Comments on the OECD's Discussion Draft on BEPS Action 7: Preventing the Artificial Avoidance of PE Status

Brussels, 12 June 2015

To: Head, Tax Treaties, Transfer Pricing and Financial Transactions Division

Submitted by email to:

Dear Marlies,

EBIT welcomes this opportunity to provide comments to the OECD on the revised Discussion Draft on “BEPS Action 7: Preventing the Artificial Avoidance of PE Status” (hereinafter “the Discussion Draft”).

  • EBIT generally supports the OECD’s ongoing work to address artificial avoidance of PE status and to eliminate double non-taxation through PEs and ensure that tax is properly charged on activities that are valuable and integral to MNC’s business operations. Whilst we support many of the recommendations and clarifications in the Discussion Draft, we also have a number of serious concerns.
  • EBIT’s Members note in particular that the OECD’s latest revised proposals replace the OECD’s earlier alternative approaches on a number of significant PE issues by a set of specific and definitive proposals which are largely focused on widening the scope of the dependent agent rule (and narrowing the scope of the independent agent rule) and narrowing the scope of the specific activity PE exemptions, and preventing abuse of the PE rules by segregating activities across associated entities through the proposed anti-fragmentation rule. All in all, these new proposals will in EBIT Members’ views very significantly expand the scope of existing PE rules.

A. Artificial avoidance of PE status through commissionaire arrangements and similar strategies

  • EBIT’s Members are concerned about the proposed amendment to the Commentary on Article 5 (5), which deems a PE to exist if a dependent agent concludes contracts or negotiates material terms of the contract. In particular, in the new paragraphs 32.4, 32.5 and 32.6, this will mean that agents are regarded as falling within the rule when their activity does not involve a substantive role in the process for the conclusion or negotiation of contracts. We consider this to be contrary to the revised Article 5 (5) itself. EBIT also urges the OECD to provide further clarification in the Commentary that, in all cases, the agent must always play a real and substantial role in the process of negotiation or conclusion of contracts if the threshold test in Article 5 (5) is to be met.
  • We agree that transfer pricing should be the preferred methodology to address any BEPS concerns arising from limited risk distributors and welcome the clear statement in the Discussion Draft that changes to the dependent agent rules are not intended to apply. However, in our view the proposed wording could still lead tax administrations to take different approaches, with some arguing that the wording of Article 5 (5)(d) catches low risk distributors and others interpreting the treaty in line with the Commentary. The OECD should provide clear guidance on this important point through an amendment to Article 5.
  • EBIT welcomes the revision to the independent agent test that an agent acting exclusively for one unrelated principal is not automatically excluded from qualification as an independent agent. This test should focus on whether the agent is in any particular case carrying on the business of the principal. The focus should also be on the nature of the agent’s activities, not the relationship between agent and principal (i.e. whether or not related), otherwise businesses would be forced to choose an unrelated party over its own affiliated entity to avoid adverse tax consequences where both agents are fulfilling the same function.
  • EBIT is also concerned that there is still no clear definition in the OECD guidance of what constitutes an independent agent, and the proposed revisions are in our view not helpful in this respect. In our view, to take away the economic independence test will put far greater emphasis on the “legal” independence test (which looks at the operational independence of the agent however) and will shift the focus from an “entity” perspective in relation to the agent to an “individuals” perspective, namely on the people carrying on the agency functions. EBIT’s Members urge the OECD therefore to clarify the exact requirements of the independent agent test (and their rationale).

B. Artificial avoidance of PE status through the specific activity exemptions

  • We welcome the fact that the OECD acknowledges in paragraph 21 in the Commentary to Article 5 that the activities listed in Article 5 (4) are in many cases, but not all, of a preparatory or auxiliary character. We note that the Discussion Draft clarifies that Working Party 1 has concluded that situations that give rise to BEPS concerns need to be addressed, but we are concerned that the proposed measures will go much further in practice and will also affect traditional businesses with a dramatic increase in PEs as a result, and alter the existing international standards on the allocation of taxing rights on cross-border income.
  • However, EBIT is very concerned that the scope of the “anti-fragmentation” rule appears to have been significantly stretched since the earlier Discussion Draft. The new examples, in particular, suggest a shift from an anti-avoidance measure to the introduction of a “force of attraction” where a group has incorporated a subsidiary company in a jurisdiction, and, as a consequence, this would shift the balance of taxing rights between source and residence states.
  • Also, since a PE has until now been recognised per legal entity, the establishment of a PE at the level of “connected entities” is in our view a sea-change to the current rule (in principle valid in all areas of international tax), and one which will trigger many issues that must be carefully considered before implementation (e.g. assume that related companies A and B each perform activities in State C, which are viewed as a “cohesive business operation”, the question arises under which Tax Treaty will the PE qualification be done? Even if Tax Treaty A-C and Tax Treaty B-C include the revised wording, both legal entities will be considered to have a PE in State C, If A already has a PE in C, will B then be “contaminated” for PE qualification purposes. Also, in the above example, how will the taxable income be calculated? Will this be done in the same way as when the PE were recognised by legal entity (i.e. on a stand-alone basis)?
  • The first sentence of proposed paragraph 30.1 states that the purpose of the anti-fragmentation rule is to “prevent an enterprise or a group of connected enterprises from fragmenting a cohesive business operation into several small operations in order to argue that each is merely engaged in a preparatory or auxiliary activity”. In our view, this rule aims to prevent MNCs taking all the aggregated activities performed in a jurisdiction out of the scope of local taxation by fragmentation, in a situation where those activities, taken together, would exceed the PE threshold (and so be subject to tax). We agree that (with appropriate safeguards) effective rules to prevent this should encompass activities performed in a jurisdiction by all group companies resident outside that jurisdiction (not just activities performed by a single non-resident enterprise). However, EBIT Members consider that great care needs to be taken before proposing the automatic inclusion, in that aggregation, of activities performed by related parties resident in the jurisdiction. This is because an incorporated company carrying out business in its jurisdiction of residence/incorporation will almost certainly exceed the PE threshold by itself. Including the activities of that company in an anti-fragmentation rule pre-judges the outcome of whether the threshold is reached, and turns the anti-fragmentation rule into a force of attraction principle. It effectively means that where an MNC has a subsidiary in a source state, any activity at all, however minor, which is carried out in that state by a related party will trigger a PE of that related party.
  • The new proposal in our view creates a presumption that there is a PE and makes it necessary to consider the test in every situation or, in the absence of qualifying for the 5 (4) exemption, would at least lead to a greater incidence of disputes over whether the related party’s premises constitutes a fixed place of business at the taxpayer’s disposal in the source territory. Tax administrations could easily inappropriately interpret against the taxpayer. EBIT Members therefore recommend that an anti-fragmentation rule only apply to the aggregated activity of non-resident related parties in order to retain the coherence of an anti-fragmentation rule as a threshold test. If the activity of residents is to be included as well, EBIT Members call on the OECD to include a carve-out where the combined activity of the non-residents is minimal compared to the (taxed) activity of the residents. This carve-out would help prevent the proliferation of low value PEs. If the anti-fragmentation test is done without carving out the resident companies, it will end up in many cases in a PE. .In our view, the problem with the anti-fragmentation rule is that with an end-to-end global supply chain, any two parts of it which happen to be in the same country will almost certainly be presumed by a tax administration to being complementary and cohesive. The anti-fragmentation rule should not operate to add two separate business operations together in a territory just because they are part of an MNC that has an end-to-end supply chain. This is to be distinguished from the situation where a distribution entity is broken down into its component parts, say, with a view to splitting out bits that might qualify as incidental when there is no rational business reason for not having them in the same legal entity in that territory other than the potential tax benefit.

    In practice, it precludes an MNC from having a consignment manufacturing arrangement (CMA) with affiliates in a country where the produced goods are on sold, notwithstanding the exception of Article 5 (4). This new PE definition, together with the cohesive character of the activities, may lead to artificial and inefficient structuring of global supply chains.
  • Also, the concept introduced in the sentence in the example on page 33 of the Discussion Draft: “the definition of permanent establishment is not limited to situations where a resident of one Contracting State uses or maintains a fixed place of business in the other State; it applies equally where an enterprise of one State uses or maintains a fixed place of business in that same State” seems to be a new definition which EBIT urges the OECD to clarify

E. Profit attribution to PEs and interaction with action points on transfer pricing

  • EBIT Members are gravely concerned that the OECD’s proposal for follow-up work on the attribution of profits issues will in practice make it very difficult to address PE threshold and PE profit attribution issues together, which we strongly believe is required. We are particularly concerned that the ongoing work on Article 5 which involves lowering the existing PE threshold will lead to a large number of new low-value PEs where is no additional profit at stake - meaning significant compliance burden, and also many potential disputes over profit allocation (with resulting risk of double taxation), compared with what may be referred to as the corresponding transfer pricing arrangements that would typically otherwise apply in the absence of a PE. EBIT welcomes the commitment to further work on profit allocation and encourages the OECD to take that work forward swiftly. However, if the work on the profit attribution issues cannot be accelerated, we would recommend that final decisions on Article 5 be taken only when the results of the work on the profit attribution are available.
  • EBIT trusts that the above comments are helpful and will be taken into account by the OECD in finalising its work in this area. We are happy to discuss and remain committed to a constructive dialogue with the OECD.

Yours sincerely,

European Business Initiative on Taxation – June 2015