http://decisions.fca-caf.gc.ca/fca-caf/decisions/en/item/66815/index.do 
Swirsky v. Canada[1] (February 7, 2014) is a decision of the Federal Court of Appeal affirming a decision of the Tax Court:
http://decision.tcc-cci.gc.ca/tcc-cci/decisions/en/item/31114/index.do 
The underlying facts involved a loss attribution plan that was attacked by the Minister:
[1] In three transactions which took place in 1991, 1993 and 1995, the appellant transferred shares of a family-owned corporation to his wife. His wife incurred interest and carrying costs in connection with loans taken out to finance the transactions. Over a period of time, the appellant claimed losses arising from the shares transferred to his wife. The amount of the losses was equal to the interest and carrying charges paid in connection with the loans.
[2] The appellant claimed the losses pursuant to subsection 74.1(1) of the
Income Tax Act, R.S.C. 1985, c. 1 (5th Supp.) (Act), which attributes income and losses on property transferred from one spouse to the other back to the transferor spouse.
[3] The Minister of National Revenue reassessed the appellant for his 1996 to 2003 taxation years to disallow his claim for the losses on the basis that the appellant’s wife did not use the borrowed money for the purpose of earning income. It followed that there were no losses on the shares that could be attributed back to the appellant.
[4] A judge of the Tax Court of Canada found that the appellant failed to show that his wife had a reasonable expectation of earning income from the shares at the time she acquired them. Because she did not have a reasonable expectation of income when she acquired the shares, she was not entitled to deduct the interest and carrying charges when computing her income from the shares. It followed from this conclusion that the Minister did not err when he disallowed the appellant’s claim to the losses (2013 TCC 73).
[5] The Judge went on to deal with two alternate arguments advanced by the Minister. First, the Judge found that subsection 74.5(11) of the Act did not preclude the attribution of the losses on the shares to the appellant. Second, the Judge found that the general anti-avoidance rule (GAAR) found in section 245 of the Act did not apply to the facts before him.
On one level the decision simply affirms the Tax Court judge’s conclusion that the wife had no reasonable expectation of profit at the time she acquired the shares:
[8] … The appellant asserts that the Judge erred in law by relying inordinately upon the wife’s subjective intention and not enough upon objective manifestations of purpose.
[9] I reject that assertion. At paragraph 30 of the Judge’s reasons, he wrote that in Ludco the Supreme Court held that the test for determining the purpose for interest deductibility was “whether, considering all of the circumstances, the taxpayer had a reasonable expectation of income at the time the investment was made”. He went on to note that the Supreme Court found that a taxpayer’s subjective intention, while relevant, was not conclusive on the question of purpose. In my view, this recitation of the applicable law is that in Ludco and far from over-emphasizing the wife’s subjective purpose, the Judge considered and gave weight to a number of objective manifestations of the purpose for which the shares were purchased:
- There was no evidence that the family corporation ever paid dividends prior to 1999 (reasons, paragraph 32).
- Prior to the transactions, the appellant’s family was supported by shareholder loans paid by the family corporation. Those loans were later transformed into bonuses paid to the appellant. The bonuses were not related to shareholdings (reasons, paragraph 34).
- In the years immediately after the transactions, family expenses continued to be paid by the family corporation. Such payments were treated as loans to family members, regardless of whether they held shares in the corporation (reasons, paragraph 36).
- The family corporation did not have a dividend policy or plan in place to pay dividends on the shares after their acquisition by the appellant’s wife (reasons, paragraph 46).
- The loan transactions were set up so the appellant’s wife, as borrower, would never have to pay interest or carrying charges out of her own pocket (reasons, paragraph 44).
- It could be inferred that the appellant’s wife had a reasonable expectation of receiving a capital dividend. After 1999, the next dividend was paid in 2003 (reasons, paragraph 47).
[10] In my view, these findings of fact were well-founded in the evidence and adequately support the Judge’s conclusion that the shares were not acquired with an objectively reasonable expectation of earning income.
[11] In oral argument, counsel for the appellant argued that the Judge also erred in his application of the test set out in
Ludco by failing to view the relevant transactions through the lens of the family context, including the fact this was a family business. In my view, the Judge did not err as appellant’s counsel argued. Throughout his reasons the Judge was mindful of the nature of the family corporation and the evidence about the family context in which the transactions took place.
What is perhaps more interesting is that the court specifically declined to enter into the GAAR analysis undertaken by the Tax Court judge:
[12] In light of my conclusion that the Judge committed no error of law and no palpable and overriding error of fact or mixed fact and law, it is not necessary for me to address the Judge’s alternate conclusion about the application of the GAAR and I decline to do so. Accordingly, these reasons should not be taken to endorse the Judge’s analysis of this issue, particularly his statement at paragraph 71 of his reasons about the relevant onus of proof.
Accordingly the appeal was dismissed with costs.
Comment: Paragraph 71 of the trial decision that drew the Court of Appeal’s attention was as follows:
[71] The purpose of a transaction is a question of fact and the ordinary rules of onus apply. Since the GAAR was first used by the Minister at the confirmation stage in this case, the respondent is required to show that the primary purpose of the three share dispositions was to obtain a tax benefit.
The trial decision also contains the following paragraph dealing with onus:
[53] The respondent did not rely on subsection 74.5(11) in reassessing Mr. Swirsky or in confirming the reassessments. This argument was raised for the first time in the Reply to Notice of Appeal. While it is possible to raise a new argument at that stage of the proceedings, the onus of proving any of the facts required to support that argument will be on the respondent: Canada v. Anchor Pointe Energy Ltd.[ 2007 FCA 188] Therefore the respondent is required to show on the balance of probabilities that one of the main reasons for the transfer of the Torgan shares was to reduce tax. For the following reasons, I find that the respondent has not met that onus.
The Anchor Pointe decision referred to is a decision which specifically held that the onus was on the taxpayer to demolish assumptions raised by the Minister at the confirmation stage:
[39] With respect, I disagree with this position. First, I do not think that the confirmation of Justice Rip’s finding in paragraph 27 of his reasons in Anchor Pointe Energy Ltd. v. Her Majesty The Queen 2002 DTC 2071 (T.C.C.) goes beyond an acknowledgment of the Crown’s obligation to plead accurately its assumptions of fact. As previously mentioned, Rothstein J.A. reiterated that assumptions of fact can be made at the confirmation stage of the assessment, but never addressed the issue of the onus of proof of these assumptions.
[40] Second, the position taken by the motions judge ignores the second meaning of assessment under section 152 to 177 of the Act, i.e. the product of the assessment as opposed to the process. Either at the initial or at the objection stage, the Minister is attempting to determine the tax liability, and quantum, of the taxpayer. He is entitled throughout this period, until his final determination, to rely upon facts newly discovered or revealed by the taxpayer, and assume them. Nothing in the meaning of assessment requires or permits that some facts be assumed by the Minister, others not, and that, as a result, two categories of assumptions of fact can be created with a different onus for each one. In my respectful view, this runs contrary to the rationale behind the onus of proof, especially in this case where the Minister would have to prove a negative, when all the evidence is in the hands of the taxpayer.
The Supreme Court denied leave to appeal Anchor Pointe.
It is not clear from either the Tax Court decision or the Court of Appeal decision whether the Minister pleaded as an assumption of fact that “the primary purpose of the three share dispositions was to obtain a tax benefit”. If such an assumption was pleaded then it would seem that the Tax Court judge’s conclusion in paragraph 71 was erroneous in light of the decision in Anchor Pointe. If not, then it would appear that his conclusion was correct and the Court of Appeal’s concerns may have been unfounded.
[1] 2014 FCA 36.