The following is based on an article first published in Tax Journal on 26 February 2016.
In its decision in David Stephen Sanderson v HMRC  EWCA Civ 19, the Court of Appeal dismissed Mr Sanderson’s appeal against decisions of the FTT and UT and confirmed that HMRC was able to make use of the discovery assessment provisions, contained in section 29 TMA 1970, in order to assess him to tax where the officer could not reasonably be expected to know there was an insufficiency of tax even where details of the pre-DOTAS scheme were disclosed in the relevant tax return.
Mr Sanderson was issued with a discovery assessment pursuant to section 29 TMA 1970 in relation to capital gains tax which arose in 1998-99, as a result of his participation in a tax planning scheme (the Scheme). The Scheme, which attempted to create capital losses, was successfully challenged by HMRC and failed to achieve the intended fiscal consequences.
Between 1999 and 2007, a specialist HMRC team consisting of members of Special Compliance Office (SCO) and Special Investigation Section carried out an in-depth investigation into the Scheme. In the days before the disclosure of tax avoidance schemes (DOTAS) regime, HMRC was obliged to identify taxpayers who had participated in the Scheme through a manual review of all tax returns which had been submitted to it in respect of the years concerned in which more than £200,000 had been claimed as a capital loss.
In July 1999, the Office of Supervision of Solicitors (OSS) provided a list to SCO, containing the names and addresses of individuals who had paid to acquire losses through the Scheme. The list included Mr Sanderson’s details. This information was recorded on an HMRC database consisting of individuals who were under investigation by SCO.
SCO proceeded to obtain Mr Sanderson’s file from his district tax office and reviewed it. Following this review, the file was returned to the district office, SCO noting that Mr Sanderson had failed to lodge his 1997-98 and 1998-99 tax returns. SCO requested that the returns be sent to them when they were received by the district office. This failed to happen.
Mr Sanderson’s tax return for 1998-99, due to be filed with HMRC by 31 January 2000, was eventually received by HMRC on 24 February 2003. This gave HMRC 14 months from receipt to issue an assessment to tax under the normal time limits.
In his return, in the “white space” further information section, Mr Sanderson gave a clear indication of the source of the losses and the name of the relevant trust. His return was not sent to SCO as had been requested and no formal enquiry was opened or assessment issued. HMRC admitted, during the hearings below, that if it had searched for Mr Sanderson’s file he would have been subjected to an enquiry under section 9A TMA 1970, within the relevant statutory time period.
In the meantime, HMRC issued a closure notice against the trustees of the Scheme in November 2003, which reduced the £1bn loss claim, which had been submitted by them, to nil. SCO then wrote to all of the taxpayers connected to the Scheme, including Mr Sanderson, indicating the terms of a settlement offer. However, no enquiry had been opened into Mr Sanderson’s return and he did not respond to this offer. Finally, in January 2005, 12 months later and two years after submitting his return, HMRC made its purported “discovery” and issued a discovery assessment to Mr Sanderson. Mr Sanderson challenged the validity of the discovery assessment and appealed to the FTT. He was unsuccessful both before the FTT and on appeal before the UT. He appealed to the Court of Appeal.
Court of Appeal decision
By the time the matter reached the Court of Appeal, the issue was whether the second condition imposed by section 29(5) TMA 1970 for the exercise by HMRC of the power to issue a discovery assessment had been satisfied. So far as relevant, section 29 provided at the relevant time as follows:
“(1) If an officer of the Board or the Board discover, as regards any person (the taxpayer) and a year of assessment –
- that any income which ought to have been assessed to income tax, or chargeable gains which ought to have been assessed to capital gains tax, have not been assessed, or
- that an assessment to tax is or has become insufficient, or
- that any relief which has been given is or has become excessive,
the officer … may, subject to subsections (2) and (3) below, make an assessment in the amount, or any further amount, which ought in his … opinion to be charged in order to make good to the Crown the loss of tax.
(3) Where the taxpayer has made and delivered a return under section 8 or 8A of this Act in respect of the relevant year of assessment, he shall not be assessed under subsection (1) above —
- in respect of the year of assessment mentioned in that subsection; and
- in the same capacity as that in which he made and delivered the return,
unless one of the two conditions mentioned below is fulfilled.
(5) The second condition is that at the time when an officer of the Board —
- ceased to be entitled to give notice of his intention to enquire into the taxpayer’s return under section 8 or 8A of this Act in respect of the relevant year of assessment; or
- informed the taxpayer that he had completed his enquiries into that return,
the officer could not have been reasonably expected, on the basis of the information made available to him before that time, to be aware of the situation mentioned in subsection (1) above.”
In order to determine what could be reasonably expected, the courts have deemed the analysis to centre on a “hypothetical” rather than actual officer, of general competence, knowledge or skill (HMRC v Lansdowne Partners LLP  STC 544).
In determining what it was reasonable for the officer to know, the Court of Appeal construed the bounds of relevant information narrowly. Mr Sanderson’s file had been obtained by
SCO, and they had asked for Mr Sanderson’s tax return to be sent to them once received by his district tax office. Further, SCO had written to Mr Sanderson regarding the settlement opportunity. In addition, Mr Sanderson had made the following extensive disclosure in his return:
European average rate option (Trade No. 82831)
I am entitled to the loss of £1,825,663 by virtue of the provisions of TCGA 1992 s.71(2). The loss is part of a loss of £1,000,000,000, which accrued to the Trustees of the Castle Trust on 8th April 1997, on the disposal of a European Average rate Option (Trade No. 82831) relating to shares in Deutsche Telecom.
Beneficial interest in the Castle Trust
On 24th November 1998, I purchased for a fee (part of which is contingently payable) from the Trustees of the Charter Trust 2.273% of their beneficial interest in the Trust Fund of the Castle Trust. The interest determined on 25th November 1998, when I became absolutely entitled to receive from the Trustees of the Castle Trust the sum of £16.04.”
In order to ascertain whether there was sufficient information available to HMRC, the test is whether there was enough information available for the officer to decide to raise an assessment. HMRC argued, and the Court of Appeal accepted, that it was not enough for the disclosure made by the taxpayer to simply cause an officer to ask further questions. The officer must “be made aware of an actual insufficiency” of tax (Langham v Veltema  STC 544).
Mr Sanderson argued, following the logic in Charlton v HMRC  STC 866, that the information contained in the disclosure which was contained in his return would have been sufficient for the hypothetical officer to infer that he was one of a number of individuals participating in a tax avoidance scheme. This, however, was rejected by the Court on the basis that unlike in Charlton where the disclosure of the Scheme Reference Number would mean that a form AAG1, containing full details of the scheme, had been provided to HMRC by the scheme promoter, there was no such number or form in this case.
The Court concluded that it would have been “entirely speculative” for the hypothetical officer to conclude that another part of HMRC may have information on the Scheme. Mr Sanderson’s appeal was therefore dismissed.
The unusual facts of Sanderson may be sufficient to render it a peculiarity, as similar arrangements would now be notifiable under the DOTAS regime and in such circumstances the taxpayer would be able to rely upon the Charlton decision to prevent HMRC from issuing a discovery assessment. However, this is the latest in a number of recent discovery cases in which the courts have attributed to the hypothetical officer very little knowledge, notwithstanding extensive information having been made available to HMRC.
In this instance, Mr Sanderson had provided details of the Scheme in his tax return. His name was included on a list of people who had participated in the Scheme which had been supplied to SCO in 1999. His file was located and obtained by SCO, who issued instructions that his tax return should be forwarded to them on receipt by the district. Given this background, one would have thought the hypothetical officer could reasonably be expected to have been aware that there were chargeable gains which ought to have been assessed to capital gains tax.
In dismissing Mr Sanderson’s appeal, the Court of Appeal has further eroded the level of knowledge which may be imputed to the “hypothetical officer”. It would appear that disclosing details of a tax avoidance scheme is not, of itself, sufficient to prevent HMRC making a discovery assessment pursuant to section 29 TMA 1970. In addition, taxpayers cannot assume that the hypothetical officer will be deemed to be aware of information held by others within HMRC.
The “discovery” in this case appears to have been the realisation by HMRC that due to an oversight on its part, the investigation into Mr Sanderson had not been pursued. Without the decision in Charlton (which was expressly endorsement by the Court of Appeal), HMRC would no doubt seek to argue in similar circumstances that the disclosure of a Scheme Reference Number would be insufficient to prevent it from issuing a discovery assessment. Were it to be successful with such an argument, section 29 would be rendered all but meaningless.
The Sanderson decision is available to view here.