For this special edition of Kroll's Transfer Pricing Times, we asked regional leaders across our transfer pricing practice for their response to six key questions posed by and addressed in Guidance on the Transfer Pricing and Implications of the COVID-19 Pandemic released by the Organization for Economic Cooperation and Development (OECD). Leaders covering the United States, Canada, Latin America, Europe and the Asia Pacific provided their thoughts on how local tax authorities may perceive the OECD paper as well as some practical considerations for taxpayers.
What Practical Approaches May Be Available to Address Information Deficiencies, Such as “outcome-testing” Approaches?
Due to the time lag in availability of the financial data of comparables, there are three potential options for making financial adjustments to compensate for the impact of the COVID-19 pandemic:
- Adjust the results of the tested party as far as possible to eliminate the impact of COVID-19 for the period in question, likely FY2020, and compare the adjusted results of the tested party to the range of the comparable companies’ results which will likely be from a multiple year period before the impact of COVID-19, such as FY2017-2019.
- Determine the actual impact of COVID-19 on the sales, gross profit and operating expenses of the tested party, and adjust to the same extent the financial data of the comparable companies from the pre-COVID-19 period, such as FY2017-2019. This would likely lead to a lower range against which the tested party’s actual performance in FY2020 can be compared for the purposes of determining compliance with the arm’s-length principle.
- Carry out the financial analysis of the tested party and comparables over a longer multiple-year period of review, say five years or more, to even out the relative impact of COVID-19 on FY2020. This more basic approach is not considered the best approach due to the exceptional (once in a century) circumstances of COVID-19. It is, however, specifically identified as an option by the Singapore tax authority.
Another approach is to look at the impact on company profitability of past pandemics or economic shocks; however, the economic impact of COVID-19 is unique in many ways and still evolving in many places. Therefore, observable data from previous and arguably lesser/different epidemics/pandemics or economic shocks may not be particularly informative. In addition, the availability of benchmarking data on commercial databases is evolving rapidly in Asia, so coming up with a suitable set of comparables over an extended period would be practically difficult outside of Japan, Korea and Australia.
The Australian Taxation Office (ATO) said it will seek to understand the financial outcomes the business would have achieved “but for” the impact of COVID-19. The emphasis is on gathering evidence to support any impacts of COVID-19 on the business. This may include a variance analysis of budgeted (pre-COVID-19) versus actual results as well as a broader analysis of how the relevant industry has been affected. These impacts and any changes in business strategies should be documented as they are considered and implemented.
The Internal Revenue Service (IRS) expects taxpayers to analyze intercompany transactions based on the actual results for the taxable year under review and to consider various comparability factors in that analysis. One of the comparability factors mentioned is significant economic conditions that could have an impact on the agreed upon pricing. As such, it will be important that a comparability analysis review the impact the COVID-19 pandemic had on the parties to the controlled transaction compared to the comparable uncontrolled transaction or comparable companies.
Canada’s regime allows, and in fact expects, an “outcome testing” approach that may take place only after the tax year has closed, with transfer pricing adjustments allowed before the filing of the tax return. Many taxpayers are also reviewing quarterly year to date financial information from comparables, to consider adjustments before year end.
In general, the documentation rules across the region expect documentation on an “outcome testing” basis. There has always been a disconnect between transfer pricing documentation and income tax filings as tax returns are usually filed between March and April, while transfer pricing documentation deadlines in the region are generally June or later. This will present significant challenges because tax authorities will expect comparable financials to include COVID-19 affected results, and those results will mostly be unavailable at the time of the income tax filing. Most regulations allow for the use of extraordinary or business cycle adjustments to comparable companies. It is expected that many taxpayers will either adjust pre-pandemic comparable information based on tested party outcomes or file amended tax returns once post-pandemic financial information is available.
Tax authorities will attempt to be pragmatic and flexible in allowing taxpayers to rely on information that becomes available after the end of the tax year and before the tax return is filed. How exactly this “outcome testing” approach will work in practice remains to be seen. Its effectiveness in relieving taxpayers from double taxation will depend heavily on a smoothly functioning and highly coordinated post-COVID-19 Mutual Agreement Procedure (MAP) process, as noted in paragraph 21 of the OECD’s guidance, so taxpayers have cause for concern here.
Can Loss-making Comparables Be Used?
There is no explicit requirement to exclude loss-making companies in a comparability analysis in the United States. However, an example from the IRS’s Transfer Pricing Documentation Frequently Asked Questions (“IRS TP FAQs”) released on April 14, 2020, indicates that the IRS expects robust transfer pricing documentation to explain how unforeseen business circumstances experienced by the tested party could have caused operating losses. 1Without this type of support, the IRS may question the reliability of a comparable analysis that supports a tested party with losses.
Canada has no specific prohibition against relying on loss-making comparables, even for “limited risk” arrangements, and from a practical perspective one would expect that 2020 benchmarking sets are more likely to include some loss-making companies. However, supporting unusual 2020 results through multi-year averaging may not be feasible in Canada, where the Canada Revenue Agency’s (CRA) administrative policy defaults toward single-year testing.
Some countries, such as Guatemala and Ecuador, explicitly demand the exclusion of comparable companies with losses. Other jurisdictions like Panama, Colombia and Mexico do not explicitly require loss making comparables be excluded but usually reject them in practice, and in general, tax authorities in the region do not appreciate the use of such comparables. However, in their regulations, Guatemala and Ecuador also allow the use of comparables with losses if the reason to use those comparables is extensively documented. Other jurisdictions will follow the same rationale.
Typically, there is a reluctance among Asia Pacific tax authorities to allow the use of loss-making comparables. In certain cases, such as in Indonesia, the tax authority can go further to reject any comparable that has made a loss in any one year of the review period, or is lacking data in any one year in the review period. Other jurisdictions may allow loss makers but see them only as outliers and expect that at least the lower quartile would be positive. Certainly, for practical risk management purposes, taxpayers should try to avoid reliance on loss-making comparables. If this is not possible, then taxpayers can use on loss-making comparables and be prepared to argue for their inclusion in later discussions with the relevant tax authority, relying on the commercial reality that independent parties sometimes make losses without ceasing to be a going concern, as well as supportive comments by the OECD on this topic.
In principle, yes, loss-making comparables can be used. In practice, taxpayers will need to demonstrate that the loss-making comparables and the tested party have truly comparable risk profiles, which will be a challenge for most taxpayers, especially those with limited risk tested parties. Independent parties dealing at arm’s length could very rarely be classified as limited risk entities. “Full risk” entities are on firmer ground. The clarity of the relevant intercompany agreement and the consistency with which the parties adhered to that agreement pre- and post-COVID-19 will be the focus of investigation.
Can Entities Operating Under Limited Risk Arrangements Incur Losses?
Early indication from the IRS’s Advance Pricing and Mutual Agreement (APMA) group is that with detailed and relevant data, such losses may be appropriate in the pandemic year(s). However, it remains to be seen how this could play out in a regular audit cycle as opposed to a mutual agreement context such as an Advance Pricing Agreement (APA).
Listed below are certain important considerations in determining whether a taxpayer can support an operating loss for a controlled transaction when assuming limited risks:
- Identifying the actual factors that resulted in losses (e.g., limited risk entity non-operational for a period of time due to local authorities attempting to control the pandemic?)
- Consistency of actual dealings with terms and conditions outlined in the intercompany agreement
- Operating results of the counter party to the transaction (e.g., did the entrepreneurial entity also incur losses?)
The term "limited risk" is not explicitly stated in most regulations for countries in Latin America. However, in practice, local tax authorities expect to see profits by those entities categorized as limited risk in the intercompany agreements or previous transfer pricing documentation. This is likely to be the most contested issue by tax authorities arising from the pandemic for.
The acceptance of losses for limited risk entities depends on the facts and circumstances of each case and in how such have been documented in previous transfer pricing analyses and intercompany agreements. Taxpayers should carry out a detailed review of their related party arrangements (including legal agreements, but also conduct of the parties and what independent parties would have agreed to in similar circumstances) to determine, with appropriate evidence, the degree to which related entities bear risks. If it is found that entities are truly no risk in legal form and (more importantly) commercial substance, and this can be supported by referring to how independent parties would structure their arrangements, then such entities should not incur losses. However, in the majority of cases, it is likely that related entities will be found to be risk bearing at least to some extent and, as such, it is possible that they may incur losses, or share in the losses of the value chain. This conclusion can be supported by observations of short-term loss-sharing behavior by independent parties seeking to keep their global value chains intact during the COVID-19 pandemic.
Potentially, yes loss making comparables can be used, if the available arm’s length data supports that outcome under the facts and circumstances. Even “limited risk” companies are not absolutely risk free. Intercompany contracts should be reviewed, and could be a barrier to short-term deviations on profit outcomes.
Again, in principle, yes, entities operating under limited risk arrangements can incur losses, but only to the extent that losses arise from risks against which they were not contractually protected. The allocation of risks set out in intercompany agreements will be an important guide, as well as how such allocations were described in previous transfer pricing analyses. To the extent that an intercompany agreement “guarantees” a certain return to an entity, tax authorities following the OECD guidelines will expect that contractual guarantee to be respected.
Under What Circumstances May Arrangements Be Modified to Address the Consequences of Covid-19?
Independent parties dealing at arm’s length will, in times of economic hardship, renegotiate to protect and preserve a long-term mutually beneficial arrangement. While that statement appears obvious and empirically true, tax authorities have often struggled to accept this behavior is consistent with the arm’s-length principle in related party transactions. The OECD’s guidance in paragraphs 42-46 adopts a skeptical tone, perhaps reflecting tax authorities’ fear that temporary measures that adversely affect their tax base might become permanent in a future that they cannot anticipate. According to this latest guidance, in practice, related-party arrangements may be modified to address the consequences of COVID-19 only where the taxpayer can provide clear evidence that independent parties in comparable situations engaged in comparable transactions would and have renegotiated their arrangements. This burden of proof on the taxpayer feels quite high.
The guidance here also seems to conflict with guidance given earlier in paragraph 30 of the OECD’s Guidance: “One potential solution to the uncertainty caused by the COVID-19 pandemic would be to allow for the inclusion of price adjustment mechanisms in controlled transactions.” It is likely that this conflict arises from tax authorities’ supposition that a price adjustment clause like that anticipated in paragraph 30 might work to their advantage, “providing flexibility” as the guidance puts it, whereas a contractual renegotiation anticipated in 42-46 might not. Compiling evidence from comparables to support a COVID-19 motivated modification of a related party arrangement is a worthwhile exercise for taxpayers and their advisors to commence as soon as possible. It is also possible that tax authorities may consider guidance suggesting that third parties at arm’s length would have renegotiated contracts (e.g., paragraph 10 stating “…it is important to consider any changes in the economically relevant characteristics, including the terms and conditions of the agreement, and whether at arm’s length, unrelated parties would have tried to renegotiate those terms and conditions”) as an invitation to alter the pricing of existing long-term intercompany agreements. One area where this could be applied is in intercompany financing arrangements, where the low interest rate environment in the wake of the pandemic might lead tax authorities to claim that borrowers at arm’s length would have refinanced, and in turn make an adjustment to intercompany interest rates on existing agreements. The language in the intercompany loan agreement is important here. For example, if the agreement allows the borrower to prepay the intercompany loan anytime without penalty, it could be hard to argue that a borrower at arm’s length would not have taken advantage of that, but if the agreement does not allow prepayment or includes a prohibitive make-whole penalty, then it can be argued the borrower would have had no choice but to adhere to the existing contract.
It is unclear how tax authorities will react to this as no public statements have been made. Some jurisdictions require filing intercompany agreements with tax authorities, especially in the commodity areas. Any changes to those agreements will be highly visible.
Arrangements between related parties can be modified, or rather should be modified, if it can be shown that independent parties would have done so in the same or similar circumstances. For example, if an independent distributor would seek discounts on the supply of goods, and an independent supplier would agree to such discounts to ensure the continued survival of the distributor and hence the entire value chain, then such discount could or should be granted between related party suppliers and distributors. This is driven by the facts and circumstances in each case. For taxpayers with an APA, this could potentially result in a breach of the critical assumptions in the APA. In those circumstances, taxpayers are encouraged to proactively engage with the ATO as soon as they become aware that a breach of the critical assumptions in the APA has occurred or is likely to occur. This could include:
- Business as usual;
- Renegotiating the APA; or
- Suspending or modifying the APA for a set period.
Modifications to existing arrangements will likely be respected if it can be demonstrated that uncontrolled parties would likely agree to the modified terms and conditions under similar circumstances. Further, such modifications should align to the actual facts and circumstances involving the controlled transaction.
Before any modifications to agreements are made, taxpayers and their advisors should review the terms outlined in the existing agreement to determine whether there is already some flexibility to make certain changes.
The basic requirement of Canada’s transfer pricing regime is that the terms and conditions of a related-party transaction should be the same as would have applied if the same parties had been dealing at arm’s length. There are no specific rules that require consistency of policy or outcome over time, and modifying existing arrangements should be supportable so long as it can be demonstrated that arm’s-length parties would have done so – which, of course, depends on the facts and circumstances.
How Should Operational or Exceptional Costs Arising From Covid-19 Be Allocated Between Related Parties?
The IRS has not specifically provided any guidance on how costs arising from a global pandemic or otherwise should be handled between parties to a controlled transaction. There are two areas to consider.
The first is an analysis of the relevant functions, assets and risks related to the transaction, which should provide guidance on which related party should ultimately bear these non-recurring costs:
- Which party assumes the risks, controls the decisions and ultimately benefits from the activities performed in connection with these costs
- Evidence of how third parties would share these types of costs under similar circumstances
- What the intercompany agreement states related to the type of costs included in the intercompany charge or details related to the responsibilities of each party
Any such non-recurring costs should be identified, and sufficient support be available to justify which party(ies) should ultimately bear these costs in advance of preparing the transfer pricing documentation for the year.
The second area that needs to be considered is how similar types of costs are reflected in comparables. For example, the taxpayer may conclude that certain non-recurring costs related to the pandemic will be included in the cost base of the service fee. However, the comparables used to support the profit markup charged as part of the service fee may categorize these costs below the operating profit line. Adjustments to the comparable companies would need to be made to include such costs, which could ultimately lower the range of results used to support the service fee.
Canada’s regime offers no specific guidance on this issue yet. Where Canadian companies have excluded COVID-19 related costs from cost-based transfer pricing methods on the position that they are “exceptional”, “non-recurring” or “extraordinary” (OECD, paragraph 36), there will be even more pressure on those companies to retain the benefits of government assistance as an offset. Some taxpayers may take a broad view on the types of costs that qualify as exceptional/non-recurring/extraordinary and include efficiency-related costs such as labor variances (e.g., where a manufacturer’s COVID-19 safety measures temporarily reduced labor productivity per unit).
It is unclear how tax authorities will react to the treatment of exceptional costs related to COVID-19. No major pronunciation has been made. However, experience shows that tax authorities reject the allocation of such costs initially, but courts may side with the taxpayers in most cases. There is likely to be significant controversy on this matter.
Reference should first be made to the legal agreement to ascertain what has been agreed between the parties. This can provide guidance but is not determinative, as the real question is what would third parties have agreed to in similar circumstances? Again, this depends on the facts and circumstances.
Intercompany agreements have never assumed such importance. Allocation of COVID-19-related costs should follow the allocation of risks, which should be set out in intercompany agreements. The OECD guidance basically says: “do what the intercompany agreements tell you to do and what third parties would do/are doing in comparable situations.” Which doesn’t provide very helpful guidance. The guidance not to treat recurring costs as exceptional or non-recurring costs seems equally self-evident. It would be more helpful to have guidance on where to find the all-important evidence of how independent parties allocate exceptional non-recurring costs between them. In general, it can be expected that taxpayers with vaguely written or non-existent intercompany agreements will struggle to defend their cost allocation positions to their tax authorities.
Does the Receipt of Government Assistance Affect the Price of Controlled Transactions and Allocation of Risk Within Those Transactions? (or Perhaps How Should It)
Given the CRA’s administrative policy of following OECD guidance to the greatest extent that local laws allow, and that the Canadian legal transfer pricing regime is very flexible on how arm’s-length prices are determined (without detailed regulations that dictate methodology, data uses, etc.), it is likely that the only area in which the CRA’s response may diverge materially from the OECD’s suggestions from December could be in the area of government assistance.
Nothing in Canadian law dictates how government assistance should affect transfer prices, but as a matter of administrative policy the CRA has issued TPM-17 which argues that when using a cost-based method, “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.” This policy pre-dates the COVID-19 crisis and was originally intended for items like scientific research and experimental development tax credits. This summer, a CRA official confirmed that the same policy would be applied to programs related to the COVID-19 crisis, such as the Canada Emergency Wage Subsidy (CEWS)—and that the CRA expected that no reliable data would be available to counter their default policy of leaving the government assistance benefit with the local Canadian taxpayer. The full text of that response is below:
“The purpose of the TPM-17 is to ensure that when a cost-based method is being used as a proxy for observing an arm’s length price for goods or services in a related party cross border transaction, any government assistance is properly accounted for. Where government assistance has been received that assists in paying some of the costs of either labor or capital items (depending on the relevant cost base) that assistance will not usually be used to reduce the cost base for the purpose of determining the value provided by the party in the transaction.
The COVID-19 related government assistance is a temporary emergency measure implemented to support Canadians and Canadian businesses facing hardship as a result of the global COVID-19 outbreak. Given the nature of the assistance and circumstances under which it is provided, it is unlikely that market evidence would exist to support sharing this type of emergency assistance. Accordingly, COVID-19 related government assistance would be expected to form part of the tax base of the recipients.”
It is unclear how tax authorities will treat COVID related government subsidies in the context of transfer pricing. No major pronunciation has been made. Due to the difficult economic situation generated by COVID-19, governments focused their assistance to people and, in general, provided minimal assistance to corporations. For example, it is widely known that Mexico provided no aid to corporations, while other countries mainly focused on payroll assistance.
We can expect tax authorities to challenge taxpayers they suspect have passed on government assistance via transfer pricing.
The ATO is one of the few tax authorities to issue specific guidance on the treatment of COVID-19 related government support payments in transfer pricing arrangements.
The ATO expects that the benefit of the payments received will be retained by Australian entities and should not result in a change to the transfer price. For example, in the case of a service provider paid on a cost-plus basis, that cost should be gross cost before the government subsidy.
As a result, such government assistance would or should have no impact on the pricing of controlled transactions.
The IRS has not broadly provided guidance related to this issue, but it has been mentioned by APMA to consider in APA filings that involve the 2020 tax year. At minimum, taxpayers must be aware of the impact any government assistance has had on the financial performance of the parties to the controlled transaction. The recommended approach would also likely be similar to how extraordinary costs should be addressed. The taxpayer should review the intercompany agreement terms to determine what (if anything) is said regarding government incentives. Second, evidence should be gathered that identifies instances where unrelated parties address how government assistance credits are treated.
1Transfer Pricing Documentation Frequently Asked Questions, Internal Revenue Services – Treaty and Transfer Pricing Operations, April 14, 2020. https://www.irs.gov/businesses/international-businesses/transfer-pricing-documentation-frequently-asked-questions-faqs