Baizanis, Georgios v Snap Innovations Pte Ltd and another [2024] SGHC 200

The Plaintiff, Georgios Baizanis, is a cryptocurrency investor. The 1st Defendant (“D1”), Snap Innovations Pte Ltd, is a company providing information technology and computer-related services, with a focus on technological solutions for brokers and traders. Its sole shareholder is Dr Ting Shang Ping (“Dr Ting”). At the material time, the 2nd Defendant (“D2”), Bernard Ong, was reflected on D1’s website as a “director” of D1, although he was not registered as such with the Accounting and Corporate Regulatory Authority (“ACRA”).

This case arose from the Plaintiff’s investment in a scheme known as “Cryptotrage”, which involved arbitrage trading in cryptocurrencies on the Binance Exchange. According to the Plaintiff, the Cryptotrage scheme was operated by D1 through its employees in Vietnam. The Plaintiff claimed that pursuant to the scheme, he deposited cryptocurrencies into accounts on the Binance Exchange that were operated by D1’s office in Vietnam. The Plaintiff professed that he initially invested only small amounts in the scheme, as he wanted to limit his exposure should any governance failures within D1 occasion the misappropriation of client funds. The Plaintiff alleged that he was nevertheless enticed into increasing his investments by one Wu Zhongyi (“Zee”), who was reflected on D1’s website at the material time as D1’s director in Vietnam. Zee did so by offering the Plaintiff what appeared to be a corporate guarantee by D1, which would indemnify the Plaintiff against losses arising from fraud. The corporate guarantee was embodied in a document called the “Service Agreement”, which was allegedly signed on D1’s behalf by both D2 and Zee. After obtaining the comfort of the Service Agreement, the Plaintiff then increased his investments in the Cryptotrage scheme.

On 9 February 2021, Zee disappeared, apparently after having misappropriated the cryptocurrencies deposited by investors (including the Plaintiff) in the Cryptotrage scheme. When the Plaintiff attempted to call on D1’s indemnity under the Service Agreement, D1 denied responsibility for the Cryptotrage scheme (which D1 claimed was entirely Zee’s operation). D1 also disavowed the Service Agreement, saying that D2 and Zee were both “independent contractors” of D1 with no authority to sign the Service Agreement on D1’s behalf. As for D2, he denounced the Service Agreement as a forgery, claiming that his signature had been copy-pasted on it.

The Plaintiff commenced this action against D1 and D2 (whom I collectively refer to as the “Defendants”) to recover the value of the cryptocurrencies which he claimed to have invested in the Cryptotrage scheme and lost through Zee’s fraud. The Plaintiff raised various causes of action:

(a) He sued D1 for breach of the Service Agreement, claiming that the corporate guarantee therein obliged D1 to indemnify the Plaintiff for the losses arising from Zee’s fraud. The Plaintiff’s cause of action was premised on his contention that both D2 and Zee had actual, or at least ostensible, authority to enter into the Service Agreement on D1’s behalf.

(b) In the event of the guarantee in the Service Agreement being unenforceable as against D1, the Plaintiff sought in the alternative to hold D2 liable for a breach of warranty of authority.

(c) The Plaintiff also sued both Defendants for allegedly breaching their duty to properly supervise Zee, thereby allowing the latter to perpetrate the fraud that led to the Plaintiff’s loss.

I dismiss the Plaintiff’s claims and set out my reasons for doing so.

The Facts

The bulk of the factual narrative below emanated from the Plaintiff. This is unsurprising, given the Defendants’ position that they had no dealings with the Plaintiff prior to Zee’s disappearance.

The Plaintiff’s factual evidence

The Plaintiff’s account of what transpired can be canvassed in three parts, namely:

How the Plaintiff came to invest in the Cryptotrage scheme

The Plaintiff explained that he came to be acquainted with Zee sometime around the end of December 2018, through the messaging platform, Telegram. Zee was based in Vietnam, where he and his team managed the Cryptotrage scheme. The scheme involved the generation of profits from arbitrage trades in cryptocurrencies. Those trades were performed by bots, ie, trading programmes which communicated with the cryptocurrency exchange on which the trades were performed. Traditionally, the process of searching for arbitrage opportunities was a manual one – this was both tedious and time-consuming. With the advent of bots, the process became easier and incurred significantly reduced latency. In the months following his acquaintance with Zee, the Plaintiff began investing small amounts in the Cryptotrage scheme.

The statement of accounts for the Plaintiff’s investments in the Cryptotrage scheme came in the form of “daily reports”. These reports comprised tabulated numbers that had apparently been prepared by Zee’s team in Vietnam and were disseminated via Telegram messages. The daily reports contained little by way of elaboration as to what the numbers in the tables meant, or how they were arrived at. However, this appeared to pose no issues for a seasoned cryptocurrency investor such as the Plaintiff.

The Plaintiff’s initial foray into the Cryptotrage scheme seemed profitable, prompting him to contemplate increasing his investments as well as bringing his friends on board as fellow investors in the scheme. However, before embarking on that commitment, the Plaintiff thought it prudent to conduct due diligence on Zee and ascertain exactly who Zee worked for. The Plaintiff thus posed his inquiries to Zee, who responded that he was a director of D1. Zee explained that he and his team, which was known as “Snap Vietnam”, operated out of Vietnam. Zee also sent the Plaintiff a document called the “Deck”, which contained a set of presentation slides that listed Snap Vietnam’s key personnel and explained the Cryptotrage scheme. The first substantive page of the Deck, which described the credentials of both Zee and one Rick Nguyen (“Rick”), is extracted below:

To confirm the veracity of the information within the Deck, the Plaintiff decided to reach out to D1 directly. On 7 February 2019, the Plaintiff called one Michael Lim, who was an employee of D1 at the time. Michael Lim verbally confirmed to the Plaintiff over the phone that Zee and Rick were working for D1. That same day, the Plaintiff followed up with an e-mail to Michael Lim, attaching a copy of the Deck and providing Michael Lim with an explanation of what the Cryptotrage scheme was about. In his e-mail, the Plaintiff expressed his concerns about how the Cryptotrage scheme was not mentioned anywhere on D1’s websites and explained that he consequently wanted to ensure that the persons he was dealing with (ie, Zee and Rick) were indeed persons from D1’s Vietnam office and not impersonators perpetrating a scam. The Plaintiff’s e-mail also asked if the Cryptotrage scheme was indeed a product of D1. An extract of the Plaintiff’s e-mail to Michael Lim is set out below:

Michael Lim replied to the Plaintiff’s e-mail a few hours later, with a rather short response confirming that Zee and Rick were from D1’s Vietnam office. Michael Lim’s e-mail is extracted below:

Michael Lim’s e-mail also re-attached the first substantive page of the Deck which set out the credentials of Zee and Rick (the contents of which have been extracted at para 10 above), which the Plaintiff had sent to Michael Lim in the preceding e-mail.

In March 2019, the Plaintiff visited Vietnam to conduct on-site due diligence in respect of Snap Vietnam’s trading operations. There, he met Zee and various other members of the Snap Vietnam team, including Rick, one Fong Chee Keong (“CK”), who was allegedly Zee’s right-hand man, and one Martin, who was allegedly Snap Vietnam’s chief trader. The Plaintiff claimed that these persons could no longer be traced at the time of trial.

According to the Plaintiff, he was told by Zee, Rick and CK that D1 was using a Vietnam-incorporated company, Click Staff Company Limited (“Clickstaff”), as a conduit for executing D1’s operations in Vietnam. The Plaintiff was introduced to the sole shareholder and director of Clickstaff, Nguyen Thanh, who also served as Snap Vietnam’s Human Resource Manager. During the trial, Nguyen Thanh testified remotely from Vietnam, giving evidence on the Plaintiff’s behalf. In doing so, Nguyen Thanh gave various examples of D1’s involvement in Snap Vietnam’s business, including the following:

(a) D1, through Snap Vietnam, ventured into and funded the development of trading bots for arbitraging cryptocurrencies.

(b) Dr Ting facilitated D1’s business in Vietnam by getting Clickstaff to purchase various equipment to facilitate Snap Vietnam’s operations.

(c) D1 also used Clickstaff to recruit staff for Snap Vietnam, with Zee personally negotiating the salaries of Vietnamese staff engaged by Clickstaff. While Clickstaff would pay those salaries, it was reimbursed by either Zee or CK, who made payment to Clickstaff on D1’s behalf.

The Plaintiff also claimed that Zee told him that D2 was Zee’s “boss and CEO”. Following from this, the Plaintiff performed searches on the internet and discovered that:

Signing of the Service Agreement

After a few weeks of investing in the Cryptotrage scheme, the Plaintiff was invited by Zee to increase his investments. The Plaintiff was told that this would enhance his financial returns, as higher investment levels meant higher profit-sharing percentages. However, the Plaintiff remained wary of doing so, as higher investment levels necessarily carried the risk of higher losses in the event of Snap Vietnam’s staff absconding with his cryptocurrencies. Zee, Rick and Martin sought to assure the Plaintiff that safeguards were in place to prevent unauthorised withdrawals. However, this did not suffice to assuage the fears of the Plaintiff, who refused to raise his investments unless D1 executed a corporate guarantee indemnifying him against losses which might arise from internal fraud. The Plaintiff was eventually informed by Zee that Zee himself could sign such a corporate guarantee on D1’s behalf. However, the Plaintiff told Zee that he also wanted an additional “authorised director” from D1 to sign the corporate guarantee. Zee thus informed the Plaintiff that he would get D2, who was his “boss” and “CEO”, to sign on the corporate guarantee as well.

The Plaintiff thus set about personally drafting the Service Agreement, the key terms of which are extracted below:

The operative clauses of the Service Agreement are highlighted below:

(a) Clause 1 contemplated that D1 (which was referred to by the Service Agreement as “Cryptotrage”) owed various obligations that would be performed through Clickstaff (which was referred to by the Service Agreement as the “child company” of D1).

(b) Clause 1.1 set out D1’s primary obligation, which was to “manage the funds” of the Plaintiff and “his Group” (ie, friends that the Plaintiff would co-opt to invest in the Cryptotrage scheme) through arbitrage trades in cryptocurrency.

(c) Materially, cl 1.2 embodied the corporate guarantee sought by the Plaintiff. This clause contemplated that in the event of “internal fraud” by any staff of the Cryptotrage group of companies, the “parent company” (which was presumably a reference to D1) had to replace any cryptocurrencies “stolen by fraud” within five business days.

The Plaintiff prepared a printed copy of the Service Agreement that he had drafted (which I will refer to as “the Original Draft”) and brought the same to Vietnam on 7 May 2019 to have it signed. At the Vietnam office, the Plaintiff met Zee, who scanned the Original Draft and printed out the scanned electronic file to produce what I will refer to as “the Scanned Draft”. The Plaintiff and Zee appended their wet-ink signatures on both the Original Draft and the Scanned Draft, with Zee signing in his capacity as “Director, Vietnam Operations”. The Plaintiff then took photographs of the signed Original Draft, which he exhibited in his Affidavit of Evidence-in-Chief (“AEIC”). As for the Scanned Draft, the Plaintiff surmised that after this was signed by Zee and himself, Zee had scanned the signed copy and sent the electronic copy to D2 via Telegram.

The Plaintiff also explained that D2 had been travelling to various countries at the time, but D2 eventually returned to Singapore where he signed the Service Agreement:

(a) On 21 May 2019, D2 travelled to Vietnam where he discussed the Service Agreement with Zee. The Plaintiff knew this because he had sent a Telegram message to Zee that day, asking about the status of the Service Agreement’s execution. Zee had replied that D2 was coming to Vietnam and would discuss the Service Agreement with Zee that very evening. Zee’s reply was corroborated by D2’s passport, which showed that D2 was in Vietnam on 21 May 2019.

(b) On 24 May 2019, the Plaintiff sent another Telegram message to Zee seeking a status update, to which Zee responded that he was following up with D2 on the matter. Zee further said that D2 was in Hong Kong and would sign the Service Agreement on Monday, 27 May 2019. Again, Zee’s response was corroborated by D2’s passport, which showed D2 to be in China from 25 to 26 May 2019.

(c) The Plaintiff’s case was that D2 ultimately returned to Singapore on 27 May 2019, where he signed the Service Agreement in the early hours of the morning, before leaving again for Malaysia. Again, this was consistent with D2’s passport, which showed that D2 left Singapore for Malaysia on 27 May 2019.

The Plaintiff contended that D2 physically executed the Service Agreement, by appending his wet-ink signature on it while he was in Singapore; there was no suggestion by the Plaintiff of D2 having signed the Service Agreement electronically. I will refer to the putative paper copy of the Service Agreement bearing D2’s wet-ink signature as the “Executed Paper Copy”.

On the morning that D2 purportedly executed the Service Agreement (ie, 27 May 2019), at 8.28am, Zee sent the Plaintiff a Telegram message forwarding what purported to be an electronic copy of the Executed Paper Copy bearing D2’s signature and D1’s corporate stamp, with an accompanying message that D2 had signed and stamped the Service Agreement. I will refer to this electronic copy of the Service Agreement, which bore D2’s signature and D1’s corporate stamp and which was forwarded by Zee that morning, as the “E-Copy”. In his message, Zee also offered to let the Plaintiff collect a duplicate copy of the signed Service Agreement from D1’s Singapore office. An extract of the Telegram message sent by Zee to the Plaintiff is set out below:

The foregoing extract shows two distinct messages in the chain, which I have demarcated into two separate boxes bounded by dashed lines – I refer to these as the “upper box” and “lower box”. From the extract, one sees that:

The words “Deleted Account” appeared at three points within the conversation chain – I have marked these in the extract above with the boxed numbers “1”, “2” and “3”. On the face of the extract, it is unclear if the three deleted accounts belonged to a single person. The Plaintiff’s IT forensic expert, Wilfred Nathan, testified that while the term “Deleted Account” may appear at various points of a Telegram conversation chain, the deleted account denoted by that term could potentially belong to different persons (each of whom have deleted their respective Telegram accounts) at different points where that term appears in the chain. Relying on Wilfred Nathan’s analysis, the Plaintiff advanced the following positions:

(a) Deleted Account “1”, which sent the message in the upper box to the Plaintiff at 8.28am, belonged to Zee.

(b) Zee had, by the message in the upper box, forwarded to the Plaintiff a message which Zee himself had received from Deleted Account “2”. That message from Deleted Account “2”, which had been sent to Zee just one minute earlier (as denoted by the timestamp “08:27:18”), enclosed the E-Copy.

(c) Deleted Account “3”, which sent the message in the lower box to the Plaintiff (also at 8.28am) was the same as Deleted Account “1”, ie, both were Zee’s account.

The key question thus arose as to whom Deleted Account “2” belonged. This was not a point that Wilfred Nathan was able to resolve by his analysis. However, it was the Plaintiff’s evidence that when he first viewed the conversation chain on 27 May 2019, Deleted Accounts “1” to “3” had yet to be deleted. At that point, the Plaintiff had noticed an icon with the letter “B” appearing in relation to Deleted Account “2”. The Plaintiff inferred that the icon denoted an account belonging to D2, whose first name (“Bernard”) began with the letter “B”. On that hypothesis, Deleted Account “2” must have belonged to D2, meaning that D2 had sent a message enclosing the E-Copy to Zee on 27 May 2019 at 8.27am (as denoted by the timestamp “08:27:18”). About a minute after that, at 8.28am, Zee had:

While the Plaintiff did not personally witness D2 executing the Service Agreement, he argued that the court should accept Zee’s message (in the lower box) as evidence that D2 did sign it. This was because Zee’s communications had all been corroborated by D2’s passport entries (see para 20 above). That in turn meant that Zee had intimate knowledge of D2’s travel itinerary and must consequently have known if and when D2 signed the Service Agreement.

Zee and the Plaintiff then arranged for the Plaintiff to collect a hard copy of the Service Agreement from D1’s office in Singapore. In turn, the Plaintiff asked his friend, Frederic Schmidt (“Schmidt”), to do so on his behalf.

On 2 June 2019, Zee sent a Telegram message to D1’s employee, Kenneth Chan, enclosing an electronic copy of the Service Agreement purportedly bearing D2’s signature. In that message, Zee asked Kenneth Chan to print the same and pass the printout to Schmidt. Kenneth Chan complied and passed the printout to Schmidt, when the latter turned up at D1’s office the next day. Schmidt subsequently passed the printout to the Plaintiff, who adduced the same in evidence as exhibit “P1”.

The Plaintiff’s position was that the E-Copy which Zee forwarded to him via Telegram on the morning of 27 May 2019 was identical to the electronic copy that Zee sent to Kenneth Chan on 2 June 2019. This meant that P1, being a printout of the electronic copy sent to Kenneth Chan, was for all intents and purposes also a printout of the E-Copy sent to the Plaintiff on 27 May 2019. As the Defendants offered no credible evidence to challenge the Plaintiff’s position on this point, I have proceeded on the basis that references to:

are all synonymous. Subsequent references to P1 and the “E-Copy” should therefore be understood in that light.

The most crucial section of P1 was its third page, a snapshot of which is extracted below:

As seen from the extract above, the third page bore the signatures of the Plaintiff and Zee (with Zee signing off as “Director, Vietnam Operations”) – the signatures would have been appended by Zee and the Plaintiff when the latter visited the Vietnam office on 7 May 2019 – see para 19 above. Critically, the third page also bore what appeared to be D2’s signature and, to the right of that, D1’s corporate stamp. The handwritten words “Bernard Ong. Managing Director, Snap Innovations (Singapore)” also appeared below D2’s purported signature and the corporate stamp.

The Plaintiff claimed that he was initially under the impression that P1 was the Executed Paper Copy, ie, bearing D2’s wet-ink signature, as D1’s corporate stamp was in colour and not black-and-white. He claimed to have discovered that P1 was a printout of an electronic copy only after this suit was commenced. The Plaintiff suspected that the Executed Paper Copy remains in the possession of either D1 or D2, who have both refused to pass him the same.

The Plaintiff professed that he always expected to receive an “original” copy of the signed Service Agreement, ie, where the signatures therein were in wet ink. While at the Vietnam office on 7 May 2019, he and Zee had taken the trouble to append their wet-ink signatures on both the Original Draft and the Scanned Draft (see para 19 above). This was to allow both the Plaintiff and D1 to each have an “original” document, ie, where the signatures were in wet ink. If D2 had thereafter appended his wet-ink signature on both these drafts, this would have given rise to two documents where the signatures of all three signatories (ie, Zee, the Plaintiff and D2) were in wet ink. From the Plaintiff’s perspective, it was important for him to possess an “original” document for his records. Thus, when Zee told the Plaintiff that he would send the signed Service Agreement to Singapore for execution by D2, the Plaintiff presumed that Zee would courier both the Original Draft and Scanned Draft (each bearing the wet-ink signatures of Zee and himself) to Singapore, where D2 could then pen his wet-ink signature and affix D1’s wet-ink stamp. As it turned out, Zee did not do that but (as alluded to at para 19 above) merely scanned one of the two drafts and sent the scanned electronic copy to D2. If D2 had printed that electronic copy and signed on the printout, the Executed Paper Copy would have borne the wet-ink signature of only D2 – the signatures of Zee and the Plaintiff would have merely been printed images. Unfortunately, the document which the Plaintiff eventually obtained (ie, P1) was not even that. It was a printout of an electronic copy where the signatures of all signatories (including D2) and D1’s corporate stamp were merely printed images and not in wet ink.

Theft of the Plaintiff’s cryptocurrencies and the aftermath

After the Service Agreement was purportedly executed, the Plaintiff and his friends began increasing their investments in the Cryptotrage scheme. The Plaintiff deposited three types of cryptocurrencies into the scheme: the first was USDt, followed by BUSD (from November 2019 onwards) and BNB (from July 2019 onwards).

At the time, Snap Vietnam had two accounts on the Binance Exchange which respectively held what was termed “Tier 2” and “Tier 3” status. Zee wanted to raise the quantum of BNB deposits in each of these accounts to 11,000 BNB, as crossing that threshold would qualify the accounts for upgrade to “Tier 9” status. Holders of Tier 9 accounts enjoyed large discounts on trading fees and received referral fees from the Binance Exchange. At Zee’s behest, the Plaintiff deposited a total of 22,000 BNB into the two accounts, thereby enabling them to reach Tier 9 status. In return, the Plaintiff was paid half the Tier 9 benefits reaped.

In the months that followed, the Plaintiff’s Cryptotrage trades were profitable and he was able to make various withdrawals of his gains. In October 2019, a company called Torque Group Holdings Limited (“Torque”) was incorporated in the British Virgin Islands (“BVI”), with D2 as its registered director and Chief Executive Officer (“CEO”) and Zee as its Chief Technology Officer (“CTO”). Zee informed the Plaintiff that Torque was a “spinoff” from Cryptotrage that operated an online cryptocurrency trading platform employing the same arbitrage algorithm as that used in the Cryptotrage scheme. The Plaintiff accordingly invested cryptocurrencies on the Torque platform as well.

The Plaintiff alleged that Snap Vietnam had, unbeknownst to him, started to consolidate its Cryptotrage trades such that they would be conducted from only one of the two Tier 9 Binance accounts – I refer to this remaining account simply as “the Binance Account”. The Binance Account also came to be used for trades on the Torque platform, meaning that cryptocurrencies invested by investors on the Torque platform were deposited into the Binance Account. The Plaintiff viewed this as “commingling” of the cryptocurrencies which he had invested in Cryptotrage with those invested by Torque’s investors – something which he had neither consented to nor even been made aware of. He surmised that the commingling was to allow Torque to reap the pecuniary benefits of the Binance Account’s Tier 9 status (referred to at para 33 above).

On 9 February 2021, Zee sent a Telegram message to D2, in which Zee confessed to engaging in unauthorised futures trading using cryptocurrencies in the Binance Account. Zee explained that the bots were unable to sustain the profits, thereby prompting him to resort to futures trading which, though more profitable, was also risky and ultimately incurred losses that grew progressively bigger. The Plaintiff claimed that Zee had also absconded with the Plaintiff’s cryptocurrencies from the Binance Account. Following this development, D2 issued a statement to Torque’s investors, alerting them to significant trading losses incurred by Torque arising from Zee’s trades. Zee has been untraceable since.

The Plaintiff claimed that up until the unravelling of Zee’s fraud, he was able to make withdrawals from what he had deposited in the Cryptotrage scheme. This lulled him into thinking that there was nothing amiss and that the daily reports (mentioned in para 9 above) showing the daily balances of his cryptocurrency deposits were in order. After Zee’s actions came to light, the Plaintiff realised that those daily reports were “completely bogus”, as Zee had been using the Plaintiff’s cryptocurrencies to engage in risky and highly-leveraged trades. The Plaintiff reckoned that Zee’s fraud must ultimately have led to the depletion of all the cryptocurrencies which the Plaintiff had deposited under the Cryptotrage scheme.

As investigations into Torque commenced, the Binance Account was frozen until further notice, thereby preventing the Plaintiff from withdrawing whatever might have been left of his investments under both the Cryptotrage scheme and the Torque platform. It was only at this point that the Plaintiff realised that the Binance Account had been used by Torque. The Plaintiff thus approached CK, urgently seeking redress. CK arranged a conference call with D2, which was held on 16 February 2021 and attended by D2, the Plaintiff and CK. The call was secretly recorded by the Plaintiff, who adduced transcripts of the call at trial. The transcripts capture D2 saying that he could not recall signing the Service Agreement. The transcripts also reflect that when the Plaintiff persisted in pressing for a remedy, D2 capitulated and suggested that a repayment plan could be explored.

On 26 February 2021, D2 applied to the BVI courts for an order to appoint liquidators for Torque on an urgent basis. This led to the appointment of Torque’s provisional liquidators who, on 3 March 2021, sent a circular to Torque’s creditors and customers notifying them of the appointment and informing them that the Binance Account was now under the liquidators’ control.

The Defendants’ factual evidence

D2 maintained that he never executed the Service Agreement. In other words, the Executed Paper Copy never existed, which in turn meant that the E-Copy (and hence P1) could not possibly have been derived from it. D2 maintained that the E-Copy (and hence P1) was forged and made the following points about the features in P1’s third page (extracted at para 29 above):

D2 claimed that he came to learn about the Service Agreement only at the conference call with the Plaintiff on 16 February 2021 (referred to at para 38 above), after Zee’s disappearance. D2 also made a police report about P1 having been forged.

In relation to D1’s corporate stamp, Dr Ting corroborated D2’s position by explaining that D2 had no access to it, and thus could not have applied the stamp to the Service Agreement. D1’s corporate stamp was always securely kept in the Operation Manager’s room under lock and key, accessible only to two persons, being the Operation Manager and Dr Ting himself. Dr Ting also added that the corporate stamp was an ordinary rubber stamp which could be procured in “a day or two”, meaning that anyone other than D2 could have made a similar-looking stamp and applied the same to the Service Agreement.

The Defendants also explained that while D2 and Zee may have been given the titles of “director”, they were in truth merely independent contractors engaged by D1. Dr Ting explained that it was D1’s practice to give “nice-sounding corporate titles” to its independent contractors, so that they would “appear credible” when promoting D1’s products and services to potential customers. This was corroborated by D2, who explained that both he and Zee were independent contractors who had been conferred the title of “director” to enable them to perform their duties.

Framing of the Issues

Having laid out the parties’ respective versions of the facts, I now summarise the claims raised by the Plaintiff, and the Defendants’ defences thereto. That will in turn set the backdrop for framing the issues in this case.

The Plaintiff’s claims against D1

The Plaintiff submitted that both D2 and Zee had both actual and ostensible authority to enter into the Service Agreement on D1’s behalf and that D1 was consequently bound to honour its terms.

In relying on the Service Agreement, the principal clause which the Plaintiff sought to invoke was cl 1.2, the relevant portion of which read:

As seen from the clause, D1’s obligation to reimburse the Plaintiff would be triggered in the event of “internal fraud” by staff, where any cryptocurrencies were “stolen by fraud”. The Plaintiff maintained that two occurrences in the present case amounted to fraud falling within the purview of cl 1.2:

The Plaintiff claimed that following the events above, D1 was obliged under cl 1.2 to indemnify him for the quantum of cryptocurrencies reflected in the daily report issued on “the day before … such adverse event”. As Zee’s fraud came to light on 9 February 2021, the Plaintiff took this to be the day on which the adverse event occurred, meaning that the daily reports to be used for ascertaining the size of D1’s indemnity obligation were those issued the day before, ie, on 8 February 2021. The 8 February 2021 daily reports in turn set out the Plaintiff’s balances as at the end of 7 February 2021. According to these daily reports, the Plaintiff held balances in three cryptocurrencies: USDt, BUSD and BNB. He listed the balance for each of these cryptocurrencies as at the end of 7 February 2021, as reflected in the 8 February 2021 daily reports, and ascribed a US dollar value to each balance, as per the table below. He then summed up all three dollar values to peg his total loss at US$9,122,044:

The Plaintiff claimed that D1’s refusal to honour cl 1.2 of the Service Agreement and indemnify him for this sum, being the Plaintiff’s loss from the fraud, constituted a breach of contract.

Apart from relying on D1’s alleged breach of the Service Agreement, the Plaintiff also claimed that D1 owed a “fiduciary, contractual or tortious duty” to the Plaintiff to supervise Zee. The Plaintiff claimed that D1 failed to do so, thereby allowing Zee’s fraud to injure the Plaintiff.

The Plaintiff’s claims against D2

The Plaintiff also claimed against D2 for breach of warranty of authority. In this respect, the Plaintiff’s case was that:

(a) D2, in holding himself out on his LinkedIn profile as D1’s “Managing Director”, as well as in allowing himself to be held out on D1’s website as D1’s “director”, warranted that he had the authority to act on D1’s behalf;

(b) the Plaintiff entered into the Service Agreement in reliance on that warranty; and

(c) D2 breached that warranty by now denouncing the Service Agreement.

While not explicitly framed as such in his pleadings or submissions, this claim for breach of warranty of authority was advanced as an alternative to the Plaintiff’s principal claims against D1. In other words, it would only be if the Plaintiff failed to establish that D2 (or Zee, for that matter) had the authority to bind D1 to the Service Agreement that the issue of breach of warranty of authority by D2 would come into play.

As with D1, the Plaintiff similarly claimed that D2 owed the Plaintiff a duty to supervise Zee. The Plaintiff claimed that this was a “tortious duty”, arising from D2’s capacity as D1’s managing director and Zee’s supervisor in Snap Vietnam, as well as from D2’s capacity as CEO of Torque. The Plaintiff claimed that D2 breached this duty by the following derelictions:

(a) D2 delegated to Zee the authority to run and manage Snap Vietnam, thereby allowing Zee to surreptitiously misappropriate the Plaintiff’s cryptocurrencies.

(b) D2 allowed Zee to commingle the cryptocurrencies which the Plaintiff had deposited for the Cryptotrage scheme with those of Torque’s investors. As Torque’s CEO, D2 should have suspected that something was amiss upon discovering cryptocurrencies in the Binance Account belonging to persons other than Torque’s investors.

The defences

In furtherance of their position that the Executed Paper Copy (bearing D2’s wet-ink signature) never existed, the Defendants emphasised that the Executed Paper Copy was never adduced in court and argued that the Plaintiff should consequently be regarded as having failed to prove it. As for P1, this was a printout of the E-Copy, which in turn purported to be a copy of the Executed Paper Copy. The Defendants argued that the Plaintiff failed to discharge his burden of establishing the E-Copy as an authentic copy of the Executed Paper Copy. With authenticity not having been established, the E-Copy (and hence P1) was consequently inadmissible as evidence of the Executed Paper Copy.

D1 further submitted that even if D2 had executed the Service Agreement, neither D2 nor Zee possessed actual or ostensible authority to sign it on D1’s behalf. Notwithstanding their titles of “director”, both men were in truth merely independent contractors with no authority to bind D1 to the Service Agreement.

As regards the Plaintiff’s claim that D2 breached his warranty of authority, D2 maintained that he never professed to be an agent of D1. In any case, no such warranty could have been made by him to the Plaintiff, given that both men never met prior to the signing of the Service Agreement. D2 pointed out that while the Plaintiff may have culled some details about him from online sources, the Plaintiff made no serious effort to verify them with proper sources.

D1 also argued that even if the Service Agreement was validly concluded on its behalf, the guarantee in cl 1.2 was never triggered. This clause would have come into operation only in instances of “internal fraud by any staff of the Cryptotrage group”, where digital assets were “stolen by fraud”. D1 argued that both instances of fraud alleged by the Plaintiff – (a) the commingling of the Plaintiff’s cryptocurrencies with those of Torque’s investors; and (b) Zee’s theft of the cryptocurrencies – were insufficient to trigger cl 1.2, for the following reasons:

(a) As regards the commingling of cryptocurrencies, D1 argued that this was not an instance of “fraud” as there was no evidence of Zee having ever been prohibited from using the Binance Account to house deposits from investors outside the Cryptotrage scheme (such as Torque’s investors). In any case, the particulars of the alleged commingling were never pleaded.

(b) As regards Zee’s alleged theft of the cryptocurrencies, the Defendants argued that this did not constitute “internal fraud by any staff” of D1, given that Zee was not D1’s “staff” but merely an independent contractor. More importantly, there was no evidence to show just how much of the cryptocurrency deposits in the Binance Account had been depleted or withdrawn by Zee, nor to indicate whether the cryptocurrencies allegedly depleted or withdrawn in fact belonged to the Plaintiff (as opposed to, say, Torque’s investors). That being the case, the Plaintiff’s cryptocurrencies could not be regarded as having been “stolen by fraud”.

The Defendants also rejected the suggestion that they were under any duty to supervise Zee. D1 argued that the Plaintiff’s AEIC and submissions completely failed to disclose any legal basis for such a duty. D2 argued that it was not incumbent on him to supervise Zee, as neither Zee nor D1’s operations in Vietnam were under his charge.

Finally, the Defendants contended that the Plaintiff failed to properly quantify his damages:

(a) Firstly, there was no evidence as to how much cryptocurrency deposits remained in the account after Zee’s alleged fraud. The Defendants contended that if substantial amounts remained, the Plaintiff’s claim in this action would be premature. The Plaintiff had filed a proof of debt with Torque’s liquidators in respect of his cryptocurrencies in the Binance Account. Allowing his claim against the Defendants at this juncture could result in him acquiring a windfall, should he eventually recover some or all of his cryptocurrencies from Torque’s liquidators .

(b) Furthermore, the Plaintiff had calculated his losses based on the cryptocurrency balances as set out in the daily reports for 8 February 2021. However, those daily reports were unreliable, given that the Plaintiff himself pleaded that they were “completely bogus”.

Issues to be determined in this case

Against the backdrop of the parties’ respective claims and defences, as set out above, the following issues arise for my determination:

The following sections will address each of these issues in turn.

Issue 1: Whether the Service Agreement was proven to have been signed by D2

The Executed Paper Copy was never adduced in court. The Plaintiff claimed that it remains within the possession of the Defendants, who have refused to pass the same to him. The Plaintiff had served on D2 a notice to produce the Executed Paper Copy but D2 (unsurprisingly) failed to do so. The Plaintiff thus resorted to proving the existence of the Executed Paper Copy by adducing P1, ie, a printout of the E-Copy sent by Zee. A fundamental premise of the Plaintiff’s case was thus that the E-Copy (and hence P1) was a faithful reproduction of the Executed Paper Copy.

Having carefully considered the Plaintiff’s evidence, I find that he has failed to establish that the Service Agreement was signed by D2. For the Plaintiff’s case to succeed, he must at the very least show that the Executed Paper Copy, bearing D2’s wet-ink signature, exists. I am not persuaded that the Plaintiff has succeeded in doing so, for the following reasons:

(a) Firstly, the Plaintiff failed to abide by the relevant evidential procedures for proving the Executed Paper Copy. Given that primary evidence of the Executed Paper Copy was not forthcoming, he needed to demonstrate that the statutory requirements for adducing secondary evidence of the same were fulfilled – this he failed to do.

(b) Secondly, even if the Plaintiff is entitled to adduce secondary evidence of the Executed Paper Copy, the putative secondary evidence in this case – the E-Copy (manifested in paper form as P1) – was not shown to be authentic. The E-Copy and P1 were consequently inadmissible in evidence.

I explain these findings below.

Whether the Plaintiff failed to abide by the evidential procedures for proving the Executed Paper Copy

Section 66 of the Evidence Act (Cap 97, 1997 Rev Ed) (“EA”) provides that documents must be proved by way of primary evidence (unless an exception in s 67 of the EA applies, in which case the relevant document may be proved by way of secondary evidence). In this respect, s 64 of the EA explains what constitutes primary evidence:

Thus, primary evidence of the Executed Paper Copy (and its contents) would have been the Executed Paper Copy itself.

If one accepts the Plaintiff’s case that:

that would mean that the E-Copy constitutes secondary evidence of the Executed Paper Copy. This much is clear from a reading of s 65 of the EA, which states:

The Plaintiff nevertheless sought to argue that the E-Copy is primary evidence. In doing so, he alluded to Explanation 3 to s 64 of the EA, which states that:

I reject this contention as Explanation 3 is clearly inapplicable to the present context. The E-Copy sent by Zee was, by the Plaintiff’s case, a copy of the Executed Paper Copy. Yet, there was no evidence showing how Zee had derived the E-Copy from the Executed Paper Copy, and thus nothing to show that the E-Copy did “reflect” the Executed Paper Copy (whether “accurately” or at all).

The Plaintiff further submitted that the E-Copy was primary evidence by virtue of s 116A of the EA. For ease of reference, the relevant parts of s 116A of the EA are set out below:

I find the Plaintiff’s argument difficult to comprehend. For a start, he did not attempt to explain which subsection in s 116A of the EA he was relying on. Given the dearth of elaboration, I can only assume that he was referring to subsection (1), given that subsections (2) and (3) appear to have no application to the present facts. Even on that premise, I find it difficult to see how subsection (1) has any bearing on the Plaintiff’s contention that the E-Copy was primary evidence of the Executed Paper Copy. Applied to the present context, all subsection (1) does is to presume that the communication of the electronic record (ie, the E-Copy in this case) over Telegram was accurate. There is nothing controversial about this – even the forensic experts from both sides found no reason to question the integrity of the transmission process over Telegram. However, accurate transmission simply meant that the E-Copy received by the Plaintiff and Kenneth Chan over Telegram was the same as what Zee had sent over Telegram. That conclusion bears no relation to the crucial link which the Plaintiff must establish, as between the E-Copy sent by Zee over Telegram and the Executed Paper Copy from which the E-Copy was purportedly derived. Accurate transmission of the E-Copy from Zee over Telegram could not, in and of itself, transform the status of the E-Copy into primary evidence of the Executed Paper Copy. I fail to see how the operation of s 116A of the EA can possibly culminate in that result.

The upshot of the above is that the E-Copy was, at best, secondary evidence of the Executed Paper Copy. That being so, what would have been the appropriate route by which the E-Copy could be admitted (as secondary evidence) under the EA? Some guidance can be found in the remarks of Belinda Ang J (as she then was) in Jet Holding Ltd and others v Cooper Cameron (Singapore) Pte Ltd and another [2005] 4 SLR(R) 417 (“Jet Holdings”) (at [146]):

At first blush, it would appear that a possible pathway by which the E-Copy could be admitted as secondary evidence of the Executed Paper Copy lies in s 67(1)(a) of the EA. That provision allows secondary evidence to be given of the existence, condition or contents of a document that is admissible in evidence, if the original appears to be in the possession or power of one of the following three categories of persons:

and that person does not produce the original despite having been served with a notice in s 68 of the EA. Given the Plaintiff’s endeavour to use the Executed Paper Copy as evidence against the Defendants, whom he claims to be withholding the Executed Paper Copy from him, the relevant limb of s 67(1)(a) of the EA would potentially have been limb (i). Alternatively, if the Executed Paper Copy was somehow sent to Zee, another potentially relevant limb would have been limb (ii) (on account of Zee being “out of reach”).

I venture the observations above with a highly tentative tone because, apart from the fact that they are made without the benefit of argument, it was not the Plaintiff’s case that s 67(1)(a) of the EA allowed the E-Copy to be admitted as secondary evidence. Rather, he insisted that ss 65 to 67 of the EA were “not relevant” because the E-Copy was conferred the status of primary evidence by operation of s 116A of the EA. The Plaintiff declared that “once the Plaintiff satisfies the requirements of section 116A of the EA, sections 65 to 67 [of the] EA falls away.” In short, he had placed all his eggs in the s 116A basket and jettisoned any prospect of leveraging on s 67(1)(a) of the EA. In light of this, the evidence needed to establish the circumstances described in limbs (i) or (ii) of s 67(1)(a) of the EA was either not led or – if it had been led – lay buried in the documents without having been unearthed and deployed in aid of arguments that may have persuaded me to admit the E-Copy under those provisions. At this juncture, it is appropriate to echo one of the holdings in Jet Holdings (at [149]–[151]):

Given the Plaintiff’s misplaced reliance on s 116A of the EA to adduce the E-Copy as primary evidence, as well as his failure to demonstrate why the statutory requirements for admission of secondary evidence were satisfied, the conclusion must be that evidence for proving the Executed Paper Copy and its contents has not been properly admitted in this case.

Whether the Plaintiff’s documentary evidence was authentic

If I am wrong in my conclusion above, and the Plaintiff is entitled to adduce secondary evidence of the Executed Paper Copy, that evidence would have to be the E-Copy (manifested in paper form as P1). Even then, the Plaintiff would in my view have failed to establish the authenticity of the E-Copy (and hence P1), meaning that that this piece of documentary evidence would have been inadmissible in any case.

The law draws a distinction between adduction and authentication of documentary evidence. Just because a document is adduced in court (whether by way of primary or secondary evidence), this does not mean that it must be regarded as authentic. In CIMB Bank Bhd v World Fuel Services (Singapore) Pte Ltd and another appeal [2021] 1 SLR 1217 (“CIMB v World Fuel”), the Court of Appeal held (at [52]–[54]):

The Court of Appeal in CIMB v World Fuel also cited with approval (at [51]) the following remarks by Ang J in Jet Holdings (at [146]):

In the present case, the Defendants challenged the authenticity of P1 (and, by extension, the E-Copy), maintaining that D2 never signed the Service Agreement. In doing so, the Defendants put the authenticity of the E-Copy and P1 into issue. As observed in CIMB v World Fuel (at [36] and [54]):

Consequently, even if the E-Copy (and hence P1) is admitted as secondary evidence, the Plaintiff must still defend the challenge to its authenticity.

If the E-Copy (and hence P1) is found not to be authentic, it will not be admissible in evidence. In Mustaq Ahmad (alias Mushtaq Ahmad s/o Mustafa) and another v Ayaz Ahmed and others and other appeals [2024] SGHC(A) 17 (“Mustaq Ahmad”), the Appellate Division of the High Court stated (at [169]) that “authenticity is a necessary condition of admissibility”. The Appellate Division also endorsed (at [169]) the following observations by Vinodh Coomaraswamy J in Super Group Ltd v Mysore Nagaraja Kartik [2018] SGHC 192 (at [53]):

With the above principles in mind, I now assess the issue of authenticity of the E-Copy and P1.

The burden of proving authenticity versus the burden of proving forgery

In general, a party who asserts that a document is authentic bears the burden of proving that document’s authenticity: see CIMB v World Fuel at [37] and [70]. In Mustaq Ahmad, the Appellate Division held (at [160(b)]):

The basis for placing the burden of proof on the party advocating that the document is authentic lies in s 105 of the EA, which sets out the default rule that the burden of proof as to any fact lies on the person wishing the court to believe in that fact’s existence: CIMB Bank Bhd v Italmatic Tyre & Retreading Equipment (Asia) Pte Ltd [2021] 4 SLR 883 at [73].

However, the practical application of that rule may prove challenging in a case such as the present, where authenticity has been challenged on the basis that the document concerned was forged. While the law is clear that the burden of proving authenticity lies on the party asserting that the document in question is authentic, it is also clear that the burden of proving forgery lies on the party asserting that the document was forged. In Alwie Handoyo v Tjong Very Sumito and another and another appeal [2013] 4 SLR 308 (“Alwie Handoyo”), the Court of Appeal held (at [157]):

Consequently, where authenticity is challenged on the ground that the document was forged (which I imagine is not an uncommon ground for challenging authenticity), the issue of the document’s authenticity and the question of whether the document was forged are apt to be closely intertwined. That would in turn make it difficult to tell where the Plaintiff’s burden (to prove authenticity) ends and where the Defendants’ burden (to prove forgery) begins. This difficulty appears to have been recognised by the Court of Appeal in Sudha Natrajan v The Bank of East Asia Ltd [2017] 1 SLR 141 (at [42]):

I would suggest that when a party seeks to tender a document which the opposing party alleges to be forged, the assessment of the competing evidence could possibly be structured according to the following steps, to ensure that the relevant burden of proof is properly allocated:

(a) One first needs to ask if the tendering party has adduced sufficient evidence to establish the document’s provenance. If there is such evidence and the evidence would – if uncontroverted – suffice to establish the document’s provenance on a balance of probabilities, the tendering party will have established authenticity unless the same is challenged by the opposing party.

(b) If the opposing party challenges the document’s authenticity, the initial area of focus should be on that portion of his case which does not go toward alleging forgery, eg, evidence which seeks to cast doubt on the tendering party’s account of how he came to possess the document. At this stage of the analysis, the court should assess if, notwithstanding the countervailing evidence from the opposing party, the tendering party’s evidence as to the document’s provenance still suffices to discharge the latter’s burden of proving authenticity on a balance of probabilities. If the answer is ‘no’, the tendering party’s case on authenticity fails even without delving into the issue of forgery.

(c) If the answer to the question posed at step (b) above is ‘yes’, authenticity will be considered as having been prima facie established, unless the opposing party proves his allegation of forgery. Once the analysis reaches this stage, the opposing party must discharge his burden of proving forgery on a balance of probabilities. To do so, his evidence would have to go beyond merely “poking holes” in the tendering party’s case on provenance – something which would already have been done at step (b) above without success. Rather, the focus must shift to that aspect of the opposing party’s evidence which relates specifically to his allegation of forgery, eg, handwriting expert testimony indicating that a signature was irregular; computer forensic analysis showing that tampering occurred; circumstantial evidence suggesting that someone had both reason and opportunity to manipulate the document; etc.

The practical application of the framework above may prove challenging, especially when it is not entirely clear whether the opposing party’s evidence should be regarded as merely discrediting the claimant’s case as to provenance, and therefore dealt with at step (b), or as alleging forgery and thus dealt with at step (c). Be that as it may, given how the burden of proof has been assigned at law, it remains necessary to properly earmark who needs to prove what.

Useful guidance can be gleaned from the Court of Appeal’s decision in Wibowo Boediono and another v Cristian Priwisata Yacob and another and other appeals [2018] 2 SLR 481. In that case, the appellants alleged that the respondents owed a debt which was evidenced by a signed note and an undated cheque. The appellants claimed that one of the respondents signed the cheque as security for the debt. The respondents contested this, alleging (inter alia) that the signatures on the cheque and note were forgeries. The lower court dismissed the appellants’ claims on this point, holding that they failed to discharge their burden of proving that the cheque and note were authentic. This finding was reversed on appeal, where the Court of Appeal remarked (at [52]–[54]):

One sees from the passage above that the Court of Appeal spoke about the respondents’ burden to prove forgery only after alluding to the appellants having adduced evidence as to the documents’ provenance. The Court of Appeal observed that the cheque indisputably originated from the second respondent’s chequebook. It is also evident from the case that the signature concerned looked like the respondent’s. On the back of those facts, the Court of Appeal found that the provenance of the cheque had been established to a prima facie standard. It was against this backdrop that the Court of Appeal proceeded to consider if the respondents had discharged their burden of proving forgery.

Reverting to the present case, I am of the view that even at step (a) of the framework in para 78 above, the Plaintiff has failed to discharge his burden of establishing the authenticity of the E-Copy. There was nothing to show how Zee got his hands on the E-Copy, and still less to show that the E-Copy was indeed derived from the (putative) Executed Paper Copy. The only evidence tying the E-Copy to D2 was Zee’s Telegram message (at para 22 above) in which Zee claimed that the Service Agreement had been signed by D2. That statement by Zee was hearsay, but even if it could be admitted (eg, under the hearsay exception in s 32(1)(j)(ii) of the EA, on account of Zee being untraceable), the weight to be given to that statement would be negligible. By the Plaintiff’s own account, Zee was a fraudster who had absconded with millions of dollars in digital assets. In my view, Zee’s statements do not suffice to establish a prima facie case as to the provenance of the E-Copy.

Even if I am wrong and Zee’s hearsay statement can be regarded as prima facie establishing the E-Copy’s authenticity, I take the view that the Plaintiff still fails at step (b) of the framework in para 78 above. In other words, the Defendants have, quite apart from their evidence that the E-Copy was forged, successfully cast doubts on the E-Copy’s provenance:

(a) Firstly, I agree with D2’s submission that the Plaintiff’s entire hypothesis of how P1 came to be handed to Schmidt made no sense. It was the Plaintiff’s case that D2 was in Singapore during the early morning hours of 27 May 2019, during which he executed the Service Agreement. The Plaintiff postulated that after execution, the Executed Paper Copy was scanned to produce the E-Copy, which D2 sent to Zee over Telegram that very morning at 8.27am. About a minute after that, at 8.28am, Zee forwarded D2’s message (which enclosed the E-Copy) to the Plaintiff over Telegram, via the message in the conversation chain extracted at para 22 above. This meant that the very last location of the Executed Paper Copy must have been D1’s office in Singapore, where D2 allegedly appended his wet-ink signature. There was no suggestion of the Executed Paper Copy having been couriered to Zee in Vietnam after it was signed by D2, or to anywhere else for that matter. That being the case, when Schmidt came to D1’s Singapore office to collect a copy of the signed Service Agreement on 3 June 2019, one would have expected Kenneth Chan to simply pass him the Executed Paper Copy bearing D2’s wet-ink signature. Why was that not done? Instead, Kenneth Chan had to wait for the E-Copy to be sent via Telegram from Zee in Vietnam, notwithstanding that the Executed Paper Copy would already have been sitting in the Singapore office. There was no apparent reason for such contrived circularity. It was not as if D2 needed to keep the Executed Paper Copy (bearing D2’s wet-ink signature) for his own records, given that record-keeping would have entailed D2 maintaining a copy of the contractual documents bearing the wet-ink signature of the counterparty (ie, the Plaintiff) rather than that of himself.

(b) Secondly, the evidence suggests that D2 could not have affixed D1’s corporate stamp on the Service Agreement. D2’s uncontroverted testimony was that he did not have access to D1’s corporate stamp. Dr Ting similarly testified that only Dr Ting and D1’s Operation Manager had access to the corporate stamp, which was kept under lock and key. I also note that it was not the Plaintiff’s case that D2 had purchased a duplicate of D1’s corporate stamp (which had a rather generic design) to apply to the Service Agreement.

In light of these factors, the conclusion which I would draw at step (b) of the framework in para 78 above is that the Plaintiff’s evidence as to the provenance of the E-Copy has been sufficiently discredited, to the point that the Plaintiff must be regarded as having failed to discharge his burden of proving the authenticity of the E-Copy (and hence of P1). That conclusion can be sustained without having to venture into the Defendants’ evidence of forgery.

In seeking to establish the authenticity of the E-Copy, the Plaintiff (yet again) sought to rely on s 116A of the EA. As explained at para 65 above, I rejected the Plaintiff’s attempt at using s 116A to confer upon the E-Copy the status of primary evidence. As regards his attempt to now invoke s 116A of the EA to establish the E-Copy’s authenticity, I similarly find his arguments to be misconceived. He contended that s 116A of the EA creates a presumption that the Service Agreement is authentic, submitting as follows:

As alluded to above, I am prepared to accept that s 116A(1) of the EA operates to create a presumption that the electronic record was accurately transmitted over Telegram. However, that only meant that the E-Copy received by the Plaintiff and Kenneth Chan in their respective devices over Telegram faithfully corresponded to what was sent by Zee from his device. That had no bearing on the issue of the E-Copy’s authenticity, which would centre on whether the E-Copy sent from Zee’s device faithfully corresponded to the Executed Paper Copy, from which the E-Copy was allegedly derived.

If I am wrong in my analysis above and the Plaintiff can be regarded as having established a prima facie case that the E-Copy is authentic, notwithstanding the doubts over its provenance, the next step would be to address the Defendants’ claim of forgery, as per step (c) of the framework at para 78 above. At that stage, the burden would fall squarely on the Defendants to prove, on a balance of probabilities, that the E-Copy (and hence P1) was forged.

In that respect, I find that the Defendants have succeeded in establishing forgery. My finding is broadly premised on two grounds:

I elaborate on each of these grounds in turn.

Evidence of the documentary experts on the issue of forgery

D2’s documentary expert was James Tan, a Senior Forensic Consultant with Infinity Forensics Pte Ltd. The Plaintiff’s documentary expert was William Pang, a handwriting analyst from HFDE Services Pte Ltd.

Both experts approached the issue of whether the E-Copy was forged primarily by examining P1, being the physical manifestation of the E-Copy. Both experts focused heavily on features within the third page of P1, particularly (a) D2’s signature; and (b) the handwritten words “Bernard Ong. Managing Director, Snap Innovations (Singapore)”. The third page of P1 has been reproduced at para 29 above, but I set out below an enlarged snapshot of the relevant section of that page, for ease of visualisation:

I have encased the ending portion of the handwritten word “Innovations” within the green rectangle, as this was the subject of some focus in the Defendants’ expert evidence. As for D2’s purported signature, one sees a stroke extending upwards and touching the letter “p” in the typed word “Operations” and another stroke extending downwards and touching the letter “a” in the handwritten word “Bernard”. I have circled both these points of contact in red, as they were the subject of some focus in the Plaintiff’s expert evidence.

(1) James Tan’s expert opinion

James Tan took the view that P1 was “modified”, with D2’s signature and the purported handwritten words on the third page having been copy-pasted. His conclusion was based principally on two observations.

Firstly, James Tan found it significant that the black dot situated to the right of the handwritten word “Innovations” fell on the right side of the vertical black bar, at the third page of P1. To him, this suggested that the black dot, and by extension the handwritten words and signature purportedly penned by D2, were copy-pasted. To illustrate James Tan’s point, I have magnified the section encased within the green rectangle in the snapshot extracted at para 86 above:

James Tan took the view that the black dot was part of the handwritten word “Innovations” – this was evident from the similarity in colour between the handwritten word and the dot. As such, if the Executed Paper Copy ever existed, it would have borne not only D2’s wet-ink signature and the handwritten words, but the black dot as well – all of which, according to the Plaintiff’s case, were penned by D2 on the third page. Critically, James Tan postulated that it was the process of scanning the Executed Paper Copy to create the E-Copy which generated the vertical black bar in the E-Copy, with that bar reflecting the right-most edge of the Executed Paper Copy’s third page as the latter was being scanned. This meant that it would have been impossible for any feature on the third page of the Executed Paper Copy (such as the black dot) to manifest itself within the E-Copy as being situated to the right of the vertical black bar. The very fact that the black dot appeared in the E-Copy (and hence in P1) as lying to the right of the vertical black bar suggested that the black dot could not have been captured from scanning the Executed Paper Copy. Rather, the black dot (and by extension the handwritten words and D2’s signature) on the E-Copy must have been pasted on an electronic copy of the Service Agreement, as follows:

(a) A copy of the Service Agreement (without D2’s signature and handwriting) must have first been scanned, thereby creating an electronic copy in which the vertical black bar was generated by the scanning process.

(b) Thereafter, D2’s signature, as well as the handwritten words and accompanying black dot, must have been pasted onto that electronic copy, thereby creating the E-Copy in which the black dot was positioned to the right of the vertical black bar.

Secondly, James Tan observed that the image of the corporate stamp in P1 was relatively sharp, while the signature of D2 was blurry in comparison. This further supported the inference that D2’s signature in P1 was not authentic.

(2) William Pang’s expert opinion

In contrast, William Pang took the view that there was “no indicia of forgery” in P1. He concluded that neither D2’s signature nor the handwritten words on the third page of P1 was copy-pasted. He supported this conclusion with the following observations:

(a) There were no “suspicious lines” suggestive of the signature having been copy-pasted.

(b) A signature that was copy-pasted would typically be transposed onto an area where it could be clearly displayed. However, D2’s signature on the third page was appended at a spot where it overlapped with the typed word “Operations” above it and the handwritten word “Bernard” below it, as seen in the two red circles within the magnified image at para 86 above.

William Pang also sought to bolster his opinion by relying on what transpired when the Plaintiff’s solicitors tried to procure specimens of D2’s signature and handwriting from D2’s solicitors. Specifically, during the interlocutory exchanges, the Plaintiff’s solicitors wrote a letter to D2’s solicitors asking for, inter alia:

The date 27 May 2019 is significant because that was the date on which, according to the Plaintiff’s case, D2 had signed the Service Agreement (see para 20(c) above). Any specimen of D2’s signature or handwriting existing as at that date would show what D2’s signature and handwriting looked like when he allegedly executed the Service Agreement. However, D2’s solicitors replied to this request in the negative, by way of a letter stating that D2 did not have possession, custody or power over the documents sought. Given this negative reply, the Plaintiff had to procure samples of D2’s signature from other sources. The Plaintiff ultimately managed to obtain copies of five documents signed by D2 on various dates post-27 May 2019, which were then furnished to William Pang as specimens for the latter’s forensic analysis. In his expert opinion, William Pang referred to the negative reply from D2’s solicitors and concluded that if D2 himself did not possess any specimens of his own signature that were penned before 27 May 2019, this meant that there would have been no source from which D2’s signature could have been copied at the point the Service Agreement was signed, thereby weakening the suggestion that his signature was copy-pasted on the E-Copy.

A substantial portion of William Pang’s expert observations centred on highlighting similarities between the specimens of D2’s signature (procured by the Plaintiff) on the one hand, and what was alleged by the Plaintiff to be D2’s signature and handwriting in P1 on the other. In reliance on these similarities, William Pang drew the following conclusions:

(a) As regards the signature on the third page of P1, there was “very strong support” that it was penned by the person who signed the specimen signatures (ie, D2). In particular, there were various similarities in characteristics exhibited by both the signature in P1 and the specimen signatures.

(b) As regards the handwritten words “Managing Director, Snap Innovations (Singapore)”, it was “likely” that those were also penned by the person who signed the specimen signatures (ie, D2). William Pang drew support for this conclusion from two traits exhibited in D2’s specimen signatures:

(i) The first was the signature’s long descending stroke – looking at the signature in the extract at para 86 above, this would be the stroke dipping down and touching the word “Bernard” (see the lower red circle).

(ii) The second comprised the two dots to the right of the signature.

William Pang noted that both traits could also be discerned in the handwritten words “Managing Director, Snap Innovations (Singapore)”, in that the person writing these words also had a habit of using long descending strokes and ending each sentence with a dot.

(3) My conclusion on the expert evidence

I begin with William Pang’s evidence that D2’s signature and the handwritten words on the third page of P1 were likely penned by the person who signed the specimen signatures, ie, D2.

I am prepared to agree that the signature on the third page of P1 was that of D2. However, that conclusion is beside the point, given that the thrust of D2’s defence was not that the signature in P1 did not belong to him, but that it was copy-pasted.

As for William Pang’s suggestion that it was “likely” that the handwritten words “Managing Director, Snap Innovations (Singapore)” on the third page of P1 were penned by D2, I find this to be unpersuasive. I bear in mind the guidance offered by the V K Rajah JA in Sakthivel Punithavathi v Public Prosecutor [2007] 2 SLR(R) 983 (at [76]):

To recapitulate, it was D2’s position that the handwritten words on the third page of P1 were not in his handwriting. I agree with the Defendants’ submission that to test that position, the handwritten words in P1 should have been compared against samples of D2’s handwritten words. However, that was not the approach taken by William Pang, who compared the handwritten words in P1 with samples of D2’s signature. Given that it was difficult to discern any legible letter of the alphabet from D2’s signature, what William Pang embarked on was clearly not a like-for-like comparison. Furthermore, the similarities were cursory at best. William Pang relied on two traits in D2’s signature: (a) the single stroke dipping downwards; and (b) the two dots at the end, observing that the handwritten words similarly had some characters with long descending strokes and ended with a dot after each line. In my view, it does not take an expert to appreciate that features such as long descending strokes and ending dots are generic handwriting traits. It might have been different if William Pang had offered his expert views not only on the presence of the descending strokes and ending dots in the handwritten words, but also on any similarities in appearance which these features, as they appeared in D2’s specimen signatures, bore to the corresponding features in the handwritten words. No such analysis was performed, notwithstanding that William Pang embarked on such an analysis when comparing D2’s specimen signatures with D2’s purported signature in P1. In my judgment, the parallels drawn between D2’s specimen signatures and the handwritten words “Managing Director, Snap Innovations (Singapore)” fell far short of justifying William Pang’s conclusion that the handwritten words were “likely” to have been written by D2.

In any case, given D2’s defence that his signature was copy-pasted, the more germane question is not whether the signature and handwritten words in P1 belonged to D2, but whether they could have been copied from other sources and then pasted on the third page of P1. William Pang reckoned that there was unlikely to have been any copy-pasting because of the absence of “suspicious lines”, as well as the fact that D2’s purported signature in P1 overlapped with the typewritten word “Operations”. However, these attributes are, in my view, inconclusive. William Pang conceded in cross-examination that handwritten characters can be manipulated to render their background transparent, thus allowing them to be pasted over existing text on a document. That being the case, his observations about the overlapping portions of D2’s signature and the absence of suspicious lines were insufficient to rule out the possibility of copy-pasting having taken place.

I next move to William Pang’s opinion that D2’s signature could not have been copy-pasted because there were no samples of D2’s signature in existence prior to 27 May 2019 (when D2 allegedly signed the Service Agreement), from which a copy could have been made. To recapitulate, the Plaintiff’s solicitors had tried to procure specimens of D2’s signature and handwriting by asking D2’s solicitors for documents or contracts:

D2’s solicitors replied by way of letter dated 30 March 2023 stating that D2 did not have such documents in his possession, custody or power. Focusing on that negative reply, William Pang opined as follows:

With respect, this was an entirely factual inference which lay squarely within the remit of the court, and not the expert. As a cautionary illustration of why experts should not stray beyond their designated remit, it is obvious that William Pang’s inference above was based on a misapprehension of the facts. D2’s solicitors did not say that there were no documents prepared before 27 May 2019 bearing D2’s signature. They were instead saying that there were no documents prepared before 27 May 2019 bearing D2’s signature that appeared together with the words “Managing Director, Snap Innovations (Singapore)” and his initials – this being what the Plaintiff’s solicitors had requested for in the first place. If the Plaintiff’s solicitors had phrased the documentary request more generically – by simply asking for specimens of D2’s signature without any qualification as to what the signature should appear together with – the response from D2’s solicitors may well have been different.

I now move to James Tan’s expert evidence. He concluded that P1 was modified, with a key plank of his assessment being that the black dot which formed part of the handwritten word “Innovations” fell to the right of the vertical black bar (see para 88 above). I do not accept this aspect of his opinion. It is apparent from James Tan’s analysis that his conclusion rested on the twin assumptions that:

(a) the vertical black bar was absent from the paper copy of the Service Agreement which D2 allegedly penned his handwritten words and signed, meaning that the Executed Paper Copy did not have the vertical black bar; and

(b) the vertical black bar could only have been generated by the process of scanning the Executed Paper Copy to create the E-Copy.

Under those assumptions, it would admittedly have been impossible for any feature within the third page of the Executed Paper Copy captured by the scan (eg, the black dot) to show up to the right of the vertical black bar in the E-Copy, seeing as how that bar marked the right-most edge of the Executed Paper Copy’s third page as it was being scanned.

However, there was another plausible factual permutation which James Tan failed to consider, under which the vertical black bar was already present on the paper copy of the Service Agreement which D2 physically executed. At trial, the Plaintiff proffered this very factual permutation as an explanation for why the black dot lay to the right of the vertical black bar. As described at para 19 above, the Plaintiff claimed that Zee had scanned the Original Draft of the Service Agreement and printed the scanned electronic copy to produce the Scanned Draft, after which Zee and the Plaintiff appended their wet-ink signatures to both the Original Draft and Scanned Draft, in the Vietnam office. Building on this, the Plaintiff surmised that:

(a) the vertical black bar could have been present on the printout of the Scanned Draft, having been generated by the process of scanning the Original Draft; and

(b) what was then sent to D2 via Telegram for the latter’s execution was likely an electronic copy created by scanning the signed (as in, signed by the Plaintiff and Zee) Scanned Draft, which bore the vertical black bar, rather than by scanning the signed Original Draft, which had no black bars.

Given the above, if D2 applied his pen to a printout of an electronic copy of the signed Scanned Draft, D2 could have inserted the black dot to the right of the vertical black bar.

It should be noted that the Plaintiff’s explanations in the preceding paragraph were proffered extremely late in the day. They were not alluded to in the Plaintiff’s pleadings, nor fleshed out in his AEIC. In fact, it was only after William Pang was extensively cross-examined by D2’s counsel on the first day of trial, as regards the complete absence of any black bars on the Original Draft, that it became apparent to the Plaintiff that the presence of the black bars in P1 was going to pose an issue to his case. Consequently, on the third and fourth days of the trial, the Plaintiff provided the explanations in sub-paragraphs (a) and (b) of the immediately preceding paragraph. The Plaintiff sought to explain the belatedness of his explanation, saying that his insights arose only after “the very stark observations” made by counsel for D2 in the course of cross-examining William Pang. Despite having been raised at the eleventh hour, I find the Plaintiff’s explanation (that D2 may have applied his pen to a paper document which already bore the vertical black bar) to be plausible and I accept it. Accordingly, I agree with the Plaintiff that the black dot’s location vis-à-vis the vertical black bar does not conclusively prove James Tan’s copy-pasting hypothesis.

However, there was a further plank to James Tan’s expert opinion, which I accept as being indicative of D2’s signature on the Service Agreement being forged. As alluded to at para 89 above, James Tan took the view that D2’s signature in P1 was blurry and of poor quality, in contradistinction to the sharpness of D1’s corporate stamp. The difference in resolution is not particularly obvious from the snapshot extracted at para 86 above but would be obvious even to the naked eye upon examination of the physical exhibit P1. To me, this is significant. The Plaintiff’s version of the facts could possibly explain why the sharpness of the image of D2’s signature within P1 differed from the sharpness of, to take an example, the Plaintiff’s and Zee’s signatures: D2 did not sign on the physical copy of the Scanned Draft bearing the wet-ink signatures of Zee and the Plaintiff, but on a printout of an electronic copy of the Scanned Draft (which Zee sent to D2 over Telegram – see para 16 above). This meant that when the Executed Paper Copy (allegedly bearing D2’s wet-ink signature) was scanned to create the E-Copy, the signatures of the Plaintiff and Zee would at the very least be undergoing a second round of digitisation, while D2’s signature would be undergoing only its first. However, this could not be used to explain the difference in sharpness between D2’s signature and the corporate stamp. The Plaintiff’s case was that D2 executed the Service Agreement in the early morning of 27 May 2019 while he was in Singapore, before sending the E-Copy, bearing both D2’s signature and the corporate stamp, to Zee via Telegram at 8.27am that same day – see the Telegram message extracted at para 22 above. This must mean that if the signature and stamp were indeed appended by D2, both must have been appended at the same time. Thereafter, the number of rounds of digitisation to which both D2’s signature and the corporate stamp may have been subjected, before they were finally transposed to what is now seen in P1, would have been no different. In light of that, I agree with James Tan that the difference in resolution between D2’s signature and the corporate stamp, as they appear in P1, is troubling.

The Plaintiff offered no explanation for this discrepancy in visual resolution. The best riposte that he could muster was a complaint as to how James Tan’s observations on the sharpness of the corporate stamp was not raised in his expert report. However, this was in my view an insufficient objection:

(a) Firstly, the main plank of James Tan’s opinion centred on the location of the black dot vis-à-vis the vertical black bar. As observed at para 100 above, it was extremely late in the day when the Plaintiff pulled the rug from under this aspect of James Tan’s expert evidence. Specifically, the suggestion that D2 had signed on a printout of the Scanned Draft (which would have already contained the vertical black bar) was advanced by the Plaintiff only on the third and fourth days of the trial. No trace of that explanation could be found in the Plaintiff’s pleadings or his AEIC, which meant that at the time James Tan was preparing his expert report and AEIC, a critical facet of the Plaintiff’s case was missing. Under such circumstances, one would be hard put to deny the Defendants the opportunity to seek further elaboration from James Tan while he was on the stand, as regards the authenticity of D2’s signature in P1.

(b) In any case, this aspect of James Tan’s testimony (regarding the sharpness of the corporate stamp) did not emerge from questioning by the Defendants’ counsel. Rather, it was the Plaintiff’s counsel who elicited that evidence from James Tan during cross-examination. At no point in the questioning did the Plaintiff’s counsel object to the evidence that he was drawing from James Tan; nor did the Plaintiff seek the court’s permission to recall William Pang so that the latter might contradict James Tan’s evidence on this point.

Based on the above, I am inclined to accept James Tan’s expert view that D2’s signature on P1 was copy-pasted, meaning that P1 (and by extension the E-Copy) was consequently a forgery.

Circumstantial evidence on the issue of forgery

My conclusion that the E-Copy (and hence P1) was forged is also premised on the circumstantial evidence, which I regard as consistent with D2 being unlikely to have signed the Service Agreement. The circumstantial evidence encompasses the following factors:

I canvass each of these factors below.

(1) No commercial rationale to enter into the Service Agreement.

In assessing an allegation that a party’s signature on a contract has been forged, one of the factors that the court will take into account is whether any commercial rationale existed for that party to have entered into the underlying transaction: Alwie Handoyo at [176]–[177]. In the present case, there was no commercial rationale for D1 to have entered into, or for D2 to have signed, the Service Agreement:

From D1’s perspective, the Service Agreement clearly embodied onerous obligations. It exposed D1 to unlimited liability for governance failures in relation to not only its own staff, but also staff from a wholly unrelated company (ie, Clickstaff, in which D1 owned no shares). The Plaintiff failed to point to any convincing quid pro quo that might explain why D1 would undertake such a burden. While the Plaintiff asserted that the Service Agreement was intended to entice him into increasing his investment in the Cryptotrage scheme, he offered no evidence to show how D1 benefitted from – or was even connected to – the scheme. There was no evidence of profits from the Cryptotrage scheme flowing back to D1. The Plaintiff asserted that Dr Ting received substantial “unauthorised payments” from Torque, but this was unsupported by any documentary proof.

From D2’s perspective, it would similarly have made little commercial sense for him to sign the Service Agreement. D2 explained that signing the same would render him responsible for having bound D1 to serious obligations, in favour of someone whom D2 had never even met. D2 reaped no personal benefit from exposing himself in that manner. I find D2’s contention to be a cogent one.

The Plaintiff contended that D2 did have an interest in concluding the Service Agreement with the Plaintiff, as this would provide the assurance needed by the Plaintiff to increase his cryptocurrency deposits in the Binance Account. The increased deposit level would in turn allow the Binance Account to achieve Tier 9 status, thereby unlocking various financial benefits for the account holder (referred to at para 33 above). Given that Torque also used the Binance Account for its trading platform, the benefits from Tier 9 status would flow to Torque, and thereby benefit D2 personally (by virtue of his link with Torque). I find this argument to be speculative. By the Plaintiff’s own case, the Service Agreement was concluded in May 2019, which was some four months before Torque even came into existence. In any case, no evidence was provided to show how the benefits from the Binance Account’s Tier 9 status had been shared with Torque.

(2) The ill-defined nature of the Service Agreement

Furthermore, it is clear from the face of P1 that its clauses were as scant on details as they were imprecise in terminology. Notwithstanding the hefty burdens which it purported to impose, the entire Service Agreement was embodied in a smattering of clauses spanning over two-thirds of a single page. It also curiously referred to Clickstaff as D1’s “child company” when D1 in fact owned no shares in Clickstaff.

These features would have sounded a series of alarm bells, in the face of which one must then ask if a director in D2’s shoes would, in the ordinary course of business, have simply signed off on the Service Agreement and thereby bind his company to such ill-defined obligations, without even seeking the benefit of legal advice. It seems to me that the answer would very likely be “no”.

(3) The call between the Plaintiff and D2 on 16 February 2021

I also find D2’s conduct after the discovery of Zee’s fraud to be consistent with D2’s claim that he was unaware of the Service Agreement.

As I alluded to at para 38 above, the Plaintiff initiated a call with D2 and CK, on 16 February 2021, after Zee’s disappearance. The Plaintiff claimed that during the call, D2 had agreed to explore a repayment plan with the Plaintiff. With respect, that only told part of the story. It is important to note that the call was recorded by the Plaintiff secretly, meaning that while the Plaintiff knew that D2’s words were being captured for posterity, D2 did not. Evidently, the Plaintiff was hoping to catch D2 at his most candid. Despite this, the Plaintiff was unable to point to anything within the transcripts of the call supporting his allegation that D2 had signed the Service Agreement. Instead, what the transcripts showed was D2 repeatedly indicating that he had no recollection of ever signing the same. The following is an example of what D2 said:

When it became apparent during the call that D2 was unfamiliar with the Cryptotrage scheme, the Plaintiff had tried to explain, but the transcripts showed no indication of the Plaintiff’s explanation ringing any bells for D2:

Even after the Plaintiff’s explanation, D2 maintained that he had no recollection of the Service Agreement and even queried if that document could have been forged by Zee. The following sets out D2’s response:

Notwithstanding D2’s professed lack of knowledge about the Cryptotrage scheme, the Plaintiff persisted in pressing D2 for redress, saying:

It was only at this point that D2 expressed openness to a repayment plan, saying:

I note that despite the conciliatory tone struck by D2, there was still nothing in his words demonstrating any semblance of familiarity with the Cryptotrage scheme or the Service Agreement.

I am therefore of the view that the transcripts of the call do not support the Plaintiff’s claim. If anything, they corroborate D2’s case that he knew nothing about the Service Agreement.

(4) The police report filed by D2

D2’s professed lack of knowledge about the Service Agreement was also consistent with his subsequent conduct in filing a police report on 26 March 2021, in which he affirmed that his signature in P1 was forged.

Conclusion on whether the Plaintiff’s documentary evidence was authentic

To summarize my findings above on the issue of authenticity, I am of the view that that the Plaintiff has failed to establish the authenticity of the E-Copy (and hence of P1). I further take the view, after having considered (a) the expert testimonies and (b) the circumstantial evidence, that the Defendants have successfully proven that D2’s signature on the E-Copy was copy-pasted, meaning that the E-Copy (and hence P1) was forged.

Since the E-Copy and P1 have not been shown to be authentic, they are inadmissible in evidence.

Conclusion on whether the Service Agreement was proven to have been signed by D2

To recapitulate, the Plaintiff has failed to demonstrate why the E-Copy (and hence P1) may be adduced as secondary evidence of the Executed Paper Copy. Furthermore, even if he was entitled to adduce the E-Copy as secondary evidence, he has failed to establish the authenticity of the E-Copy (and hence of P1). Relatedly, the Defendants have succeeded in establishing that P1 (and by extension the E-Copy) was forged.

In light of these findings, both the E-Copy and P1, which were the very centrepieces of the Plaintiff’s evidential display, are inadmissible. This created a massive gap in the Plaintiff’s case that the Executed Paper Copy exists. As such, the only conclusion which I can reasonably draw is that the Plaintiff has failed to prove that D2 signed the Service Agreement.

Before leaving this section, I should add that the difficulties associated with proving authenticity could have been avoided had the Plaintiff exercised greater diligence after he was informed by Zee (on 27 May 2019) that the Service Agreement had been signed by D2. As explained at para 31 above, it was evidently important to the Plaintiff that he be given an original of the Service Agreement (ie, one bearing the wet-ink signatures of D2, Zee and himself) for his records. That being the case, once the Plaintiff received P1 from Kenneth Chan, he should have properly examined the document to ensure that it contained D2’s wet-ink signature. Had he done so, it would have been readily apparent to him that P1 was not the Executed Paper Copy. As James Tan rightly pointed out, the signatures, handwriting and initials in P1 were of poor quality. Even the Plaintiff’s own expert, William Pang, confirmed that “[v]isually, [P1] appears to be a reproduction” and that P1 was a “low resolution document” with many rough edges which indicated that it had been photocopied a few times. When D2’s solicitors conducted an inspection of P1 at the Plaintiff’s counsel’s office, they too could see that the document was not an original and had sent a letter to the Plaintiff’s counsel conveying that fact.

Given that P1 was obviously just a reproduction, the Plaintiff ought to have taken the simple step of verifying P1’s authenticity with D2 or Dr Ting. Had that been done, this entire ordeal could have been avoided altogether. The Plaintiff’s excuse for not doing so was that he had thought P1 was an original document, as it was in colour and not black-and-white. This excuse held no sway. In an age where colour printers are ubiquitous, the Plaintiff ought to have known better than to derive comfort solely from the fact that P1 was in colour. He was content to simply file P1 away and proceed to invest millions of dollars in the Cryptotrage scheme, purportedly on the faith of a highly questionable document. Throughout the many months that followed, while the Plaintiff was happily reaping profits from the Cryptotrage scheme, he could have paused at any juncture to take stock and check with D2 or Dr Ting about P1’s authenticity. Again, he did not do so. The Plaintiff met with D2 for the very first time only in February 2021, ie, close to 21 months after the Service Agreement was allegedly signed by D2, by which time the orchestrator of this debacle had long vanished. In the premises, it is clear to me that as between the parties to this suit, the Plaintiff must shoulder the consequences of his own inaction; I do not think that it is reasonable for him to now lay the blame at the Defendants’ feet.

Issue 2: Whether D2 and Zee had the authority to enter into the Service Agreement on D1’s behalf

In light of my finding that D2 has not been proven to have signed the Service Agreement, the issue of whether he and Zee had the authority to enter into the Service Agreement on D1’s behalf falls away. Nevertheless, given the extensive submissions by the Plaintiff on how D2 and Zee possessed both actual and ostensible authority to enter into the Service Agreement on D1’s behalf, as well as the Defendants’ submissions in response, I make the observations below.

Actual authority

In Hely-Hutchinson v Brayhead Ltd [1968] 1 QB 549 (“Hely-Hutchinson”), which was cited with approval by the Court of Appeal in Banque Nationale de Paris v Tan Nancy and another [2001] 3 SLR(R) 726 at [64], Lord Denning MR explained the concept of actual authority within the corporate context as follows (at 583):

In the present case, there was no resolution from D1’s board of directors authorising D1’s entry into the Service Agreement. Nevertheless, the Plaintiff argued that both D2 and Zee possessed actual authority to enter into the Service Agreement on D1’s behalf, on account of their positions within D1. Specifically, the Plaintiff’s contended that D2 and Zee were both “unregistered directors” and consequently endowed with authority by two sources, namely:

In arguing that D2 and Zee were “unregistered directors” of D1, the Plaintiff highlighted that a person need not be registered with the ACRA as a director to be regarded as a director for the purposes of the CA. The Plaintiff relied on the interpretation of the word “director” in s 4(1) of the CA:

What s 4(1) of the CA means is that apart from individuals who are formally appointed by the company as de jure directors (with the appointment effected in accordance with the company’s constitution and the individual being registered with the ACRA as a director), individuals not formally appointed as such may under certain circumstances still be treated as directors. Such individuals include “shadow directors” and “de facto directors”. I note that the case law has drawn a distinction between these two categories of persons:

(a) I begin with the shadow director. The definition of “director” in s 4(1) of the CA includes individuals “in accordance with whose directions or instructions the directors or the majority of the directors of a corporation are accustomed to act”. This description has come to be understood as a reference to shadow directors. In Raffles Town Club Pte Ltd v Lim Eng Hock Peter and others (Tung Yu-Lien Margaret and others, third parties) [2010] SGHC 163 (“Raffles Town Club”), Chan Seng Onn J explained (at [45]):

Shadow directorships are thus marked by “a pattern of behaviour” on the part of the board in complying with the relevant individual’s directions. In Sakae Holdings Ltd v Gryphon Real Estate Investment Corp Pte Ltd and others (Foo Peow Yong Douglas, third party) and another suit [2017] SGHC 73 (“Sakae Holdings”), Judith Prakash JA (as she then was) described the shadow director as “the puppeteer pulling the strings from above” (at [33]).

(b) As regards the de facto director, Chan J explained in Raffles Town Club (at [50]) that such an individual:

The notion of a de facto director is thus conceptually distinct from a shadow director, in that the former exerts his influence on the company by the assumption of directorial responsibilities, while the latter exerts his influence on the company through the de jure directors: Hans Tjio, Lee Pey Woan & Pearlie Koh, Corporate Law (Academy Publishing, 2015) (“Corporate Law”) at para 08.062. Whether an individual falls into one category or the other “is a question of fact and degree”: Sakae Holdings at [33].

In the present case, the Plaintiff did not explain what he meant by the term “unregistered directors”, which he used to describe D2 and Zee. There was nothing in the pleadings, evidence or submissions alluding to any pattern of behaviour by D1’s board in acting on the directions of either D2 or Zee, such that both men should be considered shadow directors. The Plaintiff’s submission could therefore be construed, at least by the process of elimination, as meaning that D2 and Zee were de facto directors of D1.

In Raffles Town Club, Chan J (at [58]–[59]) endorsed the guidelines laid out in Gemma Ltd v Davies [2008] BCC 812 for determining if de facto directorship exists. These guidelines can be summarised as follows:

(a) To establish that an individual is a de facto director of a company, it is necessary to plead and prove that he undertook functions in relation to the company which could properly be discharged only by a director.

(b) It is not necessary for an individual to be held out as a director in order for him to be regarded as a de facto director. “Holding out” is not a sufficient condition, as what matters is not what the individual called himself but what he did. However, “holding out” may be important evidence supporting the conclusion that the individual acted as a director in fact.

(c) It is necessary for the individual alleged to be a de facto director to have participated in directing the affairs of the company on an equal footing with the other director(s), and not in a subordinate role.

(d) The individual in question must be shown to have assumed the status and functions of a company director and to have exercised “real influence” in the corporate governance of the company.

(e) If it is unclear whether the acts of the individual are referrable to an assumed directorship or to some other capacity, the individual is entitled to the benefit of the doubt, but the court must be careful not to strain the facts in deference to this observation.

Bearing the above guidelines in mind, it is clear that the evidence fell far below what was needed to establish D2 and Zee as de facto directors of D1. The fact that D1’s website “held out” D2 and Zee as directors was insufficient, as it was also necessary to consider what both these individuals did vis-à-vis D1. In this respect, there was nothing to show that D2 or Zee served any functions within D1 falling within the purview of a formally appointed director.

I begin with why I think D2 could not have been a de facto director. Firstly, both the Plaintiff and his witness, Nguyen Thanh, claimed to have been told by Zee that D2 was Zee’s “boss”. However, putting aside the point that Zee’s statements constituted hearsay (as well as the fact that those statements, even if admissible, would likely hold little weight given that the Plaintiff himself maintained that Zee was a fraudster), Zee’s alleged statements were bereft of details that could explain exactly which matters D2 was supposed to have been the “boss” of. Secondly, while the Plaintiff suggested that D2 was in D1’s office regularly, this shed no light on whether D2 had performed any directorial duties in D1 while at the office. In any case, D2 explained that his regular presence in D1’s office was the result of an office-sharing arrangement between D1 and D2’s other unrelated business. Thirdly, there was little to no evidence on the extent of D2’s involvement in D1’s operations in Vietnam. The lack of clarity as to what transpired in Vietnam is seen in the following:

(a) Nguyen Thanh testified that he had once been invited by Zee to attend a meeting with D2, when the latter visited Snap Vietnam’s office. However, there was nothing in Nguyen Thanh’s evidence about whether the meeting (which lasted for all of 20 minutes) demonstrated D2’s alleged de facto directorship in D1.

(b) Snap Vietnam’s technical lead, Basu, had also seen D2 in Vietnam “a few times”. However, Basu did not explain what responsibilities were discharged by D2 in Vietnam. In his Telegram chats with the Plaintiff, Basu confirmed that all his communications were with Zee, while his interactions with D2 were social in nature.

(c) Crucially, both Nguyen Thanh and Basu had never seen any documents bearing D2’s signature.

Clearly, the evidence fell far short of establishing D2’s status as a de facto director. I should add that there was also no evidence of D2 ever having signed an employment contract with D1, or of D2 ever having been remunerated by D1 for his alleged directorial services. The Plaintiff claimed that D2 received “some kind of profits” but adduced no evidence to support that assertion.

As for Zee, I similarly do not think that he was a de facto director of D1. The Plaintiff claimed that Zee made business decisions relating to D1’s operations in Vietnam, including the acceptance of deposits, negotiating “profit-cut fees”, and approving withdrawals of investors’ funds. Nguyen Thanh also testified on the Plaintiff’s behalf to say that Zee was in charge of engaging (through Clickstaff) Vietnamese staff for D1’s operations in Vietnam and approving their monthly salaries and expenses. However, I am not convinced that those activities may be regarded as having been conducted by Zee on D1’s behalf (as opposed to Zee acting on his own account). While D1 admitted to having made some payments to Clickstaff, it explained that those payments were for D1’s “crypto mining rig business and software development services”, and not for the Cryptotrage scheme. In fact, D1 disclaimed any involvement with the Cryptotrage scheme. In my view, D1’s denial on this front was consistent with the evidence:

As such, the operation of the Cryptotrage scheme in Vietnam could not be conclusively attributed to D1. I am therefore unable to confidently ascribe the full extent of Zee’s managerial activities in Vietnam (which, according to the Plaintiff, had been heavily centred on administering the Cryptotrage scheme) to D1’s business. There being insufficient evidence to even link Zee’s activities in Vietnam to D1, questions as to whether Zee orchestrated those activities as a de facto director of D1 fall away.

If I am wrong in my analysis above and both D2 and Zee may be regarded as de facto directors of D1, the next step would be to consider the Plaintiff’s submission (at para 125 above) that both men had been endowed with actual authority to bind D1 to the Service Agreement, by virtue of:

(a) s 25B of the CA; and

(b) D1’s constitution.

I begin with the Plaintiff’s argument that as (de facto) directors of D1, both D2 and Zee were statutorily clothed with authority by the operation of s 25B of the CA. I set out the relevant portions of this provision below:

I do not think that s 25B of the CA assists the Plaintiff’s submission on actual authority. At the threshold, the Plaintiff failed to explain why s 25B of the CA should be interpreted as applying not only to de jure directors but to de facto directors as well. In any event, that is not something which I need to comment on given that substantively, the Plaintiff’s reliance on s 25B of the CA is flawed. That provision serves to relieve parties transacting with a company from having to check whether the directors’ authority to transact on the company’s behalf is curtailed by the company’s constitutional documents or processes. It is evident from a plain reading of this provision that it is directed at powers to bind the company which are already in place, with subsection (1) focusing on precluding those powers from being negated by restrictions internal to the company (at least so far as the third party is concerned). Contrary to the Plaintiff’s suggestion, s 25B of the CA does not purport to be a standalone source of power that endows directors with authority to bind their companies to all manner of contracts, when no such authority is discernible at law to begin with.

In this regard, an analogy can be drawn with the “indoor management rule” (also referred to as the “Turquand rule”, having been derived from the case of Royal British Bank v Turquand (1856) 6 E&B 327), which s 25B of the CA “partially reflects”: Corporate Law at para 07.096. The indoor management rule essentially entitles an outsider to assume that the company’s internal processes have been duly complied with unless he has been put on enquiry to the contrary: Corporate Law at para 07.091. The rule thus relieves outsiders transacting with a company from the burden of having to check for non-compliance with the company’s internal processes that might otherwise defeat the authority of the agent transacting on the company’s behalf. The indoor management rule similarly does not function as a standalone source of authority – the authority which the indoor management rule seeks to shield from defeat must still be independently derived from the legal principles of agency. As observed by Dawson J in the Australian High Court’s decision of Northside Developments Pty Ltd v Registrar-General and others (1990) 93 ALR 385 (at 420):

Similarly, I take the view that s 25B of the CA cannot be relied upon by the Plaintiff as the basis for imputing actual authority to D2 and Zee.

The final argument raised by the Plaintiff for establishing actual authority rested on D1’s constitution. Specifically, Article 74 of D1’s constitution reads:

The Plaintiff relied on this article to argue that D1’s directors (including what he termed as “unregistered directors” such as D2 and Zee) had the power to bind D1 to the Service Agreement.

I fail to see how Article 74 is relevant to the Service Agreement. That article refers to the raising of funds by D1’s directors and makes no mention of directors having the power to grant broad indemnities against unlimited losses, such as those expressed in the Service Agreement.

For the reasons above, I reject the Plaintiff’s contention that D2 or Zee were endowed with actual authority (implied or otherwise) to bind D1 to the Service Agreement.

Apparent or ostensible authority

I turn now to the Plaintiff’s submission that D2 and Zee were also clothed with apparent or ostensible authority to sign the Service Agreement on D1’s behalf.

The terms “apparent authority” and “ostensible authority” are synonymous and have on occasion been used interchangeably: see Next of kin of Ramu Vanniyar Ravichandran v Fongsoon Enterprises (Pte) Ltd [2008] 3 SLR(R) 105 at [16], citing Walter Woon on Company Law (Tan Cheng Han SC gen ed) (Sweet & Maxwell, 3rd Ed, 2005) at para 3.24. Rather than having to keep spelling out both terms in this judgment, I have employed only the term “ostensible authority” in the discussion that follows.

In Freeman & Lockyer v Buckhurst Park Properties (Mangal) Ltd [1964] 2 QB 480 (“Freeman & Lockyer”), Lord Diplock (at 505–506) laid out four conditions that must be fulfilled for a contract to be enforceable against a company, notwithstanding the putative agent of the company having no actual authority to enter into that contract on the company’s behalf. The four conditions, which were cited by the High Court in Skandinaviska Enskilda Banken AB (Publ), Singapore Branch v Asia Pacific Breweries (Singapore) Pte Ltd and another and another suit [2009] 4 SLR(R) 788 (at [80]), are restated below:

In Guy Neale and others v Ku De Ta SG Pte Ltd [2015] 4 SLR 283 (“Guy Neale”), the Court of Appeal remarked that taxonomically, the doctrine of ostensible authority is a species of estoppel (at [93]):

As regards the first of the four Freeman & Lockyer conditions (ie, the making of a representation), the principal’s very act of appointing an agent to a specific office carries the potential to constitute a representation by the principal that the agent has the relevant authority. Specifically, when the agent is appointed to an office which usually carries with it actual authority to bind the principal to the transaction in question, the world at large would naturally assume that the agent – by virtue of holding that office – possesses that authority. One may then imply from that appointment a representation by the principal that the agent is now endowed with actual authority to bind the principal to the transaction. In Hely-Hutchinson, Lord Denning MR made this point in relation to the office of a managing director (at 583):

In the present case, the Plaintiff contended that such a representation was in fact made by D1, via the following platforms:

(a) D1’s website, which held out both D2 and Zee as having been appointed to the office of “director”.

(b) D2’s LinkedIn profile, which represented that D2 had been appointed to the office of “Managing Director”. As D2’s evidence was that the LinkedIn profile had been created with inputs from D1’s “marketing guys”, the Plaintiff contended that the representation could be regarded as having emanated from D1 itself.

I will examine the Plaintiff’s claim of ostensible authority in respect of D2, before considering the same in respect of Zee.

Whether D2 had apparent or ostensible authority

Although D2’s LinkedIn profile had stated that he was D1’s managing director, that could not in my view constitute a representation sufficient to ground a finding of ostensible authority. It is well-established that “an agent who has no authority (whether actual or ostensible) to perform a certain act cannot confer upon himself authority to do that act, by representing that he has such authority”: Skandinaviska Enskilda Banken AB (Publ), Singapore Branch v Asia Pacific Breweries (Singapore) Pte Ltd [2011] 3 SLR 540 (“Skandinaviska (CA)”) at [38]. That principle flows as a matter of course from the second of the Freeman & Lockyer conditions, laid out in para 142 above. LinkedIn profiles are typically under the control of the account holder, and not the company that the account holder professes to work for or represent. There is nothing to stop an individual from using his LinkedIn profile to portray himself to the world as holding all sorts of high-sounding positions. It was not open to the Plaintiff to blissfully assume that D2’s LinkedIn profile was a representation emanating from D1 (or anyone else other than D2, for that matter). If one is planning to undertake a significant investment to the tune of millions of dollars, on the faith that the purported agent holds the office which his LinkedIn profile professes him to hold, the sensible thing to do would be to call the organisation (in which the office is purportedly held) to seek confirmation. The Plaintiff took no such precautionary steps.

The Plaintiff nevertheless argued that the representation in D2’s LinkedIn profile could be attributed to D1 on account of D2’s testimony that the profile had been created with the input of D1’s “marketing guys”235 – for completeness, I should add that D2’s evidence on this point was denied by D1. In my view, this did not advance the Plaintiff’s submission that D2 had ostensible authority:

(a) Firstly, the revelation about the involvement of D1’s “marketing guys” in crafting D2’s LinkedIn profile emerged only at trial. At the point when the Plaintiff entered into the Service Agreement, purportedly in reliance on the LinkedIn profile’s contents, there was nothing within the Plaintiff’s field of view to suggest that those contents had been blessed in any way by D1. Indeed, the LinkedIn profile alluded to D2’s positions in multiple organisations, and not just D1.

(b) Furthermore, D2’s evidence that the profile had been created with inputs from D1’s “marketing guys” was scant. His evidence, which was given in the course of his cross-examination, contained no elaboration as to who the “marketing guys” were, or whether they were even D1’s employees. More crucially, there was nothing to shed light on the extent of the authority (if any) conferred on the “marketing guys” to act on D1’s behalf, in relation to providing input on D2’s LinkedIn profile. Plaintiff counsel also did not seek further details. The second of the Freeman & Lockyer conditions (laid out at para 142 above) is clear: for a representation to found ostensible authority, it must be made by someone with “actual” authority to manage the principal’s business. The evidence relating to the involvement of the “marketing guys” fell far short of what was required to make out that condition.

I next move to D1’s website, which held out both D2 and Zee as holding the office of “director”. D1 argued that a representation on the principal’s website, to the effect that an agent holds a specific office within the principal’s organisation, cannot constitute a representation by the principal as to the agent’s authority. D1 argued that that there are no authorities suggesting that such reliance can be placed on a principal’s website, for the purposes of establishing ostensible authority. In my view, D1’s objection needs to be unpacked, as it calls two possible representations into question:

As regards the first possible representation, I am of the view that if the principal’s official website states that the agent holds a certain office within its organisational structure, that is a representation made to the world and which the principal must be held to. D1 argued that there was no evidence as to who was responsible for D1’s website. I would roundly reject this as an unacceptable position. There was no dispute that the website belonged to D1, nor any suggestion of the relevant webpages having been uploaded without authorisation. Under such circumstances, D1 should not be allowed to shirk responsibility for the contents of its own website.

As regards the second possible representation – ie, that the agent had the authority by virtue of his office to bind the principal to the transaction at issue – whether such a representation can be implied from the facts requires careful consideration of various factors. A representation that an agent is a director in a company does not necessarily equate to a representation that he has the authority to enter into the transaction at issue. Much will depend on whether the office which the agent occupies (eg, director), or which the principal holds the agent out as occupying, can reasonably be perceived as usually possessing the authority to bind the principal to the kind of transaction at issue.

In Blasco, Martinez Gemma v Ee Meng Yen Angela and another and another matter [2021] 3 SLR 1360 (“Blasco”), the High Court had occasion to consider the scope of the authority attendant upon the office of managing director. Kannan Ramesh J (as he then was) cited (at [23]) the remarks of Lord Denning MR in Hely-Hutchinson (at 583) that when a person is appointed as a managing director, that person:

Thus, the fact that a person is appointed to the office of managing director does not mean that he is cloaked with ostensible authority to bind the company to every conceivable class of transactions (see also Guy Neale at [89]). Rather, subject to any other representations that may have emanated from the company, ostensible authority covers only those contracts which a managing director would usually have the authority to transact on the company’s behalf.

Thus, in Blasco, one of the issues was whether a senior company officer had the authority to borrow money on the company’s behalf from certain private investors. On the issue of actual authority, Ramesh J rejected the notion that the office of managing director would, as a universal proposition, confer implied actual authority to borrow money or give security on behalf of the company. Instead, he opined that whether such authority existed hinged on a context-driven inquiry that would turn on a range of factors (at [32]):

Ramesh J was further of the view that since the office held by the agent could not be regarded as carrying actual authority to borrow money or give security on the company’s behalf, it followed that the appointment of the agent to that office could not confer upon him ostensible authority to perform such tasks: see Blasco at [44].

It is thus clear that just because the principal appoints an agent to, or holds out an agent as occupying, an apex office within the principal’s organisation, this cannot in and of itself imply a representation by the principal that the agent is cloaked with authority to bind the principal to every conceivable class of transactions. It is important to juxtapose the transaction at issue against the office held by the agent and ask whether that transaction can reasonably be understood as falling within a class of transactions which someone holding that office would usually be authorised to conclude on the principal’s behalf. If the answer is in the negative, potential counterparties cannot (in the absence of any other relevant indicators emanating from the principal) blithely construe the agent’s holding of that office as a representation by the principal that the agent has the authority to enter into the transaction at issue.

The Plaintiff insisted that a representation by D1 that D2 held the office of director was sufficient in and of itself to constitute a representation that D2 had the authority to bind D1 to the Service Agreement. In support of this, the Plaintiff referred to the case of Heperu Pty Ltd v Morgan Brooks Pty Ltd (No 2) [2007] NSWSC 1438 (“Heperu”). In that case, the principal, Morgan Brooks Pty Ltd (“Morgan Brooks”), was a mortgage originator which had appointed one Mr Cincotta as its manager to source for borrowers seeking mortgage loans. Morgan Brooks’ website listed its various officeholders and specifically named Mr Cincotta as one of its managers. Notably, Morgan Brooks issued a name card to Mr Cincotta describing him as “Director” – this name card was later furnished by Mr Cincotta to the plaintiff, when both men met. Mr Cincotta successfully persuaded the plaintiff to invest funds through Morgan Brooks, after which the plaintiff signed what he thought was an investment contract with Morgan Brooks. Unfortunately, the funds that the plaintiff thought were being invested through Morgan Brooks were misappropriated by Mr Cincotta (who was ultimately made a bankrupt). When the plaintiff sued Morgan Brooks for the return of his investment funds, the latter contended that Mr Cincotta had no authority to enter into the investment contract on Morgan Brooks’ behalf. Morgan Brooks maintained that it was a mortgage originator and, as such, the provision of investment advice or the arrangement of deposits of investor funds were patently beyond its ordinary course of business (at [43]). Notwithstanding, the court held that Mr Cincotta had ostensible authority to enter into the investment contract on behalf of Morgan Brooks (at [84]).

Heperu was endorsed by the High Court in Viet Hai Petroleum Corp v Ng Jun Quan and another and another matter [2016] 3 SLR 887, which observed (at [35]) that the stipulation of an agent’s office on his name card could potentially constitute a representation by his principal (which issued the name card to him) as to the agent’s authority to enter into the transaction concerned. The Plaintiff sought to extrapolate this observation into a broader proposition that “the usual scope of authority of a “director” is the authority to enter into a contract and bind the company”.

With respect, the Plaintiff’s submission failed to attribute proper weight to a critical aspect of the decision in Heperu. In that case, a key submission by the company was that the investment contract which had ostensibly been entered into on its behalf patently fell outside of its ordinary business of mortgage origination. However, the court observed (at [29] and [76]) that this argument was undermined by the company’s own website, which described itself as “an established Fund Manager within the Banking & Finance Industry”. To the world at large, someone in the position of the agent, who held the office of “Director” within a company that had styled itself as a “Fund Manager within the Banking & Finance Industry”, could very reasonably be perceived as having the authority to bind the company to the investment contract at issue (at [77] to [79]). The court had thus scrutinized the nature of the business which the company held itself out as conducting, to see if the contract purportedly entered into on the company’s behalf was of a kind typical to the ordinary course of that business. The court did not – contrary to what the Plaintiff’s submission envisaged – mechanistically latch on to the title of “director” (as reflected in the agent’s name card) as the basis for finding the existence of ostensible authority.

To recapitulate, I had observed (at para 153 above) that in assessing if the principal’s act of appointing an agent to a specific office, or of holding the agent out as occupying that office, suffices to constitute a representation cloaking the agent with ostensible authority to enter a transaction, one needs to juxtapose that transaction against the nature of the office. In particular, one needs to ask if the transaction belongs to a class which someone holding that particular office would usually be authorised to conclude on the principal’s behalf. The case of Heperu is instructive in demonstrating that when one embarks on that assessment, a relevant factor to consider would be the nature of the business which the principal carries out, or at least professes to carry out. If the transaction clearly falls outside the ordinary course of that business, then subject to any other relevant factors, this could make it less reasonable for outsiders to simply assume that the transaction (being out of the ordinary) is one which someone holding the agent’s office has the usual authority to conclude. In turn, that could militate against the conclusion that the appointment of the agent to the office concerned sufficed to constitute a representation cloaking him with ostensible authority to enter into the transaction at issue.

Applying the above principles to the present facts, it is clear to me that while D1’s website may have represented to the world that D2 was a director of D1, this fell far short of cloaking D2 with ostensible authority to enter into the Service Agreement on D1’s behalf:

(a) Firstly, the nature of the Service Agreement was exceptional, to say the least. It embodied a corporate guarantee by D1 that would expose it to unlimited liability, by obliging D1 to replace any cryptocurrencies which had been stolen, with no cap on quantum. Furthermore, the replacement had to be effected within an extremely tight turnaround timeframe of five business days. This bespoke obligation clearly posed tremendous business risk, with the potential to jeopardise D1’s very viability. The commercial hazard was compounded by the fact that the corporate guarantee required D1 to underwrite losses flowing from the fraud of not just its own employees, but the employees of the “Cryptotrage group of companies”. This would cover fraud by any employee of Clickstaff – a company in which D1 held no ownership stake (Clickstaff was wholly owned by Nguyen Thanh) but which had nevertheless been curiously termed by the Service Agreement as the “child company” of D1. Crucially, there was no evidence showing what benefit D1 stood to gain from providing such a corporate guarantee. The Plaintiff claimed that the Service Agreement was meant to induce him to deposit more cryptocurrencies into the Cryptotrage scheme, as this would facilitate the attainment of Tier 9 status by the Binance Account and thereby allow the Cryptotrage scheme to enjoy the financial benefits attendant upon that status (alluded to at para 33 above). Yet, there was nary a wisp of a paper trail linking the operation of the Cryptotrage scheme to D1.

(b) Secondly, unlike in the case of Heperu, the Service Agreement centred on an activity that patently fell outside the scope of D1’s ordinary business. D1 provided information technology services, with particular focus on solutions assisting brokers and traders in their brokerage and trading activities. At the material time, D1’s website stated that it dealt with “AI Products”, “Blockchain Products”, “Software” and “Trading Software”. D1’s focus on technology was echoed in its ACRA business profile, which described D1’s activities as “Other Information Technology and Computer Service Activities (eg, Disaster Recovery Services)”. The Plaintiff sought to capitalise on the fact that D1’s website also indicated that it offered products such as “Broker Solution”, “Trading Solution” and “Crypto Solution”. However, I fail to see where this submission led to. It is true that D1’s technological solutions had focused on a specific business sector, ie, that involving brokerage, trading, arbitrage and cryptocurrency transactions. However, the Cryptotrage scheme entailed a fundamentally different business proposition, involving as it did the onboarding of investor deposits, as well as the management of and trading in client funds. There was no digital or physical publication which the Plaintiff could point to that described D1 as being in the business of accepting (or being licensed by any financial regulator to accept) funds from investors or engaging in trades on their behalf. The Plaintiff himself had been alive to the dearth of any links between the Cryptotrage scheme and D1. It was clearly not lost on him that this was a red flag, given how he was prompted to reach out to Michael Lim to surface his concerns about the absence of any mention of the Cryptotrage scheme on D1’s website (see para 11 above). It is also noteworthy that despite the Plaintiff’s e-mail to Michael Lim explicitly seeking confirmation that the Cryptotrage scheme was indeed D1’s product, the latter’s e-mail in reply (extracted at para 12 above) completely sidestepped any mention of the scheme.

In light of the unusual obligations that the Service Agreement placed on D1 and the fact that the Cryptotrage scheme clearly fell beyond D1’s ordinary course of business, I take the view that D1’s act of holding out D2 as a “director” could not, without more, imply a representation that D2 had the usual authority to execute the Service Agreement. Under the circumstances, it would not be reasonable for a counterparty to unthinkingly take an individual’s appointment as a director to imply that the individual possessed the usual authority to enter into a transaction such as the Service Agreement. This was not a run-of-the-mill commercial deal. Any prudent person looking to extract an indemnity of this order from a company would have sought confirmation from the company’s board of directors. The conclusion must be that D2 had no ostensible authority to execute the Service Agreement on D1’s behalf.

Whether Zee had apparent or ostensible authority

I next consider whether Zee had ostensible authority to enter into the Service Agreement on behalf of D1.

In seeking to establish Zee’s ostensible authority, the Plaintiff relied on the e-mail exchange with Michael Lim on 7 February 2019. As explained at para 11 above, the Plaintiff had written an e-mail to Michael Lim asking if Zee and Rick were indeed from D1’s Vietnam office. The Plaintiff’s e-mail set out a description of the Cryptotrage scheme and enclosed the Deck in which Zee was described as “Director Vietnam” who “Oversees a team of 20 Developers, 6 Traders” (see extract of the Deck’s slide at para 10 above). In response to the Plaintiff’s e-mail, Michael Lim replied with an e-mail confirming that:

The Plaintiff relied on the fact that Michael Lim’s e-mail reply (extracted at para 12 above) had re-attached the first substantive page of the Deck, which set out the credentials of Zee and Rick, as meaning that Michael Lim had effectively confirmed that Zee was a director of D1.

In my view, Michael Lim’s reply e-mail adds little value to the analysis. While the e-mail may have re-attached the first substantive page of the Deck, it said nothing to elaborate on the contents thereof. All that Michael Lim’s e-mail did say was that Zee was from D1’s Vietnam office. I find it difficult to comprehend how such a terse response could be extrapolated into a representation that Zee had the authority to sign the Service Agreement. No insights could be gleaned from Michael Lim’s e-mail as to whether Zee had any authority to enter into transactions relating to the Cryptotrage scheme on D1’s behalf, or whether the scheme was even part of D1’s business to begin with. I should add for completeness that as at the point of trial, Michael Lim could not be found and could thus not be called to testify.

The Plaintiff also relied on D1’s website, on which Zee was held out to be “Director, Snap Vietnam”. I have explained at para 158 above why the fact that D1’s website held out D2 as a director did not suffice to clothe him with ostensible authority to enter into the Service Agreement on D1’s behalf. Those reasons apply with equal force to Zee.

I conclude with the observation that any notion of Zee having possessed ostensible authority by virtue of D1 holding out Zee as one of its directors is somewhat undermined by the Plaintiff’s testimony that he was uncomfortable with Zee being the only signatory to the Service Agreement (see para 16 above). It was precisely to placate the Plaintiff’s concerns that Zee offered to co-opt his “boss” (ie, D2) into signing the same. As explained at para 142 above, the third of the Freeman & Lockyer conditions requires that the contracting party be induced by the representation to enter into the contract. By the Plaintiff’s own admission, the representation that Zee held the title of director was not enough to induce the Plaintiff to enter into the Service Agreement.

In light of the analysis in this section, I find that Zee, like D2, had no ostensible authority to enter into the Service Agreement on D1’s behalf.

Issue 3: Whether D2 was liable for breach of warranty of authority

I next consider the Plaintiff’s claim that D2 was liable for breach of warranty of authority.

The nub of the Plaintiff’s cause of action can be summarised as follows. By way of his LinkedIn profile, D2 held himself out as D1’s managing director, thereby effectively warranting that he was authorised to act on D1’s behalf. On the faith of that warranty, the Plaintiff executed the Service Agreement, while under the impression that he was transacting with D1 through the agency of D2. Thereafter, thinking that he had obtained the security of the corporate guarantee under the Service Agreement, the Plaintiff was emboldened to increase his exposure to the Cryptotrage scheme, which naturally caused him to suffer larger losses when Zee perpetrated his fraud. If D2’s warranty is found to be untrue, such that D1 is not bound by the Service Agreement, the Plaintiff would have lost his right to be indemnified by D1 for those losses. In that eventuality, the Plaintiff contended that his losses may properly be attributed to D2’s breach of warranty.

In response, D2 argued that he never made any representations to the Plaintiff as to his authority to act on behalf of D1. Both he and the Plaintiff were complete strangers, having never met prior to the conference call on 16 February 2021 (alluded to at para 38 above). This meant that D2 could not have made any representations to the Plaintiff prior to the latter signing the Service Agreement.

The elements required to establish an action for breach of warranty of authority are stated in Peter Watts & F.M.B. Reynolds, Bowstead and Reynolds on Agency (Sweet & Maxwell, 22nd Ed, 2021) (“Bowstead and Reynolds”) at para 9-060. In Fong Maun Yee and another v Yoong Weng Ho Robert [1997] 1 SLR(R) 751 (“Fong Maun Yee”), the Court of Appeal endorsed those elements (at [53]), which I cite below:

Reverting to the present case, the Plaintiff’s claim for breach of warranty of authority is in my view without merit.

The typical factual matrix under which a cause of action for breach of warranty of authority may arise is when the purported agent warrants that he is authorised to transact on the putative principal’s behalf and this induces the claimant to transact with the purported agent, during which the claimant labours under the impression he is actually transacting with the putative principal (through the purported agent). If the warranty is unfounded and the purported agent was not empowered to bind the putative principal to the transaction after all, the claimant may seek recourse against the purported agent for loss arising from the transaction with the purported agent (as was the case in Fong Maun Yee). The present case is nevertheless atypical, in that the Plaintiff was not induced by the warranty of authority from the purported agent (ie, D2) to transact with the latter. In fact, there was never any such transaction between the Plaintiff and D2: while the Plaintiff executed the Service Agreement, that document was never signed by D2 (whose signature on P1 was forged). The only person whom the Plaintiff did transact with was Zee.

The Plaintiff sought to rely on Chu Said Thong and another v Vision Law LLC [2014] 4 SLR 375; [2014] SGHC 160 (“Chu Said Thong”) (at [230]). That case involved a fraudster who claimed to have been assigned an option to purchase a property. In truth, the option was fabricated and the property owner had never granted any such option. Under the guise of various fake personalities, the fraudster approached the plaintiffs and offered to assign the option to them. He also set the stage by approaching the defendant law firm to engage them to act in the transfer of the property to the plaintiffs. When one of the plaintiffs called the defendant law firm to seek verification, the latter (through its secretary) made two crucial representations (at [50(b)]), ie:

The Plaintiff did not explain how the factual matrix in Chu Said Thong supported his claim for breach of warranty of authority. It is unclear if the Plaintiff was seeking to draw any parallels, eg, by relying on how the purported agent in Chu Said Thong (ie, the defendant law firm) was held liable to make good the loss sustained by the plaintiffs not only from the latter’s dealings with the purported agent, but also from their direct dealings with a third-party fraudster (the latter being the $105,200 that the plaintiffs paid directly to the fraudster for the option). In the present case, the Plaintiff is similarly seeking to make D2 (as the purported agent of D1) liable for loss sustained by the Plaintiff in dealing directly with an alleged third-party fraudster (ie, Zee).

It is apt at his juncture to highlight that a claimant can bring an action against the purported agent for breach of warranty of authority, in respect of loss suffered through the claimant’s dealings with a party other than the purported agent, if the purported agent’s warranty pertained to the authority of that other party: Bowstead and Reynolds at para 9-064. Thus, if the purported agent falsely warrants to the claimant that a third party was authorised to transact on the putative principal’s behalf, and this induces the claimant to deal with that third party on the faith that the claimant is actually dealing with the putative principal (through the third party), the purported agent is potentially liable to the claimant for loss flowing from those dealings. Nevertheless, that is not a principle which the Plaintiff can avail himself of on the facts of this case, where there was no warranty of any sort emanating from D2 (whether in his LinkedIn profile or otherwise) touching on Zee’s authority. There was no warranty from D2 that could be said to have induced the Plaintiff into thinking that in transacting with Zee, the Plaintiff would be transacting with D1 (or D2 for that matter). In that respect, the case of Chu Said Thong is entirely distinguishable. In that case, the representations emanating from the purported agent (being the defendant law firm) alluded not only to the authority of the purported agent itself but also to the authority of the third-party fraudster – the defendant law firm expressly represented to the plaintiffs that the fraudster had been issued the option by the property owner (see paragraph 172(b) above).

The Plaintiff’s claim for breach of warranty of authority could thus not be sustained.

Before leaving this section, I should point out that even if the claim was well-founded, the Plaintiff did not sufficiently prove the full extent of his loss purportedly arising from D2’s breach of warranty of authority. While this observation could have been canvassed in the section below touching on the Plaintiff’s failure to properly quantify his loss (under the heading “Issue 6”), it would be more convenient to deal with it here. The objective of awarding damages for a breach of warranty of authority is to put the innocent party, as far as possible, in the position which he would have been in had the warranty of authority been true: Ku Yu Sang v Tay Joo Sing and another [1993] 3 SLR(R) 226 at [51]. The warranty at issue thus pertains to the purported agent’s authority to bind the putative principal to the contract and not the putative principal’s ability to fulfil that contract. As such, if the putative principal would in any event have been unable to perform the contract to any significant degree and also lacks the resources to satisfy any judgment for breach of contract, the innocent party may potentially not be allowed to recover the full extent of what he expected to reap from the principal performing the contract: see also Tan Cheng Han SC, The Law of Agency (Academy Publishing, 2nd Ed, 2017) at paras 09.067–09.068. In the present case, there was no evidence to indicate the extent to which D1 would have been able to honour the corporate guarantee in the Service Agreement. There was no evidence as to D1’s balance sheet or cash flow to establish that D1 was in any position to make good on such a large payment.

Issue 4: Whether D1’s obligation to indemnify was triggered

Given my finding that the Service Agreement was not authentic, it is not strictly necessary for me to determine whether its clauses could be construed as supporting the Plaintiff’s claim. I will nevertheless state my opinions on this, given the substantial arguments advanced by the parties.

The principal clause in the Service Agreement relied upon by the Plaintiff in seeking an indemnity from D1 is cl 1.2, which reads:

The Plaintiff contended that the following two occurrences could be regarded as “fraud” within the meaning of cl 1.2, thereby triggering D1’s obligation under that clause to indemnify the Plaintiff for the ensuing losses:

(a) Zee’s commingling of the cryptocurrencies belonging to that of the Plaintiff (which were meant for the Cryptotrage scheme) with that of Torque’s investors, which occurred when Zee allowed the Binance Account to be used in connection with the Torque trading platform. This ultimately led to the Plaintiff being unable to withdraw whatever remained of his cryptocurrencies in the Binance Account when the account was frozen pending investigations into Torque’s affairs.

(b) Zee’s depletion of the cryptocurrencies invested by the Plaintiff in the Cryptotrage scheme, either through Zee’s fraudulent trades or his outright theft.

I begin with the first alleged occurrence of fraud, pertaining to the commingling of cryptocurrencies. To be clear, the Plaintiff’s case was that the very act of commingling his cryptocurrencies in the Binance Account with that of Torque’s investors would, in and of itself, constitute “fraud” under cl 1.2.

D1 objected to the Plaintiff’s assertion that his digital assets in the Cryptotrage scheme had been commingled with that of investors under the Torque platform as that point was not sufficiently pleaded. In response, the Plaintiff contended that the allegation clearly emerged from paras 31(b) and (c) of his amended Statement of Claim, which read as follows:

I am inclined to agree with D1 on this point. I do not think that these sub-paragraphs within the Plaintiff’s pleadings provided sufficient particulars of the allegation of commingling. All that sub-paragraph (b) stated was that Zee had used the assets in the Cryptotrage scheme to engage in unauthorised trades. Sub-paragraph (c), on the other hand, spoke of unauthorised transfers of the funds from the Cryptotrage scheme into accounts managed by Torque – this was inconsistent with the Plaintiff’s current position that cryptocurrencies from Torque’s investors had been deposited into the Binance Account, which was by then already in use for the Cryptotrage scheme.

In any case, the Plaintiff failed to demonstrate why the commingling of cryptocurrencies was in any way wrong, much less a “fraud”. There was no indication – whether in the Plaintiff’s AEIC, the Service Agreement, or in any other document – of any restriction prohibiting Zee from depositing the cryptocurrencies of Torque’s investors (or of anyone else, for that matter) into the Binance Account. So long as the deposits were properly segregated into sub-accounts within the Binance Account, with clear records to track the balances belonging to different investors, it seems that there would have been no harm arising from the sharing of the Binance Account. On the contrary, depositing the cryptocurrencies of Torque’s investors into the Binance Account appeared to be a natural step forward because:

(a) Torque was supposedly a “spin-off” from Cryptotrage; and

(b) Increasing the level of cryptocurrency deposits in the Binance Account helped to maintain the Tier 9 status that it enjoyed. As explained at para 33 above, that status conferred financial benefits such as lower trading fees, which would result in lower costs to investors of the Cryptotrage scheme (including the Plaintiff himself).

I thus agree with the submission of D1 that the Plaintiff cannot rely on the commingling of cryptocurrencies as an instance of fraud that triggered cl 1.2 of the Service Agreement.

Next, I examine whether the acts by which Zee had allegedly depleted the Plaintiff’s cryptocurrencies in the Binance Account amounted to “internal fraud by any staff” which resulted in those cryptocurrencies being “stolen by fraud”. D1’s submissions on this point were as follows:

(a) There could not have been any “internal fraud by any staff” as Zee was not “staff” but an independent contractor of D1.

(b) There was no evidence that the cryptocurrencies allegedly depleted by Zee’s unauthorised trades or theft were those invested under the Cryptotrage scheme (as opposed to those of Torque’s investors).

D2 similarly argued that cl 1.2 was never engaged because the fraud committed by Zee was in relation to only the digital assets managed by Torque and not those invested under the Cryptotrage scheme.

Even if I accept the Plaintiff’s argument and adopt a broad interpretation of the word “staff” to include Zee (notwithstanding D1’s contention that Zee was nothing more than an independent contractor), the Plaintiff still failed to show how cl 1.2 of the Service Agreement would have been brought into play. In particular, the Plaintiff failed to offer sufficient particulars of the alleged “internal fraud” and explain how his cryptocurrencies had been “stolen by fraud”, these being the triggers specified in cl 1.2. As D1 pointed out, there was no evidence showing just how much of the cryptocurrency deposits within the Binance Account had been depleted or withdrawn by Zee, or demonstrating that any cryptocurrencies misappropriated by Zee were indeed attributable to the Cryptotrage scheme (rather than Torque’s investors). The Plaintiff did not even see fit to call Torque’s liquidators as witnesses to shed light on these important questions. I thus agree with the Defendants that there is insufficient evidence for me to affirmatively rule that the Plaintiff’s cryptocurrency deposits in the Cryptotrage scheme had been the subject of fraud.

For the reasons above, I find that even if the Service Agreement were valid and binding upon D1, the Plaintiff would nevertheless have failed to prove that cl 1.2 of the Service Agreement came into operation. D1 would thus have been under no obligation, by virtue of that clause, to indemnify the Plaintiff for the sum which the latter now claims.

For completeness, I do not find that cl 1.1 was breached as no contractual obligation appears to be housed within that clause.

Issue 5: Whether the Defendants owed the Plaintiff a duty to supervise Zee

The Plaintiff’s alternative claim was that the Defendants each owed him a duty to supervise Zee, which they breached by not exercising proper oversight, thereby allowing Zee to perpetrate the fraud which caused the Plaintiff loss.

Whether D1 had a duty to supervise Zee

The Plaintiff submitted that D1 had a duty to put in place safeguards to supervise Zee’s activities in Vietnam and to prevent commingling of the Plaintiff’s cryptocurrencies in the Binance Account with those belonging to Torque’s investors. According to the Plaintiff’s Statement of Claim, that duty was a “fiduciary, contractual or tortious duty owed to the Plaintiff in law and in fact”.

I begin with the observation that the Plaintiff seemed to have abandoned this submission in his Closing Submissions, where he stated rather cryptically:

However, in his subsequent Reply Submissions, the Plaintiff resuscitated this head of claim by asserting that the “authorities are clear” that D1 owed a duty to third parties in tort, and that this duty was breached by D1’s neglect in allowing Zee to misappropriate the digital assets in the Binance Account. In reply, D1 maintained that it was under no duty to supervise Zee because the latter was merely an independent contractor and not a member of D1’s staff.

In Skandinaviska (CA), an argument similar to that canvassed by the Plaintiff was raised and rejected. In that case, the defendant company’s employee, Chia, had forged the board resolutions of the defendant company which he then used to dupe the plaintiff bank into extending credit facilities to (ostensibly) the defendant company. Chia drew huge sums from those facilities, which he then misappropriated. The plaintiff bank claimed that its losses arose from the defendant company breaching its duty to properly supervise Chia (at [97]). The Court of Appeal rejected this argument and held that the requisite degree of proximity was absent, as between the plaintiff bank and the defendant company, to satisfy the test set out in Spandeck Engineering (S) Pte Ltd v Defence Science & Technology Agency [2007] 4 SLR(R) 100 (“Spandeck Engineering”) for establishing a common law duty of care in negligence (at [99]–[103]). Amongst other things, the Court of Appeal in Skandinaviska (CA) noted (at [103]–[104]):

The Court of Appeal also touched on the issue of causation, remarking that even if a duty of care existed, the loss did not arise from a breach of that duty. Rather, the proximate cause of the plaintiff bank’s loss was its own negligence in readily accepting a forged document without performing the necessary verifications:

In the present case, the Plaintiff simply glossed over the test in Spandeck Engineering, asserting that:

There was no elaboration accompanying this one-liner.

Under the circumstances, I am compelled to conclude that the Plaintiff’s submission that D1 owed him a duty of care to supervise Zee is devoid of merit.

Whether D2 had a duty to supervise Zee

I now turn to the Plaintiff’s claim that a duty was also owed by D2 to supervise Zee, which D2 breached.

The Plaintiff contended that D2’s duty to supervise Zee arose by virtue of his position as managing director of D1 and CEO of Torque, both of which gave him the ability to prevent Zee’s fraud. The Plaintiff further claimed that D2 breached that duty when he failed to exercise due diligence or control over Zee, thereby allowing the latter to siphon away the Plaintiff’s cryptocurrencies. In particular, the Plaintiff argued that D2 displayed Nelsonian blindness in the following areas:

(a) D2 knew that Zee, as a director of Snap Vietnam, was involved in the Cryptotrage scheme, but still allowed Zee to concurrently hold the office of CTO in Torque; Zee was thus allowed to wear two hats and thereby place himself in a position of conflict of interest.

(b) D2 knew, or ought to have known, that cryptocurrencies invested in the Cryptotrage scheme were being commingled with that of Torque’s investors, but nevertheless acquiesced to this state of affairs.

In reply, D2 maintained that he was not at any material time responsible for D1’s Vietnam operations. More importantly, he was merely an independent contractor of D1 and thus could not have possibly owed any duty to supervise members of D1’s team.

It is well established that directors owe fiduciary duties and legal responsibilities to their companies: see the High Court’s remarks in Vita Health Laboratories Pte Ltd and others v Pang Seng Meng [2004] 4 SLR(R) 162 (“Vita Health”) at [14]. The duties owed by a director include the duty to properly supervise the subordinates to whom the director has delegated his functions: see Vita Health at [21]. The views in Vita Health were re-affirmed by the Court of Appeal in PlanAssure PAC (formerly known as Patrick Lee PAC) v Gaelic Inns Pte Ltd [2007] 4 SLR(R) 513 at [130].

The Plaintiff failed to explain why any such duty which a director may owe to his company to properly supervise his subordinates should be extended, such that the duty is also owed to an outsider such as the Plaintiff. The Plaintiff cited the case of TV Media Pte Ltd v De Cruz Andrea Heidi and another appeal [2004] 3 SLR(R) 543, where the Court of Appeal lifted the corporate veil and imputed liability for a company’s tortious acts to the director responsible for the company’s breach of duty. In that case, the company was held liable in negligence for selling slimming pills that ultimately damaged the plaintiff’s liver. In finding that the company’s director should be held personally liable as well, the court held (at [145]):

That case does not assist the Plaintiff, who failed to point to any tortious liability of the corporate entities involved, for which D2 could be made personally liable by way of piercing the corporate veil:

(a) In respect of D1, I have already explained at para 192 above that the Plaintiff failed to justify the existence of any duty owed by D1 to the Plaintiff at law, in relation to the supervision of Zee, the breach of which could have given rise to an actionable tort.

(b) In respect of Torque, the Plaintiff similarly failed to explain the basis for imputing any duty owed by Torque to the Plaintiff, which was breached by Torque. In fact, the Plaintiff’s submissions did not allude to any duty owed to him by Torque at all.

I should also add that in respect of D1, even if there had been any tortious liability on the part of the company, the facts of the present case do not come anywhere close to the threshold for lifting the corporate veil and imputing D1’s tortious liability to D2. Unlike the director in TV Media, D2 was not in any way D1’s “controlling mind and spirit”. As explained at para 131 above, D2 did not even meet the test for being a de facto director. There was no evidence of D2 having exercised any significant responsibility over D1’s business or over Zee’s activities in Vietnam (other than Zee’s statement that D2 was his “boss”).

For these reasons, I find that D2 owed no duty to the Plaintiff to supervise Zee. Any action for breach of D2’s duty to supervise Zee must therefore fail.

Issue 6: Whether the Plaintiff properly quantified his loss

Even if the issue of liability had been decided in the Plaintiff’s favour, I would nevertheless find that the Plaintiff has not properly quantified his loss. Specifically, he failed to prove the quantum of cryptocurrencies which he allegedly lost following Zee’s fraud.

The Plaintiff sought to recover the value of the cryptocurrencies which he claimed to have deposited in the Cryptotrage scheme. He alleged that he had deposited three different cryptocurrencies which, at the time Zee’s fraud was discovered, collectively added up to US$9,122,044 in value. As explained at para 47 above, the Plaintiff derived this sum by adding up the US dollar value of the balance of each of his three cryptocurrency deposits, existing as at the end of 7 February 2021, as encapsulated in the 8 February 2021 daily reports:

The Plaintiff’s methodology for calculating his balance for each cryptocurrency can be gleaned from the daily reports for 8 February 2021:

(a) The Plaintiff’s claim for 656,422 USDt was supported by a daily report that Snap Vietnam sent via Telegram on 8 February 2021. That report enclosed a table, extracted below, purporting to set out his daily USDt balances from 30 January to 7 February 2021:

A perusal of the table above indicates that the Plaintiff took the finishing balance for 7 February 2021 (ie, 659,997.160 USDt, as indicated in the final row under the “Finish” column) and subtracted the corresponding “Trading Fee” (ie, 1,979.99148 USDt, as indicated in the adjacent cell) and Snap Vietnam’s 20 per cent cut (ie, 1,594.5875 USDt, as indicated in the final row under the “Our Profit 20%” column), to arrive at the final round figure of 656,422 USDt.

(b) The Plaintiff’s claim for 3,073,178 BUSD was supported by a daily report that Snap Vietnam sent via Telegram, also on 8 February 2021. That report enclosed a table, extracted below, purporting to set out his daily BUSD balances from 4 to 7 February 2021:

A perusal of the table above indicates that the Plaintiff likewise took the finishing balance for 7 February 2021 (ie, 3,075,512.2426 BUSD, as indicated in the final row under the “Finish” column) and subtracted the applicable fees (ie, 2,334.6453 BUSD, as indicated under the “Management and fees (20%)” column) to arrive at the final round figure of 3,073,178 BUSD.

(c) The Plaintiff’s claim for 21,772.06 BNB was supported by daily reports that Snap Vietnam had sent via two Telegram messages, both dated 8 February 2021. The daily reports in both messages collectively enclosed five tables, each purporting to set out the daily BNB balances of the Plaintiff and (what appeared to be) various other investors, for the period spanning 4 to 7 February 2021:

The titles of the five tables suggested that they pertained to accounts belonging to persons with the names “Daryl”, “Chia HL”, “Fotis” and “George”, with the name “George” (which appeared in two out of the five tables) presumably referring to the Plaintiff. A perusal of the final row of each table indicates that the Plaintiff derived his final figure of 21,772.06 BNB by taking the finishing balance for 7 February 2021 (under the “Finish” column) and subtracting the applicable fees (under the “Management and fees (50%)” column) before summing up the result across the five tables, ie, (10,198.27549 – 20.4670) BNB + (754.7250 – 1.5147) BNB + (8,834.9008 – 17.7309) BNB + (306.1170 – 0.6143) BNB + (1,721.9921 – 3.4559) BNB ≈ 21,772.06 BNB.

Preliminarily, D1 submitted that the Plaintiff should not be allowed to claim for losses in all three categories of cryptocurrencies (ie, USDt, BUSD and BNB) because the ambit of the Service Agreement was restricted to covering only the Plaintiff’s USDt deposits. That being the case, argued D1, the corporate guarantee under the Service Agreement could not extend to losses in BUSD and BNB. D1 sought to justify its submission as follows:

(a) When the Service Agreement was executed on 27 May 2019, the Plaintiff had deposited only USDt for the Cryptotrage scheme. His deposits of BUSD and BNB were made later, meaning that these latter two classes of cryptocurrencies were not contemplated – and consequently not covered – by the Service Agreement.

(b) Further, the Service Agreement was clearly focused on cryptocurrency trades. Specifically, cl 1.1 of the Service Agreement laid out D1’s obligation “[t]o manage the funds of The Client or his Group under the scope of performing Cryptocurrency Arbitrage trades”. The BNB deposits thus fell outside the scope of the Service Agreement because BNB had been deposited by the Plaintiff not for the purpose of trading but for “stacking”. “Stacking” entailed accumulating the BNB deposits in an account on the Binance Exchange to levels that would allow that account to achieve Tier 9 status, thereby triggering the benefits of referral fees and discounted trading costs alluded to at para 33 above. D1 contended that BNB had never actually been traded in any arbitrage transactions, pointing to the further and better particulars furnished by the Plaintiff wherein he stated that “the BNB Cryptotrage scheme was a purely stacking project without any trading in the BNB funds invested by the Plaintiff”. If BNB had never been traded, argued D1, BNB deposits should be regarded as falling outside the ambit of the Service Agreement and thus outside the guarantee in cl 1.2.

D1’s attempt to limit the categories of cryptocurrencies covered by the Service Agreement can be dealt with briefly. In short, I disagree that the Service Agreement covered only USDt and not BUSD or BNB deposits:

(a) First, cl 1.2 employed the term “Funds managed under Cryptotrage books”, without drawing any distinction between cryptocurrencies managed at the time the Service Agreement was signed and cryptocurrencies flowing in after the Service Agreement had been signed. This clause was therefore wide enough to cover the subsequent BNB and BUSD deposits.

(b) Second, I disagree with D1’s submission that the Service Agreement covered only cryptocurrencies that were traded. The use of the word “trades” in cl 1.1 is neither here nor there. There was nothing in the Service Agreement that explicitly required the cryptocurrencies to be “traded” before they were covered by the guarantee. The Plaintiff’s BNB deposits, even if not traded, still served the pivotal function of stacking the digital assets in the accounts on the Binance Exchange to a level that unlocked Tier 9 status. That in turn generated earnings and savings that significantly boosted the Cryptotrage scheme’s profitability. Given the central role played by the Plaintiff’s BNB deposits, it is not apparent to me that the contracting parties would have intended for BNB to fall outside the protection which the corporate guarantee in cl 1.2 offered.

Having said that, I hold the view that the Plaintiff nevertheless failed to satisfactorily establish his loss. This was due in large part to the fact that his attempts at quantification hinged principally on the daily reports, which I find to be riddled with serious evidential shortcomings.

Firstly, the Plaintiff was effectively using the daily reports to prove the truth of what those daily reports purported to assert (ie, the daily balances of the cryptocurrency concerned). This was plainly hearsay evidence. Without the benefit of argument, it would appear that a possible way of overcoming this evidential hurdle might have been for the Plaintiff to invoke s 32(1)(b) of the EA, which allows for the admission of statements made in the ordinary course of a trade, business or profession. To invoke that exception, it would have been incumbent on the Plaintiff to at least present some evidence shedding light on, for instance, who prepared the daily reports and the processes attendant upon their preparation. As it were, nothing of the sort was adduced by the Plaintiff.

Secondly, there was a dearth of evidence as to the source from which the numbers within the daily reports were taken. While it is possible to make some sense of the tables in the daily reports and thereby use them to derive the components of the US$9,122,044 claimed by the Plaintiff (as per the exercise at para 202 above), question marks still remain as to where the constituent numbers within these tables were derived from:

(a) No explanation was provided as to the source of the starting balances in the earliest daily reports. If payments had been made to purchase the initial cryptocurrency balances, there were no receipts to document them. As D1 pointed out, the Plaintiff failed to provide any accounts from the Binance Exchange, or verifiable transaction IDs or hashes of his deposits.

(b) One sees from the daily reports that the finishing balance for each day would, after making the necessary deductions (such as Snap Vietnam’s fees), constitute the next day’s starting balance. Gains from the next day’s trades would then be added to the next day’s starting balance to yield the next day’s finishing balance. The process of accounting for the daily gains and deductions in this manner culminated in the finishing balances for 7 February 2021, which was used by the Plaintiff to calculate his claim figure of US$9,122,044. However, he failed to provide details of the underlying daily trades, which would have been a key driver of the finishing balance captured every day.

The Plaintiff asserted that the court should simply treat the daily reports as “akin to a bank statement”. However, without the necessary evidence to address the gaps above, the tabulated figures in the daily reports were nothing more than a series of inscrutable black boxes.

Thirdly, as regards the BNB daily reports, the Plaintiff’s claim incorporated balances that appeared to be in the names of persons other than himself (ie, “Daryl”, “Chia HL” and “Fotis” – see para 202(c) above). He did not explain why he should be regarded as having the standing to include those balances in his claim.

Finally, the very authenticity of the daily reports was at issue. By his own pleadings, the Plaintiff described the daily reports as “completely bogus”. I set out below the relevant sections from his Statement of Claim:

At trial, the Plaintiff attempted to qualify his pleaded case by saying that it was only the 7 February 2021 daily reports (which would have reflected the daily balances up to and including 6 February 2021) that were bogus. He explained that prior to that day, all his withdrawal requests had been honoured by Snap Vietnam, thereby implying that the daily reports for those prior days must have been in order. However, it is difficult to see how this explanation assisted his attempt at quantification:

(a) Firstly, as stated at para 207(b) above, it is apparent from the daily reports that the finishing balance for any one day would, after the necessary deductions constitute the next day’s starting balance. Gains from the next day’s trades would then be added to the next day’s starting balance to yield the next day’s finishing balance, and the cycle is repeated. As seen at para 202 above, the Plaintiff purported to quantify his loss by using the finishing balances for 7 February 2021, as captured in the 8 February 2021 daily reports. For reasons just explained, those balances would have been derived by iteratively accounting for the finishing balances of the preceding days, including those for 6 February 2021 as captured in the 7 February 2021 daily reports. If the 7 February 2021 daily reports were indeed bogus (as per the Plaintiff’s revised contention), that necessarily puts into question the integrity of the 6 February 2021 finishing balances captured therein. As these balances would have fed directly into the 7 February 2021 finishing balances that was used to calculate the Plaintiff’ loss, his calculations would still be tainted by the bogus figures.

(b) Secondly, it was the Plaintiff’s case that Zee had been conducting unauthorised futures trading with the Plaintiff’s cryptocurrencies, with the result that the daily reports may have been inaccurate because of their failure to reflect those trades. The Plaintiff discovered that Zee had been performing the unauthorised trades for months leading up to February 2021, meaning that the trades were being carried out for a significant duration, leading up to when the 8 February 2021 daily reports were prepared. That would similarly render the daily reports for the days leading up to (and including) 8 February 2021 suspect, thereby undermining the Plaintiff’s calculations.

(c) Thirdly, the fact that the Plaintiff was able to make some withdrawals of his investments in the Cryptotrage scheme prior to 7 February 2021 would at best indicate that there were some balances remaining of the Plaintiff’s cryptocurrency deposits in the Binance Account up to that day. This had no bearing on whether the figures in the daily reports accurately reflected the quantum of those balances.

The Plaintiff was also unable to adduce any evidence to show how much of his cryptocurrencies in the Binance Account had been lost following Zee’s alleged unauthorised trades and theft. Notably, the Plaintiff could not confirm how much of his cryptocurrencies remained in the Binance Account, which had since come under the control of Torque’s liquidators. The Plaintiff had submitted a proof of debt with Torque’s liquidators seeking a sum of approximately US$14.8 million, which included not only his investments under the Cryptotrage scheme but also under the Torque platform (since both used the same Binance Account, which has since come under the control of Torque’s liquidators). Against that backdrop, the Defendants highlighted fears that the Plaintiff could get a windfall, particularly if Torque’s liquidators eventually return a sizeable portion of the Plaintiff’s cryptocurrency deposits to him after the Defendants have already paid substantial damages. The Plaintiff sought to downplay this concern by alleging that he will likely “only end up with 0.2 per cent or 0.02 per cent” (he was unable to recall the exact amount). However, there is nothing to substantiate these percentage figures. As alluded to at para 184 above, the Plaintiff made no attempt at obtaining evidence from Torque’s liquidators in relation to the assets that remained in the Binance Account, ostensibly because he did not wish to “harass” them while they did their work. However, his failure to even try to procure such evidence meant that the risk of overcompensation remained very much a live one.

In sum, even if the Plaintiff succeeded in establishing the Defendants’ liability (which he has not), I would nevertheless conclude that the Plaintiff has failed to satisfactorily quantify his loss. There is simply no reliable basis on which the Plaintiff can lay claim to the quantum of cryptocurrencies as set out in para 47 above, which he sought to recover by this suit.

Conclusion

For the above reasons, I dismiss the Plaintiff’s claims against the Defendants. In summary:

(a) The Plaintiff failed to adduce sufficient evidence to prove the Executed Paper Copy and its contents. Specifically:

(i) The Plaintiff failed to abide by the evidential procedures for adducing the E-Copy (and hence P1) as secondary evidence proving the Executed Paper Copy.

(ii) Even if the E-Copy could be adduced as secondary evidence, the Plaintiff failed to establish its authenticity (and hence the authenticity of P1), especially given the dubious circumstances surrounding the E-Copy’s provenance. Additionally, the Defendants were able to show that D2’s signature on P1 was likely copy-pasted and thereby establish that P1 (and hence the E-Copy) was forged. Given that the E-Copy and P1 had not been shown to be authentic, they were consequently inadmissible.

The above findings meant that the Plaintiff failed to prove that D2 had signed the Service Agreement.

(b) Even if it were accepted that D2 had signed the Service Agreement, neither D2 nor Zee possessed any actual authority to do so on D1’s behalf. Both men also had no ostensible authority to execute the Service Agreement on D1’s behalf.

(c) The claim against D2 for breach of warranty of authority failed. This was not a typical scenario where the Plaintiff, in reliance on D2’s purported warranty of authority, transacted with D2 while under the erroneous impression that the Plaintiff was dealing with D1 through D2. There was no such transaction with D2, given that D2 never signed the Service Agreement executed by the Plaintiff. The person whom the Plaintiff did transact with was Zee, but D2 never made any warranty about Zee having authority in any shape or form, which the Plaintiff could claim to have relied on when dealing with Zee.

(d) Even if D2 had successfully bound D1 to the Service Agreement, the operative clause obliging D1 to indemnify the Plaintiff for his losses (ie, cl 1.2) was not triggered. The Plaintiff failed to sufficiently demonstrate that there was “internal fraud” leading to the loss of his cryptocurrencies, or that his cryptocurrencies had been “stolen by fraud”, as contemplated by that clause. That D1’s obligation to indemnify could not have come into operation even if it existed meant that the Plaintiff’s claim against D1 for breach of contract had failed.

(e) The Plaintiff’s claim against the Defendants for the alleged breach of their duty to supervise Zee was devoid of merit. The Plaintiff failed to show that such a duty even existed to begin with.

(f) Finally, the Plaintiff failed to properly quantify his loss. His attempts to do so hinged heavily on the daily reports, which were riddled with serious evidential shortcomings. In any case, the Plaintiff has filed a proof of debt with Torque’s liquidators to claim whatever was left of his cryptocurrencies in the Binance Account following Zee’s fraud, but failed to adduce any evidence substantiating his assertion that the amount which he is likely to recover will be negligible.