Public Prosecutor v Loh Cheok San [2023] SGHC 190

This appeal concerns the application of the one-transaction rule and the totality principle. The present case involves two amalgamated charges under s 124(4) of the Criminal Procedure Code (Cap 68, 2012 Rev Ed) (“the CPC”) of abetting the offence of cheating by way of conspiracy under s 420 read with s 109 of the Penal Code 1871 (Cap 224, 2008 Rev Ed) (“the Penal Code”). The Respondent had not only (a) conspired with six others and his employer, Vermont UM Bunkering Pte Ltd (“Vermont”), to cheat its customers into making excess payments for marine fuel oil (“the Second Charge”), but had also (b) further conspired with his manager and another employee to cheat Vermont of parts of these ill-gotten gains (“the First Charge”). The Respondent pleaded guilty to and was convicted on both charges. He was sentenced to 35 months’ imprisonment for the First Charge and 65 months’ imprisonment for the Second Charge, with both sentences ordered to run concurrently. It is against this aggregate sentence which the Appellant appeals. The Appellant is not appealing against the individual sentences which were meted out for the two charges.

Background facts

The Respondent, a 51-year-old Singaporean, was employed as a cargo officer by Vermont, a company in the business of oil trading and ship bunkering. His role was to oversee the operation of bunker barges in supplying fuel to vessels.

From 2014 to 2016, the Respondent, members of Vermont’s senior and middle management, together with the company, engaged in a scheme to cheat Vermont’s customers by way of “buyback” transactions. These transactions operated in the following manner. The Respondent and his co-conspirators would first target vessels which possessed excess or remaining marine fuel oil in their tanks. They would then collude with the chief engineer or the captain of such a vessel to supply less marine fuel oil to the vessel than what was ordered by the vessel’s owner. This was essentially done by “buying back” the excess marine fuel oil held by the vessel before “adding” the full amount ordered by the vessel’s owner. The agreed price for the “buyback”, which was paid to the chief engineer or the captain of the vessel, would usually be lower than the market rate. The fuel oil which was “bought back” could then be sold by Vermont at a higher rate, allowing it to profit from the difference.

The Respondent’s role in this scheme was to inform his bunker manager whenever there was an opportunity for a “buyback” transaction, and to facilitate the transfer of cash between the chief engineer or captain and the bunker manager. Over a span of about two years, the Respondent participated in 52 of such “buyback” transactions. This resulted in Vermont’s customers being cheated of a total of approximately USD$3,645,976. Of this sum, the Respondent earned a “commission” of at least $43,600.

In addition to this scheme, the Respondent also conspired with the bunker manager and another cargo officer to cheat Vermont of part of the moneys gained from the “buyback” transactions. They did so by falsely representing to Vermont in each transaction that either the price or quantity or both the price and quantity of the fuel which was “bought back” were higher than they actually were. These over-declarations allowed the Respondent and his accomplices to pocket the excess sums beyond what was actually paid to the chief engineers and/or captains, and they thereby induced Vermont to pay them more commission than they were “entitled” to. Vermont suffered a “loss” of about USD$980,000, of which the Respondent gained approximately USD$314,961 (or $410,712).

These facts form the basis for the two charges against the Respondent:

(a) The First Charge, under s 124(4) of the CPC, of abetment by conspiracy to cheat Vermont by over-reporting the price and quantity of marine fuel oil actually “bought back” under s 420 read with s 109 of the Penal Code; and

(b) The Second Charge, under s 124(4) of the CPC, of abetment by conspiracy to cheat Vermont’s customers through the scheme involving “buyback” transactions under s 420 read with s 109 of the Penal Code.

Decision below

The Respondent pleaded guilty to both charges and was sentenced to 35 months’ imprisonment for the First Charge and 65 months’ imprisonment for the Second Charge. Both custodial sentences were ordered to run concurrently.

The District Judge began by observing that the incidents underlying the Second Charge “gave rise” to the incidents pertaining to the First Charge, and thus both offences could be said to stem from or relate to the same 52 transactions. There was also proximity in time, proximity of purpose, proximity of location, and continuity of design between the corresponding incidents underlying the two charges.

Turning to the legal interests which were violated in the present case, the District Judge acknowledged that the interests protected by the two charges were different. He also agreed that the two offences involved different victims. Nevertheless, he found that the two offences were “inextricably related”. This provided a strong indication that the general rule in Public Prosecutor v Raveen Balakrishnan [2018] 5 SLR 799 (“Raveen”), ie, that the sentences for unrelated offences should run consecutively, should not be engaged.

In support, the District Judge appeared to take the view that Vermont did not have a legitimate interest deserving of protection under the First Charge. I reproduce the relevant portions of his decision here:

The District Judge also considered the fact that the Respondent has been charged, convicted and sentenced for both offences as already addressing the different legal interests in respect of both offences. He also added that the acknowledgment that both offences concern different legal interests does not, by itself, address the question of whether the two legal interests must invariably be protected by ordering both sentences to run consecutively.

Taking into account the Respondent’s guilty plea and his finding that the aggregate sentence did not run afoul of the totality principle, the District Judge ordered both sentences to run concurrently.

The parties’ cases

The Appellant’s case

The Appellant argues that the global sentence of 65 months’ imprisonment is manifestly inadequate and should be enhanced for three reasons.

First, the District Judge had erred in ordering the two custodial sentences to run concurrently. According to the Appellant, the two offences did not form part of the same transaction and the Respondent must be separately punished for each offence by running the two sentences consecutively. Even if the two offences formed part of the same transaction, the Appellant maintains that consecutive sentences should be imposed to reflect the Respondent’s enhanced culpability and the need for general and specific deterrence.

Second, the District Judge had failed to give sufficient weight to the totality of the Respondent’s offending conduct. The Appellant submits that the Respondent’s conduct had caused substantial harm to Vermont’s customers and the public interest, resulted in significant pecuniary gain to the Respondent, involved a premeditated and well-planned scheme, and was committed as part of a criminal syndicate. These aggravating factors justify an upward calibration in the aggregate sentence. The Appellant also argues that the District Judge had placed “undue weight” on the Respondent’s guilty plea.

Third, the Appellant takes the position that a global sentence of 100 months’ imprisonment, which involves running the two sentences consecutively, is more consistent with sentences meted out in previous cases.

The Respondent’s case

On the other hand, the Respondent argues that the aggregate sentence of 65 months’ imprisonment is not manifestly inadequate.

First, the Respondent submits that the District Judge had correctly applied the one-transaction rule. The two offences were “intimately related” in so far as the subject matter of the First Charge, the excess payments made by Vermont to the Respondent and his co-conspirators, was essentially derived from the subject matter of the Second Charge, the excess payments made by Vermont’s customers to Vermont. The two offences also involved the same victims – Vermont’s customers. Accordingly, the two offences should be treated as part of the same transaction. The Respondent also argues that there is no basis for deviating from the rule that the sentences for offences forming part of the same transaction should be ordered to run concurrently.

Second, the Respondent points to several factors which he deems relevant to the determination of an appropriate global sentence. These include the Respondent’s cooperation in the investigations of the Corrupt Practices Investigation Bureau, the fact that the Respondent played a minor role compared to his co-conspirators, and the Respondent’s relatively insubstantial benefit from the illegal acts. A consideration of these factors leads to the conclusion that the sentence of 65 months’ imprisonment is not manifestly inadequate.

Third, the District Judge had failed to consider the relevant sentencing precedents, which stress that the actual role played by each conspirator and the personal benefit derived by them must be considered. To this end, the Respondent stresses that extra care must be exercised in referring to past cases concerning cheating offences involving single offenders.

Issues to be determined

In my view, addressing the following two issues will assist in deciding whether the total sentence of 65 months’ imprisonment should be disturbed on appeal:

(a) Whether the one-transaction rule applies in the present case; and

(b) How, if at all, the totality principle affects the final sentence.

Preliminary issue: The threshold of appellate intervention in sentencing

The parties agree on the well-established principles governing appellate intervention in a trial judge’s sentencing decision. Specifically, the appellate court will not ordinarily disturb the sentence imposed by the trial judge except where it is satisfied that (Haliffie bin Mamat v Public Prosecutor and other appeals [2016] 5 SLR 636 (“Haliffie”) at [71]):

(a) the trial judge erred with respect to the proper factual basis for sentencing;

(b) the trial judge failed to appreciate the materials placed before him;

(c) the sentence was wrong in principle; or

(d) the sentence was manifestly excessive or manifestly inadequate, as the case may be.

The following analysis, pertaining to the two issues stated at [21], will focus on grounds (c) and (d) – whether the aggregate sentence of 65 months’ imprisonment is wrong in principle or manifestly inadequate. Where the exercise of the sentencing judge’s discretion is contrary to principle, the appellate court must reconsider the sentence afresh on the basis of the correct facts and/or principles. If a higher or lower sentence is more appropriate on this basis, the appeal ought to be allowed: Kavitha d/o Mailvaganam v Public Prosecutor [2017] 4 SLR 1349 (“Kavitha”) at [15]. Even if the sentencing judge has not erred in principle, a sentence may be manifestly excessive or inadequate if it requires substantial alterations rather than minute corrections to remedy the injustice: Haliffie at [72].

Issue 1: Whether the one-transaction rule applies

The one-transaction rule states that where two or more offences are committed in the course of a single transaction, all sentences in respect of those offences should generally be run concurrently rather than consecutively: Mohamed Shouffee bin Adam v Public Prosecutor [2014] 2 SLR 998 (“Shouffee”) at [27] citing Public Prosecutor v Law Aik Meng [2007] 2 SLR(R) 814 (“Law Aik Meng”) at [52]. The question of whether the various offences form part of a single transaction depends on whether they constitute a “single invasion of the same legally protected interest”: Raveen at [39]. In deciding this question, the court may also consider the proximities in time and place, continuity of action and continuity in purpose or design with respect to the offences. Ultimately, the one-transaction rule is a rule of fairness which rests on the notion that an offender should not be doubly punished for what is essentially the same conduct, even though that conduct may disclose several distinct offences at law: Raveen at [69].

That being said, the one-transaction rule is not mandatory and may be departed from in order for the court to arrive at a just sentence: Seng Foo Building Construction Pte Ltd v Public Prosecutor [2017] 3 SLR 201 at [67]. For instance, the court may order multiple sentences to run consecutively even where the offences form part of a single transaction. This may be done to give effect to a particular sentencing interest such as deterrence or to adequately capture the enhanced culpability of the offender: Shouffee at [81(b)]. Should the court decide to deviate from the rule, it should state the reasons or considerations for doing so. This pre-empts any contention that its decision was made in ignorance of the rule: Shouffee at [46].

The one-transaction rule should also be understood alongside the complementary principle set out in Raveen that the sentences for unrelated offences should run consecutively. This principle and the one-transaction rule are “two sides of the same coin”. In other words, to say that two offences are “unrelated” means that they are not “part of a single transaction”; conversely, to describe them as “part of a single transaction” means they are not “unrelated”: Raveen at [69]. Offenders will thus be prevented from receiving unwarranted discounts for what are essentially separate criminal courses of conduct.

Turning to the present case, I take the view that the District Judge had erred in his application of the relevant legal principles, regardless of whether the analysis is undertaken in the context of the one-transaction rule or the complementary principle on unrelated offences in Raveen. I acknowledge that the one-transaction rule and the rule in Raveen should be viewed as guidelines instead of determinative or comprehensive rules to be rigidly applied: see Shouffee at [37]. This provides the court with the necessary flexibility to achieve the ends of justice within the context of each case. I am therefore concerned not with the fact that the District Judge had exercised his discretion in the course of his decision, but whether the reasons he relied on were defensible or correct in principle.

The crux of the issue lies in whether there had been a single invasion of a legally protected interest. I agree with the Appellant that different legal interests were violated by the two offences. This was not lost on the District Judge, who recognised that the legal interests at play were “conceptually different”. This finding would have presumably led the District Judge to the conclusion that the two offences did not form part of a single transaction or, on the flipside, that the two offences were unrelated. As such, consecutive sentences would have been warranted.

However, the District Judge held that the two offences were nonetheless linked and that their respective sentences should run concurrently. This was because the two offences stemmed and related to the same set of 52 transactions. He also placed much emphasis on the fact that Vermont was not a “real” victim and that it could not “morally claim to be entitled to have a legitimate interest to be protected”. Because of these reasons, the District Judge “respectfully decline[d] to exercise [his] discretion” to protect Vermont’s legal interest in the First Charge and did not order the sentences to run consecutively.

The analysis of the District Judge, as summarised in the preceding paragraphs above, does not, with respect, stand up to scrutiny.

First, it is unclear whether he considered the two offences to form part of the same transaction, in the context of the one transaction rule. On one hand, his acknowledgment that separate legal interests were violated by the two offences indicates that both offences did not form one transaction (or, in other words, that they were unrelated). On the other hand, the District Judge ordered the sentences to run concurrently by declining to exercise his discretion. This implies that the District Judge viewed the offences as one transaction (or as related offences) such that the default position, if he declined to exercise his discretion, would be for the sentences to run concurrently under the one transaction rule. Moreover, the District Judge’s pronouncement that the two offences were “inextricably related” suggests that they formed part of the same transaction: see Raveen at [44], [53], [54] and [69]. Despite this apparent contradiction, I accept that a holistic reading of the District Judge’s decision suggests that the District Judge had found the offences to be unrelated but had nonetheless exercised his discretion to depart from the general rule in Raveen.

This brings me to my second point. I do not agree that the reasons proffered by the District Judge are sufficient to either justify a departure from the general rule on unrelated offences in Raveen or to support a finding that the two offences formed part of the same transaction. The foundation of the one-transaction rule was articulated in Shouffee at [31]:

It is therefore clear that the court may find that two offences form part of the same transaction (and are, in other words, related) even when they violate different legal interests. Instances in which the court may make such a finding include where a rigid focus on the diversity of interests violated may lead to the counterintuitive outcome that the accused would have been better off committing a more serious offence: see Shouffee at [38]. However, I am not satisfied that the ostensible factual links between the two offences, as identified by the District Judge above at [29], are sufficient to support a finding of one transaction or, on the flipside, that the general rule on unrelated offences should be departed from.

I accept that it was only possible for Vermont to have been cheated of those sums because Vermont had cheated its customers in the first place. It is also true that the two offences were proximate in time and location. However, these factors do not, by themselves, justify a finding that the offences comprised a single transaction. Employing the example raised by the High Court in Shouffee at [33] of a date rapist who rapes and steals from his victim, the two offences would be indisputably proximate in time and location. The circumstances surrounding the first offence of rape may also have afforded the date rapist the opportunity to carry out the second offence of theft. Following the analysis in Shouffee, however, the court may nonetheless find that the two separate interests which have been implicated warrant separate punishment in such a scenario. Indeed, in most situations where the one-transaction rule is potentially applicable, the offences will inevitably possess some form of factual link: see, eg, Ewe Pang Kooi v Public Prosecutor [2022] SGHC 300 at [51]–[54]. This may include the proximity of time and/or location, and it will thus be unsurprising if the commission of one offence affords the offender the opportunity to carry out another.

On a broader note, I disagree with the District Judge that the offences should be viewed as “inextricably related” on the facts. Here, the two schemes underlying the two offences clearly involved different courses of conduct. The conspiracy underlying the Second Charge was a sophisticated one involving multiple stakeholders, including a director of Vermont. In order to avoid detection, the co-conspirators would even pump the volume of marine fuel which was “bought back” into Vermont’s barge tank before supplying the full volume ordered by the vessels’ owners. This caused the flow meter (introduced by the Maritime and Port Authority of Singapore) to generate inaccurate data regarding how much marine fuel was supplied. Conversely, the conspiracy underlying the First Charge was a relatively simple one carried out by employees of Vermont who saw their chance to make a quick buck by reporting inaccurate figures. The schemes were therefore carried out by separate (albeit overlapping) groups of individuals and differed in complexity. In my view, this also demonstrates that continuity of design did not exist between the corresponding incidents underlying the two charges.

The conspiracies also targeted different victims – the Second Charge targeted Vermont’s customers while the First Charge targeted Vermont. While I accept that Vermont’s customers were the only entities who had lost moneys which rightfully belonged to them, it does not follow from these contentions that Vermont was not a “real” victim. Vermont was dishonestly induced to make additional payments to the Respondent and his co-conspirators for the First Charge through the false representations which they made to Vermont. Vermont was therefore a victim for the purposes of the First Charge. As the two schemes were aimed at separate victims, I cannot agree with the District Judge’s finding that both offences shared the proximity of purpose. Therefore, given that the two offences stemmed from two distinct courses of conduct with the respective objectives of targeting different victims through different means, the one-transaction rule is not engaged to prevent the Respondent from being doubly punished for the two criminal courses of action.

The District Judge also seemed particularly swayed by the notion that Vermont had no moral claim for protection under the law because it had itself conspired with the Respondent to commit another crime. This calls to mind the age-old principle of ex dolo malo non oritur actio – no court will lend its aid to a man who founds his cause of action upon an immoral or illegal act (Moore Stephens (a firm) v Stone Rolls Ltd (in liquidation) [2009] AC 1391 at [26] citing Holman v Johnson (1775) 1 Cowp 341 at 343). While this principle bears relevance to disputes arising in the civil context, it has little application in the realm of criminal law: Rex v Tan Ah Seng [1935] MLJ 273 (“Tan”).

In Tan, the offender was charged with criminal breach of trust. He misappropriated $40 which he had received for the purpose of renting a house to be used as a brothel. Although the money was entrusted to the offender for a criminal purpose, the court held that the word “property” in the relevant provision was wide enough to encompass such money. The court focused on the wrongful gain which had accrued to the offender and found “no reason… to allow a man to escape punishment for one crime because he has conspired with the complainant to commit another”. These pronouncements were endorsed by the Court of Appeal in Raj Kumar s/o Brisa Besnath v Public Prosecutor [2021] SGCA 88 (“Raj”) at [8]. In particular, the Court of Appeal in Raj held at [9] that a bare possessory right to the property in question was sufficient to satisfy the relevant legal element for the crime of criminal breach of trust.

I am cognisant that the principles espoused in Tan and Raj pertain to the question of criminal liability. In other words, the two cases support the proposition that a man should not escape criminal liability for a crime simply because he has conspired with the complainant to commit another. Nevertheless, these principles provide a useful guide for the present case, albeit in the specific sentencing context of deciding whether the only two sentences which an accused is facing should be run consecutively or concurrently. This is because the practical effect of the District Judge’s decision is to exempt the Respondent from serving his sentence for the First Charge simply because the victim for that charge was a co-conspirator in the Second Charge. In my view, this results in the same undesirable outcome which the principles in Tan and Raj seek to address. The application of these principles to the present case indicates that the focus should remain on the criminality of the Respondent's conduct, regardless of whether Vermont was a “real” victim for the purposes of the First Charge.

This is in line with the fact that, on a more fundamental basis, the criminal law is primarily concerned with the punishment of the offender for his own criminal conduct (Guay Seng Tiong Nickson v Public Prosecutor [2016] 3 SLR 1079 at [69]), rather than the protection of the victim’s rights. Vermont, as a co-conspirator to the Second Charge, may not have suffered any wrongful loss, as conceived in a moral sense. This does not, however, detract from the fact that the Respondent had committed two distinct offences by engaging in two separate conspiracies targeted at different victims. As such, even leaving aside my earlier finding at [36] that Vermont is a victim for the purposes of the First Charge, the two sentences should nevertheless be run consecutively. This is because to hold otherwise would be to effectively allow the Respondent to escape punishment for his criminal conduct underlying the First Charge.

Accordingly, I do not agree with the reasons provided by the District Judge which go towards his finding that the two offences formed part of the same transaction or alternatively that a departure from the general rule on unrelated offences in Raveen was warranted. As such, the sentences for these offences ought to run consecutively. The imposition of concurrent sentences for the two offences would result in the perverse and unjust outcome of the offender not having to bear any real consequences for his additional offending: see Public Prosecutor v Yap Pow Foo [2023] SGHC 79 at [124] citing Muhammad Sutarno bin Nasir v Public Prosecutor [2018] 2 SLR 647 at [22].

There is also no further basis for the court to exercise its discretion to depart from the rule on unrelated offences here. For the avoidance of doubt, the aggravating and mitigating factors which the Appellant and the Respondent have raised in their submissions were similarly placed before the District Judge. The District Judge would therefore have considered these factors in arriving at the individual sentence for each charge, which the parties have not appealed against. As these factors have already been given effect to at an earlier stage of the sentencing analysis, I am careful not to accord them undue weight in relation to how the sentences should run: see Raveen at [91]–[92]. As such, these factors are not directly germane to the issues in this appeal. I note, however, that the District Judge found the Respondent’s plea of guilt to be a pertinent factor which “tilt[ed] the balance further in favour of a concurrent sentence”. I disagree. The plea was entered more than four years after the Respondent was first charged in court. It is thus not representative of the Respondent’s remorse with respect to the two offences and will not, in any case, warrant the running of the two sentences concurrently.

In the alternative, even if I am wrong in concluding that the two offences did not comprise a single transaction, a departure from the straightforward application of the one-transaction rule is warranted on the facts. As explained earlier at [25], the one-transaction rule is not a mandatory rule and may be departed from in order for the court to arrive at a just outcome. In Shouffee, the High Court held at [41] that:

As noted at [35]–[36] above, for the First Charge, the Respondent had entered into a separate conspiracy (a) involving a different group of individuals, (b) of a different complexity, (c) with a different objective and (d) targeting a different victim. These factors reflect an enhanced level of culpability arising from the conspiracy underlying the First Charge, which will not be adequately encapsulated in the sentence meted out for the Second Charge alone. I am therefore satisfied that the imposition of consecutive sentences is appropriate on the present facts.

As a corollary to my conclusion on this issue, the Respondent faces a global sentence of 100 months’ imprisonment.

Issue 2: Whether the totality principle affects the final sentence

Given my finding that the District Judge’s decision on the preceding issue is contrary to principle, I turn to consider the appropriate aggregate sentence afresh on the basis that the two sentences should run consecutively: see Kavitha at [15]. This raises the further question of whether the totality principle applies to recalibrate the global sentence of 100 months’ imprisonment.

The totality principle serves as a “final check” to ensure that the aggregate sentence is proportionate to the overall criminality presented and is not excessive: Raveen at [65]. This principle comprises two limbs (Raveen at [73]):

(a) to examine whether the aggregate sentence is substantially above the normal level of sentences for the most serious of the individual offences committed; and

(b) to examine whether the effect of the aggregate sentence on the offender is crushing and not in keeping with his past record and future prospects.

If the court finds that the totality principle has been engaged, it may opt for a different combination of sentences to run concurrently or consecutively, or adjust the individual sentences imposed: Shouffee at [81(i)].

On the first limb, I must consider whether the global sentence of 100 months’ imprisonment is substantially above the normal level of sentences for the crime of engaging in a conspiracy to cheat under s 420 read with s 109 of the Penal Code. Both the Appellant and the Respondent raise precedents which, according to them, support their respective submissions on the appropriate global sentence. These precedents are instructive in elucidating the normal level of sentences imposed for this crime.

In response to the precedents submitted by the Appellant, the Respondent argues that the court should take “extra care” in referring to cheating offences which involve single offenders and should take guidance primarily from cases involving conspiracies to cheat. This contention holds little merit.

The court must, of course, be careful in referring to appropriate precedents. This ensures that the yardstick for comparison is borne out of cases involving similar circumstances: see Shouffee at [56]. Yet, precedents involving cheating offences under s 420 of the Penal Code can be of utility to the present case even if they were carried out by a single offender. A conspiracy is, at its heart, an agreement between persons to engage in a common criminal object: Goldring, Timothy Nicholas v Public Prosecutor [2015] 4 SLR 742 (“Goldring”) at [47]. Where these persons act together in pursuance of this common criminal object, every act done in furtherance of this object by each of them will be taken, in law, to be done by all: see United Kingdom Law Commission, Conspiracy and Attempts (Law Com No 318, 2009) at para 1.59 (Chairman: The Right Honourable Lord Justice Munby) citing R v Macklin (1838) 2 Lewin 225 at 226; Er Joo Nguang and another v Public Prosecutor [2000] 1 SLR(R) 756 (“Er Joo Nguang”) at [30]. Parties to a conspiracy are therefore jointly engaged in the criminal enterprise as if they had each entered into it by themselves, although their respective punishments may differ depending on their level of involvement and culpability: see Er Joo Nguang at [26]. This is affirmed by the fact that an accused person facing a charge read with s 109 of the Penal Code for abetting an offence by conspiracy may be punished with the very punishment provided for that offence.

In my view, when the court deals with charges involving abetment by conspiracy, suitable precedents should be identified with reference to the facts of the overall criminal enterprise and not simply whether they involved a single offender or multiple offenders working in tandem. Such an approach was adopted by the High Court in Goldring, where it considered precedents relating to single offenders even though the charge in that case involved a conspiracy to cheat. Similarly, in Public Prosecutor v Lam Leng Hung and other appeals [2017] 4 SLR 474 at [396]–[397], the High Court identified starting points for the offence of engaging in a conspiracy to commit criminal breach of trust based in part on the available precedents involving single offenders. It is therefore on this basis that I proceed to consider the precedents submitted by the parties for the purposes of identifying the normal level of sentences for cheating offences under s 420 of the Penal Code.

The first precedent raised by the Appellant is Public Prosecutor v Gene Chong Soon Hui [2018] SGDC 117 (“Gene Chong”). The offender was a finance manager who created fictitious transactions and misrepresented that payments were due to cheat his company of about $3.2m. No restitution was made. He faced 165 cheating charges and pleaded guilty to 15 proceeded cheating charges involving about $1m in total. The 15 charges took place over the course of about five years. The District Court sentenced the offender to 33 months’ imprisonment for each cheating charge, with a global sentence of 132 months’ imprisonment.

The second case relied on by the Appellant is Public Prosecutor v Neo Aileen [2013] SGDC 315 (“Neo”). The offender had cheated the victim of USD$1m by offering her an attractive exchange rate. No restitution was made. The offender pleaded guilty to two cheating charges involving USD$500,000 each. She was sentenced to 39 months’ imprisonment for each charge, with an aggregate sentence of 78 months’ imprisonment.

The overall criminal enterprises in the two cases above share similar features with the facts of the present case, such as the quantum of losses suffered by the victims and the absence of restitution. It should, however, be observed that the offenders in those cases obtained higher pecuniary benefits than the Respondent – a fact to which I shall return to below.

On the other hand, the Respondent relies heavily on the case of Tay Huay Hong v Public Prosecutor [1998] 3 SLR(R) 290 (“Tay”). The offender was charged with criminal conspiracy to cheat a company of a total sum of US$3.75m. He obtained a pecuniary benefit of USD$300,000. Restitution was subsequently made of this sum. He claimed trial and was sentenced to 20 months’ imprisonment for the single cheating charge. I do not place much weight on Tay as an indicator of the normal level of sentences imposed for an offence under s 420 of the Penal Code for two reasons.

First, the cheating offence in Tay was committed under the Penal Code 1871 (Cap 224, 1985 Rev Ed). Under that version of the Penal Code, the punishment for an offence under s 420 was a maximum imprisonment term of seven years with a liability to a fine. The maximum imprisonment term was however increased to ten years in 2008 for cheating offences committed on or after 1 February 2008. It is trite that sentencing precedents relating to an earlier version of the same offence that has a different prescribed punishment are of low precedential value: Kow Keng Siong, Sentencing Principles in Singapore (Academy Publishing, 2nd Ed, 2019) (“Kow”) at para 13.131. Specifically, for offences under the earlier version of s 420, the court would have calibrated the sentences with reference to the maximum sentence of seven years’ imprisonment. It therefore stands to reason that the sentences in these earlier cases would have been correspondingly lower.

Second, in arriving at the overall sentence, the trial judge in Tay expressly considered the fact that full restitution was made by the offender. Restitution is usually a relevant sentencing consideration which carries mitigating value. This is because it provides some evidence of remorse, good character or reformation: Public Prosecutor v Fernando Payagala Waduge Malitha Kumar [2007] 2 SLR(R) 334 at [50]. As this mitigating factor is absent on the present facts, the utility of Tay as a sentencing precedent is accordingly limited.

The Respondent also raised the case of Gan Chai Bee Anne v Public Prosecutor [2019] 4 SLR 838 (“Anne Gan”). I do not, however, find this case to be helpful for the purposes of this issue as it concerned charges under s 6(c) of the Prevention of Corruption Act (Cap 241, 1993 Rev Ed) instead of the cheating offence under s 420 of the Penal Code.

Having considered the precedents raised by the parties, I accept that the normal range of sentences imposed for a single cheating charge involving similar circumstances to the present case is between 33 and 39 months’ imprisonment. This is supported by other precedents involving conspiracies to cheat. For example, in Public Prosecutor v Koh Seah Wee and another [2012] 1 SLR 292 (“Koh”), the two offenders, who were working in the Singapore Land Authority (“the SLA”), conspired to cheat the SLA of about $12m over three years. About $9m was recovered in money and assets, leaving a total loss of between $3m and $4.5m – a sum comparable to the present case. The two offenders pleaded guilty to, inter alia, cheating offences under s 420 as well as s 420 read with s 109 of the Penal Code. The High Court imposed a sentence of four years’ (or 48 months’) imprisonment for each cheating offence with respect to both offenders. This included the second accused who was the junior employee but played a vital role in the conspiracy. In my respectful view, the sentence imposed in Koh affirms my view that the normal range of sentences applicable in this case is 33 to 39 months’ imprisonment, given that the case at hand does not possess the additional aggravating factor of defrauding public bodies of public funds.

Comparing the aggregate sentence of 100 months’ imprisonment with the range of sentences normally imposed for a charge of cheating under s 420 of the Penal Code, I am of the opinion that the aggregate sentence is substantially above the range of 33 to 39 months’ imprisonment. The sentence of 100 months’ imprisonment, which stands at about two and a half to three times the identified range, will have a compounding effect on the severity of the sentence which is disproportionate to the Respondent’s overall criminality: see Raveen at [16].

I am also cognisant of some differences between the present case and the relevant precedents (Gene Chong and Neo) from which the above range is drawn. On one hand, the Respondent in this case had gained a lower amount of pecuniary benefit than the offenders in those precedents (as observed at [54]). On the other hand, the sums cheated in the present case exceed the losses suffered by the victims in those cases.

In addition, the present case differs from the relevant precedents in that it involves amalgamated charges arising from 52 distinct transactions. I note that the doubling of the court’s sentencing jurisdiction for amalgamated charges as provided for in s 124(8) of the CPC only applies to acts which take place on or after 31 October 2018: see s 124(10) of the CPC read with Criminal Justice Reform Act 2018 (Commencement) (No. 2) Notification 2018 (No. S 721). As the acts underlying the present charges were committed between 2014 and 2016, s 124(8) of the CPC does not operate. The range of sentences normally imposed may potentially be adjusted upwards for amalgamated charges framed in respect of acts committed on or after 31 October 2018. That being said, I accept the general proposition that a person who commits multiple acts of offending should, all else being equal, be treated more harshly than one who commits a one-off act, even if both courses of conduct result in the same outcome: Public Prosecutor v Song Hauming Oskar and another appeal [2021] 5 SLR 965 at [69]–[70].

Keeping in mind the overall criminality of the Respondent’s conduct across the two amalgamated charges, I find that an aggregate sentence of 80 months’ imprisonment is appropriate. In the light of my earlier finding that the two sentences should run consecutively, I reduce the aggregate sentence from 100 months’ imprisonment to 80 months’ imprisonment by recalibrating the sentence for the Second Charge to 45 months’ imprisonment.

For completeness, even if I am incorrect in my earlier finding that the two sentences should run consecutively, I am satisfied that the totality principle should operate to increase the accused’s global sentence from 65 months’ to 80 months’ imprisonment. In Anne Gan, the High Court explained at [20] that the totality principle may serve to boost sentences if they would otherwise result in a manifestly inadequate overall sentence:

Given the totality of the Respondent’s criminality as summarised at [44], the global sentence of 65 months’ imprisonment would, in my view, be manifestly inadequate. A boost to the sentence to 80 months’ imprisonment would therefore be warranted to ensure that the sentence is proportionate to the facts and circumstances of the present case.

Finally, I do not consider the sentence of 80 months’ imprisonment to be crushing or not in keeping with the Respondent’s past records or future prospects. A crushing sentence reflects “the destruction of any reasonable expectation of useful life after release”: Kow at para 27.155 citing R v Yates (1985) VR 41 at 48. The aggregate sentence of 80 months’ imprisonment, taking into account the Respondent’s age, will not significantly diminish such expectations.

Conclusion

Flowing from my findings on the two issues above, I find that the sentence imposed by the District Judge is wrong in principle, and a proper application of the one-transaction rule and the totality principle results in a global sentence of 80 months’ imprisonment. I therefore substitute the sentence of 65 months’ imprisonment for the Second Charge with 45 months’ imprisonment and order the sentences for the two offences to run consecutively. In my judgment, this outcome avoids the unjust result of the Respondent escaping punishment for his added culpability of engaging in the second conspiracy to cheat Vermont, while ensuring that the global sentence is proportionate to the totality of the Respondent’s offending conduct. This conclusion also demonstrates that the aggregate sentence of 65 months’ imprisonment imposed by the District Judge was manifestly inadequate. The appeal is thereby allowed to this extent.