This investigation is part of the recent targeting of allegedly problematic tax rulings by the Commission, and was in fact opened on the same day as the Apple and Starbucks investigations (for my comments on Apple see here). Although the facts naturally differ, the legal issues at stake in all three cases are similar; thus many paragraphs are identical in the three decisions of the Commission to open proceedings. However, it is the Fiat investigation that has proven to be the most tempestuous, since the Commission had to issue an information injunction against Luxembourg in order to get the information it needed to launch its probe.
The arm’s length principle
Just like Starbucks and Apple, the problems for FFT lie in a tax ruling that validated a so-called ‘advance pricing agreement’ (APA). An APA is a contract between a tax authority and a taxpayer stipulating the pricing method the latter will apply to its affiliated-company transactions when calculating its taxes. Transactions within multinational groups need to comply with the ‘arm’s length principle’ as defined in Article 9 of the OECD Model Tax Convention, which requires that their terms don’t differ from those between unrelated companies. Were it otherwise, a multinational would be tempted to manipulate the prices of intra-group transactions, thus shifting profits to low (or no) tax jurisdictions. The 2010 OECD Transfer Pricing Guidelines provide five different methods of applying the arm’s length principle to intra-group transactions in order to avoid such a situation. Luxembourg did in fact rely on one of the OECD’s methods in order to calculate FFT’s taxable basis, but the Commission still concluded that the FFT APA did not comply with the arm’s length principle for reasons that will be elucidated below.
FFT and the Fiat Group
FFT is a member of the Fiat Group. This group is composed of Fiat S.p.A., incorporated in Italy, and all the companies Fiat S.p.A. controls. Fiat S.p.A. fully owns FFT, in a direct (40%) and indirect (60%) manner. Within the Fiat Group, all funding, corporate finance, bank relationships, foreign exchange and interest rate risk management, cash pooling, money market operations etc. are performed by treasury companies. FFT provides treasury services and financing to the Fiat group companies based (mainly) in Europe (excluding Italy), and also manages several cash pool structures for them. In plain English, Fiat subsidiaries across Europe both lend to and borrow from FTT large sums of money, amongst other things.
The FFT APA
In September 2012, the Luxembourgish tax authorities issued a tax ruling approving FFT’s transfer pricing arrangements, since they found that FFT’s remuneration complied with the arm’s length principle. The FFT APA (binding for five years) stated that
“[t]he transfer pricing study determines an appropriate remuneration on the capital at risk and the capital aimed at remunerating the functions performed by the company of EUR 2.542 million on which a range of +/- 10% is envisaged.” .
As mentioned above, FFT did rely on an OECD transfer pricing method to calculate its remuneration. More specifically, it relied on the transactional net margin method (TNMM), the method also used in the Apple and Starbucks APAs. According to this method, in order to examine whether a taxpayer’s remuneration complies with the arm’s length principle, one compares a specific net profit indicator (e.g. return on assets) that the taxpayer realises from a controlled transaction to the same indicator in a comparable internal or external transaction. In the FFT APA, equity was chosen as the net profit indicator, and the approximated arm’s length rate of return on equity was estimated through the CAPM financial model.
Most importantly, the Commission in paragraph 48 outlined four steps for estimating FFT’s remuneration: 1) estimation of the capital at risk (in application of the Basel II criteria); 2) identification of the capital used to perform the functions and to support the financial investments; 3) estimation of the expected return of capital by using the CAPM to remunerate the capital at risk and identification of the return to reward the capital used to perform the functions; and 4) estimation of the overall profitability to be left to FFT to remunerate the functions performed and the risks borne by FFT.
The Commission’s State aid focus
In order to constitute State aid, a measure has to fulfill four conditions: it must confer an advantage on an undertaking, this advantage must be selective, it must be granted by the state and through state resources, and it must distort competition and affect intra-Union trade. The fulfillment of the two final conditions is, according to the Commission, ‘relatively straightforward’ (para 54). The rulings were issued by the Luxembourgish tax authorities and the reduction in tax they allegedly entailed resulted in a loss of tax revenue, that would have otherwise been available to Luxembourg. State resources were therefore used. Furthermore, Fiat is active across the EU, and any state support is bound to strengthen its position in relation to its competitors. Consequently, the Commission focused on the two conditions it found most contentious, namely the existence of an advantage and selectivity.
The conferral of an advantage – mis-stating the private investor test once more?
Just like it did in the Apple opening decision, the Commission in para 62 seems to misstate the private investor test. More specifically, it focuses on what kind of tax arrangements a private operator, here FFT, would propose, and not on what a private operator, here Luxembourg, would accept. However, the private investor test is not used to compare the actions of the aid recipient with a market operator, but the actions of a Member State with a market operator of a comparable status operating in normal conditions of market economy (see e.g. para 29 in C-142/97 Belgium v Commission; for fiscal State aid see paras 89-92 in Case C‑124/10 P Commission v EDF). Thus, whenever the Commission talks about what a private operator would ‘declare’, we must read this as meaning ‘what tax declaration a private operator would accept’, and then compare that with the arrangements Luxembourg accepted in the FFT APA.
The nexus between State aid and the ALP
Coming to the gist of the State aid analysis, the sentence that first implicitly links the State aid conditions to the arm’s length principle (ALP) and thus to the OECD Guidelines is to be found in para 60:
“In order to determine whether a method of assessment of the taxable income of an undertaking gives rise to an advantage, it is necessary to compare that method to the ordinary tax system, based on the difference between profits and losses of an undertaking carrying on its activities under normal market conditions” .
The underlying premise here seems to be that companies are normally taxed on the basis of economic data that are the result of non-manipulated transactions. Thus, in the case of multinationals, market conditions can only be arrived at through transfer pricing established at arm’s length. If the arm’s length principle is not respected, an advantage is conferred on the taxpayer, and this triggers the application of Article 107 TFEU.
Compliance with the arm’s length principle?
The Commission concluded that the FFT APA did not comply with the arm’s length principle (para 81). The reasons given are manifold and mostly technical.
Firstly, the APA sets FFT’s tax base at 2.542 million euros – give or take 10%. This means that its tax base is virtually fixed for the duration of the tax ruling and can only range somewhere between 2.288 and 2.796 million euros. Thus, even if FFT’s activities doubled, there are no provisions in the tax ruling guaranteeing that its tax assessment would adjust accordingly.
Secondly, the OECD method used was in concreto inappropriate. It is here that the plot thickens. In para 62 the Commission had warned that ‘depending on the facts and circumstances of the taxpayer, not all methods approximate a market outcome in a correct way.’ It then went on to claim that according to the CJEU’s case law the use of direct methods for setting an appropriate level of profits is preferred (para 65). However, the only case it cited to support its assertion is the Forum 187 case and in particular para 95 therein, which was also relied upon in the Apple opening decision. In my view, this paragraph can neither support nor refute such claims. It only says that:
“[i]n order to decide whether a method of assessment of taxable income […] confers an advantage on [the recipient], it is necessary […] to compare that regime with the ordinary tax system, based on the difference between profits and outgoings of an undertaking carrying on its activities in conditions of free competition.”
The Commission conceded that the selection of ‘capital employed’ as the net profit indicator in the TNMM was not problematic per se, and is in fact in principle compatible with the OECD guidelines. Nevertheless, it asserted that in the case of FFT the two components that determined FFT’s arm’s length remuneration were set at too low a level. These components are: (i) the amount of capital remunerated; and (ii) the level of remuneration applied to this capital amount.
First: why was the amount of capital remunerated too low? Many reasons were given but the main ones are that, under the FFT APA, there were big subtractions from FFT’s capital basis, and the required equity remuneration was only applied to a small proportion of FFT’s capital. Moreover, only 9.9% of equity was found to be ‘equity at risk’, and the Commission also doubted whether the CAPM method could be applied to a lower capital base than total equity. Furthermore, the way ‘equity at risk’ was calculated in the FFT APA could lead to an unacceptably low taxable basis. The equity at risk should be calculated by analogy to the Basel II framework, according to which banks are required to hold capital in proportion to their ‘risk weighed assets’. Without good justification, the FFT APA, when calculating FFT’s regulatory capital need, disregarded all assets other than third party assets. In effect this posits that there is no credit/counterparty risk in operations with intra-group companies, and that is why these assets generate no capital needs. Given that FFT’s intra-group exposure accounts for around 70% of its total assets, the Commission found this arrangement difficult to understand: a prudent independent market operator would set aside capital for loans granted to group companies, even if there was a good chance that the group would support one of its members in case of default.
Second: why was the level of remuneration applied to this amount too low? Again, many reasons were given but the main ones are that FFT was compared to undertakings to which it was not really comparable. The appropriateness of the APA’s comparables was thus disputed. Moreover, Luxembourg uncritically accepted numbers that, in the CAPM equation for calculating the acceptable rate of remuneration, this would lead to a more favourable treatment to FFT than could possibly be justified.
What about selectivity?
The Commission only devoted one paragraph (82) to selectivity. Still, it managed to confuse me. In the first half of the paragraph it reminds us that a deviation from administrative practice gives rise to a presumption of a selective advantage, and that rulings that deviate from that practice have the effect of lowering the tax burden of the undertakings concerned. The Commission says that they ‘have’ this effect, not that they ‘might’ have this effect. I wonder, does a tax ruling that deviates from previous practice and results in a higher tax liability for an undertaking compared to its ‘peers’ amount to State aid? Evidently not. Apart from this, the Commission also asserted that since the ruling deviates from the arm’s length principle as regards FFT, the advantage is selective. However, this cannot be decisive – the wording here is careless. The Commission seems to forget that selectivity is an exercise in comparison: if the Luxembourgish authorities deviated from the arm’s length principle to the same extent in all their dealings with undertakings comparable to FFT then the advantage is not selective and there is no State aid, even if the aforementioned principle was not respected!
The Fiat investigation is yet another reminder that the intricate transfer pricing arrangements that multinationals have put in place in order to minimise taxes will continue to give State aid lawyers a headache; proficiency in international taxation is swiftly becoming a must. Still, should the recent investigations into tax rulings reach the European courts (and I am sure they will), their ‘doctrinal’ implications – especially as regards the advantage and selectivity conditions – will become clearer, thus ensuring a higher level of tax predictability for the taxation of multinational companies.
Dimitrios Kyriazis is reading for an MPhil in Law at the University of Oxford and is a Light Senior Scholar at St Catherine’s College. His thesis examines the interaction between EU State aid rules and national tax measures, exploring, in particular, the extent to which EU State aid policy can serve as a tool against harmful tax competition between Member States. In addition, Dimitrios also teaches European Union Law, and is one of the conveners of Oxford’s EU Law Discussion Group.
Image: Fiat 124 Special Sedan. [Alden Jewell/Flickr]