The Tax Residence of Special Purpose Companies
Many financial structures involve the use of a special purpose company (SPC) – a company set up specifically to play a role in the structure. Normally the SPC has a limited degree of activity (for example, entering into transaction at the outset and continuing to hold assets and have liabilities), but rarely undertaking other transactions while the structure is in place. Very often, SPCs are established outside the UK in order to prevent UK tax being chargeable on profits which arise, or which might be deemed to arise, in relation to the transactions undertaken by the SPC. In such a case, it is critical that the overseas SPC is not resident in the UK for tax purposes. Although there is long-standing case law on the question of tax residence of companies, it is rare for cases to come before the courts. That case law was recently applied by the High Court and the Court of Appeal when HM Revenue & Customs (HMRC) challenged the tax residence of an SPC used in a structure designed to mitigate a potential UK tax charge. The HMRC challenge was based on the assertion that a Dutch SPC was being run onshore in the UK and was therefore tax resident in the UK. The decision in favour of the taxpayer was reached notwithstanding that the inherent nature of the SPC and the limited activities undertaken by it meant that its board of directors was necessarily carrying out few of the functions which would normally be associated with the management of a company.
The tax residence of an SPC in a low tax jurisdiction is often crucial in structuring a finance transaction. If the SPC were, in fact, resident in a jurisdiction where it would have significant tax liabilities, the resulting tax leakage would often mean that the transaction would be uneconomic. The difficulty with establishing the tax residence of an SPC is that, by its very nature, its participation in a transaction is limited. For example, it would usually be involved at the outset of a transaction, when it is asked to participate in a structure that has been determined by other parties. Once the structure has been established, the SPC could be described as doing no more than receiving and making payments (often through a security structure which means it has little discretion in the calculation or determination of the payment amounts) and playing no active role until the structure is collapsed or runs its course. Again, it is unlikely to be the SPC that starts the unwinding process. This would more likely be initiated by, for instance, an arranger.
In Wood v Holden, a direct tax avoidance motive was clear. This is not always the case. SPCs are often used where, in economic terms, there is no mischief but rather a tax mismatch which could give rise to a timing cost rather than an absolute cost. Alternatively, SPCs can be used for VAT planning purposes, although this may be restricted by the European abuse of rights doctrine.
It should be noted that it is not yet known whether Wood v Holden will progress to the House of Lords; it is therefore possible that the House of Lords could reverse the decision of the Court of Appeal. However, the Court of Appeal emphatically upheld the decision of Park J, and it would be surprising for the House of Lords to overturn the decision.
There are two tests which must be looked at when determining whether a company is resident in the UK for tax purposes. The first (the ‘common law test’) is what was referred to by the High Court judge as the ‘common law of corporate residence’. As the judge noted, this expression is somewhat misleading as some of the concepts are based on statute. The second test (the ‘tiebreaker test’) is the test set out in s 249 Finance Act 1998. This test supplements and can override the common law test where there is a double taxation treaty in place between the UK and the country in which the SPC is also claiming to be resident. If there is no double taxation treaty with the country in which the SPC is claiming to be resident, the tiebreaker test is not relevant.
The first test to consider is the common law test: if the SPC is not resident in the UK under the common law test, it will never be resident in the UK. If the SPC is resident in the UK under the common law test, the tiebreaker test will need to be considered.The flowchart shows the process which must be followed in arriving at a conclusion.
As will be seen, the tiebreaker test was not determinative in Wood v Holden, but some guidance was given on its application by Park J in the High Court.
The basic rule of UK law corporate residence is that a company is resident in the UK if either it is incorporated in the UK or the ‘central management and control’ of the business of the company is carried on in the UK.
Where a non-UK resident SPC is required in a financing transaction, the SPC should not be incorporated in the UK because of the provision that any company incorporated in the UK is deemed automatically to be resident in the UK under the common law test1 (subject to very limited exceptions).
The second limb of the test is the part that was relevant in Wood v Holden, as the court was asked to decide whether the central management and control of the Dutch SPC was in the UK. The ‘central management and control’ test originates in cases dating from the late 19th century, but the main authority for the test is the decision of the House of Lords in De Beers v Howe2. In this decision, the Lord Chancellor (Lord Loreburn) said that the test ought to approximate as nearly as possible to the test used for determining the residence of an individual. He said that:
“[a] company cannot eat or sleep, but it can keep house and do business. We ought, therefore, to see where it really keeps house and does business. An individual may be of foreign nationality, and yet reside in the United Kingdom. So may a Company… a Company resides, for purposes of Income Tax, where its real business is carried on … and the real business is carried on where the central management and control actually abides.”
This can only be determined on the facts of each case. In De Beers, the central management and control of the business was in London, where the main board of the company met on a consistent basis, despite the fact that the business of the company was carried out where its mines were located, in southern Africa. Since the De Beers judgment, there have been several cases on the point, both in the UK and in other Commonwealth countries that have the same domestic law (such as New Zealand and Australia). The courts have concluded that the central management and control of a company, in normal circumstances, is equivalent to the decisions made by the company at board level; in other words, the location where the final decisions that bind the company are considered and made.
One case where the court reached a different view was Unit Construction Co Ltd v Bullock3, which was relied upon by HMRC in Wood v Holden. Where the functions of the board of Unit Construction (which was registered in Kenya) had effectively been usurped by the board of directors of its parent company (which met in London), London was the place of central management and control. No decisions were actually being taken in Kenya at all. All decisions of the company were taken in London by the board of directors of the parent company. Notwithstanding that this may have been outside the constitution of Unit Construction, the House of Lords found that this was where the central management and control was being exercised.
The courts have also held (in a case known as Untelrab4) that, even though the board of a company in fact follows instructions given to it, it does not necessarily follow that the control and management lies with the person giving the instructions, so long as the board does exercise its discretion and would refuse to carry out an improper or unwise transaction.
The tiebreaker test is set out in s 249 Finance Act 1994. This is relevant where a company is resident in the UK under the common law test, but is regarded also as being resident in another jurisdiction under the domestic laws of that state and there is a double taxation treaty between the UK and that state which contains a tiebreaker clause. Examples would include two jurisdictions where SPCs are commonly used (for instance, on securitisation transactions) – Ireland and the Netherlands. It is not relevant where there is no full double taxation treaty in force with the offshore jurisdiction – as will be the case with many tax havens, such as the Channel Islands and the Cayman Islands. If the test applies, the company is to be regarded as resident where its place of effective management is situated.
The concept of ‘effective management’ is not in origin a UK concept; it stems from the wording of the OECD Model Double Tax Treaty. The commentary to the OECD model treaty does not have much further detail as to the way in which the test should be applied, but certain countries have made their views known as to its application. For example, New Zealand has expressed its view that the place of effective management is situated where the practical day-to-day management of a company is situated, regardless of where the overriding control is exercised.
There is little UK case law on the meaning of ‘place of effective management’, although the Court of Appeal in Wood v Holden were very clear that the wording meant that it was necessary to show an actual physical place in the UK where the effective management was carried out in order to satisfy the test – it was not sufficient to show that the management was being carried out somewhere within the UK. The main source of information on how the ‘place of effective management’ test is treated by the UK authorities is the guidance published by HMRC5 stating its view of the test. HMRC states that, in its view, the ‘place of effective management’ of a company is a different test from its central management and control. The example given by HMRC is of a company which is run by executives based overseas, but with a final directing power which is exercised by non-executive directors who meet in the UK. In such circumstances, HMRC’s view is that the central management and control of the company will be in the UK but that the place of effective management of the company is overseas. Therefore, if the place of effective management is in a jurisdiction with a double tax treaty with the UK containing a tiebreak clause, s 249 FA 1994 would mean that the company was resident for all UK tax purposes in the other jurisdiction. In the case of a SPC, where the only decisions relating to the activity of the company are taken by the board, it is unlikely that there would be a difference between the results given by the common law test and the tiebreaker test.
It is little surprise that HMRC sought to challenge the planning used in Wood v Holden. It was not disputed that the structure was established with the specific purpose of avoiding a substantial charge to capital gains tax in the UK (on the sale of shares) for which the taxpayer (Mr Wood) would have been liable without the scheme. It is not necessary to understand the details of the way in which the scheme worked in order to examine the decision of the court. What is of importance is that, in order for the scheme to succeed, one of the companies in the structure (E BV) had to be resident in the Netherlands and not in the UK. E BV was a SPC which undertook only the specific transactions assigned to it; it was prevented by agreement from conducting any other business than that which it undertook.
In effect, HMRC had two arguments. The first was that the burden of proof was on Mr Wood to show that E BV was not resident in the UK, and that he had not succeeded in showing that this was not the case. The second argument was that the board of E BV did not properly consider the transactions put before them and were acting as a ‘rubber stamp’ for the decisions that had been taken by Mr Wood and PwC, Mr Wood’s advisers who were the architects of the scheme.
In response, Mr Wood argued that, although the burden of proof was on him to show that E BV was not resident in the UK, he had produced evidence that E BV was incorporated in the Netherlands and that the decisions of the board of E BV approving the steps described above were taken in the Netherlands. On the basis of this evidence, the common law test resulted in the conclusion that E BV was resident in the Netherlands. This meant, Mr Wood argued, that the burden of proof had now shifted to HMRC who had to produce additional evidence that the central management and control of E BV was in the UK.
Mr Wood also argued that the fact that the board of E BV were undertaking a ‘rubber stamping’ exercise did not mean that they did not have the ultimate say in whether E BV should take part in the transactions. In fact, on the evidence, it was clear that the board of E BV was being asked to take the decisions in each case.
At first instance, the Special Commissioners decided that, although the board of E BV had considered the transactions, they had not given sufficient thought to the precise terms of the transaction to indicate that they had properly approved the transaction. On this basis, the Special Commissioners held that they had not seen sufficient evidence from Mr Wood to discharge his burden of proof of showing that E BV was not resident in the UK.
The decision of the Court of Appeal was so much in line with the principles set out in the High Court judgment of Park J that references here to the Court of Appeal decision are essentially also references to the High Court decision. The Court of Appeal decided that the arguments made on behalf of Mr Wood both succeeded.
First, the court held that Mr Wood had shown enough evidence to show that E BV should be regarded as resident in the Netherlands on the basis that it was incorporated there and the board had taken all the major decisions on behalf of the company there. The court held that this was sufficient to discharge Mr Wood’s burden of proof in rebutting the presumption created by the HMRC’s capital gains tax assessment that E BV was not resident in the UK.
Second, the court held that HMRC had not shown sufficient additional evidence to prove that E BV was resident in the UK. The court held that the fact that the board of E BV had not paid much attention to the price at which it had acquired the relevant shares and had not considered in detail the terms of the sale documentation did not mean that the board was not taking the decisions on behalf of E BV in the Netherlands.The fact that the board simply followed the advice of PwC was again considered not to be relevant, on the basis that this was advice (and was expressed as such). There was nothing which forced the board to follow the advice. In theory, even if the decisions taken by the board could have been challenged by the shareholders of E BV as being in breach of the duties of the board, this did not affect the fact that it was the board that was taking the decisions in the Netherlands. The court was swayed to a certain extent by the fact that the board, being a trust company which formed part of the ABN AMRO group, would not ‘rubber stamp’ a proposal which was illegal or which would be likely to prejudice the reputation of its parent company. This followed the line taken by the Special Commissioners in Untelrab. It did not matter that E BV took very few decisions and that the few decisions taken by the company were in accordance with the scheme proposed by PwC. It was the board that was taking the ultimate decisions on behalf of E BV and it was taking them in the Netherlands.
As mentioned above, the court did not feel the need to deal explicitly with the question of whether the place of effective management of E BV was in the Netherlands. However, the court observed that, the ‘place of effective management’ of a company had to be a specific place and not just within a country. In this, HMRC were not helped by the decision of the Special Commissioners who held that ‘the Appellants have failed to satisfy us that the central control and management of E BV was not in London’, despite the fact that Mr Wood himself lived in Lancashire, and the relevant offices of PwC were situated in Manchester and Birmingham.
This contrasts with the position in Unit Construction, where the facts were that the board of Unit Construction had stepped aside and allowed the board of its parent company to take over control of the company. Here, because the board could (and, in the view of the courts, would) have refused to commit E BV to do anything that was illegal or which was clearly inappropriate for E BV to undertake, it had not ceded control to Mr Wood or to PwC.
The statements made by the court as to the burden of proof show that, as a minimum, any structure that involves a SPC should ensure that there is sufficient evidence available to rebut any initial presumption that the decisions involving that SPC are being taken in a different jurisdiction. Therefore, evidence should be retained to show that proposals were put to the board of the SPC in the relevant jurisdiction, and that those proposals were properly considered and accepted.
This burden of proof was reasonably easy to satisfy in Wood v Holden, where E BV only took two major decisions, very close together. However, if an SPC is being used in an ongoing structure (for example, a conduit financing), it will be even more important to ensure that the same level of evidence of consideration by the board is available for each transaction undertaken by the SPC to reduce the risk that it can be said that a participant in another jurisdiction is in fact directly or indirectly controlling the SPC. It will also be helpful if the board is aware of its duties and responsibilities: one of the factors that helped the court was that the board of E BV were employees of a highly reputable trust company and knew that they should not take decisions without due consideration.
Having set up the SPC, it is important that the directors are not ‘mindless’ in approving transactions. It was clear from the High Court judgment that the advisers of a SPC were perfectly entitled to advise the SPC to carry out a particular transaction, but it was also made clear that the line should not be crossed into directing the SPC to enter into the transaction. It was helpful in Wood v Holden that PwC had a separate engagement letter with E BV under which it was engaged to provide advice in connection with the sale of the shares in it owned as part of the scheme. It was also helpful that PwC proffered their advice as advice on the face of it and requested that the board of E BV consider and, if appropriate, sign the relevant documents.
In Wood v Holden, it was also helpful that Mr Wood himself showed little interest in the tax structuring. From the evidence, it appeared that he did not have sufficient knowledge of how the structure worked to be regarded as instructing E BV to do his bidding.
However, where financing structures are set up for UK headquartered sophisticated groups who have their own tax managers who understand (and possibly devise) the structure, care will need to be taken to ensure that HMRC would not have a stronger argument that the UK resident participant in the transaction has usurped the functions of the SPC’s board. In these circumstances, any UK participant must exercise a high degree of self-restraint. It must allow the board of the SPC time to consider the risks and rewards of the transaction. In practice, it will be helpful if it can be shown by contemporaneous evidence (by way, for example, of a board paper) that any particular risks to which the SPC will be exposed (such as credit or currency exposure) have been considered and, if appropriate, are the subject of indemnification and/or mitigation.
Until it is known whether the House of Lords will consider this case, any conclusions must be tentative. However, the case does confirm that, provided the structuring is done properly, it is possible for an SPC to be set up as non-UK tax resident company, even if it conducts little or no activity and is taking part in a structure that was devised in the UK for the ultimate benefit of participants in the UK.The important proviso is that the directors of the SPC must themselves decide that the SPC will participate and on what terms to do so.
1 Section 66 Finance Act 1988
2 De Beers Consolidated Mines Ltd v Howe (Surveyor of Taxes) [1906] AC 455
3 Unit Construction Co Ltd v Bullock (Inspector of Taxes) [1960] AC 351
4 Untelrab Ltd v MacGregor (Inspector of Taxes) [1996] STC
5 HMRC Statement of Practice 1/90, paragraph 22