The Union Castle Mail Steamship Company Ltd v HM Revenue and Customs & Ors

These appeals are concerned with corporation tax and derivative contracts. Although this is a combination likely to appear daunting, the issues of statutory construction that arise are not complex but they are important. The most important is whether, as counsel for the appellant taxpayers graphically submitted, Parliament has “surrendered to accountants” the determination of the taxable profits and allowable losses resulting from derivative contracts.

The issue is whether the “derecognition” in the accounts of the appellant companies of 95% in one case, and 100% in the other, of the value of derivative contracts held by them respectively gave rise to an allowable loss for the purposes of corporation tax. This issue turns on the proper construction and application of schedule 26 to the Finance Act 2002, which contains an exhaustive code for the taxation of profits arising from derivative contracts.

The appellants are The Union Castle Mail Steamship Company Limited (Union Castle) and Ladbrokes Group Finance plc (Ladbrokes).

Union Castle is a wholly owned subsidiary of Caledonia Investments plc (Caledonia), a publicly quoted company with investment trust status. HMRC disallowed a deduction of £39,149,128 made by Union Castle in its corporation tax return for the year to 31 March 2009, claimed as a result of a derecognition of 95% of derivative contracts held by it. Its appeal against the closure notice issued by HMRC was dismissed by the First-tier Tribunal (the FTT) in a Decision dated 27 July 2016. Its appeal to the Upper Tribunal (the UT) was dismissed, albeit on different grounds. It appeals to this court with permission granted by the UT.

Union Castle’s appeal to the FTT was designated as a lead case for two other appeals, one being the appeal by Ladbrokes. HMRC had issued closure notices disallowing deductions in its 2008 and 2009 tax computations for lossesresulting from the derecognition of derivative contracts. There were common issues and, in addition, an issue that applied only to Ladbrokes (the Gateway issue). On the Gateway issue, the FTT held in favour of Ladbrokes, with the result that its appeal was allowed. The UT reversed that decision. Ladbrokes appeals, with permission granted by the UT. I deal with Ladbrokes’ appeal at the end of this judgment.

The facts as they relate to Union Castle were not in dispute before the FTT or the UT and were set out in an agreed statement. They were helpfully summarised by the UT in their Decision at [10] which I gratefully adopt:

“(1)

Prior to 21 November 2008 Union Castle had issued share capital consisting of 502 shares of £1 each, fully paid, held by Caledonia.

(2)

From about May 2007, the board of Caledonia wished to implement a hedging strategy, using put options against a FTSE index. The board was concerned about a possible substantial fall in UK equity markets.

(3)

The board was concerned that purchase of such put options might prejudice Caledonia's investment trust status. Accordingly it was envisaged that Union Castle might purchase the put options instead.

(4)

Between 20 June and 31 December 2007, five FTSE put options at an aggregate cost of £10 million were acquired by Union Castle, and a further put option was acquired in January 2008 at a cost of £2 million.

(5)

In July 2008, accounting guidance for investment trusts and venture capital trusts clarified their right to invest in derivatives, such that it appeared that Caledonia could safely hold such investments in its own name.

(6)

During the financial year ending 31 March 2009, some of the put options were exercised and further put options were purchased. As at 31 October 2008 Union Castle held three put options and three put spreads ("the Contracts").

(7)

On 19 November 2008, Caledonia's audit committee considered novating the Contracts from Union Castle to Caledonia but realised that this would crystallise a tax charge in Union Castle owing to the current value of the Contracts. The committee therefore considered the possible issue by Union Castle of a new kind of share capital to Caledonia with dividend rights, whereby the economic benefit of the Contracts would effectively be transferred to Caledonia. They noted that this would oblige Union Castle to write off the value of the Contracts, thereby crystallising a tax loss.

(8)

On November 2008, Union Castle made a bonus issue to Caledonia of 5020 "A Shares", ten for every one existing ordinary share held by Caledonia.

(9)

The A Shares carried a right to receive a dividend equal to 95% of the cash flows arising on the close-out of the Contracts, such dividend to be paid within five business days following receipt by Union Castle of the cash flows.

(10)

As a consequence of issuing the A Shares, Union Castle was required to "derecognise" 95% of the value of the Contracts for accounting purposes, amounting to £39,149,128.

(11)

Between January and August 2009 Union Castle closed out the Contracts for aggregate proceeds of £25,042,545 and paid dividends to Caledonia in a sum equal to 95% of those cash flows.

(12)

On the issue of the A Shares, the following debits and credits were recognised by Union Castle:

Cr Financial asset £39,149,12825

Dr income statement £39,149,128

Cr share capital £5,020

Dr share premium £5,020

(13)

The A Shares were added to Caledonia's investment ledger as a new security, with no cost attributed, but they were ascribed at fair value, reflecting the "pass-through" right to 95% of the future cash flows from the derivatives. Caledonia did not include an entry in its income statement, but reallocated a part of the fair value from the Ordinary Shares in Union Castle to the A Shares.

(14)

Union Castle agreed for the purpose of the proceedings that its accounting treatment in accordance with GAAP should more appropriately have debited the value of the cash flows to the statement of changes in equity rather than to income.”

Despite the agreement recorded in sub-paragraph (14), expert evidence was adduced by both sides before the FTT which addressed, among other issues, whether the debit was more appropriately taken to profit and loss account, rather than to the statement of changes in equity. The FTT accepted the evidence of Union Castle’s expert that the former was the more appropriate treatment, although the latter could not be said to be wrong. By reason of the relevant provisions of schedule 26, this makes no difference to the outcome of the appeal and it is unnecessary to consider it further.

The rights attached to the A Shares as regards income were expressed as follows: “out of the profits available for distribution the holders of the A Shares shall be entitled to be paid a dividend equal to 95% of each of the option cash settlements (if any) received by the Company under” and there were then identified each of the relevant derivative contracts. The rights further provided that “unless the Company has insufficient profits available for distribution and the Company is thereby prohibited from paying dividends by the [Companies Act 2006], the dividends payable on the A Shares…shall be paid without undue delay and in any event within five business days following receipt of each of the option cash settlement amounts”. The dividends were payable without the need for any resolution of either the directors or the company in general meeting.

These rights acknowledged that dividends are not lawfully payable except in accordance with Part 23 of the Companies Act 2006 (CA 2006). Part 23 requires, among other things, that a company’s accounts must show distributable profits at least equal to a proposed dividend.

In order to ensure so far as possible that Union Castle would have distributable profits available to pay the dividends on the A Shares, Caledonia provided a letter dated 21

Union Castle, as a company incorporated under the Companies Act 1985, was required to prepare and file annual accounts. The directors had to be satisfied that such accounts “give a true and fair view of the assets, liabilities, financial position and profit or loss” of Union Castle: section 393(1) CA 2006. A company, with some exceptions, may elect to prepare its accounts either in accordance with section 396 CA 2006 or in accordance with international accounting standards: section 395(1) CA 2006. If accounts are prepared in accordance with section 396, they must be prepared in accordance with requirements laid down in regulations and, in order to comply with section 393(1), it is generally taken that they will give a true and fair view if they are prepared in accordance with the Financial Reporting Standards issued by the Financial Reporting Council (or until July 2012 the Accounting Standards Board), unless exceptionally the directors consider that departure from those standards is necessary in order for the accounts to show a true and fair view: see the discussion in GDF Suez Teeside Ltd v HMRC [2017] UKUT 68 (TCC) (GDF Suez) at [62]-[69]. In their totality, these requirements constitute UK GAAP.

In the case of accounts prepared in accordance with international accounting standards (IAS), those standards are defined by section 474(1) to mean the standards from time to time adopted by the European Commission in accordance with EC Regulation 2016/2002. The standards so adopted are the IAS, or International Financial Reporting Standards (IFRS), promulgated by the International Accounting Standards Board. IAS 1 (Presentation of Financial Statements) sets out the overall requirements of financial statements. The financial statements must “fairly present” the financial position, financial performance and cash flows of a company and the application of the various IAS is presumed to achieve a fair presentation. Where compliance with a particular IAS would be so misleading as to conflict with the purpose of financial statements, the directors must depart from it, with appropriate disclosure of the nature, reasons and impact of the departure.

As a listed company, Caledonia was required to prepare its group accounts in accordance with IAS, and as a subsidiary Union Castle also prepared its individual accounts in accordance with IAS.

This appeal is in large part concerned with the effect of IAS 39 (Financial Instruments: Recognition and Measurement).

Accounts properly prepared either in accordance with section 396 or in accordance with international accounting standards are said to be prepared in accordance with generally accepted accounting practice (GAAP). This is reflected as regards the taxation of companies’ profits and gains in section 50 of the Finance Act 2004:

“(1)

In the Corporation Tax Acts “generally accepted accounting practice” means

(a)

in relation to the affairs of a company or other entity that prepares accounts in accordance with international accounting standards (“IAS accounts”), generally accepted accounting practice with respect to such items;

(b)

in any other case, UK generally accepted accounting practice.

(2)

In the Corporation Tax Acts “international accounting standards” has the same meaning as in regulation (EC) no 1606/2002 of the European Parliament and the Council of 19 July 2002 on the application of international standards.”

For convenience, I will refer to accounts prepared in accordance with either IAS or UK GAAP as GAAP-compliant accounts.

In what follows, I set out or summarise the relevant paragraphs of schedule 26, as they applied in the tax year 2008/09, largely as the UT did in its Decision at [12] – [20].

Paragraph 1(1) states:

Paragraph 1(2) provides that Schedule 26 is the exclusive provision for charging such profits to corporation tax.

Paragraph 2 defines “derivative contracts” for these purposes, and the definition includes options, futures and contracts for differences. There is no dispute that the contracts in question of Union Castle and Ladbrokes were derivative contracts, as defined, to which IAS 39 applied.

Paragraph 14 provides (so far as material):

The remainder of paragraph 14 then makes different provision depending on whether or not the company is a party to a derivative contract for the purposes of a trade carried on by it.

Paragraph 15 of Schedule 26, which is central to the disputed issues, provides (so far as material):

Paragraph 17A(1) provides:

Paragraph 17B(1) provides:

Part 6 of Schedule 26 contains special computational provisions. Of these, paragraph 25A, headed "Debits and credits recognised in equity or shareholders' funds", states:

The issues arising on schedule 26, as they were argued before both Tribunals, were as follows:

Reversing the FTT on the “loss” issue, the UT held that the debit attributable to the derecognition was a “loss” within the meaning of schedule 26. On the “arise from” issue, the UT reversed the FTT’s decision that any loss did arise from derivative contracts for the purposes of paragraph 15. Agreeing with the FTT, the UT held that, if (contrary to their decision on the “arise from” issue) the relevant loss did arise from the derivative contracts, the resulting debits would have fairly represented those losses. Thus, although the UT held in favour of Union Castle on the “loss” and “fairly represent” issues, it held in favour of HMRC on the “arise from” and dismissed Union Castle’s appeal.

HMRC ran an alternative case against Union Castle before the FTT and the UT that, if HMRC failed under schedule 26, the issue of the bonus shares by Union Castle fell within the scope of the transfer pricing rules in schedule 28AA to the Income and Corporation Taxes Act 1988, thereby reducing to nil the amount deductible for the debit. In view of the decisions of both Tribunals in favour of HMRC under schedule 26, this issue was not determinative of the appeals. Nonetheless, both Tribunals quite properly decided this issue, in case it should be material on an appeal. Disagreeing with the FTT, the UT held that the issue of the shares was a “provision” within schedule 28AA which was therefore capable of applying in Union Castle’s case. In view of the conclusion to which I (and, as I understand it, the other members of the court) have come on the issues under schedule 26, it is not necessary to decide HMRC’s alternative case under schedule 28AA and we did not hear argument on it.

As will later appear, the “fairly represent” issue cannot, in my judgment, be divorced from the “loss” issue and the “arise from” issue. The task for the court on these issues is to construe and apply paragraphs 14 and 15, having regard to the other paragraphs of schedule 26, particularly paragraphs 17A and 17B.

Paragraph 14(1) contains the general provision that for the purposes of the charge to corporation tax imposed by paragraph 1, the profits and losses arising from the derivative contracts of a company are to be computed “using the credits and debits given for the accounting period in question by the following provisions of this Schedule”. Leaving to one side the Gateway issue, these words lead to paragraph 15 and, in particular for present purposes, to paragraph 15(1).

Paragraph 15(1) provides that “[t]he credits and debits to be brought into account” in respect of a company’s derivative contracts are “the sums which, when taken together, fairly represent, for the accounting period in question (a) all profits and losses of the company which (disregarding any charges or expenses) arise to the company from its derivative contracts and related transactions”.

In determining “the amounts to be brought into account by a company for any period for the purposes of this Schedule”, paragraph 17A(1) provides that, subject to the provisions of the schedule (including, in particular, paragraph 15(1)), the amounts shall be “those that, in accordance with generally accepted accounting practice, are recognised in determining the company’s profit and loss for that period”. The effect of these words is determined by paragraph 17B(1) which provides that any reference in the schedule to “an amount being recognised in determining a company’s profit or loss for a period is to an amount being recognised for accounting purposes” in the company’s profit and loss account or in any of the other accounts or statements listed in paragraph 17B(1).

The effect of paragraphs 17A and 17B is to take as a starting point the credits and debits in a taxpayer company’s GAAP-compliant accounts. Importantly, however, this is qualified by the opening words of paragraph 17A: “[s]ubject to the provisions of this Schedule (including, in particular, paragraph 15(1))”. For the reasons which I will develop later, this qualification subjects the credits and debits as shown in GAAPcompliant accounts to the “fairly represent” requirement in paragraph 15(1).

The “loss” issue arises because HMRC submit that, even leaving aside the “fairly represent” requirement, there was no “loss” for the purposes of paragraph 15(1) arising from the derecognition. The mere fact that an accounting debit was created upon the issue of the A Shares did not necessarily mean that the debit represented a “loss” or “expense” within the meaning of paragraph 15(1). For that purpose, it was necessary to analyse the true nature of the transaction leading to the derecognition in order to determine whether, as a matter of law, it produced a loss.

The FTT accepted HMRC’s submission that there was no loss, as Union Castle was entitled to exactly the same amount on close-out of the options after the issue of the A Shares as it was before their issue. It received the full cash benefit under the closedout options and distributed 95% of it by way of dividend on the A Shares. The issue of the A Shares and the consequent derecognition involved no real loss for Union Castle.

The UT reversed the FTT on this issue. It accepted that HMRC may be right that the issue of the A Shares did not involve a loss in relation to the relevant options, at least in common parlance, but said at [31] that the issue was the meaning of “losses” in paragraph 15(1) where the credits and debits to be brought into account in computing profits and losses are, by virtue of paragraph 17A, those recognised in GAAPcompliant accounts.

I agree with the UT’s conclusion on the “loss” issue and with their reason as just summarised. If the “fairly represent” requirement is ignored, a company’s profits and losses are under these paragraphs determined by the entries in the company’s GAAPcompliant accounts. That is the purpose of paragraphs 17A and 17B. It is not the use of the words “profits” and “losses” in paragraph 15 which requires a further assessment of their character but the “fairly represent” requirement. It is legitimate to ask why the “fairly represent” requirement should be included at all if such further assessment was in any event required.

In the rest of [31] and in [32]-[37] the UT gave further reasons for their conclusion, which are best discussed in the context of the “fairly represent” issue.

It was common ground below and before us that the words “fairly represent” in paragraph 15(1) did more than govern just the immediately following words “for the accounting period in question”. As observed by the UT at [47], it could not be regarded as “a criterion relating merely to attribution or timing of losses and profits”, relying on the decision of the Supreme Court in DCC Holding Ltd v HMRC [2010] UKSC 58, [2011] 1 WLR 44 and the decision of a differently constituted UT in GDF

DCC Holding and GDF Suez were decisions on the loan relationship code in the Finance Act 1996 but the parties accepted that there was no relevant distinction for present purposes between the loan relationship code and the derivatives contracts code in schedule 26. The UT took those decisions as authority for the proposition that “fairly represent” enables HMRC to prevent a mismatch in accounting treatment between the accounts of a parent company and the accounts of a subsidiary in relation to the same transaction. While it was arguable that such mismatches may not be confined to cases of parent and subsidiary, there was no authority that “fairly represent” encompassed any other category of case. Accordingly, the UT considered that the question on the authorities was whether there was any kind of accounting mismatch in the case of Union Castle and it concluded there was none.

Since the UT gave its Decision in the present case, this court has given judgment in GDF Suez: see [2018] EWCA Civ 2075, [2019] 1 All ER 528. In his judgment, with which Lord Kitchin and Asplin LJ agreed, Henderson LJ traced the genesis and development of the relevant provisions of the loan relationship code, which are precisely mirrored in the equivalent provisions of schedule 26. Section 84 of the Finance Act 1996 mirrors paragraph 15 of schedule 26. Both were amended by the Finance Act 2004 to remove the words “in accordance with an authorised accounting method”, and the provisionsdealing with authorised accounting methods were also deleted. Also deleted were section 84(2) and paragraph 15(3) which provided that the reference to profits and losses arising to a company included any profits and losses which in accordance with generally accepted accounting practice were carried to or sustained by any reserve maintained by the company. These deletions removed a direct express link between GAAP and section 84 and paragraph 15, but they were replaced by the indirect link introduced by section 85A and paragraph 17A respectively.

Section 85A and paragraph 17A were amended by schedule 6 to the Finance Act 2006, which was expressly enacted to counter tax avoidance. The words “(including, in particular, [paragraph 15(1)/section 84(1)]” were inserted after “Subject to the provisions of this [Schedule/Chapter]”. Commenting on this amendment at [43], Henderson LJ said:

“I would be inclined to infer that Parliament’s purpose must have been to make it clear that the “fairly represent” requirement in s.84 (1) is a separate and potentially overriding condition which has to be satisfied, once the initial computation in accordance with UK GAAP has been performed.”

The computation in accordance with GAAP-compliant accounts is performed for profits and losses arising from derivative contracts under paragraph 17A and for loan relationships under section 85A(1). Henderson LJ went on to say at [43]:

“the requirement to “fairly represent” the profits, gains and losses arising to the company will not necessarily be answered by saying that they are recognised in accordance with UK GAAP, because s.84(1) would then add nothing of substance to s.85A(1), and there would be no point in making the latter provision expressly subject to the former.”

Having summarised the decisions below and the competing submissions on appeal to this court, Henderson LJ reiterated at [88] that the purpose of the amendments made in 2006 “must have been to make it clear, for the avoidance of any doubt which might otherwise have arisen, that the fair representation requirement in section 84(1) was a separate and overriding condition which had to be satisfied in computing the credits and debits to be brought into account by a company in respect of its loan relationships”. At [91], Henderson LJ said there was little point in expressly making section 85A(1) [paragraph 17A(1)] subject to section 84(1) [paragraph 15(1)] if Parliament had intended that the fair representation requirement should always be assessed by reference to the same accounting criteria as those mandated by sections 85A and 85B [paragraphs 17A and 17B].

Henderson LJ rejected as wholly unpersuasive the submission that the requirement had only a limited attribution function and said at [93]:

“The objection that Parliament would have formulated specific guidance on the application of the fair representation test, if it was intended to be an overriding requirement of a substantive nature, is at first sight more compelling, particularly when it is remembered that the test was until 2004 explicitly linked to “an authorised accounting method”. Nevertheless, I do not think that the objection is well-founded, although it was persuasively advanced by Mr Ghosh. The concept of fairness is central both to the development and application of accounting standards, and to any process of judicial appraisal by a court or tribunal. In itself, the concept needs no elucidation, but rather provides a touchstone which is well suited to application by accountants, lawyers and judges, bringing their professional experience and expertise to bear in widely differing factual contexts.”

While this court upheld the UT’s decision in GDF Suez, it did so on the basis of a much broader reading of the effect of the words “fairly represent”. That reading specifically rejected the view of the UT in GDF Suez that they could not be regarded as a freestanding criterion, which was not bound by the accounting treatment of profits and losses and could indeed override such treatment. The same applies in the present case. It is not enough to enquire whether there is any accounting mismatch between the accounts of a parent and its subsidiary that requires an adjustment to be made to the debits and credits in Union Castle’s accounts.

It has long been established that in the ordinary way the computation of the profits and losses of a business for tax purposes is to be undertaken in accordance with the ordinary principles of commercial accountancy, what is now called GAAP. For more recent affirmations of this principle, see Gallagher v Jones (Inspector of Taxes) [1994] Ch 107 (CA) and HMRC v William Grant & Sons Distillers Ltd [2007] UKHL 15, [2007] 1 WLR 1448. But, as Sir Thomas Bingham MR acknowledged in Gallagher v Jones at p.134, this general principle must give way to any express or implied statutory rule. The “fairly represent” requirement is such a rule.

In the recent decision of this court in HMRC v Smith & Nephew Overseas Ltd [2020] EWCA Civ 299, further consideration was given to the “fairly represent” requirement in the loan relationship code. The case concerned exchange losses and was decided on the basis of the special provisions introduced to deal with such losses, which are not applicable to the present case. Submissions were made as to whether, if applicable to exchange losses, the losses in issue in that case satisfied the fairly represent test. Giving the leading judgment, Rose LJ carefully considered Henderson LJ’s judgment in GDF Suez and in a passage at [42], with which Coulson LJ specifically agreed, said:

“I agree with HMRC’s submission that the presence or absence of a tax avoidance purpose should not be determinative. Although the Court in GDF Suez explained how the amendments to the loan relationships regime in 2004 and 2006 were prompted by the desire to close loopholes and prevent tax avoidance, the wording of the statute does not refer to tax avoidance as a yardstick. It is not correct to give the ‘fairly represent’ test a limited meaning by regarding tax avoidance as the paradigm situation where the test would not be met. The test may well be failed in a case where there is an avoidance motive but where the more specific provisions directed at preventing avoidance do not, for whatever reason, apply. However, the override is not limited to that situation since it is intended to operate in favour of the taxpayer as well as in favour of HMRC. It may lead, for example, to profits being left out of account for tax purposes even though they are included in the company’s accounts in accordance with GAAP. I also agree that the presence or absence of an ‘asymmetry’ of the tax treatment of a transaction when looked at from the perspective of the counterparties is not a factor that need be present in every case where the override is triggered. It so happens that asymmetry was a factor both in GDF Suez and in the earlier case of DCC Holdings (UK) Ltd v Revenue and Customs Commissioners [2010] UKSC 58, [2011] 1 WLR 44. That does not mean, in my view, that the absence of an asymmetry in any subsequent case militates against the override being triggered. Finally, I agree with Mr Gibbon [counsel for HMRC] that the hurdle of ‘manifest absurdity’ which the Upper Tribunal appears to have applied before triggering the ‘fairly represent’ override is too stringent test. The true analysis is that section 84(1) is engaged wherever fair representation would not otherwise be achieved.”

I endorse the view there expressed as to the role of the “fairly represent” requirement or “override”. Neither asymmetry of tax treatment nor a tax avoidance purpose is necessary for its operation. Indeed, it is neutral in its application for and against the taxpayer. It is therefore strictly irrelevant that both Union Castle and Ladbrokes sought to take advantage of the same tax avoidance scheme marketed by Deloitte. Nonetheless, such schemes are more likely to explore the scope for arbitrage between

It might be thought that there was some tension between this approach and the fact that neither Union Castle nor Ladbrokes, in preparing their accounts, departed from specific accounting standards in order to show a true and fair view or a fair presentation. However, no such tension exists, nor to be fair did Mr Peacock submit that it did. The true and fair or fair presentation “override” forms part of GAAP. It involves a departure from a particular accounting standard but not a departure from GAAP. By contrast, where applicable, the statutory “fairly represent” requirement in paragraph 15(1) does mandate a departure from GAAP.

The “fairly represent” test requires the court to examine the totality of the relevant transaction in its factual context. Having regard to the purpose of schedule 26, which is to determine the profits and losses arising from derivative contracts for the purpose of charging the net profits to corporation tax, the court must decide whether the accounting debit fairly represents a loss. In this connection, it should be noted that the purpose of accounting standards, which is in general terms to assist actual and potential investors, and others dealing or proposing to deal with companies, in their assessment of the financial position and performance of a company, may not always match the purpose of determining profits and losses for tax purposes. For this reason, the “fairly represent” requirement appears in paragraph 15(1) and in the equivalent provisions dealing with loan relationships.

Union Castle submitted that the correct focus of the paragraph 15 analysis was the derecognition, not the subsequent payment of the dividend on the A Shares. HMRC was wrongly conflating the two. The derecognition related to the fair value of the derivative contractson 21 November 2008. Through the issue of the A Shares, Union Castle lost access to 95% of the cash flows from the contractsand hence lost 95% of the fair value of the contracts, which was reflected in the derecognition. This gave rise to an actual diminution of Union Castle’s resources and in its net asset value. Economically, Union Castle no longer held the risk and reward in relation to the contracts, which had been passed to Caledonia. At that time no profits had been received, and therefore there could not be any application of profits at that time. When subsequently Union Castle received the close-out proceeds, it did not at that point choose to pay out 95% of the proceeds, because it had already committed to doing so by the issue of the A Shares.

These submissions were advanced specifically in the context of the “loss” issue, which I accept is determined by the accounting treatment, but they naturally fed into and formed an essential part of Union Castle’s submissions as regards the “fairly represent” requirement, particularly as informed by Henderson LJ’s judgment in GDF Suez.Mr Peacock accepted that the existence of an accounting loss did not determine whether it “fairly represented” a loss, but it did mean that there needed to be a good and particular reason to justify overriding a “loss” as shown in GAAP-compliant accounts, on grounds of fair representation. No such reason existed in the present case.

The relevant facts of the present case are, as it seems to me, as follows. First, at all material times, both before and after the issue of the A Shares, Union Castle was the beneficial owner of the derivative contracts. At the time of the issue of the A Shares,

Second, the obligation imposed on Union Castle is not an absolute obligation to pay 95% of the close-out proceeds but is an obligation to pay a dividend on the A Shares equal to that sum. This obligation not only presupposes that Union Castle remains beneficially entitled to the derivative contracts and their close-out proceeds, but also requires for its lawful performance that Union Castle has distributable profits at least equal to 95% of the close-out proceeds when payment of the dividend is due. Given the cost of the derivative contracts, their close-out would be highly unlikely to result in distributable profits equal to 95% of the close-out proceeds. Other profits would be required for the dividend to be paid. The purpose of the Caledonia agreementwas to ensure, so far as possible, that this requirement would be fulfilled. In fact, Caledonia was the holder of the A Shares, so the purpose of its undertaking to make the capital contributions was to ensure that it received the dividends on those shares.

Third, although not essential to the analysis, the right to the dividend was attached to shares which were issued as bonus shares to Caledonia as the holder of Union Castle’s ordinary shares.

These factors demonstrate, in my judgment, that the debit required by IAS 39 to be made in Union Castle’s accounts by the derecognition did not, as a matter of legal analysis or economic reality, fairly represent a loss to Union Castle for the purposes of paragraph 15(1). Union Castle lost no asset nor incurred any liability other than a liability to pay a dividend on shares, such shares being issued for no consideration to its holding company. The payment of a dividend is not a loss. It is the very reverse of a loss: it is the distribution of profits. The difference between a loss and a distribution of profits is expressly recognised by section 830(2)of the Companies Act 2006, which defines a company’s distributable profits as its accumulated, realised profits, so far as not previously utilised in a distribution, less its accumulated realised losses. If the payment of a dividend is not a loss, I am unable to accept that an obligation to pay a dividend, in this case out of future profits, can be a loss.

Nor am I able to accept that the UT’s analysis of the effect of the derecognition on Union Castle can apply in the context of the “fairly represent” issue. Addressing the “loss” issue, the UT said at [31] that a loss arose on derecognition “because the economic value to the company of the Contracts no longer exists (as to 95% of the value in Union Castle’s case) and the company’s worth has gone down”. That is unobjectionable if restricted to an analysis of the effect of the derecognition on the entries in Union Castle’s accounts. But viewed more broadly, it would be wrong to say that Union Castle’s value had gone down or that the economic value to Union Castle of the derivative contracts no longer existed. It may be thought that the only diminution in value that resulted from the issue of the A Shares was in the value of

In my judgment, Mr Ghosh on behalf of HMRC was right in his submission that the issue of the A Shares was in substance an election by Union Castle that it would in the future distribute by way of dividend to its parent company 95% of the close-out proceeds of the derivative contracts, as and when they were received. I am unable to see that Union Castle thereby incurred anything that could fairly be described as a loss for the purposes of paragraph 15(1) of schedule 26.

One means of cross-checking this conclusion is to consider the means by which Union Castle in fact ensured that it had the distributable reserves necessary to pay the dividends on the A Shares as and when the derivative contracts were closed out. In January 2009, two options were closed out for net cash settlements totalling £13,734,511, giving rise to an obligation to pay a dividend of £13,047,785 on the A Shares.

In order to pay this dividend, Union Castle had to comply with the requirements of Part 23 of the Companies Act 2006. Section 836 required that its “relevant accounts” should show that it had distributable profits at least equal to the proposed dividend. Section 836(1) requires that whether a distribution may be made is to be determined by reference to, among other items, “profits, losses, assets and liabilities” as stated in the relevant accounts. Those accounts are the company’s last annual accounts, except that where the distribution would contravene Part 23 it may be justified by reference to interim accounts: section 836(2). Section 838(1) requires, in the case of a private company like Union Castle, interim accounts to be such accounts as enable a reasonable judgment to be made as to the amounts of the items mentioned in section 836(1).

In January 2009, the last annual accounts of Union Castle were for the year ended 31 March 2008, which showed distributable reserves of just under £5 million, which was inadequate for the proposed dividend. Accordingly, interim accounts were prepared. The accompanying notes stated that they had been prepared in accordance with IFRS. The income statement included the debit of £39,149,128 resulting from the derecognition of the derivative contracts. Together with other items, this produced a loss for the period since 1 April 2008 of just under £3.4 million, resulting in reserves of approximately £1.57 million after crediting the reserves as at 31 March 2008.

There was therefore, on that basis, a significant shortfall in the distributable profits required to pay the proposed dividend. This was overcome by writing back the debit of £39,149,128, described in the income statement as “[a]ccounting debit disregarded for distributable earnings”. The effect was to pay the dividend out of profits constituted by the reversal of the very debit said to be a loss. This appears to be a remarkable piece of “now you see it, now you don’t” accounting.

The same process was repeated for the subsequent dividends on the A Shares.

If an accounting debit can be disregarded for the purpose of determining a company’s profits available for distribution, I am unable to understand how the debit can nonetheless fairly represent a loss for the purposes of paragraph 15(1). In truth, it was not a loss but only, as the accounts described it, an “accounting debit”.

The UT held that, although the debit giving effect to the derecognition constituted a loss for the purposes of paragraph 15(1), the loss did not “arise to the company from its derivative contracts and related transactions”, so that the debit was not brought into account under paragraph 15(1). Union Castle appeals against that decision.

The UT rested its decision on three principal grounds. First, paragraph 15(1) uses different expressions to denote a connection. In addition to “arise to the company from”, it refers to credits and debits “in respect of” its derivative contracts and to charges and expenses incurred “under or for the purposes of” its derivative contracts. As a matter of language, “arises from” bears “a narrower meaning and implies a direct causal connection between losses (or profits) and derivative contracts” than, in particular “in respect of”: paragraph [39].

Second, the UT placed reliance on the reference to profits and losses that “arise to the company from its derivative contracts and related transactions”. Paragraph 15(7) defines “related transaction” as “any disposal or acquisition (in whole or in part) of rights or liabilities under the derivative contract”. It followed that profits and losses that arise from related transactions (as defined) do not arise from the derivative contracts themselves. The UT said at [40]:

“Since related transactions are defined as disposals in whole or in part of rights or liabilities under the derivative contracts, it would be very surprising if the draftsman had assumed that something remoter from the derivative contracts themselves (viz. an agreement to transfer a sum of money equivalent to the economic benefit of the contracts) was something that “arises from” the derivative contracts”.

Third, a purposive analysis showed that schedule 26 aimed to tax derivative contracts, and related transactions, not transactions that do not affect the value of the derivatives themselves: paragraph [42].

Mr Peacock for Union Castle challenged the UT’s reasoning on a number of different grounds.

As regards the first ground for the UT’s decision, Mr Peacock accepted that “in respect of” denotes a broader connection than “arise from” but he submitted that it did not follow that “arise from” requires a direct causal link between losses or profits and the derivative contracts. The second ground relied on by the UT, the inclusion of “related transactions”, did not dictate this result. It was appropriate to include express reference to transactions involving the acquisition or disposal of rights or liabilities because the resulting losses or profits are rightly said to arise from the acquisition or disposal, rather than the derivative contracts themselves. Without express reference to them, these losses or profits would not come within the derivative tax code at all. By contrast, the loss resulting from the derecognition arose from the derivative contracts themselves. IAS 39 required that 95% of the value of those contracts be derecognised and that loss arose from the contracts. This was an adjustment to the value of the contracts, because in economic terms their value to Union Castle had declined as a result of the issue of the A Shares.

Mr Peacock also challenged the third ground relied on by the UT. Its reasoning would exclude losses or profits arising from any turning to account of derivative contracts that fell short of an acquisition or disposal of the whole or part of the rights or liabilities under the contracts. He suggested that this would mean that profits and losses from a number of ordinary commercial transactions would not be within the charge to corporation tax, and he instanced sub-participation arrangements. As a separate point, Mr Peacock also drew attention to the meaning of “from” in other contexts, such as emoluments “from” employment, which show a much wider meaning than the direct causal link decided by the UT in this case.

I should say, first, that consideration of the meaning of words such as “from” in very different contexts does not assist the construction of “arise from’ in the context of schedule 26, nor is it assisted by instancing wholly different transactions without a full examination of their terms and of the corporation tax provisions, wherever they appear, applicable to them.

A fall in the market value of a derivative contract causes a loss which is properly described as arising from the contract itself. Its value has declined. Union Castle argues that, likewise, a derecognition mandated by IAS 39 reflects a decline in the value of the derivative contract to the company and also arises from the contract. This submission depends on showing that the relevant debit did not arise from the issue of A Shares, but arose from the derecognition. The difficulty for Union Castle is that the derecognition and the issue of the A Shares are inseparable. The issue of the A Shares had the effect, by reason of the derecognition mandated by IAS 39, of reducing the carrying value of the derivative contracts by 95%. In my judgment, it is not tenable to say that the derecognition arises from the derivative contract, as opposed to the issue of the A Shares.

Mr Peacock was unable to suggest any satisfactory reason why “related transaction” had been drafted to refer only to disposals and acquisitions of rights or liabilities under derivative contracts if paragraph 15(1) was intended to extend to profits or losses arising from a wider class of transaction. In my judgment, it is the definition of “related transaction” that, as a matter of construction, clearly demonstrates its intended scope. The first and third grounds relied on by the UT are consistent with the restricted ambit shown by the use and definition of “related transaction” and indeed gain much of their force from it.

I would therefore dismiss Union Castle’s appeal on this ground.

Ladbrokes entered into similar transactions to Union Castle. In November 2008, it made a bonus issue of A Shares to its parent company. The A Shares carried the right to a dividend equal to amounts payable on certain swap contracts. In consequence,

The position of Ladbrokes differs from that of Union Castle because, when it derecognised the swap contracts, it did not recognise the debit in its profit and loss account or any of the other statements listed in paragraph 17B. Instead, the debit was recognised in equity. Accordingly, paragraph 25A of schedule 26 applies, which provides:

“Where in accordance with generally accepted accounting practice a debit or credit for a period in respect of a derivative contract of a company –

(a)

is recognised in equity or shareholders’ funds, and

(b)

is not recognised in any of the statements mentioned in paragraph 17B(1),

the debit or credit shall be brought into account for that period for the purposes of this Chapter in the same way as a debit or credit that, in accordance with generally accepted accounting practice, is brought into account in determining the company’s profit or loss for that period.” [The reference to “Chapter” would seem to be mistake for “Schedule”.]

This gives rise to the Gateway issue, which is whether a debit falling within paragraph 25A must meet the requirements of paragraph 15 before it can be brought into account for corporation tax purposes under schedule 26. Ladbrokes argue that paragraph 15 is not applicable to such debits and that they are brought into account under a combination of paragraphs 14 and 25A.

The FTT held in favour of Ladbrokes. They accepted its submission that paragraph 25A achieved for the debits covered by it what paragraph 15 achieved for the debits to which paragraph 17B applied, namely (i) that the debits had to be in respect of derivative contracts and (ii) that they had to be appropriate to the right accounting period. It followed that paragraph 15 did not apply to debits covered by paragraph 25A.

The UT allowed HMRC’s appeal against this decision: see the UT’s Decision at [73][83]. It held that paragraph 25A equates the position of credits and debits recognised in equity or shareholders’ funds with those credits and debits recognised in the profit and loss and other statements to which paragraph 17B applies. This was the effect of the words “in the same way” in paragraph 25A. Paragraph 15 sets the requirements for all credits and debits to be brought into account, namely that they should fairly represent for the accounting period in question profits and losses which arise from the company’s derivative contracts and related transactions. It sets out the credits and debits to be brought into account, while paragraph 14 is concerned with the computation of profits and losses using those credits and debits. The UT did not agree with the FTT’s acceptance that paragraph 15 had the limited roles for which Union Castle contended. Although the UT reached their conclusion as a matter of the natural and ordinary meaning of the relevant provisions, they considered that there could be no purpose in a more favourable and less stringent treatment for credits and debits that were recognised in equity or shareholders’ funds than those recognised in the statements to which paragraph 17B applies.

Ladbrokes challenges the UT’s decision and reasoning on several grounds. First, if it were correct, there is no discernible reason why recognition in equity or shareholders’ funds was not simply added to paragraph 17B, rather than being the subject of a different paragraph in a different part of the schedule. Second, unlike paragraph 17A, paragraph 25A was not amended to make it expressly subject to paragraph 15. Nor, if paragraph 15 in any event qualified the rest of the schedule, was there any need to amend paragraph 17A to make it subject to paragraph 15. The right conclusion is that Parliament intended to make paragraph 17A, but not paragraph 25A, subject to paragraph 15. Third, paragraph 14 specifically refers to the “the following provisions [plural] in this Schedule”, indicating that paragraph 15 is not the universal gateway. Paragraph 15 identifies credits and debits to be brought into account in certain circumstances, while it is the purpose of paragraph 14 to bring credits and debits (whether identified in paragraph 15 or elsewhere) into account. Paragraph 25A provides a separate mechanism for determining other credits and debits to be brought into account. It is therefore paragraph 15 (read with paragraphs 17A and 17B) and, separately, paragraph 25A which identify the credits and debits to be brought into account and paragraph 14 which determines how they are to be brought into account. The words “in the same way” in paragraph 25A refer to the “how” of paragraph 14, not to the “which” of paragraph 15.

I do not accept these submissions. As both parties accept, the critical words in paragraph 25A are “in the same way”. A debit or credit recognised in equity or shareholders’ funds is to be brought into account “in the same way as a debit or credit that, in accordance with generally accepted accounting practice, is brought into account in determining the company’s profit or loss for that period”. Debits and credits brought into account in accordance with GAAP in determining the company’s profit or loss are the subject of paragraphs 17A and 17B. Those debits and credits are brought into account if they satisfy the requirements of paragraph 15. The debits and credits subject to paragraph 25A must be brought into account “in the same way”. They too must satisfy the requirements of paragraph 15.

As Mr Ghosh submitted, there is no wording in paragraph 15 that limits its application to those debits and credits covered by paragraph 17B. It applies generally to “[t]he credits and debits to be brought into account in the case of any company in respect of its derivative contracts”, which are echoed in paragraph 25A as well as in paragraph 17A.

The UT was right to say that paragraph 15 fulfils a role which is wider than paragraph

Ladbrokes relied on the use in paragraph 25A of the phrase “in respect of a derivative contract” as opposed to the narrower formula “arise from” in paragraph 15, as showing that paragraph 25A was intended to cast a different and wider net that paragraph 15. This is, however, untenable given that paragraph 14 expressly applies, and is restricted, to “profits and losses arising from the derivative contracts of a company”.

Mr Peacock was unable to suggest any policy that might justify a more generous treatment of debits to equity and shareholders’ funds under paragraph 25A than the treatment of the debits within paragraph 17B.

For these reasons, I consider that the UT was correct in its conclusion on the Gateway issue and I would dismiss Ladbrokes’ appeal.

The overall result, if my Lords agree, is that both appeals are dismissed. In the case of Union Castle’s appeal, I agree with the UT’s conclusions on the “loss” and “arise from” issues, but I have come to a different conclusion on the “fairly represent” issue and I would dismiss the appeal on that ground as well as on the “arise from” issue.

After distributing our judgments in draft to the parties, the court received a submission from Union Castle seeking an order for the matter to be remitted to the

I consider that, having dismissed the appeal before us, we should remit the matter to the FTT for it to decide (i) whether Union Castle should now be permitted to seek to amend its tax computation in this way and (ii) if so, whether it has a good case for such an amendment. Having not heard argument on either issue, I express no views on them, nor do I express any view on Union Castle’s submissions as to the consequences of our decision.

I agree.

I also agree.