BlackBerry’s Transfer Pricing in Light of BlackBerry Limited v. The King

This is a guest post by Harold McClure, a New York City-based independent economist with 26 years of transfer pricing and valuation experience.

BlackBerry Limited v. The King (2023 TCC 137) represents an attempt by the Canadian Revenue Agency (CRA) to tax profits earned by BlackBerry’s US subsidiary under Canadian Foreign Accrual Property Income (FAPI) rules. The Canadian tax law issues are not this economist’s forte, but transfer pricing was raised by one of BlackBerry’s expert witnesses in a way that raises interesting questions.

This article considers the potential wider transfer pricing implications of the arguments put forward on behalf of BlackBerry in that litigation, including in particular the impact on the group’s US transfer pricing positions. It presents a possible model for the allocation of BlackBerry’s income for fiscal year ended February 28, 2011 (which was BlackBerry’s best year) and considers the implications for the group’s effective tax rate. It also addresses the fact that a portion of BlackBerry’s R&D was conducted by US employees under a contract R&D arrangement, and suggests some key considerations in conclusion.

Background to the Canadian FAPI Rules

Let’s begin with an explanation from Dominion Tax Law[1] of how the FAPI rules allow the CRA to tax foreign income:

[I]f subsidiaries of Canadian corporations earn ‘passive income’ (everything that is not ‘active business income’ or “ABI”), the Canadian parent corporation has to pay Canadian tax on said foreign passive profits. BlackBerry Canada paid BlackBerry US to do research, and since doing research for a fee is service income, which is deemed by the FAPI rules to be passive income not ABI, the US profit of a US company is bizarrely taxable in the hands of the Canadian parent. Normally, the Canadian parent would get tax credit in Canada against the tax on FAPI for all of the tax paid in the US on the US profit by the US subsidiary. As a result, because of such tax credits, usually no net tax is payable on the FAPI attributed to the Canadian parent when the foreign subsidiary operates in a jurisdiction with a roughly comparable level of corporate tax. However, in this case, BlackBerry US received US research and development (“R&D”) tax credits, wiping out the US tax on the US profit, so there was technically no US tax paid. Consequently, the Canada Revenue Agency (“CRA”) contends that BlackBerry Canada has to pay tax in Canada on the FAPI attributed to the ‘passive income’ profits of BlackBerry US.

To counter the application of the FAPI rules, the representatives of BlackBerry hired a Canadian transfer pricing economist to explore the implications of alternative contractual structures. Dominion Tax Law comments as follows:

The first report was by Mr. Brad Rolph, a transfer pricing economist, partner with Grant Thornton Consulting and, for most of the past decade, National Leader of Grant Thornton’s transfer pricing practice in Canada. In a nutshell, his report was arguing that BlackBerry Canada had several other ways of arranging its affairs in compliance with [the Income Tax Act (Canada)] regarding its R&D activities and the economic exploitation of the consequential intellectual property created, and each of them would have resulted in less tax payable in Canada.

However, BlackBerry deliberately chose the structure that would not harm Canada’s tax base, and indeed resulted in more tax payable in Canada than all of the legally acceptable alternatives. The conclusion BlackBerry wants the TCC to draw from this is that hitting it for extra FAPI tax when it was already being a ‘good Canadian corporate citizen,’ and paying more tax in Canada than it would otherwise have to, would somehow be wrong.

This argument strikes me as a dangerous way of trying to avoid a relatively minor tax issue, as it may invite a challenge from the IRS. However, there may well be additional considerations in play which I am not privy to.

A Pure Distribution Model

To put the issues in context, consider that BlackBerry had nearly $20 billion in total revenue during fiscal year ended February 28, 2011, with 40% of these sales in the US, 10% in the UK, 7% in Canada, and 43% in other nations. BlackBerry sourced its products from third party vendors, at a cost approximating 53% of sales. The cost of providing technical services was approximately 3% of sales, and ongoing R&D represented 6.5% of sales. Selling costs represented 12.5% of sales, so consolidated operating profits represented 25% of sales.

Table 1 presents the allocation of consolidated profits between the Canadian parent and the distribution affiliates in a model where the Canadian parent bears the cost of providing technical services and all R&D expenses, with the distribution affiliates bearing the selling costs. The model suggests that 92% of worldwide income would be attributed to the Canadian parent if all intangible assets were owned by the parent and the operating margins for the distribution affiliate were only 2%. The 40-F filings for BlackBerry indicate that indeed only 8% of BlackBerry’s worldwide income was allocated to the foreign affiliates.

The transfer pricing issue can be viewed through the filter of the gross margin for the distribution affiliates, with Table 1 presenting three possible alternatives.

Table 1: BlackBerry Allocation of Income and Effective Tax Rate Under Three Distributor Margins

Table 1 assumes that the intercompany policy grants the distribution affiliates with a 14.5% gross margin, which would be consistent with a 2% operating margin and a markup over selling expenses = 16%. Under this policy, the distribution affiliate’s share of worldwide income would be only 8%, consistent with the information presented in BlackBerry’s 40-F filing.

The income tax section of this filing also noted that the Canadian tax rate was only 30.5%, which may have been lower than the tax rates for its distribution affiliate. Table 1 also assumes that the tax rate for its foreign affiliates is 34.5%, so every 1% increase in the operating margin of the distribution affiliates increases worldwide taxes by $8 million. Had all income been allocated to the Canadian parent, income taxes would have been only $1525 million. Since 8% of consolidated profits were allocated to the distribution affiliates, worldwide taxes were $1541 million. Under this allocation of income, the effective tax rate (ETR) would be 30.82%. We should note that actual income taxes paid were lower for two reasons including a one-time goodwill write-off and tax credits for R&D.

The IRS and other tax authorities might argue that the appropriate gross margin should be 15.5%, which would imply a 3% operating margin and a 24% markup over selling costs. Under this alternative policy, the share of income accruing to the distribution affiliates would rise to 12%, lowering the share of income according to the parent corporation falling to 88%. Under this allocation of income, ETR rises to 30.98%. Table 1 also considers the implications of an intermediate policy where the gross margin = 15%, implying a 2.5% operating margin and a 20% markup over selling costs.

This model assumes that BlackBerry would prefer less income allocated to the foreign distribution affiliates. BlackBerry would also want to avoid any double tax disputes where the CRA insisted on a lower gross margin but the foreign tax authorities insisted on a higher gross margin. The model is also predicated on the assumption that the foreign affiliates, including the US affiliate, are merely routine distributors that do not own valuable intangible assets.

The R&D Issues

The 40-F filing for BlackBerry noted that some of its R&D personnel are employees of the US affiliate. The economist report prepared by the representatives of BlackBerry suggested that the US affiliate incurred approximately 22% of overall R&D expenses under a contract R&D arrangement, where the parent paid the US subsidiary 108% of US incurred R&D costs. Table 2 assumes that US incurred expenses were $210 million during fiscal year ended February 28, 2010, and $290 million during fiscal year ended February 28, 2011. Under the intercompany policy, US profits would average $20 million per year for its contract R&D operations.

Table 2: US Contract R&D

If the US R&D operations were structured in a clearly articulated contract R&D contractual arrangement, the only potential dispute between the IRS and the CRA would be over the markup. The IRS might argue for a higher markup, while the CRA might argue for a lower markup. BlackBerry likely had some form of benchmarking report under the Transactional Net Margin Method to support its 8% markup. Changes in this markup would have only a trivial impact on the allocation of income, given the modest level of US-incurred R&D expense.

The economist report prepared on behalf of BlackBerry’s position in the FAPI dispute suggested that the R&D operations could have been structured in various alternative ways. For example, the US affiliate could have conducted its R&D outside of a contract R&D arrangement and later transferred the value of its intellectual property to the Canadian parent at fair market value.

The determination of this fair market value would be a challenging economic analysis, which the economist’s report did not attempt to perform. This value would also be considerable given the high profits during this period and BlackBerry’s stock market valuation. The market capitalization for BlackBerry reached $50 billion for a while but quickly fell below $10 billion in subsequent years. BlackBerry had a first mover advantage in the development and sales of smartphones. Even though Apple launched its iPhone in 2007, this new product did not do well in sales to enterprises despite its success in terms of consumer sales. However, Google’s Android technology and its relationships with phone manufacturers leap-frogged the suite of BlackBerry smart phones.

Another approach could have been a cost sharing arrangement between the Canadian parent and the US affiliate. Give US sales represented 40% of worldwide sales, the US affiliate would likely have made cost sharing payments to the Canadian parent but would have enjoyed a substantial share of worldwide profits as co-owner of the worldwide intangible assets.

The IRS might be tempted to assert that one of these alternative structures be used to perform any transfer pricing analysis, as the implication would be that the US share of worldwide income would be much higher than the allocation, under the view that the US affiliate is a mere distributor performing a modest amount of contract R&D. Such an attempt by the IRS would certainly be controversial, as it is similar to its failed attempts in the Westreco v. Commissioner (60 T.C.M. 824, 1990) litigation. In that case, the IRS originally asserted that any intangibles created by US R&D belonged to the US affiliate of Nestlé, but the taxpayer had a clear contract R&D agreement where the US affiliate received total costs plus a modest markup. The IRS position was then limited to challenging the markup, but the taxpayer provided a convincing economic analysis that this markup was reasonable.

BlackBerry appears to have similarly structured a clear contract R&D contractual arrangement, where the US affiliate received total cost plus an 8% markup. It does seem strange, therefore, that its own representatives are posing possible alternatives. To be fair, however, the Canadian FAPI rules are unusual in that they attempt to tax the modest income of its US affiliate, and it would be dangerous to speculate about the pros and cons of different approaches without the full context.

Conclusion

The above discussion illustrates how attempts to mitigate the impact of local tax rules can have wider implications as regards global transfer pricing risks. Although the vagaries of litigation may suggest specific tactics to address specific situations, they need to be tested in the context of the group’s global position. The foundations of a robust global strategy include a macro analysis of the allocation of profits, as well as the micro delineation of individual transactions, taking into account non-TP considerations such as R&D tax credits, withholding taxes, sales taxes and customs duties.

[1] Dominion Tax Law’s case comment on BlackBerry Limited v. The King 2023 TCC 137.