Court of Appeal · Decided 26 October 2015

Tomolugen Holdings Ltd and another v Silica Investors Ltd and other appeals

Feedback & support

[2015] SGCA 57Case no. Civil Appeals Nos 123, 124 and 126 of 2014

What happened, in plain English

There is no plain-English summary of this judgment yet.

Ask about this judgment

Laws this judgment relies on

Grounds of decision

Official text from eLitigation

Sundaresh Menon CJ (delivering the judgment of the court):
Introduction
[1] Before us are three appeals against the decision of a High Court judge (“the Judge”) who dismissed applications to stay the court proceedings in Suit No 560 of 2013 (“the Suit”) in favour of arbitration. In the Suit, the plaintiff seeks relief under s 216 of the Companies Act (Cap 50, 2006 Rev Ed) (“the Companies Act”) for oppressive or unfairly prejudicial conduct towards it as a minority shareholder. There are altogether eight defendants. They include the company of which the plaintiff is a minority shareholder, as well as other shareholders and current or former directors of either that company or its related companies. The stay applications were made by the defendants pursuant to s 6 of the International Arbitration Act (Cap 143A, 2002 Rev Ed) (“the IAA”) and the court’s inherent power of case management.
[2] These appeals raise issues concerning the arbitrability of disputes over minority oppression or unfairly prejudicial conduct, as well as the proper approach to adopt when dealing with potentially overlapping court and arbitral proceedings. They draw out a tension that will mount as commercial transactions and the disputes which they spawn grow in their sophistication and complexity. This tension is one that is likely to arise whenever disputes straddle court and arbitral proceedings, with the two sets of proceedings engaging common, although not necessarily identical, issues and parties. The tension lies in the court’s desire, on the one hand, to avoid the complications inherent in having to resolve a dispute across two different fora and, on the other hand, its recognition that it must conform to the statutory mandate laid down in s 6 of the IAA to direct that any dispute concerning a matter which is the subject of an arbitration agreement governed by the IAA is to be resolved by arbitration. Where these two considerations pull in opposite directions, the latter necessarily prevails because of the mandatory terms of s 6. The question which then arises for the court is how it can best manage the proceedings that it has control over in order to ameliorate the complications that are bound to arise by reason of the overlapping proceedings.
Factual background
Overview of the material facts
[3] The plaintiff in the Suit is Silica Investors Limited (“Silica Investors”), a minority shareholder which holds approximately 4.2% of the issued share capital of Auzminerals Resource Group Limited (“AMRG”), the eighth defendant in the Suit. AMRG is the company whose affairs Silica Investors alleges have been conducted in a manner that is oppressive or unfairly prejudicial towards it.
[4] AMRG is a listed Singapore company that develops, explores and exploits mines in Australia. Its wholly-owned subsidiary, Solar Silicon Resources Group Pte Ltd (“SSRG”), is also involved in mine development and exploration, and owns Australian mining licences and exploration permits.
[5] Silica Investors first acquired shares in AMRG from Lionsgate Holdings Pte Ltd (“Lionsgate”), the second defendant in the Suit. The share sale agreement between Silica Investors and Lionsgate (“the Share Sale Agreement”) contains an arbitration clause. Lionsgate currently remains a shareholder of AMRG and holds approximately 9% of its issued share capital.
[6] Lionsgate is wholly owned by Tomolugen Holdings Limited (“Tomolugen Holdings”), the first defendant in the Suit. Tomolugen Holdings is the majority shareholder of AMRG and holds approximately 55% of its issued share capital directly. The third to seventh defendants in the Suit are shareholders and current or former directors of Lionsgate, AMRG or SSRG; together, they hold just over 3% of the issued share capital of AMRG.
[7] Lionsgate made an application pursuant to s 6 of the IAA for a stay of the court proceedings against it in the Suit. Lionsgate said that a part of the dispute in the Suit fell within the scope of the arbitration clause in the Share Sale Agreement and the court proceedings against it in respect of that part of the dispute therefore had to be stayed in favour of arbitration; it further submitted that the remainder of the court proceedings against it should be stayed as well in the interests of appropriate case management pending the resolution of the arbitration. The other seven defendants named in the Suit (collectively, “the remaining defendants”) also filed stay applications that were contingent on the success of Lionsgate’s application. They argued that if the court proceedings between Silica Investors and Lionsgate were stayed in favour of arbitration, then the rest of the court proceedings in the Suit should also be stayed pending the resolution of that arbitration. Their stay applications were made pursuant to the court’s inherent power of case management.
[8] The Judge dismissed all the stay applications. His decision is reported as Silica Investors Ltd v Tomolugen Holdings Ltd and others [2014] 3 SLR 815 (“the HC Judgment”). Civil Appeal No 126 of 2014 is Lionsgate’s appeal against the Judge’s decision, while Civil Appeals Nos 123 and 124 of 2014 are appeals by three of the remaining defendants. Silica Investors is the respondent in all three appeals.
[9] In the two sections that follow, we shall set out in greater detail the Share Sale Agreement between Silica Investors and Lionsgate, as well as the allegations made in the Suit. This will provide the factual context for the analysis of the issues raised in these appeals.
The Share Sale Agreement
[10] The Share Sale Agreement between Silica Investors and Lionsgate consists of an original agreement dated 23 June 2010, which was modified by a supplemental agreement dated 5 July 2010. At the time the Share Sale Agreement was concluded, Lionsgate was known as Tomolugen Pte Ltd. This change of name is immaterial for present purposes, and we use “Lionsgate” to refer to the entity both before and after its change of name. Silica Investors and Lionsgate were the only parties to the Share Sale Agreement. Under that agreement, Lionsgate agreed to sell Silica Investors up to 2.5m AMRG shares at a price of AU$4.00 per share. The amended completion date was 14 July 2010.
[11] The Share Sale Agreement contained terms and warranties relating to the management of AMRG and its assets and liabilities. These included the following:
(a) Lionsgate agreed to support the passage of an AMRG board of directors’ resolution for a nominee of Silica Investors to be appointed as a director on AMRG’s board: cl 2.5(a) of the Share Sale Agreement.
(b) Lionsgate warranted that the accounts of AMRG and its related companies gave a true and fair view of the state of affairs of those companies: cl 8.2 of the first schedule to the Share Sale Agreement.
(c) Lionsgate warranted that at the completion date, AMRG and its related companies would have settled or discharged “all current liabilities, inter-company loans (in excess of S$250,000 per company) and shareholders’ advances and all obligations of AMRG [and its related companies]”: cl 8.3 of the first schedule to the Share Sale Agreement.
[12] The Share Sale Agreement contemplated the listing of SSRG on a recognised stock exchange by a specified date. It also contained provisions for the distribution of SSRG shares in specie amongst the shareholders of AMRG should the listing of SSRG succeed, as well as exit provisions for Silica Investors in the event that the listing of SSRG failed.
[13] In addition, the Share Sale Agreement contained an arbitration clause (viz, cl 12.3) which provided that “any dispute arising out of or in connection with this Agreement, including any question regarding its existence, validity or termination” was to be referred to and finally resolved by arbitration in Singapore in accordance with the Arbitration Rules of the Singapore International Arbitration Centre (“the SIAC”) for the time being in force. We shall discuss the arbitration clause and its effect in greater detail later in this judgment.
The allegations made in the Suit
[14] Silica Investors commenced the Suit in June 2013, and its only complaint there is that AMRG’s affairs were conducted in a manner which was oppressive or unfairly prejudicial towards it as a minority shareholder of AMRG. On this basis, it seeks wide-ranging relief, including a share buy-out order and orders regulating the conduct of AMRG’s affairs; in the alternative, it seeks an order that AMRG be placed in liquidation.
[15] Silica Investors’ complaint of oppression or unfair prejudice is supported by allegations that the Judge classified into four broad categories. We shall use these same categories in this judgment since they were adopted by the parties on appeal. The first category, we shall refer to as “the Share Issuance Allegation”. This concerned an issuance of 53,171,040 AMRG shares on 15 September 2010. The share issuance allegedly had the effect of diluting Silica Investors’ shareholding in AMRG by more than 50%; it was also purportedly done in breach of AMRG’s memorandum and articles of association. AMRG claims that it issued the shares to discharge debts which it owed for certain mining licences (“the Solar Silica Assets”) that had been transferred to SSRG. Silica Investors, on the other hand, says that those debts were fictitious, and that there was no commercial justification for the share issuance.
[16] The second category of allegations, we shall refer to as “the Management Participation Allegation”. Silica Investors claims that there was an understanding or a legitimate expectation on its part that it would participate in the management of AMRG. In support of this allegation, Silica Investors points to cl 2.5(a) of the Share Sale Agreement, which we referred to at [11(a)] above. Silica Investors claims that it has been denied its entitlement to participate in the management of AMRG.
[17] The third category of allegations, we shall refer to as “the Guarantees Allegation”. Silica Investors claims that on 11 July 2012, AMRG executed guarantees securing the obligations of Australian Gold Corporation Pte Ltd (“Australian Gold”). Australian Gold is apparently unrelated to AMRG, but 97% of its shares are owned by AMRG’s majority shareholder, Tomolugen Holdings. The third, fourth and fifth defendants were or became directors of Australian Gold either at the time or soon after the guarantees were executed. Silica Investors alleges that it was not in AMRG’s interest to take on the liabilities under those guarantees.
[18] The fourth and final category of allegations, we shall refer to as “the Asset Exploitation Allegation”. It is similar to the Guarantees Allegation in so far as it concerns the incurring of financial liabilities by AMRG for the benefit of its majority shareholder, Tomolugen Holdings. Specifically, Silica Investors alleges that AMRG expended resources to identify and develop mining assets in Australia. These assets are now apparently owned by an entity which, although not related to AMRG, is 90% owned by Tomolugen Holdings. Accompanying this assertion are claims of non-disclosure by AMRG’s board of directors of material information relating to these assets.
[19] The four categories of allegations appear unconnected. But, what binds them together is that the matters which they concern were all allegedly orchestrated by Mr Roger Thomas May, the seventh defendant in the Suit, who was a shadow director of AMRG at the material time. It is said that Mr May acted in connivance with those directors of AMRG who had been appointed by Tomolugen Holdings and Lionsgate to conduct AMRG’s affairs in a manner which was oppressive or unfairly prejudicial towards Silica Investors as a minority shareholder.
The decision below
[20] The Judge held (at [18] of the HC Judgment) that for the purposes of s 6 of the IAA, he had to identify the “essential dispute” in the Suit and determine whether that fell within the scope of the arbitration clause in the Share Sale Agreement. He found (at [37]) that the essential dispute was whether the affairs of AMRG had been conducted in a manner that was “oppressive, ie, commercially unfair or unfairly prejudicial … towards [Silica Investors] as a minority shareholder”. He held (at [56]–[58]) that this dispute did fall within the ambit of the widely-drafted arbitration clause because there was a sufficient factual connection between at least two of the four allegations (namely, the Share Issuance Allegation and the Management Participation Allegation) and the Share Sale Agreement.
[21] However, the Judge did not grant Lionsgate’s stay application because he held that the dispute was non-arbitrable. He considered that a claim for relief under s 216 of the Companies Act straddled the line between arbitrability and non-arbitrability, and that in most circumstances, such a dispute would fall on the side of being non-arbitrable (see [141]–[142] of the HC Judgment). He was satisfied that on the facts of the case, the dispute was non-arbitrable (at [143]). In particular, he noted (likewise at [143]) that the relief sought in the Suit was of a kind which an arbitral tribunal could not grant; further, the dispute involved parties who were not bound by the arbitration clause in the Share Sale Agreement. The Judge’s dismissal of Lionsgate’s stay application meant that the remaining defendants’ stay applications fell away, and the Judge did not consider them.
[22] We shall elaborate further on the Judge’s reasons for his decision as well as the parties’ arguments in the course of our analysis of the issues that arise for our consideration.
The issues before us
[23] Three issues arise for our determination in these appeals:
(a) first, whether a dispute over minority oppression or unfairly prejudicial conduct is arbitrable;
(b) second, whether the court proceedings between Silica Investors and Lionsgate, or any part thereof, fall within the scope of the arbitration clause in the Share Sale Agreement; and
(c) third, in the event that the court proceedings between Silica Investors and Lionsgate (or any part thereof) are covered by the arbitration clause and are stayed in favour of arbitration, whether the remainder of the court proceedings (whether against Lionsgate or against the remaining defendants) should also be stayed pending the resolution of the arbitration.
[24] We shall address these issues in the aforesaid order. We are aware that the second issue is both logically and legally anterior to the first, in that the question of arbitrability is a live one only if the dispute in the Suit, or a part of it, falls within the scope of the arbitration clause in the Share Sale Agreement. But, since the Judge’s decision turned on his finding of non-arbitrability, we think it convenient to deal with the issue of arbitrability first. Before we do so, however, we shall first consider a threshold question as to what standard of review the court should adopt in an application for a stay under s 6 of the IAA. This question (“the threshold question”) was raised by Prof Lawrence Boo, who appeared as amicus curiae at our request.
The threshold question: the standard of review in a stay application under the IAA
The backdrop to the threshold question
[25] It is often said that an arbitral tribunal’s jurisdiction is based on the consent of the parties as manifested in the arbitration agreement concerned. But, it is also established that an arbitral tribunal has the jurisdiction to determine the existence and extent of such consent. The arbitral tribunal’s jurisdiction to determine its own jurisdiction – known as kompetenz-kompetenz – cannot be based entirely on the parties’ consent. This form of jurisdiction necessarily precedes and exists independently of such consent. The arbitral tribunal may, in the exercise of its kompetenz-kompetenz, conclude that there was never any consent by the parties to refer their disputes to arbitration, and as a consequence, that it had no jurisdiction to begin with.
[26] The potential circularity that this entails is broken by Art 16 of the UNCITRAL Model Law on International Commercial Arbitration (“the Model Law”), which has the force of law in Singapore: s 3 of the IAA. Article 16 embodies the kompetenz-kompetenz principle, and it confers on an arbitral tribunal the jurisdiction to “rule on its own jurisdiction, including any objections with respect to the existence or validity of the arbitration agreement” [emphasis added].
[27] The kompetenz-kompetenz principle has the potential to give rise to friction with the regime set out in the IAA (and the Model Law), which mandates a stay of court proceedings in favour of arbitration when the conditions for the grant of a stay are satisfied. Specifically, under s 6 of the IAA, the court must stay court proceedings relating to “any matter” that is covered by an arbitration agreement upon an application for a stay by a party to that agreement. The only exceptions are where the court is satisfied that the arbitration agreement is “null and void”, “inoperative” or “incapable of being performed”: s 6(2) of the IAA. This regime is substantially similar to that found in Art 8(1) of the Model Law, but with some differences in phraseology. We set out the material portions of both provisions below. Sections 6(1) and 6(2) of the IAA state:
Article 8(1) of the Model Law states:
[28] A court hearing a stay application under s 6 of the IAA will necessarily have to take a view on the existence and scope of the arbitration agreement in question. Only then can it decide whether a stay must be granted under s 6. The friction which we have alluded to in the previous paragraph may arise because any determination made by the court on the existence and scope of the arbitration agreement may well intrude into the remit of the arbitral tribunal’s kompetenz-kompetenz.
[29] The threshold question is one that invites consideration of how a balance may be struck between the court and the arbitral tribunal, both of which may confront the issue of whether the arbitral tribunal has jurisdiction. In our judgment, that balance may be struck by adjusting the standard of review which the court adopts when hearing an application for a stay under s 6 of the IAA. In this regard, the first of two prevailing views is that the court should only undertake a prima facie review of the existence and scope of the arbitration clause. Thus, if the court is satisfied on a prima facie standard that the conditions for the grant of a stay (ie, that there exists an arbitration clause which is valid and which covers the dispute at hand) have been met, it should grant the stay and defer to the arbitral tribunal the determination of whether those conditions have indeed been satisfied. We refer to this as the prima facie approach. This approach preserves the arbitral tribunal’s kompetenz-kompetenz to examine the existence and scope of its jurisdiction afresh and determine it fully.
[30] The second prevailing view is that the court should undertake an actual determination of the existence and scope of the arbitration agreement when it hears a stay application under s 6 of the IAA. On this view, the court grants a stay if, and only if, it is satisfied that the requirements for the grant of a stay have in fact been met. We refer to this as the full merits approach. This approach might have the advantage of expedience in that it allows the court to pronounce with finality on an arbitral tribunal’s jurisdiction in the first instance, instead of deferring the question to the arbitral tribunal, only to face the prospect of the same question coming back to the court in the event of an appeal against the arbitral tribunal’s jurisdictional ruling, or if the unsuccessful party in the arbitration resists enforcement of the arbitral award or applies to set it aside on the basis of the arbitral tribunal’s lack of jurisdiction. There appears to be a clear split between these two views in both the case authorities across various jurisdictions as well as the literature (see United Nations Commission on International Trade Law, UNCITRAL 2012 Digest of Case Law on the Model Law on International Commercial Arbitration (2012) at paras 29–32).
[31] The standard of review that a court in Singapore should apply ultimately turns on the interpretation of s 6 of the IAA. Section 6 has a strong association with and was based on Art 8 of the Model Law. We shall accordingly begin our analysis by reviewing the travaux préparatoires of the Model Law and the position in other jurisdictions before returning to discuss the appropriate approach for the Singapore courts.
The travaux préparatoires of the Model Law
[32] In our judgment, the travaux préparatoires of the Model Law are inconclusive as to the appropriate approach to take with regard to the threshold question, as both the prima facie approach and the full merits approach can and have been credibly advanced based on different parts of these materials. Proponents of the full merits approach rely on the Report of the Working Group on International Contract Practices on the Work of its Fifth Session, UN Doc A/CN/9/233 (“Report of the Fifth Session”). At para 77 of the Report of the Fifth Session, the UNCITRAL Working Group on International Contract Practices (“the Working Group”) apparently expressly considered and rejected the prima facie approach:
[33] In his book International Commercial Arbitration, Volume I: International Arbitration Agreements (Wolters Kluwer, 2nd Ed, 2014) (“Gary Born”), Gary B Born suggests (at p 1083) that this is “[t]he most direct evidence” that the Model Law’s drafters rejected the prima facie approach. This view is shared by Howard M Holtzmann & Joseph E Neuhaus, A Guide to the UNCITRAL Model Law on International Commercial Arbitration: Legislative History and Commentary (Kluwer Law, 1989) (“Holtzmann & Neuhaus”) at p 303 and Peter Binder, International Commercial Arbitration and Conciliation in UNCITRAL Model Law Jurisdictions (Sweet & Maxwell, 3rd Ed, 2010) at p 125. The use of the word “finds” in Art 8(1) of the Model Law is also said to support a full merits review by the court: Gary Born at pp 1082–1083.
[34] On the other hand, proponents of the prima facie approach rely on developments in the drafting of the Model Law that occurred at the sixth and subsequent sessions of the Working Group. At the sixth session, the Working Group proposed the introduction of a fourth paragraph to the Article in the draft text of the Model Law (“the draft Model Law”) that then articulated the kompetenz-kompetenz principle (now Art 16 of the Model Law). The proposed fourth paragraph read as follows (Report of the Working Group on International Contract Practices on the Work of its Sixth Session, UN Doc A/CN/9/245 (“Report of the Sixth Session”) at para 66):
This proposed paragraph was eventually rejected. The Working Group thought that it was not sufficiently clear, especially in its interface with the then draft Article governing applications for stays of court proceedings (now Art 8 of the Model Law): Report of the Sixth Session at paras 68–69. The hesitance to incorporate the proposed fourth paragraph in the form set out above resulted in the Working Group requesting the UNCITRAL Secretariat (“the Secretariat”) to revise the text of that draft Article.
[35] In lieu of such a revision, the Secretariat proposed instead the introduction of an entirely new Article (“the draft Art 17”) in the composite text of the draft Model Law that was presented to the Working Group at its seventh session: Composite Draft Text of a Model Law on International Commercial Arbitration, UN Doc A/CN.9/WG.II/WP.48. The draft Art 17 read as follows:
[36] The draft Art 17 was eventually rejected by the Working Group at its seventh session. The prevailing view of the Working Group was that this proposed Article might enable parties to engage in tactical delays and did not cohere with the kompetenz-kompetenz principle articulated in Art 16 (see Report of the Working Group on International Contract Practices on the Work of its Seventh Session, UN Doc A/CN/9/246 at para 55):
[37] Proponents of the prima facie approach say that the decision not to incorporate the draft Art 17 is significant in that it expressed the intention that the arbitral tribunal, rather than the court, should have the first say on the existence and scope of the arbitration agreement in question, and thus, on whether the arbitral tribunal has jurisdiction. They also contend that this view is reinforced by the rejection of subsequent suggestions to reincorporate the draft Art 17 into the draft Model Law.
[38] The subsequent suggestions for the reintroduction of the draft Art 17 came from the International Bar Association (“the IBA”) and the UK government. In the Secretariat’s Analytical Compilation of Comments by Governments and International Organizations on the Draft Text of a Model Law on International Commercial Arbitration, UN Doc A/CN.9/263 (“Secretariat’s Analytical Compilation of Comments”) at p 30, the IBA recorded its dissatisfaction with the rejection of the draft Art 17:
[39] Similar sentiments were also expressed by the UK government: Second Addendum to the Secretariat’s Analytical Compilation of Comments, UN Doc A/CN.9/263/Add.2 at p 5. These proposals to reintroduce the draft Art 17 were considered but eventually rejected: Report of the United Nations Commission on International Trade Law on the Work of its Eighteenth Session, UN Doc A/40/17 at paras 160–161. Instead, a compromise was eventually reached by the introduction of what is now Art 16(3) of the Model Law, which allows a discrete ruling by an arbitral tribunal on its jurisdiction to be referred to a court with supervisory jurisdiction for a decision.
[40] The proponents of the prima facie approach contend that these developments in the drafting of the Model Law evince the intention that jurisdictional issues be deferred to the arbitral tribunal, with the court having the final say only through an appeal from the arbitral tribunal’s jurisdictional ruling: Frédéric Bachand, “Does Article 8 of the Model Law Call for Full or Prima Facie Review of the Arbitral Tribunal’s Jurisdiction?” (2006) 22(3) Arb Int’l 463 at p 473. The upshot is what has been termed by Fouchard Gaillard Goldman on International Commercial Arbitration (John Savage & Emmanuel Gaillard eds) (Kluwer Arbitration, 1999) at pp 400–401 as the “negative effect” of the kompetenz-kompetenz principle, which is that arbitral tribunals should be the “first judges of their jurisdiction” and should be able to decide the existence (or otherwise) of their own jurisdiction prior to any court or judicial authority. Gary Born asserts that this reasoning is “inconclusive” and “irrelevant”, and that “the deletion of the [draft Art 17] … [is] in no way decisive or even relevant as to the meaning of Article 8(1)”: Gary Born at p 1084. We are less inclined to dismiss the rejection of the draft Art 17 as irrelevant. The very reason for the proposed introduction of this draft Article was the lack of clarity in the interface between the draft versions of what are now Art 8 and Art 16 (see [34]–[35] above). That interface was a problematic one, and the draft Art 17 was one of the solutions that the Working Group considered in an attempt to address it.
[41] In our judgment, the travaux préparatoires of the Model Law are ultimately somewhat equivocal as to the appropriate approach to take for the purposes of the threshold question, although it seems to us from the historical analysis we have undertaken that, perhaps, the record offers a little more support for the prima facie approach. That said, as we noted at [32] above, there is certainly material to support both the prima facie and the full merits approaches. We venture to suggest a historical reason for this seeming lack of certainty in the travaux préparatoires. When the draft version of what is now Art 8 of the Model Law was formulated at the fifth session of the Working Group, it was consciously framed in terms that were “in line with the corresponding provision in the 1958 New York Convention” and “for the sake of consistency with that important Convention”: Report of the Fifth Session at para 76. In other words, the Working Group endeavoured to stay faithful to Art II(3) of the Convention on the Recognition and Enforcement of Foreign Arbitral Awards Concluded at New York on 10 June 1958 (“the New York Convention”), which states:
[42] When Art II(3) of the New York Convention was formulated in the 1950s, it sought principally to achieve the limited goal of preventing Contracting States from refusing to recognise the validity of arbitration agreements (see New York Convention: Convention on the Recognition and Enforcement of Foreign Arbitral Awards of 10 June 1958 Commentary (Reinmar Wolff ed) (Hart Publishing, 2012) at pp 153–155). The kompetenz-kompetenz principle is not expressed in the New York Convention and does not appear to have been considered in its travaux préparatoires. While the kompetenz-kompetenz principle has an impressive vintage in international law tribunals (see Lord Collins of Mapesbury JSC’s account of it in Dallah Real Estate and Tourism Holding Co v Ministry of Religious Affairs of the Government of Pakistan [2011] 1 AC 763 at [79]–[81]), its acceptance in international commercial arbitration circles appears less rooted. Indeed, even at the time of the drafting of the Model Law, some three decades after the New York Convention was concluded, the kompetenz-kompetenz principle was still “thought to be somewhat controversial” (see Holtzmann & Neuhaus at p 479) and had yet to be accepted into all national laws (see Analytical Commentary on Draft Text of a Model Law on International Commercial Arbitration, UN Doc A/CN.9/264 at p 37).
[43] The drafters of the New York Convention were thus not exercised by the need to deal with the competing claims of the court and the arbitral tribunal to rule on the latter’s jurisdiction; they had the rather more limited aim of obliging national courts’ recognition of arbitration clauses. Commentators appear largely united in the view that Art II(3) of the New York Convention is silent or, at best, equivocal as to the standard of review which a court should adopt when it is faced with a stay application: see Gary Born at pp 1054–1055; Julian D M Lew, Loukas A Mistelis & Stephan Kröll, Comparative International Commercial Arbitration (Kluwer Law International, 2003) at para 14-54; and Recognition and Enforcement of Foreign Arbitral Awards: A Global Commentary on the New York Convention (Herbert Kronke et al eds) (Kluwer Law International, 2010) at pp 108–109.
[44] To summarise our review of the travaux préparatoires of the Model Law, the kompetenz-kompetenz principle as encapsulated in what is now Art 16 of the Model Law was deliberated at the sixth session of the Working Group: Report of the Sixth Session at para 58. The introduction of the kompetenz-kompetenz principle added a further dimension to the draft Model Law, which was that of having to allocate the jurisdiction to determine an arbitral tribunal’s jurisdiction as between the court and the arbitral tribunal. That was not considered when Art II(3) of the New York Convention was drafted. The adherence to Art II(3) of the New York Convention in what is now Art 8 of the Model Law despite the emergence and incorporation of the kompetenz-kompetenz principle as embodied in Art 16 of the Model Law might account for the seeming ambivalence that is found in the travaux préparatoires where the threshold question is concerned. In a sense, Art 16 was forward-looking, while Art 8 had its eye fixed on the past.
The position in other jurisdictions
[45] We now consider the position in other jurisdictions. In this regard, we shall confine our discussion to cases from England, the Hong Kong Special Administrative Region (“Hong Kong SAR”) and Canada. The latter two are Model Law jurisdictions, while England is not.
England
[46] Although England has not enacted the Model Law, its case law is still persuasive for two reasons. First, the UK’s Arbitration Act 1996 (c 23) (“the UK Arbitration Act 1996”) was heavily influenced by the Model Law. The UK Departmental Advisory Committee on Arbitration Law (“the UK Advisory Committee on Arbitration Law”), in its report on the draft Bill published in July 1995 which was subsequently enacted as the UK Arbitration Act 1996 (see the 1996 Report on the Arbitration Bill (Chairman: The Rt Hon Lord Justice Saville) (“1996 Report on the Arbitration Bill”) (reprinted at (1997) 13(3) Arb Int’l 275)), stated at para 4 that “at every stage in preparing a new draft Bill, very close regard was paid to the Model Law, and it will be seen that both the structure and the content of the July draft Bill … owe much to this model”. Second, the stay provisions in s 9 of the UK Arbitration Act 1996 are substantially similar to those in s 6 of the IAA. Section 6 of the IAA in fact mirrors very closely the wording used in s 1 of the UK’s Arbitration Act 1975 (c 3) (“the UK Arbitration Act 1975”), which was the predecessor provision of s 9 of the UK Arbitration Act 1996. Section 6 of the IAA requires the court to stay proceedings unless it is “satisfied” that the arbitration clause is null and void, inoperative or incapable of being performed, as does its English counterpart. This is in contrast to Art 8 of the Model Law, which employs the word “finds” instead of “satisfied”. We reproduce ss 9(1) and 9(4) of the UK Arbitration Act 1996 below for convenience:
[47] The UK Arbitration Act 1996 also expressly recognises the kompetenz-kompetenz principle. Section 30 thereof states that the arbitral tribunal may “rule on its own substantive jurisdiction”, including questions as to:
[48] The English courts generally adopt the full merits approach when hearing stay applications under s 9 of the UK Arbitration Act 1996, subject to a residual discretion to stay the court proceedings under the court’s inherent jurisdiction so as to allow the arbitral tribunal to make a determination on its own jurisdiction instead. The English position was summarised by Aikens LJ in Joint Stock Company “Aeroflot Russian Airlines” v Berezovsky [2013] 2 Lloyd’s Rep 242 (“Aeroflot v Berezovsky”) at [73]–[79] as follows:
(a) There is a burden on the party asserting: (i) that there is a concluded arbitration agreement; and (ii) that it covers the disputes which are the subject of the court proceedings to prove that that is indeed the case: Ahmad Al-Naimi (T/A Buildmaster Construction Services) v Islamic Press Agency Inc [2000] 1 Lloyd’s Rep 522 at 525 per Waller LJ. If that party cannot prove either limb (i) or limb (ii), then the court has no jurisdiction to grant a stay under s 9(1) read with s 9(4) of the UK Arbitration Act 1996.
(b) If the court cannot decide limb (i) or limb (ii) above in a summary fashion on the written evidence, there are two courses open to it: Birse Construction Ltd v St David Ltd [1999] BLR 194 at 196 per Judge Humphrey Lloyd QC. First, the court can direct an issue to be tried pursuant to Part 62 of the English Civil Procedure Rules (“the English CPR”), which gives the English courts the express power to “decide [the] question”, amongst others, of whether an arbitration agreement has been concluded (see [49] below); or, second, it can stay the court proceedings under its inherent jurisdiction (but not pursuant to s 9(1) read with s 9(4) of the UK Arbitration Act 1996) for the putative arbitral tribunal to decide the issue of the existence and scope of the arbitration agreement.
(c) If the court adopts the former course, it must be satisfied of the existence and scope of the arbitration agreement on a balance of probabilities: Aeroflot v Berezovsky at [73]. If the court adopts the latter course, it must be satisfied that there is an arguable case that the arbitration clause is valid: JSC BTA Bank v Ablyazov [2011] 2 Lloyd’s Rep 129 at [31]–[33] per Christopher Clarke J.
(d) Once the party applying for a stay has established the existence of an apparently concluded arbitration agreement that covers the matters in dispute in the court proceedings, it is for the party resisting the stay application to “satisf[y]” the court that the arbitration agreement is “null and void”, “inoperative” or “incapable of being performed”. This must be proved on a balance of probabilities: Aeroflot v Berezovsky at [77].
(e) The court can, in theory, order a trial of an issue to determine whether the arbitration agreement was “null and void”, “inoperative” or “incapable of being performed”. But, if the evidence and possible findings going to that issue also impinge on the substantive rights and obligations of the parties, the court is unlikely to so order unless the trial can be confined to a relatively circumscribed area of investigation: A v B [2007] 1 Lloyd’s Rep 237 at 261 per Colman J.
[49] With regard to Part 62 of the English CPR, which we referred to in the previous paragraph, r 62.8(3) thereof makes express provision for the English courts to determine issues concerning the existence or scope of an arbitration clause. It states:
Hong Kong SAR
[50] Unlike the English courts, the courts of Hong Kong SAR generally adopt the prima facie approach when hearing applications for a stay of court proceedings in favour of arbitration. In Private Company “Triple V” Inc v Star (Universal) Co Ltd & Another [1995] 2 HKLR 62 (“Triple V”), a decision of the Hong Kong SAR Court of Appeal, Litton VP (with whom Liu JA and Keith J agreed) said at 65, quoting the judge in the court below:
[51] Triple V was cited with approval by the Hong Kong SAR Court of Appeal in PCCW Global Ltd v Interactive Communications Service Ltd [2007] 1 HKLRD 309. Tang VP, who delivered the judgment of the court in that case, cited the passage in Triple V quoted above, and distinguished the English position by emphasising the absence of an equivalent of r 62.8(3) of the English CPR in the Hong Kong SAR’s civil procedure rules. Tang VP said at [60]:
The prima facie standard is well settled in Hong Kong SAR and has also been applied by its courts at first instance: T v TS [2014] 4 HKLRD 772 at [16] and Ling Yan Temple Ltd v Ng Yook Man [2010] HKEC 734 at [18].
Canada
[52] Turning to the position in Canada, the leading decision there is that of the Supreme Court of Canada in Union des consommateurs v Dell Computer Corp [2007] 2 SCR 801 (“Dell Computer”). The majority in that case likewise preferred the prima facie approach, but with a caveat, which we shall come to in a moment. Dell Computer was a decision concerning Quebec’s Code of Civil Procedure (Cap C-25), the relevant provisions of which read as follows:
[53] Deschamps J (with whom McLachlin CJ, Binnie, Abella, Charron and Rothstein JJ concurred) observed that the prima facie test was “increasingly gaining acceptance around the world”: Dell Computer at [83]. She said that any question relating to an arbitral tribunal’s jurisdiction should, as a general rule, first be resolved by the arbitral tribunal itself. The exception was where the dispute as to jurisdiction could be resolved purely as a matter of law. Deschamps J thought that this exception was justified for several reasons, namely (Dell Computer at [84]):
[54] However, Deschamps J also made it clear that as long as the challenge to an arbitral tribunal’s jurisdiction required the production and review of factual evidence, the court should refer the issue to the arbitral tribunal itself. The same was said for questions of mixed law and fact, unless those questions required only superficial consideration of the documentary evidence already in the record: Dell Computer at [85]. Deschamps J also said that before the court departed from the general rule that the arbitral tribunal should first hear disputes as to its own jurisdiction, the court should be satisfied that the challenge to the arbitral tribunal’s jurisdiction was not a delay tactic and would not unduly impair the conduct of the arbitral proceedings: Dell Computer at [86]. Deschamps J considered and rejected a suggestion that a distinction should be drawn between cases where the validity of the arbitration clause was in question, as opposed to cases where only the scope of the arbitration clause was in question.
[55] The dichotomy between validity and scope formed the basis of the position taken by the dissenting judges in Dell Computer, Bastarache and LeBel JJ. They thought that the question of an arbitral tribunal’s jurisdiction should not be deferred to the arbitral tribunal where the validity of the arbitration agreement was in question (Dell Computer at [176]), save that in circumstances which “truly merit the label ‘international commercial arbitration’, it may be more efficient to submit all questions regarding jurisdiction for the arbitrator to hear at first instance”: Dell Computer at [178].
[56] The prima facie approach has also been adopted and applied in Canadian provinces other than Quebec. The leading cases in British Columbia and Ontario are the decisions of their respective Courts of Appeal in Gulf Canada Resources Ltd/Ressources Gulf Canada Ltée v Arochem International Ltd (1992) 66 BCLR (2d) 113 (“Gulf Canada”) and Dalimpex v Janicki (2003) 228 DLR (4th) 179 (“Dalimpex”). Both Gulf Canada and Dalimpex were cited with approval by the majority in Dell Computer at [82]. All three cases were applied recently in New Brunswick in Harrison v UBS Holding Canada Ltd (2014) 418 NBR (2d) 328, where the New Brunswick Court of Appeal accepted the prima facie approach, and held that the judge at first instance erred in applying a balance of probabilities test to determine the existence and applicability of an arbitration clause on an application for a stay of court proceedings in favour of arbitration (at [26]).
The position in Singapore
[57] Against that background, we turn to the position in Singapore. There are three Singapore decisions which have addressed the threshold question, and all of them have adopted the prima facie approach. The first is the decision of this court in Sim Chay Koon and others v NTUC Income Insurance Co-operative Limited [2015] SGCA 46, an ex tempore judgment where we observed at [5], in the context of a stay application under s 6(1) of the Arbitration Act (Cap 10, 2002 Rev Ed), that the prima facie standard was the applicable one.
[58] The second is The Titan Unity [2013] SGHCR 28, a decision of the learned Assistant Registrar (“AR”), Mr Shaun Leong Li Shiong. There, the plaintiff bank provided financing to a company (“Onsys”) for the purchase of a cargo of fuel. The plaintiff, as the lawful holder of the bills of lading, brought a misdelivery claim against the owner of the vessel on which the cargo was shipped as well as against the alleged demise charterer of the vessel (“Oceanic”). Oceanic applied for a stay of the court proceedings under s 6 of the IAA, arguing that the arbitration clause in the time charterparty between it and Onsys had been incorporated into the bills of lading on the basis of which the plaintiff brought its claim. AR Leong examined a number of the authorities discussed above and concluded (at [34] of The Titan Unity):
AR Leong thought that the prima facie threshold had been crossed on the facts of the case, and granted the stay.
[59] The third Singapore decision is Malini Ventura v Knight Capital Pte Ltd and others [2015] SGHC 225 (“Ventura v Knight Capital”), a judgment of Judith Prakash J which was issued shortly after we heard oral arguments in the present appeals. In Ventura v Knight Capital, the defendant lenders commenced arbitration against the plaintiff (“Ms Ventura”), a purported guarantor for a loan under a deed of guarantee containing an arbitration clause, after the borrower had defaulted in making repayment of the loan. Ms Ventura’s position was that the deed of guarantee was a nullity because she had never signed it. She commenced an action in the High Court to restrain the lenders from continuing with the arbitral proceedings against her.
[60] The lenders applied under s 6 of the IAA for a stay of the court proceedings. They urged the court to adopt the prima facie approach and defer the question of the existence of the arbitration agreement to the arbitral tribunal (which had already been constituted). Ms Ventura, on the other hand, argued that in situations where a party “denied ever entering into the alleged arbitration agreement so that its very existence [was] brought into question”, the court should adopt the full merits approach and determine the question of whether any arbitration agreement had been reached based on the usual civil standard of a balance of probabilities after a full trial: Ventura v Knight Capital at [24].
[61] Prakash J described the issue before her as “the chicken and the egg question”: Ventura v Knight Capital at [19]. She began by considering the position under the Model Law. She observed that under the Model Law regime, the court’s consideration of whether an arbitral tribunal had jurisdiction “must come after the tribunal’s own examination of the issue” (at [28]). She also observed that under the Model Law, an arbitral tribunal’s powers were wide and extended to ruling on the very existence of the arbitration agreement in question. Prakash J next turned to consider the position under s 6 of the IAA. She held that the IAA had been enacted to incorporate the Model Law as part of Singapore law, and strictly circumscribed court intervention in arbitral proceedings. Prakash J concluded that the regime in force under the IAA “gives primacy to the tribunal although, of course, the court still has an important role to play” (at [36]). She thought that “it would satisfy the rights of both parties if the party applying for the stay was able to show on a prima facie basis that the arbitration agreement existed” (likewise at [36]). She therefore adopted the prima facie approach and, in doing so, she distinguished the position in England. She recognised that ss 6(1) and 6(2) of the IAA were worded similarly to ss 9(1) and 9(4) of the UK Arbitration Act 1996, but pointed out that “the context of the two sections is not the same” (at [35]). Prakash J alluded to the “logical discomfort” of adopting the prima facie approach in cases where the very existence of the arbitration agreement was in question (at [37]). She thought, however, that that discomfort had to be put aside in the light of the widespread acceptance and effect of the kompetenz-kompetenz principle.
[62] On the facts of the case before her, Prakash J held that the lenders had established a prima facie case that the deed of guarantee had been signed by Ms Ventura despite her protestations otherwise. Prakash J therefore stayed the court proceedings in favour of arbitration.
[63] The prima facie approach was also the view urged upon us by the amicus curiae, Prof Boo. We agree that a Singapore court should adopt a prima facie standard of review when hearing a stay application under s 6 of the IAA. In our judgment, a court hearing such a stay application should grant a stay in favour of arbitration if the applicant is able to establish a prima facie case that:
(a) there is a valid arbitration agreement between the parties to the court proceedings;
(b) the dispute in the court proceedings (or any part thereof) falls within the scope of the arbitration agreement; and
(c) the arbitration agreement is not null and void, inoperative or incapable of being performed.
[64] Once this burden has been discharged by the party applying for a stay, the court should grant a stay and defer the actual determination of the arbitral tribunal’s jurisdiction to the tribunal itself. The court will only refuse to grant a stay when it is clear on the evidence placed before it that one or more of the above three requirements have not been satisfied. The arbitral tribunal’s determination of its jurisdiction will nonetheless remain subject to overriding court supervision in the form of an appeal under s 10(3) of the IAA against the arbitral tribunal’s jurisdictional ruling, or in proceedings for setting aside or refusing enforcement of the award rendered by the arbitral tribunal (see, respectively, s 24 of the IAA and Art 34 of the Model Law, and s 31 of the IAA).
[65] We part company with the English position and adopt the prima facie approach for the purposes of the threshold question essentially for four reasons. First, the prima facie approach coheres better with what we consider was envisaged by the drafters of the IAA. The earliest iteration of the IAA (viz, the International Arbitration Act 1994 (Act 23 of 1994) (“the original IAA”)) was enacted in 1994, and it drew heavily from the recommendations made in the Report of the Sub-committee on Review of Arbitration Laws (1993) (Chairman: Giam Chin Toon) (“1993 Report on Review of Arbitration Laws”). That report included a draft Bill, which was considered and adopted with amendments by the Singapore Academy of Law’s Law Reform Committee, and this subsequently resulted in the enactment of the original IAA in 1994 (see the remarks of Assoc Prof Ho Peng Kee, the then Parliamentary Secretary to the Minister for Law, at the second reading of the International Arbitration Bill 1994 (Bill 14 of 1994) (“the 1994 International Arbitration Bill”): Singapore Parliamentary Debates, Official Report (31 October 1994) vol 63 (“Singapore Parliamentary Debates vol 63”) at cols 627–628).
[66] The 1993 Report on Review of Arbitration Laws stated at paras 37–38:
It is implicit, if not explicit, in this passage that the arbitral tribunal is to be the first arbiter of its own jurisdiction, with the court having the final say. This point is underscored by an earlier recommendation made in the 1993 Report on Review of Arbitration Laws at para 23 to confine the scope of curial intervention to three narrow and well-defined areas:
These two passages, in our judgment, support the construction of s 6 of the IAA as requiring the prima facie approach where the threshold question is concerned.
[67] Second, to require the court, on a stay application under s 6 of the IAA, to undertake a full determination of an arbitral tribunal’s jurisdiction could significantly hollow the kompetenz-kompetenz principle of its practical effect. The full merits approach has the potential to reduce an arbitral tribunal’s kompetenz-kompetenz to a contingency dependent on the strategic choices of the claimant in a putative arbitration. If the claimant decides to pursue its claim by arbitration, the arbitral tribunal will determine any challenge to its jurisdiction, and thus, its kompetenz-kompetenz will be given full vent. But, if the claimant decides to pursue its claim by bringing proceedings in court (instead of by recourse to arbitration), the court will be seized of jurisdiction, and will be able (and, indeed, on the full merits approach, obliged) to make a full determination on the existence and scope of the arbitration clause; this will deprive the putative arbitral tribunal of its kompetenz-kompetenz. In our view, the strength of the kompetenz-kompetenz principle cannot depend on the arbitrary choice of the claimant as to whether it will pursue its claim by way of court proceedings or by way of arbitration. That undermines the principles of judicial non-intervention and kompetenz-kompetenz which were at the forefront in the drafting of the Model Law and the enactment of the original IAA (see Assoc Prof Ho’s remarks at the second reading of the 1994 International Arbitration Bill: Singapore Parliamentary Debates vol 63 at cols 625–626). We should point out that the strain which the English position puts on these principles of judicial non-intervention and kompetenz-kompetenz has not escaped criticism (see Arbitration Law (Robert Merkin gen ed) (informa, Looseleaf Ed, 15 August 2011 release) at para 8.21, as well as David Joseph QC, Jurisdiction and Arbitration Agreements and their Enforcement (Sweet & Maxwell, 2nd Ed, 2010) (“Jurisdiction and Arbitration Agreements”) at pp 346–347). This difficulty is avoided if the prima facie approach is adopted.
[69] Finally, contrary to what the English Court of Appeal suggested in Aeroflot v Berezovsky at [73], we do not think the word “satisfied” in s 6(2) of the IAA suggests that the court is required to conduct a full merits review when it is faced with the threshold question. The UK Advisory Committee on Arbitration Law, in the 1996 Report on the Arbitration Bill (cited at [46] above), stated at para 54 that the language used in what is now s 9 of the UK Arbitration Act 1996 was “the language of the Model Law and of course of the [New York Convention], presently to be found in the [UK] Arbitration Act 1975”. The drafters of the UK Arbitration Act 1996 thus did not draw a distinction between the term used in the English legislation (“satisfied”) and that used in the Model Law (“finds”). We have expressed the view above that the Model Law, as can be seen from its travaux préparatoires, is equivocal on the appropriate approach to take with regard to the threshold question, and thus, we do not think the use of the word “satisfied” brings the matter further one way or another. If anything, a textual analysis of s 6(2) of the IAA, which replaces the word “finds” in Art 8(1) of the Model Law with “satisfied”, would favour a prima facie threshold rather than a determination on a balance of probabilities.
[70] We therefore hold that in considering a stay application under s 6 of the IAA, the court need only be satisfied to a prima facie standard that the requirements for the grant of a stay under that section have been met before ordering a stay. Our subsequent discussion of the issues in these appeals should therefore be construed in this light by any arbitral tribunal that might in due course hear the dispute in the Suit.
Whether a dispute over minority oppression or unfairly prejudicial conduct is arbitrable
The concept of arbitrability
[71] We turn now to the question of arbitrability. The absence of arbitrability has come to be associated with that class of disputes which are thought to be incapable of settlement by arbitration. The concept of arbitrability has a reasonably solid core. It covers matters which “so pervasively involve ‘public’ rights and concerns, or interests of third parties, which are the subjects of uniquely governmental authority, that agreements to resolve … disputes [over such matters] by ‘private’ arbitration should not be given effect”: Gary Born at p 945. However, the outer limits of its sphere of application are less clear. Lord Mustill and Stewart Boyd QC, for instance, suggest that “[i]t would be wrong … to draw … any general rule that criminal, admiralty, family or company matters cannot be referred to arbitration”: Michael J Mustill & Stewart C Boyd, The Law and Practice of Commercial Arbitration in England (Butterworths, 2nd Ed, 1999) at pp 149–150.
[72] Arbitrability may be relevant to an arbitration at both its initial and its terminal stages. At the terminal stage, an award rendered on a non-arbitrable dispute is liable to be set aside or refused enforcement: Art 34(2)(b)(i) of the Model Law and s 31(4)(a) of the IAA. It must follow from this that the non-arbitrability of a dispute will be a ground for refusing a stay of court proceedings in favour of arbitration. Indeed, in the court below, this was assumed to be the case by the parties as well as by the Judge. After all, it would be pointless for the court to stay court proceedings in favour of arbitration in cases where the applicable law does not permit the subject matter of the dispute to be resolved by arbitration, since to grant a stay in such circumstances would be to compel an arbitration which may lead to an award without force or legal value: Larsen Oil and Gas Pte Ltd v Petroprod Ltd (in official liquidation in the Cayman Islands and in compulsory liquidation in Singapore) [2011] 3 SLR 414 (“Larsen Oil v Petroprod”) at [26].
[73] This leads us to the relevance of arbitrability at the initial stage of an arbitration. Section 6(2) of the IAA, which sets out the circumstances in which the court may refuse a stay of court proceedings in favour of arbitration, does not include the non-arbitrability of the subject matter of the dispute as one of the grounds for refusing a stay. Instead, s 6(2) states that a stay may only be refused if the arbitration agreement in question is “null and void”, “inoperative” or “incapable of being performed”. The exclusion of non-arbitrability in express terms might seem to give rise to an “apparent inconsistency” (see Holtzmann & Neuhaus at p 303). But, as noted in the same work, this has “little, if any, effect” in practice for two reasons: first, Art 1(5) of the Model Law preserves the force of the State’s laws on arbitrability; and second, “the term ‘null and void’ would ‘normally’ include considerations of arbitrability” (see Holtzmann & Neuhaus at p 304).
[74] We accept that a dispute concerning a non-arbitrable subject matter would fall within one or more of the three exceptions in s 6(2) of the IAA that permit a court to refuse a stay of court proceedings in favour of arbitration. The English Court of Appeal expressed a tentative preference for fitting a non-arbitrable dispute within the “inoperative” limb of s 9(4) of the UK Arbitration Act 1996 in Fulham Football Club (1987) Ltd v Richards and another [2012] Ch 333 (“Fulham FC v Richards”) at [35]–[36]. In our judgment, an arbitration agreement would be either “inoperative” or “incapable of being performed” in relation to a dispute which involves a subject matter that is not arbitrable. The real issue then is as to the proper ambit of arbitrability.
[75] The concept of arbitrability finds legislative expression in s 11 of the IAA, which reads as follows:
It is evident from this that the essential criterion of non-arbitrability is whether the subject matter of the dispute is of such a nature as to make it contrary to public policy for that dispute to be resolved by arbitration. Beyond this, the scope and extent of the concept of arbitrability has been left undefined, as a consequence of which, it falls to the courts to trace its proper contours (see the 1993 Report on Review of Arbitration Laws at paras 26–28; Larsen Oil v Petroprod at [24]).
[76] In our judgment, the effect of s 11 of the IAA is that there will ordinarily be a presumption of arbitrability so long as a dispute falls within the scope of an arbitration clause. This presumption may be rebutted by showing that (Larsen Oil v Petroprod at [44]):
(a) Parliament intended to preclude a particular type of dispute from being arbitrated (as evidenced by either the text or the legislative history of the statute in question); or
(b) it would be contrary to the public policy considerations involved in that type of dispute to permit it to be resolved by arbitration.
[77] In Larsen Oil v Petroprod, this court found that the above conditions were satisfied, and the presumption of arbitrability was accordingly rebutted. In that case, the respondent (“Petroprod”), which was in liquidation, commenced proceedings against the appellant (“Larsen”), with whom it had a management agreement. Petroprod sought to avoid, as unfair preferences or transactions at an undervalue, payments that it had made to Larsen under the management agreement. Relying on an arbitration clause in that agreement, Larsen applied for a stay of the court proceedings. Petroprod resisted the stay application on the grounds that the dispute did not fall within the scope of the arbitration clause, and that in any event, the dispute was not arbitrable. Petroprod succeeded on the first point, which was sufficient to warrant the High Court’s dismissal of Larsen’s stay application; but, on appeal by Larsen, we nonetheless proceeded to consider whether claims made pursuant to the avoidance provisions in ss 98 and 99 of the Bankruptcy Act (Cap 20, 2009 Rev Ed) read with s 329(1) of the Companies Act were arbitrable, and affirmed the High Court’s decision that they were not. We explained that allowing the arbitration of claims which arose upon insolvency would run contrary to the objectives of the insolvency regime (Larsen Oil v Petroprod at [45]):
[78] Another area widely regarded as non-arbitrable is the liquidation of an insolvent company by the court. In Four Pillars Enterprises Co Ltd v Beiersdorf Aktiengesellschaft [1999] 1 SLR(R) 382, we observed in passing (at [23]) that a winding-up order was a relief which was not available in arbitration, and that “[o]nly the court could grant such relief”.
[79] The issue of whether a dispute over the liquidation of a company is arbitrable was given fuller treatment in A Best Floor Sanding Pty Ltd v Skyer Australia Pty Ltd [1999] VSC 170 (“Skyer Australia”), a first instance decision of the Supreme Court of Victoria. The applicant in that case (“AB Floor”) and the respondent (“Skyer”) merged their respective businesses by establishing a joint venture company (“Harvest”), with each of them holding 50% of Harvest’s shares. The joint venture agreement between AB Floor and Skyer provided that “any dispute, difference or question … touching [on] the association [between AB Floor and Skyer] or … the dissolution or winding up thereof” was to be settled by arbitration: Skyer Australia at [6]. Skyer made an application in its capacity as a contributory to wind up Harvest. In response, AB Floor applied for a stay of Skyer’s application in favour of arbitration.
[80] Warren J held that the arbitration clause was null and void, and refused to grant the stay. She said at [13] and [18]:
[81] More recently, in In re Cybernaut Growth Fund LP Cause No FSD 73 of 2013 (23 July 2013) (unreported) (“Cybernaut”), Judge Andrew Jones QC, sitting in the Financial Services Division of the Grand Court of the Cayman Islands, held that a petition for the liquidation of an exempted limited partnership was non-arbitrable. The learned judge stated that two reasons militated against the conclusion that such a petition was arbitrable (at [7]):
[82] The first of the two reasons given by Judge Jones QC for his finding of non-arbitrability in part concerns the inadequacy of the remedies that an arbitrator may grant, as compared to the remedies that may be within a court’s jurisdictional competence to grant. This factor was also relied on by the Judge in this case in dismissing the stay applications before him. We agree that it is beyond the powers of an arbitrator to liquidate a company, but we do not agree that an arbitrator’s lack of power to grant a particular relief sought should necessarily dictate the outcome of the inquiry into the arbitrability of the dispute in respect of which that relief is sought. The simple point is that it just does not follow from the limitations which might apply to an arbitrator’s powers that the dispute itself is thus unsuitable for resolution by arbitration. We shall develop this point further at [96]–[100] below.
[83] As for Judge Jones QC’s second reason for ruling in Cybernaut that the dispute before him was non-arbitrable, we do regard that as compelling. In our judgment, a dispute over the liquidation of an insolvent company is non-arbitrable because such a liquidation is a process in which the greater public beyond the parties to the dispute have an interest. A creditor’s winding-up petition seeks “an order to put a company into an insolvent liquidation that will affect the interests of all creditors as well as of all members”: Gamlestaden Fastigheter AB v Baltic Partners Ltd and others [2007] Bus LR 1521 at [32] per Lord Scott of Foscote. Because of this, a person who presents a winding-up application is required to advertise the application under the Companies (Winding Up) Rules (Cap 50, R 1, 2006 Rev Ed) (“the Companies (Winding Up) Rules”). The advertisement plays a crucial role in protecting the interests of parties whose rights stand to be adversely affected by the grant of a winding-up order, and who therefore have a right to be heard. In Re a Company (No 007923 of 1994) [1995] 1 WLR 953, Nourse LJ said at 958F–958H that an advertisement served the purpose of giving notice primarily to creditors and contributories who were entitled to be heard on the winding-up application, and secondarily, to those who might trade with the company in the window between the filing of the winding-up application and its final determination, and “who might thus be adversely affected by the provisions of the insolvency regime”.
[84] In our judgment, a claim for relief under s 216 of the Companies Act stands on a different footing from the liquidation of an insolvent company or avoidance claims that arise upon insolvency because the former generally does not engage the public policy considerations involved in the latter two situations. There is certainly nothing in the text of s 216 to suggest an express or implied preclusion of arbitration. Nor does the legislative history and statutory purpose of the provision suggest that a dispute over minority oppression or unfair prejudice is of a nature which makes it contrary to public policy for the dispute to be adjudicated by an arbitral tribunal.
[85] Section 216 of the Companies Act first featured as s 181 of the Companies Act 1967 (Act 42 of 1967), the earliest iteration of what is now the Companies Act (as defined at [1] above). Section 216 has been preserved almost unchanged since its introduction, save for (among other minor amendments) an amendment in 1984 vide the Companies (Amendment) Act 1984 (Act 15 of 1984), pursuant to which the court was conferred the power to sanction a derivative action: see s 216(2)(c). Section 216 of the Companies Act was modelled on s 210 of the Companies Act 1948 (c 38) (UK) (“the UK Companies Act 1948”), which, in turn, was enacted pursuant to a recommendation by the UK Committee on Company Law Amendment (“the Cohen Committee”) in 1945. The Cohen Committee’s recommendation was its response to a perceived need to “strengthen the minority shareholders of a private company in resisting oppression by the majority”: Report of the Committee on Company Law Amendment (Cmd 6659, 1945) (Chairman: Mr Justice Cohen) (“Report of the Cohen Committee”) at para 60.
[86] The Report of the Cohen Committee (at paras 59–60) gave two examples of such “oppression by the majority”. The first was the hardship occasioned by reason of restrictions on share transfers in private companies when a minority shareholder passed away. Such restrictions allowed the majority (who were also usually directors of the company concerned) to force the executors of the minority shareholder’s estate to sell the deceased’s shareholding to the majority at extortionate prices. The second example concerned the payment of excessive remuneration to directors, which resulted in there being little or nothing left for dividend distributions to shareholders. The Cohen Committee proposed that the court be given an “unfettered” discretion to impose on the disputing parties whatever settlement it considered just and reasonable in such circumstances. This proposal gave rise to s 210 of the UK Companies Act 1948. The subsequent 1962 UK Committee on Company Law Amendment (more commonly known as “the Jenkins Committee”), when reviewing s 210, emphasised that it was intended to deal with management disputes between shareholders of privately-run companies: Report of the Committee on Company Law Amendment (Cmnd 1749, 1962) (Chairman: Lord Jenkins) at para 199.
[87] Section 216 of the Companies Act has undoubtedly increased in scope and prevalence from the two narrow instances of minority oppression that were initially referenced by the Cohen Committee in 1945. But, the essence of a claim for relief under this provision remains the same. A modern statement of the nature of such a claim is found in a famous passage of Lord Hoffmann’s judgment in O’Neill and Another v Phillips and Others [1999] 1 WLR 1092 (“O’Neill v Phillips”) at 1098–1099 in relation to s 459 of the Companies Act 1985 (c 6) (UK), the then English equivalent of s 216 of the Companies Act:
This exposition was endorsed by this court in Lim Swee Khiang v Borden Co (Pte) Ltd [2006] 4 SLR(R) 745 at [82].

The judgment continues past this page. Read the rest at eLitigation.

Tomolugen Holdings Ltd and another v Silica Investors Ltd and other appeals

Laws this judgment relies on

Grounds of decision

Introduction

Factual background

Overview of the material facts

The Share Sale Agreement

The allegations made in the Suit

The decision below

The issues before us

The threshold question: the standard of review in a stay application under the IAA

The backdrop to the threshold question

The travaux préparatoires of the Model Law

The position in other jurisdictions

England

Hong Kong SAR

Canada

The position in Singapore

Whether a dispute over minority oppression or unfairly prejudicial conduct is arbitrable

The concept of arbitrability