THE FACTS
Following an audit of GEII Rivoli Holding's accounts for the 2013 and 2014 financial years, the tax authorities questioned the deductibility of the difference between the 5.08% rate applied and the 2.79% corresponding to the value mentioned in 3° of 1 of Article 39 of the CGI.
During the litigation phase, the company produced an initial analysis based on the RiskCalc tool developed by Moody's, identifying the risk rating that could have been assigned to it, as well as a rate range established by reference to those obtained by fifteen non-financial companies, belonging to heterogeneous business sectors.
A second corroborative analysis was submitted to the Paris CAA, based on the calculation of two financial ratios, one of which, known as the "loan-to-value" (LTV) ratio, was based on bond market data taken from the Standard & Poor's Capital IQ financial database.
THE RULE
A trend in case law built up around the 2020s has redrawn the contours of proof with regard to the deductibility of rates charged to majority shareholders.
Specifically, the borrowing company may rely on the rates of bank loans granted, under arm's length conditions, to companies in the same non-financial sector, which have obtained credit ratings close to that which can be determined for it, even though these other companies may belong to heterogeneous business sectors.
The borrowing company may also take into account the yield on bonds issued by companies in comparable economic conditions, where such bonds represent a realistic alternative to an intra-group loan.
THE JUDGES
The TAA de Paris in 2021, then the CAA de Paris in 2022, rejected the company's claims and confirmed the rectifications made.
Firstly, the judges noted that, in order to justify the 5.08% rate applied to its parent company, GEII Rivoli Holding had produced a report using the RiskCalc tool developed by Moody's, which identified the risk rating that could have been awarded to the company, i.e. Baa1. However, this risk rating had been obtained without entering the applicant's business sector in the RiskCalc tool. Thus, the CAA was able to dismiss this method as inconclusive on this ground, without committing an error of law, since such a circumstance meant that the company's particular economic situation was not taken into account.
Secondly, in rejecting the corroborative method proposed by the company, the CAA considered that the company did not justify that a bond issue would have constituted a realistic alternative to an intra-group loan.
Finally, the CAA considered that the company had not been provided with any precisely identified comparables whose relevance it would have been able to assess.
THE CONSEIL D'ÉTAT'S SOLUTION
The EC accepted the first argument of the lower courts, rightly considering that the company's sector of activity is an important parameter to be taken into account when calculating the credit rating on the RiskCalc tool.
However, it rejected the rest of the arguments, thus validating the company's economic and statistical demonstration. More specifically, the EC emphasized:
- "The size of a company is not in itself such as to hinder access to this market, and that the realistic nature, for a company having recourse to an intra-group loan, of the alternative hypothesis of a bond issue can only be assessed in the light of the specific characteristics of this company and of the transaction, the rates observed on this market having to be adjusted if necessary".
- The arm's length rate put forward by the company as corresponding to its level of risk was based on the use of rate curves established on the basis of all the transactions recorded, for loans of the same duration contracted by companies with the same risk profile, and it was not argued that the recording of transactions in this database was unreliable".
OUR ANALYSIS
THE RISKCALC TOOL IS USEFUL, BUT NOT ALL-POWERFUL
Developed by Moody's, the RiskCalc tool has gained legitimacy among tax judges since the Studialis ruling by the Paris CAA in 2020 (no. 18PA01026).
This tool can be used to determine a borrower's risk rating, which is the first essential step in demonstrating the arm's length nature of a rate charged to majority shareholders. However, this tool requires a detailed analysis of the borrower's intrinsic parameters, both quantitative and qualitative, including in particular its business sector.
This latter indicator has a major influence on the past and future prospects for growth, profitability and therefore risk, of the players making up a given market. If this essential criterion was not included, the analysis produced initially could not be relevant or complete, as it necessarily misunderstood the company's economic situation.
It is interesting to note, however, that neither the contemporaneity of the analysis nor the relevance of the tools cited were discussed, thus validating, and no doubt definitively, the praetorian trend initiated by the aforementioned Studialis, BSA de la CAA de Versailles (n°20VE03249), and Willink du Conseil d'Etat (n° 446669) rulings.
Above all, it should be noted that the demonstration that finally won over the Conseil d'Etat was based on an alternative financial ratio known as "loan to value" (LTV), which relates the level of indebtedness to the value of the company's real estate assets. In this case, this indicator led to an estimate, based on a comparison with the ratios of listed French and European property companies, that the financial rating it could have obtained would not have exceeded BBB, i.e. a level close to that initially proposed by RiskCalc.
In this case, the LTV ratio had been calculated taking into account a financial debt corresponding exclusively to the loan whose rate had to be assessed. One might have thought that the calculation was flawed because it was circular. However, by focusing on the principal loan (the purpose and amount of which were not in dispute), without taking into account the interest (the rate of which was at the heart of the debates), the ratio was indeed relevant and valid.
THE CONSECRATION OF THE BOND MARKET
In its July 2019 Wheelabrator opinion, the Conseil d'etat paved the way for a pragmatic approach, in line with OECD practice, to the taxpayer's demonstration of the "arm's length" nature of an interest rate charged in the context of intra-group financing, allowing in particular the use of bond benchmarks.
However, this opinion, as well as subsequent rulings, seemed to contain a reservation, by making reference to the bond market conditional on demonstration that "these loans constitute, in the hypothesis under consideration, a realistic alternative to an intra-group loan". In other words, the taxpayer appeared to have to prove that issuing bonds was a realistic alternative to taking out a conventional loan with a bank or credit institution.
In recital 10, however, the CE seems to reinforce the burden of proof on the tax authorities. The judge considers that "the realistic nature, for a company having recourse to an intra-group loan, of the alternative hypothesis of a bond issue can only be assessed in the light of the specific characteristics of the company and the transaction, with the rates observed on this market having to be adjusted, where necessary, to take account of the specific features of the company in question". In order to disregard the reference to the bond market, it would seem that the tax authorities would have to demonstrate that, given its specific and intrinsic parameters, this option would have no purpose, or would be inappropriate.
In our view, such proof is impossible.
BENCHMARKS FOR ALL?
While the two-step economic analysis now seems to be well recognized by the tax judge, both in its credit risk calculation component and in its search for comparables on bond markets, it should be remembered that this approach is only valid if the lender is a majority shareholder within the meaning of Article 212-I. Minority shareholders cannot use this analysis to justify a different rate from that referred to in article 39-1-3 of the CGI (see CAA Versailles, Sté Financière Lilas, n°19VE00546). This trend further reinforces the difference in treatment between taxpayers.
