In a recent landmark case involving basic transfer pricing principles, Canada v. Cameco Corporation, 2020 FCA 112, the Canadian Federal Court of Appeal sided with the taxpayer. The Court rejected an argument by the Crown that would have applied “realistic alternatives”-like principles to effectively disregard and recharacterize certain related party purchase and sales transactions. For international observers, the case is worth studying, if for no other reason than to understand the government’s aggressive arguments. In light of the codification of the realistic alternatives principle in IRC § 482, the IRS might now be emboldened to make similar arguments in the US.
The dispute centered around Cameco, a Canadian corporation in the uranium business. The Canadian Revenue Agency (“CRA”) had reallocated to Cameco nearly 500 million CAD of profits from Cameco Europe AG (SA, Ltd) (“CEL”), a Swiss subsidiary, which had earned the profits from agreements to purchase uranium. The Crown argued that the profits should be reallocated to Cameco on the theory that Cameco, instead of CEL, would have bought and sold the uranium itself, if Cameco and CEL were arm’s length parties. Cameco estimated that more than 2 billion CAD was at stake, when later years were taken into account.
On appeal, the Court focused on how to interpret the recharacterization rule in section 247 of the Canadian Income Tax Act and concluded that the relevant question is whether arm’s length parties would enter into these transactions—not whether Cameco, in particular, would have. The Court’s decision comports with long-standing arm’s length principles on which US taxpayers rely, despite fears that some—including the OECD—are veering away from these principles.
Factual background. CEL earned the profits at issue from selling uranium it bought under several agreements it acquired from another Cameco subsidiary in 2002. When the agreements were originally signed in 1999, they included options to purchase uranium. In 2001, these agreements were amended so that the subsidiary became obligated to purchase uranium. In Cameco Corporation v. The Queen, 2018 TCC 195, the Tax Court found that, as of 1999, the agreements had negligible value, and any value they had derived from the option feature. The Tax Court further found that when CEL acquired the agreements in 2001, the agreements had become worthless as a result of the loss of the option feature. The uranium market, however, unforeseeably improved after CEL acquired those agreements; CEL subsequently realized substantial profits.
The facts were not truly in dispute on appeal. Although the Court of Appeal observed that several of the Crown’s legal arguments were backdoor attacks on the findings of fact, the Crown did not appeal any factual findings from the 69-day hearing.
Transfer pricing dispute. On appeal, the dispute centered around paragraphs 247(2)(b) and (d), which provide for transfer pricing adjustments for transactions that no arm’s length parties would enter into, other than for a tax benefit. The Crown argued for a subjective test under 247(2)(b) and (d). Invoking concepts loosely similar to the “realistic alternatives” principle in US law, the Crown contended that under a subjective test all CEL’s profits should be reallocated to Cameco because Cameco would not have entered into any of the relevant transactions with any arm’s length party, and would have instead kept the business opportunity for itself.
The Court concluded that the Crown’s interpretation was contrary to the plain language of the statute, which requires an objective test based on hypothetical arm’s length parties. According to the Court, the proper analysis hinges on the price hypothetical arm’s length parties would enter into the relevant transaction. It appears that for the Crown to recharacterize a transaction under 247(2)(b) and (d), it would have to meet the high hurdle of showing that no hypothetical parties dealing at arm’s length would enter into the relevant transaction.
Beyond the plain language of the statue, the Crown’s interpretation would result in improperly ignoring the separate and legitimate existence of CEL, akin to the parallel principle in the US from Moline Properties. If the Crown’s interpretation of 247(2)(b) were to stand, the Court reasoned, then every Canadian corporation wanting to do business in a foreign jurisdiction through a foreign subsidiary would be subject to the CRA’s reallocation of all the profits back to Canada. After all, no corporation that would want to do business in a foreign jurisdiction would sell that opportunity to an unrelated party.
OECD parallels. The Court notes that its interpretation, requiring the Crown to respect Cameco’s legitimate structure and transactions with subsidiaries, is in keeping with the OECD Transfer Pricing Guidelines. But contrasting from the 2017 OECD Guidelines—and similar to US law—the Tax Court emphasized the legal arrangements, including contractual allocations of risk, over DEMPE-like functions. The Court of Appeal pointed out that it is not clear that either Cameco or any of its subsidiaries added any value to the uranium; CEL just bought the uranium and sold it at a higher price. Although the issues in the case predate the 2017 Guidelines, taxpayers should still take comfort that the Court’s interpretation of 247(2)(b) does not depend on the OECD Guidelines. Indeed, the Court explained that the OECD Guidelines are sometimes helpful, but they are not controlling over Canadian statutes.
As the OECD, through proposals like Pillar One, shows its willingness to bend arm’s length principles, taxpayers can find comfort in decisions like this that show the international taxation rules they have relied on for decades have not been blown up just yet.