Cheon Fat Limited v The Commissioners for HMRC

We were provided with a bundle of documents of 505 pages together with some additional pages which had been erroneously excluded when the bundle was prepared and some additional materials from the Appellant. The evidence justifying the making of the Assessments was limited to the following:

(1)

VAT return and corporation tax return data from commencement of trading in 2012 through to APE 31/03/17 and VAT accounting period 12/16.

(2)

The note of an initial scoping visit made by HMRC to the Appellant’s premises on 24 April 2016.

(3)

Test eat reports and the associated officer notebooks for test eats undertaken on 4. 8, 12, 20 and 28 May 2016.

(4)

Typed (but undated and unsigned) notes made of an unannounced visit made by HMRC on 25 November 2016 (this note was not shared with the Appellant at the time of the visit).

(5)

Spreadsheets setting out the information extracted from the Appellant’s business records for 4, 8, 12, 20 and 28 May 2016 (Meal Ticket Analysis).

(6)

Test eat reports and the associated officer notebooks for test eats undertaken on 26 July 2017.

(7)

Handwritten notes for an unannounced visit on 26 July 2017 together with a) a receipt for records taken on that date and subsequently returned on 27 July 2017; b) Z readings for that day’s trading and c) cash up note signed on behalf of the Appellant.

(8)

Assessment calculation spreadsheets (Assessment Analysis) including data of credit card transactions said to have been extracted from evidence provided by Worldpay.

(9)

The note of a meeting on 5 March 2018.

(10)

Witness statements and oral testimony received from Officers Beard and Khan and Mr C Chan (director of the Appellant). Mr Chan gave his evidence through an interpreter.

We were not provided with any of the following evidence:

(1)

The source information provided by the merchant acquirer;

(2)

Either the originals or copies of the documents from which Meal Ticket Analysis were derived.

No satisfactory explanation was provided for the absence of the merchant acquirer data. Given the transposition errors by HMRC which were identified shortly before and during the hearing whilst we had no question as to the integrity of the information, how HMRC had transposed and used it could not be verified by us or by the Appellant (who was never provided with the source information either). This information was a critical component of the VAT Assessments and thereby the corporation tax assessments and we would have been somewhat uncomfortable in determining quantum without that source data. In the end, however, and as set out below, we considered that a more appropriate basis of assessment was to uplift only the figures of cash declared and thus the quantum of the assessments does not now depend on the credit card information with the consequence that its absence is immaterial.

A key issue between the parties concerned the whereabouts of the source documents for the Meal Ticket Analysis. The note of the unannounced visit on 25 November 2016 does not have an identified author, it is undated and unsigned. It was not prepared by either Officer Beard or Khan who both became involved in this matter after the investigation was completed and the Assessments raised. The officers involved in the original investigations have retired. However, significant parts of the note were put to Mr Chan in examination in chief and cross examination from which we conclude that it is a sufficiently accurate record of the events on that evening.

That note records (and Mr Chan confirmed) that during that visit he produced and gave to the attending officers business records (including meal tickets, purchase invoices and bank statement) for May 2016 and 25 November 2016 together with an A4 purple book in which he recorded his DGT. The note records that a receipt was given for the records uplifted but none was produced. Mr Chan (quite reasonably) could not now recollect whether he had signed or been provided with a receipt. It is not known whether HMRC took copies, but none are now available. The note also records that the attending officers agreed that the purple book would be returned on the following Monday (28 November 2016). There is no record in the note confirming that the documents were returned and certainly no signed receipt evidencing that they were so returned.

It is now claimed by Mr Chan that the records were never returned with the consequence that the Appellant could not verify or effectively challenge the Meal Ticket Analysis. We note that Mr Chan’s witness statement made no such claim. Nor was the asserted absence of the records noted in the meeting note of 5 March 2018 or in subsequent correspondence despite that meeting note recording the presentation of the Meal Ticket Analysis and referring to the DGT figures within the purple book.

When Mr Chan was giving evidence, Mr Sims asked Mr Chan how the VAT return for the period October – December 2016 had been prepared if HMRC had not returned the purple book. Mr Chan explained that as HMRC had only uplifted meal tickets for May 2016 and 25 November 2016 his accountant had been able to prepare the VAT return from the meal tickets.

We have carefully considered the evidence that is available to us and have concluded, on the balance of probabilities, that the records (consisting of meal tickets, purchase invoices and the purple book) were returned to the Appellant business. The meeting notes records that Mr Chan requested that the purple book be returned as soon as possible. We find it highly unlikely that the visiting officers would have recorded and then ignored the promise to return the records. We also consider it unlikely that had the records not been returned that it would not have been raised as an issue by Mr Chan immediately.

The return of the original records does not, however, explain the absence of copies of them. Whilst it is conceivable that HMRC produced the Meal Ticket Analysis from original sources without taking copies we consider it significantly more likely that copies were taken from the originals and that the spreadsheet was then produced from the copies. It, however, appears that those copies have been either lost or destroyed by HMRC. We make no finding that the copies were wilfully destroyed. We consider it significantly more likely that once the Meal Ticket Analysis was produced the copy records were considered of less importance and were destroyed or “lost” when the Northampton tax office was closed. Their absence means that the detail of the Meal Ticket Analysis cannot be verified to source data.

The Appellant essentially asked us to draw an adverse inference against HMRC for failure to produce underlying evidence which could have assisted the Tribunal. They referred to the Upper Tribunal judgment in Kyriakos Karoulla v HMRC [2016] UKUT 255 (TCC). It was asserted that the duty of candour on HMRC required them to disclose to the Appellant and include in the Tribunal bundle all relevant evidence.

Whilst we agree that HMRC should, pursuant to the duty of candour, provide all evidence they have to assist the Tribunal, we are satisfied that other than the credit card data from Worldpay (considered above), HMRC had no other evidence in their possession, custody or power which could have been disclosed. HMRC cannot produce evidence they do not have, and we therefore conclude it inappropriate to draw any adverse inference in this case. Regarding the Worldpay data because it is not necessary to support the assessments in the quantum we have determined appropriate given the other evidence and for the other documents because HMRC do not have them and the Meal Ticket Analysis, taken with the over evidence is sufficient to determine an appropriate quantum.

We considered the corporation tax and VAT return data.

A very consistent pattern of trading was evident from this information. In the four years ended 31 March 2013 – 2016 turnover in the accounts ranged from a low of £317,010 (for APE 31/03/14) and a high of £338,879 (for APE 31/03/16). The largest trading loss was £11,628 (for APE 31/03/15) and the largest profit was £7,791 (for APE 31/03/16).

Over the period of trading the VAT returns revealed net declared sales starting in 09/12 of £68,322 and ending with £86,441 in 12/16. The December VAT quarters were consistently (and perhaps not unexpectedly) the highest in each year with £86,441 being broadly consistent with other years in the period from 2012 – 2016. Takings in other periods were between the low in 09/12 and a high in 03/15 at £84,946.

The scoping visit report reveals that the business was open for two sessions seven days per week: 2 hours at lunchtime Monday to Saturday and 3 hours on Sunday; and 5 hours in the evening Sunday to Thursday and 6 hours Friday and Saturday

The records and notebooks were reasonably detailed. They recorded the number of people in the restaurant when the test purchasers arrived, persons arriving and leaving. Also recorded were details of the meals ordered and other observations throughout the evening.

Taken with the initial scoping visit in April we understand that there were 22 tables in the restaurant. We imagine that the majority of those tables were for 2 people with larger tables being created from these. However, by reference to the observation carried out on 26 July 2017 it is apparent that capacity of the restaurant exceeded 46.

The numbers observed on each of the visits was as follows:

Date	Period	Number of customers observed including those present on arrival and those subsequently arriving (where officers saw different numbers both numbers are shown)	Take away
Sunday 24 April 2016	12:10 – 13:25	2	1
Wednesday 4 May 2016	18:30 – 20:00	12	4
Sunday 8 May 2016	18:40 – 20:20	24 (25)	4
Thursday 12 May 2016	18:25 – 19:40	14	1
Friday 20 May 2016	18:55 – 20:45	54	6 (one saw the driver on 3 occasions)
Saturday 28 May 2016	18:30 – 19:50	62 (65)	6
Wednesday 26 July 2017	12:30 – 13:40	10	`
Wednesday 26 July 2017	20:05 – 22:00	50	2

Number of staff observed working was as follows:

Date	Number observed
Wednesday 4 May 2016	4
Sunday 8 May 2016	7
Thursday 12 May 2016	4
Friday 20 May 2016	10
Saturday 28 May 2016	10 (one recorded 8)
Wednesday 26 July 2017	6

The evidence from the test purchasers was consistent that the Appellant used pre-printed meal ticket books which were white. They bore the business name, address and VAT number and were preprinted with consecutive red numbers. Meal tickets were undated.

The majority of test purchasers were not able to observe other parties paying. However, on 28 May 2016 the observations indicated a 50:50 split of the transactions viewed between cash and credit in terms of payment type per transaction rather than value.

All test purchases were paid in cash. Each of the test purchasing groups (though not necessarily each member of the test purchasing pair) identified the meal ticket used in respect of their meal.

The menu reveals that the all you can eat buffet was £14.80 Sunday – Thursday and £16.80 on Friday and Saturday, set meals were £10-13.50 pp, a lunchbox was £5 and a la carte main dishes £5.00 - £6.50, starters £3.20 - £6.50 but mainly around £5, rice £2.50-3. Drinks prices were not listed but from the test purchases soft soda drinks were £1.80 per drink and a Tsing Tao beer was £2.80.

We were provided with detailed spreadsheets which we were told, and accept, recorded the information from each of the meal tickets retained by the Appellant and uplifted on 25 November 2016 and on which VAT was declared.

This analysis demonstrated that only the test purchases on 12 and 28 May 2016 were recorded.

The cash to card ratio for each evening was (as set out in the table below – percentages may vary slightly from recorded cash and card amounts due to adjustments permitted by HMRC arising from calculation errors in the records):

Date	Cash amount	Cash %	Card Amount	Card %
Wednesday 4 May 2016	£20.60	4.03%	£490.60	95.97%
Sunday 8 May 2016	£126.60	13.55%	£777.90	85.25%
Thursday 12 May 2016	£155.30	22.09%	£547.60	77.91%
Friday 20 May 2016	£240.90	13.21%	£1,583.10	86.41%
Saturday 28 May 2016	£314,20	14.56%	£1,850.50	85.44%

The Meal Ticket Analysis shows a disturbing pattern of the use of meal tickets. By way of example, on 20 May 2016 the Meal Ticket Analysis indicates that tickets 49 and 50 (the end of one book) and tickets 1 – 22 of the next book were used during the course of the day’s trade. However, by reference to the credit card transaction data rather than the tickets being used sequentially as each customer requested a bill the tickets were used in the following order 49, 2, 4, 6, 3, 5, 16, 17, 14, 12, 11, 13, 19, 9 22, 7, 18, 8, 20, 15, 10, 21 (cash transactions were 50 and 1). The time between 2 and 3 is 50 mins, and between 11 (used first) and 10 (used later) is 1 hour and 47 mins.

The Meal Ticket Analysis of card transaction data for 20 and 28 May 2016 reveals the following as to the pattern of bills paid:

Date	Lunchtime/ 17:30 – 18:30	18:31 – 20:30	20:31- closing
4 May 2016	6 (all of which were lunctime)	1(test eat missing)	4
8 May 2016	2	4 (test eat missing)	6
12 May 2016	2	3 (inc. test eat)	6
20 May 2016	none	5 (test eat missing)	16
28 May 2016	2	10 (inc. test eat)	12

From meal tickets uplifted for the period 1 – 28 May 2016, HMRC also produced a schedule of “missing meal tickets” these showed that the first ticket on the 1^st was numbered 39, though what then appeared to be 7 whole and two part books there were 71 missing tickets plus the one per day that was used as a summary for the day. Missing tickets therefore represented 19.6% of the purportedly used tickets.

The Meal Ticket Analysis indicates that there are credit card transactions which are not recorded. However, we note that there is no evidence whether a declined transaction generates a transaction number.

HMRC attended at the premises on two separate occasions in order to observe and record the cashing up process and takings for that day. On the first, Friday 25 November 2016, takings for the day were recorded as £2927.40 by reference to the credit card transactions taken and the cash in the till (£730.50). The cash in the till was adjusted for £80 float (representing the amount recorded as stated by Mr Chan when asked what the float was) and £163.00 wages paid in cash to two employees. The total meal tickets for that day plus take away sales was £2975.70. The difference between the two was explained by Mr Chan as arising from customers who disputed payment of the 10% service charge. The cash to credit ratio for that night was calculated at 28.62:71.38%.

The second unannounced visit was on Wednesday 26 July 2017. On this occasion Mr Chan was not present as he was seriously unwell. Mr Fam was managing the restaurant and responsible for cashing up. The cashing up process identified £733.50 cash which needed to be adjusted again for an £80 float (as confirmed by Mr Fam), the payment of £69 as cash wages to staff and £7.80 for sundry purchases made with cash from the till. Total cash take was recorded as £730.30. The credit card terminals were noted by HMRC as recording credit sales as £687.60. In fact the readings from the terminals which were in the bundle show that the total credit card sales for that evening were £817.90.

From the observations on this evening it is apparent that the restaurant was busy with 50 diners observed 6 arriving after 20:05 and a further 10 at lunchtime. These observations would indicate an average meal value per head of £25.80 including lunchtime trade.

When calculating the cash to card ratio HMRC used the incorrect card takings figure of £687.60 and also, and somewhat bizarrely, inverted the ratio. The revised calculations submitted and on which we are asked to determine then used the correct figures and ratios derived from the cash up data.

We have noted that neither Officers Khan nor Beard were, in any way, involved in the investigation or the raising of the assessments. They became involved after the retirement of the officers involved and upon the closure of the Northampton VAT office. Officers Beard and Khan could only therefore explain the basis of the calculations as they understood them. Neither had seen any more information than that contained in the Tribunal plus the Worldpay data.

We found them both to be honest witnesses. Both identified errors in the respective VAT and CT calculations and Officer Khan identified that the discovery assessment for APE 31/03/17 was not validly raised. They were open about the mistakes identified and assisted the Tribunal in these regards; however, they added little else to the evidence before us and on which we needed to determine the appeal.

We found Mr Chan to be a straightforward witness. He answered all questions put to him, but we cannot accept all of his evidence as there were inconsistencies within his oral testimony and in respect of the testimony and the documentary evidence available.

Mr Chan was and remained adamant throughout that there was no cash suppression in the business and that he accurately recorded all takings in the purple book. However, whist he could not remember whether both he and the officer had counted the cash taken on the evening of the first unannounced visit, he accepted that the cash takings recorded on that day were correct including the whole days takings once the cash he had used to make certain wages payments had been added back. He also accepted that he said that the float on that day was £80 plus a separate float of £80 in coins albeit that he said that the float was not necessarily always £80 and on some days could have been higher.

Further, in an open question put to him in evidence in chief, he estimated that credit card takings represented 60-70% of the Appellant’s daily takings. He accepted in cross examination that an average cash percentage of takings at 8.68% as calculated from the declared takings over the four-year period was “too low”.

He could not explain why the test eat purchases on 4, 20 and 28 May 2016 had not been recorded nor was there any explanation of the missing meal tickets on the missing meal ticket schedule referred to at paragraph 47 above.

Mr Chan made no attempt to explain the disturbing pattern of meal ticket use against credit card transaction numbers and times, his answers to all questions regarding this were that he could not understand how it had happened.

For the first time in his oral evidence, he vaguely indicated that cash may have been taken by his staff but there was no other evidence to corroborate the assertion.

As regards the pattern of trade, he accepted in cross examination that the trade as recorded from commencement to the end of the assessed periods was broadly consistent. He could not evidence the previous claim that there was a difference in trading pattern or recorded takings following his appointment as a director in February 2015.

From this evidence we must make certain key factual determinations:

(1)

Whether takings were suppressed

(2)

Over what period

(3)

Whether suppression was a deliberate act

HMRC’s revised VAT Assessments are calculated taking the known credit card sales derived from the Worldpay data and then applying a calculated cash uplift of 35.04%. The uplift is calculated taking the total cash takings for 25 November 2016 and 26 July 2017 as compared to total takings for those days.

Comparison of the Appellant’s declared credit card sales to the Worldpay data indicated that there was limited under declaration of credit card sales. As HMRC used the Worldpay data to calculate the assumed cash suppression at least a proportion of the identified credit card suppression was included in the assessment figures.

We have considered the VAT Assessment schedules which reveal the limited credit card sales suppression. Whilst we are satisfied that the schedules indicate this limited suppression, we do not consider it appropriate to uphold an assessment based on such suppression. There were more than de minimis calculation errors in the Assessment spreadsheets and absent disclosure of the Worldpay data we consider it inappropriate to conclude what, if any, credit card suppression there was. Accordingly, we conclude that VAT assessments based on an uplift to the Worldpay data is inappropriate.

However, on the evidence we consider that there was considerable cash suppression. We base that conclusion on the following evidence:

(1)

The test purchases made by HMRC on 4, 8 and 20 May were not recorded.

(2)

In our view the Meal Ticket Analysis indicates that at least two meal ticket pads were in use at any one time and that the business chose to declare one sequence of numbers and even then only incompletely.

(3)

There was no explanation for the missing tickets in the sequence.

(4)

These factors taken together indicate a concerted arrangement for the suppression of cash sales.

(5)

Mr Chan confirmed that he estimated that between 30 – 40% of sales were cash and a declared cash figure of 8.68% was too low.

(6)

There was no explanation for the anomalies in transaction times.

(7)

The contemporaneous notes of the cash up visits both record that an £80 float was used. We accept that the float may, on occasions though not regularly, have been greater than £80 but we consider that on both 25 November 2016 and 26 July 2017 the float was £80 as stated and recorded as there was no reason for Mr Chan or Mr Fam to have indicated a sum which was different from that used on the day. Changes in float value do not therefore address the significantly larger cash takings on the days of the unannounced visits.

(8)

From the accepted fact that cash in the till was used to pay wages, and for shopping we consider that there was a lack of control or adequate record keeping as to how cash was used within the business.

(9)

The restaurant was regularly busy in the evenings even during the week providing opportunity for suppression.

We find that takings were suppressed over the full period for which we have the Appellant’s corporation tax and VAT return data. Trading was consistent and followed a reasonable seasonal pattern. There was marginal growth over the full period of trading as might be expected but there is certainly no evidence that there was a marked change in the declarations made by the business at any point and particularly not by reference to Mr Chan’s appointment as a director on 1 February 2015.

We do not consider that because HMRC chose not to raise assessments for either VAT beyond 12/16 or corporation tax beyond APE 31/03/17affects the reasonable conclusion reached for all periods prior to 12/16. As regards corporation tax HMRC conceded the APE 31/03/17 discovery assessment on the basis that they had all the information and knowledge they needed to have opened an enquiry into that year but did not do so (despite being in the middle of the enquiry into APE 31/03/16) with the consequence that they did not meet the requirements for raising a discovery assessment. The same would apply for APEs 2018 and 2019.

We make no comment on whether HMRC could have raised VAT assessments for later periods but their decision or failure not to do so cannot affect the clear and consistent pattern in trading over the periods which are subject to this appeal.

We set out below our conclusions on what quantum of tax should have been assessed but we find it impossible to conclude that the suppression we have identified was anything other than deliberate because of its scale (as set out below of the order of £2000 per week).

Further, by his own admission Mr Chan said that cash sales represented 30-40% of trade and that cash takings below 10% were too low and yet for the entire period in which he was a director, declarations were consistently made on the basis of cash takings which were, on average below 10%. We do not see how anyone alive to their duties as a director and making the declarations necessary to allow VAT and corporation tax returns to be signed could have failed to notice the inaccuracies in the turnover recorded and accordingly the behaviour giving rise to the declared inaccuracies must be considered to be deliberate.

On the basis that we have concluded that there was a deliberate suppression of takings throughout the period of trading and in all periods assessed by HMRC both for VAT and corporation tax purposes HMRC were entitled to assess by reference to a period exceeding 4 years. All the assessments are therefore raised within the relevant statutory time limits. The Appellant is also liable to penalties on the basis of deliberate conduct and hence in the range 30 – 70% as the suppression was not concealed from HMRC.

The critical question is whether the assessments are overstated.

The Appellant contends that using only the two data points from the two cash up visits to assess and in particular over such a lengthy period is statistically unsound.

HMRC’s revised VAT Assessments indicate £692,203.09 net under declared sales over 18 VAT quarters. The Appellant asserts that at an average calculated meal value of £19.50 per person the VAT Assessments it is entirely unrealistic requiring approximately an additional 151 customers per week. The Appellant contends that such level of turnover and the associated profits arising from it is entirely unrealistic for the Appellant’s business.

The Appellant contends that if we take the £19.50 average meal value (reduced over the period of the assessment to allow for inflation from 2012 – 2016) and apply it to the totality of the evidence derived from the various HMRC observation of the Appellant’s business and in particular the volume of customers observed scaled up for the full period of trading the quantum of the assessment should plainly be reduced. No clear figures for reduction were provided by the Appellant, however, the grounds of appeal appeared to accept suppression of 5% (but no indication as to 5% of what).

As the Appellant rightly contended (and it was not disputed by HMRC) in light of the direction provided by the Court of Appeal in Pegasus Birds Ltd v HMCE [2004] EWCA Civ 1015 where, as here, we are faced with a best judgment assessment it is our primary task to find the correct amount of tax due on the evidence. However, as per the judgment of the High Court in Van Boekel [1981] STC 290 it cannot be the role of the Tribunal (as it is not the role of HMRC) to do the work of the taxpayer. We must fairly assess the evidence and determine if the sums assessed are reasonable on that evidence. In reaching our decision as set out below we have considered all the evidence. We have probably gone considerably further than we are required to do in terms of evaluating and determining a fair assessment of the quantum of tax which should be due. In doing so we do not impugn (save for the obvious mistakes) the exercise undertaken by HMRC. As indicated below the assessment raised came within the range of what might be considered to be reasonable on the evidence. However, in our view it was at the upper end of the reasonable range and overall we considered it more likely that the amounts we have determined are closer to the correct figures.

In this regard and given the absence of the underlying data from Worldpay we consider that the quantum of the assessments should be determined by reference to the Appellant’s acceptance that cash takings were not the 9% declared but of the order of 30-40%.

We therefore take as our starting point the declared cash totalling £151,623.23. We gross up that figure on the assumption that the correct cash proportion is 30, 32.5, 35, 37.5 and 40% as follows:

Assumed cash percentage	Calculation	Assumed cash
30%	Cash/0.0868*0.3	£524,043
32.5%	Cash/0.0868*0.325	£567,713
35%	Cash/0.0868*0.35	£611,383
37.5%	Cash/0.0868*0.375	£655,054
40%	Cash/0.0868*0.4	£698,724

We note that HMRC’s assessment is at the top end of this range.

By reference to these sums and taking account of the Appellants declared credit card sales of £1,664,212.20 (as opposed to the Worldpay data) total turnover for the 18 VAT periods would be:

Assumed cash percentage	Assumed cash	Declared credit sales	Total assumed sales
30%	£524,043	£1,454,935	£1,978,978
32.5%	£567,713	£1,454,935	£2,022,648
35%	£611,383	£1,454,935	£2,066,318
37.5%	£655,054	£1,454,935	£2,109,989
40%	£698,724	£1,454,935	£2,153,659

In each instance that represents a suppression of total sales as follows:

Assumed cash percentage	Declared sales total	Suppressed sales	Suppressed sales percentage declared total
30%	£,1727,784	£372,420	21.55%
32.5%	£,1727,784	£416,090	24.08%
35%	£,1727,784	£459,760	26,61%
37.5%	£,1727,784	£503,431	29.14%
40%	£,1727,784	£547,101	31.67%

These figures substantially exceed the 5% indicated by the Appellant. The bottom end of the range is consistent with the 19.6% of missing meal tickets and the top end of consistent with the unrecorded test purchases.

We therefore determined to test these figures by considering the number of additional customers which would be required to generate the additional sales. As indicated above the Appellant urged us to conclude that the average meal price was £19.50. Having considered the test purchase data, the bill values declared on the days of the test as revealed by the Meal Ticket Analysis, the average meal value calculated for 26 July 2017, the prices on the menu and making an assumption that most customers do not drink only soft drinks, (and some will drink large quantities of alcohol) we consider it reasonable to impute an average meal price of £24 over the period (which takes account of inflation from 2012 through to the average noted on 26 July 2017).

The number of additional customers per week on the basis of the permutations of suppression are:

Assumed cash percentage	Suppressed sales	Number of customers (total additional sales divided by (18 periods, x 13 weeks per period x £24)
30%	£372,420	66
32.5%	£416,090	74
35%	£459,760	81
37.5%	£503,431	89
40%	£547,101	97

Having carefully considered all the evidence we consider that calculating suppressed sales by grossing up the declared cash take such that cash represents 35% of takings as set out above results in a reasonable and fair estimation of suppression. We do not consider that the increase in customers which are implied are unrealistic or unreasonable on the whole of the evidence available to us given the capacity of the restaurant and the observations of HMRC’s officers on how busy the restaurant was on each of the days of observations or by reference to the pattern of business revealed and discussed above.

The additional VAT due across the 18 VAT periods is thereby £76,626. That amount is to be allocated to VAT periods by applying the factor of 3.456 to the Appellant’s declared sales in each prescribed accounting period.

By their skeleton, the Appellant contends that the losses in cash identified by HMRC are to be explained by staff theft such that there should be no corporation tax cost to the business. This allegation of theft by employees amounting to a little short of £2,000 per week was raised for the first time in the skeleton which was served on the morning of the hearing rather than 21 days prior to the listing. It was unevidenced. The matter does not appear to have been reported to the police, there was no evidence that employees had been taken to task over the alleged theft any disciplinary action.

The only basis on which this assertion can realistically even be predicated is that no officer of HMRC saw Mr Chan receive or pocket the cash. However, we are not specifically concerned with Mr Chan but with the Appellant corporate entity. Even if Mr Chan did not take or benefit from the cash (see below section 455) it does not mean that it was stolen from the business in the sense required to be excluded from the profits calculation.

Further, it would be expected that if staff were stealing from the business the sums would be taken on an individual meal by meal basis where the meal bill was paid in cash and not from the takings at the end of the night. Had such theft been occurring then it would be expected that it would have occurred prior to cashing up and that would indicate that the takings for 25 November 2016 and 26 July 2017 were also subject to suppression which would increase the assessments.

We consider the assertion as to theft by staff to be a spurious allegation feebly aimed at reducing the corporation tax charge and we do not accept it, particularly as it was raised within 24 hours of the hearing and no evidence provided to support the suggestion.

The Appellant also contended that HMRC were not entitled to rely on the principle of continuity so as to conclude, for corporation tax purposes at least, that the suppression demonstrated in 2016 and/or 17 could be presumed for earlier years. In this regard they referred to the judgment of the Upper Tribunal in Stirling Jewellers (Dudley) Limited v HMRC [2020] UKUT 0245 (TCC). On the basis of our factual findings, we do not consider the Stirling judgment to be relevant. Here HMRC do not need to apply a presumption of continuity per se as such a presumption is only needed where there is no positive evidence of continuity. Here, there is evidence that turnover was consistent throughout the period assessed such that it is highly likely that cash sales were similarly consistently suppressed.

The Closure Notice and the Discovery Assessments have been calculated by reference to the suppressed turnover calculated for VAT purposes. On the basis that we have determined that the VAT Assessments are overstated the Closure Notice and Discovery Assessments will be similarly overstated. When determining the amendment for the Closure Notice and raising the Discovery Assessments HMRC made no adjustment for increased costs.

Officer Khan explained that HMRC would usually only adjust the profits calculation where suppressed sales had been identified if the taxpayer provided additional information as to what were also suppressed costs. The Appellant had been given the opportunity to comment on the proposed assessments and had not provided evidence of additional costs. Absent any such information Officer Khan considered no adjustment to profits appropriate.

We agree. Had the Appellant admitted the suppression it might have then also been able to identify additional costs. It has not and it is therefore reasonable that the assessments be adjusted only to reflect the amendments we have made to the VAT Assessments.

We have not undertaken the calculations ourselves and leave that to HMRC. In accordance with the decision we have reached.

We were told, and accept, that it is HMRC’s practice to assume that where the takings of a limited company are suppressed they have been extracted by the participants (be that the shareholders or the directors). HMRC do not undertaken any tracing activity or look to establish which participants have benefitted on the basis that it is a reasonable assumption to have made.

The Appellant contends that HMRC were not able to identify Mr Chan as the person in charge on any of the covert visits and it is therefore unreasonable to impose a section 455 charge.

That argument completely misses the objective underpinning section 455. That section taxes the participants on the basis of extracted profits. No corporation tax assessment is raised where suppression is as a consequence of theft from the company by someone other than a participant. Where there is no evidence of such theft and no alternative explanation provided there is no evidence on which to set aside HMRC’s assumption and the section 455 aspects of an assessment should stand.

Accordingly, save for the adjustments now required consequent upon our conclusion on quantum of suppression the section 455 assessments are appropriate.

The penalties essentially flow from our conclusion that the suppression was deliberate.

As the penalties are calculated by reference to the potential lost revenue HMRC will need to amend them in line with our above conclusions on quantum.

We have considered whether further allowance should be given for telling. When questioned Officer Khan indicated that he considered that the percentage reduction applied to the penalties was broadly reasonable. He accepted that he may have given an element of reduction for telling but considered that full reduction for helping to have been generous.

As HMRC acknowledge there is an overlap between telling and helping. Both telling and helping permit an up to 30% reduction each in the penalty (as between the maximum and minimum permitted – in this case between 70% (maximum) and 30% (minimum) of the potential lost revenue). HMRC’s guidance explain that telling includes “admitting the failure, disclosing the failure in full and explaining how and why the failure occurred” proceeding to explain “the person needs to show a positive approach to tell what has happened not just reacting to questions” and “the extent is whether everything is disclosed to us … overall, the telling … must be positive and as complete as possible.” Helping is explained as giving reasonable help in quantifying the amount of tax, positive assistance, active engagement and volunteering of information.

The Appellant has been given half of the maximum discount for helping and giving when the two are taken together. We consider that reasonable. The Appellant did co-operate with the investigation and was not obstructive. However, apart from an acceptance that 5% of the assessment should stand (which could have been the service charge element) there was never an admission to suppression until the acceptance in cross examination that 30-40% of sales were in cash. Whether the discount could have been differently allocated is not a matter which justifies further analysis. The Appellant has been given the level of mitigation we consider appropriate.

For the reasons we have stated we allow the appeal in part:

(1)

The appeal in respect of the discovery assessment and associated penalty for APE 31/03/17 is allowed in full.

(2)

The appeal in respect of the VAT Assessments is allowed in part. The quantum of the assessment for each prescribed accounting period is to be calculated applying a factor of 3.456 to the declared cash turnover. Which results in a reduction of for the overall VAT Assessments from £147,728.47 (original assessment)/£136,709 (revised assessment) to £76,626.

(3)

The appeal in respect of the VAT Penalties is allowed in part. HMRC had already reduced the VAT Penalty Assessments from £67,213.92 to £53,715.08 having conceded that the errors in respect of the service charge and calculation errors were careless. Further reduction will be required in consequence of the reduced VAT Assessments.

(4)

The appeal in respect of the Closure Notice and Discovery Assessments and associated Penalties is allowed in part. HMRC are to reduce them by reference to the adjusted turnover computation arising from our decision on the VAT Assessments.

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.