COSCO Shipping Specialized Carriers Co, Ltd v PT OKI Pulp & Paper Mills and others [2024] SGHC 92

HC/SUM 2676/2023 (“SUM 2676”) is an application by COSCO Shipping Specialized Carriers Co., Ltd. (“CSSC”), the Claimant in HC/ADM 50/2022 (“ADM 50”), for an anti-suit injunction against the 1st Defendant, PT OKI Pulp & Paper Mills (“OKI”).

The initial hearing of SUM 2676 took place before me on 27 September 2023. I dismissed the application on 27 December 2023 and gave oral reasons for my decision. Counsel for CSSC subsequently wrote in on 10 January 2024 to request that I hear further arguments pursuant to O 18 r 28 of the Rules of Court 2021 (the “FA Request”). I allowed that request and heard the further arguments on 7 February 2024 (the “FA Hearing”).

Having considered the facts and submissions in their totality (including the arguments made at the FA Hearing), I affirm my earlier decision to dismiss SUM 2676. These are the full grounds of my decision.

The facts

The parties

CSSC is a company incorporated in China. It is in the business of operating and managing specialised vessels under the wider COSCO Shipping group. CSSC was at all material times the owner of the vessel “LE LI” (the “Vessel”).

The 2nd Defendant, COSCO Shipping Specialized Carriers (Europe) B.V. (“COSCO Europe”), is a company incorporated in the Kingdom of the Netherlands. CSSC and COSCO Europe are related entities, but nothing turns on this.

OKI is a company incorporated in Indonesia that is in the business of manufacturing paper pulp and paper products. One of its manufacturing facilities (the “Mill”) is located in Palembang, Indonesia. OKI also claims to own and operate a nearby port facility (the “Terminal”). Part of the Terminal consists of a jetty (the “Jetty”) located approximately 2,050 metres offshore. A trestle bridge (the “Trestle Bridge”) connects the Jetty to the mainland. The pulp and paper products manufactured at the Mill are trucked to the Jetty and loaded onto vessels berthed there.

The contractual arrangements between the parties

By a contract of affreightment and an accompanying addendum both dated 6 April 2021 (collectively, the “Head COA”), the Vessel was chartered by CSSC (as shipowner) to COSCO Europe (as head charterer).

The Vessel was sub-chartered by COSCO Europe to OKI under a further contract of affreightment and an accompanying addendum, both also dated 6 April 2021 (collectively, the “Sub-COA”).

On or about 31 May 2022, a cargo of approximately 27,000 air-dried metric tonnes of bleached hardwood kraft pulp acacia PEFC (the “Cargo”) was loaded on board the Vessel while she was berthed at the Jetty. The Cargo was variously destined for Changshu Port in China and Kunsan Port in South Korea.

Nine bills of lading (the “B/Ls”) were issued by CSSC (as carrier) to OKI (as shipper) in respect of the Cargo. The B/Ls were all dated 31 May 2022.

The Incident and the Indonesian Proceedings

On 31 May 2022, shortly after the Vessel had cast off from the Jetty and was departing from the Terminal with the assistance of tugs, she made contact with the Trestle Bridge. This caused a section of the Trestle Bridge spanning some 220 metres to collapse. I will refer to these events as the “Incident”.

On or about 26 October 2022, OKI commenced proceedings against CSSC in the Kayu Agung District Court, Indonesia to claim for losses it allegedly suffered in consequence of the Incident (the “Indonesian Proceedings”). OKI initially estimated its losses to amount to US$592,787,794.00, although that estimate was subsequently revised downwards to US$269,307,341.00.

The court proceedings in Singapore

On 4 August 2022, CSSC commenced this action (ie, ADM 50) to limit its liability arising out of the Incident, pursuant to Part 8 of the Merchant Shipping Act 1995 (2020 Rev Ed) (the “MSA”). Various applications and cross-applications have since been brought in connection with ADM 50, one of which is SUM 2676. Rather than wade into the procedural minutiae, I will only set out such details as are necessary to contextualise my decision in SUM 2676.

On 25 August 2022, and following service of the originating claim on COSCO Europe, CSSC applied by way of HC/SUM 3219/2022 (“SUM 3219”) for, among other things, the grant of a decree of limitation (the “Limitation Decree”).

OKI filed its Notice of Intention to Contest (“NIC”) on 11 October 2022. The hearing of SUM 3219, which was initially fixed for 12 October 2022, was adjourned in light of OKI’s indication that it intended to challenge the court’s jurisdiction to entertain ADM 50.

On 24 November 2022, OKI proceeded to apply by way of HC/SUM 4238/2022 (“SUM 4238”) for, among other things, declarations that:

I heard SUM 3219 and SUM 4238 on 17 April 2023. I dismissed SUM 4238 on 22 May 2023; my decision in respect of that application is set out in COSCO Shipping Specialized Carriers Co, Ltd v PT OKI Pulp & Paper Mills and others [2023] SGHC 149. SUM 3219 was further adjourned to be heard on a later date.

Shortly after I delivered judgment on SUM 4238, OKI applied on 28 July 2023 by way of HC/SUM 2302/2023 (“SUM 2302”) for leave to withdraw its NIC. SUM 2302 was heard on 15 September 2023 by an Assistant Registrar (the “AR”), who allowed OKI’s application. OKI thereafter filed the Notice of Withdrawal of its NIC on 16 September 2023. CSSC appealed against the AR’s decision by way of HC/RA 197/2023 (“RA 197”). I heard RA 197 on 25 September 2023 and dismissed it on the same day.

On 25 August 2023, CSSC – having since come to learn of the Indonesian Proceedings – applied by way of SUM 2676 for, among other things, an anti-suit injunction to enjoin OKI from pursuing the Indonesian Proceedings. The application and supporting affidavits were electronically served on OKI through their solicitors, Messrs Clasis LLC (“Clasis”), on 4 September 2023 prior to OKI’s withdrawal from this action on 16 September 2023.

By the time I heard SUM 2676 on 27 September 2023, OKI was no longer a party to the proceedings (although it had previously filed affidavits and submissions in response to SUM 2676, without prejudice to its jurisdictional challenge and application to withdraw its NIC). Accordingly, I heard SUM 2676 as though it were an ex parte application. For completeness, COSCO Europe – who were represented by Messrs JLex LLC – attended but otherwise played no active role in SUM 2676. COSCO Europe did not object to SUM 2676. After hearing CSSC’s arguments, I reserved judgment.

I next heard SUM 3219 on 5 October 2023 and granted CSSC the Limitation Decree it sought, albeit on amended terms.

As I mentioned at [2] above, I dismissed SUM 2676 on 27 December 2023 and provided oral reasons for my decision. CSSC subsequently made the FA Request on 10 January 2024. I allowed the request and heard CSSC’s further arguments on 7 February 2024.

The arbitral proceedings in Singapore

On 26 October 2022, CSSC commenced arbitral proceedings in Singapore against COSCO Europe pursuant to an arbitration agreement in the Head COA (which I discuss at [37]–[39] below). In its Notice of Arbitration to COSCO Europe, CSSC sought (among other things) a declaration that CSSC had not breached its obligations under the Head COA, as well as various reliefs in respect of its liability arising out of the Incident. I was given to understand that that arbitration is ongoing.

On 19 September 2023, CSSC commenced separate arbitral proceedings against OKI in Singapore. The circumstances said to have entitled CSSC to commence those proceedings are likewise discussed in further detail below. In that arbitration, CSSC has sought as against OKI declarations of non-liability and various reliefs in respect of CSSC’s liabilities/losses arising out of the Incident, among other things.

At the FA Hearing, I was informed by counsel for CSSC, Mr Toh Kian Sing SC, that OKI had brought a jurisdictional challenge before the Court of Arbitration of the Singapore International Arbitration Centre (the “SIAC”) in respect of the arbitral proceedings brought by CSSC against it. For that reason, an arbitral tribunal had yet to be constituted.

Summary of CSSC’s grounds for an anti-suit injunction

Before me, Mr Toh advanced four grounds in support of CSSC’s application for an anti-suit injunction:

General principles on the grant of anti-suit injunctions

Anti-suit injunctions granted in exercise of the court’s equitable jurisdiction

The overarching principles governing the court’s equitable jurisdiction to grant anti-suit injunctions are settled (see Lakshmi Anil Salgaocar v Jhaveri Darsan Jitendra [2019] 2 SLR 372 at [49]):

In deciding whether to grant an anti-suit injunction, there are five factors that fall to be considered (see BCS Business Consulting Services Pte Ltd and others v Baker, Michael A (executor of the estate of Chantal Burnison, deceased) [2023] 1 SLR 1 (“BCS”) at [34]):

Subsequent references in these written grounds to factors (a) – (e) should be understood as references to the abovementioned factors in that order.

Although it has sometimes been said that the five factors must be considered in the round, it is now clear law that factor (e) supplies an independent ground for the grant of a “contractual” anti-suit injunction. If the applicant can show that foreign proceedings have been instituted in breach of an arbitration or exclusive jurisdiction agreement, anti-suit relief should ordinarily be granted unless there are strong reasons not to do so: Sun Travels & Tours Pvt Ltd v Hilton International Manage (Maldives) Pvt Ltd [2019] 1 SLR 732 at [67]–[68]; VKC v VJZ [2021] 2 SLR 753 (“VKC”) at [16].

Factors (a) to (d), on the other hand, are more relevant to the grant of “non-contractual” anti-suit injunctions: VKC at [18].

Anti-suit injunctions granted in exercise of the court’s inherent jurisdiction

Unlike Grounds 1 to 3 advanced by OKI (which sought to invoke the court’s equitable jurisdiction to grant anti-suit injunctions), Ground 4 was an appeal to the court’s inherent jurisdiction.

It is settled that there exists “the inherent power of the forum court to protect the integrity of its processes once set in motion” through the grant of anti-suit injunctions: BCS at [54]. For convenience, I will refer to these injunctions as “protective anti-suit injunctions”.

Unlike anti-suit injunctions granted in exercise of the court’s equitable jurisdiction (which are intended to protect the injunction applicant’s rights), the object of protective anti-suit injunctions is to safeguard the forum court’s processes from interference: BCS at [54].

The principles that govern the grant of protective anti-suit injunctions may be summarised as follows:

Ground 1: Whether the Indonesian Proceedings had been brought in breach of the Arbitration Agreement

CSSC’s submissions

CSSC’s claim for injunctive relief was advanced primarily on the ground that the Indonesian Proceedings had been commenced by OKI in breach of a valid and binding arbitration agreement between CSSC and OKI.

I noted at the outset that there was some ambiguity as to the existence of the arbitration agreement. Let me explain why. The B/Ls, which were in the well-known CONGENBILL 94 form, contained an incorporation clause (the “Incorporation Clause”) on the reverse which reads:

There was, however, nothing on the face of the B/Ls to indicate if the Incorporation Clause referred to the Head COA or the Sub-COA.

To overcome this ambiguity, CSSC relied on the principle that where it is unclear which charterparty an incorporating clause refers to, the court should presume that the terms of the head charterparty are incorporated “since [that] is the contract to which the shipowner, who issues the bill of lading, is a party”: Pacific Molasses Co. and United Molasses Trading Co. Ltd. v Entre Rios Compania Naviera S.A. (The “San Nicholas”) [1976] 1 Lloyd’s Rep. 8 at 11.

On that footing, I was prepared to interpret the Incorporation Clause as referring to the Head COA, so that cl 61 of the Head COA (the “Arbitration Agreement”) was incorporated into the B/Ls. That clause reads:

It was thus necessary for me to decide if the Indonesian Proceedings were brought in respect of a dispute that arose “out of or in connection with” the B/Ls.

OKI’s opening salvo was that its claim in the Indonesian Proceedings (ie, for damage caused to the Trestle Bridge) was a pure tort claim that could not be said to have arisen “out of or in connection with” the B/Ls. CSSC unsurprisingly objected to OKI’s characterisation of its own claim; CSSC’s case was that OKI had deliberately pleaded its claim in tort so as to circumvent the Arbitration Agreement. Thus, before I could even consider the scope of the Arbitration Agreement, I had to first determine the anterior threshold question of how OKI’s claim should be characterised for the purposes of the analysis.

Where the proper characterisation of a legal claim is contested, the ordinary starting point is to first determine the law that the court should apply to dispose of that question. There were, in my view, only three possibilities in the circumstances of this case: English law, Singapore law, or Indonesian law. It was not argued (or even suggested) that any other system of law had any bearing on the parties’ dispute. I was cognisant that in such circumstances, the court can simply fall back to the “default rule” and apply the lex fori as a matter of practicality: Ollech David v Horizon Capital Fund [2024] SGHC(A) 8 at [54]–[56]. As Lord Leggatt put it in FS Cairo (Nile Plaza) LLC v Lady Brownlie [2021] UKSC 45 (at [113]):

In this regard, it was plain to me that as a matter of Singapore law, OKI’s claim could only be characterised as a tort claim. If there were any means by which OKI’s claim could be recast as a breach of the contract of carriage evidenced by the B/Ls, it was not raised in argument and I was unable to discern any such means for myself.

In any event, it was unnecessary for me to arrive at a conclusive determination as to which of the three laws should be applied for the purposes of characterising OKI’s claim. This was because the same result would have followed even if I applied English or Indonesian law.

English law was in the running for two reasons. First, it was expressly designated in the Arbitration Agreement as the substantive law governing the Head COA; since the Arbitration Agreement set out in cl 61 of the Head COA was incorporated into the B/Ls, English law also therefore governed the contract of carriage evidenced therein. Second, there was no express choice of law to govern the Arbitration Agreement itself; in the absence of such an express choice, the parties are presumed – under the prevailing choice of law framework laid down in BCY v BCZ [2017] 3 SLR 357 – to have impliedly chosen the governing law of the contract to also govern the Arbitration Agreement: Anupam Mittal v Westbridge Ventures II Investment Holdings [2023] 1 SLR 349 at [62].

Neither CSSC nor OKI adduced any evidence on how OKI’s claim would be characterised as a matter of English law. The Court could therefore take judicial notice of the fact that there was no material difference in approach between Singapore law and English law: Chan Chin Cheung v Chan Fatt Cheung and others [2010] 1 SLR 1192 at [33]. Alternatively, it would also have been open to this Court to apply the presumption of similarity as between Singapore and English law: The “Chem Orchid” [2015] 2 SLR 1020 at [159]. On that basis, I was satisfied that OKI’s claim would similarly be regarded as a pure tort claim under English law.

I turn now to Indonesian law, which was both the lex loci actus – ie, the law of the place where the act giving rise to the dispute occurred – and also the lex fori in respect of the Indonesian Proceedings.

OKI adduced expert evidence on Indonesian law from Dr H. Zahrul Rabain, S.H., M.H. (“Dr Zahrul”) in support of, among other things, its contention that under Indonesian law, OKI’s claim against CSSC was in truth a tort claim (or the Indonesian equivalent of a tort claim). Dr Zahrul opined in his report that “OKI’s cause of action against CSSC in Indonesia is in tort” pursuant to certain provisions of the Indonesian Codes.

CSSC’s Indonesian law expert, Mr Mokki Arianto (“Mr Arianto”), took the view that Dr Zahrul had referred to the wrong provisions in his analysis. Importantly, however, nothing in any of the expert reports prepared by Mr Arianto challenged or refuted Dr Zahrul’s (and OKI’s) position that OKI was, in the Indonesian Proceedings, pursuing a claim in tort (or the equivalent thereto under Indonesian law). From Mr Arianto’s first report, it appears that he assumed – apparently on CSSC’s instructions – that OKI’s claim in the Indonesian Proceedings had been brought in breach of the Arbitration Agreement. For that reason, there was no discussion or analysis by Mr Arianto on the characterisation of OKI’s claim under Indonesian law. I was accordingly prepared to accept that under Indonesian law, OKI’s claim would likewise be characterised as a tort claim.

For the foregoing reasons, I took the view that OKI’s claim had to be characterised as a pure tort claim for the purposes of deciding if the Indonesian Proceedings had been commenced in breach of the Arbitration Agreement.

CSSC’s fall-back position was that even if OKI was seeking to vindicate a non-contractual claim in the Indonesian Proceedings, the dispute was nevertheless to be regarded as one that arose “out of or in connection with” the B/Ls. To buttress this argument, Mr Toh referred me to various features of the contractual arrangement and the commercial context to make the point that the parties intended for all disputes between them – whether contractual or otherwise – to be referred to arbitration in Singapore.

Mr Toh also pointed out that CSSC was looking to rely on certain contractual provisions which were said to offer CSSC a defence against OKI’s claim. The potential availability of those contractual defences, so the argument went, was a factor weighing heavily in favour of the conclusion that the dispute forming the subject matter of the Indonesian Proceedings was one that arose “out of or in connection with” the B/Ls.

The applicable principles

I took as my starting point the established principle that an arbitration agreement “is to be construed like any other commercial agreement, with a view to giving effect to the intention of the parties as objectively expressed in it”: Insigma Technology Co Ltd v Alstom Technology Ltd [2009] 3 SLR(R) 936 at [30].

An important aspect of this interpretive exercise is the well-known rule of construction that has come to be known as the Fiona Trust or “one-stop shop” presumption: Tomolugen Holdings Ltd and another v Silica Investors Ltd and other appeal [2016] 1 SLR 373 at [124]. In Fiona Trust & Holding Corporation and others v Privalov and others [2007] UKHL 40, Lord Hoffmann explained (at [13]) that:

Inherent in this formulation is a limit to the Court’s generosity in construing arbitration agreements. The ambit of an arbitration agreement must ultimately correspond to the parties’ intentions, which must in turn be objectively ascertained having regard to the words used and the circumstances in which they were agreed upon. As the Court of Appeal noted in Rals International Pte Ltd v Cassa di Risparmio di Parma e Piacenza SpA [2016] 5 SLR 455 (at [34]):

The approach to determining if a stand-alone tort claim falls within the scope of an arbitration agreement

To be clear, I did not for a moment doubt the correctness of the Fiona Trust presumption (which is firmly embedded in our jurisprudence), but it offered me little concrete assistance in determining if a stand-alone tort claim fell within or without the scope of an arbitration agreement. On this point, I found the case of Eastern Pacific Chartering Inc v Pola Maritime Ltd [2021] 1 WLR 5475 (“The Pola Devora”) instructive.

The brief facts of The Pola Devora are these. The claimant shipowners (“Eastern”) and the defendant charterers (“Pola”) entered into a time charter (the “Time Charter”) for the vessel “Divinegate” on 18 September 2019. The Time Charter included an exclusive jurisdiction clause which read:

The Time Charter ran its course and the “Divinegate” was eventually redelivered to Eastern. A dispute subsequently arose as to whether Pola was liable for certain sums of hire that had been left unpaid. Pola asserted that it was entitled to set off those sums because of various expenses it allegedly incurred over the lifetime of the Time Charter. Eastern’s case was that Pola had no such right of set off.

Eastern then proceeded to arrest the “Pola Devora” in Gibraltar in the belief that the vessel was beneficially owned by Pola. Pola contended that it was merely the time charterer of that vessel, and that the owner was in fact a third party (“Pola Rise OOO”). Eastern eventually released the vessel after it was supplied with a copy of the time charter between Pola and Pola Rise OOO, but Eastern did not concede that Pola was not the beneficial owner of the Pola Devora, or that it had wrongfully arrested that vessel.

Eastern subsequently commenced proceedings in England against Pola in respect of its claims for unpaid hire. In those proceedings, Pola brought a counterclaim in tort against Eastern for damages in respect of the arrest of the “Pola Devora” (which Pola alleged was wrongful). Eastern applied to strike out that counterclaim on grounds that the English courts had no jurisdiction over the matter (or, alternatively, that the English courts should decline jurisdiction in favour of the Gibraltar courts). In response, Pola argued that the claim was one falling within the scope of the exclusive jurisdiction clause in the Time Charter.

Ms Patricia Robertson QC, sitting as a Deputy High Court Judge, agreed with Pola. There were two points in Ms Robertson QC’s decision that I found relevant. The first was the learned judge’s observation that “[t]he language “in connection with” is naturally to be read as, if anything, wider than “arising under”, or variants on that phrase” (at [37]). She thus concluded that whether or not Pola’s counterclaim was one “arising out of” the Time Charter, Pola had to at least show that they were claims “in connection with” the Time Charter (at [24]).

The second point of note was the approach taken by the learned judge in determining if a pure tort claim fell within the scope of an exclusive jurisdiction clause (at [37]):

I will refer to the emphasised parts of the extract above as the “Causative Connection Test”.

On the facts before her, the learned judge concluded that Pola’s counterclaim in tort was causatively connected to the legal relationship constituted under the Time Charter (at [39]):

In my view, the learned judge’s analysis was cogently reasoned and I found the Causative Connection Test to be an attractive and practical formulation that gives effect to the Fiona Trust presumption. Indeed, the passages I have cited above clearly indicate that the learned judge had the Fiona Trust presumption in mind when she formulated the Causative Connection Test.

To persuade me that the Causative Connection Test was “not the definitive or exclusive test in determining if a non-contractual claim falls within the scope of an arbitration clause”, Mr Toh referred me to the case of Sea Master Special Maritime Enterprise v Arab Bank (Switzerland) Ltd [2022] EWHC 1953 (Comm) (“Sea Master”) in CSSC’s FA Request and at the FA Hearing. Specifically, he submitted that there were at least two other approaches to the question:

I had no difficulty in accepting that the Causative Connection Test was not the only approach to determining if a tort claim falls within the scope of an arbitration agreement. However, I did not understand Mr Toh to be suggesting that it was incorrect or improper to apply the Causative Connection Test – that would have been a somewhat surprising submission to make given that CSSC itself initially cited The Pola Devora to persuade me that there was a sufficiently causative connection between OKI’s claim in the Indonesian Proceedings and the parties’ legal relationship under the B/Ls. Indeed, I was not pointed to anything in Sea Master or any other case that challenged the validity or correctness of the Causative Connection Test.

Turning to the alternative approaches Mr Toh referred to, they were, in my view, either inapplicable or unworkable vis-à-vis OKI’s claim. The Parallel Claims Test contemplates a situation where an event (or series of events) gives rise to concurrent liability in contract and tort. To the extent that the tort claim mirrors the contractual claim, the tort claim must, in substance, be treated as one arising “out of or in connection with” the contract. On these facts, there was no suggestion that the Incident gave rise to concurrent liability in tort and contract on the part of CSSC, such that OKI had a parallel claim in contract. Accordingly, the Parallel Claims Test was inapplicable here.

As for the Closely Knitted Test, it considers if the contractual and non-contractual claims arose out of – or were “closely knitted together” on – the same facts so that “the agreement to arbitrate on one could properly be construed as covering the other”: The Playa Larga at 183. Although it does not require any concurrence of contractual and non-contractual liability, the Closely Knitted Test is similar to the Parallel Claims Test insofar as they both consider the extent to which the non-contractual claim may be recast as a contractual claim.

In The Playa Larga, the claimant-purchasers (“Iansa”) entered into a contract for the sale of sugar with the defendant-sellers (“Cubazucar”). Iansa paid for but did not receive the full amount of sugar it contracted for, and so a claim in the tort of conversion was brought by Iansa against Cubazucar. The contract of sale provided for “[a]ll disputes arising out of [the] contract” to be settled by arbitration. The question arose as to whether Iansa’s claim in conversion was a dispute “arising out of” the contract.

At first instance, Mustill J answered that question in the affirmative (The Playa Larga at 182):

Mustill J’s reasoning and conclusion were upheld on appeal by Ackner LJ: The Playa Larga at 183.

I was also mindful of the court’s observation in The Pola Devora (at [46]) that the Closely Knitted Test is stricter than the Causative Connection Test insofar as the former requires a near-total overlap of the facts said to ground the contractual and non-contractual claims. This observation was repeated in Sea Master (at [111]).

In Sea Master, the question arose as to whether the defendant-shipowner’s counterclaim for “reasonable remuneration and quantum meruit” could properly be characterised as “arising out of or in connection with” certain switch bills of lading (at [103]). That question was first answered by the arbitral tribunal in the affirmative (at [111]):

These conclusions were upheld by Pickens J in the High Court (at [112]–[116]).

Returning to the facts of the present case, it was clear to me that OKI’s tort claim could not be recast as a claim for breach of the contract of carriage evidenced by the B/Ls. For that reason, I was unable to see how the analysis could be advanced by resorting to the Parallel Claims Test or the Closely Knitted Test. On the other hand, I remained persuaded that the Causative Connection Test offered an appropriate means for determining – at least, on a presumptive basis – if OKI’s tort claim was one “arising out of or in connection with” the B/Ls.

The relevance of contractual defences to the tort claim

I turn now to address the significance of CSSC’s putative contractual defences. Mr Toh submitted that in determining whether a dispute arose “out of or in connection with” a contract, the court must consider not only the principal claim but also the contractual defences that were raised (or would foreseeably be raised) against that claim, if any.

In deciding if court proceedings have been brought by a party in breach of an arbitration agreement, the analysis proceeds in two steps:

I agreed that the contractual terms said by CSSC to provide it with a defence to the principal claim(s) brought by OKI in the Indonesian Proceedings were matters that had to be considered for the purposes of the analysis. In Mozambique, Lord Hodge observed that in considering the substance of the dispute, regard must also be had to defences that were raised (or would foreseeably be raised) (at [73]):

However, to appreciate the true significance of an asserted defence to the determination of whether court proceedings have been commenced in breach of an arbitration agreement, it was necessary in my view to distinguish between two distinct lines of inquiry that were thrown up as a result of the interplay between those defences, the principal claim(s), and the arbitration agreement:

In answering the first question, I began by recalling the foundational question that guided the overall inquiry: on a true construction of the Arbitration Agreement, was OKI’s tort claim in the nature of a claim that the parties intended to resolve by arbitration?

As I noted at [75] above, the Court is entitled to consider the full contractual matrix – which includes any contractual terms said to provide a defence – in construing the Arbitration Agreement. I also noted at [54] above that regard must be had to the context in which the contract (here, as evidenced by the B/Ls) was entered into. That would simply be an application of the trite principle that an arbitration agreement must, like any other contractual term, be interpreted having regard to the document as a whole and the relevant context: Zurich Insurance (Singapore) Pte Ltd v B Gold Interior Design & Construction Pte Ltd [2008] 3 SLR(R) 1029 at [131].

The surrounding terms of the contract may inferentially clarify if the parties intended that a particular type of claim should be resolved by arbitration. If, for example, the terms relied on by CSSC clearly exclude CSSC’s liability to OKI for damage caused to the Trestle Bridge in the performance of the contract of carriage, that would clearly show that the parties contemplated the possibility of such incidents occurring and that they wished to contractually allocate that risk in a particular way. On that basis, the Court may thus conclude that pure tort claims for damage caused to the Trestle Bridge are in the nature of claims that the parties intended to settle by arbitration. That would suffice to displace any contrary indications flowing from an application of the Causative Connection Test.

If, on the other hand, the relevant contractual terms shed no light on whether claims for allision damage fall within the scope of the Arbitration Agreement – or indeed, point away from that conclusion – then the Court may well infer that the principal tort claim is not in the nature of a claim that the parties intended to arbitrate upon. In my judgment, that was in substance the approach taken in Aggeliki Charis Compania Maritima SA v Pagnan SpA (The “Angelic Grace”) [1995] 1 Lloyd’s Rep. 87 (“The Angelic Grace”), a case which Mr Toh relied on extensively in argument.

In The Angelic Grace, the claimant-shipowners voyage chartered the “Angelic Grace” to the defendant-charterers. The charterers nominated a discharge port in Chioggia, Italy and thus became entitled under a special “Chioggia Lightening Clause” in the voyage charterparty to use Chioggia roads for cargo lightering operations. The charterers subsequently called for discharge of the cargo carried by the vessel into the “Clodia”, which was an unpowered open “floating elevator” which they owned. During the cargo discharge operations, the weather deteriorated and the master of the “Angelic Grace” decided to move the vessel. In carrying out that manoeuvre, the “Angelic Grace” collided with the “Clodia”. The charterers subsequently commenced proceedings against the owners in Venice for the damage suffered by the “Clodia”.

The voyage charterparty provided for all disputes “arising out of [the] contract” to be resolved by arbitration in London. The owners therefore commenced arbitration against the charterers in London. In the arbitration, the owners brought various claims against the charterers, which included contractual claims (ie, for breach of a safe anchorage warranty, or an implied term that the charterers would take reasonable care not to damage the “Angelic Grace” during lightering operations) as well as a claim in tort (ie, that the charterers were at fault for the collision, which also resulted in damage to the “Angelic Grace”). The charterers filed a substantive defence in the arbitration, albeit under protest against the tribunal’s jurisdiction.

Separately, the owners applied to the English High Court for an anti-suit injunction to restrain the charterers from maintaining the Venetian proceedings. At first instance, Rix J applied the Closely Knitted Test (which I discussed at [67] above) and decided that the charterers’ tort claim in the Venetian proceedings was a matter within the scope of the arbitration agreement (see Aggeliki Charis Compania Maritima SA v Pagnan SpA (The “Angelic Grace”) [1994] 1 Lloyd’s Rep. 168 at 174):

Rix J’s decision was upheld by the Court of Appeal. Leggatt LJ observed (The Angelic Grace at 91) that:

In my judgment, Rix J and Leggatt LJ were plainly influenced by the fact that there were special terms in the voyage charterparty to indicate that the parties contemplated the possibility of collision claims and wished to allocate that risk by contract. Importantly, it is implicit in Leggatt LJ’s decision that the mere assertion of a contractual defence or cross-claim was not ipso facto sufficient to warrant the conclusion that the charterers’ tort claim fell within the scope of the arbitration agreement. The court had to consider the substance of the claims and defences in light of all the facts to ascertain if the dispute was properly referrable to arbitration. That was precisely the approach endorsed by the UK Supreme Court in Mozambique, in which Lord Hodge commented that the court is “not tied to the pleadings but should look to the substance of the claims and likely defences” (at [85]).

I turn now to the second line of inquiry, ie, whether the defence itself is a matter that falls within the scope of the arbitration agreement. It should be apparent from my analysis thus far that this question cannot be answered by examining the relevant contractual terms in vacuo. The significance of those terms can only be deduced by considering the words used in relation to the facts said to justify their invocation.

The case of Mozambique bears out this principle. The brief facts of Mozambique are these. The Republic of Mozambique (“the Republic”), acting through various state-owned entities, borrowed money from several banks under various loan agreements. Those loans were secured by sovereign guarantees (the “Guarantees”), which guarantees were governed by English law and provided for the exclusive jurisdiction of the English courts. The loans were taken out to finance the Republic’s purchase of equipment and services under various contracts with several of the defendants (the “Supply Contracts”). The Supply Contracts were governed by Swiss law and contained arbitration agreements providing for arbitration in Geneva, Switzerland.

Proceedings were subsequently brought in the English courts by the Republic, who claimed that it was the victim of a complex conspiracy perpetrated by the defendants. It was alleged that as part of the conspiracy, large bribes were paid by certain defendants to corrupt Mozambican officials. The Republic claimed that as a result of the conspiracy, it was exposed to potential liabilities amounting to approximately US$2 billion under the Guarantees (in addition to having suffered other macro-economic losses).

In response, a group of defendants (collectively referred to as “Privinvest”) applied under s 9 of the UK Arbitration Act 1996 (the “UK AA”) for those proceedings to be stayed. Privinvest argued that the Republic’s claims fell within the scope of arbitration agreements in the Supply Contracts.

Lord Hodge undertook an extensive review of the authorities and concluded that the Republic’s claims were not matters within the scope of the arbitration agreement. I have referred to some of the principles relied on by his Lordship in reaching that conclusion (at [74]–[75] above), and it is not necessary to say more on that part of the judgment for present purposes.

Here, I focus on Privinvest’s “partial defence on quantum”, which they argued justified a partial stay (if not a full stay) of the proceedings. Privinvest disputed the Republic’s quantification of its claims and asserted that “it provided valuable goods and services and that the Republic has squandered them and sabotaged the project for reasons of internal politics” (at [96]). This defence, Privinvest argued, was a matter so closely connected to the Supply Contracts that it fell within the scope of the arbitration agreements contained therein, such that the English proceedings should be stayed fully (or, in the alternative, to the extent they related to the defence) (at [101]).

This argument was rejected by Lord Hodge. The Republic’s principal claims were found not to fall within the scope of the arbitration agreement, and Lord Hodge noted that it was against those claims that Privinvest was raising its “partial defence on quantum” (at [106]). Having regard to the subsidiary nature of that defence, Lord Hodge concluded that Privinvest’s dispute on quantum was not sufficiently connected to the Supply Contracts to warrant a stay of any kind (at [107]):

As Lord Hodge’s judgment demonstrates, it does not follow from the contractual nature of an alleged defence/cross-claim that the defence/cross-claim is therefore a matter within the ambit of the arbitration agreement. It bears repeating that the touchstone is whether the parties objectively intended for such a matter to be settled by arbitration, having regard to the surrounding facts and contractual terms considered in the round.

To conclude, the identification of “matters” and where they stand vis-à-vis an arbitration agreement have been framed as a two-stage test that proceeds linearly (see [74] above). At the second stage, the relevant matters – which include defences, whether raised or which may foreseeably be raised – must be considered (a) alongside each other; and (b) against the backdrop of all the relevant facts and contractual terms. Otherwise, much like the proverbial elephant and the blind men, any inferences drawn as to the parties’ objective intentions will necessarily be clouded to the extent that the court is looking at an incomplete picture.

The Indonesian Proceedings had not been brought in breach of the Arbitration Agreement

With the foregoing principles and considerations in mind, I remained convinced of my earlier conclusion that on a prima facie review of CSSC’s case, the dispute forming the subject matter of the Indonesian Proceedings was not a dispute “arising out of or in connection” with the Arbitration Agreement in the B/Ls.

OKI’s claim was not causatively connected to the legal relationship constituted under the B/Ls between OKI and CSSC

As I stated at [49] above, it could not seriously be disputed that OKI’s claim had to be characterised as a stand-alone tortious claim for damage caused by CSSC to the Trestle Bridge. It was also not seriously argued that OKI had any contractual claims against CSSC under the B/Ls in respect of the Incident; there was, for example, no suggestion or evidence of any claims by OKI for loss or damage to any of the cargo shipped onboard the Vessel as a result of the Incident. In the premises, it was appropriate to apply the Causative Connection Test to determine if OKI’s tort claim was prima facie a matter within the scope of the Arbitration Agreement.

It was not disputed that under the B/Ls, CSSC stood as carrier and shipowner vis-à-vis OKI as shipper. This relationship, which I will refer to as the “carrier-shipper relationship”, is the quintessence of bills of lading and the contracts of carriage evidenced therein.

To succeed at this stage of the inquiry, it was therefore incumbent on CSSC to show that:

CSSC did not attempt to characterise OKI’s tort claim as being causatively connected to the carrier-shipper relationship. The pertinent question, therefore, was whether CSSC could succeed on the second limb.

(1) No shipowner/jetty-owner relationship was constituted under the B/Ls

CSSC submitted at the FA Hearing that the terms of the B/Ls (including the terms of the Head COA incorporated into the B/Ls) constituted a relationship between CSSC qua carrier/shipowner and OKI qua owner of the Jetty (a reference that was presumably intended to encompass the damaged Trestle Bridge). I will refer to this as the “shipowner/jetty-owner relationship”. To make good this argument, CSSC emphasised two points:

In my judgment, neither point lent any credence to the assertion that a shipowner/jetty-owner relationship had been constituted under the B/Ls. The Load Port Clause reads:

The words “Oki Sea Port” were, in my view, only used for convenience in designating or describing the Terminal as the loading port. I could not agree with the suggestion that the clause “recognised” OKI as the owner of the Trestle Bridge, Jetty, or Terminal. Even if I assumed that it “recognised” OKI as the owner of those properties, I was not persuaded that any legal significance could or should be attached to it.

Turning to Mr Toh’s second point, it is a fact that the allision occurred at a time when CSSC was performing its obligations under its contract of carriage with OKI. Mr Toh suggested that the overlap in time was significant because the parties intended to resolve by arbitration any claims by OKI against CSSC arising in the course of CSSC performing its obligations under the B/Ls. However, I found that to be a question-begging response because it assumed that the overlap in time ipso facto created a sufficient connection between the claim and the parties’ contract. In my judgment, something more was required than a mere temporal overlap in events.

Overall, there was nothing in the B/Ls or the associated contracts of affreightment to suggest that the parties intended to accord weight to the fact that OKI was the owner of the Terminal, Jetty, or Trestle Bridge. It is on this point that the present case parted ways with the facts of The Angelic Grace. As I noted at [85] above, the voyage charterparty in The Angelic Grace specially provided for claims resulting from collisions arising in the course of lightering operations. The natural conclusion, therefore, was that the charterparty constituted a legally significant relationship between the claimant qua owner of the “Angelic Grace” and the defendant qua owner of the “Clodia” (which was deployed for the lightering operation).

(2) No shipowner/mill-owner relationship was constituted under the B/Ls

CSSC further submitted that, in addition to claiming under the B/Ls as owner of the Jetty, OKI was also claiming as owner of the Mill. I will refer to this as the “shipowner/mill-owner relationship”.

As a preliminary point, I was mindful that nothing turned on whether the shipowner/mill-owner relationship existed. Even if I agreed that such a relationship was constituted under the B/Ls, it was not suggested that OKI’s claim could be described as causatively connected thereto. The thrust of CSSC’s argument was that the existence of this relationship would reinforce its contention that there also existed the shipowner/jetty-owner relationship.

To demonstrate the existence of the shipowner/mill-owner relationship, Mr Toh advanced three main arguments. He first referred me to three terms in the Head COA that were allegedly incorporated into the B/Ls. These terms, it was argued, indicate that under the B/Ls, OKI was intended to stand qua owner of the Mill vis-à-vis CSSC.

The first of those terms was cl 40, which relates to cargo nominations. The relevant parts of cl 40 read:

Mr Toh submitted that this term “envisages a series of shipments between the parties taking into account OKI’s capacity as the mill owner and its ability to ship certain quantities of cargo over a period of time”. I could not agree with this submission. Clause 40 was an ordinary cargo nomination clause and nothing more. It required the charterer (which, under the Head COA, was COSCO Europe) to provide estimates and/or notice of cargo quantities based on information that was available to it. Even on the assumption that OKI was the party obliged to provide those estimates/notices to CSSC, there was still nothing in that clause to suggest that OKI was the owner of the Mill, or that the parties were contemplating OKI’s productive capacities qua Mill-owner.

The second contractual term Mr Toh referred me to was cl 51, which is a force majeure clause. The relevant parts of cl 51 read:

Clause 51 only makes a bare and passing reference to a “mill”. There was nothing in the clause to even associate OKI with it, or to indicate that it was the Mill that was being referred to.

The third contractual term Mr Toh referred me to was cl 66, part of which excludes the owner’s liability for certain types of damages. That part reads:

It was apparent to me that cl 66 is a generically worded exclusion clause that does not even refer to the Mill. There was simply nothing in cl 66 that spoke to the legal relationship between OKI and CSSC. Mr Toh submitted that this clause would “conceivably exclude CSSC’s liability for OKI’s claims (qua jetty owner or mill owner) for business interruption”. Whether cl 66 had that effect was irrelevant because I was not concerned with the substantive merits of CSSC’s putative defences.

I turn now to address Mr Toh’s second main argument. Mr Toh referred me to the heads of loss claimed by OKI in the Indonesian Proceedings, which he said indicated that OKI was purporting to advance the claim qua Mill-owner and Jetty-owner.

In an e-mail dated 10 June 2022 sent by Clasis to CSSC’s solicitors, Messrs Rajah & Tann Singapore LLP (“Rajah & Tann”), Clasis summarised the heads of loss OKI was claiming for in the Indonesian Proceedings. Those heads were described as:

Whether OKI will in fact be entitled to recover all those losses in the event it establishes its principal claim was not a question that I was concerned with. However, it was clear to me that the scope of OKI’s recovery – specifically, its entitlement to recover for consequential losses – had to be a result that flows from OKI’s pleaded claim. Mr Toh’s submission flipped that remedial logic on its head by asserting that OKI’s pleaded losses determine the capacities in which OKI was claiming against CSSC. In my view, that approach was wrong and for that reason, I was of the view that Mr Toh’s submission was without merit.

The third main argument advanced by Mr Toh was, in effect, an appeal to the commercial context of the parties’ dealings. Specifically, it was contended that “the relationship envisaged by the Head COA and Sub COA was a long-term logistical arrangement” that “contemplated OKI’s various capacities as cargo owner, mill owner (from which the cargo was manufactured) and jetty owner (from which the cargo was shipped)”. Mr Toh submitted that the Head COA and Sub-COA were “elaborate, lengthy contractual documents with provisions that were customized with this long-term logistical arrangement in mind”. Mr Toh further submitted that the parties’ use of the standard CONGENBILL form could not detract from that reality.

This argument was neither here nor there. Even if I assumed Mr Toh’s submissions accurately reflected the nature of the commercial relationship between OKI and CSSC, I was concerned to deduce the legal relationship between the parties under the contract of carriage they had entered into (and within which the Arbitration Agreement was contained). The context was an undoubtedly important factor but ultimately, the terms of the contract had to speak for themselves. For the reasons I have already given, there was nothing in the B/Ls and the associated contracts of affreightment to show that any relationship other than the carrier/shipper relationship was formed thereunder.

For all the foregoing reasons, there was, in my judgment, no basis to conclude that OKI’s tort claim was “causatively connected” in any meaningful sense to any legal relationship established under the B/Ls between CSSC and OKI. The presumptive inference, therefore, was that the parties could not have intended for OKI’s tort claim to be covered by the Arbitration Agreement.

CSSC’s putative defences did not reveal an intention that a claim of the sort brought by OKI should be resolved by arbitration

The next question was whether my prima facie conclusion at [115] above could be displaced by CSSC’s reliance on its putative contractual defences. In this regard, Mr Toh brought to my attention four terms in the Head COA that, according to him, offered CSSC contractual defences to OKI’s claim. I address each of them in turn.

(1) OKI’s alleged breach of the Safe Port Warranty

Box 8 of Appendix 1 read with cl 6 of the Head COA allegedly required OKI to nominate “1 Safe Berth / Safe Anchorage, 1 Safe Port”. I will refer to this as the “Safe Port Warranty”. CSSC argued that the Terminal was not a safe port and CSSC incurred losses or liabilities because of this alleged breach by OKI.

At the FA Hearing, I enquired of Mr Toh if OKI was even under an obligation to CSSC to nominate a safe port given that (a) the relevant terms were in the Head COA (which was only binding between CSSC and COSCO Europe); and (b) the Sub-COA interposed between CSSC and OKI. Mr Toh argued that those terms in the Head COA were incorporated into the B/Ls and thus, the reference therein to “Charterers” should be understood, by means of verbal manipulation, as a reference to OKI. In the alternative, it was submitted that the obligation could, in any event, be implied into the B/Ls.

Leaving aside this difficulty, there was in my view nothing special about the Safe Port Warranty to suggest that the parties contemplated an allision between the Vessel and any property forming part of the Terminal. It might well be that certain features rendering the Terminal unsafe – and which therefore speak to a breach of the Safe Port Warranty – also caused or contributed to the Incident (and to be clear, I am only hypothesising for the sake of argument). However, that factual overlap was not a factor that shed any light on the parties’ objective intentions as regards possible allision claims.

(2) The exclusion of liability under cl 66 of the Head COA

Mr Toh submitted that “[o]n quantum, CSSC could also rely on Clause 66 of the Head COA and/or Sub COA” as incorporated into the B/Ls. The language of cl 66 was set out at [109] above. I also observed that cl 66 was a generically worded exclusion clause that says little apart from purporting to exclude CSSC’s liability for certain heads of damages. As with the Safe Port Warranty, there was simply nothing special about cl 66 to suggest that liability for damage caused by an allision was within the parties’ contemplation at the time the contract of carriage was concluded. This, again, was to be contrasted with The Angelic Grace, where the salient terms of the charterparty catered specifically for collision liability.

(3) The defence of negligent navigation

I will address the third and fourth defences raised by CSSC together because they are identical in substance. Both defences flow from cl 63 of the Head COA, the relevant parts of which read:

Clause 63 performs the dual function of:

Both defences are identical in substance, and I will refer to them interchangeably as the “negligent navigation defence”.

As a matter of ordinary practice, the negligent navigation defence (also sometimes referred to as the “nautical fault exception”) is commonly raised against claims for or as a result of damage to cargo or the ship. In my view, rational commercial parties in a shipper/carrier relationship that (a) contemplated the risk of allision damage to property other than the ship or cargo laden on board it and (b) intended to contractually allocate that risk would not (c) be content to rely on the negligent navigation defence to achieve their ends. The express exception in cl 63 does not, in my view, communicate such an objective intention. The Hague-Visby Rules generally regulates the rights and obligations as between carriers and cargo interests. Even if it turns out that Art IV r 2(a) or the express exception in cl 63 can be profitably raised against OKI’s claim for damage to the Trestle Bridge – and here, I express no view on the matter – the point remains that the language of neither contractual defence supported the inference that OKI’s claim was one that, on an objective analysis, the parties contemplated and intended to settle pursuant to the Arbitration Agreement.

Whether CSSC’s putative defences were themselves matters falling within the scope of the Arbitration Agreement

So far, I have only considered the extent to which CSSC’s putative defences indicate that OKI’s claim was a matter that was intended to be resolved by arbitration. There is still the residual question of whether the defences were themselves matters falling within the scope of the Arbitration Agreement.

As the case of Mozambique suggests (see [92]–[93] above), the court’s decision to stay proceedings depends on whether the defences are (a) mere “subordinate” issues that the parties could not have intended to hive off to arbitration; or (b) substantive issues that the parties intended should be resolved by arbitration, which may in turn suggest that the entirety of the dispute should be stayed in favour of arbitration. Parity of reasoning dictates that the same question arises where applications for contractual anti-suit injunctions are concerned.

As I explained at [93] above, this inquiry had to be undertaken bearing in mind the nature of OKI’s principal claim in the Indonesian Proceedings. For reasons already given, I was not persuaded that the parties objectively intended for OKI’s claim to be settled by arbitration. With that in mind, I was of the view that CSSC’s asserted defences under cll 63 and 66 – which are true defences inasmuch as they purport to exclude or limit liability – are subordinate issues that the parties could not have intended to refer to arbitration separate from the resolution of the principal claim.

The Safe Port Warranty stands on a different footing. I was prepared to assume that whether the Safe Port Warranty had been breached was a contractual dispute that the Arbitration Agreement was intended to cover, even if the Safe Port Warranty was not a factor supporting the inference that OKI’s claim should similarly be resolved by arbitration. On that hypothesis, arbitration in Singapore would be the appropriate method for resolving any claims for breaches of the Safe Port Warranty in the event CSSC decides to pursue them.

I was mindful of the principle that contracting parties generally intend for all their disputes to be resolved by a single tribunal. However, the Fiona Trust presumption is but one of several interpretive devices that the Court may use to achieve its primary goal, which is to deduce and give effect to the parties’ objective intentions (see [54] above). The evidence before me indicated that OKI’s claim was not one that the parties objectively intended to refer to arbitration, and I did not think that collateral claims by CSSC relating to the Safe Port Warranty were weighty enough to displace that conclusion. In my judgment, it would be consistent with the parties’ objective intentions for OKI’s tort claim to be decided in the Indonesian Proceedings and CSSC’s cross-claim for breach of the Safe Port Warranty to be decided by arbitration in Singapore.

Conclusion

To conclude this section, I was not satisfied that OKI’s tort claim was causatively connected in any meaningful way to the legal relationship created by the B/Ls. The presumptive inference, therefore, was that the parties did not intend for claims of that sort to be resolved pursuant to the Arbitration Agreement. Having considered the defences CSSC said it intended to raise against OKI’s claim, I was unconvinced that there was anything in those defences to displace my prima facie conclusion.

I was also not convinced that those defences – with the exception of the alleged breach of the Safe Port Warranty – were themselves matters within the scope of the Arbitration Agreement. As I mentioned above at [127], I was prepared to assume that the alleged breach of the Safe Port Warranty was a matter properly referrable to arbitration. However, that was not a sufficient basis upon which I could conclude that the Arbitration Agreement was intended to provide for the resolution of OKI’s principal claims along with all the alleged defences CSSC was seeking to raise against it. This was an occasion where the Fiona Trust presumption had to give way to the facts, which demonstrated that certain matters were not intended to be resolved by arbitration. OKI’s tort claim belonged to that class of matters.

For these reasons, I remained of the view that there was no breach by OKI of the Arbitration Agreement that needed to be restrained by the grant of a contractual anti-suit injunction.

Ground 2: No exclusive jurisdiction agreement had been concluded between the parties

CSSC argued in the alternative that the Indonesian Proceedings had been brought in breach of an alleged exclusive jurisdiction agreement in favour of Singapore. Mr Toh submitted that the putative exclusive jurisdiction agreement had been concluded in a series of e-mails between Rajah & Tann, CSSC’s Protection & Indemnity Club (“West of England P&I”), and Clasis. These e-mails were exchanged shortly after the Incident occurred and related to negotiations on security for OKI’s claims arising therefrom.

By an e-mail dated 1 June 2022, CSSC sent a draft letter of undertaking (“LOU”) in West of England P&I’s standard form to OKI. That draft provided for disputes thereunder to be referred to the High Court of England and Wales.

By an e-mail dated 2 June 2022, Clasis sent an amended draft of the LOU to West of England P&I. Relevant for present purposes is the following statement in Clasis’ e-mail:

West of England P&I responded to Clasis’ e-mail on the same day to say that:

On 20 June 2022, Rajah & Tann wrote to Clasis to “accept” the latter’s proposal that the dispute be determined by the Singapore courts. The relevant parts of Rajah & Tann’s e-mail reads:

Clasis then responded to Rajah & Tann’s e-mail the next day (ie, on 21 June 2022) refuting the existence of any agreement as to jurisdiction:

On the evidence as laid out above, I was not persuaded that any jurisdiction agreement had been concluded between CSSC and OKI, let alone one agreeing to the exclusive jurisdiction of the Singapore High Court. An offer is “an expression of willingness to contract on specified terms made with the intention that it is to become binding as soon as it is accepted by the person to whom it is addressed”: Chitty on Contracts (Hugh G. Beale gen ed) (Sweet & Maxwell, 34th Ed, 2022) at para 4-003. When Clasis “propose[d] that parties agree that the matter be determined by the Singapore Courts” in its e-mail dated 2 June 2022, that proposal was made by way of certain amendments to the draft LOU attached to that email. Clasis explicitly informed West of England P&I that the revised draft was “subject to [OKI’s] final approval”. It was not suggested that OKI ever agreed to that draft generally, or to the revised dispute resolution mechanism specifically. Thus, there was simply no offer by OKI to enter into an exclusive jurisdiction agreement that was capable of acceptance by CSSC – at best, Clasis’ statements amounted to what the law would consider an invitation to treat. That being the case, it was impossible to find that any exclusive jurisdiction agreement had been concluded between the parties. Accordingly, CSSC could not succeed on Ground 2.

Ground 3: There was no basis for granting a non-contractual anti-suit injunction

CSSC submitted in the alternative that a non-contractual anti-suit injunction should be granted to restrain OKI from pursuing its claims in the Indonesian Proceedings. Those proceedings, according to CSSC, were vexatious and oppressive.

As I explained at [29]–[31] above, whether the Indonesian Proceedings were oppressive and vexatious was but one factor (albeit an important factor) that CSSC had to establish if I were to grant the non-contractual anti-suit injunction sought. I also had to be satisfied that OKI was amenable to the jurisdiction of the Singapore courts; that Singapore was the natural forum for the parties’ dispute; and that no injustice was likely to result if I granted (or declined to grant) the anti-suit injunction.

Having considered the facts and arguments in light of these requirements, I was not minded to allow CSSC’s prayer for a non-contractual anti-suit injunction. I explain why below.

OKI was amenable to the jurisdiction of the Singapore courts

The first factor I had to consider was whether OKI was amenable to the jurisdiction of the Singapore courts. As the Court of Appeal held in Koh Kay Yew v Inno-Pacific Holdings Ltd [1997] 2 SLR(R) 148 (“Koh Kay Yew”) (at [17]):

In my judgment, OKI was amenable to the jurisdiction of the Singapore courts. I have already set out the relevant procedural background (see [13]–[22]) above) and would highlight the following points as material to the question at hand:

In the circumstances, I was satisfied that OKI was amenable to the Court’s jurisdiction. It was irrelevant that OKI had filed its NIC only with a view to challenging the Court’s jurisdiction. As the passage from Koh Kay Yew makes clear, amenability to jurisdiction can be established through the proper service of documents. In this case, OKI was properly served.

Singapore was not the natural forum for the dispute

It is settled law that the natural forum for determining a tort claim is prima facie the place where the tort occurred: Rickshaw Investments Ltd and another v Nicolai Baron von Uexkull [2007] 1 SLR(R) 377 at [37]. The rationale for this rule is that the law of the place where the tort occurred (or the lex loci delicti) will ordinarily be the law that governs the resulting tort claim(s).

However, this is only a prima facie position and the court must look to any other substantial features of the case to determine if the presumption should be displaced – including if it was purely fortuitous that the tort should have occurred where it did: JIO Minerals FZC v Mineral Enterprises Ltd [2011] 1 SLR 391 (“JIO Minerals”) at [106].

In this case, the Incident occurred in Indonesia and so Indonesia was the presumptive natural forum. That the tort occurred in Indonesia could not be regarded as a result of mere fortuity. The B/Ls and the contracts of affreightment designated the Terminal (which is located in Indonesia) as the loading port. The damaged property (ie, the Trestle Bridge) is likewise situated in Indonesia. As Mr Toh himself emphasised, the Incident occurred “in the course of [CSSC] performing its obligation under the contract of carriage” (ie, to carry the Cargo from Indonesia to China and South Korea).

CSSC advanced the interesting argument that in this case, Singapore (and not Indonesia) should nevertheless be regarded as the natural forum because certain forum mandatory statutes under Singapore law apply to the dispute. Specifically, it was contended that the Carriage of Goods by Sea Act 1972 (2020 Rev Ed) (the “COGSA”) and the MSA “have the force of law in Singapore” and “[a]ccordingly, these provisions apply mandatorily such that Singapore (instead of Indonesia) must be regarded as the natural forum for the dispute”.

I seriously doubted the contention that the COGSA and the MSA (specifically Part 8 thereof, which provides for a shipowner’s right to limit its liability for certain claims by reference to the tonnage of the vessel concerned) are forum mandatory statutes properly so-called. It was noted in JIO Minerals that forum mandatory rules serve a “regulating or protective function”, as opposed to a merely “supplementary or facilitative” function (at [104]). In Goldilocks Investment Co Ltd v Noble Group Ltd [2018] 5 SLR 425, Aedit Abdullah J considered that the Securities and Futures Act 2001 “could, given its regulatory nature, possibly operate as a forum mandatory statute, as with penal, revenue, and other public laws” (at [13]).

In contrast, the COGSA and the MSA are generally concerned with the balance of civil rights and obligations and thus lack the public flavour that the cases seem to require of forum mandatory statutes.

Even if I assumed that the COGSA and the MSA are forum mandatory statutes, there was in my view a more fundamental difficulty with Mr Toh’s argument. In JIO Minerals, Andrew Phang JA (as he then was) observed that the applicability of a forum mandatory rule may be a relevant factor in deciding if Singapore is the natural forum for a dispute. However, it is not the existence of the forum mandatory rule per se that is significant. Instead, the court must consider the substantive effect that an application (or non-application) of that rule is likely to have on the proceedings and the parties’ rights. In JIO Minerals, the point was expressed as follows in relation to the Misrepresentation Act (Cap 390, 1994 Rev Ed) (at [97]):

Here, there was no indication that the COGSA was relevant to the parties’ dispute. Proceedings in respect of that dispute were pending in Indonesia. The B/Ls are expressly governed by English law and were issued in respect of cargo shipped from Indonesia to China and South Korea. The parties (ie, OKI and CSSC) are domiciled in Indonesia and China, and COSCO Europe in the Netherlands. Singapore was only referred to in the Arbitration Agreement, which provides for an SIAC-administered arbitration in Singapore.

The B/Ls contain a clause paramount that would, pursuant to s 3(4) COGSA, trigger the operation of the COGSA if the dispute were before the Singapore courts. The part of cl 63 constituting the clause paramount reads:

Section 3(4) COGSA provides:

Importantly, however, the dispute was not before the Singapore courts. Mr Toh’s submission amounted to this: if the dispute were before the Singapore courts, the COGSA would apply and therefore, CSSC should not be deprived of the procedural or substantive advantages it would otherwise enjoy thereunder. In my view, that line of argument was entirely circular. It begged the question by assuming a prior basis for the operation of the COGSA when there was in fact none. CSSC could not pull itself up by its own bootstraps in that manner.

In addition, I was also mindful that if Mr Toh’s argument was correct, it would have led to the somewhat absurd conclusion that Singapore would be the natural forum for every dispute arising out of any bill of lading expressed as subject to the Hague-Visby Rules, irrespective of the parties’ express choice of law and jurisdiction.

Turning to the MSA, I start with the observation that the Indonesian Proceedings and CSSC’s limitation action in ADM 50 are separate proceedings in two different fora. The relevant question at hand was whether Singapore was the natural forum for the dispute forming the subject matter of the Indonesian Proceedings. As the case of Evergreen International SA v Volkswagen Group Singapore Pte Ltd and others [2004] 2 SLR(R) 457 (“Evergreen”) plainly demonstrates, it is the features of the principal dispute that are germane to the natural forum inquiry.

In Evergreen, a collision occurred in Singapore’s territorial waters between the “Ever Glory” (which was owned by the plaintiffs) and the “Hual Trinita”. The plaintiffs proceeded to obtain a limitation decree and constitute a limitation fund in Singapore. The defendants (who were the owners or insurers of the cargo carried on the “Hual Trinita”) boycotted the limitation action; they went on to arrest the “Ever Reach” (which was a sister vessel of the “Ever Glory”) in Antwerp and commenced an action in tort against the plaintiffs in Belgium. The plaintiffs applied to the Singapore High Court for an anti-suit injunction to restrain the defendants from pursuing the Belgian action.

In concluding that there was an “overwhelming case” that Singapore was the natural forum for the parties’ dispute (at [31]), Belinda Ang J (as she then was) noted (at [28]) that:

In Evergreen, Singapore was clearly the centre of gravity for the parties’ principal dispute and therefore had the strongest claim to being the natural forum. That the plaintiffs had accrued limitation rights in Singapore was a factor relevant only for the purposes of deciding if the Belgian proceedings were oppressive or vexatious, and whether the defendants would be deprived of any legitimate juridical advantages if the anti-suit injunction was granted.

On that basis, I saw no merit in CSSC’s argument that the MSA should be taken into account in identifying the natural forum for the parties’ dispute. As far as the natural forum inquiry goes, the relevant facts of the present case (see [147] above) are similar to those of Evergreen, albeit that they weigh in favour of diametrically opposed outcomes. In this case, the centre of gravity for the parties’ dispute was Indonesia (being the place where the tort occurred). Indonesia had to be regarded as the natural forum for the dispute.

The Indonesian Proceedings were neither vexatious nor oppressive

My conclusion that Indonesia (and not Singapore) was the natural forum for the parties’ dispute was sufficient to dispose of CSSC’s prayer for a non-contractual anti-suit injunction. Nonetheless, for completeness, I will briefly set out CSSC’s submissions in respect of the other requirements and my conclusions on them.

Even if I was wrong and Singapore was the natural forum for the parties’ dispute, it remained for CSSC to demonstrate that it was vexatious and oppressive for OKI to pursue its claim in Indonesia. There is no forensic test for determining if foreign proceedings are vexatious or oppressive, although the decided cases lay down various considerations that the court should have regard to. A useful, non-exhaustive summary of these considerations was set out in VEW v VEV [2022] 2 SLR 380 (“VEW”) (at [44]):

In my oral judgment dismissing SUM 2676, I expressed the view that if the Indonesian Proceedings had not been commenced in breach of the Arbitration Agreement, it would follow that OKI was entitled to do so (so far as the contract of carriage was concerned). The assertion that the Indonesian Proceedings were brought by OKI to oppress or vex CSSC would be weakened to that extent. CSSC did not challenge that specific conclusion in its FA Request or at the FA Hearing.

CSSC’s main argument was that the Indonesian Proceedings represented a collateral attack on its right to select the forum for limitation of liability and avail itself of its tonnage limitation rights in Singapore. It was in that context that Mr Toh relied heavily on the case of Evergreen.

As would have been apparent from my analysis of Evergreen (at [157]–[158] above), CSSC stood in a very different position from the injunction applicant in that case. I have already pointed out the first critical point of departure: Singapore was the natural forum for the parties’ dispute in Evergreen. In this connection, Ang J noted that “[o]ther than arresting the Ever Reach in Belgium, the defendants [had] no connection whatsoever with Belgium. The plaintiffs [were] a company incorporated in Panama. The operators of Ever Glory [were] based in Taiwan” (at [13]). Furthermore, the collision occurred in Singapore’s territorial waters. The same could not be said here.

Ang J’s finding that Singapore (and not Belgium) was the natural forum for the parties’ dispute in Evergreen – although a conceptually distinct step in the overall analysis – was simultaneously also a factor supporting the conclusion that the Belgian proceedings were vexatious and oppressive. It shored up the legitimacy of the plaintiffs’ choice to commence limitation proceedings in Singapore, which choice the injunction respondent had intentionally attempted to frustrate: Evergreen at [42], [50], and [55]. To that extent, the cynicism in the defendants’ conduct was readily discernible in Evergreen. I did not think that CSSC had the benefit of making such an argument in the case before me.

The second critical point of departure was this. In Evergreen, the defendants had commenced the Belgian proceedings only after the plaintiffs obtained a limitation decree and constituted a limitation fund in Singapore. That latter step was said to “[provide] the necessary focal point for consideration of the issues at hand” (at [42]). With that in mind, Ang J concluded (at [47]) that:

In this case, the Indonesian Proceedings were commenced on 26 October 2022, which was some time before the Limitation Decree was granted on 5 October 2023. At the time the Indonesian Proceedings were commenced, CSSC had no accrued limitation rights – at best, they had an expectation that those rights would accrue in due course.

Mr Toh calibrated his submissions to cater for that nuance. He emphasised that SUM 3219 was filed before the Indonesian Proceedings were instituted, and that but for OKI’s allegedly dilatory tactics in ADM 50, CSSC would in all likelihood have obtained the Limitation Decree before 26 October 2022. On that footing, Mr Toh invited me to conclude that OKI had taken deliberate steps to render CSSC’s limitation action nugatory and to coerce CSSC into litigating in Indonesia. That, it was argued, was where the vexation and oppression lay.

I was not persuaded by these arguments, which Mr Toh had in fact previously raised at the hearing of RA 197. In my earlier decision on RA 197, I concluded that there was nothing procedurally abusive about OKI’s conduct in ADM 50:

For these reasons, I did not think that CSSC’s argument could be advanced by reference to what OKI did (or did not do) in relation to ADM 50.

In the premises, the real question was whether OKI’s commencement of the Indonesian Proceedings on 26 October 2022 amounted to an illegitimate attempt at frustrating or subverting CSSC’s choice of limitation forum, contrary to the principle established in The Volvox Hollandia [1988] 2 Lloyd’s Rep. 361 (“Volvox Hollandia”).

In Evergreen, Ang J explained (at [47]) that:

In my judgment, the foregoing passage demonstrates that Ang J’s finding of oppression and vexation was one that was made on the back of very strong facts before her. Foremost on the learned judge’s mind was the fact that the limitation decree had been made, the limitation fund constituted, and liability admitted – all that remained was for the limitation fund to be distributed. Ang J was also conscious that there were other parties claiming against the limitation fund; that was not the case here. It was in light of those facts that considerations of comity had to yield to the need to protect the injunction applicant’s established rights (at [55]). In my view, Evergreen was not authority for the wide proposition advanced by CSSC that the Volvox Hollandia principle takes effect from the moment limitation proceedings are put in motion by a shipowner, such that foreign proceedings cannot henceforth be commenced against that shipowner without breaching that principle.

The case of Seismic Shipping Inc and another v Total E&P UK plc (The “Western Regent”) [2005] 2 Lloyd’s Rep. 359 offers a useful counterpoint. In that case, the first and second claimants were respectively the owner and demise charterer of a seismic survey vessel. As the vessel navigated the North Sea with six streamers in tow, one of the streamers made contact with a buoy positioned at a well head in an offshore oilfield owned by the defendant. The buoy was dragged along with the vessel, and this resulted in damage to the well head.

The claimants commenced limitation proceedings in England on 5 November 2004. The defendant, for its part, commenced proceedings against the claimants in the Southern District of Texas, Galveston Division on 24 January 2005 for losses it allegedly suffered in consequence of the incident. Under Texan law, the claimants’ liability would be limited to the vessel’s post-allision value and not the 1976 Convention limits prevailing in the UK. The claimants applied to the English High Court for an anti-suit injunction restraining the defendant from pursuing the proceedings in Texas, as well as for summary judgment against the defendant. The defendant cross-applied for a declaration that the court had no jurisdiction to entertain the claimants’ action or, in the alternative, an order dismissing the claim. On 10 February 2005, the claimants paid a sum of money into court in constitution of the limitation fund; a limitation decree was subsequently issued on 22 March 2005.

In dismissing the claimants’ application for an anti-suit injunction, Clarke LJ started from the position that “[t]he purpose of an injunction is not to ensure that an English judgment is recognised by a friendly foreign state but to prevent unconscionable conduct” (at [48]). Clarke LJ took the view that there was nothing unconscionable about the defendant’s prosecution of the Texas proceedings. In reaching this conclusion, Clarke LJ placed weight on the fact that those proceedings had been stayed by the Texas court pending the final determination of the limitation action in England (at [49]). Furthermore, judicial comity required the court to hesitate long before “[interfering], albeit indirectly, with the process of the Texas court and the right of the Texas court to decide in accordance with its own laws whether or not to recognise and enforce the English limitation decree” (at [50]). Finally, the learned judge was also guided by the fact that the Texas proceedings had been commenced before the decree was issued (at [55]):

I was not referred to any authority that purported to draw a bright line as to when foreign proceedings can no longer be legitimately commenced without infringing the Volvox Hollandia principle. In my view, that is because a finding of oppression and vexation is one that the court can only make having regard to the particular facts of each case.

Overall, I was not persuaded that OKI’s commencement of the Indonesian Proceedings at so early a stage in the Singapore limitation proceedings could be regarded as an illegitimate attack on CSSC’s right to choose its limitation forum. In reaching this conclusion, I had regard to the following features of the case:

Mr Toh also submitted that Indonesia would not recognise or enforce the Limitation Decree or any limitation fund constituted in Singapore. This, in his submission, was a juridical advantage that OKI was cynically attempting to avail itself of to CSSC’s prejudice.

Even if I took CSSC’s assertion at face value, it is for the Indonesian courts to decide what effect the Limitation Decree is to be given: The Western Regent at [50]; MSC Mediterranean Shipping Company SA v Stolt Tank Containers BV (The Flaminia) [2022] EWHC 835 (Admlty) at [103]. Even if it were true that the Indonesian courts would refuse to recognise the Singapore limitation decree, that fact was not ipso facto sufficient to establish vexation and oppression in the present case. That consideration weighed heavily in favour of granting the anti-suit injunction in Evergreen but as I noted (at [172] above), the decision was ultimately one that was made having regard to the particular circumstances of the case. Ang J explained (at [54]) that:

No injustice was likely to result from my refusal to grant the anti-suit injunction

The final factor (ie, whether injustice is likely to result if the anti-suit injunction is granted or not granted) is related to the two preceding factors. In VEW, the Court of Appeal made the following observations (at [45]):

CSSC’s submissions on this point were closely allied with their submissions on why the Indonesian Proceedings were oppressive and vexatious. For reasons set out at [160]–[179] above, I was not persuaded that CSSC would be deprived of any legitimate juridical advantages by reason of my refusal to grant an anti-suit injunction.

I was not addressed on the injustice, if any, that OKI was likely to suffer if I allowed CSSC’s prayer for an anti-suit injunction. In any event, as there were numerous other grounds upon which CSSC’s application had to fail, any consideration of this point would have been academic.

For all the foregoing reasons, I maintain my earlier decision to dismiss CSSC’s application for a non-contractual anti-suit injunction.

Ground 4: The Indonesian Proceedings did not undermine the integrity of the Singapore courts’ jurisdiction, processes, and judgments

I turn now to address Ground 4, which was OKI’s prayer for a protective anti-suit injunction. I have set out the relevant principles pertaining to such injunctions (at [32]–[35]) above.

Mr Toh submitted that a protective anti-suit injunction was necessary to protect the Singapore-seated arbitrations and the tribunal’s jurisdiction, over which the Singapore courts retained supervisory jurisdiction. In my judgment, that argument had to fail in light of my conclusion above at [131] (ie, that the Indonesian Proceedings have not been commenced in breach of the Arbitration Agreement).

Mr Toh submitted in the alternative that a protective anti-suit injunction was also necessary to protect the limitation proceedings in Singapore, which were unlawfully threatened by OKI’s commencement of the Indonesian Proceedings. I rejected that argument for the reasons set out at [163]–[179] above. As I noted at [177(e)], there was an imminent – if not already pending – challenge against the Indonesian courts’ jurisdiction to entertain OKI’s claim. At the same time, CSSC had yet to admit its liability to OKI in ADM 50. In those circumstances, it would have been premature for me to conclude that the Indonesian Proceedings amounted to an attack on the Singapore courts’ processes. As The Western Regent exemplifies, it would not be in the interests of judicial comity to pre-empt the Indonesian courts’ decision on its jurisdiction by granting CSSC a protective anti-suit injunction. In the absence of any other basis for asserting that the Indonesian Proceedings undermined the Singapore courts’ processes, I declined to grant the protective anti-suit injunction.

There was no basis for allowing OKI’s Alternative Prayer

Finally, CSSC invited me, by way of its Alternative Prayer, to grant a partial anti-suit injunction in the event I was not minded to allow the primary relief it sought (see [27] above). As would be apparent from the conclusions I have reached above, I did not see anything improper about OKI’s commencement of the Indonesian Proceedings. I again reiterate the point that the Indonesian courts had (at least up to the time of the FA Hearing) yet to dispose of OKI’s jurisdictional challenge in the Indonesian Proceedings, and that it was not for this Court to rush in pre-emptively (see [177(e)] and [186] above). There was accordingly no reason to restrict – whether partly or wholly – OKI’s conduct of the Indonesian Proceedings.

Conclusion

For the reasons set out above, I was of the view that there were no grounds for granting CSSC the injunctive relief it sought. I therefore maintain my earlier decision to dismiss SUM 2676.