HMRC v Charlton, Corfield and Corfield [2012] UKUT 770 (TCC)

Two holdings with very different fates. Charlton’s ruling that a DOTAS scheme reference number on a return engages the s29(5) bar is one of the few reliable taxpayer wins in discovery law. Its suggestion that a discovery goes stale with time was buried by the Supreme Court in Tooth. Cite the first; never the second.

The Facts

The taxpayers participated in a marketed capital gains tax avoidance scheme which was subsequently found not to work. The arrangements fell within the disclosure of tax avoidance schemes rules in Part 7 of the Finance Act 2004.

The promoters accordingly notified HMRC of the scheme on form AAG1, and HMRC allocated it a scheme reference number. The promoters passed that number to the users, and the taxpayers entered it on their self-assessment returns in the box provided for the purpose, as the DOTAS rules require.

HMRC did not open enquiries into the returns within the s9A window. Once the window had closed, and once the scheme had been shown not to work, HMRC raised discovery assessments under s29 TMA 1970.

The Statutory Framework

Two provisions were in play:

• s29(1) TMA 1970: the requirement that an officer “discovers” an insufficiency. What does “discovery” require?
• s29(5) TMA 1970: the bar where, on the basis of the information made available before the enquiry window closed, an officer could reasonably have been expected to be aware of the insufficiency. The information pool is defined in s29(6), as explained in Langham v Veltema.

The Ratio Decidendi

Why the DOTAS holding matters

This is a rare instance of a taxpayer winning a s29(5) argument. The reasoning is compelling: HMRC designed the disclosure regime precisely so that it would know about avoidance arrangements. A taxpayer who complies with it has given HMRC exactly the notice the system was built to generate. Having received that notice, HMRC cannot say it could not reasonably have been expected to be aware of the insufficiency.

Contrast this with Veltema, where the taxpayer disclosed a valuation figure but nothing that alerted the officer to an insufficiency. A DOTAS number is different in kind: its entire function is to flag that the taxpayer is claiming a tax result HMRC is likely to dispute.

The standard of awareness

The Upper Tribunal also addressed how sure the hypothetical officer must be. The officer need not be aware of the insufficiency to the point of certainty; awareness that the assessment was insufficient, in the sense of being more likely than not, is enough. This is a taxpayer-friendly gloss on Veltema and remains good law.

The Staleness Passage: Obiter, and Now Disapproved

At paragraph [37] the Upper Tribunal said that a discovery must have “the quality of newness”, and went on to suggest in the latter part of that paragraph that a discovery could cease to be one through the passage of time, that it could go stale.

Practitioners built a substantial body of argument on that sentence between 2013 and 2021. It was applied and developed in Pattullo v HMRC [2016] STC 2043 and in a long line of FTT decisions.

The Supreme Court in Tooth held that there is “no place for the idea that a discovery which qualifies as such should cease to do so by the passage of time”, and expressly disapproved the latter part of paragraph [37] of Charlton. Because HMRC’s appeal in Tooth failed on the deliberate inaccuracy point, those remarks were strictly obiter, but in Harrison the Upper Tribunal held they were intended as binding general guidance and applied them.

The Line of Authority

• Cenlon Finance Co Ltd v Ellwood [1962] AC 782. The foundation. Discovery needs no new fact; what matters is that it newly appears to the officer that there has been an undercharge.
• Langham v Veltema [2004] EWCA Civ 193. The s29(5) test: the taxpayer must have clearly alerted the officer to the insufficiency.
• Sanderson v HMRC [2016] EWCA Civ 19. The Court of Appeal described the hypothetical officer as one of general competence, knowledge and skill, not expected to resolve difficult points of law, and confirmed that the taxpayer must have supplied information justifying the conclusion of an actual insufficiency.
• HMRC v Tooth [2021] UKSC 17. Discovery is subjective and officer-specific; there is no collective HMRC knowledge; staleness is not a concept known to the statute; and a return must be read as a whole.

Practitioner Application

Using the DOTAS holding

• Check whether a scheme reference number was entered. Where the client used a disclosed arrangement and reported the SRN on the return, Charlton gives a direct s29(5) argument. Retrieve the filed return and the AAG4 or equivalent.
• Establish what HMRC did with it. The SRN is fed into HMRC’s systems. Subject access and Freedom of Information requests can show when the return was risk-assessed and by whom.
• Fix the timing. The question is what was available before the enquiry window closed. Disclosure after that date does not assist.
• Do not overreach. The holding is about SRNs and the entries associated with them, not about avoidance generally. A scheme that fell outside DOTAS, or where the SRN was not reported, is back in Veltema territory.

Handling the staleness point

• Do not plead it. Delay alone no longer defeats a discovery assessment.
• Redirect the evidence. The timeline material gathered for a staleness argument is still valuable, for s29(5), for HMRC’s burden on careless or deliberate behaviour under s29(4), and where delay is genuinely egregious, for a public law challenge based on conspicuous unfairness.
• Watch for HMRC over-reading Tooth. The disapproval was confined to the temporal decay point. The rest of Charlton, including the standard of awareness and the DOTAS holding, stands.

Frequently Asked Questions