On 1 August, just before the summer break, HM Revenue & Customs (HMRC) published a "Technical Note" and draft legislation (together, the "Note") entitled "Tax Treaties Anti-avoidance". The Note opens a very time-limited consultation which closes on 22 September 2011.

The broad effect of the draft legislation (if enacted in its present form) would be to prevent taxpayers from relying on the provisions of a UK double tax treaty. It applies where there is a "scheme" (which is extraordinarily widely defined – see below) with a main purpose or one of its main purposes being to obtain a benefit under the treaty.

Taken at face value, this legislation could easily block commercially standard structures intended to manage tax efficiently, provide financing or manage assets. This means that some cross-border financing arrangements will have to be restructured, and all will have to be reconsidered.

The Note is explicit – this legislation is intended to override the provisions of double tax treaties, past, present and future. When the legislation comes into force (likely to be summer 2012), it will remove treaty protection from structures already in place, as well as blocking it for future structures. The Note specifically suggests that HMRC intend to apply this legislation "in practice" to counter treaty shopping.

A link to the Note can be found here.


The Note is published at a time when HMRC has had very recent success in the Courts in defeating treaty-based planning. On 25 July 2011, the Court of Appeal ruled in HMRC's favour in the joined cases in Huitson1 and Shiner2, holding that expressly retrospective legislation to block double tax treaty planning was enforceable despite being retrospective. On 12 July 2011, the Supreme Court refused permission for an appeal by the taxpayers in Bayfine3, finalising the Court of Appeal's decision that the non-discrimination article of the UK/US double tax treaty did not operate to completely eliminate any tax liability on a particular item of income. On 22 December 2010, the Supreme Court refused permission for an appeal by the taxpayer in Smallwood4, finalising the Court of Appeal's decision that purposive interpretation of the UK/Mauritius double tax treaty was enough to block private client capital gains tax planning.

This legislation also comes at a time when the Fiscal Affairs Committee of the OECD is considering amendments to the Commentary on the OECD Tax Treaty Model, intended to tie down the meaning of "beneficial ownership" as used in the interest and dividend articles of the Model.5 Many countries have concluded treaties which do not contain specific "anti-conduit" rules: an example of this is the treaty in issue in the Indofood6 case. Such countries often try to apply a strained meaning to "beneficial ownership" to combat conduit companies and treaty shopping, instead of adopting the alternative course of re-negotiating the treaty.

The UK, by contrast, applies a fairly consistent policy, in negotiating treaties, of including specific "anti-conduit rules". This renders obscure the policy behind HMRC's proposals in the Note, representing as they do a "catch-all" approach.

Proposed legislation

The legislation deals separately with UK residents and non-residents. It proposes one provision for each, to sit at opposite ends of the part of TIOPA dealing with "interpretation of double taxation arrangements". All the other rules in that part (sections 130 to 133) are genuinely to do with interpretation of treaty rights. These new provisions - although entitled "interpreting provision" - represent an override instead.

This legislation is intended to be introduced in Finance Act 2012. It will apply for UK residents to any income or capital gains arising on or after the date the legislation comes into force (ie, a date yet to be determined in summer 2012) and for non-residents to income arising after that date. It will apply regardless of when the structure was established. It will also apply to double tax treaties yet to be negotiated.

Non-residents (proposed new section 133A TIOPA)

The proposed legislation (in so far as it affects non-UK residents) applies to what the legislation calls UK source income. This is defined to mean (1) dividends paid by a company resident in the UK, or (2) income arising in the UK which falls within the articles of the relevant double tax treaty dealing with "interest" or "income from debt-claims", royalties, or other income (see, for instance, article 20 of the US/UK double tax treaty). Significantly, the proposed legislation would not apply to non-residents' business profits, income and gains from immovable property or other kinds of "non-passive" income.

It is proposed that, if all of the following conditions are satisfied, the recipient cannot rely on the double tax treaty to apply a reduced rate of UK tax (or exemption) on the income if:

  • a "scheme" is or has been "put in place";
  • the relevant provision of the double tax treaty only applies because the "scheme" exists (i.e. the treaty provision would not have applied absent the "scheme"); and
  • the main purpose or one of the main purposes of "a person" in "putting the scheme in place" is to ensure that the double tax treaty applies to the income.

There are two important points to note.

First, "scheme" is extremely widely defined (in the proposed section 133A(8)). It includes any arrangement or understanding. The current draft legislation also seems to contemplate that a "scheme, arrangement or understanding" may be "put in place" by a single person. There are substantial logical problems with this "bird's nest" of a definition, but overall it seems that the concept of scheme is likely to cover many transactions involving any sort of structuring.

Second, although decided cases give a fairly clear meaning to main purpose, HMRC take a number of different approaches to this question. In practice, where HMRC seek to apply the legislation, they sometimes advance the view that it is not the main purpose of the overall commercial operation that matters but the main purpose test of choosing one means of achieving that commercial aim rather than another7. HMRC's view is not a fair statement of the legal position and is a point which we believe will be cleared up in the Tribunals in due course.

HMRC may however seek in practice to apply their wide interpretation to the proposed legislation (if enacted), and argue that the simple fact of choosing to finance through, or hold IP or other assets in, a country which has negotiated a "good" treaty with the UK (instead of, say, one with no treaty with the UK) will bring a structure within this legislation. This is capable of going far wider than existing anti-conduit rules. More importantly, it will apply to all jurisdictions whatever the terms of the relevant treaty.

The proposals are particularly problematic because they cast a very wide net and will need, in practice, to be narrowed by HMRC guidance. In turn, it is very difficult to force HMRC to adhere to previously adopted positions so there may not be consistency under the proposed rules.

The Note indicates (Chapter 1, paragraph 2) that the legislation is aimed at attempts "to access the benefits of double tax treaties when they are not properly due". It also indicates that the legislation applies only to "arrangements that constitute an abuse of the treaty" (Chapter 1, paragraph 4). However, it is far from clear what the first of these statements means. It is the terms of a treaty, correctly construed, which determine whether treaty reliefs are due. As part of the process of construction it may be legitimate to consider the OECD Commentary: this too will depend primarily on the terms of the treaty and the Commentary. But there is no clear policy basis for introducing some additional requirement, of the sort denoted by the word "properly" in the Note. The second statement is not necessarily true: the proposed main purpose test is capable of being applied much more widely than, say, the EU law Cadbury Schweppes8 abuse of law test.

UK residents (proposed new section 129A)

Unlike the provisions for non-UK residents discussed above, the provisions for UK residents apply to all forms of income, and chargeable gains as well. It disapplies the benefit of a treaty if broadly the same three "avoidance conditions" (ie, the conditions in section 133A) are met.

To be precise, the legislation applies whenever a treaty provides an exemption or a reduced rate of UK tax for a particular item (notably, there is no reference to credit for withholding tax or underlying tax on non-exempt dividends: this may be on the assumption that unilateral relief would be available in any event, but unilateral relief is not generally available where a treaty exists, and the point requires additional clarification). The legislation also applies whenever a treaty prevents an item being deemed to be income/gains of a UK resident.

The most notable point of this part of the legislation is what it does to the test for residence. The test for tax residence will be the UK domestic test without any reference to the standard tie breaker test that applies in existing treaties. HMRC state (Chapter 2 paragraph 7) that this is "to prevent UK residents manipulating the residence tie-breaker provisions in the UK's double tax treaties" (a clear nod to Smallwood), but the legislation goes much further.

Why is additional legislation being proposed?

We understand that the Government has been alerted to a number of tax avoidance schemes and arrangements which rely on the terms of a double tax treaty for success. It considers that attempts to shut down such schemes have failed and it now regards a more widely applying avoidance rule to be necessary. However, this legislation goes much wider than marketed schemes.

It is surprising that the Government did not wait until the committee investigating the viability of a general anti-avoidance rule (GAAR) had reverted (expected in late October) with its comments and, as we suspect, draft legislation. If a GAAR is also introduced in Finance Act 2012, UK taxpayers with any international structuring requirements will effectively have two new sets of broad anti-avoidance rules to consider, as well as the relevant provisions and the meaning of "beneficial ownership" in the relevant treaty, before they can be satisfied that relief under a double tax treaty is available. We also understand that HMRC is reviewing their position on the main purpose test generally, with a publication expected later in the year. There is simply too much movement in this area.

Next steps

In terms of next steps, there are three key issues to consider.

Practical structuring

Structures that have been put in place at a time when they were "acceptable" or at the very least effective may now (with effect from July 2012) become ineffective. Those structures need to be reconsidered. There are likely to be solutions.

For arrangements already in place, and which are to stay in place, documenting a purpose will be crucial. Experience tells us that it gets harder and harder to document purpose as time goes by. For many existing arrangements put in place several years ago, it is likely to be a difficult task to prove whether they were put in place without a main purpose of obtaining the benefit of a double tax treaty.

Parties will have obligations under existing agreements that may well now leave them exposed. Parties to cross border agreements will want to ensure that any gross up clauses address the situation where HMRC successfully argues that a treaty benefit is not available and in consequence, a different rate of withholding may apply from that agreed upon in the contract.

These provisions also need to be addressed for structures which are not yet in place but are currently being put in place and are intended to last into summer 2012 and beyond. That consideration should start now.


This legislation is likely to cause real difficulties. We will be making representations. Corporates, banks and asset managers should consider making their own representations as well.

Compliance with international obligations - the UK's international competitiveness

There may also be concerns that in such cases the provisions would be incompatible with the Interest and Royalties Directive.9 The Directive permits Member States to apply domestic provisions required for the prevention of fraud or abuse, or withdraw the benefit of the Directive but only where "the principal motive or one of the principal motives for the relevant transaction is tax evasion, tax avoidance or abuse". In this context, it is not HMRC's or the UK Courts' view of what constitutes "tax evasion, tax avoidance or abuse" that matters, but the Commission's and, particularly, the ECJ's.

Depending on the fact pattern, there are also questions over whether this legislation would be enforceable given the UK's binding obligations under Freedom of Establishment (for EU transactions) and Free Movement of Capital, and particularly the World Trade Organisation Treaties (for non-EU transactions). The fact that it changes the law applying to pre-existing structures raises questions of improper retrospective legislation. However, this would have to be investigated on a case-by-case basis.

One aspect of current fiscal policy is the attraction of business to the UK. The proposed provisions on non-UK residents (in particular) risk hampering the achievement of that objective.

It also remains to be seen what legislation of this sort will do the UK's programme of treaty negotiations. HMRC has negotiated inclusion of provisions of the nature proposed by the legislation into specific treaties: for instance, it is used in the definition of "conduit arrangements" in article 3(1)(l) of the UK/Switzerland treaty. But these main purpose provisions do not appear in all treaties, and no UK treaty contains main purpose provisions of such broad application.

However, the new provisions are expressed to apply to all treaties. Presumably, this extends to treaties even where the other contracting state has in negotiations specifically rejected (or limited the application of) provisions of this kind. These rules also apply to treaties not yet in force. Further, given that the new provisions are expressed to be for "interpretation" of treaties, it may be arguable that any anti-abuse provisions in a treaty, covering the same ground as the proposed new provisions, would as a matter of necessary implication exclude the application of the new provisions.

If you have any questions or comments or would like assistance in drafting a response to this consultation, please contact us.