- Treaty Abuse: BEPS Action 6
- September 17, 2015 | Author: Blaise L. Marin-Curtoud
- Law Firm: Jones Day - London Office
- Action 6 (Treaty Abuse) is a key element of the OECD's BEPS Project. Action 6 targets, in summary, "treaty shopping", i.e., where a person in country A, which is not, in principle, eligible to benefit from a given tax treaty with country B, invests through an entity in country C to benefit from the treaty. More generally, Action 6 intends to prevent the granting of treaty benefits in inappropriate circumstances.
The Original Action 6 Proposals
The original Action 6 proposal (published in September 2014) contained two approaches to the treaty shopping issue (either or both approaches could have been included in the treaties).
One approach was based on the so-called "principal purpose test" ("PPT") where the treaty benefit would be denied if it was viewed as one of the principal purposes of the relevant transaction; the second approach was a "limitation on benefits" rule ("LOB"), following essentially the typical clause included in treaties signed by the US, i.e., that only certain defined tax residents are eligible to the treaty benefit.
Each approach was criticised for its supposed weakness: the PPT was viewed as creating too many uncertainties for the market (i.e., subjectivity test), and the LOB was viewed as lacking flexibility and reflecting US domestic policy concerns. The inclusion of both approaches was a way to compromise and progress on the matter.
The Revised Proposal
Following extensive consultation and discussion, a new draft was released (May 2015) and a simplified LOB version was proposed to be used in combination with the PPT (as an alternative to the full LOB).
The simplified LOB would cover: individuals, governments, publicly traded entities, entities 50 percent or more beneficially owned by the above persons, active business, derivative benefits (i.e., entities owned 75 percent or more by equivalent beneficiaries) and competent authority discretionary relief (the relief being available where there are clear non-tax reasons, a potentially difficult hurdle, although the authorities are asked to deal "expeditiously" with the requests).
The idea behind a simplified LOB is that it may be more easily included in a "multilateral instrument", and the OECD members would then fine-tune it in their bilateral negotiations for their respective treaties. Also, a simplified LOB should be simpler than a full one, to the extent that the PPT could take care of the remaining complexities (e.g., no base erosion test would be needed, since the PPT should exclude the relevant arrangements; also, the intermediary entities may be looked through if the relevant contracting states so decide). Still there are questions as to whether the LOB (even in its simplified format) and the PPT should be tied together, or whether the contracting states should be free to use the method they prefer.
Although the simplified LOB is supposed to be more practical for a number of situations, it is still not helpful for certain vehicles (e.g., securitisation entities) as they would need to be able to identify their beneficial owners (which is, generally, not practical).
For the regulated collective investment vehicles ("CIV"), it was concluded that the flexible findings of the 2010 OECD Report, "The granting of Treaty Benefits with respect to the income of collective investment vehicles", remain relevant (the practical implementation of the 2010 Report being facilitated if the recommendations in the TRACE project (Treaty Relief and Compliance Enhancement) are also implemented). There are still doubts as to whether the proposed language, in respect of CIVs, provides enough strong support for the implementation of the 2010 Report, and to what extent the TRACE project should be part of the implementation package.
While it is generally recognised that at least certain non-CIV funds, such as the securitisation entities, do carry an economic importance, certain member states still feel that non-CIV funds may be used for treaty shopping purposes. Thus, the non-CIV funds may, ultimately, have to rely on the discretionary relief. The situation of the non-CIV funds (which is also problematic under the PPT) may continue to be discussed after September 2015 and beyond as part of the "multilateral instrument" efforts up to December 2016.
For pension funds, it is proposed that they should qualify for the relevant treaty where 90 percent of their beneficiaries are residents of either contracting state, or of a third state if they are entitled to a treaty between the source state and such third state and would be subject to the same or lower WHT rates.
Regarding the PPT, and the related uncertainties, it has been proposed that four new examples be added to provide illustrations on the application of the rule. However, these examples may be viewed as being fairly simplistic, with the consequence that they do not necessarily provide actual guidance. It has been suggested that further examples should be added, including in respect of CIVs (i.e., they should not be challenged on the basis of the PPT given that their first and foremost reason for existence is not tax). It is also suggested that more effort should be put into finding a less subjective criterion for PPT purposes. Separately, it is proposed that advisory panels be formed within the states to advise on the application of the PPT. It has been also suggested that the non-application of the LOB rules should be viewed as a positive factor for the PPT analysis.
Separately, the US Delegate to the Action 6 Working Party has made a proposal to exclude from treaty relief entities benefitting from a "special tax regime" (i.e., regimes resulting in a low effective tax rate), with certain exceptions such as pension funds, charities, etc. The proposal comes very late in the process, and it does raise various issues in terms of combination with the other proposed rules. There is also a question as to whether the point should not be dealt with under other BEPS actions (e.g., Action 5 on harmful tax practices).
A final version of Action 6 is due to be released in September 2015.