Root2tax Ltd and Root3tax Ltd v Revenue and Customs (INCOME TAX/CORPORATION TAX : Anti-avoidance) [2017] UKFTT 696 (TC)

FTT-Tax

Root2tax Ltd and Root3tax Ltd v Revenue and Customs (INCOME TAX/CORPORATION TAX : Anti-avoidance)

[2017] UKFTT 696 (TC) · 2017-03-02

[1]This decision relates to two applications by HM Revenue and Customs, or HMRC. The first is for an order that certain arrangements are notifiable for the purposes of Part 7 of the Finance Act 2004 (‘FA 2004’), the Disclosure of Tax Avoidance Schemes, or DoTAS, provisions. The second, made as an alternative should the first fail, is for an order that the arrangements are to be treated as notifiable. Applications of this kind are rare, but not quite novel: in 2009 HMRC made an application similar to the first of those before me to a Special Commissioner, whose decision is reported as Revenue and Customs Commissioners v Mercury Tax Group Ltd [2009] STC (SCD) 307 . The arrangements in issue in that case were, however, very different from those with which I am concerned and the relevant law has since been extensively amended; accordingly I am able to derive little assistance from that case. The current provisions have been considered, in different contexts, in other cases, notably by the Administrative Court in Walapu v Revenue and Customs Commissioners

[2016] EWHC 658 (Admin)

[2016] STC 1682

. The issue in that case, too, was rather different from that before me but some assistance can be derived from the judgment, as I shall later explain.

[2]In much of the documentation produced for the hearing the arrangements are described as ‘the Alchemy scheme’, a name derived from some of the material produced by those who devised or marketed it. I will use the terms ‘the Alchemy scheme’ or ‘the scheme’ as convenient shorthand, while intending them to be neutral.

[3]The essential structure of the Alchemy scheme is straightforward. An individual (‘the user’), typically but not invariably a director or key employee of a small company, enters into a spread bet contract with a counterparty, an established spread-betting business at arm’s length, though aware that it is entering into a pre-conceived arrangement. The bet relates to the performance of a basket of hedge funds over a given period: if the funds rise in value to a stated level or above by the end of the period, the user wins, and if not he loses the bet. In the ‘vanilla’ version there is that simple binary outcome; more complicated variants allow for sliding scales of gains and losses but the principles are identical and there is no need to say any more about those variants. At the same, or about the same, time as he enters into the spread bet the user enters into a hedging contract, commonly but not always a call spread option (‘CSO’), with the same counterparty. The hedging contract mirrors, though not always exactly, the spread bet; the critical and invariable feature is that the outcome of the CSO is also dependent on the performance of the same basket of hedge funds, but in reverse: thus if the user wins on the spread bet he is certain to lose on the CSO, and vice versa. The spread bet and the CSO are arranged so that they mature at the same time.

[4]The sizes of the bets differ from user to user, as do their precise terms. Thus, to take one simple example, the identity of the securities in the basket will differ from one iteration to another. In a hypothetical but illustrative example a user might enter into a spread bet by which he could win £95,000 if the basket rises in value to the requisite level, or lose £11,000 (the amount of his ‘stake’) if it does not, and into a CSO by which he could lose £106,000 offset by the payment to him by the spread betting house of a premium, for the grant of the CSO, of £6,000. Thus in that example if the basket of funds rises in value to the agreed extent the user gains £95,000 but loses £100,000 net, making an overall loss of £5,000. If the basket does not rise, the user loses £11,000 on the spread bet, but gains £6,000 on the CSO, again a loss overall of £5,000. He is, therefore, in a small net loss situation, while the spread betting house is certain to make a gain of the same amount, in the example £5,000, whatever the outcome. That amount is, in effect even if not in form, the spread betting house’s fee for participation in the arrangements. The terms of the agreements between the user and the spread betting house also provide that the latter is not required to pay out on the one unless it receives payment on the other. I should add that, as Mr Patrick Way QC, leading Mr Imran Afzal for the respondents, pointed out, the premium paid by the spread betting house to the user for the CSO may attract capital gains tax, depending on the user’s personal circumstances. This is not, however, a material factor and can be disregarded in what follows.

[5]The user does not remain exposed to the certainty of a loss but invites his employer or, in some iterations, an employee benefit trust (‘EBT’) established by the employer, to relieve him of the CSO. In the ordinary case the employer has profits available to be paid out, whether by way of dividends or as earnings, or there are funds available within the EBT. The employer or EBT agrees to assume the CSO (or its equivalent), and does so by way of novation. At this stage, usually a few days after the user has entered into the two contracts, when the period remaining to maturity is relatively long and the outcome of the two contracts is uncertain, it is expected that the spread bet will have a modest positive value in the user’s hands, and the CSO a modest negative value. Thus by relieving him of the burden of the CSO, for no payment, the employer or the EBT has conferred an employment-related benefit on the user, which he declares and on which he pays the appropriate tax. The scale of the benefit is determined by reference to a recognised method, the Black-Scholes Model. Although HMRC reserve their position about the suitability of that method, it is not a matter about which I am concerned in the context of these applications.

[6]Nothing more happens until the contracts mature, typically a few months later. Ideally, from his perspective, the user will have won on the spread bet, and the employer or EBT will have lost on the CSO. Thus the employer or EBT makes a payment to the counterparty, which in turn makes a payment of about the same amount to the user. In this case, and if the scheme works as intended, the user will have received a significant sum which, as betting winnings, is not taxable; he will have suffered tax, if at all, only on the much smaller value of the benefit to him of being relieved of the CSO. It is, of course, in the nature of gambling that outcomes are not certain and in some cases the user loses the spread bet while the employer or EBT wins the CSO. I will return to instances of this kind when I deal with the evidence; for the purposes of an initial summary this eventuality can be left to one side.

[7]I was provided with several examples of actual implementations of the scheme illustrating the differences of detail between them, but I do not think it necessary in this decision to deal with those differences except at a fairly high level of generality. Neither party suggested that differences of detail might dictate whether one variant was, and another was not, notifiable, or that they might have any other significance for present purposes. The one variant I should mention, though only for completeness, is that in some cases the employer took out a loan which was guaranteed by the user, and ultimately repaid by him from his winnings on the spread bet; in this variant the user did not receive, or at least retain, an immediate cash sum, but received an increase in the balance of his director’s loan account. Although the mechanics of this variant were more complicated it did not seem to me, and the parties did not argue, that it should be distinguished in some way.

[8]As their names suggest, the respondents are, or until the second respondent entered into liquidation were, related companies. The applications are made against them because HMRC consider that they were the ‘promoters’ of the Alchemy scheme, in the statutory sense to which I shall come: in summary, they devised, or participated in the devising of, the Alchemy scheme and, directly or indirectly, marketed the scheme to potential users. Much of the correspondence referred to them indiscriminately and it did not become clear to me why there were two similarly named companies and, save for one minor point to which I shall come, there did not seem to be any material difference between them with respect to their activities. Again, neither party suggested that the outcome of these applications for one should differ for the other. The applications and the relevant legislation

[9]The first, and as I have said preferred, application is for an order that the arrangements I have briefly described are notifiable. That application is based upon FA 2004 s 314A, which is as follows:

‘(1) HMRC may apply to the tribunal for an order that— (a) a proposal is notifiable, or (b) arrangements are notifiable. (2) An application must specify— (a) the proposal or arrangements in respect of which the order is sought, and (b) the promoter. (3) On an application the tribunal may make the order only if satisfied that section 306(1)(a) to (c) applies to the relevant arrangements.’

[10]The alternative application is based upon FA 2004 s 306A which, so far as material, is as follows: “(1) HMRC may apply to the tribunal for an order that—(a) a proposal is to be treated as notifiable, or

(b) arrangements are to be treated as notifiable. (2) An application must specify— (a) the proposal or arrangements in respect of which the order is sought, and (b) the promoter. (3) On an application the tribunal may make the order only if satisfied that HMRC— (a) have taken all reasonable steps to establish whether the proposal or arrangements are notifiable, and (b) have reasonable grounds for suspecting that the proposal or arrangements may be notifiable. (4) Reasonable steps under subsection (3)(a) may (but need not) include taking action under section 313A or 313B. (5) Grounds for suspicion under subsection (3)(b) may include— (a) the fact that the relevant arrangements fall within a description prescribed under section 306(1)(a); (b) an attempt by the promoter to avoid or delay providing information or documents about the proposal or arrangements under or by virtue of section 313A or 313B;

(c) the promoter’s failure to comply with a requirement under or by virtue of section 313A or 313B in relation to another proposal or other arrangements.’

[11]The two provisions, both introduced by the Finance Act 2007, are plainly similar in their aim, described by Green J in Walapu at [147] as ‘to enable HMRC to learn about and challenge’ avoidance schemes, or suspected avoidance schemes. Notification also now carries with it the potential trigger for an accelerated payment notice, in accordance with Chapter 3 of Part 4 of the Finance Act 2014. The conditions which must be met differ between the two provisions, and I shall focus initially on those applicable to a s 314A application, returning to the s 306A conditions towards the end of this decision.

[12]I am concerned in these applications only with ‘arrangements’, and not with a ‘proposal’. It is not seriously disputed that the steps undertaken by users of the scheme are ‘arrangements’ in the ordinary sense of that word, and that they also come within the definition provided by FA 2004 s 318:

‘“arrangements” includes any scheme, transaction or series of transactions’

. As I shall explain later, Mr Way’s argument focused on a different feature of the statutory framework. The mere fact that a series of transactions amounts to ‘arrangements’ does not render them notifiable; they must possess the additional, or distinctive, qualities set out in s 306(1), to which both s 314A and s 306A refer. That subsection is as follows:

‘(1) In this Part “notifiable arrangements” means any arrangements which— (a) fall within any description prescribed by the Treasury by regulations, (b) enable, or might be expected to enable, any person to obtain an advantage in relation to any tax that is so prescribed in relation to arrangements of that description, and (c) are such that the main benefit, or one of the main benefits, that might be expected to arise from the arrangements is the obtaining of that advantage.’

[13]The Regulations to which sub-s (1)(a) refers are the Tax Avoidance Schemes (Prescribed Descriptions of Arrangements) Regulations 2006 (‘the 2006 Regulations’). Regulation 10, in the form in which it was at the relevant time, provided as follows:

‘(1) Arrangements are prescribed if the arrangements are a standardised tax product. But arrangements are excepted from being prescribed under this regulation if they are specified in regulation 11. (2) For the purposes of paragraph (1) arrangements are a product if— (a) the arrangements have standardised, or substantially standardised, documentation— (i) the purpose of which is to enable the implementation, by the client, of the arrangements; and (ii) the form of which is determined by the promoter, and not tailored, to any material extent, to reflect the circumstances of the client; (b) a client must enter into a specific transaction or series of transactions; and (c) that transaction or that series of transactions are standardised, or substantially standardised in form. (3) For the purpose of paragraph (1) arrangements are a tax product if it would be reasonable for an informed observer (having studied the arrangements) to conclude that the main purpose of the arrangements was to enable a client to obtain a tax advantage. (4) For the purpose of paragraph (1) arrangements are standardised if a promoter makes the arrangements available for implementation by more than one other person.’

[14]It is common ground that reg 11 is of no relevance to these applications. There is a further difference between s 314A and s 306A in that in order to come within the former the relevant arrangements must meet all of the criteria set out in paras (a) to (c) of s 306(1), whereas the latter merely indicates that the para (a) criterion may be a relevant factor. Nevertheless, there is some duplication, or at least similarity, between paras (b) and (c) and reg 10(2)(d). As will become apparent, I shall not need to determine for the purposes of these applications whether this difference between the two provisions is material.

[15]Regulation 18, entitled ‘Employment income provided through third parties’, is also of relevance, because of HMRC’s view that the Alchemy scheme is a means of making a payment to an employee of what would ordinarily be employment income in a manner which (if the scheme works) avoids the payment of income tax and national insurance contributions. ITEPA, to which the regulation refers, is, of course, the Income Tax (Earnings and Pensions) Act 2003. I have added annotations in square brackets to the text of the regulation as set out below for ease of understanding:

‘(1) Arrangements are prescribed if— (a) Conditions 1 and 2 are met and Condition 3 is not met; or (b) Conditions 1, 2 and 3 are met and at least one of Conditions 4 and 5 is met. (2) Condition 1 is met if the arrangements involve at least one of the following— (a) a relevant third person taking a relevant step under section 554B [earmarking of sum of money or asset]; (b) any person taking a relevant step under section 554C [payment of sum or transfer of asset] or 554D [making asset available]; or (c) B [a current, past or prospective employer] taking a step under section 554Z18 [earmarking] or 554Z19 [provision of security]. (3) Condition 2 is met if the main benefit, or one of the main benefits, of the arrangements is that an amount that would otherwise count as employment income under section 554Z2(1) is reduced or eliminated. (4) Condition 3 is met if, by reason of at least one of sections 554E to 554X [exclusions] or regulations made under section 554Y [further exclusions], Chapter 2 of Part 7A does not apply. (5) Condition 4 is met if the arrangements involve one or more contrived or abnormal steps without which the main benefit in paragraph (3) would not be obtained. (6) Condition 5 is met if the arrangements involve— (a) a relevant step being treated as taking place; and (b) Chapter 2 of Part 7A applying as a consequence of sub-paragraph (a). (7) In this regulation— (a) references to sections or Parts are to those in ITEPA unless otherwise stated; (b) “B” has the meaning given for Part 7A by sections 554A(1)(a) and 554Z17(7) read together [see note to sub-reg (2) above]; (c) “contrived or abnormal” has the same meaning as in section 207 of the Finance Act 2013; and (d) “relevant third person” has the same meaning as in section 554A(7).’

[16]I have included reg 18 only for completeness since, although Ms Nathan addressed me on it, Mr Way’s sole argument in relation to it was that condition 2 is not engaged (for the reasons to which I come later). I do not, therefore, need to deal with the regulation in any detail, or to explore the various statutory provisions to which it refers. The only remaining statutory provisions I need to address contain definitions, but it is more convenient to set out the evidence before doing so. The evidence

[17]I heard the oral evidence of Mr David Hole, an officer of HMRC working as a technical adviser, with special expertise in employment income, in the counter-avoidance directorate and responsible for initiating the enquiries which led to the making of the applications, Mr Steve Goringe, a chartered accountant employed by HMRC, also in the counter-avoidance directorate and responsible for an analysis of the arrangements as they had been implemented by some users, and of Mr Blair Forsyth, a lawyer by background and a director of the first respondent, Root2 Tax Ltd. All had made witness statements which set out the greater part of their evidence in chief. I had also significant volumes of documentary evidence, including examples of the documents by which the arrangements were implemented by a variety of users in the differing versions which I have briefly described. There was little dispute about the evidence of Mr Hole and Mr Goringe, and I can deal with it quite briefly; more needs to be said about Mr Forsyth’s evidence.

[18]Mr Hole’s part in the investigation which led to the making of the applications began when he was asked for his technical advice on what HMRC then understood of the Alchemy scheme. He realised that it was a means of extracting funds from owner-managed companies or EBTs without incurring a tax charge, but he did not know at that stage how it achieved, or was intended to achieve, that objective. Enquiries were put in train, and they led to a meeting between HMRC and representatives of the respondents in July 2014; Mr Hole was not present at the meeting and was reliant on HMRC’s minutes of it. By this stage HMRC had reached the view that the Alchemy scheme, or more precisely the arrangements of which it consisted, were notifiable; the respondents argued that they were not. Later that year Mr Hole became more closely involved in the enquiries, and eventually he took charge of them.

[19]He and his colleagues corresponded with the respondents’ agents, and made various further enquiries, including internet searches. They learnt that the Alchemy scheme had been used on many occasions, and they formed the view that it appeared to fall within the statutory description of ‘notifiable arrangements’. The respondents and their agents remained adamant that the arrangements were not notifiable, and it became clear that further discussion would not resolve the disagreement. In July 2015 Mr Hole sent formal notices to the respondents, pursuant to s 313A of FA 2004 (which I do not need to set out), requiring them, if they continued to maintain that the arrangements were not notifiable, to give a statement of their reasons for so maintaining, and to do so by 28 August 2015. The deadline was later extended, but Mr Hole formed the view that the respondents were providing only partial information, despite his identification of the information he required and the documents he wished to see, some but not all of which were provided. I shall return to this part of his evidence, so far as it is necessary to do so, when I come to the parties’ submissions on HMRC’s alternative application. Mr Hole eventually came to the conclusion that he was not going to achieve more by continuing the correspondence or by pursuing his s 313A enquiry, and the applications now before me were made in June 2016.

[20]Mr Goringe undertook an examination of self-assessment returns and companies’ filed accounts in order to identify users of the Alchemy scheme. He did so by searching (by computer) for certain standard phrases which users had adopted when completing their returns, or which had been adopted in the accounts. Once he had identified users of the scheme he compiled spreadsheets based on the further information he was able to glean from the returns and accounts they and their employers, or EBT trustees, had submitted. He found that in many cases the scheme had worked as intended: the user had won the spread bet, and the company or EBT had been obliged to pay the counterparty. What he also discovered was that on many occasions when the user had lost the spread bet and the company or EBT had correspondingly won on the hedging contract the exercise had been repeated, in some instances several times, until the desired outcome was achieved—that is, the user won overall, after taking account of wins and losses, the amount of money, or approximately the amount of money, he would have won had the result of the first iteration been as intended.

[21]In his second witness statement Mr Forsyth took issue with some of Mr Goringe’s findings. Mr Goringe accepted as he gave his evidence that he had made a number of errors, including some of those identified by Mr Forsyth, but he made the point that he had access to substantially less complete information than Mr Forsyth. I do not intend to dwell on this disagreement, since it relates to matters of detail and not of principle. The only point of importance lies in Mr Goringe’s agreement with Mr Way that he could not show that repetition had taken place on every occasion on which the user had lost the initial spread bet, or that in all cases where there had been repetition the eventual outcome was as intended, and he thus could not exclude the possibility that in some cases the user had lost rather than gained from his use of the scheme. That was a point made by Mr Forsyth with which Mr Goringe did not take serious issue. Nevertheless, it was clear from Mr Goringe’s analyses that it was in only a small minority of such cases that repetition had not taken place until the desired result was achieved.

[22]Mr Forsyth’s evidence was that he had qualified as a solicitor after leaving university, but had specialised in tax advisory work (I deduce not as a practising solicitor) since 2008, becoming one of the founder shareholders of the first respondent when it was incorporated in 2011. He did not describe in his witness statement what, if any, relationship he had formerly had with the second respondent, but it appears from the evidence as a whole that it was similar. The first respondent was originally known as Root3 Tax Limited, but in 2012, for reasons of which I am unaware, the companies exchanged names. Although, as I understood Mr Forsyth’s evidence, one of the two companies focused on users intending to dispose of their CSOs or similar hedging contracts to companies and the other on users intending to dispose of them to EBTs, there did not appear to be any other material difference in their respective activities until the second respondent entered into liquidation.

[23]Mr Forsyth agreed that the respondents’ business was the provision of tax advice, and that they, and he personally, had been involved in shaping the Alchemy scheme. The basic structure had, he said, been devised by others before the respondents’ advice was sought, and it was at that stage, when the respondents had been able to consider and advise on it, that the scheme had been refined. He was at pains to say that the respondents merely provided tax advice about the scheme to prospective users, but he accepted as he gave his oral evidence that it was the respondents who identified a suitable counterparty, initially a company which traded as Heronden, and that the respondents did not merely devise, or refine, the Alchemy scheme, but remained involved in its various implementations.

[24]At the outset all, or almost all, of the spread bet and hedging contracts were made between the users and Heronden. At a later stage, and because too many users were losing their spread bets and incurring substantial losses, apparently because the baskets of securities chosen by Heronden often did not perform as hoped, some different counterparties were found, again by the respondents. Heronden did not drop out altogether, but the terms on which the contracts with it were made were changed. The contractual documents, that is the spread bet and hedging contracts as well as the novation agreement, were prepared, Mr Forsyth said, by the counterparty. In most, if not all, cases the basis was the commonly used ISDA form, though with supplementary documents specific to the individual contract. Some other documentation, such as board meeting minutes, and some standard-form letters, were provided by the respondents. In addition the users and the trustees of the EBTs typically took financial advice from a firm, Aston Collie, identified to them by the respondents; I understood that Aston Collie had some link to Heronden.

[25]Much of Mr Forsyth’s witness statement was devoted to descriptions of various iterations of the scheme, and was designed to identify both differences in the manner in which the scheme had been implemented by the users, and that the spread bets and hedging contracts were entered into at commercial rates. Since, as I have said, nothing turns on differences of implementation and HMRC do not disagree that the terms of the contracts with Heronden or, later, the other counterparties, were commercial I do not need to deal with this aspect of Mr Forsyth’s evidence, save to say that I have no reason to doubt what he said in both respects. However, it is not accepted by HMRC that the differences of implementation are material, such as to indicate that each iteration was not simply an implementation of a standard scheme, and Ms Aparna Nathan, who appeared before me for HMRC, spent a significant amount of time cross-examining Mr Forsyth about the documentation used in the various implementations, copies of which were included in the hearing bundles, and about the manner in which the respondents participated in the transactions.

[26]I regret to say that I found much of Mr Forsyth’s evidence on these points disingenuous, if not evasive. Taken at face value, it would lead to the conclusion that a succession of individuals, who had happened on the idea of taking simultaneous or near simultaneous spread bets and hedging contracts and whose employers, or EBTs of which they were beneficiaries, had fortuitously offered to relieve them of the CSOs, had chosen to instruct one or other of the respondents to advise on the tax consequences of their doing so, and that the respondents’ sole part was to provide that advice, albeit refining the scheme in the process. Mr Forsyth repeatedly failed to answer Ms Nathan’s questions about the common form of the documents used in each iteration, by focusing instead on details which would inevitably differ from one iteration to another, such as the monetary amounts, the identity of the basket of securities or the identity of the counterparty. He was similarly evasive about the part played by the respondents in some ‘webinars’ by which the Alchemy scheme was promoted to prospective users and their advisers; I was wholly unconvinced by his attempts to distance the respondents, and himself, from them.

[27]There is in my view no doubt, from the documentary evidence together with the concessions Mr Forsyth was compelled to make when confronted with it, that if the respondents did not devise the Alchemy scheme they played a leading part in refining it to the point at which it could be marketed, that they participated in the formulation of the marketing material including the ‘webinars’, even if they did not deliver the latter, and that while the counterparty may well have used its own standard documentation for the spread bet and CSO, all of the other documentation—request to the employer or EBT to take over the CSO, board or trustee minutes, instructions to Aston Collie, among others—was prepared by the respondents, and that each of the documents was, as reg 10 puts it, in ‘standardised, or substantially standardised’ form, ‘determined by the promoter, and not tailored, to any material extent, to reflect the circumstances of the client’.

[28]Mr Forsyth was unable to explain, at least to my satisfaction, why a rational user might enter into the two contracts, a spread bet and matching CSO, and retain them both until maturity when, whatever the outcome, he was certain to suffer a loss if he did so. His response was that any investor, when entering into a spread bet or other risky arrangement, would commonly hedge the investment in order to minimise his exposure, a proposition which, in general terms, I can accept. Mr Forsyth’s attempt then to explain why, having protected himself in that way, the user might almost immediately divest himself of the protection by transferring the hedge to someone else was, in my view, incomprehensible. He was resistant to the suggestion by Ms Nathan that no prospective user would enter into a spread bet and matching CSO without the certainty that his employer or an EBT would relieve him of the latter, and made the point that in a few cases that had not happened. The documentary evidence, however, showed that in each of those cases the novation had not occurred because of a clerical or similar error. In my view it is an irresistible conclusion that the user entered into the two contracts with the certainty, accident aside, that he would be relieved of the CSO, and that the notion that the CSO was entered into for any other purpose is fanciful. Mr Forsyth also suggested that a company which took over hedging contracts in materially identical terms from each of its two director-shareholders knew nothing of the transactions until they had been entered into, an assertion which Ms Nathan, in my view rightly, described as sophistry. The statutory definitions

[29]The first of the definition provisions to which I have referred is to be found in FA 2004 s 307(1), which supplies the meaning of ‘promoter’ for the purposes of the DOTAS provisions:

‘For the purposes of this Part a person is a promoter— (a) in relation to a notifiable proposal, if, in the course of a relevant business, the person (“P”)— (i) is to any extent responsible for the design of the proposed arrangements, (ii) makes a firm approach to another person (“C”) in relation to the notifiable proposal with a view to P making the notifiable proposal available for implementation by C or any other person, or (iii) makes the notifiable proposal available for implementation by other persons, and (b) in relation to notifiable arrangements, if he is by virtue of paragraph (a)(ii) or (iii) a promoter in relation to a notifiable proposal which is implemented by those arrangements or if, in the course of a relevant business, he is to any extent responsible for— (i) the design of the arrangements, or (ii) the organisation or management of the arrangements.’

[30]Mr Way accepts that if the Alchemy scheme amounts to notifiable arrangements (which he does not accept) the respondents must be regarded as promoters of it within the statutory meaning. That concession seems to me to be both appropriate and inevitable, not only because of my findings of fact but also because of what appears at s 307(6):

‘In the application of this Part to a proposal or arrangements which are not notifiable, a reference to a promoter or introducer is a reference to a person who would be a promoter under subsections (1) to (5) if the proposal or arrangements were notifiable.’

[31]I do not therefore need to say any more about the meaning of ‘promoter’. The second relevant definition, provided by s 318, is of ‘advantage’, as it is used in both s 314A and s 306A: ‘“advantage”, in relation to any tax, means—(a) relief or increased relief from, or repayment or increased repayment of, that tax, or the avoidance or reduction of a charge to that tax or an assessment to that tax or the avoidance of a possible assessment to that tax,

(b) the deferral of any payment of tax or the advancement of any repayment of tax, or

[32]This provision is at the heart of Mr Way’s argument, to which I now come. Submissions and discussion

[33]I shall focus in this section of my decision on the real battleground between the parties, which resolved into three fundamental but interlinked issues. The core of the respondents’ argument is that the DOTAS provisions are not engaged at all, because the arrangements did not give rise to a tax advantage: the user could not have undertaken the transactions I have described in any other way, and the tax to which they give rise has been declared and paid. That argument depends for its success on an answer favourable to the respondents to the question which represents the second issue: what is the tax advantage, if any, which falls within the scope of the legislation, as applied to these transactions? That question cannot be answered without addressing the third issue: which of the various transactions fall within the scope of the ‘arrangements’ to which the statutory provisions apply? Logically those issues must be approached in reverse order.

[34]Before moving on I should deal briefly with HMRC’s argument, in respect of their preferred application, that all of the hallmarks which reg 10 of the 2006 Regulations sets out are present: the arrangements are standardised tax products, with substantially standardised documentation devised by the respondents as promoters and requiring little or no tailoring for each implementation, and they are designed to enable a person (the user, as I have termed that person) to gain a tax advantage, namely the payment of what would otherwise be regarded as employment income without the burden, or most of the burden, of the tax which that income would ordinarily bear. The Alchemy scheme is also caught, they say, by para (c) of the definition in s 318, in that it is designed to avoid the deduction of tax in accordance with the PAYE arrangements. I should mention for completeness that there was some discussion at the hearing about the question whether the companies which participated in the Alchemy scheme claimed relief, for corporation tax purposes, of the amount of the benefit to the employee of the novation, or the higher amount they were compelled to pay if the CSO was lost, but although Ms Nathan suggested that a claim for relief for the higher amount might itself amount to an ‘advantage’, that suggestion was not at the forefront of her argument and I shall leave it out of account for the purposes of these applications.

[35]Although, in his skeleton argument, Mr Way disputed HMRC’s claim that the Alchemy scheme is a standardised tax product he did not pursue the argument orally with any vigour. It will be apparent from what I have already said that I agree with Ms Nathan on this point. Even a cursory perusal of the documents shows a recurring pattern with little variation, apart from dates, names, amounts and similar details, from one iteration to another. It is also apparent that the documentation required minimal tailoring to each user. Mr Way’s response, as it is put in his skeleton argument, is that ‘[e]ach individual would discuss and negotiate specified tailor-made documentation for himself’. In my judgment that statement significantly overstates the position; as I have said, the dates, names, amounts and similar details differed from one iteration to another but neither Mr Forsyth nor Mr Way was able to identify any shaping beyond that to fit the needs of an individual user.

[36]I will come to the question whether there was, or was intended to be, a ‘tax advantage’ within the statutory meaning shortly, but again Mr Way did not seriously challenge the proposition that the aim and, if it succeeded, the result of the scheme was that the user received, largely tax free, a significant sum while his employer or an EBT of which he was a beneficiary suffered an approximately equal loss. Thus the outcome, if not the form, of a successful implementation was that the user received a payment at the expense of his employer or the EBT. I should add, at the risk of repetition but in case there is any residual doubt, that I am satisfied, in reaching that conclusion, that none of the users entered into the spread bet and CSO without the certainty, accident aside, that his employer or an EBT would relieve him of the CSO, and that all concerned knew and intended that, assuming the result of the two contracts was as hoped, the user would benefit while the employer or EBT would suffer a corresponding loss.

[37]I was referred by Ms Nathan to authority, particularly Pilkington Brothers Ltd v Inland Revenue Commissioners

[1982] 1 WLR 136

, 55 TC 705 and Snell v Revenue and Customs Commissioners

[2007] STC 1279

, in order to demonstrate that it is the totality of the arrangements, rather than any discrete part of them, which must be considered. As Mr Way did not challenge the premise I do not think it necessary to say more about those authorities. Mr Way’s argument, rather, was that while the entry into the spread bet, the CSO and the novation could fairly be described as ‘arrangements’, the final step in the ‘arrangements’ was the novation and that what followed, that is the win or loss on the spread bet and CSO, could not reasonably be regarded as part of the arrangements, but, as he put it in his skeleton argument, ‘flowed from the rights and obligations under the instruments held by different persons and were dependent on the performance of the underlying index’.

[38]That argument, in my judgment, does not address the provisions of reg 10 of the 2006 Regulations correctly. If one assumes, as I am willing to do for present purposes, that Mr Way is right to confine the ‘arrangements’ to the transactions ending in the novation, the question posed by reg 10(3) is whether the informed observer would conclude that (to repeat the provision for convenience) ‘the main purpose of the arrangements was to enable a client to obtain a tax advantage’. If one assumes for the purposes of argument that the outcome represented a tax advantage, it is difficult if not impossible to detect any other purpose to the arrangements than the attainment of that advantage. As I have already said, no rational investor would enter into a spread bet and matching CSO of the kind utilised in the Alchemy scheme without the near certainty that he would be relieved of the latter, and it is equally clear that the aim, even if it was not always achieved, was to put money in his hands at the ultimate expense of the employer or EBT.

[39]That conclusion brings me to the question whether the scheme, if it worked as intended, led to a tax advantage in the statutory sense. I was referred by both parties to what was said by Lord Wilberforce in CIR v Parker

[1966] AC 141

on the meaning of the same phrase as it was used in s 43(4)(g) of the Finance Act 1960, which defined a ‘tax advantage’ in these terms:

‘a relief or increased relief from, or repayment or increased repayment of, income tax, or the avoidance or reduction of an assessment to income tax or the avoidance of a possible assessment thereto, whether the avoidance or reduction is effected by receipts accruing in such a way that the recipient does not pay or bear tax on them, or by a deduction in computing profits or gains.’

[40]That definition is not identical to that provided by s 318 of FA 2004, but it is obviously very similar. At p 178 Lord Wilberforce said:

‘The paragraph, as I understand it, presupposes a situation in which an assessment to tax, or increased tax, either is made or may possibly be made, that the taxpayer is in a position to resist the assessment by saying that the way in which he received what it is sought to tax prevents him from being taxed on it; and that the Revenue is in a position to reply that if he had received what it is sought to tax in another way he would have had to bear tax. In other words, there must be a contrast as regards the “receipts” between the actual case where these accrue in a non-taxable way with a possible accruer in a taxable way, and unless this contrast exists, the existence of the advantage is not established.’

[41]HMRC’s position, shorn of detail, is straightforward: it is that the Alchemy scheme is a device designed to extract from a company or EBT what would otherwise be, or be treated as, employment income in the guise of betting winnings and that, to borrow Lord Wilberforce’s words, ‘if he had received what it is sought to tax in another way he would have had to bear tax’—that other way being as the employment income the receipt truly is. Mr Way’s response is that what the user receives could never be categorised as employment income because his winnings from the spread bet are derived from an asset which he had acquired in his personal capacity and had owned at all times. It was not possible to say, as HMRC appeared to do, that the winnings represented some sort of benefit conferred on the person by his employer or the EBT trustees. Nor could it be said that the payment made, in this case, by the employer or EBT to the counterparty conferred some benefit on the person. The benefit to him was the novation; anything which came later depended on the performance of the basket, had no connection with his employment and did not affect him.

[42]An analogy advanced by Mr Way was the receipt by an employee of shares from the employer; that receipt was a taxable emolument: see Weight v Salmon (1935) 19 TC 174 ; but anything which followed, such as dividends derived from the shares or their proceeds of sale if disposed of, was not. That proposition was made clear by the House of Lords in Abbott v Philbin

[1961] AC 352

in which the question was whether an employee who was granted an option to buy shares in his employer company should be taxed in the year in which he acquired the option (1954-55) on the value of the option over the price paid for it, or in the later year (1955-56), when he exercised the option, on the benefit to him of buying the shares at less than market value. At p 367, Viscount Simonds, who was in the majority, said that:

‘… the argument for the Crown appeared to demand for its success that the grantee of the option did not acquire a perquisite at the date of the grant. There could not be one perquisite at the date of the grant and a second perquisite when the shares were taken up. Therefore the Crown's case, in my opinion, fails at the initial step. But there are other grave difficulties in the way of its success. The taxable perquisite must be something arising “therefrom,” i.e., from the office, in the year of assessment. I do not want to embark on the notoriously difficult problem as to the year to which for the purpose of tax a payment should be ascribed, if it is not expressly ascribed to any particular year. But I do not find it easy to say that the increased difference between the option price and the market price in 1956 or, it might be, in 1964 in any sense arises from the office. It will be due to numerous factors which have no relation to the office of the employee, or to his employment in it. The contrast is plain between the realised value, as it has been called, of the option when the shares are taken up (though the realisation falls short of money in hand) and the value of the option when it is granted. For the latter is nothing else than the reward for services rendered or, it may be, an incentive to future services. Unlike the realised value it owes nothing to the adventitious prosperity of the company in later years. On this ground also I should reject the claim of the Crown.’

[43]What HMRC were attempting to do, said Mr Way, was to compare the tax charge which arose on the novation with the tax charge that would have arisen if, instead, the employer or the EBT had made a direct payment to the individual. That approach cannot be reconciled with the explanation of the concept of a tax advantage provided by Lord Wilberforce in Parker ; it is not a comparison of like with like. The novation is the only transaction by which value was transferred from the employer or EBT to the individual. The spread bet winnings, if the bet was successful, did not come from the employer or EBT, which had never been a party to that transaction. The only benefit conferred on the individual by the employer or EBT was that of relieving him of a potentially onerous asset, but there was no certainty that the user would win the spread bet and that the CSO would result in a loss, and it did not always happen. An equation of an uncertain situation of that kind and a straightforward payment from the employer or an EBT to the individual was not possible.

[44]If, instead, the spread bet resulted in a loss, the individual would have to pay the counterparty from his own resources and, because the transaction was entirely one into which he had entered on his own account, he would have no recourse to the employer or the EBT trustees. Mr Way accepted that a loss was not the desired result, but he emphasised that it had occurred on numerous occasions. The individual whose spread bet was unsuccessful suffered real economic loss since in addition to losing his stake under the spread bet he would have to pay income tax (based on the value of the benefit at novation) and potentially capital gains tax on the granting of the CSO (depending on the sums and the individual’s circumstances). It was true that in some cases when the individual lost the spread bet the transactions were repeated, but that was not always the case and there were some instances where the individual was left with a substantial loss.

[45]Again, I do not think Mr Way is addressing the right question. What is in issue is not whether the user could have disposed of the CSO to his employer or an EBT in a different way which would have attracted a higher tax charge, nor is it whether the Alchemy scheme works, in the sense that neither the user nor the employer or EBT can be compelled to pay more tax. The question, as the legislation makes clear, is not whether the user is enabled to obtain a specific tax advantage, nor whether the scheme succeeds in its aim, but whether, as reg 10(3) of the 2006 Regulations puts it, the ‘informed observer [would] conclude that the main purpose of the arrangements was to enable a client to obtain a tax advantage’ [emphasis added]. For the reasons I have already given it does not seem to me to be seriously arguable that the arrangements in this case do not meet that description. The only reasonable conclusion to which the hypothetical informed observer could come is that the purpose of the scheme is to extract money from a company or EBT without incurring the tax charge which would be incurred were that money to be paid as employment income, and thereby to secure a tax advantage. Whether the scheme succeeds in that aim is beside the point.

[46]I do not agree with Mr Way that any analogy is to be drawn between a case such as this and the grant of share options as in Weight v Salmon and Abbott v Philbin . It does not seem to me that the grant of share options from which the recipient may or may not benefit at some time in the future, and the relieving of an employee of a burdensome or probably burdensome contract as part a scheme designed to put untaxed money in his hands are comparable, even if in each case a tax charge is incurred on the present-day value of the benefit to the employee. For similar reasons I disagree with Mr Way about the application to this case of Condition 2 of reg 18 of the 2006 Regulations; the focus is not on the novation alone, but on the arrangements taken as a whole and the ‘informed observer’s’ perception of them. Conclusions on HMRC’s preferred application

[47]I am satisfied that:(a) the Alchemy scheme amounts to ‘arrangements’ in the statutory sense;

(b) the scheme enables, or might be expected to enable, a person to obtain a tax advantage;

(d) the arrangements are in a standardised form and have substantially standardised documentation, in a form determined by the respondents, requiring minimal tailoring for each user;

(e) the arrangements have been made available for use by more than one person; and

(f) the respondents are the ‘promoters’ of the scheme in the statutory sense.

[48]It follows that the arrangements are notifiable and I therefore make the preferred order sought by HMRC. The alternative application

[49]I do not need to deal with this application, but will nevertheless make some comments in case I am found elsewhere to have fallen into error in respect of the preferred application. As I have said, the test is different, but the only real point of contention is whether, as s 306A(3) puts it, HMRC ‘have taken all reasonable steps to establish whether the proposal or arrangements are notifiable’ and ‘have reasonable grounds for suspecting that the proposal or arrangements may be notifiable’. It is clear from my earlier findings of fact that the second limb of that test is satisfied, and I do not need to say any more about it.

[50]It was plain both from his witness statement and oral evidence and from the correspondence that Mr Hole and his colleagues had attempted to engage with the respondents and their agents in order to ascertain what the arrangements were, how they were implemented, who had taken advantage of them, and with what result. I formed the impression that Mr Hole had been very patient in the face of an unwillingness to be forthcoming; it would be unfair to describe the responses he received as stonewalling, but they can only be considered to represent an attempt to deflect him. I am left in no doubt that HMRC have taken all the reasonable steps contemplated by s 306A(3) and, had I not made the preferred order, I would make the alternative order. Appeal rights

[51]By virtue of art 3(a)(i) of the Appeals (Excluded Decisions) Order SI 2009/275 any decision of this tribunal about the applicability of ss 306A and 314A is an excluded decision for the purposes of s 11(1) of the Tribunals, Courts and Enforcement Act 2007 and there is accordingly no right of appeal against this decision. COLIN BISHOPP TRIBUNAL JUDGE RELEASE DATE: 11 SEPTEMBER 2017

Cited in 10 later judgments

and 1 more case

Revenue & Customs v Curzon Capital Ltd (INCOME TAX/CORPORATION TAX : Anti-avoidance) 2019