By Craig A. Sharon, Esq.
Ernst & Young LLP, Washington, DC
Transfer pricing professionals have been trying to absorb a series of Organisation for Economic Co-Operation and Development (OECD) reports over the past six months: (1) a report addressing base erosion and profit shifting (the BEPS Report); (2) a plan laying out the action items and timelines for addressing the issues identified in the BEPS Report (the BEPS Action Plan); (3) a revised discussion draft on the transfer pricing aspects of intangibles (the Intangibles Report); and (4) a "white paper" on transfer pricing documentation (the White Paper).2 These reports have been accompanied by a dizzying array of analysis, explanation, and prognostication from OECD officials, Treasury/IRS representatives, outside advisors, taxpayers, and others in a seemingly non-stop round of speeches, conferences, webcasts, articles, and other discourse.
So let me add to the chatter by focusing on a fundamental, oft-mentioned topic: Is the OECD moving away from the arm's-length standard as the international benchmark for transfer pricing? Indeed, is the OECD, as some commentators have suggested, in the early throes of adopting, or setting the stage to adopt, formulary apportionment as an alternative to the arm's-length standard? On the latter question, it's a stretch in my view to argue that the OECD is wavering in its long standing opposition to formulary apportionment, particularly given the flat-footed arguments against it in the BEPS Action Plan. Still, taxpayers should be concerned about the OECD's direction because there's a fair amount of room between current mainstream thinking about the arm's-length standard and formulary apportionment. Within that broad spectrum, it's clear that the OECD, with the support of both developed and developing countries, is proposing to change the prevailing application of arm's-length principles. That shift raises several questions:
On the first question, it appears that the OECD would now like to apply the arm's-length principles set forth in the OECD's 2010 report on the attribution of profits to permanent establishments (PEs) under Article 7 (Business Profits) of the OECD Model Tax Treaty (the Authorized OECD Approach, or AOA) to transactions between associated enterprises under Article 9 (Associated Enterprises) of the OECD Model. Under the AOA, in allocating income between an enterprise and one of its PEs, a taxpayer is first required to perform a functional analysis that identifies the economically significant activities and responsibilities undertaken by the PE. That analysis focuses primarily on the "people functions" that run the business, assume and control risk, manage important assets, and control the capital that funds the business. After the key people functions are identified, the AOA then requires taxpayers to apply vaguely defined "mechanisms" to attribute the risks, assets, and capital of the business between the enterprise and the PE consistent with the location(s) of the key people functions.
The recent OECD reports make clear that the OECD now has the same kind of "activities-based" test in mind for associated enterprises:
The OECD's new emphasis on and prioritization of people functions represent a significant change in traditional transfer pricing functional analysis, which has historically focused in equal part on the respective functions, assets, and risks of the relevant related parties, as revealed in their operations, intercompany agreements, and other legal arrangements. Except in extraordinary circumstances, a taxpayer's chosen legal forms and intercompany agreements have been respected and given effect. That's important because returns to assets and allocations of risk within a multinational enterprise generally flow from the company's legal structure and the terms of its intercompany agreements. Under the OECD's new formulation, this historic approach would give way to an AOA-type analysis that looks first to functions performed, with assets and risks following the key functions in unspecified ways, but seemingly based on either speculative theories about how unrelated parties would structure a transaction at arm's length (i.e., general AOA principles) or objective, formulaic-type factors such as sales, headcount, or tangible assets (i.e., OECD surrogates for "economic activity"). In the absence of nearly identical comparable third-party transactions (i.e., an unrelated transaction with no material differences in functions, assets, and risks from the controlled transaction) - a rare occurrence in transfer pricing - legal constructs, even if properly executed and followed to the letter, would be overridden by the people functions test, possibly resulting in a change in the deemed owner of assets and/or bearer of risks determined under a taxpayer's intercompany arrangements.3
Ironically, the OECD seems to have come full circle. The AOA applies the intercompany transfer pricing rules expressly "by analogy" to PEs, while the OECD is now proposing to apply the intercompany rules by reference to the AOA -- an analogy within an analogy. One wonders, though, how far the double analogy can be stretched or, more precisely, how far the OECD intends to take it.
The OECD has been careful not to sound as if it's trying to revolutionize existing core transfer pricing principles, based as they are on the arm's-length standard. Consistent with that, the OECD does not believe that its new emphasis on people functions represents a radical change in the standard.4 To the contrary, the BEPS Report and the BEPS Action Plan summarily reject formulary apportionment and acknowledge that the existing framework works well for the most part:
As described in the BEPS Action Plan (e.g., at p.10), the stated transfer pricing target of the BEPS project is the high mobility of intangible property, risk, and capital and the increasingly integrated nature of global businesses that allow the separation of assets and risks from functions. These characteristics create "technically legal" planning opportunities for multinationals to shift substantial income to affiliates in low-tax jurisdictions having limited functionality. Adding fuel to the fire, these "artificial," contract-based allocations are supported by one-sided pricing methods that focus on easier-to-value people functions and leave the bulk of the returns to intangible assets, risks, and funding with the non-tested party.
If the OECD is proposing only that the days of allocating substantial chunks of system profit to a "P.O. Box" or a "guy and his dog" located in a tax haven are over and that, regardless of legal formality, some reasonable amount of functional substance is required within a low-tax entity to respect an intellectual property (IP) migration, risk transfer, or other "high-risk" transaction, then most taxpayers, including those who have relied heavily on intercompany agreements to determine transfer pricing outcomes (e.g., principal structures and cost-sharing arrangements), should be fine. The current OECD guidance, although imprecise, does not set a particularly high legal threshold for satisfying such a substance test. Under Chapter IX of the OECD Guidelines dealing with restructurings, a taxpayer need not perform the key functions on its own, but instead can outsource such functions to affiliates or third parties, provided the taxpayer retains the authority and has the expertise to "manage and control" its subcontractors. As a practical matter, that standard can be satisfied with only a few well-placed managers. In recognition of this and in response to Chapter IX of the OECD Guidelines, taxpayers that may have started with limited local substance in low-tax jurisdictions have been adding employees in recent years. In light of this trend, the OECD's new people functions test could be viewed as only reinforcing or reaffirming what most taxpayers are already doing.
Notwithstanding OECD representations about the limited scope of the BEPS project, there is ample reason to believe that the OECD is focused on much more than requiring taxpayers in low-tax jurisdictions to add some minimal level of substance to support their legal arrangements. Neither the people functions test nor the other transfer pricing items in the BEPS Action Plan reflect such a limited purpose, nor are they likely to have such a limited effect. In my mind, the odds are high that the proposed changes, in whole or part, will lead to less consensus, more uncertainty, and new transfer pricing controversy - all results that are contrary to the stated goals of the BEPS project.
Here's why, in my view:
As noted in a previous Commentary,6 I'm all for having the OECD manage the BEPS project, and while both the scope of the Action Plan and the compressed timetable are bold and impressive, the breadth of the Action Plan, at least as it applies to transfer pricing, may be too ambitious and the implementation challenges too daunting to achieve the project's stated goals. Who knows whether or not a broad international consensus based on the OECD's new proposals is possible and, if so, where that consensus will fall in the extreme spectrum between current arm's-length principles and formulary apportionment, as supplemented by overriding undefined special measures? And equally important, if there is no new consensus, what will happen to the current consensus as governments splinter, with each free to adopt the BEPS concepts that they find most appealing? Perhaps underappreciated to date, the downside risks of a failed BEPS project (e.g., double-tax chaos) are just as great as, if not greater than, the OECD's perception of the downside risks of the status quo, which only further behooves taxpayers, tax authorities, and other interested parties to work cooperatively together to find common ground.
In the meantime, multinationals will have to go about their business. That creates an immediate question: To what extent should the OECD's people-first functional analysis be viewed as a new arm's-length benchmark or as a clarification of the existing standard? The question is critical as a practical matter. If the new focus on people functions is deemed to create a new arm's-length standard, it would presumably apply only on a prospective basis and will be accompanied by transition rules that account for a taxpayer's historic transfer pricing model (e.g., possible buy-ins/buyouts based on historic IP investments and risk allocations). On the other hand, if the new emphasis on functions is treated simply as a clarification of the current rules (i.e., as just providing additional detail on the substance required to support a taxpayer's chosen transaction form), taxpayers will need to be concerned about the possible application of the "clarified" rule to open taxable years. The Obama Administration's proposed legislation regarding the definition of "intangibles" for purposes of §367(d) is not a helpful precedent. Regardless, I can imagine individual tax authorities taking different positions on the prospectivity of the new principles. In fact, the BEPS discussion may already be influencing the course of current audits. That cannot be good for the avoidance of double taxation. The OECD may need to lay down a few markers as the BEPS process proceeds to help both taxpayers and tax authorities cope with the interim uncertainty. Otherwise, a project intended to promote investment, jobs, and growth may retard it for years to come.
This commentary also will appear in the October 2013 issue of the Tax Management International Journal. For more information, in the Tax Management Portfolios, see Maruca and Warner, 886 T.M., Transfer Pricing: The Code, the Regulations, and Selected Case Law, Gleicher, 892 T.M., Transfer Pricing: Competent Authority Consideration, Culbertson, Durst, and Bailey, 894 T.M., Transfer Pricing: OECD Transfer Pricing Rules and Guidelines, and in Tax Practice Series, see ¶3600, Section 482 - Allocations of Income and Deductions Between Related Taxpayers, and ¶7160, U.S. Income Tax Treaties.
1 The views expressed herein are those of the author and do not necessarily reflect those of Ernst & Young LLP or any other member of Ernst & Young Global Limited.
2 The OECD reports referenced in this Commentary can be found at www.oecd.org/tax.
3 Presumably, the OECD will continue to support the benchmarking of intercompany arrangements against nearly identical third-party transactions when such transactions exist (i.e., when the functions, assets, and risks of the comparable transaction mirror the functions, assets, and risks of the controlled transaction). The OECD is obviously most concerned about differences in function and is effectively proposing to align a taxpayer's assets and risks with its functions, regardless of the terms of any relevant upfront agreement, such as a cost-plus contract research and development arrangement. See, e.g., Examples 12 and 13 in the Intangibles Report, pp. 60-62.
4 See, e.g., "BEPS Project Not Intended to Eviscerate Arm's-Length Principle, OECD Official Says," 2013 TNT 149-5 (8/2/13).
5 Many commentators on the BEPS project, including the U.S. Department of Treasury, have stated a preference for a CFC-type solution as an alternative to a changed arm's-length standard, but those comments seem to overlook the fact that the CFC solutions proposed to date focus only on intangible-related income (and affect only U.S. companies) and that the identification of such intangible income will likely require the application of arm's-length principles (possibly in combination with formula-like allocation factors) to separate the "tainted" income from other kinds of income. For this reason, the definition of intangible property and the activities-based transfer pricing principles incorporated into the BEPS Action Plan and/or the Intangibles Report may have broader application than anticipated.
6 "Initial Observations on the OECD BEPS Project," 42 Tax Mgmt. Int'l J. 363 (6/14/13).