Policy Proposals  Business Law   Comments on the Public Discussion Draft on BEPS Action 3 (Strengthening CFC Rules)

April 30, 2015

Dr. Achim Pross
Head of the International Co-operation and Tax Administration Division, CTPA
Organisation for Economic Co-operation and Development

Comments on the Public Discussion Draft on BEPS Action 3
(Strengthening CFC Rules)

Keidanren hereby submits its comments on the Public Discussion Draft "BEPS Action 3: Strengthening CFC Rules" published by the OECD on April 3, 2015.

1. Introduction

Keidanren supports the OECD's work on Action 3 from the perspective of developing truly effective and efficient controlled foreign company (CFC) rules as measures against base erosion and profit shifting (BEPS), and of creating a level playing field for companies. For instance, the Double Irish Dutch Sandwich structure is regarded as a typical BEPS technique. While this structure is made up of various elements, it has been argued that had there been robust CFC rules in the parent jurisdictions, the structure's effects could have mostly been nullified. As a representative of the business community of Japan which reputedly has one of the world's strictest CFC regimes, we believe that creating a level playing field for companies requires, at first instance, for jurisdictions with inadequate CFC rules to revise such rules. As a following step, it is hoped that efforts will be made to minimize differences among the CFC rules of individual jurisdictions to the extent possible, giving primary consideration for the need to prevent and eliminate any double taxation resulting from subjecting taxpayers to the CFC rules of more than one jurisdiction. At the same time, a certain degree of relaxation of CFC rules is a highly feasible option for jurisdictions with over-inclusive rules.

On the other hand, CFC rules in various forms have been already introduced in numerous jurisdictions and firmly established as part of their systems. Therefore, simply trying to standardize national rules in a uniform manner will not work. For example, the European Union nations have to ensure the consistency of their CFC rules with the European Court of Justice's case law, as mentioned in the Public Discussion Draft. Given that a large number of OECD countries are also members of the EU, it is expected that the OECD's recommendations have to be made with the EU in mind. Still, thorough examination is required as to whether those recommendations are suitable for the existing rules of non-EU countries as well.

With respect to the definition of CFC income, we have no objection to the proposed policy that a partial-inclusion system should be employed for the purpose of efficiently combatting BEPS. Nevertheless, a wider variety of approaches, including an entity approach, should be allowed as a means to achieve policy objectives. In particular, an entity approach that refines the range of application of CFC rules on an entity basis prior to applying substantial analysis provides a simple solution for both tax administrations and taxpayers because the approach allows an easy determination of whether a company is subject to a CFC rules.

In the current environment in which CFC rules in various forms are enforced globally, what is required of the OECD is not to recommend a single best practice, but to define essential principles of CFC rules that are truly effective in combatting BEPS. Furthermore, any CFC rules enforced must be clear and workable, take into consideration companies' compliance costs, and ensure that they are not applied subjectively.

Based on the viewpoints stated above, we present below our comments. It should be noted that these comments are the result of an examination conducted in a short period of time to meet the deadline. The Japanese business community continues to have a strong interest in CFC rules, and hopes that there will be more opportunities to voice our views and opinions in the future.

2. Approaches to Defining CFC Income

(1) Evaluation on Categorical Approach

In comparison with the excess profits approach set out in the Public Discussion Draft, the categorical approach would be consistent with the purpose of CFC rules, as it is designed to categorize the types of income subject to the rules and capture the very income that raises BEPS concerns. The OECD's proposal that active income be excluded from, and passive income be included in, CFC income is understandable as a basic concept. When conducting a substance analysis to determine whether income is active or passive, it is desirable to take into account the business substance of the analyzed company and, if the company presents little BEPS risk, to keep the analysis as simple as possible by using objective factors relating to the company so as to alleviate the administrative and other burdens. However, in the case of companies in the financial and leasing sectors whose income tends to be deemed passive due to the nature of their business, individual jurisdictions should endeavor to harmonize their criteria for determination taking into consideration that such income would have been earned from an active business.

Considering these points and the need for an approach that is as objective as possible, it may be appropriate to adopt a method similar to an employees and establishment analysis when carrying out the substance analysis referred to in paragraph 89.

Paragraphs 106 and 112 of the Public Discussion Draft propose that a substance analysis be conducted with all sales and services income treated as passive. However, sales and services income account for the majority of income of a company having genuine business substance. Hence, we disagree with this proposal as it would excessively widen the scope of income subject to CFC rules in the name of combatting BEPS and substantially increase the administrative burden on companies. We support an active presumption for services and sales income, unless specific thresholds are breached by taking into consideration the active conduct of business by the companies.

Insurance income demands careful consideration based on its nature when determining whether such income is passive or active. Attention should be paid, at least, to the fact that income from certain intragroup transactions and reinsurance deals, among other things, should sometimes be treated as active income due to the uniqueness of the insurance market.

Additionally, whereas no recommendation has been made in the Public Discussion Draft, we cannot find any rational reason why capital gains arising from corporate reorganizations and other events should be treated differently from income such as dividends and interest. For example, in a globalized economy, it is not unusual for a MNE to acquire a foreign MNE in pursuit of growth and competitiveness. In this situation, the acquiring MNE may wish to restructure the corporate holding structure of the acquired company and its subsidiaries in various jurisdictions to create synergies among member companies. It is obvious that there is a bona fide business reason for the acquiring MNE to transfer shares of some subsidiaries to other constituent entities, but in a certain jurisdiction the capital gain arising from that transfer might be classified as CFC income depending on the situation. It should therefore be recommended that capital gains arising from active business operations not be treated as CFC income.

(2) Evaluation on Excess Profits Approach

The excess profits approach, which relies on subjective criteria for determining factors such as the rate of return and eligible equity, is likely to become a source of disputes between tax administrations and taxpayers. In addition, the rate of return creates a risk of tax administrations imposing taxes based on an assumption not in accordance with the substance of business activities. For these reasons, we are not in support of the proposed excess profits approach being recommended as a best practice.

It appears that the excess profit approach is proposed mainly with a view to capturing IP income shifted from a parent jurisdiction to a low tax jurisdiction. However, the proposed Option 1 of the potential special measures in the Discussion Draft on "Risk and Recharacterisation" is also aimed at the taxation of excess returns derived from exploiting IP assets transferred to a low tax jurisdiction. Although the relationship between the two proposed measures is currently unclear, multiple taxation of the same profit should be avoided and we request the OECD to ensure that the final recommendations on Action 3 (CFC) and Actions 8-10 (Transfer pricing) will be mutually consistent.

3. Clear Prioritization of Building Blocks

For the purpose of reducing the administrative burdens on tax administrations and taxpayers, it is extremely important to clearly prioritize the building blocks of CFC rules. This means that the issues able to be determined based on objective factors should be dealt with first while those requiring substantive judgments should be examined last. Therefore, the Public Discussion Draft should explicitly recommend that priority between building blocks be given to the issues that can be determined based on objective factors such as the definition of a CFC, threshold requirements, and the definition of control.

4. Specific Issues

(1) Secondary Rules

We feel that the possibility of secondary rules on page 3 of the Public Discussion Draft could be disruptive. These rules could be construed as permitting third countries to introduce a form of taxation other than the CFC rules proposed in the Public Discussion Draft. While the explanation is not sufficiently detailed for us to judge, generally speaking, it is desirable to avoid a complicating of rules.

(2) Definition of a CFC

Paragraph 34 of the Public Discussion Draft proposes that permanent establishments (PEs) be taken into account for CFC rules in certain cases. However, there are varying national treatments of what constitutes a PE and different views between taxpayers and tax administrations. For that reason, it is desirable to take a PE into account for CFC rules only when it has been registered in the jurisdiction or its treatment has been otherwise clarified.

Another concern is in relation to the acquisition of a company as part of reorganization and other schemes. There have been cases where an acquired company's subsidiaries are captured by the CFC rules as a result of a mechanical application of the CFC rules of the acquiring company's jurisdiction, irrespective of whether there has been any intention to engage in a BEPS technique. We believe that it is not appropriate for the CFC rules to be applied in this manner as no tax base has been eroded. Consideration should be made for corporate acquisitions, for example by allowing a certain grace period during which time the acquiring company can reorganize its group structure without being subject to CFC rules.

(3) Threshold Requirements

We agree with the Public Discussion Draft's recommendation that a low-tax threshold be introduced, but emphasize that such a threshold must be clear and simple.

When establishing a low-tax threshold based on the tax rates of individual countries, that threshold should be meaningfully low in order to focus on CFC jurisdictions where the BEPS risk is expected to be extremely high. Whereas the Public Discussion Draft simply recommends that a low-tax threshold be set at 75% or lower of the statutory corporate tax rate, this does not make complete sense in the current situation in which some countries continue to keep their effective corporate tax rates high. It would be preferable for some countries to set a threshold at 50% of the statutory corporate tax rate of the parent jurisdiction.

In conjunction with this, in order to reduce the burden of examining the effective tax rates of the relevant countries, a white list approach should be introduced that allows the exclusion from such examination of countries posing little BEPS risk.

Paragraph 51 of the Public Discussion Draft states that there is no general recommendation for or against de minimis thresholds. However, from the perspective of enhancing measures against BEPS and reducing administrative burdens, consideration should be given to introducing a de minimis threshold in combination with an anti-fragmentation rule, as suggested by the Public Discussion Draft.

(4) Definition of Control

We basically agree with defining "control" as holding more than 50% control, as recommended in paragraph 65 of the Public Discussion Draft. In principle, it is desirable that the determination of control is made uniformly based on ownership on the last day of the fiscal year. Our concern relates to paragraph 65, which states that the interests of unrelated resident parties are also to be counted in determining the level of control. Assume a CFC is owned by a resident taxpayer and a non-resident partner entity with each holding a 50% stake. In this case, a determination of whether the CFC falls under the definition of control would require an investigation into the shareholders of the partner entity, including unrelated parties. Such an investigation, however, would be very difficult, especially in the event of the partner entity or its shareholders being listed companies that have a diverse range of shareholders. Therefore, the scope of investigations and examinations required of a company should be clarified.

(5) Rules for Computing Income

Paragraph 131 of the Public Discussion Draft recommends that a calculation of a CFC's income be made in accordance with the rules of the parent jurisdiction. However, there are cases where recalculating a CFC's income pursuant to the rules of the parent jurisdiction would impose an excessive burden on the taxpayer that has many CFCs subject to this process. From this perspective, attention might also be paid to certain situations that require the application of the third option of paragraph 132, under which taxpayers are allowed to choose either the computational rules of the parent jurisdiction or those of the CFC jurisdiction.

(6) Rules for Attributing Income

Paragraph 147 presents two approaches to determining how much of income to attribute: one being income based on ownership on the last day of the year, and the other being income based on the period of ownership. We consider it appropriate to attribute income based on ownership on the last day of the year, rather than that based on the period of ownership, from the perspective of accurately calculating and attributing taxable income and reducing administrative burdens.

(7) Rules to Prevent or Eliminate Double Taxation

For CFC rules, preventing and eliminating double taxation is of paramount importance. From this standpoint, paragraph 157 of the Public Discussion Draft refers to a foreign tax credit as a means to eliminate double taxation. Yet, it is more important to work out ways to prevent double taxation from arising in the first place. We are also concerned that relief for double taxation provided for in existing CFC rules is inadequate.

In this regard, paragraphs 159-162 of the Public Discussion Draft recommend that, in the event of a CFC being subject to the CFC rules of multiple jurisdictions, the rules to apply first be those of the jurisdiction whose resident shareholder is closer to the CFC in the chain of ownership. The problem is that, when taxation methods, exemption criteria, and other conditions vary among the relevant jurisdictions, it is highly difficult to identify which part of the income has been taxed under CFC rules, resulting in a high chance of double taxation occurring. To avoid this, consideration may need to be given to an approach under which, in the case of a CFC being subject to the CFC rules of multiple jurisdictions, the jurisdiction whose resident shareholder is closer to the CFC in the chain of ownership is given the sole taxation right.

Sincerely,

Subcommittee on Taxation
KEIDANREN