The initial contact by HMRC was made by Officer Harding. The correspondence before us shows that her correspondence was with AH and members of the accountancy firm at which he worked.

The first visit was made by Officer Harding to the accountancy firm on 26 February 2015.

Officer Harding did review some records for 05/14 after that visit (ie for a period before the Appellants acquired the business). That review included reviewing purchase invoices and her letter of 28 April 2015 referred to the analysis of the purchase invoices for 05/14 (in the absence of retail scheme calculations and gross takings records supported by till rolls) supporting 75.44% of sales being standard-rated. Officer Harding also referred to the standard-rated percentage in 11/13 having been 82%.

There was further correspondence between Officer Harding and the accountants and on 7 January 2016 Officer Harding sent her findings for the 05/14, 08/14 and 11/14 returns:

On 28 January 2016, AH emailed HMRC copies of some of the requested Costcutter reports (for December 2013 and January, February, April and May 2014). He said that VAT returns were based on the point of sale retail scheme, with the split of standard-rated to zero-rated sales and output tax due informed by the Costcutter reports from the tills.

There was an exchange of emails about the missing Costcutter reports (for March, June, July and August 2014) to support the actual daily gross takings and querying a low output tax declaration in 11/14. AH said they were not available – the previous owners did not have them and they cannot be obtained from Costcutter. Being asked further about this, AH said that the till was installed and maintained by Costcutter but the printouts that were downloaded from the computer belonged to the trader.

Officer Fellows had taken over from Officer Harding after the end of January 2016, and in May 2016 Officer Fellows told AH that if the detailed till audit trail to support the monthly readings was not available it may be possible to extract the data from the till.

Officer Fellows visited AH on 26 May 2016 (the Appellants themselves were not present). Officer Fellows was not given any information about the sales figures or how the VAT workings were done at that meeting. There was a record of the figures they had entered on SAGE for each month, but there was no back-up data (eg no till rolls, no audit trail).

On 29 June 2016 AH emailed Officer Fellows stating that the records she had asked for (sales transactions) do not exist, as they were originally maintained on a computer belonging to the previous partnership. He said that the Appellants were happy for her to interrogate their computer system. He added that he had now drafted the August 2015 accounts, which showed an unexplained cash difference of £114,722. He said that after discussion with his client this had been accepted by him (ie the client) as probably relating to sales, as he had no confidence in the till recording system, leading him to invest in a brand new system. He had included 5/6 of this amount, £95,602, as sales in the accounts and the VAT amount of £19,120 as due to HMRC.

On 1 July 2016 Officer Fellows asked AH about the tills used (to enable her to make a decision on till interrogation) and AH replied later that day saying the response from the client was “Tills are Toshiba A10 (two tills, one has just had its hard drive replaced a few months ago) and the version of software running is IT Retail Systems Version 2.1 …which I believe is bespoke to my needs now”. AH went on to say that sales were recorded on a computer system having the Nisa software which the client did not have access to and he had to get someone to print these for him.” AH also said that client said that the reports “did not mean much to him and he had little or no confidence in the summaries”. This email also said that the Appellant was lax and no reconciliations were attempted, which meant that he, AH, had discovered the cash difference.

On 15 July 2016 AH told Officer Fellows that he estimated that for the year to August 2015 the ratio of zero-rated to standard-rated products was approximately 15:85. He suggested that the VAT due on the undeclared sales was £16,252 to be split equally over the four quarters.

In her response later that day Officer Fellows indicated that 85% appeared to be higher than achieved in earlier returns, where the range was 35% to 75%. She asked for the output figures on the VAT returns, without incorrectly including lottery commissions and PayPoint activity, before she decided if further checks were required. Officer Fellows then set out questions arising from her checks on purchases, eg duplicate entries. Giving evidence, Officer Fellows said she was looking at the details on SAGE – she did see purchase invoices but did not audit each invoice.

Officer Bray visited the premises with Officer Tilt on 7 September 2016 to copy the data from the tills. Mr Rai, Mr S Rai and AH were present. The Appellants had given permission in advance, and gave permission whenever requested by Officer Bray during the visit.

Officer Bray copied the data from two tills onto a USB stick and then used that stick to copy the data onto an HMRC-encrypted laptop at the premises. The USB stick was used for the data transfer so that there is no need to attach or link the two computers. That data was later converted by HMRC specialists into a readable format and Officer Bray used that to produce the Data Analysis and the Bursha Data Report.

The parties disagreed as to what happened to the USB stick that was used to copy the data from the tills onto Officer Bray’s laptop. HMRC’s position was that it was handed to one of the Appellants (for this purpose, being Mr Rai or Mr S Rai, not AH as their representative) at the visit; whereas the Appellants denied this, and said HMRC should have it.

The evidence before us was:

We recognise that there was no evidence before us from Mr S Rai or AH, and that the events in question were several years ago. Considering all of the evidence, we prefer Officer Bray’s evidence as to the general procedure in relation to USB sticks being left with the trader and his recollection of events on the day. We find that the USB stick was given to either Mr Rai or Mr S Rai at the premises that day, and HMRC did not take it with them.

After the visit, Officer Bray sent a copy of the uplifted data to HMRC’s Data Conversion & Audit Team (“DCAT”) on 25 November 2016, asking them to carry out a conversion into conventional text and tables. It was sent to DCAT via the network (even though they were operating in the same room). DCAT provided him with the converted data on 1 December 2016. This was provided to him as a set of IDEA files.

Officer Bray used this converted, readable, data to produce the two documents which he sent to Officer Fellows. Giving evidence he explained that IDEA is a strong audit tool that automatically keeps a history of the audit trail, and records every step taken with the data. He then exported the sheets onto Excel. He did then make some further additions within Excel – he prepared a small table showing the estimated underdeclared VAT for each quarter, and inserted some totals columns into the Z readings to show the total value of VATable trade departments and non-trade departments. We accept this evidence.

Although the Data Analysis report is more detailed than the Bursha Data Report, and the former was used to produce the latter, we start by summarising the Bursha Data Report as that sets out the range of data retrieved and conclusions drawn therefrom. We do not lose sight of the Appellants’ submissions as to the source of the data, namely whether it was from their tills and, if so, was the software giving data from their shop. We address those challenges having set out the data retrieved.

The Bursha Data Report includes:

The Data Analysis was produced in the form of Excel sheets. In summarising these sheets, we have also included some of the evidence of Officer Bray. The sheets include:

Officer Bray emailed these reports to Officer Fellows on 7 December 2016.

We find that the data which was emailed to Officer Bray by DCAT, analysed by him and formed the basis of the Bursha Data Report and the Data Analysis was from the two tills at Bursha Foods which had been extracted by HMRC on 7 September 2016. We find that the process of analysing the data and producing the data sheets was performed carefully and did not introduce errors or mistakes into the data, notwithstanding that Excel sheets are editable. We make these findings on the basis of Officer Bray’s evidence, the fact that the reports identify two tills, one in more continuous use than the other (with the left till being Till 1), the end date of the data being the date of HMRC’s visit to Bursha Foods, and the list of departments being used (which Mr Rai agreed looks like a standard list for a convenience store).

These findings do not of themselves address Mr Doshi’s separate challenge as to whether the data on the tills was that of Bursha Foods given that there had been a change of software when the Appellants changed suppliers and a replacement hard drive. We address that, and make findings in relation thereto, in the Discussion.

The Bursha Data Report and Data Analysis were prepared for internal HMRC use. In February 2017 Officer Fellows told Officer Bray that she would like to provide the reports to the Appellants’ agent. Officer Bray produced amended versions of these reports – he redacted references to HMRC strategies and its intelligence on till software providers, and added the term “sample” sales. These revised files were emailed to Officer Fellows on 13 February 2017.

These revised files were emailed by Officer Fellows to AH on 17 February 2017.

Officer Fellows issued the VAT assessments on 17 July 2017, some of which were amended on 28 July 2017. Those assessments (as amended) are detailed at [20] above. A copy of them was sent to Mr Doshi, who was by this time instructed to act for the Appellants.

There was a meeting on 18 August 2017 between Officer Fellows, Officer Bray and Mr Doshi. At that meeting Mr Doshi accepted that HMRC had obtained till information from his client’s tills on 7 September 2016. Officer Bray explained the methodology he had used. In preparing his reports. Officer Bray reminded Mr Doshi that his client had been given the USB stick, which contained an exact copy of the information HMRC had taken away on that visit.

There was further correspondence between HMRC and Mr Doshi. Then, in an email of 5 December 2017, Mr Doshi told Officer Fellows:

Following further emails, HMRC issued the penalty explanation letters on 29 March 2018 (which are considered in the context of the Discussion on penalties below). The notices of penalties were issued on 22 May 2018. The penalties issued are at [20] above.

There was further correspondence between Mr Doshi and HMRC and the Appellants gave notice of appeal to the Tribunal on 6 April 2019. The hardship application was granted on 24 July 2019.

The issues are set out at [23] above. We address first those relating to the output tax assessments and then the penalties (both deliberate and careless in respect of those assessments).

The jurisdiction of the Tribunal in relation to an assessment under s73(1) VATA 1994 has been considered extensively in the authorities. The correct approach to take was considered by the Upper Tribunal in Mithras (Wine Bars) v HMRC [2010] UKUT 115 (TCC) and was common ground between the parties. In summary:

We have set out in our findings of fact the enquiry undertaken by HMRC. During that time they were not provided with till rolls or the documents/numbers which were used by AH to prepare the VAT returns on behalf of the Appellants.

Officer Fellows has explained the basis on which the assessments were issued (both on 21 July 2017 in a letter to Doshi & Co, and confirmed in evidence at the hearing):

We consider that the approach taken by Officer Fellows was more than reasonable. The assessment for 11/14 was based on the information which was given to her by the Appellants’ then accountant. The assessments for the following four periods were based on the data extracted from the Appellants’ tills and which had been carefully analysed by Officer Bray. Not only do we consider that Officer Fellows made the required value judgment honestly and bona fide, we do not consider that Officer Fellows could have adopted any other approach.

Mr Doshi submitted that there must be doubts as to the accuracy of the Data Analysis given the changes to the till software and the replacement hard drive, and that Officer Fellows had not taken account of other information which had been available to her. We reject those submissions:

HMRC have established that the assessments were issued to the best of their judgment.

We have full appellate jurisdiction in respect of the Appellants’ appeal against the amount of the assessments, and can consider all of the evidence and submissions before us for this purpose. We do not lose sight of the fact that the Appellants bear the burden of establishing that the amount of each or all of the assessments should be reduced.

Both parties acknowledged that we were being presented with a binary choice – between the amounts declared by the Appellants in the VAT returns and the assessments which had been issued. In the course of the submissions and evidence at the April 2024 Hearing we were taken to some of the correspondence between the parties, and that included some emails from Mr Doshi to HMRC in which he proposed compromises based on adjustments from the accounts. Mr Doshi did not submit that we should consider whether those variations represented the correct amount of tax.

We consider first the VAT returns as filed, then Mr Doshi’s challenges to the amount of the assessments which have been issued for the VAT periods in issue.

The output tax declared in the VAT returns for the business for the periods in issue is set out at [52] above. In relation to these returns:

In this situation, there is minimal evidence before us supporting the accuracy of the VAT returns as submitted.

We have set out at [92] the basis on which Officer Fellows calculated and issued the assessments. We assess the challenges put on behalf of the Appellants before then reaching our decision on the correct amount of tax on the basis of all of the evidence and submissions before us.

Mr Doshi submitted that the Tribunal cannot be 100% confident that the data it has in front of it is the correct data copied from the two tills on 7 September 2016 and relates to Bursha Foods given that there had been changes of software and the hard drive in Till 2 had been replaced. Mr Doshi submitted that the conclusions based on the Data Analysis were illogical and unrealistic for the business of Bursha Foods. Mr Doshi’s submissions also criticised the evidence before this Tribunal (namely the absence of the USB stick and the failure to produce a pdf of the data in the form extracted from the tills, ie before it was analysed and the sheets extracted in the form of the Data Analysis).

We have already found that the USB stick was given to either Mr Rai or Mr S Rai at the premises on 7 September 2016. After that date, it was no longer in HMRC’s possession and cannot be produced by them. It was for the Appellants to retain that stick and, once it became evident to them that they no longer had it, to decide what further actions to take, which could have included instructing a data specialist to copy the data from the tills again (although we recognise that the replacement of Till 1 after the visit would have prevented the Appellants from obtaining a copy of some of the data extracted by HMRC).

We have also found that the data analysed by Officer Bray was from the two tills at Bursha Foods which had been extracted by HMRC on 7 September 2016 and that the process of analysing the data and producing the data sheets was performed carefully and did not introduce errors or mistakes into the data.

We consider here Mr Doshi’s separate challenge as to whether the data on the tills at the premises was that of Bursha Foods given that there had been a change of software when the Appellants changed suppliers. We have already found that the hard drive on Till 2 was replaced in May 2016 and the software on both tills had been changed before that date, ie before the date on which HMRC visited the premises and extracted data from the tills. Mr Doshi’s submission was essentially that the data extracted by HMRC could have been from different businesses if the software had previously been used in another store.

Addressing these submissions about the changes to the tills and software (before considering separately the submissions about the transactions which were said to have been recorded on the tills):

It was not helpful that the Appellants were not able to adduce evidence regarding the exact timing of the change of suppliers, or the date on which the software had been changed in the tills. In the absence of such evidence, and taking account of the two reports themselves and Officer Bray’s evidence, we provisionally concluded that it was more likely than not that the data extracted from the tills was that of Bursha Foods, notwithstanding changes in software or the hard drive which had occurred. We place particular emphasis on the continuity of the data and departments for this purpose, and that the pattern of how the two tills were used accords with the evidence of Mr Rai. We then considered whether the submissions and evidence as to the transactions which were shown in the Data Analysis supported this conclusion or whether they were evidence to the contrary.

Mr Doshi put forward various challenges to the Data Analysis at different times (whether in his written submissions, in opening at the April 2024 Hearing, during examination of Mr Rai and cross-examination of Officer Bray and also in closing submissions). We have taken them all into account and address them below (albeit we have approached them in categories rather than each individual submission) before making our assessment based on all of the evidence in its entirety:

We have remained mindful of both the burden of proof and the standard of proof, seeking to assess the whole picture presented to us. We consider carefully the evidence as to the numbers of customers, the Camelot data and the average spend per customer that is implied by the Data Analysis. We agree with Mr Doshi that some aspects could be viewed as unrealistic. However, we were troubled by how little documentary evidence we were given as to the business of the Appellants to enable us to assess fully these challenges. The VAT periods in issue are clearly several years ago, and we accept that Mr Rai does not have the manuscript records he made at the time of the takings (and that he kept no other records at that time). However, whilst Mr Rai described the Data Analysis as being woefully incorrect, he did not produce documentary evidence to support or explain this statement. Not only have the Appellants not produced any copies of that data from, eg, Costcutter or Nisa, but also they have not sought to show us how the business operated after this period, eg producing till records for a later period, with copies of purchase invoices and sales data, stock ordering, bank records, to present a picture of the level of activity of the shop.

Having considered all of the evidence (including our consideration of the changes to the tills and software) we were not, on balance, persuaded that the data that was extracted was from a different business. Whilst we have reached this conclusion as to the source of the data, we do take account of the challenges put (eg as to number of transactions) when assessing the correct amount of tax.

The challenge to the assessment for 11/14 included Mr Doshi submitting that HMRC should not have accepted the explanation being provided by AH and should not have issued an assessment in reliance on that explanation.

Mr Rai’s evidence was that there had been no discussion between AH and him regarding this and he had not accepted (as AH had said in his email of 29 June 2016) that the difference related to unreported sales.

The difficulty is that AH was acting for the Appellants at that time, was authorised to correspond with HMRC on their behalf, and whilst Mr Rai says that this cash difference did not exist, we were not shown any financial information to assist with making findings in this regard. AH had not been called as a witness in this appeal, so the accusations being made by Mr Rai could not be put to him. Furthermore, Mr Doshi’s submissions were that we should prefer the amounts declared in the VAT returns to those assessed, yet the VAT returns were prepared by AH.

Mr Doshi submitted that Officer Fellows had checked purchase invoices and had only identified a few duplicate claims for input tax. She had not challenged the ratios of purchases between zero-rated and standard-rated. Mr Doshi submitted that it was therefore not logical for Officer Bray to come up with standard-rated sales at 73.6% to 83.2% when the standard-rated purchases were 55%. Furthermore, if standard-rated purchases were increased in the relevant periods to match those assessed as outputs, then the input tax would increase, either reducing or wiping out the assessment.

Giving evidence, Officer Fellows emphasised that she had checked only a sample of the purchase records. In cross-examination, she said that purchases would not necessarily correlate to sales in any particular VAT period. We consider this as a submission, and in principle we agree.

The evidence to which we were taken showed that just as there were different proportions of standard-rated sales in the Data Analysis on which the assessments for 02/15 to 11/15 were based, in other periods the standard-rated purchases had fluctuated, eg in 05/14 Officer Harding had identified that the purchase invoices (in the absence of retail scheme calculations and gross takings records) supported 75.44% of sales being standard-rated. This was not one of the VAT periods in issue, but it does illustrate the variation (which we would expect to exist).

Mr Doshi’s submissions included the effect of making further adjustments to the proportion of standard-rated sales, and corresponding adjustments to the input tax claimed for each period. However, these submissions were not grounded in documentary evidence of the underlying transactions.

Mr Doshi challenged the VAT scheme which had been used by HMRC to issue the assessments. This particular challenge was first made in his cross-examination of Officer Fellows – it had not been made in the grounds of appeal, written submissions, in oral opening or been raised as questions to the first two witnesses, Mr Rai and Officer Bray.

We considered that whilst the challenge had not been particularised, it was broadly within the scope of the Appellants’ pleaded challenge to the amount of the assessments. On that basis, we considered that the submissions could be made, but the timing of them has potentially affected the evidence (in particular as questions were not put to Mr Rai or Officer Bray in cross-examination or re-examination).

Mr Doshi submitted that HMRC had impermissibly used different retail schemes when issuing the assessments – the 11/14 assessment was based on apportionment methodology (consistent with the approach taken by Officer Harding for assessments for earlier periods), whereas the assessments for 02/15 to 11/15 were based on a point of sale scheme (as they were based on the till data).

Giving evidence Officer Fellows said that she had been told by AH that the Appellants had moved to a point of sale scheme. We accept this, and it is consistent with Mr Rai providing till data to AH for AH to use to prepare the VAT returns.

The assessments for 02/15 to 11/15 were thus issued on the same basis as that which was being operated by the Appellants. Officer Fellows agreed that she had used an apportionment approach for 11/14, explaining that this was the evidence she had – there was no till data, which would be required to make point of sale calculations.

Whilst Mr Doshi submitted that businesses must continue to adopt their chosen retail scheme in successive periods, he did not explain to us why HMRC, when issuing assessments and having limited information, are not permitted to use whichever method is workable in the light of the information available. We have not placed any weight on this submission.

Mr Doshi submitted that we cannot be 100% confident that the amounts assessed by HMRC are correct. We agree, but that is not the relevant standard of proof. We are required to reach a decision as to the correct amount of tax on the balance of probabilities.

We are not persuaded that we should prefer the amounts declared in the VAT returns, given the lack of documentary evidence supporting those (whether the figures provided by Mr Rai to AH or AH’s workings) and the accusations which have been made against AH.

In this situation, we have considered carefully the assessments issued by HMRC and all of the evidence before us as to how they were calculated and the challenges which have been put by Mr Doshi. We recognise that there are some matters which raise questions, in particular the customer count in the Data Analysis, the average spend which results from this and the apparently high proportion of sales of Cigarettes. However, we were not satisfied that there was sufficient evidence before us which would support any specific adjustments or corrections being made to the amounts assessed. On balance, in the light of all of the evidence, we have concluded that each of the assessments assessed the correct amount of tax.

The appeal against the assessments is dismissed.

HMRC issued penalty assessments for the underdeclarations of output tax in the five periods in issue. HMRC had issued penalties for 11/14 and 11/15 on the basis that the conduct was careless but for 02/15 to 08/15 on the basis of deliberate conduct (whilst pleading in the alternative that the conduct was careless).

The Appellants have appealed against the decision that a penalty is payable, against the decision as to the amount of the penalty and HMRC’s decision not to suspend the penalties. We have jurisdiction to affirm or cancel HMRC’s decision to issue the penalties, and to affirm the amount or substitute another amount (that was within HMRC’s power).

We set out below the basis on which the penalties were issued according to the penalty explanation schedules, including the reductions allowed for quality of disclosure, and then reach our conclusions.

The penalty explanation schedule for the periods 02/15, 05/15 and 08/15 explains the basis on which HMRC considered the behaviour was deliberate:

A 40% reduction for the quality of disclosure was applied (0% for telling, 10% for helping and 30% for giving). HMRC say that the Appellants did not accept that the sales value was incorrect or explain how the errors arose, the Appellants’ accountant has discussed the findings but did not provide positive assistance to quantify the assessment, and the Appellants did give access to the records at the accountant’s premises and allowed HMRC to extract data from the till.

There was no reduction for special circumstances.

The penalty explanation schedule sets out HMRC’s conclusion that the behaviour was careless on the basis that cash differences had been identified by the accountant in the preparation of the accounts for 30 August 2014 and he estimated that there were further differences for 30 August 2015. The amount of sales understated for 11/14 was based on one quarter of the cash differences calculated. The behaviour is careless because of the lack of controls in place to record and report the correct amount of sales.

A 90% reduction was allowed for quality of disclosure (20% for telling, 40% for helping and 30% for giving). This was explained as the accountant accepted that the output tax had been understated but could not explain how the errors had occurred. He provided help to estimate the level of adjustment required to sales, and responded to requests for information.

HMRC have considered special reduction but considered no circumstances to warrant such a reduction.

HMRC said they can only suspend a penalty if they can set conditions to help avoid penalties in the future and if we think you can meet these conditions. Here, they cannot suspend the penalty as the VAT registration has been cancelled so no future returns will be submitted

The penalty explanation schedule sets out HMRC’s conclusion that the behaviour was careless:

A 50% reduction for the quality of disclosure was applied (10% for telling, 10% for helping and 30% for giving). HMRC say the accountants have accepted the basis of the calculation for this period, the Appellants’ accountant has discussed the findings but did not provide positive assistance to quantify the assessment, and the Appellants did give access to the records at the accountant’s premises and allowed HMRC to extract data from the till.

HMRC have considered special reduction but considered no circumstances to warrant such a reduction.

HMRC said they can only suspend a penalty if they can set conditions to help avoid penalties in the future and if we think you can meet these conditions. Here, they cannot suspend the penalty as the VAT registration has been cancelled so no future returns will be submitted

In his written submissions, Mr Doshi had submitted that there should be no penalties, that any inaccuracies were neither deliberate nor careless, with the only concession at that time being that the input tax inaccuracies were careless. The appeal against those penalties has been withdrawn. However, at the April 2024 Hearing Mr Doshi accepted that the conduct of the Appellants in respect of the underdeclarations of output tax was careless. This was the basis on which HMRC had issued the penalties for 11/14 and 11/15.

We consider first whether HMRC have established that the inaccuracies in the 02/15, 05/15 and 08/15 VAT returns were deliberate.

The Tribunal in Auxilium Project Management Ltd v HMRC [2016] UKFTT 249 (TC) explained a deliberate inaccuracy in the following terms:

This explanation has been approved by the Upper Tribunal in CF Booth Ltd v HMRC [2022] UKUT 217 (TCC) at [36]-[37]. We adopt and apply that approach.

The penalty explanation schedule based the conclusion that the conduct was deliberate on the accountants having prepared the VAT returns using figures provided by the Appellants, and that the shortfall was consistently great such that the Appellants must have deliberately provided suppressed numbers.

Mr Lindsay submitted that the divergence meant it was the inevitable conclusion that amounts were deliberately suppressed. Being asked to explain HMRC’s decision to impose penalties on the basis that the conduct was careless for two of the relevant periods, Mr Lindsay referred to the fact that for 02/15, 05/15 and 08/11 HMRC had the Data Analysis from the tills for the entirety of each of these three periods, showing the level of suppressions.

We recognise that it is entirely possible that inaccuracies may be the result of different conduct in different periods, such that penalties should be assessed on different bases, or no penalty issued at all. Here, there was no evidence that the conduct of the Appellants had changed in relation to different periods in issue. We do not regard the reason given by Mr Lindsay for the penalties being imposed as explaining HMRC’s decision in this regard – the existence of the till data explains the basis of calculation of the assessments by HMRC, not the Appellants’ conduct in relation to the preparation and submission of the VAT returns.

Here, we have found that Mr Rai provided gross takings numbers from the tills and Costcutter reports generated by the tills to AH to enable AH to prepare the VAT returns. Mr Lindsay submitted that there has been no explanation for the inaccuracies in the VAT returns. We disagree – Mr Rai’s evidence was that he had not seen the VAT returns before their submission, and the lack of copies of information provided to AH, together with Mr Rai’s accusations against AH, means that the explanation being put forward is essentially that Mr Rai provided correct information to AH but that information was not then used to prepare accurate VAT returns. We make no findings as to the conduct of AH. However, on the basis of the evidence before us, we are not satisfied that the Appellants knew that the VAT returns being submitted by AH were inaccurate. Accordingly, the conduct was not deliberate.

Mr Doshi has accepted that the Appellants’ conduct was careless. We agree, and we base this conclusion on the Appellants’ failure to keep any records, or to require that VAT returns be shown to them in draft beforehand (as we consider that they could reasonably have been expected to identify whether the sales figures shown on returns matched those they had provided to AH). This conclusion applies to all five periods for which penalties were issued.

HMRC has applied different mitigation to the penalties assessed on the basis of its conclusions as to the quality of disclosure:

The reasons for these differences were set out in the relevant penalty explanation schedules.

Whilst the Tribunal has jurisdiction to substitute its own decision as to the appropriate level of mitigation, we do not consider it would be fair or in the interest of justice to reduce the level of mitigation which has been allowed – this had not been pleaded by HMRC and the Appellants did not therefore address this possibility in submissions. Instead, we focus on the Appellants’ submissions that higher mitigation should be allowed.

We take particular account of the following:

We consider that the permission for HMRC to extract data from the tills was significant, but so was the failure to produce VAT workings. Overall, and having considered all of the correspondence between AH and HMRC to which we were taken, we allow mitigation of 60% in total for telling, helping and giving. We would allow the same level of mitigation for each of the five periods in issue (as we do not consider that the quality of disclosure varied during this time, and the absence of till data for 11/14 was not the result of actions of the Appellants); however, we have not, for the reason set out above, reduced the 90% level which HMRC had decided to allow for 11/14.

We do not consider that HMRC’s decision in relation to the absence of special circumstances was unreasonable.

HMRC has refused to suspend the penalties. The penalties imposed for 02/15 to 11/15 were ineligible to be suspended under paragraph 14 as only penalties for careless inaccuracies can be suspended. On the basis of our decision, all five of the penalties are potentially eligible to be suspended. However, we agree with HMRC that as the Appellants’ VAT registration number has been cancelled, the power should not be exercised as paragraph 14(3) provides that HMRC may suspend a penalty only if compliance with a condition would help the taxpayer to avoid becoming liable to further penalties for careless inaccuracy. There will be no further VAT returns and therefore no possibility of inaccuracy or careless inaccuracy.

The appeal against the penalties is allowed in part.

As set out above:

As regards the appeals which remained in issue before us:

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.