Discovery And All That

 In Discovery, Tax Hell

As is said on the main discovery page of this website: For HMRC the term ‘discovery’ is very serious. HMRC have a limited window to launch an investigation and they would dearly like to open up earlier years to have a good dig around in your past, but they can only go this once they have made a discovery – so you can see why it’s so important to them.

Here Ray McCann, a partner at Joseph Hage Aaronson digs down deeper into what a discovery is, what the implications are and how to avoid it. If you are new to discovery as a subject I suggest you go here. But if you know the basics read on.

The interaction between discovery, time limits and DOTAS


  • HMRC’s efforts to recoup back taxes are increasing.
  • There are strong statutory protections intended to give taxpayers certainty.
  • HMRC may assert careless or deliberate behaviour where time limits are involved.
  • Discovery may, in effect, be no more than a suspicion on HMRC’s part.

HMRC have ramped up their attack on tax avoidance and taken a keener interest in the filing positions of high net worth individuals and large corporates. As a result, a detailed knowledge of when the department can pursue a past liability or make an “assessment where a loss of tax is discovered” has become a must-have skill for advisers. Knowledge of those provisions is not enough, however, because often HMRC are pursuing unpaid tax relating to years of assessment long since thought closed so a strong understanding of time limits and, crucially, culpability are also essential skills.

In broad terms, the statutory provisions that allow for an assessment in respect of the past are generally referred to as “discovery” and, until the administration of companies and individuals was separated, the statutory provisions were contained within TMA 1970, s 29. Today s 29 applies to income tax and capital gains tax only; FA 1998, Sch 18 para 41(2) applies to companies. As the provisions are broadly similar, for ease TMA 1970, s 29 should be regarded as encompassing all such provisions as they apply to individuals, partnerships and companies.


Section 29 provides that HMRC can assess a loss of tax arising from the fact that there has been:
  • no assessment;
  • an inadequate assessment; or
  • excessive relief.
The powers given to HMRC are broad but not unlimited. There are important safeguards for taxpayers which include:
  • no assessment can be made if the return was filed on the basis of a “practice generally prevailing”; and
  • one of two conditions must exist before an assessment can be made:
    • either the understatement was brought about through carelessness or because it was deliberate; or
    • at the time any enquiry time limit expired or any enquiry was closed, HMRC could not reasonably be expected to be aware of the understatement based on the information provided by the taxpayer at that time.

Most disputes involving HMRC powers to make “discovery assessment” entail reasonable expectation, although the question of whether a prevailing practice exists has also been relevant in the context of employee benefit trusts.

What, though, is the meaning of “discover”? It may be worth challenging HMRC on the basis that no discovery has been made but, in my view, other than in the most exceptional situation this will prove a fruitless exercise.

It is clear that a mere change of view, even by the same inspector, can amount to a discovery. In any event, ultimately the tribunals are likely to side with HMRC on this issue if only because “discover” has, over the years, been afforded such a broad meaning.

There are good public policy reasons why this should be so. Despite the spooky little HMRC eyes that stare at you from billboards, in reality the department looks at only a tiny proportion of the information that it receives. Of that it is mostly seen by only junior or inexperienced staff, to whom much of what appears in a self assessment will mean little. In many cases, by the time a specialist or more experienced inspector takes a look, the enquiry window will have closed.

The disclosure of tax avoidance schemes (DOTAS) regime and the requirement to “tick a box” if the transfer of assets rules were engaged are partly explained by that reality and if HMRC had to look at everything nothing would get done. However, as the Upper Tribunal’s decision in CRC v Charlton [2013] STC 866 made clear, even if DOTAS reference numbers are evident in the tax return HMRC might still fail to open an enquiry and later seek to rely on s 29.

Fair dealings

As a result, in most cases it will be more worthwhile to concentrate on whether HMRC can act on any discovery it claims to have made because it is here that the concept of “fair dealings” comes into play.

In short, fair dealings is a form of bargain. It involves the taxpayer taking reasonable care to make an honest self assessment that provides HMRC with appropriate information as to what is on his self assessment. In return the taxpayer obtains certainty that his tax position is final when an enquiry window closes or a closure notice is issued.

When tax planning is involved, HMRC have sought recently to deviate from this bargain. It is increasingly the case that they will pursue tax avoidance treating it is as tax evasion, including the use of code of practice 9 where time limit issues arise. HMRC are also more likely to assert that a taxpayer was careless even when professional advice has been taken.

It should not be a surprise if HMRC seek to leverage the fact that an appeal to the tribunal is costly, time-consuming and stressful and inevitably carries uncertainty especially when tax planning is involved.

Defending clients

How do you defend the client? It is critical to be clear about what you are dealing with. The temptation to fire off some robust detailed argument as to why HMRC are wrong must be resisted if only because, in so doing, you may provide HMRC with information that strengthens their position. Instead spend time with the client to ensure that the facts are clear. Since the issue may involve a transaction that is long forgotten, this may be difficult.

Once it is clear or reasonably clear that no tax evasion is involved, it is crucial to understand how HMRC are restrained by time limits. Absent careless or deliberate behaviour, they are out of time after four years.

As a result, when tax schemes are involved HMRC will increasingly assert that the taxpayer has been careless so that the time limit is increased to six years. However, this is still a relatively short period and, in a recent case, HMRC asserted that a 20-year limit applied “as the taxpayer had deliberately entered into a tax avoidance scheme”. For this reason it is necessary to understand the general principles under which an underpayment would be regarded as careless or deliberate.

Despite the view of some inspectors, a loss of tax is “deliberate” only if the taxpayer has done something bad that he knows is bad. As long as the taxpayer genuinely believed that the entries on the self-assessment return were correct, any understatement that may later be shown cannot be deliberate. Excluding deliberate understatement puts HMRC out of time after six years.


Whether the taxpayer was careless is more nuanced and taxpayers who do not seek the assistance of a reputable adviser greatly increase the risk that HMRC would take the view that any understatement resulted from careless behaviour.

If a taxpayer did take advice, consideration needs to be given to its quality. The “cowboy promoters”, as the government has labelled them, store up problems for their clients and for any adviser who acted as an introducer or intermediary. HMRC described many of the tax outcomes designed by promoters as too good to be true; this may be so but is ultimately irrelevant as to whether the planning meets the statute.

More importantly, too many clients swallowed what they were told about the merits of the scheme. To date, taxpayers have had mixed outcomes in resisting penalties imposed by HMRC since reliance on professional advice does not allow a taxpayer to abandon all personal diligence. Generally, advice given to a client by a reputable adviser should, even if the advice is later shown to be incorrect, protect the taxpayer from a discovery assessment outside the four-year limit and penalties.

White space

There are two other main issues to consider where tax schemes are involved: was the scheme within DOTAS and what, if any, entries appeared in the white space?

In light of the Charlton decision, when a DOTAS scheme reference number was correctly included on a self-assessment return, there should be no question of HMRC having a discovery position if they fail to open an enquiry.

A scheme reference number on the return will, however, be of little help if the taxpayer has been careless, where the tax scheme is shown to be fraudulent or where the taxpayer knew that it was suspect. HMRC are pursuing a number of tax schemes on these lines.

Then there is white space. At times the only difference between white space and outer space is that Albert Einstein could make sense of outer space. Some may disagree but, in general, white space entries will be of little protection if the explanation required to ensure that no discovery assessment could be made would be so extensive as to make an enquiry all but inevitable.

But white space was generally intended to show that HMRC could reasonably be expected to be aware of the understatement. This brings us to the aspect of discovery that has been so difficult over the years, whether HMRC should reasonably have been aware of the potential understatement?

Reasonably aware

To be precise, s 29 provides that no discovery assessment can be made unless the “[HMRC] officer could not have been reasonably expected, on the basis of the information made available to him […] be aware of the [understatement]”. It is frankly impossible to be precise as to where this line is and each case will turn on its own facts. But there are some general points that assist:
    • It is not necessary to argue that an officer looked at the return, so HMRC are not home and dry merely because the return was not examined.
    • Any information contained in the return for the year in question includes anything also contained in a return for the two immediately preceding years.
    • For this purpose the information is anything contained in a return or claim or any document, accounts or particulars provided during an enquiry.

The information does not need to amount to a clear statement that there is an understatement but it must be sufficient that a reasonable inference on the part of the officer would be that an understatement existed.

The hypothetical HMRC officer who is assumed to have looked at the information will also be regarded as having a skill level appropriate to the transaction in question.From this it is clear that, if HMRC have not been told anything, a defence against a discovery assessment will rely entirely on time limits. But if at least some information has been provided there is a chance of success. How big a chance will depend on the nature of the understatement.

In 2004 Langham v Veltema [2004] STC 544 raised considerable alarm in the profession leading to a suggestion that a taxpayer would be protected from a discovery assessment only where he flagged up to HMRC that an underpayment actually existed.In reality, the bar is nowhere near this high and not enough attention was paid to the fact that, in Veltema, the valuation was understated by about half. It is likely that, had that value (even though still below a properly considered value) been closer to the correct value, the outcome would have been different.

However, Veltema did make clear that, whatever else is required, HMRC’s ability, in effect, to look through walls is not required.

It also made clear that the onus is on the taxpayer to ensure that relevant information is included in the first place. But the “reasonable” inference test is important and, too often, HMRC seek to enforce discovery assessments when it should be clear that that it is not met.

No more than a suspicion

More recent cases are Hankinson v CRC (No 3) [2012] STC 485 and Charlton. The latter was a successful outcome for the taxpayer. It provides an excellent summary on how to approach the questions that arise when considering whether information provided with a return is sufficient for an officer to make a reasonable inference that there is a tax understatement.

Hankinson is also of some value in clarifying that HMRC do not need to be satisfied that all the requirements of s 29 are met before they can issue a discovery assessment and that most of the “machinery” in s 29 involves matters that are to be decided on appeal against the assessment.

In short, no one will defeat a discovery assessment by arguing that HMRC did not consider whether s 29(5) or s 29(6) are satisfied at the point the assessment is issued. The reason for this is straightforward: in almost all cases “discover” means “suspect”. Indeed the entire approach to discovery provision is easier if this substitution is made, since in many cases that is all HMRC has – a suspicion.

So where are we today? HMRC’s focus remains on tax schemes but their relentless quest for revenue will mean that discovery issues are increasing. The battleground will focus on what information was included with the return but, even where significant information was provided, it will not necessarily protect the taxpayer from a discovery assessment.

HMRC are pursuing many past liabilities in circumstances where the statutory protections afforded to the taxpayer should apply, so a robust defence should still be possible. Care is needed if there was a previous enquiry and regard should be had to professional conduct in relation to tax.

This feature originality appeared in Taxation magazine.

Recommended Posts

Leave a Comment


Start typing and press Enter to search