BAILII [Home] [Databases] [World Law] [Multidatabase Search] [Help] [Feedback]

First-tier Tribunal (Tax)

You are here: BAILII >> Databases >> First-tier Tribunal (Tax) >> Homsub Ltd v Revenue & Customs (VAT - Best Judgment - Flawed methodology) [2019] UKFTT 536 (TC) (14 August 2019)
Cite as: [2019] UKFTT 536 (TC)

[New search] [Contents list] [View without highlighting] [Printable PDF version] [Help]

VAT – Best Judgment – Flawed methodology.




Appeal number: TC/2017/00188























Sitting in public at Taylor House, London on 05 August 2019.


Mr. L. Ahmed for the Appellant

Ms K. Ellis, litigator of HM Revenue and Customs’ Solicitor’s Office, for the Respondents



1.            The appellant company is a franchisee in respect of Subway products, essentially being hot and cold food, which can be consumed either on or off their premises. The appellant operated from five different stores in five different towns, with the various outlets being within the appellant company’s business and accounting for VAT being within one single company VAT return each quarter.

2.            The respondents became concerned that the appellant’s percentage split between supplies subject to VAT and supplies not subject to VAT were at a different level to those which it thought might be appropriate. In those circumstances it arranged to visit each of the five outlets on 25 November 2015 to invigilate the sales at each store. The HMRC officers kept records, store by store, and they have been produced in schedule form at pages 106 – 133 in our bundle of documents.

3.            On 19 February 2016, subsequent to the invigilation exercise, the respondents issued a Notice of Assessment in respect of the appellant’s VAT periods ended 28 February 2012 - 30 November 2015 under reference 60215900034. The assessment was in the sum of £78,318 and, as the respondents subsequently observed when a statutory review was undertaken, the assessment in respect of part of that time span was statute barred.

4.            On 14 September 2016 the respondents wrote to the appellant to inform it that it considered it appropriate to apply the results of the invigilation to the appellant’s VAT quarters ended 28 February 2016 and 31 May 2016. It is thus revised the amount of VAT said to be outstanding to the sum of £91,709.

5.            On 26 September 2016:

(1)    the respondents amended the assessment under reference 60215900034, to the sum of £78,640, and

(2)    issued an assessment in the sum of £13,069 in respect of the appellant’s VAT quarters ended 28February 2016 and 31 May 2016, giving the reference 60915900196.

6.            The appellant then requested a statutory review which was concluded by a Review Decision dated 17 November 2016.

7.            The result of the review was that the Reviewer concluded that the assessment of 26 September 2016 included the appellant’s VAT periods ended 31 May 2012 and 30 August 2012 for which assessment was out of time and so they had to be withdrawn. The Reviewer also concluded that the assessments needed to be reconsidered, upwards, because when the assessments were made it became clear that the assessing officer had used “the net figure as declared on the VAT returns and applied the relevant percentage to that, rather than using the gross sales declared.”

8.            The Reviewer also concluded that the assessment had been made to best judgement.

9.            The appellant appealed to this Tribunal, in time, and there has been a decision made by the respondents that the VAT need not be paid as a condition of this appeal proceeding.


Best Judgment

10.        The respondents purport to have made the assessments referred to above, using “best judgement” further to s73(1) Value Added Taxes Act 1994.

11.        When we consider whether one or more assessments have been made to best judgement, we take our guidance from the decision of Woolf J inVan Boeckel v C & E [1981] STC 150andRahman (No. 2) v HMRC [2003] STC 150. It is plain from a consideration of those decisions that six distinct principles emerge. They are that :

(1)    The respondents must be in possession of some material upon which a best judgement assessment can properly be based,

(2)    The respondents are not required to undertake the work which the taxpayer would ordinarily undertake so as to arrive at a conclusion about the exact amount of tax due,

(3)    The respondents are entitled to exercise their best judgement power by making a value judgement on the material available,

(4)    This Tribunal should not treat an assessment as invalid simply because it takes a different view as to how the best judgement could or should have been applied to the material available to the respondents. Before the Tribunal interferes, it needs to be satisfied that the purported best judgement assessment was wholly unreasonable.

(5)    The Tribunal is to start by assuming that the respondents have made an honest and genuine attempt to arrive at a fair assessment.




12.        The methodology adopted by the respondents to arrive at the figures in the assessments was as follows. The invigilators had recorded, in respect of each outlet for 25 November 2015, each sale made and annotated it with whether it was “eat in” or “take out”. In a different column in the annotation it was recorded whether the food was “hot” or “cold”. Those differences needed to be recorded because of the different VAT treatment in respect of hot food and cold food on the one hand and eat in and take out food on the other hand.

13.        Then the respondents added up, outlet by outlet, the number of transactions upon which VAT was due and those upon which VAT was not due. When the respondents undertook that exercise it resulted in them concluding that VAT was due on the following percentages of the sales made on that day :

Lee on Solent : 84%,

Woolston : 80%,

Portsmouth : 88%,

Swaythling : 88%, and

Bitterne : 89%.


14. When we read the papers and the parties’ respective Skeleton Arguments for this appeal, we were immediately concerned that the assessment methodology adopted by the respondents was significantly flawed. A simple count of transactions that did attract VAT and those which did not attract VAT might be capable of being appropriate in certain kinds of business where, for example, VAT is due at one rate on one service and at a different rate on a different service. It is accepted by all concerned that of the various methodologies that could properly be adopted, there is always an element of imprecision when undertaking anex post factoexercise. However, the complication in the instant case is that at any of the appellant’s outlets a customer might come into the shop to purchase several items that did attract VAT accompanied by other items that did not attract VAT. For example, a customer might buy cold sandwiches, hot coffee, a slice of cake and a bag of crisps. Two of those items would attract VAT whereas two would not. Assessment by reference to the comparative value of sales upon which VAT was and was not due would not have been a difficult auditing exercise to undertake because the till gavedetails of the make-up of all of the sales so that it would have been possible to distinguish goods on which VAT was chargeable and those upon which it was not due.

14.        Before proceeding further, we should record that in this appeal there has been no suggestion by the respondents that any of the appellant’s tills were being improperly manipulated or that sales were not being recorded through the till. To put it bluntly, it has not been suggested that this is a case of “fiddling the till.”

15.        The appellant complained that the methodology adopted by the respondents was flawed in the sense that it was not sufficiently refined to give rise to a reasonably reliable overall picture. The appellant argues that the exercise should have been undertaken by reference to transaction values, rather than the number of transactions. In other words, the respondent says that the respondent should have looked at the value of supplies made which did attract VAT as compared to the value of supplies made which did not attract VAT.

16.        At the outset of the hearing we raised our concerns about the respondents seemingly flawed and potentially misleading methodology, with Ms Ellis. Although this is not an appeal in which we have received any expert evidence we suggested to Ms Ellis that an expert statistician, forensic accountant or forensic mathematician would be alarmed to find that the methodology used by the respondents was considered to be either acceptable or such as to give rise to a reasonably reliable result. After this matter had been considered for some time, we adjourned for about 20 minutes to allow Ms Ellis to take advice and/or instructions on this issue. When we reconvened Ms, Ellis informed us that the respondents accepted that the underlying methodology, which had resulted in the appealed assessments, was flawed. In our judgement that methodology was flawed to such an extent that it would be wholly unreasonable, and unfair to the appellant, to base a best judgement assessment thereon. Ms Ellis did not argue otherwise.


Meal Deals

17.        By reference to the Review Decision it can be seen that the true area of concern on the part of the respondents was that Subway sometimes had promotions called “Meal Deals” whereby several products would be bundled together for a single headline price. In this appeal the meal deal’s relevance is limited to 1 January 2015 onwards because, we accept, prior to that time no relevant meal deal offers had been offered/sold.

18.        The appellant’s case is that from 1 January 2015 onwards a meal deal offer was available to customers whereby for the all in price of £3 a customer could purchase a sandwich (hot or cold) and a drink (which could be a fizzy drink or hot beverage upon which VAT would be due or bottled water upon which no VAT would be due). If the meal deal involved hot food, then it would be subject to VAT except for any part of the meal deal that involved bottled water instead of a fizzy drink or a hot beverage.

19.        The respondents’ complaint is that because of the way in which the appellant’s till was set up, it treated £2.99p of each meal deal as attributable to the sandwich (no VAT if cold) and only £0.01p to the accompanying drink which, if subject to VAT, would mean that the VAT thereon would be one fifth of one penny.

20.        There are two competing positions. The appellant takes the position that it is entitled to run its business as it sees fit and to make such commercial decisions as best suit its business. In effect the appellant says that he is entitled to sell loss leaders, as do many major retailers, or to sell stock at less than cost price if that somehow serves the best commercial interests of the business. In our judgement this is not a true loss leader situation. This is a situation where goods are packaged together to be sold at a single all in price. In those circumstances we agree with the respondents that we must look at the reality of the transaction when apportioning the part of the money paid by the customer between the various components within the package of goods sold.

21.        The exercise of apportionment is itself fraught with difficulty. Where one of the components has been bought in by the appellant, from a wholesaler, then there is a fixed starting point to ascertain the cost to the retailer of that particular product. However, in the case of shops selling sandwiches which are not bought in from a third party, it is far more difficult to arrive at a proper proportion of the overall price which has to be attributed to that food element. That is because one has not only to look at the cost price of the ingredients, which may be fairly modest, but also at the labour element to make and package the product. The appellant argues that labour is by far the largest cost component within the cost of a sandwich and the overall meal deal package.

22.        After we had identified these various issues the parties took some 20 – 30 minutes to consider whether an agreement could be reached as to a fair overall percentage uplift in respect of the proportion of supplies attracting VAT. When we returned, we were told that the parties were unable to reach any agreement and that each of them preferred the Tribunal to decide the issue.

23.        We invited each party to make representations as to how we should go about that task. We think we do her no injustice if we say that on behalf of the respondents Ms Ellis’ position was that she was unable to give any significant assistance and felt that the matter should be left entirely within our hands.

24.        However, having looked afresh at the decision inPegasus Birds Ltd v HMRC [2004] EWCA Civ 1015we note that in the judgement of Lord Justice Carnwath he referred with approval to the two-stage test adumbrated inRahman v C & E [1998] STC 826where it had been common ground between the taxpayer and the Commissioners that the Tribunal should adopt a two-stage approach which had been described thus:

“…. The practice is to consider these cases into stages: (1) consideration whether the assessment was made according to the “best judgement of the commissioners”; if not, the assessment fails and stage (2) does not arise; (2) if the assessment survives stage (1), consideration whether the amount of the assessment should be reduced by reference to further evidence of further argument available to the Tribunal ………”

25.        Given that we have found that the assessment fails the best judgement analysis because, as conceded by the respondents, it was methodologically flawed to the extent that the method actually used was not one that could amount to the respondents making a genuine and reasonable attempt to arrive at a reasoned assessment, on the foregoing authorities we need to proceed no further. That is because, as identified inRahman (No. 1),which was cited with approval inPegasus Birds, if we conclude that the assessments were not made to best judgement, that is the end of our function in this appeal.

26.        However, we should record the following given that we invited the parties to address us on what approach we should take if we had to undertake an assessment to arrive at the correct amount of additional tax payable, if any.

27.        We have no doubt, as noted above, that any calculation should be based on the values of transactions rather than numbers.

28.        On behalf of the appellant Mr Ahmed believed that the appropriate percentage uplift to take account of meal deals would be about 2% and submitted that in the exercise that we had to undertake it would be fairer to look at the results over a one-month period rather than those for a single day. He very fairly said he did not know whether that would or could result in a more or less favourable outcome to his client.


29.        In circumstances where the respondents do not attack the integrity or reliability of the sales recorded by the appellant’s tills, but only the way in which each till was set up when it came to recording drinks included within certain meal deal packages, we would have considered it appropriate to take the appellant’s position, as declared in each relevant VAT return, as the reliable starting point.

30.        It was a matter of agreement that those various returns showed that, on average, over the relevant period, the percentage of sales said to attract VAT were as follows:

Lee on Solent : 62%,

Woolston : 67%,

Portsmouth : 82%,

Swaythling : 68%, and

Bitterne : 59%.


31.        From the foregoing it will be apparent that there is a considerable differential between the appellant’s various outlets. The appellant contended, and we accept, that there will be variations from outlet to outlet. He explained that where an outlet is near to a large student population it is usual for a much higher percentage of turnover not to be subject to VAT because students tend to take food away rather than to eat in. He also recognised that there will be seasonal variations given that customers are more likely to want hot food during the winter months and cold food during the summer months. Thus it can be appreciated that there are numerous imponderables to be taken into account when undertaking an ex post facto exercise.

32.        Our methodology would be to take our best assessment of the proportion of the cost attributable to the food component of a cold food takeaway compared to the fizzy or hot drink component (which would attract VAT). There was no documentary evidence about the cost price of any fizzy drinks sold by the appellant, but he put the cost price at no more than 20p – 30p per serving because the drinks are dispensed from a pressurised tap rather than being sold in cans.

33.        The more difficult part of any assessment using the appropriate methodology would be to arrive at a fair and proper assessment of the cost price of the cold food component of a meal deal, including the labour element.

34.        We also record that we accept that meal deals only came on stream from 1 January 2015 and thus any best judgement assessment would not apply to any VAT period prior thereto.

35.        For the reason set out above this appeal is allowed and each of the assessments is quashed. Based upon the authorities cited above it is not our function to go on to undertake any kind of assessment to ascertain what, if any, additional VAT might be due.

Right to apply for permission to appeal

36.        This document contains full findings of fact and reasons for the decision. Any party dissatisfied with this decision has a right to apply for permission to appeal against it pursuant to Rule 39 of the Tribunal Procedure (First-tier Tribunal) (Tax Chamber) Rules 2009. The application must be received by this Tribunal not later than 56 days after this decision is sent to that party. The parties are referred to “Guidance to accompany a Decision from the First-tier Tribunal (Tax Chamber)” which accompanies and forms part of this decision notice.





Release date: 14 August 2019


BAILII: Copyright Policy | Disclaimers | Privacy Policy | Feedback | Donate to BAILII