Determining the applicable law of an arbitration agreement when there is no express choice of a governing law – Enka Insaat Ve Sanayi A.S. v OOO Insurance Company Chubb [2020] UKSC 38.

This brief note considers aspects of the recent litigation over the identification of an unspecified applicable law of an arbitration agreement having an English seat. Though the UK Supreme Court concluded that the applicable law of the arbitration agreement itself was, if unspecified, usually to be the same as that of the contract to which the arbitration agreement refers, there was an interesting division between the judges on the method of determining the applicable law of the arbitration agreement from either the law of the arbitral seat (the view favoured by the majority) or from the applicable law of the underlying contract (the view favoured by the minority). As will become clear, the author of this note finds the views of the minority to be more compelling than those of the majority.

In a simplified form the facts were that, in February 2016, a Russian power station was damaged by an internal fire. ‘Chubb’, insurer of the owners of the power station, faced a claim on its policy. In May 2019, Chubb sought to sue ‘Enka’ (a Turkish subcontractor) in Russia to recover subrogated losses. Enka objected to these Russian proceedings claiming that under the terms of its contract of engagement any such dispute was to be arbitrated via the ICC in England: in September 2019, it sought declaratory orders from the English High Court that the matter should be arbitrated in England, that the applicable law of the arbitration agreement was English, and requested an English anti-suit injunction to restrain Chubb from continuing the Russian litigation.

Neither the arbitration agreement nor the contract by which Chubb had originally engaged Enka contained a clear provision specifically and unambiguously selecting an applicable law. Though it was plain that the applicable law of the underlying contract would, by the application of the provisions of the Rome I Regulation, eventually be determined to be Russian, the applicable law of the arbitration agreement itself could not be determined as directly in this manner because Art. 1(2)(e) of the Regulation excludes arbitration agreements from its scope and leaves the matter to the default applicable law rules of the forum.

After an unsuccessful interim application in September 2019, Enka’s case came before Baker J in December 2019 in the High Court. It seems from Baker J’s judgment that Enka appeared to him to be somewhat reticent in proceeding to resolve the dispute by seeking to commence an arbitration; this, coupled with the important finding that the material facts were opposite to those that had justified judicial intervention in The Angelic Grace [1995] 1 Lloyd’s Rep 87, may explain Enka’s lack of success before the High Court which concluded that the correct forum was Russia and that there was no basis upon which it should grant an anti-suit injunction in this case.

In January 2020, Enka notified Chubb of a dispute and, by March 2020, had filed a request for an ICC arbitration in London. Enka also however appealed the decision of Baker J to the Court of Appeal and duly received its requested declaratory relief plus an anti-suit injunction. The Court of Appeal sought to clarify the means by which the applicable law of an arbitration agreement should be determined if an applicable law was not identified expressly to govern the arbitration agreement itself. The means to resolve this matter, according to the court, was that without an express choice of an applicable law for the arbitration agreement itself, the curial law of the arbitral seat should be presumed to be the applicable law of the arbitration agreement. Thus, though the applicable law of the underlying contract was seemingly Russian, the applicable law of the arbitration agreement was to be presumed to be English due to the lack of an express choice of Russian law and due to the fact of the English arbitral seat. Hence English law (seemingly wider than the Russian law on a number of important issues) would determine the scope of the matters and claims encompassed by the arbitration agreement and the extent to which they were defensible with the assistance of an English court.

In May 2020, Chubb made a final appeal to the UK Supreme Court seeking the discharge of the anti-suit injunction and opposing the conclusion that the applicable law of the arbitration agreement should be English (due to the seat of the arbitration) rather than Russian law as per the deduced applicable law of the contract to which the arbitration agreement related. The UK Supreme Court was thus presented with an opportunity to resolve the thorny question of whether in such circumstances the curial law of the arbitral seat or the applicable law of the agreement being arbitrated should be determinative of the applicable law of the arbitration agreement. Though the Supreme Court was united on the point that an express or implied choice of applicable law for the underlying contract usually determines the applicable law of the arbitration agreement, it was split three to two on the issue of how to proceed in the absence of such an express choice.

The majority of three (Lords Kerr, Hamblen and Leggatt) favoured the location of the seat as determinative in this case. This reasoning did not proceed from the strong presumption approach of the Court of Appeal (which was rejected) but rather from the conclusion that since there had been no choice of applicable law for either the contract or for the arbitration agreement, the law with the closest connection to the arbitration agreement was the curial law of the arbitral seat. As will be seen, the minority (Lords Burrows and Sales) regarded there to have been a choice of applicable law for the contract to be arbitrated and proceeded from this to determine the applicable law of the arbitration agreement.

The majority (for the benefit of non-UK readers, when there is a majority the law is to be understood to be stated on this matter by  that majority in a manner as authoritative as if there had been unanimity across all five judges) considered that there was no choice of an applicable law pertinent to Art.3 of Rome I in the underlying contract by which Enka’s services had been engaged. It is true that this contract did not contain a helpful statement drawn from drafting precedents that the contract was to be governed by any given applicable law; it did however make many references to Russian law and to specific Russian legal provisions in a manner that had disposed both Baker J and the minority in the Supreme Court to conclude that there was indeed an Art.3 choice, albeit of an implied form. This minority view was based on a different interpretation of the facts and on the Giuliano and Lagarde Report on the Convention on the law applicable to contractual obligations (OJ EU No C 282-1). The majority took the view that the absence of an express choice of applicable law for the contract must mean that the parties were unable to agree on the identity of such a law and hence ‘chose’ not to make one. The minority took the view that such a conclusion was not clear from the facts and that the terms of the contract and its references to Russian law did indicate an implied choice of Russian law. As the majority was however unconvinced on this point, they proceeded from Art.3 to Art.4 of Rome I and concluded that, in what they regarded as the absence of an express or implied choice of applicable law for the contract, Russian law was the applicable law for the contract.

For the applicable law of the arbitration agreement itself, the majority resisted the idea that on these facts their conclusion re the applicable law of the contract should also be determinative for the applicable law of the arbitration agreement. Instead, due to the Art.1(2)(e) exclusion of arbitration agreements from the scope of the Regulation, the applicable law of the arbitration agreement fell to be determined by the English common law. This required the identification of the law with which the arbitration agreement was ‘most closely connected’. Possibly reading too much into abstract notions of international arbitral practice, the majority concluded that, in this case, the applicable law of the arbitration agreement should be regarded as most closely connected to the curial law of the arbitral seat. Hence English law was the applicable law of the arbitration agreement despite the earlier conclusion that the applicable law of the contract at issue was Russian.

As indicated, the minority disagreed on the fundamental issue of whether or not there had been an Art.3 implied choice of an applicable law in the underlying contract. In a masterful dissenting judgment that is a model of logic, law and clarity, Lord Burrows, with whom Lord Sales agreed, concluded that this contract contained what for Art.3 of Rome I could be regarded as an implied choice of Russian law as ‘… clearly demonstrated by the terms of the contract or the circumstances of the case’. This determination led to the conclusion that the parties’ implied intentions as to the applicable law of the arbitration agreement were aligned determinatively with the other factors that implied Russian law as the applicable law for the contract. Russian law was (for the minority) thus the applicable law of the underlying contract and the applicable law of the ICC arbitration (that, by March, 2020 Enka had acted to commence) was to take place within the English arbitral seat in accordance its English curial law. Lord Burrows also made plain that if had he concluded that there was no implied choice of Russian law for the contract, he would still have concluded that the law of the arbitration agreement itself was Russian as he considered that the closest and most substantial connection of the arbitration agreement was with Russian law.

Though the views of the minority are of no direct legal significance at present, it is suggested that the minority’s approach to Art.3 of the Rome I Regulation was more accurate than that of the majority and, further, that the approach set out by Lord Burrows at paras 257-8 offers a more logical and pragmatic means of settling any such controversies between the law of the seat and the law of the associated contract. It is further suggested that the minority views may become relevant in later cases in which parties seek a supposed advantage connected with the identity of the applicable law of the arbitration. When such a matter will re-occur is unclear, however, though the Rome I Regulation ceases to be directly applicable in the UK on 31 December 2020, the UK plans to introduce a domestic analogue of this Regulation thereafter. It may be that a future applicant with different facts will seek to re-adjust the majority view that in the case of an unexpressed applicable law for the contract and arbitration agreement that the law of the seat of the arbitration determines the applicable law of the arbitration agreement.

As for the anti-suit injunction, it will surprise few that the attitude of the Court of Appeal was broadly echoed by the Supreme Court albeit in a more nuanced form. The Supreme Court clarified that there was no compelling reason to refuse to consider issuing an anti-suit injunction to any arbitral party who an English judge (or his successors on any appeal) has concluded can benefit from such relief. They clarified further that the issuance of an anti-suit injunction in such circumstances does not require that the selected arbitral seat is English. The anti-suit injunction was re-instated to restrain Chubb’s involvement in the Russian litigation proceedings and to protect the belatedly commenced ICC arbitration.