Federal Deposit Insurance Corporation as Receiver for Amcore Bank NA & Ors v Barclays Bank Plc & Ors

This is the judgment of the Court following the sixth Case Management Conference (“CMC”) in this long-running case. The CMC essentially concerned the application of the claimant (as the receiver of 19 closed US banks) for disclosure by the defendant banks of transaction data requested by the claimant’s LIBOR expert, Dr Friederiszick of E.CA. It was contended that he required this data to perform his methodology on the “But-For USD LIBOR issue”. Having heard submissions over two days, I gave a ruling at the end of the hearing that the application was refused, on the basis that the disclosure sought was neither appropriate nor proportionate, let alone necessary or required for Dr Friederiszick to produce an expert report on the basis of his preferred methodology. I indicated that I would give my reasons for that ruling in due course. These are those reasons.

The claim concerns alleged “lowballing” of USD LIBOR by the defendant banks between August 2007 and December 2009. The claimant alleges that the defendant banks colluded and agreed including with the British Bankers’ Association (“BBA”, represented in these proceedings by the third and fourth defendants) in making LIBOR submissions that were artificially low and, in doing so, breached Article 101 of the TFEU and Chapter 1 of the Competition Act 1998 and/or that they are liable in tort under US state law for fraudulent misrepresentation. It is alleged that the 19 closed banks suffered loss as a result of the suppression of USD LIBOR by reducing their receipts from their banking businesses. The trial of this claim is fixed for 19 weeks from 23 February 2026.

The proceedings were commenced in March 2017 following a decision of the US Court in the LIBOR Multi-District Litigation (in which the claimant is one of a large number of plaintiffs) that it did not have jurisdiction over non-US banks including the defendant banks. The proceedings have a complex history but the salient aspects can be summarised as follows.

The eleventh defendant, UBS, applied unsuccessfully to strike out the competition claims on grounds of limitation. It took until July 2020 for that application to be determined. Snowden J (as he then was) dismissed the application.

Accordingly, the first CMC did not take place until March 2021 before Miles J. He was concerned about the length of time it had taken to progress the proceedings even since July 2020. At [41] of his judgment, he said:

“It does appear to me, looking at the evidence of Ms Vernon [the solicitor for the claimant], that though some work has been done since the end of July 2020, not a great deal has been done to progress the case. I have gained the impression that more needs to be done for this case to progress. If it is a matter of resourcing, then it seems to me that more resources need to be devoted to the case. I accept that it is a very large case, but that is in part because the Claimant has chosen to bring proceedings on behalf of such a large number of Claimants. The Claimant necessarily has to accept that it will devote sufficient resources to the case to enable it to proceed properly. If one chooses to bring proceedings in this jurisdiction, one must abide by the rules of the jurisdiction. Part of the overriding objective is for cases to be conducted as expeditiously as possible, conforming to the other requirements of justice.”

The defendant banks gave disclosure in May 2021 of their borrowing and lending data in respect of millions of transactions. In the case of nearly all the banks this covered the period from January 2005 to May or June 2012, so for the alleged suppression period (pleaded as from August 2007 to the end of 2009) as well as more than two years before it and two and a half years after it. Although the claimants and its experts have had that data for more than three and a half years, it appears that they have not engaged in any detailed analysis of it. Mr Andrew Green KC for the claimant accepted in his reply submissions that the regression analysis which his expert proposes had not yet been commenced.

Due to developments in the litigation in the US, the claimant applied at the second CMC before Sir Anthony Mann in February 2022 for a temporary stay of these proceedings, which he granted until 31 July 2022, with a third CMC to take place on or after 1 October 2022. The third CMC was in fact heard by Zacaroli J (as he then was) in December 2022. He noted the complexity of the case and the number of issues. He also noted that it was hoped that disclosure would be given on all issues in the next nine months after which it was incumbent on the claimant to particularise its case by way of an amended pleading. He ordered that any further application for specific disclosure should be made by 13 October 2023.

Extensive further disclosure was given by the defendants between July and October 2023. The claimant’s solicitors sent a lengthy letter seeking disclosure of transaction data on 6 October 2023 and then issued an application for that disclosure on 23 October 2023, one of four applications by the claimant considered by Miles J at the fourth CMC in December 2023. It is clear from Ms Vernon’s witness statement in support of that transaction data application that what was being sought at the behest of E.CA was data from the beginning of 2002 until the end of 2016. She emphasised that the disclosure to date was uneven between the defendant banks and that uniformity was important. In revised relief sought at the hearing before Miles J, rather than seeking the disclosure, the claimant sought information about what data was held by the banks and what further information they had. At the hearing, the defendant banks pointed out that they had already disclosed vast amounts of transaction data covering some six and a half years.

Miles J was not prepared to grant even that more limited relief. He set out his reasons at [133] to [149] of his judgment ([2024] EWHC 85 (Ch)). He agreed with the bank defendants that the application was premature and that there should first be discussions between the experts to determine exactly what further data (if any) is needed by the experts and the costs of retrieving it.

Miles J ordered a fifth CMC to take place in May 2024 to deal with any other specific disclosure applications. That CMC was heard by Marcus Smith J on 14 and 15 May 2024. The transaction data application, which had been dismissed by Miles J, was not renewed before Marcus Smith J, but, at the outset of the hearing, counsel then appearing for the claimant, Mr Richard Blakeley KC, explained that the disclosure process in relation to transaction data was ongoing and that the requests for disclosure were expert-driven in the sense that, as the judge put it, what was being sought by way of transaction data had the imprimatur of the claimant’s experts in that they were saying: “we can establish a but-for LIBOR position if we have certain material”. The judge went on to ask whether the experts had agreed a methodology for establishing the but-for LIBOR position and Mr Blakeley KC said there had been no attempt to agree a methodology. He said that the claimant had suggested that the expert issues be determined at the sixth CMC in November 2024, but the defendants wanted to have amended pleadings first. The judge expressed concern that this was being allowed to drift and left too late for a trial in February 2026. On the second day of the CMC, he commented that, when he had read the material for the hearing, he had wondered when he was going to get an application to adjourn the trial.

What he in fact did was to set out in the Fifth CMC Order a timetable for agreement by the experts of the methodology for evaluating whether USD LIBOR was suppressed as alleged or statements of the rival methodologies, together with the specification of any data requests supported by evidence from the experts explaining why the data requests were being pursued by reference to that expert’s methodology. The date for responses to data requests was set by the judge as 18 October 2024 and any extant data requests were to be determined at this sixth CMC. Marcus Smith J also set a longstop date for exchange of experts’ reports on alleged USD LIBOR suppression of 28 February 2025, but he expressly provided in the Order for those expert issues and dates to be considered again at the subsequent CMCs. It is worth noting that none of the parties sought to dissuade the judge from adopting the procedure set out in the Fifth CMC Order.

On 28 June 2024, the claimant served a methodology statement from Dr Friederiszick of E.CA as contemplated by [15] of the Order of Marcus Smith J. He explained how his methodology would involve a regression analysis. He considered that the transaction data produced to date by the defendant banks was not sufficient to permit the analysis he envisaged, which would cover before, during and after the alleged suppression period of 2007 to 2009. He was seeking transaction data back to February 2002, five years before the alleged suppression period and for a further three years after the Wheatley Review was published in July 2013, that is until August 2016, nearly seven years after the end of the alleged suppression period. He also sought transaction data relating to other financial instruments than the direct comparator categories, specifically repurchase agreements (repo transactions). He noted that, in the alleged suppression period, there was a lack of direct comparator data for a substantial proportion of days due to the general lack of liquidity on the interbank market.

The defendants served a responsive methodology statement from their expert Dr Padilla of Compass Lexecon on 23 August 2024. In summary, his methodology compared USD LIBOR submissions to actual interbank transactions by the relevant defendant bank on a given day in a given tenor and therefore did not require any disclosure of transaction data beyond that given already.

The experts met on 4 and 10 September 2024. Immediately after the latter meeting the claimant served draft Re-Re-Amended Particulars of Claim. Of relevance for present purposes is the deletion of [42] of the original pleading. This had provided that the expert evidence of alleged LIBOR suppression would be based on the bank defendants’ actual costs of borrowing:

“The FDIC-R will rely on, inter alia, the difference between: (i) the Bank Defendants’ (and/or the other Panel Banks’) USD LIBOR submissions during the Suppression Period on the one hand; and (ii) on the other hand the actual costs of borrowing to the Bank Defendants (and/or the other Panel Banks) during the Suppression Period by reference to the costs of funds in transactions the Bank Defendants (and/or the other Panel Banks) in fact entered into and/or expert evidence as to the Bank Defendants’ (and/or the other Panel Banks’) actual (or likely) costs of borrowing in the interbank market (or markets) for the borrowing and lending of USD.”

This has been deleted and the claimant has substituted a reference to Dr Friederiszick’s methodology statement dated 28 June 2024.

The claimant served a supplementary methodology statement from Dr Friederiszick on 8 October 2024, together with a letter from its solicitors setting out a series of data requests to each of the defendant banks. These sought transaction data for the direct comparator categories (deposits, certificates of deposit and commercial paper) for longer periods than already disclosed although the “clean periods” either side of the alleged suppression period were shorter than Dr Friederiszick had originally sought, now being from 1 January 2004 and up until 31 December 2013. For the whole of the relevant period transaction data was also sought (to the extent not already disclosed) for other financial instruments, specifically repos and floating-rate notes. The request also sought data for puttable instruments and callable instruments for the whole of the relevant period, but that aspect of the request was abandoned before the current CMC.

A further statement from Dr Padilla was served on 5 November 2024 responding to the data requests and the defendant banks served evidence detailing the work and the cost which producing the additional data would involve. The claimant then produced a Reply statement from Dr Friederiszick dated 14 November 2024 in [17] of which he stated, significantly:

“A simplified form of the USD LIBOR-submission regression analysis could in principle be conducted using only the disclosure already provided of direct comparator transaction types and… covering a shorter “clean” period (August 2004 to December 2012) than the one for which I requested data. However, the relevant question for me as an independent expert seeking to answer the questions in the most appropriate way is not whether the statistical technique I have proposed can in principle be conducted in a less robust fashion using the existing data, but on whether the additional data requested will render the approach materially more robust and reliable.”

Dr Friederiszick noted at [21] that the claimant’s solicitors had informed him that some bank defendants might be able to disclose data more quickly than others. He saw no difficulty in using the existing data disclosed as a starting point and refining his analysis as additional data became available. However, he would require the final dataset and resolution of all clarification questions 2-3 months before the submission of his expert report. Later, at [54] he said that if no further transaction data were produced by the defendant banks at all, he would require at least 4 months in which to produce his report. If all the additional transaction data requested were produced, he estimated that the time required to produce his report would be up to six months.

Before considering the relevant submissions of the parties and setting out my conclusions, it should be noted that the bank defendants sought to argue in their evidence and in correspondence that the methodology of Dr Padilla, who does not require any more transaction data, should be preferred to that of Dr Friederiszick. However, as I pointed out in argument, the question of whose methodology is to be preferred is a matter for the trial judge and cannot be resolved at an interlocutory hearing. Ultimately, despite submissions in reply made by Mr Green KC that the bank defendants’ submissions were still made on the basis that the methodology of Dr Padilla is to be preferred, I do not consider that I was being invited to choose between methodologies. In any event, in reaching the conclusion that no further disclosure of transaction data should be ordered, I have made no assumptions as to which methodology is to be preferred.

It is also to be noted that prior to the hearing of the sixth CMC one of the bank defendants, UBS, had agreed to provide disclosure of certain further transaction data. This is reflected in the order made as a result of the CMC which I have already approved.

I propose to summarise in broad outline the rival submissions on the issue of whether disclosure of further transaction data should be ordered. On behalf of the claimant, Mr Andrew Green KC submitted that the defendant banks’ contention that the additional transaction data sought by Dr Friederiszick was not necessary for his regression analysis which he could conduct on the material he had, was misconceived. It involved rejecting his independent expert opinion that the additional data would increase the robustness of his analysis and it ignored the enormous data gaps in the currently disclosed data for the alleged suppression period.

Mr Green KC also submitted that the defendant banks’ argument that the additional data categories, repos and floating rate notes, were insufficiently relevant was misconceived, because it was Dr Friederiszick’s expert opinion that he needed this data for his methodology particularly given the paucity of existing data in many days of the alleged suppression period. Furthermore, the 2013 Wheatley Review had identified these as types of transactions to which LIBOR submitters should have regard.

Mr Green KC submitted that the Court should accept Dr Friederiszick’s expert opinion that the additional data is required to produce his methodology in a robust manner unless it was absolutely clear at this interlocutory stage that, for the reasons given by Dr Padilla, Dr Friederiszick’s views are wrong.

He submitted that, in considering whether it is reasonable and proportionate to require the bank defendants to disclose the further data, the bank defendants were wrong to suggest that they had only had since 8 October 2024 to respond to the data requests. It was important to bear in mind that the bank defendants had known about Dr Friederiszick’s data requests for five months since his methodology report of 28 June 2024 and some of the banks had used that time to collect the data sought but still refused to provide it. He took the Court to the evidence filed on behalf of the first defendant, Barclays, which showed that they had collected a certain amount of data which could be produced in weeks but they refused to do so. He also referred to the latest evidence on behalf of the NatWest ninth and tenth defendants which is that they have extracted data from their Wall Street system in respect of six of the seven categories sought which could be produced by mid to late January, at a cost of £60,000 and yet they refused to produce it because to do so would be disproportionate.

The eleventh defendant, UBS, had collected the requested data but in its evidence maintained that its production was unnecessary and disproportionate and increased the risk of data mining inherent in Dr Friederiszick’s methodology. However, in the week before the hearing, UBS had agreed to produce the data. It was notable that Barclays and NatWest were not prepared to adopt the same approach in relation to data they had already collected and that the other banks were not adopting the cooperative approach contemplated by Miles J at the fourth CMC. Mr Green KC asked rhetorically why some of the banks had not yet started collecting the data when they had had Dr Friederiszick’s methodology report for five months. For example, Lloyds, the second, seventh and eighth defendants were saying their records were in such a state that it will take five months to know whether there is anything to disclose and a further four to six weeks to disclose any material. They should have started looking in June when they received Dr Friederiszick’s methodology report. Mr Green KC also submitted that the banks could deploy more resources to shorten the time it would take to produce the data.

Mr Green KC submitted that the additional data sought was required because of the significant gaps in the data produced so far. Dr Friederiszick’s opinion is that he needs 30 transactions a day within the three direct categories of transactions for his regression analysis, failing which he will need to look for data in the indirect categories. This was challenged by Dr Padilla, who considers that only at least one transaction by the same bank in the same tenor in the 24 hours before the LIBOR submission is required. Dr Padilla produced a table of the percentage of USD LIBOR submissions with comparable borrowing transactions on that basis (at least one in the previous 24 hours) for each bank in the alleged suppression period which purported to show for the one month and three month tenors averages of 98% and 95%. However, Mr Green KC submitted that this included non-London and non-interbank transactions which were not directly referable to the LIBOR question. Dr Friederiszick had conducted an analysis stripping out those non-London, non-interbank transactions set out in Table 1 of his supplementary methodology statement. This showed that on average in the one month tenor there was not a single London interbank transaction in 26% of the days in the alleged suppression period. For the three month tenor there was no London interbank transaction on 43% of the days. The figures for these gaps increased to 65% and 88% respectively for the six month and twelve month tenors. However, as I pointed out in argument, the percentages of the gaps in data were much greater for UBS and Deutsche Bank than the other defendants, skewing the average.

Dr Friederiszick considered that the incentive to lowball would have been greater on the days when there was low liquidity, during periods of financial distress in the market. This was emphasised by Mr Green KC as a factor supporting the need for additional data to fill the gaps in the present disclosure. He submitted that Dr Padilla did not seem to take issue with this point. He was critical of the contention in the skeleton argument for the fifth defendant, Rabobank, that this was a point against further disclosure since the market dislocation in the financial crisis called into question the probative value of financial instruments which are not directly relevant and the ability of Dr Friederiszick reliably to estimate a simple unsecured interest rate from a non-simple and/or secured instrument. Mr Green KC submitted that it simply could not be right that one excluded disclosure of transactions involving such secured instruments.

Mr Green KC submitted that the Court could not at this interlocutory stage simply reject Dr Friederiszick’s evidence that he needed the additional transaction data to increase the reliability and robustness of his regression analysis. Mr Green KC emphasised that the whole point of a regression analysis is that one tests objectively and by reference to data in the clean period what are the relevant correlations. It will analyse the available data to establish which independent variables have the strongest relationships with the dependent variable, the LIBOR submission, in the clean period and then use those relationships to predict but-for LIBOR in the alleged suppression period.

He relied upon the ruling on disclosure given in the Competition Appeal Tribunal (“CAT”) by Green J (as he then was) in Peugeot SA v NSK Ltd [2017] CAT 2 at [7]:

“even though the Tribunal cannot at this stage say with confidence on the facts of the case that the disclosure sought will be relevant and useful it is possible in the abstract to form a view that as a category it is capable of being relevant and useful and that is sufficient to justify ordering disclosure.”

He also referred the Court to the relevant passage in the final report of the Wheatley Review in Box 4B under [4.8], specifically the third bullet point:

“LIBOR submissions should be determined based upon the following hierarchy of transaction types. Submitters should use their experience of the inter-bank deposit market and its relationships with other markets to develop their LIBOR submission. Greatest emphasis should be placed on transactions undertaken by the contributing bank.

1 Contributing banks’ transactions in:

 the unsecured inter-bank deposit market;

 other unsecured deposit markets, including but not limited to, certificates of deposit and commercial paper; and

 other related markets, including but not limited to, overnight index swaps, repurchase agreements, foreign exchange forwards, interest rate futures and options and central bank operations.”

At [4.7] the report stated:

“…the Wheatley Review recommends that, in advance of the agreement of a more detailed code of conduct, LIBOR submitters should refer to the suggested submission guidelines set out in Box 4.B in the determination of their LIBOR submissions.”

Submissions about the practicalities of the defendant banks producing the additional data sought were made on behalf of the claimant by Mr Daniel Cashman. He dealt first with the remaining points of dispute with the eleventh defendant UBS, of which there are two. The first concerns data for 2013 relating to certificates of deposit, commercial paper and floating-rate notes which would take two months to collect and produce. UBS says this would be an expensive exercise, but he submitted that could not be a sufficient basis for the Court to be satisfied that it is disproportionate. The second concerns repos held on UBS’s Martini database which UBS contends it is not necessary to disclose. The data could be collected and produced in two months. Mr Cashman submitted that it should be disclosed.

In relation to the ninth and tenth defendants, NatWest, he repeated the point made by Mr Green KC referred to at [23] above, that the data could be produced by mid to late January at a cost of £60,000. He submitted that mid to late January was not too late and the cost was not excessive. The claimant did not pursue disclosure of repos against NatWest since very limited (if any) further data was available.

In relation to the first defendant Barclays, data for certificates of deposit, commercial paper and floating rate notes had already been harvested and could be produced in two months, so the position was similar to UBS and the data should be disclosed. The position on repos was that the bulk of the data was on the iRepo database and could be produced relatively quickly, so Barclays should give that disclosure.

In relation to the sixth defendant Deutsche Bank, data on directly relevant transactions for 2004 had already been harvested and could be produced within two weeks, so he submitted that should be disclosed. Additional data for certificates of deposit and floating rate notes for 2004 and the last eighteen months of the overall period could be produced in six to seven weeks which is still January so that data should be disclosed. Repos could be disclosed by mid-February. Overall Deutsche Bank now said it would take 85 to 100 person days to search for and extract the data sought.

In relation to the fifth defendant, Rabobank, data for direct comparator instruments for 2004 would require the restoration of data from storage tapes which would take two to three weeks. Likewise, data on floating rate notes for 2004 to 2012 is held by KPMG on behalf of Rabobank and could be extracted in two to three weeks. Some data on repos is also held by KPMG and could be extracted in the same time frame and some data is on the ARTS system and would take a month to extract. KPMG estimates that to prepare the data requested by the claimant would require two to three months from the date the extracted data was provided by Rabobank, but that time frame had not been reduced even though the claimant had reduced the amount of data it was seeking. Mr Cashman repeated the point made by Mr Green KC that KPMG could reduce the period further by deploying more manpower. Rabobank also made a point about the overall cost of the exercise which could be some £300,000, but Mr Cashman submitted this needed to be kept in perspective and Rabobank had probably already spent that resisting the requests.

Finally, in relation to Lloyds/HBOS, Mr Cashman pointed out that, at the time, this was two separate panel banks, Bank of Scotland, the second defendant and Lloyds, the seventh and eighth defendants. Lloyds would need to restore data held on back up tapes. This would cost about £100,000 which he submitted was proportionate, but the real issue is that it is said that the exercise would take some five-six months. The claimant was sceptical about that, but Mr Cashman repeated the point made by Mr Green KC, if it was going to take that long why had Lloyds not started the exercise sooner. He asked the Court to order the work to be commenced with a liberty to apply to these defendants if there were problems, to be supported by proper evidence.

The lead submissions on the transaction data issue were made on behalf of the defendant banks by Mr Conall Patton KC, counsel for Rabobank. He began by helpfully taking the Court through the procedural history of this litigation (which I have summarised earlier in this judgment). He then summarised the relevant legal principles in relation to disclosure: necessity, relevance and proportionality as set out in [40] of Rabobank’s skeleton argument. He pointed out that those principles apply just as strongly where the source of a request for disclosure is an expert. He referred to the decision of the CAT (Roth J and Hodge Malek QC) in Ryder Ltd v Man SE [2020] CAT 3 where at [40(3)] the CAT said:

“We recognise of course that these are very large damages claims. However, any estimate will still be reached through averages, extrapolations and aggregates. It does not mean that every logical avenue that might be relevant can be explored, or that all data which is arguably relevant must be provided. As observed by Birss J in Vodafone v Infineon Technologies AG [2017] EWHC 1383 (Ch), at [31]: “while of course more [disclosure] can be better …it is relevant to ask how much more would it be and how much better would it make the result.” The decision as to what disclosure to order is appropriate is informed by the views of the economic experts but it is not determined by what data they would like to have or what method they would like to use. It is for the Tribunal to decide.”

At [40(5)] the CAT continued:

“It is not therefore simply a question of relevance, as some of the skeleton arguments we received seemed to suggest. Disclosure will only be ordered in relation to a specific category of documents if the Tribunal is satisfied the documents sought are relevant and that disclosure would be necessary and proportionate. The Tribunal will not make an order merely because it determines that the documents are relevant to the issues.”

Similarly in a ruling on disclosure in the Umbrella Interchange Fee litigation, the CAT (Sir Marcus Smith President presiding) said ([2024] CAT 37 at [16]):

“We consider that sufficient data has already been assembled through the very considerable efforts of all parties to enable pass-on properly and fairly to be tried without the additional data from World Remit or Pets at Home. Whilst we do not doubt that this additional data is data that would be of material benefit to at least some of the experts retained by the parties, we do not consider that it is so material as to oblige the Retailer Claimants to continue to seek to provide this data. The purpose of the exercise has never been either to conduct a sampling exercise with the necessary volume of participants that entails, or to provide complete coverage of all sectors which may be relevant to the claim. All of the parties have more than enough to do in order to prepared for Trial 2, and we consider that the provision of additional data from World Remit and Pets at Home now constitutes a distraction. Furthermore, there is real doubt as to whether the data could be provided in time to enable its effective use by the experts, at least not without disrupting the timetable to trial in a significant way.”

Mr Patton KC submitted that these decisions were inconsistent with the claimant’s submission in its skeleton that, subject to any point about disclosure of the data sought being impracticable or wholly disproportionate, that should be the beginning and the end of the matter. Nothing in the Peugeot case suggested a contrary approach. That was a case where there had been engagement between the experts at an early stage as to what data should be produced. What Green J was not saying was that, as soon as an expert claims that certain data is necessary, the Court should simply rubber stamp what the expert demands. This is clear from the later passages in his judgment at [26] to [29] where the defendant was ordered to produce an explanatory witness statement before any disclosure was ordered.

Mr Patton KC then turned to the bank defendants’ points as to why the Court should not accede to the claimant’s application. The first point, which he accepted was a relatively limited one, was that this was not a case in which the Court was being told by both experts that this data was needed. It was a relevant factor that one independent expert, Dr Padilla, was saying that it was not needed and that there is another perfectly sensible way of dealing with this. The whole point of the procedure which Marcus Smith J put in place was to see if it was common ground between the experts that this data was really needed or just a preference of one of the experts, which was the position as it turned out.

He submitted that the deleted [42] in the claimant’s pleading (referred to in [14] above) speaks volumes. For some seven years the claimant’s case was that it would rely on the difference between the banks’ LIBOR submissions during the alleged suppression period and the actual cost of borrowing to the banks during that period, by reference to the costs of transactions which the banks actually entered into and/or expert evidence as to the banks’ actual or likely borrowing costs. This was no doubt pleaded as one would expect with expert input at the time. Thus the claimant had recognised until very recently that the natural and obvious way to test whether there was lowballing is to compare the LIBOR submissions with the actual borrowing transactions in the interbank market entered into by the banks. In their evidence the defendant banks had said that they would be inviting the Court to infer that the reason that paragraph was deleted was because the data the claimant has had for more than three years does not support a case of persistent suppression. This point had not been answered or rebutted.

Furthermore, in the parallel proceedings in the United States in which essentially the same case is being pursued, the expert process has been concluded without the need for any of the additional data now sought. The experts have been able to conduct their analysis on the data which the defendant banks have also disclosed in these proceedings. This point has also not been answered by the claimant.

Mr Patton KC’s second point was that, even on Dr Friederiszick’s own expert methodology, this additional data is not necessary. He pointed out that regression analysis is most commonly used in competition cases such as where there is an alleged price fixing cartel, but this is not such a case. It is about the banks’ perceived cost of borrowing for which as Dr Padilla points out there is no body of economic theory dealing with this situation, unlike in competition cases. The claimant does not say what variables it is going to use as a matter of economic theory to enable it to predict what LIBOR would have been. The position is left completely open-ended. What emerges from Dr Friederiszick’s methodology statement is that he is seeking the data in order to see if it reveals what factors contributed to the formation of USD LIBOR. In other words, he is asking for the data to see if he can come up with a theory. What emerges from his methodology statements is that Dr Friederiszick has not in fact begun his analysis despite the claimant having had the disclosed data for more than three years. Accordingly this is not a case in which he has done an analysis which arguably reveals lowballing was going on but has insufficient data to make good that case.

On the basis of what Dr Friederiszick now says in [17] of his Reply statement (quoted at [16] above) he not only has performed a complete volte-face from what he originally said, but it is clear he could have started his analysis three years ago on the basis of what has been disclosed to date and seen where it got him. There is no explanation as to why that has not been done. What Dr Friederiszick was saying was: “give me as much data as possible and then I can see where I want to take it”, but as I had pointed out in argument, that is not how litigation is conducted in this jurisdiction. Mr Patton KC submitted that this fed directly into the point made by Marcus Smith J at the last CMC about data mining. It was not appropriate to give the data and then find out what the methodology is. Rather the methodology should be set out, then the Court can assess whether additional data is needed.

Mr Patton KC’s third point was that LIBOR is concerned with simple unsecured borrowing and the defendant banks have already disclosed data about millions of such transactions. What was being sought by way of floating rate notes and repos was data about transactions that are not simple and are secured so that was the wrong sort of data for the issue before the Court. In relation to Box 4B in the Wheatley Review final report, he said that there was a hierarchy of transaction types. Most significant were transactions in the unsecured interbank market and other unsecured deposit markets including certificates of deposit and commercial paper. All this has already been disclosed. Third in the hierarchy is other related markets such as repos which are thus seen as much less relevant evidence. The defendant banks were not saying those transactions were irrelevant, just that where the most relevant transactions have been disclosed and the claimant has not begun the analysis, those other transactions are of insufficient relevance.

Mr Patton KC said that Mr Green KC had sought to make much of the alleged vast gap in the transactions already disclosed. This was not a case of a gap in disclosure. It was just that the banks did not enter into borrowing transactions on particular days during the financial crisis, so that there was no data. However, Dr Padilla’s table referred to at [25] above showed that, for most of the tenors for most of the banks, the percentages of at least one transaction on each day in the alleged suppression period were high. It was true that for the 12 month tenor the percentages were much lower, but Mr Patton KC submitted that this was neither here nor there since the claimant’s case was one of persistent systemic suppression of LIBOR. No-one was suggesting that the pattern of behaviour was different for the 12 month tenor than the other tenors, so that disclosure of more transactions in that tenor was not required, because on the claimant’s case there was systemic lowballing across the board. He also made the point, by reference to Table 3 to Dr Padilla’s Response, that the volume of transactions in the alleged suppression period was in the thousands of billions of dollars.

He submitted that the claimant suggested there was a data gap by stripping out of Dr Padilla’s table any transaction which it did not regard as a London or an interbank transaction. That was an extremely odd logic because the transactions Dr Padilla included are all simple unsecured borrowing transactions comparable with the transactions the LIBOR question is asking about. It is true some are not London transactions or interbank transactions, but he says in so far as that makes a difference that is easily adjusted for. By contrast what the claimant does is take out those transactions which are much more like a LIBOR transaction and then say there is a data gap, which is then used as a launching pad for seeking data for transactions which are extremely different from LIBOR. Furthermore, Dr Friederiszick does not say he is going to ignore the non-London, non-interbank transactions.

Mr Patton KC submitted that the fact that there were days in the alleged suppression period where there was no borrowing, so no comparator, should not make a difference to the claimant’s overall theory of systemic suppression if the theory holds good. A pattern will be established from the days when there is data and one can interpolate for the days where there is no data. The solution is not to use data for transactions like floating rate notes and repos which are dissimilar to LIBOR transactions. In any event, if those dissimilar transactions were disclosed, the experts would engage in highly technical analyses as to how you get from the data in those transactions to a number comparable with simple borrowing, assuming it can be done at all. He submitted it would lead to lengthy cross-examination of the experts at trial on the correct methodology for that calculation which would be a completely satellite debate from the real issue as to whether there was lowballing.

On the date range of data now sought from January 2004 to December 2013, although narrower than originally sought, Mr Patton KC submitted that it was still unnecessarily wide. It was accepted that there was a need for a clean period before and after the alleged suppression period if one were going to do a regression analysis, but the question is how long. He reminded the Court that the pleaded suppression period was from August 2007 to the end of 2009. There was no pleaded case of suppression at an earlier point in time. To the extent that it was now said that there had been regulatory findings against Rabobank and Deutsche Bank in 2005 and 2006, this was sporadic rogue trading which could just as easily have resulted in a higher rate as a lower one, depending on the motive of the particular trader. It was not systemic suppression of LIBOR as now alleged, which was tied in with the financial crisis which did not start until August 2007. Furthermore, the claimant has had the transaction data back to January 2005 for three and a half years, but has not analysed it to demonstrate systemic lowballing prior to 2007. There was nothing to support Dr Friederiszick’s assertion that lowballing in 2005 and 2006 cannot be precluded.

So far as the period after the alleged suppression period is concerned, the claimant pleads that the suppression continued until “at least the end of 2009” but although the claimant has had seven and a half years since the proceedings were issued, it has never particularised any case of persistent suppression after 2009. The claimant has had the transaction data until at least April 2012 for three and a half years but all that is now said in Dr Friederiszick’s methodology statement is that the claimant alleges that suppression might have continued until the end of 2011. That allegation is unsupported by any analysis of the data. The position overall is that the claimant already has ample clean periods either side of the alleged suppression period.

Mr Patton KC then dealt briefly with the issues on proportionality for Rabobank, addressing the points made by Mr Cashman set out at [35] above. In relation to the point that Rabobank’s time and cost estimates had not reduced although the claimant had narrowed its requests, Rabobank’s solicitors had written before the hearing to say the estimates remained unchanged and on the morning of the second day of the hearing, to say that on the basis of what KPMG had told them, more manpower could be thrown at the task of data retrieval which would bring the period required down by three weeks but increase the cost by £30,000. However, the exercise overall would still take some three months which plainly has implications for the trial timetable. He also pointed out that to retrieve data a series of steps have to be gone through, as set out in his solicitor’s witness statement, even if more people are deployed, meaning there is an irreducible minimum period of time the exercise will take.

Submissions on the timetable points were then made on behalf of the defendants by Mr Richard Handyside KC, leading counsel for the Lloyds defendants. He reminded the Court that in addition to expert evidence on LIBOR suppression, it was contemplated that there would be expert evidence in four other fields, competition economics, American tort law, American civil procedure/limitation and quantum. There was an awful lot to get through in time for the trial in February 2026 and it was critical that the trial date was maintained. The defendant banks maintained that the date for exchange of alleged suppression reports should be 28 February 2025, the backstop date set by Marcus Smith J. Their timetable also set 7 November 2025 as the last date in relation to experts’ reports (supplemental quantum reports), again the longstop date set by Marcus Smith J.

In contrast, the claimant’s date for exchange of alleged suppression reports was 17 April 2025, with the expert timetable running through to 28 November 2025 for supplemental quantum reports. Mr Handyside KC submitted that there would be a significant problem with the timetable particularly in relation to alleged suppression reports if the Court ordered disclosure of the additional data sought. It was unclear by what process of reasoning the claimant had arrived at the 17 April date. The claimant’s draft order seeks disclosure of the additional data by 14 February which is wholly unrealistic so far as Lloyds is concerned and, in any event, it does not appear that 17 April is a realistic date by which Dr Friederiszick can produce his report. That allowed only two months from the 14 February date for disclosure whereas in correspondence only a month before the hearing the claimant’s solicitors had been saying he needed three months from disclosure of the additional data. Furthermore, in his Reply statement dated 14 November, Dr Friederiszick was now saying if the additional data were produced he would need up to six months to produce his report. It was unclear whether he was saying six months from now or six months from production of the data, although Mr Handyside KC submitted it was likely to be the latter given the volume of data being sought. Even if it were the former, his report could not be produced until May or June 2025, always assuming the defendant banks could meet the 14 February deadline, which Lloyds cannot.

Dr Friederiszick and Dr Padilla are also the competition economics experts. On the claimant’s now proposed timetable, supplemental alleged suppression reports would be due on 30 May 2025 with a joint report by 27 June 2025. The claimant was proposing to serve its competition economics report by 31 July (the bank defendants’ proposed date is 9 May). Mr Handyside KC submitted that it was not realistic to expect the same experts to be working on two different reports, both substantial pieces of work, at the same time. Furthermore, the claimant’s position is that the alleged suppression expert evidence has to be completed before the quantum reports could be produced. In all likelihood the claimant’s own backstop date for all expert evidence of 28 November would need to be pushed back, which simply does not work for a heavy trial listed for 19 weeks in February 2026. He submitted that something has to give, which should be the wide ranging late disclosure application for additional data.

Mr Handyside KC then addressed Lloyds’ disclosure, noting that during investigations by the US authorities between 2010 and 2014 into, inter alia, its USD LIBOR submissions, Lloyds collated 32 million documents and manually reviewed nearly 2 million of them. This exercise had gone wider than the alleged suppression so the documents disclosed to the US authorities were used as the starting point for disclosure in the parallel US proceedings. It had been agreed that in the present proceedings, Lloyds would disclose the so-called 2016 Lloyds US production set (i.e. the disclosure given in the US proceedings), essentially a so-called “lift and drop” and this was done on 12 May 2021. It was agreed with the claimant that it would review that disclosure and, if it thought additional disclosure was required, formulate a request, but no additional requests were made for some two and a half years.

In the US proceedings, there was a dispute as to the applicable date range and the US Court determined that the appropriate date range was August 2007 to May 2010, not to 31 October 2011 which the US plaintiffs had sought. Lloyds gave the resulting additional disclosure (the so-called 2023 Lloyds US production set) on 8 September 2023 and that disclosure was also given to the claimant in the present proceedings. The position thus is that the second, seventh and eighth defendants have disclosed transaction data for money market and interbank borrowing and lending trades for the period August 2005 to June 2012, two full years before the alleged suppression period and two and a half years after it. Even on Dr Friederiszick’s table stripping out non-London and non-interbank trades, the figures for the one month and three month tenors (the most important because they were most often used as reference rates in commercial lending transactions) still shows strong percentages of days on which data is available for Lloyds of 99% and 83% respectively. The figures for HBOS are lower, but still have at least one transaction in those two tenors on more than two thirds of the days in the alleged suppression period.

The first time that the claimant sought to raise a request for further transaction data was in a 22 page letter from its solicitors sent just before midnight on Friday 6 October 2023, just a week before the deadline for the issue of applications for the fourth CMC. The application notice was then issued on 23 October and responded to by a statement from Mr Bristow of Hogan Lovells, Lloyds’ solicitors from which it would have been clear to the claimant that the data request would require at least two to three months’ work to restore obsolete databases before any extraction could take place, that this was a provisional estimate and that the overall work would cost hundreds of thousands of pounds and take many months.

Mr Handyside KC pointed out that, as already recorded at [8] above, the application for disclosure was not pursued as such at the fourth CMC before Miles J, but rather the claimant sought witness evidence from the defendant banks as to what data they all held. Miles J dismissed that application. Mr Handyside KC submitted that the suggestion now made by the claimant that it was incumbent on Lloyds after that hearing to continue investigations into the costs and feasibility of producing the additional data was wholly wrong. After that CMC, the next step was for the experts to hold discussions as to what was relevant and necessary and there was no reason for the defendant banks to incur significant costs investigating the availability of data which might well not ultimately be required.

By the time of the fifth CMC before Marcus Smith J, there had been some exchanges between the experts but no agreement. He then made the detailed order referred to at [11] above. After extensions of time thereafter, the date for the claimant to make data requests was 8 October 2024 and that was the date when the requests now before the Court were made. In response on 5 November, Mr Bristow served a detailed witness statement explaining the considerable challenges Lloyds would face in seeking to produce any of the data now sought. In summary, he identified the three stages required: first to identify the back ups of the data sought which is stored on physical tapes, second to restore those back ups so that the data is recoverable in a usable format and capable of extraction and third the specific transaction data now requested would need to be identified, extracted and produced. Mr Bristow also explained that the tapes in question were used for disaster recovery purposes. He then explained in considerable detail the work that would need to be done at each of the three stages. Overall the task would take six months. Mr Handyside KC submitted that the Court should accept the evidence about timescales in Mr Bristow’s evidence. Although it had been suggested by Mr Cashman that this was out of kilter with other banks, this was not so. The evidence from Clifford Chance for NatWest if the claimant had pursued the request for repos against them described a not dissimilar series of stages to those described by Mr Bristow and that the exercise would take about the same length of time.

Mr Handyside KC submitted that it would be disproportionate to require Lloyds to undertake this work and pointless given that it would not be available in time for Dr Friederiszick to use it in his report. He expressly accepts in his supplementary methodology statement that he does not require every bank to produce the additional data in order to undertake his analysis.

Mr Adrian Beltrami KC then made submissions on behalf of Barclays. He submitted that Barclays has already produced an enormous quantity of data on over a million trades from 1 January 2005 to 31 May 2012 for the most directly relevant comparators. On average, the claimant has over 150 transactions per day and a total value of US$8 billion for transactions in the alleged suppression period disclosed by Barclays. The alleged data gap, as explained by Mr Patton KC and specifically in relation to Barclays, does not exist because it has been manufactured by excluding, in an illogical way, non-London and non-interbank transactions, in other words the second category in the hierarchy in Box 4B of the final report of the Wheatley Review. None of this made any sense.

He submitted that [42] of the claimant’s original pleading gave the impression that the allegation of material and sustained suppression was based on work done on the numerous transactions already disclosed by the defendant banks in 2021 but it now appears that the work had not been done. However, that pleaded case had been deleted and replaced by reliance on the product of unspecified expert evidence which has not even been started. He submitted that this should not be put forward as the basis for the disclosure of more data to find out whether the claimant has a case.

He emphasised the time and the cost of the exercise which would now be required. He also submitted that the additional data being sought was of less utility than the data the claimant already has. This was not only because, as Mr Patton KC had explained, it is for a less relevant period and includes less relevant transactions but because, at least so far as Barclays is concerned, it is likely to be incomplete and potentially duplicative particularly in relation to floating rate notes and repos, for the reasons explained in the witness statement of Ms Bickerton, Barclays’ solicitor. This fed into a broader point that the evidence being sought from Barclays is not the same as that sought from the other banks. The claimant originally contended that was a real problem, telling Miles J at the fourth CMC that it was necessary for the data sets from the defendant banks to be complete and uniform, as was emphasised several times in Dr Friederiszick’s methodology statement. However, in his Reply statement, Dr Friederiszick now says that there is no fundamental difficulty if the data is not complete or uniform. This change of position is unexplained. When Dr Friederiszick says something is “necessary” it does not mean that it is necessary, just that it is something he would like.

Mr Beltrami KC submitted that the lateness of the application was entirely the fault of the claimant. When Barclays gave its disclosure of data in May 2021 it said in terms that any focused and specific requests for further disclosure should be made once the claimant had reviewed the data disclosed. No application was made for two and a half years.

Ms Nehali Shah made submissions on behalf of the sixth defendant, Deutsche Bank. The disclosure it had made in 2021 included data relating to repos and similar transactions, at least for the alleged suppression period. The claimant still seeks disclosure from it of deposits, certificates of deposit and commercial paper for 2004 and for July 2012 to December 2013 and of repos for 2005 to 2006 and 2010 to 2013. She submitted on the basis of Deutsche Bank’s evidence that, although data for 2004 could be produced within two weeks, production of data for the later periods would take a lot longer, at least three months to extract, process and produce the data. She explained the various stages that would be required as set out in the evidence. The stages of the expert evidence process could simply not be completed in time for a trial in February 2026.

Even though data for 2004 could be disclosed in short order, Deutsche Bank resisted giving such disclosure essentially for the reasons given by Mr Patton KC. Ms Shah submitted that it was difficult to see what the claimant would do with data for 2004 from one bank.

Mr Paul Luckhurst made brief submissions on behalf of UBS which had agreed to produce data it had collected. He explained that this had been agreed because the ease and limited cost of production outweighs the time and expense of further argument on the point and the data in question includes for the period 1 January to 8 August 2005, a period for which the majority of the other bank defendants have provided some data. So the decision to provide the data was a pragmatic one, not because UBS accepts that there are problematic gaps in its transactional data. He submitted, as had Mr Beltrami KC, that there are no massive data gaps and, even if there were, it would be perverse to fill them with repos rather than for example unsecured New York interbank borrowing.

Accordingly the issues in dispute between the claimant and UBS are whether UBS should disclose data for 2013 for deposits, certificates of deposit and commercial paper and whether it should disclose data for repos. He submitted that collecting and producing the 2013 data would take two months and be an expensive exercise. The claimant already has data from UBS for a clean period after the alleged suppression period of three years which is ample.

On the repo data, Mr Luckhurst submitted that UBS has not previously collected this data, including in the US proceedings. It would again take some two months to collect it at considerable expense. As Mr Patton KC had submitted, repos are not an ideal comparator for LIBOR. Their addition to the already extensive data sets would increase costs and complexity with little or marginal analytical benefit.

Mr Laurie Brock made submissions on behalf of the ninth and tenth defendants, NatWest. The only dispute remaining between them and the claimant concerns data from their money market trading platform called the Wall Street system for 2004 and April 2012 to December 2013. He submitted, as had counsel for the other bank defendants, that the clean periods before and after the alleged suppression period were already ample and no further disclosure was necessary.

Mr Duncan McCombe made submissions on behalf of the third and fourth defendants, the BBA parties. No data requests were made against them but they oppose the claimant having further disclosure, adopting Mr Patton KC’s submissions as well as taking three threshold points. He submitted that the expert case the claimant now sought to run was outside the scope of the permission granted by Marcus Smith J which at [13] of his Order was permission to rely on the evidence of an economist “on the issue of whether and the extent to which USD LIBOR was suppressed as alleged”. At that time the allegation as to how the suppression should be calculated was in [42] of the pleading but that has now been deleted. Dr Friederiszick now expressly disavows the comparison to actual borrowing costs contained in the original pleading which was before Marcus Smith J. Accordingly his proposed methodology is not within the permission granted. I can deal with this point now, shortly. Ingenious though it is, I do not accept it. As I pointed out to Mr McCombe, he is an outlier on this, as none of the bank defendants is contending that the proposed methodology is outside the permission for expert evidence. As I have already indicated it will be for the trial judge to decide which is the correct methodology.

Mr McCombe’s second point was that the claimant’s new case in [42] of the amended pleading is inadequately pleaded. He emphasised the different functions of statements of case and expert evidence, relying on what was said by Norris J in Pacific Biosciences of California Inc v Oxford Nanopore Technologies Ltd [2018] EWHC 806 (Ch) at [46]:

“expert evidence and particularity in the statement of case serve two different functions. The function of the expert evidence is not to advance a claimant's case. The function of the expert evidence is to provide opinion evidence, agreeing or disagreeing with allegations which are contained in the claimant's case. It is important that the distinction between the two is maintained.”

Mr McCombe submitted that the simple reference in what is now [42] of the amended pleading to Dr Friederiszick’s methodology statement was inadequate. He drew attention to other paragraphs of the amended pleading where the deliberate suppression of LIBOR is pleaded by reference to actual borrowing costs. I see the force of this point, but none of the defendants has sought to strike out [42] or served a Request for Further Information in relation to it. As matters stand, any inadequacy in the pleading is not a reason, per se, to refuse the additional data requests.

His third threshold point was that Dr Friederiszick’s methodology statement is vague and does not identify clearly a methodology. The claimant’s additional data requests were engaged in what he described as “hit and hope” litigation which should not be encouraged. As I have already indicated, it does not seem to me appropriate at this interlocutory stage to determine which is the correct methodology and I do not propose to do so.

In my judgment, the additional disclosure now sought is not appropriate or proportionate let alone necessary or required for the claimant’s expert Dr Friederiszick to produce an expert report based on his preferred methodology of regression analysis, for a number of reasons. The starting point is that, as long ago as May 2021, the defendant banks gave disclosure of their borrowing and lending data, covering millions of directly relevant transactions, in the case of nearly all the banks for the period from January 2005 to May or June 2012, so for the alleged suppression period (pleaded as from August 2007 to the end of 2009) as well as more than two years before it and two and a half years after it. The exceptions were Lloyds/HBOS who disclosed data from January 2006 to June 2012 and UBS who disclosed data from August 2005 to May 2012.

What was clearly contemplated by the defendant banks and would be expected by the Court is that the claimant and whatever expert it had instructed would then engage in an analysis of that data and, to the extent that more data was required to enable the expert to produce a report, then make an application supported by evidence explaining in detail why the voluminous data already disclosed was not sufficient to enable a report to be produced. However, that is not what occurred. It would appear that although the claimant has had the data for three and a half years it has not engaged in a detailed analysis of it. What is said by Dr Friederiszick at [52] of his methodology statement in June last year is that the currently available transaction data “is not sufficiently uniform or comprehensive to permit the analysis I envisage” from which it would appear that he has not yet commenced that analysis. This was in effect admitted by Mr Green KC in his reply submissions.

Rather what Dr Friederiszick was saying when this application was issued was that in order to conduct the analysis which he was envisaging, it was “necessary” for him to have transaction data not just for the most direct comparators (deposits, certificates of deposit and commercial paper) but for other types of financial instruments, specifically floating rate notes and repos, for a fifteen year period from February 2002 to August 2016. That period has now been narrowed to January 2004 to December 2013, but still covers 10 years. Significantly, as set out at [16] above, Dr Friederiszick now accepts in his Reply statement that he could conduct his regression analysis using the transaction data already disclosed, albeit that he contends that the additional data will make his analysis more robust and reliable. The problem is though that he has not attempted to conduct the analysis using what he has already, so one simply does not know whether a regression analysis on the basis of it would be sufficiently robust and reliable. It gives one no confidence that the additional data sought is really “required” or “necessary”, words repeatedly used on behalf of the claimant, as opposed to being something that, as Mr Beltrami KC put it, Dr Friederiszick would like to have.

As I said more than once during the hearing, I have a distinct sense of the expert tail wagging the court dog here, which is not how litigation in this jurisdiction is conducted. It is for the Court, not the expert, to decide what disclosure is relevant, necessary and proportionate as the passages from the CAT decisions relied on by Mr Patton KC set out at [37] to [39] above make clear. There is nothing to the contrary in the judgment of Green J in Peugeot. The fact that an expert would like additional disclosure or that, as Dr Friederiszick says elsewhere, “ideally” he should have additional data for an extended period does not mean that the Court will order its disclosure. The position might be different if the expert had tried to do a regression analysis with the existing data but been unable to do so and this had been fully explained to the Court, but that is emphatically not the position here.

Mr Green KC sought to address this point in his reply submissions by emphasising that the Court should proceed on the assumption that Dr Friederiszick’s methodology is a valid one which he is entitled to pursue and the relevance and necessity of the additional data sought have to be assessed on that basis and are demonstrated, so that all that remains for consideration is proportionality. As I have indicated, at this interlocutory stage, I do proceed on the basis that the methodology is a valid one, but contrary to Mr Green KC’s submission, it simply does not follow that, because the expert says that the additional data is needed to conduct a robust regression analysis, the Court just has to accept his opinion on that issue. That really is the tail wagging the dog. The Court is entitled to examine critically whether the additional data is really necessary despite what the expert says. That is not to criticise his methodology, contrary to what Mr Green KC submitted in his reply submissions.

Mr Green KC also sought to make much of the alleged data gaps in the transaction data already disclosed in support of the claimant’s case that the additional data was necessary, but I consider that there is nothing in this point, essentially for two reasons. First, I agree with Mr Beltrami KC that the gaps as identified in Dr Friederiszick’s Table 1 to his supplementary methodology statement are accentuated by excluding non-London and non-interbank transactions. This is illogical since such transactions in New York and elsewhere are still more direct comparators with USD LIBOR than dissimilar transactions such as floating rate notes and repos. As Mr Beltrami KC said, they are thus in the second category in the hierarchy of relevance in Box 4B of the Wheatley Review final report: “other unsecured deposit markets, including but not limited to, certificates of deposit and commercial paper”. Although Mr Green KC sought to argue the contrary in reply, it does seem to me that what the claimant is doing is seeking to exclude the second category in support of an application for disclosure of data in the third category: “other related markets” which by definition are less relevant than data in the first and second categories. As Mr Beltrami KC said, this makes no sense.

When the non-London and non-interbank transactions in that second category are included, as they are in Dr Padilla’s table 1, the percentages of days in the alleged suppression period where there was at least one transaction on the day in question are high, particularly in the one month and three month tenors, which as Mr Handyside KC said are the most important because they are most often used as reference rates in commercial lending transactions. I do not consider that it is necessary for the claimant to have the extensive data in relation to the dissimilar transactions in order for Dr Friederiszick to conduct a regression analysis.

The second reason why I consider that there is nothing in the alleged data gap point is that the claimant’s case is one of sustained persistent suppression of LIBOR during the period from August 2007 to the end of 2009. If that case is made out by reference to the voluminous data for the days for which transaction data is available, the fact that data is not available for some of the days in the alleged suppression period should not matter. The point can be illustrated by a simple example of a given week in the period. If data is available for days 1, 2 and 3 and days 6 and 7 and that data shows a pattern of persistent lowballing, the fact that data is not available for days 4 and 5 will not weaken the claimant’s case that there was persistent lowballing throughout that week since the Court will readily infer that the same pattern of lowballing occurred on days 4 and 5. Disclosure and analysis of dissimilar transactions such as floating rate notes and repos will not be necessary to establish systemic suppression of LIBOR if that was occurring.

It is also striking that, until the recent amendment of the claimant’s pleading, [42] of its Particulars of Claim (as set out at [14] above) sought to put its case on lowballing on precisely that basis of the difference between the defendant banks’ USD LIBOR submissions and the actual costs to the defendant banks of borrowing in the interbank market (in other words the actual transactions entered for which data has been disclosed). That case would not have required further disclosure, particularly of the dissimilar transactions. That case has now been deleted and substituted by a general reference to Dr Friederiszick’s methodology statement. It is that methodology statement which seeks to rely upon the other dissimilar transactions and Dr Friederiszick in his supplementary methodology statement criticises Dr Padilla for adopting the approach which was previously, until its deletion, the claimant’s own case. As I said during the hearing it is not for the Court to determine at this interlocutory stage whose methodology is the correct one, but the fact that for some seven years the claimant’s own case was that straightforward one pleaded until its deletion does call into question how “necessary” or “required” the additional disclosure now sought really is. It is no answer to that point that the defendants have consented to the amendment. There is also considerable force in the defendant banks’ submission that it is to be inferred that the reason for the deletion is that the claimant has realised that it cannot make out its case of persistent systemic suppression of USD LIBOR on the basis of the difference between the defendant banks’ USD LIBOR submissions and the actual costs to the defendant banks of borrowing in the interbank market. However, I do not need to decide that question which is plainly one for trial.

Turning to the issue of date range, as just noted, the claimant’s pleaded case is of systemic persistent suppression of USD LIBOR in the period from August 2007 to the end of 2009. There is no pleaded case of such systemic persistent suppression prior to August 2007 and although the reference to the end of 2009 is qualified by “at least”, there has been no attempt by the claimant to plead any case of systemic persistent suppression after the end of 2009. The data already disclosed by all but two of the defendant banks is for the period from January 2005 to the summer of 2012. It thus includes ample “clean periods” of more than two years before and after the alleged suppression period. Even in the case of Lloyds/HBOS and UBS there are clean periods of over a year before the alleged suppression period and, in any event, as already noted, UBS has agreed to provide additional disclosure. The claimant sought to rely upon regulatory findings against Rabobank and Deutsche Bank of manipulation of LIBOR in 2005 and 2006 but, as Mr Patton KC pointed out, this was not systemic suppression of LIBOR which coincided with the financial crisis but sporadic rogue trading which could just as easily have resulted in a higher rate as a lower one, depending on the motive of the particular trader.

So far as the claimant’s request for transaction data beyond 2012, now just for 2013, is concerned, that would give a remarkable four year clean period after the end of the alleged suppression period. I consider that data further away in time from the alleged suppression period will be less useful because of the changing nature of the market and of LIBOR submission processes over time. Overall, I do not consider that the claimant has made out a case that disclosure of data for a wider period than already disclosed by the defendant banks is necessary.

Furthermore, I do not consider that disclosure by the defendant banks of additional data relating to other types of transaction such as floating rate notes and repos is necessary, particularly given Dr Friederiszick’s acceptance in his Reply statement that he could conduct a regression analysis on the basis of the data already disclosed, in circumstances where he has yet to conduct such an analysis. I also agree with Mr Patton KC that the disclosure of these dissimilar secured transactions would necessitate the experts engaging in highly technical analyses to compare the data in respect of those transactions with simple unsecured borrowing which is the direct comparator with USD LIBOR. Unless those technical analyses were agreed between the experts, this would lead to the likelihood of extensive cross-examination at trial about what is essentially a satellite point. As I have already said, analysis of these dissimilar transactions would not be necessary to establish systemic suppression of USD LIBOR if that was occurring.

Accordingly, I would refuse the application for disclosure of additional data on the ground that it is not necessary for the determination of the issue of whether there was systemic suppression during the alleged suppression period, before even considering the questions of proportionality and the likely effect of ordering such disclosure on the timetable to trial. However, the evidence in relation to those questions serves to reinforce the conclusion that the additional disclosure should not be ordered.

In relation to the time which the defendant banks would require to produce all the additional data sought and the cost which would be involved, I do not propose to repeat all the arguments made by their counsel on that question, which are summarised above. It suffices to say that I see no reason not to accept the defendant banks’ submissions and evidence on that issue. The clear picture which emerges is that, with the exception of Lloyds/HBOS, the exercise would be unlikely to be completed by the defendant banks until February or March 2025 at the earliest. Lloyds/HBOS would not be able to complete the exercise until about May 2025. The claimant sought to run a number of arguments to address this conundrum, although it is notable that it did not suggest that the defendant banks’ evidence was inaccurate or incorrect. The point was taken that, in effect, the defendant banks were the authors of their own misfortune because they should have started the exercise of extraction and production of data sooner as some banks appear to have done. With respect, this point is hopeless. Why, I ask rhetorically, should the defendant banks waste time and money in a disclosure exercise when the claimant’s application had been dismissed by Miles J at the fourth CMC in December 2023 and Marcus Smith J had laid down a procedure at the fifth CMC under [19] of which any data requests were not to be made until 20 September 2024?

It was then said by Mr Green KC that the time for extraction and production of data could be foreshortened by the defendant banks and their consultants deploying more manpower, but as was pointed out on behalf of the banks, the various steps required to produce the data are sequential and deploying more manpower would only make a marginal difference.

In his reply submissions, Mr Green KC sought to address the concerns about the effect of the additional disclosure exercise sought on the trial timetable by submitting that the bank defendants should produce whatever they could by mid-February 2025 and that Dr Friederiszick would then produce his report by 17 April 2025. This submission was reiterated by Mr Alex Barden on behalf of the claimant in submissions on timetabling at the end of the hearing. So far as the disclosure is concerned, it seems to me that the suggestion that the claimant would proceed on the basis of whatever can be produced by mid-February 2025 rather defeats the contention that all the additional data sought is required or necessary for Dr Friederiszick to prepare his expert report, as did the related suggestion by Mr Cashman at the end of the hearing, that the defendant banks’ concerns about timing and cost could be alleviated by the Court only ordering the disclosure of some of the data sought.

The impact of ordering the further disclosure and the time it would take to produce it on the rest of the timetable leading up to trial is obvious. Even leaving Lloyds/HBOS out of account, the additional data could not have been produced, if ordered at the sixth CMC, until February or March 2025. Dr Friederiszick contemplates that he might then have clarification requests and, in his Reply statement, says that he would require up to six months to produce his expert report if all the additional data were disclosed which strongly suggests that if the disclosure had been ordered he could not have produced his report until sometime in July or August 2025, five to six months after the longstop date of 28 February 2025 set by Marcus Smith J at the fifth CMC. Mr Green KC’s and Mr Barden’s submissions that, if the bank defendants gave what disclosure they could by mid-February 2025, Dr Friederiszick could produce his report by 17 April 2025 simply flies in the face of the length of time Dr Friederiszick says he needs in his Reply statement. I cannot accept that submission which is unreal. In my judgment, in the real world, whatever additional disclosure were ordered (even if limited to what the bank defendants could produce by mid-February 2025), it would be unlikely that Dr Friederiszick could produce his report before July or August 2025 on his own timings. That would inevitably put back the dates for supplemental alleged suppression reports and a joint report probably to dates in October and November 2025. It would also affect the competition economics reports from the same experts and the quantum reports which are in part dependent on those reports. It follows that the service of experts’ reports would be unlikely to be completed until just before this 19 week trial is due to start in February 2026. As I said at the end of the hearing, an inevitable consequence of ordering the additional disclosure would be that the trial date would have to be adjourned, which is not something which any of the parties wants.

For all those reasons, the application for disclosure of additional data is dismissed save to the extent that NatWest and UBS have agreed to provide certain disclosure. This conclusion is reflected in the Order sealed on 16 December 2024 which I have already approved.