First-tier Tribunal (Tax)

[2009] UKFTT 216 (TC)

TC00166

Appeal number MAN/00/0136

VAT – assessment – penalty (S 60 VAT Act) – appeals dismissed.

FIRST-TIER TRIBUNAL

TAX

1. PIZZA PLACE

2. REZA TAHERI-KHAMN

3. EBRAHIM KHAKNEJAD-KERMANI

Appellants

- and -

THE COMMISSIONERS FOR HER MAJESTY'S

REVENUE AND CUSTOMS (VAT) Respondents

TRIBUNAL: Judge Richard Barlow

Member Alban Holden

Sitting in public in Manchester on 8 and 9 November 2008 and 2, 3, 4, 5, and 6 March 2009 (Parties closing submissions in writing completed 6 May 2009).

Mr Tahir Nawaz chartered accountant of T Nawaz & Co for the Appellant

Mr Andrew Noble of counsel instructed by the General Counsel and Solicitor to HM Revenue and Customs for the Respondents

© CROWN COPYRIGHT 2009
 


DECISION

1. The tribunal was constituted as mentioned above on the hearings in March 2009 but on the two days in November 2008 Miss Christine Owen also sat as a member. When the matter resumed in March 2009 Miss Owen was unable to sit because of illness. With the consent of both parties the hearing proceeded before the chairman and Mr Holden only. The tribunal had reached no conclusions about the case in November 2008 nor had it discussed the evidence that had been heard then and Miss Owen took no further part after that date.
2. Mr Reza Taheri-Khamn was referred to as Mr Taheri at the hearing and we will adopt that practice in this decision. Mr Ebrahim Khaknejad-Kermani was referred to as Mr Kermani and we will refer to him in that fashion. Pizza Place is the trading name of a pizza takeaway business operated in partnership by Mr Taheri and Mr Kermani at Fartown Huddersfield and was the name under which the partnership was registered for VAT with effect from 3 December 1993 until the partners sold it in June 1999.
3. The appeals are against assessments for the periods from 4 December 1993 to 31 May 1999 in a total sum of £71,831 made on 7 September 1999 in the name of the partnership and a penalty under section 60(1) of the VAT Act 1994 issued on 31 January 2000 addressed to the partners in the sum of £61,056 being 85% of the tax assessed (with some rounding) which the respondents allege was evaded through the dishonest acts of the partners.
4. We heard evidence from customs officers Mrs Angela McCalmon, Mr David Pyrah, Mr Dean Foster, and Mr Matthew Priestman and from Mr Martin Scoffield, concerning the computerised till for which software was supplied by a company for which he had worked. With Mr Nawaz's agreement, we read the statements of other customs officers whose evidence was not disputed and had before us a number of documents put in evidence by the respondents. We heard evidence from Mr Kermani and Mr Taheri and had before us documents put in evidence by them. Both parties made written submissions after the conclusion of the evidence and we have taken into account those submissions when reaching our decisions.
5. We make the following findings of fact.
6. The respondents' enquiries began in February 1999 with two test purchases which were used as a reason to visit the appellants' premises for reconnaissance but which did not form any part of the reasons for or quantification of the assessment.
7. On Friday 26 February 1999 the respondents' officers made an unannounced visit to the Pizza Place near the end of the day's trade and collected diaries for 1998 and 1999 in which the appellants had recorded their takings showing, for each day of trading, a "taking" figure and a "ZR" figure. It was the appellants' case that the "taking" figure was the correct amount of total sales and the "ZR" figure was the correct amount of zero-rated sales. The diaries were presented to the appellants' accountant who completed their VAT returns by deducting the zero rated figure from the total and calculating the output VAT included within the remaining figure, as well as dealing with input tax.
8. In 1998 the weekly total of sales was also recorded in the diary. The Z readings taken weekly in 1998 were in some cases attached to the diary and the totals agree with the weekly total recorded in the diary which itself agrees with the addition of the relevant individual days.
9. The appellants said that, at some point in time, they began to take daily Z readings. They claimed to have done this from about 20 February 1999, shortly before the customs officers' unannounced visit, but we note that the recording of weekly figures in writing in the diary ended on 27 November 1998, though there is a Z reading attached to the diary for the following week for a sum that appears to be for a weekly amount. We certainly cannot understand why the appellants would go to the trouble of counting the cash every night, as they claim to have done, but only taking a Z reading weekly to see if it agreed with the cash as counted. Any discrepancy would then be more difficult to reconcile than if the readings were taken at the same time as the cashing up. It is the appellants' case that the Z reading did always agree with the cash as counted and manually entered in the diary whether the reading was taken weekly or daily.
10. On the unannounced visit the appellants took a Z reading from the till under the supervision of the customs officers at the close of business shortly after midnight and the amount was £612.50, which agreed with the takings in the till, being the takings for most of Friday 26 February though the Z reading, having been taken after midnight, recorded the date as 27 February. Subsequently the appellants produced a second Z reading for the earlier part of that day taken at 19.05 hours for an additional £52.20 which Mr Kermani said he had taken because he had to leave the premises. Given that the cash for the later part of the day agreed with the Z reading he must also have taken the cash with him. His explanation for taking the extra Z reading was that he decided to leave the premises to take his daughter to casualty following a fall. His explanation for taking the cash was that he wanted to buy vegetables for the business. Those two statements appear to us to be in conflict especially as at 7pm at night with a trip to casualty as, presumably, a first priority it seem improbable that he would be able to buy vegetables that day anyway.
11. The appellants were interviewed on 14 April 1999 and this additional Z reading and a figure of £52 was mentioned then. It was later that the Z reading itself was produced. The takings of £664.70 which the appellants therefore said was the figure for Friday 26 February was a higher amount than the declared figures for Fridays for at least the preceding 18 weeks, mostly by quite a wide margin. The appellants sought to explain that by reference to a fish and chip shop nearby having closed which boosted their trade.
12. On 2 and 4 March Mr Pyrah, who was an officer familiar with the details of the workings of computers, took information from the appellants' computerised till and recorded it electronically on floppy discs for later examination and analysis.
13. The information from the till included files known as audit.dat and log.dat. Audit.dat recorded every transaction in detail but when a Z reading was taken the information was deleted from the audit.dat file and a summary of it was transferred to log.dat. Mr Pyrah's evidence was that that is what appeared to have happened based on his analysis of the information and Mr Scoffield, whose evidence we deal with below, confirmed that was the way the programme worked. Mr Pyrah recovered the log.dat details for the period from 29 November 1998 to 27 February 1999. Some information, amounting to 5,000 bytes out of 722,000 had been lost, but he found no evidence of a computer virus. The log.dat information includes a total of the numbers of items sold in a number of categories such as BG, KB and PZ which he concluded were references to burgers, kebabs and pizzas which we regard as an entirely reasonable assumption especially as PZ is the most numerous item and the business was mainly a pizza shop and we did not understand the appellants to contest that conclusion. The value of the items sold is also recorded.
14. From that information Mrs McCalmon made a comparison of the values of goods recorded in the log.dat file as sold on each day between 29 November 1998 and 27 February 1999 and compared it to the daily declarations as recorded in the diaries. The log.dat takings, if that is what they were, showed a suppression of 52% of the true takings when compared with the declared takings. The assessment is simply made by the application of that suppression rate from the start of the business. Mrs McCalmon justified that starting date on the basis that the declared takings of the business had been reasonably steady throughout that period and there seemed to be no reason to assume that the suppression had started at any particular point or that the rate of suppression had either risen or fallen markedly.
15. It is apparent that the correctness of the assessment is fundamentally dependent upon the log.dat data being truly a record of what was entered into the till.
16. In 1999 Mr Scoffield was the technical services manager of the company which supplied the programme of which audit.dat and log.dat are part. It was supplied to a third party who supplied it to the appellants in this case. Mr Scoffield holds a BSc in Maths and Physics and both of those courses involved computing. Through his employment he was familiar with the workings of the PizzaPos software programme of which log.dat and audit.dat are part and which the appellants had been using. The purpose of log.dat was mainly to help large firms with their stock control, which we accept was irrelevant to the appellants, but that does not mean that it is irrelevant for the purposes of this case because in order to do that it records sales details.
17. Mr Scoffield no longer works for those employers and has no connection with the third party that actually sold the programme to the appellants and he has no connection with the appellants. We accept that his evidence was entirely truthful and is that of an independent expert whose evidence was not undermined by cross examination.
18. In fact the appellants were sold the till programme, as they said and we accept this was so, because it enabled them to store the names and addresses of regular customers which helped them to avoid hoax orders and made their delivery system more efficient.
19. Mr Scoffield explained that each time a Z reading is taken that transfers the audit.dat data to the log.dat file in the summary form we have described and that from that point it no longer remains in the audit.dat file. Z readings are not taken automatically so that if a trader chooses to take one only weekly the data would be transferred weekly from the audit.dat file to the log.dat file.
20. When a Z reading is taken the machine usually prints out a copy as well as recording it electronically and it was certainly the appellants' case that that happened at the Pizza Place as some such printouts were attached to the 1998 diary as we have already said. Why they did not keep all their Z readings was not explained.
21. Mr Scoffield was shown a weekly Z reading printout for the week ending 24 October 1998 which is stapled to the dairy for that week and which is the same amount as the total takings recorded in the diary (£1924.15). He was shown a similar Z reading report timed at 00.21 on 26 February 1999 which was found at the premises on the day of the unannounced visit and which is the takings for the single day's trading on the day before the visit and which agrees with the diary entry for that day, the 25 February. It is £447.10. In respect of both of those Z readings Mr Scoffield said that it was impossible for them to have been taken from the same till as that from which the log.dat data produced by the respondents was taken because they do not agree with the log.dat figures for that week or that day respectively. Once a Z reading is taken the entries whose total it represents cannot then be included again in any other Z reading. The same applies to any other Z readings that do not agree with the log.dat. They cannot have been taken from the same till. The evidence of Mr Pyrah was that the log.dat data he produced and which Mr Scoffield was examining was that taken from the appellants' till or rather the computer function associated with it.
22. We therefore find as a fact that the Z readings taken from the appellants records do not agree with those in the log.dat taken from the appellants' till and that, if the till was working correctly, that means that the takings declared by the appellants were less than the true takings i.e. those recorded in the till.
23. The main issue between the parties is whether the till was working correctly. The appellants state that they have declared the correct amount of tax and that the sums recorded in their diaries are the true takings from which their accountant completed their VAT returns. Those diary figures agree with the Z readings they took and so they contend that the till memory, the log.dat in particular, was malfunctioning though somehow the Z readings were accurate. They accepted that the other functions of the till such as recording the customers' addresses functioned correctly.
24. The appellants' evidence was that the till had given a lot of trouble. They produced a letter from the supplier, Steven Oliver, who said that since the supply of the Pizza.Pos system the following work had been done: four replacement keyboards, a replacement printer, a replacement power supply and removal of viruses on two occasions. Mr Oliver also said he had trained the staff a number of times. We note that he is unspecific about dates and it is also the appellant's case that they had to replace the computer following a break-in on Christmas day 1997, so only events after that date would be relevant to the information we have for the purposes of this case. Mr Nawaz in his final submissions told us that the appellants had probably bought the till in 1995 so some of the repairs may well be irrelevant.
25. It is relevant to consider the significance of the fact that the data in the computer memory contains what appear to be references to other traders. The appellants bought a programme for use exclusively in their business and should have been supplied with a clean copy of the programme which was set up for their use. The references to the other traders suggests that rather than receiving that clean copy they were sold a copy of a programme that had been pirated from similar programmes already sold to other traders and which had traces of those traders left in it. However, there is no evidence that the data input by the appellants while they operated the programme was in any way affected by that fact.
26. Mr Scoffield said that hardware problems like a new keyboard and printer would not have caused the log.dat to malfunction and that if a virus had affected it, it would have been very likely to have affected all the computerised functions. The appellants agreed that the addresses of customers were unaffected and we note that it is the appellants' case that the printout of the Z reading gave a correct figure repeatedly but that somehow the transfer of that information to the log.dat memory malfunctioned. It is difficult to see how that could be the case because on two occasions namely a Z reading taken on 22 February 1999 for £400.95 and the Z reading of £612.05 on the day of the unannounced visit were both reproduced on the log.dat file. The £400.95 was declared in the diary and the £621.05 agreed with the amount found in the till in the supervised cashing-up on the unannounced visit.
27. It is the appellants' case that the log.dat data is corrupted though not the Z readings taken from the same records and they offer no explanation for the two days when the log.dat data was not corrupted. We accept of course that if the computer was malfunctioning it could work correctly on occasions but a conclusion that it did is difficult to reconcile with the fact that other functions of the till and computer were working correctly and Mr Scoffield's evidence about the fact that a virus would be most likely to have affected all functions.
28. It is also difficult to see how the audit.dat data could be transferred to the log.dat file daily, as it was, during the period when only weekly Z readings were being taken according to the appellants. The taking of a Z reading was what triggered the transfer of data. There is no reason to think that data could be transferred without a Z reading even if the transfer corrupted the data at the point of transfer. If data was being transferred without Z readings being taken it might be expected that it would have been transferred at random times either several times a day or a few days apart but in the main it was being transferred consistently daily contrary to the appellants' assertions that they took the Z readings only weekly for part of the period in question.
29. The respondents' case is that the appellants must have had access to another till that could produce the false Z readings and that the log.dat data tells the true story taken from the till actually in use. There is no direct evidence of the existence of such a till but the respondents say that the circumstantial evidence is compelling. We note first that unlike some Z readings the ones which were produced simply state the total amount taken and there is no breakdown of numbers of items sold. It would not therefore be a laborious process to key in items into another till that would then produce a Z reading of any figure that was desired. A single large entry would achieve that result or rather two entries as the Z readings distinguish between deliveries and cash sales. The inconsistencies in the appellants' contentions between the correct operation of the till in some respects together with its supposed multiple errors in other respects are significant in our opinion.
30. Of course, the correct operation of the till on the two days mentioned above might be said to corroborate either party's case because, if the respondents' are right, why would the appellants have declared the correct figure when they declared £400.95 if their practice was to declare false figures on other occasions. That would not explain how it came about that the takings on the day of the unannounced visit were correctly transferred into the log.dat file.
31. Whilst we agree that a malfunctioning machine might occasionally function correctly we would have decided on a balance of probabilities that the log.dat data is an accurate record on the evidence and findings we have made above.
32. However, there is one more piece of evidence which we regard as so highly significant that it re-enforces our conclusion that the log.dat data is reliable to a degree that makes us certain of that fact to a degree well above that of making a finding on a balance of probabilities.
33. The diary for 1998 contains Z reading printouts for the weeks ending on Fridays 20 November, 27 November and 4 December. For each day within those three weeks the daily amounts of takings are written into the diary. For the weeks ending 20 November and 27 November the weekly takings on the Z readings agree exactly with the amounts written in the diary. For the week ending 4 December the diary appears to record a figure £15.20 less than the Z reading.
34. As we have already indicated, the Z readings all record the date and time when they were taken and they are taken shortly after midnight so the dates when they were taken should be the early hours of Saturday in respect of weeks ending Friday so the relevant dates actually printed on the Z readings should be 21 and 28 November and 5 December. In fact they are 21and 27 November and 5 December.
35. The Z reading for the middle week therefore was taken after the end of the Thursday's trade of that week but, as it agrees with the diary entries for the whole of that week including the Friday's trade, which could not have been legitimately known until 24 hours later that demonstrates that, for that week at least, the Z reading was demonstrably false and had been deliberately manipulated to agree with preconceived trading figures that were themselves false.
36. We would add that, as the following week's declared takings also agreed with the diary entries and the Z reading (or rather only differed by £15.20 being much less than a day's takings), there was no question of the misplaced day being included in a larger declaration for the next week.
37. Also the times and dates on the preceding week and succeeding week's Z readings confirm that there is no evidence that the time and date for the middle week of the three was incorrectly recorded by a machine error.
38. Effectively, the respondents' case is that this aspect of the evidence confirms that the appellants were operating a deliberate system of under-declaration and that they made an error in the operation of that system which has demonstrated beyond doubt that it existed.
39. We agree.
40. It follows that an assessment was justified and we hold that the method of assessment adopted by the respondents arrives at a very accurate figure in this case as it is based on a fairly long period in which we have found that the true figures (in the log.dat) are known. There being no reason to suggest that the period in question was not representative, which we hold to be the case in the absence of any obvious marked fluctuation in trade according to the declared figures and in the absence of any evidence that there was any such marked fluctuation, we uphold the assessment in full.
41. Both partners admitted that they took an equal role in the business and were actively involved in its day to day operation. It is inconceivable that some third party would have manipulated the takings records in the way that we have found that they were and even if one of the partners had been responsible for carrying out the manipulation the other must have known about it and acquiesced in it. We therefore find that both of the partners are were involved in dishonest conduct for the purpose of evading VAT and are thereby liable jointly and severally for the penalty under section 60 of the VAT Act 1994. No reason emerged to suggest that the mitigation should be increased.
42. We make no award of costs. The respondents failed to deal efficiently with the disclosure of information and documents some of which we had to direct during the hearing should be disclosed, being such as should have been disclosed much earlier. The appellants on the other hand have denied dishonesty and have been found to have been dishonest and have failed completely so far as the appeals are concerned so no award of costs would be appropriate for them either.
43. The appeals are dismissed. The assessments of tax and penalty are upheld in full.

RICHARD BARLOW

TRIBUNAL JUDGE

RELEASE DATE: 24 August 2009

MAN/00/0136