Claim of the rose

IT'S A TRUTH universally acknowledged that a woman in possession of a good income must be in want of a tax avoidance scheme. Philippa D'Arcy certainly was. Better known by her maiden name as Philippa Rose, she founded her executive search company, the Rose Partnership, in 1981 at the precociously young age of 22, and it is now recognised as probably the leading headhunter firm for City high-fliers. According to the Evening Standard 'City Rich List', during 2001-02 the highest paid director of the company (presumably her) took a £2.1 million salary, which helped to boost her estimated worth to £20 million, putting her 50th equal in the list with Nicola Horlick.

On 10 January 2002 Mrs D'Arcy received a letter from her tax adviser, Mr Philip Shirley, setting out the details of a gilt 'repo' scheme. The scheme was a new version of the old accrued income scheme, and depended on the making of a tax-free capital gain between the sale cum div and the purchase ex div of the same gilts, together with the deductibility for income tax purposes of the manufactured interest payment.

Scheme details

This is the way the scheme worked — I have simplified the numbers a little to make it easier to follow. The transactions were carried out through a broker, NCL, on an execution-only basis.

On 13 February 2002, acting on Mrs D'Arcy's behalf, NCL became the temporary holder of £33.5 million of gilts from Royal Bank of Scotland (RBS), an independent third party. It did this under a 'repo' agreement by agreeing to buy the gilts for settlement on 14 February and to sell them back on 20 February for almost the same price. Because the company went ex div during this period, NCL would have to pay £1.5 million to RBS as 'manufactured interest'. If this had been the whole transaction, Mrs D'Arcy would have got her deduction for a payment of manufactured interest, but would have been assessed to income tax on the interest which she would have received because she held the securities before they went ex div. This would also have been about £1.5 million.

However, at the same time NCL sold the securities on 13 February for £33.3 million (including £1.5 million of accrued interest) for settlement on 14 February, and repurchased them on 20 February 2002 for £31.8 million, ex div and including £50,000 of accrued interest. The repurchase was not agreed in advance, and both these transactions were with the best offer available on the market. It happened that both sale and purchase were with JP Morgan Stanley ('JPMS') but this was not a requirement of the scheme. The resultant movement of the gilts is shown in the box below.

The position now is that, instead of receiving interest of £1.5 million, Mrs D'Arcy has — in commercial terms — made a capital gain of £1.5 million. This, however, is precisely why the accrued income provisions of TA 1988, ss 710-728 were introduced; to tax this capital gain (which would otherwise be exempt) as if it was income. As will be seen later, Mrs D'Arcy had an intriguing argument why the provisions should not apply in her case.

Getting closure

The Revenue opened an enquiry into the return for 2001-02. No details are given in the decision of when it started, but a closure notice was issued on 15 February 2005, which concluded that the disposal, acquisition and repo over the gilts 'should be regarded as a composite whole which was circular and self-cancelling'. As a result, under the Ramsay principle, the claim for a deduction of the £1.5 million manufactured interest was disallowed, and an amendment made by HMRC to the self assessment. It was the appeal against both the conclusion and the corresponding amendment (under TMA 1970, s 31(1)(b)) that was heard by the Special Commissioner.

The first issue raised was a procedural one based on the closure notice. In their skeleton argument, HMRC said they no longer wanted to run the Ramsay argument of circularity, because they now appreciated that the sale and repurchase being through JPMS was coincidental. Instead they wanted to claim that the provisions relied on by the taxpayer to escape the accrued income scheme did not allow her to do so.

For the taxpayer, it was argued that where an appeal was against a conclusion, the subject matter of the appeal was that conclusion, and it was not possible for HMRC to introduce a new argument, nor for the Commissioner to determine the correct amount of tax due from scratch. The power in TMA 1970, s 50(7) to increase or decrease an assessment should be seen as a power to amend the assessment to give effect to the conclusion as agreed or varied. If not, it negated the structure of the legislation, which brought the enquiry to a formal close with the issue of the closure notice.

HMRC, on the other hand, argued that the power to amend assessments was the same as it always had been, and that there was good precedent for the right to raise a new issue on appeal. Any problem with a taxpayer being 'ambushed' with unexpected arguments should be dealt with through case management by the Special Commissioner. The power to raise new arguments was particularly necessary now that it was not possible for HMRC to raise alternative assessments, as had been common prior to self assessment.

Procedural decision

The Special Commissioner, Dr John Avery Jones, said that the provisions for amending assessments in s 50, the key wording of which dates from before self assessment, did not match up very well with the new provisions in s 31(1) specifying what could be appealed against. However, since the self assessment is in relation to the total income of the taxpayer, the logical conclusion of HMRC's argument was that if they had issued a closure notice about a trading profit which was then appealed by the taxpayer, they would be able to raise completely different issues in the appeal — for example about rental income.

He therefore concluded that, when the appeal is against HMRC's conclusion at the end of an enquiry, the appeal is against what is set out in the closure notice, and the Commissioners retain their broad power to vary or confirm the assessment, but only in relation to the issues raised by the conclusion. In this case, the conclusion related to the purchase and sale of the gilts, and therefore new issues could be introduced in relation to those transactions, but could not have been in relation to different transactions.

In passing, Dr Avery Jones said that he saw no reason why alternative amendments should not be raised in a conclusion, with the amendment giving effect to the one HMRC preferred. So, for example, a conclusion could be stated in a closure notice specifying that the taxpayer was liable either to income tax or capital gains tax on a particular sum, and that HMRC believed the former to be the case and had amended the assessment accordingly.

The limitations which the closure notice applies to the arguments that can subsequently be raised only re-emphasises the problem with the latest 'interventions', since they have no similar safeguards. However, the idea that HMRC can reach a conclusion that is not a conclusion at all, but is instead several alternative conclusions of which they prefer a particular one, does not fit well into the intended finality of the self assessment code. In particular, it is to be hoped that HMRC do not try to take advantage of the suggested opportunity to issue alternative conclusions by adding a catch-all about 'I therefore conclude that your taxable income for the year is £X' to all conclusions in an attempt to achieve the right to raise arguments about any and all entries on the return during the appeal even if they have not been the subject of the enquiry.

The scheme

The basis of the scheme was the exemption in s 715(1)(b), which applies if 'on no day … the nominal value of securities held by him exceeded £5,000'.

Clearly the gilts involved here were well in excess of £5,000, but the argument turned on the definition of 'held' in s 710(7):

'A person holds securities:

a) at a particular time if he is entitled to them at that time;

b) on a day if he is entitled to them throughout the day or he becomes and does not cease to be entitled to them on the day'.

A combination of subsections 6 and 8 also provides that the acquisition, or entitlement, takes place on the day of the contract, not the day of settlement. So, contrary to the 'real' movement shown in the previous box, the deemed dates of the transfer are all 13 February except for the purchase from JPMS which takes place on 20 February.

The taxpayer's argument was that, under these deemed provisions, there is no day when she is entitled to gilts throughout the day — the acquisition on the first leg of the repo on 13 February is matched with the disposal on the same day to JPMS, and the gilts purchased from JPMS on 20 February have already been 'sold' on 13 February. Alternatively, looking at it from a common-sense point of view, the gilts bought from RBS under the repo on 13 February for settlement on 14 February were sold on the same day, and the gilts bought on 20 February from JPMS were sold on the same day under the second part of the repo. The accrued income rules therefore should not apply, and the gain remains an exempt capital gain. It should be noted that the effect of this now would be to deny relief for the manufactured interest deduction, which can only be offset against manufactured interest receipts or accrued interest — the rules appear to have been altered in response to this scheme.

HMRC put forward two arguments to attack the scheme. The first was purposive — s 715 was intended to prevent holders of small amounts of securities from being charged under the accrued income scheme, not to exempt large dealings within the course of a day. Alternatively, on a close reading of the legislation, Mrs D'Arcy did not 'become and not cease to be entitled to' the gilts on 20 February, as required by s 710 (7) because she had already ceased to be entitled to them on 13 February.

It was perhaps a tactical error to try and run a legalistic interpretation at the same time as a purposive one, although it is easy to understand, in the current climate of a developing principle, why it seems best to cover all the bases. Dr Avery Jones applied, in effect, a purposive interpretation to s 710(7) and looked at the 'reality' that Mrs D'Arcy was not entitled to any gilts at the end of 20 February, so could not be said to have 'become and not ceased to be entitled to' them on that day.

Purposive construction

That left the main purposive interpretation argument to be answered. This was dealt with fairly quickly by Dr Avery Jones. The provision used the concept of holding securities 'on a day', and Mrs D'Arcy did not do so, according to the definition. He went on to say:

'If the draftsman had been concerned about dealings of larger amounts in the course of a day he would not have used the concept of entitlement on a day. But there was no reason for him to have been concerned with dealings during a day because no income accrues during a day'.

With respect to Dr Avery Jones, I am not sure this is the right way to formulate the purposive problem. In interpreting a relieving provision, surely the question to ask is 'what transactions are intended to be relieved?' The answer would be 'transactions undertaken by holders of small amounts of stock'. That said, it would appear that the use of 'on any day' in the provision is a drafting error, and the better provision to have used would have been 'at any time'. Interestingly, in a subsidiary argument, Dr Avery-Jones looks at Simon's Taxes to decide what the purpose of a section is. Had he used the same source here, in paragraph A.7506, s 715(1)(b) is stated as applying when the nominal value 'does not exceed £5,000 at any time' during the relevant period, not 'on any day'.

Another purposive approach would be to consider the purpose of the accrued income legislation as a whole. The purpose is clearly to catch large transactions where an exempt capital gain is achieved by selling cum div and repurchasing ex div. That said, the legislation still has to be capable of interpretation in line with this purpose. It would be possible, though not entirely easy, to argue that the concept of a 'day' as defined in s 710 was only for the purposes of counting days in order to do the calculations of accrued interest.

Circularity

Perhaps a better purposive approach is to run the Ramsay argument against the deductibility of the manufactured payment which HMRC backed away from. They appear to have backed away because it was a coincidence that JPMS were on both sides of the intermediate purchase and sale, but the intention was always to deal in the market at the best price available as part of a circular transaction that would have no lasting consequences outside tax.

In Barclays Mercantile v Mawson [2005] STC 1 the judgment quotes with approval the following statement from W T Ramsay [1981] STC 174 at 182.

'To say that a loss (or gain) which appears to arise at one stage in an indivisible process, and which is intended to be and is cancelled out by a later stage, so that at the end of what was bought as, and planned as, a single continuous operation is not such a loss (or gain) as the legislation is dealing with, is in my opinion well, and indeed essentially, within the judicial function.'

It is surely possible to argue that the tax relief for manufactured interest payments was meant to operate in a real world where a taxable interest payment had been received that did not, in economic terms, 'belong' to the recipient, and which had to be paid over in the form of manufactured interest. The deductibility of the payment was intended to match the taxability of the receipt. Artificially creating a manufactured interest payment and then artificially matching it with an exempt gain, in a transaction which collapses to nothing within seven days, is not what the legislation was designed for.

Unacceptable avoidance?

However, if the reasoning in the decision is right, where does that leave us? If the scheme had not been stopped, there would only have been two things preventing every taxpayer in the country from completely obliterating their income tax liability. The first would have been the availability of gilts to repo around the ex div date, and the second would have been the costs of the scheme. Put bluntly, taxpayers with large enough incomes could pay a fee to cancel their tax liability, but taxpayers with 'normal' incomes would find it uneconomic.

Obviously the loophole was stopped as soon as it was identified, and the introduction of the disclosure of tax avoidance scheme (DOTAS) revisions will make the identification and closure of schemes faster, and hence less profitable for advisers. But on the other hand, is it really sensible, or indeed defensible, to have a tax system which allows the expenditure of so much energy on the creation and then the combating of entirely artificial 'loss-buying' schemes?

There is surely a very significant difference between setting up a transaction, a business or an investment in the most tax-efficient way possible, and the creation of a completely artificial tax-avoidance transaction either to stand alone and create a loss, or to negate the tax consequences of a real transaction. For some time, the judiciary seemed to be creating a doctrine that tried to recognise that difference, and to cope with the particular problems that tax law creates. By turning their back on this in recent cases, and insisting that the purposive principle of interpretation is 'one size fits all', they would appear to be doing the tax system a disservice.

Issue: 4069 / Categories: Comment & Analysis , Admin