Market leading insight for tax experts
View online issue

The Ramsay principle: where are we now?

printer Mail
Speed read

Under the Ramsay principle of statutory interpretation, it is only ‘realistic’ to look at the overall effect of the series if the statutory provisions require one to focus on the broad practical position. In the cases involving expenditure which are analysed in this article, the courts have not expressly stated that the statute focuses on a realistic view of the facts, thereby permitting a court to go behind the agreed consideration. These cases are not wrongly decided, but the courts should have expressly stated that the statute permitted them to go behind the parties’ allocation of expenditure, not only where that agreement is dishonest or a sham, but also where it is artificial.

In a previous article in this journal I expressed the view that the courts are too willing to apply the Ramsay principle to taxing statutes. They presume that the principle applies, and in so doing they fail to focus on the requirements of the relevant statute. This is evidenced by their ready adoption of the ‘neat apothegm’ (per Moses LJ in PA Holdings [2011] EWCA Civ 1414) that ‘the ultimate question is whether the relevant statutory provisions, construed purposively, were intended to apply to the transaction, viewed realistically’. Viewing the facts ‘realistically’ means, for example, looking at the overall effect of a series of transactions intended to operate together, rather than at the effect of each transaction viewed in isolation from the others in the series. But it is only ‘realistic’ to look at the overall effect of the series if the statutory provisions require one to focus on the broad practical position. Thus, the courts are failing to ask a penultimate question, whether the Ramsay principle applies in the first place.

In this article I shall illustrate my point by reference to recent tax avoidance cases in which the taxpayer has claimed to have incurred expenditure qualifying for an allowance or deduction, but the court has rejected the claim on the ground that, although the transaction was not a sham, the expenditure was actually incurred on something else. I do not say that those cases are all wrongly decided; my criticism is that the courts have adopted the apothegm, and taken a ‘realistic’ view of the facts, without first focusing on the requirements of the relevant statute.

Tower MCashback

The first case is Tower MCashback LLP v HMRC [2011] UKSC 19. There, the taxpayer claimed to have incurred £27.5m of capital expenditure on the provision of plant within CAA 2011 s 11, by paying that amount as consideration for the acquisition of certain novel software from MCashback. Despite its novelty, the taxpayer did not obtain an independent valuation of the software, and was willing to pay the £27.5m price for it because a bank had agreed to fund 75% of the price by way of an interest-free non-recourse loan. The bank also did not obtain a valuation, and was willing to make the loan because MCashback had agreed that 75% of the £27.5m paid by the taxpayer would be applied in providing security for the bank loan, on the same interest-free, non-recourse, terms. Thus, as well as acquiring the software, the taxpayer also obtained ‘soft finance’ from MCashback. The Special Commissioner held that the transactions were not a sham, and the High Court held that it followed from this that the £27.5m consideration was expenditure incurred on the provision of the software within the meaning of s 11. In other words, the High Court thought, the contractual attribution of the £27.5m to the software was determinative. The Supreme Court disagreed, holding that on a realistic view of the facts found, 75% of the £27.5m was incurred on the soft finance rather than on the software. What did the Supreme Court mean by a ‘realistic view’ of the facts? Emphasis was placed on the fact that the market value of the software was much less than £27.5m, but this could not, of itself, justify the conclusion that part of the £27.5m was really incurred on the soft finance. After all, the taxpayer might have believed that the software was genuinely worth more, and so struck a bad bargain. But in fact the taxpayer was indifferent as to the value of the software, as evidenced by his failure to obtain a valuation. Thus, the reason why at least part of the £27.5m was incurred on the soft finance was because the contractual attribution to the software was wholly artificial. In my view the Supreme Court ought to have stated expressly that s 11 focuses on a realistic view of the facts in the sense that the court may go behind the agreed consideration not only where it is dishonest or a sham but also where it is artificial.

The Supreme Court might have justified that approach by analogy with an earlier decision of the House of Lords, in Stanton v Drayton [1983] 1 AC 501. There, the taxpayer company had acquired an asset for a purchase-price of £4m to be satisfied by the issue of new shares in the company. But the market value of the shares when issued was only £3m. The Revenue argued that the value of the ‘consideration given for the acquisition’ of the asset within the meaning of (what is now) TCGA 1992 s 38 was therefore only £3m. However, the House of Lords held that the Revenue was not entitled to go behind the agreed consideration in a case where (per Lord Fraser) ‘the transaction is not alleged to be dishonest or otherwise not straightforward’. In my view, a transaction is ‘not straightforward’ where the parties’ agreement is artificial.


The second case is Audley v HMRC (TC01084). There, the taxpayer established a settlement whose trustees issued him with a loan note at an issue price of £2m, although as the taxpayer knew the market value of the loan note was only £37,500. The question was whether the whole of the £2m was paid ‘in respect of the acquisition of’ the loan note within the meaning of the relevant taxing statute, FA 1996 schedule 13. The tribunal held that the transaction was not a sham but that, applying the Ramsay principle, on a realistic view of the facts only £37,500 of the £2m was so paid. In my view the tribunal should have held, first, that the words ‘in respect of the acquisition’ should be given a broad practical meaning, so as to focus on a realistic view of the facts in the sense that the court may go behind the contractual consideration for the acquisition of a loan note insofar as the consideration is artificial. The tribunal should then have held that the excess over £37,500 was indeed artificial because on the one hand the taxpayer knew that the loan note was worth only £37,500, and on the other hand the trustees were not dealing with the taxpayer at arm’s length because they issued the loan note at his behest and in all probability would have done so for only £37,500 if the taxpayer had so required.


The third case is Drummond v HMRC [2009] STC 2206. There, the taxpayer bought five life insurance policies for a consideration of £1.96m and immediately surrendered them for their surrender value of £1.75m, as part of an unsuccessful tax avoidance scheme. The taxpayer argued that all of the £1.96m was ‘consideration given wholly and exclusively for the acquisition’ of the policies within the meaning of TCGA 1992 s 38. But the Special Commissioner held that viewing the facts realistically, only £1.75m, the surrender value, was so given, the excess being a fee given for participating in the scheme. The taxpayer appealed but the Court of Appeal held that the appeal was a ‘hopeless endeavour’. In my view, the fact that the taxpayer paid more than the surrender value does not mean that the excess was a fee. The court should have held that s 38 focuses on the parties’ agreement except where the agreement is a sham or is artificial. Here, the taxpayer agreed to pay £1.96m under a bargain at arm’s length which was not artificial because the taxpayer believed that ownership of the policies would bring substantial tax advantages.


The final case is Mayes v HMRC [2011] EWCA Civ 407, where the taxpayer purchased a second hand life insurance policy for £133,000 when its surrender value was only £1,800. He did so because he was be able to obtain a very valuable income tax advantage, called corresponding deficiency relief, on the surrender of the policy. HMRC contended that only £1,800 was given for the policy, the rest being a fee, and they relied on Drummond. The taxpayer contended that Drummond was inconsistent with Stanton v Drayton. I agree with the taxpayer: as mentioned above, Stanton v Drayton decided that the court cannot go behind the agreed consideration if it is honest and not artificial. In Mayes the parties’ agreement was honest and not artificial because the policy was worth £133,000 to the taxpayer having regard to the tax advantage which ownership would bring.

Suppose that the relevant statute, purposively construed, entitles HMRC to re-allocate consideration where the parties’ own allocation is artificial. In my view, in such a case the statute positively requires such a re-allocation to be made: HMRC does not have any choice in the matter. In other words, where it suits him the taxpayer can insist on the re-allocation even though he was one of the parties to the transaction.

In Booth v Buckwell, the taxpayer sought to go behind an apportionment of consideration on the ground that it was artificial. However, the court held that the parties ‘cannot subsequently seek to re-allocate the consideration for tax purposes. They have chosen to carry through the transaction in a particular manner, and the taxation consequences flow from the manner adopted. The Crown’s position may well be different in certain cases. After all, the Crown was not a party to the transaction.’ With respect, this makes no sense. The only reason why HMRC is not bound by the parties’ allocation is because the statute requires a different allocation to be made, and the court must give effect to that requirement whether HMRC likes it or not.