On 25 March 2015 the Court of Cassation of France handed down a new decision dealing with the so called “unilateral”, “optional”, “hybrid”, and “asymmetrical” jurisdiction clauses giving choice to one contractual party where to bring action against the other. Earlier in 2012, the same court issued the much discussed Rothschild decision which stated that such jurisdiction clauses are to be deemed invalid as per French law and implicitly supported a line of case law from other countries which treated the issue in the same manner. Now, with its newest judgment, the Court of Cassation in fact confirmed its previous case law and even widened its scope.

The unilateral jurisdiction clauses (see generally D. Draguiev, Unilateral Jurisdiction Clauses: The Case for Invalidity, Severability or Enforceability (2014) 31 Journal of International Arbitration Issue 1, pp. 19–45) are widely used in a number of business sectors, primarily in credit and financing agreements. Such clauses provide that one of the parties to the agreement is granted an option to choose the forum where the dispute between the parties shall be decided, while the other party is entitled to bring action only in a particularly specified forum (e.g. particular arbitral institution or national court). The option is usually granted to the financing party which is entitled to choose to sue in a forum different from the one available to the borrower, and usually such clauses are drafted so that the financing party can bring its claim in any other jurisdiction. The rationale behind the architecture of a unilateral clause is that the lender can pursue the borrower and/or its assets in any jurisdiction it deems appropriate without restriction while the borrower may commence proceedings only in a particular forum.

Some jurisdictions see such clauses as fitting squarely into party autonomy and accept them without much ado. English courts are a good example in this regard (e.g. Mauritius Commercial Bank Ltd. v. Hestia Holdings Ltd. & Sujana Universal Industries Ltd. [2013] EWHC 1328). A number of other jurisdictions have, however, voiced concerns about unilateral clauses (e.g. France, Russia, Bulgaria, etc.). The main objection against unilateral clauses is that they contravene the basic procedural principle of equality of parties. Indeed, one party to the clause is granted under the agreement better opportunities to bring claims against the other. Some jurisdictions consider that this is permitted as it is based on consent while other deem this contrary to the balance between the parties.

The 2012 judgment of the Court of Cassation of France ruled in the case of Mme ‘X’ v. Banque Privée Edmond de Rothschild (the Rothschild case) that a unilateral clause is invalid. The agreement between a French national (natural person), Mme X, and Banque Privée Edmond de Rothschild provided that the disputes shall be referred exclusively to Luxembourg courts while Mme X decided to bring claim in a French court arguing that the jurisdiction clause was invalid. The Court of Cassation confirmed this. A number of reasons support the conclusion of the French court, including that the clause to opt where to bring claim is “potestative” – such clauses are triggered by a party’s own conduct, i.e. are discretionary and dependent only upon the party’s own choice. It was the bank’s choice where to bring claim and this was found “potestative” since the bank could unilaterally impose its forum selection on the lender. Furthermore, the dispute arose in a consumer scenario which was not specifically noted by the Court of Cassation but yet it might have had importance as consumers are treated as “weaker” and “protected” parties.

The judgment of 25 March 2015 concerned a French company and a Swiss bank (Credit Suisse). Under the jurisdiction clause, the French company was restricted in bringing action to the courts in Zurich while the Swiss bank could sue in any other competent court. The French claimant commenced proceedings in France in spite of the express forum selection. The forum selection (prorogation) clause was governed by the 2007 Lugano Convention, because the cross-border dispute involved a Swiss party. The main question turned to be: would Art. 23 of the Lugano Convention accommodate such unilateral jurisdiction clause? If so, the claim should have been brought in Switzerland, as this was what the parties chose; otherwise, granted that the clause could be found invalid, there was no prorogation under Art. 23 of the Lugano Convention and instead the default jurisdictional rules should apply. The Paris Court of Appeal held that Art. 23 of the Lugano Convention should apply and that there was express choice of exclusive court jurisdiction of Zurich courts, therefore French courts lacked jurisdiction. The Court of Cassation reversed.

The Court of Cassation succinctly commented that the clause providing for a right of the bank to commence proceedings before “any other competent court” was not based on objective elements and could not meet the requirement for legal certainty and predictability which jurisdiction clauses under Art. 23 of the 2007 Lugano Convention should have. The judgment does not speak further of the nature of the “objective elements” but it implicitly refers to the case law of the Court of Justice of the European Union (CJEU). In a judgment in the case C-387/98 the CJEU emphasised that the jurisdiction (prorogation) clause should state the objective factors on the basis of which the parties have agreed to choose a court or the courts to which they wish to submit disputes which have arisen or which may arise between them. Both Art. 23 of the Brussels I Regulation, now Art. 25 of the Brussels Ibis Regulation, and Art. 23 of the Lugano Convention are part of a single complex jurisdiction regime and are underpinned by one and the same principle: the parties should have sound understanding in which court they should appear if a dispute arises. This is even more important when the dispute has cross-border reach so there should be greater certainty in which countries the competent courts are located. The CJEU sought to imply that a valid jurisdiction clause has to provide sufficiently clear and precise hallmarks and criteria which courts the parties envision as competent under their disputes. If this is placed within the context of the Credit Suisse judgment, this might mean that a clause stating that a claim can be brought in “any competent court” (i.e. effectively everywhere under the respective applicable jurisdictional requirements) is too wide, too uncertain, and too imprecise. This would, then, contravene the requirement for predictability.

There are a number of implications arising out of the Credit Suisse Court of Cassation decision. First, it seems that reliance on unilateral jurisdiction clauses is far from providing certainty. The Rothschild judgment was widely criticised and this gave the impression that it is by no means final and should be revised. There were decisions in favour of unilateral clauses (e.g. in Spain: Decision of 18 Oct. 2013 of Court of Appeal of Madrid) after the ruling in Rothschild and this further supported the view that the reasoning in Rothschild is flawed and isolated. Instead, it was upheld. Second, the judgment of 25 March 2015, unlike Rothschild, deals with a purely commercial dispute in business-to-business settings. Third, the Credit Suisse judgment widens the effect of Rothschild. Typically such a clause has the following structure: one part of it provides for exclusive dispute settlement in a particular arbitral institution or national court if the claim is brought by one of the parties; the second part provides that the other party is not bound by the exclusivity and might, at its own preference, bring suit in any other court having jurisdiction to hear the matter. The very rationale of a unilateral clause is that one of the parties safeguards that it will be able to commence proceedings in any country where it considers that the debtor is located and/or has assets. If the second part of the clause is curtailed, the possibility that a lender traces and attempts to enforce against debtor’s assets might be sufficiently diminished.

The main argument in various decisions against unilateral clauses was that the parties under them are not balanced in their rights. Now, the new Court of Cassation judgment provides a completely new argument against unilateral jurisdiction clauses: that they are not sufficiently precise and predictable. What the Credit Suisse decision suggests is that the sweeping wording of a jurisdiction clause in its second part, e.g. “any court shall be competent”, does not meet the requirements for validity of the clause and renders it invalid. This puts a significant question mark as to how to draft a unilateral clause which aims to safeguard a lender or another “stronger” party and at the same time avoid including wide-reaching and undetermined phrases such as “any court”, etc. This has the potential to bring confusion to the entire area of law concerning prorogation clauses as well.


________________________

To make sure you do not miss out on regular updates from the Kluwer Arbitration Blog, please subscribe here. To submit a proposal for a blog post, please consult our Editorial Guidelines.

This page as PDF

One comment

  1. I would not be really prone to flag this case as related to arbitration — if not for other reasons because … it’s not an arbitration matter. Moreover, the French Cour de Cassation has decided back in 1970’s that an “asymmetrical arbitration clause” is valid and enforceable — see “Cour d’appel d’Angers”, of 25.09.1972, “La Gazette du Palais”. year 1973, 1.º Semestre, jan.-fev., p. 210 and decision of the “Cour de Cassation, Premiere chambre civile”, of 15.05.1974: Bull. civ. I, n. 143.
    For the time being, the validity of unilateral arbitration clauses is not disputed in France.
    There are also decisions in the US and UK following this understanding.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.