A landmark new decision on how the residency of a trust is determined

By Nicolas F. Baass  LL.B., LL.M. (Tax) and Kim G C Moody  CA, TEP

This blog deals with a landmark new Tax Court decision released last week.  Given its importance, we spill a lot more ink than our usual blog entries.  Accordingly, be warned that this entry is lengthy.

It could reasonably be expected that a topic of such importance as the residence of trusts would be explicitly provided for in the Income Tax Act (“ITA”).  After all, the ITA purports to be a complete code on income taxes and as such, this glaring oversight is rather surprising.1  Consequently, Canadian taxpayers have been left largely without guidance as to how one may determine trust residence.  For many years taxpayers looked to the Federal Court’s decision in Thibodeau Family Trust v. The Queen2 for guidance as to the residence of trusts.

In Thibodeau the Court was called on to determine the residence of a trust.  The Minister was seeking to tax gains generated on the disposition of shares held by a trust which purported to be a Bermuda resident trust.  The Court noted the absence of a statutory formula to determine trust residence and opted to establish a judicial formula “applicable to the facts of this case alone”.3  In its analysis the Court states that the corporate test of mind and management for determining corporate residence is not applicable to trusts as trustees cannot delegate their authority and cannot adopt a policy of inactivity with respect to their fiduciary duties.4  The Court then goes on to analyze certain criteria that may be applicable in formulating a judicial formula of a trust’s residence. As we outline below, the reliance on Thibodeau is now in question.

24     The facts of this case for selecting as the criteria to propound such a formula are: (1) that the majority of the trustees of the Trust are and were at all material times resident in Bermuda; (2) that the trust document of constitution permitted a majority decision on all matters of trustees’ discretion; (3) that the Department of National Revenue by the assessment discounting the value of the said shares in Thibodeau Express Limited held by the Trust by 10% as being a minority interest have admitted that Leo J Thibodeau, the Canadian trustee of the Trust and the majority shareholder of the shares of Thibodeau Express Limited at the material time did not control the shares held by the Trust and was not, as pleaded, “the guiding mind and will of the Thibodeau Family Trust” at the material time; if it were otherwise, all the shares in Thibodeau Express Limited would be controlled by Leo J Thibodeau and there would be no basis to discount as was done by the assessment, the value of the shares held by the Trust; (4) that the res of the Trust in the main was in Bermuda at all material times; (5) that some of the beneficiaries resided the whole or part of the year, 1972, in the United States; (Canada was their residence during the other times); and (6) that the Thibodeau Family Trust did not in 1972 or at any other relevant time carry on business in Canada.

25     Selecting these two of the said criteria so as to found such a judicial formula to decide the said issue of residence in this case, namely (1) that the majority of the trustees were resident in Bermuda in 1972 at all material times, and (2) that the trust document permitted a majority decision on all matters of Trustees’ discretion, the finding of fact is that the Thibodeau Family Trust in the year 1972 had a sole residence in Bermuda for the purpose of Canadian income tax. [Emphasis added]

From the above paragraph 24 of the Thibodeau decision, it appears that the judge considered the following criteria in determining the residence of a trust:

(1) The residence of the majority of the trustees;
(2) The fact that the trust documents permitted a majority decision on all matters of trustees’ discretion;
(3) Who is the “guiding mind and will” of the trust;
(4) The location of the property of the trust;
(5) The residence of the beneficiaries; and
(6) Where the trust was carrying on business.

In the subsequent paragraph the Court chooses to select the residence of the trustees and the fact that the trust documents permit majority decisions on all matters of Trustees’ discretion.  Based on this, the Court concluded that the trust was a Bermuda trust.  While the ultimate test applied by the Court was that of the Trustee’s residence, the Court did mention that the criteria of “guiding mind and will” should be considered.  The second aspect of the test enunciated by the Court, the fact that the trust documents permit majority decisions, appears to have been ignored by subsequent commentators and court decisions.

The Thibodeau decision has for many years been interpreted by tax practitioners as standing for the proposition that the residence of a trust is determined by the residence of the trustees.  The CRA seemed to partially confirm this in Interpretation Bulletin IT-4475 in which the residence of a trust is to be determined as follows:

Normally residence of a trust is dependent upon residence of the trustee or trustees who can exercise management and control of the trust. In some situations the facts may indicate that a substantial portion of the management and control rests with some other person such as the settlor or the beneficiaries. In these situations the residence of this other person may be considered to be the determining factor for the trust regardless of any contrary provisions in the trust agreement. [Emphasis added]

Strangely enough, the issue of the residence of a trust did not arise again until more than 30 years later in 2009 with the new Garron Family Trust v. The Queen6 decision released on September 10, 2009.

The Garron decision introduces what appears to be a unique approach to determining the residence of a trust.  The facts of this case are relatively straightforward and we shall not dwell on them at length.  In essence, the taxpayer executed a transaction commonly known as a “Barbados freeze”.  Shares of two Canadian corporations were frozen in favour of two Barbados Trusts (the “Trusts”).  Following the freeze transaction the Barbados Trusts disposed of the shares, generating capital gains of over $450,000,000.  The trusts claimed the exemptions under Article XIV(4) of the Canada – Barbados Tax Treaty7 which provides that gains from the alienation of property is only taxable in the state of which the alienator is a resident.  Obviously the Trusts claimed to be residents of Barbados where the gains would be non-taxable under Barbados law.  The Minister took the position that this Article of the Treaty did not apply as the Trusts were resident in Canada.  Issues with respect to the application of section 94, subsection 75(2) and GAAR were also raised, but will not be discussed as the focus of this blog is the residence of trusts.

Justice Woods of the Tax Court starts her analysis as to the appropriate test for trust residence by stating that Thibodeau does not set out the test of residence solely on the residence of the trustee.  Justice Woods points out that the Court in Thibodeau had restricted the application of its judicial test to determining residence of a trust only to the facts of the particular case.  She also comments on the fact that the Court in Thibodeau explicitly found that the Canadian trustee did not control the trust.  Justice Woods finishes her analysis of Thibodeau by rejecting the Federal Court’s contention that the corporate formula respecting residence (i.e. management and control) cannot apply to trusts as trustees cannot delegate their authority.  Justice Woods points out that it cannot be assumed that trustees will never be in breach of their fiduciary obligations.  Based on the foregoing, the Court concludes that there is a lacuna as to the issue of trust residence.

While it is difficult to find fault in Justice Woods dismissal of the Thibodeau case in light of the foregoing arguments, it is difficult to understand why the Court did not simply attempt to analyze the same criteria as those examined by the Court in Thibodeau.  Instead of adopting such an approach, the Court decides to innovate and adopt the test of central management and control applicable to corporations.  The Court’s reasoning for adopting this test is dramatically short on analysis.  The Court only spends four short paragraphs (159 to 162) in rationalizing this decision, content simply in stating that the corporate residence should apply to trusts because:8

Although there are significant differences between the legal nature of a trust and corporation, from the point of view of determining tax residence, the characteristics are quite similar. The function of each is, at a basic level, the management of property.

Second, adopting a similar test of residence for trusts and corporations promotes the important principles of consistency, predictability and fairness in the application of tax law.

One could question what characteristics are similar between a trust and corporation from a point of view of determining tax residence?  Aside from the fact that the two function to manage property there is no elucidation as to these characteristics.  Furthermore, would consistency and predictability not be better promoted by applying the test which the tax community has accepted for the past 30 years, namely the residence of the trustees?  It must be said that Justice Woods does not supply very compelling arguments for the adoption of the management and control test.  What is equally surprising is that Justice Woods does not contemplate any other potential methods of determining trust residence.  At the very least, such factors as what law the trust was settled under, the location of the trustee’s offices and such other factors might have been considered and rejected before reaching a conclusion.

Justice Woods pursues her analysis by asking what management and control means and by applying her findings to the case.  The essential determination of facts by the Court was that the beneficiaries managed the affairs of the trusts and that the trustees were only marginally implicated in the affairs of the trust.  Based on this, Justice Woods finds that the central management and control of the trusts rests with the Canadian resident beneficiaries and as such that the trusts were resident of Canada.  Strangely enough, this approach is hard to reconcile with Justice Woods’ following statement:

To my knowledge, the management and control test has to date only been applied in a corporate context. A review of judicial decisions in this area suggests that management and control has usually been found to reside in a board of directors, even though the directors may be under significant influence from shareholders or others.

By analogy one may be tempted to think that management and control of a trust could be found to reside with the trustees, even though the trustees may be under significant influence from beneficiaries or others.  Indeed, one would think that even though the trustees were under the influence of the beneficiaries, the management and control of the trust would ultimately still remain with the trustees.  Perhaps Justice Woods’ reasoning would have been more convincing if she had invoked the concept of trustee de son tort or de facto trustee, thereby imputing trustee status to the individual, or individuals, who truly manage and control the trusts.

Given the amounts at stake in the Garron case it is hoped the taxpayer will appeal the Court’s findings.  This would be welcome as the question of trust residence has long been in a state of limbo.  A decision from a higher court would greatly help in crystallizing a definite rule on the residence of trusts.

Given the above, the following is an incomplete list of comments and/or questions that arise from Garron:

  1. Justice Woods seems to use the fact that trustees will not always adhere to their fiduciary obligations as the key point to enable her to utilize a “central management and control” test.  This is puzzling.  Trustees have a fiduciary obligation to manage the trust assets on behalf of the beneficiaries.  If the trustees fail in their duties, the beneficiaries can seek legal remedies against the trustee.  Accordingly, given that trustees must manage trust assets, the fact that trustees may fail in their fiduciary obligations does not seem to be a complete analysis so as to enable a central management and control test to determine trust residency.
  2. Justice Woods seems to have been troubled by the fact that the Barbados trustees took direction form the beneficiary(ies) in the management of the trust assets.  That fact alone, however, should, arguably, be fine given the fact the trustees must manage trust assets on behalf of the beneficiaries or what is in the best interest of the beneficiaries.
  3. If Garron is the correct new standard for determining trust residency, what kind of activity will constitute central management and control being in Canada?
  4. Would the result of Garron have been different if there were two Barbados trustees, Mr. Garron as a trustee and the trust deed required that two out of three trustees consent to make decisions?  Likely not, but an interesting question.
  5. Over the years, many planning arrangements have been implemented to arguably change or establish the residency of a trust in a target area such as Alberta or offshore.  Many of these plans will need to be revisited in light of Garron to ensure that the planning is still sound.
  6. On its face, it is difficult to reconcile this tax decision with trust law.  Trustees must manage trust property as part of their duties.  Accordingly, does not management reside with the trustees?  If not, what is the legal relationship then?  Is the relationship argued to be an agency relationship, a bare trustee relationship, a sham? This does not appear to have been considered by Justice Woods.
  7. While not discussed above, Justice Woods appears to have been troubled by the fact that the Barbados trustee was owned by an accounting firm and she did not see, based on the evidence, any particular expertise of the Barbados trustee in trust law/matters.  We query why this is even relevant.  If relevant, will this now require trustees to be well versed or experts in trust law?
  8. Again, while not discussed above, Justice Woods appears to have been troubled that the owner of the Barbados trustee – the accounting firm – was also involved in the tax planning for Mr. Garron.  We again query why this is relevant in the determination of tax residency.

Stay tuned… there is sure to be lots of follow-up on this important matter.

1. For example, in the United States the Internal Revenue Code (“IRC”) explicitly sets out the criteria for the residence of trusts at sections 7701(a)(30), (31) and Federal Regulation 301.7701-7.2. Dill v. R. [1978] C.T.C. 539.

3. Ibid, paragraph 23.

4. Ibid. paragraph 22.

5. Interpretation Bulletin IT-447 – Residence of a Trust or Estate, May 30, 1980.

6. Myron A. Garron and Berna V. Garron, as Trustee of the Garron Family Trust v. Her Majesty The Queen.

7. Agreement Between Canada and Barbados for the Avoidance of Double Taxation and the Prevention of Fiscal Evasion with Respect to Taxes on Income and Capital (the “Treaty”).

8. At paragraphs 159 and 160 of the decision.