The Law Governing Arbitration Agreements in England after UniCredit

This post was written by Manuel Penades, who is Reader in International Commercial Law at King’s College London. It is the first contribution to the EAPIL online symposium on the UK Supreme Court Judgment in UniCredit Bank GmbH v RusChemAlliance LLC [2024] UKSC 30. The other contributions, by Faidon Varesis and Etienne Farnoux, can be found here and here, respectively. Readers are encouraged to participate in the discussion by commenting on the posts.

1. The Relevance of the Law Governing Arbitration Agreements to English Antisuit Injunctions

The law governing arbitration agreements has recently attracted significant attention in English arbitration law. Compared to other jurisdictions, the issue has become disproportionally complex and the UKSC has done little to improve the situation. The relevance of the topic is not limited to cases where the jurisdiction of arbitrators is in question. Cases such as UniCredit Bank GmbH v RusChemAlliance LLC [2024] UKSC 30, Enka Insaat Ve Sanayi AS v OOO Insurance Company Chubb [2020] UKSC 38 or Sulamérica Nacional de Seguros SA v Enesa Engenharia SA [2012] EWCA Civ 638 show that the law governing arbitration agreements impacts directly on the court’s powers to enforce contractual promises to arbitrate. These three cases concerned the possibility to issue an antisuit injunction preventing a party from litigating claims allegedly subject to arbitration agreements. In Enka, the applicable law was critical to define whether the dispute in question was covered by the scope of the arbitration agreement. In Sulamérica, the effectiveness of the arbitration agreement was dependant on the law applicable to it. In both cases, the seat of the arbitration was located in England.

2. The Law Governing the Arbitration Agreement in UniCredit

The question in UniCredit was whether the English courts had jurisdiction to order a Russian company not to pursue court proceedings in Russia against a German bank when the parties had agreed to settle any dispute between them by arbitration in Paris under the ICC Rules and had chosen English law to govern their contracts (bonds in this case).

The English court’s equitable jurisdiction to grant an injunction ‘in all cases in which it appears to the court to be just and convenient to do so’ arises from section 37(1) of the Senior Courts Act 1981. Yet, when the defendant is not in the jurisdiction and the seat is not in England and Wales, the power of the court to issue an antisuit injunction is dependant on two requirements. First, that the claim falls under one of the gateways provided in para 3.1 of Practice Direction 6B allowing service of the claim against the foreign defendant out of the jurisdiction. Second, that England and Wales is the ‘proper place’ in which to bring the claim. The UKSC addressed both matters in UniCredit but this post is only concerned with the first requirement, as the second is discussed in this symposium by Faidon Varesis.

The gateway on which Unicredit relied is available when ‘a claim is made in respect of a contract […] governed by the law of England and Wales’ (para 3.1(6)(c) PD6B). The bonds stated that they were governed by English law but RusChemAlliance argued (and Unicredit did not contest until the last instance, see [19]) that the contract with which the application for antisuit injunction was concerned was the arbitration agreement, whose breach the injunction sought to impede. The bonds did not contain any selection of applicable law specifically addressed at the arbitration agreement. The question then was whether English law governed the arbitration agreement pursuant to the general choice of law provided in a different clause of the bonds.

To answer this question, the UKSC revisited the test that was established just four years ago in Enka. The relevant passages of that test read ([170]):

iv) Where the law applicable to the arbitration agreement is not specified, a choice of governing law for the contract will generally apply to an arbitration agreement which forms part of the contract.

v) The choice of a different country as the seat of the arbitration is not, without more, sufficient to negate an inference that a choice of law to govern the contract was intended to apply to the arbitration agreement.

According to these rules, the general choice of English law to govern the bonds was equally applicable to the arbitration agreements contained in them. The principle of separability did not prevent this result as, in the words of the UKSC in Enka, ‘separability is not a principle that an arbitration agreement is to be treated as a distinct agreement for all purposes but only that it is to be so treated for the purpose of determining its validity or enforceability [and] thus, does not require that an arbitration agreement should be treated as a separate agreement for the purpose of determining its governing law’ ([41]).

RusChemAlliance, however, argued that this conclusion should be displaced by one of the two exceptions that the UKSC had established in Enka, particularly [170(vi)(a)]:

vi) Additional factors which may, however, negate such an inference and may in some cases imply that the arbitration agreement was intended to be governed by the law of the seat are: (a) any provision of the law of the seat which indicates that, where an arbitration is subject to that law, the arbitration agreement will also be treated as governed by that country’s law.

According to RusChemAlliance, the fact that under French law (as the law of the seat) the arbitration agreements in the dispute at hand would be governed by the French rules applicable to international arbitration agreements meant that the choice in the bonds in favour of English law could not be extended to the arbitration agreements. The UKSC disagreed, and concluded that assuming this level of legal foresight by the mere choice of Paris as seat went ‘beyond what it may in practice be realistic to expect’ [52]. The ‘natural’ [22] interpretation was to conclude that the parties wanted English law to govern every clause in the bonds. Therefore, English law applied to the arbitration agreements and the case fell within gateway 3.1(6)(c) PD6B. More generally, the UKSC conceded that the exception in [170(vi)(a)] of Enka created significant uncertainty and decided to remove it.

3. The Lessons from UniCredit

Two main lessons follow from this part of the decision in UniCredit.

3.1. The residual importance of implied choice of law for arbitration agreements

First, it confirms the residual role of implied choice of law, which in effect is squeezed out of the common law doctrine of the proper law of the contract for arbitration agreements. Enka and Kabab-Ji SAL (Lebanon) v Kout Food Group (Kuwait) [2021] UKSC 48 had already ruled that a choice of law to govern the contract should be generally construed as applying to the arbitration agreement set out (or incorporated by reference) in a clause of the contract. Yet, they had not excluded the possibility of an implied choice of law for arbitration agreements (in line with the traditional threefold test of the common law doctrine of the proper law). In fact, the extension of the choice of law from the contract to the arbitration agreement could have been seen as an implied choice of law inasmuch the UKSC acknowledged that such extension was ‘an inference’ which could be displaced in the exceptions provided in [170(vi)] to ‘imply that the arbitration agreement was intended to be governed by the law of the seat’.

The UKSC went a step further in UniCredit and declared that ‘if it were necessary or relevant to characterise the choice of law for the arbitration agreement signified by such a governing law clause as “express” or “implied”, I think it would be more apt to call it an “express choice” because it is identified by interpreting the express terms of the contact and is not based on any implied term’ ([27]). That is, what in the past could have been seen as an implied choice of law based on the terms of the contract or the circumstances of the case is now an express choice of law, which has simply been extended to cover the arbitration agreement using ‘the rules of contractual interpretation of English law as the law of the forum’ ([21], in line with [170(iii)] of Enka).

Enka also confirmed that ‘where there is no express choice of law to govern the contract, a clause providing for arbitration in a particular place will not by itself justify an inference that the contract (or the arbitration agreement) is intended to be governed by the law of that place’ ([170(vii)] of Enka). A reader could have reasonably understood this rule simply to mean that, in the absence of choice of law for the matrix contract, the choice of seat does not imply a choice of the seat’s law to govern the arbitration agreement. In fact, the UKSC had defined the implication of terms as a ‘process of inference’ in [35] of Enka.

With this background in mind, there was every reason to believe that the use of the term ‘inference’ in [170(v), (vi) and (vii)] of Enka referred to the possibility to construe a choice of law for the matrix contact or a choice of seat as implied choices of law for the arbitration agreement.

The result after UniCredit, read together with Enka, however, is that the ‘inference’ in [170(v) and (vi)] is, in reality, an express choice. The UKSC had already anticipated this reading in [39] of Kabab-Ji. Implied choice will, therefore, be reduced to two narrow scenarios.

One concerns instances where: 1) there is no express choice of law to govern the contract, 2) the parties have chosen an arbitral seat, and 3) something else in the arbitration agreement (whatever that might be) clearly indicates an undeclared intention to have the contract to arbitrate governed by the law of the seat. In practice, however, these cases will probably avoid the embroilment of the implied choice test and be resolved under the closest and most real connection test, which will generally also lead to the law of the seat (as per [170(viii)] of Enka).

The other possible avenue for implied choice will be the instances where parties select the law governing the matrix contract impliedly. The majority in Enka did not discuss the possibility to ‘infer’ that an implied choice of law for the main contract extends to the arbitration agreement. This is a significant gap between steps (vii) and (viii) in [170] of Enka and later decisions by the UKSC have failed to fill it. It was only the minority in Enka that addressed this possibility. In their view, it would be a ‘natural, rational and realistic’ inference and ‘simply the correct objective interpretation of the parties’ main contract and arbitration agreement’ ([228]). It is reasonable to expect this question to reach the UKSC in the future.

It is difficult to envisage a viable argument in favour of implied choice of law for arbitration agreements outside the discussed scenarios.

3.2. The uncertainty around the new section 6A of the Arbitration Act 1996 and a proposal to resolve it

The proposed choice-of-law rule for arbitration agreements in the Arbitration Bill provides that

6A Law applicable to arbitration agreement

(1) The law applicable to an arbitration agreement is—

(a) the law that the parties expressly agree applies to the arbitration agreement, or

(b) where no such agreement is made, the law of the seat of the arbitration in question.

(2) For the purposes of subsection (1), agreement between the parties that a particular law applies to an agreement of which the arbitration agreement forms a part does not constitute express agreement that that law also applies to the arbitration agreement. […]

After years of uncertainty and a lengthy review process, the new rule offers the possibility of a fresh start. The new section 6A eliminates implied choice of law entirely and replaces the closest and most real connection test with a hard-and-fast rule in favour of the law of the seat. The Bill also refines the wording of section 6A(2) to avoid litigation about the meaning of an express choice of law concerning the arbitration agreement. Yet, the UKSC has muddled the waters in UniCredit. By confirming that a generic express choice of law for the matrix contract, even without the definition of ‘Agreement’ as in Kabab-Ji, is an express choice for the arbitration agreement, it has opened the floodgates to the type of litigation that the introduction of section 6A tries to avoid. As argued in my written evidence before the House of Lords Committee tasked with the Arbitration Bill, what will follow from the adoption of the Bill is a new type of litigation concerned with whether an express choice of law is ‘sufficiently’ express to satisfy section 6A.

The arbitration agreements in UniCredit serve as illustration. The contracts provided that ‘This Bond and all non-contractual or other obligations arising out of or in connection with it shall be construed under and governed by English law’. This wide choice of law clause in not just ‘an agreement between the parties that a particular law applies to an agreement of which the arbitration agreement forms a part’ (section 6A(2)). The clause goes beyond the agreement (the bonds) and reaches ‘all non-contractual or other obligations arising out of or in connection with’ them. Would this be express enough for the purposes of section 6A(1)(a)? Importantly, in the transition between the Second Consultation and the Final Report, the Law Commission excluded the requirement that the choice of law clause for the arbitration agreement had to be ‘in the arbitration agreement itself’, which means that it should not be impossible to have express choices that are not in the arbitration agreement or that do not refer to it expressly.

This uncertainty could have been largely avoided by the UKSC.

The Court could have decided that the type of ‘inference’ in UniCredit (which followed [170(v)] of Enka) was just a case of implied choice of law for the arbitration agreement. This finding would not have operated under section 6A, which excludes implied choice. Alternatively, the UKSC could have clarified that, while such ‘inference’ might be deemed an express choice under the common law doctrine of the proper law, this would no longer apply under section 6A given the exclusion in section 6A(2). Yet, the Court refused to make such reassuring statements and, instead, declared that ‘depending on what the word “expressly” is taken to add to the word “agree”, this [ie, section 6A(2)] would not by itself alter the law as stated in Enka’ ([28]). What is more, the UKSC insisted on the idea, already mentioned in Enka, that ‘it does not matter’ whether a choice of law is express or implied, as ‘the distinction is of no legal significance’ ([27]). The reality is that, while express and implied choice might produce the same effects in common law, the distinction is of critical legal significance under section 6A. Express choice will be effective whereas implied choice will not. Turning its back to the imminence of section 6A and refusing to avoid unnecessary confusion was the wrong path for the UKSC. The reputation of English arbitration law and London as arbitral seat will suffer for it, and so will the parties who choose to arbitrate in London.

The paradox over the last years is that, while the proposal behind section 6A was motivated (in a significant part) by the ‘complex and unpredictable’ choice of law test for arbitration agreements after Enka (see Law Commission, ‘Review of the Arbitration Act 1996: Final Report and Bill’, para 12.20), that test has been simplified after UniCredit by eliminating the problematic exception in [170(vi)(a)] of Enka. In parallel, the definition of express choice of law confirmed in UniCredit will complicate the interpretation of section 6A and be a source of major uncertainty, which will require further court intervention.

While unlikely, this concern would justify a reconsideration of section 6A by Parliament before the passing of the Bill. A viable compromise would be to replace the current wording with the following:

(2) For the purposes of subsection (1), agreement between the parties that a particular law applies to an agreement of which the arbitration agreement forms a part constitutes express agreement that that law also applies to the arbitration agreement.

This wording would capture the essence of the common-sense approach insisted upon by the UKSC in Enka, Kabab-Ji and UniCredit while reaching a compromise between most of the voices that have participated in the review process. It would preserve party autonomy as well as the requirement of an express choice (hence eliminating the uncertainties around implied choice of law). It would also satisfy those who invoke the appropriateness to align the contract with the arbitration agreement while maintaining the default rule in favour of the law the seat.

4. The Jurisdiction of English Courts to Issue Antisuit Injunctions Concerning Arbitration Agreements Seated Outside of England & Wales When Solely the Matrix Contract is Governed by English Law

The analysis in UniCredit confirmed that, subject to the proper forum test, English courts have jurisdiction to issue antisuit injunctions regarding arbitrations seated abroad when the contract contains an English law clause and the parties have not indicated that the arbitration agreement is governed by any other law. The jurisdictional hook was the law governing the arbitration agreement.

The UKSC, however, did not exclude the possibility that this jurisdiction might also exist when the matrix contract alone, and not the arbitration agreement, is governed by English law. Unicredit only suggested this argument late in the appeal before the UKSC and the judgment did not address it. There is little doubt that this possibility will be explored in future cases, and rightly so.

While arbitration agreements might be separable from the matrix contract for validity and choice of law purposes, it is hard to deny that Unicredit’s request for an antisuit injunction before the English courts was made in respect of the contract that contained those agreements. In the context of the choice of law analysis, the UKSC declared in UniCredit that ‘even if the obligations created by the arbitration agreement were regarded as separate from the bond contract for this purpose, they are on any view “obligations arising … in connection with” the bond’ ([31]). It would be surprising if the expressions ‘in connection with’ and ‘in respect of’ received such disparate interpretations by English courts that one led to the inclusion of arbitration agreements contained in the document, whereas the other excluded them.

Further, case law has given a broad interpretation to the words ‘in respect of’. In Albon (t/a N A Carriage Co) v Naza Motor Trading SDN BHD & Anor [2007] EWHC 9 (Ch), Lightman J decided that ‘the formula of words in CPR 6.20(5) “in respect of a contract” [now para 3.1(6)(a) PB6B] does not require that the claim arises under a contract: it requires only that the claim relates to or is connected with the contract.  That is the clear and unambiguous meaning of the words used’ ([27]). The judge reinforced this reading by reference to the decision of Mann CJ in Trustees Executors and Agency Co Ltd v Reilly [1941] VLR 110, 111, who stated that

The words ‘in respect of’ are difficult of definition, but they have the widest possible meaning of any expression intended to convey some connection or relation between the two subject matters to which the words refer.

Deripaska v Cherney [2009] EWCA Civ 849, [67], has confirmed this broad interpretation.

5. Conclusion

Pacta sun servanda lies at the core of UniCredit. This makes it an appealing decision which confirms England as a stronghold for arbitration and paves the way for a new field of arbitration-related litigation before English courts. Yet, it also complicates unnecessarily the success of the review of the Arbitration Act 1996. The UKSC should have achieved the former without causing the latter.