Cabo Concepts Limited v MGA Entertainment (UK) Limited & Anor.

The four-week trial in this matter was due to commence on 27 June 2022. On 6 June 2022, just 3 weeks before the trial, the Defendants (“MGA”) informed the court that they had missed something in the region of 84,000 documents during the data collection process underlying their disclosure. This led to an Order of 9 June 2022 by Bacon J (the assigned trial judge) adjourning the trial and providing for the service of evidence explaining what had gone wrong with disclosure and why MGA should not be required to pay the costs thrown away by the adjournment (“the June Order”).

The new trial is now fixed to commence on 1 October 2024, with an increased time estimate of six weeks.

Following the service of substantial quantities of evidence, I heard argument on 20 July 2022 as to the consequential matters arising from the adjournment of the trial. I made various orders at the hearing, including orders that essentially required the repetition of MGA’s disclosure exercise in conjunction with an independent e-disclosure provider, together with orders designed to set a new timetable leading up to trial (“the July Order”).

However, there was insufficient time to give judgment in respect of three important issues that were raised: first, whether the Claimant (“Cabo”) should be entitled to recover its costs thrown away by reason of the adjournment of the trial on an indemnity basis; second, whether MGA should be subject to an “unless order” in respect of compliance with its disclosure obligations arising under the July Order; and third whether Cabo should be entitled to an order for costs on account in relation to its costs thrown away by reason of the adjournment and, if so, at what level.

In respect of the first and third issues, I indicated at the end of the hearing that I would make an order for costs on an indemnity basis and that I was prepared to award costs on account of 45% of Cabo’s total costs incurred in preparation for the trial (those costs having been identified in the seventh witness statement of Mr Spector, Cabo’s solicitor, dated 14 July 2022 in the sum of approximately £1.3 million). These are my written reasons for that decision.

In respect of the second issue, I indicated to the parties that I wished to consider the issue further. Having now done that, I also set out my decision on that issue in this judgment.

In this matter, Cabo, a UK toy start-up, claims damages and declaratory relief against MGA, a leading supplier of toys around the world, in respect of alleged breaches of statutory duty (including abuse of a dominant position and unjustified threats of patent infringement proceedings) which are said to have caused the failure of Cabo’s business.

The nature of the claim and, in particular, the nature of the conduct of which Cabo complains, is of importance in understanding the significance of the disclosure failures that emerged shortly before trial. In a nutshell, Cabo alleges a secret anti-competitive campaign on the part of MGA to stifle the launch by Cabo of collectable toys marketed under the “Worldeez” brand which were likely to compete with one of MGA’s own blockbuster brands. Cabo only became aware of the conduct of MGA when certain emails dating back to May 2017 between Mr Laughton (senior vice president of MGA UK) and The Entertainer, a leading toy retailer in the UK, were passed to Cabo by representatives of The Entertainer.

Key disclosure custodians, including Mr Laughton and Mr Larian (MGA’s Chief Executive Officer), are said to have been the main protagonists in the collusive and unlawful conduct alleged against MGA, which unlawful conduct was principally implemented and/or recorded in emails sent by MGA and telephone calls made to leading toy retailers. It therefore appears to be no exaggeration to say that competition law infringements and unlawful threats under IP law are capable of being evidenced by individual emails. It is also no exaggeration to say that, with the exception of the emails that were passed to them by The Entertainer, Cabo had no visibility around the conduct of MGA. In the circumstances, it was inevitable that the proper conduct of the disclosure process by MGA would be of the utmost importance.

Against that background, the revelation shortly before trial that MGA’s disclosure exercise was defective was of the utmost concern to Cabo. As Mr Spector explains in his sixth statement dated 8 June 2022, Cabo reluctantly concluded that despite the passage of 5 years since the events with which its claim was concerned, there was now no prospect of a fair trial and an application to vacate the trial was necessary. Although MGA was initially ‘neutral’ in the face of this application, it ultimately consented, acknowledging that the trial had to be vacated because (as Fieldfisher explained in a letter to the court of 8 June 2022): “…it appears that, pending further explanation [as to the deficiencies in the disclosure exercise], the parties and the Court cannot have full confidence in the exercise conducted by [MGA]”.

Further to the June Order, a substantial volume of evidence has been served by MGA designed to identify and explain the deficiencies in its disclosure. I shall return to some of the detail in a moment, but for present purposes I note that it now appears to be common ground that approximately 40% of documents were missed by MGA at the harvesting stage (just over 1 million documents were harvested with something in the region of 800,000 documents having been missed), that nearly half of all potentially relevant documents were never even reviewed and that a number of warning signs were (inadvertently) overlooked. MGA presently believes that a major cause of the deficiencies was an indexing error in Microsoft Outlook when harvesting former employees’ emails, meaning that larger data sets, in particular, failed to filter correctly. MGA has also identified issues with the harvesting of current employee’s emails, albeit that the lack of any audit trail means that it has not been able to explain why the data was incomplete.

There is no suggestion that the deficiencies in disclosure were deliberate, but there is no question that they were serious. By way of example, the original document harvest for Mr Laughton produced 204,950 documents whereas a recently conducted re-harvest has produced 657,996, an increase of over 200%. The deficiencies led, at the eleventh hour, to the collapse of the trial and to Cabo finding itself in the unenviable position of having another two years to wait for determination of its claim.

Pursuant to the July Order, the harvesting exercise and application of keyword searches will be repeated by an independent e-disclosure provider engaged by MGA for that purpose. Any additional documents which the independent provider identifies as responsive to keyword searches (i.e. documents not already reviewed) will then be reviewed by Fieldfisher and disclosed to Cabo. Once the exercise has been completed, reports will then be filed by the independent provider and by Fieldfisher identifying the documents harvested and confirming compliance with the terms of the July Order. No doubt very substantial further costs will be expended on all sides.

I now turn to deal with the outstanding three issues requiring determination.

The court’s jurisdiction to make an award of costs on an indemnity basis arises from CPR 44.3(1)(b). Whereas costs on the standard basis must be proportionate and any doubts as to whether the costs were reasonably and proportionately incurred must be resolved in favour of the paying party, costs on the indemnity basis are not subject to the requirements of proportionality and any doubt as to whether costs were reasonably incurred must be resolved in favour of the receiving party. In deciding what order to make about costs, the court will have regard to all the circumstances of the case including the conduct of the parties (see CPR 44.2(4) and (5)).

Subject to one caveat, the principles to be applied by the court on an application for indemnity costs are not in dispute. They were clearly articulated in the leading authority of Excelsior Commercial & Industrial Holdings Ltd v Salisbury Hammer Aspden & Johnson [2002] EWCA Civ 879 by Lord Woolf CJ at [31]-[32] and by Waller LJ at [39]. In summary, the court has a “wide and generous discretion” in making orders about costs. An order for indemnity costs will be justified where either the conduct of the parties or “other particular circumstances” of the litigation (or both) are such as to take the situation “out of the norm”. As Lord Woolf observed, “that is the critical requirement”. In Esure Services Ltd v Quarcoo [2009] EWCA Civ 595, at [25], Waller LJ explained that the word “norm” was not intended to reflect whether what occurred was something that happened often so that in one sense it might be seen as “normal”, but was “intended to reflect something outside the ordinary and reasonable conduct of proceedings”.

As Lord Woolf CJ made clear in Excelsior, there are an infinite variety of situations which may justify the making of an indemnity order. It is not necessary for some sort of moral lack of probity or conduct deserving moral condemnation on the part of the paying party to be established. An award of indemnity costs is not penal but compensatory, the question in all cases being, what is fair and reasonable in all the circumstances of the case (see Catalyst Investment Group Ltd v Lewinsohn [2009] EWHC 3501 (Ch) per Barling J, citing Reid Minty (A Firm) v Taylor [2001] EWCA Civ 1723, at [20]-[28]).

The caveat to which I referred above arises because Ms Wakefield QC, on behalf of MGA, submits that if conduct is to justify an award of indemnity costs, it must be “unreasonable to a high degree” - see Noorani v Calver [2009] EWHC 592 (QB) per Coulson J (as he then was) at [8]:

“Indemnity costs are no longer limited to cases where the court wishes to express disapproval of the way in which litigation has been conducted. An order for indemnity costs can be made even when the conduct could not properly be regarded as lacking in moral probity or deserving of moral condemnation: see Reid Minty v Taylor [2002] 1 WLR 2800). However, such conduct must be unreasonable “to a high degree. ‘Unreasonable’ in this context does not mean merely wrong or misguided in hindsight”: see Simon Brown LJ (as he then was) in Kiam v MGN Limited No2 [2002] 1WLR 2810”.

Mr Bacon QC, acting together with Ms Kreisberger QC on behalf of Cabo, says that this is an inappropriate gloss on the test of “outside the norm”, that the Court of Appeal in Excelsior made it clear that they were moving away from such a test (as articulated in Kiam) and that insofar as it is a test that has since been applied, the courts have fallen into error.

On a careful reading of paragraphs [30]-[32] of Excelsior, I consider that Mr Bacon goes too far in his analysis. Lord Woolf CJ expressed the view in those paragraphs that he had “no difficulty” with the propositions identified in Kiam, in the context of that case.

Further and in any event, Teare J was faced with, and rejected, the same argument in Suez Fortune Investments Ltd v Talbot Underwriting Ltd [2019] Costs LR 2019. In that case, Teare J pointed out at [7] that there was a long line of authority to the effect that where it was said that conduct was unreasonable, it must be unreasonable to a high degree to justify an order for indemnity costs. At [8] he said this:

“It was suggested that the requirement that conduct must be unreasonable to a high degree was not stated in the CPR and that this gloss on the CPR was therefore wrong in principle. However, the requirement is, I think, a necessary corollary of the scheme of the CPR. Having regard to the importance ascribed to the principle of proportionality in the CPR, where unreasonable conduct is relied upon as justifying costs on the indemnity basis, and hence removing the need for the costs to be proportionate, the conduct must be unreasonable to a high degree. Otherwise due regard would not be had to the importance of proportionality in the scheme of the CPR”.

However, he went on to observe, having regard to the broad test in Excelsior, which requires the court to have regard to “all the circumstances”, that it would be wrong to say that indemnity costs are only appropriate where there is unreasonable conduct to a high degree, but (at [10]) that “where conduct is relied upon as justifying an order for indemnity costs it must be unreasonable to a high degree”. At [11] he concluded his analysis of the competing arguments in these terms:

“In the light of the wide nature of the discretion to order costs on the indemnity basis I accept the submission made by counsel for the Underwriters that there may be an “aggregation of factors” which justify an order for costs on the indemnity basis, one of which may be unreasonable conduct though not to a high degree. What matters is whether, looking at all the circumstances of the case as a whole, the case is out of the norm in such a way as to make it just to order costs on the indemnity basis. That is the approach in Excelsior; see also ABCI v Banque Franco-Tunisienne [2003] EWCA Civ 205 at para 70 per Mance LJ”.

With respect, I agree. I shall adopt the same approach in considering the submissions made in this case.

Ms Kreisberger relies upon five overarching themes which, taken together, justify an order for indemnity costs. She described these themes in her submissions as follows: (i) MGA’s original insistence on e-disclosure being conducted in-house; (ii) inadequate supervision of the e-disclosure process by Fieldfisher and its in-house document review provider, Condor Alternative Legal Solutions (“Condor”); (iii) the suite of technical failures that occurred during the e-disclosure process; (iv) the failure to identify “red flags”, i.e. indicators that the disclosure exercise was defective (Ms Kreisberger also described this as “turning a blind eye” to the deficiencies); and (v) the defective nature of the re-harvesting process which has been undertaken since the deficiencies in MGA’s disclosure exercise have come to light, together with MGA’s subsequent conduct in the lead up to this hearing.

MGA accepts that it must pay the costs thrown away as a result of the adjournment of the trial on the standard basis, but it contends that, whilst its original harvesting of documents was admittedly deficient, there is nothing in its conduct, or in the conduct of its advisers, which takes this case “out of the norm” of the ordinary and reasonable conduct of proceedings. Further it submits that there are no other particular circumstances to be taken into account.

I must therefore turn to consider each of the themes identified by Ms Kreisberger, although I shall deal with the first two themes together as they are effectively two sides of the same coin.

In my judgment, MGA’s disclosure exercise took the wrong course from the outset.

At the first CMC on 25 February 2021, Mellor J ordered that the Disclosure Pilot would apply and in the lead up to the second CMC the parties liaised over the approach to be taken by MGA to its disclosure exercise. In a letter dated 22 July 2021, Cabo confirmed that it was content for MGA’s in-house IT team to capture all documents from repositories and upload them onto the Relativity platform, but said that in circumstances where it was clear that MGA’s IT team had no knowledge or experience of disclosure in England “we still consider that there should at least be supervision from e-disclosure specialists in England who understand the legal requirements in this jurisdiction”. Cabo accordingly proposed a paragraph in the draft Order for the second CMC to this effect. This was resisted by MGA, which sought to address the point in a letter from Fieldfisher of 23 July 2021 by confirming that “…the disclosure process will be supervised by e-disclosure specialists in the UK who understand the relevant legal requirements in the jurisdiction. These will include both Fieldfisher lawyers and one or more e-disclosure specialists from Condor ALS, Fieldfisher’s in-house document review provider…”.

At the hearing of the second CMC on 27 July 2021, Cabo persisted in its request for independent supervision but (as is clear from the transcript of his judgment), having been shown the letter of 23 July 2021, Recorder Douglas Campbell QC accepted MGA’s submission that an order making provision for independent supervision was unnecessary, where he had been assured that UK e-disclosure specialists were going to supervise the process.

Regrettably, it appears to be common ground that, notwithstanding the assurance provided to the court on 27 July 2021, the technical process of harvesting documents was not subject to any such supervision and Ms Wakefield acknowledged during her submissions that there was no technological direction as to best practice:

This lack of supervision appears to lie at the root of the problems that then occurred, not least because, as it turns out, MGA’s in-house IT team did not have the necessary levels of experience or knowledge of best practice required for the conduct of a substantial disclosure exercise of this type, involving in excess of 1 million documents. Unfortunately Fieldfisher did not question the information they received from MGA as to the expertise of its IT team and nor did they instruct an e-disclosure expert (whether in the shape of Condor or anyone else) to consider the approach that MGA intended to take to the harvesting of documents and whether that approach was in accordance with best practice. In my judgment this would have been an obvious precautionary measure in a case where the client had no experience of English litigation, where the disclosure exercise was very substantial and where Cabo had already identified concerns around MGA’s in-house expertise and had put down a marker as to the importance of the disclosure exercise being conducted properly.

There was dispute at the hearing between the parties over the expertise of various members of MGA’s IT team, but in my judgment, Fieldfisher ought to have applied a more rigorous approach to examining the assurances from its client that it had the necessary expertise. It is apparent from the first witness statement of Mr Tiongco, that prior to 2017, MGA had always used external eDiscovery providers and that he only implemented an in-house eDiscovery system within MGA in 2017. Since then, MGA’s IT team had been running eDiscovery in house and had upgraded their software to the Relativity One platform, together with receiving training on Relativity administration and processing. However, as Mr Tiongco accepts, this case was the first UK e-disclosure exercise he had worked on and it was also the largest harvesting exercise he had worked on since MGA started to manage e-disclosure in house in 2017. As he also accepts, “[t]he only case larger in scale was the proceedings between MGA and Mattel in the US in about 2010, although the eDiscovery process there (including the harvesting of documents) was managed by an external third party” (emphasis added). Mr Tiongco worked together with Mr Key, a network engineer, who also accepts in his first statement that this was the first eDisclosure exercise he had conducted under UK law. The IT team’s work was overseen by Mr Laurence Cheng, since February 2021 MGA’s Senior Litigation Counsel. However, these proceedings were his first experience of litigation in the UK and the first time he had managed an eDisclosure exercise for MGA. He did not provide “any specific input” and he relied upon Mr Tiongco’s expertise to carry out the disclosure exercise and Fieldfisher’s advice as to compliance with disclosure obligations under UK law.

It is clear that had anyone with independent expertise in eDisclosure been engaged by Fieldfisher to consider the appropriateness of the approach that MGA was intending to take to the harvest of documents (including its use of Outlook software and its decision to apply a filter at the point of harvesting, contrary to the recommendation of Mr McSweeney), it would quickly have emerged that the approach proposed by MGA did not accord with best practice:

Ms Wakefield placed considerable reliance upon Mr Tiongco’s evidence that he had “no concern whatsoever” about MGA’s ability to carry out the disclosure exercise in house. However, to my mind (and in light of Mr Nikel’s evidence) this merely serves to highlight his lack of experience. Mr Nikel’s view is that “[i]t is obvious” that MGA are “not sufficiently experienced in the limitations of Microsoft Outlook and Microsoft 365 software and have not demonstrated the required level of attention to detail and necessary controls needed to perform such a significant discovery exercise”. In his tenth witness statement, Mr Pimlott says that he considers it to have been both reasonable and unremarkable in the circumstances of this case that Fieldfisher relied on the technical skills and expertise of MGA’s in-house e-discovery team, but I disagree. It would only have been reasonable if Fieldfisher had ensured (by obtaining independent external assistance from an experienced e-disclosure provider) that MGA’s internal IT team had both the expertise necessary to conduct the task and the understanding of best practice and quality control needed to carry it out properly to the standards expected by an English court.

In my judgment, the failure to ensure sufficient oversight of the e-disclosure process in circumstances where the court had been assured that such oversight would be provided (and so had declined to make the order for independent supervision as sought by Cabo), together with MGA’s insistence that the e-disclosure exercise should be carried out in-house, was unreasonable and out of the norm. The circumstances of this case rendered the disclosure exercise of acute significance to Cabo and I reject MGA’s attempts to characterise its failings as purely a series of inadvertent technical defects. To my mind that is to ignore the omissions made early in the process which left MGA’s insufficiently experienced, unsupervised, in-house team to conduct the disclosure exercise, using inappropriate and ill-advised methods which did not accord with, or apply, best practice. Insofar as may be necessary, I also consider that this conduct was unreasonable to a high degree.

It is common ground that various technical failings occurred during the disclosure process. I have already touched upon some in the previous section, but they appear to have included (i) the use of Microsoft Outlook, software which is not designed for a disclosure exercise of this sort and has various well-known limitations; (ii) the inappropriate use of Microsoft 365, a cloud-based communication and collaboration platform created by Microsoft which (whilst it is commonly used in e-disclosure) also appears to have limitations when it comes to searching data; (iii) the failure to follow guidance from Microsoft as to how to conduct date range searches in mailboxes held in Microsoft 365, a failure which led to the “Creation Date” field being used instead of the Sent or Received dates, leading to incorrect results (described by Ms Kreisberger as “a rooky error”); (iv) the decision to filter the data prior to upload in circumstances where MGA wrongly believed that full dataset exports would result in significant extra expense in data hosting charges; and (v) the failure to exercise appropriate levels of quality assurance and control.

These technical failings were inadvertent. From the evidence, MGA’s IT team believed that they had the necessary expertise and training to carry out the required e-disclosure exercise and they were not disabused of that belief, as they should have been, for the reasons I have already explained.

Nevertheless, technical failures of this magnitude are not the norm. The norm involves disclosure exercises being carried out without extensive and serious defects; the norm involves parties ensuring (particularly in a case involving such large amounts of documentation) that best practice is followed in conducting the harvesting of documents so as to avoid the risk of errors and so as to comply with their obligations under the CPR; the norm does not involve the collapse of a substantial trial at the last moment brought about by reason of the disclosing party not being in a position to confirm to the court that it can have full confidence in the disclosure exercise. To my mind such events fall outside the ordinary and reasonable conduct of proceedings.

Further and in any event, in conjunction with the lack of supervision and oversight to which I have already referred, the technical failures are part of a narrative of failings which should not have occurred if MGA and its advisers had approached the task of disclosure with appropriate care and attention. In that context, also, they are, in my judgment, out of the norm.

Cabo relies on a number of occasions when it says that the issues surrounding MGA’s disclosure exercise could and should have been spotted. It says that the failure to react to these “red flags” was “out of the norm”. MGA accepts that there were a small number of occasions when issues with its disclosure could, with hindsight, have been identified, but it contends that, while it is regrettable that these “red flags” were not followed up, reasonable decisions were made at the time.

During her submissions, Ms Kreisberger focused in particular on three examples, and I shall consider each of these in turn.

First, while preparing Mr Laughton’s statement, Mr Pimlott records in his eighth statement that Fieldfisher discovered that a key email from Mr Laughton to Mr Arora of B&M Stores, a toy retailer, which was referred to in paragraph 27(i) of the Particulars of Claim was not in MGA’s disclosure and had not been harvested from custodians. Mr Pimlott says in his evidence that Fieldfisher were “not unduly alarmed by this”, knowing that “disclosure is an imperfect process and errors occur”. He says that one errant email “did not set ringing alarm bells that the entire harvesting process was flawed”.

Nevertheless, Fieldfisher raised the issue in an email to MGA on 2 December 2021, a query they followed up again on 10 January 2022 when they received no initial response. Mr Key then responded that the missing email was a “mystery” to him and he has explained in his evidence that there was no way for him to know at what stage the email had been missed, that he had no reason to suspect any widespread issues with data collection but that he suggested in his reply that “the only way to verify the missing email would be to run the entire data collection process for Andrew Laughton again and to run a new search”. Mr Pimlott’s evidence is that he was “somewhat surprised by this” but nevertheless he took no further action. Instead, Fieldfisher’s response was to thank Mr Key for his explanation and to instruct: “No need for now to re-load his .pst files, but we will let you know if we need anything else in due course”. Mr Pimlott explains that he thought re-harvesting Mr Laughton’s data would not necessarily reveal the missing email if it had been deleted (he was aware of a litigation hold for relevant custodians but not that MGA had a general litigation hold in place that prevented deletion of all emails by the relevant custodians since 2010).

In his eighth statement, Mr Pimlott says that he has considered whether, if Fieldfisher had pressed to identify the missing email, the deficiencies in MGA’s disclosure exercise would have come to light sooner. He does not in fact answer that question, but instead seeks to justify his decision not to press the matter further, saying that he considered that a re-harvesting of Mr Laughton’s documents would be “disproportionate” and would unjustifiably divert his team’s resources at a very busy time in the case.

On balance, I consider this to be a surprising decision. It is obviously important not to apply hindsight to the conduct of Fieldfisher, who no doubt were still labouring under the impression that MGA had conducted a comprehensive disclosure exercise. Indeed in his tenth witness statement Mr Pimlott says that Fieldfisher had seen no signs of other documents missing, which he describes as “relevant context”.

However, in my judgment, the identification of an email (sufficiently important that it had been relied upon by Cabo in its pleading) that had been neither disclosed nor harvested from the relevant custodian should have caused Fieldfisher, at the very least, to re-run the data collection exercise for Mr Laughton. Given Fieldfisher’s responsibility to ensure the preservation of data, Mr Pimlott ought to have been well aware of the detail of the litigation holds put in place by his client and, had he been so aware, there would have been no question of forming the view that the re-run might not reveal the missing email.

In my judgment there was no sound basis on which Mr Pimlott could conclude that the missing email, described by MGA as “a mystery”, was of minimal significance and could effectively be “parked” without further investigation. There was no explanation whatever for it not having been captured in disclosure. It was plainly a red flag and it should have been investigated.

Second, as Mr Pimlott explains in his eighth witness statement, shortly before the Defence was served in the Autumn of 2020, an email was sent to Fieldfisher by Mr Laughton attaching a small number of emails relevant to the proceedings, some of which were privileged. It seems that at the time of disclosure, no cross check was undertaken to ensure that documents provided by MGA to Fieldfisher had been captured in disclosure (and the documents attached to Mr Laughton’s email were therefore not cross checked with the disclosure). Mr Pimlott only discovered this on 30 May 2022 in advance of the PTR when he asked a member of his team to check that the emails attached to Mr Laughton’s original email were in MGA’s disclosure. It transpired that not only should three of the attached emails have been disclosed, but they also could not be located in the overall pool of documents originally harvested.

Mr Pimlott explains in his evidence that when Fieldfisher was undertaking the disclosure exercise in these proceedings it considered whether to review documents provided by MGA at the time of preparation of the Defence. However, Fieldfisher took the view that this was not necessary in light of the Extended Disclosure Order and what they understood to be the “comprehensive ‘bottom up’ exercise tailored for the issues in these proceedings” that was being undertaken by MGA. Furthermore, Fieldfisher concluded that the documents provided to them were unsearchable such that any cross checking exercise would be slow and laborious. In the circumstances, a cross check would not be proportionate.

On balance, and again being careful not to apply hindsight, I do not regard this decision as particularly surprising. The counsel of perfection would have been to carry out a cross check, but where Fieldfisher thought that MGA was carrying out a comprehensive disclosure exercise which should capture all the documents provided at the time of preparation of the Defence, it is perhaps understandable that the decision was taken not to incur the additional costs of so doing. Seen against that background, this was certainly (as Ms Wakefield submits) a missed opportunity, but it was not an unreasonable decision. Fieldfisher’s confidence in the disclosure exercise was entirely misplaced for reasons I have explained, but Fieldfisher made a pragmatic decision based on its knowledge at the time.

Third, it has transpired during the analysis conducted after the June Order that a so-called “batching error” occurred at the time of the review by Fieldfisher of documents that were considered likely to be of particular importance (because they had responded to so-called “Worldeez” search terms). On 24 August 2021, Fieldfisher emailed MGA’s IT team to ask them to split the population of 1,282 documents responsive to these search terms into three batches, so that they could be reviewed by different team members. Mr Key attempted to do this using what he refers to as the ‘List Search’ function in Relativity, but, due to an overlap between the batches, 389 documents were inadvertently omitted. Mr Pimlott explains in his ninth witness statement that this error was not identified by his team, not least because there was a degree of overlap between the batches which resulted in the overall number of documents across the three batches appearing to be correct.

However, separately, Mr Key informed Fieldfisher that, in order to facilitate the equal split, he had to omit 10 documents from the batches. Mr Pimlott’s evidence is that his team cannot recall what, if any, steps were taken to review these 10 missing documents. He says this:

“If someone had looked back at the original Worldeez search terms saved search in order to find 10 additional documents, they undoubtedly would have realised that many more than 10 documents had not been reviewed and the full set of un-coded documents would then have been reviewed. The most likely explanation therefore is that, as an oversight, on completing their review of the three batches my team forgot that Mr Key had said that his three batches omitted 10 documents and believed that once those batches had been reviewed that the exercise was complete”.

Although this appears to have been an isolated incident relating only to the batching of 1,282 Worldeez documents, this admitted “oversight” in relation to so many potentially relevant documents was extremely unfortunate and plainly should not have occurred. It was particularly unfortunate given that the documents concerned were reviewed by Fieldfisher (rather than MGA) “on the basis that they were inherently more likely to be responsive” and indeed it seems that the 10 documents that Mr Key expressly identified as having been left out of the process were from Mr Larian.

Mr Nikel’s view is that in the conduct of this exercise no quality control or quality assurance process was being employed either by MGA or by Fieldfisher. His opinion is that this was a “red flag that was missed and seems to have been passed over by Mr Pimlott’s team”. I agree. Ms Wakefield submits that this issue has only recently been discovered and could not have resulted in the adjournment – however, I do not think it is possible to make such assertion without knowing what was in the missing documents. As Ms Wakefield realistically accepted, if they had contained new information which caused Cabo to apply for an amendment to its claim, then that could well have resulted in an adjournment to the trial. Further and in any event, the “batching error” is another example of a failure in the overall disclosure process conducted by MGA.

In my judgment, the failure to investigate the red flags in the first and third examples fell outside the ordinary and reasonable conduct of proceedings. Even if the failure to follow up those red flags is not conduct which, on its own, can be characterised as conduct unreasonable to a high degree, when combined with the conduct and the circumstances to which I have already referred, I again regard it as part of a narrative of failings relating to disclosure which plainly fall outside the norm.

Since the disclosure issues came to light, Fieldfisher has instructed MGA to conduct a full re-harvest of all its custodians’ documents. Instead of using an external e-disclosure provider, the exercise has again been conducted in-house by Mr Tiongco and Mr Key, albeit that it was partially observed by Mr Seigle-Morris. A very substantial number of new documents have been harvested.

Cabo points to various flaws in this process, identified by Mr Nikel in his report, including that a field called ‘Creation Date’ was used in an attempt to filter existing employees’ emails but did not work; that a proper audit trail is still lacking; that three emails from Mr Laughton have still been missed (albeit that Mr Seigle-Morris regards such discrepancies as “not uncommon” in the context of the total of 853,155 messages that were identified); and that a consistent method of deduplication may not have been applied.

In response to Mr Nikel’s report, Mr Pimlott’s tenth statement dated 18 July 2022 confirms that the harvesting exercise has been redone and asserts that “Mr Alex Seigle-Morris, an independent eDisclosure specialist at TLS has overseen this process and has applied quality control checks to verify the robustness of the harvest. He has expressed the view that the re-harvest is robust”. Unfortunately, within 24 hours, this assertion was shown to be inaccurate. In a third statement from Mr Seigle-Morris served on 19 July 2022, he directly contradicts Mr Pimlott, saying this:

“…I acknowledge that the data collection and processing carried out by MGA’s IT team for the re-harvest was not in accordance with best practice for e-disclosure. I did not have oversight over MGA’s export of the re-harvested data or de-duplication. The only action taken by MGA’s IT team of which I had direct sight was when we tested the ‘Created Date’ field to cull documents…”

Mr Seigle-Morris also confirms that in light of information from MGA to the effect that they did not include unindexed items when they ran the exports of current employee mailboxes from Office 365, “I can no longer say with certainty that the data provided for the current employees constitutes a comprehensive set…”.

Mr Seigle-Morris’ evidence in his third statement led to an eleventh witness statement from Mr Pimlott dated 19 July 2022 retracting, albeit in unapologetic terms, the sweeping statements he had made in his tenth statement. Save for his reference to an early draft of Mr Seigle-Morris’ third statement, no explanation is provided as to how he could have gained such an erroneous understanding of the level of Mr Seigle-Morris’ oversight or of Mr Seigle-Morris’ views as to the robustness, or otherwise, of the re-harvest.

Ms Kreisberger prays this late evidence from MGA in aid in support of the proposition that MGA has never really understood the seriousness of its failures on disclosure, that it has shown no real contrition and that it has made no proper attempt constructively to progress the litigation. She submits that it was Cabo’s enquiries in advance of the trial that led to the discovery of the deficiencies in disclosure. She also relies on the fact that MGA left it to Cabo to apply for an adjournment once the “bombshell” had dropped over disclosure and that thereafter MGA acted unreasonably in refusing an extension of time to Cabo for service of its evidence for this hearing (notwithstanding that Cabo had granted an extension of time of 10 days to MGA for the service of MGA’s evidence), necessitating the filing of an application for an extension of time by Cabo. Finally she says that MGA inappropriately threatened an adjournment in the lead up to this hearing and ignored a direction given by Bacon J on 15 July 2022 to the effect that all outstanding issues would be resolved at the hearing on 20 July.

Ms Wakefield opposes this characterisation of events, pointing out that Mr Pimlott explains how the deficiencies in disclosure came to light in his eighth witness statement and that this was not prompted by Cabo. She disputes the suggestion that MGA has been lackadaisical in its dealings with Cabo or with the court and she contends that MGA has sought at every turn since discovery of the disclosure issue to keep the court updated. She accepts that the re-harvest has not been done in accordance with best practice but she contends that the position has now been reached where all mail boxes of former employees have been uploaded and “it is very probably the case” that this also applies for existing employees; essentially she says that MGA has “worked tirelessly” to rectify the deficiencies in disclosure and that I must take this into account in the exercise of my discretion on costs. Ms Wakefield submits that there was no tactical attempt to avoid this hearing but that there was concern at the prospect that Cabo would serve expert evidence which MGA would not have time to answer, hence the suggestion that MGA would seek an adjournment (albeit that in the event the decision was made to go ahead with the hearing).

These various submissions from the parties were based on extensive references to correspondence and emails in the lead up to the hearing and, in MGA’s case on reference to a “Timeline” setting out how MGA’s disclosure issues were discovered and the correspondence that then ensued. However, the import of the correspondence and the accuracy of the Timeline was hotly disputed.

In the end I do not consider it to be necessary to resolve each and every one of the numerous complaints made by Cabo against MGA under this particular heading – it is not unusual for parties to disagree about conduct in the lead up to a hearing, and this case is no different. Furthermore, I am not convinced that ultimately it much matters which party finally prompted the identification of the failings in MGA’s disclosure exercise; the important thing is that they were identified and are now going to be addressed pursuant to the terms of the July Order.

However, I accept (as does MGA) that the re-harvesting exercise has been deficient and, in the circumstances, I cannot see that I can properly take it into account as weighing heavily in the balance against an order of indemnity costs, as MGA invites me to do. True it is that MGA has sought quickly to put matters right, but regrettably it has not taken the right course in doing so; having fallen into error in not instructing an independent e-disclosure provider first time around, I would have expected that once deficiencies of this scale were identified, MGA would then have appreciated the need to retain such a provider to proffer advice and supervision in the context of the repeated exercise.

Mr Pimlott appears to have understood the importance of this, making a point of confirming that Mr Seigle-Morris had overseen the re-harvest and had applied quality control checks to verify its robustness in his tenth statement. Unfortunately however, this evidence turned out to be inaccurate and it appears to me to be illustrative of a lack of proper attention to detail, not least given what has gone before. Despite the obvious significance of this issue, Mr Pimlott provides no adequate explanation for this erroneous evidence and it is hard to understand how the partner with joint conduct of the proceedings on behalf of MGA could have made such an error, regardless of the pressurised circumstances in which Mr Pimlott was no doubt operating.

In my judgment, this conduct (by which I mean the deficient re-harvesting and the subsequent evidence of Mr Pimlott) is perhaps best seen as part of a continuing failure adequately to grapple with the need for proper supervision and oversight of the disclosure process. Belatedly Mr Pimlott appears to have appreciated that need, but failed to put in place adequate measures to provide for it. I consider that this conduct was unreasonable (and unreasonable to a high degree) and separately, or in conjunction with the other matters I have identified, it was out of the norm.

For the reasons set out above and looking at all of the circumstances of the case as a whole, I consider that MGA’s conduct in connection with the disclosure exercise was out of the norm in that it was outside the “ordinary and reasonable conduct of proceedings” (a formulation expressly applied by Newey LJ in Whaleys (Bradford) v Bennett [2017] EWCA Civ 2143, at [22]). Accordingly it is just to make an order for indemnity costs.

Cabo seeks an order that, unless MGA complies with the terms of the July Order as to disclosure, MGA’s Defence will be struck out and Cabo shall be entitled to enter judgment.

It is common ground that the court has a broad discretion to make conditional, or “unless” orders under CPR 3.1(3). That jurisdiction was set out by Bryan J in JD Classics v Hood [2021] EWHC 3193, at [135]-[142].

In summary, the broad jurisdiction to make conditional orders is “to enable the court to exercise a degree of control over the future conduct of the litigation”. Thus CPR 3.1(3) is concerned “with the basis on which the proceedings will be conducted in the future and that remains the case even when the condition is imposed in order to make good the consequences of some kind of previous misconduct” (Huscroft v P&O Ferries Ltd [2011] 1 WLR 939 at [17] per Moore-Bick J). Because a conditional order striking out a statement of case is “one of the most powerful weapons in the court’s case management armoury”, it should not be deployed unless its consequences can be justified and the court should consider carefully whether the sanction is appropriate in all the circumstances of the case (per Moore-Bick J in Marcan Shipping (London) Ltd v Kefalas [2007] 1 WLR 1864 at [36]). Accordingly, “before exercising the power, the court should identify the purpose of imposing a condition and satisfy itself that the condition it has in mind represents a proportionate and effective means of achieving that purpose” (Huscroft at [19]). The court is “entitled to take into account the effect of making or not making the order sought on the overall fairness of the proceedings and the wider interests of justice as reflected in the overriding objective” (JSC BTA Bank v Ablyazov [2010] EWHC 2219 (QB) at [38]).

The failures on which Cabo relies in seeking this order have been addressed in the context of my discussion on the subject of indemnity costs and for that reason, I can deal with this issue quite briefly.

In short, Cabo contends that by reason of (i) MGA’s failures in relation to its initial disclosure exercise; (ii) its flawed attempts to re-harvest its documents; and (iii) its subsequent failure to conduct itself constructively, this is an appropriate case for an unless order. Ms Kreisberger points to the “human cost” of the lost trial, namely that the individuals who founded Cabo served their Particulars of Claim approximately 2 years ago, have had to suffer the loss of their business and the stress of legal proceedings in the meantime, only then to find that, through no fault of their own, they have been denied a trial within a fair space of time but must now wait for a period of another two years for their final reckoning.

Whilst I have every sympathy with Ms Kreisberger’s client, I do not consider this to be an appropriate case for an unless order. Having regard to the guidance in the authorities identified above, I do not consider MGA’s past failures in relation to disclosure (both in respect of the initial exercise and the recent re-harvesting) to justify such an order when seen in the context of the terms of the July Order. That order provides for the appointment of an independent e-disclosure provider by MGA together with the subsequent undertaking by that provider of all technical aspects of a re-harvesting exercise (as set out in Schedule 1 to the July Order). There is no reason to suppose that an independent provider will fail to carry out that exercise properly or that there will be any further cause for complaint on the part of Cabo. The review of all new documents (not previously reviewed) will be carried out by Fieldfisher and not by MGA. Cabo has not suggested that it has no faith in such a review.

Furthermore, in the context of this application, it seems to me to be of importance that the failures on the part of MGA were not deliberate and that, in light of the terms of the July Order, there is no reason to believe that they will occur again. Cabo has not suggested otherwise. Whilst there has been debate before me as to the conduct of MGA in the lead up to this hearing, I do not consider that I need to make any findings about that conduct, save that I accept that MGA has sought to provide evidence to the court to explain the deficiencies in its disclosure (as it was required to do by the June Order) and has shown itself to be anxious to remedy the deficiencies (albeit to date in a somewhat misguided fashion). I do not consider any of MGA’s conduct to be consistent with an intention to flout the July Order and thus I do not consider it necessary to impose an unless order in order to ensure compliance with the July Order. Given that the trial has been adjourned until 2024, there is no question of making an order that is designed to protect the trial date and Cabo does not suggest that it is needed for that purpose.

Of course I have regard to the individual circumstances of Cabo’s founders, but in my judgment those circumstances, whilst unfortunate, do not make it proportionate or fair to make an unless order given all the other circumstances of this case to which I have referred. If, following further disclosure being provided by MGA, any issues arise with that disclosure, these will need to be dealt with in the ordinary way, by requests for specific disclosure and the like. I agree with Ms Wakefield that it would be disproportionate for MGA to find itself subject to the possibility that its Defence could be automatically struck out in the event that a further issue on disclosure were to arise.

Accordingly, I reject the application for an Unless Order.

Finally I now turn to set out my reasons for ordering that Cabo should have its costs on account in the sum of 45% of its total legal costs incurred in preparation for the trial.

It is common ground that the Court should not undertake a summary assessment of Cabo’s costs thrown away under CPR 44.6(1)(a). If the Court orders a detailed assessment under CPR 44.6(1)(b), then rule 44.2(8) provides that the court will order MGA “to pay a reasonable sum on account of costs, unless there is good reason not to do so”. There is accordingly a presumption that a payment on account of a reasonable sum will be made, subject to an exception.

It was accepted by Mr Bacon on behalf of Cabo during the hearing that “costs thrown away” does not, as a matter of principle, mean the costs of the entirety of the work done to date, or in all of the preparations for the trial. On the contrary, it is only the cost of the work that has been done and which will have to be repeated for the relisted trial which will be recoverable (Fern Trading v Greater Lane [2021] EWHC 1939 (Comm) per HHJ Pelling QC at [28]). If an element of costs incurred remains for the benefit of a party at a subsequent hearing it will not have been thrown away.

The assessment of the quantum of costs thrown away will ordinarily be an exercise for a costs judge who will have before him or her the information necessary to decide which element of the costs has been thrown away by the adjournment of the trial. However, CPR 44.2(8) requires me to order the payment of a reasonable sum unless there is a good reason not to do so.

Cabo advanced evidence in Mr Spector’s seventh statement in support of this application, which asserted that its total wasted costs amounted to £1,285,431.49, being £977,000 for counsel’s brief fees, £245,290 for SCW’s fees of trial preparation and £63,141.49 for miscellaneous disbursements including expert fees, Veritas witness training and Opus2 bundling. Mr Spector said in terms in his statement that all of these costs incurred in the lead up to trial “are now thrown away because the trial has had to be vacated”. However, as I have already indicated, this stance was not (perhaps unsurprisingly) maintained at the hearing. Indeed Mr Bacon acknowledged the principles I have referred to above and accepted in terms that the full quantum of Cabo’s costs should not be the benchmark for an order. He also indicated that Mr Spector now accepts that he has overstated the position in his statement.

Against that background, Ms Wakefield says that were it not for the approach adopted by Cabo, she would have accepted the principle of a payment on account and the argument would have been only as to the appropriate amount. However, she takes great exception to Mr Spector’s seventh statement, pointing out that it is “simply not true”, that Cabo should have known it was not true and that it is difficult not to form the view that its purpose was as an “opening gambit” in a negotiation rather than as realistic and credible evidence. She submits that this is inappropriate in the context of an application of this sort, that it has the effect of inflating the quantum of the payment on account that Cabo is seeking to recover from MGA (Cabo seeks 75% of the total costs incurred in its skeleton argument) and that this is itself “a good reason” not to award costs on account.

I have considered this submission with care, and there is no doubt that (as has been accepted) Mr Spector’s seventh statement overstates the position. This should not have occurred. It means that the court does not have a genuine or realistic estimate from Cabo as to the level of costs that have been wasted by reason of the adjournment of the trial.

However, on balance, in circumstances where Mr Spector has acknowledged his error, and given that his evidence does at least establish the total amount of costs incurred by Cabo in the lead up to trial, I do not consider that I should take the approach of refusing to make any order. It seems to me that this would be contrary to the overriding objective of dealing with cases justly, including compensating innocent parties who are not to blame for the adjournment of a trial. A refusal to make an order for costs on account at this stage would mean that Cabo would be kept out of those wasted costs for at least another two years. In circumstances where the adjournment of the trial has occurred by reason of events that are not of Cabo’s making, I consider that would be an extremely harsh outcome.

Furthermore, I consider that where the rules require me to identify “a reasonable sum”, I am in a position to do that having regard to the evidence.

Cabo has incurred very substantial costs in the lead up to trial (as evidenced by Mr Spector’s statement) and where that trial was adjourned with less than three weeks to go before it was due to commence and a new trial will not be taking place for a further two years, it is plainly inevitable that a substantial proportion of those costs (which I have determined should be awarded on an indemnity basis) will be wasted:

Doing the best I can in all the circumstances, which include that there has been no detailed assessment such that there is a significant degree of uncertainty as to the amount of costs that have been wasted by reason of the adjournment, any sum I identify will have to be an estimate. Christopher Clarke LJ arrived at a similar conclusion in Excalibur Ventures LLC v Texas Keystone Inc [2015] EWHC 566 (Comm) at [23]-[24]:

“[23]…Any sum will have to be an estimate. A reasonable sum would often be one that was an estimate of the likely level of recovery subject…to an appropriate margin to allow for error in the estimation. This can be done by taking the lowest figure in a likely range or making a deduction from a single estimated figure or perhaps from the lowest figure in the range if the range itself is not very broad

[24] In determining whether to order any payment and its amount, account needs to be taken of all relevant factors including the likelihood (if it can be assessed) of the claimants being awarded the costs that they seek or a lesser and if so what proportion of them; the difficulty, if any, that may be faced in recovering those costs; the likelihood of a successful appeal; the means of the parties; the imminence of any assessment; any relevant delay and whether the paying party will have any difficulty in recovery in the case of any overpayment ”.

Having regard to that guidance, I consider that a payment on account of 45% of the total sum incurred in costs by Cabo in respect of the aborted trial, namely £578,444.17, being 45% of £1,285,431.49, is a reasonable sum.

In particular I have regard to the fact that the costs of taking this case to trial were considerable and that it is inevitable that a very substantial part of those costs, which I have awarded on an indemnity basis, has been wasted by reason of the late adjournment (as discussed above). Owing to the nature of Cabo’s evidence I have been forced to estimate the full extent of the wasted costs but I have little doubt that they will exceed, probably to a significant extent, the figure of £578,444.17 that I have ordered. For convenience I have taken a percentage of the total costs figure, but I have done so only with a view to identifying a reasonable figure having regard to the likely level of recovery of wasted costs. MGA is well able to pay that figure by way of costs and no concerns have been expressed by MGA as to the potential for it to be unable to recover those costs in the unlikely event of an overpayment. Whilst an attempt might be made to appeal my decision on indemnity costs, MGA has accepted the principle that it is required to compensate Cabo for the costs thrown away by the adjournment and I consider that the figure I have ordered provides an appropriate margin for error.

I would like to express my gratitude to all counsel for their extremely helpful written and oral submissions.