The Commissioners for HMRC v GE Financial Investments

The FTT’s decision ([2021] UKFTT 0210 (TC), Judge Brooks) sets out at [3] and [4] a detailed statement of agreed facts. The critical aspects have already been summarised, but the agreed facts also include that the LP was dissolved in 2009 and that the stapling provision was then removed from GEFI’s Articles. GEFI was also re-registered as an unlimited company. Further, some details are set out of a failed attempt by the US and UK tax authorities to reach agreement as to GEFI’s residence for Convention purposes and an unsuccessful attempt by GEFI to procure agreement between the two authorities under the mutual agreement procedure in Article 26.

The FTT’s decision also sets out at [8] more detail of the LP’s activities, taken from a chronology that had been sent to HMRC. This included additional detail about Loans 1 to 5 and also payment instructions under which cash was moved, in particular partnership distributions to enable tax liabilities to be met, and details of resolutions passed by GEFI Inc to ratify past activity. The judge then went on to describe, with extensive quotations, contemporaneous exchanges which in the judge’s view indicated the “extent of the participation in the loan transactions by the various entities and the individuals involved” ([9]).

The judge then considered expert evidence on US tax law and Delaware law. Each party engaged an expert on the former and GEFI also put forward expert evidence on the latter. No oral evidence was provided by the experts. As far as tax is concerned a joint memorandum confirmed that there were no material points of disagreement between the two experts. The findings made about US tax are set out above.

The expert evidence on Delaware law confirmed among other things that the LP was a separate legal entity, that its general partner was its agent and liable for its obligations, and that a limited partner was not so liable unless it was also a general partner or participated in the control of the partnership business. As the UT noted, however, there is no indication that there is a legal requirement for a Delaware limited partnership to conduct business.

On Issue 1, the judge concluded at [61]-[66] that GEFI was not US resident for the purposes of Article 4(1) of the Convention. The judge accepted HMRC’s submissions that, for residence to be established, Article 4(1) of the Convention required both liability to worldwide taxation and a “connection” or “attachment” to the country concerned. Worldwide taxation was a necessary but not a sufficient condition. Further, the share staple did not represent a connection with the United States, as opposed to a connection to GEFI Inc. In reaching this conclusion the FTT took account of the differences of treatment it had identified between US domestic corporations and stapled foreign entities for federal income tax purposes.

The judge also decided that GEFI was not carrying on business. In reaching that conclusion the judge referred to a number of cases that had considered the meaning of “business” in various contexts, including Ramsay v HMRC [2013] UKUT 226 (TCC), [2013] STC 1764 (“Ramsay” – not to be confused with the much more famous House of Lords decision with that name). He agreed that GEFI’s objects should be taken into account (contrary to HMRC’s submission that the sole focus was the LP) but considered that they were not determinative. The judge went on:

The judge then gave an example that a director of GEFI Inc had been content to sign a loan agreement for a substantial sum, only asking for the reason “seemingly as an afterthought and ‘out of curiosity’”. The judge also referred to the contemporaneous emails indicating that it was members of GE’s tax department rather than directors of GEFI Inc that had been concerned with the loan arrangements and their terms, a comment by one member of the tax team, a Mr Tomasetti, indicative of a lack of business purpose in relation to steps taken in 2004 and a query raised within the tax team about replacing a director of GEFI Inc. The judge then went on:

As already indicated the judge went on to consider what was then Issue 2(b) and concluded it in GEFI’s favour.

The UT reversed the FTT’s decision on Issue 1, making it strictly unnecessary to consider Issue 2. However, taking account of the possibility of a further appeal the UT helpfully did consider what was then Issue 2(a), and concluded that the FTT was entitled to reach the conclusion that GEFI was not carrying on business.

After considering the facts and the FTT’s decision, the UT discussed UK and US domestic law, in the latter case by reference to the expert reports rather than the FTT’s findings, concluding at [76(5)] that:

The UT went on to consider the MTC and OECD Commentary on it in detail (including the historical development of the residence article), before turning to the Convention. It concluded at [113] that there was no support for the view that only particular types of attachment were sufficient for a person to be regarded as resident in a Contracting State for MTC purposes.

As to the Convention, the UT rejected HMRC’s argument that, unlike incorporation, share stapling did not provide the requisite connection to establish US residence. It took account of the Memorandum of Understanding accompanying the US/Netherlands Treaty referred to in the FTT’s findings (see [‎32] above) as “contextual background”, although it said at [123] that it placed little weight on it.

The UT also considered Crown Forest and Alta Energy in some detail. It concluded at [130] that Crown Forest was authority for the proposition that “full” taxation was the only connecting factor to establish residence under the US/Canada treaty. It considered that Alta Energy was to a similar effect, establishing that the OECD test rested solely on domestic law tests of residence, such that if a person was liable to full taxation under domestic law then they were resident for treaty purposes, formal criteria such as incorporation being commonly used in addition to factual criteria such as place of management ([136], [137] and [151]).

Having determined that persons deemed to be resident for domestic law are treated for MTC purposes as resident for treaty purposes, that incorporation is commonly treated as establishing residence despite being a “formal” rather than a factual or territorial criterion, that the stapling provisions had a similar effect to a provision deeming a stapled entity to be US incorporated, or alternatively share stapling was a “criterion of a similar nature” without the need for an additional legal connection, and that the United Kingdom was to be presumed to be aware of the share stapling rule when it entered into the Convention, the UT concluded that GEFI was a resident of the United States for Convention purposes ([138]-[152]).

As to whether GEFI carried on business in the United States, the UT considered the authorities in some detail. It concluded at [232] that, on a fair reading of the FTT’s decision, the FTT had considered the principles to be derived from them. It noted that the FTT took account of GEFI’s objects but commented that it would have been entitled to disregard them because of the role of the LP, and decided at [238] that what critically weighed in the balance for the FTT was what GEFI (or more accurately the LP on its behalf) had actually done. Read as a whole, the FTT also did not wrongly apply the US test of carrying on business ([240]). The FTT had been entitled to accept HMRC’s submission that GEFI acted as a conduit ([243]). Any errors made were not material ([245]).

Before going on I should make one point about nomenclature. I have, like the tribunals below, used the term “worldwide” as a convenient way to refer to the taxation imposed by the United Kingdom and the United States on their residents. It should be understood to mean a form of tax that is unrestricted by the source of the income, as opposed to taxes imposed on the income of non-residents, which are generally linked to a source in the relevant State. The distinction could be seen as being between tax by reference to the attributes of the person and tax by reference to the location of the source. The use of the term “worldwide” gives rise to no issue on the facts of this case, because it reflects the tax systems of both parties to the Convention. But as the OECD Commentary recognises some States adopt a more territorial approach even as regards their own residents, and in such cases a different label such as “residence-basis taxation” may be more appropriate.

The correct starting point in determining Issue 1 is to consider the words used in the Convention to define “resident of a Contracting State”. By way of recap, Article 4(1) provides:

It is incontrovertible that GEFI cannot in fact be regarded as falling within any of the enumerated criteria of domicile, residence, citizenship, place of management or place of incorporation as far as the United States is concerned. This is irrespective of whether those concepts are interpreted in accordance with a) UK law, as Ms McCarthy submitted was the effect of Article 3(2), set out at [‎25] above, since we are concerned with the application of the Convention for UK tax purposes, b) US law on the basis that Article 3(2) is subject to the proviso “unless the context otherwise requires” and Article 4(1) expressly refers to the “laws of that State”, or c) as Mr Baker submitted, in accordance with a combination of UK law, US law and the OECD Commentary.

If US law was relevant in applying Article 4(1) to the facts of this case then, in the absence of any evidence of US legal consequences of stapling beyond the effect of s.269B, we must assume that the relevant legal principles are the same as those applying in this jurisdiction. The critical point is that GEFI is in fact UK incorporated. It is liable to tax in the United States “by reason of” the application of s.269B, which provides that what it explicitly recognises to be a foreign corporation (that is, one not created or organised under US law) “shall be treated” as if it were a domestic corporation. It is not liable to tax by reason of actual incorporation in the United States or by virtue of any of the other enumerated criteria listed in Article 4(1). Although at one point the UT appeared to suggest that GEFI should effectively be assumed to be incorporated in the United States by virtue of s.269B, I did not understand Mr Baker to rely on such an analysis and I cannot see that it would have any evidential support.

Even if s.269B provided in terms that, where it applied, the stapled foreign corporation was to be deemed to be a domestic corporation, and (contrary to Ms McCarthy’s submission) US tax principles were to be applied in interpreting the concept of “place of incorporation”, I cannot see that it would assist GEFI. In Fowler, the Supreme Court considered whether UK tax legislation which “treated” Mr Fowler as self-employed for income tax purposes when working as a deep-sea diver in the UK continental shelf applied for the purposes of the UK/South Africa double tax convention in circumstances where he was in fact in employment. Arguably Fowler was a stronger case for applying the statutory fiction because it was clear in that case that Article 3(2) required UK tax principles to be applied when interpreting terms such as “employment”: see Lord Briggs’ judgment at [9]. Nevertheless the Supreme Court held that the fiction of self-employment did not apply for treaty purposes. Rather, the relevant articles had to be “applied to the real world” and this was unaffected by Article 3(2), which instead required the relevant terms to have their “settled meaning…viewed from the perspective of UK tax law” ([30]-[32]).

As Ms McCarthy submitted, both the provision considered in Fowler (s.15 Income Tax (Trading and Other Income) Act 2005) and s.269B are the type of deeming provisions that deem something to be what it clearly is not (akin to deeming black to be white). Neither s.15 nor s.269B is the sort of provision that might deal with uncertainty at the margins, where there would be much more scope for saying that the treatment or characterisation provided by the provision should apply when interpreting the treaty. This point was picked up by Lord Briggs at [25] where he said:

On the face of it, therefore, GEFI can be regarded as US tax resident for Convention purposes only if the requirements of s.269B amount to a criterion of a similar nature. That is indeed GEFI’s fallback argument (which I address at [‎121]-[‎123] below), but its primary argument is that all that the underlined words do is to establish that taxation on a worldwide basis, rather than a source basis, is required.

In my view that is not the natural meaning of the words used. Article 4(1) defines a resident of a Contracting State as a person “liable to tax… by reason of” having a particular status (domicile, residence etc.). The words “by reason of” make clear that liability to tax is the consequence of having the requisite status. It does not say that residence status is the result of, or equates to, being liable to tax. Further, the status in question comprises a list of specific connecting factors following by an express ejusdem generis provision. The fact that a list is included suggests that the individual items referred to in it were regarded as having significance. The choice of the words “of a similar nature”, meaning (broadly) of the same kind or genus, is also relevant. Each of the listed factors is a type of substantive factual or legal connection between the person concerned and the State in question, strongly indicating that for another factor to be something of a similar nature it would also need to be a connection that has a similar character or quality.

I agree with Ms McCarthy that if all that was intended to be covered was anything that resulted in worldwide taxation under domestic law then there would be no need for any specific list. The underlined text would much more straightforwardly have referred simply to unrestricted or worldwide taxation. If a list was nonetheless included then the words “of a similar nature” would have read something like “to the same effect” or “having the same consequence”.

It is also relevant that the Convention includes place of incorporation and citizenship in the list of connecting factors. This is a departure from the MTC, which refers to “…domicile, residence, place of management or any other criterion of a similar nature…”. The additional reference to place of incorporation reflects a reservation by the United States which is set out in the OECD Commentary. Neither the addition nor the reservation would be required if the list was no more than a proxy for worldwide taxation.

Turning to the context, the primary context is of course the other terms of the Convention, starting with Article 4 itself. I agree with Ms McCarthy’s submission that they tend to support HMRC’s case that the focus is on determining residence by reference to factual connecting factors, or what she termed for ease of reference a “relevant local connection”. Article 4(2) provides that US citizenship or a “green card” is not enough to establish US residence in the absence of another connection such as a permanent home. The “tie-breaker” for dual-resident individuals in Article 4(4) also operates by reference to the relative strength of connections. Article 4(6) effectively switches off the domicile of dependency that US nationals obtained by marrying UK domiciled men before 1974, so allowing their domicile to be determined by reference to their actual circumstances.

At first sight the second sentence of Article 4(1) also supports HMRC’s interpretation, but in fact Ms McCarthy did not seek to rely on it. The reason that it seems to assist is that it expressly provides that persons subject only to source taxation are not to be treated as resident. That appears to lead to a strong inference that the first sentence is doing more than simply saying that residence covers anyone who is subject to worldwide taxation, because if it was just saying that then the second sentence would be redundant.

However, the history of the MTC equivalent of Article 4(1) is that the second sentence was not initially included. When it was included it was in the following terms:

The OECD Commentary explains (para. 8 of the 2000 version) that its inclusion was intended to apply to persons such as foreign diplomats who are considered to be resident in a Contracting State but are subject to tax only from sources within the State.

I would make two observations on this. First, the introduction of the second sentence is suggestive of Ms McCarthy’s approach: if the first sentence is longhand for worldwide taxation rather than requiring a real local connection such as residence, why was the second sentence needed? Secondly, and more importantly, we must interpret the Convention. The MTC, its history and the OECD Commentary cannot be more than guides. The second sentence in Article 4(1) of the Convention is materially different from its MTC equivalent. While it might indeed apply to foreign diplomats it is unarguably wider because it also applies to profits of a permanent establishment. The only rational reason for that is that the parties considered that it was necessary to depart from the MTC by spelling out that taxation of profits of a permanent establishment in a State is not a sufficient local connection to amount to residence for Convention purposes. If they intended that worldwide taxation was the test of residence then the carve out for permanent establishments would be otiose.

As already explained, the words of the Convention must be interpreted not only in their context but in the light of the Convention’s object and purpose. However, the object and purpose are primarily gleaned from its text, read in the light of the surrounding circumstances: see [‎41] and [‎42] above.

It is trite that the general object of a double tax convention is to relieve double taxation, whether by restricting taxing rights and/or by requiring one State to give credit for tax paid in the other. Article 4 of the Convention is an important provision because the State in which a person is resident for Convention purposes may be granted the primary or sole taxing right, with the taxing right of the other State in respect of income and gains arising in its territory either being restricted or removed entirely.

The function of double tax conventions in apportioning taxing rights between the State of residence and the State of source, and a distinction that is generally drawn between “active” and “passive” income in making that apportionment, was considered in an illuminating passage in Alta Energy in the context of the Canada/Luxembourg treaty (from the judgment of Côté J with whom the majority agreed):

The provisions of the Convention illustrate both the significance of residence and the distinction noted in Alta Energy between active and passive income.

The starting point is Article 1(1), which makes clear that (unless specifically provided otherwise) the Convention applies to residents of one or both States. That is important. The Convention confers material benefits on taxpayers to whom it applies, and the Contracting States have agreed that those benefits should be available only to their respective residents. I pause here to note that, if GEFI’s argument were right, an entity based anywhere in the world the shares in which were stapled to a US entity would have to be granted the benefits of the Convention.

Turning to the specific provisions, and most pertinently on the facts of this case, Article 11 of the Convention entirely removes the source State’s right to tax interest income arising to a resident of the other State unless it is attributable to a business carried on through a permanent establishment in the source State (the classic example of “active” rather than passive income). If a business is carried on through a permanent establishment, then the source State has the primary taxing right under Article 7. Other examples are the dividend provisions (Article 10), where the taxing rights of the source State in respect of residents of the other State are restricted by limiting the rate of any withholding tax, royalties (Article 12) where a similar approach is taken to Article 11, capital gains (Article 13) where again the source State’s taxing rights in respect of residents of the other State are removed except in specified circumstances, and employment (Article 14) which also restricts the ability of the Contracting States to tax each other’s residents.

The importance of the residence concept means that both States have a clear interest in delineating the scope of the other State’s ability to determine who falls within the concept of Treaty residence, because where a person is resident in that other State the source State may well have to cede or at least restrict domestic taxing rights. (Indeed, in the particular context of the Convention the United Kingdom’s interest is arguably heightened for corporate entities because of the fact that the tie-breaker in Article 4(5) depends on mutual agreement, and if agreement cannot be reached it must allow a credit for US tax paid on US source income against UK tax on the same income: [‎22] above.)

Ms McCarthy submitted that if GEFI’s primary argument was right then there would be nothing to prevent a US law being introduced to the effect that all companies with blue logos should be taxed as domestic corporations, which cannot have been intended. While Mr Baker derided this as a ridiculous example, it is not without force and it could readily be replaced by a more realistic example, such as a rule that subsidiaries of US corporations should be taxed as if they were domestic corporations.

Mr Baker’s answer to this was essentially that s.269B was an existing provision of the Internal Revenue Code when the Convention was agreed, and both Contracting States are taken to be aware of each other’s tax systems. Further, future changes would potentially breach the requirement in Article 26 of the Vienna Convention that the parties perform a treaty in good faith and/or would lead to the Convention being terminated or renegotiated.

These arguments do not really answer Ms McCarthy’s point. The first assumes that GEFI is right in its interpretation of Article 4(1), namely that s.269B has the effect of treating it as US resident for Convention purposes. As to the second, it is far from clear to me that a change in the US tax system – which would doubtless be prompted by much broader policy considerations than the impact on a particular double tax convention – would engage Article 26. And a treaty cannot sensibly be interpreted on an assumption that any difficulties would lead to its termination. Rather, it seems to me that a much more sensible interpretation is that reflected in the words used in Article 4(1), namely that it is necessary to identify certain connecting factors between the person and the State in question that justify that State in imposing tax.

A further point that it is convenient to address at this point is GEFI’s argument that a conclusion that s.269B did not make a stapled entity US tax resident for Convention purposes would create much greater uncertainty for the taxpayer since the tie-breaker in Article 4(5) would not be engaged and it would be left without any treaty protection. It would be a “second class resident”. In contrast Ms McCarthy submitted that if all that mattered was worldwide taxation then substantial uncertainty would be created as to the application of the Convention by forcing more taxpayers into having to try to seek agreement under Article 4(5).

I did not find this debate to be of material assistance. If HMRC is right then GEFI is solely UK tax resident for Convention purposes and, subject to Issue 2, should not be liable to US tax on its US source interest income at all, due to the effect of Article 11(1). This is because it would have a full entitlement to the benefits of the Convention. This is despite the “savings clause” in Article 1(4) which permits each State to tax its own residents as if the Convention had not come into effect. That only applies to persons who are resident “as determined under Article 4”, whereas in this scenario GEFI would not be US resident under that Article. In contrast, and absent agreement under Article 4(5), a company that is dual resident for Convention purposes is able to claim only limited benefits under the Convention, albeit that the particular benefit that GEFI seeks happens to be one that would be available.

The position in which GEFI may find itself if HMRC’s appeal is upheld, namely that it is still liable to US tax and must attempt to obtain unilateral relief in the United States for the UK tax, would have nothing to do with the operation of the Convention as interpreted by this court. Rather, it would be the result of the operation of US domestic law, namely the treaty override in s.269B(d) (see above) and/or any different view that the United States may adopt about the correct interpretation of the Convention.

The tie-breaker in Article 4(5) does not assist GEFI. Unlike the 2000 version of the MTC, where the corporate residence tie-breaker turned on the place of effective management of the company, Article 4(5) depends on agreement and, in default of agreement, requires the United Kingdom to allow a credit for US tax paid on US source income against UK tax on the same income ([‎22] above). This rather reinforces the United Kingdom’s interest in the circumstances in which companies can be treated as US tax resident for Convention purposes. Further, it is hard to see how agreement would be likely to be reached other than by a determination of the relative strength of (real) connecting factors. A pure clash of domestic regimes imposing worldwide tax – which is what GEFI says is the only relevant factor – is unlikely to be a promising starting point for a negotiation.

Turning to the MTC more generally, the UT found significant support for its approach in the MTC and OECD Commentary: indeed that was its starting point. I accept that the MTC provides important context and helps explain why, for example, some terminology may be included even if it is not of direct relevance to the tax systems of the two Contracting States in question, so as to minimise departures from the MTC. However, I would reiterate that what must be interpreted is the treaty in issue, not the MTC. For example – and as discussed above in the context of the second sentence of Article 4(1) – it might be that variations to the MTC that are reflected in the treaty could affect its interpretation more generally. Further, the Commentary is just that. It should not be read as if it were a statute and it is important to read comments in it in their context.

Both parties relied on aspects of the 2000 version of the Commentary on Article 4. Read as a whole, I do not consider that it supports GEFI’s interpretation. For example, it refers at para. 3 to the fact that domestic laws “generally” impose comprehensive taxation based on “personal attachment” to the State, possibly extending for example to seafarers on ships with their home harbour in the State (which I would regard as a proxy for a personal attachment). Paragraph 8.2 reflects Article 4(1) in referring to liability to tax “by reason of various criteria”. Paragraph 10 explains how the position of dual-resident individuals is resolved under Article 4(2) by giving the “attachment” to one State a preference over the attachment to the other. Paragraphs 21 to 24 comment on the corporate residence tie-breaker, which under the 2000 version of the MTC was as already mentioned determined by the place of effective management of the company, emphasising the importance of the place where key management and control decisions are taken.

Mr Baker placed particular reliance on para. 4, which states:

I do not consider that these comments bear the weight that Mr Baker sought to place on them. Paragraph 4 is of course accurate as far as it goes but it simply makes the uncontroversial point that whether a person is in fact “resident” in a State is to be determined under that State’s domestic law, rather than under the terms of the treaty. It does not say that any rule that imposes worldwide taxation amounts to residence: if anything it indicates the opposite by saying that it is up to States to determine “the conditions under which” a person is to be regarded as tax resident and “consequently” fully liable to tax. This clearly encompasses an ability to deal with cases at the margins, such as the seafarer already alluded to, but there is no indication that it goes beyond that.

The reference to deeming in para. 8 must be read in its context. There is an obvious link between diplomats (and others sent abroad on government service) and their sending State. Again, this is effectively a case at the margins where the Commentary makes clear that it is within the power of Contracting States to determine the precise scope of residence. It does not indicate that full tax is enough or that anyone subject to full tax is resident: rather, full liability to tax follows from the relevant “personal attachment”. Like para. 4 this reflects the words used in Article 4(1).

In summary, I do not consider that the OECD Commentary supports GEFI’s case. I would also note that if GEFI were right then it is somewhat surprising that the straightforward proposition that anyone subject to worldwide taxation in a State is a resident of it was not spelled out.

One further point to make on the MTC relates to the corporate residence tie-breaker. If Article 4(1) of the Convention is considered from the perspective of the 2000 version of the MTC, then its use of place of effective management as the tie-breaker also rather reinforces HMRC’s interpretation. Essentially, it aims to determine the State with the closest real connection.

Crown Forest concerned whether the US/Canada double tax treaty could apply to limit the rate of Canadian withholding tax on barge rental payments made by the taxpayer to a Bahamian company, Norsk. Norsk had its sole place of business in the United States but was taxed there as a foreign corporation on the profits effectively connected to its US business, not including the rental payments. The Supreme Court of Canada rejected the argument that Norsk was a US “resident” for treaty purposes. While it had a “place of management” in the United States and that was one of the listed factors under Article 4 (as in the MTC and the Convention), that was merely one factor contributing to the determination that it carried on business there rather than being the basis of its tax liability. Further, its US business did not amount to a criterion “of a similar nature”, because all the other criteria were grounds for taxation of worldwide income rather than merely source income.

Mr Baker placed particular reliance on the following statement by Iacobucci J in considering that further point, giving the judgment of the Court:

As Mr Baker had to accept during oral submissions, the ratio of Crown Forest is that the equivalent in the US/Canada treaty of the first sentence of Article 4(1) of the MTC refers to liability to tax on a comprehensive or worldwide basis rather than taxation on a more limited basis. As far as the MTC is concerned that is uncontroversial, and indeed it explains why the introduction of the second sentence into Article 4(1) of the MTC (a sentence which in fact was not included in the treaty considered in Crown Forest) was not considered to alter the meaning of the first sentence. As Ms McCarthy submitted, worldwide taxation is necessary, but Crown Forest does not decide whether it is sufficient. That issue simply did not arise and the Court did not address it.

To the extent that any assistance can be gleaned from Crown Forest, it tends to assist HMRC’s case. Iacobucci J emphasised at [25] the need to show that liability to tax arose “by reason of” a listed ground, which “connotes the existence of some sort of causal connection or, in the least, some relationship of proximity”, whereas Norsk’s US place of management was not the basis of its liability. A similar point is made at [33] and [34]. In considering the parties’ intentions, Iacobucci J commented at [42] that he did not believe that the designers of the treaty would have envisaged that Norsk “ought to benefit from the preferential tax treatment accorded to residents”. At [57] an academic commentary is cited which refers to a person being considered resident in a country that “asserts an unlimited right to tax … that is, a right based upon the taxpayer’s personal connection with the country”. The conclusions at [68] record at (3) that:

Alta Energy was a majority decision of the Supreme Court of Canada that concerned a Luxembourg incorporated company that sought the benefit of the Canada/Luxembourg treaty to exempt gains on sales of property from Canadian tax, on the basis that its legal seat was in Luxembourg and that it was therefore resident there for treaty purposes. Article 4(1) of the treaty materially mirrored the MTC. In the Supreme Court the Minister of National Revenue accepted that Alta Energy was a resident of Luxembourg for the purposes of the treaty (see at [17], [60] and [92]) but, as mentioned at [‎42] above, argued that the GAAR could be invoked because the transaction abused the treaty, which should be read as requiring that residence in a Contracting State necessitated “sufficient substantive economic connections” to it. The Court held that the treaty could not be interpreted in that manner.

Again, this decision has no direct application to this case. It was a case about whether the GAAR applied to an entity that was admittedly treaty-resident. It does not determine the requirements for being treaty-resident in the first place. And as with Crown Forest, to the extent that it is relevant it tends to assist HMRC. For example, in discussing Article 4(1) Côté J emphasised at [53] that a person is resident if they are liable to tax by reason of “one of the connecting factors listed”, and stated at [54] that:

Further, at [67] Côté J commented as follows:

The majority concluded that the definition of residence in Luxembourg law (which granted residence to entities with their legal seat there) did not depart from accepted usage. Rather, it was consistent with international practice, which used a mixture of place of incorporation or legal seat on the one hand and “real seat” aimed at identifying the place of effective management on the other. If the parties had intended to depart from established criteria where an entity had minimal economic connections with a State and only a “formal, legal attachment”, then that would have been made clear ([60]-[62] and [67]).

Mr Baker relied on a discussion at [55] which refers to paragraph 4 of the OECD Commentary (see [‎91] above):

However, all the Court was saying here is that the definition of residence is left to domestic law. Indeed, the first sentence indicates that the Court recognised the two components of a) liability to tax, and b) residence.

Ms McCarthy referred to two other foreign authorities, namely a decision of the Indian Income Tax Appellate Tribunal in Deputy Commissioner of Income-Tax v General Electric Co. Plc. UK, IT Appeal Nos. 2361 to 2363 (Cal.) of 1996 and Cross Objection No. 62 (Cal.) of 1962 and a decision of the Portuguese Supreme Administrative Court in Case 0882/10 dated 12 January 2011. The reasoning in both is more directly in point than Crown Forest and Alta Energy, although the Indian decision is of a first instance tribunal and in that respect obviously carries less weight than a decision of a senior court. I do not need to consider these cases and do not rely on them. I simply note that the reasoning in each appears to be more consistent with HMRC’s case than with GEFI’s.

The parties relied on three academic commentaries, namely Klaus Vogel on Double Taxation Conventions, now in its 5^th edition, 2022 (“Vogel”), a chapter by Marcel Widrig entitled “The Expression ‘by reason of his domicile, residence, place of management…’ as applied to companies” in Guglielmo Maisto’s Residence of Companies under Tax Treaties and EC Law, published in 2009 (“Widrig”), and Corporate Residence and International Taxation by Robert Couzin, published in 2002 (“Couzin”).

I do not agree with GEFI’s argument that, properly understood, the academic commentaries support its position.

The leading, and most up to date, commentary is Vogel. In its discussion of the words “any other criterion of a similar nature” in Article 4(1) of the MTC at pp.287-290 (paras. 60-67) it describes references to domestic laws in the MTC (and UN equivalent) as therefore “not unqualified, (i.e., not all connecting factors giving rise to full liability to tax are sufficient)” and then says at p.288:

This rather speaks for itself, and is supported by a passage at p.289, para. 64, that states:

However, what Mr Baker relied on was passages that follow at paras. 65 and 67 which discuss nationality/citizenship and place of incorporation. Vogel takes the approach that these do not create territorial links because they are “purely legal”. He submitted that the inclusion of those concepts in the Convention therefore made Vogel’s territorial interpretation inappropriate, leaving the functional interpretation which reflected GEFI’s position.

With respect, this ignores Vogel’s rationale, which includes accepting that “not all connecting factors giving rise to full liability to tax are sufficient” and that the list of connecting factors should not be treated as “mere examples”. It also fails to take account of the express reference to deeming provisions being insufficient. Further and importantly, Vogel’s discussion relates to the MTC, where the enumerated factors are domicile, residence and place of management. Unlike the MTC the Convention expressly refers to citizenship and place of incorporation, making clear that such links are sufficient.

While Vogel presumably has in mind factual links such as a place of management, for what it is worth I have some difficulty in viewing place of incorporation as not being a link of a territorial nature. A company is a legal construct, and the fact that it derives its existence from the laws of a particular State seems to me to be a rather obvious link with that State. In contrast, at least in English law the country where an individual is domiciled might in some circumstances have a tenuous link with the individual. I also note that, from an English law perspective, a company’s place of incorporation can be regarded as its domicile: see for example Gasque v Comms. of Inland Revenue [1940] 2 KB 80 (a surtax case holding that a company’s domicile is its place of registration, meaning the jurisdiction in which it was formed), and in a more general private international law context Dicey, Morris & Collins on the Conflict of Laws, 16th ed, Rule 185, 30R-001, which states that the domicile of a corporation is the country under whose law it is incorporated. It is also worth noting in passing that the corporate tie-breaker in Article 4(3) of the 2017 version of the MTC, which has changed to a provision requiring mutual agreement, expressly refers to place of incorporation as one of the relevant factors that the Contracting States should take into account in their discussions, alongside place of effective management. The fact that Article 4(1) was not altered at the same time to include an express reference to place of incorporation seems to suggest that it was not considered necessary. The inclusion of place of incorporation in the revised Article 4(3) is also a clear indication that the place of incorporation is regarded internationally as an appropriate connecting factor.

However, that debate is of no material relevance to the application of the Convention because it expressly refers to place of incorporation in Article 4(1). The important point is that, applied to the Convention, Vogel’s analysis is consistent with the conclusion that HMRC’s interpretation is to be preferred.

Widrig takes a similar approach to Vogel. The author comments at 8.1 that the MTC “lists a number of connecting criteria” typically used to define residence. He explains at 8.2.4 that “other criteria of similar nature” must be “interpreted in the context of the other examples of connection criteria”. Any other criterion leading to unlimited tax has to be similar to one of those. He comments that the three listed criteria (domicile, residence and place of management) “have to a certain extent a local connection”, such that other criteria had to have “a certain local character of a factual nature”. Like Vogel, Widrig goes on to state that the law of incorporation of a company is not “similar” because it “lacks the effective personal attachment”, but the author notes that both the United States and Canada have placed reservations on Article 4 in that respect and on that basis place of incorporation could be considered as being of a similar nature (pp.281-282). The conclusion at 8.5 (with which I agree) states that:

Couzin is similarly supportive of HMRC’s position. Commenting on the MTC, the author states at section 3.1.2:

In discussing the listed connecting factors, the author observes at 3.1.3 that the ejusdem generis qualification “implies some limitation” on the application of national law, irrespective of whether Article 3(2) applies to the enumerated criteria. I would agree with that. Whatever legal system is used to interpret those individual concepts, the words “of a similar nature” can only be interpreted in accordance with the Vienna Convention. Couzin also recognises the possibility that domicile may cover place of incorporation at 3.1.3.1 and concludes at 3.1.3.2 that a domestically incorporated company should in any event be regarded as liable to tax by reason of its residence where the domestic law has that effect. In considering the words “any other criterion of a similar nature” Couzin states at 3.1.3.5:

Couzin also considers Crown Forest in some detail, and concludes at 3.1.4.1 that:

While I agree with Mr Baker that the final sentence may be read as a comment about terminology, recognising that concepts like “full tax liability” do not work well for States that impose tax on a territorial basis (see [‎60] above), I disagree that Couzin’s distinction between residence and source taxation in this section supports GEFI’s case. Rather, it is clear that Couzin does not agree that all that is required is a criterion that serves as a basis for taxation. Instead, the criterion must be based on some level of attachment between the taxpayer and the State.

The FTT recorded evidence from HMRC’s expert that the United States and Netherlands appeared to have agreed in a Memorandum of Understanding that a stapled entity should be treated as US tax resident for the purposes of the US/Netherlands double tax convention (see the final paragraph of [‎32] above and the FTT’s decision at [30]). However, as far as the Convention is concerned I consider that this is best seen as no more than the unilateral opinion of one of the parties, which is not a relevant aid to interpretation: [‎38] above. The fact that a third party may have agreed to that stance in the context of a different treaty, in circumstances that are unknown, makes no difference. At the most GEFI might argue that the published existence of the Memorandum of Understanding prior to entry into the Convention was part of the surrounding circumstances that could be taken into account under Article 32 of the Vienna Convention, but as a unilateral opinion I consider this doubtful. The UT placed little weight on it. I would not be inclined to place any weight on it.

GEFI’s alternative case is that the requirements of s.269B amount to a criterion “of a similar nature”.

I do not accept this. The US connections required by s.269B are limited to a) stapling of more than 50% by value of the foreign corporation’s shares to those of a domestic corporation, and b) direct or indirect ownership as to 50% or more by US persons. Both of these requirements relate to the ownership and control of the relevant company. Neither requires any form of link between the company itself and the United States, whether a formal legal one (such as incorporation, the location of its registered office or similar) or a factual one (such as place of management). The facts that the entity to which the company is stapled is itself US incorporated and that both entities are ultimately US owned cannot suffice. In contrast, the criteria specified in Article 4(1) all describe legal or factual connections between the entity itself and the relevant Contracting State of a kind that may justify worldwide taxation.

The UT referred at [148] to GEFI’s “substantive economic ties” to the United States, but any such ties it had are irrelevant to the application of s.269B and, as Crown Forest illustrates, economic or other links not falling within the criteria in Article 4(1) cannot provide a different route to treating an entity as US resident for Convention purposes.

In conclusion on Issue 1, HMRC’s interpretation is to be preferred based on the ordinary meaning of the words used in their context and in the light of the object and purpose of the Convention. In summary, the first sentence of Article 4(1) requires both the existence of a local connection falling within the enumerated criteria or which is of a similar nature to those criteria, and that that connection attracts worldwide taxation. Further, GEFI’s status as a stapled entity under s.269B does not amount to a criterion of a similar nature.

Accordingly, the FTT reached the correct conclusion on Issue 1. As to its somewhat compressed reasoning, there is some validity in Mr Baker’s criticism of the FTT’s reliance on the second sentence of Article 4(1), albeit that the explicit reference to profits of a permanent establishment in the version in the Convention weakens that criticism: see above. Mr Baker also fairly criticised the FTT’s reliance on its findings that US tax differences exist between domestic and stapled corporations in concluding that there was no relevant connection with the United States. The conclusion is correct but the existence of differences is not relevant to it.

It also follows that the UT made an error of law in reaching the conclusion that GEFI was US resident for Convention purposes. I will not comment further on its detailed reasoning, other than to note that using the MTC and OECD Commentary as its starting point, rather than the words of the Convention itself in the light of its object and purpose, may have increased the risk of error. I should also observe, however, that although it makes no difference in light of the decision of this court, I have some concern that the UT exceeded its jurisdiction in purporting to take a different view of some aspects of the expert evidence to that of the FTT (see [‎54] above). The UT (and this court) is fixed with the lower tribunal’s findings on the evidence unless its decision is set aside due to an error of law. It is only at that point that the UT or this court has power to make its own findings on the evidence: ss.12 and 14 Tribunals, Courts and Enforcement Act 2007.

On this issue, I agree with the UT that the FTT made no material error of law. Ultimately the question was an evaluative one for the FTT to determine, and it was entitled to conclude as it did. Indeed, based on my much more limited understanding of the evidence I see nothing surprising about its conclusion that GEFI was not carrying on business.

It is common ground that Article 3(2) requires the concept of “carrying on business” to be applied to GEFI by adopting the meaning of that concept for the purposes of UK taxes to which the Convention applies: see [‎27] above. It is also common ground that “business” is wider than “trade” for UK tax purposes.

However, it is worth noting two points. First, it is not obvious to me that UK tax law principles should be applied without any reference to the broader principles that apply in interpreting the Convention, and in particular the need to consider the meaning of words in their context. The context here includes a clear distinction between interest attributable to a business carried on in a Contracting State and interest derived in other circumstances: essentially the difference between active and passive income discussed in Alta Energy (see [‎77] above). However, I will not develop this point further given the common ground between the parties in this case.

Secondly, there is relatively limited modern case law authority on the meaning of business in UK tax law because the concept is not generally of central relevance to the imposition of tax on income. For example, non-UK resident companies conducting business in the United Kingdom are subject to corporation tax if they carry on a trade (rather than a business) here (s.19 Corporation Tax Act 2009), and a company falls within the charge to tax as soon as a chargeable source of income is acquired: s.1167 Corporation Tax Act 2010. In contrast there are numerous cases that consider whether a particular activity amounts to a trade.

The leading modern case on the meaning of “business” for tax purposes is American Leaf Blending Co v Director-General of Inland Revenue [1979] AC 676 (PC) (“American Leaf”), an appeal from the Federal Court of Malaysia. GEFI also placed reliance on the much earlier Court of Appeal decisions in Inland Revenue Comrs v Korean Syndicate Ltd [1921] 3 KB 258 (“Korean Syndicate”) and Inland Revenue Comrs v Westleigh Estates Company Limited [1924] 1 KB 390 (“Westleigh”), and on the House of Lords decision in South Behar Railway Company Ltd v Inland Revenue Comrs [1925] AC 476 (“South Behar”).

American Leaf concerned a Malaysian entity that abandoned a loss-making tobacco business and made five successive lettings of various parts of its premises, including a warehouse and factory. It sought to set its prior losses against the rental income, which depended on the source of the rents being a business. Lord Diplock gave the judgment of the court. He rejected the view of the special commissioners that the fact that letting property was included in the company’s objects was conclusive in favour of there being a business, noting that a comment of Pollock MR in Westleigh that apparently supported that proposition was made in the context of a company carrying out one of its principal objects, and adding (at p.683):

Lord Diplock went on to distinguish private individuals where there was “no presumption” that the receipt of rents from property amounts to carrying on business, and a company incorporated for the purpose of making profits for its shareholders, where “any gainful use to which it puts any of its assets prima facie amounts to the carrying on of a business”, adding that it was not easy to envisage circumstances that would displace that prima facie inference where a company’s property is let out for rent (p.684). He then said:

While the question was one of fact, the Privy Council concluded that the only reasonable conclusion was that the company carried on business, nothing in the evidence rebutting the prima facie inference.

Mr Brinsmead-Stockham submitted that the FTT had made fundamental errors of law. First, it failed to apply the “strong presumption” or inference that a company in GEFI’s position was carrying on business (the epithet “strong” is derived from Westleigh at p.415 and was not disputed by HMRC). Instead, he submitted that the FTT simply applied a list of factors derived from the VAT case of C&E Comrs v Lord Fisher [1981] STC 238, which were applied in the Upper Tribunal decision in Ramsay, without recognising that those cases related to individuals (albeit that Mr Brinsmead-Stockham recognised that the factors are relevant). Secondly, the FTT reached its conclusion based on a lack of activity without recognising that no significant (or any) activity was required. Thirdly, it did not engage with three of the four higher court decisions on which GEFI relied. More specific errors were a misunderstanding of Lord Sumner’s reference to “domestic operations” in South Behar, an incorrect application of the US tax law test of business and a misunderstanding at [90] about which loan agreements were covered by the Memorandum of Acknowledgement.

I do not consider that any of these alleged errors involves a material error of law, and this is also not a case that gets anywhere near succeeding in an Edwards v Bairstow challenge on the basis that no reasonable tribunal could have reached the conclusion that it did on the evidence.

Arnold LJ rightly reminded Mr Brinsmead-Stockham of what Lady Hale said in AH (Sudan) v Secretary of State for the Home Department [2007] UKHL 49, [2008] 1 AC 678 at [30] (in relation to the Asylum and Immigration Tribunal):

The FTT received detailed submissions from GEFI about the correct legal test to apply, by reference to American Leaf and the other cases on which GEFI relies. A tribunal does not make an error of law by not naming or citing from particular cases, particularly when it is clear that it has focused on more recent authorities that consider them. Further, the FTT’s decision expressly refers to the “strong” inference that a company is carrying on business when it puts its property to gainful use in citing another case at [75], and other citations at [76] and [78] capture the same point. It cannot be assumed that when it reached its conclusions the FTT somehow forgot a point that it had repeatedly made.

More fundamentally, a strong presumption or inference that a company is carrying on business does not mean that it necessarily is. Ultimately, GEFI’s position is that because the LP held some very large loans GEFI must therefore be taken to have been carrying on business. That does not follow. A multi-factorial evaluation is required: it all depends on the particular facts. For example, in American Leaf Lord Diplock laid stress not only on the lettings themselves but on related activities such as negotiations with different tenants and the removal of machinery. As these additional activities rather illustrate, there may well be a qualitative difference between what is involved in letting property to tenants and simply receiving an income stream from loans. In South Behar the company in question had at its risk funded the construction of a railway by agreement with the Secretary of State, and the income in question arose from a later agreement under which the Secretary of State took over the operation of the railway in exchange for an annuity rather than the previously agreed share of earnings. The decision was that the company still carried on its railway-related business. That is important. It was not a decision that it had a new business of earning annuities. In Korean Syndicate the company was formed to acquire, work and turn to account mining concessions. It acquired a large concession in Korea but, after considering working it itself, ultimately contracted with another entity to exploit it in return for a percentage of profits. Lord Sterndale MR expressly distinguished the company’s objects from the case of “a person who holds certain investments and merely draws the interest from them, or of an owner of mines who simply leases them in consideration of the payment to him of royalties” (p.272).

Expanding on that last point, GEFI placed reliance on the focus in American Leaf and the earlier cases on whether the activity was pursuant to the objects of the company. Mr Brinsmead-Stockham suggested that Lord Diplock’s comment on what Pollock MR said in Westleigh (see [‎132] above) amounted to saying that a company undertaking its principal object must be carrying on business. I would not read Lord Diplock’s judgment as intending to establish any such rule, but in any event it is of limited relevance here. The issue for decision is whether GEFI carried on business through a US permanent establishment. Its only US activity was its interest in the LP, in which as the limited partner GEFI had no active participation. The appropriate focus is therefore on the LP, and whether it was carrying on business through the activities of its general partner (see [‎30] above). The main purpose for which the LP was formed was to “hold directly or indirectly financial receivables and other assets…” (see [‎9] above), and there was no indication of any legal requirement for it to carry on business ([‎48] above). The LP’s constitution was equally consistent with acting as a pure passive holding vehicle and with something that amounts to a business.

As to whether the FTT applied a US tax law test, it explicitly recognised at [73] that it was required to consider the meaning of business as understood in UK law, and proceeded to refer at [73] to [80] (the remainder of its section under the heading of “law” in relation to this issue) only to UK case law. The FTT’s reference at [84] to “passive, sporadic or isolated activity” does reflect a comment by GEFI’s expert Mr Shashy to that effect, but that represented Mr Shashy’s own characterisation of the LP’s activities rather than a reiteration of the test under US tax law. The FTT evidently agreed with the description. The other alleged errors, relating to South Behar and the Memorandum of Acknowledgement, are clearly not material. As to the former, what the FTT was saying at [88] (see [‎51] above) was that the initial formation of the LP and capital contributions were the kind of “domestic operations” referred to by Lord Sumner in South Behar at p.487, where he referred to the domestic operations of declaring and paying dividends, remunerating directors and presenting reports. The FTT was however entitled to conclude that subsequent payment instructions did not convert the LP’s activities into a business, in contrast to South Behar where the receipt of annuities was found to be sufficient to continue an existing business.

Turning to the FTT’s overall evaluation, the critical factual conclusions are at [84], which bears repeating:

The FTT then expanded on this with some examples from the documentary evidence and more factual detail about what the LP actually did, before concluding at [91] that GEFI was not carrying on business through the LP, referring again there to the lack of strategic direction by GEFI Inc’s directors.

This conclusion does not involve any material error of law. In essence, the LP acted merely as a passive holding vehicle for some loan receivables. Their size makes no difference: the test is a qualitative one. The board of GEFI Inc, as the LP’s general partner, did not make strategic decisions and in fact had “very little involvement” ([90]). While the involvement of GE’s tax department was neither surprising nor a reason to treat the LP’s activities as somehow being “denatured” from being a business as a result of any tax motivation (which would be an incorrect approach), the FTT made no such error. It was fully entitled to focus on the lack of activity by the general partner in reaching its evaluative conclusion.

In conclusion I would allow HMRC’s appeal on Issue 1 and dismiss GEFI’s appeal on Issue 2, with the result that the UT’s decision is set aside on Issue 1 and upheld on Issue 2. The effect is to reinstate the FTT’s conclusion that the United Kingdom is not required to confer double tax relief in respect of US tax paid on GEFI’s interest income for the periods in dispute.