27/02/18

European Commission publishes State aid decision on Amazon

On 26 February 2018, the Commission published the non-confidential version of its October 2017 decision ordering Luxembourg to recover State aid from Amazon (the Decision). Luxembourg has in the meantime challenged the Decision before the EU General Court. The publication of the Decision sheds further light on the Commission’s reasoning, in particular as regards the key criteria of advantage and selectivity. While an individual State aid decision does not have a direct impact on other taxpayers, businesses can now better assess their own State aid exposure in light of the Commission’s reasoning.

Facts of the case

In the case at hand, a company fully taxable in Luxembourg (LuxOpCo) paid from May 2006 to June 2014 a royalty to a Luxembourg partnership (LuxSCS) for the use of certain intangibles (technology, marketing-related intangibles and customer data). The Luxembourg tax authorities at the time confirmed by means of a tax ruling that the royalty was in line with Luxembourg transfer pricing rules. The Decision argues this royalty exceeded the arm’s length value, so that the tax base of LuxOpCo was unduly reduced. 

Advantage – more uncertainty in transfer pricing? 

State aid is defined as a measure granted by the State or through State resources, which distorts or threatens to distort competition and affects intra-EU trade by favouring certain undertakings or the production of certain goods. 

The Decision develops two lines of reasoning on the existence of an advantage. In the primary one, the Decision makes an extensive functional analysis based on numerous documents obtained during the investigation, including from the U.S. Tax Court case between Amazon and the IRS (see here the U.S. Tax Court opinion of 23 March 2017). The Commission considers that LuxSCS did not have any function, risk and asset in relation to Amazon’s European business, nor to the development of the intangibles. While Amazon claims these intangibles were essentially developed in the U.S., the Commission allocates functions, risks and assets to LuxOpCo. 

Although both the Commission and the OECD prefer the CUP method under their respective guidance, the Commission subsequently dismisses the CUP method, rejecting the comparability of agreements that Amazon considers sufficiently established by the U.S. Tax Court to value a sub-set of the same intangibles in the U.S. Tax Court case. Instead, the Commission asserts the transactional net margin method (TNMM) should apply, with LuxSCS as tested party making the less complex contributions as compared to LuxOpCo. It is remarkable that the Commission relies on and extensively refers to the 2017 OECD TP Guidelines, thus effectively enforcing non-binding guidance with retroactive effect. Based on its TNMM analysis, the Commission considers that the royalty paid by LuxOpCo should only cover the costs incurred by LuxSCS to develop and maintain the intangibles as well as a minor mark up on LuxSCS’ related expenses. 

In the subsidiary line of reasoning, the Commission points out what it considers methodological mistakes in the transfer pricing analysis. The Commission argues that LuxOpCo performed more than just routine functions. It also questions the use of operating expenses rather than total costs as profit level indicator to benchmark the profitability of LuxOpCo under the TNMM, and also the introduction of a floor and a cap to LuxOpCo’s remuneration. 

Selectivity – comparison with all taxpayers or to group companies only? 

On selectivity, the Decision applies three lines of reasoning. In its primary line, the Commission presumes selectivity, claiming that an individual measure (here, the advance tax confirmation of 2003) giving an advantage is automatically selective. This reading of the MOL case is yet to be confirmed by the EU Courts. 

As a subsidiary line, the Decision applies the usual 3-step selectivity test: first defining the reference framework (i.e., the “normal” application of the tax rules), then identifying if LuxOpCo is better treated than other undertakings in a similar legal and factual situation (within the same reference framework) and thirdly determining whether the difference in treatment (if any) is justified by the nature of the system. The Decision, as in other recent tax State aid decisions, chooses the general corporate income tax system as reference framework, even though it questions the correct application of the Luxembourg transfer pricing rules in LuxOpCo’s specific case. The Luxembourg transfer pricing rules are not questioned as such. The Decision compares LuxOpCo to any other corporate taxpayer, rather than to the sole group companies party to intragroup transactions and thus subject to transfer pricing rules. Because other taxpayers could not allegedly reduce their tax liability by paying an excessive royalty, the Commission considers that LuxOpCo received a selective advantage that is not justified by the nature of the tax system. 

As a third line of reasoning, the Decision takes Luxembourg transfer pricing rules as reference framework, albeit without the related administrative practice. However, it barely develops any reasoning and instead relies on the alleged existence of an advantage to conclude that the selectivity test is also met. 

Potential actions to take 

The Decision does not directly apply to other taxpayers or groups, whether in Luxembourg or in another Member State. The reasoning shows, however, that the Commission continues to use State aid rules as a tool to push forward tax reforms in the EU and address what is perceived as unfair tax competition. 

As the Commission appears to insist on applying the 2017 OECD TP Guidelines with retroactive effect, companies engaged in intragroup transactions should make sure they have adequate transfer pricing documentation and review whether the allocation of functions, assets and risks correctly reflects economic substance. As part of a longer term transfer pricing strategy, restructuring could be opportune. Complying with the 2017 OECD TP Guidelines should minimize the State Aid risk going forward. 

Next steps 

Several tax State aid cases are still in the formal investigation procedure stage: McDonald’s and Engie in Luxembourg, Inter Ikea in the Netherlands, the Gibraltar tax ruling regime and the UK CFC financing exemption. Appeals with the EU General Court are pending in the Apple (Ireland), Starbucks (the Netherlands), Fiat (Luxembourg), Amazon (Luxembourg) and Belgian excess profit ruling scheme cases. The Commission continues to look into the tax practices of the EU Member States and is expected to open more investigations in the coming months.

dotted_texture