Harding v HM Revenue & Customs

This is an appeal from the Special Commissioner (Mr Charles Hellier, sitting alone) promulgated on 15^th March 2007, whereby he dismissed an appeal by Mr Nicholas Harding from an assessment to capital gains tax for the year ended 5^th April 1996 in respect of a gain which the Commissioners for Inland Revenue (now HMRC) decided had arisen on his redemption of certain loan notes (“the Loan Notes”) for £1,925,718 on 1^st July 1995.

There is no issue as to the facts. The only question on this appeal is whether the Loan Notes were, or were not, qualifying corporate bonds (“QCBs”) at that time, within the meaning of section 117 of the Taxation of Chargeable Gains Act 1992 (“the 1992 Act”), and related provisions, in the form then in force. If the Loan Notes were QCBs in July 1995, this appeal succeeds. If they were not, it must be dismissed.

Section 117(1) provides as follows:

The Loan Notes were in standard form and, by reference to Conditions which governed the whole series of which they formed part, contained an option for the holder of each Loan Note exercisable during the ten day period following the giving of a Redemption Notice, to have the Loan Note redeemed in US dollars, Canadian dollars or German deutschmarks, at a defined exchange rate, close but not identical to the exchange rate prevailing at redemption. Condition 4 provided expressly that in the event that the holder should fail to exercise that option within the ten day period for exercise to which I have referred, the option would lapse. Although the Loan Notes were governed by German law, expert evidence demonstrates that time was of the essence for the purposes of exercise of the option, as it would have been under English law.

Mr Harding was issued the Loan Notes by Computer 2000 AG in exchange for shares which he had until then held in Frontline Distribution Limited (“Frontline”) on 13^th January 1995, and on the same day gave notice pursuant to the Conditions to redeem the Loan Notes on 1^st July 1995. He did not exercise his currency conversion option, and accordingly it lapsed on 23^rd January 1995.

In order for an asset to be QCB, it must satisfy the conditions set out in section 117, one of which is that it must be “a security” … “which is expressed in sterling and in respect of which no provision is made for conversion into, or redemption in, a currency other than sterling.”: see section 117(1)(b). I shall refer to this as “the (b) condition”. It is common ground that, as at the date of issue on 13^th January 1995, the Loan Notes were securities, and satisfied all the other requirements of section 117, other than the (b) condition, but that, because of the currency conversion option which I have described, it was not on that date a QCB. Mr Harding’s case, both before the Special Commissioner and on this appeal, is that because his currency conversion option lapsed due to non-exercise on 23^rd January, the Loan Notes no longer fell foul of the (b) condition on 1^st July 1995, so that they were QCBs on the date of redemption. HMRC’s case is that a loan note or other security which contains a currency conversion option on issue, does not cease to fall foul of the (b) condition merely because that option has lapsed by the time of the redemption or other disposal (such as by sale or gift) of the security.

The short question of construction which is raised by this appeal is therefore whether a security in which a currency conversion option has lapsed, becomes, at the moment of lapse, “a security … in respect of which no provision is made for conversion into, or redemption in, a currency other than sterling.”

The reason why this short point of construction is of such importance to the parties is that, upon issue to Mr Harding, the Loan Notes had rolled-over into them a very substantial capital gain which had accrued by reason of the large increase in the value of the shares for which the Loan Notes were exchanged. If HMRC’s case on construction is correct, then that rolled-over gain, together with any additional gain between the issue and redemption of the Loan Notes, became chargeable to tax on 1^st July 1995. If Mr Harding’s case on construction is correct, then the rolled-over gain simply disappeared from tax altogether when his currency conversion option lapsed on 23^rd January, with the delightful consequence (for him) that it will never be taxable at all. Since Mr Southern who appeared for Mr Harding could not point to any other circumstances in which Parliament had consciously legislated for a rolled-over gain to disappear altogether from tax, and could suggest no reason why it should have been consciously intended as the consequence of the specific language of section 117(1)(b), it is, or at least became, common ground before me that if this appeal should succeed, Mr Harding would obtain a windfall benefit as the unintended result of a drafting anomaly.

Although the question of construction with which this appeal is concerned is a short one, turning as it does on the meaning of one short phrase, the parties’ arguments both before the Special Commissioner and before me have ranged far and wide, it being common ground (albeit for different reasons) that the search for the correct meaning of section 117(1)(b), in relation to the consequences of the lapse of a relevant currency option, may be enlightened by an appreciation both of the history of the relevant provisions, an understanding of their purpose, and a setting of section 117 in its context. For Mr Harding, Mr Southern also invited me to draw inferences from amendments made in 1997 to the scheme of which section 117 forms part, the result of which was, from November 1996, to prevent the total escape from taxation of the rolled-over gain. The purpose of a part of those amendments, so Mr Southern submitted, was precisely to close off the anomaly of which, prior to amendment, Mr Harding claims the benefit.

The identification of QCBs as a class of asset qualifying for special capital gains tax treatment was originally enacted by section 64 of and Part I of Schedule 13 to the Finance Act 1984. It is an example of legislation with an expressly stated purpose. Section 64(1) provides as follows:

Sub-section (2) then provides a narrower definition of a QCB than was in force by 1995, in which what are now sub-sub-sections (a) and (b) are expressed in substantially identical terms in sub-sub-sections (b) and (c), but which are preceded by the requirement that as from its issue, the relevant security has been quoted on a recognised stock exchange in the United Kingdom, or dealt with on the Unlisted Securities Market, or was issued by a body which at that time had some other quoted share, stock or security.

In Weston v. Garnett [2005] STC 1134, Buxton LJ offered this fuller explanation of the purpose of the introduction of QCBs in the 1994 Act, by reference to notes in Current Law Statutes:

The capital gains tax treatment of gilt-edged securities referred to in section 64(1)(a) of the Finance Act 1984 is described in section 67 of the Capital Gains Tax Act 1979. In summary, a gain accruing on the disposal by any person of a gilt-edged security shall not (subject to irrelevant exceptions) be a chargeable gain except where the disposal occurs within twelve months of the acquisition of the security.

By section 139 of the Finance Act 1989, a new type of asset was added to the class of QCBs. This was the “deep gain security” provided for in Schedule 11 to that Act. The identification of deep gain securities is a complex and difficult subject in its own right, well beyond the scope of this judgment. Mr Southern tried valiantly to persuade me that an analysis of this structure was material to the true construction of section 117. I mean no discourtesy by saying, without elaboration, that I was not persuaded.

By 1995 the requirement that, to qualify as a QCB, a security had to be listed, or issued by a company with other listed shares or securities, had been removed. It follows that the original purpose, identified by Buxton LJ, of stimulating the British bond market, and limiting the special treatment by way of exception to bonds genuinely traded in that market has been outgrown. Nonetheless, the purpose of excluding loans that give the loan creditor an actual or potential interest in the debtor company or its performance remains, as does the exclusion of participation in anything other than sterling bonds, both denominated and redeemable as such. This follows from the preservation without significant amendment of what is now section 117(1) (a) and (b) from 1984 through to 1995 and beyond.

Mr Southern submitted that the obvious purpose of the exclusion of securities with provision for conversion into, or redemption in, a currency other than sterling was because Parliament consciously wished to ensure that forex gains and losses should remain liable to tax, whereas gains and losses merely incident to the making of normal commercial loans should not. To the extent that it matters, I am not inclined to accept this analysis. If the original purpose of what is now section 117(1)(b) was to only exclude from tax investments ordinarily obtainable in the British market so as to promote tax-efficient competition with gilts, then I see no reason why the extension of the same treatment to off-market bonds should, without any corresponding amendment to that sub-sub-section serve any different purpose. Nonetheless, I accept that the practical effect of that sub-sub-section was to exclude securities incorporating forex risks and advantages from immunity from capital gains tax as QCBs.

Before looking more closely at the definition of QCB, it is necessary to explore in a little depth some of the tax consequences of a corporate security qualifying, or as the case may be, not qualifying as a QCB. At its simplest, the consequences are simple. A person who acquires, either by original issue or by purchase a QCB is exempt from chargeable gains on it, and is not entitled to credit for losses incurred, in each case between his acquisition and his disposal of it, whether by redemption, sale or gift. By contrast, the holder of a security which is not a QCB (“a non-QCB”) is not so exempt.

More complicated consequences follow from the acquisition or, as the case may be, the disposal of QCBs and non-QCBs by way of exchange in the context of corporate reconstructions and take-overs. For present purposes, the relevant context is a takeover in which a chargeable asset in the target company (such as an ordinary share) is exchanged for a security (QCB or non-QCB) in the acquiring company. Taking the exchange of a share for a non-QCB first, the general rule is that the accrued gain in the old asset (the share) is, to use the applicable slang, “rolled-over” into the new asset (the non-QCB). The mechanism by which roll-over is achieved involves two statutory fictions. The first is that the exchange is not treated as involving any disposal of the old asset or any acquisition of the new asset. The second is achieved by treating the new asset as if it had been acquired at the same time, and for the same price, as the old asset. In its simplest form this mechanism is applied where there is a reorganisation of the share capital of a single company, by section 127 of the 1992 Act. In relation to take-overs, section 135(3) applies the same mechanism (with any necessary adaptations) as if the two companies concerned were a single company reorganising its share capital.

The effect of such a roll-over is that the person exchanging his old asset is temporarily relieved from paying tax on the latent capital gain at the time of its exchange for the new asset. That rolled-over gain becomes taxable, together with any gain in the value of the new asset, upon a chargeable disposal of the new asset, such as a sale or redemption of it for cash.

Since the essential feature of a QCB is that a disposal of it for cash is not chargeable to tax, it was necessary to make special provision for an exchange of an old chargeable asset (such as a share) for a QCB, in the context of a take-over. The relevant special provision is contained in section 116. By sub-section (5), section 127 is dis-applied, so that there is no roll-over. By sub-section (10) such an exchange gives rise to no immediately chargeable disposal of the share but the chargeable gain or allowable loss which would then have arisen is calculated, and then charged to tax or (if a loss) allowed at the time of the disposal of the QCB, and at the rate then in force. This is generally known as a “frozen gain” mechanism.

In relation to the latent gain inherent in the old asset immediately prior to exchange, the frozen gain regime achieves in relation to an exchange for QCBs substantially the same result as the roll-over arrangement achieves in relation to an exchange for non-QCBs. In both cases the latent gain in the old asset is taxed upon the disposal of the new asset. As might be expected, gains or losses attributable purely to the new asset are, if it is a QCB, left out of account. The key to the successful operation of the frozen gain regime is that it displaces the ordinary rule that no chargeable gain or allowable loss occurs on the disposal of the QCB.

A feature of the frozen gain regime as it stood in 1995 was that it depended upon the occurrence of “a transaction” of such a description that, apart from section 116, the roll-over provisions of sections 127 to 130 would have applied: see section 116(1)(a). This requirement creates no difficulty where the new asset qualifies both upon its acquisition and upon its disposal, as a QCB. It can only be acquired by means of a transaction. Indeed, the whole of the roll-over scheme and the frozen gain regime substituted for use in connection with QCBs works without anomaly, provided that there is no change in the status of the new asset between acquisition and disposal (whether it is a QCB or a non-QCB).

The anomalies in the present case arise from Mr Harding’s contention that the Loan Notes, which it is common ground were non-QCBs upon acquisition, changed their status upon the lapse of the currency conversion option on 23^rd January 1995, ten days after acquisition and just over five months before redemption. If that change in status occurred, then it is also common ground first, that the roll-over regime was dis-applied because there is no statutory provision for a charge to tax upon the disposal constituted by the redemption of the (by then) QCB. Secondly the frozen gain regime is dis-applied because the Loan Notes were not QCBs upon acquisition, and did not change their status so as to become QCBs by virtue of any transaction. In that context, there appears to have been an inconclusive debate before the Special Commissioner as to whether, even if the lapse of the currency conversion option were to be regarded as a transaction, any provision in sections 116 to 137 would have triggered the calculation of a frozen gain. Since in my judgment it is plainly correct that the mere lapse by non-exercise of a currency conversion option cannot be a transaction, that debate need not be resolved.

The reason why the substantial capital gain latent in the shares which Mr Harding exchanged would fall out of tax altogether is not merely because the disposal of the QCB (if such it was) in 1995 would give rise to no chargeable gain, but also because the exchange of the shares for the Loan Notes in January (when they were by common consent non-QCBs) would not give rise to a charge to tax either: see section 127, as applied by section 135, which together require the new asset to be classified as at the time of the take-over, rather than (if different) at the time of its subsequent disposal.

Of course, if it could be discerned from the definition of a QCB, read as a whole, that the status of a relevant asset as a QCB or non-QCB was immutable throughout its life, no such difficulties would arise. But that begs the very question raised by the present appeal in relation to section 117(1)(b).

Section 117 sets out, in substance, four conditions which must be satisfied if an asset is to qualify as a QCB. First, it must be a “security” as defined in section 132(3)(b), which provides as follows:

It is plain that, used that way, the word “security” identifies an asset in the nature of an investment, rather than property deposited, pledged or charged to secure repayment of a debt. It is common ground that the Loan Notes were a security, but there was a keen debate between counsel as to whether the definition focussed upon the chose in action (i.e. the underlying asset) or upon the document or documents constituting title to it, and recording its precise terms.

“Security” in section 117(1) is clearly used as meaning something distinct from the debt on it referred to in sub-sub-section (a). The distinction between “security” and “debt on a security” was examined in depth by Chadwick LJ in Weston v. Garnett (supra), in particular in the following passage in paragraph 28 of his judgment:

Mr Gibbon for HMRC made much of this point. He submitted with force that if the “security” was the Loan Note document itself then it could not be said that it was a “security in respect of which no provision is made for conversion into, or redemption in, a currency other than sterling” merely because the currency conversion option set out in the Conditions of the Loan Note had lapsed. The Special Commissioner adopted a similar analysis when, at paragraphs 85 and 88 of the Decision, he concluded that:

In the Lewis case Robert Walker J relied, at page 513, on a passage in the judgment of Lord Migdale in Cleveleys Investment Trust Co v. IRC [1971] SC 233, at 243, in which he said:

Tempting though it is to accept this submission (not least because it conveniently puts paid to the anomaly to which Mr Harding’s case gives rise), I have not been persuaded by it. In my judgment the proposition that the “security” in section 117(1) is a reference to the document which evidences the debt was tested to destruction in W T Ramsay Limited v. IRC [1982] AC 300; [1981] STC 174, in which one of the questions was whether a loan (called “loan 2”) was a “debt on a security” within the meaning of the Finance Act 1965, Sch 7, para 11(1). The legislation under review used the concepts of “security” and “debt on a security” in substantially the same way as does section 117, albeit for a different purpose. At page 184 Lord Wilberforce said this:

Nor do I read Chadwick LJ’s judgment in Weston v. Garnett (supra) as suggesting, still less constituting binding authority, that “security” in section 117 means the document, rather than the chose in action evidenced by it. In that case, the assets in issue were two successive loan notes and the terms of the first gave the holder a right to exchange them for the second. The second notes (but not the first) contained a share conversion option, and the question was whether the debt on the security constituted by the first loan notes at all times represented a normal commercial loan within the meaning of section 117(1)(a). The existence of the share conversion option in the second loan note was fatal to the “normal commercial loan” test if, and only if, it was a term of the debt on the security constituted by the first loan notes. The taxpayer’s submission was that the right to convert into shares was conferred only by the second loan notes, and therefore did not affect the question whether the first loan notes were QCBs.

In drawing a sharp distinction between the security and the “debt on a security” which is the focus of section 117(1)(a), Chadwick LJ was seeking to demonstrate not that the security was the document of title to the loan note, but simply the loan was to be distinguished from the chose in action constituted by the loan note, taken as a whole. The issue would have been the same whether the security was a loan note, stock, bond or any other qualifying security. The fact that the phrase “loan note” both in that case and this case includes in the word “note”, something commonly used to describe a document, is in my judgment pure coincidence. There is for example a well understood distinction between a stock (meaning a chose in action) and a stock certificate (meaning the documentary evidence of title). The same distinction exists as a matter of principle between a loan note and the document evidencing it, even though the language does its best to obscure that distinction.

The second condition forming part of the definition of QCB in section 117 is the requirement in sub-section 1(a) that the debt on the security “represents and has at all times represented a normal commercial loan”. I shall refer to it as “the (a) condition”. For present purposes it is unnecessary to investigate further the substantive requirements of that condition. They were pithily summarised by the Special Commissioner in paragraph 54 of the Decision as follows:

For present purposes the real point of the (a) condition is that the debt on the security must satisfy the normal commercial loan requirement “at all times”. The use of the past sense “has” as the means of expressing the “at all times” requirement is in my judgment a reflection of the fact that the question whether or not an asset is a QCB crystallises for most practical purposes upon disposal.

Nonetheless, it is not only upon disposal that the question whether or not an asset is a QCB matters. As I have illustrated, where it is acquired in exchange for a chargeable asset (such as a share) the status of a security as a QCB or non-QCB on acquisition is crucial to the question whether any accrued gain or loss in the share is rolled-over into the new asset, or frozen. Thus, if the new asset is issued upon such an exchange, so that it has no previous life prior to that exchange, it is at first sight rather meaningless to ask whether the underlying debt has “at all times represented a normal commercial loan”. Nonetheless, if the terms which regulate the debt include, for example, a right to convert it into equity exercisable only five years later, it will not be a normal commercial loan on day one: see Weston v. Garnett. For present purposes, what matters is that the phrase “has at all times” puts beyond doubt the proposition that the existence in the terms regulating the debt of any bell or whistle which takes it out of the category of a normal commercial loan will deprive the security of the status of QCB even if that bell or whistle has lapsed by the time of the disposal of the security.

The third condition for the existence of a QCB is the (b) condition itself. It differs from the (a) condition in three respects. First, it is concerned with a single bell or whistle, namely the presence or absence of a currency conversion or redemption provision. Secondly it calls for an analysis of the security, rather the debt on the security. Thirdly, it contains no express “at all times” time requirement, and is governed by the present rather the past tense, both in the phrase “is expressed in sterling” and in the phrase “in respect of which no provision is made”.

The first of those distinctions provokes the question why it was thought necessary to separate the (b) condition from an otherwise compendious prohibition of bells and whistles in the (a) condition. One answer may I think be that it was felt that the presence of a currency conversion option might not deprive an otherwise compliant debt from the status of a normal commercial loan. The equally plausible alternative is that the (b) condition was separately included simply because it reflected a different aspect of the policy, namely the requirement of a status of “Britishness”, a conclusion which can be extrapolated from Buxton LJ’s analysis of the underlying purpose in Weston v Garnett.

It is even harder to understand the second distinction. If one asks what is it that must not be convertible into or redeemable in a currency other than sterling, it is surely the debt, just as much as the security of which it forms the essential part. That conclusion is if anything reinforced by section 117(2)(b), which excludes from the relevant disabling currency redemption provisions one which merely swaps sterling for the relevant foreign currency at the exchange rate prevailing at the moment of redemption. Again, it seems to me entirely plausible that the focus on the security rather than the debt in the (b) condition was also the result of a separate drafting input, occasioned by the need to address a different policy requirement, and therefore that the form of words used differs to an extent by accident rather than by design.

In my judgment, the critical question in this appeal is whether the third distinction (i.e. the change in tense and the omission of an express “at all times” requirement) is also more the result of accident than the reflection of a conscious intention that the (b) condition should operate in a wholly different manner to the (a) condition.

The sheet anchor of Mr Southern’s case is that the use of the present tense (‘is’) in both limbs of the (b) condition requires the question whether that condition is satisfied to be addressed separately and distinctly, and therefore with potentially different results, on every occasion on which the question whether the security is, or is not, a QCB matters. Unlike the (a) condition (which can only admit of a uniform answer whenever in its life the security is tested), the use of the present tense in the (b) condition at least admits of the possibility that: (1) a security may change its status (QCB or non-QCB) during its lifetime, and if so: (2) that it can do so without the occurrence of a transaction within the meaning of section 116(1).

Taking those two questions in turn, there certainly are instances contemplated by other parts of section 117 in which a security may change its status (as between non-QCB and QCB) during its lifetime. For example, sub-section (10) provides that a security issued by a member of a group of companies to another member of the same group is not a QCB “except in relation to a disposal by a person who (at the time of the disposal) is not a member of the same group as the company which issued the security;”. Thus a security which is a non-QCB on an intra-group issue may become one if either the holder leaves the group before disposal or if, after sale to a non-group company or to an individual, that person disposes of it again. Similarly, under sub-section (7), a corporate bond which, because it was issued on or before 13^th March 1984 was a non-QCB on issue, may become a QCB if it is acquired after that date, subject to certain stated exceptions. The date of issue requirement is, incidentally, the fourth condition. It is expressed so as to separate mere corporate bonds (which are non-QCBs) from qualifying corporate bonds.

In the present case, the issue is not so much whether a “corporate bond” as defined by section 117(1) is or is not a qualifying corporate bond, so the sub-section (7) provision for change of status is only indirectly in point. Nonetheless, a common feature of both sub-sections (7) and (10) is that the change of status (in both cases from non-QCB to QCB) only occurs by virtue of a transaction. Furthermore the transaction necessary to trigger sub-section (10) may, but not necessarily will, be a transaction within the meaning of section 116(1) sufficient to trigger the frozen gain provisions which would prevent the anomaly of a chargeable gain falling altogether out of tax. The potential for anomalies arising from section 117(10) probably depends on the intricacies of the taxation of corporate groups, into which minefield counsel before me declined to tread. I shall therefore say no more about it.

It is however clear, not least because despite invitation counsel could not identify any potential example, that nothing in section 117 outside the (b) condition contemplates or permits a security changing its status as between non-QCB and QCB otherwise than by means of a transaction, in circumstances where any inherent or rolled-over gain therefore passes out of tax altogether due to the inapplicability of the frozen gain regime, and the inability to identify a chargeable event upon the disposal of the QCB after acquiring that status. The possibility was touched on that an asset could acquire or lose the status of a QCB by moving in or out of the definition of security in section 117(1). For as long as the security status depended on the issuer being quoted or listed on an exchange, that seems to me to have been a real possibility, but since by 1995 any company qualified as the potential issuer of a security as defined, the possibility seems to me to have become altogether remote by the relevant time.

Mr Southern placed considerable reliance, in support of his submission that section 117(1)(b) necessarily contemplated non-transactional changes of status, upon what appears to have been Parliament’s assumption to that effect when amending the QCB statutory regime in the Finance Act 1997 with effect from November 1996, and therefore only after the dates relevant for present purposes. There are two relevant amendments. The first, to section 132 of the 1992 Act, widened the definition of “conversion of securities”, so that the relevant part of it now reads (with square brackets round the sections inserted by amendment):

The effect of those amendments, from November 1996, is to ensure that if a non-QCB is capable of changing its status without a transaction into a QCB, that will trigger the frozen gain regime, such that any gain which would have been chargeable if on the same date the non-QCB had been disposed of, will be frozen and taxed upon the subsequent disposal of the QCB at the rates then in force. Furthermore, a Revenue press release seeking to explain those amendments stated by way of example that:

Basing himself on the use of “may” as underlined in the above quote, Mr Gibbon cautioned me against placing too much reliance upon an after the event assumption by Parliament (or more realistically the draftsman) as to the effect of an earlier statutory provision. The result of the amendment was to close off the anomaly contended for by Mr Harding in this case, but by a different route (namely the frozen gain regime) than that which would apply to the gain which had occurred prior to the lapse of the currency option if the anomaly posed by his case does not exist. In such circumstances it would be taxed by virtue of the roll-over regime.

I have largely identified or at least hinted at the parties’ rival submissions in the foregoing analysis, but it is convenient at this stage to summarise them. For Mr Harding it was submitted that once it is appreciated that the definition of QCB is such that a security is in principle capable, unlike a leopard, of changing its spots during its lifetime so as to move between QCB and non-QCB status in both directions and, by reference to section 117(1)(b), by the operation of its terms without a transaction, the anomaly that a rolled-over gain existing prior to that change of status is thereby altogether lost to taxation is simply an anomaly which is incurable by any legitimate method of construction, and only (as in fact occurred) by an amendment of the surrounding scheme which left that propensity for non-transactional change of status intact, but deemed a necessary transaction to have occurred.

Mr Southern submitted that the omission of “at all times” and the use of the present tense throughout the (b) condition could not sensibly be described as irrelevant or accidental. It compels a separate analysis of the status of the security on every relevant occasion and, crucially for present purposes, a separate analysis on acquisition and disposal.

Mr Southern sought to justify that outcome of the (b) condition on economic grounds. He submitted that the essence of a QCB was to confer relief from tax on gains to the extent that they were capable of being affected by equity risks and rewards (the (a) condition) and forex risks and rewards (the (b) condition), Since, once a forex option lapsed, the security was no longer capable of having its performance affected by forex risks and rewards, it was appropriate and intended that its holder should thereupon obtain the benefit of that exemption from tax. On that view, he submitted that it was therefore irrelevant that, once it lapsed, the currency conversion option continued to appear as a printed term in the Conditions to the Loan Notes. Both English and German law provided that, upon lapse, the option no longer constituted a term of the chose in action. He relied on a footnote supplied at the Special Commissioner’s request after the hearing by the German law expert to the effect that, upon lapse, the option should be treated as if it had never been included.

The Special Commissioner did not accept Mr Southern’s submission that, once it lapsed, the currency conversion option could not affect the continuing value of the Loan Notes. He relied upon examples to prove that point that had not been discussed during the hearing, and which, on this appeal, Mr Gibbon did not seek to support. Mr Southern’s conclusion was that, shorn of this attempt to provide a commercial rationale, the Special Commissioner’s conclusion that a spent option continued to be a disqualifying provision for conversion into or redemption in a foreign currency within the meaning of the (b) condition was simply an exercise in unjustified formalism.

Mr Gibbon’s submissions for HMRC started with the proposition, which Mr Southern did not challenge, that all legislation, including revenue statutes, are to be construed purposively. He relied in particular on the following paragraph of the judgment of Robert Walker LJ in Billingham v. Cooper [2001] STC 1177:

Secondly, by reference to the materials to which I have already referred, he defined the purpose of the QCB legislation in the manner in which I have described. If the (b) condition were construed in the manner contended for by Mr Harding, he submitted that the consequential anomaly of allowing accrued, including rolled-over, gains on obviously chargeable assets to fall entirely out of tax was so extraordinary that it compelled a search for something other than the narrow literal construction.

That approach to construction was adopted by Neuberger J in Jenks v. Dickinson [1997] STC 853, in another case about anomalies arising from a particular construction of the QCB regime. The particular issue was about the appropriate extent to give to a retrospective deeming provision, but the statements of principle appear to be of general application, and are encapsulated in the following passages from the judgment. The first is a citation from the judgment of the Privy Council in Mangin v. Commissioner of Inland Revenue [1971] AC 739 at 746 which, in the relevant respect, was not subsequently criticised by the House of Lords in IRC v McGuckian [1997] STC 908:

The second was a citation from the speech of Lord Reid in Luke v IRC [1963] AC 557, at 577:

In his own words, Neuberger J applied those principles to the anomaly with which he was faced as follows:

Neuberger J cautioned himself, as shall I, by the following necessary restraint:

Mr Gibbon submitted that the anomaly arising from the wholesale escape from tax of all gains (including rolled-over gains) which had accrued by the time of the lapse of the currency option was so great as to compel a search for an alternative construction to that contended for by Mr Harding, even if Mr Harding’s case reflected more precisely the literal meaning of the words in the (b) condition. He supported the Special Commissioner’s conclusion that the references in the (b) condition to the security rather than the debt justified a relatively formalistic approach, based upon the presence or absence of an offending foreign currency conversion provision in the written Conditions to the Loan Notes regardless whether it had lapsed. He submitted that Mr Harding’s case, based as it was on taking snapshots of the security at different dates logically led to the conclusion that a currency conversion option which had yet to become exercisable should also be excluded, an analysis which, had Mr Southern adopted it, would have led to the identification of a frozen gain on exchange of Mr Harding’s shares for the Loan Notes. He pointed to the artificiality of Mr Harding’s construction which would by its nature apply so as to exclude from consideration a currency conversion option which lapsed only moments before redemption, rather than (as in the present case), some five months earlier. He criticised the case based upon the later amendments as irrelevant and unhelpful, and concluded that the construction adopted by the Special Commissioner avoided the anomaly of enabling accrued gains to escape taxation, without creating what he described as any counter-mischief.

In my judgment a cardinal feature of the task of construction in the present case is the anomaly arises from Mr Harding’s construction which, by permitting a security to change after acquisition from a non-QCB to a QCB before disposal but without any transaction, thereby enables substantial accrued gains to fall altogether out of tax. It is one which is not created by any other provision in section 117 (since all the other potential changes of status are triggered by transactions). The most egregious example of the anomaly is where the non-QCB has, because of its status as such, rolled-over into it a substantial chargeable gain already accrued on the shares for which it was exchanged. In that context it falls fairly and squarely foul of Neuberger J’s analysis in Jenks v. Dickinson, and it gives rise to an apparently irresistible temptation for tax avoidance. The holder of shares replete with chargeable gains may, rather than by selling them and paying the tax, exchange them for a security which is only not a QCB because of a currency conversion option, and then by declining to exercise it convert the security into a QCB which is redeemable tax-free. In the present case the application of any Ramsay test to that analysis was disabled by the undoubted fact that two of Mr Harding’s colleagues accepted loan notes in identical form to his in exchange for their shares, and both exercised their currency conversion options on redemption.

The anomaly is not however confined to cases where there is a rolled-over gain. The object of the statutory distinction between the QCBs and non-QCBs is that only the former should escape tax on their gains. The logical corollary, if a non-QCB is inherently capable of changing into a QCB without a transaction, is that one would expect tax to be payable in respect of gains made during the period when the security did not qualify as a QCB. But Mr Harding’s construction avoids even that consequence. Provided that, however shortly before its disposal, the security is converted, non-transactionally, into a QCB, all earlier gains, whether rolled-over or accrued during the life of the security, fall out of tax.

The mischief which causes the anomaly is simply the propensity of a security which is a non-QCB only because it falls foul of the (b) condition to acquire that desirable status when any relevant currency conversion option lapses. Nothing else in section 117 gives rise to that anomaly, and it is therefore to be questioned whether Parliament can possibly have intended that the (b) condition should do so. Provisions by way of options for currency conversion (rather than a provision under which the conversion is automatic) in a bond otherwise expressed in sterling, are likely to be the norm rather than the exception. This is not therefore one of those cases where the offending anomaly arises only on a relatively unusual set of facts.

I accept Mr Southern’s submission that, literally construed, and in particular by reference to the differences in time-language to which I have referred, by comparison with the (a) condition, the (b) condition does appear to have the consequences for which Mr Harding contends. I also accept that his construction does not have the counter-mischief of failing to tax or relieve gains or losses flowing from benefits and risks inherent in the forex market once the option has lapsed. No such potential for gains or losses has been identified, other than that identified by the Special Commissioner, which Mr Gibbon did not seek to defend. I also accept, contrary to Mr Gibbon’s challenge, that there is nothing linguistically inappropriate in treating the (b) condition as requiring the analysis as at the disposal date to take account of a currency conversion condition which has yet to become exercisable in the future. Such a provision, although not yet exercisable, plainly remains an important term of the bargain at the disposal date, and may at least in theory affect the value of the security before it becomes exercisable, for example on a sale rather on a redemption. In fact, the need to avoid the security becoming a deep gain security means (as was appreciated by the draftsman in the present case) that the effect to which forex risks and benefits could be allowed to alter the redemption price (and therefore the value) was strictly limited, as is reflected in the small 2% difference between the cap and the floor inserted into the currency conversion option in paragraph 4.6 of the Conditions.

Nonethless, the question remains whether there is some legitimate alternative construction of the (b) condition which avoids the glaring anomaly presented by the literal construction, nonetheless implements the general purpose of the QCB regime, and which carries no counter-mischief of its own. I have not been persuaded that the formalistic approach adopted by the Special Commissioner with Mr Gibbon’s continuing support offers the gateway to a different construction. In my judgment that approach is dependant upon treating references to the security as if they referred to the documents, rather than to the underlying chose in action or asset, an approach which I consider to be misguided for the reasons which I have already given in detail. Indeed, as I have sought to explain, the choice to aim the (b) condition at the security rather than the debt seems to be more likely to have been accidental than one invested with any deliberate purpose.

In my judgment however, the alternative construction, by which an otherwise relevant currency conversion option does not fall out of view merely because it has lapsed by the date of disposal, is both an available construction, and the construction which ought to be preferred. My reasons follow. First, it plainly avoids the glaring anomaly, and no counter-mischief has been suggested. Secondly it does so by bringing the (b) condition into conformity with the rest of section 117, by preventing the lapse of a relevant currency conversion option from causing a non-transactional change of status. Thirdly I find it impossible to believe that the draftsman who framed the (b) condition, or Parliament when it passed it, consciously intended to introduce a propensity for non-transactional change of status, simply by using the phrases “is expressed” and “in respect of which no provision is made”, and by leaving out the phrase “at all times”. No conceivable purpose can have existed for introducing by a sidewind a propensity for non-transactional changes of status, merely by requiring the chose in action to be examined on every relevant occasion to see what are its terms. At the moment of redemption, the Loan Notes in the present case contained a lapsed currency conversion provision. In my judgment the language of the (b) condition contains no deliberate pointer to the relevance or otherwise of its lapse, and may perfectly reasonably be construed as applying as much to a lapsed provision as to a currently exercisable provision or (which Mr Southern concedes) to a provision which has yet to become exercisable at the relevant date. The answer to the question which of those three different types of provision fall foul of the (b) condition is to be arrived at by consideration of the underlying consequences. The only answer which is free of anomaly is that all three types, including for present purposes a provision that has lapsed, do so.

I should add that in his reply Mr Southern pointed me to provisions in the Finance Act 1989 (which therefore followed the original drafting of what is now section 117, but which were in force by 1995) which, he submitted, showed that where Parliament wished the question to turn on the terms of issue, so as to exclude any consequence of subsequent lapse, it did so. The relevant provisions to be found at paragraphs 2(3) and (3)(1)(b) of Schedule 11 to the 1989 Act dealing with the qualifying conditions of a deep gain security. I accept that the draftsman in 1989 made the position clearer than is to be found in the 1984 draft of the (b) condition which survived to 1995 and beyond. But it by no means follows that a lack of clarity, by comparison with a substantially later provision, is a powerful indication of an intention to do the opposite, especially where, as Mr Southern was constrained to concede, Parliament cannot possibly have deliberately or consciously intended to enable a non-QCB to change its status to that of a QCB without a transaction, merely upon the lapse of a currency conversion provision.

I recognise the force of Mr Southern’s submission arising from the amendments to the QCB regime implemented with effect from November 1996. It does appear that the draftsman (and/or those instructing him/her) perceived a risk that a construction of the type advanced by Mr Harding in relation to the (b) condition might prevail, and that they dealt with the risk by an indirect route, rather than by meeting it head-on. The consequence is that, since November 1996, the propensity for a non-transactional change of status between non-QCB and QCB is expressly recognised by the amendments, and in particular by the amendment to section 132(3)(a)(ia), although it is only when read together with the contemporaneous press release that it is apparent that the relevant risk was perceived as arising from the (b) condition itself.

Nonetheless, I am not persuaded that the indirect curing of a perceived risk as to a serious anomaly after the relevant date is a sufficiently persuasive indication as to the intention of Parliament (originally in 1984) to displace the conclusion to be arrived at by a process of construction directed to avoiding glaring anomalies in the first place. In my judgment that part of the 1997 amendment was, upon a long and considered appreciation of the true construction of section 117(1)(b), simply unnecessary.

The consequence is that in my judgment the Loan Notes in the present case were not, either when issued or when redeemed, securities “in respect of which no provision is made for conversion into, or redemption in, a currency other than sterling” within the meaning of section 117(1)(b) of the Taxation of Chargeable Gains Act 1992. This appeal must therefore be dismissed.