This website uses cookies

This website uses cookies to ensure you get the best experience. By using our website, you agree to our Privacy Policy

Andrew Park

Tax Partner, Price Bailey

Quotation Marks
The tax saga of John Hargreaves certainly got noticed. Few individuals receive tax assessments for £84m of tax

A taxing battle: Hargreaves v HMRC [2022] UKUT 34 (TCC)

A taxing battle: Hargreaves v HMRC [2022] UKUT 34 (TCC)


Andrew Park critically reviews the legal battle between Matalan founder Hargreaves and HMRC

The tax saga of John Hargreaves certainly got noticed. Few individuals receive tax assessments for £84m of tax. Fewer still succeed in overturning such vast assessments at the First-Tier Tax Tribunal (FTT). However, in the interim between the FTT decision and the appeal to the Upper Tribunal – J Hargreaves v HMRC [2022] UKUT 34 (TCC) - the legal ground shifted beneath the taxpayer’s feet.

What were the facts?

The entrepreneur behind Matalan had sold c.£231m of shares in the company in May 2000 – at a time when he considered himself to have recently become a non-UK resident denizen of Monaco. Most of the potential liability – some £80m of it – was to Capital Gains Tax and, since he did not believe he had such a liability and had no unrelated gains to declare, he did not complete the CGT pages of his UK self-assessment tax return for the year.

Mr Hargreaves’s 2000/01 tax return was not picked up for enquiry by HMRC within the statutory 12-month window after he filed the return on 31 January 2002. Accordingly, when it later came to HMRC’s attention there was a potential deficiency and it embarked on an investigation, HMRC was required to raise a 'discovery assessment'. When HMRC finally did so on 9 January 2007 it was outside the statutory four-year window after the end of the tax year in question to raise a discovery assessment on an 'innocent error' and relied instead on being within the six-year limit for raising a discovery assessment on an error brought about carelessly.

What was his case?

Mr Hargreaves’s appeal to the FTT in 2019 was multi-pronged. His leading counsel argued:

  1. 'Staleness point' – any discovery made by HMRC was 'stale' by the time HMRC finally acted on it to raise a discovery assessment because HMRC had waited some three years after the potential deficiency came to its attention. In any event, no discovery was made, because dependent upon the negligence, information and practice conditions in s29 Taxes Management Act 1970 (TMA 1970): 
  2. 'Negligence condition' – the situation was not brought about by the negligent conduct either of Mr Hargreaves or PwC – his then agent.
  3. “Information condition' – a discovery could not in any case be made because HMRC had all relevant information in the submitted tax return, had its opportunity to open an enquiry within the 12-month window and failed to do so.
  4. 'Practice condition' – based on generally accepted practice, Mr Hargreaves was non-UK resident and non-UK ordinarily resident at the time he disposed of the shares.

What happened next?

The FTT allowed Mr Hargreaves’s appeal on the staleness point, but rejected his arguments no discovery had been made based on failure to meet the discovery conditions. In the FTT’s view a valid discovery was made but it was made at least three years before the assessment was raised. The FTT concluded: “the discovery has lost its quality of newness and become stale by the time the assessment was made. Accordingly, the assessment cannot stand.”

HMRC duly appealed the FTT decision on the ground the discovery was not stale, and Mr Hargreaves appealed on the other three grounds no discovery had been made lest HMRC succeed on the staleness point.

It was at this stage – between the FTT decision and the Upper Tribunal hearing the appeal – that the foundations of Mr Hargreaves’s initial success shifted decisively under his feet. The concept of ‘staleness’ in discovery cases was something which only started to emerge in 2008 when some tax tribunals began to accept the inference that for something to be a ‘discovery’ within the context of the longstanding legislation in TMA 1970 common-sense interpretation dictated it needed to be new and fresh – or else what was being discovered to justify raising an assessment outside the enquiry window.

Over time, a rule of thumb emerged that, if HMRC failed to raise a discovery assessment within anything approaching three years of making a discovery, then it had no valid fresh discovery with which to raise an assessment.

To many of us, this seemed an equitable and valuable taxpayer protection from being under investigation without end – with all the stress and professional costs and distraction from the rest of life this involves. However, it was not until 2021 a case finally came before the Supreme Court involving staleness.

Time for change?

The case was CRC v Tooth [2021] STC 1049 – and it brought the whole edifice crashing down. The Supreme Court decided the notion of staleness simply did not exist within the context of tax discoveries – it never had and why should it, given the Supreme Court considered existing taxpayer protections were enough.

At a stroke, the Tooth decision bound the Upper Tribunal and swept away the basis on which Mr Hargreaves had won at the FTT – accordingly, there was no option but to agree with HMRC that its appeal on the staleness point be allowed. This then left Mr Hargreaves with the three alternative fall-back arguments which had not previously convinced the FTT – in which regard:

  • The negligence condition - the Upper Tribunal decided the FTT was entitled to its conclusion there was a prima facie case Mr Hargreaves had been negligent in not seeking further advice from PwC. That then shifted the burden across to Mr Hargreaves to displace the prima facie case. At the time the return was submitted there was a respectable body of opinion to the effect Mr Hargreaves could have rightly claimed to be non-resident and the Upper Tribunal was not satisfied the FTT could fully support its findings on the Negligence Condition without properly considering what advice PwC might have given and whether PwC might have concluded non-residence had Mr Hargreaves sought further advice from them.
  • The information condition – the proposition all necessary information was available within the enquiry window was rejected because the Upper Tribunal found the taxpayer could not demonstrate a hypothetical HMRC officer would have awareness of an actual insufficiency during the normal enquiry window rather than, at best, a possible one. This was because the taxpayer had not completed the Capital Gains Tax pages of the tax return and was irrespective of whether the officer could have realised the taxpayer was UK resident.
  • The practice condition – this revisited some of the same key ground as in the pivotal Gaines-Cooper case – R (Gaines Cooper) v CRC [2011] UKSC 47 - that went to the Supreme Court in 2011. Mr Hargreaves attempted to distinguish from Gaines-Cooper in that his matter in 2000/01 predated what was settled at the Supreme Court for later years. However, the Upper Tribunal found practice in 2000/01 similarly required people to do more than merely mechanically comply with the day count rules in HMRC’s then IR20 guidance and required people to make a “clean break” or leave the UK for a clear and settled purpose as demonstrated by multi-factorial analysis.

Since the Upper Tribunal endorsed the FTT’s conclusions on the practice condition, a valid discovery assessment then just relied upon one of the two other two conditions being met. Since it agreed the information condition was met, the FTT’s error of law on the negligence condition had no material relevance on the ultimate decision about whether the discovery assessment has valid. Accordingly, the Upper Tribunal found for HMRC the assessment was valid and opted to remake the error it had identified in the FTT’s conclusions on the negligence condition.


Even if the outcome is unsurprising after the Supreme Court had swept away the staleness concept in Tooth, Hargreaves is an interesting and useful exploration of the bases required for HMRC to make valid discoveries. In this author’s experience, advisers normally focus closely on the characterisation of behaviours for the purposes of the discovery assessment time limits at s36 TMA 1970 and pay the explicit taxpayer protections provided by the validity conditions at s29 TMA 1970 rather less attention. This serves as a reminder the validity conditions must be explicitly and carefully considered whenever HMRC makes discovery assessments and a full review of the facts might raise some unexpected and nuanced questions. This case did not ultimately turn on whether the advisor might have reached the same conclusion had the taxpayer sought further advice but another one might.

Andrew Park is tax investigations partner at Andersen in the United Kingdom: