The law relating to discovery assessments seems to provide an inexhaustible supply of ‘interesting’ aspects – and an equally endless supply of tribunal cases. There have been some recent decisions about whether inattention on the part of HMRC can cause a discovery to ‘lose its newness’ and become stale, preventing HMRC from raising an assessment even before the end of the limitation period. But even more interesting, is the case of G Clark v HMRC [2017] UKFTT 392 (reported in Tax Journal, 26 May 2017).
What happened here is that HMRC issued a discovery assessment on Mr Clark because it took the view that there was a liability to tax which had not been assessed. However, in due course, it concluded that there were actually no grounds for the loss of tax it thought existed.
However, HMRC found another area where it thought there was a loss of tax and said that the earlier discovery assessment was sufficient to cover this as well.
Er, no – said Mr Clark. The awareness of the hypothetical HMRC officer of a possible loss of tax can only refer to that loss of tax.
The tribunal explained that for the purposes of determining whether a discovery assessment has been validly made, it is necessary to identify the loss of tax that has been asserted by the officer; and then to test whether the further conditions in TMA 1970 s 29 have been met by reference to that loss of tax. That sounds pretty much what Mr Clark was arguing.
However, the tribunal went on to say that it is entitled to apply the law to the facts as it finds them and to form its own view. It does not have to confine itself to the reasons advanced by HMRC, nor indeed is it constrained by the arguments of either party, before or at any stage in the proceedings. In this case, there was a discovery and there was an assessment, and the tribunal saw no reason why they should not be linked up.
That is all very well, but surely this means that HMRC can issue a discovery assessment and, if it is found to be wrong or unfounded, it can thrash around indefinitely to find something (anything) to which the assessment might apply.
That must be a recipe for abuse. It enables HMRC to keep the time limit for an assessment open indefinitely, simply by raising a discovery assessment which is groundless in the hope that something will turn up eventually.
The tribunal did suggest that there was a limitation on the scope of the discovery assessment. It said that it cannot extend to a loss of tax for which no valid assessment was capable of being made by reason of a specific prohibition under s 29 – for example, if HMRC was already aware of the relevant facts. Unfortunately, a review of earlier cases on the subject indicates that such limitation is largely illusory.
Accordingly, it is likely to become seriously difficult to close down a discovery assessment in future because, while it is alive, it can cover anything.
I wonder what the will of Parliament might have been on this point.
The law relating to discovery assessments seems to provide an inexhaustible supply of ‘interesting’ aspects – and an equally endless supply of tribunal cases. There have been some recent decisions about whether inattention on the part of HMRC can cause a discovery to ‘lose its newness’ and become stale, preventing HMRC from raising an assessment even before the end of the limitation period. But even more interesting, is the case of G Clark v HMRC [2017] UKFTT 392 (reported in Tax Journal, 26 May 2017).
What happened here is that HMRC issued a discovery assessment on Mr Clark because it took the view that there was a liability to tax which had not been assessed. However, in due course, it concluded that there were actually no grounds for the loss of tax it thought existed.
However, HMRC found another area where it thought there was a loss of tax and said that the earlier discovery assessment was sufficient to cover this as well.
Er, no – said Mr Clark. The awareness of the hypothetical HMRC officer of a possible loss of tax can only refer to that loss of tax.
The tribunal explained that for the purposes of determining whether a discovery assessment has been validly made, it is necessary to identify the loss of tax that has been asserted by the officer; and then to test whether the further conditions in TMA 1970 s 29 have been met by reference to that loss of tax. That sounds pretty much what Mr Clark was arguing.
However, the tribunal went on to say that it is entitled to apply the law to the facts as it finds them and to form its own view. It does not have to confine itself to the reasons advanced by HMRC, nor indeed is it constrained by the arguments of either party, before or at any stage in the proceedings. In this case, there was a discovery and there was an assessment, and the tribunal saw no reason why they should not be linked up.
That is all very well, but surely this means that HMRC can issue a discovery assessment and, if it is found to be wrong or unfounded, it can thrash around indefinitely to find something (anything) to which the assessment might apply.
That must be a recipe for abuse. It enables HMRC to keep the time limit for an assessment open indefinitely, simply by raising a discovery assessment which is groundless in the hope that something will turn up eventually.
The tribunal did suggest that there was a limitation on the scope of the discovery assessment. It said that it cannot extend to a loss of tax for which no valid assessment was capable of being made by reason of a specific prohibition under s 29 – for example, if HMRC was already aware of the relevant facts. Unfortunately, a review of earlier cases on the subject indicates that such limitation is largely illusory.
Accordingly, it is likely to become seriously difficult to close down a discovery assessment in future because, while it is alive, it can cover anything.
I wonder what the will of Parliament might have been on this point.
