French Courts Weigh In on Intragroup Financing

This article was originally published by Tax Notes International on 22 July 2024 and is reproduced with permission from the publisher.

Introduction

Intragroup financing transactions have become a regular topic of discussion with the French tax authorities (FTA) and one of the main sources of transfer pricing adjustments in France.

This can be explained by two main factors: The burden of proof is borne by taxpayers, and the FTA takes a strict position.

Burden of Proof

Under article 212, I, a, of the French tax code, it is up to taxpayers to demonstrate the arm's length nature of the interest rates paid to related entities when they exceed the rate set in article 39, 1, 3°, of the code (the safe harbor rate).

The figure shows the evolution of the safe harbor rate, which was particularly low during the 2013-2022 period (less than 3%) and has significantly increased since.

FTA Position

When taxpayers use a higher interest rate than the safe harbor rate, the FTA adopts a quasi-automatic rejection of the transfer pricing analyses provided by the taxpayers to justify that an interest rate complies with the arm’s-length principle, arguing that:

1. the scoring models used to assess the borrower’s creditworthiness are not Reliable;¹ and
2. the financial transactions selected as comparable are not relevant (mainly because they have different amounts, or because of the different industry, size, or geographic location of the borrowing companies).

Many disputes have arisen over the evidence that can be provided by French companies to establish the arm’s-length nature of an intragroup interest rate. Over the years, this long line of caselaw has led to the development of a number of clear principles, such as the importance of the borrower’s risk profile in the analysis,² the possibility of justifying the arm’s-length nature of an interest rate by any means (in particular, by relying on the yields observed in the bond market),³ and the relevance of the scoring models to assess a company’s credit risk.⁴ Two recent court decisions, one by the Conseil d’Etat (France’s highest administrative court)⁵ and another by the Paris Administrative Court of Appeal,⁶ brought additional clarity regarding the types of evidence that taxpayers can produce to demonstrate that an interest rate is at arm’s length.

Evolution of the Safe Harbour Rate in France

Key Takeaways of the Decisions

Credit Risk Analysis

• The borrower’s creditworthiness can be determined based on reliable scoring models or the analysis of specific financial ratios, if relevant in the case at hand (for example, the loan-to-value ratio for the financing of a real estate acquisition).
• The industry in which the borrower operates is a key qualitative criterion that needs to be considered in the credit risk analysis.

Relevance of the Comparable Used

• The size of the borrowing company is not relevant to determining if a bond issuance could have constituted a realistic alternative to an intragroup loan.
• Corporate bond yield curves (average rates updated daily by sector, rating, maturity, and currency) can be considered as relevant elements of comparison, even if the underlying bonds used by the database to construct those yield curves are not provided by the taxpayer.
• The taxpayer may rely on financial transactions of different amounts; or
  concluded by companies of various sizes, operating in different business sectors; or
  geographic locations, as long as these characteristics are already taken into account at the stage of the credit risk analysis.

Sté GEII Rivoli Holding

Background

To finance the acquisition of a EUR28 million office building in Paris, GEII Rivoli Holding SAS entered into an intercompany loan agreement, bearing interest at a rate of 5.08% per year (higher than the safe harbor rate of 2.79%).

To justify the arm’s-length nature of this intragroup interest rate, the company provided two different analyses:

• Analysis 1: Determination of the borrower’s credit risk (Baa1) using the scoring model developed by Moody’s Analytics (RiskCalc) and a computation of a range of interest rates by reference to the rates obtained by 15 nonfinancial companies operating in different industries, with ratings between A3 and Baa3.
• Analysis 2: Assessment of the credit risk using two different financial ratios (including the loan-to-value ratio) and the determination of an arm’s-length range based on the corporate bond yield curves provided by the S&P Capital IQ database.

Both analyses were rejected by the FTA. The Administrative Court⁷ and the Administrative Court of Appeal⁸ sided with the FTA and disregarded the transfer pricing studies provided by the taxpayer.

Conseil d’Etat Decision

The Conseil d’Etat rejected the first analysis but accepted the second.

• Rejection of Analysis 1: The borrower’s credit risk was determined without considering the company’s business sector. The use of scoring models such as RiskCalc is possible,⁹ but the credit risk analysis must take into account the borrower’s industry.
• Acceptance of Analysis 2: A bond issuance could have constituted a realistic alternative to the intragroup loan under review, and the rejection of the rates derived from corporate bond yield curves is not justified.

Willink SAS

Background

In 2011 Willink SAS issued two intercompany convertible bonds with a maturity of 10 years, carrying interest at an annual rate of 8%. To demonstrate the arm’s-length nature of this interest rate, the company provided the FTA with the following analysis:

• Assessment of the borrower’s creditworthiness using Moody’s Analytics RiskCalc (B1). The reasonableness of such a credit risk was corroborated at a later stage using another scoring model (Credit Analytics developed by S&P Capital IQ).
• Identification of comparable uncontrolled financial transactions in the Moody’s database and computation of an arm's length range of spreads.
• Comparison with the implied spread attached to the convertible bonds issued by Willink SAS, computed as the difference between the annual rate of 8% and a relevant 10-year risk-free rate for the French market.

The FTA rejected the transfer pricing study produced by the company. Both the Administrative Court¹⁰ and the Administrative Court of Appeal¹¹ sided with the FTA. The Conseil d’Etat ruled in favor of the taxpayer and censured the decision of the Administrative Court of Appeal.¹²

Administrative Court of Appeal Decision

The referring Administrative Court of Appeal confirmed the following:

• the relevance of the scoring models to assess the credit risk of a borrowing company;¹³ and
• the fact that the financial transactions selected as comparable by the taxpayer cannot be rejected on the sole ground that they are of different amounts or concluded by companies of different sizes operating in different industries or geographic locations, provided these characteristics are already taken into account in the credit risk analysis.¹⁴

Conclusion

Although these court decisions bring additional clarity to determining the arm’s-length nature of interest rates in intragroup financing transactions, there is still a significant level of uncertainty. The authors anticipate that it will continue to be one of the main matters giving rise to transfer pricing litigation in France.