The Kennel Club has criticised the Dangerous Dogs Act on the basis that the legislation ignores the most important factors that lead to biting incidents, and says that it is unacceptable to ban an entire breed based on the actions of a single animal. The proposed diverted profits tax (DPT) appears to suffer from similar defects: whilst there have undoubtedly been incidents of multinationals paying less tax than the public or Parliament would like, this does not seem to be the best way to deal with them. It is an extraordinarily complex set of provisions, released halfway through the base erosion and profit-shifting (BEPS) process, and expected to raise a relatively modest £350m per year. Is it needed, and if so, will it work in its present form?
We have been given nearly 30 pages of legislation, and 50 pages of guidance, to introduce yet another TAAR, when we already have the GAAR on the statute book. Paragraph B5.3 of the GAAR guidance explicitly states that the GAAR can be applied to counteract ‘abusive arrangements which try to exploit particular provisions in a double tax treaty’ - so why not use it? It seems that the GAAR was thought to be untried and too uncertain, and to take too long (and perhaps too much effort) to apply. And so the volume of complex legislation, including the ‘artificial transfers of profits’ measure introduced only last year, grows ever larger.
Alternatively, if the arrangements are truly artificial, and in substance a company does have a permanent establishment (PE) or additional profits in the UK, why not challenge them on the basis of whether ‘the relevant statutory provisions, construed purposively, were intended to apply to the transaction, viewed realistically’ (Ribeiro PJ in Arrowtown [2003] HKCFA 46)?
Introducing an extra 30 pages of legislation seems to be the easier option for HMRC, but surely existing rules should be applied robustly before asking for yet more weapons to be added to the armoury – and yes, like many in the tax profession, I would support HMRC having the resources it needs to do the job properly.
It is interesting that in the same week as the Autumn Statement, China announced that it would ‘comprehensively monitor the profit levels of foreign companies to make sure there is no base erosion and profit shifting’. This does seem to be a more measured approach.
Assuming that the measure is needed, will it work? There are two major problems: scope and calculation.
On scope, the obvious point is that this may breach our treaty obligations, and particularly the right to freedom of establishment within the EU. However, it appears to have been very carefully drafted to stay (just) on the right side of the line, notwithstanding Cadbury Schweppes (C-196/04). The first technical argument is that it is explicitly drafted as a new tax, and not as a profits tax akin to corporation tax. It is not within CTSA (although there is an obligation to notify chargeability). In Bricom Holdings Ltd v CIR 70 TC 272, the Court of Appeal held that the charge on apportioned profits under CFC rules did not breach treaty obligations, because it merely required the calculation of a notional sum. However, if that is the only defence HMRC has against a treaty or EU challenge, I would not like to rely on it: in substance this is a tax on profits of a company which does not have a UK taxable presence, or which has understated its profits despite complying with OECD transfer pricing principles. For HMRC to seek to rely on form rather than substance would not be an edifying precedent.
The stronger argument is that this is aimed at structures which are abusive and wholly artificial. In Cadbury Schweppes, the CJEU held that:
‘A national measure restricting freedom of establishment may be justified where it specifically relates to wholly artificial arrangements aimed at circumventing the application of the legislation of the member state concerned.’
The background note (but not the legislation itself) says that the main objective ‘is to counteract contrived arrangements’ which ‘result in the erosion of the UK tax base’. As a minimum, this needs to be clearly and explicitly stated in the legislation itself, both to reduce the risk of challenge under EU law and also to minimise unnecessary compliance burdens on those who are not within the intended scope.
HMRC does not (yet) have a general treaty anti-abuse rule, although Action 6 of BEPS proposes that one should be introduced. However, their track record before the UK courts, where HMRC has estimated an 80% success rate in recent avoidance cases, as well as the Halifax principle in the EU suggests that, provided it is sufficiently narrowly drawn, HMRC can probably defend this measure.
But if it is so narrowly drawn, how much tax will it really raise? A well-advised multinational, which has properly implemented the advice it was given, will surely ensure that no-one in the UK ‘has, and habitually exercises … an authority to conclude contracts on behalf of the enterprise’ (OECD Model Treaty art 5(5)), and will also ensure that its transfer pricing methodology stands up to scrutiny and does not have ‘insufficient economic substance’.
The rules seek to attack two specific situations. The first is where a foreign company makes sales to UK customers and ‘a person’ is carrying on activity in the UK in connection with those sales. If it is ‘reasonable to assume’ that the activity has been ‘designed’ to ensure there is no UK PE, then the profits attributable to the ‘avoided PE’ will be calculated and taxed. At least one of two conditions must be met: the ‘tax avoidance condition’ which is a relatively standard main purpose test in relation to the arrangements, and the ‘mismatch condition’. This requires there to be a significant tax difference, and also a lack of economic substance. It must be ‘reasonable to assume’ that the transaction was ‘designed’ to secure the tax reduction: either the tax benefit is greater than any other benefit, or there is a mismatch between the contribution of economic benefit by staff and the tax benefit claimed.
All of this sounds fine in theory, but will require very detailed fact-finding to establish. Why not just attack the arrangements under existing transfer pricing rules? This appears to be a straightforward attempt by the UK to make a unilateral amendment to the definition of a PE. This is unlikely to be popular internationally, undermines the BEPS process and risks retaliation from countries such as India. Australia has already announced that it will consider similar measures. The whole point of the OECD Model Treaty definition of a PE is that it is a factual test, used on a common basis internationally: eroding this consensus may well lead to chaos, and a real risk of double taxation.
The second situation is where there is an existing UK entity, but HMRC considers that it is not reporting enough profit. This is, perhaps, both simpler and more powerful than the ‘avoided PE’ provision, since it is grounded in transfer pricing rules and requires an adjustment where a ‘material provision’ results in an ‘effective tax mismatch’ and there is ‘insufficient economic substance’. (It is hard to resist saying, yet again, that surely existing transfer pricing rules should deal with this.) The scenario envisaged appears to be a case where a royalty is paid out of the UK to a tax haven, where there is little substance: that could be valid under OECD arm’s length principles, but fall foul of this rule. Furthermore, the new law appears to allow HMRC to look at the global arrangements and not just the immediate royalty to, for example, Ireland: clearly HMRC is willing to take on the role of global policeman in these circumstances.
The economic substance test has real difficulties for innocent but complex groups. Whilst it may be possible to calculate whether there is an ‘effective tax mismatch’ (broadly, a tax rate less than 80% of the UK rate), it will be almost impossible to determine, on a transaction by transaction basis, how the tax benefits compare to the economic benefits. It is not enough to say that groups only need to worry if transactions have been ‘designed’ to achieve the benefit: as there is no over-riding purpose test, they have to consider whether it is ‘reasonable to assume’ that that is their intended effect. Even worse, they must notify chargeability if transactions ‘might’ come within these provisions: it is not far-fetched to assume that all multinational groups will make multiple notifications. Do we need a clearance system, perhaps?
Note also the subtlety in the way that the attack will be made. HMRC issues a warning, against which the company can make representations only on limited grounds. A formal notice is then issued, and over a 12 month review period detailed discussions can be held, but the tax has to be paid in the meantime – a ‘super accelerated payment notice (APN)’. Delaying the company’s right of appeal until after the review period has expired appears vulnerable to challenge under judicial review. The objective seems to be to bully companies into restructuring and reporting higher UK profits, rather than risking a DPT charge at 25%.
In summary, it is clear that the government is worried about abusive multinationals and feels the need to ‘do something’. It is much less clear that these proposals are the best thing to do: they add to complexity and compliance burdens, and will be difficult to enforce. This is not to say that the government should do nothing: it is entirely reasonable that the policy should be for multinationals to pay UK tax on their economic activities, and for anti-avoidance rules to ensure that this is the case.
What is less reasonable is to introduce such complex provisions in such a rush, particularly given the BEPS timetable. There are those who think BEPS will not work and that we cannot wait to introduce the diverted profits tax: I do not agree. This measure will make BEPS more difficult to achieve, and it risks a whole raft of unilateral measures being introduced by other countries.
Given the length of time it took to introduce the CFC rules, and the level of consultation that took place there, there is surely no immediate urgency for this measure (which will have to be rushed through in the pre-election Finance Bill without the opportunity for MPs to consider it properly) – unless, as cynics might say, the pressure is that of a general election in five months’ time. Why not withdraw this measure, consult properly and reintroduce it in 2016? – either as part of a BEPS package, or separately if BEPS stalls. I say we should give BEPS a chance.
The Kennel Club has criticised the Dangerous Dogs Act on the basis that the legislation ignores the most important factors that lead to biting incidents, and says that it is unacceptable to ban an entire breed based on the actions of a single animal. The proposed diverted profits tax (DPT) appears to suffer from similar defects: whilst there have undoubtedly been incidents of multinationals paying less tax than the public or Parliament would like, this does not seem to be the best way to deal with them. It is an extraordinarily complex set of provisions, released halfway through the base erosion and profit-shifting (BEPS) process, and expected to raise a relatively modest £350m per year. Is it needed, and if so, will it work in its present form?
We have been given nearly 30 pages of legislation, and 50 pages of guidance, to introduce yet another TAAR, when we already have the GAAR on the statute book. Paragraph B5.3 of the GAAR guidance explicitly states that the GAAR can be applied to counteract ‘abusive arrangements which try to exploit particular provisions in a double tax treaty’ - so why not use it? It seems that the GAAR was thought to be untried and too uncertain, and to take too long (and perhaps too much effort) to apply. And so the volume of complex legislation, including the ‘artificial transfers of profits’ measure introduced only last year, grows ever larger.
Alternatively, if the arrangements are truly artificial, and in substance a company does have a permanent establishment (PE) or additional profits in the UK, why not challenge them on the basis of whether ‘the relevant statutory provisions, construed purposively, were intended to apply to the transaction, viewed realistically’ (Ribeiro PJ in Arrowtown [2003] HKCFA 46)?
Introducing an extra 30 pages of legislation seems to be the easier option for HMRC, but surely existing rules should be applied robustly before asking for yet more weapons to be added to the armoury – and yes, like many in the tax profession, I would support HMRC having the resources it needs to do the job properly.
It is interesting that in the same week as the Autumn Statement, China announced that it would ‘comprehensively monitor the profit levels of foreign companies to make sure there is no base erosion and profit shifting’. This does seem to be a more measured approach.
Assuming that the measure is needed, will it work? There are two major problems: scope and calculation.
On scope, the obvious point is that this may breach our treaty obligations, and particularly the right to freedom of establishment within the EU. However, it appears to have been very carefully drafted to stay (just) on the right side of the line, notwithstanding Cadbury Schweppes (C-196/04). The first technical argument is that it is explicitly drafted as a new tax, and not as a profits tax akin to corporation tax. It is not within CTSA (although there is an obligation to notify chargeability). In Bricom Holdings Ltd v CIR 70 TC 272, the Court of Appeal held that the charge on apportioned profits under CFC rules did not breach treaty obligations, because it merely required the calculation of a notional sum. However, if that is the only defence HMRC has against a treaty or EU challenge, I would not like to rely on it: in substance this is a tax on profits of a company which does not have a UK taxable presence, or which has understated its profits despite complying with OECD transfer pricing principles. For HMRC to seek to rely on form rather than substance would not be an edifying precedent.
The stronger argument is that this is aimed at structures which are abusive and wholly artificial. In Cadbury Schweppes, the CJEU held that:
‘A national measure restricting freedom of establishment may be justified where it specifically relates to wholly artificial arrangements aimed at circumventing the application of the legislation of the member state concerned.’
The background note (but not the legislation itself) says that the main objective ‘is to counteract contrived arrangements’ which ‘result in the erosion of the UK tax base’. As a minimum, this needs to be clearly and explicitly stated in the legislation itself, both to reduce the risk of challenge under EU law and also to minimise unnecessary compliance burdens on those who are not within the intended scope.
HMRC does not (yet) have a general treaty anti-abuse rule, although Action 6 of BEPS proposes that one should be introduced. However, their track record before the UK courts, where HMRC has estimated an 80% success rate in recent avoidance cases, as well as the Halifax principle in the EU suggests that, provided it is sufficiently narrowly drawn, HMRC can probably defend this measure.
But if it is so narrowly drawn, how much tax will it really raise? A well-advised multinational, which has properly implemented the advice it was given, will surely ensure that no-one in the UK ‘has, and habitually exercises … an authority to conclude contracts on behalf of the enterprise’ (OECD Model Treaty art 5(5)), and will also ensure that its transfer pricing methodology stands up to scrutiny and does not have ‘insufficient economic substance’.
The rules seek to attack two specific situations. The first is where a foreign company makes sales to UK customers and ‘a person’ is carrying on activity in the UK in connection with those sales. If it is ‘reasonable to assume’ that the activity has been ‘designed’ to ensure there is no UK PE, then the profits attributable to the ‘avoided PE’ will be calculated and taxed. At least one of two conditions must be met: the ‘tax avoidance condition’ which is a relatively standard main purpose test in relation to the arrangements, and the ‘mismatch condition’. This requires there to be a significant tax difference, and also a lack of economic substance. It must be ‘reasonable to assume’ that the transaction was ‘designed’ to secure the tax reduction: either the tax benefit is greater than any other benefit, or there is a mismatch between the contribution of economic benefit by staff and the tax benefit claimed.
All of this sounds fine in theory, but will require very detailed fact-finding to establish. Why not just attack the arrangements under existing transfer pricing rules? This appears to be a straightforward attempt by the UK to make a unilateral amendment to the definition of a PE. This is unlikely to be popular internationally, undermines the BEPS process and risks retaliation from countries such as India. Australia has already announced that it will consider similar measures. The whole point of the OECD Model Treaty definition of a PE is that it is a factual test, used on a common basis internationally: eroding this consensus may well lead to chaos, and a real risk of double taxation.
The second situation is where there is an existing UK entity, but HMRC considers that it is not reporting enough profit. This is, perhaps, both simpler and more powerful than the ‘avoided PE’ provision, since it is grounded in transfer pricing rules and requires an adjustment where a ‘material provision’ results in an ‘effective tax mismatch’ and there is ‘insufficient economic substance’. (It is hard to resist saying, yet again, that surely existing transfer pricing rules should deal with this.) The scenario envisaged appears to be a case where a royalty is paid out of the UK to a tax haven, where there is little substance: that could be valid under OECD arm’s length principles, but fall foul of this rule. Furthermore, the new law appears to allow HMRC to look at the global arrangements and not just the immediate royalty to, for example, Ireland: clearly HMRC is willing to take on the role of global policeman in these circumstances.
The economic substance test has real difficulties for innocent but complex groups. Whilst it may be possible to calculate whether there is an ‘effective tax mismatch’ (broadly, a tax rate less than 80% of the UK rate), it will be almost impossible to determine, on a transaction by transaction basis, how the tax benefits compare to the economic benefits. It is not enough to say that groups only need to worry if transactions have been ‘designed’ to achieve the benefit: as there is no over-riding purpose test, they have to consider whether it is ‘reasonable to assume’ that that is their intended effect. Even worse, they must notify chargeability if transactions ‘might’ come within these provisions: it is not far-fetched to assume that all multinational groups will make multiple notifications. Do we need a clearance system, perhaps?
Note also the subtlety in the way that the attack will be made. HMRC issues a warning, against which the company can make representations only on limited grounds. A formal notice is then issued, and over a 12 month review period detailed discussions can be held, but the tax has to be paid in the meantime – a ‘super accelerated payment notice (APN)’. Delaying the company’s right of appeal until after the review period has expired appears vulnerable to challenge under judicial review. The objective seems to be to bully companies into restructuring and reporting higher UK profits, rather than risking a DPT charge at 25%.
In summary, it is clear that the government is worried about abusive multinationals and feels the need to ‘do something’. It is much less clear that these proposals are the best thing to do: they add to complexity and compliance burdens, and will be difficult to enforce. This is not to say that the government should do nothing: it is entirely reasonable that the policy should be for multinationals to pay UK tax on their economic activities, and for anti-avoidance rules to ensure that this is the case.
What is less reasonable is to introduce such complex provisions in such a rush, particularly given the BEPS timetable. There are those who think BEPS will not work and that we cannot wait to introduce the diverted profits tax: I do not agree. This measure will make BEPS more difficult to achieve, and it risks a whole raft of unilateral measures being introduced by other countries.
Given the length of time it took to introduce the CFC rules, and the level of consultation that took place there, there is surely no immediate urgency for this measure (which will have to be rushed through in the pre-election Finance Bill without the opportunity for MPs to consider it properly) – unless, as cynics might say, the pressure is that of a general election in five months’ time. Why not withdraw this measure, consult properly and reintroduce it in 2016? – either as part of a BEPS package, or separately if BEPS stalls. I say we should give BEPS a chance.
.tmb-editorspic.jpg?sfvrsn=df04ab27_1)
