Andrew Hickman
Head of Transfer Pricing Unit, Centre for Tax Policy and Administration
Organisation for Economic Co-operation and Development

Keidanren hereby submits its comments on the Public Discussion Draft "BEPS Action 8: Revisions to Chapter VIII of the Transfer Pricing Guidelines on Cost Contribution Arrangements" published by the OECD on April 29, 2015.

Cost contribution arrangements (CCAs) have been used by some multinational enterprises to effectively transfer intangibles from the country in which they are headquartered to a lower-tax jurisdiction and thereby unjustly reduce the total tax burden of the corporate group. This is pointed out as one of the causes of base erosion and profit shifting (BEPS). Keidanren supports the OECD's work on the revisions to Chapter VIII of the Transfer Pricing Guidelines, from the perspective of preventing base erosion and creating a level playing field for companies.

Of particular importance is the descriptions of risks in this Public Discussion Draft. As seen in the "cash box" scheme, when an entity merely plays the role of providing funding for a research and development (R&D) project and has no capability or authority to control the risks associated with the CCA, it is not appropriate for the entity to receive benefits from the CCA, including the acquisition of an interest in the resulting intangibles. We therefore concur with the essence of the proposed revisions. Japan's administrative guidelines on transfer pricing also provide that "degree of contribution shall be assessed as low when the corporation or the foreign-related party merely bears the cost of the formation and so on of the intangible properties that are highly expected to be the source of income".

However, as for the proposed rule of measuring contributions made by CCA participants at value rather than at cost, we are uncertain whether it will work in practice, even though it seems to be theoretically correct in light of consistency with the guidance on intangible transactions not covered by a CCA, would be effective in preventing BEPS, and is conceptually comprehensible.

In Japan, although many companies frequently conclude R&D outsourcing contracts with associated enterprises, entering into CCAs for transfer pricing purposes is not that common. There are a number of conceivable reasons for this. One is that, in many corporate groups, their intellectual property rights tend to be managed almost exclusively by the parent company domiciled in Japan. Another reason is the absence of incentives to transfer intangibles to a lower-tax jurisdiction without a legitimate business reason. But, the dominant reason seems to be the uncertainty perceived in the method of CCAs whereby costs are allocated in accordance with the participants' proportionate shares of expected benefits.

If, despite such circumstances, contributions were to be measured not at cost but at value (i.e., the arm's length price) with the exception of some low value-added services, companies might accordingly suffer a heavier administrative burden and undergo more disputes with tax administrations over the validity of the value measured. Tax consequences would become particularly unpredictable in cases where the proportionate shares of the benefits produced by the intangibles resulting from a CCA turn out to be different from the initial expectations or where participants in a CCA spread over multiple (three or more) countries. These might make the already high hurdle of using CCAs even higher going forward.

To enhance competitiveness, companies are considering the implementation of various types of R&D projects. Their future projects may include joint R&D projects that are run by the parent company and its overseas subsidiary or that use the intangibles of a multinational enterprise the company acquired. It is therefore fairly possible that companies will enter into CCAs in the future. Paragraph 6 of the Public Discussion Draft states that an advantage of CCAs lies in the simplification of multiple transactions for transfer pricing purposes, such as by offsetting payments against receipts and vice versa with regard to royalties associated with the cross-licensing of intangibles. Care needs to be taken to ensure that the revisions to Chapter VIII do not inadvertently deter companies from using CCAs and jointly engaging in open R&D activities on a global scale.

We therefore suggest that, in the process of finalizing the revisions to Chapter VIII, additional consideration be given, in particular, to the following matters:

The first is the creation of safe harbor provisions, which is significant in two aspects. The one aspect relates to the fact that the revisions to Chapter VIII are being made as part of the BEPS Project. Given this background, we consider it rational to limit the subjects of heavily detailed transfer pricing analysis, especially the measurement of contributions at value, to those CCAs that pose the high risk of BEPS. They may include CCAs with an entity domiciled in a lower-tax jurisdiction, and "important" CCAs involving the intangibles required to be reported on in the master file. Requiring each and every CCA to be subjected to measurement at value is excessive. At least, measurement at cost should be permitted in certain cases. As described in paragraph 23, in a case where the contribution consists of low value adding services, such a contribution should be measured at cost.

The other aspect relates to the possibility that questions may subsequently arise as to the appropriateness of the value of contributions measured, depending on the level of benefits resulting from the CCA. Even in that situation, the value measured and the benefits expected initially should be respected as much as possible. Subsequent corrections, if at all necessary, should be limited to highly exceptional cases, and the conditions those cases must satisfy need to be clearly defined. (It is also conceivable to take such measures as deeming deviations up to a certain percentage to be acceptable in principle.)

CCAs have the very nature of an advance agreement between the parties for transfer pricing purposes, stipulating the value to be contributed and the benefits to be received. Given that nature, CCAs are considered to have an affinity to advance pricing arrangements (APAs). We thus suggest that the Transfer Pricing Guidelines include a statement that recommends obtaining bilateral and multilateral APAs.

The second matter to which consideration should be given is the methods of measuring value. This Public Discussion Draft only states in paragraph 7 that Chapter VI, which deals with intangibles, and other chapters of the Transfer Pricing Guidelines should be referred to, and the existing Guidelines provide no clear guidance either. This is not necessarily adequate from the perspective of encouraging the use of CCAs. In the case of joint R&D projects involving intangibles, identifying comparable transactions as a benchmark is especially difficult due to the unique nature of each project. Although it may be one of the options to use, for example, the discounted cash flow method for measuring value, problems remain as to arbitrariness and subjectivity. In such cases, we assume that other approaches, including a combination of multiple methods, need to be considered, but hope that these Guidelines will recommend the best practices.

From that standpoint, we highly appreciate the addition of Examples 1 to 5 by this Public Discussion Draft, but it should be even more useful to include a greater number of examples in these Guidelines. As partly detailed later, additional examples are needed to illustrate the treatment of cases in which the proportionate shares of expected benefits have subsequently changed, of cases of non-recognition, of cases in which a CCA has been terminated without producing a successful outcome, and of cases of buy-ins and buyouts, among other things.

The third matter is the interactions with other chapters of the Transfer Pricing Guidelines. This Public Discussion Draft pays attention to consistency with the revisions to Chapters I (The Arm's Length Principle) and VI (Special Considerations for Intangibles) made as part of the BEPS Project. A question however arises as to how the interaction with Chapter II (Transfer Pricing Methods) would change if contributions to CCAs were to be measured at value rather than at cost. Take paragraph 2.138 of these Guidelines that explains how to split combined profits under a residual profit split method. While the paragraph states that "an allocation key based on expenses may be appropriate" that include R&D expenses, we are concerned about how much impact the revisions to the guidance on CCAs would have on the current transfer pricing practices as a whole, including those being implemented in compliance with Chapter II.

The fourth is balancing payments. In paragraph 27 of this Public Discussion Draft, those payments are defined as an adjustment to the value of participants' contributions under a CCA. Example 4, however, seems a little bit confusing, because the balancing payment from Company A to Company B (equal in present value to $220 million per annum in years 6 to 15) appears to be an adjustment not to the value of Company A's contributions but to its "profits," the term used synonymously with benefits. We hope that this example will explicitly explain the mechanism of triggering balancing payments, that is, the system under which a change in the shares of benefits triggers a revision to the shares of contributions.

The fifth is the treatment of non-recognition as a CCA participant. As was discussed concerning the Public Discussion Draft on "Risk, Recharacterisation and Special Measures", the consequences of non-recognition are not always clear. Take Example 5 of this Public Discussion Draft, in which Company A is, under the CCA, anticipated to provide funding of $100 million per annum for the first five years of the project and to receive profits of $330 million per annum in years 6 to 15. While the example simply concludes that Company A cannot be regarded as a participant in the CCA, further explanation is needed to illustrate what adjustments for tax purposes are required of Companies A and B, respectively, in that case.

Last but not least, the OECD has repeatedly argued in relation to BEPS Actions 8 to 10, including in this Public Discussion Draft, that transfer pricing outcomes should be in line with value creation. We have no objection to this argument per se. However, we are gravely concerned that no progress has been seen on dispute resolution mechanisms despite the fact that, without consensus among the member countries as to what constitutes value, discussions are moving forward on taxation methods, such as the introduction of non-recognition and special measures, and the expanded application of profit split methods.

Keidanren believes that in general, the source of value creation of many corporate groups lies in the parent company's R&D function, rather than the overseas subsidiaries' activities for marketing products and enhancing intangibles, at least in the manufacturing industry. We emphasized this point in our comments concerning the Public Discussion Draft on "Profit Splits", and hereby confirm that this position of ours remains unchanged (i.e., CCA should not be used in a way that inappropriately allocates income to an entity which does not contribute to value creation). We strongly hope that, as the deliverables of BEPS Action 14, the OECD will present truly effective dispute resolution mechanisms that include the introduction of a mandatory and binding arbitration provision.

Sincerely,

Subcommittee on Taxation
KEIDANREN