BEPS 5 Years Later: Still No Convergence on Transfer Pricing
The 2015 base erosion and profit-shifting report on transfer pricing modernized the OECD’s guidance in many key technical respects, but its success as the basis for a globally consistent approach to transfer pricing is doubtful.
Consistent with the BEPS action plan, one of the stated goals of which was “to assure that transfer pricing outcomes are in line with value creation,” the actions 8-10 report’s revisions to the OECD transfer pricing guidelines focused largely on intangible property transactions and the allocation of risk among multinational group members. Unlike the 2010 OECD guidelines, which did not address risk in much detail, the 2015 report established a process for assessing the contractual allocation of risk — applicable also to the risk of intangible development — and determining which entity or entities are entitled to the upside or downside associated with that risk. The revisions were incorporated into the 2017 edition of the guidelines.
Many of the most significant and controversial revisions relate to the BEPS report’s introduction of control over risk as a critical factor in assessing the allocation of risk and the returns from intangibles. Under the revised guidance of chapter 1.D, a transfer pricing analysis must begin with the accurate delineation of the taxpayer’s transaction and a review of the relevant intercompany contracts. Provided that the contracts accurately reflect the economic substance of the transaction, the contractual allocation of risk among group members should be respected as long as the contractual risk bearer exercises control over the relevant risk and has the financial capacity to bear it.
The actions 8-10 report’s new risk analysis framework is generally deferential to multinational groups’ intercompany contracts, and it does not prevent the sharing of control with other group entities or the outsourcing of “day-to-day risk mitigation” functions to related or unrelated parties. However, it does require that the entity claiming any upside returns has the capability and authority to decide which risks to take on and how to manage them, and that it actually exercises its decision-making authority. If a contractual risk bearer does not satisfy this control threshold, the guidance provides that tax authorities may reallocate the risk — along with the upside and downside consequences of that risk — to the party that does exercise such control.
Although the new control-based analysis introduced in chapter 1.D can in principle apply to any economically relevant risk, the actions 8-10 report explicitly incorporates the analysis into its revised chapter on intangibles. Although the new chapter 6 of the OECD transfer pricing guidelines notes that any contribution to the value of an intangible must be compensated on arm’s-length terms, it provides that an entity must exercise control over the relevant risks to be entitled to the upside returns associated with the intangible. The guidance says the necessary control typically relates to the risks associated with “development, enhancement, maintenance, protection, and exploitation” of the intangible, language that gave rise to the “DEMPE” acronym that some Working Party 6 delegates believe has taken on an unintended life of its own.
Together, the actions 8-10 report’s changes restrict the returns that a typical “cash box” — a group entity that does nothing more than fund intangible development that other group members perform and control — can receive. Pure cash boxes are entitled to “no more than a risk-free return” on their invested capital, while entities that control investment risk without controlling any specific operational risks should receive “an appropriate risk-adjusted return.” The 2015 report offered little guidance on how these returns should be determined, however, and drew criticism for elevating the importance of functions and disregarding contributions of assets.
Other important additions to chapter 6 address the different categories of intangible property, recognize that intangible valuation methods — like the U.S. cost-sharing regulations’ income method — may be more reliable than traditional transfer pricing methods in some cases, and introduce a special approach to transfers of hard-to-value intangibles that broadly mirrors the U.S. periodic adjustment rules. The report also amended the chapter on cost contribution arrangements — a term that includes the U.S. regulations’ concept of a cost-sharing arrangement — to align with the revised chapter 6.
Clarifications and Continuity
Despite these changes, the BEPS report on actions 8-10 retains most of the core concepts and principles laid out in the 2010 OECD transfer pricing guidelines. As subsequently noted by many involved in drafting the report, the report reaffirms a commitment to the arm’s-length principle as the fundamental concept underlying article 9 of the OECD model tax convention.
Any deviations from a theoretically pure version of the arm’s-length principle, including the framework for allocating risk, were part of a “pragmatic fudge” intended to preserve a workable system and were far less radical than they could have been, according to Andrew Hickman, head of the OECD’s transfer pricing unit during the report’s drafting. Similarly, Michael McDonald of EY, formerly a U.S. Treasury Department financial economist and a chair of Working Party 6, has stressed that the actions 8-10 report — read in its entirety — was a clarification of existing principles rather than a radical shift in approach. According to McDonald, the BEPS project’s theme of aligning profit with “value creation” is simply another way of stating the arm’s-length principle, and only a selective interpretation of the report’s language supports an exclusive emphasis on functions and a complete disregard of intercompany contracts.
The interpretation of the actions 8-10 report as a continuation of, and not a departure from, the existing system has been echoed by some of the report’s critics from civil society and academia. One such critic, Reuven Avi-Yonah of the University of Michigan School of Law, told Tax Notes that the BEPS project failed to make any meaningful improvements to the transfer pricing rules. “I think that was a failure, as even the OECD recognizes to some extent since they moved beyond it in pillar 1” of the OECD’s ongoing project to update the international taxable nexus and profit allocation rules, Avi-Yonah said.
According to McDonald, however, the generally conservative approach reflected in the text of the actions 8-10 report and later incorporated into the OECD transfer pricing guidelines was sound. McDonald said the changes succeeded in providing greater clarity on many key points and tie all of the guidelines’ chapters together within an internally consistent analytical framework. “I am happy with — indeed proud of — the revised language incorporated into the OECD transfer pricing guidelines as a result of the BEPS project,” McDonald said.
“The revisions clarified that the taxpayer’s transaction should first be properly understood, and only then properly priced; reiterated that contractual terms provide the important starting point for understanding the transaction, while also providing important stress tests of contractual terms based on an evaluation of substance and control over risks; [and] provided a systematic way of evaluating risks,” McDonald said. They also “provided a clearer discussion of intangibles that effectively presaged the definitional modification to intangibles in the [Tax Cuts and Jobs Act]; provided substantially enhanced valuation guidance that directly address common BEPS opportunities, including the income method and guidance on enhancements to intangibles legally owned by others, as well as introducing a ‘commensurate with income’ concept that is tied to information asymmetry; [and] provided clearer guidance on cost contributions arrangements,” he noted.
As a U.S. delegate, McDonald further added that the improvements to OECD guidance did not come at the cost of any significant friction with U.S. law, noting that “I was convinced then, and remain convinced, that the OECD guidance sits quite comfortably with the [section] 482 regulations.”
Same Text, Different Interpretations
However, McDonald’s assessment of countries’ implementation of the language used in the actions 8-10 report is a different matter. “I am significantly less happy with — indeed alarmed by — some interpretations and applications of the guidance in practice. These include the use of the guidance to blur the fundamental principle that risks and functions can be separated under the arm’s-length principle [and] the use of the guidance, in particular the guidance on risk, to gratuitously recharacterize legitimate transactions,” McDonald said.
The actions 8-10 report ultimately made little change to the 2010 OECD guidelines’ strict standard for a tax agency’s recharacterization, or “nonrecognition,” of the structure of the taxpayer’s transaction and its substitution with a transaction structure that the agency considers more appropriate. The 2015 report instructs tax authorities to limit their review to the pricing of the taxpayer’s transaction, and to disregard the structure of the transaction — regardless of whether the transaction would take place between unrelated parties — only if it is inconsistent with the economic substance or is commercially irrational. The principle generally aligns with the United States’ approach to recharacterization, as reflected in its negotiation position during the BEPS project and in reg. section 1.482-1(f)(2)(ii).
But whatever consensus may have been reached on recharacterization by Working Party 6 delegates in 2015 seems to have been short-lived. Australia’s tax agency in particular has made it clear that it has every intention of applying its “reconstruction powers” to transactions that unrelated parties would not undertake. Poland has expressed a similar position, and some U.K. officials have in effect described the controversial diverted profits tax as a way of circumventing the high recharacterization threshold.
McDonald also expressed concern regarding countries’ misapplication of the concept of DEMPE functions, criticizing “the elevation of 'DEMPE' — a fairly obvious valuation principle simply stating that contributions to intangibles should be appropriately remunerated — to a powerful acronym and independent source of guidance.” He also objected to what he sees as the inappropriate “conflation of DEMPE functions and control over risks [and] the related improper conflation [of] DEMPE functions and entitlement to residual returns.”
Some jurisdictions, like Hong Kong, have gone as far as to incorporate the acronym, or the terms it represents, into their domestic legislation. And Mexican and U.S. officials have acknowledged that disagreement on the relevance of DEMPE functions to the selection of transfer pricing method has complicated mutual agreement procedure negotiations between the two countries’ competent authorities. Although the language in the actions 8-10 report does not specify the form of compensation that a party contributes — for example, by performing DEMPE functions — to managing a risk that it does not assume, Mexican officials believe the guidance entitles that party to share in the upside risk as determined using a profit split.
And regarding profit splits, the durability of the consensus behind the OECD’s 2018 profit-split implementation guidance, which Working Party 6 delegates struggled with long after the actions 8-10 report’s release, was discounted by U.S. officials almost as soon as it was published. Christopher Bello, then-branch 6 chief for the IRS Office of Associate Chief Counsel (International) and another U.S. delegate to Working Party 6, said in 2018 that the profit-split guidance is likely to encounter the same problems noted by McDonald regarding the actions 8-10 report.
“When you compare how the guidance turned out with how many of the people in the room wanted it to turn out, I think it actually turned out pretty good,” Bello had said. “There's a difference between [asking], ‘What does the guidance actually say?’ — and I think that what it actually says is right — compared to [asking], ‘What are some countries going to claim that it says or want to think that it says?’”
And while it may take time for any pointed interpretive differences in the recently finalized guidance on financial transactions to become apparent, the stark disagreements that delayed agreement on finalized guidance suggest that similar problems may arise regarding the returns to entities that control investment, but not operational, risk. Although the financial transactions guidance does address the concept of a “risk-adjusted return,” McDonald generally criticized “the failure to properly recognize the appropriate ex post returns to financial investment risk-taking.”
Thus, despite all of Working Party 6’s technical achievements in preparing the 2015 actions 8-10 report and releasing a steady flow of post-2015 transfer pricing guidance on complex issues, the prospect of countries consistently acting in accordance with this guidance is doubtful. The dissatisfaction that fueled the original BEPS actions 8-10 reemerged in the form of pillar 1 of the OECD’s proposed international tax reforms, which would explicitly replace the arm’s-length principle — at least for a portion of deemed residual profit (amount A) — with a formulaic variant of the residual profit-split method for many of the world’s largest and most profitable multinationals.
But according to Brian Jenn McDermott Will & Emery, a former Treasury deputy international tax counsel and another former U.S. delegate to Working Party 6, it may be unrealistic to expect any more under actions 8-10 in light of political constraints. The actions 8-10 report went as far as the United States was willing and able to go under the arm’s-length principle, prompting other countries to look elsewhere for solutions, he said.
“Countries have been looking for — and in some case adopting — new tools outside of the transfer pricing guidelines and the arm’s-length principle to address what they see as undesirable outcomes from a policy perspective," Jenn said, pointing to pillar 1 and pillar 2, digital services taxes, diverted profits taxes, Australia’s multinational antiavoidance law, and the U.K.’s offshore receipts tax. “The action has moved elsewhere. That is to say, for the foreseeable future — and setting aside how amount B [under pillar 1] might be implemented — I don’t anticipate important new developments in the transfer pricing guidelines.”