Allergan PLC & Ors v The Competition and Markets Authority

There is a simple central question at the heart of this appeal and these applications: whether the CMA properly put its case to the witnesses called by the commercial parties and appellants (together the companies), when the companies appealed the CMA’s infringement decision of 15 July 2021 in Case 50277 (the Decision) to the Competition Appeal Tribunal (the Tribunal).

The companies’ appeals from the Decision have already given rise to three lengthy decisions of the Tribunal (referred to as H1, H2 and H3 respectively) on 18 September 2023 (H1), 29 September 2023 (H2), and 8 March 2024 (H3).

H1 found there to have been infringements of the Chapter II prohibition on abusing a dominant position in the market for 10mg and 20mg hydrocortisone tablets. There is no appeal in H1 before us, so I shall say little more about it.

H2 found there to have been infringements of the Chapter I prohibition on agreements between undertakings, having as their object the prevention, restriction or distortion of competition in the market for 10mg hydrocortisone tablets. The Tribunal in H2 did not, however, simply uphold the CMA’s Decision and find that all the companies’ grounds of appeal from it failed. It added some important caveats. It said at [24]-[26] that it was deciding the appeals on the basis of the evidence that it had heard, but that it was concerned that the CMA’s case had not been fully put to the companies’ witnesses (primarily Mr John Beighton), seemingly on the premise that issues of their dishonesty arose. It then provisionally dismissed the appeals on a factual basis that went beyond the Decision. The main additional findings in H2 were: (i) at [144(8)], that there was a conversation between Mr Amit Patel (of Auden) and Mr Beighton (of AMCo – as to which see [18] below) in which an agreement was reached, not only that AMCo would stay out of the market, but also that AMCo would take Auden’s product and sell it at around the prevailing market price (as set by Auden), (ii) at [153(2)(i)], that Auden, Waymade and AMCo behaved dishonestly in concluding the 10mg Agreement (defined in H2 at [7(1)]) as an infringing agreement made between Auden and AMCo between 23 October 2012 and 24 June 2016), and (iii) at [153(2)(iii)(a)], that “Mr Beighton was dishonest at the time of the conclusion of the Second Written Agreement (an agreement for the supply of 10mg hydrocortisone tablets made between Auden and AMCo on 25 June 2014); and that he lied about it in the witness box”. At [157]-[159], the Tribunal in H2 decided that it was in no position finally to determine the companies’ appeals. Instead, it directed further argument as to the implications of the CMA’s case not having been fully put. It said that “[t]he CMA [had] made a decision which we consider on the merits to have been correct” and “[h]ad Mr [Robert] Sully and Mr Beighton not been called, then we are entirely satisfied that the appeals ought to be dismissed for the reasons we have given”.

At the 2-day “due process” hearing on 26 and 27 October 2023, nearly 5 months before H3 emerged, the Tribunal significantly widened the enquiry that it had instigated. It did not confine itself to hearing submissions as to whether the CMA’s case had been fully put, and as to the implications of the CMA’s case not having been fully put. It re-opened the question of the nature of the CMA’s case as to the 10mg Agreement, and as to dishonesty. At the start of the hearing, the President said:

In H3, the broadened enquiry was consolidated. The Tribunal said at [4] and [16] that H2 had been provisional because “the Tribunal was concerned that a central aspect of [it] was never put to [Mr Beighton and Mr Sully]”. As a result, the companies’ appeals against the Decision succeeded, and the provisional findings in H2 could not stand, because of “a failure, on the part of the CMA, to put the adverse findings in [the Decision] to [Mr Sully and Mr Beighton]”, which “fatally undermine[d] the conclusion, otherwise open to the Tribunal … that there was sufficient material to uphold [the Decision] when considering (in substance) the documentary evidence alone”.

At [19(11) and (12)] of H3, the Tribunal made clear that the main thing that had needed to be put was the “existence of the collateral understanding”, which Mr Beighton had denied. The Tribunal accepted that Auden, the de facto monopoly manufacturer of 10mg hydrocortisone tablets, had transferred significant value to AMCo, the purchaser of those tablets that was threatening to enter the market as a manufacturer pursuant to a marketing authorisation received in September 2012 (the Marketing Authorisation). The “collateral understanding” to which H2 was referring was an alleged understanding, outside the terms of the Second Written Agreement between Auden and AMCo, that AMCo would not actually enter the market as a manufacturer during the term of that agreement. The Tribunal thought the inference from the significant value transfer was strong, but that it “needed to be assured that alternative explanations from the witnesses could not hold water”. Mr Beighton had answered questions from the Tribunal as to AMCo’s reasons for the arrangement between AMCo and Auden that could not “be dismissed out of hand”. The Tribunal would “never know … how Mr Sully and Mr Beighton would have defended themselves … from the inference of an anti-competitive collateral understanding”. In reality, however, the Tribunal thought that the real problem arose from the failure to put these matters to Mr Beighton, rather than Mr Sully, who was the in-house lawyer.

H3 then included lengthy expositions of: (a) the CMA’s case in H2 ([56]-[84]), (b) the case that was put by the CMA in H2 ([85]-[114]), (c) the questions put by the Tribunal in H2 ([115]-[124]), and (d) the closing submissions in H2 ([125]-[135]). At [131]-[133], the Tribunal made it clear that it did not understand how the CMA could have put its case without alleging that Mr Beighton was lying and was dishonest. H3 concluded at [145] by saying again that the failure of due process that it had identified fatally undermined its substantive decision in H2. The companies had put forward a positive case that, no matter how apparently plausible the inference of collateral understanding, there was no such collateral understanding. Mr Beighton gave coherent evidence that: (i) it would have been in AMCo’s interests to enter the market the moment it could, and (ii) the Second Written Agreement between Auden and AMCo was no more than a stop gap enabling AMCo to supply the market whilst it secured its own independent supply. The Tribunal could not properly disbelieve him without full cross-examination.

The CMA appeals from the decision in H3 on the grounds that the Tribunal had no basis for overturning its findings in H2 that the companies had flagrantly infringed the Chapter I prohibition by entering into the 10mg Agreement. Its case had been properly put to the companies’ witnesses. The CMA contends that it had not found dishonesty in the Decision and had not alleged dishonesty in response to the companies’ appeals and had no need to do so (something that the Tribunal repeatedly acknowledged). The CMA’s case had always been clear. It was expressed in the Decision at [6.10]-[6.17]: (i) AMCo was a potential competitor to Auden, (ii) the parties reached an understanding that AMCo would not proceed to place its own product on the market, (iii) in return for the (continuing) transfer of value from Auden. That was what was called in H2 and H3 the “10mg Agreement”.

In the course of the CMA’s oral opening, this court pointed out that, even if the CMA were right on these points, it still needed to appeal the parts of H2 that had: (i) found dishonesty (and possibly the finding that the Second Written Agreement was a true sham, and that there had been an oral conversation between Mr Amit Patel and Mr Beighton), (ii) declined finally to determine the companies’ appeals, and (iii) held that a further substantive hearing was required. The CMA ultimately filed, in the course of the hearing before us, an Appellant’s Notice against the decision in H2, for which it sought permission at the hearing.

At a preparatory hearing before this court, we directed that the companies’ applications for permission to appeal from the provisional findings against them in H2 should be heard at the CMA’s substantive appeal against H3 on the basis that, if permission were given, the substantive appeals would be heard at the same time. The companies raised many of the points, which they had raised in their applications for permission to appeal, also by way of respondents’ notices filed in answer to the CMA’s appeals. They suggested that the order in H3 should be upheld for the additional reasons in those respondents’ notices. I have not further distinguished between the points raised in the respondents’ notices and in the applications for permission to appeal. They are dealt with substantively together in section H of this judgment.

Accordingly, we have been dealing with essentially 3 substantive issues: (i) the CMA’s substantive appeal against H3, (ii) the companies’ applications for permission to appeal against H2, and (iii) the CMA’s application for permission to appeal parts of H2.

The companies raised a number of fundamental criticisms of the H2 decision, but they broadly submitted that important pieces of evidence had not been challenged by the CMA. First, Mr Beighton had said that he wanted to enter the market, because he would, in a normal generic market, have obtained 50% of that market (40,000 boxes per month), which would have been more profitable than the deal he had with Auden which only reached 12,000 boxes per month in 2014. Secondly, Mr Beighton was not challenged on his evidence that the Second Written Agreement was a stop gap, whilst he sorted out AMCo’s production issues, and that AMCo always intended to enter the market with its own product when it could. Moreover, Mr Beighton had continued to try to source 10mg hydrocortisone tablets from Aesica even after the Second Written Agreement. Thirdly, Mr Beighton had taken legal advice as to the commercial terms agreed in writing. He ought to have been directly challenged on clause 2.2 in the Second Written Agreement that allowed AMCo to enter the market on 3 months’ notice to Auden, and whether he (Mr Beighton) agreed not to exercise that right, and whether there was, in fact, any broader or different understanding beyond the written terms. Fourthly, Mr Beighton ought to have been challenged on whether he cancelled AMCo’s production of 10mg hydrocortisone tablets from Aesica because of the unwritten collateral understanding that he would not exercise his written clause 2.2 right to enter the market at all during the 2-year term of the Second Written Agreement. Fifthly, Mr Beighton ought to have been asked about the companies’ case that Auden was simply (lawfully) unilaterally incentivising AMCo not to enter the market. Sixthly, Mr Beighton ought to have been challenged as to his dishonesty or serious misconduct, as H3 said. The companies’ case was that the Tribunal was right to find that these procedural defects meant that the companies’ appeals from the Decision had inevitably to succeed. There was no need for a retrial. If the process were held to be so unsatisfactory that the companies’ appeals could not be simply allowed, then the companies’ alternative case was that a retrial was an unfortunate necessity. The companies also argued that the collateral understanding could not anyway persist beyond the time when Mr Amit Patel left Auden, and Actavis took over its business on 29 May 2015; a corporate understanding can only be held by human persons. Finally, the companies challenged the Tribunal’s findings as to object infringement and market definition.

Against this shortly stated background, I have decided that: (i) the CMA’s appeal from H3 should be allowed, (ii) the CMA’s application for permission to appeal from H2 should be granted, and its appeal should be allowed, (iii) the companies should not be granted permission to appeal from H2, (iv) the Tribunal’s provisional findings in H2 that the companies’ appeals from the Decision should be dismissed, should be finalised and reinstated, and (v) the Tribunal’s additional findings in H2 going beyond the Decision as to dishonesty and other matters should be overturned.

There are essentially 5 reasons for these conclusions. First, it was not appropriate for the Tribunal in H2 to proceed as it did. There was no need for a further hearing. It should have determined the companies’ appeals from the Decision on their merits on the evidence that it had heard. The parties had had their opportunity to present their evidence and to cross-examine the witnesses. There was no need for a second bite at the cherry. Secondly, even if a limited further hearing about whether cross-examination on dishonesty had been needed (which it was not), it was not appropriate for the Tribunal in H3 to engage in an entirely fresh examination and analysis of the CMA’s case. Thirdly, the CMA’s case was always clear. It did not involve any allegation of dishonesty, and the CMA was entitled to push back against the Tribunal’s insistence that it did. Fourthly, the CMA properly cross-examined the companies’ witnesses on the case it had held was established in the Decision and that it advanced in H2. Fifthly, none of the companies’ criticisms of the Tribunal’s core findings in H2 has any substance or real prospect of success on appeal.

Despite the fact that this appeal took 4½ days to argue and we have been provided with several thousands of pages of evidence and authorities, I do not regard the issues as complex. I mean no criticism of the Tribunal when I say that it followed an inappropriate procedure. I think it was doing what it thought right, but somehow lost sight of some of the essential realities of this kind of appeal from the CMA’s decisions. This judgment will proceed to deal with matters in the following order: (i) an outline of the essential facts, (ii) the Chapter I prohibition, (iii) the role of allegations of dishonesty in alleged infringements of the Chapter I prohibition, (iv) whether there was actually any confusion about the CMA’s case, (v) whether the CMA’s case was fully put to Mr Beighton, (vi) the appropriate conduct of an appeal from a decision of the CMA finding there to have been an infringement of the Chapter I prohibition, (vii) the additional points raised by the companies, and (viii) disposition of the CMA’s appeals and the applications for permission to appeal.

This section is a much slimmed-down version of the essential findings of fact made by the Tribunal at [7] and [27]-[153] of H2. It also includes some additional documents relied upon by the parties in argument.

The Decision found that three parties were involved in the relevant infringements, namely the 20mg Agreement (see [19] below) and the 10mg Agreement. They were Auden (as manufacturer), Waymade (originally as prospective market entrant), and AMCo (which took over from Waymade on 31 October 2012). This judgment does not need to be more specific as to the entities involved, save in respect of the period following 29 May 2015, when Mr Amit Patel left Auden, and Actavis took Auden over.

In essence, the 20mg Agreement was concluded in July 2011 between Auden (Mr Alan Barnard under the management of Mr Amit Patel) and Waymade (Mr Brian McEwan under the management of Mr Vijay Patel). Waymade had cleared all regulatory requirements to manufacture 20mg hydrocortisone tablets in March 2011. Auden sold Waymade 1,000 boxes per month at £4.50 per box and then repurchased 800 boxes from Waymade at £34.50 per box, on the basis that Waymade would not enter the market. Auden’s payments to Waymade in respect of 20mg hydrocortisone tablets between 11 July 2011 and 30 April 2015 amounted to over £1.8 million. H2 concluded at [110] that there was no alternative explanation for the 20mg Agreement: “Auden [was] paying Waymade to stay out of the market and maintain the existing (high) prices”.

The 20mg Agreement is the background to the 10mg Agreement. It is important to note, however, that 10mg hydrocortisone tablets accounted for 96% of hydrocortisone tablets dispensed between 2012 and 2017. Between July 2011 and September 2012, Auden supplied Waymade with 10mg hydrocortisone tablets at market rate (between £31.50 and £34.50 per pack). In September 2012, as I have said, Waymade obtained its own Marketing Authorisation for 10mg hydrocortisone tablets, thus becoming a potential competitor to Auden. The alleged 10mg Agreement was unwritten and entered into between the same people at Auden and Waymade by 23 October 2012. The substance of the 10mg Agreement was essentially the same as the 20mg Agreement. H2 concluded at [113] that “[f]or exactly the same reasons as we have articulated in relation to the 20mg Agreement, the 10mg Agreement can only have been an agreement between Auden and Waymade for Waymade to be paid to stay out of the market”. Between November 2012 and June 2016, Auden paid AMCo some £21 million in relation to 10mg hydrocortisone tablets. This was what H2 referred to as the “value transfer”. A week after the initial 10mg Agreement was allegedly concluded, AMCo replaced Waymade as party to it. The complex series of transactions that gave rise to that change do not matter, but from that date onwards, Cinven effectively owned AMCo. Cinven sold AMCo to Advanz on 21 October 2015. Accordingly, for our purposes, Cinven and Advanz are parties to these appeals as owners from time to time of AMCo. Allergan, through its generics business, Actavis, acquired Auden on 29 May 2015. Accordingly, for our purposes, Allergan and Actavis are parties to these appeals as owners from time to time of Auden.

The companies placed reliance on a due diligence report from Deloitte dated 23 October 2012 relating to Cinven’s intended acquisition of Waymade. Deloitte recorded that, based on information provided by Waymade, there would be a “new product launch” of 10mg hydrocortisone tablets in the UK in 2013. The report commented that the launch would take “market share from the incumbent supplier … the current market for Hydrocortisone tablets is supplied solely by [Auden]”.

The First Written Agreement was ultimately signed on 25 February 2014, but had an effective starting date of January 2013 and a 15-month term ending on 31 March 2014. It provided for sales of 6,000 boxes of 10mg hydrocortisone tablets by Auden to AMCo at £1 per box. It was drafted by AMCo’s solicitors, Pinsent Masons, who apparently never asked why the product was being sold at a discount of 97% to its market price (then, in excess of £30 per box). The negotiations were fraught and complicated by two features: first, the difficulties that AMCo were having in sourcing their own product from Aesica, and secondly, the attempts by Auden to persuade AMCo (or its parent) to buy it (Auden) out.

The Second Written Agreement has been regarded by all parties as rather more significant than the First Written Agreement. Again, it was drafted with the assistance of Pinsent Masons. Its commercial background was AMCo’s heightened realisation that the 10mg hydrocortisone tablets that it was hoping to source from Aesica were “skinny label” (known as the “orphan drug” problem). AMCo’s Marketing Authorisation only permitted it to be produced with a label that indicated it was for “replacement therapy in congenital adrenal hyperplasia in children” and not in adults. By mid-2014, AMCo’s customers had told them they were not (or were anyway less) interested in a reduced indication children’s 10mg hydrocortisone tablet (even though the tablet itself was identical). Ultimately, H2 decided at [62] and [115(3)(i)], using its test of market definition, that the 10mg “skinny label” product acted as a competitor to a 10mg “full label” product, and that they were substitute products. It is perhaps worth noting that H2 decided at [61]-[62] that Pinsent Masons’ competition advice to AMCo on this point, to the effect that the two products were not competitors, was wrong. Advanz has, as I have said, sought permission to appeal the finding as to market definition.

The Second Written Agreement had an effective date of 25 June 2014, and a term of two years. By clause 2.2, AMCo was obliged to “procure all its requirements in the Territory for hydrocortisone product(s) in tablet and capsule formulation from Auden on an exclusive basis”. There was, however, a proviso allowing AMCo to manufacture hydrocortisone tablets provided that it did not do so “without [giving] Auden at least three months’ written notice of its intention to do so”. Clause 17.2 allowed Auden, if notified of AMCo’s “intention to commence supply of its own version”, an option to terminate on three months’ written notice to AMCo. The Second Written Agreement provided for the sale by Auden of 12,000 packs per month at £1.78 per pack. On the basis of Ms Kelly Lifton’s evidence, H2 rejected at [67]-[69] the CMA’s suggestion that AMCo had slowed down production of its own 10mg hydrocortisone tablets. The delays were Aesica’s fault. That said, however, the Tribunal thought that “AMCo’s position in promising to stay off the market (which is what the CMA found) would have had far more traction if AMCo had actually been able to enter the market”.

The negotiations for the Second Written Agreement were dealt with in H2 at [80]-[86]. The CMA emphasised in oral argument some of the contemporaneous emails. I set these out in chronological order.

On 19 April 2014, Mr Beighton emailed Mr Guy Clark (also in AMCo’s office) as follows:

On 28 May 2014, Mr Beighton emailed Mr Amit Patel of Auden as follows:

On 13 June 2014, Mr Sully emailed Mr Beighton and Ms Jane Hill as follows:

On 15 June 2014, Mr Beighton emailed Ms Jane Hill, Mr Peter van Tiggelen, Mr Sully and Mr Guy Clark (all of AMCo) as follows:

The Second Written Agreement was signed, as I have said, on 25 June 2014. On that date, Mr Karl Belk wrote internally to Mr Guy Clark, Ms Jane Hill and others, copied to Mr Beighton and others, as follows:

Later, on 25 June 2014, Mr Guy Clark of AMCo wrote to Mr Beighton as follows:

On 27 June 2014, Mr Karl Belk of AMCo emailed Aesica to say that it was “with disappointment and regret that I must write to inform you that our Hydrocortisone tablet project will be suspended for the UK territory”.

On 28 June 2014, Mr Beighton wrote to his launch team as follows:

Advanz directed the court to a series of emails between AMCo and Aesica in mid to late 2014 in which the possibility of AMCo procuring its own hydrocortisone products was discussed.

Section 2 of the Competition Act 1998 (the 1998 Act) imposes the Chapter I prohibition on agreements, decisions and concerted practices which prevent, restrict or distort competition, as follows:

In BAI and EU Commission v. Bayer AG (Joined Cases C-201 P & C-301 P), 6 January 2004 (Bayer), the CJEU explained how a tacit acceptance, beyond simply unilateral conduct, could be sufficient to support an anti-competitive agreement at [100]-[102] as follows:

The Tribunal stressed repeatedly that there could be no infringement of the Chapter I prohibition without something crossing the line between the parties to the agreement or understanding (see [15(1)] and [93] of H2, and [24(1)] of H3). H3 found at [85] that the Decision had decided that there was a collateral understanding between Auden and AMCo that crossed the line, and that that understanding ought to have been put to AMCo’s witnesses. H3 concluded at [84] that such a case had not been put.

The Tribunal acknowledged in both H2 (see [24(3)], [24(5)] and footnote 182) and in H3 (see [38], [122(2)], and [134]), that dishonesty was not a requirement of an infringement of the Chapter I prohibition. An infringement of the Chapter I prohibition was a question of strict liability. I agree.

Section 36(3) of the 1998 Act provides that: “[t]he [CMA] may impose a penalty on an undertaking … only if [the CMA] is satisfied that the infringement has been committed intentionally or negligently by the undertaking”. But even an intentional infringement (as found by the Decision at [6.930] and by the Tribunal in H2 at [15(4)] and [151]) is not the same as one that is undertaken dishonestly.

[6.884-5] and [6.922-3] of the Decision held as follows:

At [143] in H2, however, the Tribunal held that the Second Written Agreement was a sham in the true sense, because the terms only reflected a part of the deal between Auden and AMCo. H2 decided that both written agreements were “dishonest shams” [153(2)(iii)(c)].

The CMA’s case as to both the 20mg Agreement and the 10mg Agreement was set out in the Decision at [6.1]-[6.18]. Those paragraphs started at [6.3] by explaining that: “[b]etween 11 July 2011 and 30 April 2015 Auden and Waymade shared a common understanding that: (a) Auden would supply Waymade with 20mg hydrocortisone tablets on terms that amounted to monthly payments (or ‘value transfers’) to Waymade; and (b) In exchange for these payments, Waymade would not enter the market independently with its own 20mg hydrocortisone tablets”. The Decision recorded at [6.7] that: “[n]o party or individual has given a credible explanation for these payments, other than that they were to buy off Waymade’s entry”. At [6.8]-[6.10], the CMA explained how, between 2008 and 2012, Waymade had obtained its Marketing Authorisation for its own 10mg hydrocortisone tablets. Auden supplied Waymade at market rate until then, when Waymade presented a competitive threat to Auden. Within a month, Auden and Waymade entered into the 10mg Agreement as follows:

At [24(2)] of H2, the Tribunal recorded the way in which the CMA’s counsel had described its case in closing submissions as follows:

At [91] of H2 the Tribunal considered that the Decision articulated “with sufficient clarity the nature of the 10mg Agreement which it [found] objectionable”.

Mr Tristan Jones KC, counsel for the CMA, explained the CMA’s case again on the first day of the due process hearing on 26 October 2023 as follows:

The CMA’s reply skeleton before us described its case at [10] in these terms:

The Tribunal seems to have struggled with aspects of the CMA’s case that I have articulated above. First, it found it hard to accept that the case did not involve an element of dishonesty or “naughtiness” (see [5] above, [24(2)] and [151]-[153] in H2, and [38], [122] and [131]-[134] in H3). Secondly, it did not accept that, when the CMA used the term “sham” to describe the written agreements in the Decision, it was not using that term in the sense explained by Diplock LJ in Snook v. London & West Riding Investments Ltd [1967] 2 QB 786 at 802 (“Snook”) (see [15], [143] and [157] of H2). A subsidiary element of that position is the Tribunal’s refusal to accept that the CMA’s allegation of “intention” did not also imply misconduct of some kind. Thirdly, as [145] in H3 explains, the Tribunal did not accept that, once the companies’ witnesses had put forward a coherent case that: (i) there was no collateral understanding, (ii) it was always in AMCo’s interests to enter the market, (iii) the Second Written Agreement was no more than a stop gap, they could be disbelieved without express cross-examination and challenge to these points. It was this latter problem that the Tribunal held at [145] in H3 fatally undermined the CMA’s case.

In my judgment, the CMA and the Tribunal were largely passing each other like ships in the night. The CMA stuck to its guns from the Decision to the due process hearing as to the nature of the case it was alleging, whilst the Tribunal consistently thought that the CMA’s case, as I have recorded it above, inevitably involved allegations that AMCo’s representatives, at least, had been guilty of both misconduct and dishonesty.

The truth is that the CMA’s case did not involve misconduct or dishonesty of any party or witness. An infringement of the Chapter I prohibition, as I have said, imposes strict liability. The parties did not disagree as to the law. What needed to be proved was an agreement or understanding between undertakings which had as its object the prevention, restriction or distortion of competition. The agreement or understanding could be inferred from all the facts, but something had to cross the line between the undertakings, even if it was only a nod or a wink or a tacit understanding (see Bayer at [36] above and CMA v. R (VW AG) and BMW AG [2023] EWCA Civ 1506, at [57]). It was not enough for one party unilaterally to seek to incentivise the other to act uncompetitively (in this case, not to enter the market).

As I see it, the CMA’s case was, and is, crystal clear. The disagreements between the parties and between the CMA and the Tribunal were about what, in terms of fairness and due process, had to be put to the witnesses before adverse inferences could be drawn against them. It is to that aspect of the case that I now turn.

I have framed this question by reference to Mr Beighton, but not Mr Sully, for two reasons. First, the parties’ arguments before us focused on Mr Beighton, not on Mr Sully. Secondly, the high point of the findings in H2 relate to Mr Beighton’s misconduct, not Mr Sully’s. In [153] of H2, the Tribunal found Mr Beighton to have been dishonest and to have lied in the witness box, but not Mr Sully. Although it finds the written agreements, which Mr Sully drafted, to have been “dishonest shams”, he “had no involvement that we have found in the negotiating of price and/or quantity of product supplied, and it is these aspects which drive the dishonest bargain that was reached. … It is the enormous margin that Auden appears to have gifted AMCo that is key, and this was not negotiated by Mr Sully”. The Tribunal, after “many days considering the evidence”, concluded that it would have been a counsel of perfection to have expected Mr Sully to have asked either himself or Pinsent Masons about this.

Thus, if the appeal is to be determined, one way or the other, as the parties submitted it should, on the basis of the findings in H2, it will be on the basis of the way in which Mr Beighton’s cross-examination was conducted, not Mr Sully’s.

The starting point for any analysis of Mr Beighton’s evidence is his witness statement. There is not space in this judgment to recite it all. But it is worth summarising its contents very briefly. First, it is to be noted that Mr Beighton was a very senior pharmaceutical executive, who had worked in the industry for 39 years. He had worked for Smith Kline Beecham, then as UK Managing director of Teva UK Limited, before becoming CEO of Goldshield Group Limited. When Goldshield rebranded as Mercury Pharma Group Limited, Mr Beighton oversaw its refocus, stabilization and growth, until it was acquired by Cinven on 31 August 2012. On 15 March 2013, he was appointed as CEO of AMCo, where he stayed until the end of 2015. Mr Beighton now holds non-executive roles in the pharmaceutical industry. He was an expert in the generics market, including being involved in the leadership of the British Generics Manufacturing Association.

At [4] of his witness statement, Mr Beighton explained, significantly in my view, that there was not, in fact a 10mg Agreement. His whole 41-page statement is directed at explaining, in some detail, why the “inferred 10mg Agreement” alleged by the CMA had not existed, and the written agreements were not shams. After reading his statement, it is hard to conclude that Mr Beighton did not understand full well what was being alleged against him by the CMA’s Decision. His conclusions at [114] and [115] make this point transparent:

In addition to Mr Beighton’s statement prepared for the appeal to the Tribunal, Mr Beighton had submitted to interviews under caution with the CMA during its investigations in October 2017. Mr Beighton was cross-examined before the Tribunal by Ms Marie Demetriou KC, for the CMA, for half of 23 November 2022 and most of 24 November 2022. That included a series of detailed questions put by the Tribunal itself towards the end. The transcript runs to nearly 300 pages.

Mr Beighton’s cross-examination started with questions about what had happened after Mercury was acquired by Cinven in August 2012. He accepted that he was involved in the acquisition of AMCo “from more or less that time”. Mr Beighton was asked about the Deloitte report (see [21] above) and the due diligence materials for the acquisition. Mr Beighton said that Cinven’s focus, in acquiring AMCo, was on the launch of a new 10mg hydrocortisone product.

There followed some cross-examination about whether Mr Beighton was curious about why Auden was selling 10mg hydrocortisone tablets to AMCo at such a huge discount to the market price. Mr Beighton understood that involved a transfer of value to AMCo. He said that he assumed that he had asked Mr Vijay Patel and Mr Brian McEwan (both of AMCo) why Auden had been prepared to do that deal, but could not recall the conversation. He assumed that Mr Vijay Patel said that he had persuaded Mr Amit Patel (of Auden) to do so. Mr Beighton said that he “wanted to understand that the deal was going to continue”. He did not delve further. Mr Beighton was then asked a series of questions about Mr McEwan’s and Mr Vijay Patel’s interviews with the CMA. Mr Beighton accepted that it seemed from the latter interview that both Mr Vijay Patel’s and Auden’s understanding of “this very beneficial advantageous deal” was “that Auden [believed] that if they do not provide all this money to AMCo, AMCo [would] launch its own product”. At the conclusion of that section of cross-examination, it was put to Mr Beighton that he understood at the time that Mr Brian McEwan and Mr Vijay Patel knew that the price only dropped to £1 when Waymade obtained its Marketing Authorisation. Mr Beighton denied that he did. After detailed questions about the market in 2012, Ms Demetriou put to Mr Beighton that he was unable to point to any explanation as to why Auden would have entered into the supply contract with AMCo. His answer is recorded as inaudible. It was at that point that the Tribunal said that it had some questions on that aspect. It was agreed that those questions should be reserved for later.

Mr Beighton then accepted that AMCo was a competitor of Auden in the 10mg hydrocortisone market, but an unequal one because of the orphan designation issue. Mr Beighton was then asked how the early increase in supplies of 10mg hydrocortisone tablets from Auden had been negotiated. He denied knowing or remembering. He did, however, say that he had in his head “when we moved from 6,000 packs to 12,000 packs and the discussions that I had with Mr [Amit] Patel at that point”. When challenged on the early increase, he said:

Mr Beighton prevaricated about whether he was in charge when the supply was increased from 2,000 to 6,000 boxes, but said: “[t]his is not to say I did not meet [Mr Amit Patel] at some stage, because I did on two or three occasions, but I do not remember specifically discussing 6,000 packs”. Although he said he was not trying to avoid responsibility, there was a lot going on at the time. He ultimately accepted that he would have known at the time about the very beneficial increase in volumes (for AMCo).

There followed an important passage in Mr Beighton’s cross-examination, where he mentioned some of the points made by the companies in answer to the CMA’s appeal:

The passage just cited is significant, because it makes perfectly clear that Mr Beighton understood during his cross-examination (as he had in his statement) where the CMA was coming from. He knew that he was being asked why Auden would agree to supply more and more product at cost price if there was no “pay for delay” agreement. He denied such an agreement. He thought, in hindsight, that Mr Amit Patel had made a mistake in “doing this deal with us”. Mr Beighton’s answer, when it was put that Mr Amit Patel was only paying because he knew that AMCo could launch its product, was: “[n]o, my point is, why would that make any difference?”, and “unless he thinks I am completely bonkers, why would I not launch my product as soon as I got access to 40,000 packs a month? I promise you that the economics of this I would have -- are hugely in favour of launching my own product”. This part of the cross-examination concluded with this:

In relation to the negotiations for the First Written Agreement, it was put to Mr Beighton that the rationale for the Aesica production was as a back-up in case the supply from Auden failed. He denied that, and said that he never stopped pushing the Aesica product forward, despite documents put to him that seemed to show the Aesica product being regarded by some in AMCo as a back-up. Cross-examination followed about the effect that the possibility of Cinven acquiring Auden had on the negotiations for the written supply agreements. After several documents written by others were put to Mr Beighton, there were the following exchanges:

The sub-text to this cross-examination is, in my judgment, crystal clear. The CMA was saying that Auden had agreed to 12,000 boxes because of AMCo’s threat to enter, and agreement to stay out of, the market. Mr Beighton would not admit that to be the case. But he did admit that he had “a strategy of bluffing [Auden] that we were ready to launch our own product and we were hoping that he would respond by giving us product”.

The second day of cross-examination began with the Second Written Agreement. Mr Beighton admitted meeting (including for lunch) and speaking with Mr Amit Patel on the telephone between April and June 2014. He said it was not his style to make notes of such conversations.

When cross-examined about his 19 April 2014 email (see [26] above), Mr Beighton said this:

When cross-examined on his 28 May 2014 email to Mr Amit Patel (see [27] above), Mr Beighton agreed that he was going to Auden to ask for 12,000 packs, because he wanted “at least to be able to sell the same number that [he] could sell with [his] own product”. Mr Beighton agreed that that meant that Auden was “foregoing the vast majority of its margin on 15% of the total market”, which was £7.2 million over the course of a year.

After some more documents were put, Mr Beighton agreed that the premise for the negotiation was that, if Auden agreed to supply AMCo with 12,000 boxes, then AMCo would not enter the market with its own product. His qualification was “but I am not saying that to him”. This was confirmed in a lengthy section of cross-examination that followed, where it was put to Mr Beighton that he bluffed Mr Amit Patel into thinking that, if he did not agree, AMCo would contest the whole market for 10mg hydrocortisone tablets. Mr Beighton effectively agreed.

Mr Beighton was then expressly cross-examined on his email of 15 June 2014 (see [29] above), where he had said that he told Mr Amit Patel that, if Auden did not supply 10,000 or 12,000 boxes, AMCo would launch its own product. Mr Beighton agreed that it looked like that was what he had said. Later he admitted that he “used the fact of having [his] own product as leverage in the negotiations” … “to get the deal over the line” and he was “keen to convey the impression to Mr [Amit] Patel that [he was] ready to launch”.

In relation to another document, Mr Beighton accepted that he was trying to convey the impression to Mr Amit Patel that AMCo’s critical card was its ability to launch its own product “because [otherwise] there was no way on earth that there was any reason for him to this deal”. It was at this stage that the Tribunal put its own questions to Mr Beighton.

In answer to the Tribunal, Mr Beighton repeated his thesis that he would have made more money by entering a market of 77,000 boxes and getting half of it than buying even 12,000 boxes from Auden. Mr Beighton accepted that if there had been a new entrant, there would have been competition resulting in lower prices. He suggested 10-15% lower, but accepted it could have gone down to just above cost, though that was less likely with only two competitors.

After some questioning about illegal agreements, there was an exchange between the President and Mr Beighton, where he accepted that the 12,000 box deal might have made sense from Auden’s point of view, as follows:

I have highlighted above the passage where Mr Beighton emphasises that there was not an agreement as alleged by the CMA. Nothing in this questioning gives me the impression that Mr Beighton was in any doubt about why he was being asked these questions and precisely what fell for decision. I will revert to that point. Despite all that, Mr Beighton never stopped reiterating that AMCo wanted to bring its product to market. The question may, perhaps, resolve into whether it would have helped if he had been asked on each occasion he said these things, whether he was lying. Interestingly, he was expressly asked at the end of the President’s questioning whether the obstacles to entry were fabricated. That was something that he, not surprisingly, denied.

After the Tribunal’s questioning, Ms Demetriou asked Mr Beighton about AMCo’s relationship with Aesica. He denied suspending the Aesica project, even though its product was not marketed when delivered in August 2014. Mr Beighton was asked about the emails dated 25 June, 27 June and 28 June 2014 (see [30]-[33] above). Mr Beighton said he did not remember seeing them and that his memory was that AMCo continued to work on the Aesica project. It could, he said, have been a short-term cancellation. He said that he recalled that AMCo had made a definite decision to prefer the Auden product over the Aesica product. AMCo decided not to launch because it could not launch. In relation to his own 28 June 2014 email to staff, he accepted that he was recognising that AMCo had chosen the Auden supply agreement over the launch of its own product. He accepted that he had been saying that it was easier commercially to sell the full label Auden product than AMCo’s own skinny label product. Finally, he accepted that, having AMCo’s own product, helped the negotiations with Auden. But Mr Beighton never accepted that he had dropped the idea of AMCo’s own launch completely.

Mr Beighton was then asked about the 3-month notice clause 2.2 in the Second Written Agreement. He accepted that Mr Amit Patel would have terminated it, had notice been given, because there would have been no benefit left for Auden. The cross-examination concluded with a lengthy series of questions about the market for the skinny label product. In the course of those questions, Mr Beighton accepted that the orphan label indication: “certainly made it more interesting for us to do a deal with - for a smaller number of packs than 40,000 with Auden”. Having accepted that there was a clear risk that AMCo would not be able to match 12,000 packs a month if it entered the market with its own product, it was put to Mr Beighton that: “the way for both sides to avoid each risk was to take the supply from Auden and not to enter independently”. He said that: “it was good for [Mr Amit Patel] and it was good for me because I was able to use this as a [stop gap] at a later date to launch our own product”. His understanding was that it was both a commercial and reputational risk for AMCo to have competed in the market with the skinny label product. Mr Beighton was then asked about his concern in January 2015 that the deal with Auden might be brought to an end when Actavis acquired it (Auden). Mr Beighton accepted that he seemed to have had a conversation with Mr Amit Patel, who told him that Actavis would continue supply on the same terms.

Mr Beighton agreed at the end of his cross-examination that, in October 2012, the market price for 10mg hydrocortisone tablets was £35 per box. It increased to £72 per box, and then started dropping once the skinny label entrants came on the market.

Following that lengthy summary of Mr Beighton’s cross-examination, I return to the question of whether the CMA’s case was properly put to him. It is plain that the CMA did not expressly suggest to Mr Beighton either that he had behaved dishonestly or that he was lying on oath. But it is equally plain that dishonesty was not a component of the case that the CMA was advancing (see [42]-[50] above), nor of the case that is enunciated in the Decision.

The central question, therefore, for this court is whether the Tribunal was right to conclude in H3 that the failure to put dishonesty, lying and to challenge Mr Beighton’s rationale truly undermined the CMA’s case fatally or at all. As it seems to me, it is necessary to deal specifically with each of the three aspects of the CMA’s case which I have suggested the Tribunal “struggled with” at [47] above. To recap, those are: (i) that the CMA’s case involved an element of dishonesty or naughtiness; (ii) that the CMA had used the term “sham” to describe the written agreements in its traditional sense of being agreements that were entered into to conceal the truth of what had been agreed, and that the finding of “intention” also involved misconduct of some kind; (iii) that Mr Beighton’s evidence that there was no collateral understanding, that it was always in AMCo’s interests to enter the market, and that the Second Written Agreement was no more than a stop gap was not properly challenged, and the Tribunal did not know what he might have said had he been so challenged. I shall deal with these three points in reverse order.

The first point to repeat is that Mr Beighton was a sophisticated witness, who repeatedly made clear that he fully understood the CMA’s case that the 10mg Agreement included an understanding that AMCo would stay out of the market. Indeed, he did all he could to reiterate his denial of such an understanding wherever he could in his witness statement and his evidence.

Secondly, the cross-examination recorded above at [54], [57], [60]-[69], and [71]-[74] put clearly to Mr Beighton that his understanding with Mr Amit Patel was that AMCo would stay out of the market as the price of the advantageous supply agreement, that AMCo was not going to enter the market whilst that supply arrangement was in place, and that the Second Written Agreement was not a stop gap, but was entered into on the understanding that AMCo was not going to enter the market. Mr Beighton agreed at [63] and [67] that he had bluffed Auden that AMCo was ready to launch its own product, and that Auden had responded by giving AMCo the product at a very advantageous price. This “bluff” was only a bluff because, as Mr Beighton also accepted, the Aesica product was (a) not ready, and (b) was unlikely to allow AMCo to contest for 50% of the market, because of the skinny label problem (contrast [60]-[61] above with [73]-[74] above).

Thirdly, in my judgment, every aspect of the Tribunal’s reasoning in [145]-[146] of H3 is contestable and seems to be based on a flawed recollection of Mr Beighton’s cross-examination, some 15 months earlier.

Mr Beighton’s so-called “positive case” that it was in his interests to enter the market the moment he could was indeed challenged, and he accepted, as I have just said, that the 12,000 boxes that Auden agreed to supply reflected the reality of the risks that AMCo posed to Auden. It was repeatedly put to Mr Beighton that the deal was on the basis that AMCo would not enter. It is true that he often said that he never gave up the idea of doing so. But that was simply an answer that the Tribunal had to evaluate against the other available evidence and inferences.

Also, it was wrong for the Tribunal to suggest that the other aspects of Mr Beighton’s positive case were not challenged, and that this undermined the CMA’s case. The Tribunal accepted at [145(4)(iii)] that it had no problem with the inferences both it (in H2) and the CMA (in the Decision) had drawn against Waymade and Auden. The Tribunal’s problem arose from the fact that AMCo had advanced a positive case before the Tribunal, and from the fact that H2 had found at [144(8)] that there was a collateral understanding between Auden and AMCo. That collateral understanding, to the effect that AMCo would stay out of the market in return for the supply of 12,000 boxes per month at cost, was, according to the Tribunal, not put to Mr Beighton (see [145(4)(iv) and (v)] in H3). The Tribunal thought at [145(4)(vi)] that Mr Beighton never gave his answer as to why Auden might have agreed to such a course, because he was not asked. That was the Tribunal’s fundamental error. It was put to Mr Beighton that Auden had agreed to supply 12,000 boxes on these terms, because AMCo agreed to stay out of the market (see the cross-examination summarised at [62], [63], [66], [68], [71] and [74] above). I shall deal with other aspects of Mr Beighton’s allegedly positive case in section H of this judgment.

In addition, the point about the Second Written Agreement being a stop gap fell to be evaluated alongside the more important evidence about whether or not the understanding was indeed that AMCo would not enter the market.

Finally, it is worth mentioning the Tribunal’s concern about evidence that the understanding crossed the line. [108]-[109] of H3 suggests lines of cross-examination that ought to have been put to Mr Beighton, including for example: “what could possibly have impelled Auden to give more product at this price to AMCo?”. These questions seem to ignore, as I have already said, that the CMA repeatedly put to Mr Beighton that the only reason for the supplies on such advantageous terms was AMCo’s agreement to stay out of the market. That was the whole thrust of the cross-examination about Mr Beighton having bluffed Mr Amit Patel into thinking that, if Auden did not agree to supply product on advantageous terms, AMCo would contest the whole market for 10mg hydrocortisone tablets. It was a bluff, because Mr Beighton knew he could not do so. But Mr Amit Patel only agreed to supply on the basis that AMCo would not carry out its threat to enter the market.

I conclude that, in my judgment, the Tribunal was mistaken to think that Mr Beighton’s evidence was not challenged when he said that: (i) there was no collateral understanding, (ii) it was always in AMCo’s interests to enter the market, and (iii) the Second Written Agreement was a stop gap. The stop gap point was part and parcel of whether there was, in reality, an understanding that AMCo would not enter the market.

This point can be readily answered by reference to [6.922] in the Decision where the CMA explained that its “description of the supply deals as a sham simply [meant] that the CMA has found their true purpose to be for Auden/Actavis to pay AMCo, rather than simply to give it product to sell as in a genuine bona fide distribution deal”. The Tribunal suggested at [15(5)] in H2 that the CMA could not disavow the true meaning of “sham”, because that is what the CMA had actually found in the Decision. It seems to me that that was a misunderstanding. As I have already explained, the CMA’s case was clear throughout. It was not that the Second Written Agreement was entered into in order to conceal the truth. It was simply that there was an understanding between Waymade, and then AMCo, on the one hand and Auden, on the other, that, whilst Auden supplied product to AMCo on advantageous terms, AMCo would not enter the market.

There is nothing in the Decision to suggest that the CMA was alleging that AMCo’s, and later, Advanz’s intentional infringement of the Chapter I prohibition implied that they had been guilty of dishonesty or misconduct. This conclusion follows from the answer that I shall give in the next sub-section.

The Tribunal expressed the view at [151] in H2 that it was impossible in this case to find an intentional infringement of the Chapter I prohibition without making an implicit finding of dishonesty. I disagree. If the Tribunal were right, it would complicate many, if not most, competition cases. The Chapter I prohibition imposes strict liability. It prohibits certain anti-competitive arrangements and agreements, whatever the motives for them. Some may be motivated by ignorance of the law. Others may come about by the actions of disparate actors within an undertaking. It would make the CMA’s task impossible if it were required to put or prove dishonesty, even where that state of mind might plausibly be implied. The legislation does not require it and nor, therefore, should the Tribunal. It is true, as the Tribunal held, that a nod or a wink or an understanding must “cross the line” between the parties to an anti-competitive arrangement. It is true also that the subjective intentions of the parties may be relevant to whether or not, objectively viewed, an infringement of the Chapter I prohibition is to be inferred. None of that means, in my judgment, that the CMA is obliged to allege dishonesty in any particular case.

I do not find much assistance from the Tribunal’s use of the term “naughtiness”. If it simply means misconduct, then that term would be better used. The term misconduct may be a kinder one than dishonesty, but my preference would be to be clear as to what is being alleged. If dishonesty is not alleged, then the fact that an anti-competitive arrangement may properly be characterised as some kind of misconduct does not, in my view, assist the analysis. An intentional infringement attracts quasi-criminal penalties. That should be enough for undertakings to understand that the conduct in question is unlawful.

Accordingly, I would endorse the CMA’s consistent approach to this case. It had no need to allege or prove the dishonesty of any particular actor within the undertakings in question, including Mr Beighton. The consequences of this reality are dealt with in the subsequent sections of this judgment.

To be clear, I disagree with the Tribunal’s perspective in [38], [122] and [131]-[134] in H3 that dishonesty had to be put to Mr Beighton. The above summary of how Ms Demetriou cross-examined Mr Beighton demonstrates precisely how a case of anti-competitive conduct can and should be put to a recalcitrant witness without needing to go so far as putting that they were dishonest. I disagree with the Tribunal that there was any real doubt about what Mr Beighton might have said, had he been specifically challenged about whether his answers were true. He stuck very closely to the story he was telling in his witness statement, save where the cross-examination forced him to admit inconsistencies. As I have said, he was a sophisticated witness who was well aware of the consequences of what he was saying.

I will deal with the question of whether or not it was necessary expressly to suggest to witnesses that they were lying in section H of this judgment.

Section 46(1) of the 1998 Act provides that a “party to an agreement in respect of which the CMA has made a decision may appeal to the Tribunal against, or with respect to, the decision”. Such decisions include one as to whether the Chapter I prohibition has been infringed. Paragraph 3(2) of schedule 8 to the 1998 Act allows the Tribunal to confirm or set aside the decision, remit the matter to the CMA, give such directions, or take such other steps, as the CMA could itself have given or taken, or make any other decision which the CMA could itself have made.

Paragraph 7.78 of the Tribunal’s “Guide to Proceedings” 2015 states as follows:

It is against this background that this court needs to consider whether, as the CMA submits, the Tribunal adopted an inappropriate procedure when it released H2 to the parties, but ordered a further hearing to consider whether the CMA’s case had been fully put to the companies’ witnesses.

In my judgment, that procedure was inappropriate in all the circumstances of this case. The process of deciding, on the evidence, whether each of the companies’ appeals from the Decision should succeed was a unitary one. It was the Tribunal’s duty to hear the evidence that was called, to listen to the parties’ submissions, having made appropriate case management directions, before deciding whether each of the appeals fell to be dismissed or allowed or whether some other order ought to have been made in respect of them under paragraph 3(2) of schedule 8 to the 1998 Act.

In this case, the veracity of the witnesses and the explanations they gave for their conduct, looked at against the backdrop of the contemporaneous and other evidence, was intimately bound up with the question of whether each of the companies had infringed the Chapter I prohibition. It was not appropriate to order a further hearing whilst also stating unequivocal conclusions, epitomised by the finding at [156] of H2 that the 10 mg Agreement was a “by object infringement of the Chapter I prohibition”, whose “object was flagrantly anti-competitive and the anti-competitive effects significant, in that an abused monopoly position was maintained and supported”. It was necessary for the Tribunal to decide whether its concerns about the way the CMA’s case had been put affected those conclusions, before stating them.

The process that the Tribunal adopted was also unjust, because it allowed the parties a second bite at the cherry of argument, when they thought they had addressed the Tribunal on all relevant points. In the result, as I have already pointed out, the two days of further argument expanded from a concern about whether the CMA’s case had been fully put, and as to the implications of the CMA’s case not having been fully put (see [4] and [5] above), into fundamental questions as to the essential nature of the CMA’s case as to the 10mg Agreement, including detailed points about dishonesty. It was particularly remarkable that such argument could have occurred following a finding by the Tribunal, whether expressed to be provisional or not, that Mr Beighton was dishonest and had lied in the witness box ([153(2)(iii)(a)] of H2). The situation was compounded by the fact that, on any analysis, that had never been the CMA’s case.

If the Tribunal had held residual and legitimate concerns about whether the CMA’s case had been properly put or even about whether dishonesty had been cross-examined upon, after the conclusion of the argument, it had two clear choices. First, it could have written to the parties seeking further submissions either orally or in writing on the point. Conceivably, such submissions could even have resulted in witnesses being recalled for further questioning. But it was not appropriate to deliver a decision, provisional or otherwise, before raising and resolving those concerns. Secondly, it could have decided the matter on the basis of its view as to whether there had been a fatal flaw in the presentation of the CMA’s case. Presumably, in such a situation, the Tribunal might have decided to allow the companies’ appeals if it had formed the view that there had been such a flaw, or to dismiss them if it had formed the view that there had been no such flaw. On either analysis, it would have been open to the losing parties to seek permission to appeal to this court on the basis of the Tribunal’s alleged error of law. It was not open to the Tribunal to deliver a provisional decision, re-open the argument on fundamental points, and then to reverse its own decision.

The procedural error made by the Tribunal has left this court in a difficult position. I will return to how the matter should now be resolved, after I have dealt with the companies’ applications for permission to appeal the Tribunal’s reasoning in H2.

The points raised by the companies in their Appellants’ Notices challenging the provisional decision in H2 can be summarised as follows (see [13] above). H2 did not deal properly with, and the CMA did not properly challenge, Mr Beighton’s evidence that: (i) AMCo always intended to enter the market to obtain 50% of it, (ii) the Second Written Agreement was a stop gap whilst AMCo readied itself to enter the market, (iii) he had continued to try to source 10mg hydrocortisone tablets from Aesica even after the Second Written Agreement, (iv) he had taken legal advice as to the commercial terms agreed in writing, and there was no broader or different understanding beyond the written terms, and (v) he had not cancelled AMCo’s production of 10mg hydrocortisone tablets from Aesica because of any unwritten collateral understanding that he would not exercise his written clause 2.2 right to enter the market at all during the 2-year term of the Second Written Agreement. The sixth point was that the CMA had not properly cross-examined Mr Beighton as to alleged dishonesty and serious misconduct. The seventh point was as to the CMA’s failure to challenge whether Auden was unilaterally incentivising AMCo not to enter the market. The eighth point made by the companies was that the collateral understanding could not anyway have persisted beyond the time when Mr Amit Patel left Auden on 29 May 2015, and Actavis and Allergan took over its business. Advanz also challenged the Tribunal’s finding of a “by object” infringement and its evaluation of the market, suggesting that the 10mg skinny label product was not a competitor to a 10mg full label product, and that they were not substitute products.

In my judgment, none of these points, nor any of the different ways of putting them in the companies’ applications for permission to appeal H2 (or in their various respondents’ notices), raises any arguable point of law, or has any real prospect of success. As I have explained in section F of this judgment, the CMA’s case was fairly put to Mr Beighton.

On AMCo’s ability and intention to enter the market in order to acquire 50% of it, Mr Beighton ultimately accepted that the 12,000 boxes reflected the commercial realities of AMCo’s potential to enter the market (see [60]-[63], [65]-[71] and [74] above). It was simply not the case, as Advanz submitted, that the cross-examination omitted to challenge Mr Beighton on his evidence that he was always intending to enter the market and gain 50% of the market share. The evidence was far more nuanced than that. It was certainly a reasonable interpretation of the outcome of the questioning that I have summarised that Mr Beighton accepted that, when he entered into the First and Second Written Agreements, AMCo was not in a position to enter the market to gain 50% of it. He was seeking to bluff Mr Amit Patel into thinking that AMCo could, so as to get the maximum possible supply of product at cost price. The entire cross-examination was premised on the basis that the deals that AMCo and Mr Beighton reached were on the tacit understanding that AMCo would not, whether or not because it could not, enter the market whilst the advantageous supply terms continued.

The companies’ second point was that Mr Beighton’s evidence that the Second Written Agreement was a stop gap, whilst AMCo readied itself to enter the market, was not properly challenged. It was also put the other way round, so as to regard the Aesica product as a back-up in case the Auden supply failed. Whichever way one views Mr Beighton’s evidence, I have explained in detail why this was a bad point at [78]-[85] above.

The third point was that Mr Beighton had not been challenged on his evidence that he had continued to try to source 10mg hydrocortisone tablets from Aesica even after the Second Written Agreement. There was indeed evidence to that effect in some of the documents. I do not, however, see how that fact affects the thrust of the cross-examination that I have summarised. Mr Beighton was questioned about his approach to the Aesica product as described at [62], [63], [65], [73] and [74] above. The fact that AMCo may have been keeping Aesica’s skinny label product on the boil does not, in any way, depreciate the case that the CMA put, namely that the deal that AMCo made with Auden was on the basis that AMCo would not enter the market whilst the supply continued.

The fourth point concerned legal advice given by Pinsent Masons to AMCo about the legality of the commercial terms agreed in the Second Written Agreement. It is suggested that Mr Beighton was not challenged on whether there really was a broader or different understanding between AMCo and Auden beyond what Pinsent Masons had drafted. In the light of what I have already said, this argument adds nothing to the companies’ case. Either the written terms were all that was agreed or there was an additional tacit understanding. The latter suggestion was what Mr Beighton understood the CMA’s case to be, and was, as I have explained, fully put to him. It was not necessary to explore whether Pinsent Masons did or did not understand the full extent of what had been tacitly agreed.

The companies’ fifth criticism of the way Mr Beighton was cross-examined is another way of putting the same point. Again, either there was a tacit understanding that clause 2.2 would not be exercised or there was not. Mr Beighton understood the CMA’s case to be that he had tacitly agreed with Mr Amit Patel that he would continue to buy from Auden without entering the market whilst supplies continued on the same terms. Mr Beighton’s own statement at [115] puts the point very well: “the idea that somehow [we] would or could have colluded with Auden to take some discounted product to stay out of the market instead of launching a properly competitive product of our own would just not have been feasible”. It was properly put to him that that is what had actually happened. Clause 2.2 was not an absolute answer to the existence of a tacit understanding. It was just one piece in the factual jigsaw. As explained in [73] above, Mr Beighton was challenged about the cancellation of the Aesica product.

The companies’ sixth point was as to whether Mr Beighton had been properly or adequately challenged on dishonesty and lying. I have already explained at [88]-[91] above why I do not think it was incumbent on the CMA to cross-examine Mr Beighton on the basis that he had behaved dishonestly and was guilty of misconduct. For related reasons, I do not think that the CMA was obliged to cross-examine Mr Beighton expressly on the basis that he was lying in the witness box (as the Tribunal found he was at [153(2)(iii)(a)] in H2). On one analysis, Mr Beighton agreed with much of what was put to him. He obviously did not admit the tacit understanding, but, as I have said, he understood that that was the case against him. Nothing would have been gained by asking whether he was lying. He would have almost certainly denied that he was, and repeated his nuanced answers to the questions. As I have also explained, the Tribunal was, I think, wrong to suppose that there were any answers he might have given that would have extended its understanding. I have considered the Tribunal’s detailed analysis in H3 and proposed hypothetical questions at [88] in H3. With respect to the Tribunal’s obviously painstaking consideration, I think the H3 judgment is significantly overthought. It is true that Mr Beighton put forward a carefully tailored analysis leading to the conclusion that he had had no tacit understanding with Mr Amit Patel, which had crossed the line. But it is equally true that Mr Beighton understood full well the CMA’s case that he had reached such a tacit understanding with Mr Amit Patel, and that the only rational economic explanation for the value transfer that had occurred was that he had. Any number of accusations of lying would not have bridged that gap. The Tribunal had to make up its mind on the evidence.

Cross-examination is allowed, even required, in order to ensure that all sides to contested litigation have the opportunity to challenge evidence that they dispute, and to explain the parts of their own evidence that is disputed. The process should not become over-complex or a trap for the unwary.

This point is unarguable in the light of the inferences already reasonably drawn by the CMA in the Decision and the Tribunal in H2. As my summary of Mr Beighton’s cross-examination in section F above shows, there was ample material, both documentary and evidential, from which it could properly be inferred that there was indeed a tacit understanding between Auden and AMCo that, in return for the advantageous supply arrangements, AMCo was staying out of the market.

The companies’ eighth point challenged the Tribunal’s conclusion at [103(2)] and [136]-[138] in H2 that the infringement of the Chapter I prohibition can have continued after Mr Amit Patel left Auden on 29 May 2015 and Actavis/Allergan took over Auden. The companies’ central point is that no common understanding can survive when one of the individuals who was party to it leaves the undertaking in question, and (as the Tribunal found at [137(2)]) no “later acquirer of Auden [could] be criticised for failing to appreciate that there was an improper agreement between Auden and AMCo”. The companies submit that, as a matter of law, a common understanding can only persist, for the purposes of competition law, where there is a continued meeting of minds. I do not agree. Evidence of the subjective state of mind of parties to an anti-competitive agreement is not required. An anti-competitive agreement can be inferred from all the circumstances and the evidence. In this case, there was abundant evidence from which the continuation of the existing anti-competitive arrangement could be inferred, without proof that the directors of Actavis knew the details of how it had been reached before they bought the supplier.

First, Mr Beighton accepted in cross-examination that it seemed from the documents that he had had a conversation with Mr Amit Patel in which he had been told that Actavis would “continue with supply on the same terms”. Secondly, the hugely favourable supply terms continued unchanged after the Actavis takeover and Mr Amit Patel left Auden. Indeed, the market price of 10mg hydrocortisone tablets peaked in early 2016 at £72 per box, yet Actavis was still selling the product to AMCo at £1.78 per box. Thirdly, of course, AMCo did indeed stay off the market. The Tribunal was right to point out that: “[t]he whole point of the undertaking as a “unit of account” in competition law is that liability … operates at the level of the undertaking”. I would endorse the Tribunal’s factual approach when it said at [138]:

The final point raised by Advanz relates to the connected questions of: (a) whether there can be a “by object” infringement, and (b) whether the skinny label and full label products were true substitutes competing in the same market, when the customers for the two products were potentially or actually different.

The problem with both these arguments is that they have been determined on the facts against the companies. Even if they can be said to raise potential legal arguments, that does not make the challenge to the Tribunal’s clear findings arguable. At [62] and footnote 94 in H2, the Tribunal decided that the two products were substitutes, as had been decided in H1, and as was supported by evidence of the kind recited at [128] and [129] (and [130]) in H2. It is not clear what facts, as found, the companies wish to challenge. Secondly, and perhaps even more importantly, if there were in fact an understanding between Auden and AMCo that AMCo would keep its skinny label product off the market, there was obviously a “by object” infringement. As the General Court said in Toshiba v. European Commission (Case T-519/09) 21 May 2014 at [231]: “the very existence of the Gentlemen’s Agreement provides a strong indication that a competitive relationship existed”, and “it is unlikely that they would have entered into a market-sharing agreement if they had not considered themselves to be at least potential competitors”. Thirdly, as H2 found, there was a downward competitive effect on prices when the skinny label product did actually enter the market. At [115(3)(v)] of H2, the Tribunal said: “[o]ver time, because “skinny label” was priced at below “full label”, and because multiple entrants competed against each other, an inevitable downward pressure on both “skinny label” and “full label” prices manifested itself”. The cross-examination that I have summarised in section F above makes all that clear.

In my judgment, Advanz’s attempt to appeal the object infringement and market definition points is hopeless.

I would not, therefore, grant the companies permission to appeal H2 on any of the grounds that they advance. They raise no arguable point of law under section 49(1)(c) of the 1998 Act.

For the reasons I have given, the CMA’s appeal against the Tribunal’s decision in H3 should be allowed. That decision was summarised in [145] and [146] of H3 as follows:

The mistakes that the Tribunal made in H3 were, in outline, in thinking that the CMA’s case: (i) inevitably involved allegations of misconduct and/or dishonesty and/or written agreements that were dishonest shams, (ii) was unclear, (iii) had not been properly or fully put to the witnesses, and (iv) was fatally undermined by these alleged errors. The Tribunal also fell into error in deciding that the appropriate consequence of what it had found was that the companies’ appeals from the Decision had to be allowed.

It is now necessary to consider what outcomes follow from these determinations. It seems to me that those outcomes are dependent, at least in part, on the outcomes of the applications for permission to appeal H2 made by each of the companies and the CMA.

The CMA’s new Appellant’s Notice seeks to appeal the findings in [157]-[160] of H2 to the effect that the Tribunal was “not in a position finally to determine the [companies’] appeals and that a further substantive hearing was required”. The CMA alleges that the Tribunal in H2 wrongly said that: (i) the Decision expressly or impliedly found that the written supply agreements disguised the true nature of the 10mg Agreement (see [7(5)-(6)], [7(10)] and [12]), (ii) a consequence of the Decision was that human actors must have understood that the written agreements “hid the true purpose of the arrangement” ([24(4)]), (iii) the CMA’s use of the word “sham” and an allegation of intentional infringement inevitably implied dishonesty by humans including Mr Beighton ([24(1)-(3), [151], [152], and [153(2)]), (iv) it was necessary to find Mr Beighton personally dishonest to uphold its findings in H2, and (v) because allegations of sham, intention and dishonesty had not been put to Mr Beighton, it could not finally determine the appeals without a further hearing ([157] and [160]). The Tribunal was, in fact, fairly able to dismiss the companies’ appeals on the basis of the evidence, which it did at [26] and [156]. The CMA asked this court to confirm the Decision relating to the 10mg Agreement under paragraph 3(2) of schedule 8 to the 1998 Act.

For the reasons I have given, the CMA must be granted permission to appeal H2. It seems to me that the CMA’s grounds of appeal succeed. The Tribunal was in a position, after the argument in H2, finally to determine the companies’ appeals from the Decision. It should have done so by dismissing the appeals for the reasons it gave, excluding those that concerned dishonest shams, dishonest intentions, the dishonesty of parties in general and Mr Beighton in particular, and the precise inferred conversation found to have occurred between Mr Amit Patel and Mr Beighton at [144(8)] of H2. The Tribunal was wrong to deliver a provisional judgment and, having done so, to adjourn for a further “due process” hearing. This court should confirm the Decision relating to the 10mg Agreement under paragraph 3(2) of schedule 8 to the 1998 Act. There is no need for a re-hearing of the companies’ appeals, because the Tribunal’s factual and legal determinations were clear, save for the procedural errors I have already explained.

That leaves the companies’ applications for permission to appeal the Tribunal’s provisional decisions in H2, which for the reasons given in section H above must be dismissed.

In conclusion, I would make it clear why there was no need for the Tribunal to make an inference of a specific conversation with specific content between Mr Amit Patel and Mr Beighton (see [144(8)] of H2). The evidence that I have summarised in this judgment and that is summarised in H2 itself allowed the Tribunal (and allows this court) to draw the clear inference that there was a tacit understanding between Auden and AMCo, both before, at the time of and after the Second Written Agreement that, whilst Auden continued to supply AMCo with 10mg hydrocortisone tablets at highly advantageous prices, AMCo would not enter the market with its own product. Moreover, the CMA’s case was sufficient to establish an infringement of the Chapter I prohibition, without the need to find that there had been an inferred conversation between specific actors at Auden and AMCo in which it had been expressly agreed that, in addition to AMCo staying out of the market, it would “take Auden’s product and sell it at around the prevailing market price (as set by Auden)”.

Accordingly, the CMA’s substantive appeal against the Tribunal’s decision in H3 will be allowed. The CMA will be granted permission to appeal from the Tribunal’s decision in H2, and its appeal will be allowed. The companies’ applications for permission to appeal the Tribunal’s decision in H2 will be dismissed. There will be an order, in substitution for the decisions in H2 and H3, that will be made by this court finally dismissing the companies’ appeals from the CMA’s Decision under section 15(3) of the Senior Courts Act 1981 and paragraph 3(2) of schedule 8 to the 1998 Act.

I agree.

I also agree.