Bharat Patel v The Commissioners for HMRC

Ground 1 is that the FTT erred in law in its stage 3 evaluation by failing to recognise that, at least in relation to the pre-2015 VAT assessments, Mr Patel’s underlying time-limit case under section 73(6) VATA was very strong. That, it was submitted, was clear without any detailed investigation and the merits of his case should therefore have carried significant weight in the balancing exercise. To understand the arguments surrounding the time limit issue we need to turn to the relevant legislation and case-law.

Section 73(6) VATA provides the assessment shall not be made after the later of:

“(a)

2 years after the end of the prescribed accounting period; or

(b)

one year after evidence of facts, sufficient in the opinion of the Commissioners to justify the making of the assessment, comes to their knowledge…”

There was no real dispute as to the correct approach to the one-year limb in section 73(6)(b) as explained in the relevant authorities:

The appellant submitted that, applying that agreed approach, the FTT should have treated his time-limit case (at least for the pre-2015 assessments) as very strong and therefore a significant factor at stage 3 of the Martland balancing exercise. His case was that the last substantive provision of evidence to HMRC occurred before 24 December 2014 and that nothing material was provided after that. Accordingly, the assessments issued in 2016 for periods ending in 2013 and 2014 were outside the one-year limb (and, for those periods, also outside the two-year limb). He contended that subsequent requests by HMRC for meetings or documents, and the absence of any response, did not of themselves supply later evidence capable of restarting the one-year period on the facts of this case. He also emphasised that HMRC had not, before the FTT, identified any specific later evidence within the final year which (in the assessing officer’s opinion) justified the assessments. Once the taxpayer had identified what appeared to be the last relevant evidence, it was for HMRC to explain why that earlier evidence was not sufficient. On that basis, he submitted that the FTT’s statement that it could not reach a clear view and that the case was not “obviously overwhelmingly” in his favour showed a failure to take proper account of the strength of the time-limit point.

HMRC submitted that no error of law was disclosed because the FTT was required only to form a general impression of the underlying merits and to avoid a “mini-trial”, and an appellant tribunal should be slow to interfere with the FTT’s evaluative discretion on a late-appeal application. HMRC contended that, on the material before the FTT, it was open to conclude the case was not obviously overwhelmingly in Mr Patel’s favour because the contemporaneous correspondence showed continued non-engagement through 2015 and 2016 (non-attendance at the planned December 2015 visit and failure thereafter to provide an alternative date or make records available), and such non-engagement could itself constitute evidence capable of justifying an assessment. Relying on Rasul, HMRC argue that non-attendance at a meeting can amount to relevant evidence of fact and may support an inference that the taxpayer has nothing further to say. HMRC also submitted it was relevant that the way in which the appellant had put his case below was different to the one now being raised. In any case, HMRC submit the FTT had before it material which undermined the appellant’s case that he had told HMRC at the December 2014 meeting there was no more information to give (inconsistencies with contemporaneous correspondence and the FTT’s unchallenged view of the appellant’s lack of reliability as a witness). It was therefore well within the range of permissible outcomes for the FTT to treat the merits on the time limit issue as not obviously decisive at the permission stage.

The issue is whether the FTT erred in law by failing to recognise that the merits of the appellant’s underlying appeals (in respect of the pre-2015 VAT assessments) were very strong having regard to the one year time limit under s73(6) VATA. For the reasons which follow we do not consider this ground is made out. We start by reminding ourselves that the question for us is not whether the Upper Tribunal would characterise the merits in the same way, but whether the FTT’s approach to that question was legally flawed. As the case-law makes clear (see [17] above), the FTT was required only to form a general impression of the merits without undertaking a detailed investigation. It was not required to carry out a full examination of the time-limit issue.

On behalf of the appellant, Mr Chacko submitted that this was a case where it could straightforwardly be seen that the appellant’s case on time limits was overwhelmingly strong: the “last piece of the puzzle” was the meeting before 24 December 2014 by which time it was apparent that the taxpayer had no further material to provide, so the one-year period in section 73(6)(b) necessarily began then. On that basis, and in the absence of any positive evidence from HMRC identifying later facts within the final 12 months, the time limit argument was said to be effectively unanswerable and capable of resolution without detailed inquiry.

However, in agreement with Mr Abernethy, who appeared for HMRC, we consider it relevant, in assessing whether the FTT erred in its legal approach, to revisit the way the time-limit point was put before the FTT. In the written material before the FTT, the appellant’s formulation of the “one year” point in relation to the pre-2015 assessments was in terms that:

“they exceeded the rule of 1 year from facts December 2014 (By 24 Dec 15 can have assessed all VAT quarters between 24 Dec 11 to 24 Dec 15) however assessments were not done by that date”.

The thrust of the appellant’s post-hearing submissions to the FTT was that HMRC had seen “some” records at (or by) the December 2014 visit, and that, given HMRC’s knowledge of the business, the officer could and should have made best-judgment assessments then (or by the end of 2015), with the possibility that HMRC could “always raise another assessment thereafter” if further information emerged. That formulation directed attention to whether the officer ought reasonably to have assessed earlier on the basis of the partial material then available.

That contention is materially different from the way the point was developed before us. The argument now advanced focuses on identifying, as a matter of section 73(6)(b), the date on which the last piece of evidence of sufficient weight to justify the assessment was communicated to HMRC, and then measuring the one-year period from that date. By contrast, the argument before the FTT was framed in terms of whether HMRC could and should have assessed earlier on a best-judgment basis. As explained in Rasul (see [88]) those are not the same questions. The fact that the evidence might reasonably be regarded as of sufficient weight so as to found the basis for a valid best judgment assessment does not mean a decision not to regard the evidence as being of sufficient weight is one which is necessarily wholly unreasonable or perverse.

The argument now pressed also places different weight on Mr Patel’s evidence about the December 2014 meeting. In his witness statement he said that at that meeting he gave access to all the records he had, that the officer said some were missing, and that he distinctly remembered saying he had no other paperwork to give. On the present formulation, the “last piece of the puzzle” is said to be the fact (known to HMRC by December 2014) that no further documentation would be forthcoming, so that the one-year period could not be kept open merely by HMRC continuing to request information or meetings. That is materially different from the best-judgment submission advanced below, which proceeded on the basis that some records had been provided and were sufficient to justify earlier assessment, and implicitly invited a conclusion that it would have been unreasonable to wait.

Taking the argument as it was put to the FTT first, it was open to the tribunal to conclude that it could not reach a clear view on the likely strength of the time-limit contention without fuller consideration of the evidence and arguments. It was entitled to regard a detailed assessment of the issue as inappropriate at the permission stage. In that context the FTT would have been able to infer from HMRC’s subsequent correspondence that HMRC had not understood the taxpayer to be saying that all the information that would be produced had been produced. The FTT would not have been bound to accept the version given by Mr Patel or to have ascribed it more weight, as compared with the contemporaneous documents, given the lapse in time when it was served and also in the light of the FTT’s finding (at [54](f)) that the appellant was not an “entirely reliable witness”.

Even if we were to assume, in the appellant’s favour, that the FTT should have addressed the argument as it is now put, the FTT would not, in our judgment, be duty bound to have concluded that the merits on the time limit argument were overwhelmingly strong. Mr Chacko referred us to a number of decisions at first-tier tribunal level that made the point that the fact HMRC had requested further information from the taxpayer or a meeting with them did not, without more, amount to new evidence. While those observations reflect a concern that HMRC should not circumvent the protection afforded by the one-year time limit, they do not eliminate the need for a fact-sensitive evaluation of whether subsequent events amount to “evidence of facts”.

It remained open to HMRC to contend that the subsequent attempts at meeting and getting information in 2015 were capable of amounting to “evidence of facts”. If so a further question would arise as to whether the officer’s view that evidence of facts obtained before the one year preceding the assessment was not of sufficient weight to justify an assessment could properly be characterised as irrational or perverse. The chronology recorded by the FTT included later communications and attempts by HMRC to arrange inspection and obtain records in 2015 and 2016. In those circumstances there remained scope for argument, without deciding the point, as to whether the relevant “last evidence” for section 73(6)(b) purposes was fixed at the December 2014 meeting or whether later events (including later non-engagement in response to specific requests) could bear on when, in the officer’s opinion, sufficient evidence existed to justify assessment. The observations in Rasul illustrate how non-attendance, depending on context, can have evidential significance. While the FTT, correctly in our view, did not find the appellant’s case hopeless, it was well able to conclude that its merits were not overwhelmingly strong. The fact that the appellant had filed evidence but HMRC had not did not mandate the FTT to find that the appellant’s case would therefore succeed. The FTT would be entitled to take into account the documents that HMRC had already provided, that if the matter proceeded to a substantive hearing HMRC might rely on witness evidence at that stage, and that the evidence of either party might when tested in cross-examination and evaluated alongside the totality of evidence not necessarily be accepted.

We are not therefore persuaded that the FTT erred in its assessment of the merits, which in line with authority was to be conducted without the need for detailed investigation. Its conclusion that the case was not obviously overwhelmingly in the appellant’s favour was within the range of conclusions reasonably open to it on the material and submissions before it. Ground 1 is therefore not made out.

The appellant’s submissions on Ground 2 were that the FTT erred at stage 3 of the Martland balancing exercise by placing weight on the prejudice to HMRC of “reopening” a dispute HMRC regarded as closed, because (he contends) an out-of-time VAT assessment is a nullity and can be challenged outside the tribunal system, including by way of defence in collection proceedings (or by a claim for wrongly demanded tax). He relied on Lord Advocate v Shanks as authority that an out-of-time VAT assessment may be treated as a nullity and its invalidity may be raised in enforcement proceedings. On that basis, the FTT is said to have taken into account an irrelevant (or legally mistaken) consideration, or to have failed to take into account the proper relevance of the availability of a challenge in another forum when assessing finality and prejudice.

In Lord Advocate v Shanks (a decision of the Second Division, Inner House of the Court of Session), the Commissioners of Customs and Excise had made a global VAT assessment which they accepted was out of time under the relevant statutory limit. The Commissioners nevertheless brought an action in the Sheriff Court to recover the sum assessed as tax due, relying on the “deemed due” provision in relation to collection (Paragraph 4(9) of Schedule 7 of the 1983 Value Added Tax Act which referred to assessments made under Paragraph 4(1) of that Schedule) where no tribunal appeal had been brought. The sheriff granted decree in favour of the Commissioners, but on appeal (to the sheriff principal) it was held that an assessment made out of time was not a valid assessment and could not be treated as “deemed due” so as to preclude a validity challenge in recovery proceedings. On a further appeal to the Inner House, their Lordships each agreed an out-of-time assessment was a nullity and upheld the decision under appeal. The out of time assessment was not an assessment within the meaning of Paragraph 4(1) and therefore did not fall within the “deemed due” provision of Paragraph 4(9). The assessment’s invalidity could therefore be raised by the taxpayer as a defence in the recovery action notwithstanding the absence of a tribunal appeal.

Mr Chacko submitted that the ability to argue validity in collection proceedings was further supported by a legislative comparison between the VAT legislation on appeals and the direct tax legislation. The VAT legislation had no equivalent to s34(2) Taxes Management Act 1970 (which required that time-limit challenges to direct tax assessments had to be made by way of appeal). He also submitted that HMRC’s suggestion that Shanks could be disregarded because it was a decision of the Scottish courts was wrong in a tax context (relying on R (Jwanczuk) v HMRC [2025] UKSC 42 for the proposition that, in revenue cases, UK courts and tribunals generally follow superior court decisions from other UK jurisdictions on the construction of tax statutes). On that basis, he said the FTT’s reasoning that HMRC should not have to expend resource on a matter which was “closed” proceeded on a mistaken view of the law, giving a factor (prejudice to HMRC) significance it should not have received.

Mr Abernethy disputes the possibility of validity challenges based on time limits in collection proceedings. He referred to Autologic Holdings v IRC [2005] UKHL 54 (at [13]) and the Court of Appeal’s application of that approach in Knibbs v HMRC [2019] EWCA Civ 1719, arguing the exclusivity principle meant the taxpayer had to use the statutory remedies and cannot indirectly challenge assessments in civil proceedings. HMRC also relied on various insolvency cases (Vieira v HMRC [2017] EWHC 936 (Ch), HMRC v Chamberlin [2011] EWCA Civ 271 and HMRC v Harris [2011] EWHC 3094 (Ch), and the Insolvency Practice Direction) to submit that, in practice, a bankruptcy court will not go behind a tax assessment on the basis that it is out of time, save in exceptional cases such as fraud, collusion or a glaring miscarriage of justice. HMRC again highlight that the point was not raised before the FTT in the form now advanced, and that it was not an error of law for the FTT to omit a point not clearly articulated unless it “leapt from the evidence” as obviously significant (referring to Pawar v HMRC [2025] UKUT 00309 (TCC) at [51] and Market and Opinion Research International v HMRC [2015] UKUT 0012 (TCC) at [62]). HMRC submitted that even if, contrary to the above, they could, despite there being no appeal, face the validity issue in collection proceedings, the point was not material to the overall analysis because it would, if anything, reduce the prejudice in that the taxpayer could still run the validity point elsewhere.

It is helpful to start with some opening observations about Stage 3. As explained in Martland (at [45]), the tribunal’s task is to exercise judicial discretion taking account of all relevant factors rather than to work through a checklist. As regards how a tribunal should approach considerations of prejudice, beyond noting the case-law on consideration of merits (see above at [17]) and the need to weigh the importance of complying with time limits, Martland did not prescribe a particular detailed approach. The tribunal must (as with any case of course) engage with the core points of argument and evidence raised by the parties, but it is not required to undertake a free-standing investigation of potential arguments that have not been advanced. The way the issues are framed and the evidence and submissions are put forward will therefore matter. With that in mind we note the following.

First, it is necessary to be clear about the way in which Shanks was deployed before the FTT. The point made was not that, because an out-of-time VAT assessment might be challenged in enforcement proceedings, HMRC could not properly rely on finality or on the dispute being “closed” when prejudice was assessed. Rather, the appellant relied on Shanks in support of a different proposition: that if (and to the extent that) a VAT assessment is made out of time it is “void” or a “nullity”, with the result that the usual tribunal time limits did not apply (or could be treated as inapplicable) to an appeal against such an assessment. The FTT was not asked to decide, as a factor in prejudice, whether the availability of a possible collection-proceedings defence meant that HMRC could not say the matter was closed or would suffer prejudice by a late tribunal appeal. It follows that the FTT’s failure to address Shanks in the manner now advanced does not, of itself, disclose an error of law.

Second, even if one assumes in the appellant’s favour that the s73(6) time limit argument is a strong one that was capable of being deployed in collection proceedings it does not follow that the prejudice to HMRC is thereby minimal. If permission were granted, HMRC would be required to prepare for and defend a tribunal appeal. That would not necessarily be confined to the narrow validity point but could extend more broadly to issues of quantum if pursued. By contrast, if the time-limit issue were to be raised in collection proceedings, the dispute would be confined to the narrower issue of whether the assessments were issued within the statutory time limit. On that basis, the appellant’s submission does not demonstrate that prejudice to HMRC is minimal. Even if HMRC might have to address the time-limit issue elsewhere the FTT was entitled to conclude that there is an additional and not insignificant prejudice in requiring HMRC to defend a full tribunal appeal.

We add also that the appellant’s argument on prejudice, if correct, would cut both ways. If a strong time-limit argument can be raised outside the tribunal in collection proceedings, the refusal of permission to appeal out of time in the tribunal would mean the appellant could defeat the assessments in the courts. Mr Chacko’s response was that refusal of permission would still prejudice him because, even if a validity challenge were available in collection proceedings, he would lose the opportunity to dispute quantum before the FTT. That is a valid consideration in principle, but it does not ultimately change the outcome when the respective prejudice is balanced: HMRC would still suffer significant prejudice (if the possibility of the time limit issue being used in collection proceedings were to be used as a basis for granting permission to appeal out of time) whereas the appellant would not (he would still be able to argue about the time limit issue in the collection proceedings before the courts).

The stage 3 exercise does not in any case require the tribunal to determine how litigation might unfold in another court or to resolve the scope of such proceedings. That would involve a degree of speculation inconsistent with the established principles already referred to which discourage entering into a detailed investigation of the substantive merits of the dispute. The tribunal’s task is to evaluate the practical consequences of granting or refusing permission on the application before it.

Here, the FTT rightly concentrated on the consequences in terms of the prejudice of a grant or refusal of permission which were obvious and inherent namely of the taxpayer losing the right to appeal or conversely of HMRC having to defend an appeal in the FTT that it would otherwise not have to.

Accordingly the FTT, in dealing with an application to make an appeal late was not, when weighing the respective prejudice to the parties, required to determine the status of Shanks and posit how that would be applied in any potential recovery proceedings that were initiated, and in relation to which assessment validity issues were raised despite no appeal on that in the FTT having been timeously pursued. By the same tokenit follows that we, in hearing an appeal against such decision, do not consider this the appropriate forum in which to resolve how a first instance court dealing with recovery proceedings would approach any argument about the validity of an out-of-time VAT assessment or determine whether they would follow Shanks. (Shanks it will be noted arose out of collection litigation rather than in the context of a tribunal appeal).

Finally, while we accept that language, routinely referred to in HMRC’s submissions and in FTT decisions on permission to appeal late applications, regarding HMRC being able to regard a matter as “closed” should be used with care where an appellant asserts that validity could be raised elsewhere, the FTT’s reasoning, properly read in context, was addressing the practical prejudice in requiring HMRC to devote resources to a substantive tribunal appeal long out of time, and the broader public interest in compliance with statutory time limits. On the appeal as advanced below, that was a legitimate consideration when analysing prejudice within the stage 3 balancing exercise.

We therefore conclude that Ground 2 is not made out.

For the reasons above we dismiss the appeal.

By way of postscript we would add that, even if there was an error in the FTT not regarding the appellant’s merits as overwhelmingly strong, and in failing to recognise that HMRC could still litigate VAT assessment validity (but not quantum) in collection proceedings, and the FTT decision were then set aside, it should not be assumed that an FTT (or we, if we were in the position of remaking the decision) would have granted permission to appeal late.

The FTT made unchallenged findings on delay and on receipt of the assessment letters. It found that, on the balance of probabilities, the assessment letters were likely to have been sent to Mr Patel’s address and received by him. It treated the appeals as having been made in February 2018, at least a year after the last assessment date, and rightly noted the delay was serious and significant, going on to conclude that the reasons advanced for the delay were weak. No appeal is brought against those aspects of the FTT’s findings and analysis. In that context, the stage 3 balancing exercise remained one in which it was open to an FTT (or the Upper Tribunal if it were to remake the decision) to conclude that the combination of substantial delay and the prejudice on HMRC of defending a late tribunal appeal (which arose even if the time limit issue could be disputed in collection proceedings) justified refusal of permission. The outcome of the application did not simply turn on the balancing of respective prejudice and the strength of the underlying merits.

In other words even if the appellant were successful in his grounds such that the decision were remade we consider it unlikely the result would be different.