Evonik UK Holdings Limited & Ors v Commissioners of Inland Revenue

The Claimant (Evonik) is a member of the Franked Investment Income Group Litigation Order (the GLO). I will assume that any reader of this judgment has a general understanding of the legal issues to which the GLO gave rise. Since that litigation has been proceeding for over 20 years now, it is not practicable to provide any digestible potted summary of the litigation as a whole. Rather, I give the barest summary in deliberately broad-brush terms sufficient to put in context the issue that I must now decide.

One aspect of Evonik’s claim was for compensation for the time value of money forgone when it paid Foreign Income Dividends (FIDs). Until the FID regime was abolished, on paying a FID, Evonik was obliged to account for advance corporation tax (ACT). It was, however, able to reclaim that ACT when its liability for mainstream corporation tax (MCT) fell due. Litigation in the GLO established that some of the ACT paid on FIDs was levied unlawfully and in breach of principles of the law of the European Community and the European Union (together the EU). Accordingly, it is common ground that Evonik is entitled to some compensation for the period during which it was “out of the money” as regards ACT, namely the period between the date on which ACT on FIDs fell due, and the date on which it was repaid, which is referred to in some of the authorities as the “period of prematurity”.

Evonik’s FID Claim is, accordingly, for some measure of interest on the unlawful ACT paid in connection with FIDs for the period of prematurity (its FID Claim). Evonik obtained summary judgment on its FID Claim in 2016. However, that summary judgment was on the basis that Evonik was entitled to compound interest during the period of prematurity. That summary judgment was based on the understanding of the law in 2016 which has subsequently been reversed by judgments of the Supreme Court in PrudentialAssuranceCompany Ltd v HMRC [2018] UKSC 39 (Prudential) and Test Claimants in the FII Group Litigation v HMRC [2021] UKSC 31 (FII SC3). In FII SC3, the Supreme Court set aside Evonik’s summary judgment obtained in 2016 describing that set aside as being to the extent only of the Sempra issue so that the set-aside was operative only insofar as an award of compound interest had been made.

Evonik’s position is that, in the light of Prudential and FII SC3, its compensation should be calculated by applying the following rates for the following periods:

Evonik has applied for judgment on its FID Claim to be entered on the basis of the methodology set out in paragraph ‎4. HMRC accepts that Evonik is entitled to some form of judgment on its FID Claim. However, they argue that Evonik is not entitled to judgment on the basis that it seeks for the following reasons:

It was common ground that the 2021 Order was made by consent and involved HMRC, Evonik and others compromising an application that various taxpayers had made to the court on 15 January 2021. Both parties therefore agreed that the following approach, set out at [19] of Asplin LJ’s judgment in Botleigh Grange Hotel v Revenue & Customs Commissioners [2018] EWCA Civ 1032, should be applied when construing the 2021 Order:

In this case, the document to be interpreted is a Consent Order which not only embodies an agreement between the parties but is also a formal court document which has public significance. It is important therefore, to interpret it, not only in accordance with the guidance in relation to contracts which are bilateral or synallagmatic arrangements, but also in the light of its public significance. [The judge] would not have approved the Consent Order if he had not been satisfied that it was appropriate for the Court to make such an order and that it properly addressed the relevant Companies Court procedure and issues. It must be construed in that context.

Neither side thought it necessary to engage in a review of authorities on the correct approach to contractual construction. Nor did either side make much of the “extra dimension” referred to in the extract from Asplin LJ’s judgment to the effect that the 2021 Order is a “formal court document which has public significance”. I therefore took the parties to be agreed that I should apply the following approach which is derived from the judgment in Botleigh Grange insofar as relevant to the issues of construction to which the 2021 Order gives rise:

Recital A of the 2021 Order recited, correctly, that Evonik as well as other claimants in the GLO had obtained summary judgment on their FID Claims.

Recital C of the 2021 Order referred to the following provision of an earlier order of Falk J of 3 May 2019 (the 2019 Order):

[AND UPON] the directions in paragraph 6 of the Order of 3 May 2019 that the parties seek to agree the “Established Value” of the claims listed in that order that term being defined to mean the minimum value of a claim which is established on the basis of the law as finally determined with no possibility of further appeal in the FII GLO litigation

No such Established Value had been agreed following the 2019 Order in relation to Evonik’s FID Claim because, at that point, the FID Claim was thought to have been disposed of by the summary judgment in 2016. However, by the time of the 2021 Order, the position in relation to Evonik’s FID Claim had moved on. Although the 2016 summary judgment remained in place, FII SC3 had been heard in December 2020 and Evonik’s 2016 summary judgment was under appeal in FII SC3. Both HMRC and Evonik thought that there was utility in seeking to agree Established Values as was recognised by Recital F of the 2021 Order in the following terms (with the underlining being my own, added to bring out a central aspect of both sides’ arguments):

[AND UPON] The Defendants providing their computations, or agreeing to the Claimants’ computations, of the minimum value of the Claimants’ FID Claims which were the subject of the summary judgments in part being the minimum value which has been established on the basis of the law as finally determined with no possibility of further appeal in the FII GLO litigation (the “FID Established Values”).

Paragraph 1f of the 2021 Order was a declaration (the Paragraph 1f Declaration) made by consent to the effect that the FID Established Value of Evonik’s FID Claim as at 7 March 2016 was the aggregate sum of £2,323,908.19 in respect of the claims brought by the two relevant members of the Evonik group.

It is common ground that this figure was determined by applying the Section 85 interest rate to the totality of Evonik’s FID Claim. Accordingly, it is common ground that the FID Established Value recorded in the Paragraph 1f Declaration is some £2.5m lower than the amount for which Evonik now seeks judgment in connection with its FID Claim with the difference being attributable to the fact that Evonik seeks to apply the higher Section 35A interest rate to part of the unlawful ACT in substitution for the lower Section 85 rate that was implicit in the figure of £2,323,908.19. The circumstances in which Evonik came to conclude that it should be applying the higher Section 35A rate are considered in paragraph ‎55 below.

As well as the declarations made by consent, the 2021 Order also contained operative provisions, also expressed to be made by consent. By paragraph 3 of those orders, all parties were given liberty to apply.

The parties have two very different interpretations of the concept of FID Established Value set out in paragraph ‎10 above.

Evonik emphasises that the FID Established Value is a “minimum value”. It argues that, in now seeking judgment for a higher sum than the figure of £2,323,908.19 recorded in the Paragraph 1f Declaration, it is not seeking to go behind the 2021 Order. It argues that the Paragraph 1f Declaration remains correct: £2,323,908.19 is still the “minimum value” as at 7 March 2016 of Evonik’s FID Claim which had been established on the basis of the law as finally determined with no possibility of further appeal in the GLO. Circumstances have simply moved on with the result that Evonik’s actual claim established on that basis is higher than the minimum specified in the Paragraph 1f Declaration.

HMRC argue that the FID Established Value is not just a minimum but is a minimum that is “established on the basis of the law as finally determined with no possibility of further appeal in the [GLO]”. From that, HMRC conclude that the only factor that could permit the “minimum” to be exceeded would be a decision after the 2021 Order on a point of law that remained “live” in the GLO at the date of the 2021 Order. HMRC characterise Evonik’s attempt to obtain a better outcome than is set out in the Paragraph 1f Declaration as being driven by a realisation of an error that arose from Evonik’s misapprehension about either the law or the underlying facts. In HMRC’s submission this is not something that permits the FID Established Value to be exceeded.

The parties also had different perspectives on the scope of the “liberty to apply” conferred by paragraph 3 of the operative provisions of the 2021 Order:

Both Evonik and HMRC made detailed submissions on the background to both the 2021 Order and the 2019 Order. I will, therefore, start by considering the factual background both that leading up to the 2019 Order (and the genesis of the “Established Value” concept) and that leading up to the 2021 Order. I will then set out my conclusions in the light of those findings.

I had evidence from Mr Simon Whitehead and Mr John Hayton, both of whom are solicitors at Joseph Hage Aaronson LLP (JHA), the firm that has been advising claimants in the GLO for some time. There was no challenge to the evidence of Mr Whitehead or Mr Hayton and HMRC did not rely on evidence from any witnesses. I have, accordingly, accepted the totality of the evidence of Mr Whitehead and Mr Hayton which has informed my findings on the factual background below.

As I have noted, the 2021 Order embodied the agreed compromise to an application made on 15 January 2021 by the claimants in the GLO requesting an order that HMRC be obliged to provide “Established Values” of the FID Claims.

At the time that application was made, FII SC3 had been heard (at a hearing between 7 to 10 December 2020), but the Supreme Court had not yet given judgment. The summary judgments that various claimants, including Evonik, had obtained on their FID Claims were under appeal in FII SC3. PrudentialAssuranceCompany Ltd v HMRC [2018] UK SC 39, had decided that, as a general principle, compound interest could not be awarded on a restitutionary remedy in a FID Claim in relation to the period of prematurity. However, the Claimants’ position in FII SC3 was that they should retain the compound interest that they had been awarded in 2016 because, for example, an issue estoppel applied in their favour or that it would be Henderson v Henderson abuse for HMRC to argue for simple interest only. In the alternative, the Claimants were arguing that interest in the period of prematurity should be calculated by reference to Section 35A.

HMRC’s position in FII SC3 was that interest throughout the period of prematurity should be calculated by reference to simple interest at Section 85 rates.

HMRC were also arguing in FII SC3 that all of the Claimants’ 2016 summary judgments on FID Claims should be set aside in their entirety. The Claimants, including Evonik, did not agree. They reasoned that, since HMRC accepted that, at the very least simple interest was available at Section 85 rates, the Claimants would always be entitled to judgment computed on that basis. For their part, HMRC maintained that the Supreme Court could scarcely award judgment by reference to interest at Section 85 rates since there had been no valuation of any remedy for FID Claims that was based on Section 85 interest.

Mr Whitehead’s evidence is that, in pressing for a determination of FID Established Values, Evonik was motivated by a wish to prevent HMRC obtaining any tactical advantage from the absence of any calculations using Section 85 rates. Since the Supreme Court had already heard FII SC3, Evonik and other taxpayers in the GLO prepared their calculations in some haste as they wished to minimise the scope for any such tactical advantage. HMRC similarly reviewed the calculations in comparative haste. The taxpayers’ application that led to the 2021 Order was made on 15 January 2021. By 4 February 2021, 10 sets of FID Established Values had been agreed and included in the 2021 Order.

I accept that as an accurate statement of Evonik’s subjective motivations. However, what matters is a reasonable observer’s perspective. In my judgment, a reasonable observer would conclude that the 2021 Consent Order was made in circumstances where all parties considered that there was some utility in having some computations, agreed to the extent possible, of the minimum amount of the FID Claims of various taxpayers, including Evonik, if interest for the period of prematurity were calculated only at Section 85 rates.

As I have noted, the Paragraph 1f Declaration is simply a declaration by consent as to the amount of the FID Established Value of Evonik’s FID Claim. Since that declaration was made by consent, there was no process that the court ordered HMRC and Evonik to follow when agreeing the figure. However, in practice, the figure was agreed following the application of the same process that the court had ordered in connection with the 2019 Order (see the section below): Evonik provided its computation of what it considered the “minimum” to be, HMRC provided their comments and in the light of those comments a “minimum” figure came to be agreed.

Other taxpayers, also advised by JHA, were also seeking to agree FID Established Values with HMRC. In the course of agreeing a computation of a FID Claim on behalf of Prudential plc, both JHA and HMRC came to realise on or around 27 January 2021 (before the 2021 Order was made) that, if ACT was paid on a FID and not repaid before Prudential plc issued proceedings, there was at the least a strong argument that Section 85 could not apply. That was because Section 85 applied only to “relevant payments” defined as payments of unlawful ACT that were repaid wholly or in part before a taxpayer had commenced proceedings in the High Court seeking recovery of these sums (see s85(2) of Finance Act 2019). HMRC referred to the point specifically in their following comments on calculations of the FID Established Value of Prudential’s claim:

… we note that you have calculated interest under s.85 on the £11.06m FID ACT repaid in March 2017 (when Prudential plc was awarded summary judgment on its FID claim). In HMRC’s view s.85 cannot apply because s.85(2) provides that the section applies only if the ACT was set off or repaid before the proceedings were started

As noted above, the 2021 Order drew on the definition of “Established Value” that was set out in the 2019 Order. The parties were agreed that the background to the 2019 Order is relevant on the basis that it sheds light on the proper interpretation of the concept of “Established Value” in the 2019 Order, which in turn informs the interpretation of the similar terms used in the 2021 Order.

The 2019 Order provided, so far as material as follows:

6.

The following direction shall apply so as to enable the parties to agree valuations in respect of the claims of [Evonik among others] (together the “Non Test Claimants”):

a.

In this direction, “Established Value” means the minimum value of a claim which is established on the basis of the law as finally determined with no possibility of further appeal in the FID GLO litigation.

b.

Not later than eight weeks after any order by the Supreme Court refusing permission to appeal in respect of Issues 1, 2, 4, 5, 6, 7 and 9 in the List of Issues attached to the Supreme Court’s order of 8 April 2019, the Defendants shall provide to the Non Test Claimants their computation of the Established Values of the claims of the Non Test Claimants.

c.

The parties shall use their best endeavours to agree the Established Values of the claims of the Non Test Claimants within six weeks of the Defendants having complied with paragraph b above.

d.

The Non-Test Claimants shall have liberty to apply to take any steps which they consider appropriate on the basis of such agreed Established Values.

The 2019 Order was made at a CMC before Falk J on 3 May 2019. At that time, the taxpayers had enjoyed significant success in Test Claimants in the FII Group Litigation v Revenue and Customs Commissioners [2016] EWCA Civ 1180 but HMRC were seeking permission to appeal to the Supreme Court. The grounds on which HMRC sought permission could be categorised as, on the one hand, “computational issues” going to the nuts and bolts of the quantification of the taxpayers’ restitutionary remedy and, on the other, “fundamental issues”, such as limitation and the availability of compound interest as a remedy. As at 3 May 2019, HMRC had permission to appeal to the Supreme Court on the “fundamental issues” but there was to be an oral hearing in the Supreme Court in June 2019 to decide whether they should have permission to appeal on the “computational issues”.

The taxpayers’ position going into the CMC on 3 May 2019 was that it was clear following the lengthy proceedings in the GLO up to that point that HMRC owed them something in respect of their “main claims” based on the unlawful imposition of ACT. These “main claims” were distinct from the taxpayers’ FID Claims which, as I have noted, were dealt with in orders for summary judgment made in 2016. The taxpayers were precluded by s234 of Finance Act 2013 from applying for an interim payment from HMRC save in exceptional cases. However, their position at the CMC on 3 May 2019 was that a trial, already listed to take place in the High Court in July 2019, should decide whether judgment should be entered against HMRC for “the minimum agreed value of the Claimants’ claims remaining after any partial summary judgment already awarded” (in the words of Issue 1 as formulated in a List of Issues attached to Fancourt J’s order of 22 June 2018).

The taxpayers were also concerned that HMRC were not being sufficiently diligent in addressing evidence that the taxpayers were providing in support of their main claims. Paragraph 1 of Fancourt J’s order of 22 June 2018 required the parties to use best endeavours to agree the “minimum value” of certain taxpayers’ claims and the taxpayers concerned were suggesting that HMRC had not complied with that obligation.

At the CMC on 3 May 2019 HMRC opposed the proposal for a determination of “Issue 1” in July 2019. Their first reason was that there was little utility in a High Court judge giving a judgment in part by applying the law as it stood given that the Supreme Court might change the law both on the “fundamental issues” and on the “computational issues” if it gave permission to appeal on them. HMRC were also concerned that it was procedurally unattractive to give judgment in part for a “minimum agreed value” as Issue 1 envisaged since the doctrine of merger might make it difficult for the taxpayers to request judgment in a higher amount should future developments in the GLO mean that the amount due was more than the minimum.

Falk J gave an oral decision at the CMC on 3 May 2019 to the effect that the determination of “Issue 1” should be adjourned so that it would not be heard at the forthcoming hearing in July 2019. That decision was reflected in paragraph 1 of the 2019 Order.

The provisions of the 2019 Order dealing with “Established Value” emerged after Falk J gave her oral judgment referred to in paragraph ‎34. Those provisions did not, therefore, involve Falk J determining an application made by either party for a request that “Established Values” be determined. Rather, the provisions dealing with “Established Value” came about following a dialogue between Falk J, counsel for the taxpayers (Mr Bremner KC) and counsel for HMRC (Mr Ewart KC) as to how the proceedings should be case managed given that Issue 1 was no longer to be on the agenda at the hearing fixed for July 2019. Indeed the terms of the 2019 Order were not even substantially finalised at the CMC on 3 May 2019: Falk J said that she wished to discuss the broad parameters of the order and leave the parties to fill in the detail after the hearing. Ultimately the parties did so, and Falk J signalled her endorsement of those discussions by sealing the order with which she was presented (possibly with some amendments). However, this is not a case in which there is an authoritative judgment that sets out a judge’s reasons for making the order in question.

That said, it can be seen that the 2019 Order was a somewhat different order from the 2021 Order. It involved both sides being held to the high standard of using “best endeavours” to agree a calculation, rather than the parties simply deciding, by consent, to do so. Moreover the obligation to use those “best endeavours” arose only if (as ultimately transpired) the Supreme Court refused permission to appeal on the “computational issues”, the rationale no doubt being that a refusal of permission to appeal on those issues would make the agreement of Established Values more straightforward.

Authority suggests that caution should be exercised when using the parties’ submissions as an aid to the construction of an order. In SDI Retail Services Ltd v Rangers Football Club [2021] EWCA Civ 790, Phillips LJ described this as a “difficult and dubious exercise, with parallels to admitting evidence of negotiations in construing a contract”. Underhill LJ had a somewhat different perception, but Baker LJ echoed Phillips LJ’s concern. Since in this case, there is no judgment in which Falk J set out authoritative reasons for making the 2019 Order, I consider that counsel’s submissions at the hearing before her are of even less utility as an aid to construction than they would otherwise be, since those submissions cannot serve as a setting within which conclusions within that judgment were reached. However, since both sides referred to submissions made at the CMC on 3 May 2019 I will note the following aspects of those submissions and Falk J’s response to them:

I start with the text of the 2021 Order. The FID Established Value is defined to be a minimum value that has been “established” on the basis of the law as finally determined with no possibility of further appeal in the FII GLO litigation. This phraseology can be read quite naturally as (i) establishing a hypothesis (namely that the only law to be applied is that as finally determined with no possibility of further appeal in the GLO) and (ii) enquiring as to the minimum amount of the claim that has been established on that hypothesis. This natural reading is consistent with Evonik’s interpretation. For example, on the basis of the hypothesis, Evonik might have a factually unassailable claim to 50, with the possibility of establishing further facts that could make the claim worth 100 even if the law on open issues in the FII GLO was unchanged in FII SC3. To say that the minimum value of the claim on the hypothesis is 50 does not exclude the possibility that it might turn out to be worth 100.

HMRC argue that the reference in the definition of Established Value to the “law as finally determined with no possibility of further appeal in the FII GLO litigation” does more than set up a hypothesis. They argue that it also serves to delineate those matters that can permissibly be relied upon when asserting that the “minimum” is exceeded (namely legal developments on open issues within the GLO). Put another way, HMRC argue that the 2021 Order envisages that Evonik would ultimately be entitled to obtain judgment for its FID Established Value plus whatever extra it might obtain because of the resolution of outstanding legal issues in the GLO such as compound interest and limitation issues.

However, there is a clear linguistic difficulty with that analysis. The definition of Established Value simply envisages a minimum. Neither that definition nor any other aspect of the 2021 Order distinguishes “permissible” circumstances in which the minimum can be exceeded from “impermissible” circumstances.

HMRC criticise Evonik’s interpretation that focuses on the word “minimum” as failing to give effect to the tailpiece of the definition that focuses on the “law as finally determined with no possibility of further appeal in the FII GLO litigation”. However, I consider that Evonik has much the better of the argument on the purely linguistic interpretation of the 2021 Order. If one extends the example in paragraph ‎38 by assuming that if HMRC’s arguments on limitation in FII SC3 succeeded, 50% of Evonik’s FID Claim would be statute barred, Evonik’s interpretation of the 2021 Order would lead to a conclusion that the FID Established Value is 25 and that interpretation would give effect to all aspects of the definition, including the tailpiece. By contrast, HMRC’s interpretation of the 2021 Order, as precluding Evonik from arguing that its FID Claim is worth more than 50 fails to give proper weight to the use of the word “minimum” and approaches the 2021 Order as if it involves a declaration as to the actual value of the FID Claim (on the basis of the law as finally determined in the FII GLO with no possibility of further appeal) rather than a minimum value.

I consider HMRC’s linguistic analysis to involve a strain on the natural meaning of the words used in the 2021 Order. However, Mr Ewart KC has persuaded me that HMRC’s interpretation is nevertheless linguistically permissible and so it is appropriate to continue the iterative approach by considering whether that strained interpretation is nevertheless justified by considerations of context and other matters.

HMRC place considerable emphasis on their argument that it would be contrary to common sense for them to agree to the 2021 Order, made by consent, that was asymmetric in its operation in that Evonik could, if it chose, seek judgment for an amount higher than the FID Established Value, but HMRC had no right to argue for a lower sum.

I do not accept that argument. It is clear from the background to the 2019 Order that, by May 2019, a point had been reached in the GLO litigation where all parties recognised that taxpayers were owed at least something in relation to their main claims. Even if it was asymmetric, there was some utility in trying to determine what that absolute minimum might be. The process of agreeing that absolute minimum would enable the parties to focus on what the true outstanding points of dispute were. That in turn had the potential to make the ongoing litigation more efficient and more focused, thereby saving costs. When the parties, in the 2021 Order made by consent, adopted the same concept of “Established Value” as had been used in the 2019 Order to FID Claims, a reasonable observer might well conclude that they intended this same rationale to apply to the agreement of a FID Established Value.

HMRC’s riposte is that the concept of Established Value (in the 2019 Order) was intended to be a proxy for a minimum sum on which judgment could be entered in the absence of future legal developments on open issues in the FII GLO. They rely on some of the comments made by Falk J at the CMC in May 2019 to that effect (see paragraphs ‎37.ii) and ‎37.iii)) and therefore maintain their objection that it would make no commercial sense for them to agree a minimum FID Established Value that could only ever be adjusted up and not down. However, a difficulty with that argument is that the concept of an “Established Value” emerged at the CMC in May 2019 after Falk J had expressly rejected the idea of a trial on “Issue 1” that could have led to taxpayers being awarded judgment in part. The 2019 Order accordingly envisaged that the parties would need to take some further step, once Established Values had been agreed, if those Established Values were to give rise to payment obligations on HMRC. Such further steps were not made mandatory, reinforcing the conclusion that the Established Values should not necessarily be taken to be amounts on which judgment in part would be entered in the absence of further legal developments in the GLO.

Moreover, neither the 2019 Order nor the 2021 Order contains any mechanism specifying a process for resolving a calculation of any “extra” amounts to which Evonik might be entitled following the resolution of, for example, limitation and compound interest issues within the GLO. That also serves to negative a suggestion that the FID Established Value was an amount on which judgment would be given once “topped up” by any such extra amounts.

I do understand HMRC’s point about asymmetry. However, an order such as the 2019 Order necessarily involved a degree of asymmetry. Conceptually, the parties could have been required to agree a “maximum” amount of a claim. A formulation in those terms would impose some risk on the taxpayers who would be giving up at least some potential to argue for more. Another option would be to require the parties to agree the precise amount of a claim (which would place the same risk on both parties as neither could argue for a different amount absent a determination of a live question of law in the GLO). A still further option would be to require agreement of a “minimum” amount which would put risk on HMRC who would be giving up some right to argue for less. In the circumstances prevailing at the CMC in May 2019, which included allegations that HMRC were failing adequately to respond to taxpayers’ calculations, the court was persuaded to make an order that placed risk on HMRC. Admittedly circumstances had moved on in 2021. However, when the parties agreed, by consent, to adopt a similar methodology for the determination of “FID Established Values”, they were necessarily adopting a formulation that was asymmetric and placed more risk on HMRC than on taxpayers.

HMRC make the related point that, on Evonik’s interpretation of the 2021 Order, Evonik would be able, as indeed it has done, to request judgment for a sum much higher than the declared FID Established Value, the day before judgment is due to be entered, thereby potentially reopening the whole dispute with HMRC. However, I consider this argument to involve a degree of hindsight. All substantive legal issues that arose in the GLO that go to the taxpayers’ entitlement to remedies have been resolved following over 20 years of litigation. It is understandable for HMRC to feel frustration now at the prospect of calculations already prepared having to be revisited. However, the fact that HMRC feel understandable frustration now says little about the correct construction of the 2021 Order which has to be determined in the light of circumstances prevailing when it was made. As I have explained in paragraph ‎45 above, I do not accept that the concept of “FID Established Value” was intended to set out the final amount on which judgment would be given in part (absent determination of open questions of law within the GLO itself). Its purpose rather was to set out a minimum value that could, if the parties wished, be made the subject of further applications. It may be that, over time, both HMRC and taxpayers have come to approach “Established Values” as amounts on which judgment should be entered. However, conduct or agreements after the 2019 Order or 2021 Order say nothing about how those orders should be construed.

Evonik’s approach to the construction of the 2021 Order is, in my judgment, supported by considerations of context rather than undermined by it. It is significant that the 2021 Order taps into definitions and concepts used in the 2019 Order and that the Paragraph 1f Declaration was agreed following a process similar to that the court had mandated in the 2019 Order. Accordingly, the parties were seeking to agree a minimum value of Evonik’s FID Claim based on the law as finalised within the GLO. As I have explained, that necessarily resulted in an asymmetric outcome since it constrained HMRC’s ability to argue that they should have to pay less. Parties do not routinely settle claims for the “minimum” that their opponent is prepared to agree following a reasonably rapid review. I consider that a reasonable observer would conclude that in concurring with HMRC on determination of a minimum figure, Evonik was unlikely to be forgoing any right to ask for more even if open legal issues in the GLO were resolved in HMRC’s favour.

Moreover, before the 2021 Order was made, both parties, through their legal advisers, realised that there were some potentially knotty issues surrounding the calculation of the amount of a FID Claim. The discussions about the quantification of Prudential’s claim referred to in paragraph 27 showed, for example, that it was at least possible that Section 85 interest was not available in connection with ACT repaid after proceedings started. That raised the question as to what, if any, interest would be available in connection with such ACT including whether interest accrued at the more generous Section 35A rate. I do not consider that background supports an interpretation that, by agreeing to “Established Values” that did not include amounts of Section 35A interest in this respect, Evonik should be taken to be precluded from raising this issue. That would be to impose on Evonik the risk of failing to identify, in 2021, the possibility of Section 35A interest accruing on this part of the claim. Yet, as noted, by adopting the formulation of “Established Value” used in the 2019 Order, the parties were agreeing to an asymmetric position that imposed the greater risk on HMRC rather than on taxpayers.

I am less attracted to Evonik’s argument that ultimately HMRC agreed to a consent order under which Prudential obtained interest at Section 35A rates on ACT on FIDs repaid after it commenced proceedings. HMRC expressly agreed that treatment without prejudice to its position in other cases. In any event, the settlement of the relevant part of Prudential’s claim took place in 2023, after the 2021 Order was made and I do not consider that the parties’ conduct after the 2021 Order to provide any guide to the proper construction of that order.

Nevertheless, putting together all the indications, both textual and arising from context, I conclude that Evonik’s application for judgment is not inconsistent with the 2021 Order. Accordingly, it is not necessary to consider whether Evonik needs to invoke the “liberty to apply” contained in the 2021 Order.

Before the hearing of what was to become BAT Industries and others v HMRC [2024] EWHC 195 (Ch) (FII HC3), Evonik made certain statements in an agreed statement of facts as to the value of its claim. It is common ground that in so doing, Evonik made an admission and that the higher amount for which it now seeks judgment on its FID Claim is inconsistent with that admission.

It follows that it is common ground that, for Evonik to be entitled to judgment in the amount it now claims, it needs the court’s permission to withdraw its admission. In that connection, CPR 14.5 provides as follows:

Application for permission to withdraw admission

14.5

– In deciding whether to give permission for an admission to be withdrawn, the court shall consider all the circumstances of the case, including—

(a)

the grounds for seeking to withdraw the admission;

(b)

whether there is new evidence that was not available when the admission was made;

(c)

the conduct of the parties;

(d)

any prejudice to any person if the admission is withdrawn or not permitted to be withdrawn;

(e)

what stage the proceedings have reached; in particular, whether a date or period has been fixed for the trial;

(f)

the prospects of success of the claim or of the part of it to which the admission relates; and

(g)

the interests of the administration of justice..

In my judgment, the analysis of the following factors is relatively straightforward:

The factors set out in paragraph ‎55 point firmly in favour of holding Evonik to its admission. The grounds in support are relatively weak (an earlier failure to appreciate that a higher claim could be made) and while mistakes do happen, and neither side’s conduct is bad, the admission is sought to be withdrawn just at the point at which one would expect a final judgment to be close.

The parties had very different approaches to the evaluation of competing prejudices (CPR 14.5(d)). HMRC characterised this consideration as adding little to the evaluation. They submit that, if the admission is permitted to be withdrawn, HMRC are likely (subject to the points made below) to have to pay more with the result that Evonik will receive more. HMRC would always prefer to pay less and Evonik would always prefer to receive more and therefore HMRC analyse the question of “prejudice” as involving a zero-sum game. On that analysis, HMRC consider that they would not suffer “prejudice” if they have to pay more but assert that Evonik would not suffer prejudice if held to its admission.

Evonik approaches the question of prejudice differently. It argues that, if it is released from its admission, the effect of an unfortunate mistake will be undone, it will achieve the “right” remedy and there can be no prejudice to HMRC in having to pay the “right” amount. By contrast, Evonik submits that it would suffer prejudice if it were held to its admission and had to accept a remedy that was artificially low because of a mere mistake.

I do not accept either side’s approach to this issue. I agree with Evonik that conceptually it could suffer prejudice if required to accept a remedy that is less than the amount to which it is entitled in law simply because it made an oversight. That is particularly the case given that Evonik’s admission simply operated to give the court in FII HC3 a high-level impression of the overall value of its claim. I did not refer to the value of Evonik’s FID claim in my judgment in FII HC3 and nor did any perception as to the value of that claim form part of my reasoning in a judgment that was concerned with limitation issues. The Supreme Court has been referred, when settling the order following FII SC3 to the Statement of Agreed Facts, but it appears to have had little impact on the Supreme Court’s order.

However, I agree with HMRC that litigation is not simply about determining the “right” remedy. A penny-perfect determination of remedy sometimes has to yield to considerations of proportionality and the efficient conduct of litigation. Therefore, I consider that it is quite possible for HMRC to suffer “prejudice” even if they end up having to pay the “right” amount if the process leading to that result is inefficient or disproportionate.

Ultimately, I conclude that Evonik would suffer a greater prejudice if it were held to its admission than HMRC would suffer if Evonik were permitted to withdraw that admission. I reach that conclusion largely because Evonik’s prejudice is relatively easy to discern from the evidence I have seen as it consists of obtaining judgment for its FID claim that may be lower than the actual amount due. HMRC’s prejudice, consisting of being “messed around”, is more difficult to quantify.

As regards the factors set out in CPR 14.5(f) and 14.5(g), HMRC argue that, if Evonik is permitted to withdraw its admission, there will be a live question as to what rate of Section 35A interest the court should, in its discretion, award. HMRC say that they would wish to argue that the court should select a rate that corresponds with Section 85 interest rates. Evonik’s position is that HMRC conceded, in a statement of agreed facts before the Supreme Court in 2020 that, whenever the court awards interest at Section 35A rates to taxpayers in the GLO, it should do so at the rate of the Bank of England base rate plus 2%. It also refers to a number of orders made in the GLO, including those relating to the BAT test case, have involved the award of Section 35A interest at that rate so that the measure of Section 35A interest in the GLO has been settled. It follows, argues Evonik, that its prospects of successfully obtaining judgment for the higher amount it seeks in relation to its FID Claim are high.

I see a conceptual difficulty with Evonik’s argument to the effect that, even if it is permitted to withdraw its admission, HMRC should necessarily be held to theirs. Moreover, interest under Section 35A involves the exercise of the court’s discretion as distinct from the determination of a fact which the parties are free to agree between themselves.

I heard only brief argument on the appropriate award of interest under Section 35A in relation to ACT arising on FIDs that was repaid after Evonik commenced proceedings and there was no evidence on the issue. I conclude that Evonik had the better of that argument. The court does indeed generally award Section 35A interest to large corporates at a rate of base rate plus 1% or 2% as a proxy for their assumed borrowing costs. Moreover, the obvious counter-argument to any suggestion that the court should apply a Section 35A rate of interest corresponding to the Section 85 rate is that Parliament has not legislated, as it could have done, to apply the Section 85 rate to ACT repaid after proceedings were commenced. I do not, however, consider that HMRC’s position is unarguable. HMRC accepted base rate plus 2% without knowing that Evonik would seek to apply that rate to ACT on FIDs that was repaid after proceedings were commenced.

A consideration of prospects of success therefore points in favour of releasing Evonik from its admission. However, I consider that considerations of the “administration of justice” point in the other direction. The courts and parties are now towards the end of litigation that has been going on for over 20 years. Although I accept that FII HC3 was a judgment on limitation issues only (and is itself under appeal), the reality is that, until Evonik came to the view that it had made a mistake (in June 2024), both Evonik and HMRC were close to agreeing a figure on which judgment could be entered. It is reasonable to suppose that the process of edging towards agreement on the numbers has involved a process of negotiation with each side’s position informed by the totality of that negotiation. If Evonik is now permitted, at a very late stage in the proceedings, to work loose a particular thread relating to its interest computation, I see some risk of other threads similarly being worked loose and some of the benefits of the parties’ significant agreement to date being lost. (In its application for permission to appeal submitted in response to the embargoed judgment, Evonik submitted that there is no evidence that any threads would be “worked loose”. However, HMRC’s suggestion that there should be a re-appraisal of the rate at which Section 35A interest is awarded in direct response to Evonik’s application to withdraw its admission indicates that the possibility is real).

Particularly in the context of complicated and long-running litigation such as this, I consider that the administration of justice favours holding Evonik to its admission so that HMRC and Evonik can continue their progress towards agreeing figures without a risk of “re-trading” towards the end of the process.

I have not found this issue straightforward. The considerations in CPR 14.5(a), (b), (e) and (g) point in favour of holding Evonik to its admission. The considerations in CPR 14.5(d) and (f) point in favour of permitting Evonik to withdraw it. Ultimately, I have concluded that the factors pointing in favour of holding Evonik to its admission are the weightier. Evonik’s application to be released from its admission is accordingly refused.