DB Group Services (UK) Ltd v HM Revenue and Customs

Two appeals are before us. In one, the appellants are The Commissioners for Her Majesty’s Revenue and Customs (‘HMRC’) and the respondent is UBS AG (‘UBS’) (‘the UBS appeal’). In the other, the appellant is DB Group Services (UK) Limited (‘DB’) and the respondents are HMRC (‘the DB appeal’).

UBS is a well known bank. DB was at the material time the main employer in the group headed by Deutsche Bank AG, another well known bank. Both banks have traditionally rewarded employees with bonuses. The payment of bonuses by banks, particularly since the financial crisis of 2008, has given rise to considerable public comment, not all of it favourable. These cases relate to a prior period, when both banks adopted what the Upper Tribunal described as a ‘carefully planned tax avoidance scheme which was designed to enable the [banks] to provide substantial bonuses to employees in the tax year 2003/04 in a way that would escape liability to both income tax and national insurance contributions.’ The top rate of income tax was then 40% and the perceived beauty of the schemes was that, if they worked, both the banks and the employees derived considerable tax benefits from them. The intent of the schemes was that the employees would receive shares in the nature of ‘restricted securities’, which would in consequence escape income tax under the provisions of the tax regime in Chapter 2 of Part 7 of the Income Tax (Earnings and Pensions) Act 2003 (‘ITEPA’); and the banks would pay the bonuses into the schemes without having to account to HMRC for income tax or for NICs for the relevant employees or their own liabilities on their earnings.

Both banks appealed to the First-tier Tribunal (Dr David Williams and Mr David Earle, ‘the FTT’) against HMRC’s determinations that the sums allocated to the employees as bonuses at the start of each scheme were liable to income tax under Part 7 of ITEPA as earnings from their employment and to Class 1 NICs on the same basis. The determinations were against the banks rather than the employees since the banks were in each case liable to deduct and account for any liability to PAYE income tax and Class 1 NICs on behalf of their employees. The FTT heard the appeals successively in February 2010, each appeal occupying five days.

In the UBS case, the determinations were both dated 13 October 2008 and were: (a) for £36,863,600 for income tax in respect of bonus payments of around £92m made by UBS in early 2004 to employees; and (b) for £12,717,942 for Class 1 NICs in respect of the same payments. In the DB case, the determinations were both dated 29 April 2008 and were: (a) for £36,520,000 for income tax in respect of bonus payments of around £91m made by DB in early 2004 to employees; and (b) for £12,599,400 for Class 1 NICs in respect of the same payments. The FTT released their decision on the UBS appeal on 6 August 2010 (and then re-issued it with corrections on 15 September 2010); and they released their decision on the DB appeal on 19 January 2011. The FTT dismissed both appeals, although not for identical reasons as there were factual differences between the two schemes.

With the permission of the FTT, the banks appealed to the Upper Tribunal (Tax and Chancery Chamber) (Henderson J and Mr Charles Hellier, ‘the UT’), which heard the appeals together over five days in February 2012. The outcome of the UT’s 212-paragraph decision released on 17 September 2012 was that they allowed the UBS appeal and dismissed the DB appeal: UBS AG and DB Group Services (UK) Limited v. The Commissioners for Her Majesty’s Revenue and Customs [2012] UKUT 320 (TCC); [2013] STC 68. So it is that HMRC are now appellants in the UBS case, seeking to overturn the UT’s upholding of the UBS scheme; and DB is an appellant seeking to vindicate its own scheme. HMRC in turn seeks to uphold the UT’s decision in the DB case for further reasons which differ from those given by the UT.

We heard the appeals together over three days. Mr Lasok QC (leading Mr Vallat and Ms Blackwood) opened HMRC’s appeal in the UBS case, following which Mr Prosser QC presented UBS’s responsive case. Mr Goy QC (leading Ms Shaw QC) for DB then also responded to HMRC’s case and opened the DB appeal. Mr Lasok replied generally and Mr Goy had the last word in reply on the DB appeal. I shall deal first with the UBS appeal and then the DB appeal.

Permission for HMRC to appeal in the UBS case was given by the UT on four grounds. HMRC renewed before us their permission application on two further grounds.

UBS does not argue that its scheme was other than a pre-conceived one directed at avoiding income tax and NICs. I need say no more about NICs, upon which we had no separate argument: it is agreed that the position in relation to NICs is the same as in relation to income tax. UBS’s position is simple. It says the bonuses are not taxable because the scheme brought them within the highly prescriptive provisions of Chapter 2 of Part 7 of ITEPA, inserted by the Finance Act 2003, being provisions that enabled the bonuses to enjoy the tax exemptions conferred by sections 425 and 429 of ITEPA. HMRC’s stance is that the sole reason for the scheme was tax avoidance, and the only reason why an employee would participate in it would be to avoid tax. Mr Lasok described it as a ‘cash in, cash out’ scheme: UBS would decide upon the intended bonus, which it would pay into the scheme and which the employee could take out a few weeks later. I shall later explain why I regard that as a misdescription of the scheme.

The scheme involved UBS subscribing for shares in a company and then awarding the shares to the participating employees. Normally, the award to an employee of shares in a company by reason of his employment would give rise to a charge to income tax in respect of the acquisition on an amount equal to the market value of the shares at the time. But there was no such charge if the shares were ‘restricted securities’ within the meaning of section 423(2), in which event section 425 conferred a qualified tax exemption. In principle, section 426 would, however, charge income tax on the full value of the shares on a ‘chargeable event’, which would include when the shares ceased to be ‘restricted’, as happened in these cases. But Parliament also chose to permit an escape from such charge by section 429, which provided for an exemption from it if the shares were in a company which (amongst other conditions) had not in the previous 12 months been under the employer’s ‘control’ as defined by section 416 of the Income and Corporation Taxes Act 1988 (‘ICTA’). Provided the shares fell within these exemptive provisions, there would therefore be no section 426 liability to income tax, although there would be a potential charge to capital gains tax (‘CGT’) on the disposal of the shares. For those employees who were non-UK domiciled, there would, however, be no charge to CGT unless gains were remitted to the UK. For those employees who were UK-domiciled, any charge to CGT could be reduced to 10% by business taper relief if the shares were held for two years and the employees continued to be employed by UBS.

UBS operated performance incentive schemes whereby, in or around February in each year, it made bonus awards to selected employees. The schemes were limited to employees still in UBS’s employment at the time of the award and who had not given or received notice of termination. UBS’s Executive Share Investment Plan that is now in question (‘the scheme’) was one such scheme and it was designed to enjoy the tax benefits just summarised. UBS invited selected senior employees to indicate by 12 December 2003 whether they wished to participate in it, and thus to receive a proportion of their bonus award, if any, in shares instead of cash. 426 employees agreed to participate, ten of whom had a contractual right to be paid a minimum cash bonus, whilst for the remaining 416 any bonus was discretionary.

UBS engaged Mourant & Co Trustees Limited (‘Mourant’), for a fee, to set up a special purpose vehicle for the purposes of the scheme. On 19 January 2004, an offshore special purpose vehicle company called ESIP Limited (‘ESIP’) was incorporated. Its sole function was to implement the scheme. On the same day, Mourant, in its capacity as the trustee of a pre-existing, unrelated charitable trust, the Sidemore Charitable Trust (‘Sidemore’), resolved to subscribe for 200 ordinary £0.01 shares in ESIP for £2.

On 23 January 2004, a UBS remuneration committee finalised a list of employees who would receive a bonus award for 2003. This was not communicated to the employees.

On 26 January 2004, the share capital of ESIP was reorganised and ESIP adopted a new Memorandum and Articles of Association (‘the M&A’). They provided for ESIP to have an authorised share capital of £2.6m voting ordinary shares (‘VOS’) and 100,000 non-voting redeemable shares (‘NVS’). The 200 subscriber ordinary shares already issued were reclassified as VOS.

Article 2(7) entitled holders of NVS to receive dividends and surplus assets on a winding up and to redeem all or any of their shares on 22 March 2004, 22 March 2006 or 22 June 2006 for the same amount that the holder would have received if there had been a winding-up on the relevant redemption date.

Article 2(14) provided for a forced sale of the legal and beneficial interest in the NVS in certain circumstances. It would take effect if the closing value of an index (‘the Index’, which was to be specified by the directors) on any date between the first date of issue of the NVS (28 January 2004) and 19 February 2004 (defined as ‘the Restricted Period’) was greater than the ‘Trigger Level’ (also to be specified by the directors). In such event, the legal and beneficial interest in each NVS was immediately to be sold to the trustees of a discretionary trust for UBS employees (the UBS Employee Benefits Trust Limited) for a consideration equal to ‘the Forfeiture Price’, namely one equal to 90% of their ‘Market Value’ as defined in the M&A, being their value if no restrictions applied to them. The purpose of the forced sale provision was to make the NVS ‘restricted securities’ for the purposes of section 423(2). The directors specified the FTSE 100 Index as the Index, and 4,749.15 as the ‘Trigger Level’.

Article 2(15) provided that at any time when the holder or beneficial owner of any NVS was UBS or any other group company, article 2(7) did not apply, and that the NVS carried no right to dividends or other distributions, were not redeemable and would only carry a right to receive par on a winding-up. The purpose of this provision was to ensure that during the short period from 28 to 29 January 2004, when UBS held all the NVS, it did not have ‘control’ of ESIP so as to cause a failure of one of the conditions for the enjoyment of the section 429 tax exemption.

On 26 January 2004, Sidemore resolved to subscribe for an additional 1,699,800 VOS in ESIP for a total price of £16,998. This brought its total investment in ESIP to £17,000. It held 1,700,000 VOS.

On 27 January 2004, UBS subscribed for 900,000 VOS at £9,000. The effect was that whilst Sidemore’s holding of VOS gave it shareholder control of ESIP, UBS’s holding was sufficient to enable it to block a special resolution (for example, to wind ESIP up).

On 28 January 2004, UBS offered to subscribe for 91,880 of the NVS for £1,000 each, a total of £91,880,000. That was the exact equivalent of the cash payments which the UBS employees participating in the ESIP scheme would otherwise together have received as cash bonuses. The UBS offer was, however, conditional on ESIP agreeing to: (i) set the Index and Trigger Level in the manner specified; (ii) purchase call spread options for the NVS, to neutralise the effect of the Trigger Level being reached and of there being a forced sale of the NVS (‘the hedging arrangements’); (iii) deposit the subscription price paid by UBS in an interest-bearing account until 20 February 2004; and (iv) use the subscription price paid by UBS to (a) buy shares in UBS in the last five days of February 2004 if the forced sale did not occur, or (b) keep the subscription price paid as cash if it did.

The directors of ESIP accepted and implemented those conditions. Pursuant to condition (ii), ESIP applied about 3% of the £91.88m in purchasing call options relating to the Index with an expiry date of 20 February 2004, so that if the Index exceeded the Trigger Level ESIP would make a gain, its net assets thereby increasing by about 10%. The effect of the hedging arrangements was that although the employees would be required to sell their NVS for 90% of their unrestricted value, they would not be materially worse off because the unrestricted market value of the NVS would be equal to approximately 110% of their original subscription price.

On 29 January 2004, UBS allocated 91,856 NVS to the relevant employees. On the same day, it transferred 24 NVS to the trustee of a UBS employee- benefit trust. That allocation accounted for UBS’s full holding of 91,880 NVS.

In the event, the Index did not exceed the Trigger Level, and on 19 February 2004 the NVS ceased to be subject to the forced sale provisions and the call options did not pay out: in short, the restriction was lifted. In compliance with condition (iv)(a) above, ESIP invested its net assets in the purchase of quoted UBS shares.

Articles 2(7)(e) to (g) of the M&A provided for an absolute right to redeem the NVS in March 2004, March 2006 and June 2006 at a price set by a pre-determined formula.

On 26 February 2004, UBS emailed the employees reminding them of their right to cash out in March 2004 and of the March deadline for doing so. More than half of the NVS were redeemed on 22 March 2004 at a price of £977.50 per share. Dividends were paid on the NVS in November 2004 (£13 per share) and December 2005 (£20 per share). On 22 March 2006, and once the two years necessary to take advantage of maximum CGT taper relief had ended, further NVS were redeemed for about £1,519 per share, which reflected a large increase in the value of UBS shares. On 22 June 2006, further NVS were redeemed, for £1,429 per share, reflecting a fall in the value of UBS shares. The remaining 44 NVS were redeemed in November 2006 at ESIP’s own initiative when a resolution to wind ESIP up was passed.

I shall have to refer to further legislative provisions later, but shall here refer to those at the heart of the issues.

I refer first to section 18 of ITEPA, ‘Receipt of money earnings’. This is relevant to HMRC’s submission that a charge to income tax arose upon the payment by UBS to ESIP of its subscription money for the NVS. Section 18 provides materially:

Section 19, ‘Receipt of non-money earnings’, provides materially:

I turn next to the provisions of Chapter 2 of Part 7 of ITEPA. The UT, in [11] to [19] of their judgment, provided a valuable explanation of the historical context in which Chapter 2 came to be enacted. I shall also summarise the background, but relatively shortly, and for this purpose I have gratefully drawn, in part verbatim, on the UT’s much fuller explanation. In Abbott v. Philbin [1961] AC 352, the majority of the House of Lords held that where an employee is granted a share option by reason of his employment, income tax is charged on the realisable monetary value of the option at the time of its acquisition. This was so even if the option was hedged around with conditions, provided it was capable of being turned to monetary account. As the value of such option rights was usually small, section 25 of the Finance Act 1966 reversed Abbott by imposing a charge to income tax on the exercise, assignment or release of employees’ share options, on an amount equal to the gain thus realised, whilst removing any charge to tax on the grant of the option. Section 25 was consolidated as section 186 of the Income and Corporation Taxes Act 1970. Further refinements were added by the Finance Act 1972.

The legislation so far mentioned applied only to share options. It did not apply to a share incentive scheme, under which an employee subscribed for, or was awarded, shares to which restrictions were attached for a prescribed period (for example, in relation to voting rights or dividend receipts) and which would become more valuable on the lifting of the restrictions. By section 79 of the Finance Act 1972, a charge to income tax was imposed on the appreciation in value of the shares at the end of a period defined by the earliest to happen of the lifting of the restrictions, the time when the employee ceased to have any beneficial interest in the shares and the expiry of seven years from their acquisition. There were several exceptions to this charge.

The Finance Act 1998 inserted new sections 140A to 140C into ICTA. Their effect was to remove the charge to tax in respect of the acquisition of the conditional interests in shares and to impose a new charge to tax on the market value of the shares when the condition fell away. Sections 140A to 140C were re-enacted without amendment as the original Chapter 2 of Part 7 of ITEPA, Part 7 being headed ‘Employment income: share-related income and exemptions’. Within a few months of the coming into force of ITEPA, the Finance Act 2003 substituted a new and more complex Chapter 2, with which these appeals are concerned. Chapters 1 and 2, as so substituted, contain the provisions central to the issues before us.

Chapter 1 comprises sections 417 to 421L. Section 417(1) explains that Part 7 ‘contains special rules about cases where securities, interests in securities or securities options are acquired in connection with an employment’, and that the rules are contained in various Chapters in Part 7, including Chapter 2 (restricted securities). Section 417(3) explains that Chapter 2 makes provision for amounts ‘to count as employment income’ and section 417(4) explains that Chapter 2 also provides ‘for exemptions and reliefs from income tax’. Section 420(1) defines ‘securities’ for the purposes of (inter alia) Chapter 2, as including:

Section 421 explains that in (inter alia) Chapters 1 and 2, ‘market value’ has the same meaning as it has for the purposes of the Taxation of Chargeable Gains Act 1992 by virtue of Part 8 of that Act; and that ‘where consideration for anything is given in the form of an asset (as opposed to a payment), any reference in [inter alia, Chapters 1 and 2] to the amount of the consideration is to the market value of the asset’. Section 421B provides materially:

Chapter 2, ‘Restricted Securities’, comprises sections 422 to 432. Section 422 explains that Chapter 2 applies to ‘employment-related securities’ if they are either ‘restricted securities’ or ‘a restricted interest in securities’ at the time of the acquisition. Section 423, ‘Restricted securities and restricted interest in securities’, provides:

Section 425, under the heading ‘Tax exemption on acquisition’, provides:

Section 425 therefore provides for an exemption from tax upon the acquisition of restricted securities. But section 426, under the heading ‘Tax charge on post-acquisition chargeable events’, provides for a charge to tax on the occurrence of a ‘chargeable event’. Section 427 explains what a ‘chargeable event’ is and section 428 explains the amount of the tax charge. There is no need to set those provisions out in full, but I should set out section 427(3):

The section at the heart of these appeals is section 429, ‘Case outside charge under section 426’. Its essence is that section 426 does not apply to cases covered by section 429. It is section 429 that enables the scheme to enjoy its claimed tax advantages. From the perspective of the scheme’s architects, it was a generous piece of legislation. Mr Lasok tried, but in my view failed, to explain the legislative rationale for such generosity; and Mr Prosser and Mr Goy were equally baffled by it. Section 429 provides:

The UT described the issues in both the UBS and the DB cases as having raised issues under three headings: (1) did the employees become entitled to be paid their bonuses in money before the sums allocated to them were applied in acquiring scheme shares? If yes, the bonuses were subject to tax; (2) if no, did any charge to tax arise under the provisions of Chapter 2? That raised the questions of whether the scheme shares were ‘restricted securities’ within section 423 and, if so, whether the employees were entitled to the tax exemption provided by section 429; (3) applying the Ramsay principle (WT Ramsay Ltd v. Inland Revenue Commissioners [1982] AC 300), and on a realistic appraisal of the facts, did the schemes fall outside Chapter 2 altogether?

In the UBS appeal, the FTT answered question (1) in favour of HMRC as regards the 10 participants who had a contractual entitlement to a bonus, but in favour of UBS as regards the other 416 participants who had no such entitlement. In answer to question (2), they held that the scheme shares were not ‘restricted securities’ (so that the scheme failed) but that, if wrong on that, the section 429 exemption was available. In answer to question (3), they held that the scheme anyway failed on Ramsay grounds. Thus the FTT found against the scheme both on technical and Ramsay grounds. The UT allowed UBS’s appeal and upheld the scheme.

The structure of Chapter 2 is therefore broadly, and so far as material, as follows. Chapter 2 applies to ‘employment-related securities’ if they are ‘restricted securities’ acquired by a person where the opportunity to acquire them was by reason of an employment of that person. They will be ‘restricted securities’ if they are subject to such a condition as is explained in section 423, as a result of which they may be subject to a ‘transfer, reversion or forfeiture’ in consequence of which the holder will receive less than their market value were they not subject to such a condition. Section 425 provides for a qualified exemption from tax on the acquisition of restricted securities to which subsection (1) applies. Section 426, however, provides for a charge to tax on the occurrence of a post-acquisition ‘chargeable event’ in relation to the restricted securities; and section 427 explains what the ‘chargeable events’ are, and they include ‘the employment-related securities ceasing to be restricted securities …’. Section 428 explains the calculation of the tax in such event. Section 429 explains that cases that fall within its provisions are ‘outside the charge under section 426’.

HMRC questions whether the scheme satisfies the technical requirements of Chapter 2 so as to fall within it. But even if it does, they say that the scheme is not entitled to the benefit of the tax exemptions to which it might then be thought to be entitled. That is because it is simply a tax avoidance scheme and it was not the intention of the legislature to extend the benefit of the provisions of Chapter 2 to artificial arrangements, such as the scheme, that are said to have no commercial purpose. That is HMRC’s Ramsay point. I come to each of the six grounds argued by HMRC.

Mr Lasok developed first his ‘broad’ Ramsay proposition (although no counsel actually referred us to Ramsay), and it was to the effect that the UT erred in its application to the scheme of the Ramsay doctrine. If right in this submission, the fiscal consequence was said to be that UBS would be treated as making an award of cash bonuses to the employees.

For what Mr Lasok said was the basic approach to the consideration of whether the scheme fell within the intent of Chapter 2, he cited The Collector of Stamp Revenue v. Arrowtown Assets Limited [2003] HKCFA 46, a decision of the Court of Final Appeal of the Hong Kong Special Administrative Region. The facts were miles from ours, but Mr Lasok referred us to some observations of Lord Millett, as well as to the following statement of Ribeiro PJ as to the driving principle in the Ramsay line of cases:

Mr Lasok also referred to the judgment of Arden LJ in Astall and another v. Revenue and Customs Commissioners [2010] STC 137, where she said:

Mr Lasok said the message from these statements is that legislation should be interpreted by reference to the purpose it is intended to achieve, which requires an application of it to the real world rather than to a fictitious one. He referred us to certain findings of the FTT as to the nature of the scheme. The FTT described its significant feature as an arrangement for the payment of bonus sums. They were paid into the scheme rather than directly to the employee but ultimately the employee was to get cash, and once the scheme had served its purpose the arrangement was wound up. The immediate purpose of the scheme was to take steps to remove income tax and NIC liabilities from the bonus received by the employee. The FTT then referred to the hedging arrangements in the scheme that were directed at countering the risk to the employees in case of a forfeiture. They said the aim there was to achieve a complete offset of the loss to an employee if the trigger event occurred, the reality being that the scheme was designed so that employees would not suffer any significant loss in such event. These considerations were part of the FTT’s reasoning that led it to the conclusion that the securities were not ‘restricted securities’ within the meaning of Chapter 2. Their conclusion was that the predominant reason for the scheme was the mitigation of the burden of income tax and both employer and employee NICs.

The UT, however, took a different view. They regarded the FTT’s explanation of the ‘reality’ of the scheme as little more than a restatement of its intended fiscal purpose. It was also inaccurate for the FTT to say, as they had, that the scheme provided for employees to receive the ‘same sums’ on redemption of the NVS as they would have received as earnings, because of the requirement under the scheme for the subscription money to be invested in the purchase of UBS shares. It followed that the amount received on redemption was linked to the value from time to time of the UBS shares; and, the UT pointed out, ‘for the substantial number of employees who held their NVS until the second or third redemption dates, the amount received bore no relation to the cash value of the original bonus.’

Mr Lasok submitted that that proposition unfairly summarised the FTT’s approach and said the UT’s point about the potential difference between the original bonus allocation and the actual return to scheme participators on the redemption of their NVS was unsound. The redemption arrangements were a fundamental part of the scheme. He referred to an email sent to scheme employees on 26 February 2004 reminding them in bold that the first optional redemption period would run from 1 March 2004, giving details of how to redeem and explaining that 12 March 2004 was the deadline for redeeming during this period. The focus on redemption for cash showed that the securities were not intended to function as securities in the normal sense. They were not granted as interests in companies with an independent business: ESIP had no business apart from the function it was performing for the purpose of the scheme. It was simply a vehicle for passing to the employees the cash identified for them at the outset. The scheme had no commercial purpose, only a tax avoidance purpose. Chapter 2 was not intended to apply to such a scheme. It was wrong to regard it as a scheme under which employees were being rewarded by the allocation of shares. Its sole purpose was to reward them in cash. The shares were not, therefore, ‘restricted securities’ at all, and the scheme fell outside Chapter 2. Mr Lasok did not go so far as to say that the shares issued to the employees did not exist, or were to be regarded as not existing. His point was that they did not perform the function envisaged for ‘restricted securities’ in Chapter 2. The fact that those employees who redeemed their shares in 2006 and paid CGT, subject to taper relief, was not, it was said, a relevant consideration.

Before dealing with the response to this argument, I shall also summarise Mr Lasok’s alternative Ramsay submission.

Mr Lasok’s alternative Ramsay point, for which he sought permission to appeal, is one he described as a narrower or reduced version of his primary Ramsay argument. If right on this alternative argument, but not on his broader one, he said the fiscal conclusion would be that whilst UBS had rewarded the employees with shares, they were not ‘restricted shares’ and would therefore have been taxed by reference to their market value.

The argument centred on UBS’s admission that the forfeiture restriction imposed upon the ESIP shares had no commercial purpose. It was, therefore, a commercially irrelevant provision inserted solely for the purpose of achieving the intended tax avoidance. Mr Lasok underlined that by noting that the hedging provisions in the scheme were designed to neutralise any adverse consequences of the forfeiture provision, and so served to render the forfeiture provision substantially meaningless in real terms. In substance, the narrower Ramsay argument amounted to focusing not on the alleged artificiality of the scheme as a whole, but on just one key element of it, the forfeiture provision.

In response to both Ramsay arguments, Mr Prosser explained that there was a good reason why we were not referred to Ramsay. That is because the House of Lords in Barclays Mercantile Business Finance Ltd v. Mawson (Inspector of Taxes) [2005] 1 AC 684 had since sufficiently explained what the Ramsay principle is, as recognised in Mayes v. Revenue and Customs Commissioners [2011] STC 1269. It is unnecessary to cite at length from Mummery LJ’s illuminating judgment in Mayes. The thrust of it is in the following excerpts:

We were referred to Mawson itself, from which I shall not also cite, but record that Mr Prosser drew our attention to [26] to [42] in the speech of Lord Nicholls of Birkenhead. Mr Prosser’s essential submission was that these cases show that the task for the court in a case such as this is to adopt a purposive interpretation of the relevant legislation; and the transaction in question should be looked at as a whole, as opposed to on a step by step basis. There is nothing special about tax cases, to which the approach should be the same as in non-tax cases. The task in this case is to focus on the provisions of Chapter 2, read in the context of Part 7, and identify its requirements. When that is done, Mr Prosser submitted that there is no warrant for a conclusion that Chapter 2 has no application to transactions carried out for tax avoidance purposes only, nor for Mr Lasok’s alternative submission that securities that are subjected to restrictions for tax purposes only are not ‘restricted securities’.

With a view to making this good, Mr Prosser took us on a tour through Part 7. Section 417(1), in Chapter 1, explains that Part 7 contains special rules about cases where securities or interests in them are acquired in connection with an employment. The rules are in 13 separate Chapters. Section 417(4) explains that Chapters 2, 3 and 5 to 10 make provision for ‘exemptions and reliefs from income tax’. Tax exemptions are, therefore, part of what Part 7 is about. Securities are defined in section 420(1) under seven sub-paragraphs. Nothing there suggests that the only securities contemplated are those given to an employee by way of an incentive. The definition extends to securities in companies unrelated to the employer and is, apparently deliberately, spread very wide. Mr Prosser accepted that it is legitimate in a case such as this to ask whether what the employee gets is securities or money. What it is not legitimate to ask is whether he got what he did for tax avoidance reasons, which is the broader question that Mr Lasok submitted should be answered.

Section 421B then explains that Chapters 2 to 4 apply to securities acquired by a person ‘where the right or opportunity to acquire [them] is available by reason of an employment of that person or any other person’. Section 421(3) shows that, save in specified cases, there is a presumption that such a right or opportunity to acquire securities made available by one’s employer ‘is available by reason of an employment of that person …’. Again, there is no suggestion there that what counts is whether the securities are made available for incentive purposes. Section 421E explains cases in which Chapters 2, 3 or 4 do not apply: but it does not include in the exclusions the provision of securities for tax avoidance purposes. According to HMRC, that is so obvious that there was no need to say so.

Moving to Chapter 2, and assuming that what the employees got were ‘securities’ acquired by reason of the employees’ employment, the question is whether they are in principle governed by Chapter 2. Chapter 2 provides the answer by explaining in section 422 that it applies to ‘employment-related securities’ if they are ‘restricted securities at the time of the acquisition.’ There is then a wide definition of ‘restricted securities’ in section 423. Assuming (contrary to another HMRC argument) that the NVS in ESIP satisfied the conditions of section 423, Mr Prosser’s submission was that there is no basis for Mr Lasok’s assertion that they nevertheless fell outside the grasp of section 423 because they were designed for tax avoidance purposes. Section 424 sets out the ‘Exceptions’, namely the cases when employment-related securities will not be restricted securities: again, it does not include cases where they were acquired for tax avoidance purposes. If one negotiates Chapter 2 as far as sections 426 to 429, they show that section 429 excludes restricted securities from the tax charge that would otherwise arise under section 426. Mr Prosser was unable, like Mr Lasok, to explain the rationale for section 429 and said that it ‘puzzles everyone … because [it] doesn’t make sense’. He could not discern any purpose in it save one aimed at affording an exemption to tax in cases which satisfied its conditions. Even Mr Lasok was reduced to arguing a point of construction in relation to section 429: it was no part of his argument (save as part of his broad Ramsay submission) that section 429 is not there to help tax avoiders. In fact, Parliament was quick to recognise that it may have missed a trick there, and section 86 of the Finance Act 2004 amended section 429 so as to exclude its application in tax avoidance cases. That amendment may, on one view, be said to have assumed that Mr Lasok’s broad Ramsay argument was wrong; but of course, if so, Parliament may in that respect itself have been mistaken.

In this last context, Mr Prosser also referred us to Chapter 3A in Part 7, headed ‘Securities with artificially depressed market value’. Its provisions were expressly directed to cases ‘where the market value of employment-related securities … is reduced by things done otherwise than for genuine commercial purposes.’ That again shows that, when Chapter 2 was enacted, Parliament knew perfectly well how to deal with tax avoidance schemes, and did so in Chapter 3A: but it did not do so until a year later in Chapter 2. In like vein, Chapter 3B deals with cases where the market value is artificially enhanced.

Turning to HMRC’s broad Ramsay submission, Mr Prosser said that what it comes down to is that Chapter 2 does not apply to a scheme for paying cash bonuses to employees so as to escape a charge to income tax. Insofar as it is implicit in that submission that Chapter 2 does not apply to schemes that are directed at tax avoidance, he said there is no principled justification. As to whether the UBS scheme was one for paying money to employees, it is obvious that Chapter 2 does not apply to schemes for the payment of money. The question in this case is, however, whether what the employees got was money or securities. Mr Prosser recognised that what the scheme employees got was capable of being converted or realised into money, but so will many of the securities that will qualify as restricted securities for Chapter 2 purposes – they include debentures, loan stock, certificates of deposit and other instruments acknowledging indebtedness. So the fact that the NVS issued to the scheme employees could and would be redeemed for cash in the future does not prevent them being securities.

In Mr Prosser’s submission, the answer to the ‘securities or money’ question is a fact-sensitive one. It may be that, even if shares are allocated, they will carry rights which mean in practical terms that they can simply be regarded as a right to money – if, for example, they are redeemable immediately for a fixed sum. Or they might give a right to dividends dependent on the profitability of the company, and to a right of redemption in an amount similarly dependent. It might also be relevant to consider how quickly they can be redeemed. In the UBS case, the circumstances were closer to the latter type of case. The shares did not guarantee the employees a specific, predictable return on redemption. The redemption proceeds, and dividends payable in the meantime, would depend on the performance of the UBS shares held by ESIP: upon redemption, an appropriate number of UBS shares would be sold for their then market value. The employees’ securities therefore functioned just as shares normally do.

Then it is said by HMRC that, although some scheme members did not redeem their shares for over two years, the scheme should get no credit for that because that too was motivated by tax avoidance. But all that those scheme members were doing was taking advantage of the reduction in capital gains tax that was available if the shares were held for two years. That was legitimate tax mitigation, recognised by Parliament. Mr Prosser’s answer to the broad Ramsay argument was, therefore, that what has to be looked at is not the intention of the parties to the scheme, but the nature of the securities that were provided under it. If they were covered by Chapter 2, so was the scheme, even though tax avoidance was its driver. The FTT found that these were real shares, and so they were.

In the result, the UT were, said Mr Prosser, entirely right to say there was no intellectually coherent way in which HMRC could assert that the scheme was simply about the payment of cash bonuses. No doubt a fixed amount went in; but there was no certainty what would come out on redemption.

Mr Lasok’s alternative Ramsay argument was that if the NVS were securities, they were not ‘restricted securities’. That was said to be because there was no commercial rationale for subjecting them to the ‘transfer, reversion or forfeiture’ restriction. Mr Prosser accepted that the forfeiture event in the scheme had no commercial rationale, but did not accept that that meant that the scheme did not work. Do the ‘certain circumstances’ that are a condition of section 423 mean what they say, or does section 423 only have in mind ‘certain circumstances’ other than those included for tax avoidance purposes? Mr Prosser said there was no basis for any such distinction. He accepted there must be a real, genuine possibility of the stated circumstances occurring: if they were never going to happen in the real world, a purposive interpretation of section 423 would exclude them from its contemplation. In this case, the FTT found there was a genuine possibility of a forfeiture happening on the facts of the scheme. The ‘certain circumstances’ were therefore real ones, even though their inclusion in the scheme was tax motivated.

Mr Goy, for DB, adopted Mr Prosser’s submissions on the broad Ramsay submission in response to the like case levelled against DB on the scheme in that appeal. Whether Chapter 2 applies to the schemes is a question that can only be answered by a close analysis of what it means and what it provides for. He said the Mayes case is particularly important because it was one where the mere fact that something was done to avoid tax did not mean that it could be disregarded. In this case, what is required is to look at the schemes as a whole and to take a realistic view of the facts to see whether Chapter 2 applies to them. In DB, the FTT found, in paragraph 61, that the shares were ‘real shares’, that the holders received ‘actual dividends’ and that the shares were ‘securities’. On the sale of the shares by those domiciled in the United Kingdom, CGT was paid. How can the fact that the employees received such shares therefore be disregarded? Once one arrives at the correct conclusion that the case is one about the issue to the employees of securities, the next question is whether they are ‘restricted securities’ and whether the criteria of Chapter 2 are satisfied. It would, said Mr Goy, be odd to interpret Chapter 2 as excluding tax avoidance schemes, when tax avoidance is one of the purposes for which Chapter 2 exists.

In conclusion on HMRC’s broad and narrower Ramsay arguments, I would accept the arguments advanced on behalf of UBS and DB in preference to those advanced on behalf of HMRC. There is no dispute that the scheme was a tax-avoidance scheme. It was directed at avoiding the income tax consequences that would follow from the payment to employees of bonuses in cash. Instead, therefore, UBS rewarded their employees with shares, and say that such shares were therefore employment-related securities, indeed ‘restricted securities’, whose award fell squarely within the prescriptive provisions of Chapter 2; and, to the extent that Chapter 2 afforded exemption from income tax in relation to chargeable events affecting such shares, there is no reason why UBS is not entitled to it.

I agree with Mr Prosser and Mr Goy that the first question is whether, under the scheme, what was provided to the employees by UBS was money or shares. If on a fair analysis of the scheme, the correct answer is that it was money, there can be no question of Chapter 2 having any application. Moreover, I also agree with Mr Prosser that there might be cases in which, even though what was nominally awarded were shares, an objective interpretation of the true nature of the arrangements would justify the conclusion that in fact the employee was being paid money: for example, if the shares were required to be redeemed immediately for a pre-ordained cash sum.

In this case, however, there is in my view no question of the scheme being one for the payment of money, and I regard HMRC’s endeavours to suggest that it was as misconceived. The FTT found that the NVS were real shares, some of which were held by employees for more than two years, and real dividends were paid on them. Moreover, although the employees had the right to redeem their shares for cash over a period of two years, the redemption money was not pre-ordained, but its amount varied with the fortunes of the UBS shares held by ESIP, which could have risen or fallen. The shares were therefore real shares which functioned as such. It is true that the scheme could, in substance, be regarded as one in which a fixed amount of money was invested in employment-related shares for each scheme employee. But it was not a scheme under which the employee would, upon redeeming his shares, be entitled to a payment of the like sum. What he would receive was the redemption value of his shares, which might well be a very different sum. The scheme was therefore one for the provision of shares to the employees, not for the payment of money to them.

As what was provided to the employees was shares, there is in my view no scope for arguing that the shares were other than ‘securities … acquired in connection with an employment’ within the meaning of section 417(1) and 420 in Chapter 1. They were therefore ‘securities’ for the purposes of Chapter 2; and, provided that they satisfied the conditions of section 423, they were also ‘restricted securities’ for Chapter 2 purposes. I do not understand the argument that because, so it is said, the ‘transfer, reversion or forfeiture’ circumstances had no commercial rationale, the shares could not be restricted securities. That appears simply to be another way of saying that Chapter 2 does not apply to tax-avoidance schemes. But I do not follow on what basis it can be said that the ‘certain circumstances’ referred to in section 423(2)(a) can only be ‘circumstances’ included other than for tax avoidance purposes. There appears to me to be no justification for any such distinction. I of course accept, as did Mr Prosser, that there must be a real, genuine possibility of the stated circumstances occurring: if they were never going to happen in the real world, a purposive interpretation of section 423 would exclude them from its contemplation. In this case, however, the FTT found there was a genuine possibility of a forfeiture happening on the facts of the scheme. The ‘certain circumstances’ were therefore real ones, even though their inclusion in the scheme was tax motivated. If, therefore, as follows, the NVS are ‘restricted securities’ within the meaning of section 423, from where in Chapter 2 does one derive the conclusion that Chapter 2 as a whole nevertheless cannot apply simply because the scheme is driven by considerations of tax avoidance? As the UT said, Chapter 2 contains a detailed and prescriptive code for dealing with restricted securities. I agree with the UT that the NVS were restricted securities whose taxation fate was governed by that code. I would reject HMRC’s submissions that Ramsay principles, whether broad or narrow, require these genuine, employment-related NVS to be regarded other than as genuine ‘restricted securities’ within the meaning of Chapter 2.

I would dismiss HMRC’s appeal on the broad Ramsay ground and would refuse permission to appeal on the narrow Ramsay ground.

Mr Lasok’s next point, for which he also sought permission to appeal, was based on section 18 of ITEPA, quoted in [26] above. Before both tribunals below, HMRC argued that the application by UBS of the cash equivalent to the bonus allocated to a scheme employee in subscribing for NVS in ESIP marked the time when the employee became ‘entitled to payment … of the earnings’ within the meaning of Rule 2. That argument failed and was not repeated to us. The argument that Mr Lasok advanced to us in its place was a new one based on Rule 1. It was that, accepting there was no ‘entitlement’ under the scheme documentation for the purposes of Rule 2, when UBS applied the amount of bonus earlier allocated to each scheme employee in subscribing for NVS in ESIP, it was making a Rule 1 payment of the employee’s money earnings ‘on account’ of earnings that the employee would become entitled to when the shares were later allocated to him, and so at the point of subscription such earnings were treated as received by the employee. Mr Lasok specifically did not submit that at the moment of subscription UBS was making a ‘payment … of … earnings’ for the purposes of Rule 1.

Mr Prosser pointed out that section 18 of ITEPA is in Chapter 4 of Part 2, ‘Employment Income: Charge to Tax’. Chapter 2 is headed ‘Tax on Employment Income’ and section 6(1) provides that ‘The charge to tax on employment income under this Part is a charge to tax – (a) general earnings, …’. ‘General earnings’ are defined by section 7(3) as meaning ‘earnings within Chapter 1 of Part 3 …’. Section 62, in Chapter 1 of Part 3, defines earnings materially as follows:

This case concerns bonuses, not salary, wages or fees, although a bonus is a ‘profit’ within section 62(2)(b), but Mr Prosser said they all refer to what the employee obtains. He did not suggest that this means the employee must himself actually receive the money; it will, for example, be sufficient if it is paid to someone at his direction, or if he is in control of it. In this case, however, there can be no suggestion that when UBS paid its money to ESIP in subscribing for shares, it was doing so at the direction of the employees who had agreed to take part in the scheme, or that the employees had any control over the money so paid. Moreover, when UBS asked the employees whether they would like to take part in the scheme, it expressly reserved the right not to go ahead with the scheme.

In particular, as for 416 of the employees who agreed to take part in the scheme, they had no contractual right to any bonus and could not have acquired any such right when UBS subscribed for shares in ESIP. The other 10 scheme applicants had a right to a bonus of a specified amount if they were still in employment in February 2004 and no notice to terminate their employment had been given or received. The section 18, Rule 2, argument succeeded in relation to the 10 in the FTT, but failed in the UT and was not repeated to us. Mr Prosser accepted, as I understood it, that as regards the 10, if they had received nothing by the February 2004 pay date, they would have been entitled to shares when they were subscribed for, because that is what they had signed up for. They would not, however, have been entitled to cash. In the event, the February pay date post-dated the declaration of trust in respect of the NVS.

As regards the 10, it appears to me that they had a contractual right to a particular cash bonus payment in February 2004. In the meantime, however, they agreed that their right would be satisfied by an award of shares under the scheme. That happened in January 2004, when UBS made its declaration of trust in respect of the NVS. I regard that as the beginning and the end of the story as regards the 10 for the purposes of the present submission, subject only to Mr Lasok’s point that when UBS paid the money to ESIP by way of subscription for the NVS, it was making an ‘on account’ payment of earnings to the 10 – as well of course as to the 416.

Mr Prosser accepted that when the declaration of trust was made, all scheme employees became entitled to ‘earnings’ within the meaning of section 62. He said, however, that section 18, Rules 1 and 2, were not in point, because they apply exclusively to ‘earnings consisting of money’, whereas the earnings that the scheme employees received were not money: they were non-money earnings of a money’s worth nature, and so were earnings to which section 19 applied; and section 19(4) explains that they were treated as received by the scheme members when ‘the benefit is provided’, which can only mean when the declaration of trust in relation to the NVS was made. Mr Prosser accepted that, in principle, such a benefit is taxable, the basic rule being that it is taxable on the amount that the scheme members could have got for their money’s worth. But the answer to that charge on the present facts is that the shares were ‘restricted securities’ within Chapter 2 of Part 7 and, for reasons explained, escaped a charge to income tax altogether.

In short, Mr Prosser’s answer to Mr Lasok’s section 18, Rule 1, submission is that the only relevant earnings in this case were not ‘money’ earnings at all, but were ‘money’s worth’ earnings, to which Rule 1 has no application. He also disclaimed any understanding of why HMRC favoured the ‘on account’ limb of Rule 1 rather than the ‘payment is made of … earnings’ limb, although I suspect that the answer is that the latter alternative is more obviously misconceived than the former. He also said that in any event HMRC ought not to be entitled to run the new Rule 1 point at the stage of this second appeal. There was, he said, no evidence adduced before the FTT as to whether, when it made the subscription payment, UBS was intending in any manner to discharge its liability to the scheme employees. It is also too late to remit the matter to the FTT for further evidence. HMRC are entitled to only one bite of the cherry.

The same point under Rule 1 of section 18 was also raised in HMRC’s respondent’s notice in the DB appeal, although that appeal is shorn of the minor complication that might be said to arise in respect of the UBS 10. Mr Goy adopted and supported Mr Prosser’s submissions.

In my view, there is nothing in HMRC’s section 18, Rule 1 argument. I have held in relation to the Ramsay arguments that the employees were rewarded with real shares, not money, and so I agree with Mr Prosser and Mr Goy that section 18 is irrelevant since it applies only to money earnings. In reply, Mr Lasok said that no-one seemed to have understood that he was advancing his Rule 1 argument on the assumption that he was correct on his broad Ramsay argument and that therefore the employees were to be regarded as receiving money, not shares. I admit to being one of those who had failed to understand that. However, I am satisfied that what the scheme employees received was shares, not money, and so I regard section 18 as inapplicable. I would refuse HMRC permission to appeal on this ground.

Mr Lasok advanced various submissions to the effect that the scheme anyway did not satisfy the requirements of Chapter 2 and so was not entitled to its tax exemption benefits. The first submission arose in relation to section 423, which defines ‘restricted securities’. For the shares so to qualify, they had to satisfy section 423(2)(c), which requires a comparison of two amounts. The first is what the employee is entitled to receive on a transfer, reversion or forfeiture (‘the receivable amount’). The second is the hypothetical market value of the shares at the time of the transfer, reversion or forfeiture as if there were no provision for such (‘the market value’). Mr Lasok’s argument required a focus on the bracketed words in section 423(2)(c), reading ‘(determined as if there were no provision for transfer, reversion or forfeiture) …’. Section 423(2)(c) shows that the receivable amount has to be lower than the market value, otherwise section 423(2) will not be satisfied: more accurately, it is a condition that its recipient ‘will not be entitled to [a receivable] amount of at least [the securities’] market value’. Under ESIP’s M&A, the forfeiture price was 90% of the market value estimated to be obtainable for the shares if no restrictions applied.

I should refer to certain provisions of ESIP’s M&A. Article 1.1 provided that:

The essence of Mr Lasok’s submission was that the sense of the bracketed words is that the market value to be determined for the purpose of the second calculation must be assessed on the basis that not only is there no provision for a transfer, reversion or forfeiture, but also that there is no hedging provision such as in the UBS scheme, since that provision was said to be inextricably bound up with the transfer, reversion or forfeiture provision. The effect of the hedging arrangements was that their implementation would result in the net assets of ESIP increasing by about 10%. The intention behind them was that although the employees would, upon a triggering event, be required to sell their NVS for 90% of their market value, they would not be materially worse off because the market value of the NVS would be equal to approximately 110% of the original subscription price.

The way that might translate into practical considerations was the subject of findings by the FTT. They found that the employees would be likely to receive 99.2% of their bonus entitlements in respect of the NVS if the trigger event occurred, the NVS were forfeited and the payment to ESIP under the hedging arrangements was made. The FTT also found that there was a possibility that, because of the hedging arrangements, the employees might even receive more than 100% of their bonus entitlements. By contrast, the market value of the NVS on the same date if the NVS were unrestricted, and there were no hedging arrangements, would be likely to be less than 100% of the bonus entitlements. Thus, on the latter facts, the market value was likely to be less than 100% of the bonus entitlements, whilst the possible receivable amount on the occasion of a transfer, reversion or forfeiture ranged from 99.2% to over 100% of the bonus entitlements. As a result, on the facts, and if the hedging arrangements were left out of account in the market value calculation, there was a risk that the receivable amount would be greater than the market value. Even if the lowest value for the receivable amount were to be taken, the difference between the receivable amount and the market value would be likely to be so small that it could be ignored as de minimis. It follows, said Mr Lasok, that section 423(2)(c) is not satisfied because it could not be said at the outset that the receivable amount ‘will not be’ at least the market value.

Mr Lasok supported his submission that in calculating the market value the hedging arrangements must also be left out of account by citing East End Dwellings Co v. Finsbury Borough Council [1952] AC 109, in which Lord Asquith said, at 132:

Second, he said the purpose of section 423(2)(c) was to reflect the depressing effect on the market value which would be caused by the restriction and to ensure that, if the employee was only entitled to receive the depressed value of the securities, they would still be ‘restricted securities’. The purpose of the hedging arrangements was to negate such depressing effect and to ensure that the employees did not in reality receive the depressed market value of their shares. So the hedging arrangements rectified the very problem that section 423(2)(c) was intended to address.

Third, he said that the language of section 423(2)(c) did not require only that the forfeiture provision be ignored in calculating the market value. That might perhaps, read in isolation, be regarded as the limit of the sense of the bracketed phrase. But the hedging provisions were, for the purposes of section 423(1), part of the ‘contract, agreement, arrangement or condition’ that made ‘provision’ to which any of subsections (2) to (4) applied, and it was therefore appropriate to look at all parts of the contract etc to decide what ‘provision’ was made; and, in particular, in construing section 423(2)(c) to have regard to the hedging arrangements in considering whether the condition there prescribed was satisfied.

If Mr Lasok was correct on this submission, it would follow that the shares were not ‘restricted securities’ for the purpose of Chapter 2 with the consequence that the scheme could not enjoy the benefits of Chapter 2.

The UT disagreed with the FTT’s approach, and also with Mr Lasok’s submission. They held that what section 423(2)(c) requires is to compare (a) what the employees will get on a forced sale with (b) the market value at that time ignoring the restriction. Mr Prosser submitted that the UT was correct.

The UT gave elaborate reasons for their conclusion that, contrary to the view of the FTT, the section 423(2)(c) condition was here satisfied; and Mr Prosser also advanced full reasons in support of that conclusion, which were in their essentials as follows. ‘Market value’ in section 423(2)(c) is explained in section 421, which gives it the meaning it has for the purposes of the Taxation of Chargeable Gains Act 1992 by virtue of Part 8 of that Act: that is ‘the price which those assets might reasonably be expected to fetch on a sale in the open market.’ In a case where, as here, the hypothetical sale would be subject to the restrictions, the definition would ordinarily require an assumption that the purchaser would take subject to the restrictions, which would have a depressing effect on the price the purchaser would pay and thus on the market value. The purpose of the bracketed words in section 423(2)(c) was to negate the depressing effect on the market value which would be caused by making the assumption that the purchaser would take subject to the restrictions, and so ensure that, if the employee were to receive only the reduced market value of the securities, they would remain restricted securities for the purpose of Chapter 2. In other words, in order to ensure that the securities are ‘restricted securities’, if the amount the employee will be entitled to receive on a forced sale will be less than their unrestricted market value, an adjustment has to be made to ‘market value’ as defined; and it is the bracketed words that perform that function.

I would accept that argument as a correct explanation of the function of section 423(2)(c). As to what the bracketed words mean, I regard it as plain that their ordinary and natural sense is that they require no more than that ‘the provision for transfer, reversion or forfeiture’ is to be ignored. There is no warrant for interpreting the words as if they extended also to the hedging arrangements, which as a matter of language they do not; and, as the UT said, those arrangements were anyway purely collateral and served a different purpose of ensuring that the employees would not end up significantly out of pocket if a forced sale occurred. The only purpose of the bracketed words is to negate the depressing effect of the assumption, implicit in the definition of ‘market value’, that the purchaser will take subject to the restriction. Whatever the notional market value of the securities, the employees still only get 90% of it in the event of a forced sale. As the UT said, and I agree, the conclusion that they would not be entitled to receive at least the market value of their shares is inescapable.

I would dismiss this ground of HMRC’s appeal.

Mr Lasok’s point here was that the scheme shares did not satisfy the requirements of section 429. They had to do so if the claimed exemption was to be justified, since the shares ceased to be ‘restricted securities’ on 19 February 2004, when the restricted period ended and (subject to section 429) became subject to a charge to tax under sections 426 to 428. The FTT held that section 429 would have applied if, contrary to their prior holding, the shares were ‘restricted securities’. The UT also held that section 429 applied.

As the UT noted in [113], section 429 applies if four conditions are satisfied. It is agreed that the following three are. First, that the NVS must be ‘shares … in a company of a class’ (section 429(1)(a)). Second, the transfer, reversion or forfeiture provision must apply to all the shares of the class (section 429(1)(b)). Third, all the shares of the class (other than the employment-related securities) must be ‘affected by an event similar to that which is a chargeable event’ (section 429(1)(c) and (2)(a)).

The fourth, and controversial, condition is that, immediately before the termination of the forced sale provision, the majority of the shares of the class were not held by or for the benefit of certain specified persons, including in particular ‘associated companies of’ ESIP or ‘employees of any associated company’ of ESIP (section 429(1)(d) and (4)(c) and (d)). The critical question is whether UBS was an ‘associated company’ of ESIP on 19 February 2004. If it was, section 429(4) was not satisfied.

Section 432(6) explains that, in Chapter 2, ‘associated company’ has the meaning indicated in section 421H, which in turn explains that ‘associated company’ has the same meaning as, by virtue of section 416 of ICTA, it has for the purposes of Part 11 of ICTA. Section 416 provides:

HMRC’s submission was that Mourant, as trustee of Sidemore, held the controlling VOS in ESIP; and that UBS was therefore able to exercise indirect control over ESIP through its control over Mourant. Whilst HMRC accepts that ‘control’ in section 416(2) means control at shareholder level (see Steele (Inspector of Taxes) v. EVC International NV (formerly European Vinyls Corp (Holdings) BV) [1996] STC 785, at 794, per Morritt LJ), Mr Lasok submitted that that did not preclude the conclusion that relevant control was in UBS.

The proposition that ESIP did what it was told by UBS was advanced to and rejected by the FTT, which concluded that the evidence did not show that UBS and ESIP were associated companies. I interpret them as having regarded ESIP as independent. The FTT noted in paragraph 120 of their reasons that:

The UT recorded that it had not been said that the FTT had ‘expressly misdirected itself’ on the law; and the FTT had heard evidence from witnesses of fact, including from Mr Ferrara, who had supported the view that ESIP was independent of UBS. Nor, before the UT, did HMRC refer to the transcripts of the oral evidence or the witness statements. The UT therefore wondered, at [127], how in those circumstances they could conclude that the FTT had erred in law in holding that UBS did not within the relevant one-year period satisfy the section 416(2) test of control, at shareholder level, of the affairs of ESIP.

Mr Lasok, in his criticism of the FTT’s approach, focused on the word ‘expressly’ included in the opening sentence of my preceding paragraph and said that the FTT did err in their apparent conclusion that section 416 control required a ‘necessary degree of compulsion’. In fact, the FTT said no such thing. They did, however, use such language in their separate decision in the like ‘control’ issue in the DB case, delivered five months later. As I shall explain, the UT considered in the DB case that the FTT had misdirected themselves in adopting such a ‘compulsion’ test and concluded that, on the facts (and contrary to the decision of the FTT), the relevant ‘control’ did exist. Both limbs of that approach by the UT are challenged by DB in its own appeal. Mr Lasok nevertheless submitted that the inference must be that the FTT had implicitly made the same self-misdirection in the UBS case, with the consequence that their ‘control’ decision in UBS was similarly flawed.

As for the UT’s observation that there had been no evidence from any Mourant witnesses, Mr Lasok explained that this was not because it had not been sought by UBS, but because one potential Mourant witness refused to give evidence altogether, and two others were prepared only to provide witness statements and refused to give oral evidence.

To understand the nature of Mr Lasok’s submissions under this head, I should first set out the following paragraphs of the UT’s decision:

Mr Lasok said the fact that the material may have been selective was irrelevant if there was no other material inviting a different inference. The fact that much of the material may have related to Mourant’s activities at board level was also irrelevant. He said the evidence that the UT summarised in [128] and [129] showed that UBS was directing Mourant how to exercise its shareholder control, and he sought to make that good in his oral submissions by referring us to, inter alia, minutes of an ESIP board meeting held on 28 January 2004, although it was not apparent to me how that showed that Mourant was acting in accordance with instructions from UBS.

The UT rejected the argument that the FTT had made an error of law in concluding that UBS exercised no relevant control over Mourant. They said:

Mr Prosser acknowledged that there were no Mourant witnesses before the FTT, but said there was evidence from UBS witnesses who had set the scheme up and had held meetings and discussions with the Mourant people, including evidence as to whether Mourant was simply doing what it was told. Those witnesses were cross-examined at length. HMRC’s closing case was that the evidence showed that UBS was exercising relevant control. The FTT rejected that case. HMRC’s stance before the UT, and now this court, was a different one, to the effect that in order to decide this question it was necessary merely to look at the documentary evidence, evidence which the UT had described as selective: and Mr Lasok had specifically, but perhaps understandably, not sought to support his case by reference to the oral evidence adduced before the FTT. The UT also rejected the case after merely being shown a selection of documents. This court is now being asked to say that the UT was wrong in law not to find that the FTT erred in law in its conclusion.

In my judgment, Mr Lasok’s proposition that the FTT impliedly misdirected themselves in law because they were tacitly applying a test for control that had to be underpinned by a ‘necessary degree of compulsion’ is mistaken. First, there is no warrant from anything that the FTT said in the UBS decision that they were applying any sort of ‘compulsion’ test. Secondly, as I shall explain when I come to the DB appeal, I consider, with respect, that the UT misunderstood the sense in which the FTT used the ‘compulsion’ phrase in the DB decision, and that Mr Lasok was, in consequence, seeking to build his castle on a foundation that is, in this respect, based on sand. In my judgment, the FTT made no self-misdirection in their disposition of the control issue.

There remains HMRC’s proposition that this court should find, contrary to the finding of fact by the FTT after hearing oral evidence on the issue, that the FTT erred in law in failing to hold that Mourant was merely dancing to UBS’s tune in its exercise of control at shareholder level. I regard the proposition as hopeless. What UBS wanted was clear enough. It was, however, in no position to dictate to Mourant to do its bidding and the FTT found that it did not and that Mourant exercised its shareholder powers in respect of ESIP independently. I would dismiss this ground of HMRC’s appeal.

This was Mr Lasok’s final point as to whether the scheme satisfied the technical requirements of Chapter 2. I summarised the nature and purpose of article 2(15) in [16] above. It is agreed that, were article 2(15) to be disregarded, UBS would have had control over ESIP for the purposes of section 416(2) of ICTA. Mr Lasok’s submission was that the provision was artificial and so should be disregarded. Article 2(15) provided:

Mr Lasok said that article 2(15) had no basis in commercial reality. It stripped the NVS of key rights whilst they were held by UBS. That carried with it the risk that UBS could thereby lose £60m, a risk it admitted would be commercially unacceptable if it happened, but which it was content to run because it considered the prospect of article 2(15) actually being implemented as so small. UBS’s explanation of this was that the risk applied only during the brief window during which UBS held the NVS shares, its holding of VOS was anyway sufficient to prevent a special resolution in Jersey to wind ESIP up and there was only a negligible prospect of a winding up on other grounds.

Mr Lasok’s submission to us, as to the UT, was that the courts will disregard artificial provisions in contracts that are directed at circumventing legislative provisions and which disguise the nature of the true arrangement between the parties. He invoked Street v. Mountford [1985] AC 809 and Antoniades v. Villiers [1990] 1 AC 417.

The problem with this argument is that, unlike the arrangements under scrutiny in those cases, the FTT found article 2(15) to be a genuine provision, in the sense that it represented the true arrangement to which UBS intended to submit, even though it carried the minuscule commercial risk referred to. The language of article 2(15) did not, therefore, contradict the reality of the parties’ intentions. The UT also rejected HMRC’s ‘sham’ argument, and although in [149] to [151] they placed a slightly different focus on it, they also rejected the argument as so focused. In short, they saw no basis for construing article 2(15) as meaning other than what it said. Mr Lasok repeated to us the essence of the arguments that he had advanced to the UT. Article 2(15) was, he said, a pretence included for tax avoidance purposes and should be disregarded.

This argument is also hopeless. There was no finding that the participants did not intend article 2(15) to take effect according to its terms and so the argument that it was a sham was rightly rejected. Once this point is reached, I am at a loss to know on what basis article 2(15) can be airbrushed out of the scheme.

I would also dismiss HMRC’s appeal on this ground.

I would refuse permission to HMRC to appeal on the two grounds for which it sought permission, and dismiss their appeal on the other four grounds. I would, therefore, dismiss HMRC’s appeal against the decision of the UT.

In what follows I have gratefully drawn from the judgment of the UT. The DB group was headed by Deutsche Bank AG. DB was the main employer in the UK of staff working for the group. DB had a policy of paying certain employees annual bonuses in various forms, including cash bonuses. For the tax year 2003/2004, DB implemented a tax avoidance scheme that had been devised by Deloitte and Touche LLP (‘Deloitte’). Deloitte was involved at all stages of its implementation. The scheme was considered and accepted by DB’s compensation committee in December 2003. Like the UBS scheme, it sought to take advantage of Chapter 2 in Part 7 of ITEPA. In either December 2003 or January 2004, DB’s employees were informed of the scheme and were offered the opportunity to be considered for participation in it. Those interested had to complete and return an application form by 26 January 2004. The minimum level of participation was £50,000 and the maximum was 90% of any discretionary bonus that might be awarded. The one-page summary of the scheme stated that it was one ‘that allows your non DB equity-based year-end discretionary award to be delivered to you in the form of EDSA shares’ (the scheme was then known as ‘the EDSA plan’).

On an uncertain date in late January 2004, the DB compensation committee decided on the bonus sums to be allocated to named employees. By 26 January 2004, the closing date for applications, some 300 employees had returned the necessary forms. On 27 January, Investec Bank (UK) Limited (‘Investec’), whose role broadly corresponded to that of Mourant in the UBS scheme – and which the FTT found to be independent of DB – incorporated a company in the Cayman Islands called Dark Blue Investments Limited (‘Dark Blue’), whose role broadly corresponded with that of ESIP in the UBS scheme. On 27 January, a Guernsey company, Walbrook Nominees (No 6) Limited (‘Walbrook’), agreed with DB to act as nominee for the scheme. On 28 January, Investec appointed two directors to Dark Blue.

On 29 January, DB informed Walbrook of the names of the employees for whom shares were to be beneficially held under the scheme, and told Walbrook that it would shortly receive 91,300 C1 shares in Dark Blue to be held for those employees (they were to be the ‘restricted shares’). Walbrook then executed a declaration of trust. On 29 January, Deloitte notified DB that the final funding figure for Dark Blue would be £91.3m.

On 2 February, Dark Blue’s Memorandum and Articles of Association (‘DB M&A’) were amended by special resolution. The new share structure was as follows:

The effect of article 33 of the DB M&A was that the C1 and C2 shares could not be transferred between 7 February and 1 April 2004 inclusive. The C1 shares were intended to be ‘restricted securities’ for the purposes of Chapter 2 of Part 7. The UT described the nature of the restriction as simpler, and less artificial, than in the UBS scheme: article 34 provided that if, before 2 April 2004, any individual who held, or was beneficially entitled to, C1 shares ceased to be employed by any DB company, or notice of termination was given to an employee (‘the Terminating Employee’) for any reason other than redundancy, death, disability or without cause, then:

On 2 February 2004, the following events also occurred:

On 5 February, DB subscribed for 38,042 ordinary shares in Dark Blue for the subscription price of £13,314.70 and appointed John Mooney a director of Dark Blue. On the same day, DB entered into a Shareholders’ Agreement with Investec under which it was agreed, inter alia, that:

On 6 February, the arrangements set out in the shareholders’ agreement were completed. DB paid £92.6m to Investec and Investec transferred the C1 shares to Walbrook. On the same day, Walbrook informed the relevant employees of their award of C1 shares and that it held them on bare trust for them as their nominee. The restriction on the transfer of the C1 shares was lifted on 2 April 2004; and 42% of the C1 share capital was redeemed at the first opportunity on 8 July 2004, at a price of £1,003.73 per share. The opportunities for regular redemption continued until December 2006, and all the employees had redeemed their shares by 8 December 2006, with only five of them redeeming them on that date.

I summarised in [37] above the three main questions that arose before the FTT. They answered question (1) in favour of DB. In answer to question (2), they concluded that the shares were ‘restricted securities’ and that, on a technical analysis, the section 429 exemption was available. So far so good for DB, but it then fell at the question (3) fence, in respect of which the FTT held that the Ramsay argument succeeded and so the scheme failed.

The UT disagreed with the FTT on the Ramsay argument but still dismissed the appeal since they took the view that the FTT were in error on the section 429 control argument.

I shall deal first with DB’s appeal against the UT’s decision on the ‘control’ issue, and then with the issues raised by HMRC’s respondent’s notice seeking to support the UT’s decision on other grounds.

The question is whether Dark Blue was ever an associated company of DB. If it was, section 429(4) of Chapter 2 was not satisfied, and a charge to tax would have arisen on the removal of the restrictions affecting the C1 shares. More particularly, the question is whether DB had control over the affairs of Dark Blue in the sense of the opening words of section 416(2) of ICTA, meaning control over its affairs at the level of general meetings. The FTT found that DB did not so control Dark Blue. The UT concluded that in this respect the FTT erred in law.

The UT explained, at [201] and [202], what it called the complex shareholding structure of Dark Blue. It is sufficient to quote [202]:

The UT said the FTT were right to find that DB at no time had control of Dark Blue in the formal sense of being a controlling shareholder. The UT also agreed that the FTT were right to say that the ‘carefully engineered share structure’ of Dark Blue was at the heart of the ‘control’ argument. But the UT said that was not the end of the matter: it was necessary also to consider whether DB in fact exercised control at shareholder level over Dark Blue, despite the fact that it was only a minority shareholder.

In dealing with this issue, the UT quoted paragraphs 90 to 97 of the FTT’s decision. I shall do likewise:

The UT then explained why they considered that the FTT had erred in law. They said that, in referring to ‘the necessary degree of compulsion’ in paragraph 96, the FTT had set the bar for the test of control significantly too high. I should also set out the remainder of the UT’s reasoning, since it was the focus of the arguments before us:

The sole question, said Mr Goy, is whether DB controlled Dark Blue at shareholder level; and the answer lies in whether the case falls within the type of control identified in the opening words of section 416(2) of ICTA. Mr Goy referred us to Steele’s case, supra, in which Morritt LJ said, at 794:

Mr Goy referred us first to what the FTT said at paragraphs 90 to 94, quoted in [124] above. He said the UT made no criticism of those findings. They showed that, after hearing the evidence, the FTT concluded that Investec and Dark Blue were guided closely about what they had to do and when they had to do it, but the FTT then asked themselves whether what Dark Blue did went beyond mere negotiation and recommendation so that DB alone made all the decisions. In paragraph 94, they answered that question in the negative: whilst there was close co-operation and co-ordination, there was not control.

The FTT then, in paragraphs 95 and 96, considered the involvement of Investec. It was as to this, the relationship between DB and Investec, that the UT said the FTT had, in referring to ‘the necessary degree of compulsion’, set the bar too high. Mr Goy did not suggest that, for there to be relevant control, it is necessary to have a legal right of compulsion, although he did submit that it was necessary at least to have some form of ability to require someone to do something. At the very least, he said, it will be necessary for the case to be one in which someone simply goes along with the wishes of another and does not exercise any independent thought as to whether what is to be done is in its own best interests. But that, by itself, will still not be enough, nor will mere co-operation or co-ordination be enough. Mr Goy submitted that when the FTT referred in its paragraph 96 to ‘the necessary degree of compulsion as between DB and [Dark Blue]’, they were simply attempting to identify that elusive ‘something more’ that is required in addition to merely going on with the wishes of another. The FTT were not saying that there will only be control if compliance can be enforced; and, if that hadbeen the FTT’s view, their discussion as to a practical test on the facts at paragraphs 90 to 94 made little sense. Moreover, the phrase ‘degree of compulsion’ shows that they did not in fact mean that compulsive powers were necessary. There is either a power of compulsion or there is not; there are not degrees of compulsive powers. What the FTT said was that the evidence did not show that DB’s powers went ‘far beyond’ those of a commercial entity dealing with another commercial entity such that, in reality, Investec, and therefore Dark Blue, was not in control of its own decisions.

Mr Goy emphasised that the control in question is that as to the casting of votes at a general meeting. There may well in many cases be circumstances in which A will expect shareholder B to vote in a particular way, and B does so; but there may also come a point when B will say that it is not prepared so to vote in relation to a particular matter and will point out that it is not a nominee for A and must consider its own interests. Investec, said Mr Goy, was no stooge for anybody. It of course wanted its fee and would vote in a way which was in its own interests to vote. It would not, however, have voted in a particular way because DB asked it to. Insofar as the UT regarded the shareholders’ agreement as relevant, it was not: it had nothing to do with the casting of votes at general meetings. The issue is whether, in relation to the decisions of Dark Blue in general meeting, DB could be said to have done more than proposed, advised or influenced. The FTT’s answer on the facts was what it said in paragraphs 93 and 94, namely that whilst the evidence showed close co-ordination, it did not, in the section 416 sense, show control. The FTT so found after hearing all the evidence and knowing full well that it was concerned with the implementation of a scheme in respect of which Deloitte had planned that certain steps would take place at certain times. They accepted that Dark Blue and Investec were closely guided, but concluded that such guidance did not go beyond negotiation and recommendation.

In the alternative, Mr Goy submitted that, if this court were to conclude that the FTT had misdirected themselves with regard to ‘the necessary degree of compulsion’, the UT nevertheless then also misdirected themselves by holding that the only conclusion open to the FTT was that DB controlled Dark Blue. The UT, in [205], identified the two central questions, (a) and (b), that the FTT should have asked themselves. Mr Goy submitted that the findings of the FTT showed that neither question could be answered in the affirmative. He referred us to the FTT’s findings at paragraph 42, where, looking at the position before the shareholders’ agreement was entered into, the FTT found that while ‘there was no direct agreement in place in any formal sense between DB and Investec, clearly some arrangement was in place. The order of events was not coincidental; it was a deliberate order of events.’

Mr Goy did not deny that, as the FTT found, there must have been some arrangement in place between DB and Investec. It did not, however, follow that DB controlled what Investec did as regards Dark Blue. Investec had no rights against DB, and it invested £91.3m in subscribing for C1 shares in Dark Blue on 2 February 2004 before it entered into the shareholders’ agreement with DB on 5 February. Mr Goy said the notion that, when it invested in Dark Blue, Investec was simply responding to a controlling direction from DB flew in the face of reality: it had no agreement with DB, and it could rationally only be regarded as doing what it did by consulting its own interests in doing so. In that context, Mr Goy recognised that Investec saw the prospect of earning a fee and would have been willing to implement the arrangement as it thought fit, and in its own interests. But it simply does not follow that in doing so, it was acting unthinkingly in accordance with DB’s wishes. Both DB and Investec were substantial, independent companies; and whilst they were together working towards a common goal, the suggestion that one was controlling the decisions of the other makes no commercial sense. Moreover, it is clear from the FTT’s findings in paragraphs 90 to 94 that DB had no such control over Investec. Pre-ordained it may all have been; but that did not mean that Investec was in the pocket of DB. The UT did not themselves hear the evidence, and were in no position to second-guess the FTT on this point. In particular, it was never put to Mr Beddows (of Investec) that Investec would simply do DB’s bidding.

Mr Goy also criticised the UT’s statement in [207] that Investec’s entry into the shareholders’ agreement was ‘dictated to it by DB’. Mr Goy said that proposition was contrary to common sense. There was no finding to that effect by the FTT, nor had the UT heard any evidence enabling it to substitute their own views and make such a finding. Investec entered into that agreement at a time when it had no binding commitment with DB, and it is obvious that its decision to do so would have been one made in its own interests. Nor is there anything in the agreement relating to the exercise of Investec by its voting rights at a general meeting of Dark Blue. In [208], the UT said that the FTT had ‘in effect found’ Investec to be a mere cipher. The FTT had, however, not said that, either ‘in effect’ or at all.

Finally, Mr Goy submitted that the UT fell into error in relation to section 416(3). The UT’s view, in [205], appears to have been that if DB controlled Investec in relation to the latter’s voting at general meetings of Dark Blue, then ‘the test of control in section 416(3) was satisfied’. Mr Goy submitted, and I agree, that section 416(3) is irrelevant to our facts. Investec, by virtue of its holding in Dark Blue, was formally in control of Dark Blue within the meaning of section 416(1). The relevant question for our purposes, however, is whether Investec was, vis-à-vis its role as a Dark Blue shareholder, itself in the control of DB. If it was, the opening words of section 416(2) are satisfied and DB and Dark Blue are associated companies. If it was not, those words are not satisfied. I do not understand what section 416(3) is thought to have to do with it. It is contemplating a case in which, for example, DB and X Ltd together had relevant control of Dark Blue via their combined control of Investec. On no footing is that this case. Mr Goy did not, however, go so far as to say that the UT’s error in relation to section 416(3) undermined their conclusions on the ‘control’ point to which they had otherwise anyway come.

Mr Lasok rightly did not seek to defend what the UT had said about section 416(3). He opened his substantive response by suggesting that in a simple case where A and B agree that B will do what A wants in furtherance of a tax avoidance scheme, and B then acts as a majority shareholder in C Ltd in furtherance of the scheme, A controls C for the purposes of section 416. That, he said, is essentially the scenario with which we are concerned. As to whether the UT were right to come to the conclusion they did, that requires a consideration of the decision of the FTT. Insofar as Mr Goy submitted that Investec was following recommendations, and that DB’s role did not go beyond negotiation and recommendation, Mr Lasok said the FTT made no such findings; Mr Goy’s assertions were, he said, no more than inferences from the FTT’s findings.

As to the assertion that there was no agreement between Investec and DB before the shareholders’ agreement, Mr Lasok referred to the FTT’s finding in paragraph 41 that ‘the order of events as between DB and Investec took place in accordance with coordinated plans’ and to their observation that ‘it must ask why would Investec set up [Dark Blue] in the way it did unless there was some arrangement in place as to Investec’s role in the Scheme.’ And, in paragraph 42, the FTT wrote that:

In Mr Lasok’s submission the argument on the ‘control’ issue was not capable of much elaboration. He said it was simply a question of looking at the FTT’s decision to see whether their findings support the conclusion that the control test was satisfied. Of course, the FTT found that it was not, but the implication is that the UT were entitled to conclude, on the basis of the FTT’s findings, that it was. Mr Lasok referred us, without reading them, to several paragraphs of the FTT’s decision, which he said supported the ‘control’ conclusion. I do not propose to refer to them. Their thrust is directed to the pre-ordained, regimented nature of the scheme, as to which there is no dispute.

In my judgment, and with respect, the UT went wrong on this issue. First, I consider that there was no warrant for the UT’s conclusion that the FTT, in their paragraph 96 reference to a requirement of the ‘necessary degree of compulsion’, were raising the bar too high. The UT interpreted those words as meaning an ability to ‘enforce’ compliance with (in this case) DB’s views. But the FTT plainly cannot have meant that. It is obvious that DB could not compel Investec to vote in any particular way; and if the FTT did regard the test as requiring an element of compulsion, they could have considerably reduced their discussion by simply making a finding that such element was missing.

The FTT did not, however, do that. They carefully examined the facts to see whether the case was one in which DB had relevant control and concluded, in summary, that whilst the evidence showed close co-ordination, it did not ‘in the section 416 sense, show control’. In my view, Mr Goy was right that in the phrase ‘the necessary degree of compulsion’ the FTT was doing no more than attempting, perhaps not with an ideal choice of phrase, to identify the elusive extra ingredient that needs to be found before A Ltd can be said to be in control of voting powers of a majority shareholder of B Ltd. Ultimately, whether there is such control is a fact-sensitive matter; and here the FTT found against HMRC’s control case on the facts.

I consider therefore that the UT were wrong to hold that it was open to them to consider afresh whether DB controlled Investec’s voting powers as a shareholder of Dark Blue. That was a question of fact for the FTT, and the FTT had given rational reasons for their answer to that question. Moreover, and with respect, I regard the UT’s conclusion that the only answer to the control question was that DB was relevantly in control of Investec a remarkable one, which I regard as obviously wrong. Yes, the scheme was pre-ordained and involved a co-ordinated course of action between the participants, with Investec and DB, two wholly independent companies, playing pre-ordained and co-ordinated roles, with each having its own commercial interests in bringing the scheme to fruition. It does not, however, begin to follow from this that DB was in relevant control of Investec. If A Ltd proposes to B Ltd, an unconnected and independent company, a co-ordinated course of action with a view to achieving a commercial end to the benefit of both, and B Ltd agrees to the proposal and co-operates in its implementation, it is beyond my comprehension why such a state of affairs should be thought to justify the inference that, in playing its own part in the operation, B Ltd is to be regarded as being ‘controlled’ in what it does by A Ltd. The proposition is wrong. B Ltd will, by inference, want to take part, and will do so. But there will ordinarily be no basis for an inference that the decisions it makes en route to the ultimate goal will be decisions it makes other than independently, and in its own interests, in achieving the proposed end.

The FTT understood this. The UT appear not to have done so. In my judgment, the UT fell into error in setting aside this part of the FTT’s reasoning. I would allow DB’s appeal on the ‘control’ issue. I turn to HMRC’s points on their respondents’ notice.

HMRC’s argument under this head was the same as in the UBS appeal. There were no separate features relating to DB that require additional comment. I would reject the argument for the same reasons that I consider it fails in the UBS appeal.

This is a semi-new point. It was raised in HMRC’s skeleton argument before the FTT. It was not, however, argued before the UT, where HMRC apparently confined themselves to seeking to uphold the FTT’s acceptance of their broad Ramsay argument. Mr Lasok said, however, and I agree, that for HMRC to raise it before this court will cause no prejudice to DB, as it is a question of law turning on decided facts. It is, in substance, the like alternative line of argument as that advanced in the UBS case, namely that the shares issued to the DB employees were no more ‘restricted securities’ than HMRC said the shares in the UBS scheme were. It turns, however, on the terms of the restrictions in the DB scheme. Mr Lasok advanced two reasons in support of the argument.

Article 34(a) of the DB M&A, in a part headed ‘Compulsory Transfer of C1 and C2 Shares’, provided as follows:

Mr Lasok said the only circumstances in which the triggering of the clause 34(a) condition might occur were, for practical purposes, where the employee chooses to hand in his own notice or is dismissed for misconduct, in either case within a window of about six weeks; and in both instances the condition could therefore only be satisfied by the employee doing something within his own control – either resigning, or committing misconduct. I comment that the relevant misconduct could have occurred earlier, and only been discovered by DB during the six week period, and so the suggestion that that element of the provision is entirely within the employee’s control during that period may be an overstatement. Mr Lasok said, however, that the ‘risk’ associated with the restriction of the C1 shares could be considerably minimised, or reduced to a negligible level, by an employee simply determining to remain in employment and so within the scheme.

Secondly, DB’s evidence was that the article 34 restriction on the C1 shares would result in their market value being reduced by at most 2 to 3 per cent, which HMRC submitted was de minimis and should accordingly be ignored. The result was that the restriction was said to turn on a commercially irrelevant contingency, which at most created an acceptable risk that the scheme might not achieve its tax avoidance aim. In those circumstances, article 34 should be disregarded in the assessment of whether the C1 shares were ‘restricted securities’. That would leave the scheme as a vehicle for conferring an unrestricted interest in securities upon the employees, and so would fall outside Chapter 2.

Although the first point was not argued before the UT, it is one with which the UT would have disagreed. The UT said:

Mr Goy, in response to the first head of HMRC’s argument, advanced essentially the same points. He added that there is no stress in sections 423 to 425 of Chapter 2 on why the restrictions are in place, that is whether they are there for commercial reasons or otherwise, nor whether the employee can ensure that they do not apply. The stress is simply on whether the restrictions reduce value, nothing more. I agree and would not accept HMRC’s first line of argument.

As to HMRC’s second, de minimis point, the UT was similarly unimpressed that there was anything in it. They said:

Mr Goy submitted that we should, in answer to this submission, adopt the approach of the UT. I agree. It was in my view a sound and correct approach. I would reject this argument too.

Mr Lasok also advanced this argument in relation to the DB scheme. I would reject it for the same reasons I have given in relation to the like point in the UBS appeal.

Mr Lasok submitted that, in the case of the DB scheme, there was in fact no provision under which there would a ‘transfer, reversion or forfeiture’ of the C1 shares within the meaning of section 423(2)(a), with the consequence that the C1 shares were not ‘restricted securities’ within the meaning of section 423(1). The point was that article 34(a)(ii) provided that on the occurrence of the relevant ‘circumstances’ the C1 shares will be converted into C2 shares. There is, however, no provision for their ‘transfer, reversion or forfeiture’, and instead article 34(b) provides that the C2 shares so created shall be transferred to the holders of the ordinary shares for nil consideration. Whilst therefore there is a provision in relation to the C2 shares so created for their ‘transfer, reversion or forfeiture’, there is none in relation to the C1 shares.

HMRC did not argue this in the tribunals below, but the UT raised the point themselves in [197], and disposed of it shortly. They said:

Mr Goy supported the reasoning of the UT, and I also agree with it. The consequence of the happening of the relevant ‘circumstances’ is that the C1 shares are converted to C2 shares and become subject to an immediate transfer for no consideration. I agree with the UT that that amounted to the ‘forfeiture’ of the C1 shares. That, I consider, is how the reasonable employee would regard it. Mr Lasok accepted that his contrary argument was what he called ‘an arid Chancery point’. I regard it as wrong and would reject it.

The result is that I would reject all HMRC’s submissions directed at upholding the UT’s dismissal of DB’s appeal against the decision of the FTT.

I would allow DB’s appeal against the decision of the UT. I presume that will result in an order setting aside the two determinations dated 29 April 2008 referred to in [4] above.

I would dismiss HMRC’s appeal in the UBS case. I would allow DB’s appeal in the way just described.

I agree.

I also agree.