James Keighley & Anor v The Commissioners for HMRC

The appellant accepts that he used his company credit card to fund personal expenditure. He accepts that this is the case between 2012 and 2017, and that in this regard, his behaviour has been careless.

HMRC however think differently. They think that he has used his company credit card to fund personal expenditure since 2001. In October 2019 they issued discovery assessments for £177,043. In November 2019 HMRC also issued penalty assessments based on deliberate behaviour for £59,912.

In calculating the additional income tax, HMRC have taken the personal expenditure for the years 2012 to 2017, and extrapolated backwards using RPI, under the principle of the presumption of continuity.

There are, therefore, two primary issues which we need to determine. The first is whether the behaviour which resulted in the underassessment of income tax on the appellant is deliberate behaviour or careless behaviour. The significance, of course, is that if it is the former the extended 20 year time period within which HMRC can issue a discovery assessment comes into play. If the behaviour was only careless, the time period is limited to 6 years following the end of the year of assessment to which it relates.

If we decide that the behaviour was deliberate, then we need to consider the second issue, namely whether the presumption of continuity applies at all and if it does, whether it has been properly applied in the circumstances of these discovery assessments.

The appellant also raises a subsidiary issue, namely that as a result of the compliance check in 2011/2012, which effectively gave the first appellant and the company a clean bill of health regarding the former’s use of his credit card, HMRC are effectively estopped from issuing their October 2019 discovery assessments under the provisions of SP 8/91.

The parties agree that the burden of proving that the discovery assessments were valid in time assessments rests with HMRC, who therefore have the burden of establishing deliberate behaviour. They also have the burden of establishing that as far as the penalty is concerned, there was an inaccuracy attributable to deliberate (or in default careless) behaviour. The standard of proof is the civil standard, namely the balance of probabilities.

Given that the appellant has admitted careless behaviour, and that the October 2019 discovery assessments were raised on 14 October 2019, it is HMRC’s view that they must therefore be valid for all periods from 5 April 2014 to 5 April 2017.

The law on this issue is clear and is not in dispute. To the extent that the credit card has been used for unreimbursed personal expenditure, that expenditure is treated as earnings and is thus liable to income tax and NICs.

Where HMRC discover that income which ought to have been assessed has not been so assessed, or that there has been an insufficiency in an assessment to income tax, they may make an assessment to make good that tax or that insufficiency. The general time period within which such a discovery assessment must be made is not more than 4 years after the end of the year of assessment to which it relates. This 4 year period is extended to 6 years where the loss of tax was brought about carelessly by the taxpayer, and to 20 years where the loss of tax is brought about deliberately by the person.

The meaning of “deliberate” in the context of direct tax returns was considered by the Supreme Court in its decision in Tooth v. HMRC [2021] UKSC 17. The court drew a distinction between (i) a deliberate statement which is (in fact) inaccurate or (ii) a statement which, when made, was deliberately inaccurate. At [43] the court stated that the second of those interpretations is to be preferred:

At [45] the court considered that the meaning of “deliberate” in relation to “discovery assessments” had the same meaning as used in relation with penalties:

The court summarised its conclusions at [47]:

We were also referred to Chohan Management Limited v HMRC [2021] UKFTT 196 (TC) at [112] to [113]:

On the deliberate behaviour issue, Mr Marks submitted, in summary, as follows:

The first appellant has admitted that he used his company credit card to pay for personal expenditure. This was not just the smaller items, but for family holidays, flights, hotels and meals. He made no reimbursement to the company to cover this private expenditure, nor did he make any return of the benefits himself.

His claim that he thought that the expenses had been included on his form P11D cannot be right given that he did not tell those who were responsible for completing the form, what his personal expenditure was.

He was not forthcoming during the enquiry about this, since he initially admitted personal expenditure of a small amount of £9,000 or so in 2015 which then increased to over £113,000 from 2012 to 2017.

He would have known what personal expenditure he incurred. And would therefore have known that the £9,000 or so in 2015 was not the full extent of his personal expenditure between 2012 and 2017.

This is a substantial sum. And indeed, in the 2016 tax year, the £38,000 or so personal expenses was more than he had earned as salary in that year. He must, therefore, have known that his tax returns underdeclared the personal benefit he received from the company.

There was no evidence that there was a system in place which obliged the first appellant to reimburse the company for the personal expenditure.

There is no evidence of the first appellant completing any expenses forms. He provided no evidence that he submitted any schedules or details to the company to conduct the coding out exercise which he says he thought was being carried out. In HMRC’s view, that only started in 2017, and given that the first appellant did not tell the individuals who were responsible for this coding out exercise, what the nature of the personal expenditure was, it would have been impossible for that exercise to have been carried out accurately. And the appellant knew this.

No one at the company checked his expenses or signed them off. He simply used the credit card when he liked, and the company paid the monthly payments. The first appellant knew that he was using his credit card for personal expenses and knew that he was not accounting to the company for them. He also knew that he was not returning these personal expenses on his self-assessment tax return. This is deliberate behaviour.

The 2011/2012 compliance check was into the company’s position and not into the position of the first appellant. It focussed on a number of things, not just the use of company credit cards. It was not, as submitted by the first appellant, any recognition that the way in which the first appellant and the company were dealing with the first appellant’s private expenses was an acceptable one.

On the deliberate behaviour issue Mr Smith submitted, in summary, as follows:

The company is a small family business. The first appellant is a creature of habit. Karen, who was responsible for dealing with the first appellant’s credit card expenditure would therefore have known of the first appellant’s private spending without the need to be specifically told, either orally or by some form of schedule.

The employer compliance review clearly included a review of the use of the company credit card by the directors. And gave the process undertaken by the company in 2011/2012, a clean bill of health. It was not unreasonable therefore that the process adopted before 2012 was maintained thereafter.

However, in 2012 a number of changes took place which led to the errors which occurred between 2012 and 2017. Most significantly, personnel who had dealt with the relevant administration before then (in particular Karen who had been with the company for many years) left. Following that, and the loss of her experience, things “slipped” with the result that errors were made to which the first appellant has agreed were careless on his part (he accepts that he should have checked the process on a periodic basis).

But it was entirely reasonable in the circumstances that the first appellant thought that there was a robust process in place to deal with his personal credit card expenditure which he thought had been confirmed by HMRC. Whilst it might have been misguided, he was not cavalier, nor “not bothered with establishing the correct position”. He thought he was entitled to rely on the process in place.

In order to establish that the inaccuracy in the first appellant’s returns was deliberate inaccuracy, HMRC have to show an intention on the part of the first appellant to mislead HMRC as to the truth or accuracy of the information set out in his tax return. And, in our view, this extends not just the actual knowledge regarding the truth or accuracy of information, but also where the taxpayer has blind eye knowledge as defined above.

It is our view that the first appellant had actual if not blind eye knowledge that the tax returns did not accurately record the additional remuneration arising from his use of the company credit card to satisfy personal expenditure for the period 2013-2017. We say this for a number of reasons.

It is abundantly clear that the appellant made extensive use of his company credit card to settle personal expenditure. It is clear on the evidence of the figures supplied by the first appellant’s agent for the period 2013-2017. Indeed, the first appellant accepts this, and accepts that the additional remuneration reflected by this personal expenditure was not included on his tax returns. But says that this behaviour was careless.

There is no doubt that there was an inaccuracy in the tax returns.

The level of personal expenditure, however, was extremely high both in absolute and relative terms. The first appellant does not deny that in the 2015-2016 tax year, his personal expenditure exceeded his salary. We were told that the first appellant’s tax return was drafted by the company and the company’s accountants who then returned it to the first appellant for checking and signature.

The first appellant’s evidence was that he thought that the personal expenditure was being “coded out”, by which we think he means that there was a change to his tax code so that the amount of tax that should have been collected on his personal expenditure was being collected through the PAYE system.

But we do not accept this evidence. If this was the case, then this would have been apparent to the first appellant when checking his tax return. And he would quickly have seen that no such tax credit had been given for the tax paid under the PAYE system against the tax which he had to report on the “personal expenditure remuneration” in his tax return.

And we think it unlikely that, given the level of personal expenditure, the first appellant did not notice that it had not been included in his tax return when he was checking its accuracy. We find it difficult to believe that in a year where the personal expenditure exceeded his salary, the appellant did not notice, let alone question, why the tax payable in that year did not include tax payable on the personal expenditure.

In our view the reason is that the appellant found it convenient not to do so. Had he questioned it, he would have uncovered an uncomfortable truth, namely that he was liable to pay a considerable amount of tax on the personal expenditure. It is our view that he decided to keep his head down.

There was a suggestion that there was a system in place to ensure coding out, and that system which was in place before 2012 broke down as a result of the changes of personnel in or around that time. Somewhat confusingly, the first appellant’s evidence was also that there needed to be no system in place before 2012 since he was a creature of habit and Karen would have known about his personal expenditure without needing input from him.

On this latter point we are unconvinced. We think it inherently unlikely that Karen would have known the full extent of the personal expenditure which the first appellant incurred. And he would have needed to provide considerable input into the nature of that expenditure.

But more importantly the first appellant’s evidence was notwithstanding that he knew of the changes in personnel in 2012, he simply carried on behaving towards the reporting of his personal expenditure, as he had before 2012. He did not check whether the coding out which he alleged he thought was happening before 2012, continued thereafter. He never questioned why he was not being asked to verify private expenditure which he (again somewhat confusingly) claims he was asked to verify prior to 2012.

This, again, suggests to us that the first appellant knew that he was required to report the private expenditure which he had incurred on his company credit card to HMRC, and yet knew, too, that there was no system in place to identify private expenditure and to report it to HMRC. And he never sought to question the system because it was convenient for him not to do so. This was blind eye knowledge and, in our view, resulted in inaccurate tax returns which were intended to mislead HMRC.

No corroborating oral or documentary evidence has been provided to support the first appellant’s assertion that he completed schedules, either before or after 2012 (but before 2017) which identified personal and business expenditure on his credit card. We find as a fact that no such schedules were completed between 2012 and 2017.

We do not accept, either, that because the first appellant thought that the employer compliance enquiry in 2011/2012 had given the way in which he and the company accounted for tax for personal expenditure incurred on its company credit card, a “clean bill of health”, he and the company could operate in the same way after 2012 with impunity.

This flies in the face of the first appellant’s recognition that the system changed in 2012 with the departure of key personnel, and so the analysis which he alleges was being carried out before 2012 was not being carried out after 2012 (to his knowledge). Yet he did not question why he was not being asked to clarify private expenditure. To our mind he simply kept his head down as it was expedient for him to do so.

There was absolutely nothing wrong with the first appellant using his company credit card to pay personal expenses. And this is the case even if there was no formal procedure in place either to provide prior approval for those expenses, or for reimbursement to the company once they had been incurred. But in these circumstances, it was incumbent on the first appellant, given it was only he who knew the extensive level of that personal expenditure, to ensure that, in the absence of reimbursement, the expenditure was identified to HMRC and tax paid on it. For the foregoing reasons, it is our view that the first appellant’s failure to ensure that the personal expenditure was properly and accurately reported arises from deliberate behaviour on his part.

The lead case regarding the presumption of continuity is Jonas v Bamford [1973] STC 519 in which the principle is described as follows:

“… Once the Inspector comes to the conclusion that, on the facts which he has discovered, [the appellant] has additional income beyond that which he has so far declared to the Inspector, then the usual presumption of continuity will apply. The situation will be presumed to go on until there is some change in the situation, the onus of proof of which is clearly upon the taxpayer”.

On the presumption of continuity, Mr Marks submitted, in summary, as follows:

Officer Heatley’s approach was entirely proper. She decided to use a 2014 error as the base error and scale back using RPI. This was a reasonable thing for her to do given that the appellant had admitted that he used the company credit card to pay for personal expenses before 2012, and there was no evidence to suggest that the errors in 2001 to 2011 were either absent or of a lower amount.

It is inherently unlikely that someone would go from having zero personal expenditure in 2012. Yet this is what the first appellant appears to be arguing on the basis that prior to 2012, the company could not afford it.

This position is apparently inconsistent with his assertions that prior to 2012 he was using his company credit card for personal expenditure, but the 2011/2012 compliance check gave him a clean bill of health. And that the process that had been used before then, which involved Karen, was that she would code out the private expenses, return them on his P11D, and in some way that would satisfy his reporting requirement.

The appellant has made little or no effort to ascertain the level of private spending on his credit card prior to 2013/2014. His evidence was that he had not attempted to get credit card statements before 2011.

Only the first appellant knows the true extent of his personal expenditure. HMRC suggest that it is more likely than not that the RPI adjustment on which the assessments are based, is correct.

And indeed, given that the first appellant has submitted no alternative figures, if we are to find that the presumption of continuity applies, then we are obliged to uphold the assessments in the absence of any better evidence of actual private expenditure submitted by the first appellant.

The onus is on the appellant to show a break in continuity.

The 2011/2012 employer compliance enquiry, and the conclusion of it does not provide such a break. Even though credit card statements might have been submitted to that enquiry there is no evidence that the investigating officer concluded that there was no personal expenditure and “signed it off”. Furthermore, there was also an admitted error in 2012 of over £10,000, which was whilst the compliance enquiry was ongoing and had not yet concluded. This militates against the fact that the compliance enquiry did, as asserted by the first appellant, give him a clean bill of health.

There was no break in continuity given the change of personnel which occurred in 2012. Until 2012, Karen dealt with the expenditure on the company credit cards. When she left no one filled the void. There was no evidence of this change of personnel, nor that there was an experienced team in place prior to 2012 and not thereafter.

The fact that there was a lower credit limit on the card prior to 2011 is a red herring. It does not mean a reduced ability to spend money. This credit card limit does not break continuity. Nor does the changing financial position of the company. Administrative expenses are remarkably consistent from 2001 (£2.7 million) to 2016 (£2.7 million). The financial position of the company does not break continuity.

On the presumption of continuity, Mr Smith submitted, in summary, as follows:

HMRC’s submission that the first appellant incurred no personal expenditure in 2012 is inherently unlikely, is misconceived. In 2012 the first appellant thought that the personal expenditure that he had incurred using his company credit card, was being coded out.

The first appellant made no admission at the meeting of 22 February 2019 that, as asserted by HMRC, he had used the company credit card for personal expenditure for the years 5 April 2001 to 2013. There is no such record of that admission in the notes of that meeting.

There is clear evidence that the situation that led the first appellant to make disclosures of personal expenditure on the company credit card had a definite starting point which is in respect of the year ended 31 December 2012.

The employer compliance enquiry which concluded a year-long investigation into the company’s employer compliance record had specifically looked at the issue of company credit card operation and gave the process around the operation of the company credit card a clean bill of health in that no adjustments were required in respect of it.

Furthermore, if HMRC are now saying that notwithstanding that the investigating officer had the statements, and she did not properly analyse them, this is not reflected in the 2012 meeting notes. They show that the enquiry lasted for 11 months; all systems for PAYE compliance were checked; all benefits in kind were properly reported on P11D; she specifically wanted to consider travel subsistence and entertaining.

By 2012 the credit limit on the company credit card had been increased which meant that the card had the capacity to absorb personal expenditure whereas before then any personal expenditure would have been to the detriment of the company (if the first appellant had used the card for personal expenditure, it may not have had the capacity to meet business expenditure).

The main issue which led to the inaccuracies occurring after 2012 was the change in personnel within the company’s finance team which took place in 2012.

The administrative expenses and gross profit were not remarkably consistent as asserted by HMRC and indeed those items shed little or no light on whether there has been a break in continuity.

HMRC have made no use of their third party information notice powers to seek to obtain details from the credit card companies of the first appellant’s expenditure on his company credit card.

Due to the foregoing changes, if the presumption of continuity does apply, then it ends in 2011/2012 when the changes described took place and the employer compliance enquiry ended.

The methodology adopted by Officer Heatley when making the October 2019 discovery assessments, namely taking the figures for personal expenditure provided by the first appellant’s agent, and extrapolating backwards, using RPI for periods prior to 2013, is a reasonable methodology and we find that the October 2019 discovery assessments are best judgment assessments.

The burden of establishing, therefore, that these assessments overcharge the first appellant, falls upon the first appellant. He must show that it is more likely than not that the assessments overcharge him.

Mr Smith makes a number of submissions suggesting that a clear line in the sand can be drawn in 2012, which makes it improper to use the presumption of continuity to assess personal expenditure prior to that date.

Prior to that date, the first appellant thought that his personal expenditure was being coded out. But, with the greatest of respect, this does not mean that the appellant was not using his credit card for personal expenditure prior to 2012. The evidence of the fact that there was some form of system in place, provided by the first appellant, demonstrates that he was using his credit card for personal expenses prior to 2012.

That there was this use prior to 2012 is bolstered by the submissions that the 2011/2012 employer compliance review had considered the system adopted by the company towards personal expenditure on company credit cards not just for the first appellant but for others in the company and given it a “clean bill of health”. And indeed, this was relied upon by the first appellant to counter allegations of dishonesty. But it clearly reflects an acceptance that the first appellant was using his credit card for personal expenditure prior to 2012.

We accept Mr Marks’ submission that it is inherently unlikely that having incurred personal expenditure of £25,000 or so in 2013/2014 on his company credit card, the first appellant had not used it for personal expenditure at all prior to that tax year. We think it is much more likely, and we find as a fact, that the first appellant did use his company credit card for business expenditure prior to that tax year and indeed going back to 2001.

Mr Smith submits that the main issue which led to the inaccuracies after 2012 was a change in personnel within the company’s finance team which took place in 2012. And that in some way this meant that the presumption of continuity could not be used to go back before 2012 as a method of calculating personal expenditure for years in which no figures have been provided. But for similar reasons to those enunciated above, we do not see why this breaks the continuity between 2013/2014 and 2001. There is nothing in this submission which suggests that the company credit card was not being used for business expenditure, indeed the contrary; it supports HMRC’s assertion that it was being so used, but there was simply a change of internal systems for reporting that personal use.

But the better point which we think Mr Smith makes is that the level of expenditure in 2013/2014 and the following years was considerably greater than that in earlier years, and this greater expenditure in some way arises from the change of personnel in 2012, and RPI therefore is an inappropriate measure with which to extrapolating backwards.

We do not accept this submission. We cannot see why the change of personnel has any bearing on the level of personal expenditure. It simply deals with the reporting of that expenditure. Once an agreed figure for 2013/2014 was established at £25,904, by Offcer Heatley, the fact that this might have resulted from a change in reporting requirements after 2012 due to the change in personnel does not, in any way, affect the application of the presumption of continuity. It does not provide a break. The change in personnel does not affect the figure.

Mr Smith also submitted that by 2012, the credit limit on the company credit card used by the first appellant had increased, and thus could absorb greater personal expenditure thereafter than was the case before. But no corroborating evidence was provided to support this assertion.

We do not consider whether the fact of an admission or otherwise regarding the use of the card for private expenditure at one of the meetings between the first appellant and HMRC, as being of any probative value.

The first appellant has led no evidence of what, in his view, is the more likely level of private expenditure which he incurred on his card between 2001 and 2013.

As we have said already, it is for the first appellant to provide evidence that the figures in the October 2019 discovery assessments overcharge him and evidence of what the more likely figures are. He has not done this.

The first appellant invites us to conclude that because HMRC decided not to make any adjustments in respect of the company credit card following the 2011/2012 employer compliance check (which was based on full facts in relation to the use of the company’s credit cards) HMRC are precluded from issuing the October 2019 discovery assessments in relation to periods prior to 2011/2012 on the basis of the provisions of SP 8/91.

SP 8/91, in the first appellant’s view, applies as HMRC have stated therein that they would not issue discovery assessments in cases where the particular issue has already been the subject of an enquiry like the 2011/2012 employer compliance enquiry.

In this regard paragraph 14 of SP 8/91 is relevant.

“There will also be circumstances in which the Cenlon and Olin principles are not applicable. Thus, the particular point on which the inspector subsequently takes a revised view and considers making a discovery assessment may not have been the subject of a specific agreement or, because the point was not fundamental, cannot be said to have been the subject of a specific agreement (see para 8 above). In these circumstances, a discovery assessment will not be made, provided that the inspector’s original decision, whether on a claim or on the proper amount of an assessment, was based on a full and accurate disclosure of all the relevant facts and was a tenable view, so that the taxpayer could reasonably have believed that the inspector’s decision was correct. And it follows that if the inspector’s original decision was consistent with a view of the law and practice generally received or adopted at the time, a discovery assessment would not be made where, for example, there is a subsequent change in that practice - eg, following a court decision”.

It is the first appellant’s submission that the use of company credit cards was a central tenet of the 2011/2012 employer compliance enquiry. Full information regarding that use was disclosed by the company and thus all available information was made available to HMRC at that time. This is reflected in the 2012 meeting notes as well as the information request.

And so HMRC are effectively estopped from issuing the October 2019 discovery assessments since the first and second appellants were both given a clean bill of health in relation to private expenditure on the company credit card in 2011/2012. And this applied to all years prior to and including 2011/2012.

HMRC’s view is that to the extent that SP 8/91 is relevant, it does not apply since the personal use of the appellant’s company credit card was not a central tenet of the employer compliance enquiry; and indeed, that investigation was into the company’s position and not into that of the first appellant.

Whilst, as we have mentioned, the 2011/2012 compliance check is relevant in respect of the first appellant’s subjective mental state as regards deliberate behaviour, we do not think that it operates as suggested by Mr Smith, to prevent HMRC issuing the October 2019 discovery assessments.

Firstly, we do not think we have jurisdiction to consider this point in the first place. Our view is that if the first appellant is going to allege non-compliance with a statement of practice, that should be based on a legitimate expectation that HMRC would conform to the provisions of that statement of practice and their failure to do so can only be impugned by way of judicial review. We have no such jurisdiction.

Secondly, SP 8/91 does not apply cases where there may have been fraud or negligence by or on behalf of the taxpayer. Mr Smith submits that that must be for negligence in respect of the 2011/2012 compliance enquiry. We disagree. Our view is that if a taxpayer wishes to rely on a statement of practice, they must have clean hands which they do not have if they have been guilty of fraud or negligence. That is the case whether it relates to the original enquiry or the subsequent discovery assessments.

Thirdly, paragraph 4 of SP 8/91 makes it clear that for individuals, it only applies to returns for years up to and including 1995/96. The October 2019 discovery assessments clearly relate to returns for periods which post date this.

Finally, we are not sufficiently certain that a review of the appellant’s personal use of his company credit card was the subject of a decision following the 2011/2012 compliance review, so as to fall within the ambit of SP 8/91. There is simply insufficient evidence to satisfy us that that was the case.

So, our conclusion on this issue is that HMRC are not prevented, by dint of the application SP 8/91, from issuing the October 2019 discovery assessments.

For the foregoing reasons, we have concluded that the inaccuracies in the first appellant’s tax returns which underreported his personal expenditure have arisen from his deliberate behaviour. We have also concluded that the presumption of continuity is an appropriate method for calculating the level of personal expenditure prior to the tax year 2013/2014, and the first appellant has not persuaded us that there was a break in continuity, nor of what that level of personal expenditure is more likely to have been. We have also found that the application of SP 8/91 did not prevent HMRC from issuing the October 2019 discovery assessments.

In light of the foregoing, we also uphold the November 2019 penalty assessments visited on the first appellant.

Similarly, we also uphold HMRC's section 8 decision dated 6 November 2019 for additional NICs of £63,919. However, as regards the NIC penalties set out in the decision of 25 November 2019 we take a different view.

Mr Marks accepts, in his skeleton argument, that it is possible for us to find that the first appellant acted deliberately but that the company acted carelessly. And that is our finding. This does not affect any limitation regarding the collection of the NICs, but does have an effect on the relevant penalty level.

He urges us to find that the deliberate behaviour of the first appellant "transfers to the company". But we do not find this. This is not a one man company where the guiding mind of the company can, effectively, be directly attributable to the owner manager. It is a substantial organisation with many employees and a number of systems, one of which involves the analysis and categorisaton of private and business expenditure. It is clear to us that those systems fell into slight disarray in 2012. And that the failure, by the company, to correctly categorise the credit card expenditure and thus to correctly record the private element as additional remuneration (and thus pay NICs on it) was careless.

But to establish deliberate behaviour it is for HMRC to show an intention to mislead HMRC. Or to show that the company deliberately shut its eyes to the fact that the personal expenditure might have been additional remuneration.

And whilst, as we have said above, it is our view that this was the case for the first appellant, HMRC have not shown to our satisfaction that it was the case for the company. It is not sufficient to simply say that the sins of the first appellant must be visited on the company.

The NIC penalties, therefore, should be recalculated on the basis of careless behaviour. If there is any dispute about the figure which is then assessed, then either party shall have the right to bring the matter before the tribunal who will determine the issue.

There are a number of issues which need to be considered under this broad heading.

Firstly, in respect of both accounting periods ending 2014 and 2015, whether the loan relationship debits arising from the write-off of part of the debt owed to the company by VDP were allowable deductions for corporation tax purposes for the company.

HMRC say they are not on the alternative bases that; the company and VDP were connected companies in those accounting periods (or some part thereof) and/or the debits had an unallowable purpose (together the “technical issues”).

Secondly, in respect of the accounting period ending 2014 which reflects HMRC’s foregoing position, whether the October 2020 discovery assessment is valid.

The appellant says not as there was no careless behaviour by the company or by somebody acting on its behalf. So, it was invalid and out of time.

Furthermore, the revenue amendment was, in essence, a discovery assessment which was settled by a section 54 agreement. By issuing the October 2020 discovery assessment, HMRC are seeking to relitigate a point which has already been settled (the “discovery assessment issues”).

Thirdly, penalties.

The parties agree that provided HMRC can establish that the October 2020 discovery assessment is valid, then it is for the company to establish that it was not connected to VDP, and/or the loan write-off had no unallowable purpose.

We have set out the relevant law in the appendix to this decision.

There are two technical issues. Firstly, whether the company and VDP were connected at any time during the accounting period in which the loan from the company to VDP was partially written off by the company (the “connected company issue”). Secondly whether the loan write-off had an unallowable purpose (the “unallowable purpose issue”).

Although we have set out the law in the appendix, it is worth providing a brief synopsis of how the parties agree that it operates in relation to the technical issues.

The loan relationship provisions in the CTA 2009 aligns the tax treatment of debt write-offs or impairments with their accounting treatment (on the basis that the proper accounting treatment has been adopted). Where the lender releases an unconnected borrower from all or part of a debt, the lender is able to take a tax-deductible debit against its corporation tax liability, but the borrower must account for a taxable credit.

However, where the lender and borrower are connected at some time during the accounting period in which loan is impaired or written off, the foregoing symmetry is disregarded, and whilst the borrower has no need to bring into account a taxable credit, the lender benefits from no tax-deductible debit.

In this case, it is HMRC’s assertion that the company and VDP were connected at the relevant time, and so the tax-deductible debit which the company has claimed should be written back. The company says that it was not connected with VDP and that the tax deduction was properly claimed.

If a loan relationship has an unallowable purpose, then, to all intents and purposes, the same situation arises as in a connected company loan relationship. The release of a debt is tax neutral. The borrower takes no debit, the lender accounts for no credit. In this appeal HMRC assert that the release of the loan was a related transaction and thus the foregoing treatment applies. The company denies that it was a related transaction and asserts that the company was entitled to the tax-deductible debit.

On this issue, Mr Marks submitted, in summary, as follows:

The loan was secured against the property owned by VDP. The loan agreement provided that the loan must be repaid in full on the sale of the property.

If the terms of this agreement had been adhered to, the company would have been repaid the loan in full and the amount available for distribution amongst the other lenders would have only been £236,000.

The company also lent a further £133,967 to VDP as a “trading balance”.

Once VDP sold its property, the first appellant and BM were repaid in full, Mr Clayton was paid £121,479 (leaving a balance of £37,213). Mr Murphy was paid £43,245 (leaving a balance of £14,884); £365,790 was repaid to the company, leaving a balance of £244,626, (£110,659 of which related to the loan, and the rest being the trading balance).

There was a clear decision made by VDP to only repay the company part of its loan, whilst at the same time repaying BM and the first appellant in full.

The company and VDP were connected since BM and the first appellant controlled both of them. They were controlled by the same persons.

Control here means the power of a person to secure that the affairs of a company are conducted in accordance with the person’s wishes. Here, BM and the first appellant could secure that the affairs of both companies were conducted in accordance with their wishes since they held the majority of the shares in both companies.

Affairs here does not mean all of the affairs of the company. It can apply to specific affairs, and the release of the balance of the loan falls within the ambit of this provision.

In actual fact, notwithstanding that BM and the first appellant had control by dint of their majority shareholdings, they acted together to ensure that, in regard to the loan repayments, the company and VDP agreed to the repayment schedule. This could not have happened if they did not exercise de facto control over both entities.

The company gave up its valuable rights to repayment of the balance of the loan and to repayment of the trading balance. This was to its detriment. The repayment schedule benefited BM and the first appellant. The decisions, therefore, made by BM and the first appellant had a considerable personal benefit and cannot be said to be independent decisions. Their wishes were to be repaid in full at the expense of other creditors including the company.

On this issue Mr Smith submitted in summary as follows:

Neither theoretically nor actually could BM and the first appellant secure that the affairs of both companies were conducted in accordance with their wishes.

Affairs here means every single one of the affairs of the company not just some or most of them.

For two people to control the company, they have to be shown as acting as “one” in concert or acting with one mind. The question therefore is whether the first appellant and BM acted as one unit at all times in all respects to control the company and VDP. This interpretation has been endorsed by Counsel.

The evidence shows that they clearly did not. The shareholders agreement provides that Majority Consent (90%) was required for 19 different actions of the company. BM and the first appellant therefore could not undertake these actions without the consent of SF. This is sufficient to disconnect the company from VDP regardless of the control of VDP.

There are also specific restrictions in clause 11 of the shareholders agreement which explicitly prevent BM and the first appellant from undertaking certain actions, for example making loans, transacting with associated companies, and disposal of assets over a certain value. These provisions again militate against BM and the first appellant be able to secure that the affairs of the company were conducted in accordance with their wishes.

As far as VDP is concerned, the first appellant and BM needed to operate together in order to control VDP, and the evidence shows that they did not. Whilst they acted together to (operationally) sell the property, the decision to sell was taken by all shareholders. Similarly, the decision to distribute the proceeds of sale was a decision taken by all shareholders.

Neither the company nor VDP were controlled by BM or the first appellant acting as one as required by the legislation. The test is whether they acted at all times as one in accordance with HMRC guidance.

Even though the company had security over the property, when the property was sold the security provided went with it. So, the company did not choose to abandon its position as a secured creditor. As a matter of fact, when the property was sold, the security was released as well.

It was the first appellant’s unchallenged evidence that it was SF who agreed to give up the company’s right of full repayment of the loan and the trading balance. This was in any event required by virtue of it being a transaction with an associated company of a shareholder.

There is no dispute about the tax consequences of the company and VDP being connected. Nor is there any difference between the parties as to the legislation which we should apply in determining whether they are so connected. If BM and the first appellant controlled both companies at any time in the accounting period in which the loan was partially released, then the company is not entitled to the tax-deductible debit for that partial release.

Under section 472 CTA 2009, control means the power of a person to secure that the affairs of the company are conducted in accordance with the person’s wishes by means of the holding of shares (or the possession of voting power) or as a result of any other powers conferred by the articles or other document regulating the company.

Interestingly, there is no scope within this provision to consider de facto control. The power to secure the affairs of the company must be by means of shareholdings or other documentary provision.

It is our view that whilst BM and the first appellant had control of VDP, it did not control the company, and thus the company and VDP were not connected at the relevant time.

Mr Smith makes two submissions. Firstly, that in order to control the company, a person must be able to secure that every single one of the affairs of the company are conducted in accordance with that person’s wishes.

Secondly, where two people are alleged to control the company, they must act “as one”.

On the first point, we do not agree with him that, as a general principle, if a single affair of the company is not conducted in accordance with the person’s wishes, that person has no control. It would seem odd to us that if there was a provision in a shareholders agreement which obliged the majority shareholders to take into account the wishes of a minority shareholder if the company wants to buy more than 50 paperclips in one batch, then that would prevent those majority shareholders from controlling the company. To our mind when considering the affairs of the company, one needs to look more deeply into the affairs of the company.

There must be a consideration in each particular case, which is bound to be fact dependent, as to whether the specific activities of the company in question are conducted in accordance with a person’s wishes. And, without making any hard and fast rule (which we do not need to do in this case) the relevant affairs and activities of the company which need to be considered will vary on a case-by-case basis.

Nor do we agree with him that the statutory test can be supplemented by a rule of thumb that the persons need to act “as one”. We need to consider the statutory test. We ask ourselves whether the rights attaching to the shares, the articles, or the shareholders agreement (which in our view falls within the ambit of another document regulating the company) give BM and the first appellant the power to secure that the affairs of the company are conducted in accordance with their wishes.

We also accept the argument that person for these purposes can be persons by dint of the Interpretation Act.

It is our view that the provisions of clause 11 of the shareholders agreement, which contains a raft of restrictions on the otherwise unfettered right of the majority shareholders to control the company, mean that they do not control the company. The clause 11 actions require the consent of SF. These include changing the constitution the company or its name, issuing debt instruments, forming subsidiaries, merging or winding up the business, changing the nature of the business carried on by the company, making loans, permitting transactions with associated companies or shareholders, and other similar provisions.

These matters are fundamental to the running of the company at both an operational and strategic level. They are clearly within the ambit of the affairs of the company. Since the majority shareholders do not have the power to secure that these are conducted in accordance with their wishes, it is our view that they do not have control of the company.

Irrespective, therefore, of control of VDP, the company and VDP were not connected at the relevant time.

On this issue Mr Marks submitted, in summary, as follows:

The unallowable purpose legislation applies not just to the making of the loan but also to any related transactions which means any disposal or acquisition (in whole or in part) of rights or liabilities under the loan relationship. Disposal and acquisition include circumstances where rights or liabilities under a loan relationship are extinguished by any redemption surrender or release.

The loan was such a loan relationship and therefore the write-off of part of the loan by the company was a related transaction.

The decision by the company not to enforce its security, and its rights to be repaid the loan in full, but instead allowing VDP to repay BM and the first appellant in full were not in the commercial or business interests of the company. The company could have insisted on having its secured loan repaid in full, and VDP would then have had to decide who amongst the unsecured creditors would be repaid.

It also gave up rights to its unsecured trading balance, in respect of which it received no repayment. Had the loan been repaid in full as it should have been, there would have been approximately £236,000 left over in VDP to repay unsecured creditors which would have entitled the company to be repaid 38% of its trade balance (£51,057). The company therefore should have been repaid a total of £527,057. However, it was repaid only £366,000, which is approximately £161,000 less than it was entitled to. This nearly matches the amount paid to BM and the first appellant.

The decision of the company not to enforce what was legally due from VDP following the sale of the property by VDP was not in the business or commercial interests of the company. The company therefore cannot bring into account the debit on the impairment loss as the unallowable purpose legislation prohibits it from doing so.

On this issue Mr Smith submitted, in summary, as follows:

The unallowable purposes legislation is an anti-avoidance provision designed to stop companies artificially structuring their financial arrangements. In the case of the company and VDP, the company was a reluctant financier but had no choice other than to lend the money to VDP if it wanted VDP to survive. This is not contrived intragroup financing. The arrangements between the company and VDP bear no hallmarks of avoidance.

It was Mr Sengupta’s evidence that the unallowable purpose legislation was squarely aimed at convoluted structures put in place in an attempt to circumvent the normal scheme of things when interest is paid by one company which receives a tax deduction, but the recipient is not taxed on receipt of the interest.

The question in this case is whether the loan made by the company to VDP was made for the commercial and business purposes of the company. The answer is clearly yes on the evidence of the first appellant. Furthermore, the loan enabled VDP to survive and prevented the bank taking enforcement action, and the company’s guarantee being called on.

HMRC’s interpretation of the consequences of being a secured creditor are misconceived. Once the secured assets were sold, the lender then ceased to be a secured creditor and became an unsecured creditor, something which has been confirmed by the liquidator’s report of VDP. Once the property was sold, there was no security, so the company was not entitled to full repayment of the loan as HMRC assert.

The company’s decision to write-off part of the loan was taken by all three shareholders.

Our view is that the write-off of part of the loan by the company had an unallowable purpose. It falls within the relevant statutory provisions.

Firstly, the loan from the company to VDP was a loan relationship.

Secondly, the partial write-off of the loan was a related transaction within the definition of section 304(1) CTA 2009. It was a disposal by the company of rights under the loan agreement (the right to be repaid) and falls within the ambit of section 304(2) since the company’s rights under that loan relationship were extinguished, in part, by the release of VDP’s obligation to discharge the loan in full.

Finally, the purpose of the partial write-off was not amongst the business or other commercial purposes of the company.

Mr Smith’s submissions were focused, mainly, on the justification for the making of the loan in the first place. We have not permitted HMRC to make any submissions on the issue of whether the making of the loan had an unallowable purpose. And so, these submissions are not, strictly speaking, relevant to the write-off.

The evidence is clear that the loan was secured on the property. It is equally clear that VDP had an obligation under the loan agreement to discharge the loan in full on the sale of “substantially all of the assets” of VDP. The sale of the property falls within this provision.

We do not accept the submission that at the point of sale of the property by VDP, the company had released that security, and was thus in the position of unsecured creditor. And so was no longer in a favourable position when it came to its rights to be repaid in full. This is simply not how security works in the circumstances. The purpose of the security was to ensure that the contractual promise to repay the loan in full on the sale of the property was met.

And security in this case was a legal mortgage which would, in the absence of release, run with the property. In other words, a buyer would be bound by the obligations under that mortgage if it purchased the property. And of course, no properly advised buyer would do this. The buyer would require such security to be released immediately prior to sale.

The way this works in practice (where the mortgage needs to be discharged out of the sale proceeds as was the case here, and as was the case with the bank’s security) is that a solicitor acting for the seller gives an undertaking to the buyer’s solicitor that on receipt of the purchase price, the buyer’s solicitor will use the money to discharge the debts owed to the secured creditors. Secured creditors in turn then release the charge so that the property can be conveyed, unencumbered, to the purchaser. And in these circumstances the secured creditors are paid what they are owed.

So, the company as a secured creditor was entitled to insist on repayment in full of the outstanding loan on the sale of the property. Without that, there was no obligation on the company to release its security.

However, it agreed that it would not insist on receiving the full amount due to it, and this is a release, by the company, of its right to full repayment under the terms of the loan agreement.

By doing this it clearly deprived itself of a considerable amount of money. Our understanding is that of the £476,000 secured on the property, only approximately £381,000 was repaid, the balance being written off.

In our view this was not amongst the business or other commercial purposes of the company.

The evidence shows that the company suffered financially, whilst the majority shareholders benefited (as unsecured creditors they should not have been entitled to 100% of their unsecured loans given the shortfall following the sale of the property by VDP). The evidence also shows that the reason why the majority shareholders were repaid in full is because of the work that they had done to secure a good deal for the sale of the property which resulted in a knock-on benefit for everyone.

We understand this, and this seems to us proper that the majority shareholders should receive some sort of reward for that result. But that reward has benefited those shareholders at the expense of the company. And when considering unallowable purpose, we must consider the purposes of the company and not the company’s shareholders or investors.

We do not need the first appellant’s admission that by writing off the balance of the loan, the company suffered financially. That is clear as a matter of fact. The core business of the company is selling imported doormats. It took the view that it needed to fund VDP to enable VDP to purchase a property which in turn can be rented to the company. The loan necessary to achieve that acquisition (or more importantly to refinance) was then made on arm’s length terms. We cannot see that there is an unallowable purpose here.

But by depriving itself of capital by writing off a part of the secured loan to which it was entitled to repayment in full on the sale of the property by VDP, it did something which was not amongst the business or other commercial purposes of the company. It deliberately deprived itself of working capital whilst benefiting the unsecured creditors who were also shareholders in the company.

Our conclusion, therefore, is that the partial write-off of the secured element of the loans owed by VDP to the company had an unallowable purpose and thus the tax-deductible debit which that partial write-off reflects, cannot be deducted for corporation tax purposes, by the company, for the tax years in question.

We now turn to consider the discovery assessment issues.

The law is set out in the appendix.

As for the technical issues, there are two sub issues which require resolution. Firstly, whether the company or someone acting on its behalf has acted carelessly in bringing about the underassessment of tax (the “carelessness issue”). This is important for two reasons. Firstly, where a company has submitted a tax return, HMRC cannot issue a discovery assessment unless there has been carelessness. Secondly, the period within which HMRC can issue a valid discovery assessment is extended to 6 years in the case of carelessness, which is important here in relation to the October 2020 discovery assessment.

The second strand of enquiry into the discovery assessment issues concerns whether the revenue amendment comprise a discovery assessment which was then settled by a section 54 agreement (the “section 54agreement issue”).

On this issue Mr Marks submits, in summary, as follows:

The 2014 accounts of the company which included the appellant’s debit (referred to as a bad debt provision) were submitted to Companies House on 26 June 2015.

The tax computations for 2014 confirmed that £72,520.90 and £94,579 were claimed as deductions for impairment losses (the first reflecting a write-off of the “trading balance”, the latter being an irrecoverable amount under the loan).

There is no documentary evidence to support the assertion that advice was taken before completing the accounts although the first appellant’s evidence was that he thought that it had been but could provide no such documentary evidence.

The only possible advice is the GT report. This is addressed to VDP and not the company. The company therefore cannot rely on this as being professional advice taken by the company. The final edition is dated 4 September 2015.

The GT report recommended two courses of action, one in which all debtors were paid off pro-rata and the second which paid out the first appellant and BM in full. In fact there was a third option which the report did not consider and which it should have done. GT clearly knew that there was a loan to VDP from the company. The £476,000 loan and the loan agreement are specifically referred to in the GT report. That loan agreement provides that the loan must be repaid in full upon the sale of the property. The GT report, therefore, should have considered the third option whereby the loan was paid off in full to the company as secured creditor, and the balance distributed to unsecured creditors pro-rata.

The GT report therefore has been carelessly drafted in that it does not consider the preferential rights of the company as secured creditor.

The GT report does not mention unallowable purpose either in respect of the loan or in respect of the way in which the balance of sale proceeds following the sale of the property and repayment to the bank, should be distributed. This is carelessness by GT. We should infer that they knew that the company’s loan was secured. If they did not, then this was carelessness on behalf of those instructing GT since it is clearly highly relevant to VDP’s creditors and the amounts to which they were entitled to be repaid.

On this issue Mr Smith submitted, in summary, as follows:

The company sought the advice of GT, a suitable reputable and knowledgeable adviser to whom they provided all of the relevant information.

Mr Sengupta had ultimate responsibility as partner, for the GT report. He was a long-standing and trusted adviser to the company. The report was sent in draft to Ian Brazier, the company’s finance director. GT had originally been instructed in August 2015. The report was dated 4 September 2015 but clearly a great deal of work had gone into it before then.

Although primarily responsible, Mr Sengupta delegated responsibility for drafting the report to an assistant and a senior manager who then involved GT’s treasury and loan relationship specialist.

Mr Sengupta himself, however, had specialised in mergers and acquisitions and private equity bank deals and was familiar with the issue of unallowable purpose.

It was not considered relevant either by Mr Sengupta or by the GT specialist that unallowable purpose was relevant. This reason was not included in the report was that it would simply have lengthened the report by including a matter which was irrelevant. Counsel confirmed that he, too, thought that it was irrelevant.

Although the GT report was addressed to VDP, it was clearly provided on the instructions of Ian Brazier, the finance director of the company and was used and acted upon by the company. The company could therefore rely on it, which it did, in relation to the tax consequences of repayments of the loan and the trading balance.

Carelessness, or failure to take reasonable care, is important for two reasons. Firstly, it is a gateway to HMRC raising a discovery assessment in circumstances where the company has submitted a corporation tax return for the relevant period. Secondly, it allows HMRC a 6 year limitation period to raise that discovery assessment.

Failure to take reasonable care can, in the circumstances, either be by the company or a person acting on behalf of the company.

For the purposes of this decision, we accept the submission by Mr Marks that the reasonable care which should be taken by a taxpayer is assessed by reference to a prudent and reasonable taxpayer in the position of the taxpayer in question. We extrapolate this as regards a person acting on behalf of the company by saying that the reasonable care which such an adviser should take is assessed by reference to a prudent and reasonable adviser in the position of the adviser in question.

In both cases there is an objective and a subjective element. The objective element is the prudent and reasonable taxpayer or adviser. But reasonable care must be tested against the personal or professional qualities, expertise and experience, which either the taxpayer or the adviser actually has or holds out as having. So whilst there is one standard (reasonable care) what is reasonable in one case, may not be reasonable in another, depending on the personal and professional attributes of the taxpayer and adviser.

In this case, HMRC assert that GT, acting on behalf of the company, have failed to exercise reasonable care when compiling the GT report. They also assert that the company may have also failed to exercise reasonable care in failing to tell GT of the company’s security over the property owned by VDP.

HMRC also assert that the GT report was written for the benefit of VDP and cannot be relied upon by the company. And so the company could not have acted carefully by acting upon the advice of GT. We reject this submission. As is common in the circumstances, where an adviser is instructed in relation to an intergroup transaction, and notwithstanding that there might be reliance restrictions in the document, it is commonplace for the parties to the intergroup transactions to rely on the advice provided. We find this to be the case in the circumstances of the GT report. Whilst it was addressed to VDP, it was clearly commissioned by the finance director of the company and affected both. It is wholly reasonable therefore for the company to rely on GT’s advice. By doing so we find the company exercised reasonable care.

However, and we have not come to this conclusion lightly, it is our view that GT failed to take reasonable care when compiling the GT report.

As far as the connected company position is concerned, the GT report considered this and concluded that it was “doubtful that [the company] and VDP are connected under the loan relationship rules”. It then went on to explain why (there was no individual who controlled the company and VDP through the holding of shares, powers or other documents).

So, GT not only considered the connected company position but also gave reasons as to why in their view the companies were unconnected. It so happens that their view coincides with ours, but even if it had not done so, we would have respected the technical merits of their view and found that they had exercised reasonable care in coming to their conclusion.

But we are afraid this cannot be said in respect of the unallowable purpose position.

It is Mr Smith’s submission, and Mr Sengupta’s evidence, that the reason that there is no reference to unallowable purpose in the report is that it was considered to be irrelevant to the circumstances for which the report was commissioned.

We find this surprising, to put it mildly.

The GT report is clearly based on evidence that the loan was formalised by a loan instrument in September 2010. On page 9 of the report it states “The £476K was formalized by a loan instrument in September 2010”. We suspect this is a mistake and it is intended to refer to the loan agreement of 26 November 2010. We do not know of the basis of this statement, but we take the view that a properly instructed adviser would have been provided with a copy of the loan agreement from which it is abundantly clear not only that the loan was secured, but also that it was repayable in full on the sale of the property.

Yet no reference is made to either of these provisions in the GT report. So, someone here has acted without reasonable care. It may have been the company in not providing full instructions to GT. It may be GT in failing to properly read the information that was provided to it, or, having been told something by the company, failing to verify that. But in either case, the question of whether the loan from the company to VDP was secured or unsecured, was a matter of considerable importance, as too was the obligation on VDP to discharge the loan on full on the sale of the property.

We were told that Mr Sengupta, acting as client partner has acted for the company for a number of years. And so, with the greatest respect to him, in his capacity as a partner in GT, we would have expected him to have known about, and ensured that full information was provided to his colleagues in respect of, the loan of 26 November 2010.

But even without this information, it is our view that when loans are being written off the unallowable purpose provisions are a matter of fundamental importance to any advice regarding the propriety and technical basis for asserting that a partial write-off of a loan could be taken as a tax-deductible debit in the accounts and tax return of the lender. The unallowable purpose provisions are renowned for having an extensive but somewhat undefined scope and thus impact on both the creation of a loan relationship and matters affecting it.

So, we do not think that the submission that the justification for failing to address unallowable purpose in the GT report as its being irrelevant, has merit.

We would have expected that unallowable purpose would have been addressed in the report if only to reject its application to the circumstances. If this had been the case, and depending on the rationale given for that dismissal, it is much more likely that we would have found that GT had taken reasonable care in compiling the report.

However, no mention whatsoever is made in the GT report about unallowable purpose.

And this is notwithstanding that GT is a highly reputable and experienced provider of professional services and holds itself out as having a high level of expertise. And this is endorsed by Mr Sengupta who, understandably, indicated that the technical issues associated with the proposed repayments were dealt with by specialists who reported to him, and who in turn took specialist advice from the national loan relationship specialist within GT.

If this was the case, then something seems to have slipped down a crack. We do not believe that the national loan relationship specialist within GT, properly instructed, would not have considered, let alone come to a conclusion on, the application of unallowable purpose to these loan repayments.

The failure to instruct improperly, or his or her failure to address them, leads us to the conclusion that GT failed to exercise reasonable care in the GT report on the unallowable purpose issue.

The objectively prudent and reasonable adviser which found itself in the position of GT, with GT’s expertise and experience, would have considered unallowable purpose in the context of the proposed loan repayments, and would have reflected that consideration in its report.

This should have been the case anyway, but is exacerbated by the fact that they knew, or should reasonably have known, that the loan was secured and that the company had the right to insist on full repayment of its loan as a condition for release of that security.

Our conclusion on this carelessness issue, therefore, is that HMRC have passed through the carelessness gateway which is required in order to make a discovery assessment, and the extended 6 year limitation period is engaged rather than the basic 4 year limitation period.

On this issue Mr Smith submitted in summary as follows:

The revenue amendment was held out and accepted by all parties to be an assessment in relation to the 2014 company tax return and its withdrawal by HMRC constitutes an agreement under section 54 TMA. The matter therefore is settled and the October 2020 discovery assessment is invalid. HMRC could not have made a second discovery. They cannot discover the same thing more than once.

The parties were under no misapprehension of what the revenue amendment represented. It might be a “misdescription”, but that does not mean that it was not an assessment.

Section 114 TMA, entitled “Want of form or errors not to invalidate the assessments, etc” gives both HMRC and a taxpayer a statutory right to claim that the assessment, warrant or other proceeding in question shall not be affected by reason of mistake etc.

The revenue amendment is simply a mistake of form. It is in substance an assessment dressed up as a revenue amendment. The company knew full well what the revenue amendment was supposed to be for and there was no ambiguity over what had been assessed and for what period. There was no gross error. There was simply an incorrect statutory reference which falls within the safety net of section 114 TMA.

If this is right, then HMRC’s withdrawal of the revenue amendment on 18 August 2020 created a settlement of the issue by virtue of section 54 TMA. The withdrawal of the revenue amendment cancels the assessment which in fact it was, and the matter is therefore settled as if the tribunal had cancelled the assessment. The fact that there was reference to the matter being settled by a section 54 agreement in the notice sent to the company and the agent on 27 October 2020 reflects HMRC’s acceptance that the matter had been settled in that way.

On this issue Mr Marks submitted, in summary, as follows:

The revenue amendment was of no effect as it was raised pursuant to the consequential amendment provisions in paragraph 34 of schedule 18 to the Finance Act 1998. This paragraph deals with consequential amendments following the issue of a closure notice or partial closure notice which in turn arises when an enquiry has been instigated. In this case no such enquiry had been raised into the affairs of the company for the 2014 accounting period and thus the revenue amendment was without any statutory basis.

Section 114 TMA cannot be used to correct the position. It cannot deem a revenue amendment which had no statutory basis for issue, to be a discovery assessment.

Instead, the only discovery assessment which was issued was the October 2020 discovery assessment. That is clearly described as a discovery assessment (“assessment to make good to the Crown the loss of tax”) whereas the revenue amendment is described as a “notice of amendment to a company return”.

Officer Heatley simply made a mistake, in failing to check the correct provision under which the assessment should have been made. She didn’t realise there was an error until the litigation team at HMRC pointed this out to her. She corrected the error a day later, having emailed Mr Sengupta on 18 August 2020 about the administrative mistake. In that email she said that the revenue amendment “will be withdrawn”. The October 2020 discovery assessment was the correct assessment.

There was no agreement under section 54 TMA that the subject matter of the revenue amendment or the October 2020 discovery assessment was agreed. Any section 54 agreement proposed by Officer Heatley on 27 October 2020 was rejected by the first appellant on 26 November 2020.

We need to consider two issues. Firstly, whether the revenue amendment should be construed as a discovery assessment. The company asked us to do so. Secondly it has also asked us to find, if we do find it was a discovery assessment, to go on to say that the issue with which it dealt has been settled by HMRC withdrawing from the appeal against the revenue amendment. And so, because HMRC could not make a second discovery, the October 2020 discovery assessments are not valid.

We have considered both of these, but also a third position, which has effectively been pre-empted by HMRC in their skeleton argument, and which is whether issue estoppel might apply, irrespective of whether we find that the revenue amendment was an assessment, since a section 54 agreement would conclude the issue raised in that revenue amendment irrespective of whether it should be treated as a discovery assessment.

Our view is that the revenue amendment was not a discovery assessment. What seems to have happened is that Officer Heatley has made a single discovery. This was, according to her evidence, in or around the beginning of August 2019 and that discovery was promulgated to the company by way of a notice entitled “notice of amendment to a company tax return”, issued on 15 October 2019.

Under the discovery assessment provisions in schedule 18 FA 1998, it is clear that the making and issuance of a discovery assessment is a two-stage process. Firstly, the officer makes a discovery and secondly, that discovery must be notified to the taxpayer by way of an assessment. It is this latter document which is described as a discovery assessment.

So, there is a formal process which must be undertaken in order for HMRC to issue a valid discovery assessment. It is clear from the evidence that Officer Heatley had fulfilled the first element of this process. She made a discovery.

The way in which this discovery was notified to the company was then incorrect.

In order to raise a revenue amendment, under paragraph 34 schedule 18 Finance Act 1998, a partial or final closure notice must have been given to the company. And such a notice could only be given if HMRC had opened an enquiry into the 2014 corporation tax return. No such enquiry was opened. So, there was no lawful basis on which HMRC could give the revenue amendment. It was an unlawful document.

Mr Smith submits, however, that it takes effect as a notification of the discovery, and thus is a discovery assessment. And notwithstanding the described as a revenue amendment, section 114 TMA allows the company (and indeed the tribunal) to so treat it as it is in substance an assessment and so should not be treated as void or voidable, or affected by, a mistake, defect or omission. His argument, as we understand it, is that HMRC have mistakenly promulgated the notice of assessment under the description of a revenue amendment, and that misdescription should be disregarded and the notification treated as a discovery assessment.

In his view this does not reflect a gross error. It does not fall into the category where HMRC have issued, for example, an assessment for the wrong fiscal year. It is simply a mistake in the form of notification. It is not a mistake in the substance of the discovery and notification.

Although not cited to us, we have considered the judgment of Henderson J (as he was then) in the case of Pipe v HMRC [2008] EWHC 646 (CH) (“Pipe”).

In Pipe the court had to consider whether mistakes in penalty notices could be saved by section 114 TMA. In that decision, the judge said:

“….I would emphasise that Slade LJ said in terms… That section 114(2) had no application to the facts of that case. I infer from this that the mistakes made by the Revenue must have been not only a mistake in the form of the notice of assessment sent to the taxpayer’s agent, but also a mistake in the assessment itself which was written in the assessment book. If the entry in the assessment book had been correct, and the only error was in the notice, the Revenue would presumably have been able to rely on section 114(2)(b)…”.

At face value, this would appear to support Mr Smith’s submission. There was no mistake in the discovery itself. The mistake was in the notice notifying the company of that discovery.

But the difference between Pipe and the revenue amendment in this appeal is that the form of notification in Pipe was one which was permitted by statute. All that was wrong was that it did not contain adequate information.

In the case of the revenue amendment however, it is an illegal document. It could not be what it purports to be i.e. a consequential amendment, since the statutory prerequisites for the issue of such an amendment did not exist. It could not, therefore, be legal notification of the discovery made by Officer Heatley. And this distinguishes the situation from Pipe where there was no illegality in the notification, merely a mistake of form.

In our view section 114 TMA does not extend to correct the mistake made by HMRC of notifying the discovery by way of a consequential amendment. The revenue amendment is not notification of Officer Heatley’s discovery such to bring it within the ambit of paragraph 41 of Schedule 18 Finance Act 1998.

The second issue is whether or not it took effect as a discovery assessment, the issue with which the revenue amendment dealt has been settled by dint of a section 54 agreement so that the issue cannot be re-litigated (the effect of which is that HMRC cannot rely on the October 2020 assessment as being valid, and the appellant’s appeal against it must succeed).

Sections 54(1) and 54(4) TMA both include deeming provisions. When an appeal is withdrawn by the person who has given notice of appeal (in this case the company) the parties are deemed to have “come to an agreement…that the assessment or decision under appeal should be upheld without variation”. In these circumstances, a tribunal is then deemed to have “determined the appeal and… upheld the assessment or decision without variation….”.

In this case, the section 54(4) TMA deeming provision cannot apply since it was HMRC who withdrew from the appeal, by effectively withdrawing the revenue amendment. It was not the taxpayer who withdrew from its appeal.

In these circumstances there must be an actual agreement between the parties that the assessment or decision under appeal should be treated as upheld without variation, or as varied, as discharged or cancelled.

The only documentary evidence that such an agreement was entered into is the somewhat delphic comment on the document dated 27 October 2020 entitled “Agent copy of notice of amendment to a company tax return”, that “Please see email correspondence confirming this represents cancellation of previous amendment dated 15/10/19. The appeal is determined under Section 54 Taxes Management Act 1970”.

A simple statement to this effect in this document does not, of itself, either reflect the fact that there was such an agreement or comprise a section 54 agreement of itself. One needs to consider whether there was, in fact, an agreement reached regarding the revenue amendment. The facts show that there was no such agreement reached. All that happened was, as reflected in Officer Heatley’s email to Mr Sengupta of 18 August 2020, that HMRC in preparing for the hearing, spotted an administrative mistake (the assessment was raised as a revenue amendment) and would be withdrawn. That was then unilaterally withdrawn by HMRC on 27 October 2020. There were no negotiations regarding settling the corporation tax deduction claimed by the company in its 2014 tax return which were then formalised in a bilateral agreement between HMRC and the company. There was simply the unilateral act of withdrawal by HMRC and the subsequent collapse of the appeal against the revenue amendment since it no longer had any effect.

Mr Smith submits that the appeal which was made against the revenue amendment by dint of Mr Sengupta's email of 29 October 2019 comprises an offer, which was then accepted by HMRC when it withdrew the revenue amendment. We cannot accept this. Mr Sengupta's email was simply an appeal against the additional tax of £167,767 set out in Officer Heatley's letter of 21 October 2019. Of this £167,099 related to the loan relationship issue and Mr Sengupta states "we therefore appeal against this aspect and we look forward to resolving this matter at independent Tribunal".

This is not an invitation to treat let alone an offer to settle. This is Mr Sengupta telling HMRC that he thinks they are mistaken and that the matter should be settled by an independent tribunal. It is a conventional notice of appeal. It tells HMRC that they are wrong. It is not an overture to reach a compromise.

Furthermore, HMRC withdrawing that revenue amendment cannot be seen as accepting the company's position in relation to anything. It was issued in the absence of any input from the company, and arises from HMRC's prehearing review, when it was realised that the revenue amendment had no statutory basis and could not therefore be the notification of a discovery assessment.

So, we do not think that there was any offer and acceptance, in the conventional sense of those terms, which is a prerequisite to a valid section 54 agreement.

It is our view that the circumstances fall outside the provisions of section 54 TMA. And so, the issue which was the subject matter of the discovery and subsequent revenue amendment/assessment (namely whether the company was entitled to its loan relationship debit as a deduction against its corporation tax liability for 2014) was not the subject of an agreement between the parties. As such, by issuing the October 2020 assessment, HMRC are not seeking to relitigate a matter which has already been determined either by the tribunal or by agreement.

Drawing these strands together. Officer Heatley made a discovery which she then notified to the company by way of a notice of consequential amendment. There was no statutory basis on which this notification could be given. It did not take effect as a valid notification of a discovery assessment under paragraph 41 of schedule 18 Finance Act 1998. The valid notification of that discovery was the October 2020 discovery assessment. Even if it did, withdrawal of the revenue amendment did not comprise a section 54 TMA agreement, and there was no such actual agreement made between the parties in connection with the discovery assessment or its subject matter. HMRC are not therefore estopped from litigating the October 2020 discovery assessment in these appeals.

HMRC had originally assessed the company to a penalty based on deliberate behaviour in respect of the December 2014 accounting period, but subsequently adjusted that and assessed the company to a penalty based on careless behaviour for that period.

This is clear from HMRC's letter of 23 November 2020. The letter also suggests that the penalty for the 2015 accounting period was also based on careless behaviour. Yet Mr Marks, in his closing submissions, suggested that it was based on deliberate behaviour. And so should be reduced from 42% to 19.5%. But no matter.

However, it is our view that the company has not acted carelessly in respect of either the 2014 or 2015 accounting periods.

As Mr Marks fairly points out, the situation is different when assessing a penalty under schedule 24 FA 2007 from that which is relevant to the issue of the assessment in the first place which is pursuant to schedule 18 Finance Act 1998. In the former, carelessness must be tested from the perspective of the taxpayer, whereas for the latter, as has been seen, carelessness depends on the behaviour not just of the taxpayer but also of an agent acting on the taxpayer's behalf.

As regards the October 2020 discovery assessment we have found that GT were careless.

We have also found, as regards the question of GT's knowledge that the loan was secured, that there was carelessness either by them or by the company.

But it is for HMRC to establish, when it comes to penalties, that the company has been careless. And our finding that either the company or GT was careless as regards the security for the loan is insufficient to discharge the burden of showing that the company has been careless.

Mr Marks submits that even if this was the case, the advice was not followed, and no advice was sought in the final repayment position or the deductions for corporation tax purposes actually claimed in the return.

In our view the GT report is abundantly clear that GT's view was that there was no connected parties loan relationship and thus the company could take a tax-deductible debit against its profits for the debt write-off.

On the basis of this advice the company submitted its corporation tax returns claiming the tax deduction.

So, we do not think that the company has done anything wrong here. It was party to advice given by a highly reputable firm of accountants and was entitled to rely on that advice. That advice was that it could take a tax deduction as aforesaid, which the company then did in its tax returns.

We therefore find that the company did not act carelessly in respect of the penalties to which it was assessed on this issue.

This issue is comparatively straightforward. The company claims that the first appellant had taken out and used for business purposes £11,630 of petty cash. He then spent this, wholly and exclusively, on business expenses. The entire amount, therefore, is tax-deductible. HMRC do not think that the company has provided sufficient evidence that the expenses have been incurred wholly and exclusively for the purposes of the company's trade. They have accepted that 30% are business expenditure, but the balance, £8,141, is not deductible. HMRC have also assessed the company to a penalty on the basis of careless behaviour.

The first appellant's evidence was that the money taken from petty cash was taken for him to pay for business visas, tips, taxes, lunches, drinks and general incidental expenditure when he was out of the country for weeks at a time visiting overseas suppliers, and in particular in India and Sri Lanka, where cash was required as there was no other way to pay for such expenses.

Mr Smith submits that it is for HMRC to show that this expenditure had a non-business purpose. But this is not correct. Having assessed the first appellant, the burden is on the first appellant to show that the expenditure was incurred wholly and exclusively for business purposes.

Mr Marks submits that there is no basis on which we can disturb the assessment since no satisfactory evidence has been submitted to us showing it was incorrect, nor what the correct amount is more likely to have been.

And, given that there is no documentary evidence of the matters on which the expenditure was incurred, we will never know precisely what the expenditure was on.

We accept this last point. But there is evidence, which was not accepted by Officer Heatley as to the nature of the expenditure on which we can base an alternative amount of business expenditure. That evidence comes, of course, from the first appellant. We accept his evidence that the money was used on business expenditure, not just for his expenses but also for those employees of the company who accompanied him on such trips. We also accept his evidence that paying for the matters for which he used the petty cash was important to his business relationships.

However, it is inevitable that a certain amount of the expenditure would have had a personal benefit for not just the first appellant but also for the other employees who accompanied him. It is difficult to quantify this, but we think this is likely to be 20% of the total expenditure.

It is our view therefore that 80% of the £11,630 of petty cash which is in issue in this appeal was incurred wholly and exclusively for business purposes and is thus tax-deductible for the first appellant.

We agree with HMRC that the behaviour which led to the "over claiming" of the deduction is careless as there was very little verification undertaken by the first appellant to justify the deduction.

So, as for NIC penalty for the first appellant, we direct that HMRC reassess the penalty in respect of the petty cash deduction. It should be based on careless behaviour and on 20% of the £11,630. In the absence of any agreement, then either side may apply to the tribunal for the matter be determined.

It is our decision that:

The first appellant's appeals against the October 2019 discovery assessments and the penalty assessments issued on 6 November 2019 are dismissed.

The company's appeal against the section 8 NIC decision dated 6 November 2019 is dismissed.

The company's appeal against the NIC penalty of £22,125 is allowed, but the penalty assessment is to be recalculated, by HMRC, on the basis of careless rather than deliberate behaviour. Any issues regarding this recalculation may be referred to the tribunal for resolution.

The company's appeals against the October 2020 discovery assessment (in respect of the December 2014 accounting period) and the closure notice of 11 October 2019 (in respect of the December 2015 accounting period) are dismissed.

The company's appeals against the penalties in relation to the loan relationship issue are allowed.

The company's appeal against the petty cash issue is allowed in part. 80% of the expenditure shall be treated as tax-deductible. The penalty shall be based on careless behaviour. HMRC will recompute both amounts and in the absence of agreement between the parties, the matter can be brought back before the tribunal for determination.

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.