When does delay by HMRC render a discovery assessment invalid?

07 March 2019

In the recent case of Clive Beagles v HMRC [2018] UKUT 380 (TCC), the Upper Tribunal (UT) held that a delay of nearly two and a half years between HMRC discovering that a taxpayer's self-assessment tax return was insufficient and  issuing an assessment, was too long as the discovery had become 'stale' and the assessment was therefore invalid.

This blog is based on an article first published in Taxation on 26 February 2019. A copy of that article can be viewed here


Between 2001 and 2002, Mr Beagles participated in certain tax avoidance arrangements which were designed by KPMG to create a tax deduction with no corresponding taxable amount. The arrangements were ultimately found to be ineffective in Astall v HMRC [2009] EWCA Civ 1010. 

On 29 January 2003, Mr Beagles filed his self-assessment tax return for 2001/02 and claimed a loss of over £1million under the tax planning arrangements. He provided details of the transactions which gave rise to the loss in his return. Mr Beagles also included an explanation of the arrangements in a standard form provided by KPMG.

HMRC failed to open an enquiry into Mr Beagles' return within the statutory time period. It realised its mistake in June 2004 and initially decided to defer taking any action in relation to Mr Beagles' return until it had made further progress in relation to other participants in the arrangements, in respect of whom HMRC had opened enquiries under section 9A, Taxes Management Act 1970 (TMA).

In August 2005, having received advice from leading counsel, HMRC informed KPMG that it had concluded that the tax planning arrangements were ineffective and that it intended to challenge the arrangements before the courts. On 15 January 2008, HMRC issued a discovery assessment to Mr Beagles under section 29(5), TMA. Mr Beagles appealed the discovery assessment and his case was placed on hold until determination of the other participants' appeals by the High Court and the Court of Appeal (their appeals had already been dismissed by the Special Commissioners). The Court of Appeal rejected their appeals in Astall and permission to appeal to the Supreme Court was refused in February 2010. 

On 8 February 2008, Mr Beagles appealed the discovery assessment on the basis that, amongst other things, it had become 'stale' by the time it was issued and was therefore invalid.

FTT decision

Mr Beagles' appeal was dismissed by the First-tier Tribunal (FTT). In the view of the FTT, the discovery assessment was valid as (i) a hypothetical HMRC officer would not have been aware of the insufficiency of tax in Mr Beagles’ tax return, and (ii) the discovery made by HMRC had not become stale. 

Mr Beagles appealed to the UT.

UT decision 

The appeal was allowed.

Mr Beagles argued before the UT that:

1. the FTT’s conclusion that the date of discovery was on or around 18 August 2007 was perverse, contrary to the evidence, and based on a misunderstanding of the law;

2. the FTT’s view that, even if the discovery was made some time earlier it was not stale, was perverse and wrong in law;

3. the FTT erred in its approach to section 29(5), TMA, in particular, with regard to what the hypothetical officer would have been aware of; and

4. the FTT erred in its approach to section 29(6), TMA, in particular, whether the existence and relevance of a reason for the 'market change condition' (an element of the planning arrangements), could reasonably be inferred.

Can a discovery become stale? 

HMRC submitted that it was not possible for a discovery assessment to be invalid simply because the relevant discovery was made a substantial time before the date of the assessment and that any concept of staleness which had developed in case law was not correct as a matter of law. 

The UT considered the relevant case law in which the concept of staleness has been discussed. It noted that it was first expressly referred to in Charlton v HMRC [2012] UKFTT 770 (TCC). In that case, the FTT said there might be circumstances in which, if an assessment was not made within a reasonable period after discovery, the discovery may lose its essential 'newness' so that the assessment might be invalid. These comments were, however, obiter. 

The case of Pattullo v HMRC [2016] UKUT 270 (TCC), was the first case where the question of staleness was directly in point. Lord Glennie said in that case that, on a natural reading of section 29(1) TMA, it seemed wrong not to require HMRC to make an assessment promptly once a discovery had been made, although he noted that the legislation did not make any express provision for any limitation period, except for that specified by section 34, TMA:

"It would, to my mind, be absurd to contemplate that having made a discovery of the sort specified in s 29(1), HMRC could in effect just sit on it and do nothing for a number of years before making an assessment just before the end of the limitation period specified in s 34."

The next case the UT considered was the recent case of Tooth v HMRC [2018] UKUT 38. In that case, Mr Tooth participated in the same planning arrangements as those considered in Cotter v HMRC [2013] UKSC 69. Because of the way in which Mr Tooth had completed his return, the Supreme Court's decision in Cotter meant that HMRC had not used the correct enquiry powers in his particular case and therefore had to close its enquiry into his claim without any adjustment. HMRC then decided to issue a discovery assessment.

HMRC argued that it had discovered that there was an insufficiency in Mr Tooth’s return in 2009, when it opened an enquiry into the claim. However, the discovery assessment was not issued until 2014. The FTT had found as a fact that the officer made a discovery in 2014 when issuing the assessment, but in the view of the UT that discovery was the same discovery that had been made in 2009 when the enquiry was opened. The UT, in overturning the FTT's decision on that point, was very clear that the same discovery could not be made twice, even by two different HMRC officers. As the discovery had been made in 2009, then it was clearly stale by 2014. The UT said: 

"We entirely agree with the Upper Tribunal in Charlton that on making a discovery, HMRC must act expeditiously in issuing an assessment. If, to use the words of Charlton, an officer has made a discovery, then any assessment must be issued whilst the discovery is “new”…

… Here, HMRC discovered the insufficiency in 2009. It was incumbent on HMRC, at that stage, to decide what to do consequent upon this discovery. It was not open to HMRC without more to make a discovery of insufficiency and then to “park” the question of issuing a TMA s 29 assessment until a later date."

In Mr Beagle's appeal, the UT said that it could see force in the submission that the concept of 'newness' involved in a discovery relates simply to the nature of the discovery at the time at which it is made. It was not convinced that Pattullo was wrongly decided, particularly given the earlier obiter comments to similar effect in Charlton. It therefore concluded that the concept of staleness was relevant to the principles governing discovery. 

What was the date of discovery in Mr Beagles' case?

It was common ground that HMRC had to surmount a relatively low threshold to establish that it had made a discovery. The relevant officer had to have concluded or "found out" from the evidence before him that there was an insufficiency in the return and that conclusion had to be "new". The officer had to believe that the available information pointed to a discovery, and that belief had to be reasonable. This process was comprehensively covered in the recent case of Anderson v HMRC [2018] UKUT 159 (TCC). Anderson, and the other key discovery cases cited above, also suggest that a discovery involves both a subjective and an objective element. In other words, the relevant officer must believe that the available information points in the direction of a discovery and that belief must be a reasonable one for the officer to hold.

Against the factual findings made by the FTT, the UT found that the relevant officer had made a discovery of the insufficiency of Mr Beagles' return at the latest by 1 August 2005, when he had received advice from leading counsel and was prepared to challenge the arrangements before the courts. In the view of the UT, this was a reasonable belief. The FTT's finding that the discovery took place following the decision in Astall was not consistent with its own earlier findings.

While it might be possible for an officer to discover the same insufficiency more than once for different reasons, it was not possible to make the same discovery twice for the same reasons. 

Was the discovery stale when the assessment was made? 

In Pattullo, the UT suggested that a discovery would become stale on any view after a period of 18 months. In Charlton, the UT suggested that a delay of three or four months to accommodate the final determination of another appeal which was material to liability would not cause a discovery to lose its newness. In the instant case, there was a delay of nearly two and a half years before the assessment was issued. There was no ongoing litigation at the time of the discovery. The UT recognised that HMRC had referred to the prospect of a discovery assessment in correspondence, but it could not use such a notification to justify leaving a matter open without taking any action for an unlimited period. The UT therefore concluded that the delay was material, and the discovery had lost its quality of newness by the time of the assessment. Accordingly, the assessment was invalid.

This conclusion was sufficient to dispose of the appeal in Mr Beagles' favour, however, the UT went on to consider the remaining points that had been argued before it. 

Should HMRC have been reasonably aware of the insufficiency in the return? 

Section 29(3), TMA, provides that a taxpayer who has filed a self-assessment return can only be assessed under a discovery assessment if one of two conditions is met. The first condition (fraudulent or negligent conduct on the part of the taxpayer) is found in section 29(4). That condition was not relevant in Mr Beagles' case. The second condition, found in section 29(5), requires that a hypothetical HMRC officer could not have been reasonably expected to have been aware of the insufficiency of tax from the information made available to him or her at the relevant time. In Mr Beagles' case, the relevant cut-off time was the time at which the officer ceased to be entitled to enquire into his return. For the purpose of this condition, the information that is treated as made available to the hypothetical HMRC officer is set out in section 29(6), namely, the information that is contained in the return, or accompanying documents provided by the taxpayer (or information the existence of which and the relevance of which as regards the insufficiency of tax could reasonably be expected to be inferred by the officer from the return and any accompanying documents).

The leading cases on the application of section 29(5) are Hankinson v HMRC [2012] STC 485, Lansdowne Partners LLP v HMRC [2012] STC 544 and Sanderson v HMRC [2016] STC 638, and the following can be distilled from these cases:

(1) The test in section 29(5) is applied by reference to a hypothetical HMRC officer not the actual officer in the case. The officer is regarded as having the characteristics of an officer of general competence, knowledge or skill, which include a reasonable knowledge and understanding of the law;

(2) The test requires the court or tribunal to identify the information that is treated by section 29(6) as available to the hypothetical officer at the relevant time and determine whether, on the basis of that information, the hypothetical officer applying that level of knowledge and skill could not have been reasonably expected to be aware of the insufficiency;

(3) The hypothetical officer is expected to apply his knowledge of the law to the facts disclosed to form a view as to whether or not an insufficiency exists. Importantly, the test does not require that the actual insufficiency is identified on the face of the return;

(4) The hypothetical officer must be aware of the actual insufficiency from the information that is treated as available by section 29(6). The information need not be sufficient to enable HMRC to prove its case, but it must be more than would prompt the hypothetical officer to raise an enquiry;

(5) The level of awareness is a question of judgment. The information made available must justify raising the additional assessment, or be sufficient to enable HMRC to make a decision whether to raise an additional assessment.

The UT, agreeing with the FTT (albeit for different reasons) held that the HMRC officer could not have been expected to be aware of the insufficiency of tax before the cut-off date, when he ceased to be entitled to enquire into Mr Beagles' return.

Section 29(6)

Mr Beagle's final ground of appeal was that the FTT erred in its approach to section 29(6), in particular, in relation to whether the existence and relevance of a reason for the 'market change condition' could reasonably be inferred for the purposes of the condition in section 29(5). The information that is treated as available to the officer is the information that is contained in the return or accompanying documents provided by the taxpayer. This is extended by section 29(6)(d) to information the existence of which, and the relevance of which as regards the insufficiency of tax, could reasonably be expected to be inferred by the officer from the return and any accompanying documents. The UT concluded that the 'market change condition' was too vague to be able to be inferred from the information available to the officer. It added that, in any event, there had been important judicial development on the 'Ramsay' principle in Barclays Mercantile Business Finance Ltd v Mawson [2005] STC 1 and Scottish Provident Institution v Inland Revenue Commissioners [2005] STC 15, cases decided after Mr Beagle's return was submitted to HMRC, that would have altered or impacted upon the hypothetical officer's understanding. The UT therefore found for HMRC on this point.  


Most readers would accept that HMRC requires powers to enable it, in certain circumstances, to issue assessments to recover tax where the enquiry window has closed, most obviously in the case of deliberate understatement. However, there is a widespread feeling amongst the tax community that the finality which the self-assessment regime was intended to provide is being undermined by the manner in which HMRC seeks to apply its discovery powers. This decision serves as a warning to HMRC that taxpayers cannot be left in an indefinite state of uncertainty as to whether a discovery assessment will be issued. 

Whilst providing some clarity in this area of the law, Mr Beagle's case still leaves open the important question of how long a gap can be permitted between the making of a discovery and the issue of an assessment. In Tooth, the gap was five years and in Mr Beagles' case it was two-and-a-half years. Both were found to be too long. In Gakhall v HMRC [2016] UKFTT 356 (TC), the delay was over three months and that was considered by the FTT not to be too long.  Whether a discovery has become stale will of course turn on the facts of each case. Where, for example, HMRC has not acted on a discovery for reasonable and legitimate reasons, for example, because it is awaiting the outcome of a related case, a longer period of time is likely to be permitted by the courts before the discovery becomes stale. If, however, HMRC treats the discovery process as an open-ended period of enquiry, without having regard to the need to raise an assessment to alert the taxpayer to that discovery, it is likely that the discovery will become stale within a matter of months. 

It is understood that HMRC are extremely concerned by the concept of staleness in the context of discovery assessments. HMRC have appealed this decision to the Court of Appeal where it will no doubt argue that there is no such concept. Given the importance of this issue, a decision from the Court of Appeal clarifying the position and providing some certainty would be welcome. This will not be the only important case on discovery to reach the Court of Appeal this year. The decision in Tooth will be heard on appeal next month.

A copy of the decision can be viewed here

Stay connected and subscribe to our latest insights and views 

Subscribe Here