The 2019 loan charge has attracted much attention and been the subject of detailed commentary in Tax Journal already. Given this and the ongoing public and parliamentary debate about the legitimacy of the charge, we focus here on the broader questions around statutory time limits that apply to HMRC’s ability to investigate a taxpayer’s affairs and raise assessments of tax. Central among these recently is the balance between giving taxpayers certainty as to what tax is due and when, and the pressure on HMRC to tackle perceived avoidance or underpayment with new powers in what is undoubtedly now a tougher socio-political environment for taxpayers. Key battlegrounds here are:
Before testing those battlegrounds, it’s worth recalling the ‘normal’ rules for opening enquiries into returns and/or making discovery assessments or determinations. These, of course, differ depending on what tax is in scope. The EU state aid rules also have their own (usually longer) limitation period and most readers will have had to tackle the patchwork quilt of limitation periods in other jurisdictions.
For present purposes, we focus on UK income tax, capital gains tax and corporation tax. The rules for these broadly mirror one another and we expect they will be most relevant to readers.
The window for HMRC to open an enquiry into a taxpayer’s return is short. For income tax and capital gains tax, HMRC usually has 12 months from the date the return was delivered to open an enquiry (TMA 1970 s 9A(2)). For group companies other than Companies Act 2006 small group companies, a similar 12 month deadline operates but this starts on the filing date, not the date the return is delivered (FA 1998 Sch 18 para 24). In both cases, if the return is delivered late or is amended, the enquiry window is extended. Importantly, HMRC needs to give notice of its intention to enquire. This is a factual question but is a crucial gateway to imposing further liability on taxpayers.
Where the statutory time limits for opening an enquiry have passed or an enquiry into a particular year has been closed, HMRC can still recover lost tax in those years by making a discovery assessment (or determination).
To issue a discovery assessment, HMRC must have first discovered a loss of tax, being:
a relief that has been given which is or has become excessive (TMA 1970 s 29 and FA 1998 Sch 18 para 41).
In addition, (i) that loss of tax must have arisen as a result of careless or deliberate behaviour by the taxpayer or their agent; or, (ii) at the time the enquiry window closed (or HMRC closed its enquiry into the relevant year), HMRC could not reasonably have been expected to be aware of the loss of tax – in other words, there was an error despite reasonable care being taken (TMA 1970 s 29(4) and (5) and FA 1998 Sch 18 paras 43 and 44).
The rules effectively create a three tier system. The ‘default’ time limit is four years from the end of the relevant year or accounting period in question (TMA 1970 s 34 and 34A and FA 1998 Sch 18 para 46(1)), but this applies only if there’s an error despite taking reasonable care. For careless and deliberate conduct, HMRC can go further back in time. HMRC has six years from the end of the relevant year or accounting period if the loss of tax is the result of the taxpayer’s (or their agent’s) careless behaviour and 20 years if the behaviour was deliberate (TMA 1970 s 36(1) and (1A) and FA 1998 Sch 18 para 46). HMRC can also go back 20 years if the loss of tax resulted from the taxpayer’s failure to notify HMRC of their chargeability to tax, or was attributable to arrangements which should have been notified under disclosure of tax avoidance schemes (DOTAS) or promoters of tax avoidance schemes (POTAS) (but were not) (TMA 1970 s 36(1A) and FA 1998 Sch 18 para 46).
Even if that should be relatively straightforward, recent developments have highlighted three particular complications.
Discoveries can become ‘stale’ if there is undue delay between the discovery and the assessment itself (even if it was made in time), in which case the assessment is invalidated. In Pattullo v HMRC [2016] STC 2043, the Upper Tribunal concluded that HMRC must act on its discovery ‘while it remains fresh (or before it becomes stale)’. In the recent case of Beagles v HMRC [2018] UKUT 380, the Upper Tribunal followed Pattullo. Beagles will be reviewed by the Court of Appeal, though taxpayers should have some comfort that HMRC cannot sit on newly ‘discovered’ information indefinitely.
FA 2019 introduced a new 12 year time limit (at TMA 1970 s 36A) for assessing lost income tax or capital gains tax involving offshore matters or transfers. For careless errors and those made despite taking reasonable care, the provisions increase the number of tax years potentially subject to assessment prospectively, one year at a time, until the period that HMRC can assess reaches 12 years. The existing 20 year time limit for deliberate behaviour remains.
The policy behind the new limits is that, in offshore cases, it will often take HMRC longer to investigate or longer for information to come to light under international information exchanges than in purely domestic cases. However, with some limited safeguards for information received by HMRC before expiry of the ‘old’ time limits, the period of uncertainty for taxpayers with offshore assets or income is materially increased.
Taxpayers with rulings or other arrangements potentially vulnerable to investigation under the EU state aid rules are similarly subject to longer term exposure. The general rule is that there is a ten year time limit, meaning that arrangements agreed with HMRC (or indeed positively endorsed by it in the case of the CFC finance company partial exemption) remain vulnerable to reassessment for some time. Taxpayers, however, have a shorter six year limit to claim against HMRC for breach of EU law if there has been an unlawful grant of state aid. Any challenge by way of judicial review will need to be made quickly, within three months of the relevant EC decision determining that there has been illegal state aid.
A key challenge levelled at the 2019 loan charge and other recent high-profile taxes, e.g. the diverted profits tax (DPT), is that they are retrospective. It’s important though to distinguish between the technical analysis and the substantive application to taxpayers.
In that regard, HM Treasury is right to conclude in its report published on 26 March 2019 that the loan charge legislation is not retrospective, as it is a ‘tax charge to outstanding disguised remuneration loan balances as at 5 April 2019’ and ‘does not change the tax position of any previous year or the tax treatment of any historic transaction’. Likewise, the DPT rules in FA 2015 (amended by FA 2019) introduced a new 25% tax on diverted profits with effect from 1 April 2015 (with higher rates for adjusted ringfenced oil profits and bank profits). Profits arising before 1 April 2015 were, in principle, unaffected by the new regime.
However, it’s likely that both the taxpayer facing a tax liability in the 2018/19 tax year in respect of up to 20 years’ worth of historic earnings or a multinational required to fund DPT charging notices upfront in respect of transactions entered into decades previously will feel that the substance of the tax charge is retrospective.
In the case of the loan charge, taxpayers who disclosed their use of disguised remuneration loans in their 1999/2000 tax return and had exercised reasonable care in filing their returns (e.g. by taking advice from professionals and testing that advice) would generally under normal rules have found HMRC time-barred from making an assessment into 2004/05 onwards. Even if taxpayers were careless but not deliberate in their filing, they would have had certainty after the expiry of 2005/06. However, under the 2019 loan charge, none of that matters. If any loans made on or after 6 April 1999 are outstanding as at 5 April 2019, the sum of those outstanding balances will be taxed as income or profits in the 2018/19 tax year and must be reported in that year’s tax return.
Similarly, while DPT itself cannot be imposed pre-April 2015, transfer pricing adjustments often made as part of resolving DPT challenges frequently have some degree of retrospective application. Experience suggests that this can be the case even if, at the time, the transfer pricing had been explained to (and, in some cases, accepted by) HMRC.
What if HMRC has agreed the arrangements in question or formally cleared them? Taxpayers may feel safe if they have ‘put all their cards face upwards on the table’ (see MFK Underwriting Agencies [1989] STC 873) and certainly those with advance pricing agreements that cover the years around the introduction of DPT have been more confident than those without.
However, it’s worth noting the recent decision in Oxford Instruments UK 2013 Ltd [2019] UKFTT 254 (TC). HMRC had cleared the taxpayer’s ‘tower’ structure (whereby UKCo received tax exempt dividend income which then funded tax deductible interest to USCo, creating a net tax deduction) under the anti-arbitrage rules, but then subsequently challenged the same under CTA 2009 s 441. The FTT concluded that while the scheme as a whole may have had good commercial purposes (financing the US business), UKCo’s being a party to the loan had a main purpose of securing a tax advantage. As the s 441 test was different to the anti-arbitrage test and HMRC had changed its approach to s 441’s application to tower structures, it was not inconsistent for HMRC to subsequently challenge the structure under s 441. While not technically retrospective, the FTT noted that the arrangement ‘had the apparent blessing of [HMRC but] that blessing was a mirage’.
Clearly, taxpayers should avoid seeking broader comfort by extrapolating from specific clearances (or assuming points made to one team within HMRC will land the same way with others).
Clear time limits are critical in giving certainty to taxpayers about what tax is due and when. It’s also right that HMRC has the power to investigate further when genuinely new information comes to light.
However, the loan charge and DPT rules, together with the broader shift in what is (and is not) acceptable to HMRC, suggest a pattern whereby HMRC is empowered to tax historic transactions which it wishes it had taxed in the past but for whatever reason did not, or could not. This may have public and political support in some instances (e.g. undeclared offshore profits coming to light courtesy of WikiLeaks), but any shift to a less stable, more retrospective tax system (perceived or not) will mean the UK is less attractive for individuals and businesses.
Negotiating appropriate time limits for allocating tax risk in M&A or other transactions could also become more difficult. Pressure to reduce the six-year limit seen in many deals may need to be resisted in key areas. In any event, taxpayers should be alive to the limits of any particular clearance or ruling, be careful in preserving the contemporaneous evidence in support of their position, and be mindful when taking on historic risk in new acquisitions.
The authors thank Sarah Blakelock of the same firm for her contribution to this article.
The 2019 loan charge has attracted much attention and been the subject of detailed commentary in Tax Journal already. Given this and the ongoing public and parliamentary debate about the legitimacy of the charge, we focus here on the broader questions around statutory time limits that apply to HMRC’s ability to investigate a taxpayer’s affairs and raise assessments of tax. Central among these recently is the balance between giving taxpayers certainty as to what tax is due and when, and the pressure on HMRC to tackle perceived avoidance or underpayment with new powers in what is undoubtedly now a tougher socio-political environment for taxpayers. Key battlegrounds here are:
Before testing those battlegrounds, it’s worth recalling the ‘normal’ rules for opening enquiries into returns and/or making discovery assessments or determinations. These, of course, differ depending on what tax is in scope. The EU state aid rules also have their own (usually longer) limitation period and most readers will have had to tackle the patchwork quilt of limitation periods in other jurisdictions.
For present purposes, we focus on UK income tax, capital gains tax and corporation tax. The rules for these broadly mirror one another and we expect they will be most relevant to readers.
The window for HMRC to open an enquiry into a taxpayer’s return is short. For income tax and capital gains tax, HMRC usually has 12 months from the date the return was delivered to open an enquiry (TMA 1970 s 9A(2)). For group companies other than Companies Act 2006 small group companies, a similar 12 month deadline operates but this starts on the filing date, not the date the return is delivered (FA 1998 Sch 18 para 24). In both cases, if the return is delivered late or is amended, the enquiry window is extended. Importantly, HMRC needs to give notice of its intention to enquire. This is a factual question but is a crucial gateway to imposing further liability on taxpayers.
Where the statutory time limits for opening an enquiry have passed or an enquiry into a particular year has been closed, HMRC can still recover lost tax in those years by making a discovery assessment (or determination).
To issue a discovery assessment, HMRC must have first discovered a loss of tax, being:
a relief that has been given which is or has become excessive (TMA 1970 s 29 and FA 1998 Sch 18 para 41).
In addition, (i) that loss of tax must have arisen as a result of careless or deliberate behaviour by the taxpayer or their agent; or, (ii) at the time the enquiry window closed (or HMRC closed its enquiry into the relevant year), HMRC could not reasonably have been expected to be aware of the loss of tax – in other words, there was an error despite reasonable care being taken (TMA 1970 s 29(4) and (5) and FA 1998 Sch 18 paras 43 and 44).
The rules effectively create a three tier system. The ‘default’ time limit is four years from the end of the relevant year or accounting period in question (TMA 1970 s 34 and 34A and FA 1998 Sch 18 para 46(1)), but this applies only if there’s an error despite taking reasonable care. For careless and deliberate conduct, HMRC can go further back in time. HMRC has six years from the end of the relevant year or accounting period if the loss of tax is the result of the taxpayer’s (or their agent’s) careless behaviour and 20 years if the behaviour was deliberate (TMA 1970 s 36(1) and (1A) and FA 1998 Sch 18 para 46). HMRC can also go back 20 years if the loss of tax resulted from the taxpayer’s failure to notify HMRC of their chargeability to tax, or was attributable to arrangements which should have been notified under disclosure of tax avoidance schemes (DOTAS) or promoters of tax avoidance schemes (POTAS) (but were not) (TMA 1970 s 36(1A) and FA 1998 Sch 18 para 46).
Even if that should be relatively straightforward, recent developments have highlighted three particular complications.
Discoveries can become ‘stale’ if there is undue delay between the discovery and the assessment itself (even if it was made in time), in which case the assessment is invalidated. In Pattullo v HMRC [2016] STC 2043, the Upper Tribunal concluded that HMRC must act on its discovery ‘while it remains fresh (or before it becomes stale)’. In the recent case of Beagles v HMRC [2018] UKUT 380, the Upper Tribunal followed Pattullo. Beagles will be reviewed by the Court of Appeal, though taxpayers should have some comfort that HMRC cannot sit on newly ‘discovered’ information indefinitely.
FA 2019 introduced a new 12 year time limit (at TMA 1970 s 36A) for assessing lost income tax or capital gains tax involving offshore matters or transfers. For careless errors and those made despite taking reasonable care, the provisions increase the number of tax years potentially subject to assessment prospectively, one year at a time, until the period that HMRC can assess reaches 12 years. The existing 20 year time limit for deliberate behaviour remains.
The policy behind the new limits is that, in offshore cases, it will often take HMRC longer to investigate or longer for information to come to light under international information exchanges than in purely domestic cases. However, with some limited safeguards for information received by HMRC before expiry of the ‘old’ time limits, the period of uncertainty for taxpayers with offshore assets or income is materially increased.
Taxpayers with rulings or other arrangements potentially vulnerable to investigation under the EU state aid rules are similarly subject to longer term exposure. The general rule is that there is a ten year time limit, meaning that arrangements agreed with HMRC (or indeed positively endorsed by it in the case of the CFC finance company partial exemption) remain vulnerable to reassessment for some time. Taxpayers, however, have a shorter six year limit to claim against HMRC for breach of EU law if there has been an unlawful grant of state aid. Any challenge by way of judicial review will need to be made quickly, within three months of the relevant EC decision determining that there has been illegal state aid.
A key challenge levelled at the 2019 loan charge and other recent high-profile taxes, e.g. the diverted profits tax (DPT), is that they are retrospective. It’s important though to distinguish between the technical analysis and the substantive application to taxpayers.
In that regard, HM Treasury is right to conclude in its report published on 26 March 2019 that the loan charge legislation is not retrospective, as it is a ‘tax charge to outstanding disguised remuneration loan balances as at 5 April 2019’ and ‘does not change the tax position of any previous year or the tax treatment of any historic transaction’. Likewise, the DPT rules in FA 2015 (amended by FA 2019) introduced a new 25% tax on diverted profits with effect from 1 April 2015 (with higher rates for adjusted ringfenced oil profits and bank profits). Profits arising before 1 April 2015 were, in principle, unaffected by the new regime.
However, it’s likely that both the taxpayer facing a tax liability in the 2018/19 tax year in respect of up to 20 years’ worth of historic earnings or a multinational required to fund DPT charging notices upfront in respect of transactions entered into decades previously will feel that the substance of the tax charge is retrospective.
In the case of the loan charge, taxpayers who disclosed their use of disguised remuneration loans in their 1999/2000 tax return and had exercised reasonable care in filing their returns (e.g. by taking advice from professionals and testing that advice) would generally under normal rules have found HMRC time-barred from making an assessment into 2004/05 onwards. Even if taxpayers were careless but not deliberate in their filing, they would have had certainty after the expiry of 2005/06. However, under the 2019 loan charge, none of that matters. If any loans made on or after 6 April 1999 are outstanding as at 5 April 2019, the sum of those outstanding balances will be taxed as income or profits in the 2018/19 tax year and must be reported in that year’s tax return.
Similarly, while DPT itself cannot be imposed pre-April 2015, transfer pricing adjustments often made as part of resolving DPT challenges frequently have some degree of retrospective application. Experience suggests that this can be the case even if, at the time, the transfer pricing had been explained to (and, in some cases, accepted by) HMRC.
What if HMRC has agreed the arrangements in question or formally cleared them? Taxpayers may feel safe if they have ‘put all their cards face upwards on the table’ (see MFK Underwriting Agencies [1989] STC 873) and certainly those with advance pricing agreements that cover the years around the introduction of DPT have been more confident than those without.
However, it’s worth noting the recent decision in Oxford Instruments UK 2013 Ltd [2019] UKFTT 254 (TC). HMRC had cleared the taxpayer’s ‘tower’ structure (whereby UKCo received tax exempt dividend income which then funded tax deductible interest to USCo, creating a net tax deduction) under the anti-arbitrage rules, but then subsequently challenged the same under CTA 2009 s 441. The FTT concluded that while the scheme as a whole may have had good commercial purposes (financing the US business), UKCo’s being a party to the loan had a main purpose of securing a tax advantage. As the s 441 test was different to the anti-arbitrage test and HMRC had changed its approach to s 441’s application to tower structures, it was not inconsistent for HMRC to subsequently challenge the structure under s 441. While not technically retrospective, the FTT noted that the arrangement ‘had the apparent blessing of [HMRC but] that blessing was a mirage’.
Clearly, taxpayers should avoid seeking broader comfort by extrapolating from specific clearances (or assuming points made to one team within HMRC will land the same way with others).
Clear time limits are critical in giving certainty to taxpayers about what tax is due and when. It’s also right that HMRC has the power to investigate further when genuinely new information comes to light.
However, the loan charge and DPT rules, together with the broader shift in what is (and is not) acceptable to HMRC, suggest a pattern whereby HMRC is empowered to tax historic transactions which it wishes it had taxed in the past but for whatever reason did not, or could not. This may have public and political support in some instances (e.g. undeclared offshore profits coming to light courtesy of WikiLeaks), but any shift to a less stable, more retrospective tax system (perceived or not) will mean the UK is less attractive for individuals and businesses.
Negotiating appropriate time limits for allocating tax risk in M&A or other transactions could also become more difficult. Pressure to reduce the six-year limit seen in many deals may need to be resisted in key areas. In any event, taxpayers should be alive to the limits of any particular clearance or ruling, be careful in preserving the contemporaneous evidence in support of their position, and be mindful when taking on historic risk in new acquisitions.
The authors thank Sarah Blakelock of the same firm for her contribution to this article.
.tmb-editorspic.jpg?sfvrsn=df04ab27_1)
