On April 12, 2012, the Supreme Court of Canada (“SCC”) dismissed the taxpayer’s appeal in St. Michael Trust Corp. v. The Queen. The SCC held that two Barbados trusts, which realized capital gains on the sale of shares of Canadian corporations, were resident in Canada at the time of the sale for Canadian income tax purposes and therefore subject to Canadian tax on the capital gains, rather than being resident in Barbados and entitled to rely on the capital gains exemption in the Canada-Barbados Income Tax Convention (the “Treaty”). The case is significant because the SCC confirmed the decision of the lower courts that the correct legal test for determining the residence of a trust for Canadian tax purposes is where the central management and control of the business of the trust actually takes place (rather than simply where the trustee resides).
FACTS
The trusts were established in the course of a reorganization of the share structure of PMPL Holdings Inc. (“PMPL”), which owned shares of a Canadian corporation that manufactured and assembled parts for the automotive industry. Prior to the reorganization, the shares of PMPL were owned equally by two Canadian residents: Mr. Dunin and a holding company that was wholly owned by Mr. Garron, Mr. Garron’s wife and the Garron Family Trust. The trusts were settled under Barbados law, one trust for the benefit of Mr. Dunin and his family, and the other trust for the benefit of Mr. Garron and his family. Both trusts were settled by a friend of Mr. Garron who was resident in St. Vincent, the sole trustee of both trusts was a corporation resident in Barbados which provided trustee services, and the protector of both trusts (who had the power to remove and appoint trustees) was another friend of Mr. Garron who was resident in St. Vincent.
As part of the reorganization of PMPL, the existing shareholders of PMPL exchanged their common shares for fixed value preference shares of PMPL. Each trust subscribed for shares of a newly incorporated Canadian corporation, with each corporation in turn subscribing for common shares of PMPL. The share subscriptions were transacted based on a valuation of the common shares of PMPL immediately before the reorganization of $50 million. Two years later, when each trust sold shares of its respective holding corporations to an arm’s length purchaser, PMPL was valued at approximately $532 million. Capital gains of over $450 million were realized by the trusts on the share dispositions and were not subject to Barbados income tax.
As the shares sold were taxable Canadian property, amounts on account of potential Canadian tax on the capital gains were remitted to the Canada Revenue Agency under the reporting procedures in section 116 of the Income Tax Act (Canada) (the “Act”). The trusts filed Canadian income tax returns for the year the shares were sold and sought a refund of the remitted amounts based on the capital gains exemption in the Treaty. The Minister of National Revenue (“Minister”) denied the trusts the benefit of the Treaty exemption and assessed the trusts Canadian tax in respect of their capital gains on the sale.
LOWER COURT DECISIONS
On appeal of the Minister’s assessments, the Tax Court held that the test for determining trust residence should not be based on a mechanical determination of the residence of the trustee(s), as the case of Thibodeau v. The Queen did not support such a test. Instead, the appropriate test should be consistent with the central management and control test in the corporate context, which requires a court to determine where a corporation is actually managed and controlled. The Court found on the facts that the trusts were resident in Canada when the shares were sold because general decision-making in respect of the trusts was carried out by Mr. Dunin and Mr. Garron, and not the trustee which played only a limited role executing documents and providing administrative services in respect of the trusts. As the trusts were found to be resident in Canada, the capital gains exemption in the Treaty for Barbados residents did not apply to exempt the capital gains arising on the share dispositions from Canadian tax.
The Federal Court of Appeal (“FCA”) dismissed the taxpayers’ appeal, agreeing with the Tax Court that a central management and control test should be applied in determining the residence of the trusts, and further concluding that the Tax Court made no error in determining, based on the facts presented at trial, that the trusts were resident in Canada at the time the shares were sold.
The FCA also expressed its opinion on the Crown’s alternative arguments, on which the Tax Court had also commented. First, the FCA disagreed with the finding of the Tax Court that a trust could not be said to have indirectly acquired property from a taxpayer (for purposes of paragraph 94(1)(b) of the Act) through a series of corporate transactions whereby economic value was “shifted” to the trust. Instead, the FCA held, based on Kieboom, that the trusts could be regarded as having acquired property from the taxpayer for purposes of the non-resident trust rules in section 94 of the Act. However, even if the rules in section 94 of the Act applied to the trusts in this case, the FCA nonetheless agreed with the Tax Court that the capital gains exemption in the Treaty trumped section 94 because section 94 applied only for certain purposes of Part I of the Act and was insufficient to deem the trusts to be residents of the Canada for purposes of the Treaty. Thus, if the trusts were otherwise considered to be exclusively residents of Barbados for purposes of the Treaty, the trusts would be entitled to the benefits of the Treaty as Barbados residents. As for the issue of whether the general anti-avoidance rule (“GAAR”) in section 245 of the Act should be applied in this case if the trusts were found to be non-residents of Canada, the FCA again agreed with the Tax Court that the series of transactions that resulted in the trusts becoming entitled to the capital gains exemption in the Treaty was not a misuse or abuse of the Treaty and, therefore, the GAAR did not apply.
THE SUPREME COURT DECISION
On appeal, the taxpayer argued that the test for determining the residence of a trust must be the residence of the trustee for two reasons. First, since the trust is not a separate legal person like a corporation, the central management and control test for corporations is not applicable to trusts. The SCC quickly dismissed this argument (at paragraph 10) by noting that “[w]hile a trust is not a person at common law, it is deemed to be an individual under the Act…[T]he fact that at common law a trust does not have an independent legal existence is irrelevant for the purposes of the Act”.
Second, the taxpayer argued that subsection 104(1) of the Act links a trust to its trustee(s) for all attributes of the trust, including residency, and therefore the residence of the trust must be the residence of the trustee. Subsection 104(1) states that “[i]n this Act, a reference to a trust or estate … shall, unless the context otherwise requires, be read to include a reference to the trustee, executor, administrator, liquidator of a succession, heir or other legal representative having ownership or control of the trust property …”. The SCC disagreed with the taxpayer, concluding that subsection 104(1) is not a principle of general application to trusts, and that nothing in its context suggests a legal rule requiring the residence of a trust to be the residence of the trustee. Moreover, the basic income tax charging provision in subsection 2(1) of the Act suggests the opposite by referring to a “person”, which must be read as a reference to the taxpayer whose taxable income is subject to income tax (i.e., the trust and not the trustee). In the view of the SCC, that conclusion follows from subsection 104(2) of the Act, which separates the trust from the trustee in respect of trust property.
As additional support for its conclusion that the central management and control test should be adopted in determining the residence of a trust for Canadian tax purposes, the SCC noted that there are many similarities between a trust and a corporation which justify the use of a similar test for residence, e.g., both hold assets that are required to be managed, both acquire and dispose of assets, both require banking and financial arrangements, both distribute income (either by way of dividends or distributions), and both potentially require the management of a business, and the instruction or advice of lawyers, accountants and other advisors. At a basic level, the function of a corporation and a trust is to manage property.
Furthermore, the SCC agreed with the Tax Court that adopting a similar test for trusts and corporations promotes “the important principles of consistency, predictability and fairness in the application of tax law”. As the taxpayer did not offer any good reasons for using a different test for trusts than for corporations, the SCC concluded that the central management and control test applicable to corporations should also apply to trusts.
Applying the central management and control test to the facts as found by the Tax Court, the SCC agreed that the trusts were resident in Canada. While the facts of this case led to the conclusion that the residence of the trustee was not the residence of the trusts, the SCC noted that the residence of the trustee will also be the residence of the trust in a different case if the trustee exercises central management and control of the trust and such duties are performed in the place where the trustee resides.
Although it was not necessary for the SCC to consider the Minister’s alternative arguments on the application of section 94 and the GAAR in light of its conclusion that the trusts were resident in Canada, the SCC stated (at paragraph 19) that it “should not be understood as endorsing the reasons of the [FCA] on those matters”.
COMMENTS
The SCC’s decision is not surprising, in light of the Tax Court’s carefully considered reasons for adopting the central management and control test and its findings of fact in this case. This case definitively establishes the central management and control test as the correct legal test for determining the residence of a trust for Canadian tax purposes. It also reinforces the notion that the place where the trustee resides will determine the residency of the trust if the trustee exercises central management and control of the business of the trust in that place.
On the other hand, a number of issues appear to remain unresolved by the decision of the SCC and those of the lower courts. By not providing a framework or detailed test for determining whether a trustee exercises central management and control of a particular trust, this decision of the SCC may offer limited guidance in practice. For example, if the trustee in this case had been more active in negotiating the sale of the assets of the trusts, would that factor alone have been sufficient to establish the residency of the trusts in Barbados? Moreover, where would a trust be resident if the trustee is a corporation whose board of directors is resident in Bermuda but the “back office” administration and investment decisions regarding the assets of the trust are actually carried out in Singapore?
The explicit comment by the SCC at the end of its judgment that its decision should not be construed as endorsing the reasons of the FCA on section 94 and the GAAR also raises some interesting questions. Could the Court’s comment instead be construed as admonishing the FCA for relying on Kieboom in its analysis of paragraph 94(1)(b) of the Act? Would the SCC have approached the issue of GAAR differently than both the Tax Court and the FCA?
