LAST WEEK was all about the marriage [and divorce] of transfer pricing arrangements and State aid. W
My key take-away for future transfer pricing arrangements? In designing transfer pricing arrangements, set a level of remuneration which is validated by the OECD method that achieves the maximum outcome. If it happens to have a US tax ruling at hand, don’t set remuneration levels in the EU above those applied in the US.
You will have read dozens of reports on the Commission’s Wednesday twin announcements: it ordered Amazon to repay €250 million in taxes to Luxembourg and it referred Ireland to the Court of Justice for failure to recover €13 billion of illegal State aid from Apple.
Many of the reports provided the broader picture, that of the link between the EU’s ambitious regulatory reforms aimed against tax avoidance and the Commission’s State aid tool. “The message has become ever more clear: If EU countries are not willing to play ball with proposed tax reforms, European competition authorities are willing to do it the hard way, using the Commission’s technical rules” (Politico), “The decision would come amid a renewed crackdown by the EU, which has promised to scrutinize tax arrangements between its various member states and big multinationals operating in Europe” (FT).
There will be, no doubt, articles and events questioning whether State aid is the right tool to tackle tax avoidance and whether the Commission’s chain of State aid investigations and recovery decisions has exceeded or not its area of competence and the borders of national sovereignty.
In the midst of so many elaborate discussions, as a humble lawyer, here and now I will only comment on what I found the most interesting: the application of the market economy operator (MEO) test in the area of taxation. For those less familiar with State aid law and put in very simple words: it’s a test which, if passed, indicates that there was no State aid in a given case. In the Amazon case, the Commission considered that the test was not met.
The Commission has developed a series of formulas for the application of the MEO test to various areas. These formulas may be liked by some and disliked by others but they undoubtedly add legal certainty. As the full text of the decision has not been published yet, I will assume that the final decision confirms the Commission’s line of thinking in its earlier decision to open a formal investigation in the Amazon case.
Facts in a nutshell
According to the Commission’s findings, Amazon structured its business in such a way as to concentrate its profits in a Luxembourg company, taxed in that State (Amazon EU Sarl). Amazon EU Sarl paid royalties for licensing Amazon group’s intellectual property (IP) rights to another Luxembourg company belonging to the Amazon group (Lux SCS), a so-called “transparent company” which was not taxed in Luxembourg. The payments decreased Amazon EU Sarl’ taxable base. None of those aspects were challenged by the Commission. As the Commission’s press release clearly stated “[the decision] did not question Luxembourg's general tax system as such”. So far so good.
Or perhaps not entirely – at least when it comes to the level of royalties Amazon EU Sarl paid to Lux SCS that the Commission considered to be excessively high. According to the Commission, by a tax ruling Luxembourg agreed on artificially high royalties which resulted in abnormally low taxes Amazon paid in Luxembourg. The Commission found that the profit tax missed due to the excessively high royalties represented illegal State aid granted by Luxembourg to Amazon.
The MEO
Zooming back on to the most interesting bit (to me anyway): the application of the MEO. The Commission probed whether the transaction between Amazon EU Sarl and Lux SCS was at arm’s length ie, whether the level of royalties was not artificially inflated.
The Commission did not contest, as a matter of principle, transfer pricing arrangements and tax rulings confirming them. Equally, it did not contest the “Bible” applied by tax practitioners for setting up transfer pricing arrangements, which is the OECD Transfer Pricing Guidelines.
Consequently, investigating if the particular arrangement in that case breached State aid rules, the Commission did consider each of the methods under the OECD Transfer Pricing Guidelines to approximate arm’s length pricing of intra-group transactions.
Starting from the OECD methods, the Commission sought to establish remuneration a prudent independent operator acting under normal market conditions would have paid in exchange of the IP license (the MEO test). For the purpose of the MEO test, the Commission considered that there is no choice between the various OECD methods. The Commission’s MEO test appears to pick only that method which achieves the maximum possible outcome”, which means the smallest royalty. The Commission further considers that out of those methods, that called “comparable uncontrolled price method (CUP)” would typically be the most appropriate.
Also interesting, from the press release it appears that the final decision also took into account a recent ruling of a US tax court which established a significantly lower level of payments due by Lux SCS to Amazon in the US to contribute to the costs of developing the IP.
