Intangible BEPS Risks

Trust Bloomberg Tax for the international news and analysis to navigate the complex tax treaty networks and global business regulations.

By James J. Tobin, Esq.1

Ernst & Young LLP, New York, NY

Yet another commentary where I had every intention of not writing about BEPS again and I get excited about something in the OECD reports and get sucked back into the BEPS tangled web.  This commentary focuses on the OECD BEPS report on Actions 8-10, specifically on the aspects of that report dealing with risk and intangible property (IP) which would basically attribute IP and risk returns to the related parties responsible for development, enhancement, maintenance, protection, and exploitation (DEMPE) and control of risk functions.

The report presents intended changes to the OECD Transfer Pricing Guidelines (OECD Guidelines). Since many countries subscribe to the OECD Guidelines, it has been said that the conclusions in this report immediately take effect in those countries and thus these changes have a much more accelerated effect (at least in those countries) than the other BEPS Actions. So potentially a lot more worrying than with the other Actions where I am also at odds with some aspects of the conclusions reached.2

We at EY did a survey on the effect of changes to the OECD Guidelines, which can be accessed from our website (  It was released as an International Tax Alert in August 2015, titled "Country Implementation of BEPS Actions 8-10 and 13." It concludes that of the 34 OECD countries only nine make direct reference in their domestic legislation to the OECD Guidelines and it was felt that even for a majority of these countries a change in legislation or administrative guidance would be required to introduce changes in the OECD Guidelines into domestic law. In 24 of the remaining 25 OECD countries, the OECD Guidelines are considered to have "soft law" status.  That is, they are referred to as a source of interpretation of the arm's-length principle by tax authorities or courts but they are not binding and cannot contradict existing legislative rules. Twelve non-OECD countries surveyed report specific reference to the OECD Guidelines and again, in the majority of these countries, it was felt that amendments to local legislation or administrative guidance would be necessary to adopt changes to the OECD Guidelines. In addition, the OECD Guidelines would have soft law status in many non-OECD countries.

Another aspect I find troubling is the effective date of the proposed changes. I think there can be little doubt that the positions in the report are significant changes from the existing OECD Guidelines, yet there is no discussion of the effective date aspects. There are some implicit references in the report which would lead toward the conclusion that these changes are thought to be a clarification of existing rules and I am sure some country tax authorities will try and take that view. (For example, the Swedish tax authority announced this view in December.) However, I know of no country with an articulated rule or interpretation before 2016 that took the position that an IP owner without oversight of DEMPE functions should only earn a risk-free return. So to potentially apply such a rule retroactively must be inappropriate. And I would expect the courts to agree. But how to address transition issues assuming a country adopts this view after the OECD Guidelines are changed?

Moving beyond the precedential aspect of the report, I find many aspects of the report to be either ambiguous or potentially inappropriate, at least for many fact patterns, and doubt a court would be persuaded to follow the rationales for some of the recommendations, but I obviously worry that tax authorities seeking to raise revenue would be more inclined to do so.

The basic premise of the report with respect to risk and intangible property is to ostensibly keep faithful to the arm's-length standard. However, the big change is that the economic return from intangibles and increased return from bearing risk will not merely follow the legal ownership of an intangible or the party contractually assuming risks, but rather the returns should be "appropriately" attributed to/shared between related parties who actually control and manage the risks or intangibles. This premise can be best understood and is best illustrated in the OECD report with respect to so-called "cash-box" companies which have legal ownership of IP but have no substance to actually manage the IP development or exploitation. In such case, the report concludes that the IP returns should instead be allocated to the related parties which do control and manage those functions.  The report makes clear that the reallocation of profit is for transfer pricing purposes only and not for the purpose of determining the character of the income or other consequential substantive effects. This is also one of the aspects I worry about.

A multi-step process is provided both with respect to risk and intangibles to enable the determination of how the associated profit should be allocated. The steps will require much more functional analysis of the nature and value of risk and IP inherent in a group's supply chain and of which entities in the group make contributions to that value than I believe is current practice today. Given the required brevity of this commentary, I won't go into detail on the steps. However, I will make a few observations on how the rules seem to work (or potentially not work):

  •   The report appropriately recognizes and acknowledges that bearing risk is a factor in expected returns an entity can realize and accepts that risk can be effectively allocated within a multinational group. Examples demonstrate that a commission agent might expect a lower return than a buy-sell company which takes inventory risk, and a distributor with limited risk would expect a lower return than a distributor with full risk.  The caveat being that the party purporting to bear the risk must be consistent with that characterization in its course of conduct and be in a position to control and manage that risk. An example given was a purported (the word "purported" was used several times in the report) limited risk distributor which bore risk of excess inventory where the report concludes its course of conduct was not consistent with bearing limited risk.
  •   With respect to intangibles, the operation of the principle in the report is not to reallocate beneficial ownership of the intangibles but to reallocate the income from the intangibles. Thus, for example, in the case of a low-taxed/low-substance entity owning intangible property and receiving royalty income, the royalty income itself is not reallocated. But rather the low-tax company would be required to compensate related parties who are managing DEMPE functions. The report concludes that where the legal owner – low-tax and low-substance in this case – manages or oversees no DEMPE functions, it should be left with a net return for only its funding of the intangible.  And that funding return could either be a risk-adjusted return if it had the ability to assess and manage its funding risk or a lower, risk-free return if it did not manage that risk. (And I would submit that a funding return could be quite significant when considering real world examples like venture funds investing in early stage biotech, software, etc.) Although not covered in the report, presumably the return to the DEMPE controllers should often take the form of a value-added service fee.

This construct might be relatively easy to understand and apply (not that I completely agree with it) in the context of a parent with all DEMPE functions transferring IP to a low-tax, cash-box subsidiary with no DEMPE functions. The subsidiary would still earn the IP income in the first instance (e.g., recognize royalty or other IP income) and would be required to pay a value fee to the parent so that its net income would reflect a funding return. Some real world complications I could foresee:

  •   What if control of the DEMPE functions is divided between multiple affiliates? Individually it may not be difficult to determine an arm's-length amount of the service fee that should be charged for each activity.  However, this is unlikely to produce a result that leaves low-tax IP with only a funding return unless the residual profit (or loss) from the IP is merely allocated on some formulary basis back to the DEMPE controllers. (Indeed the potential expanded use of the profit split method is the focus of follow-up work to be completed by the OECD this year, so maybe all will become clear after that work is completed?) And in any event the relative value of each DEMPE function will potentially be the source of endless debate.
  •   Adding some complexity, what if the DEMPE controller(s) are mobile? For example, the regional head of R&D is a German-based employee this year, a U.K. employee next year, and a Dutch employee the year after? Is the DEMPE profit allocation to change each year? Is there some cumulative approach which is envisioned? Where an arm's-length service fee depends on the value of the activity for the period, this would likely not be an issue. Where the fee should include some part of residual IP profit or loss, it can get too hard very quickly.
  •   How to deal with transfers of IP? Since the legal owner is still considered the tax owner of the IP, the sales proceeds presumably all initially should accrue to that owner. Therefore, a buyer, related or unrelated, should not have concern as to whether the transferor can affect a true sale/transfer. But to what extent if any should the legal owner compensate a related DEMPE controller for a portion of the gain on sale? If the transfer is to a related party, the DEMPE provider, per se, may be continuing to provide the same DEMPE functions and it's hard to see why it should also be entitled to any one time profit from a gain on a transfer. If the sale is to an unrelated party, presumably no ongoing DEMPE services will be needed. So should the DEMPE controller(s) be compensated? If so, I would think any compensation should more reflect a goodwill/going concern value with respect to its service function than the IP value itself?
  •   How to deal with losses? One can expect enthusiastic support from countries in trying to grab excess IP profit. But the report makes clear that the DEMPE controllers who manage functions and risk should be allocated residual gain or loss after a funding return for the so-called cash-box company. I see a lot of unsuccessful IP development.  I fear a low take-up for countries accepting a negative value service charge when such a loss is to be reallocated.
  •   Don't forget it is the management and control of the DEMPE functions that is deemed to merit the sharing in IP profit or loss and not the actual performance of those functions. Thus, the existence of an R&D center does not attract IP returns if the control of that R&D is exercised elsewhere. I can imagine lots of debate on this with the countries in which the R&D center is located.
  •   Moving away from the unsympathetic low-tax, cash-box company example to some other real world examples also raises lots of questions.  Assume a European pharmaceutical company buys a U.S. biotech company to acquire a promising compound for say $1 billion. Assume all future DEMPE control is done from Europe. Should the U.S. target pay a value fee and be left with only a funding return. A new way to effect an IP migration without a migration? Or could there be a partial deemed transfer of IP? Will countries really accept a value fee of all residual profit? Even those which "adopt" the OECD Guidelines? I doubt it. As I find often the case in the OECD reports, the focus is on perceived abuses with little focus on making sure the proposed "solutions" work in all cases.
  •   It's clear in the report that the IP owners or risk-taking company need not actually perform all of the DEMPE control itself but rather could outsource some of these functions to related or unrelated parties. In such case, it presumably should have the capability of managing the outsourced function. But it is unclear what level of direct activity would be required. In the case of an outsourced function, the appropriate compensation seemingly should be the arm's-length amount for that function, but one would think without an explicit sharing in residual IP or risk profit.

As with most of the OECD reports, I am concerned that a disproportionate focus is on U.S. multinationals and they will potentially suffer most in a local country enforcement activity.  I was pleased to see that in the 29 examples provided in the report with respect to intangibles there were 12 hypothetical companies named to illustrate the principles of the report. At least 10 of the names were clearly not hypothetical U.S. companies – names included Shuyona, Zhu, Osnovni, Prathamika, etc. Clear evidence that the focus is not on U.S. multinationals so at least something to take some comfort in.

Overall I found the Actions 8-10 report very troubling. By defining the portion of IP return that the company that is the legal owner is not entitled to, but leaving many open questions on which entities are entitled to that return or a portion of that return, the prospect of extended controversy around the world and high risk of double taxation seems a highly likely outcome. Lots of countries seeing a potential revenue prize to pursue. And I'd suggest the issue will be real for virtually all large multinationals and not just for the dwindling few that may have something that resembles a "cash-box." I encourage the OECD to continue its work in this area but with greater input from industry. Meeting the very ambitious timetable for the report could be said to be an achievement. But not when it produces an unworkable answer.

This commentary also appears in the April 2016 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, Chip, Culbertson, and Maruca, 6936 T.M., Transfer Pricing: OECD Transfer Pricing Guidelines, and in Tax Practice Series, see ¶3600, Section 482 — Allocations of Income and Deductions Between Related Taxpayers, ¶7160, U.S. Income Tax Treaties.


  The views expressed herein are those of the author and do not necessarily reflect those of Ernst & Young LLP.

  It should be pointed out that the report's recommendations have not yet been officially adopted as changes to the OECD Guidelines. (Proposed changes to the OECD Guidelines must be formally submitted to and approved by the OECD Council.)

Request International Tax