The Commissioners for HM Revenue and Customs v Stephen Warshaw [2020] UKUT 0366 (TCC) [2020] UKUT 0366 (TCC)
UPPER TRIBUNAL
TAX AND CHANCERY CHAMBER
[2020] UKUT 0366 (TCC)Case No UT/2019/0104
Between
The Commissioners for HM Revenue and CustomsAppellantStephen WarshawRespondent
Before
MR JUSTICE MEADEJUDGE THOMAS SCOTTDate 23 December 2020Category: Tax
[1]HMRC appeal against the decision of the First-tier Tribunal (the “FTT”) reported at [2019] UKFTT 268 (TC) (the “Decision”). The appeal raises a short but 5 important point on the statutory construction of the term “ordinary share capital” for the purposes of entrepreneurs’ relief.[2]Before the FTT Mr Warshaw appealed against a closure notice issued by HMRC which denied his claim for entrepreneurs’ relief on a disposal of shares. The relief was denied on the basis that the company in question was not Mr Warshaw’s 10 “personal company”, as required by the legislation, because certain preference shares which he held were not “ordinary share capital”. The FTT allowed Mr Warshaw’s appeal. With permission of the FTT, HMRC appeal against the Decision. The relevant legislation[3]Entrepreneurs’ relief applies a lower effective rate to certain chargeable gains. 15 References below are to the legislation as in force at the relevant period. The relief applies to qualifying business disposals, of which one is a material disposal of business assets: section 169H(2)(a) Taxation of Chargeable Gains Act 1992 (“TCGA”).
(c) 20
[4]In order to be a material disposal of shares, two conditions must be satisfied. For the purpose of this appeal, only Condition A is relevant. This is set out in section 169I(6): Condition A is that, throughout the period of 1 year ending with the date of the disposal —(a) the company is the individual’s personal 25 company and is either a trading company or the holding company of a trading group, and(b) the individual is an officer or employee of the company or (if the company is a member of a trading group) of one or more companies which are members of the trading group.[5]It is thus a requirement that, for the relevant period prior to the disposal, the 30 company is the individual’s personal company. That term is defined by section 169S(3), as follows: For the purposes of this Chapter ‘personal company’, in relation to an individual, means a company —(a) at least 5% of the ordinary share capital of which is held by the 35 individual, and(b) at least 5% of the voting rights in which are exercisable by the individual by virtue of that holding.[6]A “personal company” is therefore defined by reference to a specified percentage of both ordinary share capital and voting rights. This appeal concerns 40 4 “ordinary share capital”, which is itself defined by section 169S(5) in the following way: “ordinary share capital” has the same meaning as in the Income Tax Acts (see section 989 of ITA 2007).[7]Section 989 Income Tax act 2007 (“section 989”) defines “ordinary share 5 capital” as follows: “ordinary share capital”, in relation to a company, means all the company’s issued share capital (however described), other than capital the holders of which have a right to a dividend at a fixed rate but have no other right to share in the company’s profits… 10 Background and relevant facts[8]The facts were summarised by the FTT at [2] of the Decision as follows: The facts are not disputed and are set out in the following ‘Statement of Agreed Facts’ provided by the parties: Statement of Agreed Facts 15(1) The Appellant Mr Stephen Warshaw, was chairman of a UK-based company known as Cambridge Education Group Limited (“CEG”) which is the holding company of the Cambridge Education Group (“the Group”). Prior to 12 March 2012, he held 44,183 ordinary shares and 396,000 preference shares in CEG.(2) The majority shareholder in CEG was a private equity firm, Palamon Capital Partners. In 20 March 2012, Palamon Capital Partners undertook a reorganisation of the Group as part of a recapitalisation. As a result of this reorganisation, two new holding companies were inserted into the Group’s structures: Cambridge Education Holdings 1 (Jersey) Limited (“the Company”) and Cambridge Education Holdings 2 (Jersey) Limited (“Company 2”).(3) On 12 March 2012, Mr Warshaw exchanged all his ordinary and preference shares in 25 CEG for new shares in Company 2. On 13 March 2012, Mr Warshaw exchanged his ordinary and preference shares in Company 2 for new shares in the Company.(4) As a result of these changes, Mr Warshaw’s shareholding in the Company replicated his original shareholding in CEG. He therefore held 44,183 ordinary shares and 396,000 preference shares in the Company. On 26 March 2012, he subscribed for 24,660 B ordinary shares in the 30 Company. He became a director of the Company on 26 October 2012.(5) The rights attaching to the various classes of shares in the Company were set out in its Articles of Association. “Participating Shares” were defined as: “the Ordinary Shares, the A Ordinary Shares, the B Ordinary Shares and the C Ordinary Shares” 35(6) “Preference Shares” were defined as: “the 10 per cent cumulative preference shares of £0.01 each in the capital of the Company having the rights and restrictions set out in these articles and “Preference Shares” shall be construed accordingly”.(7) Article 2.3 stated that: 40 “… The Preference Shares shall carry no rights to participate in the profits and assets of the Company except as provided in these articles.”(8) Article 2.4(A) defined the “Preference Dividend” as follows: 5 “In priority to any other class of shares, each Preference Share shall have the right to a fixed cumulative preferential dividend (“the Preference Dividend”) which shall accrue on a daily basis from the dividend commencement date at the rate of 10 per cent per annum on the aggregate of (i) the subscription price of such Preference Share and (ii) the aggregate 5 amount of Preference Dividend that has previously compounded and not yet paid. The Preference Dividend accruing on each Preference Share shall be compounded on each anniversary of its dividend commencement date to the extent not previously paid.”(9) Article 2.4(E) stated: 10 “Aside from the Preference Dividend, no other dividends or distributions shall be made, paid or declared with respect to the Preference Shares”.(10) If the Preference shares were “ordinary share capital” (as defined in section 989 Income Tax Act 2007), Mr Warshaw held 5.777% of the “ordinary share capital” of the Company. However, if the preference shares were not “ordinary share capital”, he held only 3.5% of the 15 Company’s “ordinary share capital”.(11) On 4 December 2013, Mr Warshaw disposed of his entire shareholding in the Company for cash. He also ceased to be a director of the Company and chairman of CEG as of that date. The proceeds, costs and gains from his sale of his shareholding were: Proceeds Acquisition Cost Disposal Costs Gains Ordinary Shares £6,429,549£4,420£174,021£6,251,108 Preference Shares £77,747£39,600£2,104£36,043 B Ord. Shares £158,036£2,491£4,277£151,268 20(12) On 28 January 2015, Mr Warshaw submitted his 2013-14 self-assessment tax return. This return included a capital gains computation for the disposal of the above shares in the Company reporting a total gain of £6,438,419, and a claim for entrepreneurs’ relief in respect of the disposal.(13) On 5 October 2015, HMRC opened an enquiry into that return to look at the capital gains 25 position. The closure notice issued on 10 August 2017 stated that the capital gains arising on the disposal of shares in the Company did not qualify for entrepreneurs’ relief because the Company was not Mr Warshaw’s “personal company” (as defined in section 169S(3) Taxation of Chargeable Gains Act 1992). Mr Warshaw’s return for 2013-14 was amended in line with that decision. 30(14) Mr Warshaw appealed the decision on 4 September 2017. The decision was upheld on review on 3 November 2017. Mr Warshaw filed his notice of appeal on 1 December 2017. The issue in this appeal[9]In relation to the eligibility for entrepreneurs’ relief of Mr Warshaw’s disposal of shares, the parties agreed that the following requirements were satisfied: 35(1) Mr Warshaw was an officer of the Company.(2) The Company was the holding company of a trading group.(3) The Preference Shares did not carry any right to share in the profits of the Company other than the Preference Dividend as defined in Article 2.4(A). 6(4) Irrespective of whether the Preference Shares were “ordinary share capital”, Mr Warshaw’s shareholding satisfied the requirement as to 5% voting control in section 169S(3)(b).[10]The reasoning supporting point (4) is not discussed in the Decision. Nor is it clarified in the skeleton arguments of the parties before the FTT or this Tribunal. On 5 the face of it, it is somewhat puzzling that Mr Warshaw’s holding of Ordinary shares and B Ordinary shares could confer at least 5% of the voting rights while amounting to less than 5% of the “ordinary share capital”. However, an examination of the Company’s constitutional documents and corporate registrations shows that, while the B Ordinary shares (like the Ordinary shares) each carried one vote, their nominal 10 value was a fraction of the Ordinary shares. So, Mr Warshaw’s holding of Ordinary shares and B Ordinary shares amounted to more than 5% in number of the total issued shares of those classes, thereby conferring more than 5% of the voting rights, but less than 5% in terms of the total nominal value of the issued shares of those classes.1[11]The sole issue in this appeal, as it was before the FTT, is whether or not the 15 Preference Shares carried a “right to a dividend at a fixed rate”. If the answer is affirmative, then the Preference Shares are excluded from the definition of “ordinary share capital”, the Company was not Mr Warshaw’s “personal company”, and he was not entitled to entrepreneurs’ relief. If, as the FTT determined, the answer is negative, the contrary consequences would ensue, and HMRC’s appeal must fail. 20 The FTT decision[12]The FTT summarised the positions of the parties as follows, at [9]: In essence, Ms Choudhury contends that as the rate at which the dividend was paid remained fixed at 10%, there was a right to a dividend at a fixed rate even if the base in respect of which it was paid, 25 the compounded element, varied. Mr Ewart, however, says that because the rate of dividend is calculated by reference to any previous unpaid dividends, the preference shares did not have a right to a dividend at a fixed rate.[13]HMRC referred in support to the decisions in The Waverley Hydropathic 30 Company Limited v Barrowman (1895) 23 R 136 (“Waverley”), and Bielckus & Others v HMRC [2016] UKFTT 271 (TC) (“Bielckus”). In addition, Ms Choudhury (who, like Mr Ewart and Mr Windle, also appeared before the FTT) cited certain other provisions of the tax code which, while not directly relevant to the construction of section 989 ITA, were said to illustrate HMRC’s interpretation. 35[14]Mr Ewart relied on the “compounding” effect of the rights attached to the Preference Shares. He contended that the shares did not carry a right to a dividend at a fixed rate because the amount to which the rate was applied was not fixed. In support 1 Measurement of a shareholder’s percentage ownership of a company’s issued share capital by reference to nominal value is the generally accepted approach: see Canada Safeway Ltd v IRC [1973] Ch 374. 7 of this construction, he referred to a passage in Tilcon Ltd v Holland (Inspector of Taxes) [1981] STC 365 (“Tilcon”).[15]The FTT set out its conclusions as follows:[17]Although initially attracted by the argument advanced by Ms Choudhury, on balance and for the reasons below, I prefer that of Mr 5 Ewart.[18]Notwithstanding the similarities between the final “further provision” in The Waverley Hydropathic Company case and the condition attaching to the preference shares in the present case, I am unable to derive any assistance from the observations of the Lord 10 Justice Clerk. Not only were these in relation to different factual circumstances but they were made at a time when the principles of company law were still at a formative stage, pre-dating the seminal decision of Salomon v Salomon & Co Ltd [1897] AC 22. Neither am I assisted by the decision of the Tribunal in Bielckus in which, unlike the 15 present case and as is clear from [53] of the decision, it was not disputed that the shares carried a “right to a dividend at a fixed rate”. Also, as Ms Choudhury accepts, the other statutory provisions to which she referred are not directly relevant to the construction of s 989 ITA.[19]However, I agree with Mr Ewart that the decision of Vinelott J 20 in Tilcon v Holland does offer some support for the need to take into account both the percentage element and the amount to which it is applied to identify the rate of the dividend. Accordingly, if, as in the present case, at the time the preference shares are issued the Articles of Association provide that only one of these, the percentage element, is 25 fixed and the amount to which that percentage is to be applied may vary, those shares cannot be regarded as having a right to a dividend at a fixed rate and are therefore ordinary share capital as defined by s 989 ITA.[20]Moreover, if the approach advocated by HMRC was correct and 30 it was enough for only the percentage element to be fixed (and not that to which it is to be applied) when calculating a dividend, shares that pay a dividend at a rate of 0.01% of the taxable profits of a company or preference shares that pay a dividend at a rate of 50% of the dividend on the company’s A ordinary shares would not be ordinary share 35 capital within s 989 ITA. In my judgment this cannot be right.[21]Therefore, having concluded that the preference shares held by Mr Warshaw were “ordinary share capital” it follows that the Company was Mr Warshaw’s “personal company” and that, as such, he is entitled to entrepreneurs relief on the disposal of those shares. 40 Discussion[16]HMRC’s stated grounds of appeal were as follows:(1) The FTT erred in holding that the dividend attached to the Preference Shares was not at a fixed rate because it was cumulative. 8(2) The reasoning in Tilcon did not apply in this case.(3) HMRC’s interpretation would not lead to the consequences referred to at [20] of the Decision.[17]It is important to note at the outset that under the Articles of Association the Preference Shares carried two rights relevant to the issue in this appeal. First, the 5 dividend rights were cumulative. This simply means that if a dividend had accrued but was not paid when due, then it remained due, or accumulated. Second, any accrued but unpaid dividends were to be compounded from the date they were due, and the dividend accrued on the aggregate of the subscription price for the preference share and the amount of any dividends previously compounded and not yet paid. 10[18]HMRC’s first ground of appeal appears to elide these two rights. In fairness, so did the FTT, in the following passage in the Decision, at [8]: …Because the preference shares were cumulative, if there were insufficient reserves to pay the dividends in respect of those shares in a particular year, payment was deferred to a subsequent year. Therefore, 15 … the rate at which the dividend would be paid, 10%, would be calculated on an increased amount (ie the aggregate of(i) the subscription price and(ii) the aggregate unpaid dividends).[19]The confusion arises from the use of “therefore”. Cumulation and compounding both afford a measure of compensation for a failure to pay a dividend when due, but 20 they are distinct rights, and the latter does not necessarily follow from the former.2 In fact, Mr Ewart’s argument before the FTT relied on the effect on the rights of the Preference Shares arising from the compounding: see [14] of the Decision.[20]In our opinion, it is clear that a fixed rate dividend right does not cease to be fixed rate merely because it is cumulative. The right remains a right to a dividend at a 25 fixed rate. Indeed, the strength of the right is arguably increased by an express right of cumulation. The fact that an accumulated fixed rate dividend may be payable and therefore paid after the due date is relevant to the timing of the payment but does not have the result that either the percentage or the amount to which it is applied is not fixed. 30[21]It is apparent from paragraphs [17] and [19] of the Decision that in reaching its conclusion the FTT was persuaded by Mr Ewart’s argument that the effect of the compounding right was to render variable the base to which the 10% rate could be applied, with the consequence that the shares did not carry a right to a dividend at a fixed rate. The issue is therefore the effect of the right to compounding. The 35 remainder of this decision discusses whether the FTT’s conclusion on that issue was an error of law.[22]We begin with the authorities cited by the parties. 2 The decision in Bielckus, for example, concerned fixed rate preference shares which were cumulative but not compounding. 9[23]The Scottish case of Waverley referred to by HMRC concerned a company which was prohibited by its memorandum and articles of association from issuing preference shares. The company purported to create and allot fixed rate cumulative non-voting preference shares. The court was asked to determine whether an allottee of such an unauthorised instrument was an ordinary shareholder of the company. The 5 Court of Session held that he was not, and that he stood in the position of creditor in relation to the company. The passage on which Ms Choudhury relied was at page 140, as follows (with the emphasis from the FTT’s decision): The second parties in this case bought some years ago from the Hydropathic Company certain shares on the footing that they were 10 preference shares, and subject to various conditions. These conditions are peculiar. The fourth of these conditions is that the company has a right, in the event of a transfer, to purchase the shares at par, and at any time after ten years they have right to redeem any of such shares at par. That is a privilege given to the company which is quite inconsistent 15 with the idea that these shares are ordinary shares. Then again under the sixth condition the holders of these shares are not to be entitled to vote or be present at the meetings of the company, or to take part in the management of the company's affairs, or to participate in any of the rights or privileges of the ordinary shareholders. Now this is a 20 restriction of the rights of the purchasers of the preference shares which puts them outside of the ordinary category of shareholders altogether. There is a further provision applicable to these preference shares, and to no others, viz., that they are to receive a certain fixed rate of dividend, and if the profits are not sufficient in 25 any year to pay the dividend, the deficiency is to be made good out of the profits of subsequent years. This again places these shares in a position as unlike that of ordinary shares as possible.[24]We do not consider that this passage assists with the question of construction before us. In disagreement with the FTT (at [18]), we do not see that the fact that the 30 decision was reached shortly before Salomon v Salomon is relevant. However, it was dealing with a quite different issue, namely whether or not an illegal allotment was void in the circumstances. Moreover, there is no indication that in this passage the Court is using the term “ordinary shares” in the technical sense it is used in subsequent tax legislation3: rather, the Court refers to what it describes as “the 35 ordinary class of shareholders”. In any event, we do not accept that a statement highlighting that certain categories of preference share are very different to other classes of share, with which we agree, assists the question in this appeal.[25]We agree with the FTT that the decision in Bielckus is of no material assistance, because it concerns the separate question raised by section 989 of whether or not 40 particular shares which were agreed to carry a dividend “at a fixed rate” carried any additional right to share in the profits of the company. That is not in issue in this appeal. 3 The origins of the section 989 definition have been traced back to section 42(3) of the Finance Act 1938: Castledine v HMRC [2016] UKFTT 145 (TC) at [32]. 10[26]The FTT accepted Mr Ewart’s submission that the decision in Tilcon supported his argument that in order for a dividend to be at a fixed rate, both the rate of the dividend and the amount to which it applied must be fixed. That case concerned the provisions regarding group relief and consortium relief for companies. One issue was whether preference shares carrying a dividend which varied with the rate of advance 5 corporation tax in force for the year of payment were “ordinary share capital” within section 526(5) of the Income and Corporation Taxes Act 1970. Ordinary share capital was defined in section 526, as originally enacted, as excluding share capital “the holders whereof have a right to a dividend at a fixed rate or a rate fluctuating in accordance with the standard rate of income tax, but have no other right to share in 10 the profits of the company”. The additional restrictions introduced into the group relief rules in 1973 to apply requirements in relation to “equity holders” (in addition to ownership of a certain percentage of ordinary share capital) included within the definition of equity holder a holder of ordinary shares. “Ordinary shares” were defined for this purpose as “all shares other than fixed rate preference shares”, and, so 15 far as material, fixed rate preference shares were defined as shares which “do not carry any right to dividends other than dividends which…are of a fixed amount or at a fixed rate per cent of the nominal value of the shares”. Vinelott J stated as follows, at page 373: I accept that the holder of a share entitled to a preference dividend of 20 an amount which falls to be ascertained by reference to the rate of advance corporation tax for the year in which it is paid cannot be said to have a "right to a dividend at a fixed rate" in the natural and ordinary sense of those words. But this construction leads to results which are so patently absurd as to raise, at the very least, a doubt whether they 25 could have been contemplated by the draftsman or the legislature as following from it. The definition of "ordinary share capital" in s 526 governs the definitions of "51 per cent. subsidiary", "75 per cent. subsidiary" and "90 per cent. subsidiary" in s 532(1), and those definitions in turn govern the provisions of ss 252 and 253, which 30 relate to reconstructions; of ss 256 and 257, which relate to group income; of ss 272 to 278, which relate to grouping for the purposes of capital gains tax; of s 284(4) and (5), which deal with close companies; and of ss 483 and 484, which relate to the disallowance of trading losses. It is impossible to suppose that the draftsman intended to make 35 a bonfire of these elaborate and important parts of the tax system. Further, the application of the provisions of s 28 of the Finance Act 1973, restricting group and consortium relief, themselves depend upon definitions of "ordinary shares" and "fixed rate preference shares", which, while they differ in language from s 526, embody the same 40 concept. A fixed rate preference share is one which "does not carry any right to dividends other than dividends which (i) are of a fixed amount or at a fixed rate per cent of the nominal value of the share".[27]We are not here concerned with the ratio of Tilcon. Rather, Mr Ewart highlights the statement by Vinelott J to the effect that the two exclusions from ordinary share 45 capital of fixed rate preference shares, in the original section 526 and later in section 28 Finance Act 1973, “embody the same concept”. We understand his argument to be as follows. Mr Ewart’s central submission is that “a fixed rate” in section 989 ITA 11 means a fixed percentage applied to a fixed amount, namely the nominal value of the share. In the definition introduced by section 28 Finance Act 1973, that was in effect spelt out, because it referred to “a fixed rate per cent of the nominal value of the shares”. While not made explicit in section 526, if the two definitions “embody the same concept”, then the same construction should apply. 5[28]The FTT concluded that this passage “does offer some support” for Mr Ewart’s central submission. We agree. At a broad level, it is at the least consistent with the proposition that a rate which applies to a variable amount is not a “fixed rate”. Mr Ewart rightly accepted, however, that it does not determine the issue before us, because Vinelott J’s comment was made in the context of determining a different 10 issue to the effect of compounding, and, most importantly, it would be possible for a definition to “embody the same concept” as another definition without having precisely the same effect. However, the passage is broadly supportive of Mr Ewart’s contention.[29]While there is no direct authority on the issue in this appeal, we do have helpful 15 guidance to the approach which should be adopted in construing section 989. The provision was considered by the Upper Tribunal in HMRC v McQuillan [2017] UKUT 344 (TCC) (“McQillan”). The issue in that case, whether or not a dividend of zero was a dividend at a fixed rate, arose in the same context as this appeal, namely the definition of “personal company” for the purposes of entrepreneurs’ relief. The 20 Tribunal began by describing section 989 in the following terms, at [20]: Section 989 starts with the premise that ‘ordinary share capital’, as a defined term, includes all of a company’s issued share capital…The only exception is in respect of capital(a) the holders of which have a right to a dividend,(b) which dividend is at a fixed rate and(c) where 25 that dividend right is exhaustive of the right to share in the company’s profits.[30]In McQuillan, the taxpayers sought to establish that certain shares which they did not hold were fixed rate preference shares (and therefore not ordinary share capital), arguing that a purposive construction of section 989 was required in order to 30 be consistent with the spirit and intention of the code for entrepreneurs’ relief. The Upper Tribunal set out its views as follows, at [33] to [35]: [33] We are satisfied that no process of purposive construction can have that effect. First, it is important to recognise that the ‘purpose’ here must be the purpose of s 989 not the purpose of the entrepreneurs’ 35 relief provisions in the TCGA. We do not agree with the assumption implicit in [34] of the FTT’s decision that s 989 ITA might bear a different meaning when it is imported by cross-reference into provisions relating to a particular tax regime from the meaning it bears in the statute of which the definition section forms a part. If Parliament 40 had wished to enact a bespoke meaning of ‘ordinary share capital’ when defining an individual’s ‘personal company’ in s 169S(3) TCGA it could have done so. Thus, in some cases where ordinary share capital has been employed as the means of establishing a particular relationship, that has led to the legislature augmenting the provisions to 45 12 introduce further tests more aligned to the economic result. Ms Lemos, appearing for HMRC, referred us to one example, namely the introduction of additional requirements, running alongside the basic ordinary share capital test, for companies to be 75% and 90% subsidiaries (see s 151 of the Corporation Tax Act 2010 (‘CTA’)) for 5 the purpose of group and consortium relief. Nothing so sophisticated applies to the entrepreneurs’ relief, because Parliament chose to import a definition which, as Ms Lemos showed us, is used in many different statutory contexts relating to tax law. Although counsel for HMRC before the FTT may not have had a ready answer to the FTT’s query 10 about the wider effect impact of the construction of s 989 argued for by the McQuillans, the tribunal must take into account the risk that the repercussions of its construction of s 989 could be unexpected and far reaching. [34] In our judgment, the purpose of s 989 ITA is readily apparent 15 from the statutory language itself. It is a provision of definition which is intended to describe a clear, and readily understandable, description of the shares to which it applies. It is imported into s 169S(3) TCGA to establish a bright dividing line between those shares which will be reckoned with in assessing the extent of an individual’s interest in a 20 company for the applicable period prior to the disposal of shares or securities in a company, and shares which are not. A dividing line will have the necessary, and inevitable, consequence that some cases which are similar in economic terms will fall on one side of the line and others will fall on the other side. That, as the Upper Tribunal (Asplin J 25 and Judge Berner) said in Trigg v Revenue and Customs Comrs [2016] UKUT 165 (TCC), [2016] STC 1310, at [57], is nothing more than a normal incident of the drafting of statutory conditions defining a particular statutory concept (in this case that of ‘ordinary share capital’). As the Upper Tribunal in Trigg4 went on to say: ‘It is not for 30 the tribunal to fill any perceived gap, or to seek to equate cases on one side of the dividing line with similar cases falling on the other side by reason of similarity in effect or economic equivalence. Purposive construction cannot go so far. To construe such legislative conditions in that way would risk undermining rather than applying the distinction 35 determined upon by Parliament according to the plain words of the legislation.’ [35] The language of s 989 is, as we have found, unambiguous. The intention is equally clear and unambiguous from the language that Parliament has adopted. There is in our view no possible recourse in 40 this case to the spirit of the legislation. As the Upper Tribunal said in Trigg, at [35]: ‘There is also, in our judgment, a distinction between the policy behind, or the reason for, the inclusion of a particular provision in the legislative scheme and the purpose of that provision. Parliament 45 might wish to achieve a particular result as a general matter, and legislate for that reason or in pursuit of that policy. But if the 4 The decision of the Upper Tribunal in Trigg was reversed by the Court of Appeal on the facts ([2018] EWCA Civ 17) but no doubt was cast on these comments as to purposive construction. 13 statutory language adopted by Parliament displays a narrower, or more focused, purpose than the more general underlying policy or reason, it is no part of an exercise in purposive construction to give effect to a perceived wider outcome than can properly be borne by the statutory language.’ 5[31]We respectfully endorse the approach and observations in these passages, and return below to their relevance in this appeal.[32]Before we address the question of construction in this appeal, we can deal briefly with three points.[33]First, we were referred to a table of examples in HMRC’s Company Taxation 10 Manual5 of various shares with different rights, with in each case a statement as to whether the shares were or were not ordinary share capital within section 989 and a short commentary. While this served to illustrate that the definition may produce different results for shares which carry economically similar rights, it was common ground that it was of no material assistance to the issue in this appeal. 15[34]Second, we were referred by Ms Choudhury to various statutory definitions in other contexts which deal with ordinary shares and preference shares. It was accepted that these were not directly relevant to the interpretation of section 989, but it was submitted by Ms Choudhury that they were consistent with HMRC’s position in this appeal that the Preference Shares were not ordinary share capital. In agreement with 20 the FTT, we did not find that these other provisions illuminated the issue in this appeal.[35]Third, Ms Choudhury submitted that the compounding right attached to the Preference Shares did not prevent them from carrying a right to a dividend at a fixed rate because all that right did was to effectively provide compensation by way of 25 interest for unpaid dividends. However, we agree with Mr Ewart that there was no evidence as to the intention behind the inclusion of that right and, in any event, what matters for the purposes of section 989 is the rights attached to shares, not why those rights were attached.[36]In construing section 989, we consider that the following principles apply: 30(1) In adopting a purposive construction, the relevant purpose is that of section 989, not the legislative code in which the section falls to be considered in any particular case.(2) The purpose of section 989 is simply to act as a definition, which produces a bright line between issued share capital which is ordinary share 35 capital and that which is not. Classification of shares as ordinary share capital may be what the taxpayer seeks (as with the Preference Shares in this appeal) 5 CTM 00514. 14 or what he seeks to avoid (as with the redeemable shares in McQuillan or the deferred shares in Castledine v HMRC [2016] UKFTT 145 (TC)).(3) The definition is by its terms formalistic in nature. It looks solely to the dividend rights and any additional right to share in the company’s profits attached to a share. As stated in Trigg, it is not for the tribunal “to seek to equate 5 cases on one side of the dividing line with similar cases falling on the other side by reason of similarity in effect or economic equivalence”. Expressed another way, any process of construing section 989 should not search for whether in any particular case a classification might appear fair or just: as stated in McQuillan, a definition such as section 989 “is apt to produce results which appear unfair”: 10 [45].(4) Consistently with this point, we do not accept HMRC’s proposition that the statutory distinction between a share which is ordinary share capital and one which is a fixed rate preference share should be based on, or even informed by, whether in economic terms the share is “debt-like”. While a fixed rate 15 preference share may share some economic characteristics with a debt instrument in terms of risk and reward, it remains equity (as would be readily apparent in an insolvency) and, for a company incorporated in the UK, lawful payment of the fixed dividend (unlike an obligation to pay interest) requires the company to have available sufficient distributable profits. It is also not evident 20 why a preference share carrying a dividend right wholly or partially based on Libor (which it is common ground would be ordinary share capital) is any less “debt-like” than one carrying a dividend at a fixed rate.(5) The question of whether a share is ordinary share capital is determined by the rights attached to the share, and not the subjective intentions of the parties as 25 to its tax status, or what happens in practice. So, for instance, a share carrying a dividend right which is expressed as a fixed rate plus a contingent amount is not transformed into a fixed rate preference share for years in which the contingent amount does not in fact fall due.[37]With these principles in mind, we turn to the issue in this appeal. 30[38]The question is what is meant in section 989 by a right to a dividend “at a fixed rate”. In our opinion, “rate” in this context describes the relationship between two variables or different units of measurement, expressed as a ratio. It is clear in our opinion (and accepted by both parties), that a “fixed rate” requires the rate of dividend to be expressed as a fixed percentage or amount per share. So, a dividend right of 1% 35 plus Libor is not a right to a dividend at a fixed rate.[39]The issue is whether it is also necessary for the rate to be fixed as to the amount to which it is applied. Applying the principles which we have described, we consider that the answer must be yes.[40]We set out in an appendix to this decision the illustration which the FTT 40 included in its decision of the effect of the compounding right on the Preference Shares. Under the rights attached to the Preference Shares, the 10% rate is applied not only to the subscription amount, but also to the aggregate amount of any accrued but 15 unpaid dividends. As a result, the 10% rate applies to an amount which may vary and cannot be determined at the date of issue of the shares. As the appendix shows, if the dividends had all been paid when due, then 5 years after the date of issue the rate of dividend would be 10% of the nominal value of the shares. However, if no dividends had been paid when due, then after 5 years the dividend right would equal 14.6% of 5 the nominal value. If at any stage the arrears of dividend had been paid, the dividend right would again have become 10% of the nominal value. We do not consider that that is a right to a dividend “at a fixed rate”.[41]HMRC accept, as they must in our view, that a dividend right of 10% of the company’s profits is not a right to a dividend at a fixed rate, because although the 10 10% is fixed, the amount to which it is to be applied will vary. Ms Choudhury sought to distinguish that situation from the facts in this appeal on the basis that the Preference Shares have only two variables, namely the rate of dividend and the nominal value, whereas profits from year to year are “a third variable”, which is less “determinable and predictable” than any accumulated but unpaid dividends. In our 15 view, such a distinction would be inconsistent with the purpose of section 989 as a definition which provides a bright line. In practice, a shortfall of distributable profits may well be the very reason why a fixed dividend such as that attached to the Preference Shares which has fallen due must be accumulated. In any event, applying the principles we set out above, the degree of determinability or predictability of a 20 variation to the amount on which the dividend rate is calculated cannot in our view determine whether or not a share is or is not ordinary share capital.[42]So, we see no principled basis for distinction between a dividend expressed as a fixed percentage of profits and the dividend on the Preference Shares. We therefore agree with the FTT, who concluded at [20] of the Decision that this was the case. 25[43]We conclude that the FTT reached the right decision. The Preference Shares are “ordinary share capital”, the Company was therefore Mr Warshaw’s “personal company”, and he was entitled to entrepreneurs’ relief on a disposal of his shares. Disposition[44]HMRC’s appeal is dismissed. 30 Signed on original MR JUSTICE MEADE JUDGE THOMAS SCOTT 35 Upper Tribunal Judges RELEASE DATE: 23 December 2020 40 16 APPENDIX 5 Illustration of Effect of ‘Compounding’ Arising from Limb (ii) of Dividend Calculation in Article 2.4(A) of the Company’s Articles of Association on basis of an individual owning preference shares with total subscription price of £10,000 showing the dividends that would accrue if no dividend was paid for five years 10(1) In year 1, Article 2.4A provides for the accrual of a dividend of 10% x (subscription price (£10,000) + aggregate unpaid dividends (£0) = £1,000 This is 10% of the subscription price (nominal share capital of the shares) 15(2) In year 2, Article 2.4A provides for the accrual of a dividend of 10% x (subscription price (£10,000) + aggregate unpaid dividends (£1,000) = £1,100 This is 11% of the subscription price (nominal share capital of the shares) 20(3) In year 3, Article 2.4A provides for the accrual of a dividend of 10% x (subscription price (£10,000) + aggregate unpaid dividends (£2,100) = £1,210 This is 12.1% of the subscription price (nominal share capital of the shares) 25(4) In year 4, Article 2.4A provides for the accrual of a dividend of 10% x (subscription price (£10,000) + aggregate unpaid dividends (£3,310) = £1,331 This is 13.3% of the subscription price (nominal share capital of the shares) 30(5) In year 5, Article 2.4A provides for the accrual of a dividend of 10% x (subscription price (£10,000) + aggregate unpaid dividends (£4,641) = £1,464.10 This is 14.6% of the subscription price (nominal share capital of the shares) 35