On 9 October 2020, the Supreme Court handed down its much-awaited decision in Enka v Chubb [2020] UKSC 38, in which the central issue was how an arbitral tribunal applying English law should approach the issue of determining the proper law to be applied to questions arising in relation to an arbitration agreement. In an earlier piece, published in July 2020 shortly before oral argument on the appeal took place (see “The Proper Law of an Arbitration Agreement”, Outlook, July 2020), we considered this issue and sought to anticipate what the Supreme Court might do. Among other things, we expressed the hope that the Supreme Court might bring certainty, if not also conceptual clarity, to this previously muddled area. As we seek to develop, we feel the Supreme Court has largely achieved this, although understandably there remain at least a few points still to be resolved in future cases.

In this piece, we will first consider in some detail what the Supreme Court did decide and what it may have left open for another day. We will also reflect briefly on some of the other cases we considered in our earlier piece to see how they might stand in light of the Supreme Court’s decision. We then move on to consider a number of specific issues that either were considered by the Supreme Court or remain still to be resolved in light of the decision, as well as offering what we see as some key practical tips for practitioners going forward.

We addressed the significance of this issue as to the proper law of an arbitration agreement in our earlier piece. Accordingly, here we turn straight to the Supreme Court’s decision.

PART A: THE DECISION

The facts

The case involved a claim made under a construction contract in respect of a catastrophic fire at a power plant in Russia. Following various subrogations and assignments, the insurer of the project (Chubb) commenced court proceedings in Russia under the construction contract seeking a finding of liability against the party responsible for the design and construction of the plant (Enka). Enka commenced an English-seated arbitration and also commenced an arbitration claim in the English courts seeking an anti-suit injunction to restrain the Russian proceedings on the ground that that the construction contract contained an arbitration agreement providing for arbitration in England. The lower courts in England came to different decisions as to whether or not it was appropriate to grant the anti-suit injunction. In large part, the decisions turned on matters relating to the validity and scope of the arbitration agreement which depended on the answer to the preliminary question of which system of law (English or Russian) was to be applied to determine those questions: where, as was the case here, there was no express choice of law clause in either the construction contract or in the arbitration agreement itself, how ought the proper law of the arbitration agreement be determined and to what issues concerning the arbitration agreement should that law properly be applied?

Analysis of the decision

The Supreme Court’s decision was split 3 – 2; the majority upheld the Court of Appeal decision but on different grounds. These divisions may give the appearance that the law remains as confusing as it was. However, while the Supreme Court disagreed sharply with the Court of Appeal’s reasoning, the 3 – 2 split in the Supreme Court is much less divided than it appears. The majority and minority agreed on more than they disagreed.

There are only two main points of departure between the minority and the majority. These are:

1. The implication of the implied law of the main contract in the absence of an express choice of law clause – a difference on application of principle to the facts and not of principle itself (on the facts the minority took a broader view of implying a choice of law than the majority).
2. The principle to be applied to the choice of law with the closest connection. The majority favoured a default choice of the law of the seat. The minority did not agree there should be a default choice and held in most cases the arbitration agreement would have the closest connection with the main contract law, however determined.

It is also important to bear in the mind the split arose on the rather uncommon facts of the case – there was no express choice of law in the main contract. It was only because there was no express choice of law that:

1. The implication of an implied law for the main contract became relevant; and also
2. In the absence of an implied choice of law for the arbitration agreement, the law with the closest connection became relevant.

All five judges agreed that, had there been an express choice of law in the main contract, this choice of law would either be an express or implied choice of law for the arbitration agreement as well. In our view the Supreme Court decision settles the position for the majority of cases where the main contract contains an express choice of law clause.

The majority took a broad view on the express choice of law clause in the main contract – that it is likely to be an express choice of law for the arbitration agreement as well. On the basis of this authority, there may be more cases finding that an express main contract law is the express choice of law for the arbitration agreement as well.

All five judges agreed the law of the arbitration agreement is determined by applying English common law rules for resolving conflicts of laws (rather than the Rome I Regulation (“Rome I”). The law to be applied to the arbitration agreement will be:

1. The law chosen by the parties to govern it – either express or implied; and
2. In the absence of any choice, the system of law the arbitration agreement is most closely connected to.

There is no difference between an implied choice and an express choice. An implied choice is still a choice which is just as effective.

Whether the parties have agreed a choice of law to govern the arbitration agreement is ascertained by construing the arbitration agreement and main contract as whole, applying the rules of contractual interpretation of English law as the law of the forum.

It is generally reasonable to assume parties intend or expect their contract to be governed by a single system of law. To apply different systems of law to different parts of a contract has the potential to give rise to inconsistency and uncertainty. 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.

Exceptions to application of the main contract law

Validation principle

All five judges agreed on the application of the validation principle – i.e. the application of the law of the contract to the arbitration agreement may be negated by the validation principle where there is a serious risk that, if governed by the same law as the main contract, the arbitration agreement would be ineffective.

Law of the seat

The majority also considered that the application of the law of the contract to the arbitration agreement may be negated if the law of the seat provides that, in the absence of an express choice, the arbitration agreement will also be treated as governed by the law of the seat (e.g. Swedish Arbitration Act, section 48 – in the absence of a choice of law, the arbitration agreement is governed by the law of the seat – and the Arbitration (Scotland) Act 2010, section 6). The minority was silent on this.

Rejection of reasoning of the Court of Appeal

The Court of Appeal’s reasoning hinged on a close connection or “overlap” between the law governing the arbitration agreement and the curial law (or law of the seat). This was one of the primary reasons for the Court of Appeal’s decision that the law of the seat, rather than the law of the main contract, would normally be the implied proper law of the arbitration agreement.

The Supreme Court found (all five judges agreeing) that almost all the provisions of the Arbitration Act 1996 (“AA 1996”) relied on to support the “overlap argument” (i.e. the connection between the curial law and the law of the arbitration agreement) are non-mandatory. Where the arbitration agreement is governed by a foreign law, by reason of section 4(5) AA 1996, the non-mandatory provisions of the Act which concern arbitration agreements do not apply.

The Supreme Court found the AA 1996 contemplates and specifically provides for a situation in which the arbitration agreement will be governed by a foreign law even though English law governs the arbitration process. Therefore, no inference can be drawn that, by choosing an English seat, and with it English law as the curial law, parties are also impliedly choosing English law to govern their arbitration agreement.

Points of disagreement between the minority and majority

The minority did not agree with the default application of the law of the seat as the law with the closest connection to the arbitration agreement – they preferred the application of the law of the main contract, even if determined as a rule of law by the closest connection – with the application of the validation principle displacing the main contract law, if necessary.

The majority favoured the law of the seat as the law with closest connection because this was consistent with international law embodied in Art V(1)(a) New York Convention, enacted into English law in section 103(2)(b) AA 1996.

The majority recognised the first limb of Art V(1)(a), “law to which the parties have subjected it”, includes an implied choice.

The majority also recognised the New York Convention is to be interpreted to apply the same conflicts rule to Art II(3) on recognition of arbitration agreements, i.e. the same choice of law rule applies pre- and post-award.

Another point on which the minority departed from the majority is whether the same choice of law rules applies to the scope as to validity of the arbitration agreement – Article V(1)(c) New York Convention; section 103(2)(d) AA 1996; Article 36(i)(a)(iii) Model Law.

The majority found the general approach in conflict of laws, adopted by both the common law and Rome I, is to treat the validity and scope of a contract (as well as other issues such as the consequences of breach and ways of extinguishing obligations) as governed by the same applicable law. This makes good sense, not least because the boundary between issues of validity and scope is not always clear. Thus, it is logical to apply the law identified by the conflict rules prescribed by article V(1)(a) of the New York Convention, and section 103(2)(b) AA 1996, to questions about the scope or interpretation of the arbitration agreement as well as disputes about its validity. Hence the majority was of the view the validation principle applies to questions of validity and scope of the arbitration agreement. The minority did not agree the same choice of law applied to “scope” as to validity. Lord Burrows and Lord Sales in the minority took a more restricted approach to the ambit of the validation principle. We think the majority’s view is supported by the scheme of the New York Convention.

Article II(3) of the New York Convention (enacted as section 9(4) AA 1996) supports the majority view. This requires the court to recognise and enforce an arbitration agreement (and to stay litigation brought in breach of the arbitration agreement) unless the agreement is “null and void, inoperative or incapable of being performed”. While not express, the scope of the arbitration agreement must fall within this enquiry – if the arbitration agreement does not cover the dispute then the court is not required to stay litigation brought before it in favour of arbitration. Therefore, the same governing law (and the same means of determining the governing law) should be applied to scope as to validity.

The majority left open whether the validation principle is applicable in the default rule – where the law of the seat applies as the law with the closest connection. The international approach most commonly associated with Gary Born’s work in this area would most likely invoke the validation principle as part of an implied choice analysis, so that the law of the closest connection would not normally fall for consideration. All five judges agreed there is no sharp distinction between an implied choice and a default positive rule of law.

The majority also found the fact that the contract requires the parties to attempt to resolve a dispute through good faith negotiation, mediation or any other procedure before referring it to arbitration will not generally provide a reason to displace the law of the seat of arbitration as the law applicable to the arbitration agreement by default in the absence of a choice of law to govern it.

Re-consideration of earlier authorities

Before Enka, cases in the High Court and Court of Appeal had approached the question of the proper law of an arbitration agreement from different perspectives and by applying conflicting presumptions. For that reason, the Supreme Court noted the Court of Appeal’s exhortation that “the time has come to seek to impose some order and clarity on this area of law”^[1]. Can previous decisions be rationalised in light of Enka? Lord Burrows said, “it is very difficult to rationalise all past cases.” In our view, several previous cases can be rationalised in light of Enka applying, where there is not clear evidence of an express choice of law for the arbitration agreement, a delicate re-interpretation, focusing on the validation principle and abandoning the ‘overlap’ reasoning.

The leading Court of Appeal case prior to Enka was Sulamérica v Enesa [2012] EWCA Civ 638. In Sulamérica, the main insurance contracts contained an express choice of Brazilian law. The dispute resolution clause provided for arbitration in London. The insureds claimed that the law of Brazil applied to the arbitration agreement, with the result that arbitration was ineffective without their present consent. The Court of Appeal held that the express choice of law in the main contracts did not apply to the arbitration clause. The Court did not imply a choice of law to fill the void and went to the third stage of the test, finding English law had the “closest connection” to the arbitration agreement.

The Supreme Court focused on the Court of Appeal’s reference in Sulamérica to the “serious risk” that the arbitration agreement would be futile if it was governed by Brazilian law. By emphasising the Court of Appeal’s implicit acknowledgment of the validation principle in Sulamérica, the Supreme Court has delicately re-cast that case.

If Sulamérica were decided post-Enka, it is most likely the Court would apply the validation principle because of the “serious risk” that the arbitration agreement would be ‘significantly undermined’ and that presumption that the main contract law would apply would be rebutted in favour of implying the law of the seat as the law of the arbitration agreement. There would be no need to go to the “closest and most real connection”. The conclusion would be the same, but the reasoning would be subtly different.

Other cases which pre-date Sulamérica may also be re-interpreted in a similar way. In C v D [2007] EWCA Civ 1282, a Bermuda form insurance policy provided for New York law, together with London-seated arbitration. Applying New York law to the arbitration agreement would likely have invalidated it. Longmore LJ commented, obiter, that the law of the seat was “more likely” to have the closest connection with the arbitration agreement than with the law of the main contract. If C v D were decided afresh, the majority in Enka would no doubt agree with the conclusion, but again, English law would like be implied through the operation of the validation principle rather than the closest connection test.

Likewise, the case of XL Insurance v Owens Corning [2001] 1 All ER (Comm) 530, which was another case involving a Bermuda form insurance policy, would be decided by reference to the validation principle and a rejection of the ‘overlap’ argument. It is clear that any suggestion that the choice of a curial law (by reference to the seat) determines the law governing the arbitration agreement has been firmly discredited.

In our view, abandoning the faulty ‘overlap’ reasoning, which has been relied upon to justify selecting the law of the seat as the law with the closest connection with the arbitration agreement, and replacing this overtly with the validation principle, explains the decisions in these previous cases.

In our earlier piece, we also considered the Court of Appeal’s decision in Kabab-Ji S.A.L v Kout Food Group [2020] EWCA Civ 6; [2020] 1 Lloyd’s Rep 269.^[2] In that case, there were certain indications relied on heavily by the Court of Appeal that an express choice of law in the main contract was intended to extend to its arbitration agreement as well: the governing law provision in the main contract was broadly worded, further provisions in the main contract were suggestive of an intention that this would extend to include the arbitration agreement, and the arbitration agreement itself contained wording to the effect that a tribunal was to apply “all provisions” of the main contract. The Court of Appeal’s approach met with the approval of the Supreme Court and, if anything, it is likely that the Supreme Court’s emphasis on the significance of the proper law of the main contract will result in a more resolute approach being taken to construing main contract proper law clauses as extending to arbitration agreements contained within them. Some commentators regarded the Court of Appeal’s decision in Kabab-Ji either as being heterodox or as marking the outer limits of what was permissible as a matter of contractual construction. However, in light of the Supreme Court’s decision in Enka, it may well be that this is now to be regarded as orthodox and a straightforward application of well-versed English canons of contractual construction.

In our earlier piece, we also touched on the leading authorities in Singapore. It will need to be seen how the Singapore courts will treat the Supreme Court’s decision in Enka when the proper law issues arises again, however, there may not be all that much practical difference between it and the approach taken by the Singapore Court of Appeal in BNA v BNB [2019] SGCA 84. The Singapore Court of Appeal had expressly approved of the approach of the Court of Appeal in Sulamérica, moreover, it had highlighted (as had the High Court in BCY v BCZ [2016] SGHC 249) the significance of the proper law of the main contract, particularly where there was an express choice, albeit that this was subject to displacement if application of that system of law was likely to defeat the parties’ clear intention to submit their disputes to resolution by arbitration. Ultimately, however, the decision in BNA v BNB turned on the matter of the identification of the seat of the arbitration rather than the proper law of the arbitration agreement, so assessment of the impact of Enka will need to await an appropriate case.

PART B: POINTS FOR FURTHER CONSIDERATION

The Supreme Court did not comment on how its analysis (and reference to the choice of law regime in Rome I) might be affected by Brexit. The decision did not turn on this. Further, the Court agreed English common law rules applied to the choice of law of the arbitration agreement (this choice of law being specifically excluded from the Rome I). Where Brexit might have an impact is on the identification of the choice of law of the main contract. The Court applied Rome I, as it had to. It is conceivable Rome I may cease to apply in the future in the wake of Brexit. Even so, this is unlikely to make a material difference as the common law and Rome I choice of law rules for contracts are closely aligned.

In the next part of this article we consider four points arising from the Supreme Court’s decision:

1. Choice of law in deciding the proper law of an arbitration agreement;
2. The approach to multi-tier dispute resolution clauses;
3. Free-standing arbitration agreements; and
4. What is the practical significance of the decision for practitioners, advisers and drafters alike.

Each warrants more detailed treatment than the present format permits.

Choice of law in deciding proper law

There is at least a query whether the Supreme Court’s decision to apply English rules of contractual interpretation, as the law of the forum, to determine whether the parties expressly or implied agreed a choice of law for the arbitration agreement is restrictive and parochial. But putting that to one side, the Supreme Court’s decision leaves open whether an arbitrator sitting in England is required to take the same approach.

Under section 34(2)(f) AA 1996, an arbitrator can decide whether to apply strict rules of evidence (or any other rules) as to the admissibility, relevance or weight of any material (oral, written or other) sought to be tendered on any matters of fact or opinion. This suggests an arbitrator has latitude to choose a different choice of law approach.

Under the Model Law:

1. Article 28(2) allows the tribunal to apply the law determined by the conflicts of law rules which it considers applicable to determine the law applicable to the substance of dispute.
2. Article 19(2) gives the tribunal the power to determine the admissibility, relevance and materiality and weight of any evidence

Taking these provisions together, under the Model Law, a tribunal arguably has a wider latitude to choose a different approach to the choice of law to be applied in determining whether the parties had agreed on a choice of law.

There is an argument for the application of internationally accepted principles of construction (giving as much effect as possible to the pro-arbitration policy of the New York Convention) in determining whether parties had agreed on proper law of the arbitration agreement.

While there is an argument that arbitrators should apply the same rules as the courts of the seat to foster consistency when the seat court reviews the decision in a setting aside or challenge, applying internationally accepted principle of construction may not arrive at a markedly different conclusion on interpretation of the parties’ intent.

Multi-tier dispute resolution clauses

The Supreme Court’s obiter analysis of multi-tier dispute resolution clauses (see in particular Lord Burrows at [235]-[237]) starts from the commonsense presumption that the law of the arbitration agreement and of other aspects of the dispute resolution clause would be intended by the parties to be the same. However, negotiation, mediation and other pre-arbitration dispute resolution obligations may be alien to the law (chosen expressly or impliedly) of the arbitration agreement. For example, the parties may have specifically intended negotiation obligations to be efficacious and a pre-condition to any more formal dispute resolution mechanism by way of escalation, but the drafting may be deficient applying one possible proper law but not others.

Given the developing state of the law and fact-specific construction required for negotiation and mediation clauses, including their relationship with arbitration as the agreed option of last resort, it may well be more difficult to formulate a satisfactory default approach to multi-tier dispute resolution clauses than the Supreme Court’s short observations suggest.

The divergent English cases are illustrative of the potential complexity. For example, in Cable and Wireless plc v IBM United Kingdom Ltd [2002] EWHC 2059 (Comm) and Emirates Trading Agency LLC v Prime Mineral Exports Private Ltd [2014] EWHC 2104 (Comm), first instance judges in the Commercial Court supported a departure from the strict approach of the House of Lords in Walford v Miles [1992] 2 AC 128 which had deprecated obligations to negotiate as too uncertain to be enforced. In turn, yet more recent Commercial Court decisions have cast doubt over that shift: Emirates Trading Agency LLC v Sociedade de Fomento Industrial Private Ltd [2015] EWHC 1452 (Comm) and DS-Rendite-Fonds Nr 106 VLCC Titan Glory GmbH & Co Tankschiff KG v Titan Maritime SA Panama [2015] EWHC 2488 (Comm).

By contrast, the Singapore courts have on the whole been more willing to regard negotiation obligations as pre-conditions to agreements to arbitrate, provided the drafting is sufficiently clear and certain: International Research Corp plc v Lufthansa Systems Asia Pacific Pte Ltd [2013] 1 SLR 973; [2012] SGHC 226 (upheld on appeal: [2013] SGCA 55). The approaches in civil law jurisdictions are again varied, making this question a potentially topical one.

Free-standing arbitration agreements

Free-standing arbitration agreements are occasionally referred to as submission agreements. They are typically entered into after a dispute has arisen. The starting point is that free-standing arbitration agreements are unusual. Enka was expressly concerned with arbitration agreements contained within “main contracts” and does not lay down any principles for determining the proper law of submission agreements.

The decision in Enka is clearly premised, in the usual course, on commercial parties being presumed to intend that a clear choice of law in the main contract would generally apply also to the arbitration agreement contained within it. That analysis recognises the close connection between the main contract and the arbitration agreement, notwithstanding the principle of separability (which is limited, per Enka, to the arbitration agreement surviving the main contract in the event the main contract is invalid, in effective or has its existence challenged, as provided in section 7 AA 1996).

The approach to free-standing agreements is unlikely to lend itself to a simple presumptive rule. For example, in some cases it will be perfectly sensible to apply an express choice of law in the main contract to a free-standing arbitration agreement (on the footing this was likely to be a drafting oversight or the parties’ presumed intention). In other cases, a free-standing arbitration agreement may apply to disputes arising out of a number of contracts between the same parties, governed by different laws; here, an Enka presumption is unworkable. In any event, it is more likely that free-standing arbitration agreements would have an express choice of the law governing the arbitration agreement, avoiding these questions.

That said, in the absence of a choice of law (express or implied) the New York Convention choice of law rule would still ask first for consideration of an implied choice (looking at the circumstances of the underlying dispute and perhaps application of the validation principle) and failing that application of the law of the seat.

Practical significance for practitioners

Although the Supreme Court’s decision does not resolve all issues, as one would not expect it to, it does provide very substantial clarity and this can only assist practitioners advising parties on potential outcomes when asked to advise on the validity or effect of an arbitration clause. In particular, the combined judgment of Lords Hamblen and Leggatt (with which Lord Kerr agreed) contains, at paragraph 170, a very clear and concise summary setting out the law (as they held it to be) which may well become the starting-point when advising clients hereafter. An important point to bear in mind, in light of the Supreme Court’s decision, is that, at the end of the day, irrespective of the approach to be taken to identifying and then applying a system of law to an arbitration agreement, there are simply some arbitration agreements that are “pathological”, i.e. they are so poorly drafted that effect cannot be given to them. This is a function of poor drafting and it is not a lacuna in the law relating to arbitration that this may be the unfortunate result in a particular case.

So far as the practicalities of drafting are concerned, the decision certainly clarifies how arbitration agreements and the contracts within which they appear may be construed, and this should factor into their drafting henceforth. One of the practical impacts is that it is now more likely that an express choice of law clause in a main contract will be regarded as extended to its arbitration agreement as well, so, if parties wish for some other law to regulate their arbitration agreement (typically, the law of the seat), they would be well-advised to make this as clear as possible.

We have often heard it said, and indeed have advised ourselves, that there may be merit to including an express choice of law stipulation within an arbitration agreement itself. But some caution needs to be exercised here. It may be that ultimately the application of the law expressly chosen serves only to invalidate the arbitration agreement (cf. Sulamérica) with little room to manoeuvre for a tribunal or court seeking to give effect to the parties’ evident intention to submit their disputes to arbitration. On the other hand, not including an express stipulation in the arbitration agreement may leave room for application of the validation principle. The appropriate course to take when it comes to including an express stipulation will of course depend on the parties’ bargaining priorities.

In addition, it is not clear from Enka whether there is any room for application of the validation principle in circumstances where there is no express or implied choice of law, and yet applying the law of the seat (at the closest connection stage) will invalidate the arbitration agreement. The minority expressly said the validation principle should apply at the closest connection stage (there being no default presumption). The majority left that point open as it did not arise in the case. In light of this, it may be preferable to ensure that safe seats are selected i.e. seats which have a pro-arbitration policy and which are highly unlikely to invalidate an arbitration agreement on idiosyncratic grounds.

CONCLUSION

The decision in Enka marks a substantial development in the wider international debate on the approach to identifying the proper law of an arbitration agreement. It is unlikely to be the final word on this rich and complex subject, but it is does settle the English law approach for the majority of cases where the main contract does include an express choice of law.

That said, this article has canvassed a number of remaining questions. The proper law debate is likely to remain topical for years to come.