Introduction
The present contribution aims at discussing two fundamental concepts often coming into play in the analysis of transactions between associated enterprises involving intangible property: the notion of (i) significant people functions (“SPFs”)[1] and that of (ii) development, enhancement, maintenance, protection and exploitation of intangible property (“DEMPE”)[2].
Based on my practical experience, lately these two notions have been inaccurately used (no pun intended) interchangeably by both taxpayers and tax administrations when it comes to the performance of the functional analysis as part of the accurate delineation of the transaction according to Chapter I of the OECD Transfer Pricing Guidelines (the “TPG”), while they arguably tend to serve different purposes. What I’ll try to explain below is that based on the original thinking of the OECD’s Working Party 1 and 6[3], their introduction was meant to be complimentary as they apply in two different conceptual frameworks. Indeed, while the SPF notion is part of hypothesizing a permanent establishment as a separate and distinct enterprise for purposes of applying the arm’s length principle absent any legally binding contract between the latter and the rest of the (non resident) enterprise of which it is a part, the DEMPE notion was conceived as a delineation tool for transactions involving intangibles complementing the fundamental framework set out in Section D2 of Chapter I of the TPG.
-
Historical background of SPFs and DEMPE introduction
As prominent legal scholars have pointed out[4], while focusing on “functions” have always been central in carrying out transfer pricing analysis since 1995 (the year of publication of the first version of the OECD TPG), their meaning have been construed differently over time. Specifically, para. 1.20 of the 1995 version of the TPG highlighted that […] “compensation usually will reflect the functions that each enterprise performs taking into account assets used and risks assumed […]”. Hence, since the very first time a proper conceptual transfer pricing framework began of being developed, the performance of any function (e.g. people function or automated one) was playing a prominent role in carrying out a sound comparability analysis. However, such a generic reference to “functions” seemed a bit foggy in clearly highlighting the different conceptual framework requested from the outset by the two key provisions of the OECD Model Tax Convention, Article 7 and 9 respectively, to guarantee a sound application of the arm’s length standard.
Stated otherwise, it seemed that something else was required to elaborate on the fact that while Article 7 of the OEC MTC, in determining the taxing rights of business profits, needed an additional element to proceed in attributing profits to a permanent establishment (which, important reminder, from a legal standpoint is nothing more than an administrative ramification overseas of the rest of the enterprise of which it is a part), Article 9 moved instead from the assumption that an important comparability factor (the contractual arrangements, if existing) had to be factored in the application of the arm’s length principle. For such a reason, while both article 7 and 9 endorse a transactional approach in applying the arm’s length principle, their starting assumption has been (and, at least formally, continues to be) fundamentally different, as the relevance of contracts in Article 9 had to be inevitably replaced by a conceptual fiction in article 7 to hypothesize the application of the separate entity approach for permanent establishments as well: what gives life to such fiction is the SPF concept.
Indeed in 2008 and 2010, as part of its Authorized OECD Approach for attributing profits to a permanent establishment (“AOA”), the joint working group WP1/WP6 introduced the SPF notion […] “ in order to attribute to the PE those risks for which the significant functions relevant to the assumption and/or management (subsequent to the transfer) of risks are performed by people in the PE and also attributes to the PE economic ownership of assets for which the significant functions relevant to the economic ownership of assets are performed by people in the PE”[5]. Hence, SPF is the proxy which comes into play, absent legally binding agreements, in order to guarantee a sound risk allocation between two parts of the same legal entity: article 9 does not come into play at all, from a conceptual standpoint, when we talk about SPFs.
On the contrary, the notion of DEMPE does limit its application to transactions involving intangible property, and – based on my understanding of the historical process which led to the introduction of the acronym – complement a broader, more elaborated performance of the comparability analysis as requested by the outcome of Actions 8-9-10 of the BEPS project. Specifically, with the introduction of the notion of “accurate delineation of the transaction” in Chapter I of the OECD TPG post 2017 as an essential part of the comparability analysis, tax authorities could be entitled to reallocate risk (which is something other than “recharacterize” a transaction, in my view) in all those instances where the legal owner of intellectual property does not perform as well as other functions such as development, maintenance or exploitation of the IP. However, the flipside of any additional guidance around control over risk could lead to further confusion as tax authorizes could offer subjective interpretation of control over risk. That is why some scholars offered an interesting solution by simply conveying the message that any DEMPE analysis could function as a guardrail to guarantee whether the contractual allocation or risk is actually honored by the actual conduct of the parties.[6]
In this respect, I believe that all instances of uncertainty when it comes to apply either SPFs or DEMPE within the same framework should be cleared by carefully reading the guidance provided for by the 2017 version of Chapter VI of the OECD TPG. In particular, paragraph 6.40 states that […] “if no legal owner of the intangible is identified under applicable law or governing contracts”[…]: this is nothing more than a careful elaboration of the five comparability factors which requires starting from contractual arrangement between related parties, if existing, to then move to the assessment of all the functions performed in transactions involving IP[7].
-
Where does the confusion stem from and what’s next?
Assuming that it is now clarified that SPFs and DEMPE operate on two different conceptual levels, then where is all the uncertainty and confusion in the application of these concepts coming from? This is something true specifically for the misunderstanding around the DEMPE concept, which also tends to be confused with other important transfer pricing notions, such as “unique and valuable contributions”, which is instead referred to multiple times in the OECD guidance to explain instances of transactions between associated enterprises warranting the application of the profit split method[8]. A possible explanation can be traced back to some historical lack of acceptance of some concepts (such as the AOA, which has been expressly rejected by the United Nations Model Tax Convention) justified by an alleged unfairness of the dichotomy state of residence vs. state of source. I also believe that such an underlying feeling of dissatisfaction in the allocation of taxable profits got some tailwinds by one of the fundamental messages arising from the BEPS actions around transfer pricing, whereby the allocation of entrepreneurial risk (and therefore, residual profits) merely by contract would no longer suffice for an accurate delineation of the transaction as part of a “new” comparability analysis.
Against this background, another major source of uncertainty stemmed from the fact that the TPG as a result of the application of BEPS actions 8-10 put a significant bias on identifying critically the location of “actual” decision making” as part of the risk allocation exercise. Arguably, this is clearly something that a proper performance of the DEMPE analysis assumes, but perhaps could be better addressed by expanding on some practical scenarios around the notion of “control over risk”.
Another distinguishing factor which could be better spelled out in any future iteration of the OECD TPG revolves around “economic ownership”. Often in discussions with tax authorities and taxpayer you may improperly find a reference to “economic ownership” when the analysis of any type of transaction between associated enterprises is warranted (e.g. one involving a routine service provider of administrative services remunerated on a cost plus basis). Since 2008, the OECD clearly elaborated on the fact that economic ownership conceptually exists as a proxy of ownership for direct tax purposes insofar as Article 7 is concerned[9]; in no instances reference to economic ownership is made in the latest iteration of the OECD TPG, as indeed the latter notion is used again to hypothesize the PE as a separate and distinct enterprise and not to accurately delineate a transaction (as the DEMPE concept is instrumental to).
Eventually, while pragmatism is always warranted when it comes to transfer pricing analysis, it is also crucial that a sound tax policy debate warrants a clear and certain application of concepts useful to guarantee what all economic operators aim at: certainty in the way tax affairs are structured and taxable basis allocated among countries. To this end, it could also be a potential option on the table to have the notion of DEMPE tied with a more elaborated definition of control over risk, e.g. a DEMPE analysis would necessarily warrant exerting some sort of actual/additional control over an asset (in terms of having the power of managing it, disposing it etc) – this to further shield the relevance of DEMPE as a tool to complete the accurate delineation of the transaction as part of the comparability analysis, and not as an anti-abuse tool in disguise, which does not seem – according to my view – the sole aim of why such a notion saw the light back in 2015.
**This publication is on a personal capacity and by no means can be associated with Agoda or any other affiliated company part of Booking Holdings**
[1] See Paras.18 and 19 of the OECD (2008) and (2010) OECD Report on Attribution of Profits to Permanent Establishments, available on IBFD sources
[2] See Chapter VI of the 2022 (published on January 20) version of the OECD Transfer Pricing Guidelines, with a specific focus on para. 6.34.
[3] Working Party no.1 is involved in the update on the Commentary of the OECD Model Tax Convention, whereas the Working Party no. 6 is involved with the update of the OECD Transfer Pricing Guidelines.
[4] See C. Silberztein and M. Guillaume, Does Intangible Ownership Move with the People that perform the DEMPE functions, in International Transfer Pricing Journal 7, 2022, para 2.
[5] See paras 17 to 19 of the 2008 and 2010 AOA Report, Part I.
[6] See R.Collier and I. Dykes, On the Apparent Widespread Misapplication of the OECD Transfer Pricing Guidelines, in Bulletin for International Taxation, IBFD, January 20222, at p. 24
[7] See paragraph 6.42 of the OECD TPG, 2017 version.
[8] For some interesting observations around this point of apparent friction between the two concepts, see E. Gonen, L. Karasik and M. McDonald, Control over Risk, DEMPE functions and the Remuneration of Service Providers, in Tax Notes international, Volume 102, June 21, 2021 and 1617.
[9] See footnote 3 of the 2008 OECD AOA, page 14, whereby the OECD clarifies that the “economic” ownership of assets in the Article 7 context means the equivalent of ownership for income tax purposes by a separate enterprise, with the attendant benefits and burdens (e.g. the right to the income attributable to the ownership of the asset, such as royalties; the right to depreciate a depreciable asset; and the potential exposure to gains or losses from the appreciation or depreciation of the asset)”.
________________________
To make sure you do not miss out on regular updates from the Kluwer International Tax Blog, please subscribe here.
