On March 2, 2016, the Canada Revenue Agency ("CRA") released Transfer Pricing Memorandum 17 "The Impact of Government Assistance on Transfer Pricing" ("TPM-17"). TPM-17 outlines the CRA's guidance on the treatment of government assistance and confirms the CRA's policy: "When a cost-based transfer pricing methodology is used to determine the transfer price of goods, services, or intangibles sold by a Canadian taxpayer to a non-arm's length non-resident person and the Canadian taxpayer receives government assistance, the cost base should not be reduced by the amount of the government assistance received, unless there is reliable evidence that arm's length parties would have done so given the specific facts and circumstances." (emphasis added)
As an economist working at the CRA's Competent Authority Services Division ("CASD") when this policy was introduced, I saw first hand the arguments raised by taxpayers in their applications for Mutual Agreement Procedure ("MAP") or an Advance Pricing Arrangement ("APA") in an attempt to support the "reliable evidence" exception. Now, in light of the pandemic and the government subsidies being given by the Canadian government, the policies in TPM-17 are a hot issue in the Canadian transfer pricing landscape, as taxpayers determine how to address the Canada Emergency Wage Subsidy ("CEWS") for transfer pricing purposes.
How TPM-17 works
As per the example in TPM-17, CanCo is a subsidiary of a multinational enterprise ("MNE") and is conducting research and development ("R&D") activities on behalf of ForCo. CanCo is spending $60 on R&D and $40 on overhead expenses, for total costs of $100. Based on the functional analysis of CanCo and ForCo, it is determined that the appropriate transfer pricing methodology is the transactional net margin method ("TNMM") with a profit level indicator ("PLI") of return on total costs. Based on a comparable company search, it is determined that CanCo should earn 10% of its total costs, resulting in a total paid transfer price amount of $110 by ForCo, and a $10 operating profit to Canco.
If a $10 government subsidy was allocated to CanCo, the CRA's policy implies that the transfer price from ForCo remains $110 and that the $10 subsidy would directly offset CanCo's R&D expenditure, resulting in an operating profit of $20 for CanCo. This is the 'gross cost' approach.
The alternative approach is that the $10 subsidy (or a part thereof) would be netted out of the R&D cost base before the transfer price, resulting in $90 in costs plus a 10% return, and a total transfer price of $99 paid by ForCo to CanCo and a $9 operating profit. In this net cost approach, CanCo shares all (or part of) the government assistance with ForCo and reduces the amount payable to CanCo by $110-$99=$11.
Economic considerations – How arm's length parties would transact
The examples provided in TPM-17 present two very different views on how arm's length parties would interact in these scenarios.
The following economic considerations would support the CRA's policy:
- Let's assume that $100 in the cost base represents 10 hours of work at $10 per hour. These costs would be reimbursed completely by ForCo and the mark-up would apply to the full cost base. Netting out the costs, and applying the mark-up to the reduced base of $90, even if 10 hours are worked, implies a lower return on the total 10 hours worked.
- Again, CanCo is assuming $100 in costs to deliver on R&D activities it has secured with ForCo and, without subsidy, would receive $110 for these activities from ForCo. If CanCo receives the government subsidy, and it is netted out of the cost base, CanCo finds itself losing $1 of profit due to the reduced cost base to be marked-up. At arm's length, why would CanCo reduce its net profit by accepting the subsidy if the end result is to net the cost base with ForCo?
- From a comparability perspective, the accounting treatment of the subsidies may not be consistent between the identified comparables. Unless the comparable has access to the same subsidy, under the same circumstances, it may be difficult to derive a PLI that is reflective of the net cost base. This could result in a lower return per hours worked for CanCo in comparison to the comparables.
Arguments to support a net cost approach can be just as compelling, focussing more on the competitiveness to obtain the ForCo contract and the intent of the sponsoring government to bring, keep or maintain jobs in its jurisdiction.
- As governments generally communicate when subsidy programs are established, including their terms and conditions, ForCo would be very much aware that CanCo could leverage the subsidy to reduce its costs. Hence, ForCo could negotiate with CanCo to reduce the costs charged in order to obtain the R&D contract from ForCo. This scenario can apply if CanCo has many competitors bidding for ForCo's work. However, should CanCo have a strategic advantage to its competitors, ForCo may have less negotiation power to obtain part of the subsidy and succeed in keeping all of it.
- MNE's could also point to the overall intent of the government subsidy, notably that it is intended to attract work to that jurisdiction as further evidence that the subsidy could result in a net cost approach. From this perspective, the government is indirectly negotiating with ForCo, attracting it in its jurisdiction with an incentive that could be passed on by CanCo. As such, the government, through the subsidy, is sharing in the overall cost of the service provided by CanCo.
In the end, the CRA policy begins with the gross costs approach and it is up to the MNE to demonstrate that arm's length parties would share all or part of the subsidy.
COVID-19: Why is TPM-17 relevant today?
Since May 30, 2020, the Government of Canada has invested heavily to offset the devastating impact of the COVID-19 pandemic on various sectors of the Canadian economy through the CEWS. At June 27, 2021, 3.8M applications to CEWS have been approved, of which 5,940 claims were for an amount exceeding $1M.
While the circumstances with CEWS are different than the R&D example in TPM-17, the underlying principle would remain the same: if any of the above firms that requested CEWS were a CanCo, the CRA would expect that these firms apply a gross cost approach to determine the transfer price. Otherwise, they would have to provide reliable evidence that arm's length parties would share all or part of the subsidy with ForCo. However, since the policy behind CEWS is more akin to national unemployment insurance regimes, which are meant to be temporary in nature, as opposed to a subsidy intended to attract work to Canada, it will be difficult for taxpayers to convince the CRA to back off from the gross cost approach in TPM-17.
Further, as the comparables may not have access to the same subsidy, it is difficult to capture the impact of the various relief measures governments have put in place and their impact on the financials considered for determining CanCo's return under a TNMM. In a previous article, we discuss differences in comparability in greater detail, particularly as it applies to this situation.
The CRA's TPM-17 provides multinationals with a clear direction as to how a CanCo should reflect any subsidies received to support research and development or manufacturing activities. However, TPM-17 also recognizes that, with reliable evidence, arm's length parties could have behaved differently given specific facts and circumstances.
In any event, it is important that companies conduct thorough analysis to support their approach, especially if the functional analysis leads them to determine that they have reliable evidence that the "gross cost" approach does not reflect their particular situation. When in receipt of government subsidies, including CEWS, taxpayers should always discuss their scenario with a transfer pricing expert before accepting CRA's preferred gross cost approach. This ensures that the facts and the circumstances applicable to their situation reflect how arm's length parties would treat these subsidies.