England and Wales Court of Appeal (Civil Division) Decisions

You are here: BAILII >> Databases >> England and Wales Court of Appeal (Civil Division) Decisions >> Revenue & Customs v Mayes [2011] EWCA Civ 407 (12 April 2011)
URL: http://www.bailii.org/ew/cases/EWCA/Civ/2011/407.html
Cite as: 81 TC 247, [2011] EWCA Civ 407, [2011] BTC 261, [2011] STC 1269, [2011] STI 1444

[New search] [Printable RTF version] [Help]

		Neutral Citation Number: [2011] EWCA Civ 407
		Case No: A3/2009/2450 & 2462

ON APPEAL FROM THE HIGH COURT OF JUSTICE
CHANCERY DIVISION
THE HON MRS JUSTICE PROUDMAN
CH/2009/APP/0072 & 0063

Royal Courts of Justice
Strand, London, WC2A 2LL

12/04/2011

B e f o r e :

LORD JUSTICE MUMMERY
LORD JUSTICE THOMAS
and
LORD JUSTICE TOULSON
____________________

Between:

	THE COMMISSIONERS FOR HM REVENUE & CUSTOMS	Appellant
	- and -
	DAVID MAYES	Respondent

____________________

MR DAVID EWART QC and MR DAVID YATES (instructed by HMRC Solicitor's Office) for the Appellant
MR MICHAEL FURNESS QC (instructed by McGrigors LLP) for the Respondent
Hearing dates: 28th & 29th July 2010
____________________

HTML VERSION OF JUDGMENT
____________________

Lord Justice Mummery :

Two linked appeals

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.

Background

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 :-

"45. I sympathise with the instinctive reaction that such an obvious scheme ought not to succeed…"

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]:

"…the ultimate question is whether the relevant statutory provisions, construed purposively, were intended to apply to the transaction, viewed realistically."

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:-

"(8) … But ever since [Ramsay] the courts have tended to assume that revenue statutes in particular are concerned with the characterisation of the entirety of transactions which have commercial unity rather than the individual steps into which such transactions may be divided. This approach does not deny the existence or legality of the individual steps but may deprive them of significance for the purposes of the characterisation required by the statute. This has been said so often that citation of authorities since Ramsay's case is unnecessary."

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.

Seven Steps

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.

(1) Step 1: 2 April 2002.

A Jersey resident individual, Mr Christopher Lovell, purchased from AIG Life (part of the American Insurance Group-AIG), by means of 2 single premiums of £5,000 each, 2 AIG Bonds comprising several policies on his life. It was part of a larger purchase of AIG Bonds for use in SHIPS 2. It is agreed that the premiums were paid out of borrowed funds. It is agreed that the AIG Bonds were life assurance policies, being standard form life policies issued by AIG and marketed to potential investors as an investment product under the name Premium Access Bond (PAB) invested in the Premier Access Fund (PAF) managed by AIG. Each Bond was a cluster of 20 single premium life assurance policies, the initial premium paid under each being £250. They were claimed to be an attractive alternative to cash deposits. Money invested was available on instant access. It was claimed that there were tax advantages and a death benefit.

(2) Step 2: 6 March 2003

Mr Lovell assigned the Bonds to a Luxembourg company, January Storm Investments SA (JSI), for value. The full consideration was £256,085.

(3) Step 3:7 March 2003

JSI paid a top-up premium of £375,000 to AIG Life in respect of each policy in the first Bond and £50,000 in respect of each policy in the second Bond. The total top-up premium paid in respect of the PABs was £150m. HMRC accept that the payments were premiums. They were paid out of borrowed funds.

(4) Step 4: 31 March 2003

JSI withdrew from AIG on 31 March 2003 all of the sums paid at Step 3 on 7 March 2003. It is agreed by HMRC that the repayment to JSI was a partial surrender of the policies reflecting the amount of top-up premiums paid and the investment return from the issue of the policies up to the date of partial surrender. The income comprised of over 1 year's interest accruing on the initial investment of £5,000 and over 3 weeks' interest accruing on the additional investment. After the withdrawal, £5,000 remained in each bond at the year end i.e. 5 April 2003.

(5) Step 5: 6 November 2003

JSI assigned the Bonds to PE Shirley & Co LLP (PES), a limited liability partnership registered in England, for value.

(6) Step 6:18 December 2003

Mr Mayes appeared on the scene at Step 6, having used Barclays Private Bank as his investment adviser. PES assigned the Bonds to Mr Mayes for value, £125,949 in the case of one Bond comprising 5 policies and £7,155 in the case of the other Bond comprising 2 policies.

(7) Step 7: 13 January 2004

Mr Mayes then surrendered in whole both Bonds to AIG Life. He received in return the remaining proceeds in the Bonds (£1,780.94). He then claimed income tax relief arising from the surrender in the tax year 2003-4. The amount of "corresponding deficiency relief" claimed by him was £1,876,134. He also claimed an allowable loss for CGT purposes of £131,326 on the surrender of the life policies. His case is that the deficiency arose due to a chargeable event arising at the end of the year in which Step 3 and Step 4 occurred.

Corresponding deficiency relief appeal: outline

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.

CGT appeal: outline

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.

The ICTA provisions

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

"..for the purposes of imposing…charges to tax…in respect of gains to be treated in accordance with this Chapter as arising in connection with policies of life assurance..."

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:-

"24. …At step 3…JSI made substantial payments under the two policies. Mr Mayes says that these are clearly premiums, because they are paid to secure further benefits under the Bonds. At step 4, the withdrawal of the funds paid by way of premium at step 3 is a partial surrender of benefits under the Bonds, which gives rise to a section 540(1)(a)(v) computation which produces a large gain, because the rights to all the sums paid by way of premium other than the initial premiums of £5,000 are being surrendered, but only 1/20^th of the total amount of premiums is available as a deduction against the value of those rights. The assignment to Mr Shirley generates a computation under section 541(1)(c) which results in a loss equal to the gain made on the partial surrender, but JSI cannot deduct that loss as a corresponding deficiency under section 549(1), (a) because it does not occur at the end of final year of the policy, and (b) because JSI is not an individual. The onward assignment to Mr Mayes is of no significance to Mr Mayes' tax position. On surrender Mr Mayes makes a calculation under section 541(1)(b), which 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.) This results in a corresponding deficiency for Mr Mayes which corresponds to the gain made by JSI at step 4."

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:-

" 23. The point that Mr Furness sought to draw is this. The legislation operates in an arbitrary way in the sense that the sections about assignments make no provision for successive owners of the policy to be taxed only on the gains and losses that they themselves respectively make. It is then possible for unfair tax results to arise from transactions which have no tax avoidance element. The legislation does not admit of a purposive interpretation that might ameliorate them. It shows a lack of interest in (a) attributing gains to the person who made them, (b) not attributing gains to a person who did not make them and (c) timing the taxation of the gain fairly. Instead it operates mechanically according to a series of statutory formulae.

24. Mr Ewart submitted that the overall purpose of the legislation was to ensure that the policies were subjected to the correct amount of tax over their lifetime, rather than to ensure that individual taxpayers paid an appropriate amount. Some taxpayers end up by paying more or less tax than is strictly equitable, but over the lifetime of the policy the correct amount of tax is charged.

25 However, it seems to me that Mr Ewart's purposive construction assumes that chargeable events will result in a charge to tax and that overall at the end of the day the correct amount of tax will be exacted. Neither of these assumptions necessarily holds good."

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-

" 27. If a taxpayer takes out a policy while non-resident, and effects a partial surrender while non-resident, but then becomes UK resident before effecting a total surrender, the gain made on the partial surrender falls outside the charge to tax but he is still entitled to claim corresponding deficiency relief from UK tax on the total surrender. Conversely, if he takes out the policy and effects the partial surrender while UK resident, but then because of initially unforeseen force of circumstances becomes non-resident, he pays tax on the large gain on partial surrender, but loses the benefit of the compensating corresponding deficiency on total surrender. The former situation is capable of apparently legitimate avoidance possibilities, the latter is capable of causing hardship. Both illustrate the formulaic and prescriptive nature of the legislation.

28. The statute does not tax actual losses. If the transaction had proceeded in exactly the same way but JSI had been a UK taxpayer the gain and loss would have been offset and there would and could have been no HMRC objection."

Special Commissioner's decision

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

"80. …I am satisfied that Steps 3 and 4 were inserted in the transactions for no purpose other than the avoidance of tax. If, as Mr Ewart contends, steps 3 and 4 are elided, what actually happened? Standing back from the individual Bonds and looking at the whole transaction and the common elements present, one sees that a series of companies of which JSI was one deposited some £300,000,000 with AIG Life and then encashed the entire sum a short time later. It was known by all concerned that the entire funds were to be withdrawn within a pre-defined short period of them being deposited into the Bonds. Further, the terms on which the money was held during the period were not standard terms of the Bonds but specific and uncommercial terms to the investors and required a third party subsidy to the providers of the funds to make the scheme acceptable in market terms. And the companies that paid the funds into and then in name withdrew them from AIG Life were created from the same funds as were used by those companies, once established, to purchase the Bonds prior to these Steps. Steps 3 and 4, in the real world, left the Bonds worth in commercial terms much the same after the funds had been encashed as before they had been put in. The steps did not create any other direct value to the companies undertaking those steps. The only difference is that there was, in the view of those who undertook the scheme, a tax advantage to anyone liable to higher rate United Kingdom income tax. That, of course, was not relevant in Jersey where the initial purchaser of the Bonds lived and the main funder was based, nor in Luxembourg, where JSI was based. It only became relevant when, after both Mr Lovell and JSI had ceased to have any interest in the Bonds, they were sold by JSI to an associate of someone from the United Kingdom involved in the planning from the beginning.

81. My conclusion on all the evidence is that Mr Ewart's argument is right. The payment of sums of money to AIG Life pro rata into the Bonds was formally in the name of JSI. The subsequent pre-arranged withdrawal of all those funds pro rata from those Bonds after a brief period was in the name of JSI. But in reality the investment of funds was arranged by, the encashment of funds was ordered by, and the funds came from and went directly to, SGH. Steps 3 and 4 were pre-arranged self-cancelling steps with no commercial purpose to JSI or its immediate funders other than the tax advantage that it was intended to secure in the United Kingdom under the terms of the Chapter on the onward sales of the Bonds through a third party associated with the funding operation. They followed a pre-set time table that had been discussed by all concerned. The encashment had been arranged with AIG Life before the funds went in, and was understood by AIG Life to be part of the deal under which it received the deposit.

82. On those findings, my answer to "the ultimate question" is that to apply the sections in question to Steps 3 and 4 as evidenced above, with the result that the two steps generated a "reckonable aggregate value" such as to create a "corresponding deficiency" would not be a rational application of the Chapter and is not to be accepted as intended by the legislature. The calculations required by the Chapter are properly to be undertaken without reference to the payments in by and the payments out to, JSI. That involves no distortion or forced interpretation of the language of the Chapter. Rather, it reflects both the plans and the language used by those who took part in the actual funding operation. Nor does it involve any recharacterisation of what occurred or the imposition of some economic reality in place of what actually happened. All the other steps remain as before, and this is reflected in the history of the Bonds aside from these two Steps. The answer also accepts and reflects a coherence of the drafting of, and the purpose behind, the provisions in the Chapter."

Judgment of Proudman J

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:-

"44. … I find that a purposive construction [of Chapter I, Part XIII ICTA] does not enable the Court to disregard the additional payment of premiums and the partial surrender constituted by Steps 3 and 4. This is legislation which does not seek to tax real or commercial gains. Thus it makes no sense to say that the legislation must be construed to apply to transactions by reference to their commercial substance.

45. I sympathise with the instinctive reaction that such an obvious scheme ought not to succeed. However I cannot extract from the legislation any underlying or overriding purpose enabling me to conclude that parts of the scheme may be ignored. To do so would conflate the definition of a chargeable event with the concept of an actual charge to tax and would, I believe, revert to an acceptance of the type of submission that was roundly rejected in MacNiven. I am bound by the ratio of the decision in MacNiven and in my judgment it points only one way on the facts of this case.

46. I agree with the Special Commissioner's formulation of the statutory question to be answered in the words of Ribeiro PJ, that is to say, whether the relevant statutory provisions, construed purposively, were intended to apply to the transaction viewed realistically. He further noted the fact that steps 3 and 4 were pre-arranged self-cancelling steps with no immediate advantage to JSI or its funders other than the tax advantage they were intended to secure. He drew attention to the fact that steps 3 and 4 were arranged and funded by companies in the SG Hambros Group, the architect of the scheme. He asked the question whether a corresponding deficiency can be claimed when the deficiency related back "not to an actual chargeable event but to the coordinated series of events evidenced here." That in my judgment begged the very question that was being formulated and also conflated the concept of chargeable event with the charge to tax itself. It seems to me that the Tribunal made an error of law which, despite the circumspection and due deference I accord to the determination to Dr Williams, requires me to intervene.

47. In summary it seems to me that Chapter II of the Taxes Act adopts a formulaic and prescriptive approach. No overriding principle can be extracted from the legislation, or from the authorities, that some types of transaction should be ignored in the application of the Chapter. To say that there is no premium and no partial surrender, that those steps should be ignored, is in my judgment simply to sidestep the question of construction altogether. The pre-arranged and self–cancelling nature of the transaction was no different and no more extreme that that in MacNiven."

A. HMRC APPEAL ON CORRESPONDING DEFICIENCY RELIEF

HMRC submissions

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.

Submissions of Mr Mayes

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.

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

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

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

(4) 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.

(5) 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.

(6) 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.

(7) 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.

Discussion and conclusions

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-

"27. …too much to expect that any exposition will remove all difficulties in the application of the principles because it is in the nature of questions of construction that there will be borderline cases about which people have different views."

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.

Result of HMRC appeal

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.

B. CGT APPEAL BY MR MAYES

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?

Decision of Special Commissioner

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:

"86. … An amount is allowable to Mr Mayes for the loss on the final surrender of the Bonds. For myself, I am not entirely clear how the entire amount is allowable given the terms of the final payment if there is in reality a dual purpose behind the price he paid. And I note as a matter of fact that it was neither the seller nor Mr Mayes that determined the price he paid, but a third party to the transaction. There are therefore questions of fact, not argued before me, to be decided before the exact amount of any loss can be determined. But as I now find that there was no corresponding deficiency for income tax purposes, it appears only appropriate on these facts to accept that Mr Mayes did have a capital loss in buying the Bonds and to enquire into the detail no further at this stage. Accordingly, I allow that aspect of the appeal in principle. Again, I give liberty to apply if final agreement is not reached on this point."