Kwik-Fit Group Limited & Ors v The Commissioners for HMRC

The Appellants are all members of the Kwik-Fit group of companies (the “Kwik-Fit group”). The appeal concerns a reorganisation of intra-group debt in which they participated in 2013, following the group’s acquisition in 2011 by a Japanese company, Itochu Corporation. Under the reorganisation a number of intra-group receivables owed by the Appellants were assigned to an intermediate holding company within the Kwik-Fit group, Speedy 1 Limited (“Speedy 1”), and certain additional receivables were created in Speedy 1’s favour. The interest rates on the amounts owed to Speedy 1 were also increased.

Speedy 1 had existing tax reliefs of around £48m, in the form of non-trading loan relationship deficits (“non-trading deficits”) carried forward from earlier periods. The Kwik-Fit group anticipated that the non-trading deficits would offset Speedy 1’s interest income, while the Appellants would generate tax relief in respect of their interest liabilities which could either be used by them or surrendered by way of group relief to reduce the taxable profits of other members of the group. As the First-tier Tribunal (“FTT”) found, the effect was to use up the non-trading deficits in under three years rather than the group tax manager’s previous estimate of around 25 years.

HMRC formed the view that the reorganisation engaged the unallowable purpose rule in s.441 Corporation Tax Act 2009 (“CTA 2009”). They largely disallowed the Appellants’ claims to tax relief on the interest in respect of the accounting periods ended 31 March 2014, 2015 and 2016, capping the disallowance at the amount of Speedy 1’s non-trading deficits.

The Appellants appealed to the FTT. The FTT (Judge Jeanette Zaman and John Woodman) agreed with HMRC about the existence of an unallowable purpose but allowed the appeal in part by permitting the deduction of interest on pre-existing loans to the extent that it corresponded to the rate at which interest was charged before the reorganisation ([2021] UKFTT 0283 (TC): the “FTT Decision”). The Upper Tribunal (“UT”) (Judges Swami Raghavan and Rupert Jones) agreed with the FTT, dismissing appeals by both the Appellants and HMRC ([2022] UKUT 00314 (TCC): the “UT Decision”). The Appellants now appeal to this Court.

We are grateful for the submissions of Mr Ghosh KC, leading Quinlan Windle and Laura Ruxandu, for the Appellants and Elizabeth Wilson KC, leading Ronan Magee, for HMRC.

As explained in the recent decision in BlackRock Holdco 5, LLC v HMRC [2024] EWCA Civ 330 (“BlackRock”), the unallowable purpose rule in s.441 CTA 2009 forms part of Part 5 of that Act, which contains the provisions governing the treatment of “loan relationships” for corporation tax purposes. There is no dispute that the receivables with which we are concerned in this case are loan relationships.

In very broad terms, Part 5 provides for credits and debits from loan relationships to be determined in accordance with generally accepted accounting practice. For a non-trader like Speedy 1, a net loss from loan relationships (being an excess of debits over credits) is a non-trading deficit (s.301 CTA 2009). For the periods with which we are concerned, a non-trading deficit that could not be used against other profits or surrendered by way of group relief in the year that it arose was automatically carried forward to offset future non-trading profits of the company in question pursuant to s.457 CTA 2009. Importantly, before the rules changed with effect from 1 April 2017 there was no scope for a carried forward non-trading deficit to be surrendered to other members of the group. Instead it was “trapped” in the company in which it arose and could be used only against its own non-trading profits.

Speedy 1’s brought forward non-trading deficits were often referred to in submissions and the decisions below as Speedy 1’s losses. That is a convenient shorthand that I will also adopt where appropriate.

Sections 441 and 442 CTA 2009 relevantly provide:

“Tax advantage” is defined in s.1139 Corporation Tax Act 2010 (“CTA 2010”) in the following terms:

In summary, loan relationships debits may not be brought into account to the extent that they are attributable, on a just and reasonable apportionment, to an unallowable purpose. Unallowable purposes are purposes which are “not amongst the business or other commercial purposes of the company”. Where the main purpose, or one of the main purposes, for which a company is a party to a loan relationship consists of securing a tax advantage for it or another person, that is treated as not being a business or other commercial purpose of the company, and is therefore an unallowable purpose.

It will also be necessary to refer to the transfer pricing rules in Part 4 of the Taxation (International and Other Provisions) Act 2010 (“TIOPA”).

Section 147 TIOPA relevantly provides:

Section 148, read with ss.157 to 163, has the effect that the “participation condition” is met if, among other things, one of the persons controls the other or the persons are under common control. The concept of “potential advantage” is defined in s.155 and includes both a reduction in profits and the creation or increase in losses for tax purposes.

Section 152 TIOPA deals specifically with debt financing between corporate entities. It requires s.147(1)(d) to be read as requiring account to be taken of all factors, including whether the loan would have been made at all at arm’s length, in what amount the loan would have been made and what the interest rate and other terms of the loan would have been.

Section 174 TIOPA relevantly provides:

The relationship between the transfer pricing rules and the loan relationships code is addressed in s.446 CTA 2009, which provides:

As can be seen, the effect of s.446 is that amounts imputed under the transfer pricing rules in respect of non-arm’s length loan relationships are taxed and relieved under Part 5 CTA 2009, in the same way as if they were actual profits, losses or interest.

Neither the transfer pricing rules nor s.446 have any direct application in this case because the relevant loan relationships were found to be on arm’s length terms, but the Appellants rely on them in their submissions as part of the relevant statutory framework.

The FTT had the advantage of a detailed statement of agreed facts, which is set out in Appendix 1 to the FTT Decision. It also heard evidence from Mr Kazushi Ogura, a director of each of the Appellants and their common parent European Tyre Enterprise Ltd (“ETEL”) during the relevant period, and Mr Glenn Andrews, the Kwik-Fit group’s tax manager, both of whom were found to be reliable witnesses. As the UT observed, the FTT’s findings were detailed and comprehensive. The fact that I do not refer to the same level of detail in this judgment should not be taken to detract from the very helpful nature of the FTT’s comprehensive and clear findings, which provide the key to determining the issues in this case.

Unless otherwise indicated, references below to paragraph numbers are to paragraphs of the FTT Decision.

By the time that Itochu acquired it, the Kwik-Fit group was highly leveraged as a result of funding structures put in place by previous owners. Speedy 1’s non-trading deficits were related to that. They were attributable both to external borrowings and to shareholder and intra-group loans (para. 16). The Itochu acquisition increased the total accumulated deficit still further, to £48m by the end of Speedy 1’s 2012 accounting period, because it triggered a requirement to prepay the external debt (para. 17). The Kwik-Fit group had also accumulated a large number of undocumented intra-group borrowings attributable to group charges and re-charges and to debt re-financings (para. 18).

Mr Andrews sought advice as to how the intra-group debt might be restructured so as to simplify it and allow Speedy 1’s non-trading deficits to be utilised more quickly than his 25 year estimate. The advice was sought initially from Ernst & Young and then from PwC. PwC produced a slide pack and advice and also a shorter document which was sent to HMRC (paras. 20 and 21).

In March 2013 there was a meeting between Mr Dominic Bartley of HMRC (the Kwik-Fit group’s Customer Relationship Manager), Mr Andrews and PwC to discuss the proposals, for which Mr Bartley’s meeting note was available (para. 22). The meeting was followed by correspondence in June 2013 in which Mr Bartley was informed about the proposed rate of interest, LIBOR plus 5%, following which Mr Bartley indicated that he was happy for the reorganisation to proceed and was content with the interest rate (paras. 25 and 26). The context for the approach to HMRC was Itochu’s “risk averse” approach to the group’s tax position. The FTT found that if Mr Bartley had given an adverse response then “the Kwik-Fit Group (including the Appellants) would not have undertaken the reorganisation in the form in which it in fact took place” (paras. 97 and 98).

The proposals were then developed further in a way that included the creation of new debts that had not been envisaged in the earlier proposals, and were implemented in September and October 2013 (paras. 27 and 28).

There are two relevant categories of debts that were owed by the Appellants prior to the reorganisation, as follows (para. 60):

In addition, three new loans came into existence during the reorganisation, as follows (para. 61):

KF GB initially challenged HMRC’s disallowance of interest on the £40m debt owed by KF GB but did not pursue that challenge before the FTT. We are therefore concerned with the £16m loan note issued by KF Finance and the £19m of the loan note issued by Stapleton’s which was used to satisfy the balance of the £35m debt owed by ETEL to Speedy 1 (together, the “New Loans”). However, HMRC also relied on the creation and treatment of the £40m debt as part of the relevant factual context.

As the FTT explained at para. 62, the effect of the steps referred to at [‎27] b) and c) above was that £35m of debt that had previously been owed by ETEL was “pushed down” such that KF Finance owed £16m to Speedy 1 and Stapleton’s owed £19m to Speedy 1. Overall intra-group indebtedness was not increased but the borrowers as well as the creditor changed.

The FTT found as follows about interest rates:

Leaving the impact of the transfer pricing rules to one side, it is obvious that the higher the interest rate charged by Speedy 1 (and the lower the rate charged on any debts owed by Speedy 1), the faster the non-trading deficits would be used up. As the FTT noted at para. 142 in the context of Mr Bartley’s understanding of the position, “the interest rate set by the group as well as the quantum of loans assigned to Speedy 1 would be key drivers of the rate of use of those losses”.

In the absence of challenge by HMRC, the FTT also found as a fact that LIBOR plus 5% was an arm’s length rate of interest (para. 49). However, later in its decision it rejected the Appellants’ explanation of the increase in the interest rate on the loans owed to Speedy 1 by reference to the mandatory requirements of the transfer pricing legislation in Part 4 of TIOPA, saying this:

As to the purpose of the reorganisation, the FTT recorded at para. 48 that it was accepted by the Appellants that a purpose of the reorganisation was to accelerate the use of Speedy 1’s non-trading deficits. It further found at para. 100 that the Appellants’ directors had received and understood a memorandum prepared in June 2013 (the “June 2013 Memorandum”). The memorandum is summarised at para. 99. It referred to a “long-term aim” of simplifying the group structure by eliminating surplus entities and described the first step as simplifying the intra-group funding structure. It stated that a “significant advantage” of doing so would be permit the use of Speedy 1’s non-trading deficits “and thus reduce the group’s total tax liability”. As to “Tax impact” it said:

The FTT also found that the directors had received a detailed steps and advice paper produced by PwC in August 2013 and “understood the consequences of the transactions set out in that paper” (para. 100).

The FTT then made the following findings:

The FTT also found as follows at para. 113 about the increase in interest rate:

At para. 116 the FTT concluded that the decisions to increase the rate of interest on the Pre-Existing Loans provided strong evidence that the Appellants had acquired a new purpose for being party to them, in addition to their existing commercial purpose. This was a tax avoidance purpose, and was important as it was integral to the steps taken. It was a main purpose. As to the New Loans, ETEL had incurred the original £35m debt for commercial purposes but KF Finance and Stapleton’s did not have their own commercial purpose (see above), and the intended tax advantages were not only a main purpose but the main purpose for which KF Finance and Stapleton’s were party to them (para. 117).

The FTT then went on to consider apportionment. Based on its conclusion that KF Finance and Stapleton’s did not have their own commercial purpose in entering into the New Loans and that tax avoidance was the main purpose, it attributed all the debits on those loans to the unallowable purpose (para. 133). For the Pre-existing Loans, it agreed with HMRC’s approach on the KFG Loan (in respect of which there had been no change of creditor), namely that debits should be disallowed in respect of the increase in rate, concluding that the decision to increase the rate was “solely attributable to the tax avoidance purpose of KFG” (para. 135). For the Assigned Loans, the interest on which had been disallowed by HMRC in full, the FTT decided that this did not reflect the Appellants’ (continued) commercial purpose and instead adopted the same approach as for the KFG Loan, disallowing the interest corresponding to the increased rate. It was on that basis that it allowed the appeal in part. The FTT also agreed at para. 138 with HMRC’s approach of capping the disallowance at the amount of Speedy 1’s non-trading deficits.

The Appellants appealed to the UT on the grounds that the FTT made an error of law in deciding that there was an unallowable purpose in respect of Speedy 1’s use of its non-trading deficits, and that as far as the Appellants’ own tax positions were concerned the FTT had confused knowledge with purpose, there being no evidence that they had the requisite purpose of increasing their own deductions. In the alternative, the Appellants maintained that the FTT erred in its approach to apportionment. HMRC appealed on the ground that the FTT erred in allowing interest at the pre-increase rate on the Assigned Loans.

The UT held that the interest debits claimed by the Appellants were tax advantages. It also held that Speedy 1’s use of its non-trading deficits was a tax advantage (paras. 42 to 46 of the UT Decision). It made no difference that losses were used up as a result (para. 47). The UT also rejected arguments based on the Sema and Kleinwort Benson cases discussed below, the transfer pricing legislation and the fact that there were no findings about specific tax savings (paras. 52 to 81).

The UT went on to decide that the FTT had been entitled to reach the conclusions that it had about main purpose. The Appellants’ attempt to contrast their accepted purpose of using Speedy 1’s non-trading deficits with their disavowal of any purpose in seeking deductions for themselves flew in the face of the rationale for the reorganisation, which was “to achieve a group tax benefit”. That required not only an increase in Speedy 1’s income but the debit generated in the entity paying interest to Speedy 1 to “be capable of being put to use, by being set off against income, either in the paying entity or elsewhere in the group” (para. 105 of the UT Decision). There was no unfairness in failing to put squarely to the witnesses that the Appellants had a purpose of securing deductions, which the witnesses had in fact confirmed was a key feature of the reorganisation (paras. 108 to 110). The interlinking between Speedy 1’s non-trading deficits and deductions for the Appellants also explained the logic of HMRC’s approach of capping the disallowance at the amount of the non-trading deficits (para. 111). Further, there was no error in the inferences drawn by the FTT about the selective approach to interest rate increases (para. 115), and no evidence to indicate that the FTT should have found that compliance with the transfer pricing rules featured in the Appellants’ subjective purposes (para. 117).

The UT also upheld the FTT’s decisions about apportionment, again rejecting the Appellants’ arguments about the impact of the transfer pricing legislation and concluding on HMRC’s appeal that it was open to the FTT not to disallow interest at the pre-increase rate under the Assigned Loans (paras. 133 to 146). Finally, the UT rejected an argument based on the “double-hit” to the Kwik-Fit group of using up Speedy 1’s non-trading deficits and being denied deductions under the unallowable purpose rule (paras. 147 and 148).

The issues raised by the grounds of appeal to this Court can be outlined as follows:

HMRC has filed a Respondent’s Notice which maintains that, in addition to the reasons given by the UT, the UT should have found that the Appellants’ witnesses did give evidence that the Appellants’ purpose was to secure deductions.

Mr Ghosh’s key submissions can be outlined as follows.

It was accepted that the reorganisation had a purpose of using Speedy 1’s brought forward losses. However, that was not a tax advantage because Speedy 1 was not better off as against HMRC as a result: IRC v Sema Group Pension Scheme Trustees [2002] EWCA Civ 1857, [2003] STC 95 (“Sema”).

It was also accepted, and the FTT found, that the Appellants knew that paying increased interest would result in deductible debits, further that they wanted such debits, expected them to arise and would not have engaged in the reorganisation if they had been told that deductions would not be available. However, knowledge and expectation did not amount to a (main) purpose: cf. IRC v Kleinwort Benson [1969] 2 Ch 221 (“Kleinwort Benson”), cited in Sema.

There was no evidence that obtaining the increased deductible debits was a purpose of the Appellants, rather than simply being within their knowledge. Such a purpose was not put to the Appellants’ witnesses in cross-examination. It would be unfair to infer purpose from knowledge without that being put straightforwardly to the witnesses. Rather (and until prompted by this Court to adopt a different approach during argument), HMRC’s case had always been put on the basis that the main purpose was to secure a tax advantage for Speedy 1, being the use of its losses. The concepts of “group tax benefit” and the “interlinkage” of “twin elements” of obtaining deductions and using losses were wrongly introduced by the UT. It had not been put to the witnesses that the use of Speedy 1’s losses was no more than a means to an end of generating tax relief for the group in the form of deductible debits.

This case is important. HMRC were seeking to apply the unallowable purpose rule to commercial loans on which interest was charged at an arm’s length rate. The fact that the effect of the transfer pricing rules was mandatory, requiring interest to be charged at an arm’s length rate on the debt, was relevant statutory context. Speedy 1 had used existing, genuine, losses against its interest income. What the group had done was precisely what the statutory code envisaged.

In the alternative, the FTT and UT erred in respect of apportionment. It was wrong to treat interest on a commercial loan, charged at an arm’s length rate in compliance with the requirements of the transfer pricing rules, as infected by a main tax avoidance purpose. The fact that HMRC were prepared to “cap” the disallowance at the amount of Speedy 1’s non-trading deficits was highly relevant. The Appellants’ position, and desire for deductions, did not alter once the deficits were used up. The “capping” showed that this case was all about Speedy 1’s losses.

If it was correct that the Appellants had main purposes comprising both the use of Speedy 1’s non-trading deficits and the creation of debits then, because the former did not give rise to a tax advantage and so could not be an unallowable purpose, the tribunals’ approach to apportionment was wrong on that basis. Alternatively, if there was a single purpose involving the interlinking suggested by the UT and the use of losses as a means to an end of generating tax relief then that had not been raised with Counsel or put to the witnesses. It would require the decision to be remade and would be unfair.

BlackRock is obviously a very recent decision of this Court that considers the concept of unallowable purpose in some detail. As in that case, there was no dispute that in interpreting ss.441 and 442 CTA 2009 it is appropriate to apply the summary by Newey LJ (with whom the other members of the court agreed) in Travel Document Service v HMRC [2018] EWCA Civ 549, [2018] STC 723 at [41], in relation to the predecessor legislation in para. 13 of Schedule 9 to the Finance Act 1996. That summary is as follows:

Further, and again as in BlackRock, it is uncontroversial that what matters is the company’s subjective purpose or purposes in being a party to the loan relationship in question, and that for a corporate entity it is necessary to consider the subjective purpose of the relevant decision makers, generally the board of directors (BlackRock at [106]-[108]).

It is also worth reproducing the summary in BlackRock at [124] of key points relevant to ascertaining purpose:

Sema concerned a claim by the trustees of a pension scheme for the payment of tax credits arising on buy-backs of shares by Powergen plc. The Inland Revenue sought to counteract the perceived tax advantage under the transactions in securities rules then contained in Chapter 1 of Part XVII Income and Corporation Taxes Act 1988 (“ICTA 1988”).

The trustees’ appeal against the refusal of the tax credits was allowed by the Special Commissioners but their decision was reversed by Lightman J. This Court allowed the trustees’ appeal on the basis that they had not received an “abnormal amount by way of dividend” and so did not fall within the legislation on that basis. However, relevantly for our purposes Jonathan Parker LJ, with whom Aldous LJ and Aikens J agreed, went on to consider the concept of “tax advantage”, which was defined in s.709(1) ICTA 1988 in a virtually identical way to the definition in s.1139 CTA 2010. As is clear from Jonathan Parker LJ’s judgment at [106] his comments on that issue are therefore obiter, but nonetheless they should be afforded considerable respect. Ms Wilson also did not submit that we should not adopt them because they were wrong, rather that they were made in the context of the facts of Sema and that s.1139 is a provision that is applied in various parts of the corporation tax code, so that it should not be inappropriately cut down.

While I understand Ms Wilson’s points, for my part I would endorse Jonathan Parker LJ’s comments on the meaning of tax advantage and apply them to s.1139 CTA 2010. I do not consider that they should be confined by reference to the facts of Sema.

The issue arose in Sema because Mr Gardiner QC, for the trustees, had argued that the entitlement to a tax credit which arose by virtue of the pension scheme’s exemption from tax was not an advantage that fell within the concept of a “relief” for the purposes of s.709(1). He had also submitted that it was not a “repayment” because that presupposed a prior payment of tax, such as where tax is deducted at source.

Jonathan Parker LJ responded as follows:

Mr Ghosh also relies on further obiter comments by Jonathan Parker LJ on the question of whether the trustees had a main object of obtaining a tax advantage. Mr Gardiner had submitted that the conclusions reached below contradicted the rationale of what Cross J had said in Kleinwort Benson and were wrong given that a decision had previously been made to sell the Powergen shares for investment reasons, and the buy-backs simply provided a golden opportunity to implement that decision.

Jonathan Parker LJ referred at [113] and [114] to Lord Upjohn’s speech in IRC v Brebner [1967] 2 AC 18, 30, where Lord Upjohn had explained that the test was a subjective one to be decided on all the evidence and the proper inferences to be drawn from that evidence. He concluded that the Special Commissioners had been entitled to decide that a main object of the sales by the trustees was to obtain tax advantages. The sales were made when they were only because it was known that the tax credits were available: the tax credits were crucial because otherwise there would have been a loss. As to Kleinwort Benson, at [119] Jonathan Parker LJ agreed with the comments of Lightman J in his judgment in the High Court ([2002] STC 276) at [53], adding that:

In Kleinwort Benson the taxpayer bank bought debenture stock at a price which reflected significant arrears of interest. The outstanding interest was then paid to the bank but under the tax rules then in force it did not have to bring that receipt into its tax computation. The effect was to create a loss on the subsequent sale of the stock, which the Revenue sought to counteract under the transactions in securities rules. As in Sema, Cross J decided that the transaction did not fall within those rules so his comments on main object are obiter. The observations made by Cross J to which Jonathan Parker LJ’s judgment refers are set out in Lightman J’s decision in Sema at [52], and read as follows:

Lightman J’s comments on this which were approved by the Court of Appeal were as follows:

Mr Ghosh’s submissions were skilfully and forcefully put, but I cannot accept that the FTT was wrong to decide that the unallowable purpose rule applied to restrict the availability of relief on the Pre-existing Loans and New Loans, or that the UT was wrong not to disturb the FTT’s conclusion.

The fundamental premise of Mr Ghosh’s submissions is a distinction that he seeks to draw between the (admitted) purpose of utilising Speedy 1’s losses and the Appellants’ knowledge that deductible debits would be generated in the Appellants. In reality no such distinction can be drawn.

The relevant decision-makers in this case were the directors of the Appellants. The FTT had the benefit of clear evidence from one of their number, Mr Ogura, and from Mr Andrews, whose advice (and that of the firms he involved) the directors both understood and followed.

The FTT rightly focused on the Appellants’ decisions to participate in the reorganisation, and their purposes in doing so. Commercial decision-makers such as Mr Ogura would have no real interest in using up Speedy 1’s losses as an end in itself. It would achieve nothing of benefit to the group, for the very reason that Mr Ghosh relies on in submitting that the use of existing losses is not a tax advantage. Compared to a scenario where no reorganisation is implemented, neither Speedy 1 nor any other entity in the Kwik-Fit group would be better off vis-à-vis HMRC solely as a result of using those losses. Indeed, in the longer term the group could be worse off because the losses would no longer be available to shelter other income.

It is obvious that the decision-makers, and both tribunals, understood that the real benefit, being the one that would make the group better off as against HMRC, was not the receipt of interest income in Speedy 1 as such but the concomitant generation of interest deductions in the Appellants. It was these deductions that provided the tax saving to the group, because they were available for use either by the Appellants or by other members of the group via group relief surrenders, without needing to be used to offset any tax on the corresponding income in Speedy 1. The arrangements had the economic effect of releasing the non-trading deficits trapped in Speedy 1 for use by the group as a whole. Those losses could not be accessed directly under the rules at the time, so it was necessary to do so indirectly by creating deductions elsewhere and using Speedy 1’s losses to absorb the resulting income. It was the creation of what were, in effect, “net” deductions, being deductions where the corresponding income did not create a tax charge, that would produce a real saving for the group. This was what the reorganisation was intended to achieve, and was the reason why the Appellants took part in it.

That Mr Ogura and the other directors understood this is an obvious inference. The FTT did not need to spell the point out in quite the same way as I have done because the case was not put in that way before it. But it is, and was, obvious. It is reflected in the first paragraph of the June 2013 Memorandum, which as the FTT recorded referred to the “significant advantage” of permitting use of Speedy 1’s non-trading deficits “and thus reduc[ing] the group’s total tax liability”. The “Tax impact” paragraph set out at [‎33] above explains in terms how that reduction was to be achieved.

The point is also neatly illustrated by an email sent by Mr Andrews to Mr Ogura on 11 June 2013, shortly before the June 2013 Memorandum was finalised. The email referred to a reorganisation “so as to utilise the brought forward losses in Speedy 1”, attached a draft of the paper setting out the proposed steps and discussed professional fees, comparing them with the fact that “as a minimum we would save cash tax of £1.3m per annum for a period of 5-8 years” based on an interest rate of 1.5%, with the savings being greater if the rate was higher. This saving could only be derived from the use of the deductions available to the Appellants, in circumstances where the group did not have to concern itself with covering a tax charge on the corresponding income.

The FTT’s findings of fact reflect this. Paragraph 101(2) bears repeating, with emphasis added:

At least until prompted by this Court to consider an alternative approach, HMRC has throughout sought to maintain that the reorganisation generated two separate tax advantages which the Appellants had a main purpose in securing, one for Speedy 1 (being the use of its losses) and one being the generation of relief in the Appellants. With respect, I think that has been the cause of unnecessary difficulties in this case. It has both over-complicated the analysis and left it open to the Appellants to run the case that they have before us.

I see significant force in Mr Ghosh’s submission that, on the facts of this case, the reorganisation did not create a tax advantage for Speedy 1. Speedy 1 had existing losses which arose in an unobjectionable way. HMRC say that it is the setting of those losses against income which is the relevant tax advantage, because that avoids or reduces a charge (or possible charge) to tax within s.1139(2)(c) or (d) CTA 2010. But without the reorganisation Speedy 1 would not have a charge to tax that could be reduced, because it would not have the relevant income. In order to identify a tax advantage it is necessary to postulate the existence of income, the tax charge on which is then sought to be sheltered. Echoing Jonathan Parker LJ’s comments in Sema it is only in that sense that Speedy 1 is better off as against the Revenue.

The distinction is important because what we are considering is whether the Appellants agreed to participate in the reorganisation with a main purpose of securing a tax advantage. In the context of Speedy 1 that most naturally requires a comparison of its tax position with and without the reorganisation, not (as HMRC’s approach involves) assuming that the reorganisation has occurred but comparing the position with or without the availability of Speedy 1’s unobjectionable losses.

The tax advantage sought to be secured by the reorganisation was the generation of cash tax savings through the creation of deductible interest expense in circumstances where there was no need to worry about a tax charge on the income. The relevance of Speedy 1’s losses is just that: it makes the deductions of real value to the group because the corresponding income is sheltered by reliefs that could not otherwise be used. That is why it was entirely appropriate for HMRC to have “capped” the disallowance at the amount of Speedy 1’s brought forward losses. Once they are used up the value of the deductions to the group is offset by the need to cover the tax charge on the income.

It is right that much of the evidence, and cross-examination, in the FTT focuses on using Speedy 1’s brought forward losses. But as already discussed that was clearly with a view to the savings that would be available from the use of the deductions arising in the Appellants. It follows that, with respect, the distinction drawn by Mr Ghosh between a purpose of using Speedy 1’s losses and knowledge that a tax deduction would be available is an unreal one on the facts of this case. In effect, referring to using Speedy 1’s losses was a convenient, and understood, shorthand for a description of arrangements which had the effect of creating a real tax saving for the group.

The FTT found that there was a purpose as described at para. 101(2). In my view, and bearing in mind that the Appellants’ case was not put to the FTT in the way in which it was put to us, that paragraph and the FTT’s other findings amply demonstrate that it concluded that the generation of deductions, without tax on the corresponding income, was found to be a main purpose of the reorganisation.

It is true that the FTT also stated at paras. 109 and 110 that a main purpose of the Appellants was to secure a tax advantage for Speedy 1, being the accelerated use of its losses. But it clearly understood this in the context of the corresponding deduction in the Appellants, for example referring at para. 113 to the consequence of the increased rate being “that the Appellants benefitted from greater interest debits and Speedy 1 was able to use its NTDs over a shorter period of time” and concluding that “the only reason” for agreeing to the increase was “to secure the intended tax advantages for themselves and for Speedy 1”, and at para. 117 to the main purpose for which KF Finance and Stapleton’s were party to the New Loans being “to secure the intended tax advantages for Speedy 1 and themselves”.

The FTT was not asked to distinguish between the two asserted tax advantages and rightly considered the tax impact of them in combination. Put another way, it found a single unallowable purpose, which in shorthand terms was the “accelerated” use of Speedy 1’s losses, but in longhand comprises the generation of the deductions for the benefit of the Appellants or other group members in circumstances where the corresponding income was sheltered from tax by Speedy 1’s brought forward non-trading deficits, so in effect freeing those trapped reliefs for use by the Appellants or other members of the group.

I would prefer not to describe the use of Speedy 1’s non-trading deficits as (by itself) a tax advantage intended to be secured by the reorganisation. I consider that that introduces a complexity into the description of the Appellants’ purpose which is both unnecessary and apt to confuse. It involves the application of a technical tax analysis that reflects the detailed mechanics of how the legislation works rather than how decisions are taken and expressed in the real world. However, in my view that is a detail rather than a material flaw in the FTT’s approach.

It largely follows from what I have already said that I do not consider that there was any unfairness in the UT reaching conclusions based on the “group tax benefit”, or in this Court not basing its analysis on that part of HMRC’s case that relates to an aim of securing a tax advantage in Speedy 1. The FTT referred in terms to the benefit to the group at paras. 101(4) and 102(5), set out above. The UT rightly agreed. Further, it was obviously understood by both witnesses that it was, and could only be, the combination of the deduction in the Appellants and the sheltering of income by losses in Speedy 1 that would allow the group benefit to be obtained.

Moreover, it is clear even from the passages of the transcript on which Mr Ghosh relied that Ms Wilson put the relevant points to Kwik-Fit’s witnesses in cross-examination. It is true that there are plenty of references to utilising or accelerating the use of trapped losses, but this was in the context of the benefit to the group, which as already discussed could only be derived from the generation of interest deductions in the Appellants. For example, after agreeing that the purpose of the reorganisation was to “utilise these trapped losses in Speedy 1” Mr Ogura also confirmed that he understood that “one part of that process was that the debtor would pay more interest and it would get a corporation tax deduction for the interest that it was paying”. Even more clearly, Mr Ogura agreed with the following (emphasis supplied):

As in BlackRock, it also bears emphasising that, while ascertaining the object or purpose of something involves an inquiry into the subjective intentions of the relevant actor, it is for the fact finding tribunal to determine the object or purpose sought to be achieved, and that question is not answered simply by asking the decision maker (BlackRock at [124f)]). It was for the FTT to reach its own decision on whether there was an unallowable purpose based on all the evidence before it. The case did not depend on obtaining a concession in particular terms in cross-examination, or indeed on framing a question to a witness in a particular way.

Mr Ghosh was right to emphasise that the Pre-existing Loans had a commercial purpose, which the FTT accepted at paras. 102(1) and 112 subsisted throughout. Further, the FTT found that interest was charged at an arm’s length rate. Combined with the features that deductions for interest at an arm’s length rate are specifically contemplated by the loan relationships code, that Speedy 1’s non-trading deficits arose in unobjectionable circumstances and that the legislation specifically provides for their offset against loan relationship profits, the conclusion that the unallowable purpose rules restricted the deductions might on a superficial analysis be considered to be a surprising one.

I would agree with Mr Ghosh that in most circumstances of that kind there would be no question of the unallowable purpose rule applying. As I explained in BlackRock at [150], it cannot have been Parliament’s intention that the unallowable purpose rule will be engaged as an inevitable consequence of taking out (or, I would add, maintaining) a loan, or indeed charging interest on it at a commercial rate, subject only to consideration of whether the value of the tax benefits are sufficient to make it a “main” purpose. The mere fact that a group organises its affairs in a manner that makes use of brought forward non-trading deficits and that it expects to obtain relief for interest and other expenses of loan relationships, in each case as the legislation contemplates, cannot be enough to engage the unallowable purpose rule.

The comments of Cross J in Kleinwort Benson, set out at [‎62] above, reflect a similar point in a different context. The tax consequences he was describing for a dealer (or charity in his other example) were specifically contemplated by the legislation.

However, as this Court recognised in Sema, Cross J’s comments in Kleinwort Benson must be understood in the light of the facts of that case. Lightman J rightly emphasised at [53] of his decision in Sema (the paragraph approved by the Court of Appeal in that case) that the significance of the tax advantage to the taxpayer must be considered with care. I would add that it should also be considered in the context of the relevant legislative code. As Lightman J also explained, there is a range of possibilities. The possibilities include that the tax advantage may be a “feature” or a “relevant factor” without being a main object. (I would take the opportunity to clarify that Lightman J was not saying in the preceding sentence that anything that is more than “icing on the cake” will be a main object, rather that if it is no more than that then the answer is obvious that it will not be.) But the important point is that whether a purpose is a main purpose is a question of fact for the fact-finding tribunal, which cannot be interfered with in the absence of an error of law.

In this case, the FTT’s conclusions were based on very particular factual features:

The FTT’s analysis at para. 115 of the transfer pricing position ([‎32] above) can also not be faulted. Mr Andrews’ suggestion that loans not involved in the reorganisation were not transfer priced because it would be “tax neutral” rather makes the point for HMRC. The interest rates on loans owed to Speedy 1 was increased because a cash tax saving could be achieved – and indeed the higher that rate, the faster the saving – but a similar benefit could not be obtained in respect of other loans. Indeed, increasing the rate on the Detailagent Loan would slow the rate at which the benefit would be achieved, as the FTT explained.

It is apparent from this that the Kwik-Fit group chose not to comply with the strict requirements of the transfer pricing rules, which would have required it to adjust the rate on, among other things, the Detailagent Loan to an arm’s length rate. Against that background, its submissions based on the requirements of the transfer pricing rules ring somewhat hollow. This is not a case where, for example, the group had reviewed its transfer pricing position on intra-group debt and decided that it should adjust interest rates generally to ensure compliance. The only reason that LIBOR plus 5% was selectively applied to debts owed to Speedy 1 was to achieve the anticipated tax benefits.

The Appellants are of course correct to say that the transfer pricing rules are mandatory. If Speedy 1 had charged less than an arm’s length interest rate on the relevant loans, it should, as a “potentially advantaged person”, nonetheless have filed its tax returns as if it had. That would have enabled each of the Appellants, as “the disadvantaged person”, to claim relief under s.174 TIOPA in the form of correspondingly increased deductions. Section 446 CTA would have ensured that the amounts in question were taxed and relieved under the loan relationships regime, as if they had actually accrued. But none of that occurred, and further that approach was not applied to other loans, including the Detailagent Loan.

More importantly, what we are concerned with under the unallowable purpose rule is the Appellants’ subjective purposes. Except insofar as the group sought to ensure that the rate would not be challenged as being excessive, it is clear that the transfer pricing rules played no part in the decision making process. Those rules therefore cannot assist in determining the Appellants’ purposes. I see no statutory justification for interpreting ss.441 and 442 in a way that requires the transfer pricing rules to be taken into account when they were actually disregarded.

One aspect of Issue 1b was that HMRC had at no stage identified any tax saving achieved by the generation of the deductions in the Appellants. The Appellants say that three of the Appellants were in fact loss-making. The possibility of other members of the Kwik-Fit group making using of the deductions by way of group relief did not feature before the FTT at all, and the FTT confined its conclusions to the securing of tax advantages for the Appellants themselves and Speedy 1 (see for example para. 113, set out at [‎36] above).

This was obviously not an issue raised before the FTT, and I do not consider that HMRC was either obliged to raise it or the FTT to deal with it if it was not put in issue. But in any event I accept the submissions of Ms Wilson that there is no requirement to identify either a specific quantum of tax saving or the precise identity of the beneficiaries of that saving before concluding that an unallowable purpose exists.

First, identifying purpose requires a consideration of subjective intentions. That is necessarily a forward-looking exercise: broadly, what is sought to be achieved, not what is ultimately achieved. It is perfectly obvious that the group sought and expected to make material tax savings.

Secondly, s.442(5) provides that references to a tax avoidance purpose “are references to any purpose which consists of securing a tax advantage for the company or any other person”. This needs to be read in the context of the legislation of which it forms part. That includes the fact that, because profits and losses from loan relationships are determined in accordance with generally accepted accounting practice (and further, for a trading company, generally form part of its trading computation: s.297 CTA 2009), it may well be unclear until after the end of the relevant accounting period whether any debits will be absorbed by income in the company or whether they will generate excess reliefs available for surrender to other group companies or for carry forward. Further, the group relief position may well not be settled until well after the end of the period. This is illustrated by provisions which allow group relief claims to be made and adjusted well after the filing date for group members’ tax returns: see para. 74 Schedule 18 Finance Act 1998.

It would frustrate the obvious aim of the unallowable purpose rule if it were confined to identified amounts saved by a specified person. Rather, as Ms Wilson submitted, and bearing in mind that under s.6 Interpretation Act 1978 the singular will include the plural unless a contrary intention appears, s.442(5) should properly be read as referring to the company in question and/or any other person or persons. In this case the FTT found at para. 101 that the aim was to benefit the “whole group”. I consider that that was a sufficient identification of persons for the purposes of s.442(5). The persons in question were the Appellants and other members of the Kwik-Fit group.

Issue 2 is that the FTT and the UT erred in their application of the just and reasonable attribution provision.

The correct approach to just and reasonable apportionment pursuant to s.441(3) CTA 2009 was considered in BlackRock at [179] and [180]. In summary, it is an objective exercise which requires apportionment by reference to the relevant purposes. The exercise is a fact specific one. The facts in BlackRock did not permit any element to be apportioned to the commercial purpose ([182]). The same result was reached if a “but for” approach was adopted, as the parties had suggested ([185]).

I can see no legal error in the FTT’s approach to apportionment in this case. Mr Ghosh’s submissions proceeded on the basis that apportionment should be by reference to the relevant “main” purposes rather than any ancillary one (which in this case might include simplifying group balances: see [‎88b)] above). He also proposed a “but for” approach. As the FTT noted at para. 136, this is essentially the approach that it adopted.

KF Finance and Stapleton’s were found to have had no commercial purpose in entering into the New Loans. That entitled the FTT to attribute all the debits on those loans to the unallowable purpose. Given its findings of fact, the FTT was also fully entitled to attribute all the debits arising from the increase in rates on the Pre-existing Loans to the unallowable purpose. For the reasons already discussed the transfer pricing rules do not assist the Appellants.

However, HMRC correctly accepted that once Speedy 1’s losses were used up, it would no longer be just and reasonable to deny relief for the debits. This reflects the nature of the tax advantage sought to be secured, as already described (see [‎79] above).

In conclusion, the FTT was entitled to reach the conclusions that it did and the UT was right not to disturb them. I would dismiss the appeals.

I agree.

I also agree.