PHONES 4U LIMITED (In Administration) v EE LIMITED

On 10 November 2023 I issued judgment on preliminary issues in this action: [2023] EWHC 2826 (Ch) (“the Judgment”). The effect of the Judgment is that the claims against all the defendants are dismissed. A further hearing to address various consequential matters was held on 20 December 2023. In that hearing, I refused the Claimant’s application for permission to appeal and stated that I would give my reasons in writing. All other matters were determined at that hearing save for the question of the rates of interest to be applied to the costs of the Second and Third Defendants which were paid in euros, on which I reserved my decision. This Ruling addresses those two matters.

This Ruling uses the same abbreviations for the parties as the Judgment and all paragraph references are to the Judgment except as expressly stated.

The question of interest on costs arises as regards two distinct periods:

The power to award interest on costs prior to judgment is set out in CPR rule 44.2(g). The rate of such interest is a matter for the Court’s discretion. As Sharp LJ stated (with the agreement of Patten and Gloster LJJ) in Secretary of State for Energy and Climate Change v Jones [2014] EWCA Civ 363 at [17]:

“Ultimately, the court conducts a general appraisal of the position having regard to what is reasonable for both the paying and the receiving parties. This normally involves an assessment of what is reasonable having regard to the class of litigant to which the relevant party belongs, rather than a minute assessment which it would be inconvenient and disproportionate to undertake. In commercial cases the rate of interest is usually set by reference to the short-term cost of unsecured borrowing for the relevant class of litigant, though it is always possible for a party to displace a 'rule of thumb' by adducing evidence, and the rate charged to a recipient who has actually borrowed money may be relevant but is not determinative. See F & C Alternative Investments Ltd v Barthelemy (No 3) CA [2013] 1 WLR at paragraphs 98, 99 and 102 to 105; Bim Kemi AB v Blackburn Chemicals Ltd [2003] EWCA Civ 889 at 18 and for example, Fiona Trust & Holding Corporation v Privalov [2011] EWHC 664 (Comm).”

Generally, for a party in the UK that paid costs in sterling, the rate applied as representing a commercial rate is the Bank of England base rate plus a percentage determined according to the class of borrower to which the receiving party belongs (e.g. whether it is to be regarded as a “first class” borrower): see Note 44.2.29 to Civil Procedure, Vol 1 (the White Book).

However, where the party conducting the litigation is based overseas and incurred the costs in a foreign currency, the rate relevant to borrowing sterling is not appropriate. Most of the cases involving foreign currency concern the rate of interest on an award of damages, but the applicable principles are the same: Sec of State v Jones at [17]. As the Court of Appeal said in The “Pacific Coloctronis”[1981] 2 Lloyd’s Rep 40 at 46, addressing interest on a judgment in US dollars, prima facie the rate of interest applicable should be related to the currency of the judgment. Accordingly, where costs have been incurred in dollars by a litigant operating in the United States, the courts have generally awarded the US Prime Rate as the rate at which the receiving party should reasonably have borrowed: see e.g. Kuwait Airways Corp v Kuwait Insurance Co SAK [2001] 1 Lloyd’s Rep IR 678; Mamidoil-Jetoil Greek Petroleum Co SA v Okta Crude Oil Refinery AD [2003] 1 Lloyd’s Rep 42; Certain Underwriters at Lloyd’s London v Syrian Arab Republic [2018] EWHC 385 (Comm). (Footnote: ¹)

I was told that Counsel had not found any authority which considered the appropriate rate where the relevant currency is the euro. However, in Slocom Trading Ltd v Tatik Inc [2013] EWHC 1201 (Ch) (Footnote: ²), where judgment was given in euros, the Court awarded pre-judgment interest at 1% over the European Central Bank (“ECB”) rate. The ECB rates were indeed accepted by both P4u and Orange as the relevant basis for determination of interest as regards the costs of Orange. For DT, it was submitted that it should have the same rate of interest as all the defendants that incurred costs in sterling. However, although the fees charged to DT were calculated in sterling, they were converted into euros for each invoice, which DT then paid in euros. In my view, DT is in no different position from Orange and an appropriate rate for borrowing in euros should apply to it also.

However, the ECB had several different rates over the relevant period. Orange submitted that the relevant ECB rate is the interest rate on main refinancing operations (the “MRO rate”) which is the rate applicable to most of the ECB lending to banks. P4u’s skeleton argument contended that interest should be awarded at a percentage point over the ECB deposit facility rate, which is the rate which banks may use to make overnight deposits with the Eurosystem. The MRO rate was generally about 0.5% above the deposit facility rate. Although the judgment in Slocom v Tatik does not expressly state which ECB rate was used there, it seems clear from the figures set out at para [42] of that judgment that it was the MRO rate. (Footnote: ³) The MRO rate is a rate for borrowing, which is the criterion to be applied, and I consider that it is accordingly the relevant rate. Moreover, for the early part of the period the deposit facility rate was negative, which in my view confirms that it is not an appropriate benchmark rate for this purpose.

Mr MacLean KC in his oral submissions made clear that P4u did not object to application of the MRO rate, but that if the MRO rate was selected then there should be no uplift since it was a borrowing rate. By contrast, Orange argued for the MRO rate plus 2%. Mr Scannell KC submitted that since the MRO rate was the rate at which commercial banks could borrow from the ECB, the rate which banks would charge for lending to their customers would be higher. He also referred to Case T-201/17 Printeos SA v Commission, EU:T:2019:81. There, following the annulment of the EU Commission decision fining Printeos for infringement of competition law, the Commission repaid the fine but declined to pay interest on the basis that the performance of the fund into which the money received had been invested had been negative. The General Court upheld Printeos’ claim that this was irrelevant and that it was entitled to interest as compensation “at a standard rate for the loss of use of [the amount of the fine it had paid]”: judgment at para 67. The Court therefore upheld Printeos’ claim for interest at the MRO rate plus 2%.

I accept that there should be an addition to the MRO rate for the primary reason Mr Scannell gave. However, I gain limited assistance from the Printeos case since there the judgment addressed the question whether Printeos was entitled to interest at all, and did not consider what rate was appropriate as the Commission did not challenge the rate claimed by Printeos. (Footnote: ⁴) Neither Orange nor DT submitted any evidence as to the actual rates at which they were able to borrow over the relevant period. Both clearly come into the category of first-class borrowers. I think it is therefore appropriate to apply a conventional rate. Although in Slocom v Tatik, that was determined as 1%, above the ECB rate, that concerned interest on damages not costs and in part took account of Slocom’s status as an investment not a trading company. Here, for Orange and DT, I consider that the appropriate rate should be the MRO rate plus 1.5%.

Where judgment is given in sterling, interest is at the statutory rate, determined pursuant to the Judgments Act 1838. This applies also to interest on an award of costs, which forms part of the judgment. The Administration of Judgments Act 1970 (“AJA 1970”) s.44 gave the Lord Chancellor power by statutory instrument to vary the rate and, since last varied in 1993, it has been 8%. At least until the recent increases in the base rate, the statutory rate has been very significantly above a commercial rate for many years.

AJA 1970 s 44A provides that where judgment is given for a sum expressed in a currency other that sterling, instead of the statutory rate

“… the court may order that the interest rate applicable to the debt shall be such rate as the court thinks fit.”

For DT, Mr Leith submitted that s. 44A had no application since there was no judgment in euros only an order for costs to be subject to detailed assessment. I think that is misconceived where this assessment will lead to determination of the amount to be recovered of costs paid in euros. In my view, the order now should be that DT, like Orange, is to be paid 95% of its costs, those costs to be determined in euros by detailed assessment on the standard basis, unless agreed. Accordingly, I consider that s. 44A applies to these awards of costs and the rate of interest is therefore at large.

In Novoship (UK) Ltd v Nikitin [2014] EWCA Civ 908, at [133]-[136], the Court of Appeal held that the legislation did not indicate any policy to set a penal rate or do other than reflect the compensatory principle that the primary purpose of an award of interest is to compensate the receiving party for being kept out of its money. The successive failures to revise the statutory rate to reflect the change in commercial rates were an anomaly, but that could not justify the court, in a case to which s.44A applied, choosing in its discretion to apply the statutory rate.

I recognise, therefore, that the compensatory principle is the relevant starting point. Nonetheless, I see force in Mr Scannell’s submission that in the present case, it would be unjust to have such a strong divergence between Orange (and DT), that are being awarded their costs in euros, and all the other defendants that are awarded costs in sterling. In my judgment, in these particular circumstances, the appropriate course is to exercise my discretion to increase the uplift on the MRO rate as regards post-judgment interest as compared to pre-judgment interest, and award post-judgment interest at the MRO rate plus 2.5%.

P4u’s Grounds of Appeal set out eight grounds, and those grounds were developed in its skeleton argument (the “PTA skeleton argument”) and oral submissions. Some of those grounds involve some basic misconceptions of the law and others involve significant distortions of the approach and some of the factual findings in the Judgment, so to explain why those grounds have no real chance of success it is necessary to address some of those grounds in a little more detail than would usually be the case when refusing permission to appeal.

These grounds concern what was referred to as the Landmark Lunch, i.e. the lunch meeting between Mr Dunne, the CEO of O2, and Mr Swantee, the CEO of EE, at the Landmark Hotel in London on 19 September 2012. The background to the lunch is explained at [148]-[151]. The evidence as to what was said at the lunch is set out, along with an account of surrounding events, at [154]-[190]. My findings as to what happened at the lunch are at [194]-[200].

The allegation of infringement by a concerted practice as between EE and O2 is set out in para 136 of P4u’s Re-Amended Particulars of Claim, quoted at [69]. That of course relates to the Landmark Lunch. The allegation is elaborated at paras 137-138.

The meaning of a concerted practice which infringes art 101(1) TFEU and the Chapter I prohibition under the CA 1998 is set out at [77]-[81]. It is common ground that such infringement involves three elements: (i) two or more undertakings concerting together; (ii) conduct on the market pursuant to those collusive practices; and (iii) a relationship of cause and effect between the two.

This ground concerns element (i). The way P4u put its case is summarised at [202]. Essentially, although not so clearly expressed in its pleading, P4u alleged that a concerted practice arose in two ways: (a) O2 disclosed commercially sensitive information to EE, which therefore reduced uncertainty in EE as to the future conduct of its competitor (see the quotation from Nichicon at [81]); and (b) O2 obtained comfort from EE that EE would not take up volumes from P4u that would become available if O2 withdrew from P4u, such that O2 thereby “de-risked” its plan to withdraw.

I accepted that to give rise to an infringement an exchange of information does not have to be reciprocal: [203]. But it was indeed P4u who submitted that: “It is a question of substance whether the information is capable of removing or reducing uncertainty and facilitating practical cooperation”: [205]. In short, this depends on an assessment of the facts as to whether what was disclosed, in the particular context, had that effect: cp HSBC (too vague disclosure to have such effect) and Tate & Lyle (sufficient disclosure to have that effect): [206]-[207]. Here, on the facts, I held that the disclosure was wholly vague and would not have had that effect: [208]. That is a factual evaluation within established principles of law and I do not consider that there is any real prospect of the Court of Appeal finding that it was erroneous.

Further, as regards the alleged “comfort” given to O2, it is trite to observe that a concerted practice requires an element of concertation: see the oft-repeated definition in Suiker Unie quoted at [78], and the approval by the Court of Appeal in Argos of the statement: “it is just as essential to a concerted practice as to an agreement that there be a consensus between the two or more undertakings said to be parties to the agreement or concerted practice”, quoted at [216]. Similarly, the Advocate General in Eturas stated: “… the concept of a concerted practice does imply reciprocity. A concerted action is necessarily the result of a consensus” (quoted at [215]).

I found on the facts that Mr Swantee remained passive during Mr Dunne’s vague disclosure of what O2 had in mind, and said nothing that would encourage O2. P4u does not allege that this finding was not open to the Court on the evidence. Indeed, so concerned was Mr Swantee about what Mr Dunne appeared to be saying that, when an opportunity arose to do so, he started to make a secret recording of the conversation. I fully accept that such silence does not, in and of itself, preclude approval or support. But as the Advocate General stated in Eturas (quoted at [215] and repeated here with emphasis added):

“48.

However, the possibility of inferring tacit approval, and therefore of establishing the existence of a consensus to cooperate rather than compete, depends on the context of the communication.

49.

First, where an undertaking receives information relating to an illicit initiative and does not oppose it, its acquiescence in that initiative may be inferred from the absence of response, provided that the circumstances are propitious to the formation of a tacit consensus. The lack of opposition to an illicit communication is reprehensible because, under certain circumstances, mere lack of reaction from the addressee will lead the other party or parties to believe that the addressee subscribes to the illicit initiative and will comply with it.  Therefore, in order to infer knowing participation of the addressee in a concerted practice, the context of interaction must be such that the addressee may be deemed to appreciate that the competitor will consider its silence as an approval and will rely on mutual action, even in the absence of response.”

In the specific context in which the Landmark Lunch took place (see at [151]) I found that Mr Dunne would not have inferred any encouragement or cooperation from Mr Swantee’s silence: [217]. Again, I think that is a factual finding on the evidence which was open to the trial judge and that there is no real prospect of the Court of Appeal finding that it was erroneous. P4u seek to cast doubt on that finding by reference to what I said at [225]. But there I was indicating that I did not think that Mr Dunne was given the impression that Mr Swantee was “objecting or wholly uninterested”. That is entirely consistent with the finding that he did not think that Mr Swantee was positively supporting the course of action O2 wished to pursue or that his silence implied an agreement to cooperate.

On that basis, element (i) of a concerted practice is not made out. Since grounds 2 and 3 concern the so-called Anic presumption which, it is common ground, addresses element (iii) of a concerted practice, if there is no real chance of success on ground 1, grounds 2-3 fall away.

Nonetheless, I address grounds 2-3, just as I proceeded to address the Anic presumption in the Judgment: [231]. It is convenient to address those grounds in reverse order.

This ground concerns EE. As regards disclosure by Mr Dunne of confidential information to Mr Swantee, I accepted that if (contrary to my earlier finding) that information was sufficient to remove uncertainty about O2’s future conduct, the presumption applied. (The terms of the presumption are set out at [79]). However, I held that EE had clearly rebutted the presumption on the facts: [231]. P4u submit that this discloses an error of law on the basis that, where the concerted practice arises from a meeting, as a matter of law the only way in which the presumption can be rebutted is by public distancing and/or report to the authorities. Here, I held that the subsequent conduct of EE did not meet the high test for public distancing: [218]-[230], and of course EE never reported the matter to the authorities.

This submission was not advanced in these terms at trial, and P4u now seeks to rely on further materials to support it. It is therefore necessary to address it more fully. The Anic presumption is an evidential presumption concerning a matter which the competition authority or claimant has to prove to establish an infringement. An undertaking found to infringe is potentially exposed to a very substantial fine, such that an infringement has been held to be of a quasi-criminal character. The suggestion that, as a matter of law, provided there was a meeting then the way in which the presumption can be rebutted is narrowly restricted would obviously be capable of serious injustice. It is also contrary to authority. Hence in T-Mobile, a case strongly relied on by P4u as showing that the Anic presumption applies even if the concerted practice is alleged to derive from a single meeting, the CJEU stated, as quoted at [80]:

“Where it can be established that such undertakings successfully concerted with one another and remained active on the market, they may justifiably be called on to adduce evidence that that concerted action did not have any effect on their conduct on the market in question.”

There was no ‘public distancing’ or report to the Dutch competition authority, but the CJEU’s ruling shows that this did not preclude the undertakings from seeking to rebut the presumption. Indeed, in Anic itself, after setting out the presumption at para 121, the CJEU proceeded to consider whether Anic had rebutted it, but although that was a case where the concerted practice was based on meetings there was no suggestion by the Court that the absence of public distancing or a report to the authorities was determinative or even relevant: see at paras 126-128. In its skeleton argument, P4u now seeks to rely on Case C-455/11P Solvay v Commission, EU:C:2013:769, which was not cited at trial. But that also is inconsistent with P4u’s submission. That was an appeal regarding the Commission’s decision on the hydrogen peroxide cartel, where competitors exchanged commercially confidential information in a series of meetings. Solvay sought to rebut the Anic presumption by showing that the information exchanged did not have any influence on its conduct since its prices of the products in question had decreased and the market had remained highly competitive. Unsurprisingly, the CJEU, upholding the General Court (Footnote: ⁵), held that this alone was not sufficient to displace the presumption, stating at para 44:

“In that regard, it must be stated that probative data illustrating the competitive nature of the market and, in particular, the decrease of prices during the period concerned cannot suffice, of itself, to rebut that presumption. That data does not of itself make it possible to prove that that undertaking did not take account of the information exchanged with its competitors in determining its conduct on the market. It follows that that data does not of itself preclude the presumption that the concerted action enabled that undertaking to eliminate uncertainties regarding its conduct on the market, so that normal competition might as a result have been prevented, restricted or distorted.”

That analysis would have been wholly superfluous if the fact that Solvay had neither publicly distanced itself nor gone to the authorities was an answer to its attempt to rebut the presumption. Indeed, the CJEU does not make any reference to public distancing or a report to the authorities as relevant considerations.

Although not expressed that way, it may be that this ground of P4u’s appeal should be understood as a submission that in this regard EU law has developed since those cases, since P4u relies on what the CJEU and Advocate General said in Eturas. I accept that this is arguable (although since Eturas did not concern a meeting what was said there on this point is clearly obiter) but as the Advocate General explained, that development is based on policy reasons which apply only when an undertaking participated in a meeting or meetings at which an anti-competitive agreement is concluded as between the other participants. As he pointed out, in those circumstances the silence of an undertaking “encourages the continuation of the infringement and jeopardises its discovery”: Opinion of Szpunar AG at paras 83-84. That has no application to the present case where only Mr Dunne and Mr Swantee were present, and if EE was not party to a concerted practice there is no infringement to be discovered.

In its skeleton argument, but not in its grounds of appeal, P4u raises an alternative ground that the finding in the Judgment that the Anic presumption was rebutted cannot be supported on the evidence. That is accordingly a challenge to the factual finding at [231]. However, the argument advanced is that although the finding that EE rebutted the presumption that the exchange of 19 September 2012 influenced its conduct in 2012 or 2013 cannot be challenged, the exchange should nonetheless be presumed to have influenced EE’s decision in mid-late 2014 not to renew its agreement with P4u. This extraordinary contention was not put forward in P4u’s 170 page closing submission and Mr MacLean KC accepted, in response to a question from me, that it was never put to Mr Swantee, in the course of his extensive cross-examination, that when considering in 2014 whether EE should cease trading with P4u he had in mind what Mr Dunne had told him some 20 months earlier. Moreover, O2 had exited from P4u for new connections (in January 2013) and for upgrades (in January 2014), so none of this was by that stage confidential. Any suggestion that Mr Dunne gave Mr Swantee to understand in September 2012 that if O2 exited from an indirect retailer then it would never, irrespective of market developments in the future, decide to resume its relationship with that retailer is, frankly, preposterous.

If, contrary to my primary finding, confidential commercial information was disclosed by Mr Dunne such as to remove uncertainty in EE as to O2’s conduct, I held that the Anic presumption applies to EE as the recipient of that information but does not apply to O2 as the party disclosing that information: [232]. Ground 2 contends that this was wrong in law. That is a remarkable submission which is, unsurprisingly, contrary to the authorities. Anic itself was one of the appeals concerning the Commission’s decision on the Polypropylene cartel, where Anic had participated in a series of meetings between the major producers for over three years where the producers fixed price and sale volume targets. Although Anic had not itself agreed such targets, its participation in the meetings meant that Anic received information from its competitors as well as disclosing its own information to them, as set out at paras 96 and 110 of the CJEU judgment. The CJEU’s formulation of the Anic presumption in terms of “taking account of the information exchanged with its competitors” is a reference to the mutual exchange of information which was found to occur in that case. But if A discloses its commercially sensitive information to B, but B does not disclose its commercially sensitive information to A, and B is able to rebut the Anic presumption, it is close to absurd to suggest that there is nonetheless a concerted practice unless A shows that it did not take into account its own information. That self-evidently would not reduce A’s uncertainty as to its competitor’s conduct on the market. P4u’s reliance on the Eturas case in its skeleton in this regard is misplaced. The issue in that case was whether there was a concerted practice between the travel agents which received the message about discounts: [214].

The suggestion by P4u if the law were otherwise that would place the party disclosing the information in a better position than the party receiving the information is misconceived. If the recipient(s) can rebut the presumption and show that the information had no effect on their conduct, there is no infringement and neither discloser nor recipient is liable.

Further, and in any event, I held that even if the presumption applied to O2 (such that O2 failed to rebut it), so as to give rise to an infringement of competition law, it did not in fact influence O2’s decision to exit from P4u: [233]-[235]. That was a distinct issue in the Agreed List of Issues from the question whether there was any collusion as alleged between EE and O2: cp issue 14(d) and issue 18 in the Agreed List of Issues. There is no challenge to that finding in the Grounds of Appeal. That accordingly means that irrespective of success on Grounds 1-3, the result of the case will be the same.

Since this was pointed out in the skeleton arguments of several of the defendants, in oral argument Mr MacLean sought to challenge [233], submitting that the Anic presumption applies not only to establish infringement of the competition law prohibition of a concerted practice but also to prove the necessary causation for private damages claim. Here, it is common ground that the alleged concerted practice would be an infringement ‘by object’. Since this is not a ground of appeal, I will only say that it would have far-reaching consequences if by reason only of the legal presumption which can establish an infringement by object, that infringement is presumed to establish causation (albeit not quantum) of the particular decision (e.g. to raise prices, or in this case, to cease trading with P4u) for the purpose of a damages claim, particularly in view of the number of private actions now brought following decisions by the competition authorities. It is manifestly a very different thing to say that confidential information had the effect of removing some competitive uncertainty, and thereby infringed the pubic law prohibition, from saying that it caused the recipient of the information to take a particular decision on which the private claim is based.

This ground concerns the findings regarding Mrs Derbyshire’s email of 10 August 2014: [599]-[608]. Mrs Derbyshire was part of Mr Eyre’s team at EE, which reported to Mr Allera and Mr Milsom: [465].

To put the email in context, on 17 June 2014, Vodafone group approved the proposal from Vodafone UK to exit from P4u and sign an exclusive agreement with CPW: [382]; and then on 17 July 2014 that Vodafone-CPW agreement was signed: [384]. By early July 2014, the key aspects of an agreement between EE and CPW were agreed; and on 7 August 2014, the EE-CPW agreement was formally concluded, “subject to shareholder approval”: [553]-[556].

Mrs Derbyshire said that she had no recollection of the email and I accepted her evidence: [600]. I found as true the evidence of Mr Eyre that he had no serious expectation that Vodafone was about to pull out of P4u: [606]. I also found that there is no indication that such information was conveyed to the EE Board: [607].

I rejected the inference that P4u had sought to draw from this email, i.e. that it referred not to something that EE learnt in early August but that it was a reference back to knowledge gained from Vodafone much earlier and which had therefore influenced EE’s decision-making in June-July. (Footnote: ⁶) I found as a fact, rejecting the evidence of Mr Milsom, that the information probably came from him or possibly Mr Allera, such that by the end of the first week of August (i.e. Friday 8th August) they thought that Vodafone would shortly leave P4u: [602].

In the light of these findings, which are not challenged, my conclusion that insofar as some people at EE gained this information it had no effect on EE’s decision to leave P4u is therefore not impugned by Ground 4, even if the legal ground were correct.

The question of how EE came to this expectation in early August is therefore not critical to the outcome of the case as regards either Vodafone or EE. But in any event, P4u mischaracterises the Judgment in this regard. P4u submitted as part of its case that the court should infer from this email that EE had received such information from Vodafone. In deciding whether to accept that submission, it was appropriate to consider other possibilities: see at [86]. I do not think that a trial judge, faced with a disputed inference to be drawn from a document, is compelled to adopt either the inference put forward by the claimant or an inference put forward by the defendant. Here, I did not find that the true explanation was that the information came from CPW, but that this was a more plausible explanation than the one urged by P4u, for the reasons set out at [603]-[605], and accordingly that P4u failed to establish the inference it sought to draw. This is wholly different from the case of Al-Medenni v Mars UK Ltd [2005] EWCA Civ 1041, relied on by P4u. There, the trial judge had held that the defendant was liable not on the basis of the conduct of a specific employee of the defendant, as had been alleged, but on the basis of the conduct of someone else, which had not been alleged. That was obviously impermissible since the defendant had not come to trial to dispute the claim on that basis.

I recognise that the Judgment was very delayed, which I regret. However, I would point out that alleged collusion between major companies is generally the subject of investigation by a competition authority, leading (unless collusion is admitted) to a lengthy decision prepared by a team of officials. Because of the complexity of the matter, it is wholly exceptional for such allegations to be determined in adversarial litigation before a single judge. The main reason why it took so long to produce the Judgment is that, in order to obtain a full understanding of the evolution of the decisions of each of the three MNOs, and the influence exerted by their parent companies or shareholders, I considered it important not only to re-read the very many documents referred to in the course of the trial but also to read the multitude of additional documents referred to in the footnotes to the written closing arguments of the various parties.

P4u recognises that delay in production of a judgment is not in itself a ground of appeal: Dansingani v Canara Bank [2021] EWCA Civ 714 at [85]. (Footnote: ⁷) It means that factual findings in the Judgment may be subject to more intensive scrutiny by the appellate court. Accordingly, I address below the six specific matters on which P4u relies. However, I should note that in a trial with this volume of evidence, both oral and documentary, if the Judgment had to identify and analyse every answer or document which each party relied on, the Judgment would have been very much longer than it is already.

(a) The Landmark Lunch: P4u contend that I should have concluded that Mr Dunne took Mr Swantee’s response to his approach as a “green light” from EE. The analysis of the Landmark Lunch was based primarily on the closest contemporary documents supplemented by what the only two persons present said in their evidence at trial (written and oral), applying the approach set out at [42]-[48]. P4u contend that it was an error not to refer to what Slaughter and May said in their advice to EE. That was not evidence about what occurred at the lunch, but a solicitor’s view as to what the legal risks might be: it was not relevant to the factual findings the Court had to make. Secondly, P4u say that it was an error not to refer to a reason given by Mr Blendis in cross-examination for not sending a letter of protest after the meeting. However, in my judgment, the answers given by Mr Blendis to the series of questions on that topic were not so clear and certainly did not support P4u’s case that Mr Swantee was not being truthful in his evidence and in fact had given encouragement to Mr Dunne. In context, this was part of the evidence supporting my finding that what EE did immediately afterwards was motivated by a concern not to antagonise Mr Dunne when O2 had been trying to block the launch EE’s 4G product: [227]-[229]. Thirdly, P4u seek to draw something out of Mr Swantee’s acceptance that he could have made an “even clearer” protest to Mr Dunne. That is no more than stating the obvious; and in that same answer Mr Swantee notably said that he did not think that [in the conversation] Mr Dunne “would have got any comfort from me anyway”.

(b) Mr Dunne – Mr Laurence: this is discussed in the Judgment at [239]-[245]. P4u alleges an error in failing to take account two documents of the slide presentation circulated in advance of the O2 offsite meeting in September 2012. In fact, that slide presentation, prepared by Mr Maple, is quoted from and discussed at [129]-[131]. It is a 31-page document that looks at various potential scenarios. I regarded it as more relevant to quote from slide 5 (with reference to other MNOs following a withdrawal) and slide 15, summarising Mr Maple’s views of the necessary conditions. P4u complains about lack of reference to a line from slide 13, which is simply one of the various alternative assumptions being made in the modelling, which always required assumptions. There are other slides which are also relevant and which I similarly did not refer to expressly. Secondly, P4u allege there was a failure to consider in this context Mr Evans’ text message of 2 January 2013. That text is quoted at [198(c)]. But while I regarded it as relevant in the context of analysis of the Landmark Lunch, in my view that text could not support a case of collusion with Vodafone since “reassuring” Mr Swantee has nothing to do with Vodafone. By contrast, Mr Evans sent a full email shortly afterwards, on 10 January 2013, which I considered was much more relevant to the allegation regarding Mr Laurence. The email is therefore discussed at [244], although I then held that it did not support the alleged collusion.

(c) Mr Humm – Ms Castillo: this concerns their meeting in mid-September 2013, discussed at [287]-[293]. This ground fails to reflect the reasoning in the Judgment. The primary reason for holding that Ms Castillo would not have disclosed O2/Telefonica’s forward commercial strategy concerning P4u or given any commitment concerning P4u to Mr Humm, was that Mr Humm was in no position to give her any indication at this time as to Vodafone UK’s forward strategy, still less any commitment; and that in those circumstances it was inherently unlikely that she would give such information without receiving corresponding information or a commitment in return: see at [291]. P4u does not challenge this factual finding regarding Mr Humm; and the unlikelihood on which my reasoning was based is a point made by P4u itself: see the quotation from its pleading at [282].

As to Ms Rose’s email of 1 November 2013, that was sent to her colleagues at the very outset of negotiations between Vodafone UK and P4u, and at a time when Vodafone UK was expecting to conclude a new deal with P4u: indeed, in her email Ms Rose stated that she was at that point attracted by the idea of a new deal that would “lock in” P4u. I did not refer to it because I found that the change in Vodafone UK’s thinking about P4u occurred significantly later, following the so-called “Tim’s box” meeting of 11 March 2014: [336]-[344]. That finding is not challenged in the grounds of appeal. Accordingly, this email was not relevant: by March 2014, O2 had already come out of P4u completely since the O2 Agreement regarding upgrades had expired on 31 January 2014.

(d) Mr Dunne – Mr Whiting: this appears to refer to the conversation between them on 27 January 2014. The evidence on this is discussed extensively at [273]-[280] and my findings regarding the conversation are at [305]-[311]. Since P4u’s PTA skeleton argument makes no submissions regarding this point, it is impossible to comment or to see on what basis P4u suggests that there is a real chance that the findings would be overturned.

(e) Mr Scheen – Mr Humm: this concerns their conversation of 2 April 2014: [562]-[568]. P4u seek to challenge my finding that the probable subject which Mr Sheen was seeking to discuss was the potential for a joint acquisition of P4u. But even if P4u were correct that Mr Scheen in fact had wanted to discuss a withdrawal from P4u, the point goes nowhere since I found that because Mr Humm declined to have the secure phone call which Mr Scheen proposed, no exchange of confidential information or discussion of that subject ever occurred: [575]-[576]. Those findings are not challenged.

(f) Alleged collusive sharing of confidential information between Vodafone UK and EE in 2014. In the PTA skeleton argument, P4u rests this point on two matters: (i) EE’s modelling in April-May 204; and (ii) the blank pages in the minutes of the 21 May 2014 BRM (tendentiously described by P4u as the “missing” minutes).

As regards the evolution of EE’s internal modelling, that is analysed in some detail at [517]-526], including by reference to P4u’s allegation summarised at [519]. The critical question was not whether Mr Eyre changed his assumption about P4u in the slide deck produced on 12 May 2014 (as I found: [521]) or on 5 May 2014, as P4u submit I should have found. It is whether the change followed his discussions with Messrs Milsom and Allera on 29 April 2014, which was P4u’s case at trial: [519]. I found that it did not, for reasons set out at [520]. That did indeed involve rejecting Mr Eyre’s recollection, as expressly there stated. The finding was based on the contemporary documents in preference to the witness’ oral evidence which I found could not precisely identify the sequence of events: [519].

My findings regarding the 21 May 2014 BRM are at [527]-[533] and the blank pages in the minutes are specifically discussed at [590]-[597]. Although P4u alleges that I fail to have regard to Mr Harris’ evidence, in fact there is extensive reference to his evidence at [591]-[593]. P4u’s real complaint is that I rejected its argument that the omission was deliberate, and declined to draw the inference from the blank pages which it urged the court to draw. The reasons for that are at [593]-[595].

This is a compendious ground alleging that I disregarded inherent probabilities and the parties’ motivations and also adopted a compartmentalised approach, failing to consider the evidence in the round.

In fact, the need to have regard to inherent probabilities was emphasised at the outset: [83]; and in fact, several of the findings which P4u seeks to challenge expressly take account of inherent probabilities: see e.g. at [291]; [308]-[310]. As regards a compartmentalised approach, the PTA skeleton argument submits that because the various matters on which P4u particularly relied as regards Vodafone and EE (e.g. the Scheen-Humm conversation, the minutes of the 21 May 2014 BRM, and the meeting with Ofcom of 4 September 2014), are dealt with in a distinct section of the Judgment, described by P4u as “effectively a postscript”, they were considered only after the primary allegations of collusion had been dismissed.

If P4u seriously considers that the discussion of those matters at [562]-[619] is a postscript, it is under a massive misapprehension regarding the Judgment. Any judge, faced with a vast and complex factual scenario, has to consider how to structure their judgment. Here, I felt it was important to set out a chronological account of the evolution of the decision-making of each of the three MNOs. However, there were some specific matters that appeared to be particularly relevant to the allegations of collusion and therefore, in my view, required intensive analysis. Since that would have significantly disrupted the chronological narrative, it seemed clearer to discuss them in a special section of the Judgment. Moreover, far from meaning that they were looked at in isolation, this approach, on the contrary enabled them to be analysed in the light of what the relevant parties were saying and doing both before and afterwards. And there is indeed a section, after the discussion of these various particular matters, which deals with the allegations of collusion as regards EE “in the round”: [616]-[619].

I should add that the same approach, for the same reason, is taken in the Judgment to the Landmark Lunch, which is analysed in a separate section following the chronological account of the O2-P4u negotiations and the decision reached by O2.

Although there is a challenge to the construction of clause 13.2 of the EE Agreement, that goes nowhere unless P4u has a real chance of success in overturning the finding that EE did not act in bad faith. The allegation of bad faith is discussed at [698]-[725]. As set out at [698], that was dependent on three propositions. The PTA skeleton does not seek to challenge my conclusions on propositions (a) and (b). The appeal is directed at (c) and the decision by EE to inform P4u of the position: see at [710] and [722]. I did not hold that, as a matter of law, it could never be bad faith to disclose relevant information, but observed that as withholding of relevant information is cited as an example of potential bad faith, here it would be paradoxical to say that disclosing such information was bad faith on the part of EE. That was in view of the circumstances set out at [719].

Since bad faith is an objective test (what would be regarded as commercially unacceptable by reasonable and honest people), if there is justification for sending a letter that does not amount to bad faith, that should be determinative. Further, where a justifiable (i.e. good faith) reason was indeed the motivation of several of the directors taking the decision to send the letter, I do not see that sending the letter could nonetheless amount to bad faith on the part of a company only because other directors had a different motivation. If it were, otherwise, it would be necessary to scrutinise the mindset of each of the directors at a meeting, turning an objective test into a complex subjective assessment.

This concerns adverse inferences to be drawn from the failure to preserve documents.

As regards EE and the recording of part of the conversation at the Landmark Lunch, I rejected the allegation that the loss of the iPad and of the transcript was deliberate: [201]. It was not very clear whether a copy of the recording was ever made onto a CD: [192]. Both Mr Swantee and Mr Blendis were cross-examined about these matters, and Mr Swantee explained how his iPad would have been replaced well before the allegations in this case came to light. Therefore the failure of preservation really concerned the transcript note made by Mr Blendis. However, once I found that Mr Blendis had sought to preserve the transcript and that the failure of preservation was not deliberate and arose after Mr Blendis had left EE’s employment, it would clearly have been inappropriate to draw an adverse inference when EE went to considerable lengths to disclose everything it could about the lunch, waiving privilege on its communications with, and advice from, Slaughter and May. The suggestion that the transcript would have revealed that something said by Mr Swantee incriminated EE became, at best, wholly speculative, particularly when Mr Swantee knew that what he said was being recorded and, indeed, where he was making that recording precisely out of concern at what Mr Dunne was saying to him.

In those circumstances, it would clearly have been wholly inappropriate to draw an adverse inference. Indeed, while a section of P4u’s written closing was devoted to “Inferences from Document Destruction” (paras 62-66), that made only passing reference to EE and focused instead on Telefonica and Vodafone.

As regards Telefonica, see at [56]-[59] and [303]. Contrary to ground 8(b), I did consider whether the destruction was deliberate and rejected that allegation: [59]. The question whether in the circumstances an adverse inference should be drawn is a matter of discretion. It is important to note that the failure of preservation did not apply to O2 so, as noted at [303], internal communications from Telefonica SA or Telefonica Europe to O2 were available through O2’s disclosure. I do not see a real chance that the decision not to draw an adverse inference in this case would be overturned on appeal.

P4u’s application for permission to appeal relies on both limbs of CPR rule 52.6. For the reasons set out above, I consider that none of the grounds has any real chance of success. In those circumstances, I do not see that there is any other compelling reason to give P4u permission to appeal, and none is suggested in the PTA skeleton argument.