Tax Treaties, Transfer Pricing and Financial Transactions Division
Centre for Tax Policy and Administration
Organisation for Economic Co-operation and Development

1. General Comments

Clarifying the treatment of the transactional profit split method is one of the two most important remaining issues of the Base Erosion and Profit Shifting (BEPS) Project in our opinion, the other being the attribution of profits to permanent establishments. We therefore welcome publication of the proposed revisions to the OECD Transfer Pricing Guidelines.

The transactional profit split method is among the OECD-approved transfer pricing methods. It has already been applied by some Japanese companies to transactions with parties in certain countries, including cases where the method was applied based on mutual agreement through a bilateral advance pricing arrangement. The transactional profit split method can help provide greater certainty for taxpayers if they apply the method of their own volition based on sufficient information about a transaction, in a case where its application is appropriate.

Applying the transactional profit split method, though, entails a number of difficulties, especially in terms of accessing information on foreign affiliates, measuring revenue and costs segmented by transaction, and determining an appropriate profit splitting factor. Our experience tells that there is a material risk that arbitrary enforcement by tax administrations could lead to disputes that cannot be resolved even through mutual agreement procedures such that double taxation remains unsettled.

Another concern is that this method involves the notion of allowing the prices of foreign controlled transactions to be determined without using comparables. This notion, which is close to so-called Formulary Apportionment, could make transfer pricing rules based on the arm's length principle ambiguous. It is for these reasons that we have long voiced concern about unlimited expansion of the application of the transactional profit split method and pointed out that the refinement of guidance on the method is essential.

This Public Discussion Draft suggests that the circumstances to which the transactional profit split method is applicable remain extremely limited. We appreciate this stance. In particular, from the standpoint of continuing to maintain the most appropriate method approach, we support the statement in paragraph 18 that reads, "A lack of comparables alone is insufficient to warrant the use of a transactional profit split of actual profits." Another important point is paragraph 3, which sets out that, when applying the transactional profit split method, care should be exercised "in order to avoid the use of hindsight."

Still, the Public Discussion Draft contains multiple issues that require further explanations and/or more examples. Below are our comments on those specific issues.

2. Comments on Specific Issues

(1) Splitting of actual profits and of anticipated profits

The Public Discussion Draft highlights the contrasts between the splitting of actual profits and that of anticipated profits in a clearer way than the current guidelines, with particular emphasis on the splitting of actual profits that is explained in greater detail. Compared to the splitting of anticipated profits, the splitting of actual profits requires a greater degree of risk sharing between parties to a transaction, which arguably makes this approach all the more difficult to apply. If focusing on that effect, emphasizing the clear distinction between the two approaches can be said to be useful.

However, the appropriateness of the example in paragraph 5 is difficult to judge. This example pertains to the splitting of actual profits. The situation depicted is that Company A transfers rights in an intangible to Company B, and vice versa, with the result that each company commercializes a product using these intangibles in combination. However, from the perspective of Japanese MNE groups, which basically develop and own important intangibles at a parent company, transactions of this kind are rarely expected in practice. Also, in view of the manufacturing and sales functions that both Companies A and B appear to fulfil, the application of the transactional profit split method may not be needed in the first place; rather, other one-sided methods may provide an appropriate solution, depending on the weight of value created by each function.

At the same time, the splitting of anticipated profits requires further guidance. The Public Discussion Draft describes in paragraph 4 the case of a foreign controlled transaction where one party transfers rights in an intangible to the other. It then concludes that the price of those rights is determined by applying to the transferee's anticipated profits the transactional profit split method in conjunction with a discounted cash flow valuation technique. However, it is unclear how the profits are split, including measures to be taken in the event of ex post result deviating from the anticipated profits. As this issue seems to relate to the task of developing guidance on the commensurate with income standard, too, we hope that numerical examples will be provided.

Furthermore, we urge each country to recognize once again the significance of the statement in paragraph 1 of the Public Discussion Draft, which reads, "References to 'profits' should be taken as applying equally to losses." We have seen some jurisdictions take an inconsistent approach to transfer pricing. Specifically, in the case of a MNE group, when the group's profit margin is higher than that of a company under review, those jurisdictions apply the transactional profit split method in order to allocate more profits to the company; on the other hand, when the group operates at a loss or generates lower profit margin than the company, those jurisdictions apply the transactional net margin method so as to ensure minimum profit for the company. Such an inconsistent practice should be refrained from and clearer guidance on this issue would be desirable.

(2) Strengths and weaknesses of transactional profit split method

Paragraphs 11 to 15 of the Public Discussion Draft outline the strengths and weaknesses of the transactional profit split method on the basis of the current guidelines. We consider that the method's weaknesses, especially difficulties in its application, are well summarized. In addition to those outlined, we would like to point out as another weakness that the method tends to cause disputes between taxpayers and tax administrations over the existence of an intangible and the measurement of its value in the transaction. Further, as paragraph 15 rightly states that "in most cases a tax administration will not be able to perform the analysis or verify the information without full co-operation from the taxpayer," the method imposes a significant burden on taxpayers, which should also be recognized anew.

A weakness also exists in that different ways of applying the transactional profit split method in practice could result in considerably different arm's length prices. Precisely because of such deviation risk, the OECD refers to this approach as transactional, not corporate, profit split method. Attention needs to be paid again to that fact as well.

(3) Sharing of risks

The Public Discussion Draft proposes the "sharing of economically significant risks" by parties to a transaction be regarded as a factor indicating that a transactional profit split of actual profits may be the most appropriate method. Although this concept seems to provide a potentially useful analytical framework, it needs further improvement and clarification.

For instance, paragraph 16 of the Public Discussion Draft could allow broad interpretation of this concept by stating, "The application of a transactional profit split of actual profits reflects a relationship where the parties either share the same economically significant risks associated with the business opportunity or separately assume closely related risks associated with the business opportunity and consequently should share in the resulting profits or losses."

Firstly, whereas the paragraph refers to "share the same economically significant risks," all member companies of a MNE group share more or less risks, and what are economically significant risks partly depends on how the facts are interpreted. We are concerned that, if attention is focused on the mere fact that risk is shared, the transactional profit split method may be applied more broadly than it should be.

Rather, if the sharing of economically significant risks takes place to the extent that justifies the application of the transactional profit split method, that situation may be rephrased more specifically as the "sharing of the function of controlling economically significant risks." In that situation, it would be natural that the outcomes of the business activities are shared as well.

Secondly, as to the phrase "separately assume closely related risks," each stakeholder may differently interpret the words "closely related." For example, even if the tax administration of one country considers that the parent's development and manufacturing risks are totally different from the market risk assumed by the subsidiary, the tax administration of another country may look at the group as a whole and deem the parent and the subsidiary to separately assume closely related risks. Concrete examples of closely related risks need to be given.

Our third concern is based on the fact that companies belonging to the same MNE group naturally share in the outcomes of the business activities. In that context, if the resulting profits or losses are used as the starting point of analysis, tax administrations may readily conclude that risks are being shared. Whereas it is true that latent risks are difficult to identify, guidance should be clarified to ensure that analyses are not solely based on the outcomes of business activities. Clarification is thus needed in paragraph 9 as well that reads, "The division of combined actual profits . . . requires that the parties share in the outcomes of the business activities and risks associated with those outcomes."

(4) Highly integrated business operations

Paragraph 21 of the Public Discussion Draft suggests that, when a transaction is part of highly integrated business operations of the parties thereto, there is a possibility for those parties to share significant risks. The paragraph goes on to present the concept of sequential integration and parallel integration in a value chain, concluding that highly integrated business operations are more likely to be the case where there is parallel integration.

Suppose that a parent and its subsidiary have no commonality in the functions performed, the risks assumed, and the assets used, each of the two clearly playing a different role. In that case, sequential integration can be thought to occur in their value chain, including relationships in development and enhancement as well as in manufacturing and sales. In the sense that the application of the transactional profit split method to cases like this would be limited, the distinction between sequential integration and parallel integration is likely to be of some significance.

In some MNE groups, the parent's business division and an overseas sales subsidiary work together in managing the production, sales, and inventory processes throughout the value chain with the aim of minimizing inventory levels and maximizing earnings. This means that, conceptually, there are cases in which sequential integration and parallel integration coexist. Even in such cases, however, if the contribution of the subsidiary is not deemed equal to that of the parent, the transactional net margin method and other one-sided approaches would remain sufficient to address the situation.

(5) Unique and valuable contributions

The Public Discussion Draft states in paragraph 19 that a sharing of risks by parties to a transaction may be accompanied by the making of unique and valuable contributions by each of the parties. Then in paragraph 22, it defines "unique and valuable contributions" as contributions that are not comparable to those of uncontrolled parties and represent a key source of economic benefits.

This definition still leaves room for arbitrary interpretation. An example is the case of a sales subsidiary that performs routine functions only. Even in this case, the tax administration of the country of residence of the subsidiary may assert the existence of a marketing intangible, thereby determining that unique and valuable contributions are made.

In cases like this, it is important not to be quick to rely on the transactional profit split method; instead, the first step should be to examine existence and value of the intangible, and carefully consider the feasibility of using the transactional net margin method or its adjusted versions. We expect the OECD to continue working on the enhancement of the transactional net margin method, including comparability adjustments, while endeavoring to refine guidance on the transactional profit split method.

In the example mentioned above, even if the sales subsidiary does make unique and valuable contributions, it would be a totally different argument whether the subsidiary "shares economically significant risks" (which, in our opinion, should be referred to as "shares the function of controlling economically significant risks") with the parent. In general, the making of unique and valuable contributions does not seem to correlate positively with the sharing of economically significant risks.

(6) Group synergies

Paragraph 23 of the Public Discussion Draft reads, "There is no need to combine the total profits of the parties and use the transactional profit split method simply on account of group synergies alone." It seems that this conclusion is reasonable. However, identifying and extracting the marginal system profits arising from group synergies is difficult in practice, especially when those synergies are unquantifiable, qualitative ones. We would suggest that concrete examples be given, particularly on how to allocate such profits to group companies.

(7) Value chain analyses

The Public Discussion Draft regards value chain analyses as a tool to assist in delineating controlled transactions. We think that paragraphs 24 to 27 explain, to a certain extent, how those analyses relate to the transactional profit split method. However, each company has a different value chain. An MNE's value chain is highly complex in that it includes unrelated parties, rather than consisting entirely of associated enterprises. Attention should be paid to the fact that creating its simple model is difficult.

Value chain analyses require items that are similar to the contents of the master file, as represented by functional and risk analyses of key value drivers. We understand that the Public Discussion Draft does not go so far as to recommend that value chain analyses be mandated in each country. This point should be swiftly made clear since there is a certain jurisdiction that has already put in place the legal requirement.

(8) Profit splitting factors

We welcome expanded guidance on cost-based profit splitting factors. For example, with regard to the risk-weighting of costs, paragraph 51 indicates that risks at the development stage are higher than those at the improvement stage. This is consonant with what manufacturers actually feel in business settings.

The remaining issue of this section is to provide examples of calculations for weighting risks, adjusting employee compensation costs, and allocating location savings. Such examples are particularly needed on the issue of location savings, because paragraph 52 states that "the manner in which independent parties would allocate retained location savings would need to be reflected in the profit split." Without concrete guidance on how to treat location savings including in comparability analyses, the interpretation could vary from one country to another.

Sincerely,

Subcommittee on Taxation
KEIDANREN