In 2013, the OECD announced its Base Erosion and Profit Shifting (BEPS) project. Action 1 of the BEPS initiative related to the need to address the tax challenges of the digital economy, and although a final report on Action 1 was published in 2015, the report stressed the need for continuing work. The Inclusive Framework on BEPS, comprising 134 jurisdictions, continued its work in this area, which is currently partitioned into Pillar One, on the allocation of taxing rights and profits, and Pillar Two, on the need for a global minimum tax. On Oct. 9, 2019, the OECD Secretariat released a new proposal under Pillar One.
The Secretariat’s proposal follows on the heels of three competing proposals that were introduced by Inclusive Framework members and were the subject of a public consultation in early 2019. These are the user participation proposal, which would apply to highly digitalized businesses and allocate taxing rights and profits to jurisdictions where those businesses have users; the marketing intangibles proposal, which would apply more broadly and would allocate taxing rights and profits to market jurisdictions on the grounds that marketing intangibles exist there; and the significant economic presence proposal, which would allocate profits to a market jurisdiction using a fractional apportionment approach if a business’s presence in the jurisdiction exceeded a certain threshold.
The Secretariat’s proposal does not represent an agreed upon solution. Rather, it is a move taken by the Secretariat on its own initiative that combines key features of the competing proposals with the aim of achieving consensus. Recent draft and enacted unilateral digital service tax measures by jurisdictions such as France, India, and Italy undermine tax certainty and increase the risk of double taxation. Without consensus in the Inclusive Framework within a relatively short timeframe, it appears likely that more countries will adopt unilateral alternatives, making it increasingly difficult to navigate the global tax landscape.
The new proposal does not apply to all companies, nor does it ring-fence the digital economy and subject only highly digitalized businesses to a separate regime. Instead, it is aimed broadly at consumer-facing businesses, regardless of whether they are digitalized, and the OECD notes that some clarifying rules and carve-outs will likely be needed.
For companies that fall within its ambit, the proposal would adopt new nexus rules based on sales (with a yet-to-be-decided threshold), and it explicitly would not require physical presence in the taxing jurisdiction. This nexus rule would not be incorporated into existing permanent establishment concepts; rather, the OECD contemplates that it would be implemented via an independent treaty provision. The OECD estimates that the proposed approach would significantly increase global tax revenues and would particularly benefit low and middle income economies.
The most detailed, and perhaps the most interesting, aspect of the Secretariat’s proposal is its approach to profit allocation. The OECD explicitly acknowledges that the proposed profit allocation rule “go[es] beyond”—i.e., is inconsistent with—the arm’s-length principle that for decades has constituted the bedrock of the international transfer pricing system. The revolutionary nature of the departure is mitigated by two factors. First, all three prior proposals departed, in greater or larger measure, from the arm’s-length principle, and thus it is no surprise that the Secretariat’s work has continued in this vein—indeed, it appears unlikely that consensus would be possible without some departure from the arm’s-length standard. Second, the deviation is relatively narrowly circumscribed. The profit allocation rule operates as an overlay on the existing arm’s-length transfer pricing system and thus, assuming adequate coordination measures are designed and implemented, may be less disruptive than the prior proposals.
Profit allocation under the proposal begins with a multinational enterprise’s profit, which may (subject to further consideration) be viewed on a business segment rather than an overall basis. The proposal would then remove an amount that approximates the return attributable to the business’s routine functions, which would likely employ assumptions such as industry-specific percentages for the sake of simplicity. The remaining “residual” profit would be split between the profit attributable to market jurisdictions, and the profit attributable to other profit drivers such as intangibles and key risks. This would again rely on a simplified convention, such as an agreed percentage, rather than a transfer pricing analysis. Finally, the portion of the residual that is deemed attributable to market jurisdictions would be split among the market jurisdictions using an allocation key. The proposal refers to the result as Amount A. Although it resembles the residual profit split method in some respects, the Amount A allocation represents an override of the arm’s-length principle in both the determination of the residual profit to be split and the use of formulary apportionment in allocating such profit across jurisdictions.
Yet Amount A is only one part of the story. The Secretariat’s proposal correctly recognizes the need for coordination rules: the assignment of profits to market jurisdictions does not operate in a vacuum, but overlaps with (and may conflict with) existing transfer pricing rules that reward distribution, marketing, and other functions carried out in each jurisdiction, as well as the associated assets and risks. Therefore, the proposal provides for a return for routine marketing and distribution that takes place in a market jurisdiction, and christens such return Amount B. While Amount B reflects the presence of actual functions in a jurisdiction, and is conceived of as broadly consistent with the arm’s-length principle, Amount B would not be determined using a traditional transfer pricing benchmarking analysis, but rather by means of simplified assumptions meant to approximate the value of these routine functions. Although certain statements in the proposal, as well as comments made by the OECD during an October 9 webcast on the proposal, suggest that Amount B is broadly consistent with the arm’s-length standard, Amount B in fact represents a departure from the arm’s-length principle both in its use of simplified assumptions and in its use of direct profit allocation (i.e., allocation of profits of an entity to a jurisdiction) rather than focusing on the pricing of transactions between related entities.
Of course, not all distribution and marketing activities are routine, and taxpayers may perform many activities in jurisdictions that are unrelated to marketing and distribution. The OECD therefore recognizes that it may be necessary in some cases to include an additional Amount C, representing the incremental value of the taxpayer’s activities in a jurisdiction over the routine marketing and distribution return approximated by Amount B. According to the proposal, Amount C should be determined through a traditional transfer pricing analysis, and thus should be largely consistent with the arm’s-length standard (except, e.g., to the extent that the profits available for allocation are reduced by non-arm’s-length allocations under Amount A).
The Secretariat acknowledges that the published proposal represents an architectural framework on which to build, and that it requires significant work before it will be usable. This makes sense—it is more economical to seek consensus on a high-level approach before working out all the details. Nonetheless, it is worth remarking on some areas that may be pain points in the proposed framework.
Transfer pricing is ideally a zero sum game. The existing transfer pricing system aspires to a world in which all profits are taxed exactly once, and neither double taxation nor non-taxation occur. Overlaying a separate, non-arm’s-length system on top of the transfer pricing rules makes achieving that aim difficult where the two systems interact.
First, the second step in determining Amount A involves the removal of profits allocable to routine activities, which the Secretariat suggests would be accomplished using a simplified approach. Because this routine return would not be calculated using a formal transfer pricing analysis, the actual profit attributable to a given business’s routine activities may be higher or lower than the estimate that is employed for this purpose. If the business’s actual routine profit is higher, double taxation could result; if the actual routine profit is lower than the estimate, not all non-routine profit will ultimately be captured in Amount A. The OECD must thus strike a difficult balancing act: providing a low estimate of a routine return would increase the incidence of double taxation, while providing a more generous estimate would lessen the impact of the profit allocation rule itself, to the particular detriment of market jurisdictions without local marketing or distribution operations.
Amount B presents similar issues. A high estimate of the routine return for marketing and distribution is likely to result in increased incidence of double taxation: for businesses whose marketing and distribution functions would earn less than Amount B at arm’s-length, profits that under the transfer pricing rules are allocated to other functions would at the same time be forcibly assigned to the marketing and distribution functions by dint of simplified assumptions. Conversely, a low estimate for Amount B would likely increase the need for reliance on Amount C, which would require ad hoc analyses, undermine the benefit of simplicity that this approach is meant to achieve, and likely create more cross border controversies, since more companies and tax administrations would need to demonstrate they are entitled to returns in excess of Amount B.
Invariably, even a system of well-coordinated rules will result in disputes, and the OECD recognizes that it will be essential to develop more effective means of dispute prevention and resolution. However, it remains to be seen what these will be. The OECD’s request for public comments on this issue seeks input on stakeholders’ experiences with advance pricing agreements (APAs), the OECD’s pilot International Compliance Assurance Programme (ICAP), and mandatory arbitration under the mutual agreement procedure (MAP) articles of bilateral tax treaties. Certainly, it will be critical to ensure that arbitration and APAs are available to address profit allocation under Pillar One, and tools such as ICAP should also prove useful, at least from a risk assessment perspective. However, these mechanisms, as they currently exist, will likely not be enough.
Given that the profit allocation rules are apparently meant to apply in a fairly mechanical manner, a coordination rule will be needed where the outcome of the profit allocation conflicts with transfer pricing outcomes that are not only permitted, but required, under existing treaty provisions that apply the arm’s-length principle. While it seems likely that the OECD intends the mechanical profit allocation rules to trump the transfer pricing rules in these cases, this should be spelled out, and as the OECD acknowledges, significant work will be needed to eliminate double tax in these cases. In particular, it needs to be recognized that, while rules may be conceived of mechanically, in practice there will always be issues associated with determining an appropriate profit base and deciding which industry a taxpayers belongs to (and thus what set of fixed percentages may apply), and tax administrations’ decisions on these and other points may need to be addressed in MAP or other forums.
One of the issues that looms the largest in the OECD’s Pillar One work is seldom acknowledged head-on: where will profits come from? It may be easy to develop consensus around the idea that certain countries, and certain functions, deserve more profits, but as noted above, transfer pricing has to be a zero sum game. Since the OECD is not proposing a system that necessarily involves significant double taxation, the profits flowing to market jurisdictions under the Secretariat’s proposal will need to be taken away from somewhere else.
Complicating the issue is the fact that the situations at hand will often not be clear cut bilateral cases, where profit is taken from X and given to Y. Profits allocable to market jurisdictions might reasonably be conceived of as relating to an enterprise’s marketing intangibles. In theory, then, profits could be allocated from the owner of the intangibles to the market jurisdictions—except that, under the intangibles rules introduced under Actions 8-10 of the BEPS project, profits from an intangible are not allocable to its legal owner, but rather to the entities performing development, enhancement, maintenance, protection, and exploitation (DEMPE) functions with respect to the intangibles, so profit may be taken from X and other affiliates performing DEMPE functions and given to Y.
The illustration provided in the Secretariat’s proposal glosses over this issue. It contemplates a group of two entities—P Co, in Country 1, and Q Co, located in Country 2 and selling in Countries 2 and 3—and cursorily concludes that, because P Co owns the intangibles exploited by the group, it is therefore entitled to all non-routine profit earned by the group under the current international tax rules. This, of course, may not be true under the current system, but rather hearkens back to a pre-DEMPE age where legal ownership rules were key to profit allocation. While some simplification is desirable for purposes of illustration, this potentially erroneous assumption allows the OECD to conclude that profits to be taxed by Countries 2 and 3 under Amount A would be allocated solely from P Co, and thus obscures an important issue.
In fact, because DEMPE functions will frequently be spread across multiple entities and multiple jurisdictions, the overlay of the Pillar One profit allocation system will not simply result in the transfer of profits from one jurisdiction to another, but instead may entail a contentious dispute embroiling several countries. As noted above, the potential for disputes about the application of the profit allocation rules means that market jurisdictions will often need to be involved in such disputes as well. Broadly speaking, the fact that the OECD has not articulated a guiding principle—comparable to the arm’s-length standard—to inform these profit allocation rules will make reaching principled resolutions very difficult. This is further complicated by the fact that, where a taxpayer has global consumer-facing sales, the sheer scope of potentially interested parties is staggering. Moreover, many of the market jurisdictions involved will likely be inexperienced and understaffed to handle the inevitable controversies that will arise from the new nexus and profit allocation rules.
The existing systems for resolving double taxation disputes, including MAP cases and mandatory arbitration where the competent authorities in a MAP proceeding are unable to agree, are not well suited to resolving multilateral disputes. A traditional MAP case, which is already a significant undertaking with just two competent authorities, may prove unworkable when additional countries with competing interests are included. As for mandatory arbitration, “baseball” arbitration is generally the favored approach to arbitration in double tax cases. In baseball arbitration, both competent authorities submit resolution proposals to the arbitration panel, which will accept one of them. Baseball arbitration is therefore generally conducive to achieving a fair resolution that promotes confidence in the system, as both competent authorities have an incentive to submit a reasonable proposal lest the arbitration panel pick the competing offer. However, when three or more countries are involved, these incentives threaten to break down, and no country’s proposal may reflect a resolution that is reasonable with respect to all other countries. Multilateral baseball arbitration raises the prospect that one or more countries may come out of arbitration a clear loser, rather than all participants feeling they have reached an acceptable middle ground, and thus threatens to undermine the international appetite for binding arbitration as a dispute resolution tool.
Yet the primary alternative is reasoned decision arbitration, in which the panel considers the issue fully and reaches its own decision on the matter. While reasoned decision arbitration has the potential to produce largely fair outcomes in multilateral cases, the arbitration panel would need to spend significant time to fully consider and resolve the issue, especially given the number of competing perspectives, and this would only occur after years had already been spent attempting to resolve the issue domestically and then in a MAP proceeding. Moreover, as noted above, the Secretariat has not articulated an overarching guiding principle for profit allocation that would provide the basis for a resolution. Reasoned decision arbitration therefore may prove unworkable in light of the OECD’s goal, expressed in BEPS Action 14, of ensuring the efficient and timely resolution of disputes.
A further issue arises when one considers the details of arbitration procedures that have been agreed to between treaty partners. Frequently, these only contemplate a bilateral proceeding, in which each country will select a member of the arbitration panel, and those two members will confer and select a third member to serve as the chair. This system works well as a guard against gridlock, but is not suited to larger cases: where there are three countries, for instance, one would end up with four panelists (one for each of the countries, plus the selected chairperson), raising the possibility that the arbitration could stall in a 2-2 disagreement. Multilateral arbitration will be critical to the implementation of a solution under Pillar One, but rather than simply reaffirming arbitration, the OECD will need to devote careful attention to how it should work, and what improved procedures will be necessary.
Of course, much work remains to be done, and it will be necessary to hammer out many details of the proposal. Nonetheless, the OECD Secretariat deserves credit for continuing its work under Pillar One and encouraging the development of consensus. The proposal’s balancing act between respecting the existing transfer pricing system and adopting simplified and formulary features appears necessary for any agreement among the Inclusive Framework to emerge, and may help to stymie the adoption of unilateral measures. At the same time, that balance presents potential issues of its own, and makes it critical that the OECD complement its work under Pillar One with efficient mechanisms for preventing and resolving double tax disputes. This should not be limited to a reaffirmation of the virtues of existing tools such as ICAP, APAs, MAP, and mandatory arbitration, but should involve detailed work on how these mechanisms need to be reworked and expanded. Moreover, standards should be set that require countries to invest appropriate resources into such mechanisms, and the OECD should monitor whether such investments have been made and implemented in a manner that allows the dispute resolution procedures to operate effectively.
This column does not necessarily reflect the opinion of The Bureau of National Affairs, Inc. or its owners.
Mark Martin and Steve Blough are principals in the Economic Valuation Services group in KPMG LLP’s Washington National Tax practice, Thomas Bettge is a senior associate in KPMG’s EVS practice, and Manal Corwin is Principal in Charge of KPMG’s Washington National Tax practice.
The information in this article is not intended to be “written advice concerning one or more federal tax matters” subject to the requirements of section 10.37(a)(2) of Treasury Department Circular 230 because the content is issued for general informational purposes only. The information contained in this article is of a general nature and based on authorities that are subject to change. Applicability of the information to specific situations should be determined through consultation with your tax adviser. This article represents the views of the authors only, and does not necessarily represent the views or professional advice of KPMG LLP.