BTI 2014 LLC v PRICEWATERHOUSECOOPERS LLP & Anor

This is my judgment on the application (the “Application”) of the claimant (“BTI”) for permission to re-amend its Amended Particulars of Claim. BTI made the Application on 13 April 2022, supported by evidence in the form of a sixth witness statement of Mr Kevin Lloyd, a solicitor acting for BTI, expert evidence in the form of a skeleton outline expert report and addendum prepared by Mr Jimmy Daboo, an expert on audit practice and a skeleton outline expert report prepared by Dr Adam Love, an expert in environmental science. The First Defendant (“PwC”) objected to BTI’s application, supporting its objection by reference to evidence in the form of fifth and sixth witness statements of Mr Charles Hewetson, a solicitor acting for PwC.

By s35 of the Limitation Act 1980, any new claim made in the course of any action (e.g. by way of amendment to an existing statement of case) is deemed to have been commenced on the same date as the original action. Without more, that “relation back” rule would enable litigants to sidestep limitation rules by amending existing claims instead of making new claims. For that reason, s35(3) restricts the ability of litigants to add new claims after the expiry of a limitation period but makes an exception where such an addition is permitted by rules of court. Section 35(4) and s35(5) set out the permissible scope of rules of court in this regard. Broadly, if an amendment involves the raising of a new cause of action, it is permissible only if the new cause of action arises out of the same facts or substantially the same facts as are already in issue.

BTI commenced these proceedings within the applicable limitation period, but since it issued those proceedings, the limitation period has expired. In those circumstances, the parties were agreed that between them CPR 17.1(2), 17.3 and 17.4 require me to address the following issues:

The parties were agreed that the limitation period has expired, so Issue 1 was not in dispute. Issues 2 and 3 depend on the nature of the amendments that BTI seeks to make which I address in the next sections. Issue 4 also involves consideration of the overall history of this and other relevant litigation which I will set out later in this judgment.

The Application arises in the context of complex long-running litigation. BTI is seeking damages from PwC in respect of what it asserts to be PwC’s professional negligence relating to its audit of the financial statements of the second defendant (“Windward”) for the periods ended 31 December 2007 and 31 December 2008 (the “2007 Accounts” and the “2008 Accounts” respectively). BTI brings this claim in its capacity as assignee of Windward’s rights of action against PwC.

In 2001, Windward was a member of a group of companies that included Sequana S.A (“Sequana”). Sequana was the parent company of Windward and, through Windward, of Appleton Papers Inc (“API”). API, together with an unconnected company, NCR Corporation (“NCR”) were designated, along with other entities as Potentially Responsible Parties (“PRPs”) under the US Comprehensive Environmental Response, Compensation and Liability Act (“CERCLA”) in relation to contamination of the Fox River with polychlorinated biphenyl, a toxic substance used to manufacture a brand of carbonless copy paper. The liabilities of API, NCR and other PRPs under CERCLA comprised clean-up costs (“Remediation Costs”) and natural resource damages (“NRDs”).

For the purpose of negotiation with the US Government and other PRPs, API and NCR were grouped together and treated as a single entity responsible together for one share of Remediation Costs (the “NCR/API Share”). That treatment left unresolved the question of how NCR and API would divide up liability for the NCR/API Share as between themselves. Matters were further complicated by the fact that the BAT group of companies agreed to contribute towards the NCR/API Share of Fox River Remediation Costs.

Windward sold its shares in API in 2001. As part of that transaction, Windward gave certain indemnities in respect of API’s environmental liabilities (“Indemnity Liabilities”). The precise way in which the various liabilities were carved up is complicated but much of the detail is not necessary for the purposes of the Application. It is sufficient to note that the extent of Windward’s Indemnity Liabilities would depend on all the following factors:

The factors mentioned in paragraph 8 were not the only constituent elements in the determination of Windward’s Indemnity Liabilities. For example, NCR and API’s liability for NRDs as well as for Remediation Costs was relevant to the calculation. Moreover, NCR and API also had some potential liability for contamination of the Kalamazoo River which also drove the calculation of Indemnity Liabilities. The factors listed in paragraph 8 are simply those that are relevant in the context of the Application.

Windward purchased a fixed sum insurance policy (the “Maris Policy”) that provided it with protection of up to USD 250 million in respect of Indemnity Liabilities.

Windward’s 2007 Accounts included a provision for Indemnity Liabilities in relation to the Fox River of EUR 59.3m in excess of the Maris Policy’s value. That provision was based on (i) an estimate that Total Remediation Costs for the Fox River (excluding those relating to a part of the river known as “OU1”) would be USD 662.3m or approximately USD 623m excluding long-term monitoring costs and (ii) an estimate that the NCR/API Share would be around 60%.

Windward’s 2008 Accounts included no provision for Indemnity Liabilities relating to the Fox River in excess of the Maris Policy’s value. That nil provision was based on an estimate that (i) Total Remediation Costs for the Fox River would be around USD 640.3m and (ii) the NCR/API Share would be around 38%. This reduction in the estimate of the NCR/API Share was justified, in part, by an assumption that NCR and API would be able to avail themselves of a “divisibility defence” under CERCLA on the basis of the judgment of the US Supreme Court in Burlington Northern that was released just a few weeks before PwC completed its audit of the 2008 Accounts.

In broad summary, a “divisibility defence” under CERCLA required two limbs to be satisfied. First it had to be shown that a particular harm is theoretically capable of apportionment. Second the PRP seeking to avail of the defence must establish a reasonable basis for an apportionment on the facts such that the wrongdoer is liable only for the harm that it caused. Therefore, successful deployment of a “divisibility defence” would involve both questions of law and questions of fact. The questions of fact had the potential to be complicated, and dependent on expert scientific and engineering input.

In this litigation, BTI (as Windward’s assignee) asserts that PwC was negligent in its audit of the 2007 Accounts and the 2008 Accounts. It asserts, among other matters, that the provision for Indemnity Liabilities in both sets of accounts was inadequate and that the disclosure in both sets of accounts of the ranges of possible outcomes that was required by FRS 12 was also inadequate. As regards the 2008 Accounts, BTI alleges that PwC acted negligently in relying on the Burlington Northern judgment as a justification for a reduced estimate of the NCR/API Share.

BTI asserts that, but for PwC’s breaches of duty, Windward would not have paid substantial dividends in 2008 and 2009 because it would have concluded that it had insufficient profits available for distribution.

BTI claims damages for PwC’s alleged negligence in its audit of the 2007 Accounts and the 2008 Accounts. Two general categories of breach of duty are alleged:

As would be expected with a complicated claim such as this, the pleadings are lengthy. Relevant background facts, duties, breaches of duty and matters relating to causation are pleaded separately. I will not attempt to summarise the entirety of the APC and proposed RAPC, but rather focus on essentials which include the following so far as material for present purposes:

Miles J considered the scope of BTI’s existing pleadings as set out in the APC in the course of a case management conference on 3 March 2022 (the “CMC”). I have been provided with a copy of Miles J’s judgment given following the CMC (the “Miles J Judgment”). During the CMC, BTI sought permission to rely on expert evidence in the field of remediation practice (i.e. in the field of environmental science or engineering relating to liability under CERCLA). BTI justified that request by submitting that such expert evidence would be relevant to both the level of Total Remediation Costs that could have been expected at the time PwC were performing the audit of the 2007 Accounts and the 2008 Accounts and also to whether the liabilities of NCR/API could be regarded as “divisible” (see [13] of the Miles J Judgment).

Miles J refused permission to rely on expert evidence. He concluded that BTI’s pleading in the APC to the effect that PwC had failed to obtain appropriate independent expert advice extended only to legal advice (see [20] to [24] of the Miles J Judgment). Since only a failure to obtain legal advice had been pleaded, he concluded that the question of what an independent technical expert (i.e. an engineer or environmental scientist) would have said did not arise on the pleadings. He also decided that the likely advice of a technical expert would not, on the pleadings as set out in the APC, be relevant to “the counterfactual question of causation” or “complaints about the performance of [PwC] of its audit duties”.

PwC has consented to certain amendments. The controversial amendments can be divided into two categories to which I refer as the “Divisibility Amendments” and the “Total Costs Amendments”.

The Divisibility Amendments relate to the 2008 Accounts only. Whereas the 2007 Accounts had assumed that the liability of NCR and API in respect of the Fox River would be joint and several with the other PRPs, the 2008 Accounts assumed the opposite, namely that NCR and API would be able to rely on a divisibility defence under CERCLA.

By the Divisibility Amendments BTI seeks to plead elements relating to the “factual” component of the divisibility defence by contrast with the case pleaded under the APC which Miles J held related only to the legal aspects of that defence. For example:

The Divisibility Amendments therefore focus on the likely availability of a divisibility defence in the light of Burlington Northern, go to the proper estimation of the NCR/API Share in the 2008 Accounts only and do not relate to the 2007 Accounts. The Divisibility Amendments make no assertions about Total Remediation Costs.

By contrast, the Total Costs Amendments, relate to both the 2007 Accounts and the 2008 Accounts and involve new assertions being made to the effect that PwC underestimated Total Remediation Costs in both sets of accounts. The Total Costs Amendments include the following:

It is convenient to set out what I consider to be some differences between the Divisibility Amendments and the Total Costs Amendments. At this stage, I set out what I see as the differences at a high level. Later I will consider the significance of these differences when I come to apply the necessary legal tests in deciding whether to allow the Application.

In the APC BTI already asserts that a divisibility defence was unlikely to support the provision made for Indemnity Liabilities in the 2008 Accounts. That assertion relates to the adequacy or otherwise of the estimate of the NCR/API Share which is extensively pleaded in the APC. The APC already asserts that PwC should have obtained external expert advice on matters relating to a divisibility defence and Burlington Northern. The Miles J Judgment has determined that this pleading extends only to legal advice and neither party invites me to reach a different view. Nevertheless, the factual component of a divisibility defence is referred to specifically in the APC and, in any event, is referenced in the provisions of CERCLA itself. To a significant extent, the Divisibility Amendments deal with counterfactual matters relating to loss and causation arising out of breaches already pleaded in the APC. They seek to explain (i) that if a US lawyer had been consulted, he or she would have advised that, notwithstanding Burlington Northern, a divisibility defence could not succeed without a factual underpinning, which had to be provided by expert remediation advice as to a reasonable basis of apportionment and that (ii) if a remediation expert had been consulted, he or she would have advised that the requisite factual underpinning was not present. To that extent, the Divisibility Amendments relate back to already pleaded allegations. However, the Divisibility Amendments go further since they plead, in addition, that irrespective of what legal advice might have been forthcoming if PwC had requested it, a competent audit would involve PwC obtaining expert factual remediation advice where no such assertion has been made previously.

By contrast, in the APC, BTI makes no express assertion that there was any flaw in the estimate of Total Remediation Costs used for the purposes of provisioning, although it does allege flaws in the estimate of the NCR/API Share. Moreover, no express assertion is made that PwC needed any expert opinion on the level of Total Remediation Costs. (Some paragraphs of the APC allude to Total Remediation Costs: for example at paragraph 225.5 of the APC it is alleged that PwC failed to ask Mr Gower the right questions about “the estimate of the Fox River liability generally”. However, that goes to the enquiries that PwC should have made of management and does not suggest a need for validation by expert advice). The Total Costs Amendments are less rooted in the current express pleadings as regards the averred inadequacy of the provision for Indemnity Costs than are the Divisibility Amendments. Later in this judgment, I address BTI’s assertion that there is nevertheless a clear link with the pleadings relating to disclosures required under FRS 12.

The Court of Appeal in Mulalley & Co Ltd v Martlet Homes Ltd [2022] EWCA Civ 32 (Mulalley) has recently given guidance both as to the test for deciding whether a particular amendment pleads a new cause of action and if so, whether that new cause of action arises out of the same or substantially the same facts. In their judgment, the Court of Appeal affirmed previous guidance given in Co-operative Group Limited v Birse Developments Ltd [2013] EWCA Civ 474 (“Co-op”). I derive the following principles from the authorities:

Those, I consider are the applicable principles. Both parties referred in their written skeleton arguments and oral submissions to a number of other authorities. While it is, of course, instructive to see how the principles have been applied in different situations, it is important not to elevate into a further principle something which is, on closer inspection, simply an application of an existing principle. I can quite accept that, if the RAPC establishes an alternative route to PwC’s liability in the event that already-pleaded claims fail, that may be an indication, perhaps a strong indication, that the RAPC pleads a new cause of action (see [65] to [66] of Mulalley). However, there is no principle that any amendment to a pleading that establishes a new route to liability necessarily involves a new cause of action. Similarly, if a new breach of an already pleaded duty results in a new counterfactual case that may be of significance particularly where that counterfactual case results in a different basis on which damages would be calculated (see [67.4] of Hyde v Nygate [2019] EWHC 1516 (Ch)). However, there is no principle to the effect that the pleading of a new counterfactual necessarily involves the pleading of a new cause of action. Finally, I can accept that it may be instructive in appropriate cases to ask whether, if there were no question of limitation and the proposed amendments were the subject of a fresh action after conclusion of the first action, PwC could successfully say that the new claim is res judicata (see Murray Film Finance Ltd v Film Finances Ltd [1996] EMLR 539). That may help to inform the necessary focus on the essentials of the amendment. However, the principles I have to apply remain those set out in paragraph 28 and not those of res judicata.

Finally, BTI submits that I must consider the material facts “at the highest level of abstraction” relying on the statement of Millett LJ at 405f of Paragon Finance plc v DB Thakerar & Co [1999] All ER 440. I agree, provided it is clear what is meant by this phrase. As elucidated at [24] of Co-op I must “abstract” the relevant facts from the APC and the RAPC and the reference to the “highest level” simply emphasises the need to focus on essentials.

Miles J has already determined at the CMC that BTI’s case on the divisibility defence as currently pleaded asserts only a failure by PwC to obtain US legal advice as distinct from factual remediation advice. Neither party has invited me to reach any different conclusion for the purposes of the Application.

As I have noted in paragraph 26 above, to an extent the Divisibility Amendments plead consequences that would flow from PwC’s already pleaded failure to obtain US legal advice on a divisibility defence. I do not regard those aspects of the Divisibility Amendments as pleading any new cause of action. BTI already pleads in the APC that there is a significant factual dimension to any successful divisibility defence (see, for example paragraph 147 of the APC and the statement in paragraph 243.4.9 that the availability of a divisibility defence required a “fact sensitive analysis”). The unamended APC already asserts in paragraphs 243.4.1 and 243.4.9 that, despite Burlington Northern, there was a high likelihood that the NCR/API Share would be much higher than the 38% estimated in the 2008 Accounts. Therefore, to a significant extent the Divisibility Amendments provide further reasons why that is the case. I do not consider that in pleading further consequences that flow from the failure to take US legal advice, BTI is making assertions of fact that are, having regard to essentials, more extensive than those already made in the APC.

PwC are, however, correct to point out that the Divisibility Amendments go further than pleading consequences of a failure to take purely legal advice by pleading a separate breach of an obligation to take factual advice. That, PwC argues, is a new cause of action because it provides an independent route to establishing PwC’s liability which involves the investigation of a new counterfactual, namely what any factual advice would have said if obtained. That submission has given me pause for thought, but I have concluded that it does not properly focus on the essentials of the facts pleaded in the APC and the RAPC.

BTI’s essential case set out in the APC is that, in order properly to perform its audit of the 2008 Accounts, PwC needed expert advice on the availability of a divisibility defence under CERCLA in the light of Burlington Northern. The essential factual claim, so far as relevant to the Divisibility Amendments is that, since PwC did not have the “right kind” of expert advice, they should not have been satisfied that Burlington Northern had the effect that had been assumed in the 2008 Accounts. Moreover, because it is pleaded that there was a high likelihood that Burlington Northern had no effect the clear implication in the APC is that if the “right kind” of expert advice had been obtained, it would not have supported the position taken in the 2008 Accounts.

As a result of the judgment of Miles J following the CMC, it is now clear that the “right kind” of expert advice is, in the APC, only US legal advice. BTI wishes to cast the net wider by asserting that the “right kind” of advice includes expert factual remediation advice. However, while of course a failure to obtain factual remediation advice is different from a failure to obtain purely legal advice, in my judgment the essential factual claim is the same: PwC did not obtain necessary expert advice. The parties appear to be agreed that the availability or otherwise of a divisibility defence depends both on propositions of US law and on the existence of a factual state of affairs (a reasonable basis of apportionment) that given its nature is likely to require expert analysis. Therefore, expert legal advice on US law and expert factual advice as to a reasonable basis of apportionment are, at least in the context of a divisibility defence, simply two aspects of the same overall enquiry. Viewed against the backdrop of CERCLA and the existing APC, an allegation that PwC failed to obtain expert remediation advice is not substantially different from an allegation that PwC failed to obtain US legal advice. It follows that, as a matter of substance, this allegation adds little to the case that I have considered in paragraph 32 above.

PwC also emphasises that Miles J decided that the existing pleading in the APC did not support the admission of expert factual remediation evidence. If the Divisibility Amendments are allowed, BTI will no doubt be seeking to admit expert evidence as to what a reasonable basis for apportionment for the purposes of a divisibility defence would have been. Therefore, argues PwC, the fact that BTI’s amended case would involve the provision of expert evidence for which permission has been refused on its current case demonstrates the essential difference between BTI’s new case and its original one.

I do not accept that argument. That Miles J refused to permit expert evidence on the basis of the unamended pleadings cannot itself determine the outcome of the Divisibility Amendments which seek in part to address the very shortcoming, a failure to plead a need for factual remediation advice, that Miles J identified. Rather, the focus should be on the essential factual basis underpinning the new and old cases and I have set out my conclusion on those essentials in the paragraphs above.

In my judgment, the Divisibility Amendments plead no new cause of action.

A fair reading of the APC as a whole (and without, for the time being considering the implications of the pleading as to disclosures in the 2007 Accounts and the 2008 Accounts) does not highlight any defect in the estimate of Total Remediation Costs used in the accounts, as distinct from the estimate of the NCR/API Share. This is starkly illustrated in, for example, paragraphs 117.2, 224.3, 217 and 219 (as regards the 2007 Accounts) and paragraphs 188, 238 and 243 (as regards the 2008 Accounts). Each of these paragraphs in the APC contain a criticism only of estimates of the NCR/API Share. In the RAPC each paragraph is amended to include a criticism of the estimate of Total Remediation Costs.

In my judgment, an allegation that the Total Remediation Costs were underestimated is, in its essentials, very different from an allegation that the NCR/API Share was underestimated. There is an obvious difference between the process of estimating a share of a larger liability and estimating the total liability itself. Moreover, the likely Total Remediation Costs at the time of the 2007 Accounts and 2008 Accounts could only be sensibly estimated by reference to expert remediation evidence on, for example, the nature and scale of the works that would be needed. It is not obvious that such evidence would be needed to estimate the NCR/API Share and indeed following Miles J’s judgment, BTI will have to make its case on that issue without expert remediation evidence.

BTI argues that the approach set out in paragraph 40 above is not the correct way to approach the Total Costs Amendments. Its first argument is that, having regard to essentials, both the APC and the RAPC allege that PwC failed to audit BTI’s accounts in accordance with accounting standards and wrongly expressed unqualified audit opinions. The particulars of that breach include those summarised in paragraph 17. In BTI’s submission the Total Costs Amendments simply provide further particulars of already pleaded breaches as regards defects in the very same number, the provision for Indemnity Liabilities, and the very same audit procedures that are already heavily criticised in the APC.

BTI justified this approach by stressing the need to look at matters at the “highest level of abstraction”. However, I consider that this betrays the error of approach to which I have alluded in paragraph 30 above. BTI’s approach involves looking at the claim in the “abstract”, rationalising it as a claim involving negligent audit procedures and negligent estimation of a provision in the 2007 Accounts and 2008 Accounts, and asserting that the Total Costs Amendments simply provide further particulars of that negligence. However, ascertaining whether there is a new cause of action does not require the existing claim simply to be considered in the “abstract”. Thus, the amendments in the Co-op case cannot permissibly be approached as simply being further particulars of negligence as to the construction of a concrete floor. Instead, the requirement is that the bare minimum of essential facts be “abstracted” (i.e. taken from) the APC and compared with the bare minimum of essential facts abstracted from the RAPC. BTI’s suggested approach loses sight of the essential difference between an estimate of Total Remediation Costs and an estimate of the NCR/API Share of those costs.

Of more force were BTI’s submissions as regards its existing case on inadequate disclosure in the accounts. BTI argues that the APC already contains allegations that the disclosures in both the 2007 Accounts and the 2008 Accounts were inadequate because they failed to disclose major assumptions or a range of possible outcomes as regards the provisions for Indemnity Liabilities (see paragraphs 118 and 119 of the APC). The APC already pleads (see paragraph 227) that PwC should not have expressed an unqualified audit opinion without disclosure setting out those matters. Moreover, the APC already pleads (in paragraph 270.1 and paragraph 270.2) that if PwC had not wrongly audited the accounts, the directors of Windward would have increased the provision for Indemnity Liabilities (in 2007) or included such a provision (in 2008), failing which PwC would have declined to issue an unqualified audit opinion. Therefore, argues BTI, the Total Costs Amendments simply particularise defects in the disclosures as breaches also of the duty owed in relation to the overall provision for Indemnity Liabilities and involve no new cause of action at all. BTI supported that argument by reference to Mr Daboo’s skeleton outline report and addendum which indicated that auditing a provision for a liability, and auditing FRS 12 disclosure related to that provision were two facets of a similar task, each of which informed the other.

Attractively put though they were in both written and oral submissions of Mr Thompson QC, Mr Keller and Mr Sheehan on behalf of BTI, I reject the argument summarised in paragraph 43. The fundamental difficulty with the argument is that the APC contains no express assertion, whether in relation to the disclosures in the accounts, or in relation to the provision for Indemnity Liabilities, that PwC were under any obligation to obtain, or have sight of, independent expert evidence on the level of Total Remediation Costs.

Miles J has already concluded that the only expert advice that PwC were pleaded to need was legal advice. Legal advice alone could not determine the amount of Total Remediation Costs or range of likely outcomes for the purposes of disclosures required by FRS 12. Therefore, on the basis of the case as currently pleaded in the APC, it cannot be said that the disclosures were inadequate because they were not underpinned by external expert advice on Total Remediation Costs. Yet central to the Total Cost Amendments is the proposition that the provision for Indemnity Liabilities in the 2007 and 2008 Accounts was inadequate because it was based on a flawed estimate of Total Remediation Costs in circumstances where an alleged “proper” estimate of Total Remediation Costs could only be established by independent evidence of a remediation expert. Therefore, even taking into account BTI’s existing case on disclosure, I agree with PwC that the Total Costs Amendments require examination of a new counterfactual case namely what opinion a remediation expert would have given on Total Remediation Costs in circumstances where it has not previously been asserted that PwC needed to know the views of such an expert in order to perform its audits.

What I have just said applies, to an extent, to the Divisibility Amendments as well since they will require an examination of expert remediation evidence that is not needed on the facts as currently pleaded in the APC. However, I have already explained why, in my judgment, the Divisibility Amendments are, in substance, just a pleading of consequences that are said to flow from the already pleaded failure to obtain US legal advice on the divisibility defence in circumstances where CERCLA itself had the result that legal advice and factual advice were simply two aspects of the same enquiry. By contrast, the Total Costs Amendments involve an assertion in the RAPC of a need for PwC to obtain expert advice on Total Remediation Costs, and the expert advice that would have been given if requested, that was not alluded to in the APC or capable of being inferred from CERCLA. Those aspects of the Remediation Costs Amendments involve, in my judgment, a substantial change to the essential factual assertions that BTI makes. I do not accept BTI’s second reason why the analysis set out in paragraph 40 should not apply.

BTI suggested that I should apply the res judicata cross-check referred to in Murray Film Finance (see paragraph 29 above). That may well be a useful exercise in other cases, but I am not satisfied that it would be in this one. Neither party made any detailed submissions on how the res judicata principle might apply if, having made its claim on the basis of the APC, BTI subsequently sought to pursue the Total Costs Amendments as a new claim. As Lord Sumption noted at [17] of Virgin Atlantic Airways Limited v Zodiac Seats UK Limited [2013] UKSC 46 “Res judicata is a portmanteau term which is used to describe a number of different legal principles with different juridical origins”. It would be unwise for me to perform an analysis of how those principles might apply in circumstances where, in my judgment, a comparison of the bare minimum of factual assertions made in the APC and RAPC respectively answers the question.

I conclude that the Remediation Costs Amendments plead a new cause of action.

I will apply the following principles of law in my evaluation of Issue 3:

I consider those to be the applicable principles. As with Issue 2, both parties relied on other authorities and sought to derive principles from them but, on closer inspection, those authorities simply involved an application of the principles I have set out in paragraph 49. For those reasons, I need not deal in detail with these other authorities and I simply note the following:

Strictly, I do not need to consider Issue 3 in relation to the Divisibility Amendments because I have concluded in my analysis of Issue 2 that they do not involve the pleading of any new cause of action. However, for reasons below, even if I had concluded that the Divisibility Amendments pleaded a new cause of action, I would have concluded that the new cause of action arose out of the same or substantially the same facts as the existing claim.

Given the way BTI puts its case in the APC, there is already going to be an examination of whether a divisibility defence was likely to be available. CERCLA itself makes it clear that this examination is dependent both on legal analysis and the existence of a reasonable basis of apportionment as a matter of fact. By pleading that the prospects of a successful divisibility defence could not justify the provision for Indemnity Liabilities in the 2008 Accounts, BTI will be entitled at trial to lead evidence to suggest that the necessary reasonable basis of apportionment was not present.

It is true that BTI cannot, given the Miles J Judgment, rely on expert evidence as to the “factual” elements of a divisibility defence. Nor can BTI assert on its current case that PwC were obliged to obtain expert (non-legal) advice on those factual elements. However, the satisfaction or otherwise of the factual elements of the divisibility defence will nevertheless be an issue for determination by the court even if the evidence that BTI can rely on in support of its case is circumscribed. That leads me to conclude that any new cause of action arises out of the same or substantially the same facts as BTI’s unamended claim and that such a conclusion is consistent with the purpose of the relevant provisions set out in paragraph 49.ii) above.

PwC is correct to note that the Divisibility Amendments require the establishment of some additional facts: for example PwC’s knowledge or otherwise of the expertise of Windward’s directors, or of Mr Gower, in remediation matters (paragraphs 244.2 and 244.3 of the RAPC). There will also need to be further findings on the three apportionment reports referred to in paragraph 245.2 of the APC. However, in the context of a case that already involves detailed pleadings of facts on complicated issues, I do not regard the additional factual matters that PwC highlights as sufficiently substantial to alter my conclusion.

PwC characterises the need for new expert evidence as a “new counterfactual case” and relies on Mastercard. I have already explained why I do not regard that characterisation as entirely accurate. In any event, the presence of a new counterfactual case is not dispositive. It is simply an element in the overall determination of whether any new cause of action arises out of the same or substantially the same facts and, in my judgment, the factors to which I have already alluded mean that it does not.

BTI argues that its existing case on the inadequacy of disclosures in the 2007 Accounts and 2008 Accounts will already involve a searching examination at trial of the likely level of Total Remediation Costs. As already noted, the APC already pleads that the necessary disclosure was not present, that a range of outcomes should have been disclosed and, if PwC had properly audited the disclosures, it would have realised that the provision for Indemnity Liabilities was too low. Therefore, submits BTI, even if it is not permitted to rely on expert evidence on remediation matters, it will be entitled to explore with PwC at trial the kind of disclosures that should properly have been made under FRS 12 in the light of objective evidence as to the level of likely Total Remediation Costs. For example, in 2007 a number of PRPs sought third-party “best and final offers” (“BAFOs”) to take over responsibility for Fox River remediation and, on BTI’s case, two bidders (“Shaw” and “TetraTech”) quoted prices significantly higher than the estimate of Total Remediation Costs that formed the basis of the provision for Indemnity Liabilities in the 2007 Accounts and 2008 Accounts. BTI argues that its existing case permits it to advance the factual case that, in auditing the disclosures, PwC should have taken into account the likelihood, because of Shaw’s and TetraTech’s quotes, that Total Remediation Costs were much higher than Windward’s management were assuming. Therefore, argues BTI, the proper estimate of Total Remediation Costs is going to be litigated anyway with the result that the Total Costs Amendments arise out of the same or substantially the same facts as its existing causes of action.

I do not accept this argument. I can accept that the existing pleading relating to disclosure permits an examination of what Windward, or PwC, might reasonably have considered Total Remediation Costs to be, based on the information actually available to them. So, for example, the court may well need to consider whether the BAFOs were significant to the evaluation of an estimate of Total Remediation Costs when considering FRS 12 disclosures. The court may also be considering whether PwC’s review of information that Windward actually had on Total Remediation Costs was performed reasonably and with appropriate scepticism when considering the case on disclosure.

However, crucially, the court will not be required as part of the existing case on disclosure to consider expert (non-legal) advice which PwC did not actually have, but should (on BTI’s case) have obtained. It follows that the court will not need to make findings on what such additional expert advice would have said if it had been obtained. Yet right at the heart of the Total Costs Amendments is the proposition that the information PwC were given by Windward’s management as to Total Remediation Costs was insufficient for the purpose and should have been supplemented by independent factual expert advice. Moreover, the Total Costs Amendments make assertions as to what such independent factual expert advice would have said about Total Remediation Costs had it been commissioned. The Total Costs Amendments differ from the Divisibility Amendments in this respect because, as I have already explained, the APC already pleads a need for US legal advice on divisibility matters and the Divisibility Amendments can be understood as, in substance, setting out consequences of the failure to obtain that advice. A similar analysis cannot be applied to the Total Costs Amendments.

In my judgment, the “searching examination” of the likely level of Total Remediation Costs that BTI submits will arise even on its case as pleaded in the APC is substantially different from the kind of examination that would take place on the case as pleaded in the RAPC.

I am reinforced in this conclusion by considerations of the purpose of s35 of the Limitation Act 1980 that I have set out in paragraph 49.ii) above. PwC were entitled to assume from reading BTI’s case on the provision for Indemnity Liabilities that it was not suggested that the estimate of Total Remediation Costs was flawed even though the estimate of the NCR/API Share of those costs was. PwC would realise that the level of Total Remediation Costs was a major assumption in the calculation of the provision and so could be relevant to BTI’s criticism of the FRS 12 disclosures. While BTI’s case set out in the APC does refer to uncertainties as to the likely cost of Fox River clean up costs, and states in paragraph 64.11 of the APC (which is in a section grouped under the heading “Background to PwC’s audit”) that Windward had obtained estimates suggesting the costs might be between $400m and $1.6 billion, even in its case on disclosure, BTI does not assert a range of possible Total Remediation Costs that PwC should have disclosed under FRS 12. Nor does BTI assert that any estimate of Total Remediation Costs should have been the subject of expert advice. The amendments proposed would require PwC to investigate what expert advice would have revealed about the likely level of Total Remediation Costs. That is completely outside the ambit of, and unrelated to, the facts that they can reasonably have assumed to have investigated for the purpose of defending BTI’s claim under the APC that is rooted in alleged defects in the estimate of the NCR/API Share.

BTI observes that the question of “what estimates had been created for the duration and cost of the remediation of the Fox River and NRDs” appears as Issue 6 on the list of issues for disclosure. I can understand how BTI’s existing case on disclosure requires it to understand the extent of estimates of Total Remediation Costs which were available to Windward or to PwC. However, that issue does not mean that the court will inevitably have to consider what estimate of Total Remediation Costs PwC should have used when auditing the provisions for Indemnity Liabilities. Nor does it mean that the court will have to consider the import of expert advice that Windward/PwC did not obtain, but which BTI asserts they should have obtained.

In arguing against these conclusions, BTI submits that it would be a “refuge for the incompetent director and the incompetent auditor” only to have to answer for decisions made on the basis of information actually available to them. They refer to Mr Daboo’s clear professional opinion that an auditor would regard auditing a provision and auditing FRS 12 disclosures relating to that provision as part and parcel of the same exercise. However, that does not in my judgment alter my conclusion which is rooted on the cases as currently pleaded and the way issues arising out of those pleadings are likely to be litigated at trial.

I conclude that the new cause of action pleaded by the Total Costs Amendments does not arise out of the same or substantially the same facts as are already in issue in the existing claim.

Given my conclusions above, whether to permit the Divisibility Amendments involves an exercise of discretion under CPR 17.1(2)(b). CPR 17.4 is not engaged because I have concluded that the Divisibility Amendments add no new cause of action.

Since I have concluded that the Total Costs Amendments (i) do add a new cause of action and (ii) do not arise out of the same or substantially the same facts as the existing claim, no question of discretion arises. By virtue of s35 of the Limitation Act 1980 and CPR 17.4, I simply have no power to permit the Total Costs Amendments in those circumstances.

The exercise of discretion under CPR 17.1(2)(b) in relation to the Divisibility Amendments requires me to strike a balance between the potential injustice to BTI if the Divisibility Amendments are refused and injustice to PwC if they are allowed. Like all discretions it must be exercised judicially, having regard to all relevant factors and the overriding objective set out in CPR 1.1 and the factors listed in CPR 1.1(2).

In principle, both parties were agreed that the amendments set out in the Application should not be permitted if the new allegations they raise lack a reasonable prospect of success. However, that is academic in this case since PwC does not seek to argue that the necessary prospects of success are absent.

PwC’s arguments that I should exercise my discretion to refuse BTI permission to make the Divisibility Amendments focused on what it submitted to be the “lateness” of those amendments although PwC does not argue that the amendments are “very late” in the sense that they might cause a hearing date to be lost. As noted at [23] of ABP Technology Limited v Voyetra Turtle Beach Incorporated [2022] EWCA Civ 594, “lateness” is a relative concept in the sense that an amendment is “late” if it could have been advanced earlier. Merely labelling an amendment as “late” does not itself determine how discretion should be exercised. However, the concept of “lateness” can help to focus attention on any prejudice caused by its timing and the need for an explanation (see [24] and [30]).

Where CPR 17.4 is engaged, then considerations additional to those set out above apply. Since the grant of permission to amend under CPR 17.4 would involve PwC having to defend a new cause of action brought outside the limitation period, BTI would bear the burden of persuading the court of the justice of allowing the amendments sought (Seele Austria GmbH & Co KG v Tokio Marine Europe Insurance Limited [2009] EWHC 2066 (TCC)). That will involve an examination of the length of delay in making the amendments, the reasons for that delay and any prejudice resulting to PwC therefrom. One possible head of prejudice may consist of the effect of delay on the defence of the new claim. However, in an appropriate case it can also involve consideration of the fact that the new case has been advanced only after expiry of the limitation period.

Since PwC’s arguments as to the exercise of discretion are based on the asserted lateness of the proposed amendments and other aspects of the history of this claim, I start with a brief summary of that history.

Windward assigned its claim against PwC to BTI on 30 September 2014. BTI commenced these proceedings on 28 October 2014 and served its original Particulars of Claim on 29 May 2015.

BTI initially sought a joint trial of this claim and other related proceedings against Windward’s former directors and Sequana (the “Sequana Proceedings”). PwC opposed this suggestion, and the parties ultimately agreed that the proceedings should be tried separately.

On 27 July 2015, PwC applied to strike out the claims in these proceedings. However, by consent order, these proceedings, including PwC’s strike out application, were stayed pending resolution of the Sequana Proceedings.

On 11 July 2016, Rose J handed down judgment in the Sequana Proceedings, but gave BTI permission to appeal to the Court of Appeal against aspects of that judgment.

Following the trial of the Sequana Proceedings, on 11 January 2018, BTI gave notice to lift the stay on these proceedings and asked PwC to consent to amendments set out in draft amended Particulars of Claim. Mr Lloyd’s Third Witness Statement explained that, among other matters, the amendments “reflected a reconsideration of detail of BTI’s claims following the trial of the Sequana Proceedings, following which BTI had a better understanding of the events and was able to deploy relevant documents and evidence from the proceedings”.

In a letter dated 29 January 2018 PwC explained its view that the stay should continue given the pending appeal to the Court of Appeal. It also indicated that it regarded the amendments to the Particulars of Claim as premature given the pending appeal and that it should not be required to incur the time and expense of considering the amendments until the outcome of the appeal was known.

Both parties evidently modified their positions because, by a consent order of Marcus Smith J dated 25 June 2018, BTI was given permission to amend its Particulars of Claim and was required to file the document that became the APC by 4.30 on 9 July 2018. In addition, the stay was continued until 28 days after the handing down of the Court of Appeal’s judgment against Rose J’s judgment.

The Court of Appeal gave judgment on 6 February 2019 and the stay of these proceedings expired 28 days later. PwC’s strike out application came on for hearing and Fancourt J dismissed it on 15 November 2019. PwC appealed to the Court of Appeal against Fancourt J’s judgment and that appeal was rejected on 11 January 2021. PwC was refused permission to appeal to the Supreme Court on 15 February 2022.

BTI made the Application on 13 April 2022.

The Divisibility Amendments are “late” because they could have been made in 2018 as part of the process that led to the APC. I regard the explanation for the lateness of the amendments as relatively weak.

In arguing against these conclusions, BTI emphasises that it sues as assignee without first hand knowledge of relevant events. It criticises PwC’s refusal to provide it with access to its audit file and working papers. However, in my judgment, the strength of that explanation sharply decreased once the Sequana Proceedings were completed since BTI acquired a detailed knowledge of relevant events in the course of those proceedings. I do not consider that access to PwC’s audit file or working papers was necessary to plead an allegation of a failure to obtain expert remediation advice as well as expert US legal advice on a divisibility defence. BTI made extensive amendments to its Particulars of Claim in 2018 as a consequence of information it gathered in the Sequana Proceedings. It could have made the Divisibility Amendments at that stage as well.

I attach little significance to PwC’s letter of 29 January 2018 referred to in paragraph 76 above. These were BTI’s pleadings, not PwC’s. Whatever PwC initially thought, BTI had turned its mind to how its Particulars of Claim might be amended in 2018 and served a draft of the amended pleading on PwC. That PwC initially thought that such amendments should wait does not explain why, having chosen to amend its pleadings in 2018, BTI did not make the Divisibility Amendments at that time as well. In any event, PwC’s initial reaction was overtaken by events because, they evidently did agree to incur the time and expense of considering the amendments, even with the Court of Appeal proceedings pending, because they consented to Marcus Smith J’s order of 25 June 2018.

BTI points to PwC’s failed attempt to strike out these proceedings, culminating in an unsuccessful appeal to the Court of Appeal against Fancourt J’s decision and a further unsuccessful attempt to obtain permission to appeal to the Supreme Court. I acknowledge PwC’s actions have contributed to the fact that, even though these proceedings were commenced in 2015, the first CMC has only just taken place and the trial will not be until 2024. However, in 2018, BTI made amendments to its Particulars of Claim despite the fact that PwC had already applied to strike out these proceedings. I do not accept BTI’s submission that there was a brief “window of opportunity” to make amendments in 2018. BTI itself chose to open that window to draw on its experience from the Sequana Proceedings. In the Sequana Proceedings BTI accepted that the estimates of Total Remediation Costs used in the modelling of the provisions for Indemnity Liabilities were best estimates of those figures (see paragraph [109] of Rose J’s judgment at first instance). It had ample time by 2018 to reflect on whether it wished to maintain that position in these proceedings.

In my judgment, the reason why the Divisibility Amendments are made now, rather than in 2018, has relatively little to do with PwC’s actions. Mr Lloyd’s witness statement explains that in the period between the Court of Appeal handing down judgment in the strike-out application (on 11 January 2021) and the CMC (in March 2022), he and colleagues moved firms from Debevoise & Plimpton to Hogan Lovells and both the Divisibility Amendments and Total Costs Amendments were prompted by (privileged) discussions with CERCLA experts at their new firm. I regard that explanation, which amounts to further reflection on the claims having taken place well after the expiry of the limitation period, as weak.

In his Fifth Witness Statement, Mr Hewetson said that PwC would be prejudiced if amendments were permitted now because the memory of the auditors involved is necessarily poorer than it would have been had BTI’s full case been capable of consideration in 2015. Most of that complaint was directed at the Total Costs Amendments (see paragraphs 54 to 58 of his witness statement) but it was also directed at the Divisibility Amendments because Mr Hewetson said, in paragraph 59 of his witness statement that it has not been necessary, in the course of investigating the existing claims, to ask the auditors why they did not seek advice from an environmental science/engineering expert in the course of auditing the assumptions relating to a divisibility defence.

I regard that as relatively low prejudice. As I have explained, the Divisibility Amendments to a large extent make assertions about consequences that would have flowed if legal advice had been sought. That is concerned not with why the auditors acted as they did, but rather what would have happened if US legal advice had been sought. The Divisibility Amendments also assert that PwC failed to obtain expert factual advice. That, however, invites a largely objective analysis of whether reasonable auditors in PwC’s position would have obtained that advice. I accept that the auditors’ subjective memories might provide some relevant context for that enquiry. However, in my judgment most of the relevant context will come from documentary evidence such as PwC’s audit file which both parties acknowledge has been retained.

I therefore have a relatively weak explanation for the lateness of the Divisibility Amendments. Nevertheless, if I permit the Divisibility Amendments, PwC will suffer relatively low prejudice. By contrast, if I refuse the Divisibility Amendments, BTI will be deprived of the ability to advance a case based on a failure to obtain expert remediation advice on divisibility issues even though it has already advanced a case based on a failure to obtain legal advice and CERCLA itself makes it clear that legal advice and factual advice on apportionment go hand in glove so far as a divisibility defence is concerned. In my judgment, the balance comes down in favour of exercising my discretion under CPR 17.1(2)(b) to permit the Divisibility Amendments.

Even if I had power under CPR 17.4 to permit the Total Costs Amendments, I would not have allowed them. As I have found, BTI’s explanation for the delay in making the Total Costs Amendments is, in essence, that it thought more about its claims after the expiry of the limitation period and several years after the end of the first instance Sequana Proceedings during which BTI had accepted that the estimates of Total Remediation Costs that PwC used in auditing the provisions for Indemnity Liabilities were “best estimates”. I do not regard that as a sufficiently good reason to justify requiring PwC to defend a new cause of action after expiry of the limitation period.

That conclusion is partly a result of the weak explanation that BTI has advanced and partly a result of the prejudice that PwC would suffer if the Total Costs Amendments were permitted. Those amendments involve allegations of defects in an aspect of PwC’s work where no defects were previously alleged. Unlike the Divisibility Amendments, they cannot be explained to any material extent as matters consequential on already pleaded defects. I am prepared to accept that, to defend the new allegations comprised in the Total Costs Amendments, PwC will need to gather some evidence of auditors’ recollections of how they approached the estimate of Total Remediation Costs, even with the audit file available. Such recollections will be of materially more relevance to the Total Costs Amendments than to the Divisibility Amendments. Mr Hewetson has confirmed in his Sixth Witness Statement that Mr O’Brien and Mr Derbyshire were interviewed in 2015 when BTI first brought its claim. Mr Singla QC confirmed that Mr O’Brien and Mr Derbyshire would be giving evidence at trial (as would be expected given that they were closely involved in the audit). PwC could not have known in 2015 that they needed to ask Mr O’Brien and Mr Derbyshire about Total Remediation Costs as well as the NCR/API Share. PwC suffers greater prejudice in the context of the Total Costs Amendments, in the form of a likely dimming of Mr O’Brien’s and Mr Debyshire’s memories, than they would have suffered in the context of the Divisibility Amendments.

I give BTI permission to make the Divisibility Amendments in exercise of my discretion under CPR 17.2(1)(b). I have no power to give, and so refuse, BTI permission to make the Total Costs Amendments whether under CPR 17.2(1)(b) or CPR 17.4. I invite the parties to agree an order giving effect to this judgment and to consequential matters and will hear counsel further to the extent they cannot agree.