Supreme Court of Canada considers residency of a trust in St. Michael Trust Corp.

St. Michael Trust Corp. v. The Queen is the first appeal where the Supreme Court of Canada has considered the test for residency of a trust for tax purposes. The appeal was heard merely four weeks ago.  We were surprised when printing out the Court’s judgment that its pages could be bound by a standard staple, which is exceedingly rare for a complex tax appeal.  The importance of residency status is that a Canadian-resident taxpayer, including a trust, is taxable in Canada on worldwide income.

The St. Michael Trust case involved an offshore estate freeze.  The plan was to avoid Canadian capital gains tax on the eventual sale of a Canadian corporation, PMPL.  To achieve this, a trust was formed in Barbados (the “Fundy Settlement”), which subscribed for shares of a new Canadian holding company, which in turn held shares of PMPL.  The trustee of the Fundy Settlement was St. Michael Trust Corp., a Barbadian trust company owned by partners of a large public accounting firm.  The beneficiaries of the Fundy Settlement were Mr. Garron and his family.  A similar structure, with a holding company and a trust (the “Sommersby Settlement”) was implemented for the other shareholder of PMPL, Mr. Dunin, the beneficiaries of which were Mr. Dunin and his family.

Ultimately, the shares of the holding companies were indeed sold, resulting in capital gains of CAD $217 million for the Fundy Settlement and CAD $240 million for the Sommersby Settlement.  Both trusts claimed their respective capital gains were exempt from Canadian tax by virtue of the Canada-Barbados tax treaty.  The treaty exemption was only available if the trusts were resident in Barbados.

The trustee argued that the appropriate test for residency status of a trust is the residence of the trustees, consistent with previous case law.  The Minister of National Revenue (the “Minister”) said the test should be “central management and control”, which is the residency test used for corporations.  Before this case, there had been little judicial authority regarding the appropriate test for residency of a trust, so practitioners were hopeful that the Supreme Court would provide some helpful guidance in this critical and uncertain area of tax law.

As discussed in our past blogs, the Minister was successful at the Tax Court of Canada and the Federal Court of Appeal.  The Minister was also successful at the Supreme Court of Canada, which decided the test for residency of a trust is central management and control.  The Court described a trust’s place of central management and control as being the place “where its real business is carried on”.  In this case, the Tax Court’s finding was that the beneficiaries of the trusts exercised central management and control from Canada whereas the role of the trustee was limited to administrative services and the trustee had little or no responsibility.  Applying the central management and control test to these facts, the Supreme Court upheld the lower Courts’ decisions.

The Supreme Court agreed with the Tax Court’s earlier statement that adopting a similar residency test for trusts as for corporations promotes “the important principles of consistency, predictability and fairness in the application of tax law.”  Despite these views expressed by the Courts in this case, from a practitioner’s perspective we think this case leaves open a number of issues worthy of comment.

  1. The central management and control test is said to be the place “where its real business is carried on”.  But what does this mean in the context of a trust?  The “business” of a trust is typically to simply hold property, which depending on the situation may involve very little activity.  Unlike a corporation, there may not be day-to-day decision making required.
  2. The Federal Court of Appeal provided some guidance regarding the central management and control test.  In Justice Sharlow’s view, it is acceptable for a beneficiary to make recommendations to trustees, but if beneficiaries exercise the trust powers or trustees’ discretions, then the beneficiaries are controlling the trust.1  Did the Supreme Court intend to adopt the analysis of the Federal Court of Appeal on a wholesale basis, at least as far as the central management and control test is concerned?2
  3. If the location of central management and control changes from year to year for whatever reason, does this mean the residence of a trust also changes from year to year?
  4. A trustee has a fiduciary duty to manage the trust property.  Is the central management and control test predicated on the presumption that a trustee is generally in breach of that duty, at least for purposes of a tax appeal, until the trustee proves otherwise?
  5. Further to the above, if central management and control is found to be exercised by someone other than a trustee (for example, a beneficiary or a financial advisor), will that trustee be in breach of the fiduciary duties it owes to the beneficiaries for allowing its control to be subverted?
  6. The genesis of the central management and control test was a 1906 decision in De Beers Consolidated Mines, Ltd. v. Howe­.3  However, business was very different and much more regional in nature than it is today.  In 1906, there were no fax machines, no e-mail, no internet, aviation was in its infancy and there was no commercial air travel, and long distance telephone calls were much more expensive and less reliable.  Today, it is not uncommon for a board of directors to be comprised of individuals in various jurisdictions making decisions using electronic communication.  Arguably, the world has changed sufficiently that it is worthwhile for a Court to reconsider the principles in the De Beers case.
  7. A trust is fundamentally different from a corporation.  For example, a trustee owes fiduciary duties to its beneficiaries, whereas directors of a corporation do not owe fiduciary duties to the shareholders.  Instead, they owe a fiduciary duty to act in the best interests of the corporation.  Is the central management and control test reconcilable with this difference?
  8. Given the above, one could query whether or not this decision will help to achieve the principles of consistency, predictability and fairness in the application of tax law.  It seems to us that this case has the potential to do quite the opposite when trying to assert where a trust is resident for tax purposes.
  9. The United States has a codified test to determine whether the trust is domestic or foreign for US purposes.4  Should Canada adopt a similar model to achieve the consistency, predictability and fairness principles espoused by the Courts?

Stay tuned… much ink will be spilled on this matter in the years to come.  In the meantime, trustees will have to be mindful of the central management and control test to ensure that the trust’s residency is in the place that they desire it to be.


1. 2010 FCA 309 at paragraph 68.

2. We note the Court was explicit that it was not necessarily adopting the reasoning of the Federal Court of Appeal on certain other issues that it did not consider necessary to decide upon.

3. [1906] A.C. 455 (H.L).

4. IRC § 7701(a)(30)(E) and (31)(B).