1 The parties to this suit were presented with attractive business propositions by a third party known as Ronald Wee (“Ronald”). The plaintiff, Amberwork Source Pte Ltd (“Amberwork”), was told that Ronald’s business, Weroc Group Pte Ltd (“Weroc”), was looking to buy goods (specifically, telecommunications/fibre optic cables and related goods) for on-sale to its customers in China but lacked the funds to do so. Amberwork was thus invited by Ronald to purchase these goods and sell them to Weroc, but on deferred payment terms. On the other side of the fence, the defendants were told by Ronald that he had (via Weroc) goods to resell to customers in Singapore but needed to sell them through a reseller that was an established enterprise in Singapore. The first defendant, QA Systems Pte Ltd (“QA”), became that enterprise, and proceeded (as Weroc’s authorised reseller) to sell goods to Amberwork. Amberwork purchased those goods and resold them to Weroc for a profit.

2 This dispute has arisen because Amberwork has not been paid by Weroc. Amberwork sues QA as the first defendant for the sums Amberwork has paid on QA’s invoices, claiming that QA (as seller) has failed to deliver the goods it sold to Amberwork. It also sues QA’s director, the second defendant Yeo Chow Wah (“Sandra Yeo” or “Sandra”), for dishonestly assisting Weroc and/or Ronald to commit acts that amounted to a breach of trust. Sandra Yeo is a director and shareholder of QA, and also Ronald’s aunt. ¹

3 Both defendants disavow any liability. QA contends that it was a mere payment agent for Ronald/Weroc, that Amberwork and QA did not share any contractual relationship as they were never ad idem, and even if they were, that the sale and purchase contracts were sham transactions or tainted by illegality as they were in substance unlicensed moneylending transactions. Sandra Yeo also denies liability for the claims made against her personally.

4 The parties fiercely contest the proper characterisation of the transactions that form the subject matter of this action. The difficulty with this case is that the key protagonist, Ronald, passed away in March 2020 in unnatural and somewhat tragic circumstances. As a result, all the court is left with to piece the puzzle together are the parties’ contemporaneous communications and the opposing narratives of their witnesses.

5 The plaintiff, Amberwork, is in the business of trading cables. ²

6 In 2017, its sole director Ang Say Cheong (“Roger Ang” or “Roger”) and/or its finance manager Pua Poh Lim Pauline (“Pauline Pua” or “Pauline”) received a business proposition from Ronald. ³ Ronald was Roger’s long-time acquaintance from the cable industry. ⁴ He was also the sole director and shareholder of Weroc, a Singapore-incorporated private equity firm. ⁵

7 According to Amberwork’s pleaded case, the business proposition envisaged that: ⁶

(a) Roger/Pauline would purchase cables and related goods from an authorised reseller (the “Reseller”) recommended by Ronald and resell the same goods to Weroc;

(b) the Reseller would make the goods available for Weroc’s collection, ex-factory in China; and

(c) Weroc would then pay Roger/Pauline for the goods (within 60 days of receiving their invoice).

8 In this way, Weroc (which, according to Amberwork’s pleaded case, lacked sufficient funds to pay for the goods upfront ⁷ ) would benefit from the deferred payment terms. In consideration for this, Weroc would pay Roger/Pauline a higher price than that which Roger/Pauline would pay to the Reseller, thereby enabling Roger/Pauline to earn the difference as profit. ⁸ Roger/Pauline would not be told of the identity of the ultimate supplier of the goods in China as Ronald wished to protect those details as his trade secrets. ⁹

9 In 2018, Roger/Pauline entered into sale and purchase transactions based on this arrangement. The transactions were conducted through various entities that they controlled. ¹⁰

10 The present suit concerns two such transactions (individually, the “First Transaction” and “Second Transaction”; and collectively, the “Transactions”). The Transactions were between Amberwork and QA.

11 The Transactions relate to two invoices issued by QA to Amberwork on 10 September 2019 and 26 September 2019, for the sums of $605,132 and $80,460 respectively (individually, the “First Invoice” and “Second Invoice”; and collectively, the “Invoices”). ¹¹ The Invoices were issued on QA’s letterhead, bore its company stamp, and were signed by its representative. ¹² They were described as invoices for the supply of the items and materials that were stated therein. Both Invoices stated that the goods were to be collected ex-factory in Shenzhen, China, and stated when they were ready to be collected by. The Invoices also specified that payment was to be made by Amberwork to QA “T/T Immediate upon bill”. ¹³

12 It is undisputed that Amberwork paid the First Invoice in two tranches, on 10 September and 12 September 2019. ¹⁴ It is also undisputed that Amberwork paid the Second Invoice fully on 26 September 2019. ¹⁵ Amberwork claims that the goods stated in the Invoices were not delivered. ¹⁶

13 Almost half a year later, on 19 March 2020, Amberwork informed QA that it was cancelling its orders and requested a refund of $685,592 (ie, the total sum that Amberwork paid under the Transactions). ¹⁷ A refund was never effected. Instead, on 21 March 2020, Sandra responded by e-mail stating, among other things, that it “dutifully made the payment to Ronald … in good faith that the goods will be delivered to [Amberwork] without unnecessary delay”. ¹⁸ Amberwork interpreted this as QA’s repudiation of the agreements to supply the goods. In the present suit, it claims against QA for the sum of $685,592. ¹⁹

14 Amberwork’s case is that it is entitled to the return of the sum of $685,592. It says that contracts had been concluded between itself and QA for the sale and purchase of goods (the “Agreements”), on the back of or as evidenced by the Invoices. ²⁰ The Agreements satisfied the three crucial elements for determining the existence of a contract for the sale of goods: it provided for the identity of the parties, the price to be paid, and the specifications of the goods. ²¹ Moreover, these were legitimate transactions, structured as they were for the purposes of trade financing. ²² Consequently, Amberwork was entitled to the sum of $685,592 it paid under the Agreements as QA had repudiated the same. Further, or in the alternative, there was a total failure of consideration. ²³

15 In the alternative, if QA is found not to be party to the Agreements, it would instead be liable for fraudulently misrepresenting that it could supply the goods when it could not. ²⁴

16 Finally, even if QA was merely Weroc’s or Ronald’s “payment agent”, the wrongful retention by Weroc/Ronald of the sum of $685,592 would constitute a breach of trust. Sandra would be personally liable to account for that sum as a constructive trustee, for dishonestly assisting Ronald/Weroc in their breach of trust. ²⁵

17 QA and Sandra deny all liability for the Transactions. They say that the impugned transactions were all orchestrated by Ronald. In July or August 2019, Ronald had represented to Sandra that he had clients (like Amberwork) who wished to purchase goods from him, but who preferred to work with established counterparts. ²⁶ Weroc was unsuitable as it was only incorporated in 2017. ²⁷ Ronald proposed that Sandra use one of her companies to take advantage of this opportunity, and represented to her that he would make the necessary arrangements with Amberwork. ²⁸ Moneys received by QA from Amberwork (with QA acting as Amberwork’s agent) would be sent onward to Weroc, less an agreed administrative fee that QA would keep for its role in the transaction. ²⁹ Any communications and documents from QA to Amberwork were sent on Ronald’s instructions. ³⁰

18 There was therefore no common understanding that QA would supply or deliver the goods, and there was no contract between Amberwork and QA or any past dealings to this effect. Instead, the common understanding among all parties, including Amberwork and Roger, was that the delivery of goods would be fulfilled by Weroc and that QA’s role was limited to receiving payment from Amberwork and sending this on to Weroc less its administrative fee. ³¹

19 Furthermore, the Transactions were sham transactions which were in reality unlicensed moneylending transactions. Amberwork/Roger was in substance lending to Weroc/Ronald the sums stated in QA’s Invoices; Amberwork would obtain repayment through issuing its own invoices to Weroc and the profit element in Amberwork’s invoices was in fact interest on the loan. The Transactions could therefore not be enforced, as they were illegal and unenforceable under s 14 of the Moneylenders Act (Cap 188, 2010 Rev Ed) (“MLA”), and Amberwork was not an excluded moneylender under the MLA. ³²

20 Notwithstanding QA’s earlier contention that the common understanding of the parties was that the delivery of goods would be fulfilled by Weroc (and not QA) (see [18] above), it also avers later in its pleaded Defence that it was the only contracting party who believed the Transactions to be genuine business transactions. QA pleads this point to suggest that any contract it was party to was void and unenforceable as a result of this unilateral mistake of fact. ³³

21 In the main, two issues fall to be decided:

(a) whether QA was under an enforceable contractual obligation to deliver the goods to Amberwork; and

(b) if so, whether this obligation was breached.

22 While parties have raised other issues related to misrepresentation, dishonest assistance, etc, given my findings and conclusions below, these issues do not arise for consideration.

23 I address this issue in three parts:

(a) whether QA was merely Weroc’s or Ronald’s payment agent;

(b) whether the putative contracts between QA and Amberwork were part of sham transactions and/or unenforceable under the MLA; and

(c) whether there were validly constituted contracts between Amberwork and QA.

24 I reject QA’s submission that it was merely a payment agent or billing party with no obligation to supply goods under the Agreements.

25 In the first place, QA did not clearly plead that its role in the transactions was merely that of an agent, with no personal obligation to supply goods to Amberwork. Neither has it pleaded that it was merely a “billing party”. At paragraph 14 of its defence, QA pleads that its role “was limited to receiving payment from [Amberwork] and sending this on to Weroc less its administrative fee” ³⁴ – I do not consider this to be a sufficient pleading that QA’s role was merely that of a “payment agent” or “billing party”, or a sufficiently clear statement of what those phrases mean in terms of QA’s legal obligations and role in the Transactions. The closest the defendants’ pleaded defences come to as far as agency is concerned, is a pleading that QA was Amberwork’s agent. ³⁵ However, this contention that it was Amberwork’s agent (and not Weroc’s or Ronald’s agent) was not advanced at trial nor in the defendants’ closing submissions, and I regard the defendants as having abandoned this position.

26 In any case, quite apart from the pleading point, the defendants’ evidence does not make out any agency relationship between QA and Weroc/Ronald.

27 For one, if QA was indeed merely an agent (specifically, Weroc’s or Ronald’s agent), and genuinely understood its role to be that of an agent, then it is unusual that there are no invoices from QA to Weroc or Ronald for its “administrative fee”. Instead, it is undisputed that Weroc issued invoices as seller, to QA as buyer, for each of QA’s resale transactions with Amberwork; ³⁶ a mere payment agent or billing party would not have needed such invoices or have expected to receive such invoices from its alleged “principal”. Instead, Weroc’s issuance of invoices is more consistent with QA’s (and Sandra’s) subjective understanding from Ronald that QA would be buying goods from Weroc, for on-sale to Amberwork. Under cross-examination, Sandra was unable to explain why invoices would need to be issued by Weroc as seller, to QA as buyer, and even more perplexingly, why they had to be issued even after the moneys that QA received from Amberwork had already been paid over to Weroc (less QA’s administrative fee). ³⁷

28 The contention that QA and Sandra tended to simply follow Ronald’s instructions is neither here nor there. Such conduct is not consistent only with an agency relationship. Given that this was a business arrangement put in place by Ronald, and it was primarily Ronald who liaised with the other key players, following Ronald’s instructions would equally have just been good commercial and practical sense. More relevantly, it would also be entirely consistent with the existence of a seller-buyer relationship between Weroc and QA.

29 A third observation I would make is that QA did not describe its role as merely that of an agent in the contemporaneous documents. The Invoices and quotations issued at the time do not, on their face, state that QA was an agent. When payment was received from Amberwork, QA never suggested that it was under no obligation to deliver the goods to Amberwork as it was merely an agent. In cross-examination, Sandra admitted that she never once informed Roger or Pauline that she was merely a “billing party”. ³⁸ In her e-mail of 21 March 2020, where she responded to Roger’s allegation that the goods had not been delivered, Sandra did not say that QA was a mere agent. ³⁹ On the stand, Sandra explained this apparent silence on the basis that she had already spoken to Roger about QA’s role as a billing party over the phone. ⁴⁰

30 This evidence refers to a telephone conversation that took place on or about 19 March 2020 between Sandra and Roger, by which time Ronald’s whereabouts were unknown and he was completely unresponsive to messages from Roger and Sandra. Evidence of this conversation was given by way of an allegedly contemporaneous note. ⁴¹ As matters unfolded during the trial, Sandra conceded on the stand that the note was not in fact made contemporaneously with the telephone conversation, but was prepared nearly a month after the fact on or about 12 April 2020. ⁴² There is no other evidence of what was said during the conversation: the conversation was not recorded, nor was Sandra’s daughter (who assisted her with typing out the note ⁴³ ) present during the conversation. In the circumstances, the allegedly contemporaneous note can be given very little weight in terms of its probative value in evidencing the contents of the 19 March 2020 telephone conversation.

31 The defendants sought to apply a further spin to Sandra’s 21 March 2020 e-mail in their reply submissions. They sought to draw these words from the e-mail to the court’s attention (with emphasis added by the defendants to the italicised words below): ⁴⁴

32 Ostensibly, the defendants’ point is that these words are consistent with what a payment agent or billing party might say. However, in my judgment, they are equally consistent with the existence of a series of back-to-back sale and purchase transactions – with Amberwork making payment to QA in the Amberwork-QA leg, followed by QA making “the payment” or “all payment” to Ronald in the QA-Weroc leg. Thus, the 21 March 2020 e-mail does not assist the defendants’ case considerably, even with the further spin they sought to put on it.

33 Even as between the defendants and Weroc/Ronald, the evidence pointing to an understanding that QA’s role was merely to act as Weroc’s or Ronald’s billing agent is weak. The defendants’ closing submissions refer to paragraphs 11 to 17 of Sandra’s affidavit of evidence-in-chief (“AEIC”), but those paragraphs do not refer to any evidence (beyond an assertion from Sandra) of QA being a billing agent. When pressed on the absence of evidence in her AEIC during cross-examination, Sandra was unable to point to any specific page or paragraph from her AEIC that would change this conclusion. ⁴⁵ The defendants’ reply submissions only refer additionally to WhatsApp conversations between Sandra and Ronald which mentioned that QA could take or keep 2% as an “admin charge” or “admin fees”, and which used the word “billing” in relation to another entity. ⁴⁶ But as highlighted above at [27], it is striking that there are no invoices from QA for these “admin fees”. Without more evidence and context on what parties regarded these fees to be, I am not prepared to place much weight on two brief references to the words “admin charge” and “admin fees”.

34 In short, QA’s purported agency relationship with Weroc/Ronald has not been substantiated. QA’s conduct and the contemporaneous documents instead show that QA acted as principal and transacted as such with both Amberwork and Weroc. QA issued invoices as buyer (from Weroc) and seller (to QA) respectively, in its own name, and without any suggestion on their face that it was merely a billing party or payment agent.

35 The defendants raise two intertwined defences for why the Transactions are unenforceable (see [19] above):

(a) they were sham, round-tripping transactions involving fictitious goods; and

(b) they were unlicensed moneylending transactions that contravened the MLA, involving Amberwork as the lender and Weroc as the borrower.

36 In my judgment, both defences fail. Let me elaborate.

37 On the first defence, the defendants have failed to show that the transactions were a sham. I note from the written submissions that none of the parties has developed this argument in its own right, as distinct from the MLA argument. In my view, it is important to first decide if the legal test for whether a transaction is a sham is met – only then can one go on to consider if the transaction in question was in fact something else. It is well-established that a sham transaction requires a common subjective intention between all the parties to, inter alia, create documents that are in fact not intended to create the legal rights and obligations which they give the appearance (to third parties or to a court) of creating (Crédit Agricole Corporate & Investment Bank, Singapore Branch v PPT Energy Trading Co Ltd and another suit [2022] 4 SLR 1 (“Crédit Agricole”) at [120], citing the often quoted test laid down by Diplock LJ (as he then was) in Snook v London and West Riding Investments Ltd [1967] 2 QB 786 at 802). In this regard, “[w]hat is required for a sham is a finding the parties to the sham were dishonest in creating a pretence of a transaction in order to deceive others when there was in reality no such transaction” (Crédit Agricole at [120]). Because a finding of a sham carries with it a finding of dishonesty, there is a strong and natural presumption against holding a provision or document a sham (Chng Bee Kheng and another (executrixes and trustees of the estate of Fock Poh Kum, deceased) v Chng Eng Chye [2013] 2 SLR 715 at [51]).

38 In the case before me, there is no evidence of any such common intention existing, whether as between Amberwork and the defendants, or as between Amberwork, the defendants and Weroc/Ronald.

39 In addition, the defendants’ characterisation of the transactions as a form of “round-tripping” does not take their case very far, and there are a number of reasons for this. First, round-tripping is not in and of itself illegal or fraudulent. This was explained in Crédit Agricole at [123]:

40 Crédit Agricole is but one case in a line of cases which have recognised that round-tripping transactions are not ipso facto illegal. These cases have recently been analysed in detail in UniCredit Bank AG v Glencore Singapore Pte Ltd [2022] SGHC 263 at [34]–[74], and I see no reason to depart from that analysis.

41 To be sure, a different conclusion might be reached if these were round-tripping transactions involving fictitious goods. This distinction was explained in Goodwood Associates Pte Ltd v Southernpec (Singapore) Shipping Pte Ltd and another suit [2020] SGHC 242 at [47]–[48]:

42 In the case before me, however, there is no pleading to the fact that the transactions involved fictitious goods or that they were in this sense not real sale and purchase contracts. The evidence also does not objectively suggest that there were no goods or that the parties were (knowingly) transacting on the basis of fictitious goods. For instance, the WhatsApp conversations between Roger and Ronald clearly implied that goods in fact existed, and there is no suggestion that these WhatsApp exchanges were themselves shams intended to create a pretence. In one exchange, a concern was raised that customers might not be able to collect the goods if paperwork was delayed. In that same exchange, a suggestion was made as to whether paperwork could be settled “[w]ithout actual [taking of] goods yet”. ⁴⁷ At best, there is some oblique suggestion in the defendants’ case that there is no evidence of anyone physically seeing the goods. ⁴⁸ Reference was made to the AEIC of Bak Kai Wei (“Daniel”), who was Ronald’s personal assistant and was involved in managing Weroc’s affairs (which included the issuance of purchase orders to Amberwork and payment on Amberwork’s invoices). ⁴⁹ In his AEIC, Daniel stated that over the course of Weroc’s transactions with Amberwork, he had never seen any physical goods being transacted or delivered by Amberwork to Weroc. Furthermore, when he asked Ronald about this, Ronald represented to him that the deliveries were conducted by his counterparts in China, and that Daniel “did not have to concern [himself] with the actual transfer of goods”. ⁵⁰ In my judgment, these statements are inconclusive. The quoted statement suggests that Daniel was not involved with the actual transfer of goods. If so, it is entirely unsurprising that he had not seen any goods. As Roger explained on the stand (when discussing some earlier transactions with another Reseller), since delivery was to be ex-factory and the factory was in China, it was to be expected that Amberwork would not see the physical goods. ⁵¹ More generally, this is not an uncommon or unexpected feature of transactions governed by a string of back-to-back contracts, as illustrated in Garnac Grain Company Incorporated v HMF Faure & Fairclough Ltd and another [1966] 1 QB 650 (“Garnac Grain”). Garnac Grain concerned a genuine chain of contracts for the sale and purchase of lard. Megaw J observed at 683 that:

43 The true essence of the sham argument, as alluded to above, is that the arrangements that were put in place were intended to be used as a disguise or pretence for what they really were, ie, unlicensed moneylending transactions (see, eg, the defendants’ closing submissions at para 110). As I explain next, even on this footing, the unlicensed moneylending defence also fails.

44 The defendants contend that Amberwork cannot seek to enforce its purported contracts by virtue of s 14(2) of the MLA, which provides that:

The purported contracts would constitute contracts for a loan, as Amberwork’s payments had the practical effect of giving a loan to Ronald/Weroc at an effective interest rate of 11% to 13% over two months. ⁵²

45 The defendants also refer to the prohibition on moneylending in s 5(1) of the MLA, but the thrust of their argument focuses on s 14(2). I note for completeness that s 5(1) states:

46 Returning to s 14(2), the central issue is whether Amberwork is an “unlicensed moneylender”. In their pleadings and submissions, the defendants do not appear to rely on the presumptive provision in s 3 to establish that Amberwork is a person presumed to be a moneylender. Instead, they rely on the approach set out in GA Machinery Pte Ltd and another v Yue Xiang Pte Ltd and others [2020] SGHC 264 at [18]:

I note that QA (despite not being the purported borrower as such, since that label would only be applicable to Weroc) has not suggested that the burden ought to fall on Amberwork to instead show that it is an excluded moneylender. Nonetheless, these questions of burden of proof are not material in this case, for reasons that will become clear below.

47 Excluded moneylenders are not regulated by the MLA: see Sheagar s/o T M Veloo v Belfield International (Hong Kong) Ltd [2014] 3 SLR 524 (“Sheagar”) at [67]. The defendants submit that Amberwork does not fall within the definition of an “excluded moneylender” in s 2:

48 They centre their submission on para (e)(iii)(A) in particular, and contend that Amberwork did not lend only to corporations (particularly, Weroc), because it was actually making direct, personal loans to individuals such as Ronald. Reference is made to other transactions where Roger appears to have assisted Ronald with his own personal dealings and/or paid moneys into Ronald’s personal bank account, based on motivations similar to those underlying the arrangements in the Transactions. ⁵³

49 In my judgment, the defendants have not shown that the Transactions were actually personal loans from Amberwork to Ronald. The evidence concerning the other transactions does little to change the complexion of what were, on their face, commercial transactions in this case involving the sale and purchase of goods, structured with a view to enabling trade financing. To begin with, the examples cited by the defendants concern only a subset of the other transactions involving Ronald, most of which were between corporate entities. In addition, some of the examples of other transactions cited by the defendants in their closing submissions (at para 112) involved Roger using his personal funds – so those examples did not even involve Amberwork at all. Yet other examples were met with an explanation. For instance, according to Sandra, Ronald had explained that certain moneys for one other transaction had to be exchanged with a moneychanger (presumably into foreign currency), and so were not to be paid directly to Weroc’s account but to Ronald’s personal account. It appears that Sandra believed this explanation and had even explained to her bank that the transaction in question was a genuine one. ⁵⁴ I am thus far from satisfied that the defendants have proven that Amberwork was lending moneys to individuals and therefore not lending “solely to corporations”.

50 More fundamentally, the defendants have made no serious attempt in their submissions ⁵⁵ to explain why the court can and ought to disregard the separate legal personality of Amberwork and QA – these being the contracting parties by and to whom obligations were owed – and to instead recharacterise the Transactions as personal loans to Ronald. The foundational quality of the doctrine of separate legal personality cannot be overstated: see Goh Chan Peng and others v Beyonics Technology Ltd and another and another appeal [2017] 2 SLR 592 at [71]–[75].

51 In any case, as the Court of Appeal explained in Sheagar at [66], as far as para (e) of the definition of “excluded moneylender” is concerned, Parliament regarded corporations as a less vulnerable class of borrowers that do not need the protection afforded by the MLA as a piece of social legislation. Similarly, as noted in City Hardware Pte Ltd v Kenrich Electronics Pte Ltd [2005] 1 SLR 733 at [47] (and affirmed in Donald McArthy Trading Pte Ltd and others v Pankaj s/o Dhirajlal (trading as TopBottom Impex) [2007] 2 SLR(R) 321 at [9]):

52 In my judgment, para (e) was not intended to extend to commercial trading firms in circumstances like the present.

53 After addressing the issue of whether Amberwork was an excluded moneylender, I note that the defendants’ closing and reply submissions do not go on to positively explain why Amberwork was an unlicensed moneylender. That is a crucial link that needs to be made in order for the defendants to rely on s 14(2) of the MLA. Be that as it may, and having considered the submissions and evidence, I agree with Amberwork’s position that it was not a “moneylender” in the business of moneylending.

54 In determining whether a person is engaged in the business of moneylending, the following considerations identified in Agus Anwar v Orion Oil Ltd [2010] SGHC 6 at [9] and [11] and Ochroid Trading Ltd and another v Chua Siok Lui (trading as VIE Import & Export) and another [2018] 3 SLR 617 (“Ochroid”) at [74] are germane:

(a) whether there exists system and continuity, ie, not merely occasional loans, but an organised scheme of moneylending (based on indicators such as fixed rates, the rate of interest being dependent on a borrower’s creditworthiness, the past conduct of the borrower, and a clear and definite repayment plan); and

(b) even if no system and continuity exists, whether the alleged moneylender is one who is ready and willing to lend to all and sundry provided that they are from his point of view eligible.

55 I agree with Amberwork that neither test is satisfied in this case. I acknowledge from the authorities that a person can be carrying on the business of moneylending even if that person is selective in lending on a commercial basis to only a few or even only one trusted, regular borrower: Ochroid at [76], citing Pankaj s/o Dhirajlal v Donald McArthy Trading Pte Ltd and others [2006] 4 SLR(R) 79 at [31]. However, in my judgment, Amberwork’s transactions lack the requisite systematicity envisaged under the “system and continuity” test. For example, as Amberwork submitted, even after Weroc had defaulted on its own 60-day credit terms, there is no evidence of any penalty interest having been imposed. Nor is there any evidence in the correspondence or transaction documents of any repayment plan or scheme for accruing interest. It is also clear that Amberwork was not lending to all and sundry, but entered into a specific arrangement with Ronald/Weroc to exploit the specific business opportunity surfaced by Ronald, Roger’s long-term business acquaintance. In short, Amberwork was first and foremost a trading business and not a moneylending business.

56 For these reasons, neither the sham transaction defence nor the unlicensed moneylending defence succeeds. In my judgment, Amberwork and QA entered into the Transactions as a means of enabling trade financing, such that the goods could be sold and purchased on credit, ⁵⁶ with Amberwork taking a profit (this being the difference between what it paid to QA and what it invoiced Weroc). ⁵⁷

57 Contrary to the defendants’ submissions, this is not an absurd commercial arrangement. The defendants contend that this arrangement was commercially unnecessary, because if the goods were always ready to be collected from the factory once payment was made, and the usual course in sale and purchase agreements is for the purchaser to first make payment, then Weroc could have easily paid the factory using moneys received from its end customer. Alternatively, the Resellers could have simply agreed to accepting delayed payment from Weroc. On the defendants’ argument, either of these would have rendered Amberwork’s role nugatory. ⁵⁸

58 I disagree. These alternative scenarios respectively depend on end customers having in fact first made payment (instead of paying, for example, on credit in full or in part), and the Resellers being willing to accept delayed payment. In the event, neither of these has been established as being likely to occur. Instead, Roger gave evidence that based on his experience doing business in China, upfront payment, and sometimes even advance payment, was required before suppliers might be prepared to do business. ⁵⁹ This seems to be a credible explanation for how Ronald structured the Transactions too. In any event, even if either scenario envisaged by the defendants was likely, businesses can have a multitude of other reasons or other expenses for which obtaining credit would be commercially explicable.

59 I also accept Roger’s observation that it is not uncommon for trading companies to not reveal their suppliers, out of concern that downstream purchasers may bypass the trading company to purchase from its supplier directly. ⁶⁰ This is another possible reason for why the Transactions were structured in this manner. In response to this, the defendants point to other, earlier transactions involving Weroc and other Resellers, where the relevant factory’s address appears to have been identified. ⁶¹ They also point to an apparent concession from Roger that it would not have been easy for Amberwork to find substitute buyers for the same goods to undercut Weroc. ⁶² In my judgment, whether there was in fact a significant risk of Weroc being undercut in this case, and whether Weroc/Ronald has in fact succeeded in structuring and implementing the arrangements in a failproof manner, are not of critical importance. After all, businesses do not all approach risks in the same manner or respond to risks with the same effectiveness. What matters is whether these concerns shape the way business may be conducted by trading companies, such that it would have been unremarkable for Weroc to adopt similar practices. In my judgment, it appears likely that those in the trading business would generally be concerned with disclosing details of their suppliers, and it was unremarkable for Ronald to have also felt the same. I would add that even in an e-mail to QA dated 1 October 2019, Ronald had stated that “[Weroc] doesn’t wants [sic] you to know the exact factory location because that is a trade secret, we cannot reveal our source in case our customers approach them directly. [T]his is market practice.” ⁶³

60 Finally, the defendants challenge the existence of any contractual relationship between Amberwork and QA, on the primary basis that there was no offer and acceptance or that parties were not ad idem. ⁶⁴

61 This challenge has been mounted on two levels, each framing the “offer” and “acceptance” slightly differently:

(a) First, on the premise that the contracts were either formed through QA’s quotations (offer) and Amberwork’s purchase orders (acceptance), or Amberwork’s purchase orders (offer) and QA’s Invoices (acceptance), the difficulty is that the First Transaction’s purchase order was only sent on 10 October 2019. ⁶⁵ This was after the date of contract formation pleaded by Amberwork of 10 September 2019. As for the Second Transaction, no purchase order was even sent to begin with. ⁶⁶

(b) Second, on the alternative premise that the Invoices constituted the purported offers, the defendants highlight that Amberwork’s pleaded case was that payment was made “pursuant to the Agreements” [emphasis added]. This referred to the “agreements dated 10 September 2019 and 26 September 2019 … on terms that were stated on the [Invoices]”. ⁶⁷ The defendants argue that if payment had been made “pursuant” to already-established contracts, then this must mean that Amberwork had conveyed its acceptance of the Invoices prior to making payment. But there was no such prior communication of acceptance from Amberwork. ⁶⁸

62 I reject both arguments, and find that Amberwork has proven on balance that the Agreements were formed.

63 At the outset, I note that the defendants did not plead that there was no contract because there was no offer and acceptance or because parties were not ad idem. Such a pleading would, if it had been made, naturally proceed to identify and particularise what the purported “offer” and “acceptance” were, and why these did not reflect a meeting of minds. Instead, what was pleaded at paragraph 14 of QA’s Defence was that there was no contract involving a sale between Amberwork and QA, that any sale was instead between Amberwork and Weroc, and that this arrangement was a sham. ⁶⁹

64 Leaving this observation aside, my starting point is that, as a general matter, the courts will lean in favour of finding that a contract exists between commercial parties so as to not defeat the expectations of commercial men. As the Court of Appeal held in Gay Choon Ing v Loh Sze Ti Terence Peter and another appeal [2009] 2 SLR(R) 332 at [63], in the context of addressing a battle of forms, “[w]hat is required … is a less mechanistic or dogmatic application of [the concepts of offer and acceptance] and this can be achieved by having regard to the context in which the agreement was concluded” [emphasis in original omitted]. I also take guidance from the following summary by V K Rajah JC (as he then was) in Midlink Development Pte Ltd v The Stansfield Group Pte Ltd [2004] 4 SLR(R) 258 at [52]:

65 I begin with the argument described at [61(a)] above. I reject it as Amberwork’s case on contractual formation was never based on the purchase orders. Indeed, as a matter of principle, it need not be. In this regard, the law on offer and acceptance prioritises substance over form, and in the main, does not categorically require that a party expresses his contractual assent in one particular form or other. Neither was the presentation of purchase orders fundamental to these parties, based on their objectively-ascertained intentions. If it had been, one would not have expected to see QA proceed to issue the Invoices even in the absence of purchase orders. It is relevant to note Roger’s testimony that it was not unusual for purchase orders not to be issued. Purchase orders were “good to have … to countercheck for [Amberwork’s] own in-house bookkeeping purpose”, but there could be several reasons why a purchase order would not be issued in a given transaction, for instance because pre-existing arrangements had been made on the “back-end”, or because circumstances of urgency left no time for the purchase order to be prepared. ⁷⁰ Pauline’s evidence was likewise that what was important to securing bank financing and internal bookkeeping were the invoices and signed delivery orders, and not the purchase orders. ⁷¹ In my judgment, an artificial and rigid insistence on the prior issuance of purchase orders being a necessary formal step would be at odds with the objective intentions of parties and the commercial realities underlying how they transacted and contracted.

66 The second form of QA’s submission (described at [61(b)] above) appears to be a more direct response to Amberwork’s pleaded case. However, I am unable to agree that the implication of Amberwork’s pleadings is that Amberwork’s contractual acceptance had to come in some separate communication prior to making payment. What Amberwork pleaded was that the Agreements were dated 10 September 2019 and 26 September 2019. In my judgment, that is consistent with the Agreements being constituted in part by the Invoices sent by QA on those dates (offer) and the payments made by Amberwork on those Invoices on those same dates (acceptance).

67 There is also nothing remarkable about Amberwork describing its payments as being made “pursuant to the Agreements”. Particularly in contracts constituted in part by conduct, it is entirely logical and likely that conduct (such as the making of payment) could simultaneously represent contractual assent and contractual performance. Furthermore, in stating that the payments had been made “pursuant to the Agreements”, Amberwork’s emphasis does not appear to be on the timing of contract formation. Rather, viewed in context, its point was that the payments had been made in accordance with the Invoices’ terms – that they were of the contracted sum and made in the contractually stipulated manner. Put another way, the point was that they had been made in connection with the Agreements, and not in connection with some other arrangement or basis.

68 Pleadings aside, the contemporaneous evidence positively supports the existence of the offers in the Invoices being accepted by way of payment by Amberwork. The arrangements were such that:

(a) Ronald would contact Roger with the details of upcoming transactions and Roger would acknowledge the same; ⁷²

(b) QA would issue its quotations ⁷³ and invoices; ⁷⁴ and

(c) Amberwork would make payment of the sums stated. ⁷⁵

69 At the very least, this would constitute acceptance by conduct.

70 I also reject paragraph 20 of the defendants’ closing submissions, that there was no consensus ad idem because the Invoices did not indicate who was to take or make delivery of the goods, or where the factory from which the goods were to be collected was located. In my judgment, it was clear from both Invoices that QA took on the obligation to supply the stated items. In exchange, Amberwork had an obligation to pay for these goods. As must be the case with most manner of business, the precise details of how these obligations were to be performed and operationalised could be worked out separately. Even if some other entity or agent would make delivery or take delivery of the goods (for instance, as an agent), the legal obligation to perform would simply still rest with the contracting parties. It is the legally enforceable obligation which parties ultimately contract for.

71 To the extent that the defendants’ argument was instead one of the completeness or certainty of terms ⁷⁶ – that the identity of the party making or taking delivery had not been made apparent in the contract – they have not shown that this would have constituted a material term without which no contract could be established. As Amberwork submitted, citing Dukkar S.A v Thailand Integrated Services Pte Ltd [2015] SGHC 234 (“Dukkar”) at [6], the essential terms for any contract for the sale of goods are the identity of the parties, the price, and the specification of the goods. These were all stipulated in the Invoices. Notably, delivery was not envisaged to be carried out via delivery to a named party, but to be ex-factory, ie, made available by the seller for collection at the factory, with the buyer making its own arrangements for collection: see PT Sandipala Arthaputra v STMicroelectronics Asia Pacific Pte Ltd and others [2017] SGHC 102 at [239].

72 The defendants then say that the lack of a specific address for delivery renders the contracts uncertain. They attempt to distinguish Dukkar, as the putative contracts in this case envisage not only sale but also delivery. The problem is that delivery was to be at a location unknown to both parties. As such, performance cannot be enforced and this renders the contract uncertain. ⁷⁷ In my judgment, however, this argument is entirely contrary to the parties’ arrangements. The precise location of the factory was never once a concern to the parties, and the parties were prepared to fulfil their respective ends of the bargain on that basis. This argument is, in my view, plainly an afterthought. As emphasised in Chan Tam Hoi (alias Paul Chan) v Wang Jian and other matters [2022] SGHC 192 at [100]–[101], the law is generally anxious to uphold a contract and it usually takes a rather uncertain or incomplete contract for a court to make a finding of unenforceability (see similarly Andrew Phang Boon Leong et al, The Law of Contract in Singapore (Andrew Phang Boon Leong gen ed) (Academy Publishing, 2nd Ed, 2022) at para 03.206).

73 Likewise, to the extent that the defendants take issue with the Invoices not reflecting the “full picture”, ⁷⁸ particularly the existence of other legs in the overall commercial arrangement, it is hard to see how this affects the existence of the contracts governing this leg of the arrangement.

74 For all these reasons, I hold that QA was contractually obliged to deliver the goods as pleaded by Amberwork. QA was not merely acting as a billing party or payment agent. Neither were the Transactions a sham or unenforceable by virtue of s 14(2) of the MLA.

75 For completeness, I should note that the defendants have objected to Amberwork’s reliance on extrinsic evidence for the purposes of contractual interpretation. ⁷⁹ It suffices for me to state that the present dispute was never about competing interpretations of any particular term. The rules concerning extrinsic evidence are different in the context of ascertaining the existence of a contract. This was clearly explained by the Court of Appeal in The “Luna” and another appeal [2021] 2 SLR 1054 at [38]:

76 Having found that QA was contractually obliged to deliver the goods to Amberwork and that the Transactions are not void and unenforceable, the next issue is whether QA breached its obligations.

77 Having carefully considered the available documentary and oral evidence, I find on a balance of probabilities that the goods were more likely than not taken by Weroc or its customers in China. Significantly, given the nature of the arrangements agreed to by Weroc, QA and Amberwork, delivery by Amberwork to Weroc (or its customer) ex-factory in China would simultaneously mean, in this case, that QA would also have fulfilled its delivery obligations to Amberwork under the Agreements. This is because the Transactions were essentially structured as part of back-to-back transactions where Weroc would sell to QA, QA would resell the goods to Amberwork (with payment by Amberwork immediate upon invoice issuance, by telegraphic transfer), followed by Amberwork selling the goods back to Weroc (on 60-day payment terms). Delivery in all three legs would always be ex-factory in China.

78 Amberwork’s recourse, then, is properly against Weroc and not QA. It is not disputed that Weroc has failed to pay Amberwork. The reason for Weroc’s non-payment is not clearly before me, but this is not an issue because as far as Amberwork’s action against QA is concerned, it is ultimately unnecessary for me to reach a conclusion on why Weroc did not pay Amberwork.

79 The legal burden lies on Amberwork to prove, on the balance of probabilities, that QA failed to deliver the goods to Amberwork. Amberwork focuses much of its case on QA’s inability to produce a signed delivery order on QA’s letterhead, as evidence of proof of collection. ⁸⁰

80 The difficulty with this argument, at least in relation to the Second Transaction, is that QA has sent a delivery order to Amberwork. The reason there is no signed order is because Amberwork did not provide its signature, even after it was requested to provide one. ⁸¹ Pauline conceded, in the context of a transaction involving another Reseller, that it ought to be for the addressee to sign on the delivery order to acknowledge receipt of goods. ⁸² This would be Amberwork. Although QA did not chase for Amberwork’s signature, Amberwork cannot rely on its own failure or refusal to sign the delivery order as evidence that there was a failure by QA to deliver the goods to Amberwork.

81 Pauline’s evidence was that Amberwork did not sign QA’s delivery order because Weroc did not similarly “commit to sign” Amberwork’s delivery order. ⁸³ In my view, this was not a credible or logical explanation.

82 I disagree with Amberwork’s submission that the defendants have obfuscated the role of each party in relation to the signing of delivery orders. ⁸⁴ Leaving aside what may have been the ordinary case in other transactions, there is in my view no reason why QA’s delivery order would first have to be signed before Amberwork’s delivery order could be signed by Weroc. First, a signed delivery order is merely evidence of delivery, and not performance of a contractual obligation as such. Second, Amberwork’s submission contradicts the evidence on the Second Transaction where Amberwork declined to sign QA’s delivery order because Weroc “did not commit to sign” ⁸⁵ Amberwork’s delivery order. It appears to me that Amberwork was seeking to have its cake and eat it too. But, more fundamentally, Pauline’s explanation is untrue. The evidence, as Roger acknowledged, ⁸⁶ showed that Ronald did say that he would arrange to have the delivery order signed, and asked Amberwork to prepare a delivery order on its letterhead which Weroc would then sign to confirm that it accepted the goods in good condition. This can be seen in this WhatsApp exchange between Roger and Ronald on 26 September 2019, which is the date of the Second Invoice: ⁸⁷

83 Put plainly, Ronald had already said he would arrange for Weroc to sign the delivery order, so it was unclear what further “commit[ment]” Pauline or Amberwork required.

84 The WhatsApp exchange between Roger and Ronald goes on to suggest that Amberwork did send a delivery order to Ronald/Weroc. Roger had said, “DO send to Weroc already, pls sign & stamp and email back”. ⁸⁸ A delivery order dated 26 September 2019, on Amberwork’s letterhead and addressed to Weroc, also appears in the evidence. ⁸⁹ However, as Amberwork’s witnesses confirmed during the trial, the reality is that no delivery order was in fact sent by Amberwork. After the commencement of this suit, in response to a request from the defendants’ solicitors, Amberwork’s solicitors indicated that Amberwork did not have in its possession, custody or power any e-mail or other correspondence enclosing this delivery order. ⁹⁰ On the stand, Pauline conceded that she did not send the delivery orders for the Transactions from Amberwork to Weroc. ⁹¹ Instead, while she “didn’t send the DO to Weroc”, her position was that she “could send him anytime upon request”. ⁹² It is difficult to understand why this should have been on Weroc’s request, when Weroc was the buyer vis-à-vis Amberwork. Neither is there evidence of Amberwork chasing Weroc to sign any delivery order from Amberwork. There was therefore no factual basis, even having regard to Pauline’s expressed concerns, for Amberwork to refuse to sign the delivery order issued by QA.

85 If Amberwork’s point was instead not the absence of a signed delivery order from Weroc as such, but a delivery order signed specifically by Weroc’s customer, this concern was not apparent from the contemporaneous evidence. Nor was it a point mentioned in its pleadings or developed in its closing submissions.

86 I would also note that Weroc appears to have sent a signed delivery order back to Pauline for another transaction involving a different entity (Satoo Comtech) on 26 September 2019, ie, in the same period of time as the Transactions. While this is not conclusive to my findings in respect of the Weroc-Amberwork delivery orders, it serves to reinforce the impression that Weroc was prepared to sign on Amberwork’s delivery orders. ⁹³

87 As for the First Transaction, there is an e-mail from Pauline to Sandra stating that Amberwork needed the delivery order for the bank to approve its application for financing. This e-mail was dated 10 September 2019, 1.55pm. However, this was before payment on the First Invoice was made at around 2.40pm on the same day, and before the fact of payment was confirmed via e-mail by Pauline at 2.58pm. ⁹⁴ Following the 1.55pm e-mail, there was no further request from Amberwork for a signed delivery order. This undermines Amberwork’s claim that it placed importance on receiving delivery orders as evidence of delivery.

88 As between Amberwork and Weroc, there is in evidence a delivery order addressed to Weroc on Amberwork’s letterhead. ⁹⁵ However, the date printed on the delivery order – “18/09/2019” – had a handwritten correction from “18” to “10” (ie, changing the date to 10 September 2019). Crucially, there is no evidence of this having been sent to Weroc, be it on 10 September, 18 September, or some other date. For instance, there is evidence of an e-mail dated 10 September 2019 from Ronald to Amberwork apparently attaching Weroc’s purchase order, ⁹⁶ and another e-mail dated 16 September 2019 from Amberwork to Daniel apparently enclosing an invoice, ⁹⁷ but there is no similar covering e-mail for the delivery order.

89 In any case, and more fundamentally, the circumstances as to the issuance and signing of the delivery orders are not conclusive of whether delivery in fact occurred. No doubt, they appear important to Amberwork for Amberwork’s own book-keeping purposes, ⁹⁸ and to ensure that bank financing could be smoothly obtained by having the documents in order. ⁹⁹ But that is a different issue from whether delivery is contingent on or proven by delivery orders being issued. As I indicated above at [82], delivery orders merely provide a way of evidencing that delivery did occur. The court can and must consider all relevant evidence (or the absence of evidence as the case may be) to conclude whether delivery did in fact take place.

90 In this regard, it is notable that there is no evidence of any allegation of non-delivery being made at the material time by either Weroc or Amberwork. The tone and contents of the relevant WhatsApp exchanges between Roger and Ronald, and between Ronald and Sandra, do not give any impression that Weroc or its customer had not collected the goods. The correspondence centred on Amberwork or Roger chasing for payment under Amberwork’s invoices, and Ronald responding with various reasons for requesting more time to pay. ¹⁰⁰ Roger testified that he continued to trust Ronald, and was under the impression that Ronald/Weroc had not only obtained the goods but had also on-sold them to another customer, but was unable to pay Amberwork due to difficulties with his/Weroc’s bank. ¹⁰¹ As far as Roger was concerned, Ronald never gave any reason for Roger to believe the goods were not collected, and Roger “assumed the goods [were] collected based on the WhatsApp messages”. ¹⁰² Pertinently, Weroc never once said that it did not receive the goods. ¹⁰³ There is also no evidence of Roger asking Ronald whether he was unable to collect the goods. Pauline’s evidence was similarly that she had no reason to believe that Ronald/Weroc never obtained the goods. ¹⁰⁴ The overall tenor of the evidence is that the goods had been delivered or collected at the material time (ex-factory in Shenzhen), and that the only issue that remained was Weroc’s obligation to pay Amberwork under the second (Amberwork-Weroc) leg of the back-to-back transactions within 60 days. The same point can be made in relation to Amberwork’s issuance of a Statement of Account to Weroc for outstanding invoices and threats of legal action. ¹⁰⁵ Had delivery of the goods not taken place (which, as I explained above at [77], would occur simultaneously under the QA-Amberwork leg and the Amberwork-Weroc leg), there would be no basis upon which Amberwork could have regarded itself as being entitled to chase Ronald or Weroc for payment.

91 This is an important point to be underscored. If Amberwork took the view that it was entitled to payment from Weroc, it must be because it also took the view that it had fulfilled its delivery obligations to Weroc under its leg of the transactions. Following from that, QA must also have fulfilled its delivery obligations under its leg of the transactions to deliver the goods to Amberwork. Otherwise, if QA had not delivered, Amberwork would in turn be unable to deliver to Weroc, and it would have had no basis to seek payment from Weroc. Viewed in this light, Amberwork’s conduct was, in my judgment, ultimately more consistent with QA having fulfilled its obligations to Amberwork to deliver the goods.

92 Ultimately, inasmuch as the first (QA-Amberwork) and second (Amberwork-Weroc) legs of the transactions involved distinct contracts, the reality remained that delivery of the goods under both legs would occur simultaneously. Amberwork knew that the goods would be collected from the factory in China almost immediately. It was in these circumstances, then, that there was no communication from Amberwork to QA asserting that QA had not delivered the goods to Amberwork under the first leg, nor any assertion by Weroc or Ronald that the goods for the Transactions had not been delivered by Amberwork in respect of the second leg. It was only on 19 March 2020, close to six months after the Transactions, that Roger e-mailed Sandra for the first time, complaining about the alleged delay in delivery of the contracted goods, and asking QA to cancel the allegedly outstanding orders. As Sandra’s reply of 21 March 2020 noted, this was the first time QA received an e-mail from Roger stating that the goods had not been delivered. ¹⁰⁶

93 On the whole, based on the evidence adduced (and not adduced), I find on balance that the more reasonable and plausible conclusion to be reached is that the goods were delivered to Weroc or its customers in China, but Weroc failed to pay Amberwork. More precisely, this means that Amberwork has not met its burden of proving that the goods had not been delivered to it by QA. To be sure, the evidence from both parties has been unsatisfactory in some respects, in large part because Ronald has passed on (see [4]). Nevertheless, the legal burden is and remains on Amberwork to prove that QA breached its contractual obligations to Amberwork. In my judgment, Amberwork has failed to meet this burden.

94 I find that the Transactions were not shams but legitimate commercial transactions involving Amberwork extending trade financing to Weroc via back-to-back transactions with QA. While the arrangements may have involved an element of round-tripping, that does not, without more, make the Transactions shams. Neither were they devices to circumvent the MLA; there was no unlicensed moneylending in this case.

95 Although I find that Amberwork has established that QA was contractually obliged as a seller to deliver the goods that formed the subject matter of the Transactions, Amberwork has failed to prove that QA breached this obligation. Amberwork’s claim for breach of contract thus fails and for this reason, the claim by Amberwork is dismissed.

96 Given my finding that there were validly concluded contracts between Amberwork and QA, there is no need for me to decide on Amberwork’s alternative claims grounded on misrepresentation and dishonest assistance. Those claims, as framed by Amberwork in its Statement of Claim, were predicated on the court finding that there were no concluded contracts between Amberwork and QA.

97 In the light of my conclusions on Amberwork’s contract claim, it is also unnecessary for me to decide on QA’s defence of unilateral mistake. However, even if this had been necessary for my decision, this defence would have failed in any event. The pleaded mistake is not made out on the facts as I have found, and in any case, contradicts QA’s defence that the Transactions were shams.

98 I shall hear the parties separately on costs.