First-tier Tribunal (Tax)

[2014] UKFTT 232 (TC)

TC03372

 

 

 

Appeal number: TC/2012/02977            

 

VAT, under declaration of output tax; assessment; whether to best judgment; whether excessive; VATA 1994 ss73(1), 83(1)(p).

 

 

FIRST-TIER TRIBUNAL

TAX CHAMBER

 

 

	LUIGI PIA & SONS	Appellant
	- and -
	THE COMMISSIONERS FOR HER MAJESTY’S	Respondents
	REVENUE & CUSTOMS

 

 

 

TRIBUNAL:	JUDGE  J GORDON REID QC, FCIArb
	Dr HEIDI POON, CA, CTA, PhD

 

 

 

Sitting in public at George House, Edinburgh on 28, 29 and 31 October, 6, 17 & 18 December 2013

 

 

Philip Simpson, Advocate, instructed by Haines Watts, Chartered Accountants, Edinburgh, for the Appellant

 

Mrs E McIntyre, Officer of HMRC, for the Respondents

 

 

 

 

 

 

© CROWN COPYRIGHT 2014

DECISION

 

 

Introduction

1.             This is an appeal by a fish and chip shop business against assessments to VAT issued on the basis that sales have been significantly under-declared over a number of years.  The appeal raises the question whether the assessments were made to best judgment and if so, whether they are, nevertheless, excessive.  No novel issue of law arises.

2.             A hearing took place at George House, Edinburgh on 28, 29 and 31 October, and 6, 17 and 18 December 2013.  Philip Simpson, advocate, appeared on behalf of the appellant on the instructions of Professor McDougall of Haines Watts (HW), Edinburgh, chartered accountants. Mr Simpson led the evidence of Silvio Pia, Franco Andreucci, and Alistair Banks, the appellant’s accountant.  Mrs E McIntyre appeared on behalf of the Respondents (HMRC).  She led the evidence of George Longman and Bill Hadley.  There were also statements from several other HMRC officers who were involved in the investigations, namely, John Shearer, Colleen Craig, and Sarah Green, and Grant McNaughton. 

3.             A joint bundle of documents was produced.  All witnesses who gave oral evidence produced written statements.  These were somewhat brief given the overall length and detail of the evidence, in particular, by Mr Longman.

The Assessments

4.             Assessments, as amended, were raised on 10 January 2012 in the sum of £42,012 (covering the period between 1/4/06 and 31/3/08), and £21,610 (covering the period between 1/4/08 and 31/3/09), amounting in total to £63,622.  No surcharges, penalties or interest have been levied.  The original assessments, raised on 19 February 2010 (£59,425) and 4 June 2010 (£52,868) were subsequently reduced on internal review (see paragraphs 53, 55 and 61 below).

Grounds of Appeal

5.             The short grounds of appeal are that ‘The 2 assessments covering VAT periods 06/06 to 03/08 and 6/08 to 03/09 totalling £63,622 are based on flawed logic and are excessive.’ 

Facts

6.             The appellant is a partnership.  The four partners are Silvio Pia, Franco Andreucci and their wives.  They have carried on business as owners and operators of fish and chip shops since 1992 when the partnership was registered for the purposes of VAT.  They had business premises at 2 Polton Street, Bonnyrigg, Midlothian (the Bonnyrigg shop), and 1 Dalhousie Court, Poltonhall, Midlothian (the Poltonhall shop).  They ceased trading at the Poltonhall shop when the lease came to an end in June 2010 (outwith the periods of assessment).  Mr and Mrs Pia ran the Bonnyrigg shop and Mr and Mrs Andreucci ran the Poltonhall shop.  They were assisted from time to time by part-time staff, some of whom were family members.

7.             At lunchtimes, the Bonnyrigg shop was often very busy with school children.  In order to deal with them quickly, money from each sale was deposited in a plastic container and the individual sale was not rung up on the till at the time.  How change was given was not discussed in evidence. In his precognition, Mr Pia states that the total in the container was rung up as a single entry after the lunchtime rush.  Neither the written statements nor the oral evidence identified any entries in the till rolls or other documents produced as specifically representing these lunchtime takings at the Bonnyrigg shop being rung up as a single entry.

8.             A routine visit to the Bonnyrigg shop was carried out by Mr Longman on 7 June 2006.  In the course of the inspection he found that no audit rolls had been kept and that the cash register (till) reports were not treated as the final sales record.  He noted that various till readings were missing. He discovered that the appellant’s accountant had made the additions to the accounts as described below.  He advised the appellant inter alia of the need to have and to keep till rolls.  At this stage, Mr Longman reasonably took the view that the business records of the appellant did not disclose the true level of sales.

9.             Mr Longman also visited the appellant’s accountant on the same day.  He was informed that the turnover in the annual accounts for the years to 31 March 2003, 2004 and 2005 had been revised upwards.  Undeclared VAT had been added back into the VAT returns retrospectively in the ensuing two quarters following the year ends in question.

10.         Mr Alistair Banks of Banks Richardson Scott & Co, accountants, Edinburgh, has acted for the appellant since about 2002, writing up the books, preparing quarterly VAT returns, annual accounts, and the personal tax returns of the partners.  When preparing the appellant’s annual accounts for the year to 31 March 2003 he took the view that the appellant’s gross profit percentage (GPR) was too low; that HMRC would not accept such figures in the accounts and would initiate a tax enquiry.  He explained this to the appellant and advised that voluntary disclosure of additional sales for both income tax and VAT purposes should be made.  Mr Pia and Mr Andreucci accepted this advice although they maintained in evidence that their figures were accurate and did not accept that they had failed to disclose or account for substantial sales.    

11.         Mr Banks explained and we accept his evidence, that for 2003, before adjustment, the GPR for Bonnyrigg shop was 43.3% and for the Poltonhall shop it was 41.6%.  The accountants adjusted the sales for 2003 by the sum of £52,800 to bring the GPR to 49.1%. For the years to 31 March 2004, similar adjustment was made of £32,000 (but for the Bonnyrigg shop only) and £65,000 for 2005, ie an addition of £149,800 over three years.

12.         HMRC produced, at the request of the appellant for the purposes of this hearing, a summary of their records relating to the gross profit percentage (GPR) of fish and chip shops in Scotland.  These disclosed that the average GPR of such shops outside Aberdeen, Dundee, Edinburgh and Glasgow was 53.97% for the year 2007/08 and 55.16% for the year 2008/09.  The average in Scotland overall for these years was 54.32% and 55.22%, and for the whole of the United Kingdom, the average was 57.29% and 57.31% for these years.

13.         The GPR for the Bonnyrigg and Poltonhall shops was as follows:-

 

The second column relates to the Bonnyrigg shop, the third column relates to the Poltonhall shop.  The Poltonhall shop was closed in June 2010.

 

14.         The figures for 2003, 2004 and 2005 in the above table are after adjustments were made by the appellant’s accountants, Banks Richardson Scott & Co.  These adjustments were made to bring the gross profit percentage in line with perceived averages.

15.         In October 2006, Mr Longman wrote to the appellant in connection with the revision of the 2003, 2004, and 2005 accounts to take account of what he described as excess cash identified.  He noted that, from the records to March 2006, the recording of sales was still not adequate.  He indicated that he would return in about 12 months to review the appellant’s business records and the accuracy of its returns. He warned the appellant to ensure that it retained for six years, till rolls from all till systems and other business records.

16.         On 18 April 2008, Mr Longman visited the Bonnyrigg shop to ascertain the levels of sales indicated by current records.  A partial cash reconciliation with till readings was carried out but this has not been relied upon.

17.         At about 11pm on the same day (18 April 2008), officers Hadley and Baird made an unannounced visit to the Poltonhall shop.  The purpose was to witness a cashing up exercise.  The till X reading was £885.67.  The cash counted amounted to £1080.  That sum excluded a float of £100.  However, there was also about £40 in the till at the start of the day.  There was a discrepancy of about £155 (£1080 less £40 less £885.67).

18.         On 23 April 2008, Mr Longman visited the Bonnyrigg shop again.  He inspected the cash register (the till) and obtained till receipts. 

19.         Mr Longman’s examination revealed that the recently installed till at the Bonnyrigg shop had not been set up in accordance with the standard settings.  Such standard settings facilitate the auditing of processed transactions.  The non-standard settings which Mr Longman found facilitate the concealment of transactions and render or can render the till readings inaccurate. 

20.         Mr Longman explained the operation of the tills. In particular he explained that a Z1 report is a final daily reading showing the total takings for the day.  These were collected each night and at the end of the week put in a bundle and passed to the appellant’s accountant.  This procedure was carried out by Mr Pia at the Bonnyrigg shop and by Mr Andreucci at the Poltonhall shop.  A Z2 report showed takings over a longer period than a Z1 report, an X reading is a running total.  Mr Longman also explained how the till could operate in training mode or training clerk mode. In these modes, no transactions are recorded in the audit roll, ie those transactions would not feature in the totals of takings from day to day.

21.         The till at the Bonnyrigg shop had been purchased second-hand by the appellant on or about 20 March 2008 from ACR Retail Systems, Glasgow, when it was first brought into use at that shop.  When the appellant first started operating the new till, they did not adopt the manufacturer’s settings but used the pre-existing settings programmed by the previous owner.  Whether this was deliberate or accidental is unclear.  Neither Mr Pia nor Mr Andreucci had the skill to program or re-program the till settings in any significant way.  They were unable to explain in evidence in any comprehensive or comprehensible way what the various functions of the till were, or how in practice the till was operated and checked.  It is unlikely that either of them tampered with the settings. Whether Mrs Andreucci or Mrs Pia could do so or did so is unknown.  They did not give evidence although they attended throughout the hearing or most of it.

22.         Mr Longman reset the register to a conventional format with Mr Pia’s consent.  This enabled some previously hidden information to be printed out.  From 23 April 2008 the training mode was not used either accidentally or deliberately.

23.         On the same day (23 April 2008), Mr Longman visited the Poltonhall shop and examined the till.  He took a print of the till settings but he did not change the settings of that till at that stage. 

24.         On 29 April 2008, Mr Longman visited the Bonnyrigg shop again, took further till readings and removed till rolls for further examination

25.         On 1 May 2008, Mr Longman telephoned the supplier of the tills supplied to the appellant seeking information about the settings of tills on delivery to customers.  However, the information provided by the supplier, ACR Retail Systems, was inconclusive.  A subsequent letter by ACR dated 3 June 2008 to the appellant confirmed that, while it was their policy to zero all reports on installation, this was not done in relation to the cash register purchased by the appellant and installed at the Bonnyrigg shop.

26.         On the same day (1 May 2008) Mr Longman visited the Poltonhall shop, spoke to Mr Andreucci and reset the till there.  The till readings he had examined showed various discrepancies for which Mr Andreucci offered no explanation.  Some of the rolls had been cut between days.  Mr Longman was unable to form a view as to whether the records were complete and accurate due to the till rolls having been cut.

27.         On 23 May 2008, Mr Longman wrote to Mr Pia raising various queries about the operation of the till and its records.  He pointed out that the evening sales over the period between 20 March 2008 and 23 April 2008 were very low when compared with a later period, and the earlier (pre-8pm) evening sales were high compared with a later period.  These views are based on figures derived from the appellant’s records.  The figures are set out in a Table appended to the letter dated 23 May 2008.  Although there was some confusion about the effect of the start of British Summer time, not reflected in the till settings, this does not materially affect the conclusion to be drawn from the analysis of the figures.  He also pointed out that the till at the Bonnyrigg shop had several features, which prevented the proper use of the till for auditing purposes.  He noted that the training mode of the till had been used.  He concluded with the observation that the sales records seen for the period between the purchase of the new till and 29 April 2008 understated the true sales.

28.         Mr Longman also wrote to Mr Andreucci on 23 May 2008.  He noted a discrepancy between cash and till records identified when a spot check to reconcile cash count in the register with till readings was carried out on 18 April 2008.  He pointed out that there were a number of features of the till settings which prevented the proper use of the till for auditing purposes.  He observed that the sample of journal till rolls provided to Mr Longman by the appellant’s accountant had been cut with scissors, and information was missing.  He concluded that the till readings were an unreliable record of sales.

29.         Mr Longman visited the Bonnyrigg shop on 4 June 2008 and collected various records generated by the till there.

30.         On 6 June 2008 (a Friday) a lunchtime invigilation was carried out at the Poltonhall shop.  The takings recorded amounted to £ 63.46.  The same exercise was carried out at the Bonnyrigg shop on 11 June 2008 (a Wednesday).  The takings recorded amounted to £147.25.

31.         In July 2008, the appellant appointed HW.  By letter to Mr Longman dated 8 July 2008, HW asked for a summary of where the investigation was at present.

32.         On 7 October 2008, four test purchases were made by HMRC officers at the Bonnyrigg shop. One of those purchases was not rung up on the till and was not recorded in the appellant’s records. 

33.         On 10 October 2008 (a Friday), a lunchtime invigilation was carried out at the Bonnyrigg shop.  The takings recorded amounted to £127.60.  Two test purchases on the same day (one at each shop) were both recorded.

34.         On 15 October, 2008 (a Wednesday), a lunchtime invigilation was carried out at the Poltonhall shop.  The takings recorded amounted to £51.71.  For some unknown reason the date on the till record is 14 October 2008.

35.         By letter to Mr Longman dated 24 October 2008, HW noted that they had agreed with Banks Richardson Scott & Co that cash counts would be taken on 31 August and 30 September, and that reports thereon were expected within two weeks.  HW indicated that once Banks’ report was available, they would be in a position to respond in detail to the various points raised by HMRC in correspondence. 

36.          On 21 November 2008 (a Friday), HMRC carried out discreet observations at the Bonnyrigg shop between about 4pm and about 1am the following morning.  These observations consisted of noting the time of arrival of each customer or potential customer, the time of leaving the shop and noting any observable purchases such as package or the like.  These observations were recorded in manuscript on pro-forma sheets which were lodged as productions.  The entries (some 428) recorded by HMRC exceeded the number of till sales transactions (381) over the same period.  While part of the difference can be accounted for by, for example, observed customers being together and making one combined till recorded purchase, the observations indicate that not all the transactions were recorded by the appellant.

37.         On 26 November 2008 (a Wednesday), HMRC carried out similar observations at the Bonnyrigg shop between about 4pm and midnight.  There were about 209 observations and 193 till transactions, a difference of about 7%.

38.         On 29 November 2008 (a Saturday), HMRC carried out similar observations at the Poltonhall shop between about 4.10pm and 11.05pm.  The observations there were less accurate because some sales were home deliveries.  HMRC recognised that these were difficult to record rendering a comparison with till transactions less valuable.

39.         The results these observations did not form the basis of Mr Longman’s calculations which led to the figures in the assessments and the amended assessments.   However, he took them into account as indicating that the appellant’s records were not accurate.  He was entitled to do so.

40.         In the course of his investigations, Mr Longman examined till rolls from both shops.  He noted that several till rolls from the Poltonhall shop had been cut and several entries were missing.  Mr Andreucci cut these till rolls.  His explanation at the hearing was difficult to understand.  He said that this was done for quickness but he was unable to explain to the satisfaction of the Tribunal what he meant by this or why he did it.  Whether this of itself led to under-declaration of takings is unclear.

41.         By letter to Mr Longman dated 24 December 2008, HW intimated that the cash counts of 31 August 2008 and 30 September 2008 disclosed a discrepancy of £273 at the Bonnyrigg shop but no discrepancy at the Poltonhall shop.  HW asserted that the letter from ACR Retail Systems established that the till had not been zeroed prior to delivery, and that any conclusions to be drawn regarding totals shown in the till reports could not be relied upon as a result of the failure by ACR Retail Systems.  The letter also noted that till rolls had been cut for ease of retention and no additional information was in the “cut” parts, but the practice had ceased.  Mr Andreucci was unable to give a comprehensible explanation as to how cutting these rolls eased retention.

42.         On 4 February 2009 (a Wednesday), HMRC carried out similar observations at the Poltonhall shop between about 3.35pm and 11.55pm.  The shop opened at about 3.45pm and the first customer was noted at about 4.20pm.  These observations included some goods being taken away by car as part of a delivery service.  Overall, these observations matched the recorded till transactions.

43.         Mr Longman’s enquiries ceased for a period in 2009 while a direct tax enquiry into the appellant’s tax affairs for the year to 31 March 2005 was being carried out.  Those enquiries have since been closed and assessments have been raised.  In September 2009, Mr Longman wrote to HW inviting them to attend a meeting to discuss matters.

44.         A meeting took place on 20 October 2009, attended by inter alios the appellant’s partners, Mr Banks, Professor McDougall, Mr Longman, and Miss Craig (another HMRC officer).  At that meeting Mr Longman raised a wide range of issues and asked many detailed questions.  Various matters were discussed including unexplained variations in the pattern of trade, the presence of an alternative cash container, the excision of records from the till roll, excess cash in comparison with the till readings, unrecorded alcohol purchases, and low gross profit for businesses of the trade class.  He did not receive satisfactory replies to many of his queries.  Some till rolls were incomplete, having parts snipped off below an X reading.  No comprehensible explanation is recorded in the notes of the meeting for this.  Neither Mr Pia nor Mr Andreucci gave an adequate explanation in evidence. 

45.         On 23 October 2009 Mr Longman wrote to Professor McDougall enclosing a copy of his typed notes of the meeting for comment.  In his letter, Mr Longman raised a number of questions including questions about the numbering of till operations, missing transaction numbers and the excision of sections of till roll.  Professor McDougall responded with a range of queries of his own.

46.           HW responded on 13 November 2009 providing some explanations and raising some queries about the terms of the notes.  Further correspondence ensued and various explanations were sought by HMRC and given by or on behalf of the appellant.  Mr Longman’s investigations led him to conclude that the appellant was not recording all its takings.  In the circumstances, that was a reasonable conclusion to reach.

47.         HMRC, not being satisfied with the various explanations, issued an assessment on 19 February 2010 in the sum of £59,425 covering the period from 06/05 to 03/08.  HMRC explained the basis of the assessment in a letter to the appellant dated 22 February 2010. 

48.         Mr Longman compared sales after 8pm taken from Z readings (daily total takings) from the till at the Bonnyrigg shop during the period between 20 March 2008 and 22 April 2008, with sales after 8pm during the period 26 February 2008 and 2 March 2008 (when the old till was still in operation), and the period between 23 April 2008 and 21 May 2008 (after the new till was re-programmed by Mr Long) and the period between 22 May 2008 and 10 June 2008.  The new till was not operational between the first period (26 February to 2 March).  The till was re-set on 23 April 2008.  HMRC did not make contact with the appellant in relation to their 2008 investigations until about 18 April 2008, which is almost at the end of the period between 20 March and 22 April 2008. 

49.         This comparison led to the conclusion that sales at the Bonnyrigg shop were under declared by at least 35%.  The detail of the calculations is set out on pages 1 and 2 of his letter dated 22 February 2010.  The daily average sale values from 8pm onwards for the said periods from till roll analysis are as follows:-

          26/2/08-2/3/08                                  £293

          23/4/08-21/5/08                                £475

          22/5/08-10/6/08                                £644

          Total                                               £1,413

 

The average is £471.

 

50.         The average evening takings after 8pm over the period between 20/3/08 and 22/4/08 is £122.  The difference between the average figure of £471 and the average figure of £122 is £349.  The average overall daily declared sales through till reports for the period between 20/3/08 and 22/4/08 was £989.  £349 is about 35% of £989.  On that basis Mr Longman concluded that, in his best judgment, the declared sales should be increased by 35%.  In this analysis, Mr Longman is identifying the post-8pm time period as the focus for under-declaration, and he has used 35% as an indicator to quantify the extent of under-declaration for the Bonnyrigg shop.

51.         For the Poltonhall shop he concluded that there had been suppression of about 20% of sales.  This was based on a cash reconciliation carried out on 18 April 2008 in the presence of Mr Andreucci.  The details are set forth in the same letter dated 22 February 2010.  Mr Longman reaches that percentage by comparing the difference between the cash counted in the till of £1080, and the sales recorded by the X-reading of the till register, namely £885.87.  He rounds that difference to £200, which is slightly more than 20% of £885.87.  However, the calculation is incorrect as he failed to take into account the sum of £40 mentioned in paragraph 17 above.

52.         Mr Longman discounted a similar cash count exercise carried out at the Bonnyrigg shop on the same day as being unreliable as it was unclear what the total cash amount actually was.  This was because some cash had been removed from the till before the reconciliation had been carried out.  This is noted in Mr Longman’s letter to Mr Pia dated 23 May 2008.

53.         Using 35% for Bonnyrigg and 20% for Polton Hall to quantify the extent of under-declaration, Mr Longman calculated the weighted mean between the two shops to apply to the turnover for each year. He noted the proportion of the total turnover attributable to each shop for the years to March 2008 and March 2009.  For these years the proportion attributable to Poltonhall was just over 40% and about 33% respectively.  For Bonnyrigg it was just under 60% and about 67%.  He calculated the weighted mean for each of the two years as being 30% and 29%.  He therefore used the lower figure of 29% and applied it across-the-board, but made an assumption that 3% of turnover was zero-rated.  No one has disputed the soundness of that assumption.  Accordingly, the across-the-board calculation applied to 97% of the declared gross takings.  This, he applied in February 2010 to sales between 1 April 2005 and 31 March 2008. To reduce the risk of assessments being time-barred, an assessment was raised on 19 February 2010 for £59,425 VAT for the 3 years to 31 March 2008.  These percentages and the method by which he came to reach the figure of 29% were not disputed as arithmetical calculations.  We refer also to paragraph 61 below.

54.         Mr Longman amplified his views in a letter to HW dated 1 April 2010.  He pointed out that the discreet observations carried out on 21 and 26 November 2008 and subsequent comparison with the till rolls indicated that not all customers were recorded on the till roll both before and after 8pm.  Further correspondence ensued and in the meantime, the appellant had on 19 March 2010 requested an internal review of the assessment issued on 19 February 2010.

55.         On 4 June 2010, HMRC issued a further assessment in the sum of £52,868, covering the periods (i) from 1 April 2002 to 31 March 2005 for £31,258 and (ii) from 1 April 2008 to 31 March 2009 for £21,610. (The earlier period for £31,258 is subsequently removed, leaving the assessment standing at £21,610.)

56.         Mr Longman carried out an analysis of lunchtime takings at both shops comparing average takings with takings when HMRC were carrying out an invigilation.  For Bonnyrigg, two periods were chosen, the first being between 23 April 2008 and 27 May 2008.  The figure for average weekday lunchtime takings was £69.09; the equivalent average for Wednesdays was £53.05; and the average for Fridays was £86.11.  The invigilated figure for Wednesday 11 June 2008 was £147.25, but the date was outwith the averaging period chosen.  The range for Wednesdays was between £41 and £65.

57.         The second period for Bonnyrigg was between 25 August 2008 and 26 September 2008.  The figure for average weekday lunchtime takings was £88.55; the equivalent average for Wednesdays was £112.16 and the average for Fridays was £106.75.  The invigilated figure for Friday 10 October 2008 was £127.60, but the date was outwith the averaging period chosen.  The range for Wednesdays was between £63.90 and £113.20, although there appears to be an entirely exceptional sum of £230.12 on Wednesday 3 September 2008.

58.         For Poltonhall, two periods were also chosen, the first being between 18 April 2008 and 27 June 2008.  The average weekday lunchtime takings was £31.09; the equivalent average for Wednesdays was £27.73 and the average for Fridays was £35.82.  The invigilated figure for Friday 6 June 2008 (within the averaging period) was £63.46.  The range for Fridays was between £43.46 and £63.46.

59.         The second period for Poltonhall was between 2 September 2008 and 17 October 2008.  The average weekday takings was £37.68; the average for Wednesdays was £32.38 and the average for Fridays was £43.48.  The invigilated figure for Wednesday 15 October 2008 (within the averaging period) was £51.71.  The range for Wednesdays was between £23.62 and £51.71. We note the significance that for Polton Hall, where the dates of invigilation fell within the chosen periods for averaging, the takings on those invigilated dates in both instances denote the upper end of the range for those periods.

60.         The issue of the assessment in June 2010 was again followed by an exchange of views in correspondence. HMRC declined to vary or withdraw the assessments.  Eventually, a review was requested by the appellant.  The data on which the exercise was based, had already been intimated to HW in April 2010.  In a letter to Professor MacDougall dated 9 August 2010, Mr Longman observes that the differences between invigilated and uninvigilated sales could be due to natural variation, and I have not attempted any statistical analysis, but they are consistent with the view that sales were under recorded in the absence of HMRC staff.

61.         On review, the Reviewing Officer, DL Russon cancelled some of the assessments but confirmed the remainder.  In respect of the first assessment (dated 19/2/10), the quarters from 1/4/05 to 31/3/06 were removed (see paragraph 53); and regarding the second assessment (dated 4/6/10), the 3 years to 31/3/05 were removed (see paragraph 55).  This left intact assessments covering the 3 years from 1/4/06 to 31/3/09.  In his letter to the appellant dated 22 December 2011, Mr Russon discussed the appellant’s arguments under the following headings namely, (i) unauthorised use of till training mode, (ii) pattern of trade inconsistent with opening hours, (iii) a Cash count on 18 April 2008, (iv) storage of excised till rolls, (v) low gross profit percentage, and (vi) not all transactions recorded.

62.         In his review letter dated 22 December 2011, Mr Russon noted that the assessments were based on the following circumstances or factors, (i) voluntary additions to the daily gross takings before finalising the business accounts, (ii) the cash exercise at Poltonhall, (iii) possible use of the till training mode at the Bonnyrigg shop, (iv) low average sales after 8pm between 20/3/08 and 3/4/08, (v) cut till rolls and omitted transactions at Poltonhall shop, (vi) lunchtime invigilations at both shops disclosing higher sales than preceding uninvigilated weekdays, (vii) unrecorded test purchase at the Bonnyrigg shop, and (viii) less sales recorded than expected on two occasions at Bonnyrigg (21/11/11 and 26/11/11) when premises were under observation.  He rejected all the appellant’s arguments, concluding that Mr Longman was entitled to raise the assessments on the basis of the basket of evidence before him, and that they were made to best judgment.

Submissions

63.         Mr Simpson submitted that the assessments were not made to best judgment and there was no tax payable.  Mr Longman misunderstood the evidence about the cash count at Poltonhall on 18 April 2008.  That misunderstanding formed part of his overall calculation and the result was therefore wholly unreasonable.

64.         Mr Simpson submitted that the Poltonhall shop should be excluded from the calculations.  The lunchtime invigilations there and other evidence did not support suppression.  There were invigilations there on 6 June 2008 and 10 October 2008.  The results and the comparison with un-invigilated declared takings did not justify a conclusion that overall 20% of the sales were being suppressed.  He pointed out that Longman accepted that the observations at Poltonhall on 29 November 2008 and 4 February 2009 were inconclusive.  This is another reason for excluding Poltonhall altogether.  Moreover, he argued that in relation to the criticisms of the operation of the Poltonhall till, the evidence was that Mr Andreucci did not have the skills to manipulate its operation so as to conceal transactions.

65.         Mr Simpson submitted that there was nothing in the GPR figures to justify any conclusion about suppression of takings.  For the years in dispute, Mr Banks had no concern about the level of the business’ gross profit percentage.  Low gross profit is indicative of generous portions and a more expensive cut of fish.

66.         Mr Simpson criticised the comparison between the till log for 21 November 2008 at Bonnyrigg showing 381 logged sales and the record of HMRC’s observations on that day showing 428 entries.  There were inherent uncertainties in these observations; some customers left not apparently carrying any goods, some individuals left together which might indicate one transaction rather than two.  The difference showed that the observations disclosed about 47 more transactions than the till records; this is a short fall of about 12.3% [(428-381)/381 x 100%].  He also criticised the observation logs for Bonnyrigg on 26 November 2008.  There are 223 observations and 193 till transactions.  At worst for the appellant, this shows suppression no greater than 15% [(223-193)/223 x 100%].

67.         With reference to the evidence of alternative cash containers, that was set out in the precognitions of both Mr Pia and Andreucci and they were not challenged in cross.

68.         Mr Simpson submitted that if the Tribunal concluded that the assessments were made to best judgment, they should be regarded as excessive and reduced on the basis of one or more of the following arguments, (i) the Poltonhall element of the assessments should be removed leaving the turnover at Bonnyrigg to be increased by 35%, (ii) reduce that figure of 35% to 4.3% which is consistent with the evidence and reflects the fact that it is extremely unlikely that the appellant committed deliberate fraud; poor book-keeping is the more likely cause, (iii) a reduction in the percentage would be consistent with the differences between the HMRC observation logs and the till logs at Bonnyrigg (12% on 28 November 2008 and 15% on 21 November 2008; and it would be consistent with the small differences from the average gross profit in 2008; the appellant’s GPR in 2009 being above the average; in addition, the lunchtime invigilations do not vouch a  35% increase.

69.         Mrs McIntyre submitted that the assessments were based on a basket of concerns including (i) the additions to the accounts, (ii) missing transactions, (iii) cash count discrepancy at Poltonhall, (iv) anomalies with the till rolls, (v) differences between invigilated and un-invigilated sales, (vi) discrepancies between recorded sales and HMRC observations, (vii) an unrecorded test purchase, (viii) the programming of the new till at Bonnyrigg and the use or possible use of the till in training mode.  This was an unusual accumulation of concerns.  She submitted under reference to various documents that the assessments were made to best judgment and in any event were correct.  She accepted that if we took the view that the assessments were made to best judgment but were nevertheless incorrect the suggestion that Poltonhall be excluded from the assessment and either the same or a revised percentage be applied to the Bonnyrigg takings, was a pragmatic approach which might commend itself to the Tribunal.

Discussion

70.         One of the most striking aspects of this appeal is the evidence that for the appellant’s business years 2003, 2004 and 2005, the appellant voluntarily added a total of £149,800 to their recorded takings, while maintaining the view that their records were accurate for those years.  It is extremely difficult to accept that any businessman who honestly and diligently kept proper and accurate records would do so, whether or upon the advice of an accountant, whatever apprehensions he might have about a tax enquiry because his GPR was substantially lower than the average for his type of business.  These circumstances entitled HMRC to be suspicious of the record keeping of the appellant.

71.         The evidence of Mr Pia and Mr Andreucci was generally vague and unclear and some explanations were difficult to follow.  Our overall impression of them was that accurate record-keeping was never at the forefront of their minds.  In spite of the lengthy and detailed evidence, no clear picture emerged that any system was in place that ensured that all sales were recorded.  On the occasions HMRC were present either invigilating or checking the cashing up at the end of the day, takings were generally substantially greater than the norm, as Mr Longman pointed out in his letter dated 9 August 2010.  We do not think this is a coincidence.  Nor do we think that the presence on site of HMRC magically brings good fortune to traders, by attracting increased custom.

72.         It is not even clear to us who at each shop was and took responsibility for the accuracy of the cash collected each night compared with the recorded takings, how the float was taken into account, or where it came from.  Even the transfer of takings from the till to the safe on the premises and the removal of the cash from the safe back to the till for float, or to the bank seemed a little haphazard.  We of course make due allowance for the fact that the role of the two male partners was primarily attending to the preparation of food, and the cooking.  No doubt they were skilled and efficient at these tasks and worked hard.  Nevertheless, they were the only witnesses who gave evidence of the operations of the business.  We note that the witness statements of Mr Pia and Mr Andreucci, were identical in places and were prepared with the assistance of Professor McDougall.  We make no criticism of that but there is a stark contrast between the elegance of the written statements and the oral evidence.  All this causes us to view these written statements with some suspicion.

73.         We cannot accept the assertions made in evidence by Mr Pia and Mr Andreucci that their records have all along been accurate.  Having regard to the variety of anomalies and discrepancies identified by HMRC in the course of its investigations in 2008 and 2009 in particular, coupled with the evidence of very substantial voluntary disclosure for earlier years, their conclusion that there have been under-declarations of takings, seems reasonable.  Although Mr Banks said that the records of the appellant as presented to him from time to time were satisfactory and may even have improved over the years, that tells us very little about the critical question whether the appellant was properly recording all its sales.  The evidence we heard and accepted was largely all one way.  That evidence, reflected in our findings of fact, does not support the conclusion that sales were accurately recorded in either shop.

74.         We conclude, in the exercise of our quasi-supervisory jurisdiction that the assessments have been made to the best of Mr Longman’s judgment (VATA 1994 s73(1)).  In addressing the question of best judgment, neither party drew any distinction between the various assessments. The approach was all or nothing.  The revised assessments, beneficially for the appellant, remove certain earlier periods because of a technical breach of HMRC practice in relation to cases involving possible fraud. 

75.         The basis on which the other assessments proceeded was left untouched.  HMRC have acted honestly throughout.  No one suggests otherwise or that Mr Longman acted in bad faith or perversely, although it is argued for the appellant that Mr Longman’s use of the Poltonhall cash reconciliation was wholly unreasonable.  In our view, however, such a margin of error of calculation does not vitiate the assessments.  It may lead (as it does here) to the conclusion that an assessment does not in the light of all the material placed before the Tribunal, reflect the correct amount of tax payable.

76.         HMRC have acted fairly.  The detail of Mr Longman’s calculations set out in his letter dated 22 February 2010, demonstrate that, particularly, his averaging process and his allowance for zero-rated supplies.  That seems to be the best he could have done on his understanding of the limited information available to determine whether there has been suppression of sales and if so, by what amount.  Mr Longman’s figures are based on calculations using the imperfect material available.  His figures are plainly not arbitrary or capricious, or a mere guess.  This is therefore not one of these rare cases where an assessment is discharged because it has not been made to best judgment.  We therefore proceed, in the exercise of our full appellate jurisdiction to determine the correct amount of tax payable in the light of (a) the appellant’s contention that the assessments are excessive, and (b) the material before the Tribunal (see Mithras (Wine Bars) Ltd v HMRC [2010] STC 1370, [2010] UKUT 115 (TCC) paragraphs 8, 11, 16, 18, 21 and 23).

77.         In relation to the cash count carried out in Polton Hall on 18 April 2008, we have found that the till X reading was £885.67.  The cash counted amounted to £1080.  That sum excluded a float of £100.  However, there was also about £40 in the till at the start of the day.  The true comparison is therefore between £885.67 and £1040 (£1080 less £40).  The difference is £155, which is about 17.5% of £885.  Mr Longman’s incorrect understanding produced a figure of 20%.  He used that percentage in his overall assessment.  That was an error.  However, it does not mean that the assessments were not made to best judgment.  It means that the assessments require correction.  They are excessive.  It was also suggested by the appellant that the till float could have been as much as £50, and if we were to discount £50 instead of £40, the percentage would reduce to 16.3%.

78.         We have carefully considered the evidence. We heard lengthy evidence about the till and other records produced.  We found much of this evidence to be confusing and in places equivocal and incomplete.  We find it impossible to be sufficiently confident to make detailed findings of fact.  Overall, the evidence shows that the records produced do not disclose the full extent of the appellant’s takings. In particular, the discreet observations at the Bonnyrigg shop justify Mr Longman’s conclusion (which on the evidence before us, we endorse) that the extent of trade was not fully and completely disclosed in the declared sales.

79.         We agree with Mr Longman’s observation in his letter dated 9 August 2010.  His comparison exercise supports the view that there has been suppression of takings and demonstrates the reasonable and balanced approach taken by him in the assessment of the extent of that suppression.

80.         We cannot hold that it has been established that the VAT returns for the years of assessment are correct and complete.  The appellant has not discharged the onus to do so. We cannot identify with any precision what the correct amount of tax should be.  We are persuaded, but only just, that Mr Longman’s assessment is probably excessive having regard to the evidence and submissions presented to us. While overall, the evidence did not satisfy us that the records of sales made at either shop were accurate and complete over the periods of the amended assessments, the evidence that the records were inaccurate was stronger and more extensive in relation to the Bonnyrigg shop than the Poltonhall shop.  In the absence of a better alternative, and the fact that towards the conclusion of the hearing, the parties seemed to be moving towards the consensual position that Polton Hall could be taken out of the equation, we consider this as a feasible means to establish the ‘correct’ amount of tax insofar as it is possible  to do so.  It is not suggested that the result of the re-calculation means that there had been no under-declaration at Polton Hall, but for reasons set out below, we consider this re-calculation a reasonable basis to address the issues regarding the quantum of the assessments.

81.         We were not convinced that the evidence about alternative cash containers demonstrated that the cash collected and placed therein rather than directly into the till was always recorded.  There was no satisfactory evidence of the sums so collected.

82.         We heard some evidence about larger or more generous portions and better quality fish being the reason for a lower than average GPR.  We have made no findings of fact about this topic because the evidence was vague.  It was not clear what periods were being referred to and it is well-nigh impossible to assess what the difference is between a generous portion and a normal portion and how it might affect GPR.  The same comment applies to the quality of fish.  The price of fish is notoriously variable.

83.         While it is correct that the accounts for the years 2007/2008 and 2008/2009 show that the Bonnyrigg GPR was only just below various regional and national averages, that does not really assist the appellant in demonstrating that they have accurately recorded and accounted for their takings over the relevant periods.  The background of a massive injection of funds from some unknown source to make accounts for earlier years respectable cannot be entirely ignored.  Moreover, the evidence of invigilations, discreet observations, and the unusual operation of the till and the preservation of till records, taken together do not support the argument that the assessments are excessive and do not identify what the correct amount of tax should be, or the correct figure from which it should be calculated.  The evidence taken as a whole does not justify the conclusion that the declared takings should be increased by 4.3% or 15% as Mr Simpson suggested in his careful and thorough presentation.

84.         The approach of the appellant throughout the investigations has been, to a material extent, to raise its own queries in response to HMRC investigations rather than to offer positive information to demonstrate the accuracy of its records.  While that may be a legitimate tactic, it may not assist in demonstrating that an assessment, made to best judgment, does not identify the correct amount of tax.  That means that the basis upon which Mr Longman raised the assessments should stand except that that basis of calculation should only apply to the VAT-able goods sold from the Bonnyrigg shop.

85.         The result is that the basis upon which Mr Longman raised the assessments (subsequently amended in relation to periods they originally covered) should stand in relation to the declared takings from the Bonnyrigg shop.  Mr Longman applied his calculations to 97% of the declared takings to take account of zero-rated sales.  We agree that this should be done for the purposes of the revised calculation based on the declared takings from the Bonnyrigg shop.  Mr Longman applied the figure of 35% to the recorded takings and this should be done in relation to the Bonnyrigg shop.  It has not been proved on a balance of probabilities that this is erroneous.  Account should be taken of any change in the rate of VAT over the amended periods of assessment.

86.         We cannot with confidence adopt any other basis of calculation and no acceptable alternative has been suggested other than that the returns are accurate.  We do not accept that the returns are accurate.  As the appellant has not demonstrated, the onus being on the appellant, the accuracy of its relevant returns and has advanced no other figures beyond applying a lower percentage (which we do not accept) than the 35% identified by Mr Longman, we cannot make any other global percentage deduction, even if we were able to identify one.  That would simply be arbitrary.  HMRC are not entitled to act in an arbitrary manner when raising assessments to the best of their judgment and have not done so. Nor are we, when endeavouring to identify the correct amount of tax payable.  If our approach to the assessment of the correct amount of tax is unsound, then the only alternative is to conclude that the assessments should stand unreduced, because the appellant has failed to establish with acceptable evidence what the true amount of the supplies were during the periods of the amended assessments; and has thus failed to establish, the onus being on the appellant, what the correct amount of tax should be.  Independently of the question of onus, we have been unable to make, on the material before us, any other calculation of what the correct amount of tax payable should be.

 

Disposal

 

87.         At the conclusion of the hearing it was agreed that it would be impossible for the Tribunal to calculate precisely the correct amount of tax payable and that we should issue a decision in principle, leaving it to the parties to agree the amount of tax payable in the light of our decision.  This, we now do.

 

88.         In the foregoing circumstances, the appeal is allowed to the limited extent described above.  The amended assessments were made to best judgment.  They are nevertheless excessive and fall to be re-calculated in the manner which we have explained. 

89.         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.

Postscript

90.         The hearing in this case lasted six days.  The amount at stake was about £63,000.  The appellant produced brief witness statements late, resulting in a re-scheduling of witness evidence.  Mr Longman’s statement did not address much of the detailed documentary background or the technical evidence about the operation of tills and the analysis of till records.  The appellant’s skeleton argument did not address quantum.  There was no statement of agreed facts.  Much of the evidence was difficult to follow.

91.         This is a very inefficient way to resolve a tax dispute.  It is costly for the appellant and costly for the general taxpayer.  In future, at the Edinburgh Tribunal at least, this type of case will be carefully case managed to ensure that the substantive hearing contains only relevant and useful evidence, and that so far as possible, is concluded within a significantly shorter period than six days.

 

 

 

 

J GORDON REID

TRIBUNAL JUDGE

 

RELEASE DATE: 24 February 2014