Choice of Law for Arbitration Agreements

The highest courts of Singapore and the United Kingdom have recently issued decisions regarding the choice of law to govern arbitration agreements.

Singapore Courts’ Approach

In BNA v BNB and another [2020] 1 SLR 456; [2019] SGCA 84 (“BNA”), the Singapore Court of Appeal adopted at [44]-[48] the following three-step approach from BCY v BCZ [2017] 3 SLR 357 (HC) (“BCY”):

  • The parties’ express choice of proper law governing the arbitration agreement is first identified (the “Express Choice of Law”).
  • If there is no express choice, the court ascertains the parties’ implied choice of law based on their intentions at time of contracting (the “Implied Choice of Law”).
  • If an express or implied choice is absent, the law governing the arbitration agreement is the system of law with the closest and most real connection to it.

This is based on the framework in Sulamerica Cia Nacional de Seguros S.A. v. Enesa Engenharia S.A. [2012] EWCA Civ 638 (“Sulamerica”).

An express choice of the proper law of the main contract does not, in and of itself, also constitute the proper law of the arbitration agreement: BNA at [61].

There is a presumption that the parties’ choice of law for the entire/main/underlying contract (containing the arbitration clause) also reflects the parties’ implied choice of law governing the arbitration agreement (BNA at [61]).

However, this presumption can be rebutted. The choice of a seat that is different from the place of the law of the main contract is not by itself sufficient to displace the starting point: BNA at [62]; BCY at [65].

The inquiry of whether the presumption is displaced only becomes relevant if the law of the seat is materially different from the (I suppose, putative) law governing the arbitration agreement (lex arbitri): BNA at [63]. E.g. if the seat of arbitration is Singapore but the assumed lex arbitri by that point in the analysis is the law of the People’s Republic of China (PRC).

If parties specify only one geographical location in an arbitration agreement, and particularly where the parties express a choice for “arbitration in [that location]”, that should most naturally be construed as a reference to the parties’ choice of seat and not merely venue: BNA at [65].

Its significance lies in the fact that for legal reasons the arbitration will be regarded as situated in the state or territory of the seat, and the choice of seat also identifies the state or territory whose laws will govern the arbitral process, ie, the curial law of the arbitration. The seat will also be considered to be the jurisdiction in which the arbitral award is “made” for the purposes of the New York Convention: at [65].

The Court of Appeal held that the fact that PRC law at the material time may invalidate the arbitration agreement was not a relevant consideration in determining the proper law of the arbitration agreement. This is because there was no evidence that the parties were at least aware of the impact that the choice of PRC law could have on the validity of the arbitration agreement: BNA at [90].

English Courts’ Approach

In Enka Insaat Ve Sanayi AS v OOO Insurance Company Chubb [2020] UKSC 38 (“Enka“), the UK Supreme Court essentially adopted the same three-step framework (case summary here).

However, the majority and the dissenting minority disagreed on what happens when contracting parties did not make an express or implied choice of law for the main contract.

The majority (Lord Hamblen, Lord Leggatt, and Lord Kerr) held that generally, the arbitration agreement will be most closely connected with the law of the seat of arbitration: at [118]-[119].

This default rule is supported by the following considerations: (i) the seat is where the arbitration is to be performed (legally, if not physically): at [121]-[124]; (ii) this approach maintains consistency with international law and legislative policy: at [125]-[141]; (iii) this rule is likely to uphold the reasonable expectations of contracting parties who specify a location for the arbitration without choosing the law to govern the contract: at [142]-[143]; and (iv) this approach provides legal certainty, allowing parties to predict easily which law the court will apply in the absence of choice: at [144].

The minority (Lord Burrows and Lord Sales) held that it should be that the law with which the main contract is most closely connected governs the arbitration agreement, as this is the law with which in their view the arbitration agreement is also most closely connected: at [257].


The majority and minority in Enka considered the separability principle in their analyses, noting that the principle treats the arbitration agreement as a stand-alone agreement which may be independently valid and effective even if the main contract is not, but that this does not extend to the question of the choice of law or insulate the arbitration agreement from everything else in or which governs the main contract: at [41], [61], [63], [232], [233], [275].

The Singapore Court of Appeal in BNA however declined to consider and discuss the separability principle and its scope on the basis that it could arrive at its decision without going there: at [95]. In the court below, the High Court Judge (Vinodh Coomaraswamy J) had applied an extended scope of the separability principle to save the arbitration agreement, which may otherwise be invalidated by the applicable law of the main contract: at [27].

The Court in BNA also declined to consider and discuss the validation principle: at [95]. The Judge below considered and rejected the applicability of the validation principle in Singapore law.

The validation principle provides that “if an international arbitration agreement is substantively valid under any of the laws that may potentially be applicable to it, then its validity will be upheld, even if it is not valid under any of the other potentially applicable choices of law” (Gary Born, The Law Governing International Arbitration Agreements: An International Perspective, (2014) 26 SAcLJ 814 at [51]).

The minority in Enka considered that the validation principle is “an aspect of the general objective approach to determining the intention of the parties to a contract ut res magis valeat quam pereat (so that the main object of the agreement is upheld and not destroyed)”: at [277].

In contrast, Coomaraswamy J in BNA thought that the validation principle is inapplicable because the ut res magis general principle regarding construction of contracts is already applicable.

These principles can be understood as all simply legal fiction. If one were to ask the non-lawyer contracting parties what they really thought when they executed the contract (as I have come to realise after dealing with various clients), they would probably say they had no idea. They simply thought that well, there is an arbitration clause and there is a governing law clause, and that’s all there is to it. They would probably say that the governing law for the entire contract obviously governs the arbitration clause since it is part of the entire contract; nowhere did they say that the arbitration clause should be treated specially.

And if asked why a geographical location like Shanghai or Singapore is specified in the arbitration clause, most (assuming not advised by lawyers) would probably say, well, we followed some templates and thought it was to specify the venue; we did not know there is this thing called the seat of arbitration. If, instead, the parties were advised by competent lawyers, then it is unlikely that there would have arisen such a problem in the first place.

Such is the strange parallel legal universe when disputes present themselves before the courts. On one hand, we purport to promote party autonomy in contracting and on the other hand, we create legal fiction to twist things about to make things work according to what we think should have been the case.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.