried out. Further, they do not address the fundamental difficulty identified by the Tribunal, consisting in the mismatch or gap between (1) findings that information exchanges may occasionally have enabled participating traders to increase their spreads for particular transactions and (2) a claim predicated on an assumption that spreads were increased on every FX G10 transaction entered into by the bank employing that trader with every client of the bank throughout the period of the trader’s membership of the chatroom.

108. Speculating that something may turn up if the defendants are compelled to incur the major inconvenience and expense of a huge disclosure exercise is not a sufficient reason for subjecting them to that inconvenience and expense. There is no reason to think that statements from witnesses of fact might cure the fundamental deficiency in the case on causation. Likewise, wielding a broad axe, as the Court of Appeal suggested, is sometimes appropriate to overcome difficulties of precise quantification of loss, but it cannot render credible a case which is not based on any plausible theory of harm.

109. For our part, therefore, we do not see the force of the points about the strength of the claim made by the Court of Appeal. But, more importantly, it was inappropriate for the Court of Appeal to engage in making its own assessment of the merits when the Tribunal’s approach and decision could not be impugned - all the more so when the Court of Appeal relied heavily on material which had not been and could not have been considered by the Tribunal. The Court of Appeal should have left the evaluation of the strength of the claim to the specialist body entrusted with that task.

(e) Conclusion on this issue

110. The Tribunal was not using the choice of opt-in as a “sanction” against a non-viable claim, as the Court of Appeal suggested at para 92. The Tribunal took the applicants’ assertion that the claim would collapse if certified as opt-in at face value and viewed this as “a factor that points strongly in favour of certifying on an opt-out basis” (para 385). But in our view the Tribunal was right not to treat this factor as if it were a trump card. There are obvious dangers in allowing applicant class representatives to make such an assertion in terrorem if that stance clinches for them and for the principal members of the class the advantages of the opt-out process. The Tribunal rightly recognised that the potential collapse of the case meant that particular care was needed in assessing the strength of the claim and deciding what weight to give it: see para 118. We consider that the weight that it ultimately decided to give that factor was well within its discretion.

111. In our view, the Tribunal’s conclusion that Mr Evans’ claim was weak and that this was a powerful factor militating against certification of the proceedings on an opt-out basis was fully and convincingly reasoned and there was no ground on which the Court of Appeal was justified in interfering with it.

7. Issue 2: did the Tribunal err in its treatment of practicability?

112. The implication of rule 79(3)(b) is that, if it is practicable for the proceedings to be brought as opt-in proceedings, then generally speaking they should be. The principle is clear: if it is practicable for claimants to bring opt-in proceedings, it is unlikely to be proportionate to confer upon them the additional advantages associated with the opt-out procedure and unlikely to be reasonable to expect the defendants to have to face the additional commercial pressures to settle the claims regardless of their true merit which that would involve. As with the strength of the claim, this is only one factor (though clearly, since it is singled out in the rule, potentially an important one) which feeds into the overall evaluation which the Tribunal has to make. Its weight may vary from case to case, depending on just how difficult it might be for the claimants to bring their claims on an opt-in basis. The Tribunal noted this at para 121 of its judgment.

113. The legal significance of rule 79(3)(b) is similar to that of rule 79(3)(a). Both have to be read and applied in the context of the overriding objective of dealing with cases justly set out in rule 4 of the Tribunal Rules. There are risks of unjust outcomes on both sides. If refusing to allow opt-out proceedings means that rights are not duly enforced, that is unjust to prospective claimants. On the other hand, if allowing opt-out proceedings results in defendants paying off claims which in reality are not valid, that is unjust too. In deciding between opt-in and opt-out proceedings, the Tribunal has to seek to advance the interests of justice overall in circumstances of uncertainty.

(a) The nature of the Tribunal’s assessment

114. The open-textured nature of the question, the difficulties of assessment and the range of cases in which it arises mean that the Tribunal has a wide case-management power of evaluation and discretion in making its decision. In addition, the Tribunal has expertise in making relevant economic assessments and practical experience in managing collective proceedings which mean that it is well placed to understand and balance the risks of unjust outcomes associated with this sort of complex litigation and to make the required overall assessment of how the interests of justice are best served. The choice between opt-in and opt-out collective proceedings involves an element of trade-off between different aims, in particular the need to seek to ensure that people with valid claims should have a reasonable opportunity to vindicate them and the need to seek to ensure that defendants are treated fairly and are not subjected to tactical and commercial pressure in the proceedings beyond what it is reasonable to expect them to face, as a proportionate means of securing that opportunity.

115. Another feature of the collective proceedings regime is also relevant. The regime provides additional opportunities for claims to be brought and vindicated, but it does not guarantee that this will be possible in every situation where there has been a breach of competition law. The regime still operates on the basis of commercial factors that affect whether individual claimants or third party funders are prepared to provide the funding necessary to bring the proceedings. What the regime seeks to provide is a reasonable opportunity, not available previously, for a group of claimants to club together (or, in the case of opt-out proceedings, to be grouped together by a claims entrepreneur) to vindicate their rights in a way which would not be practical or cost-effective if they had to proceed individually. But the regime is not intended to immunise them completely from the usual background commercial considerations which have a role in determining when proceedings are thought to be sufficiently viable to be worthwhile.

116. The regime is designed to accommodate a wide spectrum of cases, with different features, to allow classes of claimants a reasonable opportunity to litigate, but not an absolute or specially privileged right to do so. At one end of the spectrum, there may be a large class of ordinary consumers affected by a breach of competition law resulting in their having to pay more for goods or services, but involving small sums for each individual which would make each individual case economically unviable given the costs involved. This was the type of situation addressed in Merricks and is the paradigm type of case in which opt-out proceedings are likely to be justified. The large numbers of claimants and the small sums involved are likely to make an opt-out procedure the only basis on which it is practicable to bring proceedings.

117. At the other end of the spectrum, the claimants who allegedly suffered loss as a result of breaches of competition law may be large commercial organisations, well capable of looking after their own interests, and the sums involved may be large enough to make it financially viable to contemplate bringing proceedings to seek recovery. That may be done by clubbing together with other similar organisations with similar claims to bring a single set of proceedings with multiple claimants, or a number of separate proceedings that are case managed together, rather than trying to bring a collective action with a class representative. In such circumstances it may not be appropriate for collective proceedings to be brought at all; and if collective proceedings are considered appropriate, it is likely to be only on an opt-in basis.

118. The present case falls between these two poles. It also demonstrates the way in which background commercial considerations continue to operate regarding decisions whether to litigate, albeit through the prism of the collective proceedings regime and in light of the additional opportunities opened up by that regime.

119. In assessing the practicability of bringing opt-in proceedings, we agree with the Tribunal (see para 122(5) of its judgment) that an objective assessment is called for, in the sense that it should not depend on evidence about the subjective state of mind of, and legal advice received by, individual potential claimants. The subjective state of mind of a claimant regarding whether it is worthwhile to join in opt-in collective proceedings will almost invariably be informed by legal advice received. That advice can be expected to cover the likelihood of the claim succeeding or settling, the amount of damages that could potentially be recovered and the likely costs and risks of litigation. Legal professional privilege will attach to that advice, but it is difficult to interpret the response from potential claimants without knowing how they have been advised. So there are legal barriers to approaching the question of practicability on a subjective basis. The Tribunal could not be confident that it has an accurate picture of a claimant’s state of mind without legal professional privilege being waived, and it would be inappropriate to place a party under pressure to waive such privilege as the price for seeking to persuade the Tribunal to certify opt-out proceedings. Moreover, where there are large numbers of potential opt-in claimants, as will be usual, it is not feasible to produce evidence of the subjective states of mind of all of them. More fundamentally, the Tribunal’s decision as to the appropriate procedure should not depend on the quality and accuracy of the advice which potential claimants have in fact received. The inference to be drawn is that rule 79(3)(b) requires the Tribunal to stand back and make an objective determination by reference to the inherent likelihood of how a reasonable person in the position of a potential claimant would assess the situation.

120. This does not require the Tribunal to assume that the position of all potential claimants is the same. It is relevant to consider the composition of the proposed class and whether it is relatively homogeneous or comprises potential claimants of different types and with different sizes of potential claim. The identification of distinct groups of claimants is a matter for evaluative assessment by the Tribunal depending on the particular facts, and is one in relation to which it again has a wide discretion. In our view, where the Tribunal identifies groups of claimants with distinct profiles relevant to the assessment to be made, it should consider the practicability of bringing an opt-in claim for each group separately; and if this yields a different conclusion for each group, it should then stand back and make an overall assessment of the balance of justice having regard to those underlying assessments. In substance, this is the approach which the Tribunal majority adopted in this case.

121. There is something of a tension between what the majority say at para 122(5) and at para 122(6) of its judgment. At para 122(5) it embraces a test of the reasonable putative class member, which might be taken to suggest that some notional average member has to be identified as representative of the whole class. But at para 122(6) the majority correctly points out that the class may be diverse, with claims of very different sizes or members with varying degrees of knowledge and experience. It goes on to say that, in considering practicability, it is then incumbent on the Tribunal to have regard to differences in class composition. We agree that this is the correct approach. We also agree that, as the Tribunal observed, this is a very fact sensitive question. If the Tribunal assesses that there are elements within the class which it is appropriate to view as distinct, ordinarily it should analyse each category by reference to a typical member within it. Then the Tribunal has to stand back to make an overall assessment of practicability which takes account of the features of those categories.

(b) The Tribunal’s assessment in this case

122. According to the Tribunal’s assessment of the evidence in the present case (at para 381 of the judgment), there is a group of financial institutions and fairly large commercial entities, likely to be sophisticated potential litigants, which have allegedly suffered what on average are significant losses. There is also a very large number of individuals and smaller entities who have allegedly suffered loss but only in very small amounts. There are members of the former group who, given their sophistication and the potential size of their claims, would be in a position to bring claims on an opt-in basis, if they so chose. It could be inferred from their unenthusiastic reaction to efforts by Mr Evans and his solicitors to interest them in bringing proceedings that, for reasons of their own, they did not wish to participate in such an action. As regards the latter group, the Tribunal accepted, at para 381(10), that it would not be practicable for them to pursue opt-in proceedings. Their claims will be small, they may well not realise that they have a claim because they made a few FX trades many years ago and, if they know that in theory they have a claim for a very small sum of money, they may not think it worth the trouble to pursue it.

123. In the light of its assessment of the composition of the class, the Tribunal majority stood back and made the overall assessment which was required. It concluded that it should not allow a sub-class of persons whose total claims were by value a tiny fraction of the aggregate claim to alter their conclusion on practicability. “That would be to allow the tail to wag the dog”: see para 381(10).

124. It is difficult to see why the financial institutions and large entities with substantial claims should be allowed to proceed by way of opt-out collective proceedings. The additional leverage that this would give them cannot be justified having regard to their particular circumstances. The Tribunal was entitled to conclude that they should not be permitted to bolster their position by pointing to and bundling themselves together with another group of potential claimants with a different profile who, for quite separate reasons, are not able to bring their claims. It would certainly not be appropriate for the Tribunal to order opt-out proceedings simply to give the large commercial operators the advantage arising from the expectation that the smaller entities for whose benefit the opt-out proceedings are supposedly being brought will increase the overall award of damages but will not ultimately come forward at the end of the day to claim their share of the award.

125. It would in principle be possible to design a class which did not include large financial institutions and for an applicant to seek authorisation to pursue collective proceedings representing such a class of smaller firms and individuals on an opt-out basis. But it is evident that such an action would not be attractive to funders. Judged from the perspective of this group of potential claimants, their inability to sue is due to background circumstances and risks from which the regime is not designed to relieve them. Their claims are simply too lacking in value to make the opportunities for suit opened up by the collective proceedings regime available to them as a class.

126. One could also imagine a case where the profile of the claims was different, with the main core of the value of the class action consisting in a mass of small claims by consumers equivalent to those in Merricks but with a small rump of bigger claims. It would be open to the Tribunal to decide that the entire set of claims should be grouped together and be brought in one set of proceedings on an opt-out basis and that the incidental advantage the claimants in the rump might thereby gain did not significantly distort the balance of procedural advantage as between the claimants and the defendant. This counter-example serves to emphasise the importance of the role of the Tribunal in exercising its discretion, and making the evaluative judgments required, in carrying out its function as gatekeeper in relation to collective proceedings.

127. We consider that the assessment of practicability made by the majority of the Tribunal in this case was within the ambit of the evaluative judgment legitimately open to the Tribunal to make.

(c) The approach of the Court of Appeal

128. In our view, the essential error made by the Court of Appeal in its analysis of practicability was that it failed to follow the sound guidance given in its earlier judgment in Le Patourel about the approach the Court of Appeal should adopt when considering evaluative judgments made by the Tribunal and instead made its own assessment of the evidence, substituting its own evaluation for that of the Tribunal.

129. In Le Patourel, at para 57, the Court of Appeal considered the circumstances in which it is legitimate for the Court of Appeal to interfere with the evaluation of facts made by the Tribunal. Green LJ said that:

“when it comes to the weighing up of the various factors relevant to the choice of opt-out or opt-in this is essentially an exercise of judgment over facts and evidence by an expert, specialist, body, that will over time accrue an increasing well of experience in how to handle these complex cases. The appellate courts recognise that the case management decisions of the [Tribunal] are exercises in pragmatism and that undue formalism and precision are not required … These considerations broaden the Tribunal’s margin of discretion or judgment. This Court should not interfere simply because it might, for the sake of argument, have drawn a different conclusion from the weighing exercise. We would expect that most opt-out/opt-in decisions will involve a weighing exercise of this nature.”

130. In Le Patourel the Court of Appeal also pointed out, at para 83, that the concept of practicability is “highly fact and context sensitive and will include, as one facet of this broad analysis, the Tribunal using its expert judgement to envisage how the costs and benefits of litigation will play out upon an opt-out or opt-in basis. The view the Tribunal takes will be a conclusion reached by an expert body with a growing depth of experience in the conduct of collective proceedings”. In that case the Tribunal had examined relevant factors such as size of class, the scale of a possible award and the impact of these on funding as important considerations. The Court of Appeal concluded, at para 74:

“The [Tribunal] came to specialist conclusions which lay squarely within its broad margin of judgement. There is in our judgment no basis in law upon which this court can properly interfere.”

131. In the present case the Court of Appeal, at paras 47 and 124, quoted these passages from its earlier judgment in Le Patourel but failed to heed them. Instead, the court embarked on its own examination and assessment of factual and expert evidence presented by Mr Evans to the Tribunal. It annexed to its judgment a table drawn from that evidence showing “a breakdown of the average claim per class member both on an aggregate basis but also divided up into categories based upon type of financial institution concerned (as class member) and annual turnover”. Green LJ observed, at para 122, that, of the total of over 18,000 financial class members, the table showed that the great majority (approaching 16,000) had modest annual turnovers of under £500,000. Whilst the average claim across the whole class was £133,805, the typical size of claim for most class members was only around £16,000. He considered that “even for the largest class members [measured by turnover] the sums at stake are relatively modest and on an opt-in basis could be dwarfed by the costs”: para 123. That seems to us debatable given that the table shows that the 342 institutions with annual turnovers of more than £50 million had a total claim valued at £1.1 billion and an average claim size of £3.4 million, while the 338 class members with annual turnovers of £10-50 million had an average claim size of £2.2 million and a total claim of £750 million. But the main point made by Green LJ was that the Tribunal had been wrong to focus on the average size of claims when “the scale of typical claims is modest and the size of typical claimants is not large”: para 127. He concluded:

“It might be that the total claim, as estimated, is between £2.1bn and £2.7bn; but it is the claim of the individual that determines whether that individual is willing to become embroiled in litigation. I would respectfully disagree with the inferences the [Tribunal] has drawn from the Table.”

132. In his overall conclusion at para 134 of the judgment (quoted at para 69 above) Green LJ repeated that, in relation to practicability, he disagreed with the inferences drawn from this data and from the evidence by the Tribunal majority. He went on to quote extensively, at paras 135-137, from the analysis of practicability in the dissenting judgment of Mr Lomas. Although Green LJ did not explain the relevance of these quotations other than “for completeness”, the implication is that he preferred the evaluation of Mr Lomas to that of the Tribunal majority.

133. There is certainly room for disagreement about what inferences should be drawn from the data that was placed before the Tribunal about claim sizes and class composition and what this showed about the practicability of bringing opt-in collective proceedings, having regard to all the circumstances, including the estimated amount of damages that individual class members may recover. That is apparent from the fact that the Tribunal was not unanimous in its view on this issue. But it was not for the Court of Appeal to undertake its own assessment and to substitute its view for that of the Tribunal because it disagreed with the inferences drawn from the evidence by the majority and preferred the opinion of the minority. Rather, it should have followed its earlier guidance given in Le Patourel and refrained from interfering with the conclusions of the specialist tribunal which, under the collective proceedings regime, has been tasked with making the relevant evaluation, unless there was an error of law in the Tribunal’s approach.

134. The Court of Appeal did not identify any such error of law. But in seeking to defend the Court of Appeal’s conclusion, counsel for Mr Evans have argued on this appeal that the Tribunal made an error of law by basing its judgment on a single average claim value of around £62,000 (with simple interest) across the whole class and failing to consider the more granular average claim size estimates for different sub-groups considered by the Court of Appeal. It is also said that the Tribunal erred by aggregating the small claim values of many thousands of class members and treating the small aggregate amount as indicating that their interests were unimportant, thereby ignoring the fact that for the vast majority of the potential class members it was not practicable to bring opt-in proceedings.

135. Although there are passages in the judgment of the Tribunal which do refer to the average size of claims across the whole of Classes A and B, we do not think it fair to infer from these references that the majority failed to consider the composition of the class and the different sizes of claim potentially available to different sub-groups within it. The Tribunal majority was plainly aware of (and referred to) the data shown in tables compiled by one of Mr Evans’ expert witnesses, Mr Ramirez. The majority expressly emphasised “the danger of averages” and that “an average may be more misleading than helpful”: see para 381(5) and (6). The judgment of the majority can be criticised for not expressly discussing what Mr Beard KC described as “the missing middle”: that is, the position of financial institutions with relatively modest annual turnovers. But it cannot be inferred that, just because a particular point is not mentioned in a judgment, it was ignored by the Tribunal. An appeal court should assume that a lower court or tribunal took into account all the evidence presented to it, unless there are clear indications to the contrary: see eg Volpi v Volpi [2022] EWCA Civ 464; [2022] 4 WLR 48, para 2(iii); Gift v Rowley [2025] UKPC 37, para 3.

136. Further, the Tribunal expressly recognised and took into account the fact that it would not be practicable for the great majority of potential class members to opt in to collective proceedings because of their characteristics and the small size of their claims. The determination that less weight should be given to the interests of such persons in the overall assessment because of the low financial value of their total claims did not involve any legal error and was a judgment which it was reasonably open to the Tribunal to make. There was no basis in law on which the Court of Appeal could properly interfere with the judgment about practicability made by the Tribunal.

8. Issue 3: other factors relied on by the Court of Appeal

137. The defendants have argued as a separate ground of appeal that the Court of Appeal erred in holding, at para 128, that the principles of “facilitating the vindication of rights” and “deterring future wrongdoers” are factors relevant in this case which point in favour of opt-out proceedings. However, there is no doubt that, as Lord Briggs noted in Merricks, para 2, these are indeed policy goals which the availability of the statutory scheme for collective proceedings is designed to further. Lord Briggs also considered that these factors fortify the approach which courts apply even in ordinary civil proceedings that, when claimants are shown to have suffered loss as a result of the defendant’s wrongdoing, the court will do the best it can on the available evidence to quantify damages and will not be put off from doing so by the difficulty or impossibility of precise calculation: paras 47-54. But these factors do not assist the Tribunal in making important decisions when operating the statutory scheme, such as a decision whether to certify collective proceedings as opt-in or opt-out. That is because, as discussed earlier, there are competing policy aims, which also underpin the regime, of protecting businesses from the burden of defending unmeritorious or inflated claims. The Tribunal is required to strike a balance in the way we have described between these interests.

138. We agree with the defendants’ submission that the Court of Appeal was wrong to suggest that “facilitating the vindication of rights” and “deterring future wrongdoers” are factors which point in favour of opt-out proceedings, at any rate without mentioning that those factors are counterbalanced by the need to safeguard the rights of defendants and that the approach of the Tribunal should be even-handed without any bias in favour of opt-out (or opt-in) proceedings. Under the statutory scheme the starting point is one of neutrality and that there is no presumption or predisposition in favour of either of opt-in or of opt-out proceedings. At para 128 of its judgment in this case the Court of Appeal appears to have lost sight of this point.

139. A similar perspective appears to have influenced the court’s approach to access to justice. At para 123 of the judgment, Green LJ said:

“With respect to the [Tribunal], it is now clear from case law that where there would be no proceedings save on opt-out terms, that is a powerful factor in favour of a claim being certified as opt-out.”

This suggests that the Tribunal failed to take account of this factor, when in fact the Tribunal specifically treated it as “a factor that points strongly in favour of certifying on an opt-out basis” (see para 110 above). The real complaint would therefore seem to be that the Tribunal did not regard the factor as conclusive. But there is no warrant for such an approach. Rule 79(3) does not direct the Tribunal to consider whether there would be no proceedings save on opt-out terms; it directs the Tribunal to consider whether it would be practicable for the proceedings to be brought as opt-in collective proceedings.

140. Reliance on “access to justice” as requiring certification on an opt-out basis does not recognise sufficiently that access to justice is something to which both claimants and defendants are entitled. The addressees of the Commission decisions have already either paid substantial fines to the Commission or been granted immunity from fines responding to the incentive to disclose the existence of the cartel to the Commission and to cooperate fully in the Commission’s investigation. The Tribunal was right to make these points at para 372(2)(ii) of its judgment. We endorse in particular this observation:

“Access to justice does not, in our judgment, mean that every case that can only be brought on an opt-out basis must be permitted to proceed on that basis. Opt-out certification is not a certification basis of last resort, in the sense that if opt-in proceedings do not work, there is effectively an entitlement to certification on an opt-out basis. Rather, ‘access to justice’ means considering – taking all of the material into account – whether certifying on an opt-out basis is appropriate.”

141. The sophistication of the collective proceedings regime shows that it was not intended simply to provide a stick with which anyone who claims, however implausibly, to have suffered loss can beat infringing undertakings into paying them substantial damages. That does not enhance the proper enforcement of the competition rules. If clearly unmeritorious claims are allowed to proceed on an opt-out basis which involves an unjustified leverage advantage for claimants of the kind we have described, the result will not be due enforcement of the competition rules but over-enforcement, contrary to the public interest.

9. Issue 4: did the Court of Appeal err in its use of the Sterling Lads ordinary decision?

142. We mentioned earlier (see para 76 above) that the Court of Appeal regarded the ordinary decision of the Commission in Sterling Lads as admissible, relevant and providing strong support for Mr Evans’ claim. The account of the basic facts of this case, set out at paras 14-32 of the judgment, was taken from the Sterling Lads ordinary decision. This was inappropriate given that the Sterling Lads ordinary decision (1) was not part of the material available to the Tribunal as it was issued after the judgment of the Tribunal was handed down, (2) was addressed to a third party (Credit Suisse), and (3) concerned a different infringement from the infringements on which the claim considered by the Tribunal was based. The last of these points was elided by referring (eg at paras 13, 14 and 25 of the judgment) to “the cartel”, as if the Three Way Banana Split, Essex Express and Sterling Lads cases all involved a single cartel. The implication that there was a single cartel was factually wrong and misleading. On the findings of the Commission, there were three different cartels, involving different chatrooms with different members and which operated for different periods.

143. At paras 96-102 the Court of Appeal explained why, in its view, the Sterling Lads ordinary decision was admissible and would have significant probative value when the matter returned to the Tribunal to reconsider whether the claim should be struck out. The final issue raised on this appeal is whether the use made of the Sterling Lads ordinary decision by the Court of Appeal was justified. We will first address, as the Court of Appeal did, the question of admissibility before commenting on the Court of Appeal’s approach to the probative value of the decision.

(a) Admissibility

144. It is a general rule of the common law that findings made by another decision-maker are not admissible as evidence of the facts found. This rule is often referred to “the rule in Hollington v Hewthorn”, after the leading case of Hollington v F Hewthorn & Co Ltd [1943] KB 587 where the Court of Appeal held that the defendant’s conviction for careless driving was inadmissible in a subsequent civil claim for damages as evidence that the defendant was negligent. As explained by the Court of Appeal in Rogers v Hoyle [2014] EWCA Civ 257; [2015] QB 265, 305, paras 39-40, the rule is founded on a principle of fairness. That principle requires that a tribunal responsible for finding facts should base its findings on its own evaluation of the evidence and not on the evaluation of someone else who is not the relevant decision-maker. The same principle applies to a person conducting an inquiry: see Three Rivers District Council v Governor and Company of the Bank of England (No 3) [2003] 2 AC 1, 238, para 5 (Lord Steyn).

145. It would be particularly unfair to treat findings made by an earlier decision-maker as admissible against a person who was not a party to the earlier proceedings and who therefore had no opportunity to influence the findings made in those proceedings by adducing evidence or advancing arguments. This fundamental objection does not apply to someone who was a party to the earlier proceedings. In some cases, such a party may be bound by findings, for example where they give rise to an issue estoppel; or it may be an abuse of process for the party to contest them. In follow-on damages claims against the addressees of a Commission decision, the Commission’s finding of infringement is binding on the Tribunal or the national court. The identification of which particular findings of fact set out in the recitals to that decision can still be challenged by the defendants and which cannot has been the subject of debate: see eg AB Volvo (Publ) v Ryder Ltd [2020] EWCA Civ 1475; [2021] Bus LR 1610. In this judgment we are addressing the different question whether the findings set out in the Sterling Lads decisions are admissible in evidence against the defendants in proceedings following on from the Three Way Banana Split and Essex Express decisions.

146. In such circumstances, where the follow-on jurisdiction is not in issue and there is no question of issue estoppel or abuse of process, the defendants are entitled to challenge the correctness of the findings and the principle of fairness considered in Rogers v Hoyle remains relevant. The court or tribunal which has to decide whether the findings of the earlier decision-maker were correct can only properly do so by making its own evaluation of the evidence on which they were based and any additional evidence. The Privy Council made this point in Calyon v Michailaidis [2009] UKPC 34; [2010] 2 LRC 280, para 27, where Lord Rodger of Earlsferry said that:

“the essential reasoning is compelling: unless the second court goes into the facts for itself, it cannot actually tell what weight it should properly attach to the previous decision. Which means that the previous decision itself cannot be relied upon.”

147. In Consumer Association v Qualcomm Inc [2023] CAT 9 (“Qualcomm”) the question arose whether the rule in Hollington v Hewthorn applies in the Competition Appeal Tribunal. The Tribunal decided that it was not bound by the rule but that it should nevertheless adopt the same principle, and that at a trial of collective proceedings it would not be appropriate to attach any weight to findings reached by other courts, tribunals or regulators: paras 22-23. The principal reason was that given in Rogers v Hoyle, namely, that:

“it is for this Tribunal to assess the evidence and make primary findings of fact. Relying upon the evaluative judgments of other decision-makers necessarily circumvents that role.”

148. The Tribunal also referred, at para 30, to the difficulty, if findings of another decision-maker were treated as admissible, of deciding what weight to give them:

“How would this Tribunal, at trial, go about assessing how much weight should be given to a particular decision of another court or regulator? That would almost inevitably involve a detailed consideration of the evidence that was before the other decision-maker and the nature of the decision-making process. It might also require an assessment of the way in which the arguments were put to that decision-making body, on both sides. There would in consequence be what HHJ Paul Matthews (sitting as a judge of the High Court) described at [51] of Crypto Open Patent Alliance v Wright [2021] EWHC 3440 (Ch) as ‘satellite litigation about the circumstances in which the earlier decision was come to, and how far it could properly be helpful in the later proceedings’.”

149. The Court of Appeal in this case took a different view. Green LJ, at para 100, thought it “well established” that the rule in Hollington v Hewthorn does not apply to the Tribunal because rule 55(1)(b) of the Tribunal Rules makes clear that the Tribunal has a wide discretion as to the evidence to be admitted. As a case supporting that proposition, he cited Qualcomm. But in Qualcomm the Tribunal held that, although not bound by the rule in Hollington v Hewthorn, it should adopt the same principle and Green LJ did not explain why he disagreed with the reasons given for that decision.

150. Instead, he considered that “[t]here is no need for the [Tribunal] to be hidebound by a common law rule on fairness”: para 101. Green LJ said that, while the Tribunal “does, of course, endeavour to secure fairness … it is a sophisticated tribunal well able to form its own view on the value, if any, of prior findings”. He expressed confidence that the Tribunal, when confronted with prior findings said to be relevant, “will carefully decide what weight can be attached to those findings”, having regard to various matters such as the nature of the earlier decision, who the earlier decision-maker was, the standard of proof applied, and the extent to which the legal analysis in the earlier decision affects the findings of fact made: para 102. Green LJ then articulated five points which seemed to him to have some force about the probative value of the Sterling Lads ordinary decision in this case: see paras 103-108.

151. The statement that there is no need for the Competition Appeal Tribunal to be “hidebound by a common law rule on fairness” is unhappily expressed. It was also unsupported by reasoning. It is one thing to say, as the Tribunal did in Qualcomm, that the rule in Hollington v Hewthorn is not binding on the Tribunal. It is another to say that the principle of fairness underpinning that rule does not, for some reason, apply to proceedings before the Tribunal. Green LJ’s discussion bypasses the latter question and simply assumes that the only relevant consideration is one of evidential weight.

152. In our view, the rule in Hollington v Hewthorn, as explained in Rogers v Hoyle, does indeed apply to the Tribunal. Rule 55(1)(b) of the Tribunal Rules is an inadequate basis on which to suggest otherwise. All that this rule says is that the Tribunal may give directions as to “the issues on which it requires evidence, and the admission or exclusion from the proceedings of evidence”. The rule does not say that, in giving such directions, the Tribunal is free to ignore a common law principle of fairness. Rule 32.1 of the Civil Procedure Rules gives a court a similar power to control the evidence adduced by giving directions as to the issues on which it requires evidence, the nature of the evidence which it requires to decide those issues and the way in which the evidence is to be placed before the court; and specifically authorises the court to use this power to exclude evidence that would otherwise be admissible. Yet no one has suggested - rightly, in our view - that this rule allows a court to admit opinion evidence that authorities such as Rogers v Hoyle would otherwise require it to exclude. It is implicit in the rule that it is to be operated in accordance with the principle of fairness. There is no good reason to interpret the Tribunal Rules differently. In any case we agree with the reasons given by the Tribunal in Qualcomm for holding that the underlying principle is a sound one which ought to be adopted in proceedings before the Tribunal irrespective of whether the authorities articulating that principle are binding on the Tribunal as a matter of precedent.

153. We also note that the rule in Hollington v Hewthorn applies here in what may be called its strong form because the appellants were not parties to the procedure which led to the Sterling Lads ordinary decision. They therefore had no opportunity to influence the findings made in that decision by adducing evidence or advancing arguments to the Commission; nor to challenge the decision by appealing to the General Court (as Credit Suisse subsequently did). It would in those circumstances be fundamentally unfair to admit the findings made by the Commission as evidence against the appellants: see Al Rawi v Security Service [2011] UKSC 34; [2012] 1 AC 531, 572-573, paras 11-13 per Lord Dyson. As Lord Dyson observed in that case at p 575, para 22, a court’s procedural discretion and powers of case management are limited by the fundamental principle of natural justice, and the same is true of the Tribunal.

154. Counsel for Mr Evans sought to distinguish the rule in Hollington v Hewthorn by submitting that it is limited to prior decisions by those acting in a judicial or “quasi-judicial” capacity and that, in issuing decisions in competition cases, the Commission does not act in such a capacity. Alternatively, it was said that such decisions are an exception to the rule in Hollington v Hewthorn and were recognised as such by the House of Lords in Crehan v Inntrepreneur Pub Co (CPC) [2006] UKHL 38; [2007] 1 AC 333.

155. The attempt to confine principles of procedural fairness to decisions made by persons acting in a judicial or “quasi-judicial” (whatever that means) capacity has long since been discredited in administrative law, having been “scotched” as a “heresy” by the House of Lords in Ridge v Baldwin [1964] AC 40: see R v Gaming Board for Great Britain, Ex p Benaim [1970] 2 QB 417, 430. There is no warrant for limiting the rule in Hollington v Hewthorn in that way. The reasons for treating findings made by an earlier decision-maker as inadmissible do not depend upon the capacity in which that person was acting. They apply to anyone who has previously expressed an opinion about what conclusions should be drawn from factual evidence. The rule is an aspect of the general exclusion of opinion evidence. The main exception is for expert evidence adduced in proceedings. Even then, only opinions on matters within the scope of the relevant expertise are admissible, and not opinions on the issues of law or fact which the court or tribunal has to decide.

156. In Crehan the trial judge, in dismissing a claim for damages for breach of European competition law, held that he was not bound by findings of fact about the UK beer market made by the Commission in decisions involving different parties. The Court of Appeal disagreed, holding that the court was precluded from rejecting the findings of the Commission by the duty of sincere co-operation under EU law. That decision was in turn reversed, and the judge’s decision reinstated, by the House of Lords. In a judgment with which the other members of the House of Lords agreed, Lord Hoffmann, at para 69, rejected the notion that a national court was required to show “deference” to a decision of the Commission by which it was not bound, and said:

“The correct position is that, when there is no question of a conflict of decisions in the sense which I have discussed, the decision of the Commission is simply evidence properly admissible before the English court which, given the expertise of the Commission, may well be regarded by that court as highly persuasive. As a matter of law, however, it is only part of the evidence which the court will take into account. If, upon an assessment of all the evidence, the judge comes to the conclusion that the view of the Commission was wrong, I do not see how, consistently with his judicial oath, he can say that as a matter of deference he proposes nevertheless to follow the Commission. Only a rule of law, in the nature of an issue estoppel which obliges him to do so, could produce such a result and the Court of Appeal accepted that there was no such rule.” (emphasis added)

157. Counsel for Mr Evans relied on the words we have emphasised in this passage to argue that a decision of the Commission is admissible generally as evidence before a court or tribunal in England and Wales and may, given the expertise of the Commission, be regarded as highly persuasive. That observation, however, was an obiter dictum unnecessary for the determination of the appeal in Crehan. Crehan was not a follow-on case, and Lord Hoffmann had made it clear in para 56 of his judgment that the case did not create a risk of the domestic court acting in breach of the duty of sincere cooperation. No reference was made to the Hollington v Hewthorn line of cases, which do not appear from the report to have been cited in argument. When Lord Hoffmann referred to the expertise of the Commission, he must, we think, have been referring to its expertise in economic analysis in general and in the operation of the United Kingdom pub and beer markets in particular: see the passage from the trial judge’s judgment quoted by Lord Bingham at para 7 of Crehan. Despite the apparent breadth of Lord Hoffmann’s dictum, it cannot be regarded as authority for treating factual findings set out in a Commission decision about how a particular market operates as admissible as evidence in proceedings between third parties. We therefore consider that it was a mistake for the Court of Appeal in the present case to say that the Sterling Lads ordinary decision was admissible before the English court, as against defendants who were not addressees of that decision, as evidence of facts found in the decision.

(b) Interlocutory proceedings

158. The argument that counsel for Mr Evans put at the forefront of their submissions on admissibility is that, even if the rule in Hollington v Hewthorn or the principle underlying it would apply at a trial of proceedings before the Tribunal, it does not (and would not in the High Court) preclude reliance on findings made by another decision-maker for the purpose of defeating a strike-out or summary judgment application or otherwise demonstrating the strength of claims at an interlocutory stage. In support of this argument, reliance was placed on cases holding that a party could rely on findings of another court for the purpose of demonstrating that there was a serious issue to be tried: see JSC Aeroflot Russian Airlines v Berezovsky [2013] EWCA Civ 784; [2013] 2 Lloyd’s Rep 242, paras 115-116; Sabbagh v Khoury [2014] EWHC 3233 (Comm), paras 202-207; Tulip Trading Ltd v Bitcoin Association for BSV [2023] EWHC 2437 (Ch), paras 17-57. In the last of these cases, at para 40, Mellor J concluded from the earlier authorities that:

“there is a limited exception to the rule in Hollington v Hewthorn which is applicable in situations where the case is at a preparatory stage yet the court has to consider what evidence at trial there might be. … [M]aterial (inadmissible at trial) can assist in identifying the evidence which can reasonably be expected to be available at trial, to which a court is entitled to have regard at the interlocutory stage.”

159. We endorse this analysis save only to observe that reliance on findings of another decision-maker for the purpose of identifying evidence which can reasonably be expected to be available at trial is not inconsistent with the rule in Hollington v Hewthorn and is therefore not strictly an exception to it. Likewise, it is not inconsistent with the principle underlying that case to rely on prior judgments or reports in so far as they record evidence of relevant facts: see eg Rogers v Hoyle, p 307, para 49. It is only in so far as such material contains opinions on matters of fact (as opposed to recording evidence) that the material is inadmissible (unless it qualifies as expert evidence).

(c) Probative value

160. Applying the principle here, we cannot see that the Sterling Lads ordinary decision against Credit Suisse either records any evidence relevant to the claims against the addressees of the Three Way Banana Split decision or the Essex Express decision or assists in identifying relevant evidence which can reasonably be expected to be available at a trial.

161. The reasons given by the Court of Appeal for regarding the decision as having “probative value” (para 103) do not, in our view, hold water. Those reasons assumed that the Sterling Lads ordinary decision involved “more or less identical facts to those arising in the instant case” (para 104) and that all the decisions of the Commission were “from the same decisional template or mould” in that “a side by side comparison … showed many paragraphs which were common across all the decisions”. This was said to show “by a process of logical reverse extrapolation and inference … what was hiding behind the short form decisions in Essex Express and in Three Way Banana Split … which might only become apparent after disclosure”: see para 105.

162. The decisions in Sterling Lads, like those in the Three Way Banana Split and Essex Express cases, involved findings of infringement by object. The argument developed by Green LJ was that, to establish infringement on this basis, the Commission accepted that it had to demonstrate, by reference to the appropriate legal and economic context, that there was a real likelihood of an actual effect on competition: para 26. In the Sterling Lads ordinary decision the Commission concluded that the exchanges of information shown to have taken place were likely to result in harm. Green LJ quoted various passages from the decision to this effect. For example, at para 32 he quoted a passage from para 473 of the Commission decision which contained the following finding (with emphasis added by Green LJ):

“The exchanges of information on bid-ask spreads in the chatroom increased transparency and reduced market uncertainties for the participating traders regarding prices. These exchanges enabled the participating traders to obtain greater certainty on the spreads they were quoting and might have informed their subsequent trading behaviour concerning spreads. The information exchanges may also have allowed them to align their spreads for particular transactions and thereby their all-in price offered to a specific client for a particular transaction. A customer who is not aware of such exchanges of non-publicly available information on spreads may have contacted more than one of the parties’ sales desks to get a price on a specific trade and may have received less competitive prices from them due to the exchanges of information on bid-ask spreads between the participating traders …”

163. The suggestion made by Green LJ (at paras 97-108 and 117) was that, although the Sterling Lads ordinary decision did not contain any findings about the actual effects of the infringements found, it contained findings that harmful effects were likely to have occurred; and this provides a basis for inferring that similar harmful effects probably occurred as a result of the infringements by object found in the Essex Express and Three Way Banana Split decisions.

164. Although the Sterling Lads ordinary decision examines the exchanges of information that occurred in the relevant chatroom in much more detail than the three settlement decisions (including long quotations of communications that took place), it is not clear to us that the Commission’s findings about the effects of those exchanges differ materially from the findings made in the settlement decisions.

165. The Tribunal did not doubt that there would be reasonable grounds for alleging that the exchanges of information admitted by the defendants enabled them to make money on specific trades. That is the possibility contemplated by the Commission in the passage we have quoted above from the Sterling Lads decision (and other similar passages). The problem which the Tribunal identified was that Mr Evans is not claiming damages for losses resulting from the manipulation of rates paid by clients on specific trades. As the Tribunal also observed, it is difficult to see how such claims could be framed as collective proceedings, given that the losses would be individual and not collective.

166. The points made by the Court of Appeal with reference to the Sterling Lads ordinary decision do not suggest any basis on which the claim pleaded by Mr Evans might be reformulated in a way that might survive a strike out application. Certainly, the Commission’s findings provide no reason to expect that analysis of the trading performance of banks at different times will show any variations which correlate in a statistically significant way with participation in the chatroom.

(d) Conclusion on use of the Sterling Lads ordinary decision

167. We conclude that the reliance placed by the Court of Appeal on the Sterling Lads ordinary decision was misconceived. Far from being a game changer as the Court of Appeal seems to have thought, the findings made by the Commission in that decision are inadmissible and irrelevant to the Tribunal’s assessment of the merits of Mr Evans’ claim against the addressees of the Three Way Banana Split and Essex Express decisions. The decision therefore would not reasonably have altered that assessment even if it had been issued before the Tribunal reached its decision on Mr Evans’ application.

10. Conclusions

168. On the issues of principle raised on this appeal, we can summarise our main conclusions as follows:

• (i) The weakness of the claim as assessed by the Tribunal – and the conclusion that no plausible case on causation had been articulated - was properly regarded by the Tribunal as a factor weighing strongly against opt-out proceedings. The merits of the claim are not a neutral factor. In the present case, this factor indicated that affording the claimants the procedural advantages and exposing the defendants to the additional burdens and “leveraging effect” of opt-out proceedings was not justified by the perceived merits of the claim.

• (ii) The Tribunal was entitled to take the view that, although opt-in proceedings would not in fact be brought, this was because a significant number of large and sophisticated financial institutions which stood to recover substantial sums of money if the claim succeeded were not interested in opting in to a collective action; and that in these circumstances, although it was not practicable for smaller entities and individuals to opt in, looking at the matter overall it could not be said that it was not practicable to bring opt-in proceedings.

• (iii) The policy of facilitating the vindication of rights and deterring future wrongdoers is not a strong factor pointing in favour of opt-out rather than opt-in proceedings. Case management by the Tribunal, in particular choosing between opt-in and opt-out proceedings, requires a balance to be struck between assisting claimants to obtain redress and protecting defendants from oppressive litigation. The proper starting point in making the determination is one of neutrality without any presumption or predisposition in favour of either opt-out, or opt-in, proceedings.

• (iv) A decision of the Commission containing findings of infringement against a third party is not admissible and those findings do not have probative value in these proceedings.

169. The Tribunal carried out a thorough analysis of all the factors listed in the Tribunal Rules as relevant in determining whether opt-out collective proceedings were appropriate. The exercise of weighing up the various factors relevant to the choice of opt-in or opt-out collective proceedings is essentially an exercise of judgment in respect of facts and evidence by an expert, specialist body. The Tribunal has a broad discretion when making the relevant judgments. The Court of Appeal should not interfere simply because it might have arrived at a different conclusion if it had been conducting the exercise. No grounds have been identified which justified the Court of Appeal in reversing the judgments and overall determination made by the Tribunal in this case.

170. For the reasons we have set out, we would accordingly allow the appeal and reinstate the Tribunal’s decision to refuse Mr Evans’ application for a collective proceedings order to be made on an opt-out basis.