1 These are the defendants’ appeals to a judge in chambers against the entirety of the decision of Assistant Registrar Desmond Chong in TA Private Capital Security Agent Ltd and another v UD Trading Group Holding Pte Ltd and another [2023] SGHCR 1 (the “GD”). ¹ AR Chong’s decision arose on the plaintiffs’ application under O 14 r 12 and O 18 r 19 of the Rules of Court (2014 Rev Ed) (the “Rules of Court”) to enter judgment against the defendants in accordance with their statement of claim. ²

2 In his decision, AR Chong declared a guarantee that had been executed by the first defendant in favour of the second defendant dated 15 April 2019 (the “Guarantee”) to be an on-demand performance guarantee (the GD at [139]). ³ He went on to strike out the defences in this action on the basis that all of the pleaded defences were wholly contrary to the plain text of the Guarantee and the other relevant contracts and documents. Accordingly, he entered judgment in favour of the plaintiffs against the first defendant in the sum of US$63.3m with interest and costs (the GD at [5] and [140]).

3 I have, in substance, dismissed the defendants’ appeals. If I had considered it necessary to characterise the Guarantee, I would not have characterised it as an on-demand performance guarantee or as any other type of performance guarantee or bond. Despite this difference of approach, however, I agree with AR Chong that both defendants’ defences should be struck out. This is because, in my view, it is not necessary to declare the Guarantee to be an on-demand performance guarantee in order to determine whether the plaintiffs are entitled to judgment. It is plain and obvious that the first defendant’s defence is devoid of any merit, whatever the true character of the Guarantee may be. As for the second defendant’s defence, it is a completely unnecessary pleading. The second defendant is named as a defendant purely because it is the assignor of the US$63.3m debt that the first plaintiff claims in this action and because it does not wish to be named as a co-plaintiff in this action. As a result, neither plaintiff makes any claim whatsoever in this action against the second defendant. It is a complete waste of time, costs and judicial resources for a defendant against whom the plaintiffs make no claim to file a defence or indeed to be compelled by the plaintiffs to file a defence.

4 The defendants have appealed against my decision. I now set out the grounds for my decision.

5 The first plaintiff is a company incorporated in the British Virgin Islands. It acts as the security agent of the second plaintiff. The second plaintiff is a fund manager incorporated in Hong Kong. One of the funds that the second plaintiff manages is a fund known as the Asian Trade Finance Fund 2 (“ATFF2”). ⁴

6 The first defendant is a company incorporated in Singapore. It is the parent company of a multinational group of companies. I shall refer to that entire group of companies as “the UD Group”. Some of the members of the UD Group trade directly in metals and metal products. I shall refer to this trading subset of the members of the UD Group as “the Operating Companies”. The Operating Companies purchased metal and metal products from the second defendant. ⁵ The second defendant is a company incorporated in the province of Ontario in Canada. ⁶

7 There are three loan agreements that are relevant to this dispute:

(a) The first loan agreement is known as “the ATFF1 Loan”. The ATFF1 loan was entered into in 2017 between a company known as Cantrust (Far East) Ltd (“Cantrust”) as lender and the second defendant as borrower. The second plaintiff has stepped into Cantrust’s shoes as the lender under the ATFF1 Loan. ⁷

(b) The second loan agreement is known as “the ATFF2 Loan”. The ATFF2 Loan was entered into in 2019 between the second plaintiff as lender and the second defendant as borrower. Under the ATFF2 loan, the second plaintiff agreed to provide an uncommitted revolving trade finance facility to the second defendant of up to US$60m. ⁸

(c) The third loan agreement is known as “the UD Loan”. The UD Loan was entered into between the second plaintiff as lender and two of the first defendant’s subsidiaries as borrowers. ⁹ One of the subsidiaries is a company incorporated in Singapore and the other is a company incorporated in Malaysia. As security for the UD Loan, the plaintiffs took charges over the following property owned by the UD Group:

(i) 2,030,500 shares in a company known as Hangji Global Ltd (“Hangji” and the “Hangji Security”);

(ii) all of the shares in a company known as Gympie Eldorado Mining Pty Ltd (the “GEM Security”); and

(iii) all of the shares in certain of the Operating Companies. ¹⁰

8 The second defendant drew down funds under the ATFF2 Loan to purchase metal and metal products from its suppliers. The second defendant then on-sold the metal and metal products to the Operating Companies. ¹¹

9 The first defendant executed the Guarantee in favour of the second defendant on 15 April 2019. By the Guarantee, the first defendant guaranteed that the Operating Companies would pay their liabilities to the second defendant. ¹²

10 It is the plaintiffs’ case that the second defendant validly assigned its rights under the Guarantee to the first plaintiff on 22 November 2019 ¹³ by way of three documents: (a) a security deed (the “Security Deed”); ¹⁴ (b) a general security agreement (“GSA”); ¹⁵ and (c) a forbearance agreement. ¹⁶

11 On 24 January 2020, the first plaintiff issued a letter to the first defendant demanding that the first defendant pay to the first plaintiff the liabilities then owed by the Operating Companies to the second defendant. Those liabilities amounted to US$63.3m as at 24 January 2020. ¹⁷ As of 31 January 2020, the first defendant had not made payment of this sum or any part of it to the first plaintiff. ¹⁸

12 The plaintiffs commenced this action in July 2020 to recover the sum of US$63.3m from the first defendant. At the time they commenced this action, the plaintiffs named the second defendant as the third plaintiff. The plaintiffs thought it necessary to do so in order to ensure that the first plaintiff’s assignor (at that time named as the third plaintiff, now named as the second defendant) was a party to this action and therefore would be bound by any judgment in it.

13 The first defendant does not dispute that the Operating Companies owed $63.3m to the second defendant as at 31 January 2020. But it raises defences that it claims have the effect of discharging its liability under the Guarantee. ¹⁹

14 There are multiple related proceedings between the parties or some combination of them in Ontario. Only two of those are relevant for present purposes. The plaintiffs rely on these proceedings to argue that both defendants are now estopped from taking positions in this action that are contrary to the findings of the court in those related proceedings.

15 The first proceeding is an action that the first defendant commenced against the plaintiffs on 12 August 2020 (“UD’s Ontario Action”). On the plaintiffs’ application, the Ontario court has permanently stayed UD’s Ontario Action. ²⁰ In UD’s Ontario Action, the first defendant sought: (a) damages for loss alleged to have been caused to the first defendant and the Operating Companies arising from the plaintiffs’ improper enforcement of the Hangji Security and/or the GEM Security; and (b) a declaration that the Operating Companies have no liabilities covered by the Guarantee. ²¹

16 The second proceeding is an action that the second defendant commenced against the plaintiffs on 14 December 2020 (“the second defendant’s Ontario Action”). ²² In the second defendant’s Ontario Action, it seeks a declaration that this action in Singapore has not been properly authorised. The second defendant’s Ontario Action has been stayed pending the determination of this action in Singapore. ²³

17 I consider the following issues in these appeals:

(a) Is the Guarantee an on-demand performance guarantee?

(b) Should the defendants’ defences be struck out?

(i) Can the GEM and Hangji securities be used to satisfy the sum owed under the Guarantee?

(ii) Can the value of the GEM and Hangji securities be set off against the sum owed under the Guarantee?

(iii) Were notices validly given of the second defendant’s assignment to the first plaintiff of the debts due to it from the Operating Companies?

(iv) Are the defendants estopped from taking positions contrary to the findings of the Ontario court?

18 I start with the issue of whether the Guarantee is an on-demand performance guarantee. It is apposite at this time to set out the operative provisions of the Guarantee.

19 Under cl 3.1 of the Guarantee, the first defendant guaranteed that the Operating Companies would duly and punctually perform their obligations to the second defendant arising from their purchase of metals or metals products from the second defendant (referred to in the Guarantee as “Materials Contract”), including the due payment of all moneys and the due discharge of all liabilities arising from those purchases. Clause 3.1 provides as follows: ²⁴

20 By cl 3.2 of the Guarantee, the first defendant undertook, as a separate and independent obligation, to pay to the second defendant as “sole or principal debtor” and “on demand” any amounts guaranteed under cl 3.1 which proved, for whatever reason, not to be “recoverable on the footing of a guarantee”: ²⁵

21 Clause 9.1 is a conclusive evidence clause. It provides that, in the absence of manifest error, a certificate by the second defendant as to the “moneys and liabilities” due by the Operating Companies to the second defendant is “conclusive and binding” on the first defendant: ²⁶

22 Clause 7.3 obliged the first defendant to make all payments under the Guarantee without set-off or counterclaim or any other deduction: ²⁷

23 The plaintiffs have made it central to their case that the Guarantee be characterised as an on-demand performance guarantee. Their case proceeds as follows. Because the Guarantee is properly characterised as an on-demand performance guarantee, the first defendant’s liability under the Guarantee is a primary liability. The first defendant cannot resist payment under the Guarantee by attacking the validity or subsistence of the debt that the Operating Companies owe to the second defendant. The first defendant will be liable under the Guarantee simply upon proof that: (a) the second plaintiff made a demand on the first defendant for payment under the Guarantee; and (b) that demand complied with the terms of the Guarantee. In the alternative, if the Guarantee is not properly characterised as an on-demand performance guarantee, the first defendant’s liability is a secondary liability. The first defendant will be liable under the Guarantee only if the second plaintiff can prove that: (a) the Operating Companies’ underlying debt to the second defendant is valid and subsisting; and (b) that the Operating Companies have failed to perform their obligation to pay that debt.

24 Given the contractual structure of the Guarantee, I do not consider it necessary to determine whether the Guarantee is properly characterised as an on-demand performance guarantee.

25 Clause 3.1 of the Guarantee obliges the first defendant to pay the debt that is due from the Operating Companies to the second defendant under the Materials Contract upon the second defendant’s demand for such payment. Assuming for the time being that the second defendant’s assignment of its rights under the Guarantee to the first plaintiff is valid, the first plaintiff has stepped into the shoes of the second defendant. From the time of the assignment, the first defendant owed its obligations under the Guarantee to the first plaintiff. The first plaintiff then demanded payment from the first defendant under the Guarantee. The first defendant accepts that, if time is frozen at the time of the first plaintiff’s demand, ie, 24 January 2020, the Operating Companies were in breach of their obligations under the Materials Contract and did owe US$63.3m to the second defendant (see the GD at [3]). ²⁸

26 There is no need to consider whether the first defendant’s obligation under the guarantee arises independently of establishing the Operating Companies’ default under the Materials Contract, whether by characterising the Guarantee as an on-demand performance guarantee or by relying on cl 3.2 of the Guarantee. Clause 3.2 applies only if, for whatever reason, some sums that are due from the first defendant to the second defendant are not recoverable from the first defendant under cl 3.1 “on the footing of a guarantee”. That situation can arise either because: (a) the Operating Companies have a substantive defence arising under the law of contract or under the general law to the second defendant’s claims under the Materials Contract and the first defendant is able under the law of assignment to assert that defence against the first plaintiff as assignor; or (b) the first defendant has a defence arising from the law of suretyship and unrelated to the Operating Companies’ defences as against the second defendant arising from the Materials Contract. The first defendant asserts no defence to the plaintiffs’ claim that falls into either category.

27 The question of whether the Guarantee is properly characterised as an on-demand performance guarantee is therefore immaterial. The first defendant accepts that the Operating Companies owed US$63.3m to the first plaintiff on 24 January 2020. Clause 3.1 is therefore engaged. There is simply no need to consider whether the first defendant owes an independent obligation as principal debtor to the first plaintiff whether by characterising the Guarantee as an on-demand performance guarantee or under cl 3.2 of the Guarantee.

28 In my view, the Guarantee is simply a contract. And so, it should be construed as a contract, applying the ordinary approach to contractual construction in light of the law on suretyship.

29 Had it been necessary for my decision, however, I would have found that the Guarantee is not an on-demand performance guarantee.

30 An on-demand performance guarantee is also known as a “performance bond”, “first demand guarantee” or “demand bond”. Whatever the name, it is a contract under which a guarantor undertakes to pay a creditor a sum of money upon the creditor’s demand. The guarantor can therefore be held liable to the creditor even without proof of any failure by the principal debtor to perform its obligations to the creditor and even if there has been no such failure (Vossloh Aktiengesellschaft v Alpha Trains (UK) Ltd [2011] 2 All ER (Comm) 307 at [46]). This is unlike an ordinary guarantee, under which the guarantor undertakes to pay the creditor only if the principal debtor fails to perform its obligations to the creditor.

31 The purpose of an on-demand performance guarantee is to protect against the failure of a contractual counterparty to perform (Wayne Courtney, The Modern Contract of Guarantee (Sweet & Maxwell, 4th Ed, 2020) (“The Modern Contract of Guarantee”) at para 13-001). A party to a contract therefore seeks an on-demand performance guarantee when it has doubts about its counterparty’s ability to perform its obligations under their contract or its ability to pay damages for a failure to perform those obligations. The on-demand performance guarantee confers two advantages upon the creditor. First, the creditor can secure payment simply upon demand, regardless of the existence or merits of any dispute between the creditor and principal debtor. Second, because the guarantor is liable to pay simply upon demand, there is a much lower risk of a dispute or litigation (Master Marine AS v Labroy Offshore Ltd and others [2012] 3 SLR 125 (“Master Marine”) at [26]).

32 Although on-demand performance guarantees have been mistaken for guarantees and likened to “exceptionally stringent contracts of indemnity” (H. G. Beale, Chitty on Contracts, Volume 2: Specific Contracts (Sweet & Maxwell, 34th Ed, 2021) at para 47-009), they are more accurately seen as being analogous to letters of credit, though they are “not equivalent to a letter of credit for all purposes” [emphasis in original omitted] (Master Marine at [26], citing LP Thean JA, “The Enforcement of a Performance Bond: The Perspective of the Underlying Contract” [1998] 19 Sing LR 389 at 400). Nevertheless, like the letter of credit, commerce and the courts treat the on-demand performance guarantee as the equivalent of cash. Provided that the demand complies with the terms of the guarantee, the only two bases on which a guarantee can resist payment are by establishing the narrow grounds of either fraud or unconscionability (Dauphin Offshore Engineering & Trading Pte Ltd v The Private Office of HRH Sheikh Sultan bin Khalifa bin Zayed Al Nahyan [2000] 1 SLR(R) 117 at [34]–[41]).

33 Because the on-demand performance guarantee is offered and taken as a substitute for cash and is therefore treated as the functional equivalent of cash, the identity of the guarantor is of utmost importance. It is for this reason that an on-demand performance guarantee is typically issued by a bank, insurance company or other financial institution (Meritz Fire and Marine Insurance Co Ltd v Jan De Nul NV and another [2011] 1 All ER (Comm) 1049 at [66]). The on-demand performance guarantee substitutes the guarantor’s creditworthiness for the counterparty’s creditworthiness. The creditor will therefore want the guarantor to satisfy two criteria. First, the guarantor must be of high creditworthiness. Second, the guarantor must have a high reputation for performing its own contractual obligations. Financial institutions satisfy both criteria. That is why the on-demand performance guarantee is typically issued by a financial institution. It is their financial strength and their reputation that gives creditors the confidence that they will get in fact get cash upon demand.

34 In Marubeni Hong Kong and South China Ltd v Mongolian Government [2005] 1 WLR 2497, the English Court of Appeal held that there is a strong presumption that parties do not intend to create an on-demand performance guarantee where a guarantee is given outside the banking context and does not contain language appropriate for an on-demand performance guarantee (“Marubeni presumption”).

35 There are differing views on whether the Marubeni presumption is part of Singapore law. In China Taiping Insurance (Singapore) Pte Ltd (formerly known as China Insurance Co (Singapore) Pte Ltd) v Teoh Cheng Leong [2012] 2 SLR 1 at [19], Chan Seng Onn J recognised that the Marubeni presumption was “important”, without expressly accepting that it was part of Singapore law. For the reasons I have given, I consider the rationale behind the presumption to be sound. I accept that the Marubeni presumption is part of Singapore law. In the present case, the Marubeni presumption is engaged because the first defendant, the guarantor, is not a bank.

36 Related to the Marubeni presumption are the factors set out in John Odgers QC (gen ed), Paget’s Law of Banking (LexisNexis, 15th Ed, 2018) at para 35.8 (the “Paget factors”). It states that an instrument will almost always be construed as an on-demand performance guarantee where it: (a) relates to an underlying transaction between the parties in different jurisdictions; (b) is issued by a bank; (c) contains an undertaking to pay “on demand”; and (d) does not contain clauses excluding or limiting the defences available to a guarantor (Wuhan Guoyu Logistics Group Co Ltd v Emporiki Bank of Greece SA (No. 1) [2013] 1 All ER (Comm) 1191 at [26]–[28]). The Paget factors also point against the Guarantee being an on-demand performance guarantee, again because the first defendant is not a bank.

37 The Marubeni presumption and the Paget factors are not intended to be conclusive in themselves. Ultimately, the characterisation of a guarantee as an on-demand performance guarantee is a matter of construction. An example of words construed as creating an on-demand performance guarantee in the banking context is found in Wood Hall Ltd v Pipeline Authority (1979) 141 CLR 443 at 447:

38 An example of words construed as creating an on-demand performance guarantee outside the banking context is found in Bitumen Invest AS v Richmond Mercantile Ltd FZC [2016] EWHC 2957 (Comm) at [19]:

39 It is clear to me that the words of cl 3 of the Guarantee cannot be construed as creating an on-demand performance guarantee.

40 The first defendant did not undertake in the Guarantee to pay the liabilities due from the Operating Companies to the second defendant simply upon demand. Instead, the first defendant’s liability under cl 3.1 arises only if there has been a failure in “due and punctual performance by the Operating Companies of all of their obligations” (cl 3.1.1). It is only in that event that the first defendant comes under an obligation to “pay [to the first plaintiff] such sum on demand” (cl 3.1.2). Further, the first defendant’s liability under cl 3.2 arises only if cl 3.1 is triggered, but for whatever reason the Operating Companies have a defence to the second defendant’s claim under the Materials Contract or the first defendant has a defence to the second defendant’s claim under cl 3.1 of the Guarantee. It is only in that event that the first defendant becomes liable to pay the first plaintiff the sum due to the second defendant from the Operating Companies “[a]s a separate and independent stipulation” (cl 3.2). As a matter of construction, it is clear that the first defendant’s obligations under cl 3 of the Guarantee do not arise unconditionally upon demand. Its obligations arise only if a specified condition precedent is satisfied.

41 The presence of cll 7.3 and 9.1 do not make the Guarantee an on-demand performance guarantee. I agree with the first defendant that cl 9.1 is conclusive only as to the quantum of money owed by the first defendant, and not conclusive as to the first defendant’s liability to pay. ²⁹ Conclusive evidence clauses will not be readily construed as being conclusive of the legal existence of the indebtedness (The Modern Contract of Guarantee at para 5-149, citing British Linen Asset Finance Ltd v Ridgeway (1999) GWD 2–78). And where a clause is found on its proper construction to cover both liability and quantum, this is a factor in favour of construing the instrument as an on-demand performance guarantee (Van Der Merwe v IIG Capital LLC [2008] 1 All ER (Comm) 435 at [32] and [51]). I cannot construe cl 9 as covering both liability and quantum.

42 As for cl 7.3, it merely prevents the first defendant from exercising any right of set-off against amounts due under the Guarantee. The only effect of this clause is to require the first defendant to litigate any alleged set-off, counterclaim or other right of deduction rather than to exercise a right of self-help against a payment due to the second defendant.

43 Even if the commercial result of an on-demand performance guarantee and the Guarantee may be the same, ie, the first defendant is liable to pay US$63.3m to the first plaintiff, the way in which the outcome is reached is significantly different. It is not the case that the first defendant’s liability to pay the first plaintiff arose merely by reason of the first plaintiff’s demand on 24 January 2020. Instead, the liability arose: (a) because the Operating Companies failed to perform their obligations to pay the second defendant (cl 3.1.1); (b) because this failure triggered the first defendant’s liability to pay (cl 3.1.2) and (c) because the first plaintiff made a demand upon the first defendant for payment under the Guarantee.

44 For the reasons I have given, I am satisfied that the Guarantee is not an on-demand performance guarantee.

45 I turn to address whether the first defendant’s defence should be struck out under O 18 r 19(1) of the Rules of Court and judgment entered against it accordingly. In my view, it should be. The first defendant’s defence is legally and factually unsustainable and thus frivolous or vexatious within the meaning of O 18 r 19(1)(b).

46 Before explaining my reasons for striking out the first defendant’s defence, I make two points.

47 First, where a plaintiff makes a claim against a defendant, there are three types of points a defendant can raise in response:

Points of the first and second type are true defences, going either to negate liability or to reduce quantum. A point of the third type is not a defence at all. The risk of over-recovery can always be addressed by the court extracting suitable undertakings from a plaintiff either when the court enters judgment or when the court lends its aid to the plaintiff by levying execution upon the judgment. The points raised by the first defendant are, at their highest, of the third type.

48 Second, where a debtor fails to pay a debt to a creditor and the creditor holds guarantees or security for the debt, the creditor has multiple and independent avenues of recourse to recover the debt (subject of course to contract or statute). Typically, the creditor can sue the debtor to recover the debt. He can sue the guarantor of the debt. Or he can have recourse to the security that the borrower or a third party has given for the debt. Unless otherwise bound by contract or statute, the creditor can choose in what order to pursue the various avenues of recourse. In the absence of a contractual right, a guarantor cannot resist liability to the creditor on the basis that creditor should have sued the principal debtor first. Likewise, in the absence of a contractual right, a principal debtor cannot resist the creditor’s claim on the basis that the creditor should have recourse to the security. Any risk that the creditor will recover more than the debt by pursuing these separate avenues consecutively can, as I have said, be addressed by the court extracting suitable undertakings from the creditor.

49 In the alternative, the points raised by the first defendant are of this type. They are not defences to liability. Instead, the first defendant simply assumes that there is a sequence of recourse that the first plaintiff must pursue in recovering the undisputed debt that the Operating Companies owe to the first defendant and that the first defendant has assigned to the first plaintiff, failing which the first defendant will not be liable under the Guarantee. There is no basis for this in the Guarantee or in any of the other relevant contracts.

50 I begin with what appears to be the first defendant’s primary defence. The first defendant relies on a series of conversations that occurred between February and May 2019, ³⁰ arguing that they amount to an agreement that the GEM and Hangji securities – given to secure only the UD Loan – can be used to satisfy the sums that the first defendant owes to the first plaintiff under the Guarantee (“Cross-collaterisation Theory”). ³¹

51 The first plaintiff’s response is as follows. First, the first defendant is not the legal or beneficial owner of the GEM and Hangji securities, and therefore cannot direct the first plaintiff to use those securities to satisfy its liability under the Guarantee. ³² Second, the UD Loan is a separate and distinct loan from the ATFF2 Loan. ³³ This is consistent with the first defendant’s admission, on multiple occasions, that the two loans are separate and distinct. ³⁴ Finally, the first defendant could not give particulars of when the agreement was allegedly reached. ³⁵

52 In oral submissions, the first defendant initially argued that parties came to a consensus – not a contract – that the GEM and Hangji securities could be used to satisfy the debt that the first defendant owes to the first plaintiff under the Guarantee. But upon further questioning, the first defendant applied – at the tail-end of counsel’s oral submissions – to amend the first defendant’s defence to plead a contract between parties to this effect. I allowed the amendment, subject to the usual costs consequences, so that I could determine the first defendant’s appeal on the best and final version of its defence.

53 This defence is nevertheless unsustainable. I do not accept that, as a matter of fact, there exists any connection between the GEM and Hangji securities and the debt that the first defendant owes to the first plaintiff under the Guarantee. The first defendant does not own the GEM and Hangji securities. I agree with AR Chong that the first defendant cannot rely on the GEM and Hangji securities to discharge or even reduce its own liability to the first plaintiff under the Guarantee (the GD at [111(b)]). At the very highest, the GEM and Hangji securities may be security for the Operating Companies’ liabilities to the first plaintiff under the ATFF2 Loan. But even then, the existence of that security affords the first defendant no defence to its liability to the first plaintiff under the Guarantee (the GD at [111(c)]). It merely raises a possibility that the first plaintiff may at some point in the future over recover.

54 Even if there exists a connection between the GEM and Hangji securities and the first defendant’s liability under the Guarantee, I do not accept that any contract exists obliging the first plaintiff to apply the GEM and Hangji securities against the first defendant’s liabilities under the Guarantee, failing which the first defendant will not be liable under the Guarantee. Much of the first defendant’s oral submissions were premised on the discussions between parties leading to nothing but a consensus ad idem falling short of a contract. Any such consensus is incapable of having legal consequences, as are necessary to afford a defence to the first defendant. The first defendant’s original framing of its case and its reframing of its case at the tail end of its submissions suggests to me that the allegation of a contract is a mere afterthought.

55 Further, even on the merits, I am satisfied that there was no such contract. I come to that conclusion for two reasons.

56 First, for a contract to be formed, there must be offer, acceptance, consideration, intention to create legal relations and certainty of terms. Although the traditional oral and acceptance analysis may be of limited assistance where continuing negotiations are concerned (Projection Pte Ltd v The Tai Ping Insurance Co Ltd [2001] 1 SLR(R) 798 at [16]–[17]), parties are still required to plead a specific date when the contract was concluded (Likpin International Ltd v Swiber Holdings Ltd and another [2015] 5 SLR 962 (“Lipkin International”) at [42]). The flexible approach in cases of continuing negotiations does not require departure from the offer and acceptance model, but simply requires a deeper appreciation of the context of the agreement, to be gained by looking at the whole course of negotiations (Gay Choon Ing v Loh Sze Ti Terence Peter and another appeal [2009] 2 SLR(R) 332 at [63]).

57 The first defendant’s case of an oral contract relies on events from February to May 2019. That spans three months. But the first defendant does not plead when, within these three months, the alleged contract was concluded. This is not sufficient. I accept that a party should be given latitude to plead with reasonable certainty the date at which an oral agreement was concluded. The party alleging it need not plead an exact date. But a pleading that asserts that an agreement was somehow entered into over a period of three months fails the reasonably certain test required for the formation of an oral agreement (Independent State of Papua New Guinea v PNG Sustainable Development Program Ltd [2019] SGHC 68 at [148]).

58 This requirement also gives commercial certainty to the parties that their continuing negotiations will not be taken to be a binding contract unless there was a clear acceptance at some point in time (Day, Ashley Francis v Yeo Chin Huat Anthony and others [2020] 5 SLR 514 (“Day, Ashley Francis”) at [49]). It is not desirable nor sufficient for a plaintiff to pool a universe of e‑mails, messages and conduct, and argue that the collective sum of these show that an agreement must have had been reached at some point in time, without making a case as to when that point was (Day, Ashley Francis at [49]).

59 The first defendant’s pleaded case of an oral contract is unsustainable. Further, given the circumstances in which it was put forward, I am entitled to assume that it cannot be improved by amendment or particulars.

60 Second, on the facts, the discussions between February and May 2019 do not reveal that any contract was concluded. The evidence shows only that the plaintiffs asked for details of the various securities. At best, all that the evidence shows is that the first defendant formed a belief unilaterally that the GEM and Hangji securities could or would be used to reduce or extinguish the debt that the first defendant owes to the first plaintiff under the Guarantee. There is simply no evidence of a consensus ad idem between the parties sufficient to afford the first defendant any sort of a defence, let alone a contract to the same effect.

61 Where a party is unsure of the most salient terms of a contract – including the date of the contract, whether the contract was oral or written and the consideration that was agreed – it is frivolous, vexatious and an abuse of process to expect the counterparty to defend a contractual claim (Likpin International at [46]).

62 Even if I accept the first defendant’s claim that there was a contract, it remains the case that it is not for the first defendant to dictate the sequence in which the first plaintiff should have recourse to its personal claim against the Operating Companies for the debt, against the first defendant on its guarantee for the debt or against the Gem and Hangji security, assuming that there was indeed a contract that they were to be used as security for the first defendant’s debt to the first plaintiff under the Guarantee. I say this because the first defendant’s defence is essentially that the first plaintiff should satisfy the sums that the first defendant owes to the first plaintiff under the Guarantee by realising the GEM and Hangji securities, instead of claiming that sum from the first defendant. It is not a defence for a debtor to complain that a creditor is taking a course of action that the debtor considers undesirable or even irrational. In so far as the first defendant’s real concern is about the first plaintiff over-recovering, that is not a defence.

63 As an alternative to the Cross-collaterisation Theory, the first defendant argues that Ontario law applies to the Guarantee ³⁶ and makes the following submissions. Under Ontario law, cl 7.3 of the Guarantee is not wide enough to exclude equitable set-off. ³⁷ Further, Ontario law allows obligations to be set off against each other even if they are not between the same parties, as long as there is a close connection between those obligations. ³⁸ There is a close connection here because the UD Loan and the GEM and Hangji securities on the one hand and the ATFF2 Loan and the Guarantee on the other hand are closely connected transactions. ³⁹ And because determining the closeness of the connection necessary for equitable set-off under Ontario law is a fact-intensive exercise, the first defendant’s defence should not be struck out and should instead be allowed to proceed to trial. ⁴⁰

64 The first plaintiff argues that regardless of whether Singapore law or Ontario law applies to the Guarantee, the first defendant cannot assert either legal or equitable set-off. Under Singapore law, both legal and equitable set-off can be contractually excluded. And both are excluded by cl 7.3. ⁴¹ Further, under Ontario law, legal set-off is excluded by cl 7.3. ⁴² And equitable set-off does not apply because the first defendant has not adduced any evidence demonstrating the necessary a close connection. ⁴³

65 This defence is unsustainable and ought to be struck out. From my review of both parties’ expert opinions on Ontario law, there is no significant difference between Singapore law and Ontario law on the exclusion of a debtor’s right of set-off, whether legal or equitable. Both systems of law accept that parties have the contractual freedom to vary or even to exclude entirely the exercise of legal or equitable set-off. Ascertaining whether they have done so is ultimately a question of construing the contractual provision said to have this effect in accordance with the usual principles of contractual construction. Under both systems of law, the starting point in the exercise of contractual construction is the words which the parties have chosen to express their agreement.

66 The words chosen by the parties in cl 7.3 of the Guarantee could not be clearer: ⁴⁴

67 It is indisputable that the parties intended cl 7.3 to exclude the first defendant’s right of set-off. I consider the wording of the clause wide enough to exclude set-off in all its varieties, both legal and equitable set-off, and both as a matter of Singapore law and as a matter of Ontario law. As AR Chong rightly acknowledged, the word “set-off” is broad enough to cover both legal and equitable set-off (the GD at [113]). The parties did not limit “set-off” in cl 7.3 to legal set-off. This must mean that they intended the word not to be limited to legal set-off (the GD at [113]). In contrast, it is usually clauses providing that amounts are to be paid “without any deduction” that may, in some contexts, be insufficient to exclude equitable set-off (The Modern Contract of Guarantee at para 11-081, citing Connaught Restaurants Ltd v Indoor Leisure Ltd [1994] 1 WLR 501).

68 The first defendant’s expert opines that cl 7.3 is not wide enough to bar equitable set-off for all purposes. But he does not substantiate his opinion by setting out for what purposes cl 7.3 bars equitable set-off and for what purposes cl 7.3 does not bar it. Nor does he cite any authority from any jurisdiction for coming to this conclusion. Like AR Chong, I reject this argument as contrived and unsustainable.

69 I turn to address the first defendant’s defence arising from the notices of assignment. Their defence proceeds in two steps as follows. First, the GSA and Security Deed were executed on 22 November 2019. This was several months after the alleged drawdowns took place between May and July 2019. The ATFF2 facility agreement requires the security to be executed before the drawdown dates. Therefore, the Security Deed and the GSA were not executed pursuant to the conditions of the ATFF2 facility agreement and the assignment is invalid. ⁴⁵ Second, even if the GSA and/or Security Deed are independent of the ATFF2 facility agreement, notice of assignment must be given to the Operating Companies in order for the Guarantee to cover their debts to the second defendant. ⁴⁶ The GSA and the Security Deed were executed after the notices of assignment were allegedly given, thus rendering the assignment invalid. ⁴⁷

70 This defence is unsustainable and ought to be struck out. At the outset, I agree with the first plaintiff that the execution of the GSA and the Security Deed are conditions precedent to the operation of the Forbearance Agreement. ⁴⁸ This is provided in cl 3.1(d), which states:

71 Although the GSA and Security Deed were executed only on 22 November 2019, the assignment covers both the second defendant’s accrued and future rights. ⁴⁹ Additionally, cll 3.1.1, 3.1.2, 4.1.1 and 4.1.3 of the Guarantee provide that the first defendant is liable for the Operating Companies’ past, present and future liabilities. ⁵⁰ There is no contractual basis for the first defendant’s submission that the first plaintiff can recover only moneys drawn down on or after 22 November 2019. In its position as the assignee of the second defendant’s rights under the Guarantee, the first plaintiff can do what the second defendant could have done pursuant to s 9.01 of the GSA and the power of attorney in cl 19 of the Security Deed. ⁵¹

72 With regard to the notice of assignment, I agree with the first plaintiff that no notice of assignment is necessary to establish the first defendant’s liability. ⁵² A failure to give notice of an assignment does not affect the contractual effectiveness of an assignment as between the assignor and the assignee (Holt v Heatherfield Trust Limited and another [1942] 2 KB 1 (“Holt v Heatherfield”) at 4). It is true that, until notice is given, the assignment remains merely an equitable assignment. But it is an assignment nonetheless. It requires nothing more from the assignor to be converted into a legal assignment by notice (Holt v Heatherfield at 4).

73 Further, the consent of the first defendant is not required for an assignment of the second defendant’s rights under the Guarantee to be valid. ⁵³ Clause 12.3 of the Guarantee expressly provides that the assignment of any rights under the Guarantee does not require the first defendant’s consent to the assignment or even notice of the assignment to the first defendant to be effective (the GD at [131]). Clause 12.3 provides as follows: ⁵⁴

74 The consent of the first defendant is expressly made immaterial to the validity of the assignment. It is true that this clause requires the second defendant to give notice of the assignment to the first defendant. But the only result of a failure to do so is that the second defendant is in breach of contract. The failure does not, under cl 12.3, invalidate the assignment or put the assignor in breach of any duty to the first defendant, contractual or otherwise. In short, it affords the first defendant no defence.

75 I therefore find that the second defendant’s assignment of its rights and liabilities under the Guarantee to the first plaintiff is an effective assignment. This defence is unsustainable.

76 The first plaintiff argues that the Ontario courts have found in UD’s Ontario Action that the ATFF2 Loan is “a different matter to which [the first defendant] is not a party”. ⁵⁵ Transnational issue estoppel therefore applies to estop the first defendant and the second defendant from taking a contrary position before the Singapore court. ⁵⁶ If they do so, it is an abuse of process and liable to be struck out. ⁵⁷

77 I do not agree with the first plaintiff for two reasons. First, one requirement of transnational issue estoppel is that the foreign judgment must be a final and conclusive decision on the merits (Merck Sharp & Dohme Corp (formerly known as Merck & Co, Inc) v Merck KGaA (formerly known as E Merck) [2021] 1 SLR 1102 at [35]). Context is key. The Ontario courts hade the finding in question in the context of a motion by the first defendant for an anti-suit injunction against the first plaintiff. ⁵⁸ And it was in the context of examining whether Ontario was the most appropriate forum (or forum conveniens) that the Ontario court held at first instance that the first defendant is not a party to the ATFF2 Loan. ⁵⁹

78 Although issue estoppel can in principle arise in respect of foreign decisions on interlocutory matters (Lakshmi Anil Salgaocar v Jhaveri Darsan Jitendra [2019] 2 SLR 372 at [101]), I do not think that this decision of the Ontario courts is correctly characterised as being a final and conclusive decision on the merits. A decision of a foreign court as to whether that foreign court is the appropriate forum, and the appropriateness of other forums, can never give rise to issue estoppel (William Jacks & Co (Singapore) Pte Ltd v Nelson Honey & Marketing (NZ) Ltd [2015] SGHCR 21 at [60]–[62]).

79 Second, I give weight to the Ontario Court of Appeal’s findings that the first instance judge did not make a final and conclusive decision on the merits. ⁶⁰ The Ontario Court of Appeal recognised that the first instance judge “was doing no more than rejecting this argument”, ie, the argument that the litigation was presumptively linked to Ontario because the plaintiffs’ claim against the first defendant arose from the ATFF2 Loan, and that the point at first instance was “that the [UD Loan] cannot be conflated with the [ATFF2 Loan] for the purpose of identifying the appropriate forum for the litigation”. ⁶¹

80 It is relevant for the Singapore court to consider what foreign law itself says about the nature of the judgment (The “Bunga Melati 5” [2011] 4 SLR 1017 at [71]–[74]). I therefore decline to treat this as a conclusive judgment on this point.

81 Therefore, I find that transnational issue estoppel does not apply here because there is no final and conclusive decision on the merits.

82 I now turn to consider the second defendant’s defence. I consider that it should be struck out. Given the factual background to the second defendant’s involvement in this litigation, the following four questions arise:

83 The present case engages all four questions. I briefly set out the procedural history. The second defendant was previously the third plaintiff in this action. ⁶² The second defendant then applied, as the third plaintiff, for leave to discontinue its action against the first defendant (at that time, the only defendant) or to stay its action against the first defendant. The court granted the second defendant leave to discontinue this action as a plaintiff but, in view of the need to have both assignor and assignee before the court, required the second defendant to remain a party as the second defendant.

84 The second defendant then proposed to the first plaintiff that it be treated as a nominal party to this action. ⁶³ The first plaintiff rejected the proposal, instead inviting the court to require the second defendant to file a defence. ⁶⁴ The second defendant therefore filed a defence. That defence was struck out by AR Chong, on the plaintiffs’ application. Although the second defendant wished to remain only a nominal party to this action, and although it advanced that intention for AR Chong to strike out its defence, the second defendant appealed against AR Chong’s striking out order to me.

85 There is no reason for the second defendant, against whom the plaintiffs make no claim, ⁶⁵ to have a defence on file. During oral submissions, counsel for the first defendant confirmed that the first defendant does not rely, for its own defence, on any of the matters pleaded by the second defendant. The second defendant’s defence is an entirely superfluous document. There was no reason for the second defendant to file the defence, for the plaintiffs to compel the second defendant to file the defence or for the second defendant to appeal against AR Chong’s decision striking it out.

86 On that ground alone, I dismiss the second defendant’s appeal. It is therefore not necessary to examine the merits of the second defendant’s defence.

87 For all of the foregoing reasons, I hold that first, the Guarantee is not an on-demand performance guarantee. I nevertheless hold that all of the first defendant’s defences are unsustainable and should be struck out. The second defendant’s defence was entirely unnecessary and should be struck out.

88 The judgment entered by AR Chong against the first defendant in the first plaintiff’s favour for US$63.3m, interest and costs stands.