Reminder of the French tax regime applicable to interest
In France, the rules concerning the payment of interest between a French company and an affiliate depend on the fact pattern:
- If the French company is the lender, and the related entity is established abroad, then usual transfer pricing rules apply. Specifically, this would mean that the burden of proof would be with the French Tax Administration (“FTA”) if it were to challenge the interest rate applied.
- If the French entity is the borrower, the deduction of the interest expense may be limited by several rules. In particular, regardless of whether the related lending entity is French or foreign, the interest rate paid by the French entity is compared to a certain legal threshold (updated quarterly; as of Q4 2021, it was circa 1.15% and has been on a downward trend these past years):
- If the rate paid by the French entity is below the legal threshold, no further effort is required on its behalf and the interest should be deductible unless other limitations apply (such as thin cap or “ATAD” rules).
- If the rate paid by the French entity is above the legal threshold, then the French entity bears the burden of proving that the rate is a market rate.
While these rules were clear, what was far less clear was what constituted acceptable proof, vis-à-vis the FTA – and ultimately the courts – that an interest rate above the threshold was indeed at market rate.
To be more precise, article 212 of the French Tax Code (“FTC”) requires that for interest paid to related parties by French entities in excess of the legal threshold to be fully deductible, the taxpayer must prove that the interest rate is at a rate he “could have obtained from independent financial establishments or organisations under comparable circumstances”. The main difficulty for taxpayers here is that until very recently, the FTA, as well as certain courts, were extremely strict in what they accepted as proof that an interest rate was at a market rate. Practically speaking, they rejected most economic analyses and only accepted a binding, contemporaneous loan offer by an unrelated bank as proof. Naturally, this was extremely difficult for taxpayers to provide, insofar as a company which is asking a binding loan offer from a bank would go for a bank loan rather than an intra-group loan.
Recent and more favorable case law for taxpayers
Fortunately, the Supreme Administrative Court has recently eased the burden of proof for taxpayers by better defining what constituted adequate proof and by opposing the overly strict “binding contemporaneous loan offer” approach that the FTA and certain lower courts had endorsed. Of particular interest is the gradual acceptance – at first by the Supreme Administrative Court, then by lower courts – of the use of financial benchmarking studies and automated tools such as scoring calculators.
In one of these cases where Fidal was involved, the Administrative Court of Appeals of Versailles ruled in favour of a taxpayer who had used RiskCalc (among other items) to establish that the interest rate it had paid was at a market rate. In this case, the court followed an interesting, three-step reasoning:
- First, the court recognised that a scoring obtained by using automated tools such as RiskCalc is inherently less accurate than the actual rating a proper rating agency would have provided through an in-depth analysis.
- Then, the court nuanced this statement by pointing out that the fact such a scoring is less accurate does not mean that it is not accurate at all and that, especially considering the high – unaffordable to many corporate entities – cost of a full-blown rating by an agency, such a scoring could not be disregarded systematically.
- Finally, the court concluded that, in the absence of any valid criticism of the scoring in question by the FTA, it was an acceptable proof and therefore ruled in favour of the taxpayer.
The third point raised by the court is the most noticeable. Indeed, in the case at hand, the taxpayer had prepared an in-depth economic analysis to support the interest rate paid whereas the FTA kept arguing, over the whole procedure, that this was not acceptable evidence without presenting, from its side, any detailed criticism against the scoring obtained by the taxpayer.
Going forward
It is important to note that while there were several such favourable examples of case law in the past months, they all have in common that taxpayers relied on other items of proof beyond just benchmarking studies and automated scorings (for example, internal CUPs such as actual loans from third parties but granted in a different period of time). This does not mean that a taxpayer could not bring an acceptable proof via benchmarking studies alone; however these cases highlight how sensitive the matter of intragroup interest rates remains, and how important it therefore is to prepare a proper documentation.
In the opposite, the FTA can no longer, as it used to, outright and systematically reject benchmarking and scoring studies. As for the courts, while they do not yet seem entirely comfortable with such clues and tools, they must now follow in the steps of the Supreme Administrative Court and look at them as a potentially sufficient evidence of a market rate which cannot be disregarded in principle.
As for the taxpayers, they will of course have to explain why these clues do evidence that the rate they are paying is at arm’s length.
Also of interest is the fact that the courts also explicitly refer to arm’s length ranges of interest rates (intervals), rather than to the interest rate which would be the only acceptable one.
In conclusion, we recommend that groups with a French entity which pays interest to a related entity prepare an in-depth demonstration of the arm’s length nature of the interest rate – including, importantly, detailed explanations of why the tools used to make this demonstration are valid and relevant.
If you wish to discuss these topics, please contact: Fidal, Paris
Read the WTS Global Financial Services Newsletter here.
