On September 21, 2012, TEI filed comments on OECD Working Party No. 6’s discussion draft regarding the transfer pricing aspects of intangible assets. The OECD discussion draft, released on June 6, 2012, sets forth proposed revisions to Chapter VI of the OECD’s Transfer Pricing Guidelines. TEI’s comments addressed the OECD’s functional analysis for allocating intangible related returns for transfer pricing purposes, noting that it appeared to move away from allocating returns based on arm’s length contracts between related parties for transfer pricing purposes. The Institute’s submission also addressed the definition of intangible assets, the approach of multi-national enterprises in organizing intellectual property on a group-wide basis, and the examples in the Annex to the OECD discussion draft.
The comments were prepared under the aegis of TEI’s European Direct Tax Committee, whose chair is Anna Theeuwes of Shell International B.V. Benjamin R. Shreck, TEI Tax Counsel, serves as liaison to the Committee and coordinated the preparation of the Institute's comments.
View full letter
In 2010, the Committee on Fiscal Affairs (CFA) of the Organisation for Economic Co-operation and Development (OECD) announced a project on the transfer pricing aspects of intangible assets, including a potential revision of Chapter VI of the OECD’s Transfer Pricing Guidelines concerning special considerations for intangibles (OECD Guidelines). The OECD published a scoping paper and held three public consultations with interested commentators. At the consultation in November 2011, representatives of the business community suggested that the OECD release an interim draft of its work for further public comment. The OECD released such a draft (Discussion Draft) on 6 June 2012, and requested public comments. In addition, the OECD announced a public consultation on this Discussion Draft, along with two other transfer pricing-related discussion drafts, to be held in Paris on 12-14 November 2012. On behalf of Tax Executives Institute, Inc., I am pleased to respond to the OECD’s request for comments on the Discussion Draft.
Tax Executives Institute (TEI) was founded in 1944 to serve the needs of business tax professionals. Today, the organisation has 55 chapters in Europe, North America, and Asia. As the preeminent association of in-house tax professionals worldwide, TEI has a significant interest in promoting tax policy, as well as the fair and efficient administration of the tax laws, at all levels of government. Our 7,000 members represent 3,000 of the largest companies in Europe, the United States, Canada, and Asia.
General Comments on the Draft
TEI welcomes the OECD’s effort to address the transfer pricing aspects of intangible assets. Transactions between related parties (however defined) have been a vexing issue for tax authorities and multinational enterprises (MNEs) alike, and thus this is an area ripe for clarification and certainty. Regrettably, while the Discussion Draft elaborates on many issues in significant detail, overall it lacks a unifying vision, pragmatism, and consistency. In addition, the draft seemingly contradicts Chapters 1 to 3 and 9 of the OECD Transfer Pricing Guidelines. In particular, the Discussion Draft does not properly reflect the manner in which MNEs derive economic value from their intellectual property (IP). For example, the comparability analysis discussed in the document refers to a particular method and is not based on a value chain (or economic) analysis.
In TEI’s view the OECD Guidelines should recognise that MNEs centralise their critical IP in a particular jurisdiction, or across a few key jurisdictions, for a variety of reasons. These reasons include: basic legal protection (including the ability to sue infringers), centralisation of IP development, funding and administration, risk hedging, a desire to clarify internal business responsibilities, and tax considerations. From a tax perspective, IP concentration simplifies transfer pricing compliance across jurisdictions, reduces the risk of double taxation, and avoids lengthy tax-related disputes over intangible assets generally. In contrast, the Discussion Draft could spawn the dispersion of IP assets and their associated returns across the entire corporate group and, therefore, across many more jurisdictions than where the MNE chooses to locate the legal and/or economic ownership of its IP. The inevitable result would be more transfer pricing disputes between MNEs, the jurisdictions in which they operate and, indeed, even between jurisdictions.
One of the OECD’s main purposes, promotion of free trade, is best served by healthy tax competition among its Member States. It is thus surprising that the OECD would seek to stanch such competition by dictating to MNEs how they should structure themselves.
TEI recommends that the OECD Guidelines strongly encourage tax authorities to conduct a holistic (or “top-down”) analysis of an enterprise’s transfer pricing practices, rather than proceeding on an asset-by-asset basis. That is, tax authorities should first understand and, absent abusive tax planning, accept an MNE’s business model and how its group IP affects transfer pricing policies. Unfortunately, it appears that the Discussion Draft would permit tax authorities to disregard contracts between related parties in an MNE, even absent tax abuse. In general, we recommend that the Draft expand its discussion of how intangible assets fit within the MNE’s global value chain, which in turn requires an understanding of the business environment in which the group operates. While an asset-by-asset (or “bottom-up”) approach may capture the individual resources (intangible or not) of an MNE, it cannot reflect the enterprise’s business as a whole, and thus does not reflect the approach taken by an enterprise from a business (i.e., non-tax) perspective.
In addition, the Discussion Draft contains a number of new provisions that will likely increase compliance costs for MNEs and lead to controversy regarding IP ownership. For example, the Draft promotes the use of the profit split method while staying silent about the transactional net margin method (TNMM). TEI recommends that the TNMM be promoted asa much simpler and more pragmatic method, especially for benchmarking routine transactions.
Further, the OECD recognises that its members will likely not reach a consensus view on the Discussion Draft. Because dissenting opinions will only add uncertainty to an already difficult area, as well as exacerbate the risk of double taxation, TEI strongly supports the development of a consensus document.
Finally, while the examples in the Discussion Draft helpfully flesh out the meaning of some parts of the OECD Guidelines, we believe the examples should be more detailed. In particular, the facts of the examples should reflect the full value chain of an MNE – that is, the “holistic view” of an MNE’s use of intangible assets across its operations, rather than an asset-by-asset approach.1
The Discussion Draft reflects certain recommendations proposed during the OECD’s 2011 consultation on the definition of intangibles. TEI supports the refinement of the definition of intangible assets to the extent it confirms that intangibles (or IP generally) should be defined as non-tangible assets that can be owned or controlled for use in commercial activities and separately identified.
However, aspects of an enterprise that are related to intangible – or tangible – assets, such as “assembled workforce”3 (sometimes referred to as “workforce in place”) should not be considered an intangible asset, because they cannot be transferred separately from other assets of the MNE. Instead, assembled workforce and similar aspects of an MNE should be considered attributes that influence the value of the MNE’s assets, but should not be intangible assets themselves. It appears that Paragraphs 25 and 26 of the Discussion Draft recognise this fact, but there is no explicit statement that assembled workforce “is not an intangible within the meaning of section A.1,” as is the case with most of the other illustrations of intangible assets in section A.4 of the Draft. Indeed, the Discussion Draft states that separate transferability is not a necessary condition for an item to be an intangible,4 suggesting that assembled workforce could be an intangible, at least under certain conditions. TEI recommends that whether assembled workforce is an intangible be explicitly stated in the guidelines.
The Discussion Draft deliberately does not categorise intangibles into certain classes, such as legally protectable assets (“hard intangibles”) and assets that are not legally protectable (“soft intangibles”).5 In addition, there is no distinction made between an intangible asset that reflects a routine or easily duplicated process on the one hand, and a unique process with high costs to replicate or other barriers to market entry on the other.
These distinctions can substantially influence the value of intangible assets, and, hence, the transfer price for those assets across the MNE. Categorising intangible assets would simplify the transfer pricing analysis (e.g., for “routine” intangibles). TEI therefore recommends that the OECD provide detailed examples of how and why MNEs have categorised such intangibles under their specific policies.
The Discussion Draft also fails to reflect the economic differences between group members of an MNE. For example, one member of an MNE may have greater bargaining power or responsibility than another, inasmuch as one member may be an investment center, another member a profit center, and yet a third member a cost center (operating on a “cost-plus” basis). If these members were unrelated parties, the disparity in bargaining power would obviously come into play when setting the price for transactions between them. Such differences are especially important in the intangibles area where, for example, the ability of one member to continue as a going concern depends upon its contractual relationship with another member, which is not an uncommon occurrence in relationships between unrelated parties. We recommend that the OECD take these factors into account in revising the guidelines.
Further, we recommend that a paragraph be added regarding the importance of “know how” to MNEs.6 Often, highly valuable intangibles can – and do – derive from a single idea, regardless of whether the original idea can be kept secret or be legally protected. The Discussion Draft should reflect this.
In addition, there is no general definition of a “marketing” intangible in the Discussion Draft even though six of the examples (numbers 3 to 8) deal with such intangibles in a distribution context. TEI recommends that the OECD Guidelines include a definition of a marketing intangible so taxpayers will know when these examples apply to their intangible assets.
Paragraph 23 of the Discussion Draft suggests that group synergies are necessarily driven by several entities. TEI submits that this ignores the reality that a parent or principal company is usually the catalyst for group initiatives, e.g., when an MNE has a central procurement entity to make use of combined purchasing power. We recommend that the Discussion Draft include an example whereby the Principal (or Parent) initiates, drives, or encourages those synergies.
Identifying the Parties Entitled to Intangibles Related Returns7
The Executive Summary of Section B of the Discussion Draft states that “transfer pricing outcomes in cases involving intangibles should reflect the functions performed, assets used, and risks assumed by the parties” and that “neither legal ownership, nor the bearing of costs related to intangible development . . . entitles an entity within an MNE group to retain the benefits or returns with respect to intangibles without more.”8 The Executive Summary notes that this statement “is consistent with other sections of the Guidelines and does not reflect an intention to depart from the principles of Article 9.”9
TEI respectfully disagrees, and believes this statement is too broad to be consistent with the arm’s-length principle. In our view, if contractual terms among related parties are consistent with this fundamental principle, they should be respected by taxing authorities regardless of the ex post results. Further, absent evidence of tax abuse, primary consideration should be given to the party that finances, bears the risk, and therefore in most cases economically owns, the IP at issue. While the Discussion Draft states that one factor in allocating intangible related returns should be the “terms and conditions of legal arrangements” among the parties10 – noting that an additional factor should be the “risks assumed, and the costs incurred” – the Draft appears to give significant – and perhaps controlling – weight to where functions are performed and where (presumably non-intangible) assets are owned. In our view, returns should be allocated to the party that bears the risk of IP development and ownership and not to the party that performs a development service for another party, if that is what the contract between the parties requires (as is the case in many contracts between unrelated parties). A parting acting as a service provider is entitled to a service fee, but not to the residual profit associated with the development of the IP.
To take perhaps the most fundamental example, employers often claim the “intangible related return” to IP developed by an employee, an unrelated party in almost every case. In this relationship, the employee is paid an agreed-upon salary and the employer reaps the rewards of the employee’s efforts – positive or negative – by bearing the cost and risks. Thus, an employee of a pharmaceutical company who develops a cure for cancer will almost certainly not reap the entire rewards from such a discovery; rather, the company will. Similarly, an advertising agency that develops an exceptionally successful slogan or trademark on behalf of an unrelated client would not receive all of the incremental economic benefits attributable to such IP. The reality of these examples notwithstanding, the Discussion Draft suggests that because the employee or advertising agency performed the “functions” in developing such a cure or slogan, the employee or agency should be allocated most (or all) of the “intangible related returns.” Because this is clearly not the case in these unrelated party contexts, it should not be the case for transfer pricing in respect of related parties.
The Discussion Draft also addresses the alignment of functional contributions and financial investment with legal rights,11 concluding that an entity contractually entitled to IP returns should either have developed, enhanced, maintained, and protected the intangible itself – or should have arranged to have such functions performed under its “control” by independent or associated enterprises (on an arm’s length basis). The alternatives in the Discussion Draft require, however, an unrealistic level of operational control by the IP owner over the intangible-creating activity or functions, thereby spreading the return from such a creation across the MNE group.
The OECD should clearly recognise that MNEs often centralise their critical IP in a single entity, or in a limited number of entities, for non-tax business reasons, such as basic legal protection (including infringement lawsuits), centralisation of IP development, funding and administration, risk hedging, and to clarify internal business responsibilities.12 Of course, tax considerations – including the effective tax rate on returns from IP – also play a part in the decision of where to locate IP ownership. We note that from a tax perspective, centralisation of ownership simplifies transfer pricing compliance around the world, minimises the risk of double taxation, and forestalls lengthy tax-related disputes over intangible assets. In contrast, a “functional analysis” of the development and ownership of intangibles – which the Discussion Draft erroneously asserts in paragraph 11 and generally in section B.1. as being mandatory – significantly complicates the management and compliance costs of transfer pricing for an MNE and will almost certainly lead to double taxation and competent authority disputes as jurisdictions bicker over which function is relatively more important.
Further, the term “control” is used throughout the Discussion Draft in respect of which party bears risk, but it is never clearly defined. More fundamentally, no matter who might control a particular risk, the arm’s-length principle should prevail. The Discussion Draft cross references paragraphs 9.23 through 9.28, which relates to which parties in fact control the relevant functions,13 but distends the importance of control in determining risk allocation among related parties from that in Chapter IX, which states that “*o+ne relevant, although not determinative, factor that can assist in this determination is the examination of which party(ies) has (have) relatively more control over the risk. . . .”14 In many cases, an MNE is willing to accept a risk in a contract with an unrelated party (especially when it is able to assess the risk and have the capacity to absorb the risk) even though it does not control the risk itself, and the Guidelines should reflect this fact.
In addition, in a global economy where talent and resources are spread all over the world, outsourcing – to related and unrelated parties – has become ubiquitous. To minimise double taxation, tax disputes, and encourage free trade, the best approach to transfer pricing is to respect the entrepreneurial decision of the IP owner to fund and steer its research and development programs. Thus, funding of intangibles by the “principal” company should remain a key criterion for IP ownership attribution for tax purposes (even with limited operational oversight by such company) since in a market economy a key criterion for determining an entrepreneurial engagement is financial risk. The OECD Guidelines should recognise that the oversight or “control” function can be – and at times (e.g., in the private equity sector) is – outsourced for a fee by the Principal to other parties.
Thus, in our view, any attempt to assess who “controls” a risk in respect of an intangible asset that goes beyond the contractual arrangement between the parties will only lead to disputes between taxpayers and taxing authorities, and, indeed, between tax authorities themselves.
Transactions involving the Use or Transfer of Intangibles15
In respect of valuing intangibles in their use or transfer, TEI maintains that a “hindsight” analysis should not be conducted. It is rarely the case for contracts between unrelated parties to contain a “re-valuation” clause addressing circumstances where the parties have gotten the price “wrong.” Thus, while many contracts involving the transfer or use of IP between unrelated parties may be contingent on production or utilisation of the relevant IP, the “price” for the IP – even if it is a “per transaction” or “per item” amount – is in most cases set in advance. Certainly, it is rare that a contract between unrelated parties will include a sort of “automatic” re-pricing clause that is triggered where it becomes apparent that one party has gotten the better side of a deal. Because the use of hindsight by tax authorities in revaluing or re-pricing the terms of a contract that were arm’s-length at the time it was negotiated is inconsistent with the arm’s-length principle, it is inappropriate.
The Arm’s-Length Principle16
Section D. of the Discussion Draft raises the bar for use of the TNMM method. Unfortunately, this will likely lead to a significant increase in compliance costs for MNEs, or force them to choose other, less reliable transfer pricing methods. In this regard, the Discussion Draft is silent in respect of the TNMM in favor of the transactional profit split method, which often results in higher compliance costs. For example, it is much easier to benchmark the return from routine transactions under the TNMM.
Paragraph 93 of the Discussion Draft states that for “know-how or trade secrets, available legal protections may have a different nature and not be as strong or last as long” as protections provided to more formal intangibles (presumably patents, copyrights, trademarks, etc.). This paragraph should be modified to reflect, among other things, that in an MNE context some intangibles, such as a trade secret, can have an unlimited life and produce non-routine returns indefinitely. (A frequent example of this is the formula for Coca-Cola.) Indeed, in some cases, an MNE may purposefully choose not to seek patent protection for technical know-how so the knowledge can remain completely secret. In these cases, the IP owner should merit permanent, inflation adjusted compensation as long as the intangibles are properly maintained and protected by confidentiality.
In respect of paragraphs 81 to 141, the OECD challenges transfer pricing methodologies that assume all residual profit from transactions after routine returns accrue to the IP owner.17 In some ways, this contradicts the TNMM, even though that methodology is heavily promoted by new Chapters 1 through 3 of the OECD Guidelines.
In addition, while the reference to the importance of a value chain analysis is welcomed (i.e., as noted, a “top down” or “holistic” approach to the IP of MNEs), the OECD should recognise that the IP centralisation policies of most MNEs almost always lead to one-sided transfer pricing methodologies. Thus, the OECD should modify the statement in paragraph 81 that a “one-sided comparability analysis does not provide a sufficient basis for evaluating a transaction involving the use or transfer of intangibles” and restore the TNMM and one-sided approaches.
In respect of paragraph 105, it is cumbersome to refer to the intangible property as defined therein as a “Section D.1.(vi) intangible.” Instead, we suggest using “Non-Comparable Intangible” or, if necessary, “Non-Comparable Value-Adding Intangible.”
In respect of paragraphs 112-13 and 135, while valuing an intangible at its replacement cost or value is sometimes inappropriate, it should be recognised that this is the method of choice when considering so-called process or routine intangibles, i.e., ones that do not drive value for the MNE as a whole. We recommend that the Discussion Draft modify paragraph 113, which currently characterises the use of this method as available in only “some limited circumstances.”
The Discussion Draft should also define what is meant by a “rule of thumb,” as expressed in paragraph 116, to avoid misunderstandings over the use of industry averages, which are a useful tool to verify assumptions, especially where no other comparable can be identified. Indeed, since the draft speaks repeatedly (and, in our view, erroneously) of industry practices in support of recharacterising transactions, it is inappropriate to completely discount the use of industry averages to verify the arm’s-length nature of a transaction. If taxing authorities are permitted to use industry practices to show that a transaction is not arm’s-length, then taxpayers should be allowed to use those same practices to show that a transaction meets this fundamental standard.
TEI submits that paragraph 151 goes too far in asking taxpayers to maintain documentation of “alternative assumptions and parameters” as part of a “sensitivity” analysis. An MNE should only be required to document the assumptions and parameters that it believes are correct and reasonably comparable and not to do what should be the job of tax authorities. The costs of conducting a sensitivity analysis of possible and undefined alternatives and assumptions would almost certainly outweigh any benefits.
Finally, paragraphs 168 through 170 discuss accounting for the tax rates of the transferor and transferee under a projected cash flow analysis of intangible valuation. This is inappropriate. Theoretically, if one unrelated party were aware of the other’s tax rate it might affect the negotiations between them, but such knowledge is rarely present. Indeed, because effective tax rates vary from year to year across MNEs and their operating groups, they likely will have nothing to do with the specific transaction being negotiated.
Comments on the Examples in the Annex
TEI recommends the inclusion in the OECD Guidelines of more realistic and sophisticated examples than those in the Annex.18 Such examples would assist taxpayers and tax authorities in applying the rules. Further, more realistic and sophisticated examples would better inform taxing authorities on modern business models of MNEs, mitigating misunderstandings between taxpayers and tax auditors, conserving time and resources for both.19
Most of the examples under chapter VI.B “Identification of Parties Entitled to Intangible Related Returns” begin with an analysis of the functions performed and then summarise how the transaction should be treated for tax purposes. This seemingly goes against the stated premises of the Discussion Draft that contractual arrangements should be the starting point of the analysis.20 TEI recommends that the examples describe the contractual arrangements and then discuss whether the contractual arrangements should be respected or disregarded, depending on the circumstances.
Moreover, some examples (such as example 20) refer to “the behaviour of independent enterprises in similar circumstances” (emphasis supplied), instead of making reference to contractual arrangements. A behavioural analysis would constitute a separate, supplementary documentary requirement that would be very difficult to satisfy and give the taxing authorities broad powers to recharacterise transactions. Thus, it is clear that documentary burdens, controversies, and double-taxation cases would increase exponentially if the draft were to continue to focus almost exclusively on functions, with reference to behaviours in the examples, without considering contractual arrangements.
We also recommend providing examples illustrating the full chain of an MNE’s operations rather than just one piece. Those examples should relate to different types of industries and business models and clearly distinguish the respective contributions to the MNE’s IP of all players (Parent/Principal/local affiliates). Where appropriate, the TNMM method should be used instead of the transactional profit split method, especially since the TNMM is the most comparable method in that context.
As for the industries to cover, we suggest
- An industry that is product-driven (i.e., one where the technology is embedded in the product itself) and where the manufacturing, supply chain and sales organisations are more routine
- An industry that sells commodities
- A services industry
The types of business models to cover should include:
- An IP centric policy versus a decentralised IP policy
- A centralised business model (Principal) versus a decentralised business model
The additional examples should also outline the evolution of a business model over time within the same MNE and its effect on IP ownership and overall compensation of the intangible.
Another example should underline the importance of central/group initiatives (standard operating procedures, group network, global communication, global ERP, etc.) that do not necessarily constitute IP expenditures per se but enhance the success of the core intangibles. In the same vein, an example should be developed whereby an MNE bundles not only the use of core intangibles, but also the use of central group services that are instrumental to the business, into a single royalty rate.
TEI appreciates the opportunity to comment on the OECD’s Discussion Draft. If the OECD believes our participation in the announced public consultation is warranted, we would be pleased to do so.21 These comments were prepared under the aegis of TEI’s European Direct Tax Committee, whose Chair is Anna M. L. Theeuwes. If you have any questions about the submission, please contact Ms. Theeuwes at +31 70 377 3199 (or An.M.L.Theeuwes@shell.com) or Benjamin R. Shreck of the Institute’s legal staff, at +1 202 638 5601 (or firstname.lastname@example.org).
1 TEI would welcome the opportunity to assist in developing any such examples.
2 See Discussion Draft, Section A.
3 See id. at paragraph 25.
4 Id. at paragraph 7.
5 Id. at paragraph 13.
6 Paragraph 16 merely describes know how without reflecting on its importance.
7 See Discussion Draft, Section B.
8 Id. at page 12.
10 Id. at paragraph 29.
11 Id. at paragraph 38.
12 Other business reasons to centralise IP include the ability to globally organise research and development and immediately disseminate upgrades around the world through a “hub and spoke” system.
13 Discussion Draft, paragraph 41.
14 Chapter IX, paragraph 9.20 (emphasis supplied).
15 See Discussion Draft Section C.
16 See id., Section D.
17 See, e.g., paragraph 108 (“caution should be exercised in adopting a transfer pricing methodology that too readily assumes that all residual profit from transactions . . . should necessarily be allocated to the party entitled to intangible related returns”).
18 TEI would be pleased to assist the OECD in drafting such examples.
19 We also note that certain examples that terms such as “goodwill” are sometimes used instead of the more proper “marketing intangibles” (however defined), such as in examples 14 and 15, which can create confusion.
20 See, e.g., Discusison Draft, paragraph 30 (“legal registrations and contractual arrangements are the starting point for determining which members of an MNE group are entitled to intangible related returns.”).
21 TEI has participated in several OECD public consultations regarding transfer pricing, including on 7 November 2011 regarding “Transfer Pricing of Intangible Property: Definitional Approach,” 22 March 2011 regarding “Valuation in a highly uncertain environment,” and 9 November 2010 regarding “Definitional and ownership issues related to intangibles developed through research and development.” TEI also filed comments with the OECD regarding the scope of the intangibles project on 14 September 2010.