In Kolassa v Barclays, another case where an investor in Eurobonds took action against the bank issuer, the ECJ allowed an investor to sue an issuer in the investor’s jurisdiction (Austria), on the basis that the investor’s securities account was in Austria and he had suffered the loss there. This was an application of the rules on tortious jurisdiction, as the ECJ held that the issuer had not freely consented to obligations owing to the investor and so the contractual rules on jurisdiction did not apply. This background makes this case doubly important: an effective jurisdiction clause may be the only way to avoid being sued in multiple investor jurisdictions, but the reasoning in Kolassa makes it difficult to see how a jurisdiction clause can be said to be agreed between issuer and investor. In particular, in Kolassa, the ECJ held that the chain of contracts between the issuer and investor meant that there was no agreement between them for the purposes of the contractual jurisdiction rules. While the test for agreement on a jurisdiction clause is slightly different to the test for contractual jurisdiction, it is still difficult to see how there can be an agreement for the purposes of the former rule but not the latter.
In this case, the ECJ attempted to set out specific requirements for a jurisdiction clause to be effective between issuer and investor. However, it is unclear how their judgment can be applied to bonds transferred in the European capital markets – the ECJ did not address the complexities of the global note structure and the interposition of common depositaries.
The ECJ held that a jurisdiction clause is ‘in writing’ if it is expressly referred to in every sale contract between the original subscriber and the ultimate investor. However, this does not reflect the realities of bond transfers. Investors are actually transferring interests held indirectly via participants in the clearing systems without extensive contractual documentation.
The alternative of a form that accords with international trade and usage may be more feasible. It will be unclear whether it might actually help until a national court pronounces on the compatibility of jurisdiction clauses with the factors set out by the ECJ.
The more serious issue is with the need for the clause to be ‘agreed’ by the issuer and investor. This is also where the tension lies with Kolassa, where there was not the requisite level of consent for the contractual jurisdiction rules of the Brussels Regulation to apply. The criterion set out by the ECJ is that the investor succeed to the same rights and obligations as the initial subscriber, by analogy with transfers of bills of lading or shares. But an investor in the capital markets typically acquires an interest in a securities account – that is, a bundle of rights against a financial intermediary which itself owns bundles of rights against further intermediaries which ultimately leads to a participant in the clearing systems. The global bearer bond itself is held by the common depositary and interests in it are recorded by the clearing systems (note that there are differences in the nature of those interests depending on whether the note is held using the classic global note structure or the new global note structure).
The result of this complexity is that, on the face of it, an investor in the notes does not succeed to the same rights and obligations as the initial subscriber. Neither the initial subscriber nor the investor are at any point the legal owner of the notes. The ultimate investor’s interest is likely to be in an account with a financial intermediary entirely independent of the bonds. It is unclear how the ECJ’s criteria for a jurisdiction clause to be ‘agreed’ could be satisfied in practice.
It is to be hoped that future judgments of the ECJ or national courts will clarify how all the criteria set out by the ECJ in this case can be satisfied by investors in the European capital markets.