Asymmetric jurisdiction clauses: French court refers questions of validity to CJEU

In recent decades the courts of some countries, including some EU member states, have questioned the validity of asymmetric jurisdiction clauses or have refused to give effect to them. These clauses (also known as unilateral or one-way clauses) give one party greater flexibility as to the forum in which they can bring proceedings. They are often used in finance transactions, to give the lender flexibility to sue the borrower in any jurisdiction where it has assets, while restricting the borrower to a named jurisdiction. For example, a clause could allow the lender to bring an action before the English courts or any court which will accept jurisdiction under its own conflict of laws rules, while requiring the borrower to bring any action before the English courts.

The English courts have repeatedly held that asymmetric clauses are valid – and, indeed, that they may be considered exclusive jurisdiction clauses for the purposes of the recast Brussels Regulation (in cases commenced before the end of 2020, in which that Regulation continues to apply to the UK). See for example our blog posts here and here.

Decisions refusing to enforce these clauses have emerged, however, from various EU member state courts, including the courts in Bulgaria, Poland and (as particularly relevant for our purposes) France, as well as courts outside the EU. So it is significant that the French court has now referred to the CJEU three questions relating to the validity of such clauses:

1. What law should govern the validity of unilateral clauses: EU law or EU member states' national laws?
2. If EU law governs this question, does EU law prohibit unilateral clauses?
3. Alternatively, if the question is governed by national laws, how should a court decide which member state's law should be applied?

We consider below the relevant French case law, the questions to be addressed by the CJEU, and why this is still relevant to the UK despite Brexit.

French case law on unilateral clauses

Since the Banque de Rothschild case (Cass., civ. 1ère, 26 September 2012, No. 11-26.022, see our blog post here), in which the highest French court refused to enforce a unilateral jurisdiction clause, such clauses have been scrutinised and often held to be ineffective by the French courts. The courts' rationale for refusing to give effect to these clauses has, however, evolved over the years.

The original Banque de Rothschild reasoning was that these clauses are contrary to article 23 of the Brussels Regulation (subsequently article 25 of the recast Regulation), which gives effect to jurisdiction agreements, because they are "potestative", meaning performance is subject to or dependent upon an event which is in the sole power of one of the contracting parties. In some civil law countries, including France, a potestative term is void for lack of mutuality of obligation. However, this reasoning is not easily reconcilable with the Brussels regime. Indeed, outside the provisions which regulate jurisdiction clauses in consumer, employment and insurance contracts, there is nothing in the Brussels Regulation (whether original or recast) which empowers courts to strike down clauses on grounds of imbalance. In reality, it is arguable that the French court was simply applying French law concepts of contractual justice to deny legal effect to these clauses.

In subsequent cases, the French courts have moved away from the consideration of whether a unilateral clause is invalid because it is potestative, focusing instead on whether the clause sets out objective factors against which the designated courts can be identified. In Danne (Cass., civ. 1ère, 25 March 2015, No. 13-27.264, considered here), the French court held that a clause which granted a bank the right to bring proceedings before "any other competent tribunal" was not enforceable on this basis. But in eBizcuss (Cass., civ. 1ère, 7 October 2015, No. 14-16.898, considered here), a clause was upheld where the choice of the party with the benefit of the clause was restricted to the courts where the other party had its registered office and the courts where any loss caused by the other party was suffered. That meant it was possible to identify, objectively, which courts might have jurisdiction in the event of a dispute.

In the light of these decisions, and in the absence of any CJEU authority on the question, there has been considerable uncertainty as to when an asymmetric jurisdiction clause will be enforced in France, or in other EU member states – as well as the risk that different member states may take different approaches. The recent reference to the CJEU on these issues may resolve this uncertainty.

Reference to the CJEU

On 13 April 2023, more than 10 years after the controversial Banque de Rothschild case, the French Cour de Cassation referred to the CJEU the following three questions, as noted above (the decision making the referral is reported here).

1. First, should the validity of a unilateral clause be governed by EU law or is it a matter for national laws? This question arises from an uncertainty around the wording of article 25 of the recast Brussels Regulation, which states that the "substantive validity" of a jurisdiction clause should be determined by the national law of the member state whose courts are chosen. The CJEU will have to interpret the width of this provision. It is accepted that grounds such as fraud, duress or mistake relate to "substantive validity", but the question is whether it should be given a broader scope, to include the effects of asymmetry.
2. Second, if the CJEU decides that EU law applies, does EU law prohibit or restrict this type of clause? This second question will require the court to grapple with the French courts' reasoning in the line of cases referred to above. In particular, it is likely to consider whether there is a requirement that jurisdiction clauses identify the designated courts by reference to objective factors. It may also address the question of whether considerations of contractual justice should be able to prevent or curtail the use of unilateral clauses, beyond the specific circumstances of consumer, employment and insurance contracts.
3. Third, if the CJEU decides that national laws apply, which state's law should be applied when the clause points at several chosen courts, or designates one court while allowing one of the parties to choose another court (and this choice has not yet been made)? The answer is far from certain.

Given the uncertainty surrounding these questions, the CJEU's ruling will provide clarity as to the approach to be adopted by EU member state courts. However, how is this ruling relevant from a UK perspective?

Relevance to the UK

At first sight, the anticipated CJEU decision might be thought to be of little relevance from a UK perspective because the recast Brussels Regulation no longer applies to the UK in light of Brexit (except where proceedings were instituted before the end of the Brexit transition period, ie before 31 December 2020). Accordingly, even if the CJEU finds that asymmetric clauses are not compatible with article 25, that will not affect the approach of the English courts, which will no doubt continue to consider such clauses valid and enforceable.

A CJEU ruling that asymmetric clauses are invalid would, however, have an impact on parties which had chosen English court jurisdiction under such a clause – as in the example above, where the borrower is required to bring any action in England but the lender can choose to sue the borrower in any court of competent jurisdiction.

In that scenario, if the borrower brought proceedings in an EU member state (eg France), the French court would presumably find that the clause is invalid and therefore accept jurisdiction if there was any ground for doing so under the recast Brussels Regulation (if the defendant was domiciled in an EU member state) or under national rules of jurisdiction (if the defendant was non-EU domiciled). At the same time, the English court would likely rule that the clause was valid and, on the lender's application, could issue an anti-suit injunction to restrain the borrower from continuing the French proceedings. (Issuing an anti-suit injunction against as EU member state court would not have been possible before Brexit, as it was considered by the CJEU to be an interference with the jurisdiction of the other EU court, and contrary to the Brussels regime. But that constraint no longer applies.)

This risk of a jurisdiction battle is exacerbated by the fact that the 2005 Hague Convention on Choice of Courts Agreements (to which both the EU and the UK are parties) only applies to exclusive jurisdiction clauses, which for these purposes almost certainly excludes unilateral clauses (see the Court of Appeal's decision in Etihad Airways PJSC v Lucas Flother [2020] EWCA Civ 1707 and our blog post here). So an EU member state court would not be required by Hague 2005 to give effect to an asymmetric clause in favour of the English court.

Even if the CJEU rules in favour of unilateral jurisdiction clauses, there is still a risk that an EU member state court would not give effect to such a clause in favour of the English court in all circumstances. This is because there is some uncertainty as to whether, under the recast Brussels Regulation, EU courts can stay proceedings or decline jurisdiction in favour of non-EU courts, as the English court now is, except where the non-EU proceedings were commenced first in time (in which case articles 33 and 34 give a discretion to stay proceedings in favour of identical or related proceedings instituted before a non-EU court). In Gulf International Bank BSC v Aldwood [2019] EWHC 1666 (QB) (see our post here), the English High Court found that articles 33 and 34 are exhaustive of a member state court's powers to stay proceedings in favour of a non-EU court. The Court of Appeal touched upon the issue without resolving it in JSC Commercial Bank Privatbank v Kolomoisky and Bogolyubov and others [2019] EWCA Civ 1708 (see our post here). So far as we are aware, there is no CJEU authority on the point.

Accordingly, in our example above, if the lender brought proceedings in the English court before the borrower commenced the French proceedings, the French court would have a discretion (but not an obligation) to stay its proceedings under articles 33 and 34. But where the French court was seised first, it is not clear whether the French court could defer to the English court, even if it wished to do so and considered the asymmetric clause to be valid and enforceable. So again, a jurisdiction battle could ensue, with the French court accepting jurisdiction despite the clause and the English court (potentially) issuing an anti-suit injunction to enforce the clause.

For these reasons, there is likely to be continued uncertainty as to enforceability of asymmetric jurisdiction clauses in favour of the English courts in some circumstances, regardless of the CJEU's decision. However, a decision in favour of validity would clearly be a helpful step toward upholding the bargain arrived at by parties choosing such clauses, and put them in a stronger position when it comes to enforcing such clauses.

Hybrid arbitration clauses

Finally, the CJEU judgment may also be relevant to hybrid arbitration clauses. The line of French caselaw following the Banque de Rothschild case related to the validity of asymmetric court jurisdiction clauses. The scope of the referral to the CJEU also appears to be strictly limited to the validity of asymmetric jurisdiction clauses. However, it is common for finance parties to have asymmetric clauses which require the borrower to arbitrate a dispute, but allow for the lender to opt to litigate in a specific court or in any court with jurisdiction (or vice versa). The CJEU response to the reference could have far reaching and potentially unintended consequences for these hybrid arbitration clauses unless these wider implications are recognised and the CJEU's judgment carefully delineated in its application.