Ng Kuan Chuan and another v Public Prosecutor and another appeal [2026] SGHC 5

This case concerns two related appeals – one by Ng Kuan Chuan (“Ng”) and the other by Leong Koon Wah (“Leong”). Both individuals (referred to collectively as “the Appellants”) were convicted of various offences after a trial before the District Judge (“DJ”) and sentenced accordingly: see the DJ’s grounds of decision (“GD”) in Public Prosecutor v Leong Koon Wah [2025] SGDC 4. Dissatisfied, Ng filed HC/MA 9001/2025/01 (“MA 9001”) appealing against both conviction and sentence, while Leong filed HC/MA 9008/2025/01 (“MA 9008”) appealing against sentence only.

I dismiss Ng’s appeal against his convictions. I also partially allow the Appellants’ appeals against sentence by reducing the imprisonment term imposed in respect of one of the charges that Ng and Leong each faced.

Background

Leong was the founder, CEO and managing director of a company called Singliworld Pte Ltd (“Singliworld”). Singliworld had marketed an investment scheme called “Singliforex”, under which persons investing in the scheme (“Investors”) would earn returns from foreign exchange (“forex”) trades that were: (a) conducted by professional traders on behalf of the Investors; and (b) matched by brokerage firms against liquidity providers in the market. Any trading gains reaped by the professional traders on behalf of the Investors would thus translate into losses to the liquidity providers, and vice versa.

The forex trades were purportedly executed through two brokerage firms:

(a) From March 2014 to the end of April 2015, the brokerage firm purportedly executing the forex trades was a Hong Kong company called Triumph Global (Asia) Ltd (“TG”). The registered directors of TG were Ng (between January 2014 and December 2014) and subsequently Leong (from December 2014 to July 2015) . During the period that TG was purportedly executing the forex trades, the Investors were told to deposit their investment capital into TG’s HSBC bank account. TG was subsequently placed on the Monetary Authority of Singapore’s Investor Alert List on 16 December 2014.

(b) In May 2015, the brokerage operations in respect of the Singliforex scheme were transferred from TG to a New Zealand company called Union Markets Ltd (“UM”). The Prosecution’s case was that Ng was a manager and Leong a shadow director in UM at the material time. When UM took over the brokerage operations, funds from fresh Investors were placed with a remittance company called “Trustworthy Services”.

The purported forex trades were conducted on, and the corresponding gains and losses purportedly arising from these trades were captured by, a trading system on the brokerage firms’ computers called “MT4”.

In reality, the Singliforex scheme was a sham. There were no professional traders trading on the Investors’ behalf. Rather, it was Ng and Leong who keyed the trades on behalf of the Investors into the brokerages’ MT4 trading systems. There were also no liquidity providers on the other side of the trades, meaning that the Investors were not trading through the brokerages against liquidity providers in the market but against the brokerages themselves. In other words, the brokerages were trading as principals against the Investors. It was thus the Prosecution’s case that the Appellants controlled both sides of the trading transaction, in that they made trading decisions on behalf of the Investors and executed the same, while at the same time executing those same trades on behalf of the counterparty (being the brokerages).

As the brokerages were trading as principals (by assuming the counter-position to the trade rather than matching it against liquidity providers in the market), this meant that any trading gains by the Investors had to be borne by the brokerages themselves, while trading losses by the Investors would translate into gains for the brokerages. By and large, the trades conducted by Ng and Leong resulted in trading gains to the Investors (and thus losses to the brokerages). The catch was that neither TG nor UM had any significant business activities outside of the forex trades and held no capital apart from that deposited by Investors. As such, when Investors who joined the Singliforex scheme earlier sought to cash out their “trading gains”, the withdrawals were funded using the capital deposited by the Investors who joined the scheme later. The DJ thus found that Singliforex was nothing more than a Ponzi scheme.

The Singliforex scheme had been marketed by way of a multi-level marketing framework under which the Investors were spurred by a system of incentives to procure others to join the scheme. As at the point of its collapse, the scheme had attracted a total of 4,264 Investors, from whom a total of about US$21m in investments had been collected. A huge chunk of this, comprising about US$13.7 million, could not be recovered. There was some evidence as to the avenues to which portions of this sum had been dissipated. For example:

(a) US$2.7m had been withdrawn from TG’s HSBC bank account and paid into Leong’s foreign personal account.

(b) $507,854.61 had been transferred to the bank account of Leong’s wife, which was beneficially owned by Leong himself.

(c) US$4m was transferred to various unrelated companies.

(d) At least US$300,000 was received by Ng, purportedly as commission for his services to the brokerages TG and UM.

(f) Withdrawals were made to pay Investors seeking to withdraw their “trading profits” under the Singliforex scheme.

(g) Payments were made towards investments that had no relation to the Singliforex scheme, eg, a US$529,900 investment into one Head Stone Group Ltd for a purported private placement project.

(h) Payments were made for various corporate expenditures, eg, the corporate secretarial fees for setting up UM.

The proceedings below

The central premise of the Prosecution’s case was that the Singliforex scheme, as marketed by Singliworld, was carried out for a fraudulent purpose within the meaning of s 340(1) of the Companies Act (Cap 50, 2006 Rev Ed) (“CA”). It was also the Prosecution’s case that both TG and UM had contravened s 82(1) of the Securities and Futures Act (Cap 289, 2006 Rev Ed) (“SFA”) by holding themselves out as carrying on the SFA-regulated activity of leveraged forex trading (pursuant to the Singliforex scheme), when they were not licenced to do so. Although TG and UM were foreign firms, the Prosecution’s case was that under s 339(1) of the SFA, their offences could be treated as having been committed wholly in Singapore despite parts of the offending acts occurring outside Singapore.

Based on the above premise, both Ng and Leong each faced the following three charges under the CA and SFA, to which they claimed trial:

(a) One charge under s 340(5) of the CA for being a knowing party to the conduct of the fraudulent Singliforex scheme – I will use the term “fraudulent trading” to refer to the offence underlying this charge.

(b) Two charges under s 331(1) of the SFA in respect of TG and UM – which I will refer to as the “SFA charge(s)” – comprising:

(i) one charge for consenting to TG’s contravention of s 82(1) of the SFA, in their capacity as TG’s director; and

(ii) another charge for consenting to UM’s contravention of s 82(1) of the SFA, in their capacity as UM’s officer (in the case of Ng) / director (in the case of Leong).

After 88 days of trial, the Appellants were each found guilty of all three charges above and convicted accordingly.

Additionally, Leong pleaded guilty to the following two charges, albeit only at the very end of the trial:

(a) One charge under s 331(1) of the SFA for consenting to Singliworld’s contravention of s 82(1) of the SFA, in his capacity as Singliworld’s director. Singliworld had contravened s 82(1) of the SFA by holding itself out as carrying on the SFA-regulated activity of leveraged forex trading (pursuant to the Singliforex scheme) when it was not licenced to do so.

(b) One charge under s 3(1) read with s 6(1) of the Multi-Level Marketing and Pyramid Selling (Prohibition) Act (Cap 190, 2000 Rev Ed) (“MLMA”) for promoting a pyramid selling scheme (being the Singliforex scheme).

The DJ imposed the following sentences on Ng and Leong:

The asterisked sentences in bold font above were ordered to run consecutively, giving rise to the following aggregate sentences:

Parties’ submissions on appeal

In MA 9001, Ng appealed against his convictions for:

(a) the fraudulent trading charge; and

(b) the SFA charge in relation only to UM (and not TG).

He also appealed against his sentences for both charges above, as well as against the sentence imposed for the SFA charge in relation to TG.

In MA 9008, Leong appealed against his sentences for:

(a) the fraudulent trading charge; and

(b) the MLMA charge.

Leong did not appeal against the sentences imposed in respect of his remaining three charges under the SFA.

Ng’s submissions

As regards the fraudulent trading charge, Ng’s case was in essence that he was not a knowing party to the fraud underlying the Singliforex scheme. Rather, he claimed to have been misled into believing that Singliworld did indeed have a team of professional traders executing trades on behalf of the Investors and that all he did was merely to abide by Leong’s instructions.

As for the SFA charge relating to UM, this alleged that Ng had in his capacity as UM's “officer” consented to UM's contravention of s 82(1) of the SFA (see [9(b)(ii)] above). Relying on s 331(5) of the SFA, which defines the term “officer” to include a “manager”, the Prosecution’s position was that Ng was UM’s “officer” because he was a “manager” of UM. Ng’s case was that he was not even an employee of UM, let alone a manager.

In relation to his appeal against sentence, Ng argued as follows:

(a) For the fraudulent trading charge, the sentence of 6½ years’ imprisonment was inconsistent with the precedents. The duration of the imprisonment term imposed was also too close to the statutory maximum of seven years as to be justifiable, given that Ng’s case was not the worst conceivable case of fraudulent trading under s 340(1) CA.

(b) In relation to the two SFA charges, Ng’s sentences should have been lower than that imposed for the corresponding charges against Leong, since Leong “was at all material times the true mastermind behind the overall scheme of offending”.

Leong’s submissions

As for Leong, his position on his sentence for the fraudulent trading charge was largely aligned with that of Ng’s, being that the sentence of 6½ years’ imprisonment was inconsistent with the precedents.

As regards the MLMA charge, Leong argued that the DJ had impermissibly double-counted certain aggravating factors when arriving at the sentence. Specifically, Leong argued that these factors had already been accorded due weight in the determination of the sentence for the fraudulent trading charge but the DJ took account of these factors again when determining the sentence for the MLMA charge. Leong further argued that the in-default imprisonment term of 18 months’ imprisonment for the penalty of $3,658,600 was excessive, given that he is an undischarged bankrupt and would “unambiguously” have no means of making payment.

My decision

I dismiss Ng’s appeal against his convictions.

I also partially allow the Appellants’ appeals against sentence by reducing the duration of the 6½-year imprisonment term imposed on each of them for the fraudulent trading charge to: (a) 4½ years for Leong; and (b) four years for Ng.

In setting out my reasons, I will deal with the Appellant’s appeals in the following order:

Ng’s appeal against his convictions

Before dealing with Ng’s substantive arguments, I first canvass various procedural objections that Ng had raised.

Procedural objections

Firstly, Ng complained that many of the Prosecution’s contentions in its submissions, as filed in this appeal, had not been put to Ng at the trial below. This, argued Ng, breached the rule in Browne v Dunn (1893) 6 R 67. However, Ng failed to identify any specific contention that the Prosecution was advancing on appeal which had not been properly put to Ng at trial, or demonstrate how Ng had been prejudiced by such omission. It is not the job of an appellate court to scrutinise an entire trial’s worth of transcripts just so as to help an appellant plug the gaps in his submissions. I thus reject Ng’s submission on this point, on account of its want of particularity. Related to this, Ng also argued that the Prosecution should not be allowed at this appeal to rely on matters raised in Ng’s investigation statements (recorded by the police), given that the Prosecution failed to properly confront Ng with the relevant portions of these statements during cross-examination, as required by s 147 of the Evidence Act 1893 (2020 Rev Ed) (“EA”). In my view, this submission also suffers from lack of particularity, in that it failed to highlight the specific portions of Ng’s statements that had not been adequately brought to his attention in cross-examination. In any case, I have looked at the critical portions of Ng’s statements that are material to this appeal – such as his admission that he had embarked on the practice of “slippage” (see [34] below) – and it is clear to me that Ng was cross-examined extensively on them.

Secondly, Ng argued that the DJ erred by “follow[ing] the Prosecution’s submissions in both spirit and form”. It is unclear exactly what Ng was seeking to suggest by this submission. To the extent that he was implying that the DJ plagiarised the Prosecution’s submissions, I am unable to discern any portion of the GD which lends itself to the suggestion that judicial copying occurred. If Ng was suggesting that the DJ modelled the flow of the GD to follow the structure of the Prosecution’s submissions, my response is that there is nothing wrong with a judge tackling the issues in the exact same order with which they were canvassed by parties, such that the sequencing of the points in the judge’s written decision bears close resemblance to that of parties’ submissions. Rather, the relevant question must be whether the judge applied his mind to the facts and circumstances of the case, such that he can be regarded as having exercised the discretion and judgment required of his judicial office: Lim Chee Huat v Public Prosecutor [2019] 5 SLR 433 (“Lim Chee Huat”) at [48]; see also Newton, David Christopher v Public Prosecutor [2024] 3 SLR 1370 at [40]. That the DJ may have accepted most of the Prosecution’s arguments does not, without more, suggest that he failed to exercise judgment on the issues before him. As explained in Lim Chee Huat (at [48]), a striking resemblance between the written judgment’s reasons and the submissions is not reason in itself to set aside a lower court’s decision, so long as the judge can be discerned as having engaged in weighing of the evidence and arguments. A reading of the GD suggests that the DJ did embark on such a weighing exercise. He took on board the Prosecution’s submissions and addressed the Appellants’ arguments, as well as explained why he preferred the Prosecution’s position.

The third procedural objection raised by Ng was that the DJ erred in allowing the admission of certain statements made by one Bu Yan Shuang (“Bu”). Bu was Ng’s fellow director in TG until she resigned from her directorship in September 2014. Bu claimed that she left after realising that TG was involved in a fraudulent scheme. The DJ had allowed the admission of Bu’s investigation statement to the police, as well as a police report which she made, in which Bu recounted Ng’s extensive involvement in TG’s operations. After giving these statements, Bu became uncontactable and could thus not be cross-examined at trial. While Ng did not dispute that Bu’s statements satisfied the criteria for the admission of hearsay evidence under ss 32(1)(j)(iii) and (iv) of the EA, he contended that her statements ought to have been excluded in the interests of justice, pursuant to s 32(3) of the EA. In particular, Ng argued that the prejudicial impact of Bu’s statements outweighed their probative value.

I reject Ng’s submission that Bu’s statements should have been excluded. Ng painted Bu to be an accomplice with an incentive to make self-serving statements inculpating Ng and thereby exonerate herself. To illustrate this, Ng pointed to the file notes of a tele-conversation which took place between Bu and the Investigating Officer on 27 May 2016, in which Bu was captured as saying that she was “willing to provide evidence in Singapore personally” and “willing to do so as she wants to clear her name”. Ng argued that the unreliability of Bu’s statements could be gleaned from how she became uncontactable, thereby reneging on her expressed willingness to give evidence in Singapore to clear her name. In my view, this was an erroneous portrayal of what happened. As alluded to at the beginning of this paragraph, Bu’s intimation that she wanted to come to Singapore and give evidence was made during her conversation with the Investigating Officer on 27 May 2016. Three days later, she did make good on that expression of intent by coming down from Malaysia to Singapore, where she filed the police report and gave the investigation statement to the police – both were recorded on 30 May 2016. This was thus not a case where Bu gave her evidence entirely from remote shores, inculpating Ng while she remained beyond the reach of the Singapore police. Instead, she personally travelled to Singapore and met with the police to give her statements and, in the process of doing so, exposed herself to being nabbed. It was only after this that she became uncontactable. I have also looked at Bu’s investigation statement and note that there were portions where, instead of distancing herself from TG’s activities, she made various concessions which might be regarded as self-inculpatory, eg, she admitted to having access to and control of TG’s HSBC bank account and to making various payments using clients’ monies.

While I do not consider the DJ as having erred in admitting Bu’s statements into evidence, I nevertheless acknowledge Ng’s concerns that the weight accorded to Bu’s statements should be appropriately circumscribed. Given Bu’s involvement in running TG, it is not possible to completely rule out the risk of her having an incentive to exaggerate the extent of Ng’s involvement, with a view to minimising the appearance of her own. Ordinarily, this concern could be met by letting Ng cross-examine Bu on the stand but that safeguard was not available here as Bu was no longer traceable by the time of the trial – see also the observations of the Court of Appeal in Public Prosecutor v Chia Kee Chen [2018] 2 SLR 249 at [57]. Still, the DJ did not appear to have placed undue weight on Bu’s statements. In fact, it is clear to me that Bu’s statements had been used to do no more than corroborate the substantive evidence against Ng (see also Public Prosecutor v Mohamad Yazid bin Md Yusof [2016] SGHC 102 at [14]−[15]) which, even without Bu’s statements, sufficed to sustain Ng’s conviction. It is to that substantive evidence which I now turn.

Ng’s conviction on the fraudulent trading charge

In relation to the fraudulent trading charge, it is not in dispute that the forex trades transacted on behalf of the Investors pursuant to the Singliforex scheme were executed by Ng on the brokerage firms’ MT4 system (see [5] above). By his own account to the police, Ng had executed the majority – at least 60% – of these trades. I also note that Ng had later changed this narrative and testified that the percentage was even higher, with him having been in charge of 90% of the trades. Ng’s defence was in essence that from around March 2014, being the commencement of the duration stipulated in the fraudulent trading charge, he had sold TG to Leong. Ng claimed that after the sale, Ng remained a director of TG in all but name – Leong (as the new owner of TG) became TG’s directing mind and will while Ng was relegated to being a “lame duck”. From that point on, Ng claimed that he had conducted the trades with TG on Leong’s instructions, with no knowledge about the fraud underlying the Singliforex scheme. In this respect, Ng raised the following contentions:

In my view, there was ample evidence for the DJ to reject Ng’s professions of ignorance as set out above. In Phang Wah v Public Prosecutor [2012] 1 SLR 646 (“Phang Wah”), the High Court remarked (at [25]) that in the fraudulent trading context, dishonesty would encompass “a situation of turning a blind eye to the obvious”. I would venture to say that the facts of this case, when put together and viewed holistically, were suggestive of Ng’s mental culpability having gone beyond that – in that Ng must have possessed actual knowledge about the fraudulent nature of the Singliforex scheme.

As an introductory observation, the evidence did not support Ng’s claim that he was merely a passive actor who executed trades on Leong’s instructions. As the Prosecution rightly pointed out in their submissions below, there were more than 7,200 WhatsApp messages involving Leong and Ng, some of which related to trading. Yet, Ng was unable to point to any pattern of Leong giving detailed trading instructions to him. In fact, the evidence suggested that Ng played an active part in not only assisting but also advising Leong with the trades. In his statement to the police, Ng explained that he had engaged in a practice which he termed as “slippage”, where he would change the price at which the trades between the Investors and the brokerage were transacted. Specifically, Ng mentioned in his police statement that Leong wanted to get better trading results for the forex trades for the Investors, so that the latter would continue putting funds into the Singliforex scheme. To this end, Ng had informed Leong that the only way to facilitate this was through the practice of “slippage”, by effecting the trades at a price that did not reflect the market price but was instead more favourable to the Investors. This clearly showed that Ng was not merely a passive executioner of Leong’s will.

More importantly, the business model on which the scheme rested must have sounded sirens that something was very wrong. Ng knew that there were no liquidity providers against whom the trades were matched. In other words, he knew that the brokerages TG and UM were in effect trading as principals against the Investors. Ng nevertheless claimed that there was nothing objectionable with this, pointing out that there are brokers in the market who do not match their client’s trades against liquidity providers and instead take up trading positions directly against their client. Ng contended that this was done by brokerages in the industry that adopt the “B-booking” model (as opposed to “A-booking” brokerages, which match clients’ trades against liquidity providers). Ng thus contended that it was perfectly legitimate for brokerages to trade against their own clients, without any liquidity providers in the picture.

With respect, the focus on the propriety of the B-booking brokerage model was nothing more than a smoke screen to shift focus away from the obvious crux of this issue. In the absence of a liquidity provider, any trading losses by the client would translate directly into trading gains for the brokerage, and vice versa. Even someone uninitiated to the world of trading can plainly see that a brokerage cannot conceivably be allowed to conduct discretionary trades on its client’s behalf, while at the same time taking an opposing position against the client within those very same trades – the conflict of interest would be massive. That was why the expert at the trial below, Adam James Reynolds, explained in his report that brokerages engaging in the B-booking model must separate the asset management arm which executes trades for the client’s portfolio, to avoid any such conflict. His report explained:

In the present case, there was no such separation. Rather, by Ng’s own case, Leong was giving trading orders to be executed on behalf of the Investors, while simultaneously giving trading instructions on behalf of the brokerage (which Ng professed to be owned and controlled by Leong). If so, it would have been obvious to Ng that Leong was controlling both sides of the transaction, dictating the price at which both client and brokerage would transact with each other. It is inconceivable that Ng, having previously worked as a futures trader in an investment bank, could have failed to apprehend the magnitude of the impropriety of what was going on. This is borne out by Ng’s own testimony that he had reminded Leong on several occasions about the need for liquidity providers to helm the other side of the trades, only to be ignored by Leong.

Despite these circumstances, Ng continued to execute the trades which he claimed to have been dictated by Leong. In fact, rather than just mechanically executing Leong’s instructions, Ng went so far as to advocate the practice of “slippage” (described at [30] above) to Leong and thereafter executed that practice every month. Ng had thus pro-actively massaged the traded prices to create the illusion of trading gains by the Investors. Under cross-examination below, Ng sought to downplay his actions by suggesting that there was nothing wrong with the practice of “slippage”, which he said was merely a means for the brokerage to narrow its bid-ask spread and thereby afford the Investors greater trading profits. This, argued Ng, was akin to the brokerage giving the Investors a “discount”. To that end, Ng argued that the fact that the MT4 system allowed for such price adjustments to be made showed that the practice could not be regarded as improper. In my view, this was an entirely bald assumption that lacked any support. There was nothing to show that the MT4 system had been designed with such inbuilt restrictions to prevent manipulative adjustments. More importantly, Ng’s attempt to paint the practice of “slippage” with a veneer of legitimacy ran counter to his statement to the police that “most broker[s] will never have slippage to let clients win” [emphasis added]. On cross-examination at trial, he also admitted that legitimate brokerages would not engage in such a practice. That Ng would advise Leong to embark on the practice of “slippage”– and thereafter execute that practice on Leong’s behalf – clearly detracted from Ng’s professions of innocence.

The inference that Ng was complicit in Leong’s schemes is strengthened by the fact that Ng, who was a signatory of TG’s sole bank account with HSBC bank, knew that TG had insufficient funds in its accounts to pay the Investors. He admitted as much while under cross-examination. Yet, as late as in end September 2014, Ng had prepared marketing materials for TG which conveyed the following message to the Investors:

Despite knowing that there were insufficient funds in TG’s HSBC bank account to make good on the “trading profits” reaped by the Investors, Ng continued to execute the trades under the Singliforex scheme, as well as facilitated the channelling of funds from TG’s HSBC bank account to Leong. Put together, Ng’s conduct was far more redolent of someone who was a knowing party to Leong’s plundering. During the appeal, Ng argued that his admissions under cross-examination – to the effect that he knew about TG’s lack of funds – were made only after the fact, ie, only after Leong’s fraud was brought to light. In my view, any suggestion by Ng that he was unaware of TG’s lack of funds at the time the fraudulent trading was taking place must be rejected. Such a claim would fly in the face of Ng’s own testimony (which Ng acknowledged was corroborated by Bu) that both he and Bu had tried to raise concerns to Leong about the insufficiency of funds in TG’s HSBC bank account. According to Ng, Leong had summarily dismissed these concerns by tersely saying that he (ie, Leong) “can handle it”, thereby culminating in a situation where there was “nothing that [Ng] could do”. Ng had thus clearly come to know about the insufficiency of funds at the material time, when the Singliforex trades were taking place, and not only after the fraud had been brought to light by investigations.

There are thus ample grounds to reject Ng’s claim that he was merely an unwitting participant who lacked knowledge of the fraud underlying the Singliforex scheme. I see no reason to disturb the DJ’s finding that Ng was a knowing party to that fraud. Consequently, Ng’s appeal against his conviction on the fraudulent trading charge is dismissed.

Ng’s conviction on the SFA charges in respect of UM

As regards the SFA charge in respect of UM, Ng’s position was that he was not an “officer” of UM, as required by s 331(1) of the SFA. As alluded to at [15] above, the term “officer” is defined by s 331(5) of the SFA to include a “manager”. Ng contended that he was not a manager – and was in fact not even an employee – of UM. Ng also argued that even if he could be considered a “manager” of UM, he did not have authority to make decisions, presumably meaning that he had no capacity to “consent” to TG’s offending acts as required by s 331(1) of the SFA.

In my view, Ng’s conviction for this charge was amply supported by the evidence. Critically, Ng’s attempt to downplay his role in UM was undermined by evidence from the Investors.

(a) One of them, Lim Kang Wee (“Lim”), was present at a meeting on 6 January 2016 attended by Leong, Ng and some of the Investors – the meeting was called after the Investors encountered difficulties in withdrawing their funds. Lim testified that during this meeting, Ng was introduced by Leong as an employee of UM, thereby contradicting Ng’s claim that he was not even employed by UM.

(b) Another one of the Investors, Tan Meng Chye (“Tan”), similarly alluded to how the Investors, after having encountered issues with withdrawing their monies in April or May 2015, met with Ng and Leong. It was Tan’s evidence that when he attended one of these meetings, Ng had been introduced by Leong as the “Lao Ban” (meaning “boss” in Mandarin) who “managed the whole thing and the operations”, while Leong professed to only oversee marketing. According to Tan, Leong informed the Investors at this meeting that Ng was the best person to explain to the Investors what had happened.

The manner in which Ng had been introduced to the Investors by Leong was clearly consonant with Ng being a person endowed with some measure of executive authority within UM. There is no evidence on record of Ng having said anything to the Investors to contradict (or at least qualify) the manner in which he had been introduced by Leong.

It is also relevant that Ng played a central role in UM’s efforts to appease Investors after they encountered issues in withdrawing their funds. According to Lim, the objective of the meeting which he had attended with Ng, Leong and the other Investors (alluded to in (a) of the preceding paragraph above) was to find a collective solution for the Investors to retrieve their monies. Notably, Lim testified that one of those solutions was to come from Ng. One of the measures by which UM had purported to allow the Investors to recover their capital and realise their “trading profits” was to give them debit cards, from which the Investors could make withdrawals – it was undisputed that these debit cards were distributed to the Investors by none other than Ng himself.

The DJ had cited various other factors in support of the conclusion that Ng was a manager of UM. In my view, while these factors may not have been particularly determinative when viewed individually, their cumulative presence certainly corroborated the conclusion that Ng played a managerial role within UM. For example, the DJ noted Ng’s key role in setting up UM and migrating TG’s business to it, as well as Ng’s access to UM’s bank account and MT4 account. Ng had also selected Trustworthy Services as the institution with which the Investors’ funds were to be deposited and was empowered to give Trustworthy Services instructions on transfers of those funds.

Having viewed the circumstances holistically, I decline to disturb Ng’s conviction on the SFA charge in respect of UM. There was sufficient evidence to support the finding that Ng was employed by UM, and in an executive capacity at that. The DJ was thus entitled to find that Ng was a “manager” of UM for the purpose of the charge.

The appeals on sentence

I now turn to the appeals against sentence. As alluded to at [20] above, I have decided to allow the appeal against the sentence for the fraudulent trading charge, by reducing the 6½-year imprisonment term imposed on each of the Appellants to: (a) 4½ years for Leong; and (b) four years for Ng.

As for the other sentences imposed by the DJ, the Appellants’ appeals against sentence are dismissed.

The fraudulent trading charge

I begin with the sentence imposed on Leong in respect of the fraudulent trading charge. I agree with the Prosecution that the facts of this case lie at the more serious end of the spectrum, given the sophistication and premeditation underlying the fraudulent scheme and the significant duration (exceeding a year) over which it lasted. I also take the view that Leong’s sentence should be pegged at no less than the 4½-year imprisonment term imposed in Phang Wah. In particular, Leong caused the Investors to lose about US$13.7m of their capital (see [7] above), as compared to Phang Wah where the High Court observed (at [73]) that “no participant in the scheme appears to have been adversely affected in any way”.

The Appellants nevertheless argued that contrary to the observation by the High Court in Phang Wah (that the scheme did not appear to have adversely affected its participants), a reading of the District Court’s judgment in Public Prosecutor v Phang Wah [2010] SGDC 505 (“Phang Wah DC”, ie, the trial court decision from which the appeal in Phang Wah arose) shows that there were significant investor losses. The Appellants relied on the District Court’s decision at [334] of Phang Wah DC, extracted below:

Based on the above passage, the Appellants calculated that the losses in Phang Wah may have reached some $47m. The Appellants thus maintained that the sentence of 4½ years’ imprisonment imposed in Phang Wah should be construed on that premise, rather than on the premise that there were no investor losses.

The Appellants’ depiction of the facts in Phang Wah appears to be erroneous. As stated at [335] of Phang Wah DC (this being the very next paragraph following that extracted immediately above), the District Court expressly observed:

The District Court’s observation was thus consistent with the High Court’s observation (at [73] of Phang Wah) that there was no evidence of any participant in the scheme having been adversely affected. For completeness, I would express my reservations about the methodology adopted by the Appellants in interpreting Phang Wah, as set out in the preceding paragraph above. The Appellants sought to demonstrate that the factual basis expressed by the appellate court in arriving at its holding (as to the appropriateness of the sentence imposed by the lower court) was inaccurate and, to that end, reverse engineered portions of the lower court’s judgment to advance an alternative factual matrix. Following from this, the Appellants proceeded to argue that the appellate court’s holding (as to the appropriate sentence) should be based on that alternative factual matrix. Clearly, this cannot be a proper means of construing an appellate court’s holding, particularly where the appellate court had explicitly grounded its ratio on facts that were plainly at odds with the alternative factual matrix which the Appellants sought to proffer. In any case, there is no need for me to say any more on this, given my view (expressed at the beginning of this paragraph) that the factual statement by the appellate court at [73] of Phang Wah did (contrary to the Appellants’ contention) accurately capture the state of facts, viz, that no participant in the scheme appeared to have been adversely affected.

Still, I take the view that the sentence for Leong should not exceed what was imposed in Phang Wah. The scale and reach of the scheme’s operations in Phang Wah, in terms of revenue generated and number of participants involved, were greater than that in the present case. Specifically, the scheme in Phang Wah generated revenue of $175m from selling a total of 25,773 packages (see Phang Wah at [12]) while the Singliforex scheme in the present case generated only US$21m from sales to 4,264 Investors (see [7] above). As regards personal gains reaped by the wrongdoer, the offender in Phang Wah secured personal benefits amounting to around US$5.1m (see Phang Wah at [12]). This was slightly greater than the gains reaped by Leong, which included the US$2.7m that had been siphoned off (see [7(a)] above) as well as the half a million dollars which became untraceable after being transferred to his wife’s account (see [7(b)] above). Juxtaposing the key facts in Phang Wah against those in the present appeal, the 6½-year imprisonment term imposed on Leong and Ng does come across as manifestly excessive. In my view, the imprisonment term for the fraudulent trading charge in the present case should not exceed the 4½ years imposed in Phang Wah.

A comparison with the fraudulent trading precedents cited by the parties also shows that a sentence of 4½ years’ imprisonment is appropriate:

Consequently, I reduce Leong’s imprisonment term for the fraudulent trading offence to 4½ years. Ng’s sentence in respect of the same offence should similarly be reduced. However, I take the view that Ng should get a slightly lower imprisonment term for the offence than that imposed on Leong. While I agree with the Prosecution that Ng played a pivotal role in operationalising the trades which lay at the heart of the Singliforex scheme, his personal gain from the scheme – standing at US$300,000 (see [7(d)] above) – was lower than Leong’s. I thus reduce Ng’s imprisonment term for the fraudulent trading charge to four years, ie, six months less than that imposed on Leong.

The SFA charges in respect of TG and UM

As regards the SFA charges in respect of TG and UM (set out at [9(b)] above), the sentences imposed below do not strike me as manifestly excessive and I thus uphold these sentences. In doing so, I make the following comments:

Leong’s sentence for the MLMA charge

This then leaves me to consider Leong’s appeal against the sentence imposed by the DJ for the MLMA charge, being 3½ years’ imprisonment and a penalty of $3,658,600 (in default 18 months’ imprisonment). At this appeal, Leong challenged two aspects of this sentence:

(a) the duration of the imprisonment term of 3½-years; and

(b) the duration of the in-default imprisonment term of 18 months in relation to the penalty order.

I cover each of these aspects sequentially below.

(1) Duration of the imprisonment term of 3½ years

As alluded to at [11] above, the DJ had ordered that the sentence for the MLMA charge run consecutively with that for the fraudulent trading charge. I note that Leong did not object to this – after all, the offence underlying each charge pertained to a different legally protected interest. Rather, his grievance was that the DJ (having ordered both sentences to run consecutively) double counted the aggravating factors common to both offences. Specifically, Leong alleged that many of the aggravating factors which the DJ had considered in sentencing Leong for the fraudulent trading charge were taken into account by the DJ again, when arriving at the sentence for the MLMA charge. These factors included the number of investors, the revenue generated, the sophistication of the scheme and the personal benefits reaped. This, argued Leong, constituted impermissible double-counting of the aggravating factors that led to a manifestly excessive imprisonment term being imposed for the MLMA charge. Leong thus asked that the imprisonment term for the MLMA charge be adjusted downwards, from 3½ years to just two years.

The rule against double counting was explained in Public Prosecutor v Raveen Balakrishnan [2018] 5 SLR 799 (“Raveen Balakrishnan”), where Chief Justice Sundaresh Menon listed (at [83]−[89]) various instances where aggravating factors may be regarded as having been impermissibly double counted in sentencing. In the present appeal before me, Leong’s submission on double counting centred on the instance where an offender is charged with two distinct offences that purportedly encompass common factual elements which could be regarded as aggravating. From Leong’s standpoint, a sentencing judge should take account of those aggravating factors for only one of the offences because taking them into account for both results in those factors being double counted to the accused person’s prejudice. This purported variety of double counting, as framed by Leong, does not fall within any of the instances listed in Raveen Balakrishnan. The question is thus whether it can be recognised as yet another instance of impermissible double counting, especially since the learned Chief Justice in Raveen Balakrishnan was careful to caution (at [91]) that the instances of impermissible double counting listed by him were not exhaustive and could not be “rigidly analogised to any set of facts”.

In my view, Leong’s submission that there was double counting in this case should be greeted with some circumspection, especially when one considers how the mischief of the MLMA offence differs from that targeted by the offence of fraudulent trading under the CA. The gravamen of the MLMA charge centred on how Leong hatched a scheme backed by a system of incentives to ensnare an ever-widening network of victims, drawing them into an unsustainable business model that carried the risk of financial ruin: see also Chua Hock Soon James v Public Prosecutor [2017] 5 SLR 997 at [42]−[50]. In contrast, the essence of the fraudulent trading charge centred on how Leong ran what was ostensibly a forex trading business when this was nothing more than a subterfuge for a Ponzi scheme. While some of the aggravating factors might appear to have been common to both offences, they operated to exacerbate each offence differently, in a manner that spoke specifically to the mischief targeted by the offence. An example of such a factor would be the sophistication of the scheme, which Leong claimed to have been impermissibly double counted. In the context of the MLMA charge, the DJ alluded to the sophistication of the incentive structure, including the complex system for calculating incentives. In contrast, when determining the sentence for the fraudulent trading charge, the DJ had regard to the sophistication of the fraudulent scheme, including how it employed overseas-incorporated shell companies such as TG and UM to serve as fake brokerages. This was a clear example of how a purportedly common aggravating factor had been taken on board for each offence in an appropriately demarcated fashion. It struck me as somewhat simplistic to simply sweep the DJ’s evaluation of such factors under the general rubric of “double counting”.

Even if one were to assume (for the sake of argument) that the aggravating factors for the MLMA and fraudulent trading charges should be treated as identical for all intents and purposes, I still have reservations about Leong’s proposal that they be taken on board when determining the sentence of only one of the two offences and discounted in the sentencing of the other. This did not strike me as a principled approach, particularly under the facts of the present case where many of the factors which Leong alleged to have been double counted were traits integral to the very essence of the impugned scheme (in relation to both its fraudulent nature and pyramid selling features). I cannot see how the court can simply excise and ignore these traits in respect of one of the offences – that would be an exercise in artificiality which alters the very character of the offending act encapsulated in the charge. Some guidance in this respect can be gleaned from the case of Chang Kar Meng v Public Prosecutor [2017] 2 SLR 68 (“Chang Kar Meng”), where the offender faced separate charges for both robbery and rape. The defence suggested that the infliction of hurt, being an implicit element of the robbery charge, should not be taken into account when sentencing the offender for the rape charge. The Court of Appeal rejected this submission, opining that the infliction of hurt could be taken into account for both charges (at [27]−[28]):

The Court of Appeal’s remarks above on the artificiality of discounting key factual traits of an offence, just because they happen to be shared by another offence that the offender is facing, are highly pertinent. Critically, the Court of Appeal observed that concerns about the sentence being unduly inflated by common aggravating factors being taken into account for multiple charges can be met by the court taking a “last look” to ensure that the global sentence does not infringe the totality principle. In the present case, it does not appear, nor has Leong advanced any substantiation for the court to conclude, that this principle has been infringed by the consecutive running of the sentences for the MLMA charge and the fraudulent trading charge.

(2) Duration of the in-default imprisonment term of 18 months

The other objection that Leong had to the sentence for the MLMA charge was the DJ’s imposition of an in-default imprisonment term of 18 months, in respect of the penalty order of $3,658,600.

The dollar value of the penalty had been derived from the Singapore dollar equivalent of the US$2.7m that was siphoned to Leong (referred to at [7(a)] above) less $67,400, the latter sum being the expenses which Leong managed to prove that he had incurred. While Leong did not challenge the quantum of the penalty, he contended that the 18-month imprisonment term imposed for default in paying the penalty was excessive. Leong highlighted that in-default imprisonment terms are meant to prevent evasion of payment and not to punish offenders who are genuinely unable to pay. In this case, Leong was bankrupt as at the time of sentencing. This bankruptcy order, argued Leong, was conclusive proof of his inability to pay the penalty. Consequently, Leong asked that the in-default imprisonment term of 18 months be lowered to just one month, so that Leong would not be excessively punished for failing to pay a sum which he clearly had no ability to pay.

I would first observe that the duration of the in-default imprisonment term of 18 months does not appear excessive in and of itself, especially since it did not appear to run afoul of the guidance by the Court of Appeal in Chang Peng Hong Clarence v Public Prosecutor [2024] 2 SLR 722 (“Clarence Chang”) at [65(a)]. While Clarence Chang involved penalties imposed under the Prevention of Corruption Act (Cap 241, 1993 Rev Ed) (“PCA”), there appears to be no reason why the guidance offered by the Court of Appeal in that case − concerning computation of the in-default imprisonment term for non-payment of penalties under the PCA − cannot be extended to penalties under the MLMA. The objective of the penalty under both regimes is the same, viz, disgorgement of pecuniary gain. In Clarence Chang, the Court of Appeal held that as a starting point, the court should calculate the period of the in-default imprisonment term in each charge by using the daily value of $1,000 for each day of in-default imprisonment. Applying that yardstick to the penalty sum of $3,658,600 in this case generates a starting duration which significantly exceeds the 18 months imposed by the DJ – rendering it difficult to construe the in-default imprisonment term imposed on Leong as excessive.

However, Leong’s complaint about the 18 months being excessive hinged not so much on it being at odds with the precedents governing computation of duration, but on Leong’s inability to pay. In other words, Leong was saying that the in-default imprisonment term was excessive because it punished him excessively for failing to pay what he lacked the capacity to pay. I have no hesitation in rejecting this ground of appeal. It is trite that when an offender relies on a fact to mitigate his sentence, he bears the burden of proving that fact (see Anita Damu v Public Prosecutor [2020] 3 SLR 825 at [28] and Kanagaratnam Nicholas Jens v Public Prosecutor [2019] 5 SLR 887 at [36]). Since Leong sought to rely on his impecuniosity as a reason to reduce a specific aspect of his sentence (ie, the in-default imprisonment term), he bore the burden of proving that he was indeed unable to pay. His failure to discharge that burden was nothing short of abject. It is undisputed that Leong did receive the US$2.7m. At the trial, he claimed that he spent the bulk of that sum on a forex investment that was unrelated to Singliworld and had spent about US$200,000 to US$300,000 to pay off the Investors. However, there was no indication of what happened to the forex investment (eg, whether it had grown or withered), nor was there any paper trail showing that the US$200,000 to US$300,000 had indeed been channelled back to the Investors. In fact, when Leong was pressed for supporting documentation to support his claims as to how he had applied the US$2.7m, he responded that he did not have them.

The Prosecution contended that even if Leong could prove that he had completely squandered the US$2.7m, leaving himself with nothing to pay the penalty, an extended period of incarceration was still warranted so as to reflect the fact that Leong – having fully consumed his ill-gotten gains – stood more culpable than someone who disgorged the same. There is no need for me to address this contention. My conclusion in the preceding paragraph – to the effect that Leong failed to demonstrate his inability to pay the penalty – suffices to justify dismissing his plea that the in-default imprisonment term be shortened.

During the hearing of the appeal, counsel for Leong stressed that his client is now a bankrupt, meaning that even if Leong had the US$2.7m tucked away somewhere, that sum of money would vest in the Official Assignee (“OA”). The upshot of this would be that even if Leong had his hands on the US$2.7m, he would still not possess the legal capacity to apply that sum towards payment of the penalty. However, Leong did not proffer any support for the assertion that the OA would indeed regard the US$2.7m – being proceeds misappropriated from the Investors – as part of the bankruptcy estate vesting in the OA. There was no suggestion of Leong having made any attempt (now that he is a bankrupt) to procure an intimation from the OA that the latter agreed with such an assertion. I should also add that when I asked Leong’s counsel to check with his client as to the amount sought by the statutory demand giving rise to the bankruptcy order, it emerged that the amount was only $36,000. This meant that if Leong had access to the US$2.7m, he would require only a miniscule fraction of that sum to satisfy the debt in the statutory demand and thereby pave the way for discharging the bankruptcy order. Absent evidence of any other provable debts, there was nothing on record standing in support of Leong’s suggestion that the US$2.7m (or at least a vast bulk of that sum) would necessarily remain locked up in the hands of the OA (assuming that this sum, being proceeds of crime, would even vest with the OA in the first place). Under the circumstances, it was clear that Leong’s bankruptcy status – to the extent that it was being raised in support of his professed inability to pay the penalty – was nothing more than a fig leaf.

Conclusion

For the reasons above, I dismiss Ng’s appeal against conviction. I also partially allow the Appellants’ appeals against their sentence by reducing the imprisonment term for the fraudulent trading charge from 6½ years to: (a) 4½ years for Leong; and (b) four years for Ng.

I also decline to disturb the DJ’s decision as regards which sentences should be made to run consecutively, as I note that the charges that were ordered to run consecutively all involved different legally protected interests: see Raveen Balakrishnan at [39] and [41]. Consequently, Leong’s aggregate sentence is reduced to 8½ years’ imprisonment and a penalty order of $3,658,600 (in default 18 months’ imprisonment), while Ng’s aggregate sentence is reduced to five years’ imprisonment and a fine of $300,000 (in default, six months’ imprisonment).