Danny Beeton and Alain Goebel of Arendt & Medernach analyze the recent decision of the EU General Court in the Amazon Luxembourg state aid tax ruling controversy, considering the implications for similar disputes and suggesting some planning points for multinational taxpayers.
On May 12, 2021, the EU General Court published its decision on linked appeals made by Luxembourg and the taxpayer Amazon against the finding of the European Commission that state aid had been granted to the company in the form of a tax ruling, resulting in non-arm’s-length transfer pricing. This decision is important because it calls into question the Commission’s established approach to such investigations.
The appeals can be found at: T-816/17 Luxembourg v Commission and T-318/18 Amazon EU v Commission. The press release on the General Court’s decision can be found here.
Background
The case relates to the 2017 decision of the Commission that the Amazon group had benefited from state aid through a Luxembourg tax ruling. The Commission’s decision in this case was that a Luxembourg company, Amazon Europe Holding Technologies SCS (LuxSCS), had entered into an intellectual property (IP) buy-in agreement and a cost sharing agreement with two U.S. Amazon companies, as a result of which it would function as the IP holding company for Amazon’s European operations in return for royalty payments.
In turn, LuxSCS would license its IP rights on a back-to-back basis to its subsidiary, Amazon EU Sarl (LuxOpCo), which would be the European headquarters, the strategic decision-maker and the principal operator of the European business. The other EU companies had service agreements with LuxOpCo under which they were remunerated on a cost-plus basis for their services of fulfillment, customer and merchant services and support services. All functions and risks relating to sales and inventories were to be the responsibility of LuxOpCo.
Amazon set the royalty paid by LuxOpCo (and then passed on by LuxSCS) so as to leave it with a net mark-up on costs of 4–6%, subject to this not being less than 0.45% and not more than 0.55% of EU revenue. The tax outcome of these arrangements was that the bulk of the European profit was transferred out of Luxembourg by the royalties and not taxed there.
The Commission investigated Luxembourg’s tax ruling because it had the following concerns:
- the ruling was made before seeing any transfer pricing analysis;
- it was not based on a recognized transfer pricing method;
- the royalty was not (as recommended in the OECD Transfer Pricing Guidelines) related to output, sales or profit;
- the ruling was allowed to continue for more than 10 years;
- the mark-up on costs allowed for LuxOpCo was too low for its functions and risks; and
- there was no explanation for the floor and cap mechanism.
Having completed its investigation, the Commission concluded that there was indeed state aid, for the following, more detailed, reasons:
- it was LuxOpCo rather than LuxSCS which performed the key functions and bore the key risks for both the IP and the European operations; and
- the transfer pricing method was wrong, in the following ways:
- LuxSCS’s mere legal ownership of the IP was not “a unique contribution”;
- LuxSCS should have been the tested party using the transactional net margin method (TNMM); and
- LuxSCS should have received a 5% mark-up on costs for its low value services.
The Appeal
The EU General Court granted leave to Luxembourg and Amazon to appeal in 2020.
Amazon argued that:
- the royalty was at an arm’s-length level;
- the functional analysis was incorrect—the choice of the TNMM to be applied to LuxSCS was an error; and
- important evidence had been overlooked.
Luxembourg argued that:
- the arm’s-length royalty rate would have been even higher than the one actually paid;
- there were many errors of fact and law, leading to an inappropriate choice and use of the TNMM; and
- no “selectivity” had been proven.
The Decision
In its decision, the General Court concludes that “none of the findings set out by the Commission in the contested decision are sufficient to demonstrate the existence of an advantage for the purposes of Article 107(1) TFEU, with the result that the contested decision must be annulled in its entirety.”
The Court’s decision was made on the basis of the following assessment:
- the Court accepted that, where not elsewhere specified, national law can be assumed to include the arm’s-length standard;
- however, the Commission must demonstrate that the taxable profit of the company would have been higher following the arm’s-length principle;
- in this case the Commission did not take into account the functions and risks of the company and was wrong to characterize the company as a mere passive holder of the intangible assets in question;
- the Commission did not demonstrate that it would be easier to find comparables, or that comparables would be more robust, for LuxSCS rather than for LuxOpCo;
- the Commission was incorrect to assert that LuxSCS’s remuneration could be calculated on the basis of the mere passing on of the development costs borne in relation to the buy-in agreement and the cost-sharing agreement without taking into account the subsequent increase in value of those intangible assets;
- the Commission erred in treating LuxSCS’s functions as low value-adding services;
- also, the Commission failed to establish that the methodological errors necessarily led to an undervaluation of LuxOpCo’s remuneration;
- even if the tax ruling was on the basis of an inappropriate profit level indicator and should not have included the ceiling mechanism, the Commission did not provide sufficient evidence to justify its inferences; and
- therefore, the elements put forward by the Commission were not capable of establishing that LuxOpCo’s tax burden was artificially reduced.
Implications for Similar State Aid Challenges
By the time of the decision of the EU General Court in the joined taxpayer and Luxembourg appeals involving Fiat Finance and Treasury on September 24, 2019, it appeared that the Commission had developed (and was able to sustain) the following set of general assumptions when investigating and deciding on state aid cases involving transfer pricing:
- that the restriction on state aid can be applied to taxation as well as expenditure;
- that the separate, “normal” tax law of each member state must always have included the arm’s-length principle by implication;
- that this arm’s-length principle can be applied by reference to international transfer pricing guidance, even if it has no legal authority in a jurisdiction;
- that today’s international guidance should be referred to; and
- that all tax rulings or advance pricing agreements (APAs) which include non-arm’s-length pricing must involve “selectivity.”
This position seems to have been largely confirmed by the decision on the Amazon appeals. However, the General Court has set a high bar for proving a selective advantage: it is not enough to demonstrate that a transfer pricing method is not the most appropriate, the alternative must also be specified and it must be shown that it can be benchmarked robustly and the necessary calculations must be presented.
In specifying the alternative transfer pricing method the Commission should base it on an accurate functional analysis which does not selectively ignore facts.
Planning Points
General recommendations for multinational taxpayers who have benefited from a formal or informal tax agreement, ruling or APA in the EU which is challenged by the European Commission are:
- check that the Commission has not overlooked important facts when reaching its functional analysis conclusions and deciding on its characterization of the parties, its choice of tested party and its transfer pricing method;
- consider the relative quantity and quality of the Commission’s comparables compared to those used for the existing calculations; and
- check whether the Commission has actually presented the necessary calculations using its proposed method, and whether they are robust.
The Commission has the right to appeal to the Court of Justice of the European Union (on points of law only) within two months and ten days.
This column does not necessarily reflect the opinion of The Bureau of National Affairs, Inc. or its owners.
Alain Goebel is a Partner and Danny Beeton is Of Counsel in the Tax Department of Arendt & Medernach.
The authors can be contacted at: alain.goebel@arendt.com; daniel.beeton@arendt.com
Learn more about Bloomberg Tax or Log In to keep reading:
Learn About Bloomberg Tax
From research to software to news, find what you need to stay ahead.
Already a subscriber?
Log in to keep reading or access research tools.