Revenue & Customs v Mayes

These are two linked second appeals under the Taxes Management Act 1970. They relate to a tax-saving thing called SHIPS 2. It was marketed by Matrix Tax Solutions, a firm of tax advisers providing introductory services, as “A Higher-Rate Income Tax Relief Structure” and as an “innovative tax solution.” A central feature of SHIPS 2 was the purchase by a non-resident company of non-qualifying life assurance policies called AIG Bonds (“AIG” standing for the American Insurance Group) followed, very soon after, by their partial surrender and a withdrawal of funds. The principal point in dispute is whether, as a matter of law and as was held by the Special Commissioner, those pre-ordained, composite, artificial and tax-motivated events in SHIPS 2 are to be disregarded for fiscal purposes.

The final destination of the voyage in SHIPS 2 was the creation, by a programme of planned events, of an amount of “corresponding deficiency relief” i.e. a loss. The sales literature of SHIPS 2 targeted taxpayers resident in the United Kingdom having high earnings or significant investment income. They were told that they could use the corresponding deficiency relief as loss, pay less income tax and claim capital gains tax (CGT) loss relief.

In both appeals the broad question is whether SHIPS 2 accomplished the tax consequences that its authors and operatives set out to achieve. In one appeal the issue is the availability of a deductible loss (corresponding deficiency relief) for income tax purposes. In the other appeal the issue is the proper amount of a payment deductible for CGT purposes.

The Special Commissioner (Dr David Williams) decided the case at first instance on 15 December 2008. He heard an appeal from a Notice of Assessment issued on 25 January 2007 in respect of the 2003/2004 tax year. Her Majesty’s Revenue and Customs (HMRC) won that appeal.

Mr David Mayes, a participant in SHIPS 2 and the lead applicant in litigation funded by 70 participants with a similar interest, appealed. At the first level of appeal on 8 October 2009 Proudman J decided that Mr Mayes won the corresponding deficiency relief issue, but that, against the wishes of Mr Mayes, the CGT issue should be remitted for further facts to be found at first instance.

A second level appeal is now brought to this court by HMRC with permission granted by Lloyd LJ on 17 December 2009. Permission was granted on the ground that the issues raised by HMRC on the interpretation of the Chapter II of Part XIII of the Income & Corporation Taxes Act 1988 (ICTA) relating to the taxation of life policies are of sufficient general importance and show sufficient prospects of success to warrant a second appeal.

Mr Mayes has brought his second appeal from the part of Proudman J’s order that remitted the case to the First-tier Tribunal (Tax Chamber) (now exercising the jurisdiction formerly in the Special Commissioner) for the determination of the facts relevant to the CGT issue. It concerns the amount paid by him for the assignment of life assurance policies under a Deed of Assignment dated 18 December 2003. The CGT question is this: for the purpose of determining, for CGT purposes, the loss suffered by Mr Mayes, were the sums paid by him under the Deed of Assignment consideration given by him wholly and exclusively for the acquisition of the life assurance policies (marketed as AIG Bonds) and therefore deductible from the surrender proceeds of the policies receivable by him? Mr Mayes’ primary case is that the whole of the consideration paid by him is deductible for CGT purposes. HMRC say that he was paying not for the Bonds, but for the tax benefits of his passage on SHIPS 2.

Mr Mayes was refused permission by Lloyd LJ on the ground that the CGT issue was a factual one that did not warrant a second appeal. Permission was granted by Sir Scott Baker at the oral hearing of a renewed application on 29 January 2010.

As the different outcomes before the Special Commissioner and in the Chancery Division might suggest, this case is not a simple one, even for the tax experts. The principal difficulty is in how judicial pronouncements on statutory construction, cast in very general terms and developed by the highest courts over the last 30 years in the context of pre-ordained, self-cancelling, composite, artificially structured transactions set up solely for tax avoidance purposes, should be applied to the particular facts of this case. An unusual feature of this case is that the series of pre-planned steps, which HMRC say include artificial, unreal and uncommercial tax-avoidance insertions, are built on or structured around a set of provisions in ICTA for the taxation of life assurance policies: and those provisions operate on a basis and in a manner that can fairly be described as artificial, unreal and uncommercial.

For reasons that will become apparent it is unlikely that this test case will be finally resolved at this level of appeal.

Mr Mayes is a UK resident. He is one of a group of participants in SHIPS 2. The only object of SHIPS 2 was to minimise the income tax liabilities of the participants as higher-rate taxpayers and their liabilities for CGT. The transactions were structured in a seven point programme, which will be explained in more detail below. Some of the steps were self-cancelling. The steps to the creation of deductible losses included the payment of initial premiums for life assurance policies; the payment of additional “top-up” premiums by a non-resident company; followed, in less than a month, by a partial surrender of the policies by that same company, thereby creating a potential for relief without triggering a charge to tax; later followed by a full surrender of the policies by the individual investor, in this case Mr Mayes, who then claimed that he was entitled to make deductions for tax reduction purposes.

Put more technically SHIPS 2 was designed to produce (a) allowable deductions from total income in the form of corresponding deficiency relief resulting on the disposal of the policies within the meaning of s.549 ICTA; and (b) capital sums allowable as deductions in the computation of gains for the purpose of CGT under s.38 of the Taxation of Chargeable Gains Act 1992 (TCGA).

In a luminous judgment Proudman J identified what she described as “the instinctive reaction” to the scheme (i.e. the likely reaction of almost every taxpayer except the participants in SHIPS 2). She said :-

HMRC persuaded the Special Commissioner that SHIPS 2 did not succeed. (If it did, about £24m would be lost in tax, though that fact is irrelevant to the point of law.) They appeal from the reversal of fortune before Proudman J. They say that their appeal raises a point of general principle: what is the current approach of the courts to the interpretation of fiscal legislation and its application to tax avoidance schemes? The intricate ICTA provisions, around which tax-planning technologists have ingeniously crafted the seven pre-planned steps in SHIPS 2, are of the kind that arouse reactions against the tax planning industry and against the tax laws upon which it depends. It seems to be a regular feature of structures like SHIPS 2 that, although they are linked to difficult and complicated legislation, they are based on remarkably simple ideas.

One reaction, based more on academic reflection rather than on gut reaction, is that “…neither justice nor reason has any place in tax law”, which “..more than any other branch of municipal law…is open to the reproach of being utterly incomprehensible by the individuals affected, and even more frequently by their legal advisers”: The Oxford Companion to Law by Professor David M Walker QC (1980) at pp 1207-8. Another passage in the same work lambasts “…the enormous volume and constantly changing detail of the chaotic and largely incomprehensible body of verbiage called the law of taxation.”

For those who have to interpret and apply the law, Philippics and denunciations of the tax system in general, or of a particular product of tax technology, are not of much practical use. On the journey through unfamiliar ICTA and SHIPS 2 terrain it is more re-assuring to keep company with clear-thinking, objective and careful expert guides, such as Mr David Ewart QC for HMRC and Mr Michael Furness QC for Mr Mayes, than to be carried away by emotive expressions of intelligent indignation.

HMRC are charged with operating a complex system for collecting tax in accordance with the law. When they are in dispute with taxpayers (and with non-taxpayers) the constitutional duty of the tribunals and the courts is plain: to construe the language of the legislation in accordance with the principles of statutory interpretation, to analyse the transactions in question in accordance with the applicable general law and to apply the correct construction of the legislation to them.

The relevant transactions may, for forensic purposes, be rudely labelled as “schemes” or “devices” or “dodges”, or be analysed less crudely as “circular” or “self-cancelling” or “pre-ordained” or “artificial”, or be paraded in the more presentable garb of legitimate “tax efficient arrangements.” However the transactions are labelled, analysed or presented, the question that the court has to decide in the contest between the state and the citizen is mainly mundane: do the tax-shy transactions actually succeed in reducing the size of the tax bill under appeal?

The answer to that question depends entirely on the language of the legislation, properly construed according to its purpose and context, and its application to the transactions, properly analysed according to their terms and context. Ribeiro PJ neatly extracted the essence of the legal techniques in Collector of Stamp Revenue v Arrowtown Assets Ltd [2003] HKCFA 66 at [35]:

If the taxpayer succeeds and HMRC and Parliament do not like the result, the law can be re-adjusted for the future in a Finance Act preceded by public debate and passed by democratic legislative processes. Even if the courts do not like the result, they have no means at their disposal to amend a law enacted by Parliament. Their sole function is to decide the case on their best understanding of the relevant transactions and the applicable law, whatever that may be. Whether or not the courts approve of the outcome is beside the point. It is not for judges to shoulder the law-making responsibilities of Parliament.

The contest in the HMRC appeal is on the courts’ approach to the construction and application of tax legislation following the leading case of WT Ramsay Ltd v. IRC [1982] AC 300, [1981] STC 174 (Ramsay) in which the “new approach” was first applied. The House of Lords have been round the Ramsay track a number of times in the last few decades. The cases cited to Proudman J and in this court cover a range of formulations of principle and of applications illustrating the general scope of the Ramsay principle: Furniss v Dawson [1984] AC 474, [1984] STC 153; Craven v White [1988] STC 476; Ensign Tankers (Leasing) Limited v Stokes [1992] 1 AC 655; IRC v McGuckian [1997] 1 WLR 991, [1997] STC 908; Frankland v IRC [1997] STC 1450; MacNiven (Inspector of Taxes) v Westmoreland Investments Limited [2001] UKHL 6, [2003] 1 AC 311, [2001] STC 237; Campbell v IRC [2004] STC 396;Carreras Group Stamp Commissioner [2004] STC 1377; [2004] UKPC 16; IRC v. Scottish Provident Institution (2004) 76 TC 538; and Barclays Mercantile v Mawson [2004] UKHL 51, [2005] STC 1.

In Carreras Group Ltd v. Stamp Commissioner [2004] STC 1377 (at paragraph 18) Lord Hoffmann compressed the true significance of Ramsay into a single paragraph:-

The polarised positions of Mr Mayes and HMRC in this case reflect the distinctions in that précis between the separate legal existence, validity and effects of the individual steps into which transactions have been divided, and a possible requirement of the potentially applicable tax legislation that the sub-divided individual steps be properly characterised as a single composite transaction, which does not, in law, have the fiscal consequences that the division into separate steps was intended to achieve, so that those steps can, for tax purposes, be ignored.

Thus, the case for Mr Mayes is that the relevant provisions in ICTA in ss. 539-554, as at the relevant date, should be applied to each of the legally valid individual steps relating to the AIG Bonds. In particular, they should be applied to the separate legal events of the payment of top-up premiums for the policies and of the partial surrender of the policies, thereby producing the fiscally favourably consequences that were the raison d’etre of SHIPS 2.

Against that, the case for HMRC is that the top-up premiums and the partial surrenders in SHIPS 2 had a real unity, which, solely for tax avoidance purposes, was denied when the participants (or rather those who designed SHIPS 2 for sale to the participants) artificially broke up the unity into single steps. The step-by-step approach was devised as a programmed means of artificially generating the corresponding deficiency relief claimed under ICTA. In applying ICTA, the court should treat those self-cancelling steps, though legally genuine, as fiscally futile. If they were disregarded for tax purposes, the end-result was that no corresponding deficiency relief was available to Mr Mayes.

I turn to the details of SHIPS 2, to an outline of the two appeals and to a summary of the ICTA provisions. A fuller discussion of the judgments below and of the submissions to this court on the two appeals follows to the finish.

SHIPS 2 was implemented in seven steps. All of them were pre-determined in a programme of events. Some of them were self-cancelling. Commercially they were pointless. They only had point in the weird world of tax planning. They were only undertaken to produce corresponding deficiency relief that would be available for saving tax.

HMRC accept that all the steps were legally genuine. They do not contend that the policies, the premiums or the surrenders were sham transactions. Nor do they contend that the Ramsay principle applies simply because the steps were pre-ordained and were inserted in SHIPS 2 in order to avoid tax. Their case has more legal sophistication and precision: it is that Step 3 and Step 4 of SHIPS 2 relied on for creating the corresponding deficiency relief do not fall within a purposive construction of the relevant ICTA provisions under which that relief is claimed and available. The main point of difference between the parties is how the Ramsay principle affects the construction of the particular provisions in ICTA and how the principle should be applied to the facts of this case relating to the top-up premiums paid for the policies and the pre-arranged partial surrender of the policies.

The seven steps were set out in a Statement of Agreed Facts and in an agreed summary in the judgment under appeal. They were amplified in submissions and explained in argument by reference to a visual aid in the form of a simple annotated diagram handed up at the hearing. (It would have been even more useful if it had accompanied the skeleton arguments as a visual aid.)

The legal arguments on the following seven steps are directed to the central issue of whether, for the purposes of ICTA, the payment at Step 3 was a premium which could be disregarded for fiscal purposes and whether the repayment at Step 4 was a partial surrender of a life assurance policy which could also be disregarded as another commercially pointless, off-shore event amounting to a fiscal nothing. If they were steps to be taken into account for a computation of a corresponding deficiency, Mr Mayes’ case is that the computation following the surrender of his policies at Step 7 is correct in showing a substantial “corresponding deficiency” for the purpose of relief under s.549 ICTA.

Step 1: 2 April 2002.

Step 2: 6 March 2003

Step 3:7 March 2003

Step 4: 31 March 2003

Step 5: 6 November 2003

Step 6:18 December 2003

Step 7: 13 January 2004

The central issue in this appeal by HMRC is whether, within the meaning of ICTA and for fiscal purposes, the payment made by JSI to AIG at Step 3 was a premium and the repayment of the same sum by AIG to JSI at Step 4 was a partial surrender of a life assurance policy. If they were, the computation of loss submitted by Mr Mayes following the surrender of his policies correctly showed a substantial “corresponding deficiency” for the purpose of s. 549.

HMRC’s case is that, on the application of the Ramsay principle of construction to the relevant ICTA provisions, Step 3 and Step 4 are to be ignored. They submit that those two Steps were a single, wholly self-cancelling, pre-ordained transaction for tax avoidance purposes having no commercial purpose whatsoever. In the result the payment at Step 3 and the repayment at Step 4 simply do not count on the true construction of the ICTA provisions. HMRC contends that the Ramsay principle is available to neutralise in this way the fiscal consequences sought in pre-ordained tax avoidance schemes.

Against that, Mr Mayes says that the terms “premium” and “surrender”, as used in the ICTA provisions, have their ordinary everyday meaning and effect. The final surrender of his policies on 13 February 2004 thus gave rise to a deficiency within s 549 ICTA entitling Mr Mayes to a deduction of £1,876,134 against any other income which fell to be taxed at the higher rate. The deficiency arose due to a chargeable event at the end of the policy year in which Step 3 and Step 4 occurred as separate transactions of payment of premiums and partial surrender of rights.

As for Ramsay Mr Mayes says that it is an aid to the construction of tax legislation, such as ICTA. He says that it is not an overriding doctrine that destroys the fiscal effects of legal steps in SHIPS 2 contrary to the proper construction of ICTA. He says that in this case there is only one possible construction of the ICTA provisions and that it leads to the result for which he argues.

On Mr Mayes’ appeal the steps relevant to the CGT issue are Step 6 and Step 7 i.e. the Deed of Assignment of the bonds to Mr Mayes for value and the subsequent surrender of the bonds by Mr Mayes to AIG Life. The issue is about the amount of relief that Mr Mayes is able to claim against gains, which are liable to CGT, for losses sustained on the disposal of the policies in 2003-2004.

Mr Mayes submits that the whole of the consideration paid by him under the Deed of Assignment (at Step 6 above) was deductible from the surrender proceeds of the Bonds receivable by him when determining the amount of any loss suffered by him for CGT purposes on that surrender; alternatively, that all of that consideration, other than that paid direct to Matrix Tax Solutions Ltd, should be so deductible. He contends that there is no need to remit the matters, as Proudman J directed, because no additional findings of fact need to be made. Reliance is placed on a general principle that the consideration is, for CGT purposes, determined in the light of the parties’ agreement and that, where the parties to the relevant transaction agree in good faith on the amount of the consideration, that is conclusive for the purposes of CGT.

Against that, HMRC submit that Mr Mayes paid £133,104.20 for policies with a surrender value of £1780.94 and that the latter sum is what he paid exclusively for the policies, the rest being paid for the tax-saving benefits of SHIPS 2. The matter has to be remitted as Proudman J ordered, as the Special Commissioner did not quantify the amount of loss that could be claimed for CGT purposes.

Mr Michael Furness QC, appearing for Mr Mayes, emphasises (and Proudman J accepted) the special nature of the regime imposed by ICTA for taxing, at the higher rate to income tax, investment in life policies by reference to chargeable events and gains. The main point in his preliminary comments on the nature of the special tax regime, in which gains were identified and a tax on gains was imposed, is its capability for producing results counter to commercial reality. The legislation could work in an arbitrary way unrelated to commercial gains and losses. Mr Furness describes it as a highly prescriptive way of exacting tax on the basis of a formulaic arithmetical approach to transactions.

Mr Furness also submits that, in the context of the Ramsay principle as an aid to construction, the terms “premium” and “surrender” in the legislation bear the meaning and have the actual consequences that they ordinarily have. The language of the legislation is paramount. It prevails over the purpose for which those particular steps were inserted in the scheme. Step 3 counts as a real premium paid for a real policy. Step 4 counts as a real partial surrender of a real policy, even though the insertion of the premium and the surrender has been made for tax avoidance purposes as self-cancelling steps in a pre-planned scheme. He contrasts this case with the context in Ramsay in which the key word to be construed in the legislation was “loss” which is capable of a real commercial connotation.

Before turning to specific provisions I should mention that on various points Mr Furness deploys in this court an impressive range of examples to illustrate the operation of the legislation. I shall not set out in this judgment the examples that he gives, or the answers that he proposes, or attempt answers of my own. As the judgment of Proudman J shows, well chosen concrete examples can provide an effective illustration of a particular point. In general, however, it is unwise for either party and, least of all, for this court to speculate too much about the way in which the law of taxation would operate in cases that are not before it. The facts of examples are not always adequate for the purpose of giving informed answers. The examples selected are sometimes slanted to vindicate the advocate’s viewpoint: and there is a real risk that judicial attempts at an answer will clutter the case law with embarrassing and ill-considered obiter dicta. That would do no good to the taxpayer, the Revenue authorities or the administration of justice when courts of the future have to wrestle with a problem presented by the actual facts of real case.

I turn to consider the overall scheme of the legislation. Proudman J’s summary of it has not been criticised. I gratefully adopt it. It will not be necessary to quote all the statutory provisions verbatim.

It is provided by the introductory s.539 that Chapter II has effect

The provisions that follow are a code for identifying and quantifying gains on life policies and for subjecting the gains to tax. As Proudman J observed, the gains to be taxed are gains attributed (“treated”) by the legislation rather than real ones.

The definitions in s.539(3) do not include “premium” or “surrender” as used in this group of sections. They bear their ordinary and natural meaning unless the context indicates otherwise. Section 540 defines the “chargeable events” in relation to a policy of life assurance. In the case, as here, of a policy that is not a qualifying policy, a partial surrender of the policy is a chargeable event by reason of s. 540(1)(a)(v) providing for “an excess of the reckonable aggregate value” over the “allowable aggregate amount” mentioned in s. 546 (2) and (3) respectively. (Section 546 provides for the calculation of the specified values at the end of each year.)

Section 541(1) provides that, on the happening of a chargeable event in relation to any policy of life insurance, there shall be “treated as a gain” arising in connection with the policy a specified amount depending on the event. Thus, in the event of the surrender in whole of the rights thereby conferred, the gain attributed is the excess (if any) of the amount or value of the sum payable or other benefits arising by reason of the event, plus the amount or value of any relevant capital payments, over the sum of the total amount previously paid under the policy by way of premium and the total amount treated as a gain by virtue of paragraph (d) of s541(1) on the previous happening of chargeable events. Paragraph (d) refers to the event of the occurrence of such excess as is mentioned in s. 540(1)(a)(v) i.e. the case of partial surrender of a policy.

Section 547 relates to the method of charging a gain to tax. It provides that, where a policy is held beneficially by an individual and a gain is to be treated as arising in connection with any policy, that gain forms part of the individual’s total income for the year in which the event happened. The income tax charge operates at the higher rate.

Section 549 deals with the circumstances in which certain deficiencies are allowed as deductions, so that a loss can correspondingly be set off against taxable income. Where an excess would be treated as a gain in connection with a policy and would form part of an individual’s total income for the year of assessment in which the final year ends, “a corresponding deficiency” occurring at the end of the final year is allowable as a deduction from his total income for that year of assessment, so far as it would not exceed the total amount treated as a gain by virtue of s. 541(1) (d).

Thus if, where a policy comes to an end by surrender, the statutory computation produces a loss rather than a gain, it is deductible. If sums are realised in respect of the rights in the policy during the life of the policy, the taxpayer may deduct a proportion of the premiums from the sum so realised when computing the gain on which he is to be taxable. The amount of a corresponding deficiency is limited to the amount of the gains previously certified under s.541(1)(d) arising on partial surrenders, which are chargeable events as defined in s. 540(1).

“Chargeable event” is a key concept. This case is concerned with non-qualifying policies. A partial surrender of a non-qualifying policy gives rise to a chargeable event at the end of the year of assessment where “the reckonable aggregate value” exceeds “allowable aggregate amount” those expressions being defined in s. 546. The legislation operates in a way that encourages the retention of life policies as long term investments. If a large part of the value of a policy is surrendered in the early years, disproportionately large gains will be attributed.

In his skeleton argument Mr Furness analyses, in the light of the statutory provisions, the actual events in SHIPS 2 under appeal as follows:-

If HMRC fail on their principal point that Step 3 and Step 4 are to be disregarded for tax purposes, I do not understand them to challenge that analysis of the way in which the relevant ICTA provisions work.

On the operation of the above provisions the judge said:-

Proudman J explained how the operation of the legislation made it difficult to give it a purposive commercial construction, which would avoid arbitrary or unfair results, or demarcate a line between legitimate avoidance and illegitimate avoidance on the basis of a pre-ordained transaction. Thus-

In dismissing the appeal relating to the amount of corresponding deficiency relief the Special Commissioner concluded as follows in a key passage:-

Having summarised the facts and the issues and the rival submissions on the legislation and having considered the authorities the judge concluded that the appeal of Mr Mayes on corresponding deficiency relief should be allowed.

The judge said that this legislation, proceeding on a formulaic approach without an overriding purpose that some types of transaction may not count, did not include terms (such as “loss” in Ramsay) capable of being construed as commercial concepts. The terms “premium” and “surrender”, which were not defined in the legislation, bore their ordinary meaning. On that approach, Step 3 and Step 4 were respectively a payment of a premium and a partial surrender of the policies, not just for life insurance purposes but for all purposes, including fiscal purposes. The tax avoidance purpose and self-cancelling nature of the steps did not by themselves entitle the court to disregard the steps, when there was nothing in the ICTA provisions indicating or contemplating that, as a matter of construction, such steps were not to count.

Proudman J stated her final conclusions as follows:-

It is submitted by HMRC that the Special Commissioner made no error of law in his decision. Particular emphasis is placed on his detailed findings of fact from which there is no appeal. Their case is that he rightly concluded that Step 3 and Step 4, which were pre-planned to be both temporary and temporally proximate, formed a single composite transaction to which the ICTA provisions could have no application. They were not separate transactions to be looked at individually. The single transaction was wholly self-cancelling; it was without any real investment, insurance or other commercial purpose or enduring effect: it had only a tax purpose, which involved producing a figure for use by the participating passengers in SHIPS 2. They were not, on a proper analysis, transactions capable of generating a chargeable event within the meaning of ICTA: it is as if Step 3 and Step 4 had wiped each other, had never happened and had no legal effect.

The Special Commissioner’s findings of fact cannot be re-opened. On the facts found no chargeable event was generated. Mr Ewart submits that Proudman J ought to have viewed the transaction realistically and interpreted the legislation purposively. She should have disregarded the additional payment of premiums and the partial surrender constituted by Step 3 and Step 4 or to deprive them of significance. It was not correct to look at each step individually. She was wrong to hold that the steps were at least, for some purposes, real transactions with real effects: AIG might have defaulted or the life assured might have died during the currency of the scheme.

The judge ought to have properly considered Mawson, which was not a case of a self-cancelling transaction. She ought to have analysed ICTA for its meaning and held that Step 3 and Step 4 were pre-arranged and self-cancelling. Although the judge relied on MacNiven as a case of a self-cancelling transaction taxed on the basis of individual steps, it was not a self-cancelling transaction. It should have been distinguished. The original loans in that case were repaid and replaced by fresh loans on different terms: the money used by the debtor to repay the loans was not simply returned to the debtor, so that it could be said that the loans had not been discharged.

The subsequent case of Edwards v. HMRC [2009] EWCA Civ 1010 handed down the day after the judge handed down her judgment is also relied on as supporting the HMRC case.

In conclusion Mr Ewart QC submits that it is hard to imagine a more artificial transaction than Step 3 and Step 4. The money was paid to AIG by JSI on terms understood and agreed by all parties that it would be paid back to JSI almost immediately. The money was not held by AIG on their normal terms applicable to the Bonds. AIG was not prepared to pay a commercial rate of interest on the money or to take any risk on the death of the life assured during the period which it had the money. The only correct characterisation of Step 3 and Step 4 is that nothing happened. The parties were in exactly the same position after Step 3 and Step 4 as they were before it took place.

He criticises as a non sequitur the judge’s reasoning that, as Chapter II did not seek to tax real or commercial gains, it applied to individual steps. He submits that this tax legislation, like other tax legislation, is designed to apply in the real world and that in this case the movement of funds in March 2003 between JSI and AIG were not “real world” transactions. The Special Commissioner found that Step 3 and Step 4 were not independent transactions: there was a single composite transaction that was wholly self-cancelling, without any commercial purpose and achieved nothing under ICTA. It therefore had no statutory consequences under the ICTA provisions.

Mr Furness says that the Special Commissioner erroneously accepted the HMRC’s submission that Step 3 and Step 4 do not count as a premium and a partial surrender, because they are self-cancelling steps in an artificial tax avoidance scheme. That was not a correct construction of the legislation. The Ramsay principle does not allow legal events to be deprived of their legal or fiscal effects simply because they are inserted for a tax saving purpose or can be described as “unreal” or “artificial”. The Special Commissioner had failed to address the ordinary legal meaning of the legislation and the terms “premium” and “surrender” or to apply that meaning to the events.

He submits on behalf of Mr Mayes that the proper analysis of the transactions under appeal is as follows.

The payments made under the two policies at Step 3 were paid to secure further benefits under the Bonds. They are therefore premiums.

The withdrawal at Step 4 of those funds paid at Step 3 was a partial surrender of benefits under the Bonds.

The partial surrender at Step 4 gave rise to a computation under s540(1)(a)(v).

That computation produced a large gain, because the right to all the sums paid by way of premium, other than the initial premium of £5,000 were being surrendered, but only 1/20^th of the total amount of the premiums is available as a deduction against the value of those rights.

The assignment to PES generated a computation under s541(1)(c) which resulted in a loss equal to the gains made on the partial surrender.

On surrender Mr Mayes makes a calculation under s541(1)(b). That results in the gain made by JSI at Step 4 being added to the total premiums and then being deducted from the aggregate of the surrender value and the capital payments made, which are the sums returned to JSI at Step 4.

That results in a “corresponding deficiency” for Mr Mayes. It corresponds to the gain made by JSI at Step 4.

As for the impact of Ramsay Mr Furness says that it is an aid or guide to the proper construction of the legislation, which depends on the language used, the purpose of the legislation and whether the step-by-step pre-ordained transaction was to be treated under ICTA as a single composite whole. Ramsay is not an overriding doctrine of tax law under which the fiscal consequences of every transaction, which is undertaken for a tax avoidance purpose and lacking a commercial purpose, must be disregarded.

Finally, Mr Furness says that the transactions under appeal should be taxed on the basis of their actual legal consequences and not on a different basis imposed by misapplication of the Ramsay principle.

The reaction to SHIPS 2 noted by Proudman J was instinctive and initial. Instinct informed by experience plays a role in decision-making, but does not relieve the court of the duty to reach a decision that is based on a proper understanding of the meaning of the legislation and of the facts that make up the transaction. Instinct has to be checked by the processes of construing the scope of the ICTA provisions and analysing SHIPS 2.

The judge carefully construed ICTA and analysed SHIPS 2. She arrived at a conclusion contrary to her initial reaction and contrary to the decision of the Special Commissioner, to which she paid proper respect as the expert fact-finding tribunal.

My first point is a general one.

The files of authorities include almost everything of note since Ramsay. I very much doubt whether, since Mawson, it really is necessary to return each time to the base camp in Ramsay and trek through all the authorities from then on. For practical purposes, it should, in general, be possible to start from the position stated in the unanimous report of the Appellate Committee in Mawson at paragraphs 26-42 under the heading “The Ramsay principle.” Mawson was obviously meant to be a significant judicial stocktaking of the “new approach” to the construction of revenue statutes first applied in Ramsay and followed subsequent cases on the Ramsay principle. The stated aim in the report delivered by Lord Nicholls was to “achieve some clarity about basic principles” whilst recognising realistically that it is no doubt-

I would prefer to incorporate, rather than replicate, in this judgment all that follows that passage in Mawson. The House of Lords has already unanimously decided and clarified in Mawson the important point of principle raised by HMRC in this appeal i.e. the scope of application of the Ramsay principle. A summary of the key paragraphs in Mawson, which themselves are a summary of the legal position in the light of the previous case law, could not begin to do full justice to the authoritative text of the Committee’s report and might give rise to the kind of further doubts that the Appellate Committee wished to dispel by their guidance. The principle is now clearer, but, as this case shows, the difficulties in its application and the scope for different conclusions remain and are probably irremovable by any legitimate judicial process.

Basing myself on the key passages in Mawson I agree with Proudman J that the Special Commissioner erred in law. In allowing the appeal by Mr Mayes, Proudman J correctly applied the new approach to construction laid down in Ramsay and clarified in Mawson, in her construction of the scope of the ICTA provisions and in her analysis of the legal and fiscal effects of STEPS 2.

First, Ramsay did not lay down a special doctrine of revenue law striking down tax avoidance schemes on the ground that they are artificial composite transactions and that parts of them can be disregarded for fiscal purposes because they are self-cancelling and were inserted solely for tax avoidance purposes and for no commercial purpose. The Ramsay principle is the general principle of purposive and contextual construction of all legislation. ICTA is no exception and is not immune from it. That principle has displaced the more literal, blinkered and formalistic approach to revenue statutes often applied before Ramsay.

The essence of the principle applicable to this case is that the ICTA provisions on the taxation of life insurance policies are to be given a purposive construction in order to determine the nature of the transaction to which they were intended to apply. The court then has to decide whether the actual transactions (in this case Step 3 and Step 4) answer to the statutory description, having taken account of their overall effect as elements of SHIPS 2 that were intended to operate together. To put it another way along the lines stated in Mawson (paragraph 32): analyse the facts of SHIPS 2, in particular Step 3 and Step 4, and then ask whether the requirements of ICTA are satisfied for the creation of corresponding deficiency relief.

Secondly, HMRC are not, in my view, in fact submitting that there is a special doctrine. They were told in Mawson that revenue jurisprudence was not “governed by special rules of its own.” (paragraph 34). Mawson made it clear that under the Ramsay principle there were two stages in the application of the statutory provisions-a purposive construction of the statute to see, on a “close analysis”, what transaction will answer to the statutory provision and a realistic analysis of the transaction to see whether it answers to that description- and that it was wrong to elide the two stages by “sweeping generalisations” about disregarding for fiscal purposes elements that were only inserted for fiscal avoidance reasons and had no commercial purpose. As I understand the HMRC case, it is that, on the application of the Ramsay principle, as a principle of statutory construction, Step 3 and Step 4 of SHIPS 2 do not fall within ICTA provisions relied on for the creation of corresponding deficiency relief.

Thirdly, on the proper construction of the ICTA provisions for the taxation of life policies, the statutory requirements as to the transactions to which the provisions were intended to apply were far removed from the kind of case in which the focus is simply on an end result, such as a loss. In this case the all- important corresponding deficiency relief available under s. 549 was the product of real premiums paid at an earlier stage for real life policies and real surrenders made at an earlier stage. Although the corresponding deficiency was created solely to save tax, that alone does not entitle the court to disregard the fiscal consequences of payment of premium and the partial surrender which led to its creation.

Fourthly, it would be an error, which the judge did not fall into, to disregard the payment of a premium at Step 3 and the partial surrender at Step 4 simply because they were self-cancelling steps inserted for tax advantage purposes. It was right to look at the overall effect of the composite Step 3 and Step 4 in the seven step transaction in the terms of ICTA to determine whether it answered to the legislative description of the transaction or fitted the requirements of the legislation for corresponding deficiency relief. So viewed, Step 3 and Step 4 answer the description of premium and partial surrender. On the true construction of the ICTA provisions, which do not readily lend themselves to a purposive commercial construction, Step 3 was in its legal nature a premium paid to secure benefits under the Bonds and Step 4 was in its nature a withdrawal of funds in the form of a partial surrender within the meaning of those provisions. They were genuine legal events with real legal effects. The court cannot, as a matter of construction, deprive those events of their fiscal effects under ICTA because they were self-cancelling events that were commercially unreal and were inserted for a tax avoidance purpose in the pre-ordained programme that constitutes SHIPS 2. It follows that a corresponding deficiency relief is available to Mr Mayes.

I would dismiss the appeal. Proudman J was right to allow the appeal by Mr Mayes on the ground of error of law in the decision of the Special Commissioner.

The issue on the appeal by Mr Mayes is whether, for CGT purposes, he suffered an allowable loss of £131,326 on the disposal of the Bonds when he surrendered them to AIG Life on 13 February 2004 for £1780.94.

Mr Mayes had paid the sum of £133,104.20 under the Deed of Assignment on 18 December 2003. Was that sum “consideration given by him …wholly and exclusively for the acquisition of the asset [the Bonds]..” within the meaning of s.38 of TCGA? The Deed of Assignment to Mr Mayes stated on its face that “the Price” paid for the Bonds was in the sum of £133,104.20. Of that sum £86,200.75 was paid directly to the owner of the Bonds, PES LLP. £46,903.75 was paid to Matrix Tax Solutions Limited. What element of the sum paid by Mr Mayes was an amount of consideration given wholly and exclusively for the acquisition of the Bonds?

The Special Commissioner held that he was bound by the judgment of Norris J in HMRC v. Drummond [2008] EWHC 1758 on this point, as was conceded by HMRC. He said:

The Special Commissioner appears to have decided, without evidential inquiry or finding facts relevant to determination of the amount deductible, that Mr Mayes had created a capital loss in purchasing the Bonds.

Proudman J held that the Special Commissioner, having correctly identified the issue, proceeded to decide “in principle” that there should be some relief, but failed to determine the main question of the basis on which that relief should be granted. The Judge’s comment was that this part of the Special Commissioner’s decision was odd in two respects.

First, Mr Mayes was capable of suffering a CGT loss regardless of whether he also obtained a corresponding deficiency: yet the Special Commissioner appeared to think that it was appropriate to find for Mr Mayes on the CGT issue, because he had found against him on the corresponding deficiency issue. As a matter of law the outcome of the CGT issue did not depend on the outcome of the corresponding deficiency issue.

Secondly, having decided the matter of a loss “in principle”, he left the amount to be decided by agreement. He made no findings of relevant fact. Yet, as the Special Commissioner recognised, his ruling involved a question of fact as to the basis of computation of loss. A decision by him was necessary on what the purchase consideration was given for. He made no finding on that.

HMRC’s case is that, on the basis of the documentary evidence, £133,104.20 was paid by Mr Mayes for his participation in SHIPS 2 and that the Bonds were worth no more than the surrender proceeds of £1,780.94 received by Mr Mayes.

Against that, Mr Mayes argues that the whole of the price was paid for the purchase of the Bonds. That was the true construction of the documentary evidence. As no additional findings of fact were needed, remission to the Special Commissioner was unnecessary. It was simply a matter of the court construing the documents against a background of agreed facts and applying the legislation. That was a matter of law which could be decided by this court.

The judge agreed with HMRC that this was not a case in which there was only one possible decision that the Special Commissioner could properly reach on the evidence. The Special Commissioner had not determined on the evidence before him what the consideration or its constituent elements were given for. That was, in the first instance, within the remit of the Special Commissioner. The matter had to be remitted to his replacement, the First-tier Tribunal (Tax Chamber), to determine that issue.

In support of the appeal by Mr Mayes it is contended that the basic facts relevant to the CGT point were as set out in an Agreed Statement of Facts. The CGT point should be dealt with by this court here and now as part of the overall appeal. This court should quantify the loss which the Special Commissioner had not quantified. This could be done as a matter of law on basis of the documents showing the sum which Mr Mayes had paid for the Bonds and the loss suffered by him on the final surrender of the Bonds.

The consideration deductible for CGT purposes under s.38 of TCGA is “the amount of value of the consideration, in money or money’s worth…given… wholly and exclusively for the acquisition of the asset together with the incidental costs to him of the acquisition.” Mr Furness argues that it is clear from the face of the Deed of Assignment between PES LLP and Mr Mayes that he acquired the policies in consideration of the whole of the stated price of £133,104.20. For CGT purposes the whole of that price less the surrender proceeds of £1780.94 received by him would give rise to an allowable loss of £131,326. The consideration was determined by the amount agreed in good faith by the parties as the consideration for the acquisition of the Bonds. The fact that part of the price was payable to Matrix Tax Solutions and not to the owner of the Bonds did not prevent that sum from being part of the acquisition cost of the Bonds. Even if that sum represented payment of commission, that was an incidental cost of acquisition that was included in what was deductible from the surrender proceeds under s.38 TCGA. Alternatively, it is submitted that at least that part of the price paid by Mr Mayes to the owner of the Bonds was relevant consideration (£86,200.75). On that basis he would have an allowable loss of £84,422.55.

In those circumstances there is no point in having another hearing and the further costs that would be involved in remitting this question to the First-tier Tribunal for determination.

I agree with the judge, who accepted HMRC’s submissions on this point. Remission of the CGT issue is necessary for an elementary reason: to resolve an issue that ought to have been resolved by the Special Commissioner either as, or as raising, a question of fact. The Special Commissioner erred in law in not resolving the amount of the consideration wholly and exclusively paid by Mr Mayes for the acquisition of the Bonds.

HMRC contend that only the surrender value of the Bonds (£1,780.94) could be said to have been paid wholly and exclusively for the acquisition of the Bonds. The excess was a percentage of the deficiency relief that Mr Mayes thought that he would enjoy for the benefits of the tax scheme. That contention raises an issue of fact. According to the decision of this court in Drummond v. HMRC [2009] EWCA Civ 608; [2009] STC 2206 at 2217 paragraph 31 the ascertainment of the consideration on the application of the “wholly and exclusively” tests in s.38(1) (a) of the TCGA is an issue of fact for the First-tier Tribunal and not a question of law. It could not be said that a remission is unnecessary on the ground that there was only one possible factual finding that the Special Commissioner could have made on the CGT issue.

The case of Spectros International plc v Madden (Inspector of Taxes) [1997] STC 114 at 136-138 is cited by Mr Furness for the proposition that, as a matter of construction and therefore of law, where, in an arm’s length transaction, two parties agree that consideration is to be allocated in a particular way, that agreement is conclusive of what constitutes consideration for CGT purposes. He submitted that his argument based on Spectros and on the earlier case of Stanton v Drayton [1983] 1 AC 501 at 510 was not advanced in Drummond and that in this case the terms of the Deed of Assignment concluded the issue as to what the consideration was paid for.

Spectros was a case in which HMRC used the proposition formulated by Mr Furness against the taxpayer. The case related to the allocation by the parties of the consideration on the disposal of an asset. The taxpayer unsuccessfully argued that CGT consequences of the agreed terms of disposal should not be followed. This is a case of what was the amount of Mr Mayes’ expenditure for the acquisition of the policies. Mr Ewart distinguishes Spectros and also submits that this court was not bound by that first instance decision, but was bound by the decision of this court in Drummond.

In my judgment, the arguments based on the authorities cited by Mr Furness should not be ruled on by this court ahead of a remitted hearing at which the legal arguments can be considered in the full context of the evidence and the facts.

I would dismiss the appeal by Mr Mayes. The judge correctly found that the Special Commissioner’s decision on this point erred in law and that remission to the First-tier Tribunal was the appropriate response in order to decide what he had not decided. At the remitted hearing there can be full argument on both the applicable law and as to what, as a matter of fact, constituted the relevant amount of the consideration paid by Mr Mayes for the acquisition of the Bonds.

I would dismiss both the appeal and the cross appeal.

I agree with the judgment of Lord Justice Mummery which sets out with great clarity why the appeal and cross appeal have to be dismissed. However, for the reasons given by Lord Justice Toulson, my concurrence is reluctant. The higher rate taxpayers with large earnings or significant investment income who have taken advantage of the scheme have received benefits that cannot possibly have been intended and which must be paid for by other taxpayers. It must be for Parliament to consider the wider implications of the decision as it relates to the way in which revenue legislation is structured and drafted.

I also agree. On the corresponding deficiency issue I add a brief summary to explain the reason for my reluctant concurrence in a result which instinctively seems wrong, because it bears no relation to commercial reality and results in a windfall which Parliament cannot have foreseen or intended.

The root problem in this case from the viewpoint of HMRC lies in the structure of the relevant statutory scheme. As has been pointed out by Proudman J and Mummery LJ, Chapter 11 of Part XIII of ICTA 1988 creates a complex set of rules for determining when a gain is to be treated as arising in connection with a life insurance policy.

Inherent in the scheme is the possibility of a disconnection between what would be regarded as a gain on an ordinary commercial view and what is to be treated as a gain for the purposes of the statute.

In some cases a taxpayer may be liable for a gain which the statute requires him to be treated as having made, although the chargeable event giving rise to the deemed gain has not caused him to make an equivalent gain in real terms.

In the present case the opposite has occurred. The inventor of SHIPS 2 has found a clever way of making the legislative structure work to HMRC’s disadvantage by devising a series of steps giving rise to a chargeable event and a corresponding deficiency, albeit that the taxpayer was no worse off commercially.

The Ramsay principle permits a purposive approach to the construction of tax legislation. Under the legislative structure with which we are concerned, the creation of a deemed gain, or corresponding deficiency, is dependent on events involving the creation and discharge or transfer of rights under an insurance policy, in particular, the payment of premium and the occurrence of a “chargeable event” as defined in s.540 (typically, the death of an assured or the maturity, assignment or surrender of the policy). The essential nature and the purpose of the scheme is that such events should of themselves trigger the calculation of a deemed gain (or corresponding deficiency, as the case may be). The test is not whether there has been an equivalent commercial gain (or loss).

In the present case it has not been suggested that the payment of premium followed shortly by a surrender of the bonds were a sham. As Mummery LJ has said, they were legal events with legal consequences. They were events which ICTA has caused to carry fiscal consequences. The particular consequences in the present case were obviously not foreseen or intended by the legislature; but legislation, especially legislation which is highly engineered, can have unintended consequences.

Unattractive as the result is for other taxpayers and the rest of society, I agree with Proudman J and Mummery LJ that the court cannot lawfully hold, as a matter of proper construction of the statute, that because the sole purpose of steps 3 and 4 was to avoid tax by the creation of a corresponding deficiency unrelated to any underlying commercial loss, those events are therefore to be treated as if they had not occurred.