First-tier Tribunal (Tax)

 

[2012] UKFTT 493 (TC)

 

TC02170

 

 

 

Appeal number: TC/ 2009/14072

 

VAT – assessment under section 73 – till readings suggesting undeclared output VAT – penalty under section 60.

 

 

FIRST-TIER TRIBUNAL

TAX CHAMBER

 

	BOUNDS GREEN SUPERMARKET  (a partnership)	Appellants
	- and -
	THE COMMISSIONERS FOR HER MAJESTY’S	Respondents
	REVENUE & CUSTOMS

 

 

TRIBUNAL:	JUDGE  CHARLES HELLIER
	CAROLINE de ALBUQUERQUE

 

Sitting in public At Bedford Square WC1 on 22 and 23 May 2012  with later evidence and written submissions

 

Altan Zorba and Naim Zorba, of A Zorba & Co, the Appellants’ accountants, for the Appellants

 

Patrick Way, instructed by the General Counsel and Solicitor to HM Revenue and Customs, for the Respondents

 

 

 

 

 

 

© CROWN COPYRIGHT 2012

DECISION

 

 

1.             Bounds Green Supermarket was a partnership between Mr. Nasir  Haidari  and Mr. Bashir Ahmad. It ran a convenience store in Haringey opposite Bounds Green railway station.

2.             HMRC  visited the supermarket on 6 November 2007. They interrogated two tills used by the business and took away a till roll they found there relating to an earlier period. Those checks and that till roll caused Christopher Nowak, one of HMRC's officers to conclude that the partnership had been understating its sales and that its VAT liability for a number of periods was higher than had been reported on its VAT returns. Assessments were issued and a penalty determination was made.

3.             The partnership appeals against the assessments for the periods 07/07, 10/07, 01/08 and 04/08, and we have taken its appeal to include one against the assessments for 07/08 and 01/09 (appeals which were notified to HMRC but not formally made to the tribunal, and to whose inclusion in the appeal HMRC and the appellant  agreed). These assessments total £82,859. The partnership also appeals against a penalty assessment of 80% of that sum, namely, £66,287.20.

The issues

4.             There were three issues before the tribunal:

(1)          whether the assessments had been issued to the best of HMRC's judgement as required by section 73 VATA 1994;

(2)          whether (and if so by how much) the assessments could be shown to be excessive;

(3)          whether the penalty was properly exigible and, if so,  whether the amount of mitigation given was appropriate.

5.             No question was raised as to whether the assessments were in time.

The Law.

6.             Section 73(1) VAT Act 1994 provides that where a person has failed to make returns under the Act or where "it appears to the Commissioners that such returns are incomplete or incorrect, they may assess the amount of VAT due from him to the best of their judgement and notify it to him.".

7.             Subsection (6) provides that an assessment must be made within the time limit provided for in section 77 and “shall not be made after the later of the following:

(a)  2 years after the end of prescribed accounting period; or

(b) one year after evidence of facts, sufficient in the opinion of the Commissioners to justify the making of the assessment, comes to their knowledge;

but (subject to that section) where such further evidence comes to the Commissioners' knowledge after the making of an assessment under subsection (1) ... another assessment may be made under that subsection in addition to the earlier assessment".

8.             Section 77 provides that:

“(1) subject to the following provisions of this section, an assessment under section 73 ... shall not be made -

(a) more than 3 years after the end of the prescribed accounting period or ... or

(b) ...

(4) subject to subsection (5) below [which is irrelevant in this case] if VAT has been lost --

(a)          as a result of conduct falling within section 60 (1) ...

an assessment may be made as if, in subsection (1) above each reference to 3 years were reference to 20 years.”

9.             Section 60 provides:

"(1) In any case where --

(a) for the purpose of evading VAT a person does any act or omits to take any action, and

(b) his conduct involves dishonesty (whether or not it is such as to give rise to a criminal liability),

he shall be liable, subject to subsection (6) below [which is irrelevant in this case], to a penalty equal to the amount of VAT evaded or, as the case may be, sought to be evaded, by his conduct….

“(7) On an appeal against an assessment to a penalty under this section, the burden of proof as to the matters specified in subsection (1) shall lie upon the Commissioners”

10.         Section 70 provides that where a person is liable to a penalty under section 60, the Commissioners, or, on appeal, a tribunal may reduce the penalty to such amount (including nil) as they think proper. Subsection (3) provides that certain matters irrelevant to this appeal shall not be taken into account in the exercise of powers under subsection (1).

11.         Section 76 provides for the assessment of penalties under section 60. Subsection (1) and (4) have the effect that such an assessment cannot be made more than 20 years after the event giving rise to the penalty.

12.         Section 80(1)(p) provides for an appeal to be made to this tribunal against an assessment or the amount of the assessment. An appeal against the assessment will be upheld if the assessment is not properly made. That will be the case if it not made to the best of the Commissioners’ judgement. The cases establish that, unless HMRC have acted vindictively or capriciously, where they do their best on the information available to them to calculate the tax due the assessment must not be struck down as not made to the best of their judgement. If the assessment is not struck down then the next question in an appeal will be whether the amount of the assessment should be upheld.

The Evidence.

13.         We had before us a bundle of correspondence, meeting notes, and copies of the till rolls obtained by HMRC's officers. We heard oral evidence from HMRC’s officers, Christopher Nowak and Mahendra Gajjir, from Altan Zorba and Naim Zorba, of A Zorba & Co, the Appellants’ accountants, and from Bashir Ahmad. All of these produced witness statements. Mr. Ahmad's partner and brother Mr. Haidari did not attend the hearing and provided no statement.

14.         During the course of the hearing it became apparent that the appellant had very little documentary or oral evidence to offer in relation to its business activity in the relevant periods. We asked about evidence of current activity and were told that CCTV footage might be available.  At the end of the hearing we directed that the appellant send a CCTV tape to HMRC, that HMRC comment on it and that the appellant be given an opportunity to address HMRC’s comments. This evidence and the parties’ submissions on it are discussed at [50ff] below.

 

Our findings of fact.

15.         The supermarket is opposite Bounds Green station. It faces stiff competition. There are eight other food and drink shops clustered round the station including a Metro Express and a Tesco Express. There is even another supermarket with the same name just around the corner.

16.         The partnership obtained a licence to sell alcohol on a 24-hour basis in January 2007, and from 1 November 2007 it opened 24 hours a day. Before that it had opened for 14 hours a day. It sells cigarettes, beer and wine and alcohol, fresh fruit and vegetables, and grocery.

17.         Mr. Ahmad told us, and we accept, that most customers make small purchases and that this is not the place where they do their weekly shop. He estimated that the average purchase was £3-£4. The evidence in the till rolls supported a figure of that order.

18.         Mr. Ahmad told us that four full-time (including himself) and four part time people staffed the shop. That was broadly consistent with a licence application which indicated that nine people were employed. We accept that evidence. At least two members of the staff were on duty at any time. At busy times two tills would be in operation and there would be a small queue at the tills. The competition means that customers would not wait for too long. Mr. Ahmad told us it took 2 to 3 minutes to serve each customer and told us of difficulties with the card machine and of the time taken discussing their families with known customers. We thought that 2 to 3  minutes was a little exaggerated and doubt that it would have taken more than one or two minutes on average to serve each customer, and this was confirmed by the CCTV footage.

19.         In reply to questions from the tribunal Mr. Ahmad gave his rough estimate of the supermarket’s sales in an average week, and its  profits (before overheads) as follows:

	Sales £	Gross Profit  %	Gross Profit £
Cigarettes	2500	5%	125
Alcohol	2500	20%	500
Fruit and Veg	2200	10%	200
Grocery	9000	25%	2250
Total	16200		3075

 

20.         Mr. Ahmad told us that his brother was responsible for the purchase of alcohol for the business and he was unsure of the mark up obtained. He thus was tentative about the figures relating to alcohol in the table above.

21.         These figures are equivalent to a yearly turnover of £840,000 and a yearly gross profit of £160,000. These figures were not inconsistent with, although they were slightly higher than, the turnover and profit shown in the partnership's accounts for 2009 of £700,000 and £130,000 respectively.

22.         Mr. Ahmad told us that, after paying staff and overheads, he and his brother took home £192 per week after tax. That is some £10,000 per annum after tax. The accounts for 2008 showed a net profit of some £60,000. We therefore did not believe this evidence. His statement made us consider that the estimates in the preceding paragraph may also have been on the low side.

The Tills.

23.         The supermarket had two tills, a CS- 6000 and a CS 240 CR. The C-6000 is more sophisticated.

24.         Mr. Ahmad told us that he printed Z reading from the tills every day or so and sent or took them to Zorba & Co. Mr. Zorba confirmed to us that he received the till rolls and that for the CS-6000 these included daily Z readings but did not include periodic ZZ reports (we shall explain the significance of these terms later). We conclude that daily rolls and Z reports were given to Zorba & Co. but that the ZZ reports were not.

25.         The CS-6000 has a row of four keys at the top of its keyboard which may be used to designate operation by different persons or "clerks". They are referred to as Clerk 4, Clerk 3, Clerk 2, and Clerk 1. The Clerk 4 key was at the left-hand end, and the Clerk 1 key at the right-hand end of the keyboard. The person operating the till would press one of these Clerk buttons and a green light lit up under or near it to indicate which Clerk was operating.

26.         We concluded that the Clerk 4 key would not be activated accidentally or as a matter of course; Clerk 1 would be the normal choice.

27.         Mr. Nowak told us and we accept that Clerk 4 was set by the programming of the till as in "training Clerk". The sales generated when a training Clerk was in operation were not printed out on the daily journal and, as a result of that, the noise of printing would be absent when Clerk 4 was being used. Mr. Ahmad said that he had not noticed a different sound when he had been using the till.

28.         Mr. Ahmad explained that the ability to use another “Clerk” was used when in the course of serving one customer he or she went to get something they had forgotten from the shop and another customer was waiting. By transferring from say Clerk 1 to Clerk 3, the waiting customer could be served without clearing the total already entered into the till for the more forgetful customer.

29.         Underneath the Clerk buttons was a keyboard for the monetary amounts, and to the right keys to describe the type of goods being sold and the method of payment,  and a subtotal key to provide the amount due by the customers and receipts.

30.         The CS-6008 till’s the programming enabled various totals and analyses to be printed off. In particular:

(1)          a daily journal of all sales was printed as sales were made. The journal print out did not include any sales made by a Clerk designated as a “training clerk”;

(2)          an "X" Reading of the sales in certain categories. Printing an  X reading did not reset the till memory;

(3)          a daily or "Z" Reading which gave the cumulative sales and number of customers since the last Z reading and cleared the relevant memory so that the following Z reading would show only transactions since the last one. This reading did not include sales made by any Clerk designated as a "training Clerk".

(4)          ZZ readings which were periodic summaries of transactions (by category if desired). including those undertaken by Clerks designated as "training Clerks". The readings showed the transactions which had taken place since the last ZZ reading;

(5)          a "post-receipt", that is to say a copy of the last receipt produced as a result of the most recent sale by any particular Clerk;

(6)          a printout of the till’s programming.

31.         Mr. Ahmad told us that the CS-6000 had been purchased by his brother second-hand from a nearby market. He said that on its acquisition they knew little about its programming. A printout produced by Mr. Nowak showing grand total sales to date made on the till enabled us to conclude that the till may have been purchased second-hand.

32.         The other, less sophisticated CS 240 CR, till kept a record of sales to date. It had a key which could be depressed for a "no sale". This opened the till but no transaction was recorded other than the fact that the "no sale" key had been used.

Reconciliations

33.          Mr. Ahmad told us that COD suppliers were paid from the till and their invoices would be kept. The till would be opened to pay for coffee and such things at lunchtime. Generally he said that every day he would reconcile the amount in the till with money paid out and sales made. He said he did these calculations using a calculator and kept no record of it.

HMRC's visit on 6 November 2007.

34.         Mr. Gajjar and a colleague made an unannounced visit to the supermarket on 6 November 2007. Mr Gajjar made some test sales on the tills and printed off a number of reports. Whilst looking for a fresh till roll to make the printouts he found an old till roll print at the bottom of the CS-6000 till.

35.         During the course of this visit his colleague found alcohol for sale which did not bear a duty stamp.

Findings from the printouts gathered at HMRC's visit.

36.         Our conclusions in this section derived principally from the evidence of Mr. Nowak whom we found expert and reliable.

37.         The CS-6000.

(1)          When a Clerk is designated as a training clerk the sales made by that clerk are not entered on the daily journal or included in the daily Z readings, although the till was programmed so that they did appear on the ZZ periodic readings.

(2)          Clerk 4 had been designated as a training clerk.

(3)          Someone had used the Clerk 4 button at 1:12am on the morning of 6 November 2007 (although for a sale of £0.10p only); it had not been used between that time and  HMRC’s visit.

(4)          The till had been programmed so that in order to print a ZZ report a pass number had to be input.

(5)          On 6 November 2007 between 1:12 am and 11:41am 174 transactions had been recorded on the till, none from Clerk 4, with a total value of £579.

(6)          A periodic ZZ report printed at 11.42am on 6 November 2007 showed the same number and value of sales as the daily Z report printed at 11.37am. Therefore no Clerk 4 sales were in the memory at 11.42 to be included in the ZZ totals. That meant that a ZZ report had been taken after 1:12am when the last Clerk 4 sale had been made (see (3) above), because only by taking such a report would the ZZ memory have been cleared of the Clerk 4 sale at 1:12am. To take a ZZ report the pass code had to be inserted. That meant that between 1:12 am and 11.37am someone must deliberately have taken a ZZ report showing the total sales made (including the Clerk 4 sales)

(7)          that ZZ report was not made available to HMRC.

(8)          Mr. Zorba's confirmation that he had received only Z reports meant that the ZZ reports were not sent to the accountants.

(9)          The old till printout found by Mr. Gajjar contained both Z and ZZ readings. This showed Z readings for total sales (excluding training clerk 4) of £1416.70 (after adding two sales between the times of the two parts of the printout), and total sales (including Clerk 4) of £3185.49 which included £1768.79 of Clerk 4 sales. There were 359 Clerk 4 transactions and 434 others. The mean amount spent by each customer was thus about £4.

(10)      The length of the period to which the readings from the old till roll related was not clear from the till roll itself, but (A) Mr. Ahmed told us that they took Z readings every one or two days, (B) the sequential numbers printed at the time Z and ZZ readings were taken showed that this old roll was the 1903rd Z reading taken from the till, and the 699th ZZ reading. The Z reading taken on 6 November was the 1936th and the  ZZ reading on that date was the 715th. Thus 33 Z readings and 16 ZZ readings separated this old roll from the 6 November roll. That put the date of the old roll in early October 2007.

38.         A day’s sales of £1416.70 would indicate sales for a quarter of £130,000. The declared turnover for the 10/07 quarter was initially £120,852 but in a later return (which may have been for both 07/07 and 10/07) was shown at £240,503. A day’s sales of £3185 would indicate a  turnover of about £290K.

39.         The CS 240 CR

40.         The daily Z printout taken on 6 November 2007 showed 53 sales (since the last Z reading) and 12 "no sales". The periodic report (for an unidentified period ending at about the same time) appears to show 2446 sales and 2763 "no sales", but another report showed sales of 92,446. There appeared to be a truncation of the number of total sales. Mr. Nowak concluded that the leading digit "9" was truncated/omitted on the periodic report and that this report was for a substantial period perhaps longer than that for which the business had owned the till. We agreed.

The accounting records

41.         Mr. N Zorba explained that in early 2008 his firm had moved to new premises. They put their archives in the basement. These included the principal accounting records of the partnership. In March 2008 a burst sewer flooded the basement and about half their records were destroyed. In 2009 the problem reoccurred and they lost the bulk of their archives. As a result he told us that none of the basic accounting records of the supermarket were available -- no till rolls, no purchase invoices, no banking records, no journals. They did however have the Sage nominal accounts records for the periods of the assessment; these they sent to the tribunal after the hearing in response to a direction from us so to do.

42.         In response to a direction from the tribunal  Zorba & Co provided a copy of a letter from loss adjusters settling their claim in respect of an “escape of water from underground pipe” on 26 January 2008, and a letter from Thames Water confirming that “your property was flooded by sewage on 23 March 2009”. We accept that flooding occurred on these dates. HMRC say that even if the flooding took place a letter from Zorba & Co to HMRC of 4 April 2008 ( after the first water leakage and before the second) says that the appellant’s records were “misplaced” during the move to new premises and makes no mention of the first flood. They say that it is not credible that all the supporting evidence was lost innocently. Zorba & Co reply that when that letter was sent they had not realised the extent of the water damage, and that after realising that the records had been destroyed by the flooding they had notified HMRC. Overall our conclusion was that it was not shown that Zorba & Co had had records which indicated that the appellant’s VAT turnover had been greater than they had reported. 

43.         Mr Zorba also told us that the partnership had ceased trading in October 2009, although it appeared that the supermarket was continuing to trade. Thus it was not possible to try to estimate the turnover of the partnership’s business by looking at records from later periods and using them as an indication of trading levels in earlier periods.

44.         The VAT returns.

45.         For the periods 07/07 and 10/07 in which the shop opened 14 hours a day the returns submitted showed output tax liabilities of £6k and £12k. For the five periods 01/08 to 01/09 in which the shop was open 24 hours a day the returns showed output tax liabilities of £13k, £10k,£12k, £9k, and £10k, that is £54k in total. The returns showed total outputs of (very approximately) twenty times output tax, indicating that about 1/3 of declared outputs were standard rated.

46.         The Assessments

47.         Mr. Nowak made his assessments thus. On the evidence of the old till roll from the CS-6000 he calculated that £1774.15 of sales had not been declared. He assumed that this represented a single day's sales. He then estimated that each day's sales should be increased by this amount. That meant an increase in each quarter’s sales of some £163,000. He took 50% of this to be standard rated sales, the remainder being zero rated, and calculated the additional VAT output tax due on these additional standard rated sales. He made an adjustment for input tax for the periods 7/07 and 04/08. And on this basis he issued assessments for 07/07, 10/7, 01/08, and 04/08 on 12 September 2008.

48.          These assessments were withdrawn on 11 May 2009 and replaced by assessments calculated on the basis that the suppressed Clerk 4 sales for each day were 55.5% of total sales (being the proportion which the Clerk 4 sales represented of the total sales shown on that till roll). They then took the output tax to be understated by this proportion. An assessment on this basis for 07/08 was made on 22 October 2008, and for the remaining periods from 07/07 until 01/09 assessments were made on 6 November 2009. These assessments total £67,784.

Later evidence: CCTV

49.         As we explain at [14] above, given the lack of evidence made available to us by the appellant, we directed that a CCTV recording be sent to HMRC and the tribunal in the hope that this might provide some objective evidence of the level of activity at the supermarket. The appellants complied with this direction and sent a CCTV recording for 5 June 2012. Although this was some years after that periods of the assessments under appeal, the place of business was the same and there was no evidence that there had been a material change in its nature or activities. It was therefore in our view relevant evidence of the level of activity at the supermarket in the period relevant to the appeal.

50.         5 June 2012 was the second day of the Jubilee public holiday. The appellant made no submissions as to whether trade on that day was greater or less than that on a normal day. We conclude that the level of trade on that day was not greater than that on a normal working day.

51.         HMRC made submissions on the CCTV recordings in accordance with our directions. Mr. Gajjar watched the whole 24 hours of the recording. He produced an analysis of the number of customers served during each successive hour of the day. The appellant did not dispute his analysis and we accept it as accurate. That analysis showed 834 customers being served during the day with a peak hourly trade being between 4 pm and 8 pm when up to 75 customers were served per hour. The slackest periods were between 2 am and 4 am, when in total only 14 customers were served. Between midnight and 11 am about one third of the total customers for the day were served (this was of interest in relation to the numbers disclosed by the tills at HMRC's visit of 6 November 2007.)

52.         Mr. Nowak and Mr. Gajjar also recorded the time taken at a till between the placing of goods on the counter and the cash drawer being closed. They timed the first 9 customers of the day. They found that those customers spent between 10 seconds and 3 minutes 49 seconds with a mean of 52 seconds per customer. Both tills were used but only two of the nine sales were made on the second till. We accept their analysis.

The parties’ arguments

53.         Mr. Way said that there was evidence on which HMRC could reasonably have based its assessment. Whether or not it was wholly accurate, it was made to the best of the officers' judgement. The burden of proof in relation to the amount rested on the appellant. It is clear that there had been intentional suppression of declared takings: that was dishonest. Section 60 was engaged. The mitigation of 20% was in line with HMRC's guidelines.

54.         Mr. N and Mr A Zorba said:

(1)          that it was accepted that sales by Clerk 4 had been omitted from the partnership’s VAT returns;

(2)          but those omissions had been inadvertent. There had been no intention to suppress sales;

(3)          the assessments were excessive. Zorba & Co offered a different approach to the estimation of its total sales in what they described as a Means Test calculation. This calculation estimated the value sales by estimating the number of customers who would be served in a day and multiplying it by an average spend per customer of £4. That calculation suggested an under declaration of some 18% in periods to 10/07, and between 7% and 18% in the periods  04/08 to 01/09. HMRC's figure of 55% was unrealistic.

Discussion

55.         It was clear to us that there had been an intentional under reporting of the partnership’s VAT liability. That was because:

(1)          Clerk 4 sales had not been included in the till rolls sent to Zorba & Co and therefore had not been included in the VAT returns.

(2)          The use of the Clerk 4 button must have been intentional.

(3)          Clerk 4  sales were likely to have been substantial.

(4)          The deliberate printing of the ZZ reports indicated that someone at the business (Mr. Ahmad or his brother) kept track of the Clerk 4 sales and knew their amount (see [38(6)] above).

(5)          Mr. Ahmad would not have been able to conduct the reconciliation of the till monies unless he knew of the amount of the Clerk 4 sales. He also must have known that the Clerk 4 sales were not contained in the journal or the Z reports. Otherwise his reconciliation would never have balanced.

(6)          The old till roll showed that it was known by someone in authority in the business (which is likely to have been Mr Ahmad or his brother) that the Clerk 4 sales were not on the Z report.

(7)          Thus it was known to the partners that the information on which Zorba & Co prepared the VAT returns did not include all the sales made.

56.         The intentional under reporting was in our view dishonest.

57.         In our view the assessments made by HMRC were made on the evidence available to them and to the best of their judgement. There was nothing to suggest that they were vindictive or capricious. There was evidence upon which they could have based the assessments.

58.         We conclude that there was intentional under declaration which was as a result of dishonest conduct by one or both of the partners in the business. Section 60 applies if (a) the taxpayer does any act or omits to take any action for the purpose of evading VAT; and (b) that conduct involves dishonesty. The failure to deliver to Zorba & Co details of sales by Clerk 4 and of sales made using the no sales button was clearly for the purpose of evading VAT: para (a) was therefore satisfied. That evasion was dishonest: para (b) is satisfied. We conclude that the partnership was liable to penalties under section 60. If we find that the amounts of tax sought to be evaded are not shown to be different from those assessed, then the amount of the penalties before mitigation must be upheld.

59.         So far as the mitigation of the penalty goes, none of the evidence before us persuaded us that there should be any further mitigation. The under declaration was deliberate and dishonest. There were few mitigating circumstances, save the later cooperation with HMRC for which a 20% reduction was generous.

60.         The only issue with which we had more difficulty was the issue of quantum: the determination of what the likely VAT liability should be.

61.         The assessments made by HMRC are subject to a number of criticisms:

(1)          they are based on one day’s till rolls. The activity on that day may have been unrepresentative of a longer period. On other days the persons on duty may have been different; the mix of items sold may have been different and may have affected the use of the Clerk 4 button;

(2)          it is implicit in the assessments that the suppression of sales was distributed between standard and zero rates in the same proportion as the declared sales; the proportion might have been different; and

(3)          it is also implicit that inputs were not also suppressed.

62.         Of these (1) and (2) could lead to a conclusion that the VAT properly due either greater or less than that assessed; (3) could only reduce the assessable amount. These are difficulties with the assessments but to our minds they do not show that the assessments were not made to the best of the officer’s judgment: they were deficiencies in the information available.

63.         The Zorba & Co alternative "means test" calculation works on the basis that, broadly, on average some 28 customers were served per hour and that each spent £4. On that basis they estimated turnover of some £2,600 for a 24 hour a day, £243,000 per quarter or £970,000 per annum. That indicates a lesser under declaration.

64.         The problem with the Zorba & Co calculations is that we had little good or weighty evidence relating to the assumptions behind their calculations from which we could conclude that their assumptions were likely to be accurate. Whilst we had some evidence of the spend of each customer, we had less which could help us come to a likely figure for how many customers were served on an average day.

65.         So far as concerns the mean spend per customer the evidence before us was this:

(1)          Mr Ahmad’s evidence that it was £3-£4;

(2)          the evidence from the old till rolls  that the average spend was £5 on a suppressed transaction and £3 on a non-suppressed transaction and on average across all CS – 6000 transactions was £4; and

(3)          the average spend from the 6 November reports was £3.30 over the period from midnight to about 11:30am .

66.         From this evidence we conclude that it is likely that the average spend on a suppressed sale was about £5, and overall that the average spend was likely to be about £4.

67.         So far as concerns the number of transactions per day we had the following evidence:

(1)          Mr. Ahmad's estimate that a customer took two minutes to be served suggested that 30 customers could be served at one till in an hour. The pictures we saw of the store showed customers within the shop but not at the till. We doubted that 30 customers an hour were served in the early hours of each morning, but when two tills were in operation in the rush hour 60 customers per hour did not seem unreasonable. The most we could conclude from this was that in a 24 hour day it was likely that no more 1200 customers were served;

(2)          the old till rolls indicate that 793 customers had been served on the CS - 6000. The period of the sales was not known but it seems more likely than not to have been about one day. Allowing for sales through the CS 240 of about a third of this number (the evidence recounted in (3) below suggests such a  split was likely, and the CCTV evidence suggested a ratio of 2:7 in the early hour of the day (and slightly greater usage might be reasonably be expected when the shop was busier), this evidence suggests that it is likely that no more than 1100 customers were served in one day.

(3)          the 6 November 2007 till readings suggest that between midnight and 11.30 am 174 customers were served on the CS-6000 and on the CS 240, 65 customers. The CCTV evidence suggested that 30 % of the sales took place between midnight and 11.00 am .This evidence points to a minimum of some 790 customers a day;

(4)          Mr Gajjar told us that while he was in the shop on 6 November he noticed about 6 customers. He was there for about an hour from 11am. That suggests a minimum of 150 customers in a 24 hour day; and

(5)          the CCTV evidence showed 834 customers on a  bank holiday.

68.         None of this was sufficient for us to conclude that it was likely that the average number of customers served was less than 1100 per day. We conclude that it is likely that it is likely that between 800 and 1100 customers were served on average each day when the shop was open 24 hours a day.

69.         In their submissions on the CCTV evidence Zorba & Co suggest that we should not compare today's level of business with that of five years ago. They say that "the basic rule of business is that growth is expected each year, [and] therefore if we were to apply an average 5% growth to each year, the reverse calculation would suggest that the average day in 2008 would consist of no more than approximately 680 people. We must also take into consideration [that] the majority of customers who visit the business are those of eastern European origin, and that their numbers have increased over the years."

70.         We are unable to accede to Zorba & Co’s submissions. First we had no evidence at all that the business had in fact grown - nor that it had  contracted – or of any reasons for any change; or  that the number of customers had changed..  Second our approach to the evidence that is available has been to consider all the sources rather than to rely solely on one.

71.          We conclude that it is not proved that it was likely that less than 1100 customers per day were served on average and that it was likely that on average more than 800 were served in a day .

72.          Coupled with our conclusion that the average spend was £4, that indicates that it is likely that daily takings were in the region of £3200-£4400. That represents weekly takings of £22,000-£31,000 - in excess of the £16,000 obtained from Mr. Ahmad's estimates at [20] above; and yearly takings of  £1.2-£1.6 million when the shop was open 24 hours a day, in excess of the turnover of some £800,000 shown in the accounts for 2009 (accounts prepared from the records presented to Zorba & Co and for a period in which the shop was open 24 hours a day.).

73.         Mr. Ahmed's estimates suggest that no more than 50% of the turnover was standard rated. The proportion declared as standard rated in the business’s VAT returns seems to have been about 1/3 (see [46] above). The Sage accounting records from Zorba &  Co indicate a split of about 2:1::zero rated:standard rated. It seems likely that more than 33% and less than 50% of the turnover was standard rated. That means the standard rated turnover was between £390,000 and £0.8 million. That would suggest annual output tax when the shop was open for 24 hours a day of between £58,000 and £120,000 per annum. Comparing that with the declared output VAT for such periods which was in the order of £50,000 per annum suggests there was under declaration of VAT of between £8,000 and £70,000 pa when the shop was open for 24 hours per day. That is an under declaration of up to more than 100% of the declared tax. It seems likely that the under declaration in the periods when the shop was open for fewer hours was the same.

74.         The assessments were made on the basis that sales equivalent to 55.5% of the declared sales had been suppressed. Given our conclusion in the preceding paragraph we do not find it proved that this was an overstatement.

75.         There is no evidence available to us as to whether input tax had been understated. We cannot conclude that it was likely that it was.

76.         Thus on the evidence available to us we cannot say that it is likely that the assessments exceed the under declarations.

Time limits

77.         Given our finding that section 60(1) applies, the back stop time limit for the making of an assessment provided for by section 77 is 20 years; but in any case the time within which an assessment may be made is also subject to section 73(6): if not made within 2 years of the period of assessment it must be made within one year after evidence of facts sufficient in the opinion of the Commissioners to make the assessment came to their attention. 

78.         The appellant raised no argument before us in relation to time limits. But we could not uphold an assessment which was not permitted by section 73.

79.         The assessment for 07/08 was made on 22 October 2008. That was plainly in time.

80.         The assessments for 07/07 to 01/09 were made on 6 November 2009. The two year limitation in section 73(6) means that the assessments for periods ending after 6 November 2007 were in time.  However the periods 07/07 and 10/07 fall more than two years before the date of that assessment. As a result an assessment for any of those periods is permitted only if it falls within section 73(6)(b) ie if it was made within 12 months after facts sufficient to make the assessments came to the Commissioners attention. The only facts relevant to these assessments appear to us to be those collected on the visit of 6 November 2007 (since after that date it appears that the only additional information obtained by HMRC was the failure or inability of the appellant to produce any records).  That was more than 12 months before the date of this assessment. These two assessments therefore appear unlawful.

Conclusions

81.         We find that the assessments were made to the best of the officer’s judgment; We find that it was not proved that they were excessive. We find that HMRC have proved that the understatements were dishonest. We are not minded to increase the mitigation applied to the penalties.

82.         We dismiss the appeals save in relation to the 07/07 and 10/07 assessments in relation to which we direct that if HMRC wish to argue that they were in time they send written submissions to the tribunal (with a copy to the appellant ) within 21 days of the issue of this decision. In the absence of such submissions the appeal in relation to those assessments is allowed. But the penalty in relation to those periods is unaffected since it is not dependant upon an assessment of the tax being made, and the time limit prescribed by section 77 for penalties under section 60 is 20 years after the event giving rise to the penalty.

 

 

 

 

 

 

Rights to Appeal

83.         This document contains full findings of fact and reasons for the relevant decision. Any party dissatisfied with a 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.

 

 

CHARLES HELLIER

TRIBUNAL JUDGE

 

RELEASE DATE:  1 August 2012