Julian Feiner (Clifford Chance) discusses the taxpayers' resounding victory at the Upper Tribunal in the company residency case of Development Securities plc and others v HMRC.
In the summer of 2004, Development Securities plc undertook a tax avoidance scheme. It set up three subsidiaries in Jersey to acquire UK property assets at an overvalue and, within 40 days, moved their residence to the UK, to achieve an increased capital loss on disposal. HMRC sought to challenge the scheme under Ramsay principles, but abandoned that argument in 2014. It claimed instead that the Jersey companies were resident in the UK, so that the scheme failed.
In the First-tier Tribunal ([2017] UKFTT 565), HMRC prevailed on muddy grounds. It seemed to justify their view of tax schemes, in Statement of Practice 1/1990, that where the tax benefit depends on a subsidiary’s appearance of non-residence without the reality, the subsidiary should be tax resident in the UK on the basis that it ‘rubber stamps’ the decision of its UK parent or advisor. HMRC had pursued this line of argument since the 1990s and had done so more forcibly since its defeat in Wood v Holden [2006] EWCA Civ 26, when the Court of Appeal ignored HMRC’s allusion to the ‘real facts’ of a capital gains tax scheme, and accepted that the local board made its own decisions.
The Upper Tribunal ([2019] UKUT 169), in 15 concise paragraphs, has delivered a stunning reality check. The FTT’s decision rested on a ‘fundamental misunderstanding’ of the transactions and the duties of the Jersey directors in relation to those transactions. Rubber stamping is always a question of fact, having regard to relevant indicators, and here the facts were that the Jersey directors knew and understood the scheme, applied their minds to it and properly considered the decisions they made. Central management and control was exercised in Jersey, not the UK.
I set out below my views on the UT decision and its implications for businesses.
Setting the scene
The company residence test of central management and control derives from the 1906 decision of the House of Lords in De Beers Consolidated Mines v Howe [1906] AC 455. It proceeds upon the analogy of an individual. A company cannot eat or sleep, but it can keep house and do business. It resides where its real business is carried on, and it is carried on where central management and control actually abides. It is a pure question of fact, determined not according to the construction of regulations or company articles, but upon a scrutiny of the course of business.
Previous UK cases relating to tax schemes provide the most relevant jurisprudence.
Wood v Holden
The Court of Appeal’s decision in Wood v Holden in 2006 is, as noted by the UT, a decision of particular significance, because it also concerned the residence of special purpose vehicles owned by a parent company. In that case, the court distinguished the facts from cases where a parent ‘usurped’ or ‘by-passed’ the local board (as was the case in Unit Construction v Bullock (1959) 38 TC 712). This was instead a case where the local board did make formal resolutions. In such cases, the relevant distinction is between a UK person merely ‘proposing, advising and influencing’ the local board and one who ‘dictates the decisions which are to be taken’. Only if decisions are dictated will the company be resident in the UK. The court found no evidence that a UK person, such as a parent or advisor, dictated the decisions that the local board was to make, even though the advisor expected the decisions to be made. It was considered improbable that the bank acting as managing director of the local board would allow its actions to be dictated by a client’s advisor; and, moreover, there was ample reason to make the decisions due to their commercial benefit.
Laerstate BV v HMRC
The FTT’s decision in Laerstate BV v HMRC [2009] UKFTT 209, which was decided in HMRC’s favour and not appealed, discerned the following spectrum of relevant cases from Wood v Holden :
(a) at one extreme, the local board signs documents mindlessly;
(b) up the scale, the local board knows what it is signing but signs without considering or having the minimum amount of information to make the decisions;
(c) next, the local board follows an outsider’s wishes after considering those wishes, and having the minimum information but less than a reasonable director would require; and
(d) at the other extreme, the local board makes an informed decision.
On this scale, (a) and (b) would constitute ‘rubber stamping’ by the local board and mean that the residence of the local company is in the jurisdiction of its parent, but (c) and (d) would not.
This scale loosely informed the FTT’s discursive analysis in Development Securities . In effect, the FTT found that the facts fit within level (b), because the only thing the local board considered was the parent’s instruction that it was lawful. It was not a case within level (c), like Wood v Holden , where the board considered a proposal, took appropriate advice and decided that it was in the best interest of the company to enter into it.
Upper Tribunal’s reformulation
The UT steered clear of the scale in Laerstate and reformulated a higher threshold for rubber stamping that is closer to, and perhaps even higher than, the concept in Wood v Holden .
That principle, set out at paras 18 to 19, is that for special purpose vehicles the relevant distinction is between influence and control. Control would be exercised if the local board abdicated its responsibility and behaved as a rubber stamp. The UT does acknowledge the existence of the Laerstate spectrum in a footnote but observes that rubber stamping is always a question of fact and degree. In the UT’s analysis, it is not possible to do more than identify what may be the indicators of rubber stamping. Those indicators are:
1. A disregard or breach of directors’ duties, including acts that are contrary to the company’s best interests: This was not established on the facts. In the UT’s view, the Jersey SPVs were not economically disadvantaged by the transaction, because the acquisitions were funded by their UK parent. Even if they were uncommercial, they were in the best interests of shareholders and therefore in the best interests of the SPVs, there being no prejudice to either the employees or creditors of the SPVs. This tightens the ‘loophole’ that HMRC thought it identified in Wood v Holden , that it may be confined on its facts to a case where there is an economic benefit to the SPV, instead of a benefit to the wider group.
2. Knowingly taking decisions without having sufficient information properly to make that decision: This similarly was not evidenced. The UT focused on the first board meeting, which lasted for five hours and involved active consideration of the issues, taking advice on stamp duty and correcting an error in the terms of a call option. The UT considered it was unnecessary for the Jersey directors to take an independent view on the strengths or weaknesses of the UK tax scheme being undertaken by the parent.
The UT confirmed that local board meetings are a sensible starting point to analyse central management and control; however, events outside the board room may be relevant. That includes events predating incorporation of the companies. The UT corrected the FTT’s erroneous quotation of the De Beers test, which would have required a focus on what took place immediately before incorporation, but endorsed the general proposition that it is appropriate to consider pre-incorporation events, including the origin of a tax scheme. It is unclear to what extent this influenced the UT’s decision. It is also unclear why the UT seemed to ignore the second decision of the board to move its residency to the UK.
Prospects of appeal
Now the question is, will HMRC appeal? Given the history, it looks likely. The dispute has spanned 15 years and rests on an important point of principle that will influence other residency cases. The current situation is unsatisfactory because we have a general concept from the Court of Appeal, 13 years ago, which has been interpreted in conflicting ways by the tribunal. HMRC will want to avoid ingraining the UT’s decision and may consider that, because it was so emphatic, the risk of an even worse outcome in the higher courts should be low.
The point of principle is to define and apply a rubber stamping threshold for SPVs, particularly in the context of tax schemes. HMRC may claim the threshold was set too high by the UT and misapplied. For example, the test should not require a breach of duty, but be limited to whether the local board considered and had information to make its decisions, in line with the Laerstate spectrum. If the threshold is lowered, the case turns on a fine balancing of the facts. The taxpayers can point to experienced professional directors, a five-hour first board meeting and consideration of technical issues. HMRC would say it was all preordained, with directors who only considered the legality of their acts: one of the directors was found to behave as a rubber stamp, while the other professional directors made minimal effort.
To succeed on appeal, HMRC may have to persuade the court to lower the rubber stamping threshold and take a ‘realistic’ view of the facts of the tax avoidance scheme. Past form, particularly in Wood v Holden , where HMRC lost in the Court of Appeal and was denied leave to appeal to the Supreme Court on a similar fact pattern, suggests that an appeal would be a doomed venture. It would require a significant change of approach, which is unlikely, though not inconceivable. HMRC has a strong track record in tax avoidance cases, but the UT has provided a clear statement that the fact the SPV was incorporated for the purposes of a tax avoidance scheme is irrelevant to the question of residency and should be left out of account.
Implications for businesses
The decision provides some certainty for businesses with offshore SPVs, which would be affected by a lower and less certain residency threshold if the FTT decision was upheld.
Uncertainties also arise in the practical application of the test. For example:
There is a growing phenomenon, in the UK and worldwide, for insurers to underwrite risks of tax residency under specific risk and warranty and indemnity policies. The progress of the case will be watched very carefully to inform the type of information and diligence that an insurer expects to see before underwriting a risk. Based on the decision, there may be greater appetite to underwrite the residency of an SPV, with residual concern about the appeal prospects.
Julian Feiner (Clifford Chance) discusses the taxpayers' resounding victory at the Upper Tribunal in the company residency case of Development Securities plc and others v HMRC.
In the summer of 2004, Development Securities plc undertook a tax avoidance scheme. It set up three subsidiaries in Jersey to acquire UK property assets at an overvalue and, within 40 days, moved their residence to the UK, to achieve an increased capital loss on disposal. HMRC sought to challenge the scheme under Ramsay principles, but abandoned that argument in 2014. It claimed instead that the Jersey companies were resident in the UK, so that the scheme failed.
In the First-tier Tribunal ([2017] UKFTT 565), HMRC prevailed on muddy grounds. It seemed to justify their view of tax schemes, in Statement of Practice 1/1990, that where the tax benefit depends on a subsidiary’s appearance of non-residence without the reality, the subsidiary should be tax resident in the UK on the basis that it ‘rubber stamps’ the decision of its UK parent or advisor. HMRC had pursued this line of argument since the 1990s and had done so more forcibly since its defeat in Wood v Holden [2006] EWCA Civ 26, when the Court of Appeal ignored HMRC’s allusion to the ‘real facts’ of a capital gains tax scheme, and accepted that the local board made its own decisions.
The Upper Tribunal ([2019] UKUT 169), in 15 concise paragraphs, has delivered a stunning reality check. The FTT’s decision rested on a ‘fundamental misunderstanding’ of the transactions and the duties of the Jersey directors in relation to those transactions. Rubber stamping is always a question of fact, having regard to relevant indicators, and here the facts were that the Jersey directors knew and understood the scheme, applied their minds to it and properly considered the decisions they made. Central management and control was exercised in Jersey, not the UK.
I set out below my views on the UT decision and its implications for businesses.
Setting the scene
The company residence test of central management and control derives from the 1906 decision of the House of Lords in De Beers Consolidated Mines v Howe [1906] AC 455. It proceeds upon the analogy of an individual. A company cannot eat or sleep, but it can keep house and do business. It resides where its real business is carried on, and it is carried on where central management and control actually abides. It is a pure question of fact, determined not according to the construction of regulations or company articles, but upon a scrutiny of the course of business.
Previous UK cases relating to tax schemes provide the most relevant jurisprudence.
Wood v Holden
The Court of Appeal’s decision in Wood v Holden in 2006 is, as noted by the UT, a decision of particular significance, because it also concerned the residence of special purpose vehicles owned by a parent company. In that case, the court distinguished the facts from cases where a parent ‘usurped’ or ‘by-passed’ the local board (as was the case in Unit Construction v Bullock (1959) 38 TC 712). This was instead a case where the local board did make formal resolutions. In such cases, the relevant distinction is between a UK person merely ‘proposing, advising and influencing’ the local board and one who ‘dictates the decisions which are to be taken’. Only if decisions are dictated will the company be resident in the UK. The court found no evidence that a UK person, such as a parent or advisor, dictated the decisions that the local board was to make, even though the advisor expected the decisions to be made. It was considered improbable that the bank acting as managing director of the local board would allow its actions to be dictated by a client’s advisor; and, moreover, there was ample reason to make the decisions due to their commercial benefit.
Laerstate BV v HMRC
The FTT’s decision in Laerstate BV v HMRC [2009] UKFTT 209, which was decided in HMRC’s favour and not appealed, discerned the following spectrum of relevant cases from Wood v Holden :
(a) at one extreme, the local board signs documents mindlessly;
(b) up the scale, the local board knows what it is signing but signs without considering or having the minimum amount of information to make the decisions;
(c) next, the local board follows an outsider’s wishes after considering those wishes, and having the minimum information but less than a reasonable director would require; and
(d) at the other extreme, the local board makes an informed decision.
On this scale, (a) and (b) would constitute ‘rubber stamping’ by the local board and mean that the residence of the local company is in the jurisdiction of its parent, but (c) and (d) would not.
This scale loosely informed the FTT’s discursive analysis in Development Securities . In effect, the FTT found that the facts fit within level (b), because the only thing the local board considered was the parent’s instruction that it was lawful. It was not a case within level (c), like Wood v Holden , where the board considered a proposal, took appropriate advice and decided that it was in the best interest of the company to enter into it.
Upper Tribunal’s reformulation
The UT steered clear of the scale in Laerstate and reformulated a higher threshold for rubber stamping that is closer to, and perhaps even higher than, the concept in Wood v Holden .
That principle, set out at paras 18 to 19, is that for special purpose vehicles the relevant distinction is between influence and control. Control would be exercised if the local board abdicated its responsibility and behaved as a rubber stamp. The UT does acknowledge the existence of the Laerstate spectrum in a footnote but observes that rubber stamping is always a question of fact and degree. In the UT’s analysis, it is not possible to do more than identify what may be the indicators of rubber stamping. Those indicators are:
1. A disregard or breach of directors’ duties, including acts that are contrary to the company’s best interests: This was not established on the facts. In the UT’s view, the Jersey SPVs were not economically disadvantaged by the transaction, because the acquisitions were funded by their UK parent. Even if they were uncommercial, they were in the best interests of shareholders and therefore in the best interests of the SPVs, there being no prejudice to either the employees or creditors of the SPVs. This tightens the ‘loophole’ that HMRC thought it identified in Wood v Holden , that it may be confined on its facts to a case where there is an economic benefit to the SPV, instead of a benefit to the wider group.
2. Knowingly taking decisions without having sufficient information properly to make that decision: This similarly was not evidenced. The UT focused on the first board meeting, which lasted for five hours and involved active consideration of the issues, taking advice on stamp duty and correcting an error in the terms of a call option. The UT considered it was unnecessary for the Jersey directors to take an independent view on the strengths or weaknesses of the UK tax scheme being undertaken by the parent.
The UT confirmed that local board meetings are a sensible starting point to analyse central management and control; however, events outside the board room may be relevant. That includes events predating incorporation of the companies. The UT corrected the FTT’s erroneous quotation of the De Beers test, which would have required a focus on what took place immediately before incorporation, but endorsed the general proposition that it is appropriate to consider pre-incorporation events, including the origin of a tax scheme. It is unclear to what extent this influenced the UT’s decision. It is also unclear why the UT seemed to ignore the second decision of the board to move its residency to the UK.
Prospects of appeal
Now the question is, will HMRC appeal? Given the history, it looks likely. The dispute has spanned 15 years and rests on an important point of principle that will influence other residency cases. The current situation is unsatisfactory because we have a general concept from the Court of Appeal, 13 years ago, which has been interpreted in conflicting ways by the tribunal. HMRC will want to avoid ingraining the UT’s decision and may consider that, because it was so emphatic, the risk of an even worse outcome in the higher courts should be low.
The point of principle is to define and apply a rubber stamping threshold for SPVs, particularly in the context of tax schemes. HMRC may claim the threshold was set too high by the UT and misapplied. For example, the test should not require a breach of duty, but be limited to whether the local board considered and had information to make its decisions, in line with the Laerstate spectrum. If the threshold is lowered, the case turns on a fine balancing of the facts. The taxpayers can point to experienced professional directors, a five-hour first board meeting and consideration of technical issues. HMRC would say it was all preordained, with directors who only considered the legality of their acts: one of the directors was found to behave as a rubber stamp, while the other professional directors made minimal effort.
To succeed on appeal, HMRC may have to persuade the court to lower the rubber stamping threshold and take a ‘realistic’ view of the facts of the tax avoidance scheme. Past form, particularly in Wood v Holden , where HMRC lost in the Court of Appeal and was denied leave to appeal to the Supreme Court on a similar fact pattern, suggests that an appeal would be a doomed venture. It would require a significant change of approach, which is unlikely, though not inconceivable. HMRC has a strong track record in tax avoidance cases, but the UT has provided a clear statement that the fact the SPV was incorporated for the purposes of a tax avoidance scheme is irrelevant to the question of residency and should be left out of account.
Implications for businesses
The decision provides some certainty for businesses with offshore SPVs, which would be affected by a lower and less certain residency threshold if the FTT decision was upheld.
Uncertainties also arise in the practical application of the test. For example:
There is a growing phenomenon, in the UK and worldwide, for insurers to underwrite risks of tax residency under specific risk and warranty and indemnity policies. The progress of the case will be watched very carefully to inform the type of information and diligence that an insurer expects to see before underwriting a risk. Based on the decision, there may be greater appetite to underwrite the residency of an SPV, with residual concern about the appeal prospects.
