Another discovery appeal bites the dust

The decision in the case of Sanderson v Revenue and Customs Commissioners [2013] (UKUT0623, published on 17 December 2013) represents yet another victory for HMRC in the long-running battle over discovery assessments.

The case concerns the 1998/99 tax return submitted by Mr Sanderson which included chargeable gains of nearly £1.8 million, but set against this, capital losses of over £2 million.  The losses had arisen as a result of Mr Sanderson’s “Beneficial Interest in the Castle Trust”, as indicated in a ‘white space’ disclosure.  The Castle Trust Scheme had become the subject of an Inland Revenue investigation in mid-1999 and a list of users which included Mr Sanderson was provided to the team of specialist investigators responsible for looking at the scheme.

Mr Sanderson’s 1998/99 return was filed late on 24 February 2003, thus extending the enquiry window until 30 April 2004.  During this period, in November 2003, HMRC issued a closure notice to the trustees of the Castle Scheme reducing the loss claim to nil.  However, Mr Sanderson’s return was not reviewed by the team until October 2004, by which point it had been concluded by the IR that the scheme was ineffective and this had been communicated to the scheme provider.  A ‘discovery’ assessment in respect of the 1998/99 tax return was subsequently made in January 2005, removing the capital losses associated with the Castle Trust Scheme.

The appeal against the assessment was heard at and subsequently dismissed by the First Tier Tribunal following which Mr Sanderson sought recourse to the Upper Tribunal where the same issues were presented, these being:

  • Whether a ‘discovery’ had been made for the purposes of Section 29(1) TMA 1970?
  • Whether there had been negligent conduct on the part of the appellant or another person acting on his behalf?
  • Whether the inspector could have been reasonably aware of an insufficiency to tax from the information made available during the period in which the enquiry window for the 1998/99 return remained open?


Previous decisions around the definition of ‘discovery’ for the purposes of Section 29(1) TMA 1970 have indicated that this has a far wider reach than simply discovering something new, and the Upper Tribunal clearly took this into consideration in deciding that a valid discovery had in fact been made.  They were satisfied that it had “newly appeared” (as quoted in the Charlton case) to the officer in October 2004 that there had been an insufficiency to tax.  That the Inspector had previously been aware as to the ineffectiveness of the Castle Trust Scheme and Mr Sanderson’s involvement, did not preclude him from ‘discovering’ the insufficiency in October 2004 when the return was first reviewed.


The Upper Tribunal followed the decision of the FTT with regard to whether the taxpayer, or those acting on his behalf, had been negligent.  HMRC contended that there had been negligence on the basis that the taxpayer, or rather his advisors, had failed to make an amendment to the 1998/99 return after it was established that the capital loss scheme was shown to be ineffective, in February 2004.  The UT judge dismissed this cross-appeal from HMRC, on the grounds that “there is no statutory provision imposing an obligation on a taxpayer to tell HMRC about something in a filed return that he subsequently finds to be erroneous. The most that can be said is that a failure to correct an error can potentially affect a taxpayer’s exposure to penalties”.

Information made available?

Finally, the tribunal found that the tax return on its own did not provide enough information to alert the officer to an insufficiency.  The appellant contended that HMRC were barred from raising the discovery assessment on the basis that Section 29(6)(d)(i) TMA 1970 applied and the Inspector should have been able to infer from the information regarding the Castle Trust Scheme supplied in the tax return, the existence of other information concerning the scheme, namely that HMRC considered it to be ineffective.

In pronouncing its decision, the tribunal made reference to Charlton:

It seems to me that, for section 29(6)(d)(i) to apply, the hypothetical officer must be able to infer (and not just guess at) the existence of specific information (albeit not its actual content) of definite relevance to the existence of an insufficiency. That was the case in Charlton: the fact that there was an SRN inevitably meant that a form AAG1 had been lodged, and that form was bound to contain information about the scheme in question. In contrast, an officer considering Mr Sanderson’s 1998-1999 return could, I think, have done no more than surmise that HMRC would somewhere have other information about the Castle Trust and that, if it did, it could cast light on Mr Sanderson’s loss claim. In my view, that is not good enough for the purposes of section 29(6)(d)(i).

The rulings regarding discovery, negligence and information made available included within this case make it very interesting reading.  It is also an indication that the room for further debate around the issue of what information is made available, and more specifically, what information may be inferred from other sources, is substantial.

Should you have a query regarding this case, or have an ongoing issue regarding a discovery assessment, please contact the TaxDesk on 0845 4900 509 and ask for John Hood.