The European Commission recently published its decision to initiate a state aid investigation into Dutch tax rulings relating to Nike. The Nike case focuses on EMEA structures (including well-known CV/BV structures) where Dutch group companies with substance pay significant tax-deductible royalties to non-taxed IP holding entities without substance (click here for our previous coverage).
The Dutch tax authorities have entered into Advance Pricing Agreements (APAs) with many taxpayers that have used structures similar to those in the Nike case. If the Commission’s position in the Nike case is continued in its final decision and ultimately upheld by the EU courts, the question arises whether the Netherlands will take action to recover state aid from those taxpayers, a possibility that was recently introduced formally in Dutch tax law.
In this article we explain the Commission’s Nike decision and its potential impact.
Our key observations
The decision focuses on transfer pricing aspects only and does not address the hybrid and other “technical” features resulting in double non-taxation as such.
As in other comparable cases (such as Amazon), the Commission challenges the transfer pricing analysis used for the rulings based on a comparison between the functions, assets and risks (FAR) of the Dutch group companies and the FAR of the IP owners. The commission also refers to OECD guidance. The Commission’s analysis raises some fundamental transfer pricing questions, including the following.
The Commission takes the view that, in the absence of employees or activities, the IP owners should be considered as “bare legal owners” and that, for that reason, tax deduction of significant royalty payments by the Dutch taxpayers is not justified. The Commission categorically dismisses the argument that (past and present) functions of the US group companies must be considered for determining the taxable profits of the Dutch group companies because the US group companies are not a party to the “tested” licence agreements. This position seems to imply that the Netherlands must tax profits that are “undertaxed” by the US, that is, profits that would be taxed in the US, through an adjustment in respect of the cost sharing agreement (CSA) / IP arrangements between the IP owners and their US group companies, if the US had applied the same (OECD) transfer pricing standards. This approach to profit allocation / transfer pricing mismatches seems to be in line with the decisions in, for instance, the Apple, Amazon and Starbucks cases. However, it is difficult to reconcile the Commission’s approach to mismatches in the context of state aid generally.
The primary line of reasoning on the application of the state aid condition – that an “advantage” must be granted – is quite clear. In short, in the Commission’s view an advantage is granted if the “ruling profits” are below the taxable profits determined based on a normal application of the relevant tax rules (that is, the at arm’s length principle under Dutch tax law).
In order for state aid to be identified, an advantage must be granted selectively. The primary position of the Commission is, with reference to case law, that the rulings must be presumed to be selective because of their individual nature. The Commission seems to argue that an actual comparative analysis with taxpayers in a comparable legal and factual situation, is not required when rulings have been issued. Nevertheless, as a fall-back, the Commission concludes that an actual comparison would also result in the conclusion that the selectivity test is met. For this purpose, the Commission first argues that the “reference group” for purposes of this comparison encompasses all Dutch corporate income taxpayers. As a second line of defence, the Commission argues that if the reference group only comprised group companies of multinationals the alleged advantages must still be considered to have been granted selectively. The references to aid scheme in the decision by themselves do not suggest that the Commission holds the view that an aid scheme can be identified from which Nike benefitted. Still, the Dutch tax authorities have entered into Advance Pricing Agreements (APAs) with many taxpayers that have used structures similar to those in the Nike case (including well-known CV/BV structures). If the Commission’s position in the Nike case is ultimately upheld by the EU courts, the question arises whether the Netherlands will take action to recover state aid from those taxpayers, a possibility that was recently formally introduced to Dutch tax law.
Both NEON and CN had significant operations in the Netherlands. NEON acted as regional headquarters of the Nike Group in the EMEA region. Its activities comprised product design, sales management, pricing and discount policies, inventory management, customer services, marketing management (including market research), local advertising and promotion (including athlete sponsorship and endorsement contracts). From 2005 to 2015, NEON operated under a licence and exclusive distribution agreement entered into with Nike International Limited (NIL), a Bermudan company. The royalties payable under this licence agreement were equal to all operating profits realised by NEON,minus a margin of 2-5% of its total revenue. In 2015, NIL was replaced by Nike International C.V. (NICV), a Dutch limited partnership. NIL/NICV had no activities other than holding most of the non-US Nike IP rights (including the Nike EMEA IP rights), comprising of trademarks, trade names and patents. NEON sold Nike products through commission agents established in the EMEA area. NEON also had a licence for non-EMEA Nike IP, which it sublicensed to related and unrelated parties outside the EMEA area. CNBV’s set up was comparable to that of NEON, but related to Converse products sold in the EMEA region. The following charts illustrate the “CV” structures.
The APAs confirmed that a margin of 2-5% of total revenues of NEON and CNBV represented an at arm’s length profit for their activities. Accordingly, the royalties due under the licence agreements with NIL/NICV and ASCV would be deductible for Dutch corporate income tax purposes. NIL/NICV and ASCV did not pay any Bermudan/Dutch tax. Hence, the effect of the structure was that, except for the margin of 2-5% of the total revenue of NEON and CNBV and the remuneration of the EMEA commissionaire agents, all other income from the EMEA sales of the Nike group was accumulated untaxed in NIL/NICV and ASCV.
The Commission’s state aid analysis
The provisional view of the Commission is that, through the APAs, the Dutch government has granted illegal state aid to NEON and CNBV. The Commission substantiates its provisional view through, by now, well-known reasoning. First, the Commission establishes that aid has been granted by a member state (the Netherlands) through state resources (a reduction of corporate income tax) and that such aid distorts, or threatens to distort, competition and trade between member states (NEON and CNBV are active in multiple EU countries). The Commission goes on to establish whether an advantage has been granted to NEON and CNBV and, if so, whether that advantage is selective in nature. The provisional view of the Commission is that both conditions have been met and that the Dutch government has therefore granted illegal state aid through the APAs.
The Commission is of the provisional view that the APAs give an advantage because the taxable profits based on the APAs are below the amounts to be reported based on a correct application of the Dutch corporate income tax rules. The Commission focuses on whether the royalties due by NEON and CBV exceed the level of royalties to be taken into account under the at arm’s length principle. The opening decision does not raise objections relating to the hybrid treatment of the CV resulting in double non-taxation of the royalty income as such or to other “technical” aspects of the IP structures.
Three alternative positions
First, the Commission argues that the transfer pricing method chosen in the APAs, the Transactional Net Margin Method (TNMM), was incorrectly applied by treating NEON and CNBV as the least complex entities, rather than NIL/NICV and ASCV. The Commission comes to this conclusion by analysing the functions performed, assets used and risks assumed by NIL/NICV and NEON and by ASCV and CNBV respectively. Based on this analysis, the Commission concludes that the activities of NEON and CNBV go far beyond the “routine distribution functions” ascribed to NEON and CNBV in the APAs, while NIL/NICV and ASCV act as bare legal IP owners to which no key value-driving functions can be attributed. The Commission repeatedly refers to the hundreds of FTE working for NEON and CNBV, and the absence of employees at the level of NIL/NICV and ASCV. Accordingly, NEON and CNBV should be entitled to a much larger share of the EMEA profits than the 2-5% of total revenue actually reported by them as taxable income for Dutch corporate income tax purposes. The Commission specifically states, as it did in the Amazon case, that functions performed by entities other than the parties to the “tested” licence agreements (NIL/NICV and ASCV as licensors and NEON and CNBV as licensees) which may have made or are making a valuable contribution, must be disregarded when analysing the transactions covered by the APAs and the NEON and CNBV license agreements. This position seems to imply that the Netherlands must tax profits that are “undertaxed” by the US, that is, profits that would be taxed in the US, through an adjustment in respect of the cost sharing agreement (CSA) / IP arrangements between the IP owners and their US group companies, if the US had applied the same (OECD) transfer pricing standards. In the Amazon case, the Commission also took this position, but pointed to a decision by the US Tax Court that confirmed that these agreements were priced at arm’s length.
Second, the Commission argues that the APAs confer an advantage on NEON and CNBV because the TNMM is not the appropriate transfer pricing methodology. The Commission considers that, while there is generally some leeway on the choice of transfer pricing method, the choice for the TNMM method in favour of the comparable uncontrolled pricing (CUP) method or, alternatively, the profit split (PS) method was not appropriate. In respect of the CUP method, the Commission’s view is based on NEON and CNBV having entered into licence agreements with third parties and other group companies outside the EMEA region. This provided for much lower royalty rates than the royalties paid by NEON and CNBV to NIL/NICV and ASCV, despite these parties performing a role similar to that of NEON and CNBV. This is a feature that is distinct from the Amazon case where the Commission found that there was no comparable third party. In respect of the PS method, the Commission argues that NEON and CNBV, in any event, carried out unique and valuable functions and if, contrary to the Commission’s primary position, NIL/NICV and ASCV were deemed to carry out unique and valuable functions as well, the only appropriate indirect transfer pricing method would be a PS method.
Third, the Commission has doubts as to the choice of the profit level indicator, a percentage of total revenue, as this includes income from sublicensing non-EMEA Nike and Converse IP. As far as we can determine from the factual description of the case, the role of NEON and CNBV in this arrangement is merely that of a conduit, hence a margin of 2-5% for this role may very well be too high, rather than too low as the Commission claims.
The Commission considers that the advantage conferred on NEON and CNBV through the APAs are individual measures applicable to these two companies only. As such, it argues that the advantage must be presumed to be selective, and it suggests that an actual comparative analysis with companies in a comparable legal and factual situation is, in general, only needed if the investigated measure regards an “aid scheme” (such as a statutory regime giving benefits to certain qualifying taxpayers). For completeness’ sake, the Commission also makes the actual comparative analysis it considers to be applicable to aid schemes only. In this respect, the Commission’s primary position is that the “reference group” of companies in a comparable legal and factual situation comprise all Dutch corporate income taxpayers. The Commission considers the reference framework to be the Dutch corporate income tax system as such, which has as objective to tax all profits to be determined in an objectified manner based on the total profits concept. The latter concept applies to all taxpayers, whether “stand-alone” or part of a multinational group, and encompasses the statutory arm’s length principle which only applies to related entities. But as a secondary position, the Commission concludes that even if the reference group comprises companies forming part of a multinational group only, the alleged advantage received by the Dutch Nike group companies has been granted selectively; that is, it was not available to all other multinational group companies. The references to aid scheme in the decision seem to have been made only in the context of how to apply the selectivity test, and by themselves do not indicate that the Commission actually holds the view that a Dutch aid scheme can be identified from which Nike benefitted.
The Commission will carry on its investigation and then issue its final decision on whether or not the relevant APAs constitute state aid. Commission policy is to issue final decisions within 18 months following the opening decision but in practice this often takes much longer.
15 October 2020
14 October 2020
14 October 2020
17 September 2020
17 September 2020
16 September 2020
15 July 2020
13 July 2020
16 June 2020
16 June 2020