Matt Andrew is head of the OECD tax treaty, transfer pricing, and financial transactions unit and Richard S. Collier is a senior tax adviser with the OECD Centre for Tax Policy and Administration.
The authors thank Ian Dykes for comments on an earlier version of this article.
In this article, the authors address the implications of the coronavirus pandemic for the arm’s-length principle in the context of the OECD’s inclusive framework.
The impact of the COVID-19 pandemic is ubiquitous. Among its many other effects, it creates challenges for the international tax system. That includes pressure on tax treaty rules (in particular, how treaties operate when the normal mobility of people is suddenly curtailed), a matter on which the OECD has issued guidance.1 Further, given that for many companies, the coronavirus has increased the importance of being able to conduct business digitally, the crisis also exerts extra pressure on the need for a solution to the taxation of the digital economy.
Those challenges also have implications for the arm’s-length principle. That is because in some cases the virus has created market conditions or events, and led to responses to those conditions and events, that are not directly addressed by existing transfer pricing guidance (specifically, the OECD Transfer Pricing Guidelines for Multinational Enterprises and Tax Administrations). That raises the question whether further guidance addressing some or all of those concerns might be useful. Unsurprisingly, the OECD has already received several requests from states and businesses to explore what additional guidance on the arm’s-length principle might be possible.
This article refers to the various questions of how the arm’s-length principle should operate in these unusual conditions created by COVID-19. Those operational issues concern the period when market conditions are in a relevant sense altered (which may vary by industry) and relate primarily to the uncertainty of how the arm’s-length principle should be applied in this period.
However, it may also be asked whether the pandemic has a wider significance for the arm’s-length principle. That might be the case if the period of disruption caused by the virus offers some lessons or insights that could alter our understanding of the principle. If that is the case, the impact of those lessons might be more widely relevant and not necessarily restricted to the period of market disruption caused by the virus.
Operational Issues Created by COVID-19
In broad terms, the transfer pricing rules based on the arm’s-length principle require a standard of income allocation that reflects the financial results that would accrue in comparable transactions entered into by third parties.2 Under the existing income allocation rules, the arm’s-length standard is of course equally relevant in the time of a financial crisis such as that caused by COVID-19. However, that is not necessarily synonymous with applying the transfer pricing guidelines (and relevant domestic law based on the arm’s-length standard) as if nothing had changed. That is because the existing guidance may be regarded as (at least partly) inadequate or uncertain as applied during this financial crisis. As noted above, clarification of the arm’s-length principle may also be needed for responses to the virus, such as how to treat extensive government support, including subsidies and loans.
It is unsurprising that the guidance on the arm’s-length principle, which is generally geared to an economic environment with normal market conditions, may need further clarification in light of COVID-19. The pandemic’s extreme economic effects may mean that guidance ceases to deliver a viable proxy for third-party outcomes, or there is uncertainty in whether it does so. That might happen when the approach dictated by the arm’s-length guidance seems to vary materially from what third parties can actually be observed to be doing, or when factors have emerged that cannot readily be accommodated under the guidance (as with government support).3
Numerous situations have been suggested as raising the possible need for additional guidance on the application of the arm’s-length principle. It may prove useful to think about those situations in terms of a more manageable number of separate categories. For example, one category might comprise all issues that arise as a direct consequence of the sharp and immediate change in market conditions, particularly in the collapse of profitability. That would include such issues as whether and to what extent limited risk distributors and other fixed-margin operators should be insulated from the effects of any sharp downturn in the profits of a business; the crisis’s impact on related-party contracts, including when there are and are not force majeure clauses; whether payment of royalties, service fees, and the like should still be made from loss-making entities; the treatment of various business costs such as restructuring and shutdown costs and stranded work force costs; and any effects on intragroup financing costs. That is obviously a broad category; any attempt to provide guidance for it is unlikely to be granular enough to address every potential issue. Rather, a more feasible approach would be to identify any general themes or observations that might then be considered when addressing individual cases on their own facts and circumstances.
A second category might address statistical and data-based questions, such as the impact of the crisis on the use of traditional benchmarking sets in any particular business, and specifically the questions of if, when, and how to adjust for material changes in commercial circumstances.4 The issues may often be complicated by the fact that sources of data used generally are historic, usually being at least one year out of date.
The treatment of government grants and subsidies that have been made available as a result of COVID-19 is another important area. Those support schemes may raise many questions regarding how they should properly be treated for arm’s-length principle purposes. For example, should the economic benefit of any government support be retained in the jurisdiction where it has been made available, or are there circumstances in which that benefit may be shared with related-party entities in other states? That question might be important when an entity receiving government support (such as wage subsidies) charges for its services to affiliates on a cost-plus basis: Should the government support be taken into account (thereby reducing the relevant cost base) or ignored?
The pandemic’s impact on agreements between taxpayers and tax authorities is also raising questions. For example, should advance pricing agreements be modified or suspended because critical APA assumptions are breached in the current environment? And to what degree should tax authorities and taxpayers demonstrate flexibility regarding critical assumption breaches? That flexibility could have a basis in the changes to market conditions created by COVID-19, given that those conditions may lead to variations in agreements between third parties. Issues of a collateral nature may also need to be considered for other agreements or arrangements if transfer prices are adjusted to respond to the pandemic. An example of that could be in pricing adjustments that affect customs duties.
Finally, guidance on best practices for administrative topics might also be useful. That could include the appropriate steps in documenting any transfer pricing changes made as a result of the COVID-19 crisis, including recording the nature of the relevant event, issue, or risk causing the change; identifying the rationale for any adjustment made; and evidencing the results or consistency of any revised approach.
It seems likely that other categories or questions will emerge as a result of the coronavirus pandemic.5 It is also likely that some of the categories will overlap. For example, in some countries, there may be circumstances in which government financial support is available to a company only if it has experienced material reductions in the performance of its financial metrics. If a company derives its revenues and profits solely from transactions with related parties, that could raise difficult questions of whether and how the adverse impact of market conditions should properly be reflected in adjustments to the company’s related-party contracts for government support purposes.
While guidance on those coronavirus-related issues may be limited, it is not by any means totally lacking. In some areas, there is useful commentary that might be applied by analogy. The OECD transfer pricing guidelines, for example, provide both a broad framework and individual sections that are relevant to some specific concerns raised by the virus. In many places, the guidelines address the practical difficulties that may arise in applying the normal approach to transfer pricing, including losses that may arise as a result of unfavorable economic conditions. The guidelines discuss the effect of extreme results on comparability considerations. They also address the effect of future unpredictable events on valuation, emphasizing the importance of resolving the position by reference to what independent enterprises would have done in comparable circumstances to take account of any valuation uncertainty in the pricing of a transaction. The guidelines suggest applying the guidance by analogy when pricing intangibles for which valuation is highly uncertain at the time of the transaction, and they contemplate situations in which fundamental changes in value would have led to a renegotiation of the relevant contract. The principles underlying the discussion seem clearly relevant to the COVID-19 situation, as is the more general approach of being guided by the axiom that independent parties act based on a consideration of the options realistically available to them.
The guidelines devote much discussion to the use of benchmarking data. That includes the preference for current period data because it “is expected to be the most reliable information to use in a comparability analysis as it reflects how independent parties have behaved in an economic environment that is the same as the economic environment of the taxpayer’s controlled transaction.” The guidelines also note that historic data may be used together with “information on economic and market changes” that may have occurred after the period for which data is available. The discussion is of course not designed specifically for the COVID-19 crisis, but the overall goal of the exercise is made clear.
The discussion of business restructurings in Chapter IX of the guidelines may also be relevant in many respects to the restructurings and reorganizations that have been caused by the COVID-19 crisis — for example, business shutdowns, redeployments of resources, costs incurred in restructuring, indemnifications, and changes to supply chains. Further, the discussion of the effect of government policies in Chapter I and of location savings and other local market features in chapters I and IX may be of some assistance in addressing the treatment of government support. The guidelines also imply that APAs should be revised or canceled when “uncontrolled economic circumstances . . . critically affect the reliability of the methodology in a manner that independent enterprises would consider significant for purposes of their transfer pricing.”
In addition to considering the relevance of the transfer pricing guidelines, it might also prove helpful to take account of country guidance and stated country positions on issues raised by the pandemic. Some guidance was produced after the 2008 financial crisis, and tax authorities made some changes in practice.6 For example, the Chinese tax authorities issued guidance that single-function, low-risk entities should not bear crisis risks and so should not report losses.7 An example of a change in tax administration practice comes from the Australian Taxation Office, which contacted taxpayers during the 2008 crisis to discuss documentation and other issues to support the arm’s-length nature of related-party transactions. Other tax authorities also sought to adjust APA proposals to take account of the crisis’s effect on business sectors or individual companies.
Specific guidance from individual tax authorities may go beyond what is available in the transfer pricing guidelines. For example, the Canada Revenue Agency has issued guidance on the impact of government assistance on transfer pricing (TPM-17). The crux of the guidance is that the benefit of any financial assistance from government agencies or programs should not normally be shared outside the state unless there is clear evidence that third parties would treat that support differently in the specific facts and circumstances.
Country guidance and practice could be relevant because it may apply in specific situations in the countries concerned but it may also suggest some options for guidance to be more widely developed and applied.
It follows that existing guidance may be relevant to arm’s-length questions raised by COVID-19 and that there may well be related issues that have been identified but for which further guidance is considered unnecessary. However, guidance will clearly be helpful on material questions raised by the crisis — that is, those that are financially important and of common relevance to taxpayers, and for which there is significant uncertainty. Examples meeting those criteria may include the treatment of government support and 2020 data issues.
Issues of Possibly Wider Relevance
The operational issues discussed above involve the immediate task of managing the application of the transfer pricing rules during the COVID-19 pandemic. However, it seems appropriate to also ask whether concerns are arising that have wider implications for the arm’s-length principle and go beyond managing its application during disruption caused by
The crisis and the plethora of challenges it brings for the arm’s-length principle arguably confirm that while the OECD transfer pricing guidelines take some account of financial crises, guidance on the arm’s-length principle has largely been developed based on the behavior of third parties in stable and profitable markets. The economic return of each party is based on its own contribution to the relevant business cycle, and the parties are assumed to be solvent and able to meet their obligations. However, when markets are in crisis, those assumptions may sometimes be less warranted because third-party behavior may become more extreme as a result of those parties trying to protect their capital, reduce costs and risk, and so forth.
A good example of that is the status of a contract between related parties. The transfer pricing guidelines tend to respect a contract as long as it is not at variance with the parties’ conduct and is not an arrangement that would not have been entered into by unrelated parties behaving in a commercially rational manner.8 The terms of the contract are otherwise generally considered the right starting point for determining the transaction to be identified and applying the appropriate transfer pricing analysis (although there are some cases in which a transaction might need to be renegotiated9).
In the COVID-19 crisis, however, there are cases in which third parties have altered or ignored the contractual terms. In the abrogation of a contract, the behavior of some third parties, particularly when taken unilaterally, has been seen to be more like a receiver in bankruptcy — for example, in cutting overhead costs; organizing a fire sale of obsolete stock; reducing stock holdings; and terminating all supplier contracts considered to be non-core to survival.10
Further, in response to the COVID-19 crisis, third parties might exceed the strict legal obligations required by their contracts. That has been seen in particular in companies with integrated business models involving third parties (such as third-party auto distributors or independent franchisees), as well as when a manufacturer or principal (for example, a car manufacturer or franchisor) goes well beyond its contractual obligations in providing financial assistance to the other party, presumably with a view to preserving its own financial viability.11
In those examples, the third party’s actions imply different behavior than is normally expected based solely on the contract terms. That in turn suggests that in a crisis, the conventional approach to respecting contracts, assessing transfer prices, and allocating profit may need to be reconsidered, and that the range of what is accepted as constituting arm’s length should be widened.12
The importance of those issues may well turn on the significance that is attributed to the crisis. Is it a unique, unprecedented, never-to-be-repeated event? Or is it instead the latest in a long line of economic crises? The more the coronavirus pandemic is seen as of relevance to recurring financial events — whether on a global, regional, or local basis — the more its lessons are potentially of more enduring interest and relevance. It might be argued that the COVID-19 crisis is so unique that its wider systemic significance is severely limited, meaning that current priorities from the perspective of the arm’s-length principle should be limited to simply addressing the immediate operational problems and then moving on as well as possible with the system and policy objectives that prevailed pre-crisis.
However, the primary current relevance of the pandemic is its impact as a financial shock or economic crisis. In that regard, any claim that it is unique or unprecedented does not seem well founded.13 True, the depth, reach, and speed of the financial shock may well eclipse previous ones, but the sad fact is that history is littered with financial crises.14 It has also been just over 10 years since the last “unprecedented” global financial crises.15
The issues discussed above may also prove of import when an entity is in crisis even though the rest of the world is not. They may be relevant to the actions of the entity concerned — or, if a party exceeds a contract’s terms to support an entity in crisis, to the actions of the entity providing the financial support — again, potentially with the same implications for what might fall in the range of arm’s-length behavior.
There may also be lessons relevant for the arm’s-length principle in the absence of a general economic crisis or an entity-specific crisis. For example, one suggestion is that the guidance in the transfer pricing guidelines on risk may need to be re-assessed in the light of what has been seen in the COVID-19 crisis. This might be on the basis that the experience from the COVID-19 crisis suggests that risk is shared across the entities in an MNE group in a more significant manner than is suggested by the guidance on risk in the transfer pricing guidelines.
The OECD is investigating arm’s-length-related COVID-19 issues and discussing with its inclusive framework and Working Party 6 how those issues may best be addressed. The goal is to secure the broadest possible support for the initiative, and delegates and companies are being canvassed regarding potential topics for priority.
There will likely be different perspectives on and priorities for developing responses — so far, the tax response to COVID-19 has been almost exclusively unilateral. Further, while companies might want fast answers, tax authorities need time to explore the questions, and they might want to launch a public consultation to do so.
The objectives of any future guidance would of course be to provide more certainty and consistency in the application of the arm’s-length principle, ideally thereby limiting the scope for dispute and double taxation.
1 See OECD, “OECD Secretariat Analysis of Tax Treaties and the Impact of the COVID-19 Crisis” (Apr. 3, 2020). See also Pascal Saint-Amans, “OECD Issues Recommendations on Implications of the
COVID-19 Crisis on Cross-Border Workers and Other Related Cross-Border Matters” (Apr. 3, 2020). The guidance addresses the implications for tax treaties of the abrupt and dramatic change in where people work. It covers the relevant impact on the creation of permanent establishments, the residence of companies, the position of cross-border workers, and the residence status of individuals. The OECD has also created an extensive database of the economic support measures that have been taken so far during the COVID-19 crisis. See OECD, “Tackling Coronavirus (COVID-19): Contribution to a Global Effort” (last updated May 29, 2020).
2 The focus of this discussion is predominantly the transfer pricing rule in article 9 of the OECD model tax treaty. The arm’s-length standard is also relevant to the profit attribution rule in model article 7, although there is less consensus on the application of that article. The definitive statement of the arm’s-length principle is in model article 9(1).
3 It is debatable whether COVID-19-related government support is a new issue in the context of prevailing transfer pricing practices. Many multinationals will have encountered the question of how to accommodate government incentives and subsidies in their transfer pricing policies and compliance practices. Further, as discussed in this article, the topic has been addressed in some tax authority guidance. However, what is clearly different in the COVID-19 crisis is the enormous scale of government support, giving the topic a much broader relevance than it previously had.
4 Database analyses are used widely and in different contexts — for example, to support a profit markup on designated costs in a cost-plus method or a net margin for an appropriate base, such as costs, sales, and assets in the application of a transactional net margin method.
5 One possible category of issues involves the identification and treatment of pro bono costs, such as those incurred in adapting production capabilities to manufacture or transport personal protective equipment. While the treatment of those costs might often have no implications for transfer pricing because they do not relate to any business undertaking and are therefore to be ignored, there may be some cases in which the question arises whether costs incurred are to be treated as genuinely pro bono or rather as a business cost by virtue of being incurred to enhance the company’s branding and standing, which could make them subject to different treatment under the arm’s-length principle.
6 As might be expected, there are various relevant parallels between the COVID-19 crisis and the 2008 global financial crisis. Both involve profound economic shock and, in response, extensive public or government support for businesses (although the sectors most affected differ). In both cases, the impact varies from jurisdiction to jurisdiction, and tax measures are seen as a central response. Similarly, both cases resulted in prodigious talks of possible tax effects and required tax policies and priorities. Although yet to be seen in full for the COVID-19 crisis, it is likely that in due course there will be extensive discussion of new tax measures that will be designed to pay for the crisis, as occurred with the global financial crisis. A large part of the tax response to the global financial crisis recognized the relevance of wider systemic issues. Several themes that emerged from that response, including calls for responses based on international cooperation, concerns about the strength of the system and how to protect it against future shocks, and the linkage of any response with an antiavoidance agenda, are relevant now.
7 See Abe Zhao, “Managing Transfer Pricing Risks by Multinational Companies in China,” 6 Int’l Transfer Pricing J. 461, 463 (Nov./Dec. 2017).
8 A transaction may be disregarded on that basis only if it is impossible to determine the appropriate price that would be acceptable to both parties, taking account of the options realistically available to them when entering into the transaction.
9 See, e.g., transfer pricing guidelines, para. 3.73.
10 The same may be seen in commercial and contract law, in which contracts and risk ratings are valid and govern third-party commercial dealings until a crisis in the business hits (such as when a party to a contract become insolvent). Then those laws and contracts are overtaken by bankruptcy and insolvency laws, which govern the situation.
11 There is some parallel between that situation and the one in which an entity incurs losses to enter new markets, increase its share of an existing market, introduce new products or services, or discourage potential competitors. See para. 1.131 of the transfer pricing guidelines.
12 One example might be the position of a limited-risk distributor. It might be argued that a commercially rational third party would not agree to limit the risk of another third party in a crisis market when there is no expectation of profit and each party seeks to limit its own capital, risk, and costs to remain solvent.
13 The status of COVID-19 as a pandemic is also not unique, with plagues and epidemics having ravaged humanity throughout its existence. In just the last 10 years, we have seen the H1N1 swine flu pandemic, the Ebola epidemic, and the Zika virus epidemic.
14 For example, see Carmen Reinhart and Kenneth Rogoff, This Time Is Different: Eight Centuries of Financial Folly (2009), which charts eight centuries of financial crises.
15 The 2008 crisis is often discussed as a wholly unprecedented event — but against the backdrop of other financial crises, that view seems incorrect. See Michael Bordo and John Landon-Lane, “The Global Financial Crisis of 2007-08: Is It Unprecedented?” NBER Working Paper No. 16589 (Dec. 2010).