agata ucedaIf time allows, I try to contribute to the OECD discussions submitting comments to their draft papers. Last November and December 2014, the OECD came out with several drafts for discussion (this earlier article tries to summarize them briefly). After reading more than 600 pages of documents during the week of Christmas holiday, I decided that I was going to pass providing comments to the paper on Action 4 interest deductions and other financial payments. After all, the proposal is basically a formula and I am a simple transfer pricing advisor, trained during 15 years to use the arm’s length principle.

For similar reasons, I decided to pass providing comments to the paper on profit splits. In those 15 years as transfer pricing advisor, I have not used the profit split method more than 10 or 12 times. Of that, 5 were in a loss situation and the exercise was to try to split losses, not profits. Of the remaining, 3 were in bilateral APAs where the allocation key used was very simple either sales or gross margin. The other 2 to 4 were very sophisticated formulas that were never implemented because the client did not understand them (and maybe even we did not understand them), the facts and circumstances of the multinational had changed by the time we had arrived to “a formula”, or the IT systems of the company could not accommodate the policy proposed. Based on 10-12 cases, I concluded that I don’t qualify as an expert on the profit split method, so I refrained from commenting. Admittedly, lack of time also played a role.

That left the paper on risks as the most important paper to provide comments. Thus, this piece summarizes the comments submitted by the writer to the OECD on the Discussion Draft “Revisions to Chapter I of the transfer pricing guidelines” (dated 16 December 2014) (“the Draft”):

As a general comment, I welcome this Draft as it is well written and timely. The Draft helps to make the link between some of the Chapters added to the OECD Guidelines in 2010, particularly Chapter 3 and Chapter 9, and the current Chapter 1, and clarifies and reinforces some points that – even though they are not new – were often overlooked in practice.

In particular, the language in section D.1 about delineating transactions and the application of the comparability factors is clear and helpful. Since the Guidelines were changed in 2010, with a new Chapter 3 on “comparability analysis”, which is often used as a synonym of benchmarking in this Chapter 3, the wording on comparability on Chapter 1 had been overshadowed. This in itself, a full Chapter on comparability/benchmarking after the Chapter on methods (Chapter 2), placed the emphasis on “benchmarking” instead of on the first and more important step of identifying the relevant commercial transactions using the “comparability factors” explained in Chapter I, Section D.

A few Specific Comments to this Section D.1

Although I agree with the general idea presented in Paragraphs 12, 13 and 14, I think it is important to mention that one of the reasons that multinationals ( “MNC”) exist is to reduce transaction related costs. If, in applying the arm’s length principle, members of a MNC need to replicate all the steps that would occur in a third party negotiation (e.g., consider all options realistically available to the parties to the transaction, assume “complete information asymmetry” between the parties, draft contracts that would be similar in content/detail to third party contracts) and at the same time they need to take into account factors that are specific to the MNC operations (like diversification or the lack thereof, implicit support and to what extent it applies, synergies and who generates them and how they can be split, etc.), the task may proof impossible to achieve.

I agree that applying the arm’s length principle should not mean an automatic benchmarking exercise and a copy-paste of information from the annual accounts. I agree also that to be able to salvage the arm’s length principle we need to apply the principles provided by the OECD Guidelines more rigorously. However, if the exercise turns out to be too complex, the reaction will be oversimplification. Copy-paste transfer pricing reports are an oversimplification. So are formulas. Also, MNCs will find themselves at a disadvantage vis-a-vis Small and Medium Size Enterprises (“SMEs”) because MNCs will have a higher administrative burden than SMEs and arguably none of the benefits of being an SME (flexibility, less stringent labour laws, less infrastructure costs, lower governance and accounting requirements and costs, etc.).

There are a number of specific transactions in a MNC in which applying the arm’s length principle needs to be done with some degree of simplification, otherwise the exercise becomes too complex, and sometimes absurd, because the comparison to what independent parties would have agreed upon is just not possible. In particular:

a) The allocation of headquarter costs and applying the benefit test. Per definition, only MNCs have these type of costs so the comparison to what third parties would have agreed upon is very artificial.

b) Similar to that, the concept of implicit support (either when calculating guarantee fees or in a similar situation). We could debate endlessly on whether a parent company will rescue a subsidiary if they were to default, and whether the answer would be different if the MNC was very centralized or not very centralized, or whether the subsidiary was core business to the MNC or not core business, etc.

The truth is that the answer is probably different in each case. Even for the same MNC the answer can be different in 2015 than it would have been in 2008, and they can even decide to rescue that particular subsidiary one time but not a second time (think of the situation between the EU and Greece at the time of writing these comments, 1st February 2015). There are just too many factors that play a role to decide if implicit support is high or low and what it means for the transfer prices of the group, not only for the granting of intercompany loans and guarantees, but also for the sale of goods or any other intercompany transaction.

c) The intercompany sale of a business (or a part of a business) and how the price of the business (or individual assets) is calculated. There are many factors that are different in a third party situation than in an intercompany situation. One very simple example is that in a third party situation, the buyer performs a due diligence (which may be more or less well carried out, very detailed or less, depending on many factors) and uses some forecasts, which maybe similar or very different from those used by the seller (depending on the amount of information they have from the seller, but also what the buyer is planning to do with the business).

Between third parties, the fair market value of the business is only a starting point for the negotiation. The respective bargaining positions (with scarcity being an important factor) play a key role. In an intercompany situation, trying for instance to introduce the concept of the “bargaining position of the parties” and the “options realistically available” is theoretically possible and intellectually challenging, but in most situation leads to very subjective or at least not univocal results.

The same way that for allocating synergies that are not “the outcome of concerted action” (as mentioned in the revised Chapter 1, D.8) we have arrived at a simplified consensus approach, that is, agreeing that they do not need to be allocated, for some of the most difficult transfer pricing questions like the ones mentioned above we may need a compromise/simplified solution, otherwise applying the ideal arm’s length principle will be impossible in practice.

When we apply the arm’s length principle today to a straight-forward intercompany sale of goods, we accept that an imperfect comparable is better than no comparable at all. We also accept that if we apply the RPM or TNMM making the comparison to independent distributors of similar goods the comparison is not perfect because independent distributors are always smaller, may have a different business strategy, or may face slightly different economic circumstances (e.g., there are many studies showing that MNCs pay higher wages than SMEs). Yet, we accept the comparison.

Why are we then trying to apply higher standards (in terms of comparability and sophistication of the analysis) to more complex transactions like intangibles and transactions that do not happen often between third parties?

I understand that there could be situations where we should be talking about mispricing, but this cannot be the general tone or assumption of the OECD Guidelines. The compliance burden cannot be the same for a company that is just trying to comply and establish a reasonable and actionable transfer pricing policy and for a company that is trying to play on the border lines. This needs to be clear in the OECD Guidelines, because, now, too many chapters and examples are based on the assumption that the objective of MNEs is to manipulate the facts to arrive at a particular result.

Comments to Section D.2, D.3 and D.4

I believe it is reasonable to assume that associated enterprises are acting collaboratively and, on a consolidated basis, do not have different risk preferences. As I mentioned above, the degree of collaboration and the extent to which different subsidiaries may have different risk preferences, depend, like the degree of implicit support, on many factors: strategy of the group, organization (matrix, regional, product lines, centralized, decentralized, etc.), management style, sector in which they operate, business model, country where the company originated from, country of operations, mode of expansion (organically, through acquisitions, mergers, etc.), even the culture of the MNC matters.

How do we take all these factors into account in a transfer pricing analysis? We simply will not be able to, other than in a very limited and imprecise manner. We will need to agree on some simplifications to address this. Like the simplification to assume shareholder costs cannot be recharged, and the benefit of implicit support cannot be charged, and synergies do not need to be charged, safe harbours, etc.

On the issue of moral hazard, it is interesting to note that the OECD appears to look at principal-agency theory with a very selective view, when in fact the role of governance appears to be largely ignored, and emphasis is placed on the “agents” (management and employees) and less so on the principals (whether that is the capital owner or the Board of Directors, etc.).

How should the observation in paragraph 67 that unrelated parties may be unwilling to share insights about the core competencies, for fear of losing intellectual property or market opportunities, affect the analysis of transactions between associated enterprises?

This point was addressed before in an earlier draft of Chapter 9, with the question whether third parties would outsource or sell their “crown jewels”. These paragraphs were dropped from the final version. The answer to this question is: it depends. It depends on the price a third party is willing to pay, it depends on the economic circumstances, and it depends on the alternatives available. We see many examples in real life of companies that have divested their core business. Nokia, which started as a wood pulp manufacturing company, is a typical business book example. They moved from wood to telecommunications, to software, to services. There are many other examples like this. Most recently, the leading manufacturer of fully electric cars, opened their patents to competitors to try to accelerate the move to electric car manufacturers. Also, if we look at the economic theory of clusters and how they develop and grow, we find many examples of cooperation versus competition between third parties. In other words, the strategy to protect certain intangible property or core competences is unique to each market player and it is not possible to try to replicate each possible option in an intragroup context.

Arm’s length principle, moral hazards, formulas, crown jewels, control over risk, DEPME functions (development, enhancement, protection, maintenance and exploitation -of intangibles, that is), special measures, very special indeed. We have started to develop a unique transfer pricing jargon and we run the risk of getting lost in semantics.


Leave a Reply

Your email address will not be published. Required fields are marked *