Implied Choice of Law for Arbitration Agreements: Modify or Abandon Completely?

The recently revised English Arbitration Act 2025 (see coverage on the Blog here) has reignited the debate on the applicable law to arbitration agreements. According to the Act, the choice of law must be made explicitly and, in the absence of such a choice, the law of the seat applies. English law has thus departed from the long-standing case law (prominently Sulamérica v Enesa and Enka v Chubb), which seeks to determine the applicable law according to a three-step test: (i) by an express choice of law by the parties, (ii) by an implied choice of law by the parties or (iii) by the law with the closest and most real connection, usually the law of the seat of arbitration.

The following post examines the second part of the three-stage test – the implied choice of law. Common law jurisprudence tends to base the implied choice of law on presumptions about the parties’ possible intentions rather than their actual intentions. Contrary to this tendency, the implied choice of law should rather be assessed in accordance with instruments of private international law, which impose stricter requirements on the implied choice of law and rely on the parties’ manifested intentions. The aforementioned English Arbitration Act reflects this high threshold for party intentions by completely abandoning the implied choice of law and referring solely to an explicit choice of law.

Implied Choice of Law in Case Law – The Generalisation of Party Intentions

The implied choice of the law applicable to the arbitration agreement has become the subject of a debate that seeks to derive the choice of law either from the law of the main contract or from the law of the seat of arbitration. In doing so, the implied choice of law has become detached from the circumstances of the individual case and the parties’ real intentions. Instead, the trend is towards making presumptions as to what reasonable parties might have hypothetically intended.

In Sulamérica v Enesa, the England and Wales Court of Appeal (EWCA) uses as a starting point “the assumption that the parties intended the same law to govern the whole of the contract” but then analyses a number of factors in order to determine the parties’ intentions. In Habas Sinai v VSC Steel Company, this starting point develops into a “strong indicator” in favour of the law of the main contract, to the extent that “[t]he choice of a different country for the seat of the arbitration [...] may not in itself be sufficient to displace the indication of choice implicit in the express choice of law to govern the matrix contract”. While contrary party intentions are possible, the hurdle to exonerate this “strong indicator” seems high. As a result, various indicia for the parties’ intentions are no longer weighted equally; rather, the court presumes a choice in favour of the law of the main contract. The decisions in Enka v Chubb equally determine the implied choice of law in the abstract. The EWCA supports a “general rule” in favour of the curial law, which can only be invalidated by “powerful reasons to the contrary”. With a different result but no less general, on appeal the Supreme Court of the United Kingdom (UKSC) considers it “natural to interpret such a governing law clause, in absence of good reasons to the contrary, as applying to the arbitration clause”, proceeds from an “assumption that a choice of law for the contract will normally apply to an arbitration clause” and speaks of a “general rule” and a “matter of principle”.

This contradiction between implied choice of law (dependent on real party intentions) and presumptions (independent of party intentions) becomes even clearer when the arguments underlying the presumptions are analysed. The doctrine of separability is invoked by both camps: those applying the law of the seat invoke the doctrine to separate the arbitration agreement from the main contract and consequently determine the applicable law independently (e.g. the EWCA in Enka v Chubb). Those favouring the law of the main contract do not see the doctrine of separability as an obstacle when extending the applicable law (e.g. the UKSC in Enka v Chubb). The fact that the doctrine of separability itself emphasises the need to determine the actual intentions of parties is often overlooked. The suffix “-ability” clarifies that separability is a possibility and not a rule (see Plavec). Although it may be a factor in determining the applicable law to an arbitration agreement, its ultimate relevance depends on the parties’ intentions.

Implied Choice of Law in Art V(1)(a) of the New York Convention

The New York Convention (NYC) provides a further point of reference for assessing the implied choice of law, despite primarily dealing with the recognition and enforcement of arbitral awards. Art V(1)(a) NYC permits the refusal of recognition and enforcement of an award if the arbitration agreement is invalid “under the law to which the parties have subjected it or, failing any indication thereon, under the law of the country where the award was made”. Thus, the NYC provides a conflict-of-law rule to determine the applicable law to an arbitration agreement, similar to the three-stage enquiry. Accordingly, the law applicable is determined primarily by the parties’ intentions and secondarily by the “law of the country where the award was made” as the default rule. While the NYC also allows for an implied choice of law, extending the above-mentioned jurisprudence to the NYC appears questionable.

Although its wording does not clarify how the implied choice of law should be understood, its underlying structure – especially with regard to the default rule – seems to prevent such a generalisation of the parties’ intentions. This is for two reasons: first, an excessive interpretation of the implied choice of law by considering the hypothetical intentions of parties undermines the scope of application of the default rule (see Chan/Teo). Second, the default rule already reflects the hypothetical intentions of commercial parties (see Born, Chapter 4; Zarra). While it remains open how the law of the seat takes account of said hypothetical interests, given the importance of party agreements in arbitration law in general and as the primary connecting factor in Art V(1)(a) NYC, it can be assumed that the default rule was not determined arbitrarily, but in accordance with these (here necessarily hypothetical) party interests. Therefore, any considerations regarding hypothetical party interests can be omitted when assessing the implied choice of law. While the implied choice of law is to be understood narrowly, the NYC itself does not prescribe an exact threshold.

Implied Choice of Law Modified?

In an attempt to derive a threshold for the implied choice of law, one can look to private international law, which regularly prescribes a more precise threshold. Subsequently, the question arises as to whether this understanding can also be applied to the implied choice of law applicable to the arbitration agreement. Art 3(1) of the Rome I Regulation contains an explicit threshold for the implied choice of law: any choice of law shall be “expressly or clearly demonstrated by the terms of the contract or the circumstances of the case”. In a deliberate departure from the predecessor provision of Art 3 Rome Convention, reasonable certainty is no longer sufficient. The reference to a clear demonstration requires a real rather than a hypothetical choice of law (see Ragno, Commentary on Art 3).

While the Rome I Regulation does not apply to arbitration agreements pursuant to its Art 1(2)(e), the considerations on the threshold for the implied choice of law may be transferred to the NYC. The European Commission’s proposal for the Rome I Regulation considered the existing international regulations for arbitration agreements (including the NYC) sufficient and therefore excluded arbitration agreements from the scope of the Regulation. However, these international regulations do not contain a threshold for the implied choice of law. In any case, there is no strict distinction between the Rome I Regulation and arbitration agreements: the choice of an arbitral seat within an arbitration agreement can be used to determine a choice of law under Art 3(1) Rome I Regulation (see the UKSC in Enka v Chubb).

An equally high threshold is reflected in other instruments of private international law: Art 4 Hague Choice of Law Principles requires the implied choice to “appear clearly from the provisions of the contract or the circumstances”. The associated commentary requires a real choice, i.e. a determination made on a case-by-case basis, and does not allow a presumed intention to suffice. Art 7 Inter-American Convention on the Law Applicable to International Contracts also follows this high threshold by requiring an “evident” choice.

Although the case of Kabab-Ji SAL v Kout Food Group suggests otherwise, it does not prevent the threshold from being transferred to the NYC. The UKSC refers to the wording of Art V(1)(a) NYC, which applies the default rule “failing any indication” as to a party agreement. The use of “indication” clarifies that the party agreement does not have to be explicit. This contention is to be followed. However, it cannot be deduced, as the UKSC assumes, that “any form of agreement will suffice”. While the wording of “indication” allows for an implied choice of law, it does not establish the threshold for such a choice, leaving room for the threshold of private international law.

Conclusion

In light of the above considerations, it seems appropriate to depart from the generalisation of party intentions in English case law. Instead, the threshold for the implied choice of law should be interpreted in light of private international law. Consequently, the three-stage test can be modified into a two-stage test: primarily, the law applicable to the arbitration agreement should be determined by a clear and real choice of law; secondarily, it should be determined by the law of the seat of arbitration. While this approach resembles the two-stage test in the English Arbitration Act 2025, which requires an explicit choice of law, it is less stringent and provides legal certainty, as the default rule is suitable to catch all cases in which the choice of law is not evident.