Professor Andreas Stephan v Amazon.Com, Inc & Ors

The Judgment of the Court (Lord Justice Snowden and Lord Justice Zacaroli):

1. These are two applications for permission to appeal a decision of the Competition Appeal Tribunal (the “ CAT ”) (Mr Justice Roth, Charles Bankes and Keith Derbyshire), in which the CAT determined to make two collective proceedings orders (each a “ CPO ”) under s.47 B of the Competition Act 1998 (“ ”). CA 1998

2. The defendants in both proceedings are companies in the Amazon group (“ Amazon ”). The claimants seek aggregate damages for alleged abuse of dominant position by Amazon contrary to the Chapter II prohibition in s.18 CA 1998 and (for the period prior to 31 December 2020) Art 102 of the Treaty on the Functioning of the European Union. The claims arise from certain practices relating to Amazon’s online store.

3. The first claim is brought on behalf of a class of consumers. The class representative is Robert Hammond (“ Mr Hammond ”). Both claims are brought, and each CPO was granted, on an “opt-out” basis.

4. The second claim is brought on behalf of a class of third-party sellers of goods via Amazon’s online store. The class representative is Professor Andreas Stephan (“ Professor Stephan ”).

5. Each claim was the subject of an earlier carriage dispute. In Mr Hammond’s case, the CAT preferred his appointment as class representative over a Ms Julie Hunter in its decision published at [2024] CAT 8 . Permission to appeal was refused by the Court of Appeal.

6. Amazon seeks permission to appeal the CPO made in both claims. The issue raised by the application for permission to appeal in each case is different, and we deal with them in turn. The appeal in Mr Hammond’s case

7. The only question raised by Amazon’s appeal against the CAT’s decision in Mr Hammond’s claim relates to the litigation funding agreement (“ LFA ”) reached between Mr Hammond and FourWorld Global Opportunities Fund Ltd (the “ Fund ”).

8. Before turning to the detail of the funding arrangements, it is important to put this aspect of the CAT’s decision into context. The question before it was whether to certify Mr Hammond as the class representative, pursuant to Rule 78 of the Competition Appeal Tribunal Rules 2015. It may do so if it is “just and reasonable” for Mr Hammond to act as the class representative: Rule 78(1)(b).

9. Rule 78(2) provides further guidance as to what the CAT must consider when determining whether it is just and reasonable to appoint a person as class representative. It shall consider whether that person:- “(a) would fairly and adequately act in the interests of the class members; (b) does not have, in relation to the common issues for the class members, a material interest that is in conflict with the interests of class members; (c) if there is more than one applicant seeking approval to act as the class representative in respect of the same claims, would be the most suitable; (d) will be able to pay the defendant’s recoverable costs if ordered to do so; and (e) where an interim injunction is sought, will be able to satisfy any undertaking as to damages required by the Tribunal.”

10. Rule 78(3) provides that in determining whether the proposed class representative would act fairly and adequately in the interests of the class members, the CAT shall take into account all the circumstances, including:- “(a) whether the proposed class representative is a member of the class, and if so, its suitability to manage the proceedings; (b) if the proposed class representative is not a member of the class, whether it is a pre-existing body and the nature and functions of that body; (c) whether the proposed class representative has prepared a plan for the collective proceedings that satisfactorily includes— (i) a method for bringing the proceedings on behalf of represented persons and for notifying represented persons of the progress of the proceedings; and (ii) a procedure for governance and consultation which takes into account the size and nature of the class; and (iii) any estimate of and details of arrangements as to costs, fees or disbursements which the Tribunal orders that the proposed class representative shall provide.”

11. It is clear from the above that the CAT was required to exercise a discretion (it “may” authorise an applicant to act as class representative) based upon a broad evaluative judgment, as to whether it was “just and reasonable” for Mr Hammond to act as class representative. An appeal against a discretionary exercise of wide case management powers by a specialist tribunal faces a particularly high hurdle. The CAT’s conclusion about the funding arrangements was but one factor within that broad evaluative judgment.

12. A defendant will have an obvious interest in the CAT’s consideration under Rule 78(2)(d) of the ability of a proposed class representative to pay any adverse costs orders. In that respect, the CAT recorded at §38(2) of its judgment that Mr Hammond had ATE cover of £15 million as regards adverse costs. The CAT concluded that arrangement was sufficient to pay Amazon’s reasonable and proportionate costs if ordered to do so, and Amazon does not seek to appeal that finding.

13. Although Amazon does not seek to challenge Mr Hammond’s arrangements to be able to pay any costs orders in its favour, it does seek to challenge the CAT’s approach to Mr Hammond’s arrangements to fund his own conduct of the proceedings. Amazon’s sole ground of appeal is that “the Tribunal erred in its approach to the suitability of the class representative’s funding arrangements at the certification stage”.

14. In this regard, at §42 of the CAT’s judgment, it was observed that LFAs for collective proceedings can give rise to a number of issues. Aside from being satisfied that the funding is adequate to enable the class representative to cover their costs budget (see above), the CAT said that it was concerned to ensure that the class representative remains in essential control of the proceedings for the benefit of the class, therefore it will in particular consider the provisions of the LFA: “(1) concerning settlement; (2) concerning termination by the funder; (3) for control of legal costs; and (4) concerning payment to the funder, including its return.”

15. Amazon’s argument focusses on issue (4). Its essential contention is that the potential return to be made to the Fund is “wholly unreasonable” and higher than was “called out” by the CAT in an earlier case as “not defensible”, and the CAT should have declined to certify Mr Hammond as class representative accordingly.

16. In any context other than collective proceedings, it might be thought anomalous that a defendant should be able to advance arguments about the detail of the arrangements for a return to be made to a funder of litigation against it. However, no-one suggested that Amazon was not entitled to raise the point. But it is important to bear in mind that Amazon’s purpose in doing so is unlikely to be motivated by altruistic concern for the members of the class of claimants.

17. The precise level of the Fund’s return depends on the stage the action reaches. The CAT acknowledged that the LFA “could potentially result in an exceptionally high return”. That potential return was, however, by the express terms of the LFA, always “subject to an order of the [CAT] to the contrary”. The CAT made no finding about the reasonableness of the return to the Fund. Its decision on this point was that “the question of whether the total Funder’s Fee is unreasonable is for more detailed consideration at another time”: §67(2).

18. Its reasoning is found principally at §64 and §66 of the judgment: “64. … At the end of the day, the amount of the funder’s fee or return which will be paid is subject to the scrutiny and approval of the Tribunal. That is in line with the approach now authoritatively set out by the Court of Appeal in its judgment in Gutmann v Apple Inc [2025] EWCA Civ 459 , where the Court emphasised the discretion of the Tribunal at the time of a judgment, and the same approach clearly applies if the Tribunal is asked to approve a settlement. The recent judgment in Merricks v Mastercard, Inc [2025] CAT 28 … demonstrates the exercise of that discretion to allow the funder considerably less than had been provided under the LFA in circumstances where the proceedings had a very poor result. That approach is particularly apposite here since, as the two sides’ competing examples demonstrated, the question whether the return stipulated under the LFA is exorbitant will be affected by the timing of the event triggering payment. …

66. We consider that we have good reason to be cautious at the certification stage. The Tribunal cannot embark on an inquiry into the funder’s assessment of risk or internal expectation of likely damages, since that would involve seeing privileged material. Nor can we look at the negotiations between the funder on the one hand and the solicitors and the PCR on the other. By contrast, after judgment or settlement, the parties would be in a position to disclose privileged material: see Merricks . And we note that the full Federal Court of Australia, in a jurisdiction with much more experience of class actions, across a wide field of claims, said in Money Max Int Pty Ltd (Trustee) v QBE Insurance Group Ltd [2016] FCAFC 148 at [11]: “…Court approval of a reasonable funding commission rate is to be left to a later stage when more probative and more complete information will be available to the Court, probably at the stage of settlement approval or the distribution of damages.”

19. Mr Turner KC, who appeared for Amazon, submitted that the CAT made an error of law, in that it applied a default rule, which it gleaned from the decision of the Court of Appeal in Gutmann (above), that consideration of the reasonableness of a class representative’s funding arrangements should be deferred until the end of the proceedings, unless there were exceptional circumstances.

20. That, he said, was evident from §64 of the CAT’s decision, read together with §66. He also relied on §17 of the CAT’s reasons for refusing permission to appeal, which was in the following terms: “The Tribunal indeed expressed some concern about the potentially very high level of return for the funder under Mr Hammond’s LFA: [67(2)]. Instead of declining to certify unless and until Mr Hammond first investigated whether he could get a LFA with a lower return for the funder, we took the course of making clear that the funder’s fee was not approved and that the Tribunal would carefully scrutinise the level of the funder’s return following an award of damages or settlement: [66] and [67(2)]. As stated at [64], that approach has been endorsed by the Court of Appeal.”

21. At §7 of Gutmann , Sir Julian Flaux, Chancellor (with whom Green LJ and Birss LJ agreed) referred to the CAT in that case relying on another decision of the CAT in Gormsen v Meta Platforms Inc [2024] CAT 11 , where it was said that the return to the funder, and questions of costs generally, are controlled by the CAT on settlement or judgment. At §9 of Gutmann , Sir Julian Flaux cited the CAT (in Gutmann ) noting that its supervisory role meant that it could revisit at the end of the proceedings whether it was prepared to endorse the payment of agreed sums to the funder, when it may have better visibility as to the proportionality of the fee in relation to the damages awarded and the complexity of the proceedings, and could require further evidence.

22. Mr Turner made the point that, in these passages, the Court of Appeal in Gutmann was merely recording what the CAT had said in that case, and did not expressly endorse what was said. We agree. Far from supporting Mr Turner’s point, however, that shows that the CAT in this case cannot have had those paragraphs in the Gutmann case in mind when referring to an approach that the Court of Appeal had “authoritatively set out”, “emphasised” or “endorsed”.

23. The CAT’s use of the phrase “at the end of the day” at §64 of its judgment was not, in our view, intended to refer to the time at which a funder’s return would necessarily be scrutinised by the CAT. It was plainly being used as a figure of speech meaning “ultimately” or “always”. The approach that the CAT was referring to as having been authoritatively set out by the Court of Appeal in Gutmann was the general proposition that the amount of funder’s fee or return “is subject to the scrutiny and approval of the CAT”.

24. That appears from that part of the decision in Gutmann where the Court’s conclusions were set out. Thus, in §82, Sir Julian Flaux said the CAT has “broad overarching powers to ensure that costs and expenses are dealt with fairly and proportionately, and in accordance with the principles of justice”. At §95, he said that the CAT may “at any stage of the proceedings make any order it thinks fit in relation to the payment of costs in respect of the whole or part of the proceedings”.

25. It is true that, at §99, Sir Julian Flaux said “Any issue as to the reasonableness of the funder’s return is to be addressed at the time of distribution”. It is important, however, to read that sentence in context. He was there addressing the argument that the class representative had fettered the class entitlement to damages by agreeing the terms of the funding agreement in that case. Sir Julian Flaux rejected that argument, making the point at the start of §99 that the arrangement made in the funding agreement “was importantly always subject to the supervisory jurisdiction of the CAT”. He then commented that, on the particular facts of that case, the reasonableness of the funder’s return would be addressed at the time of distribution.

26. For these reasons, we think it is clear that the CAT’s reference in this case to what had been endorsed by the Court of Appeal in Gutmann , was to the point that the CAT has a wide discretion to review a funder’s return at any stage of the proceedings and that it would be permissible to do that at the time of judgment or settlement.

27. Accordingly, we do not consider that it is arguable with a real prospect of success that the CAT in this case is to be taken to have misunderstood Gutmann as laying down a default rule that questions as to the reasonableness of a funder’s return were to be left to the end of the proceedings, save in exceptional circumstances.

28. Nor is it arguable with a real prospect of success that the CAT itself applied such a default rule. As appears from §64 and §66 of the CAT’s judgment set above, it took into account various factors in exercising its discretion. It cannot be said that these are factors which it was impermissible to take into account, or that its conclusion is otherwise perverse in leaving until later in the proceedings a determination as to the reasonableness of the Fund’s return, notwithstanding that it was potentially exceptionally high. If the CAT had thought that it was applying a default rule, which could only be departed from in exceptional circumstances, §66 could not have been written in the terms that it was.

29. Mr Turner submitted that the problem with an excessive rate of return is that it creates perverse incentives, and that the CAT failed to take this into account. It is inherent, unless a funder’s return is fixed by reference to the damages recovered (which it cannot be: see R (PACCAR Inc) v Competition Appeal Tribunal [2023] UKSC 28 ; [2023] 1 WLR 2594 ) that there can be a conflict of incentives between a funder and a class representative, specifically as regards decisions around settlement. That is an issue, however, which the CAT specifically addressed, at §54 to §55 of its judgment, where it concluded that the terms of the LFA provided sufficient protection in this respect. It is true that the CAT did not, at this point, refer to the potentially excessive returns, but that could not even arguably amount to an error of law. The relevance of the potentially excessive return to the Fund is that it might be more likely to create a conflict of incentives. What matters, however, is how the LFA deals with such conflict when it arises, which the CAT addressed.

30. Mr Turner further submitted that the CAT failed to address the point made in argument that the LFA had been agreed in circumstances where there was no other funder willing to offer funding, because it was made when there was an ongoing carriage dispute with Ms Hunter. The point made to us was that the CAT should have required Mr Hammond to go back into the market to see whether he could obtain a better deal once the carriage dispute had been resolved, because it would not be possible at the end of proceedings to attempt to reconstruct with any accuracy what the outcome of such testing the market would have been.

31. From the transcript excerpts we were shown, all that appears to have been submitted to the CAT was that, now that the carriage dispute had been resolved, there were prospects for funders (plural) to take the case on, the implication being that Mr Hammond ought to make fresh efforts to agree revised funding.

32. It is true to say that the CAT did not specifically identify this as a factor which weighed in the balance against deferring determination of the reasonableness of the Fund’s return. The CAT clearly had it in mind, however, making express reference to it at §67(1) of the judgment. It made further reference to it at §15 to §17 of the reasons for refusing permission to appeal, noting that the possibility of going into the market again needed to be balanced against the yet further delay that would cause in the proceedings.

33. As was made clear by Gutmann , the CAT is not bound by the terms of the agreement in fact reached between the class representative and the funder. Moreover, the question, at the point that funding arrangement is scrutinized, is not whether it was “competitive” but whether the terms, in particular the return to the funder, are reasonable in all the circumstances. To the extent that the competitive (or otherwise) circumstances in which the LFA was entered into are relevant to that determination, it was not suggested that it would be impossible to undertake that exercise at the end of the proceedings, merely that it would be more difficult. In these circumstances, we do not think it is arguable with a real prospect of success that the CAT’s decision to make a CPO in Mr Hammond’s claim involved an error of law or was irrational as a result of its rejection of the point made as to the timing and circumstances in which the LFA was agreed with the Fund.

34. For these reasons, we consider that Amazon’s ground of appeal in relation to Mr Hammond’s claim has no real prospect of success.

35. Mr Turner argued in the alternative that there is a compelling reason for the Court of Appeal to hear an appeal, because it raises important questions for the funding of class actions generally. He cited responses to a call for evidence from the Department for Business and Trade (the “ DBT ”) in August 2025, which advocated the CAT bringing forward the approval of funding terms, because this would lead to greater certainty – in particular for funders – and therefore mitigate against funders being put off altogether or increasing the cost of funding due to the additional risk that their return might later be challenged.

36. We accept that there are broader questions concerning funding of collective proceedings, which raise important issues of policy. That, however, does not provide a compelling reason for an appeal against the CAT’s case management decision in this case. The DBT has provided the appropriate forum for a wider public debate on policy in relation to funding. It would not be right for the court to seek to engage in such issues under the guise of resolving an appeal between these particular litigating parties where, for the reasons given above, there is no sufficiently arguable error of law to warrant giving permission on the merits. The appeal in Professor Stephan’s claim

37. The sole point raised in the appeal in Professor Stephan’s claim is whether the CAT erred in law in concluding that there was no actual or potential conflict of interest within the class. This is, as the CAT pointed out, a “wholly opportunistic” point taken by Amazon. It is the class members who are potentially disadvantaged by the existence of a conflict, and none of them has raised an objection. Nevertheless, particularly in an opt-out case where it is impossible to obtain the consent of the whole class, it is not suggested that it is improper for Amazon to take this point or, the point having been raised, that it need not be determined.

38. The relevant circumstances are, in summary, as follows.

39. Of the various abuses of dominance pleaded in Professor Stephan’s claim, only two are relevant for the purposes of the appeal: (1) The allegation that Amazon favoured, when selecting offers displayed as a “Featured Offer”, sellers who utilised Amazon’s fulfilment network (referred to as Fulfilment by Amazon – “ FBA ”), over sellers who utilised their own delivery network (Fulfilled by Merchant – “ FBM ”) (“ abuse 3 ”). (2) The allegation that Amazon made access to its “Prime” label contingent on the use of FBA (“ abuse 4 ”).

40. The source of the alleged conflict is that as a result of the alleged abuses, if proven, FBA sellers achieved an increase in sales (and hence profits) at the expense of FBM sellers. In fact, the position is more complicated, because many sellers used a mixture of FBA (for some sales) and FBM (for other sales). There are thus three potential groups: FBA sellers, FBM sellers and sellers using both FBA and FBM (in varying proportions). However, that complication does not change the underlying analysis.

41. In relation to abuse 3, it is important to appreciate that the claim for aggregate damages on behalf of the class will, for the most part, not be quantified on the basis of the direct effect of the abuse in terms of lost profits on sales diverted from the FBM sellers to FBA sellers. That is because, from the perspective of the class as a whole, the loss of profits by an FBM seller on a diverted sale would be counterbalanced by the profits made by an FBA seller on the sale.

42. The aggregate damages claim is primarily based on the indirect effects of the abuse. The indirect effects are said to be an increase in each of the following: (1) FBA fulfilment fees, (2) FBM fulfilment fees and (3) online marketplace fees. For the purposes of this argument, attention can be focussed on the FBA fulfilment fees and the FBM fulfilment fees. The FBA fulfilment fees are said to have increased because of the increased number of sales fulfilled by Amazon. Although the numbers of FBM sales are said to have been reduced, FBM fulfilment fees for the sales that did occur are said to have increased because third party delivery providers were unable to provide FBM sellers with the cheaper rates that would have been applicable due to economies of scale if more FBM sales had taken place.

43. In quantifying such losses, Dr Houpis (Professor Stephan’s expert) will opine on the rate at which FBM sales were diverted to FBA sales. But he will calculate the alleged overcharges on FBA fulfilment fees and FBM fulfilment fees without using that diversion rate, by way of among other things a “before and after” regression analysis.

44. As indicated above, the quantification of the claim for aggregate damages does not, for the most part, include claims for loss of profits on sales diverted from the FBM sellers to FBA sellers. That is because, generally speaking, the loss of profits on a diverted FBM sale would be counterbalanced by the profits made on the corresponding FBA sale.

45. Dr Houpis recognises, however, that there are two factual circumstances in which the quantification of the claim for aggregate damages will include lost profits on diverted FBM sales. The first is where an FBM sale was lost in favour of a sale by an overseas FBA seller (who will be outside the class). The second is where the profit margin on the lost FBM sale and the gained FBA sale would have been different.

46. Dr Houpis’s third report made clear at §§288-289 that an element of lost profits on FBM sales would be included in his estimate of aggregate damages in these two cases. His report also implied that he would employ his diversion rate in that computation. Accordingly, it is at least likely that the CAT will be required, in determining the aggregate damages at the first stage of proceedings, to make findings about the diversion rate in these respects.

47. Ms Lukacova, who represented Amazon on this application, contended that this gave rise to an actual or at least potential conflict between the class members. As developed in argument before us, we understood her point to be that in seeking to maximise the aggregate damages in respect of the two circumstances outlined above in which the loss of profits on a diverted FBM sale will not be counterbalanced by the profits gained on the FBA sale, the class representative will be motivated to argue for a higher diversion rate. However, when it then comes to the division of the aggregate damages at the second stage of the proceedings, the FBA sellers would wish to contend for a lower diversion rate so as to minimise the extent to which the FBM sellers could claim that the distribution should compensate them for loss of profits on diverted FBM sales, and/or minimise the extent to which such distribution should take into account that the FBA sellers actually profited from those diverted sales. But the FBA sellers would be bound at the second stage by any finding of the CAT as to the diversion rate at the earlier stage.

48. Mr Carall-Green, for Professor Stephan, argued that there was, on the facts, no such conflict or realistic prospect of any conflict. He said that this was because Dr Houpis’s evidence in his third report at §§280-283(a), which referred back to his second report at §24, was that, £ for £, an FBA seller’s claim for the losses that it suffered by overcharge of FBA fulfilment fees would be greater than any profit that it might have made on a diverted sale. He said that this meant that if and to the extent that the diversion rate was relevant to quantification or distribution of aggregate damages, FBA sellers would always wish to argue for a higher diversion rate.

49. Mr Carall-Green also said that this evidential point had been accepted by the CAT in its judgment on the carriage dispute ( [2025] CAT 6 at §84), by the Court of Appeal in refusing permission to appeal against that judgment, by the CAT in its judgment sought to be appealed ( [2025] CAT 42 , in particular at §144), and by the CAT in refusing permission to appeal ( [2025] CAT 66 , in particular at §§27-30).

50. In its judgment on the carriage dispute, the CAT dealt with the matter in this way, “84. Furthermore, we do not accept Ms Ford KC's argument that pursuing the logistics abuse (i.e. the favouring of products sold using FBA) is against the interests of "FBA merchants". Dr Houpis' preliminary estimate is that such merchants suffered an overcharge of about 25% on the Amazon logistics fees. Even after assuming a 50% pass-on rate to customers, he estimates that such merchants still suffered a loss of around £1 billion. Allowing that Ms Ford KC may be correct in submitting that Dr Houpis did not, for this purpose, take into account the potential reduction in the volume of sales using FBA in the counterfactual, the FBA merchants' damages from the logistics overcharge appear nonetheless to be very substantial. Ms Ford KC put forward a postulated scenario where the potential proportion of sales diverted away from FBA merchants in the counterfactual had the effect of eliminating the FBA merchants’ claim for the logistics overcharge, if the “lost profits” on those sales were brought into account. It is of course possible to come up with figures and assumptions which have that outcome, which we would add depends also on the rate of pass-through. But for present purposes, we think it is sufficient to say that having regard to Dr Houpis’ analysis, to which we were taken in some detail, we do not regard that as very plausible. It is just the sort of “speculative example” in which it is inappropriate to engage at this stage of a collective proceedings: see the observation of the Court of Appeal in [ Gutmann ] at [73]-[74].” As we have indicated, permission to appeal that judgment was refused by the Court of Appeal.

51. In its judgment sought to be appealed, the CAT returned to this finding. It said, “144. Amazon did not, as we understood it, seek to re-run the argument advanced by BIRA that there would have to be credited against the damages otherwise recoverable the ‘benefit’ to FBA merchants of preferenced sales which they would not make in the counterfactual. Even on that contention, on Dr Houpis’ estimation that benefit is greatly outweighed by the impact of higher fulfilment fees which he estimated as about a 25% overcharge for FBA, producing aggregate damages under that head alone, and after allowing for 50% pass-on, of around £1 billion. We heard nothing to impugn the view set out in the Stephan/BIRA Judgment in that regard at [84]. Amazon has not put forward any likely diversion rate that would produce a different result and, in our judgment, there is no conflict in the way that Professor Stephan, through the expert evidence of Dr Houpis, seeks to advance his case.”

52. We do not consider that there is any realistic prospect of challenging the CAT’s exercise of its discretion to make the CPO on the basis of this assessment of the evidence before it. Although Ms Lukacova indicated that Amazon did not accept Dr Houpis’ evidence, it did not produce any evidence to show that the level of the FBA overcharge was less than the profit on sales diverted from FBM sellers to FBA sellers. As a result, we consider that the CAT was entitled, at this stage, to take the view there is no realistic possibility of a conflict arising between members of the represented class which would disable Professor Stephan from advancing the case for both groups that he represents.

53. We understand, of course, that these proceedings are at an early stage, as the CAT recognised in reaching its conclusion at §84 of the judgment in the carriage dispute. If as a result of developments in the evidence, the approach taken by Dr Houpis changes or it becomes necessary for him to advance an alternative case theory in response to evidence adduced by Amazon, with the result that it might be thought that there was some conflict between the interests of the FBM sellers and the FBA sellers in relation to the level of the diversion rate, then it will be open to the CAT to revisit that question as part of its continuing supervisory role over the collective proceedings. Even then, it would not necessarily follow that Professor Stephan could not continue to represent the entire class. The CAT might, for example, consider that any potential conflict arising from the need to address separate arguments on the diversion rate could be addressed by the use of separate legal teams, experts and possibly funding, in the way suggested in UK Trucks Claim Ltd v Stellantis [2023] EWCA Civ 875 at §88 et seq. We recognise that the Trucks case involved a corporate trustee, whereas Professor Stephan is an individual, but that does not preclude the adoption of some mechanism to achieve a similar result. In any event that would be a decision to be taken then.

54. So far as abuse 4 is concerned, the CAT said, at §145, that as both FBA sellers and Amazon itself as a retailer were the beneficiaries of lost FBM sales “the point about conflict has little force as a loss by a FBM seller does not equate to a gain by an FBA seller”. Ms Lukacova suggested that although this might mean that the effect of the diversion rate might be reduced in relation to abuse 4, it was not eliminated. That might be so, but it does not rebut Professor Stephan’s answer to the alleged conflict, for the same reasons as in relation to abuse 3.

55. Amazon also contended that there is a compelling reason for an appeal to be heard, because the CAT sought to distinguish the Trucks case as one in which there was a “direct” conflict. Amazon suggested that this begged the question of what might be an “indirect” conflict, and that an appeal is necessary to clear up any legal confusion in this respect. We do not accept this.

56. The phrase “direct” conflict was used (at §146 of the CAT’s judgment) to describe the conflict of interest which arose in the Trucks case, where an argument that buyers of new trucks had not passed on part of the overcharge when they resold the truck necessarily meant that the buyers of used trucks would have suffered no loss attributable to the overcharge. This was, as the CAT observed, fundamental to the existence of the claims of one represented group of claimants. The word “direct” simply described the obvious nature of the conflict between the different class members in Trucks on the facts. We do not consider that it is arguable that the CAT was suggesting that, as a matter of law, there are two different categories of conflict of interest, one of which it would be permissible for a fiduciary to ignore.

57. For these reasons, we refuse permission to appeal in both cases.