The opt-out collective proceedings regime is a recent innovation, but there are signs already that winds of change are starting to blow in the Competition Appeal Tribunal, which may take away some of its lustre. In this article, I will consider the direction from which the winds are coming.
The CAT’s jurisdiction in respect of opt-out collective proceedings, arose in 2015, through amendments to the Competition Act 1998 made by the Consumer Rights Act 2015. The purpose was clear. Competition law infringements may cause small losses to very large numbers of people. No rational consumer will issue a claim for £20, £50 or £100. The economics do not work. The wrong, if there is one, disappears into the cracks and the profits of unlawful behaviour follow them.
The opt-out regime was designed to close those cracks. Under section 47B of the 1998 Act, claims may be combined in the Competition Appeal Tribunal and brought by a class representative. In opt-in proceedings, class members must take a positive step to join. In opt-out proceedings, UK-domiciled class members are included unless they opt out. The Supreme Court put the point crisply in Evans v Barclays Bank Plc [2025] UKSC 48: opt-out proceedings allow claims to be brought on behalf of thousands or millions of people who may not even know the litigation exists. That is their strength. It is also a potential source of danger for corporate defendants, who may find themselves engulfed in “lawfare”.
For claimants, the advantages are obvious. Scale creates value. Value attracts funders. Funders pay the lawyers and experts. A claim worth nothing in the hands of one consumer may be worth hundreds of millions, or billions, in the hands of a class representative. Consumers, claimant lawyers and litigation funders have greeted the regime enthusiastically accordingly. Defendants have been less enthusiastic. One can see why. Opt-out proceedings may pose a very large contingent liability. The mere existence of the claim may create pressure to settle.
The major decision on the correct approach to certifying a claim for opt out collective proceedings, was Merricks v Mastercard Inc [2020] UKSC 51. In broad terms, Merricks made certification easier. It emphasised that the collective regime permits aggregate damages and does not require individual assessment of each class member’s loss. It also treated the certification stage as a gateway, not a trial. The question was not whether the claim would win. It was whether it was suitable to be tried collectively. The CAT has since applied the Pro-Sys test: the methodology must be sufficiently credible or plausible, grounded in the facts, and supported by some evidence of available data.
After Merricks, CAT did not become a rubber stamp, indeed quite the contrary. Certification hearings can be delicately regarded as a necessary form of satellite litigation, fought with courteous ferocity. The Tribunal must be satisfied as to authorisation and eligibility. The proposed class representative must be suitable. The claims must raise common issues. The class must be identifiable. The case must be suitable for collective proceedings. But, after Merricks, many assumed the wind was blowing in favour of certification.
At the tail end of 2025, the Supreme Court handed down judgment in the case of Evans v Barclays Bank plc [2025] UKSC 48The claim in Evans concerned alleged losses arising out of foreign exchange infringements. The Tribunal was troubled by causation. The Commission decisions concerned information exchanges between a small number of traders, with information of short-lived commercial value. The claim sought to turn that into very large losses across a wide range of transactions. The Tribunal thought the claim was weak, although it did not strike it out immediately. It refused opt-out certification. The Court of Appeal reversed it. The Supreme Court restored the Tribunal’s approach.
The significance of Evans lies in three propositions.
First, the strength of the claim matters when deciding between opt-in and opt-out. Rule 79(3) expressly identifies it. The Supreme Court rejected the idea that strength is generally neutral. If the claim appears strong, that may support opt-out certification. If it appears weak, that may count against it, even if the claim survives strike-out or summary judgment: Evans at [99]-[101].
Secondly, the Court recognised the leverage created by opt-out proceedings. At [89], the Supreme Court referred back to Merricks and the risk that opt-out class actions confer substantial legal and commercial advantages on claimants, and burdens on defendants, capable of opportunistic exploitation. At [92], the Court sharpened the point: if claimants are given assistance to extract value from unmeritorious claims, the parties are unlikely to be on an equal footing. That is not access to justice. It is creation of settlement pressure.
Thirdly, practicability matters. At [112], the Court said that if proceedings can practicably be brought as opt-in proceedings, then generally they should be. That is not an absolute rule. But it is a strong indication. Why give claimants the additional advantages of opt-out proceedings, and expose defendants to additional commercial pressure, if an opt-in claim is a practical route?
The principles in Evans are not anti-claimant. They are directed at ensuring access to justice to both sides. The Supreme Court did not close the door to opt-out proceedings. It reminded everyone that the door has a lock.
Waterside Class Limited v Mowi ASA [2026] CAT 32 concerned an opt-out claim on behalf of consumers who had bought salmon. The allegation was that producers colluded to increase wholesale prices, and that the overcharge was passed on to consumers. The proposed class was enormous: approximately 35 to 44 million people, or 18 to 23 million households.
The case did mention Evans albeit only in passing. At [4], the Tribunal relied on Evans for the proposition that it could not attach weight to the evaluative decisions of another decision-maker, namely the Commission’s statement of objections. But the deeper significance of Waterside is not about merits. It is about economics. The CAT was concerned by the cost-benefit balance. It noted the reality that collective proceedings may principally benefit lawyers and funders. In a private action, a rational litigant watches the costs against the likely return. In a funded class action, that discipline may be weaker: Waterside at [16]-[20].
The Tribunal refused to certify the claim as it stood. The PCR had not adequately addressed what money would actually get back to the class, despite a costs budget exceeding £20 million. At [47]-[49], the Tribunal said that where the cost-benefit balance is plainly against continuation, certification would not ordinarily be appropriate.
This approach is congruent with Evans. It is not the same point, but it demonstrates a particular direction of travel. The regime is not there to create litigation for its own sake. It must produce justice for real consumers, in the real world.
Dr Maria Luisa Stasi v Microsoft Corporation [2026] CAT 34 illustrates the other side of the coin. The claim alleged that Microsoft abused a dominant position through licensing practices relating to Windows Server and cloud services. The proposed class comprised organisations which obtained licences to use Windows Server from rival cloud providers. The estimated aggregate damages exceeded £1.7 billion.
The judgment expressly adopted the Evans framework. At [34]-[37], the Tribunal explained that Evans clarified the rule 79(3) assessment: the Tribunal has a broad discretion; weakness of the claim may militate against opt-out; but in some cases the Tribunal may be unable to assess merits at certification and so may place no weight on that factor.
Microsoft relied heavily on Evans. It argued that the claim was standalone, diffuse and methodologically imprecise. It said large business customers could practicably bring opt-in claims, and proposed a bifurcated structure: larger customers opt-in; smaller customers opt-out; very large commercial entities excluded from the collective regime altogether: Stasi at [102]-[106].
After Evans, both sides claimed victory. The PCR said Evans was different: there were no comparable causation problems, the class was more diverse, and average claim values were smaller. Microsoft said Evans supported bifurcation, particularly where some class members were sophisticated businesses with potentially substantial claims: Stasi at [107]-[112].
The CAT’s decision was carefully weighed. At [113] onwards, it treated opt-in versus opt-out as an evaluative exercise. It accepted that strength and practicability mattered. But at [125] it described the claim as apparently viable, with a good prospect of success. That paragraph is important. It suggests that Evans is not as radical as some defendants might hope. A claim need not be obviously strong. It simply must not be so weak that opt-out certification becomes unfair leverage.
The Tribunal then considered whether the class could sensibly be split. It applied Evans at, but refused to turn that guidance into fragmentation. The available data did not justify Microsoft’s proposed division. Standing back, the case was not like Evans. It was not a weak claim being used to force settlement. It was a promising, difficult, early-stage claim. The Tribunal also recognised the claimant-side leverage: opt-out scale may be necessary to keep funding alive and allow many claims to be pursued at all: Stasi at [138]-[142].
Where the court fears to tread, the government may yet legislate. The DBT review may matter more than any one case. The Government’s call for evidence, published on 6 August 2025, seeks to review the opt-out regime after ten years. It expressly identifies the original objectives: redress, deterrence, and protection of defendants from unmeritorious claims. It also notes concerns about burdens on business, speculative claims, funding, certification, ADR, settlement, damages and distribution.
The direction of travel is not yet fixed. But the questions are ominous for claimants. The review asks whether the regime is proportionate, whether funder returns are fair, whether certification is predictable, whether ADR and voluntary redress should be strengthened, and what should happen to unclaimed damages. It says proposals for change will follow and be subject to consultation.
The recently announced Law Commission’s consumer class actions project sits beside this review. It is considering whether England and Wales should have a wider consumer class action regime, beyond competition law. Its stated objectives include improving access to redress, ensuring damages are distributed to the affected class, and promoting efficient litigation at proportionate cost. Work is expected to start in autumn 2026, with an initial scoping questionnaire due by 30 October 2026.
The Law Commission project will expressly have to accommodate the lessons of the DBT review. A wider consumer regime may be created. But if it is, it is likely to arrive with tighter controls on certification, funding, distribution, representative governance and settlement approval. The mood music is more regulation.
I would suggest that the Supreme Court decision in Evans has made opt-out proceedings harder to bring but only up to a point. It has moved the dial. It has made it easier for defendants to say: this claim is too weak, too speculative, too leveraged, or too capable of being brought on an opt-in basis. It has reminded the CAT that access to justice belongs to defendants as well as claimants. It has made rule 79(3) bite.
But its effect is, so far, muted. In Shotbolt v Valve [2026] CAT 4, the CAT treated a consumer claim involving up to 14 million people and modest individual losses as a paradigm opt-out case, notwithstanding Evans. In Stasi, the Tribunal certified an opt-out claim after distinguishing Evans. In Waterside, the refusal turned less on merits than on whether the projected costs and distribution plan made sense with considerations of costs benefit analysis to the fore.
The larger question remains unanswered. The CAT regime has attracted serious money, serious lawyers and serious ambition to drive forward collective redress. Whether that investment produces substantial returns for consumers and funders remains to be seen. There may be gold in prospect. Or it may, after all the expert reports, funding waterfalls and certification hearings, turn out that the only gold in prospect is fool’s gold.
A version of this article first appeared in Litigation Funding magazine here: Fools Gold