Smallwood v Revenue and Customs

This is an appeal from the Special Commissioners (Dr A.N. Brice and Dr. J.F. Avery-Jones), released on 19^th February 2008, dismissing the appeals of taxpayers against amendments to their returns for the year 2000 – 2001 which included chargeable gains of over £6m arising on a disposal of assets by trustees. In short the position is this. Mr Smallwood was the settlor of a number of shares in companies known as FirstGroup and Billiton. He settled them for the benefit of himself and his family on 24^th February 1989. The settlement gave him the power of appointing the trustees. He remained a beneficiary of the trust. The trust contained other assets, but the bulk of the assets were the two shareholdings referred to and I am not concerned with any of the others. By 2000 a Jersey company, Lutea Trustees Limited (“Lutea”) was the trustee. The shares had increased significantly in value, and the view was taken by financial advisers to Mr Smallwood and the trustees that it would be wise to reduce such a major exposure. In short, it was thought to be a good idea to sell them. A sale of those shares by Lutea would have led to a charge on Mr Smallwood (as a resident settlor having a beneficial interest under the trust) pursuant to s.86 of the Taxation of Capital Gains Act 1992 (“TCGA”). In order to avoid such a charge a scheme was devised with the following elements. First, new trustees would be appointed to replace Lutea, those trustees being offshore trustees in a country which did not tax capital gains and which has a double taxation agreement with the UK. The shares would thereby vest in the new trustees. Those trustees would sell them, and having done so they would be removed as trustees and English trustees would be appointed before the end of the fiscal year. The combined effect of those steps, if the trick worked, would be that the sale would attract no capital gains tax by virtue of detailed provisions to which I shall come in due course. That scheme was put into operation. Mauritius was chosen as the relevant new jurisdiction. On 19^th December 2000 a Mauritius company, namely KPMG Peat Marwick International Limited (“PMIL”) was appointed to be the new trustee. They sold the shareholdings in question in January 2001. On 2^nd March 2001 PMIL ceased to be trustees and Mr and Mrs Smallwood became trustees (resident in this jurisdiction). When in due course tax returns were filed, HMRC (or its predecessor) sought to charge the tax above referred to on the basis of the sale. It is that tax which is disputed.

The provision under which Mr Smallwood as settlor would be charged in relation to a transaction carried out by foreign trustees is TCGA s.86. It provides:

Where this section applies –

The conditions required by that section would have been fulfilled if Lutea had sold the shares thus:

Lutea would have been non-resident throughout the year.

Mr. Smallwood, the settlor, was resident and domiciled in the UK during the relevant year.

He has retained an interest in the settlement.

Section 86 does not apply if, for example, the trustees are resident in the UK for some part of the resident year. However, in that event s.77 applies to impose a charge. Insofar as material it provides:

The only other relevant provision of English taxation law to which it is necessary to refer is s.69, which provides:

As I have just said, the taxpayer contends that the chargeable gains which would otherwise fall within s.77(1)(b) are not chargeable gains because of the effect of the double taxation arrangements with Mauritius which are contained in the Double Taxation Relief (Taxes on Income)(Mauritius) Order 1981 (SI 1981 no.1121) which I will call the Treaty or the order. The order was made under the Income and Corporation Taxes Act 1988 s.788, and despite its name it now applies to capital gains as well as to income. S.788 permits the making of arrangements and subsection 3, as originally enacted, provides:

The order itself, in article 2, declares that it is expedient that the arrangements in the Convention set out in the schedule to the order should have effect. That is sufficient, by virtue of statute, to give it effect. The relevant provisions of the schedule or Convention are as follows.

Article 1 provides:

Article 3 contains definitions, and they in turn include:

Article 4 deals with the all-important subject of “residence” around which this case turns. It reads as follows:

Article 13 deals with double taxation in relation to capital gains. I need to set it out in full, but the important provision for the purposes of this appeal is paragraph 4. Article 13 reads:

Capital Gains from the alienation of movable property forming part of the business property of a permanent establishment which an enterprise of a Contracting State has in the other Contracting State or of movable property pertaining to a fixed base available to a resident of a Contracting State in the other Contracting State for the purpose of performing independent personal services, including such gains from the alienation of such a permanent establishment (alone or together with the whole enterprise) or of such a fixed base, may be taxed in that other State.

Notwithstanding the provisions of paragraph (2) of this Article, capital gains from the alienation of ships and aircraft operated in international traffic and movable property pertaining to the operation of such ships and aircraft shall be taxable only in the Contracting State in which the place of effective management of the enterprise is situated.

Capital gains from the alienation of any property other than that mentioned in paragraphs (1), (2) and (3) of this Article shall be taxable only in the Contracting State of which the alienator is a resident.

The provisions of paragraph 4 of this Article shall not affect the right of a Contracting State to levy according to its law a tax on capital gains from the alienation of any property derived by an individual who is a resident of the other Contracting State and has been a resident of the first-mentioned Contracting State at any time during the five years immediately preceding the alienation of the property.

The first, italicised, version of paragraph 5 is that which was in force in and for the tax year in question in this case. The underlined, non-italicised version is one substituted later. Mr Prosser QC, who appears for the taxpayer, says that the later version was introduced precisely in order to counter “round the world” schemes of the kind carried out here. He claims to derive some modest support for his interpretation of the effect of paragraph 4 from the fact that the later provision was designed specifically to counter it. However, I do not think that the subsequent legislation gives much support to arguments one way or the other about the effect of the legislation (technically a treaty) in its former form. It is the earlier words which I must construe, and subsequent amendments do not really help that exercise.

Article 24 deals with “Elimination of double taxation”. It is the case of HMRC that this is the provision that resolves the risk of any double taxation between England and Mauritius in circumstances such as those in this case. Paragraph 1 provides:

Mauritius tax payable under the laws of Mauritius and in accordance with this Convention, whether directly or by deduction, on profits, income or chargeable gains from sources within Mauritius shall be allowed as a credit against any United Kingdom tax computed by reference to the same profits, income or chargeable gains by reference to which the Mauritius tax is computed.

Subject to the provisions of the law of Mauritius regarding the allowance as a credit against Mauritius tax of tax payable in a territory outside Mauritius (which shall not affect the general principle hereof):

The United Kingdom tax payable under the laws of the United Kingdom and in accordance with this Convention, whether directly or by deduction, on profits, income or chargeable gains from sources within the United Kingdom shall be allowed as a credit against any Mauritius tax computed by reference to the same profits, income or chargeable gains by reference to which the United Kingdom tax is computed.

The taxpayer contends that the scheme as outlined above was effective. There could be no charge under s.86 because the trustees were UK resident for part of the tax year. The only applicable taxing provision could be s.77, but that did not have any chargeable gains on which to operate because Article 13(4) prevented there being any such gains. The effect of Article 13(4) was to make chargeable gains taxable only in the Contracting State in which the disponor was resident. The requirement or condition of Art 13(4) is something that has to be fulfilled and judged at a point in time – in this case, the date of the disposal. The disponor was the trustee of the settlement who, at the time of the disposal, was resident in Mauritius. It is common ground that Mauritius does not have any taxation on Capital Gains, so that gains were not taxable there. Since at that time the trustee was not resident in the UK, Art 13(4) prevented the UK from charging. Accordingly, there could be no s.77 charge either. This argument, which will be elaborated below, depends on its being correct to consider residence as a fact under Article 13(4) at a particular moment in time, namely the time of disposition.

HMRC, represented by Mr Timothy Brennan QC, claimed that Mr Prosser’s interpretation of paragraph 4 was erroneous. It is not relevant to look at the date of the disposition for the purposes of assessing the relevant residence of the disponer. That was not how residence was used in the Convention. The residence requirement was one which depended on a liability to tax which arose through residence. The trustee was, when thus viewed, resident in the UK for the period from 2^nd March 2001, because at that point the trustee (in the form of the Smallwoods) acquired a UK residence which made them liable for tax on the basis of residence; and the trustee (in the form of PMIL) had been a Mauritian resident in the preceding period, which made it liable for tax on that basis. Therefore there were successive periods of residence, each capable of giving rise to a charge to tax. Any conflicts which thereby arose were to be resolved through the mechanism of credits provided by Article 24, not via the tie-breaker provisions.

Neither party put forward a case which necessarily involved the invocation of the tie-breaker mechanisms in Article 4, but if the position was that that mechanism had to be invoked, then each party said that, on the facts, the place of effective management (“POEM”) was the one which favoured them – Mauritius for the appellant taxpayer and the UK for the respondent HMRC.

At paragraph 9 of the Commissioners’ decision they identified what they said were the two questions which were raised. They said they were:

It should be noted that HMRC says that question 1 was not the right question. At least in its present analysis, it is said that that is an irrelevant question.

In the paragraphs which follow, the Commissioners set out detailed findings of fact in relation to the circumstances in which the scheme was conceived and implemented. They had heard a significant amount of oral evidence, and had a lot of documentary evidence as well. I shall not set out, or summarise, that evidence at this stage in the judgment. Then, having set out various relevant statutory provisions, they turn to the contentions of the parties. They recorded them thus:

At paragraph 88 the Commissioners consider the question of “The Interaction of Residence and Chargeability”. Paragraph 88 records that both parties contended that during the Mauritius Period the trustee was not “liable to taxation therein [in the United Kingdom] by reason of residence” because at that time the trustee has not resided in the United Kingdom and one cannot use hindsight to say that residing in the United Kingdom in the UK Period means that the trustees are liable to tax for the Mauritius Period by reason of residence. Mr Prosser puts this in two ways: first that one should look only at the time of the alienation to determine chargeability (and also the effect of Article 13(4)), and secondly (which is also Mr Brennan’s contention) that one should look at the whole of the Mauritius Period for both purposes. The Tribunal put forward a third possibility to the parties that liability to tax is equated to Treaty Residence in Article 4(1) and in looking at the whole tax year one must necessarily use hindsight to say that by virtue of residing in the United Kingdom in the UK period the trustees are liable to tax in the tax year.

Having then set out part of the Commentary to Article 4 in the model convention upon which the United Kingdom/Mauritius arrangements were based, at paragraph 90 the Commissioners observe:

At paragraph 93 the Commissioners embark on giving their “Reasons for Decision”. They start by considering the approach to interpreting a treaty; the parties do not raise any criticisms of that approach. At paragraph 100 they observe:

In the next paragraph they observe:

Then the Commissioners go on to consider the important question of “the effect of Article 13(4)”. In paragraph 103 they agreed with HMRC’s submission that:

This paragraph seems to me to contain reasoning which is questionable both on the grounds of logic and circularity. Neither Mr Brennan nor Mr Prosser supported this reasoning.

Having thus set up the potential for dual residence, the Commissioners then set about resolving it by applying the tie-breaker. They looked at POEM. They dismissed the idea that POEM was the same as “central management and control”, a concept found elsewhere, and recorded that:

Having done that, the Commissioners then considered the facts. The material facts and findings seemed to be as follows:

There is no doubt that all the actions of PMIL in Mauritius were carried out correctly and were well-documented. The appropriate meetings were held and appropriate resolutions taken. The trust was registered, a bank account was opened, and a tax residency certificate was obtained.

Nevertheless, during the Mauritius Period the influence of Mr Smallwood and the guiding hand of two KPMG Bristol representatives (Mr Gadd and Mr Turbervill) were evident throughout.

The evidence showed that KPMG Bristol saw themselves as tax advisers to the Trustees of the Trust, being first Lutea and then PMIL. There was no formal appointment by PMIL but that was not regarded as necessary as both KPMG Bristol and KPMG Mauritius were “under the KPMG umbrella”. (See paragraph 137)

The tax paying scheme was devised by KPMG Bristol. Mr Turbervill of that firm approached PMIL and told them about the tax planning proposals, and made it clear that “the confident expectation was that the shares would be sold before 5 April 2001”. (Paragraph 138)

The Commissioners accepted evidence that there was no agreement that PMIL would behave in a certain way or make certain decisions; PMIL would only act within the context of what it was allowed to do and its duties as trustee were laid down in legislation. They also accepted evidence that the whole point of the exercise was to sell the shares and to realise the gain and to avoid tax on it. (Paragraph 139)

“140. The facts surrounding the appointment of PMIL lead us to the view that the real top level management, or the realistic, positive management of the Trust, remained in the United Kingdom. We accept that the administration of the Trust moved to Mauritius, but in our view the ‘key’ decisions were made in the United Kingdom.”

This view was said to be confirmed by subsequent events. The sale of the FirstGroup shares was not an isolated decision taken by PMIL; it had been carefully arranged beforehand by the transfer of shares to the brokers nominee account.

The Commissioners accepted the evidence of Mr Jingree of PMIL that the sale of the shares was motivated by United Kingdom tax planning reasons. However, if it had not been in the interests of the beneficiaries and the Trust, the trustees would not have sold the shares. Similarly, if the share price dropped dramatically, and if the fund manager had advised against the sale, then the trustees would not have decided to sell.

The Commissioners accepted that the decision to sell the shares was taken at a telephone meeting between the directors of PMIL on 10^th January 2001 (see paragraph 143). They also accepted that if the price of shares had fallen so as to remove any gain, then the shares would not have been sold. “Nevertheless, in our view, this was a lower level investment decision as there was no doubt that the shares would be sold; the real top level management decisions, or the realistic, positive management decisions of the Trust, to dispose of all the shares in a tax efficient way, had already been, and continued to be, taken in the United Kingdom. The ‘key’ decisions were made in the United Kingdom.” (Paragraph 143)

“We conclude that the state in which the real top level management, or the realistic, positive management of the Trust, or the place where key management and commercial decisions that were necessary for the conduct of the Trust’s business were in substance made, and the place where the actions to be taken by the entity as a whole were, in fact, determined between 19 December 2000 and 2 March 2001 was the United Kingdom.” (Paragraph 145)

Thus the tie-breaker resolved residence in favour of the United Kingdom and the Special Commissioners dismissed the taxpayer’s appeal.

Mr Prosser’s case was that the Commissioners analysed the legislation incorrectly, and one never got as far as the tie-breaker to determine residence. Alternatively if one did then one should arrive at an answer which was the opposite to that of the Commissioners. He said that the Commissioners erred in concluding that, on the facts, there was a dual residence (or potential for dual residence) at any time. What Article 13 does is to allocate the right to tax among the candidate states, and the effect of paragraph (4) is to allocate the right to tax capital gains (other than gains arising out of the transactions referred to in the earlier paragraphs) to the country in which the taxpayer is resident (as residence is defined in the Treaty). The critical time at which the snapshot is taken is the time of the alienation, (or gain, which is the same on the facts of this case). One judges residence as at that date. Once it is determined, that is an end of that particular inquiry. No other state can come into the picture as a potential state of residence, based on subsequent (or indeed prior) events. On the facts of this case that country can only be Mauritius. It matters not that the trustees became resident in the UK a short time thereafter. The snapshot had already been taken. Since there was no competition as at the date of the snapshot there was no scope for applying the tie-breaker – there was no tie to break. Thus the Commissioners were wrong to come up with an analysis in which there were two concurrent residences which competed and which had to be separated by the tie-breaker (essentially paragraph 107 of the decision below). The snapshot resulted in a picture of Mauritian residence. Article 13(4) thereby allocated the right to tax exclusively to Mauritius. Therefore the UK could not tax.

HMRC’s argument repudiates the snapshot analysis. The purposes of the Treaty, according to the heading to the Convention as set out in the Schedule to the statutory instrument, included the prevention of fiscal evasion. If the taxpayer’s arguments were right it would have facilitated, or at least promoted, evasion. Furthermore, Article 13(4), as interpreted by the taxpayer, would not have the effect of allocating taxation to a state by reference to residence; it would have taken away a right to tax by reference to residence. There was an interpretation which avoided these consequences. What Article 13(4) is intended to do is to resolve conflicts between taxation rights based on situs and those based on residence. In relation to taxation which does not fall within paragraphs (1) to (3) it does it by asking whether each state, according to its own laws, has the right to tax on the basis of residence. If the answer is No for one state and Yes for the other then (if the earlier provisions of Article 13 do not apply) the state which taxes on the basis of residence wins. If the answer is yes for both because there were successive (as opposed to concurrent) periods of residence, then there is a conflict. That is resolved by applying Article 24. One does not resolve it by applying the tie-breaker. It is, I think, implicit in these submissions that the tie-breaker would apply only if there was a factual case for saying that at a given point in time each state would, in relation to that period, say that the taxpayer was resident in its jurisdiction, but that is not the case here. In the present case one asks two questions. Did Mauritius have the right to tax based on residence: Answer, Yes (though it did not exercise it). Did the UK have the right to tax based on residence? Answer, Yes (by virtue of the statutory provisions referred to above and the residence here of the trustees from March onwards in the relevant year). There were thus two states claiming the right to tax on the basis of residence, both are entitled to do so, and Article 24 is invoked to resolve the conflict. Since there was no Mauritius tax to deduct from the UK tax, the UK tax was recoverable in full.

It will be apparent from these analyses that neither adopts the reasoning of the Commissioners as identified above. I have set out their core paragraph (paragraph 107) above. It is not clear to me how they arrive at their conclusion, which involves turning factual residence for part of the year into deemed residence (for tax purposes) for the whole of the year. Thus residence in Mauritius for part of the year turns into deemed residence for the whole of the year; and the same with UK residence. This, I think, is their concept of Treaty Residence. I am afraid I cannot see how it is justified by the wording of the Treaty. There is nothing in the Treaty that requires it; nor is there anything in the UK tax legislation which requires it either. The UK tax provisions set out above create certain tax consequences for gains in a given year if trustees, or beneficiaries, are resident here for part of the year. Where they apply, residence in part of the year gives rise to a charge to tax on gains made in another part of the year, but they do not do so by deeming the residence to be for any period longer than the actual period of residence. They do so simply by defining the gains by reference to the period in which they arise. There is a difference between those approaches, and I do not consider this distinction to be too subtle for the purposes of the Treaty (pace the Commissioners). There is no reason why the Treaty should not acknowledge it, without the need for the creation of the artificial deeming concept created by the Commissioners. There is therefore no warrant for the construction of the Commissioners, and the parties are right to disclaim it. I think that the true analysis lies elsewhere.

The difference between the parties lies in their respective approaches to Article 13. The taxpayer submits that Article 13 allocates the right to tax as between states. HMRC says that it merely defines the permitted (and forbidden) basis of taxation and allocates taxation as between situs-based rights and residence-based rights – it specifies taxation systems, if you like. It tells you which gains are taxable on a situs basis, and which are taxable on a residence basis. Having established that a residence test applies, one looks for the state or states in which the taxpayer is resident, and that state, or those states, may tax. Here there are two states, and the competition is resolved by invoking Article 24. Thus HMRC gets to its conclusion from its starting point. It is necessary to consider which starting point is correct, because if the taxpayer is correct then HMRC’s chain of reasoning does not work.

My first observation is that at one level there is an element of truth (or accuracy) in HMRC’s proposition. If the gains in question can be taxed only in the state in which a taxpayer is resident, then obviously it is posing a residence-based test rather than any other sort of test. But in my view that is just a truism, and merely describes the effect of the paragraph. It does not necessarily describe its object, which is what it would have to do if HMRC’s submissions are to work.

The wording of the paragraph does not suggest a lot of support for HMRC’s analysis. While most arguments on all sides of a debate about interpretation can be supported by the submission “If the draftsman had meant … he could easily have expressed his intention thus …”, or variants of it, and that sort of argument does not always advance the debate much, I do think that in the present case a more appropriate form of wording would have been available and used if the Treaty draftsman had intended HMRC’s effect. He would have said “shall only be taxable on the basis of the residence of the taxpayer”, or something like that. The actual wording used does not suggest that result. Paragraph (1) is permissive – gains “may” be taxed – but is plainly referring to only one state because there can only be one state of situs. Paragraph (2) is similarly permissive in tone, and does not assist on the present point. Paragraph 3 is more prescriptive – “shall be taxed only in the Contracting State …”. That again suggests one state. The wording of paragraph (4) suggests the same – “taxable only in the Contracting State of which the alienator is resident”. That smacks much more of an attempt to find one state only, and not merely a basis of taxation.

That view is supported by a review of the rest of the treaty. Articles 6 to 22 deal with categories of potentially taxable material, or potentially taxable people. Some of the permitted bases of taxation are permissive – “may” – but there are a lot of references to “only”. “Only” appears in Articles 7(1), 8(1), 14(1), 15(1), 15(2), 18(2), 19(1), 19(2) and 22(1). Putting those provisions in their contexts, and considering their wording and effect alongside the more permissive regime for other items, one gets a clear impression of an intention to find one jurisdiction only in respect of which the relevant charge to tax can be made. In respect of all of them it is hard to treat the provisions as choosing a basis of taxation, conceptually speaking, as opposed to an actual single state. It is true that the contexts are different, and not all of them are residence-related tests, but nevertheless the impression is, to my eyes clear.

This analysis supports the submissions of the taxpayer that Article 13(4) is doing the same thing. It is pointing to a single jurisdiction in which tax can be charged, and that is the state of residence. In order to make that workable one has to find a date at which residence has to be judged. There is no other realistic candidate for that point of time other than the date the gain arose (or the date of the disposition, which in this case is probably the same point of time). If there is competition between both states in relation to that point of time, then the tie-breaker applies to produce a single state in respect of which residence (in the status sense used in Article 4) exists.

This analysis is further supported by two features relied on by Mr Prosser. First, there is the old Article 13(5). Mr Prosser submitted that paragraph (5) was a “waste of ink” if Mr Brennan’s argument was correct. If all that was required was a right to tax by virtue of residence, then it served no useful purpose to refer to a state of residence some 5 years before the fiscal year of alienation. The mere fact of residence (for tax purposes) would be sufficient by itself anyway. By the time of his oral submissions Mr Brennan was minded to accept that if his argument was correct, Article 13(5) would not be necessary, but said it served the useful function of making clear that which would otherwise be implied. I do not think that Mr Brennan’s way of viewing the matter is realistic. Paragraph (5) was presumably intended to add something which would not otherwise exist. It demonstrates that the draftsman (or the Contracting States) thought that residence in those earlier 5 years would not be enough, whereas on Mr Brennan’s argument it would be. Its existence tends to support Mr Prosser’s analysis.

The other significant point favouring Mr Prosser’s analysis is Article 24, which Mr Prosser submits does not work properly (or at all) if one is trying to resolve a conflict between two charges based on residence. I think that he is right about that, and Mr Brennan went so far as to accept that its operation is “a bit obscure” in relation to resolving the competition where two states seek to charge on the basis of successive periods of residence (which is his case as to what should happen here). I do not think that “a bit obscure” really does it justice. It does not really work. The essence of the provision is as follows. Each of the UK and Mauritius will give credit for tax paid on sources from within the other territory. Thus the UK gives credit for Mauritius tax paid where the source of the tax was in Mauritius, and vice versa. Residence is used to help resolve any difficulties of where to start – see Art 24(4). It presupposes that a taxpayer is resident in one, and one only, Contracting State. The presupposition is not explicit, but looking at the mechanics it is apparent that it cannot be made to work if the taxpayer is resident in both states. So Mr Brennan’s own resolving mechanism is against him. The Commissioners thought that Article 24 could be made to work by a process of iteration. No-one sought to explain to me how that could work, and I do not see how it can. If a process of iteration is to work, one has to have a starting point. The starting point is not clear – is it to be Mauritius or the UK? That may not matter if there is a process of iteration which gives the same result no matter where one starts, but I do not understand what that process is. And if it were to work I cannot accept that this sort of process, which might be entertaining to mathematicians but not to anyone else, is a sensible basis for construing a Treaty which is supposed to have a practical application. Nor is it practicable to look to Article 27 as a provision of last resort in this sort of case, which Mr Brennan says it is. I have not set it out, but it provides for the resolution of disputes by dealings between the States. The draftsman (or the States) cannot sensibly have thought that this was a practical way of resolving an Article 24 dispute in circumstances such as these. There is an alternative approach which makes that unnecessary, and is therefore more attractive as an answer to the problem of construction. That answer is Mr Prosser’s approach.

As a factor weighing against Mr Prosser’s approach, Mr Brennan sought to show that this construction was contrary to the way in which English capital gains tax actually worked. He pointed out that capital gains tax operated on gains and losses within a tax year, not on a transaction by transaction basis. By using examples, he demonstrated that Mr Prosser’s construction would produce what were said to be anomalous results if one had gains in a Mauritius period and losses in a UK residence period (all within the same UK tax year), and vice versa. Mr Prosser’s answer to this was, in part, to rely on English authority which held that the Treaty should be interpreted “unconstrained by technical rules of English law, or by English legal precedent” (James Buchanan & Co Ltd v Babco Forwarding & Shipping (UK) Ltd [1978] AC 141 at 152). I do not think that that authority helps him. What it does is set limits on the usefulness of some of the traditional English law ways of interpreting legislation rather than barring looking at the surrounding legal playing field in order to work out what the Treaty might mean. However, having said that, I do not find the detailed operation of English capital gains tax to be much help in deciding this point. While I am prepared to assume for these purposes that one might get some assistance from an apparent complete inconsistency between an essential feature of UK taxation and the way the Treaty is said to operate, (though I am not sure even about that) the points relied on by Mr Brennan do not go that far. In the case of the sort of alleged oddities relied on by Mr Brennan, they are oddities arising from the fact that the Treaty sits as an arrangement above the taxation statutes, and inevitably has some effects on how they operate. They are the sort of things which will have to be worked out rather than clear pointers as to what the Treaty means.

Nor did I gain any assistance from Mr Brennan’s submission that, if the taxpayer were right, the Treaty had had the effect of removing a right of taxation based on residence when its intention had been to preserve rights based on residence. That seems to me to be a misreading of the situation. The Treaty creates, or provides for, a right based on residence (as residence is defined within it), but contains certain rules or structures for attributing residence and resolving potential conflicts between states which might be said to compete for residence. One of those rules or structures is the need to determine the time in relation to which residence is treated as being relevant. It operates accordingly. One applies those rules, and comes up with the taxpayer’s solution. That is not inconsistent with the purpose of the Treaty; it is implementing it. It is true that, in this case, the effect is that the gain escapes tax, but that is because Mauritius does not impose a capital gains tax in respect of those gains. The tax, when properly based on residence, is nil.

Both sides referred me to the OECD Model Convention, on which the Treaty is based, together with the official commentary on that and the additional commentary on those works of Philip Baker QC in his manual on Double Taxation Conventions. I am afraid that I did not find most of those citations helpful, because they tended not to address the points that I have to decide. Accordingly, I do not reproduce most of those citations here. However, in his commentary Mr Baker said (at pages 4-2/10 to 4-2/11):

That, of course, is less than helpful to a consideration of the present case. He goes on to give examples of an individual shifting residence during the course of overlapping tax years in two separate countries with an alienation of an asset in one state during the period of overlap, and goes on:

Unfortunately for this case, he does not hazard an answer to that question. However, his analysis does support part of my reasoning above. He distinguishes between cases where both states adopt a split year approach, and where they adopt what might be called a “deemed residence” approach. In the former case there is apparently no tie to break. That is in fact the case before me. What the Commissioners did in their decision was to create the second without any statutory justification. The statute does not deem the trustees to have been resident in any part of the tax year other than that in which they were actually resident within the meaning of the taxing statute. It seeks to charge gains made at any time in that year, which is different. The answer to Mr Baker’s unanswered question is, on the facts of this case, answer (a), for the reasons given above.

On this part of the appeal I therefore conclude as follows:

The Commissioners erred in creating a simultaneous residence for the trustees, spanning the Mauritian period.

The correct analysis is that there were three periods of successive residence in the relevant UK tax year – Jersey, Mauritius and then the UK.

Article 13(4) gives the right to tax capital gains to the state in which there was residence at the time of the disposition.

That state was, at that date, Mauritius.

Since there were no two jurisdictions vying for a claim of residence in that period, there is no tie for Article 4 to break.

Accordingly, Mauritius has the right to tax and the UK does not.

The need to determine the POEM arose on the basis of the Commissioners’ findings because they determined that there was dual residence during the period covering the disposition. That gave rise to the competition between states of residence that Article 4(3) is intended to resolve, and they resolved it accordingly. The taxpayer says that they got the answer wrong – the POEM of the trust was, during the Mauritius period, in Mauritius. The detailed analysis of the facts in which the Commissioners engaged was not, as a primary fact finding exercise, materially challenged by the taxpayer. It was the conclusions that were challenged – it was said that the only conclusion open to the Special Commissioners was that the POEM of the trustee was Mauritius.

In the light of the respective positions of the parties when measured against what I have found to be the correct way of applying the treaty, this limb of the appeal does not arise. Article 4(3) only arises if there is a relevant competition between two relevant residence cases. That competition is only meaningful if you are measuring those two cases in relation to the same point of time. Mr Brennan conceded that a POEM assessment was a snapshot assessment, contrary to his case on residence itself (though one might have to modify the metaphor to accommodate the possibility, in some cases, that you might be resolving the question across a relevant period of time longer than the period of a single and simple disposal transaction). I have held that the relevant point of time is the date of the disposal in question. At that time the only place in which the trustee was resident, in the sense provided by Article 4(1), was Mauritius. At that time there was no case for saying that there was a UK residence. The fact that the trustee subsequently became UK resident does not throw that residence back into the Mauritius period so as to generate a competition. So there was only one place of residence, and no need to invoke POEM as a tie-breaker. This would in fact be so even if Mr Brennan’s analysis were correct, since he alleges no more than consecutive, and not concurrent, periods of residence.

In those circumstances the POEM question does not arise, and I shall not decide it.

In the circumstances, therefore, this appeal succeeds and will be allowed.