Rupert Grint v The Commissioners for HMRC

This is an appeal against the decision (the “Decision”) of the First-tier Tribunal (Judge Mosedale) released on 3 August 2016 and reported at [2016] UKFTT 0537 (TC). The FTT dismissed the Appellant’s challenge to amendments made to his tax returns for the years ended 2010 and 2011 in closure notices issued by the Respondents, HMRC. The appeal is brought with permission granted by Judge Mosedale in a decision dated 6 October 2017.

The question was whether by changing his accounting date from 31 July to 5 April Mr Grint had thereby advanced his tax liability on some eight months’ worth of profit into the 2009/10 year. This would be to his advantage because a new higher rate of tax was introduced for the following tax year.

Central to the issues arising in this appeal is the concept of a “basis period” for income tax purposes. Section 5 of the Income Tax (Trading and Other Income) Act 2005 (“ITTOIA”) provides for profits of a trade or profession to be chargeable to income tax. Section 7 provides for the trading profits for a tax year to be calculated by reference to the “basis period” for that tax year. Chapter 15 of Part 2 ITTOIA sets out detailed rules for determining the basis period for a tax year. There are special rules applicable to the early years, and the final year, in which a trade or profession is conducted which are not relevant to this appeal. Rather, prior to the change of accounting date in the 2009/10 tax year, Mr Grint’s basis periods were determined by the general rule set out in section 198 ITTOIA. That provides, so far as relevant:

The term “accounting date” is defined in section 197 ITTOIA:

Since a basis period is defined by reference to an accounting date, the legislation provides for a basis period to change in some circumstances where an accounting date changes. The principal provisions governing a change of accounting date are set out in sections 214 to 217 of ITTOIA:

The 18 month test set out in section 217(3) is framed by reference to a “period of account”. The term “period of account” is defined in section 989 Income Tax Act 2007 (“ITA”):

ITTOIA is one of the Income Taxes Acts so that the term “period of account” used in section 217(3) ITTOIA bears the meaning in section 989 ITA.

Broadly, applying those provisions to this case, it was common ground that Mr Grint, a successful actor, had since about 2000 used an accounting date of 31 July each year, drawing up annual accounts to that date and that his basis period coincided with the period covered by those annual accounts. Thus, until the tax year 2009/10, he had paid tax in each year on the profit for the 12 months to 31 July falling within that tax year. For example, his accounting date in the tax year ending 5 April 2008 was 31 July 2007, his basis period was the 12 months between 1 August 2006 and 31 July 2007 and so his tax liability for 2007/08 was calculated by reference to the profits of that 12 month basis period.

What Mr Grint wanted to achieve in the tax year ending 5 April 2010 was to change his accounting date to 5 April instead of 31 July. He hoped this change would result in his basis period for the tax year ending 5 April 2010 being the period from 1 August 2008 to 5 April 2010, a period of about 20 months. This would, he believed, be achieved because, looking at the statutory provisions:

It was common ground that Mr Grint satisfied all the conditions in section 216(1) because he was carrying on a trade in the tax year 2009/10; he had been carrying on that trade for more than three years and he did not cease that trade in that tax year;

His new accounting date (5 April 2010) fell more than 12 months after the end of his basis period for the previous tax year, that is more than 12 months after 31 July 2008 which had been the end of the basis period by reference to which he paid tax in the tax year ending 5 April 2009. He therefore fell within section 216(3) rather than section 216(2).

He considered that the conditions in section 217 were met, as required by section 216(3)(a), with the result that the basis period for his 2009/10 tax year would begin immediately after the end of the basis period for the previous tax year – that is 1 August 2008 - and end on the accounting date 5 April 2010, that is just over 20 months.

If Mr Grint could achieve the result he sought then the profits of the additional 8 months from 1 August 2009 to 5 April 2010 would be earned in the basis period for the 2009/10 tax year (i.e. the period from 1 August 2008 to 5 April 2010) and so would be taxed at the tax rates prevailing in the tax year 2009/10 and not at the higher rates prevailing in the tax year 2010/11. If Mr Grint did not change his accounting date to 5 April 2010, then the profits of those additional eight months would have been earned in the basis period applicable to the 2010/11 tax year (i.e. the period from 1 August 2009 to 31 July 2010) and so would be taxed at the higher rates applicable to the 2010/11 tax year.

Mr Grint notified HMRC of his change of accounting date when he submitted his tax return for the tax year 2009/10. In the main body of the tax return he ticked the box indicating that he was self-employed and that he was attaching supplementary pages. He submitted two sets of self-employment supplementary pages, both for the tax year 6 April 2009 to 5 April 2010.

The first set of supplementary pages were referred to in the FTT’s Decision and at the hearing before us as the “2009 Return Accounts”. In the 2009 Return Accounts, in response to the question “Date your books or accounts start - the beginning of your accounting period” he put ‘1 August 2008’. In response to the question “Date your books or accounts are made up to or the end of your accounting period” he put ‘31 July 2009’. He put an X in the box indicating that his accounting date had changed permanently. The boxes for business income, expenses, net profit and taxable profit were then completed with figures which were derived in the manner we describe below.

The second set of self-employment supplementary pages submitted with Mr Grint’s 2009/10 tax return were referred to in the FTT’s Decision and before us as the 2010 Return Accounts. In the 2010 Return Accounts, in response to the question “Date your books or accounts start - the beginning of your accounting period” Mr Grint put ‘1 August 2009’. In response to the question “Date your books or accounts are made up to or the end of your accounting period” he put ‘5 April 2010’. Again, he ticked the box to indicate that his accounting date had changed permanently. He filled in the boxes for business income, expenses, net profit, taxable profit etc with figures, again derived in the manner we describe below.

The tax return was submitted online to HMRC on the date it was due, 31 January 2011.

HMRC initially accepted that Mr Grint had changed his accounting date so that his basis period on which his tax liability for the tax year 2009/10 was determined was the 20 month period from 1 August 2008 to 5 April 2010. That liability was computed by adding together the figures in the corresponding boxes in the two sets of supplementary pages submitted with the tax return. However, on 10 May 2011 HMRC officers carried out a VAT control visit of Mr Grint at the offices of his accountants Clay & Associates. The HMRC officer was shown files relating to Mr Grint. Among the files were another set of accounts referred to as the Long Accounts. The Long Accounts had been prepared by Mr Grint’s accountants for a meeting between them and Mr Grint in October 2010. The Long Accounts were produced to and signed by Mr Grint at that meeting and covered the 20 month period from 1 August 2008 to 5 April 2010.

HMRC opened an enquiry into Mr Grint’s 2009/10 tax return on 12 January 2012. The enquiry was completed and closure notices issued on 14 August 2014. HMRC’s review of the closure notices upheld them on 24 October 2014.

At the end of the enquiry HMRC concluded that Mr Grint had not established that his basis period for 2009/10 was the 20 month period from 1 August 2008 to 5 April 2010. HMRC reasoned that the “period of account” for the purposes of section 989 ITA, being the “period for which the accounts of the business are drawn up”, was the period covered by the Long Accounts (period 1 August 2008 to 5 April 2010) and was therefore longer than 18 months. That meant in turn that section 216(3) did not apply because the conditions in section 217 had not been met and it followed that section 216(3) did not operate to produce a basis period, for 2009/10, of 1 August 2008 to 5 April 2010.

HMRC has not contended that any of the other conditions in section 217 were not met; Mr Grint gave appropriate notice for the purposes of section 217(1)(a) and he satisfied Condition A because he had not changed his accounting date during the previous five years.

The dispute between the parties before the FTT focused on identifying which of various alleged “accounts” drawn up over the period by Mr Grint’s accountants fell to be treated as the “accounts” for the purpose of determining whether the period of account relevant to section 217(3) was longer than 18 months. This required the FTT to consider first what characteristics a document had to have in order to be “accounts” and then, since the FTT identified more than one document that should be regarded as “accounts”, how to decide which of the rival contenders should be regarded as fixing the period of account for the purposes of determining whether Mr Grint had met the 18 month test.

As regards which documents qualified as “accounts”, there were four sets put forward by the parties which the FTT had to consider.

The Long Accounts we have already described. They covered the period from 1 August 2008 to 5 April 2010. These were produced to Mr Grint at a meeting between him and his advisers on 18 October 2010 and were signed by him. Judge Mosedale found [33] that they were used at the meeting to discuss Mr Grint’s financial performance with him and his father. Mr Grint also knew that the Long Accounts would be used to draw up accounts covering two shorter periods as part of the process of changing his accounting date.

The Schedule Accounts drawn up by the tax department at Clay & Associates. These split the period covered by the Long Accounts into two periods: the 12 months from 1 August 2008 to 31 July 2009 (referred to as the 2009 Schedule Accounts) and the 8 months from 1 August 2009 to 5 April 2010 (referred to as the 2010 Schedule Accounts). At [23] of the Decision, the FTT found that drafts of the Schedule Accounts were prepared before the meeting with Mr Grint on 18 October 2010 but were not shown to him at that meeting. They were shown to him later when they were sent to him along with his draft tax return for the 2009/10 tax year. The way the Schedule Accounts were drawn up was simply by time apportionment, in other words, the figures shown for the various items in the 2009 Schedule Accounts were 12/20ths of the figures for the corresponding item in the Long Accounts and the figures shown for the various items in the 2010 Schedule Accounts were 8/20ths of the corresponding figures in the Long Accounts.

The 2009 Return Accounts and the 2010 Return Accounts that we have already described at [12] and [13] above, namely the self-employment supplementary pages of the tax return submitted by Mr Grint for the year ended 5 April 2010. The figures included in the supplementary pages were taken directly from the 2009 and 2010 Schedule Accounts and so were also derived from the Long Accounts by simply multiplying the figures in the Long Accounts by 12/20ths or 8/20ths as appropriate.

The New Accounts which were produced by Mr Grint’s advisers during the enquiry opened by HMRC on 12 January 2012. They covered the same two periods as the Schedule Accounts and the Return Accounts (and so the FTT described the New Accounts as comprising the “2009 New Accounts” made up to 31 July 2009 and the “2010 New Accounts” made up to 5 April 2010) but with the difference that they were prepared on the accruals basis rather than simply taking an arithmetical proportion of the figures in the Long Accounts.

In determining which of these documents were “accounts” the FTT considered expert evidence adduced by the parties. The first issue on the nature of accounts was whether it was a requirement, in order for a document to qualify as accounts, for them to be reliable. HMRC argued that there was such a requirement and that the Schedule Accounts did not qualify as accounts. The figures in them were simply arithmetically derived proportions of the period of 20 months covered by the Long Accounts and did not tell the reader anything useful about the performance of the business over the periods covered. Mr Grint’s business was one where income came in an irregular pattern, so a time apportionment did not allow the reader to see how well he was doing, year on year. Judge Mosedale rejected the submission that a set of figures had to be reliable in order to qualify as accounts. She therefore held that the Schedule Accounts were accounts. HMRC has not challenged that aspect of the Decision.

The second issue was whether the Return Accounts, which were in fact just the supplementary pages attached to Mr Grint’s tax return, were accounts. Judge Mosedale held that they were not. Mr Grint has appealed against that aspect of the Decision.

The third issue was whether the Long Accounts were accounts, or whether, as Mr Grint contended, they were prepared solely as working papers to enable preparation of the Schedule Accounts and his tax return. Judge Mosedale found that that contention was inconsistent with the evidence and held that they were accounts: see [153] – [157] of the Decision.

Therefore, pausing there, the effect of the FTT’s findings was that Mr Grint had made up “accounts” to two dates in 2009/10: the Long Accounts and 2010 Schedule Accounts were made up to 5 April 2010 and the 2009 Schedule Accounts were made up to 31 July 2009. In those circumstances, Mr Grint’s “accounting date” for the purposes of section 197 was 5 April 2010, the later of these dates. There was, therefore, a change of accounting date in the 2009/10 tax year applying the rule in section 214 since in 2008/09, Mr Grint’s accounting date was 31 July whereas in 2009/10 it was 5 April. Therefore, the conditions in section 216(1) were met and it was necessary to establish whether section 216(3) caused Mr Grint’s basis period for 2009/10 to be the period from 31 July 2008 to 5 April 2010. Given that the other relevant conditions for the application of section 216(3) were met, that issue turned on whether the 18 month test in section 217(3) was satisfied.

To apply the 18 month test in section 217(3), the FTT had to address two separate issues. First, it had to determine when the test fell to be applied since one of Mr Grint’s arguments was that the New Accounts (which were prepared after he submitted his tax return and after he notified HMRC of a change of accounting date) were nevertheless relevant. Second, applying the test at the right point in time, it had to determine which of the periods ending on the “accounting date” of 5 April 2010 was the relevant “period of account” for the purposes of that test.

Judge Mosedale held that in order to be considered for the purposes of section 217(3), the “accounts” referred to had to be in existence at the date when the notification of change of accounting date was given to HMRC in accordance with section 217(2). She therefore held that, although the New Accounts were “accounts”, Mr Grint could not rely on them because they were not prepared until after 31 January 2011 (which was when Mr Grint submitted his tax return and notified HMRC of his change of accounting date). Mr Grint has appealed against that aspect of the Decision.

To determine the “period of account” ending on 5 April 2010 relevant for the purposes of section 217(3), the FTT decided it had to identify the period, ending on that date “for which the accounts of the business are drawn up” given the definition in section 989 ITA. Having ruled out reliance on the Return Accounts because they were not “accounts” and on the New Accounts because they were prepared too late to be relevant, there were two contenders for Judge Mosedale to consider: the period covered by the Long Accounts and the period covered by the 2010 Schedule Accounts. She held that the Long Accounts were the only accounts that actually reflected Mr Grint’s financial performance over the period to which they related. She therefore concluded that the “period of account” relevant for the purposes of section 217(3) was the 20 month period covered by the Long Accounts. Since this was longer than 18 months, Judge Mosedale upheld HMRC’s conclusion that the “18 month test” in section 217(3) was not met. It followed that section 216(3) did not apply and so Mr Grint had failed to establish that his basis period for 2009/10 was the period from 1 August 2008 to 5 April 2010.

The Notice of Appeal lodged with the FTT by Mr Grint put forward four alleged errors of law on the part of the FTT. The first was, broadly, the finding that, for the purpose of ascertaining the relevant “period of account” for the purposes of section 217(3), the Long Accounts were the relevant accounts rather than the 2010 Schedule Accounts because they were Mr Grint’s general purpose tax accounts. The second was that if the test was, as the FTT found, which were “the more important” accounts then it should have found that the Schedule Accounts were more important for Mr Grint than the Long Accounts. The third ground challenged the finding that the Return Accounts were not accounts at all. The fourth ground was that the New Accounts should have been held to be relevant as there is no deadline set in section 217 by which the accounts that determined the relevant “period of account” in section 217 must be drawn up. HMRC responded to the Notice of Appeal on 1 December 2017 arguing that there were no errors of law and seeking to uphold the decision for the reasons given by Judge Mosedale. A short Reply to the Response was served on 19 December 2017.

At the start of the hearing of the appeal we raised with Mr Soares, appearing for Mr Grint, an alternative reading of section 216 and 217 which appeared to us more likely to reflect what Parliament had intended to achieve, if not what it had in fact achieved, by the provisions as drafted. That was that the “18 month test” was imposed to restrict to 18 months the length of the basis period that could arise on a change of accounting date by operation of section 216(3).

We saw the difficulties in construing section 217(3) as achieving this. The conditions for satisfying the 18 month test in subsection (3) do not refer to the “basis period” but to the “period of account”. That term is defined in section 989 ITA not by reference to the period brought into tax but by reference to the accounts drawn up by the business ending on the “accounting date”. Further, although the description of the 18 month test in subsection (3) explains when the period of 18 months ends, unhelpfully it does not say when the period begins. One might expect it to refer back to section 216(3) so as to set the start date for the 18 month period in section 217(3) as being the date “immediately after the end of the basis period for the previous tax year” referred to in section 216(3). That would mean that Mr Grint would fail the 18 month test not because he drew up the Long Accounts but because he was trying to have a basis period for the tax year 2009/10 which ran from 1 August 2008 to 5 April 2010, longer than 18 months.

Mr Soares firmly rejected this as a possible reading of the provisions. He submitted that these provisions have been in operation for many years and that HMRC have always recognised that taxpayers have the “privilege” of changing their accounting date in order to avoid profit falling into a higher rate tax year. They can do this provided that the accounts they draw up ending with the new accounting date cover a period less than 18 months, even though the figures in those accounts must then be added to the figures in an earlier set of accounts because the basis period straddles the period covered by two sets of accounts. He stressed that HMRC and the accountancy profession have always operated these provisions in a way which allows basis periods of longer than 18 months, so long as that longer period is chopped up into two smaller periods in the business’ accounts and that the smaller period ending on the new accounting date is less than 18 months. He described the result as fair because the profits earned in that 8 months (from 1 August 2009 to 5 April 2010) were earned at a time when the prevailing tax rates were the lower rates and should not therefore be taxed at the higher rate just because, but for the accounting date change, they would fall in a basis period which ended in the following tax year after rates had increased.

Ms Choudhury appearing for HMRC, agreed and did not invite us to give a purposive meaning to section 217(3) to limit taxpayers to a basis period of 18 months. She confirmed what Mr Soares told us that if the Long Accounts had never come into existence and Mr Grint had either only produced the Schedule Accounts or had only produced the New Accounts before he submitted his 2009/10 tax return, Mr Grint would probably have obtained the outcome he sought. HMRC would regard him as having met the 18 month test, even though that resulted in a basis period of 20 months and allowed income generated in 8 months rather than 6 months to be subject to the lower rates of tax applicable in 2009/10. Neither party had attended the hearing prepared to present to us a construction of the provisions different from the construction which was assumed to be correct for the proceedings before the FTT. Ms Choudhury was able only to say that the 18 month test had been imposed by income tax legislation for very many years and had never been regarded as imposing a restriction on the length of the basis period ending with the taxpayer’s new accounting date. Basis periods of up to 24 months were therefore possible under section 216(3) and were regarded by HMRC as meeting the 18 month test provided the taxpayer had accounts that were drawn up ending on the new accounting date and that covered a period of less than 18 months.

With some reluctance therefore, we will determine this appeal by resolving the grounds of appeal as put forward by Mr Grint and argued by the parties.

We consider these two grounds together as they both raise the issue of whether Judge Mosedale erred in how, when determining the relevant “period of account”, she chose between the periods represented by the Long Accounts and the 2010 Schedule Accounts, both of which she had held were properly to be regarded as accounts and both of which were drawn up to 5 April 2010.

The key passages in the Decision are paragraphs [171] onwards. Mr Grint had relied on the decision of the Court of Appeal in Jenkins Productions Ltd 29 TC 142 (1944). In that case the taxpayer company had made up accounts for the half years ended 30 June and 31 December. Those accounts had been audited and placed before the board of directors for approval. The accounts presented to the shareholders in general meeting were annual accounts to the year end 30 June. Those accounts were based on the two half-years’ figures. The Special Commissioners had upheld an assessment to Excess Profits Tax for a chargeable accounting period of three months, April, May and June 1939. Whether profits in that accounting period were “excess” as defined by the legislation depended on a comparison with the profits made in an earlier period in 1935 which would be treated as the “standard” profit. The ascertainment of the company’s standard profits for the year 1935 depended on the question whether, under section 20 of the Finance Act 1937, “the accounts of the trade or business were made up for successive periods of twelve months”. The Special Commissioners had decided that the profits for the business for the standard year 1935 must be ascertained by taking the profits for the two years ended 30 June 1935 and 30 June 1936 adding them together and dividing by two.

The company argued that it was not true to say that the accounts of its business were made up for successive periods of 12 months. If the company did not have accounts that were made up for successive periods of twelve months, then the Commissioners would have had to pick a twelve month period for the purpose of ascertaining the profits of the business in the standard year, 1935. The company wanted the calendar year 1935 to be chosen because that would exclude from the computation a substantial loss which it suffered in the first part of 1936. If the first few months of 1936 were taken into account when computing the standard profit, as HMRC proposed, those months would depress the standard level of profit in the comparator year and correspondingly increase the “excess” profit made in the quarter to end June 1939.

The Court of Appeal held that the company did make up accounts for successive periods of 12 months. The Court went on to say that if the company prepares two sets of accounts, one for successive 12 month periods and one for successive six month periods, the 12 monthly account “is one that has certain characteristics”, namely that those accounts were submitted to the shareholders in pursuance of the statutory obligation to submit accounts and they were the final accounts. Bearing those two considerations in mind, the Court held that “the yearly accounts are primarily the more important of the two sets of accounts which this Company in fact got out”. It was the 12 monthly accounts that were therefore properly to be described as the accounts of the trade or business because they were for “the company’s financial unit of time”.

Judge Mosedale’s reasoning for deciding that the Long Accounts were the relevant accounts for determining the period of account was as follows:

She held that the reference in ITTOIA to “accounts” was to “accounts of the trade” which she took to mean “something like general purpose financial statements”. The phrase did not refer to “any and every set of accounts produced by a taxpayer”: [171].

She went on to say: [178]

At [175], she disagreed with Mr Grint’s submission on the basis of Jenkins that the Schedule Accounts were “more important” because they were notified to HMRC in connection with the change of accounting date. The “more important” accounts were the Long Accounts because they were the only ones used to present to Mr Grint his financial results. They were “the only accounts which actually reflected Mr Grint’s financial performance over the period to which they related”. The Schedule Accounts were simply time apportioned figures which were convenient but which did not reflect his performance over either the 12 month period in the 2009 Schedule Accounts or the 8 month period in the 2010 Schedule Accounts. The Schedule Accounts were never actually shown to HMRC though the figures from those were transposed into the Return Accounts (that is the supplementary pages which she held were not accounts at all). The figures from the Schedule Accounts were entirely drawn from the Long Accounts. At [179], Judge Mosedale made a key finding that the Long Accounts were “more important” than the Schedule Accounts.

At [180] Judge Mosedale considered what the position would be if, instead of determining ‘the more important’ set of accounts, she instead determined which of the accounts Mr Grint produced constituted ‘the’ accounts of the business. She likened this to asking which, of the various accounts, were used to represent past transactions within a set period of time and concluded that the Long Accounts fit the description for the same reasons as set out at [(3)] above.

Judge Mosedale’s overall conclusion, set out at [182] was that the Long Accounts were the “accounts which matter” for the purposes of section 217. She outlined four respects in which the Long Accounts better fitted the description of “business accounts” than the Schedule Accounts as follows:

Mr Soares argued that Judge Mosedale has imported into the question of what are the accounts for the purposes of section 217 an additional qualification for which there was no justification, namely a requirement that the accounts are general purpose accounts or accounts relied on by the business for its general commercial or trading purposes: see [163] and [178] of the Decision. This was an error if it meant that accounts drawn up solely or mainly for the purpose of enabling the trader to complete his tax return could not be accounts within subparagraph (b) of the definition in section 989 ITA as applied in section 217(3). He argued that this would create serious problems for businesses if they drew up accounts primarily for the purpose of completing their tax returns because they then would not have any general purpose accounts and so would not have any accounts which they could use to change their accounting date.

We do not accept that Judge Mosedale was saying that if a taxpayer has a single set of accounts ending with the new accounting date, those cannot be relied on by the taxpayer to change his accounting date if they were primarily or solely drawn up for the purpose of completing the tax return. That was not the question she was addressing. Rather, Judge Mosedale was confronted with two sets of “accounts”, the Long Accounts and the 2010 Schedule Accounts, which were both drawn up to the “accounting date” of 5 April 2010 but spanned different periods. She therefore had to choose between those accounts to decide whether the “period of account” for the purposes of section 217 was the period covered by the Long Accounts or that covered by the 2010 Schedule Accounts. In deciding that the Long Accounts provided the relevant period of account, she was not making a general statement that accounts prepared for the “wrong purpose” fell to be ignored. Indeed, quite to the contrary, she held that the Schedule Accounts constituted “accounts” even though they were prepared for a specific tax purpose (see [196] of the Decision).

Therefore, the question posed in Grounds 1 and 2 is whether Judge Mosedale erred in determining whether the “period of account” as defined in section 989 ITA was the period covered by the Long Accounts, or that covered by the 2010 Schedule Accounts. The definition in section 989 has two limbs: with limb (a) applying “in relation to a person” and limb (b) applying “in relation to a trade, profession, vocation or other business”. Although Mr Grint is clearly a “person”, the need to determine the correct “period of account” arose in the context of the determination of the profits of a trade, profession or vocation, so it is limb (b) which is relevant here.

Limb (b) of the definition in section 989 refers specifically to “the accounts of the business” (in the singular) and there can be only one relevant “period of account” for the purposes of section 217. Therefore, Judge Mosedale’s task was similar to that of the Court of Appeal in Jenkins. In that case the Court had to identify “the accounts of the trade or business” and ask whether they were drawn up for successive periods of 12 months. Judge Mosedale had to decide which, of the various competing sets of accounts, were “the accounts of the business” and ask whether the period for which those accounts were drawn up was longer than 18 months.

For the reasons given in paragraphs [175] and [182] of the Decision we agree that the Long Accounts better fit the description in section 989 ITA of being “the accounts of the business”. The ‘period of account’ was the period covered by those Long Accounts. They were the accounts drawn up to discuss with Mr Grint how his career was developing and would enable him to compare his performance in that period with any earlier or later period.

Mr Soares argued that if the test is that set out in Jenkins,namely which of two rival sets of accounts are “more important”, then the Schedule Accounts win. So far as Mr Grint was concerned, the Schedule Accounts were very important because it was those that supported his change of accounting period which in turn enabled him to make a substantial saving on his taxes. We do not accept that the reference to importance in Jenkins was intended to bring into consideration a wide range of subjective factors including their potential effect on the taxpayer’s tax liability or other finances. The Court of Appeal in Jenkins did not weigh in the balance that the six monthly accounts drawn up by the company might be “important” if they enabled it to change the standard profit comparator period to one which reduced the company’s tax liability. Importance is assessed in the context of the question which is being asked which is, which are more properly described as the accounts of the business.

We therefore dismiss ground 1 of the appeal and find there was no error of law, given the way the case was argued before Judge Mosedale and before us.

The Return Accounts were, as we have described, the two sets of supplementary pages attached by Mr Grint to his tax return submitted online. Judge Mosedale held that those pages did not constitute accounts at all. She listed the characteristics that a set of numbers had to have to fall within the term accounts, a term that is not defined in the legislation. She recorded that the parties’ experts agreed that there was considerable subjectivity in the accountancy profession as to what would amount to accounts: [72]. They were agreed that the accounts must relate to an entity and should summarise the entity’s past transactions during a specified period of time. They appeared to agree that a profit and loss sheet was enough to constitute accounts but she did not have to decide whether a balance sheet was also essential as all the contenders included a balance sheet. She held that:

The Return Accounts did relate to an entity, that is Mr Grint: [86].

They summarised, or at least purported or intended to summarise, past transactions in a fixed period of time: [89].

It did not matter that they included time apportioned figures drawn from the Long Accounts and hence were unreliable: [104]. There was no requirement that figures be reliable in order to amount to accounts: [112].

The Return Accounts were not accounts because even though they contain the same information as would be comprised in a set of accounts, they were not intended as a record of past transactions in a set period of time but to return the author’s tax liability: [129] and [142]. Moreover, she said “while signed by Mr Grint, there is no evidence to show that the return was signed or adopted by him as accounts, rather than simply as his tax return”.

We agree with Judge Mosedale’s analysis and her conclusion. The self-employment supplementary pages attached to the tax return do not constitute accounts.

Mr Soares argued that this finding could create problems for a small business which inputted the figures into its tax return supplementary pages directly from the raw data without first drawing up more orthodox accounts and then taking the figures from those accounts to include in the relevant boxes in the supplementary pages. He argued that, if the supplementary pages in the tax return were not “accounts”, then such a small business would have no “accounts” at all and so the default rules set out in section 210 of ITTOIA might apply to treat the business as having a basis period very different from that assumed to be applicable in the tax return.

Ms Choudhury acknowledged that since HMRC does not require businesses to draw up accounts, such a problem might exist in theory. It was not a problem that needed to be resolved in the present appeal since there was no shortage of other sets of accounts in contention. She pointed out that Helpsheet 222 published by HMRC states that if the business does not prepare accounts for its business, it should read ‘accounting period’ to mean the period for which it provides details of its business income and expenditure and ‘accounting date’ to mean the date on which that period ends. This seems to us a pragmatic solution to a theoretical problem. It does not change the proper construction of the term “period of account” and does not affect our conclusion that Judge Mosedale made no error of law in concluding that the Return Accounts were not accounts at all.

Moreover, even if Mr Grint was correct and the Return Accounts were relevant, that conclusion alone would not result in his appeal succeeding. Mr Grint would need to go further and establish that the 2010 Return Accounts should, in preference to the Long Accounts, establish the “period of account” for the purposes of section 217. Since we agree with the FTT’s decision that it is the Long Accounts which establish “the accounts of the business” in preference to the Schedule Accounts, we see no reason why, even if they were accounts, the Return Accounts which consisted simply of attachments to Mr Grint’s tax return for 2009/10 would constitute “the accounts of the business” in preference to the Long Accounts.

The issue before the FTT and before us was whether there was a date by which the accounts referred to in section 217(3), that is the accounts which are drawn up to the new accounting date and which cover a period of account of not more than 18 months, have to exist in order for the taxpayer to rely upon them. Judge Mosedale held that on the literal interpretation of section 217 there was no express time limit as there was for the notification of the change of date to HMRC and that textual pointers went both ways. A purposive interpretation, however, led her to conclude that the accounts must exist at the time the taxpayer’s basis period was determined, in Mr Grint’s case when his 2009/10 tax return was filed on 31 January 2011.

The assumption underlying sections 216 and 217 is that the taxpayer is changing his accounting date and the new accounting date falls in the tax year in which the new basis period prescribed by section 216(3) is to apply. That is the tax year to which the return by which the change of accounting date is notified to HMRC relates. The entries of dates in those boxes do not of themselves constitute the change in accounting date, rather they are supposed to record an independently existing fact, namely that there are books or accounts drawn up to the new accounting date.

We consider that the following aspects of the statutory scheme demonstrate that, if accounts are to establish a “period of account” for the purposes of section 217, those accounts must actually be in existence when the tax return is filed:

Section 217 of ITTOIA provides that, for Mr Grint’s change of accounting date to affect his basis period, notice of that change had to be given in his tax return for 2009/10 that was submitted on or before 31 January 2011.

That same tax return also had to contain details of Mr Grint’s profit for 2009/10. If Mr Grint’s basis period had changed, that would have involved a calculation of profit for the period up to 5 April 2010.

In order to be able to complete his tax return, Mr Grint would have needed access to a determination of accounting profit for the relevant basis period when he submitted his tax return.

Therefore, the scheme of the legislation envisages that taxpayers in Mr Grint’s position notify HMRC of a change of accounting date in a tax return that itself draws on information drawn from accounts made up to the new accounting date. That is entirely consistent with a requirement that the accounts relevant for the purposes of section 217(3) must exist at the time the tax return is submitted.

If Mr Soares’ contention was right, that would reduce the 18 month test from the technical, tick box exercise that Mr Grint and HMRC both consider it to be, to a test that does not even require the box to be ticked at any point before HMRC opens an enquiry. It would make the test, in effect, a test that cannot be failed because the taxpayer would not need to have any accounts at all covering the period inserted in answer to the two tax return questions. If asked to produce the accounts to which the boxes on the tax return referred, he could simply at that point draw up accounts covering a period of less than 18 months ending with the accounting date and then rely on those accounts to satisfy the test. We find that that is not a possible construction of the provisions. Even if the Long Accounts did not exist, the New Accounts could not therefore be relied on by Mr Grint as they only came into existence after 31 January 2011 when he submitted his tax return.

Mr Grint’s appeal is dismissed.