Love ‘actually’: Employment-related loans and the Leeds Design case

The most important thing in communication is to hear what isn’t being said.

– Peter Drucker

There is much to comment on the decision of the First-tier tribunal in the Leeds Design case. However, in this short piece, I intend only to content myself to the examination of a sole point. The wide plethora of arguments did not serve to advance the taxpayer’s case. These arguments have been processed by the tribunal, its reply to them are to be found in the packaged paragraphs of its decision. The case solved. And yet. And yet there remains the curious incident of the dog in the night-time.

What were the issues arising? Before we turn to answering that question, let us first ask as to what were the issues the tribunal concerned itself with. The tribunal considered whether an undertaking, given on the part of the borrower, to pay official rates of interest could in itself be regarded as the payment of the requisite interest, so as to preclude an employment-related loan from constituting a ‘taxable cheap loan’. The taxpayer’s argument was that this ought to be the case – even in circumstances where the interest was to be simply rolled up over the course of the term of the loan rather than paid year-in year-out. The facts were complicated by the fact that the lender’s fee was expressed not as interest but rather as a discount.

The tribunal came to the conclusion that the labeling of the lender’s fee as a discount ought not in itself preclude it from constituting interest. The computation of the lender’s fee under the agreement concerned indicated that it bore all the characteristics of interest, constituting, as it did, a fee for the use of money borne by reference to time. So far, so good – for the taxpayer. Interest had accrued. However, the problem arose when the tribunal refused to accept that the accrual of interest could be viewed as the payment of interest. The case of Fenton was discussed. (I have a lot of fondness for this case for completely unconnected reasons – as the judges there accepted that the policy objective underlying the provision in question there could not be inferred – which makes it a useful addition to the quiver in the fight with GAAR). The provision concerned in Fenton was removed from the ‘taxable cheap loan’ provisions – as to how far removed, one cannot really say, in light of the accepted amorphousness of the policy objectives it embodied. In any event, in that case, the judges held that the capitalization of interest did not constitute a payment of it and, also, that ‘payment’ was an expression which derived meaning from its context. In Leeds Design, the tribunal managed an obedience to both principles. It reiterated that the expression took its meaning from its context and then also held that it meant the same as it did in Fenton, at least in the sense that capitalization could not constitute payment. In the end, one cannot take exception with this.

My own objection arises not so much from the arguments which the tribunal did address (that is, whether the compounding of interest can be viewed as a payment of it) but rather to an altogether separate point. This point was either not made to the tribunal or, if so, was considered picayune enough to be dismissed without much ado. The question when it comes to considering whether one has a taxable cheap loan, as I see it, is not so much whether interest has been paid at official rates but rather whether, under the terms of the borrowing, interest at such rates is to be paid. Assuming that this is the correct test, then Fenton and the question as to whether rolling over interest constitutes payment – questions with which both the pre-existing HMRC guidance and the tribunal concerned themselves – really fall away like fireflies in the night.

The question next arises as to how this magical switch is affected, whereby the test is adjusted from by reference to the interest actually paid to the interest contractually due under the terms of the borrowing. This is an argument which I have developed and would be happy to elaborate upon where needed to do so, so that others may rely upon it. (I have not seen any other writer espouse these views of the ‘taxable cheap loan’ provisions in their publications at the time of writing – whether in wake of the Leeds Design case or otherwise). I do not guarantee that it would be accepted but there are, in my opinion, reasonable grounds on which it is based. For the avoidance of doubt, should the argument ever be rejected by a court or precedent-setting tribunal, then I intend to leave this article posted here, as an exercise in hermeneutics. However, as promised, the weapon of choice today is the rifle and shot needs to be clean.

Section 175 ITEPA provides:

175 Benefit of taxable cheap loan treated as earnings

(1) The cash equivalent of the benefit of an employment-related loan is to be treated as earnings from the employee’s employment for a tax year if the loan is a taxable cheap loan in relation to that year.

(2) For the purposes of this Chapter an employment-related loan is a “taxable cheap loan” in relation to a particular tax year if—

(a) there is a period consisting of the whole or part of that year during which the loan is outstanding and the employee holds the employment,

(b) no interest is paid on it for that year, or the amount of interest paid on it for that year is less than the interest that would have been payable at the official rate, and

(c) none of the exceptions in sections 176 to 179 apply.

Subsection (2) provides the definition of a ‘taxable cheap loan’, without which there cannot be a charge under subsection (1). The test at (2)(b) is clearly whether there is a divergence between the official rates of interest and the interest paid for the year. My suggestion is that a reference to ‘interest paid for a year’ is to interest paid for a year! Not to interest actually paid ‘in’ a year, as appears to be argued by HMRC. However, setting aside this deeply archaic, naïve and unimaginative proclivity to go simply by the words of the legislator, there is another, more telling clue to which recourse may be had. This is found at subsection (3). Subsection (3) provides the method for calculating the ‘cash equivalent’ of the benefit:

(3) The cash equivalent of the benefit of an employment-related loan for a tax year is the difference between—

(a) the amount of interest that would have been payable on the loan for that year at the official rate, and

(b) the amount of interest (if any) actually paid on the loan for that year.

So, assuming that there is a ‘taxable cheap loan’ under subsection (2), then the computation of the charge hinges on the difference between the official rate of interest and the rate of interest which has ‘actually’ been paid. Suppose that the official rate of interest is 5% and 2% is actually paid, then the charge is 3%.

What I find telling is that (3)(b) is predicated on the assumption that there is a class (‘the class of actual payments on the loan for that year’) which forms a subset of the larger set of the class described at (2)(b) (‘the class of payments of interest on the loan for that year’). If it doesn’t form a subset of the latter, the former is at the very least something different – as the legislator is at (3)(b) simply looking at the interest payments actually made for the year (and I would respectfully submit that the former does in fact form a subset of the latter). To say that the two classes at (2)(b) and (3)(b) are the same would be to regard as otiose the very deliberate inclusion by the legislator of the word ‘actually’ at (3)(b). But the legislator is never taken to have acted in vain. And in this particular case, it seems to me to require the overlooking of a very deliberate inclusion to hold that there is not somehow a change between the two classes in his mind as he goes from one subsection to the contiguous other, as in one he looks to payments of interest for the year and in the other he considers only actual payments of interest for the year.

Now, if the class at subsection (2)(b) is not coterminous with the class at (3)(b) (and we can agree that the class at (3)(b) is rather well defined and not even I, for all my love of Hobbes, would query the concept of actuality before the First-tier tribunal), then the question arises next as to what this class is? Once we find ourselves posed with this question, the rest seems to follow rather swimmingly. The class of interest payments at (2)(b) (‘interest paid for the year’ but not ‘interest actually paid for the year’) is a reference to the contractual rate of interest – it is to the interest to be paid. There is quite simply no other candidate – and if there is, I should very much like to be told what it is.

When the realization dawns on us that the matter was really to be adjudged by reference to the contractual rates of borrowing, then those discomfiting details which so niggled at our conscious seem instead to line up in harmony. The legislator’s use of the expression ‘interest paid for the tax year’. Or ‘taxable cheap loan’. Or ‘benefit’. Everything is illuminated. But not here, not in this article. The scheme of the legislation would need to be developed into a comprehensive matrix, various objections soothed and section 191 addressed. But you are now already more than half-way there. The lawyerly expedient of taking the legislator at his word appears to bring meaning and consistency to the law.