Evans v Barclays Bank Plc & Ors [2025] UKSC 48:
Merits, Practicability and Third‑Party EU Decisions in Competition Collective Proceedings

1. Introduction

This Supreme Court decision is now a leading authority on how the Competition Appeal Tribunal (“CAT”) should exercise its “gatekeeper” role when deciding whether competition law collective proceedings under section 47B of the Competition Act 1998 should be certified on an opt‑in or opt‑out basis, and on the admissibility of European Commission decisions addressed to third parties.

The case arises out of alleged losses said to have been suffered by thousands of market participants in the foreign exchange (“FX”) markets following on from three European Commission “FX cartel” decisions (Three Way Banana Split, Essex Express and later Sterling Lads). The respondent, Mr Phillip Evans, sought to act as class representative and to bring very large “opt‑out” collective damages claims against several global banks.

The core issues were:

• Issue 1 – Strength of the claim: How should the strength of the claims influence the CAT’s choice between opt‑in and opt‑out proceedings under rule 79(3)(a) of the Tribunal Rules?
• Issue 2 – Practicability: What does it mean for opt‑in proceedings to be “practicable” under rule 79(3)(b), and how far may an appellate court interfere with the CAT’s evaluative judgment?
• Issue 3 – Policy goals: Do the goals of facilitating the vindication of rights and deterring competition law infringements generally weigh in favour of opt‑out certification?
• Issue 4 – EU Commission decisions: Can a Commission infringement decision addressed to a different undertaking (here, the Sterling Lads ordinary decision against Credit Suisse) be used as admissible, probative material in follow‑on damages actions against others?

The Supreme Court unanimously allowed the appeal, reinstating the CAT’s refusal to certify the Evans proceedings on an opt‑out basis. In doing so, it clarified four key points of general importance in the emerging UK collective proceedings regime.

2. Background: The FX Cartels and the Collective Proceedings Regime

2.1 The underlying infringements

The Evans claims are “follow‑on” competition damages claims under section 47A of the 1998 Act, based on two 2019 European Commission settlement decisions:

• Three Way Banana Split (Case AT.40135 – FOREX (Three Way Banana Split)) against UBS, RBS, Barclays, Citigroup and JPMorgan;
• Essex Express (Case AT.40135 – FOREX (Essex Express)) against UBS, RBS, Barclays and Bank of Tokyo Mitsubishi.

These decisions concerned communications in small, private chatrooms between a handful of FX traders at the addressee banks, relating to spot trading in G10 currency pairs between 2007–2013. The Commission found:

• An infringement “by object” of Article 101 TFEU and Article 53 EEA, i.e. conduct whose object was to restrict competition, so that the Commission did not assess actual effects on prices or competition.
• A “single, continuous infringement” in each case – a legal construct enabling liability for the full infringement period even though different banks joined or left the chatroom at different times.
• Collusive exchanges of commercially sensitive, short‑lived information on trading, facilitating occasional coordinated behaviour and permitting traders sometimes to widen bid‑ask spreads or favour a trader with “higher stakes at play”.

The Commission did not determine that bid‑ask spreads were widened systematically across all trades for all customers, nor quantify any market‑wide overcharge. Fines were calculated using a proxy “value of sales” based on revenues and spreads, expressly as a sanctioning device rather than an estimate of harm.

2.2 Mr Evans’ proposed collective claims

Mr Evans sought a collective proceedings order to pursue large aggregate damages claims, on an opt‑out basis, on behalf of two overlapping classes:

• Class A: persons who traded FX directly with the defendant banks in the EEA during periods when they participated in the infringements.
• Class B: persons who traded FX with other financial institutions, or with the defendants at times when those defendants were not in the chatrooms, during the cartel period – relying on a so‑called “umbrella effect”, i.e. that cartel conduct by some players relaxed competitive pressure on others, allowing market‑wide spread widening.

Crucially, the pleaded theory went far beyond the Commission’s findings:

• On the Commission’s case, traders occasionally used chatroom information to coordinate or opportunistically adjust spreads for certain trades during brief windows (minutes or hours).
• Mr Evans alleged that, for the entire period each bank had a trader in any chatroom, the bank unlawfully widened spreads on all G10 spot trades with all customers (including automated electronic trades), and that this effect then rippled across the wider market (Class B) via umbrella effects.

Indicative figures posited a value of commerce of about £41bn for Class A and £74bn for Class B, an assumed “overcharge” of 18% (based on generic cartel overcharge literature rather than any FX‑specific analysis), and pre‑interest damages of £2.155bn.

2.3 The collective proceedings framework

The modern competition collective proceedings regime was introduced by the Consumer Rights Act 2015, via new sections 47A–47C of the 1998 Act and the Competition Appeal Tribunal Rules 2015.

Key elements include:

• Section 47A: individual claims for damages for competition infringements in the CAT.
• Section 47B: “collective proceedings” combining two or more such claims, subject to a collective proceedings order (“CPO”). CPOs must specify whether proceedings are opt‑in (only those who choose to join are bound) or opt‑out (all within the class are included unless they opt out).
• Section 47C: enables aggregate damages awards for the class as a whole, without individualised loss assessment, a key feature of the regime.

The CAT Rules embed an overriding objective in rule 4: to deal with cases “justly and at proportionate cost”, including ensuring parties are on an equal footing, saving expense, proportionality, fairness and sensible allocation of Tribunal resources. Rule 79 deals specifically with collective proceedings:

• Rule 79(2): when deciding whether claims are suitable to be brought in collective proceedings, the CAT considers factors such as whether collective proceedings are an appropriate mechanism, costs and benefits, class size and nature, and suitability for aggregate damages.
• Rule 79(3): when deciding opt‑in vs opt‑out, the CAT “may take into account all matters it thinks fit” but must consider in particular:
  • (a) the strength of the claims; and
  • (b) the practicability of opt‑in proceedings, having regard to all the circumstances, including estimated individual recoveries.

The CAT also has strike‑out (rule 41) and summary judgment (rule 43) powers and may entertain such applications at the CPO hearing (rule 79(4)).

A practice direction‑level Guide to Proceedings (2015) further explains that:

• Opt‑out proceedings are generally more complex, costly and risky, so the CAT will “usually expect the strength of the claims to be more immediately perceptible” for opt‑out than for opt‑in.
• The Guide notes a “general preference for proceedings to be opt‑in where practicable”, given safeguards against abusive litigation.

3. Summary of the Judgment

3.1 What the CAT decided

The CAT (Sir Marcus Smith and Professor Neuberger, with Mr Lomas dissenting on opt‑out) held that:

• The general requirements for certification as collective proceedings were met in principle (Merricks v Mastercard established that merits are not relevant at that stage).
• However, the pleaded theory of causation was severely deficient. On the material before it — including multiple expert reports and post‑hearing written clarification — the majority considered the claims liable to be struck out, although it chose not to strike them out immediately, instead giving the applicants a “last chance” to reformulate.
• On the opt‑in/opt‑out choice, the majority held:
  • The strength of the claim was very weak and strongly militated against opt‑out.
  • It was practicable to bring opt‑in proceedings for large, sophisticated financial institutions with sizeable alleged losses, even though smaller claimants could not realistically do so.
  • Although proceedings would not in fact be pursued on an opt‑in basis (funders and key institutions were uninterested), that commercial reality did not make opt‑in impracticable in the sense required by rule 79(3)(b).
• The CAT therefore refused opt‑out certification and stayed the applications, giving time for revised opt‑in applications.

3.2 What the Court of Appeal did

The Court of Appeal (Green LJ, Flaux C and Snowden LJ) reversed the CAT’s approach. It held that:

• The CAT had been inconsistent: having decided not to strike out the claims and to allow reformulation, it should not then have treated its merits assessment as “final” for the opt‑out decision, knowing that refusing opt‑out would kill the claims.
• In most cases, the strength of the claim is a neutral factor on opt‑in/opt‑out; the CAT had not linked its view of the merits to why opt‑in was preferable.
• The CAT had erred in its analysis of practicability; statistical evidence, especially on the modest size of most claims, showed that opt‑in was in reality not practicable.
• The Court of Appeal also:
  • treated the detailed Sterling Lads ordinary decision against Credit Suisse as admissible and significantly supportive of the merits;
  • emphasised policy goals of vindication of rights and deterrence as weighing towards opt‑out;
  • remitted the merits and opt‑in/opt‑out questions to the CAT.

3.3 The Supreme Court’s holdings

The Supreme Court, in a joint judgment of Lord Sales, Lord Leggatt and Lady Rose (with whom Lord Burrows and Lord Richards agreed), allowed the appeal and restored the CAT’s decision. In essence, it held that:

1. Merits are not neutral: The CAT was entitled to regard the claims as very weak, and the weakness was a powerful factor against opt‑out under rule 79(3)(a). Strength operates on a sliding scale: the weaker the claim, the less justification there is for conferring the “leveraging” advantages of opt‑out proceedings.
2. Strike‑out vs opt‑out: There is no inconsistency in the CAT deciding not to strike out immediately, while still taking into account its (provisional but well‑developed) negative assessment of the merits to refuse opt‑out certification.
3. Practicability assessment upheld: The CAT’s evaluation that opt‑in proceedings were practicable for large institutional claimants, even though many smaller claims could not realistically opt in, fell within its wide discretionary margin. The Court of Appeal impermissibly re‑weighed the evidence and substituted its own assessment.
4. No presumption for opt‑out: Policy objectives of vindicating rights and deterring infringements do not generally tilt the scales towards opt‑out. The starting point is neutral: the CAT must balance access to justice for claimants with fairness and proportionality for defendants.
5. Hollington v Hewthorn applies in the CAT: The common law rule (excluding findings of other decision‑makers as evidence of facts) and its fairness rationale apply to the CAT. A Commission decision against a third party (here, Credit Suisse in Sterling Lads) is inadmissible as evidence of the facts found, and was wrongly treated as probative by the Court of Appeal.
6. Limited use of foreign decisions at interlocutory stage: Prior judgments or decisions may be referred to, at most, to identify what types of evidence might exist at trial in interlocutory contexts, but they cannot be treated as substantive evidence of contested facts.

4. Precedents and Legal Framework

4.1 Legislative and policy background to collective proceedings

The judgment draws heavily on the extensive pre‑2015 policy work around collective redress. Several key themes recur:

• Access to justice for consumers and SMEs: Recognised in the OFT’s 2007 consultation (Private actions in competition law: effective redress for consumers and business) and the Civil Justice Council’s 2008 report as a reason to enable collective actions where individual claims would be uneconomic.
• Fear of “US‑style class actions”: Government and policy documents repeatedly stress the need to avoid vexatious litigation, “blackmail suits” and claims run more for funders’ profit than for redress.
• Gatekeeping and merits: The Civil Justice Council recommended “positive certification” with a robust preliminary merits assessment to avoid abusive suits. The Department for Business’s 2012–2013 consultation and impact assessments emphasised “strict judicial certification” and a “preliminary merits test” to ensure only “meritorious cases” proceed.

The Supreme Court uses this history to confirm that rule 79(3)’s reference to “strength of the claims” is not accidental: Parliament and rule‑makers expected the CAT to weigh merits as a real, and sometimes decisive, factor in the opt‑in/opt‑out choice.

4.2 Merricks v Mastercard

In Merricks v Mastercard Inc [2020] UKSC 51, the Supreme Court:

• Clarified that, at the stage of deciding whether to certify collective proceedings at all, the CAT should not conduct a “mini‑trial” on merits; the threshold is relatively low and merits are not a relevant factor at that gateway stage.
• Recognised, however, the potential for opt‑out, aggregate damages proceedings to be misused and the need for “gatekeeper” functions to avoid oppressive or unmeritorious litigation (see the present judgment’s reliance on Merricks’ discussion of leverage and risk of abuse).

Evans builds on Merricks by clarifying that, although merits are not to be weighed at the “suitability for collective proceedings” stage, they are expressly relevant at the opt‑in/opt‑out stage under rule 79(3)(a).

4.3 Le Patourel v BT Group plc

In Le Patourel v BT Group plc [2022] EWCA Civ 593 the Court of Appeal:

• Confirmed there is no general presumption under the statute in favour of either opt‑in or opt‑out; the choice is discretionary, based on all the circumstances.
• Emphasised that the CAT’s opt‑in/opt‑out decision is a fact‑sensitive, evaluative judgment entitled to a broad margin of discretion; appellate courts should rarely interfere.
• Suggested that in many cases merits might be “neutral” because the CAT lacks sufficient material at certification to form a robust view.

In Evans, the Supreme Court accepts the first and second strands of Le Patourel but rejects the notion that merits are generally neutral. It confirms that, where the CAT is in fact able to form a view on strength, it should use that view in deciding between opt‑in and opt‑out.

4.4 The Hollington v Hewthorn line of authority

The judgment revisits the common law rule, originating in Hollington v F Hewthorn & Co Ltd [1943] KB 587, that:

• Findings of fact or opinions expressed by one decision‑maker (court, tribunal, regulator) are generally inadmissible as evidence of those facts in subsequent proceedings before another decision‑maker.
• The underlying rationale, articulated in Rogers v Hoyle [2014] EWCA Civ 257, is fairness: the tribunal charged with deciding the case must assess the primary evidence for itself, not be influenced by others’ conclusions.
• The concern is especially acute where the later proceedings involve parties who were not parties to the earlier decision and had no opportunity to contest the findings (see also Calyon v Michailaidis [2009] UKPC 34).

In Consumer Association v Qualcomm Inc [2023] CAT 9, the CAT itself had already decided that, although not formally bound by Hollington, it should adopt the same principle.

The Supreme Court in Evans:

• Affirms that Hollington and its fairness rationale apply to the CAT. Rule 55(1)(b) on admissibility does not license the CAT to disregard common law fairness principles.
• Holds that Commission decisions addressed to third parties cannot be deployed as evidence of facts against non‑addressee defendants in follow‑on claims.
• Clarifies that any apparent contrary dictum in Crehan v Inntrepreneur Pub Co [2006] UKHL 38 (Lord Hoffmann’s remarks about Commission decisions as “evidence”) was obiter, did not address Hollington, and cannot be read as establishing a general exception.

4.5 Other authorities

The judgment also references:

• Lloyd v Google LLC [2021] UKSC 50, for the nature and limits of representative actions and aggregate awards.
• Hughes v Richards [2004] EWCA Civ 266, on caution in striking out novel claims in developing areas of law.
• Kone AG v ÖBB‑Infrastruktur (Case C‑557/12), on “umbrella damages” in EU competition law.
• Volpi v Volpi [2022] EWCA Civ 464 and Gift v Rowley [2025] UKPC 37, on the presumption that courts consider all evidence before them unless clear indications to the contrary.
• Tulip Trading, JSC Aeroflot v Berezovsky and Sabbagh v Khoury, on the limited use of prior findings at interlocutory stages to identify potential evidence.

5. Legal Reasoning and Analysis

5.1 Issue 1 – Strength of the claim and the opt‑in/opt‑out decision

5.1.1 Relationship between strike‑out and opt‑out

The Court of Appeal saw inconsistency in the CAT’s approach: if the CAT thought the claim was so weak as to be “deserving of strike‑out” (its own language), why did it not strike it out, and how could it reasonably use that “provisional” view to refuse opt‑out?

The Supreme Court rejects this syllogism. It emphasises:

• The CAT is empowered by rules 41 and 43 to consider strike‑out or summary judgment of its own initiative in collective proceedings, given the large resource implications of such claims.
• Here, having examined very extensive pleadings and expert materials (including post‑hearing clarifications), the CAT formed the view that, even assuming the claimants could adduce evidence consistent with their theory, their pleaded theory of harm did not articulate a plausible causal link between the limited, short‑lived information exchanges and the vast market‑wide overcharge alleged.
• Nonetheless, applying the cautious approach in novel areas (Hughes v Richards), and recognising that its “final thinking” on pleading defects had crystallised only during deliberations, the CAT decided not to strike out immediately, but to give a final opportunity to reformulate.

The Supreme Court holds that this is a legitimate and fair case management choice. To say:

• “We do not yet strike this out, because we will permit one last attempt to cure defects”

does not compel the CAT to say:

• “We must ignore our assessment of the current merits when deciding whether to confer the exceptional power of opt‑out collective proceedings.”

The CAT was entitled to treat its carefully reasoned (albeit not yet procedurally final) merits assessment as highly relevant when deciding whether to expose defendants to the leverage of opt‑out litigation.

5.1.2 Why merits matter: the “leveraging effect”

The Court of Appeal had suggested that strength is “generally” neutral in the opt‑in/opt‑out choice. The Supreme Court’s central correction is doctrinal and practical:

• Rule 79(3)(a) explicitly singles out “the strength of the claims” as a factor in addition to the general considerations in rule 79(2). It would not do so if merits were generally neutral.
• The legislative and policy history demonstrates a deliberate design: opt‑out procedures with aggregate damages confer significant procedural and commercial advantages on claimants (and funders) and impose heavy burdens on defendants.

In particular, opt‑out, aggregate proceedings:

• Permit claims by thousands or millions who never actively chose to sue.
• Allow recovery without proving individual loss.
• Generate substantial exposure and defence costs, creating “leveraging” pressure on defendants to settle, even if they consider the claim weak.
• Lead to damages awards often exceeding the amounts that will ever actually be distributed, because many class members will not claim their share.

Against this backdrop, the Supreme Court holds:

“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 rooted in rule 4’s overriding objective:

• Equal footing: Weak claims should not be turbo‑charged in a way that unduly handicaps defendants.
• Saving expense and proportionality: Massive opt‑out litigation is not a proportionate use of resources where merits are poor.
• Fairness and resource allocation: CAT resources must be reserved for claims where the combination of merits and procedural vehicle is justified.

5.1.3 “Sliding scale” of strength

The Supreme Court explicitly endorses the CAT’s statement 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.”

This “sliding scale” approach means:

• Very strong claims → merits point strongly for opt‑out (subject to practicability, class composition, etc.).
• Moderately arguable claims → merits may carry limited or neutral weight, depending on how clear a view the CAT can form at certification.
• Very weak but not strike‑out‑able claims → merits point strongly against opt‑out; they may proceed at most, if at all, as opt‑in or not at all.

The Court of Appeal’s premise – that surviving strike‑out or summary judgment necessarily precludes a finding that a claim is “very weak” – is rejected as a logical error. Strike‑out/summary judgment is a low bar; the merits continuum above that threshold matters.

5.1.4 Why Evans’ case was deemed weak

The heart of the CAT’s merits concern, endorsed by the Supreme Court, is the causation gap:

• The Commission decisions established object infringements arising from limited, short‑lived, trader‑to‑trader information exchanges.
• They found that such exchanges could enable, and did occasionally lead to, specific opportunistic coordination or spread adjustments.
• Mr Evans’ pleaded theory extrapolated this to:
  • assume spreads were widened on all trades by relevant banks for the entire cartel period; and
  • invoke a market‑wide umbrella effect whereby non‑cartel banks also widened spreads.

The CAT, drawing on its economic expertise (especially Professor Neuberger’s), found that:

• The pleadings and expert materials did not articulate a coherent economic mechanism linking the occasional collusive acts in the chatrooms to a systematic, market‑wide overcharge.
• Even for Class A, the claim was not limited to trades actually affected by any identified unlawful communication; it encompassed all trades with a defendant during its participation in any chatroom.
• For Class B, the umbrella theory was thinly pleaded and did not convincingly address intra‑market competition, product fungibility and high substitutability in FX trading.

The Supreme Court endorses the CAT’s conclusion that this was not simply a “translation” problem between expert theory and pleadings. The underlying theory of harm was insufficiently grounded in a plausible causal story, rather than merely being “difficult to prove”.

The upshot: given such weaknesses, it was entirely legitimate for the CAT to decline to overlay the powerful opt‑out machinery on top of claims that might, at best, barely tip over the non‑strike‑out threshold.

5.2 Issue 2 – Practicability of opt‑in proceedings

5.2.1 Objective test and class composition

Rule 79(3)(b) asks 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.”

The Supreme Court clarifies key aspects:

• The inquiry is objective. It focuses on what is reasonably feasible for a typical member of the relevant group(s), not on subjective decisions made by particular entities, which might be driven by privileged legal advice or strategic preferences.
• The CAT is not required to treat the class as homogeneous. It may:
  • segment the class into sub‑groups with materially different profiles (e.g. large institutional traders vs small businesses vs individuals);
  • assess practicability separately for each sub‑group; then
  • stand back and make an overall judgment on whether, in the round, opt‑in can reasonably be said to be practicable or not.

The Court recognises a spectrum of cases:

• “Merricks‑type” cases: huge numbers of consumers with small individual losses – opt‑out likely the only practicable route.
• Large commercial claimant cases: sophisticated entities with large individual claims – opt‑in or individually coordinated litigation is often viable; opt‑out will be harder to justify.
• Mixed cases (like Evans): both large institutional and small individual claims. Here, the CAT must balance the interests of different segments and cannot let a low‑value “tail” dictate the procedural form for the high‑value “dog”.

5.2.2 The CAT’s evaluation in Evans

On the evidence, the CAT majority concluded that:

• There existed a substantial group of large, sophisticated financial institutions and corporates with relatively large alleged losses (often in millions of pounds per entity).
• There also existed a very large number of small entities and individuals with modest claims (e.g. in the tens or hundreds of pounds).
• For the large institutions:
  • Opt‑in litigation was inherently practicable. They had experience, resources and sufficient quantum to justify participation, if they considered the merits warranted it.
  • Their lack of enthusiasm for opting in reflected an unwillingness (given their assessment of risk, cost and value), not an impossibility.
• For the small entities:
  • Opt‑in was in practice not realistic: claims were small, long‑past FX trades hard even to identify, and the incentives to engage with litigation minimal.

However, when the CAT “stood back”, it concluded that it would be wrong to allow the relatively tiny aggregate value of the small claims segment to drive the decision to allow opt‑out primarily for the benefit of large institutions. That would, in its words, be allowing “the tail to wag the dog”.

The Supreme Court holds that this was a legitimate and reasonable evaluative judgment squarely within the CAT’s discretion.

5.2.3 Critique of the Court of Appeal’s approach

The Court of Appeal had:

• Constructed its own detailed table of claims by turnover band;
• Emphasised that most class members (by number) had modest turnovers and claims in the region of £16,000;
• Downplayed the significance of large institutional claims, characterising their potential claims as “relatively modest” and suggesting costs would dwarf recoveries even for them;
• Expressly stated it “respectfully disagreed” with the inferences drawn by the CAT from the data and preferred the reasoning of the dissenting CAT member, Mr Lomas.

The Supreme Court’s response is twofold:

1. Standard of appellate review: Le Patourel itself had emphasised that opt‑in/opt‑out is an archetypal case management and evaluative decision for a specialist tribunal, to which appellate courts should attach a “broad margin of judgment”. The Court of Appeal’s re‑weighing of evidence and substitution of its own inferences was therefore inconsistent with its own prior guidance.
2. No error of law shown: The Supreme Court rejects the contention that the CAT misdirected itself by focusing on a single “average claim” or ignoring “the missing middle”. The CAT had the granular data before it, explicitly warned against the “danger of averages” and recognised the impracticability for many small claimants. There were no clear indications that it had overlooked material evidence. Its decision to give relatively little weight to the small claims’ aggregate value was a value judgment, not a legal error.

5.2.4 Regime provides opportunities, not guarantees

A key conceptual statement in the judgment is that the collective proceedings regime:

• Expands opportunities for claims to be brought which would otherwise be commercially infeasible;
• But does not guarantee that every theoretically valid claim will be litigated or that every group of potential claimants will find the regime economically viable;
• Operates within the realities of commercial funding and risk assessment; it does not immunise claimants from those background forces.

Thus, the fact that claims will not, in practice, be brought on an opt‑in basis does not automatically make opt‑in “impracticable” or entitle claimants to opt‑out as a “last resort.”

5.3 Issue 3 – Vindication of rights, deterrence and access to justice

The Court of Appeal had placed weight on:

• Facilitation of the vindication of rights; and
• Deterrence of future wrongdoers,

as factors pointing towards opt‑out, especially where there would be no litigation otherwise.

The Supreme Court recalibrates this.

• Yes, vindication and deterrence are core policy objectives of the collective proceedings regime (as noted in Merricks).
• But they are inherently balanced against countervailing aims: avoiding unmeritorious or disproportionately burdensome litigation and protecting defendants’ own access to justice.
• There is no presumption in favour of opt‑out. The correct starting point is neutral: the CAT must choose the procedural form that best advances the overriding objective on the facts of the particular case.

The Court specifically disapproves any suggestion that:

• If no action will be brought unless it is opt‑out, this is conclusive in favour of opt‑out. It is a “powerful factor” but not a “trump card”.

The CAT properly recognised that:

• The banks had already been sanctioned by the Commission through fines or immunity arrangements.
• Allowing a very weak claim to proceed on an opt‑out basis would risk over‑enforcement, not better enforcement, which is contrary to the public interest as much as under‑enforcement.

The Court endorses the CAT’s observation:

“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.”

5.4 Issue 4 – Use of the Commission’s Sterling Lads decision

5.4.1 Why Sterling Lads was problematic

After the CAT’s judgment, the Commission adopted:

• a settlement decision in Sterling Lads against Barclays, RBS, UBS and HSBC; and
• a detailed “ordinary” infringement decision against Credit Suisse.

The Court of Appeal treated the Sterling Lads ordinary decision as:

• Admissible in the Evans litigation; and
• Having significant “probative value” in showing likely harmful effects of the FX chatroom cartels, which in turn supported the Evans claims on causation.

In doing so, it:

• Used factual recitals from Sterling Lads as the basis of its introductory factual narrative, eliding distinct cartels into a single “FX cartel”.
• Inferred that the detailed economic analysis in Sterling Lads revealed what was “hiding behind” the shorter settlement decisions in Three Way Banana Split and Essex Express, by analogy.

The Supreme Court finds this approach both incorrect and unfair.

5.4.2 Hollington and fairness apply in the CAT

Contrary to the Court of Appeal’s assumption, the Supreme Court holds that:

• Rule 55(1)(b) of the CAT Rules (on control of evidence) does not disapply common law principles of fairness, including Hollington.
• The same rationale explained in Rogers v Hoyle applies: a tribunal must reach its own factual conclusions based on evidence before it; it cannot simply “take over” another body’s evaluative findings.
• The rule applies with particular force where the party against whom findings are tendered was not a party to the prior proceedings – as here, where the Sterling Lads decision was addressed to Credit Suisse (and some other banks) but not to some of the Evans defendants, and concerned a distinct chatroom and cartel.

The Court rejects the notion that the CAT can simply be “trusted” to assign appropriate weight. The problem is not only about weight; it is about admissibility consistent with natural justice.

5.4.3 Interlocutory vs trial use of prior findings

The Supreme Court accepts the line of authority (e.g. Tulip Trading) that at interlocutory stages:

• Prior judgments, reports or decisions, though inadmissible at trial as factual evidence, may be used to show what kinds of evidence might reasonably be expected to be available at trial, thereby helping a party to show that there is a serious issue to be tried or a real prospect of success.
• They may also be referred to for factual context or as records of evidence, but not as binding or presumptively correct findings on disputed points.

However, in the present case:

• The Court of Appeal did not use Sterling Lads in this limited way. It treated the Commission’s analysis and conclusions about likely harmful effects as themselves probative of harm in different infringements against different defendants.
• The Supreme Court considers there is no clear route by which Sterling Lads would identify new, case‑specific evidence capable of curing the fundamental causation gap identified by the CAT.

5.4.4 Crehan revisited

The Court addresses the argument that Crehan v Inntrepreneur established a special status for Commission decisions as admissible evidence. It explains that:

• Crehan was not a follow‑on case; it raised different issues about the duty of sincere co‑operation under EU law (pre‑Brexit).
• Lord Hoffmann’s remark that Commission decisions were “evidence properly admissible” was obiter, did not examine Hollington, and is best read as referring to the Commission’s economic expertise rather than as creating a general evidential exception.

Therefore, there is no principled basis for treating Commission findings about a different cartel as admissible, persuasive evidence against non‑addressee banks in domestic follow‑on damages proceedings.

5.4.5 Resulting principle

The new, clear rule from Evans is that:

• Commission decisions addressed to third parties, concerning different infringements, are inadmissible as factual evidence in UK follow‑on damages proceedings against other defendants.
• They may be looked at only in the very limited interlocutory sense recognised in Tulip Trading: to indicate possible lines of evidence, not to provide evidence themselves.

6. Simplifying Key Concepts

6.1 Opt‑in vs opt‑out collective proceedings

• Opt‑in: Only those who actively sign up (opt in) are claimants. The class is usually smaller, but each member has chosen to litigate and can be assumed to have assessed the merits and stakes for themselves.
• Opt‑out: Everyone within the defined class who is domiciled in the UK is automatically included unless they take active steps to opt out. This can produce very large classes, especially for consumer or SME claims.

Opt‑out procedures massively increase defendants’ exposure and the scale of litigation. The Evans judgment underlines that this heightened leverage must be justified by the combination of:

• Non‑frivolous merits; and
• Genuine impracticability of opt‑in, especially for the core value of the claims.

6.2 “Infringement by object” vs “by effect”

• An “infringement by object” (like the FX chatroom decisions) is a type of conduct that is, by its very nature, so likely to restrict competition that the competition authority need not prove actual effects on prices or output. It is enough to show the agreement or practice had a restrictive object.
• An “infringement by effect” requires proof that the conduct actually had anti‑competitive effects.

Critically, even in follow‑on civil damages claims, a claimant must prove loss and causation. A “by object” finding does not eliminate the need to prove that the infringement caused an overcharge or other loss. Evans makes clear that simply pointing to a Commission “by object” decision does not automatically satisfy the “strength of the claim” factor under rule 79(3)(a), especially on causation.

6.3 Umbrella effect

The “umbrella effect” theory asserts that:

• Even non‑cartel firms may raise their prices when cartel members raise theirs, “sheltering” under the price “umbrella”.
• Customers who bought from non‑cartel suppliers can therefore also suffer loss caused indirectly by the cartel.

In EU law, Kone confirmed umbrella claims are, in principle, possible. But they require careful economic evidence showing that the cartel actually relaxed competitive constraints sufficiently to allow others to increase prices. In Evans, the Supreme Court agrees with the CAT that the umbrella theory, as pleaded, lacked sufficient detail and plausibility to support the broad Class B claims.

6.4 Strike‑out vs summary judgment vs merits assessment under rule 79(3)(a)

• Strike‑out (rule 41): Used where a statement of claim discloses no reasonable grounds; the test is usually akin to “no real prospect” of a valid cause of action even taking allegations at their highest.
• Summary judgment (rule 43): Granted where a party has no real prospect of succeeding on a claim or issue and there is no other compelling reason for a trial.
• Merits under rule 79(3)(a): A broader, more nuanced assessment: the CAT may consider how strong or weak the claim appears, even if it surpasses the strike‑out/summary threshold. This can influence the appropriateness of an opt‑out order.

Evans clarifies that surviving strike‑out or summary judgment does not neutralise the CAT’s ability to regard a claim as “very weak” in the rule 79(3)(a) balancing exercise.

6.5 The “leveraging effect” of opt‑out collective proceedings

The term “leveraging effect” refers to how opt‑out, aggregate damages claims can be used to exert settlement pressure:

• Defendants face potentially enormous exposure and reputational risk, often irrespective of how many class members will ultimately claim.
• Defence costs are front‑loaded and large.
• Risk aversion and capital market perceptions can drive pragmatic settlements even where merits are debatable or weak.

Evans acknowledges that this leverage can be a legitimate tool for enforcing strong claims but is dangerous if attached to very weak ones. Hence the insistence on careful merits and practicability scrutiny.

6.6 Hollington v Hewthorn, simply explained

The Hollington rule, in simple terms, says:

• “You cannot prove you are right in your case just by showing someone else has already decided you are right in some other proceedings.”

Each court or tribunal must hear the evidence and make its own findings. If you want to prove a fact, you must rely on evidence (documents, witnesses, expert reports), not on another judge’s or regulator’s conclusion about that fact — particularly where the person you are suing was not even a party to the earlier case.

Evans confirms that this principle applies just as much to Commission decisions in the CAT as to, say, criminal convictions in the High Court.

7. Impact and Future Implications

7.1 For the CAT’s gatekeeper role

Evans significantly strengthens and clarifies the CAT’s gatekeeping mandate:

• The CAT must actively consider the merits and practicability factors in rule 79(3) when choosing between opt‑in and opt‑out, and is not to treat them as neutral by default.
• It has a wide evaluative discretion on these issues; appellate courts should be slow to intervene absent a clear legal misdirection.
• The decision endorses a robust, economically literate scrutiny of pleaded theories of harm at certification, particularly in complex financial markets.

7.2 For class representatives and funders

Class representatives and their funders will have to adjust strategies:

• Merits demands are higher for opt‑out: In complex or speculative theories of harm, they should not expect opt‑out certification merely because the case survives strike‑out or is backed by a Commission infringement decision “by object”.
• Careful pleading and economic grounding: The CAT and Supreme Court will expect a coherent causal mechanism tying the infringement to the alleged market‑wide loss. Generic overcharge assumptions or sweeping umbrella theories are unlikely to suffice.
• Class design matters: Where the class mixture includes both very large and very small claims, thought must be given to:
  • whether to exclude large institutions from an opt‑out class to avoid “tail wagging dog” issues; or
  • whether to pursue separate strategies for different segments.

7.3 For defendants in competition collective actions

For defendants, Evans:

• Confirms that they may properly resist opt‑out certification on the basis that:
  • The claims, while not frivolous, are materially weak; and/or
  • Opt‑in is objectively practicable for the core, high‑value part of the class.
• Strengthens protection against attempts to buttress weak follow‑on claims by importing factual findings from foreign regulators’ decisions relating to other undertakings or infringements.
• Provides principled arguments for resisting over‑expansive umbrella claim classes where the causal theory is thin.

7.4 For the use of EU Commission decisions post‑Brexit

The case has long‑term significance for the interface between EU competition enforcement and UK private damages actions:

• Commission decisions still underpin many follow‑on actions for historic infringements, but their binding effect is strictly delimited (to addressees, infringement findings and specified legal issues).
• Evans draws a clear evidential line: Commission findings in decisions against third parties cannot be treated as evidential shortcuts in domestic litigation.
• Domestic courts and the CAT retain full autonomy and responsibility for fact‑finding on causation and loss, even in the shadow of detailed Commission analysis.

7.5 For appellate practice in collective proceedings

Evans sends a strong signal that:

• Appellate courts must respect the CAT’s specialised experience and avoid re‑trying evaluative, fact‑sensitive issues such as practicability and the weight to be given to different class segments.
• Where the CAT has articulated a coherent reasoning path within the language of rule 79 and rule 4, differences in emphasis or inference should not lead to reversal.

8. Conclusion

Evans v Barclays is a shaping decision for the UK’s competition collective proceedings regime. It does three main things:

1. Re‑centres merits and practicability: The strength of the claim and the practicability of opt‑in proceedings are real, not nominal, constraints on opt‑out certification. Weak claims and objectively practicable opt‑in routes will often point strongly against opt‑out.
2. Affirms the CAT’s wide discretions: Opt‑in/opt‑out decisions are fundamentally case‑management and evaluative; appellate courts must show deference and avoid substituting their own assessments on finely balanced factual questions.
3. Clarifies evidential boundaries: The Hollington v Hewthorn principle applies in the CAT, excluding the use of Commission decisions against third parties as factual evidence. Foreign regulators’ findings cannot be used to plug gaps in causation theories in domestic follow‑on claims.

The resulting landscape is one of guarded expansion: the collective proceedings regime remains a powerful tool for redressing competition infringements and deterring anti‑competitive conduct, but its most potent form — opt‑out, aggregate damages — is reserved for cases where both the merits and the practical realities justify it. Evans stands as a key reference point for that calibration.