Yong Kheng Leong and another v Panweld Trading Pte Ltd and another

Sundaresh Menon JA (delivering the judgment of the court):

[1] The 1 Respondent, Panweld Trading Pte Ltd (“Panweld”), is a private company limited by shares. It was incorporated in Singapore in 1985. The proceedings out of which this appeal arises concerned a claim brought by Panweld against one of its directors, the 1 Appellant, Mr Yong Kheng Leong (“Mr Yong”). It was alleged, and found by the Judge below, that Mr Yong had mismanaged the company and misappropriated some of its funds by making unauthorised payments to his wife, Mdm Lim Ai Cheng (“Mdm Lim”) who is the 2 Appellant. Mr Yong holds 20 percent of Panweld’s shares. The only other director is Mr Loh Yong Lim (“Mr Loh”) and he is the 2 Respondent in this appeal. Mr Loh was also Panweld’s only other shareholder at the material time and owned the remaining 80 percent of its shares. Mr Yong and Mr Loh had known and worked with each other for a long time. Mr Yong had been the director in charge of the day-to-day business of running Panweld since its incorporation.

[2] In March 2009, Mr Yong informed Mr Loh that Panweld needed a bank loan to secure a performance bond for a potential project as it lacked the funds to do so on its own. This surprised Mr Loh who had been under the impression that Panweld was financially healthy. He therefore engaged the firm of certified public accountants, BDO LLP (“BDO”), to undertake a forensic examination of Panweld’s accounts. These investigations revealed several alleged financial misdeeds by Mr Yong, including that his wife, Mdm Lim, had been on Panweld’s payroll and received a salary from Panweld. This had evidently been going on for 17 years from 1992 to 2009 even though, as far as Mr Loh was concerned, she was not and had never been an employee of Panweld.

[3] Panweld’s case was therefore pleaded on this basis: that Mdm Lim was never a genuine employee of Panweld and further there had never been any agreement on Mr Loh’s part for Mdm Lim to receive any salary. Mdm Lim only came to receive these payments as a result of Mr Yong’s sustained breach of the fiduciary duties he owed to Panweld as a director and his abuse of this position throughout the period in question. On this basis, Panweld contended that Mr Yong was liable to it as a constructive trustee of the sums that had been misappropriated. As regards Mdm Lim, Panweld claimed that she should be held liable on the basis of her knowing receipt of these tainted funds and/or the dishonest assistance she rendered to her husband in carrying out his scheme.

[4] Because of the long duration over which the alleged breaches had occurred, the question of limitation arose and on this, Panweld took the position that its claim against Mr Yong for breach of his fiduciary duties fell within the ambit of s 22(1) of the Limitation Act (Cap 163, 1996 Rev Ed) (which will be referred to as “the Limitation Act”) and hence was not time-barred. The same argument was made in relation to the claim against Mdm Lim.

[5] The pleaded case of Mr Yong and Mdm Lim evolved during the proceedings, and it is necessary to unpack this systematically to better understand what had transpired below. When the Defence was first filed on 11 March 2010, Mr Yong claimed that it was Mr Loh who had first placed his wife (“Mrs Loh”) and his mistress (“Sook Min”) on the payroll in March 1995, even though neither of them had ever been employees of Panweld in any meaningful sense. According to Mr Yong, he expressed concern over this and in response, Mr Loh suggested that Mdm Lim, as well, could be placed on Panweld’s payroll. It was implicit in this first version of the Defence that there had been an agreement between Mr Yong and Mr Loh that Mrs Loh, Sook Min and Mdm Lim would each be paid a salary by Panweld even though they were not in fact its employees and would not be required or expected to render it any services.

[6] Panweld filed its Reply denying the alleged agreement between Mr Yong and Mr Loh. It highlighted that Mrs Loh and Sook Min were placed on the payroll in 1995, well after Mdm Lim had started receiving salary payments in 1992. The Defence was then amended on 13 April 2011. The new case advanced on Mr Yong’s behalf was that Mdm Lim was genuinely employed as a marketing executive on a part-time basis from 1992 to 1994 and thereafter, from 1995 on a full-time basis.

[7] The Defence was revised again on 26 May 2011. This time, Mr Yong alleged that Mrs Loh and Sook Min had been employed to reduce the taxes payable by Panweld; and that Mdm Lim’s salary from 1992 to 1994, when she was allegedly genuinely employed by Panweld albeit on a part-time basis, had been paid out of Mr Yong’s bonuses and unused annual leave pay. Further, it was asserted that during the course of her full-time employment from 1995 to 2009, her salary consisted of three components: (a) a fixed salary of $1,125; (b) the transfer of Mr Yong’s car allowance of $1,000; and (c) the balance which came from Mr Yong’s annual increments and/or bonuses. Mr Yong alleged that the fixed salary component of $1,125 was 25% of the total amount paid to Mrs Loh and Sook Min in respect of their salaries and this reflected the shareholding proportion as between Mr Yong and Mr Loh. There is an element of the Defence being tailored here and even then it is not coherent: in effect Mr Yong was asserting that even though Mdm Lim was gainfully employed by Panweld, she received an amount for her work that bore no relation to the value or amount of the work she did but rather bore the same proportion paid to Mr Loh’s wife and his mistress that Mr Yong’s shareholding bore to Mr Loh’s; and everything else she received was actually money that was due to Mr Yong in respect of his car allowance and other benefits.

[8] When the case was presented in the court below, the primary contention for the Defence remained that Mdm Lim was at all material times a genuine employee of Panweld, be it part-time or full-time; and in the alternative, if she was found not to be a genuine employee of Panweld, then the payments had been made with the express approval of Mr Loh in his capacity as a shareholder of Panweld. As such, by the operation of the principle in Re Duomatic Ltd [1969] 2 Ch 365 (“Duomatic”), Mr Loh (as the only other shareholder of Panweld) having agreed to this state of affairs, there was no basis for finding an actionable breach by Mr Yong of his fiduciary duties. As to the question of limitation periods, Mr Singa Retnam (“Mr Retnam”), who appeared below for Mr Yong and Mdm Lim, accepted at the close of the trial on 9 February 2012 that if the claim against Mr Yong was made out, it would not be time-barred. Mr Retnam, however, contended that the claim against Mdm Lim, even if it were made out, would not come within the ambit of s 22(1) of the Limitation Act and would at least, in part, be time-barred pursuant to ss 6(1)(a), 6(1)(d) and 6(2) of that statute.

[9] Mr Yong and Mdm Lim also initiated a third party claim against Mr Loh on the basis that if Mr Yong was found liable to Panweld for the salary payments made to Mdm Lim, he should be entitled to an indemnity or contribution from Mr Loh to the extent these had been made with the latter’s approval.

[10] Mr Loh’s position was that the third party claim against him was misconceived in that, if he had approved the salary payments to Mdm Lim, that would then constitute a complete defence to Panweld’s claims and Mr Yong would not be liable in the first place. According to Mr Loh, there was simply no situation in which Mr Yong could be found liable if he made out his claim that he had acted with Mr Loh’s approval and so the third party claim, which rested on the proof of such approval, was wholly without merit. There was also a subsidiary point that the payments had been received by Mr Yong and/or Mdm Lim for their sole benefit and that Mr Loh had not derived any benefit from them.

[11] The Judge below found in Panweld’s favour and his decision is reported in Panweld Trading Pte Ltd v Yong Kheng Leong and others (Loh Yong Lim, third party) [2012] 2 SLR 672 (“the Judgment”).

[12] The Judge found as a fact that Mdm Lim was never a genuine employee of Panweld. The Judge further found that there was no express agreement between Mr Yong and Mr Loh for Mdm Lim to be put on Panweld’s payroll. Accordingly, Mr Yong was found to be in breach of his fiduciary duty to Panweld. As for Mdm Lim, the Judge found that she knew the funds in question had been paid to her in breach of Mr Yong’s fiduciary duty to Panweld and was therefore liable for the dishonest assistance she gave to Mr Yong, as well as her knowing receipt of the proceeds of his unlawful actions.

[13] On the question of limitations, the Judge accepted Mr Retnam’s concession that Panweld’s claim against Mr Yong fell within s 22(1) of the Limitation Act and therefore that the claim was not subject to any limitation period. The Judge accordingly held that Mr Yong was a constructive trustee of the full amount of the monies that had been misapplied, which the Judge found to be $873,959.20. As for Mdm Lim’s liability, the Judge found that the six-year limitation defence under s 6(7) of the Limitation Act applied. Thus, the amount recoverable against her was confined to the funds wrongfully paid out in the six years immediately preceding the commencement of the action (ie, $338,410 which was the total sum paid out from 2004 to 2009).

[14] As for the third party claim against Mr Loh, the Judge found that it was wholly misconceived and dismissed it. If Mr Loh had agreed to the salary payments, then Mr Yong would not be liable in the first place. On the other hand, if Mr Loh had never agreed to these payments, there would then be no basis to seek any recourse against Mr Loh for an indemnity or a contribution.

[15] In the appeal, Mr Yong and Mdm Lim raised the following arguments in the Appellants’ Case which were then further refined in the course of the oral arguments:

(a) That the Judge erred in quantifying the total sum that had been misapplied over the 17 years at $873,959.20;

(b) That the Judge erred in finding that Mdm Lim was not a genuine employee of Panweld;

(c) That the Judge erred in finding that there had been no agreement on Mr Loh’s part for Mdm Lim to receive a salary and he erred in failing to find that Mr Loh had at least impliedly assented to this by virtue of his knowledge of the relevant facts. Such assent would suffice to exclude Mr Yong’s liability on the basis of the principles laid down in Duomatic as well as in Tokuhon (Pte) Ltd v Seow Kang Hong and others [2003] 4 SLR(R) 414 (“Tokuhon”);

(d) Notwithstanding Mr Retnam’s concession below, the Judge erred in finding that the limitation period did not apply to the claim against Mr Yong, because Panweld’s claim should properly have been characterised as one for equitable compensation (instead of a constructive trust), to which the doctrine of limitation by analogy should have applied so as to impose a six-year time-bar on the claim;

(e) That the Judge erred in finding that Mdm Lim was liable on the basis of knowing receipt and/or dishonest assistance; and

(f) That the Judge erred in dismissing the third party claim against Mr Loh.

It will be evident from the recital of what had transpired before the Judge that arguments (c) and (d) above are new arguments which were not run in the proceedings below and that argument (d) at least was actually inconsistent with the position taken below.

[16] We should also mention that at the hearing before us, Mr Burton Chen (“Mr Chen”), who appeared for Mr Loh, confirmed that Mr Loh remained willing to abide by any order we might make as to the disgorging of the sums that had been paid out to Mrs Loh and Sook Min given the Judge’s finding that they too had not been genuine employees of Panweld (see [49] of the Judgment). The same undertaking was proffered to the Judge below, though he did not make any such order. We will deal with this issue at the end of this judgment.

[17] Before turning to the issues raised in this appeal, we wish to briefly reiterate two important, albeit basic, principles.

[18] First, an appellate court's power to review findings of fact will be sparingly exercised. The trial judge, having had the benefit of hearing the entire trial, would generally be in a better position to assess the veracity and credibility of witnesses where oral evidence is concerned. Therefore, parties seeking to reverse findings of fact that have been made after assessing the oral evidence will face an uphill task, and in general it will be even more so in cases such as the present where the findings of fact were made after hearing many days of evidence. An appellate court will only overturn factual findings in such circumstances if it is satisfied that the trial judge’s assessment is plainly wrong or manifestly against the weight of the evidence. These basic principles have been reiterated by this Court on many occasions (see, for example, Seah Ting Soon v Indonesian Tractors Co Pte Ltd [2001] 1 SLR(R) 53 at [22] and Tat Seng Machine Movers Pte Ltd v Orix Leasing Singapore Ltd [2009] 4 SLR(R) 1101 at [41]).

[19] The second point relates to parties bringing up new points or arguments on appeal which were not taken before or presented to the trial judge. In deciding whether to allow such new arguments to be run, the ultimate consideration is that justice must be done; but an appellate court will consider this from the perspective of both parties. Hence, a critical factor is whether the respondent would be unfairly prejudiced if new points were allowed to be taken at such a late stage. Even if the new arguments are not inconsistent with the case presented below, barring exceptional circumstances, an appellate court will not allow these to be raised if they run into gaps in the evidence in areas which could have been but were not explored below; or if the respondent is going to be prejudiced because it could have cross-examined on matters that are relevant to the new case, but did not do so because it was not relevant or necessary to do so to meet the case that was run below. This principle too has previously been stated by this Court (see, for example, Susilawati v American Express Bank Ltd [2009] 2 SLR(R) 737 at [47]).

[20] With these principles in mind, we turn to the issues raised in this appeal.

[21] This point of appeal may be quickly disposed of. It is misconceived since the quantum of the funds misapplied was derived from a Joint Expert Report that was almost entirely agreed as between the experts. The relevant portion of the Joint Expert Report is reproduced below:

[22] To the extent that the sum of $14,250 for an additional bonus payment to Mdm Lim on 21 December 2008 was not agreed because it was not stated in Mdm Lim’s IR8A form, there was documentary proof that this payment had in fact been made by Panweld to Mdm Lim. This was reflected in: (a) Panweld’s CPF Contribution Submission Form for December 2008; (b) Panweld’s salary record books for December 2008; and (c) the copy of the cheque made out in favour of Mdm Lim for the sum of $12,469.00 (after deducting her CPF contribution).

[23] As set out earlier, Mr Yong and Mdm Lim had contended before the Judge, first, that Mdm Lim was genuinely employed to serve Panweld as a marketing executive; and second, that Mr Loh had in any event expressly agreed to the payments in question being made to Mdm Lim. The Judge found that both contentions were not supported on the evidence. Mr Yong and Mdm Lim are appealing against both findings of fact.

[24] Turning to the Judge’s finding that Mdm Lim was not a genuine employee of Panweld, although arguments on this issue were made in the Appellants’ Case, this finding was not substantially challenged at the hearing before us. In any event, Mr Yong and Mdm Lim have failed to meet the threshold of satisfying this Court that the Judge’s finding on this point was plainly wrong or manifestly against the weight of the evidence. On the contrary, we find the Judge’s finding unimpeachable on the evidence.

[25] As for the finding that Mr Loh had not expressly agreed to the salary payments being made to Mdm Lim, while this might seem counterintuitive given the long period of 17 years during which the misappropriation went on, the question the Judge was presented with was whether there had been an express agreement. When the case was argued before us, Dr Tang Hang Wu (“Dr Tang”), who appeared together with Mr Retnam for Mr Yong and Mdm Lim in the appeal, had finessed the argument and presented it as one founded on implied assent. Dr Tang relied on the principle stated in Duomatic and in Tokuhon which is that subject to the usual limitations concerning the rights of third parties who deal with a company, where all the shareholders, particularly in a closed private company with a track record of informality in their dealings, assent to a particular course of dealing, even in relation to the disposal of assets, this may be effective to bind the parties. However, there are limits to this. In particular, the conduct between the parties must be such that there is sufficient basis for a court to infer: (a) that there was in fact an agreement; and (b) what the key contents of that agreement were. Herein lies the difficulty that Dr Tang faced in his efforts on behalf of Mr Yong and Mdm Lim.

[26] Undoubtedly, Panweld is a closed private company that had only two directors who were also the only shareholders at the material time. It is also undisputed that the salary payments to Mdm Lim were made transparently in that the company accountant knew about these and recorded them in the company’s books. Further, given the long duration of time over which the salary payments were made to Mdm Lim, it seems highly unlikely that Mr Loh was wholly unaware that something of this nature had been going on, especially since Mr Loh knew that Mrs Loh and Sook Min were being paid in the same way, which payments overlapped with the period of time during which Mdm Lim received her payments. However, even taking all these factors at face value, they may at their highest support the inference that there was some degree of forbearance, but there is no basis on which one could come to a clear conclusion: (a) that there was in fact an agreement; and (b) what exactly that agreement was. An appellate court is in no position to fill in these factual gaps. Indeed, it would be prejudicial to Panweld if this was done when the parties did not have the opportunity or occasion to explore these points in cross-examination before the Judge because this was not the case that was run there. In short, even if Mr Loh had some knowledge as to these payments, there is no basis (at least on the evidence that was led below) to find such an agreement or common understanding as could found a claim to any relief.

[27] We note for instance, as was pointed out by Mr Philip Jeyaretnam SC (“Mr Jeyaretnam”), who appeared for Panweld in the appeal, that the amount paid out to Mdm Lim (at the direction of Mr Yong who held 20% of the shares in Panweld) was nearly three times the amount paid out cumulatively to Mrs Loh and Sook Min (who were Mr Loh’s partners), even though Mr Loh held 80% of the shares. This is so inconsistent with the respective shareholding proportion of Mr Yong and Mr Loh that in the absence of this being explained or sufficiently explored in cross-examination, we are unable on appeal to draw any inference that the parties were proceeding on the basis of an agreement. This is unlike the position in Duomatic where Buckley J found as a fact that the two directors, who at the time were also the only ordinary shareholders of the company, had approved the accounts that showed the relevant payments having been made to each of them, after the entry in question had been explained to them by the company’s auditor (see Duomatic at 372E-G).

[28] We note that Mr Yong’s and Mdm Lim’s evidence was that it was the fixed component of Mdm Lim’s salary (ie, the sum of $1,125), as opposed to the total sum paid out, that reflected the shareholding proportion between Mr Yong and Mr Loh. However, as the Judge noted, this alleged salary structure simply does not stand up to scrutiny. When Mr Yong was challenged in cross-examination that the alleged $1,000 car allowance was not reflected in Panweld’s records, he changed his evidence and claimed that the said sum was paid from his “productivity bonus” instead. Issues of credibility aside, there was simply no sensible explanation as to why the balance of Mdm Lim’s salary would be derived from Mr Yong’s bonuses and increments. Moreover, there were differences in both the start and the end dates of the payments to Mdm Lim, Mrs Loh and Sook Min. All these factors make it impossible for us to conclude that there had been an agreement on the basis of which the parties had proceeded.

[29] For these reasons, we do not disturb the Judge’s findings of fact as to Mr Yong’s liability and accordingly uphold his decision that the payments authorised by Mr Yong in favour of Mdm Lim were made in breach of Mr Yong’s fiduciary duties as a director of Panweld.

[30] Due to the long period of time during which the misappropriation took place, it is necessary to consider whether, and if so to what extent, Panweld’s claim against Mr Yong is time-barred.

[31] As mentioned earlier, in the trial below, Mr Retnam had conceded that no defence of limitation would avail Mr Yong if the breach of fiduciary duty was made out. Dr Tang sought to withdraw this concession before us. He argued that the doctrine of limitation by analogy applies in this case to bar, at least in part, the claim against Mr Yong.

[32] Although an appellate court is not, as a matter of principle, bound by a concession of law made by counsel below, the critical question that remains is whether Panweld would be prejudiced if we permitted Mr Retnam to withdraw this concession now (see NV Multi Corp Bhd & Ors v Suruhanjaya Syarikat Malaysia [2010] 5 MLJ 573 at [16]). Before us, one of Mr Jeyaretnam’s arguments was that there was no basis at all for any limitation defence to arise because this was a defalcation by a director who was using his wife as his proxy. According to Mr Jeyaretnam, the Judge had actually made such a finding at [39] of the Judgment:

[33] Dr Tang conceded that if the facts were as Mr Jeyaretnam contended, no defence of limitation would avail his client. Dr Tang argued however that the Judge’s observations (at [39] of his Judgment) fell short of such a finding. Even if that were so, which we doubt, it would not assist Dr Tang because Panweld would then potentially be prejudiced by Mr Yong’s late attempt to change his case. As Mr Jeyaretnam submitted, had counsel having conduct of Panweld’s case below known that the defence of limitation would be in issue, he could and probably would have pursued this further in cross-examination and the Judge could and probably would have made his finding on this issue unequivocal.

[36] Although the issue of limitations was not before the Judge in light of Mr Retnam’s concession, the Judge observed that in his view the concession had been correctly made. Mr Yong was found to be a constructive trustee over the assets of the company, which he had misappropriated. But in cases such as Paragon Finance plc v D B Thakerar & Co (a firm) [1999] 1 All ER 400 (“Paragon Finance”) and Selangor United Rubber Estates Ltd v Cradock and Others (No 3) [1968] 1 WLR 1555, two distinct types of constructive trusts are identified. The first category (referred to as “Class 1 constructive trusts” or “Class 1 constructive trustees” when referring to the trustees) may potentially be denied any limitation defence, whereas those in the second category (referred to as “Class 2 constructive trusts” or “Class 2 constructive trustees”) generally may avail themselves of any defence of limitations. The Judge therefore characterised the central issue thus: what kind of constructive trustee is Mr Yong?

[37] The Judge (at [67] to [68] of the Judgment) concluded that the line between Class 1 and Class 2 constructive trustees is to be drawn according to whether the trust or trust-like relationship existed even before the wrongful transaction in question, in which case, it would be a Class 1 constructive trust (see, for example, JJ Harrison (Properties) Ltd v Harrison [2002] 1 BCLC 162 (“JJ Harrison”)). If it did not and if the constructive trust only came into existence as a result of the wrongful conduct and was in fact a remedy designed to address the consequences of that conduct, then it would be a Class 2 constructive trust. He found that Mr Yong was a Class 1 constructive trustee since he was a director of Panweld and had stood in a trustee-like position vis-a-vis Panweld’s assets even before the misappropriation took place.

[38] Before us, Dr Tang argued that even if Mr Yong was in breach of his fiduciary duties, he would be a Class 2 rather than a Class 1 constructive trustee. Dr Tang submitted that the line between Class 1 and Class 2 constructive trusteeship ought to be drawn based on whether the property in question is still being held by the defendant. On this basis:

(a) a Class 1 constructive trusteeship arises when the court finds that the defendant is in possession of the property in question and declares that he holds it on trust for the plaintiff. In such a situation, the relief that avails the plaintiff is proprietary in nature; and

(b) a Class 2 constructive trusteeship arises when the court holds that a defendant is liable to account as if he were a constructive trustee even though he does not still have the property. In reality, there is no real trust in such a case, and the relief denotes a liability to account in equity which is personal in nature.

[39] Dr Tang argued that this distinction also followed from the fact that the underlying basis for s 22(1) of the Limitation Act, which is an exceptional provision in that it excludes the ability of a defendant to rely on a defence of limitation, is that an express trustee’s possession of the trust property is never in virtue of any right of his own, but instead is always on behalf of the beneficiaries. Accordingly, time never runs against the beneficiaries of an express trust. Although s 22(1) of the Limitation Act has been extended to constructive trusts, Dr Tang submitted that the extension should not be taken too far and should, in principle, be confined to those constructive trustees who, like express trustees, retain possession of the trust property. Further, as a matter of first principles, a true constructive trust must latch on to an identifiable fund. Otherwise, the remedy cannot be said to be proprietary in nature and is merely a personal liability to make equitable compensation for the breach of fiduciary duty. Since it is undisputed that Mr Yong was never in possession of the misappropriated funds as the funds were transferred directly to Mdm Lim’s bank account, Dr Tang argued that Mr Yong does not meet the requirements for Class 1 constructive trusteeship.

[40] Further, according to Dr Tang, the applicable limitation regime in this case is to be determined by the fact that the claim should properly be characterised as one for equitable compensation. Consequently, by operation of the doctrine of limitation by analogy (which was recognised by the High Court in Katherine Tang Woon Kiang v Luk King Hung [1999] SGHC 229 (“Katherine Tang”)), the six-year time-bar under s 6(1)(a) read with s 6(7) of the Limitation Act would apply to the claim since it is essentially based on the same factual allegations as that for a common law claim in tort. The relevant provisions of s 6 of the Limitation Act are as follows:

[41] In response, Mr Jeyaretnam argued that possession of trust property should not be the determining factor as to whether the trust should be characterised as a Class 1 or Class 2 constructive trust. He submitted that such an approach would incentivise conduct designed to activate the time-bar by the simple device of parting with the property. Mr Jeyaretnam further argued that the line between Class 1 and Class 2 constructive trusteeship is more properly drawn according to whether the trust or trust-like relationship existed before the transaction in question, rather than as a remedy designed to address the conduct leading to the imposition of the trust. This approach, Mr Jeyaretnam argued, was more logical and ought to be preferred since it focuses on the fact that a person who enjoys a power to dispose of the property in respect of which he owed fiduciary duties ought to be in no better a position just because he has abused that power to separate himself from the property. Mr Jeyaretnam further argued that the doctrine of limitation by analogy does not apply in Singapore as the Limitation Act (in Singapore) is different from the corresponding English statute, and the judge in Katherine Tang had failed to appreciate this material fact.

[42] We set out our observations on the following four points:

[43] The alternative formulation of equitable compensation that was suggested by Dr Tang is inappropriate in the present scenario. The possible remedies that the Judge could, in theory, have granted include equitable compensation, rescission, account, injunctive relief, as well as other proprietary reliefs such as the imposition of a constructive trust. As a matter of principle, it would have been open to Panweld, if it had so pleaded, to elect between the available remedies (see Warman International Ltd v Dwyer (1995) 128 ALR 201). In other words, there could well have been an additional admissible claim for equitable compensation. But equitable compensation is a claim for damages for breach of a fiduciary duty; it is not the primary, much less the only remedy where one is concerned with the misappropriation of trust property. In situations like the present, case law quite clearly supports the imposition of liability on the errant director as a constructive trustee (see, for example, Gwembe Valley Development Co Ltd (in receivership) and another v Koshy and others (No 3) [2004] 1 BCLC 131 (“Gwembe Valley”) at [142] and JJ Harrison at [25]–[27]). Thus, a director who in breach of his fiduciary duties causes a loss to the company, but does not himself receive any of the company’s property, would nonetheless be liable for equitable compensation. There would be no question of any proprietary remedy being imposed in such a case because it is simply a matter of compensating the company for the loss it has suffered as a result of the breach of his duties. Where, however, a director has misappropriated the company’s property, there is no basis in principle for finding that the company’s only remedy is one for equitable compensation.

[44] We agree with the Judge that Mr Yong ought to be characterised as a Class 1 constructive trustee and not as a Class 2 constructive trustee. As a preliminary point, Dr Tang argued that this could not be a Class 1 constructive trust because there was no property on which the trust could latch. In our judgment, this argument is incorrect in principle because it confuses the remedy with the cause of action. Specifically, it mistakenly focuses on whether a trust may be imposed now and remains available as a remedy, rather than on whether there was a trust at the material time, and in respect of which, the defendant has acted in breach of his duties. The issue before us is whether there is an applicable limitation period and if so, what that may be. The answer to this depends in the first place on what the nature of the claim is and not on what the available remedies may (or may not) be at the time of judgment; if it is a claim for a breach of trust, then it remains so whether or not the trust property remains with the defendant. To take a straightforward example, if an express trustee takes trust money and fraudulently misappropriates it, he will be liable for breach of trust even if he has lost it all and there is no property left that the trust can latch on to. It may be that because the trust money is no longer available, a proprietary remedy is no longer available and the only remedy in fact available is equitable compensation, but that does not change the fact that the cause of action would be for breach of trust and the relevant limitation period would be based on that. It is undisputed here that the misappropriated funds constituted trust property in that it was the property of Panweld over which Mr Yong had the power of disposal. Indeed, in this case, he was effectively the only director running the company. A director who disposes of company property in breach of his fiduciary duties is treated as having acted in breach of trust (see JJ Harrison at [25]). Because of this, a director who obtains property of the company for himself is treated as a Class 1 constructive trustee (see JJ Harrison at [29] and [30]). The relevant property that the trust latches on to is the company’s property, over which the director has the power of disposal (see Gwembe Valley at [102], citing Taylor v Davies [1920] AC 636). This approach is also more logical as it focuses on the fact that a trustee who enjoys a power to dispose of property in respect of which he owed fiduciary duties ought not to be in a better position just because he has abused that power to separate himself from the property.

[45] This is entirely in accord with the analysis of Millett LJ in Paragon Finance at 408-409 where he said:

[46] This is the essence of the distinction between Class 1 and Class 2 constructive trustees. If a person holds property in the position of a trustee (and there is no doubt that a director is regarded as a trustee over the company’s property – see [44] and [45] above) and deals with that property in breach of that trust, he will be a Class 1 constructive trustee; whereas a wrongdoer who fraudulently acquires property over which he had never previously been impressed with any trust obligations, may, by virtue of his fraudulent conduct, be held liable in equity to account as if he were a constructive trustee. But the latter is not a case of someone who had ever in reality been a trustee of that property; and it is only by virtue of equity’s reach that such a person is regarded as a Class 2 constructive trustee.

[47] That this is the essence of the distinction can be seen by comparing two English cases. In JJ Harrison, the claimant was a company, of which the defendant was a director. A property owned by the company was valued at £8 million, but in a side letter the valuer said that it may have some development potential that had not been taken into account in the valuation. The side letter was not disclosed to the company, but the director was aware of it. A year later, he bought the property from the company without disclosing the side letter. He subsequently sold it for a profit. More than six years later, the company sued him for an account of the proceeds of sale of the land. His limitation defence failed on the ground that he was regarded as a Class 1 constructive trustee. Whereas, in Gwembe Valley, the defendant shareholder-director of the claimant company owned a majority of the shares in and controlled another company. The defendant arranged for the other company to loan money to the claimant company without disclosing his interest in the other company to the board of the claimant company. It was held that, apart from fraud, the claim would have been time-barred because the defendant’s liability to account for the secret profit was not within Class 1. As noted in Halton International Inc and another v Guernroy Ltd [2006] EWCA Civ 801 (“Guernroy”), the difference between JJ Harrison and Gwembe Valley was that the property in JJ Harrison was trust property that had been acquired by the director from the company at an undervalue; whereas, in Gwembe Valley, the defendant’s liability arose from his failure to disclose the true rate of exchange he had paid, and not because there was any misappropriation of specific property that belonged to the company (see Guernroy at [15] and [16] and Gwembe at [119]).

[48] In the present case, Mr Yong as a director of Panweld had trustee-like responsibility for its assets. He was, by virtue of his directorship, lawfully able to deal with Panweld’s assets, albeit in accordance with his fiduciary duties. When he disposed of Panweld’s assets unlawfully, whether to his wife or to himself through his wife, he was undoubtedly a Class 1 constructive trustee because he had dealt with that property in breach of the trust and confidence that had been placed in him as a director.

[49] The classification of Mr Yong as a Class 1 constructive trustee is not the end of the inquiry. The significance of this classification lies in the fact that the limitation regime that applies to claims against such trustees is the same as that which applies to claims against an express trustee. By virtue of s 22(2) of the Limitation Act, a time-bar of six years applies even in an action for breach of trust or for the recovery of trust property, subject to the two exceptions found in s 22(1). The relevant provisions are reproduced below:

[50] The rationale underlying the English equivalent of s 22(2) of the Limitation Act was explained by Millett LJ in Paragon Finance (at 409-410):

[51] Put simply, it was considered unduly harsh that trustees who had committed innocent breaches should remain liable indefinitely. Hence, there was a general stipulation of a limit – 6 years – with exceptions to apply in cases where the trustee retained possession of the property or had acted fraudulently. It may be noted in passing that these are two separate exceptions and that in cases where the claim is founded on a fraudulent breach of trust, the time-bar will not apply even though the trustee is no longer in possession of the trust property or its proceeds. In Paragon Finance, Millett LJ was of the view that Class 2 constructive trustees would not be subject to the English statutory provision that corresponds to s 22 of the Limitation Act. Rather, by virtue of the doctrine of limitation by analogy, claims against Class 2 constructive trustees would be subject to the time limits provided in the relevant statute of limitations for the corresponding cause of action at law (see Paragon Finance at 409–412). This was also the view of a differently constituted panel of the English Court of Appeal in Guernroy (at [10] and [11]). We will elaborate on the doctrine of limitation by analogy below (at [56] to [78]) but for present purposes, we record our agreement with the views expressed in Paragon Finance and Guernroy. We note that in Gwembe Valley, a different panel of the English Court of Appeal appeared to come to a different view on this point. In Gwembe Valley, it was held (at [119]) that the defendant was properly to be regarded as a Class 2 constructive trustee. However, it was also held (at [111]) that the fraud exception pursuant to which the statutory time-bar otherwise applicable to actions for breach of trust would be rendered inapplicable, would apply both to Class 1 and Class 2 constructive trustees. We do not agree with this. In our judgment, the rule that excludes the applicability of the time-bar in certain actions is an exceptional one and one of the primary reasons for drawing the distinction between Class 1 and Class 2 constructive trusts is to identify those cases where it is appropriate to expose a prospective defendant to liability for a potentially indeterminate period. The ancient rule that time does not run against an express trustee because from first to last he holds the property never in his own right but for his beneficiaries (see the extract from Paragon Finance quoted at [45] above), has been modified by s 22(2) but subject to the provisions of s 22(1). In our judgment, only Class 1 constructive trusts fall within the ambit of this provision.

[52] Having crossed the threshold of deciding that Mr Yong is a Class 1 constructive trustee and hence in principle is subject to the time-bar prescribed in s 22(2) of the Limitation Act, the next inquiry therefore is whether ss 22(1)(a) or (b) applies to exclude this time-bar. Turning first to s 22(1)(a), the concept of fraud in the context of the English equivalent of our s 22(1)(a) was discussed by the English Court of Appeal in Gwembe Valley (at [131] and [132]):

[53] We agree with and adopt this approach to assess the applicability of s 22(1)(a). Mr Yong could not have been said to have acted honestly in authorising the wrongful payments to his wife. Certainly, such payments would not be countenanced by the ordinary standards of reasonable and honest people. And Mr Yong himself could not possibly have believed that he was acting in the interests of the company by making these payments to his wife; he either knew it was contrary to the interests of the company and his co-shareholder for that matter, or was indifferent to those interests. Fraud is thus quite clearly made out on the facts and s 22(1)(a) is satisfied.

[54] As for s 22(1)(b), if, as we think is the case, the Judge had made a finding that Mdm Lim was merely a proxy (as set out in [39] of the Judgment) (see [32] to [34] above), then Mr Yong would equally be caught by this section. For all intents and purposes, and in substance, the misappropriated funds were in Mr Yong’s possession or for his benefit. It ought not to make a difference that the funds were parked with his wife. The Court in Gwembe Valley made similar observations about the irrelevance of the director having parked the secret profits in his other company (see Gwembe Valley at [137] and also Re Pantone 485 Ltd, Miller v Bain and others [2002] 1 BCLC 266 at [44]). Nonetheless, there is no need for us to reach a final conclusion on this because as noted in [53] above, Mr Yong had acted fraudulently as a Class 1 constructive trustee and that would be sufficient to deny him the limitation defence.

[55] In sum, even if Mr Retnam had not conceded the limitation defence before the Judge, or even if we were not in a position to dismiss the appeal on this point on the facts as we have done, it would not have affected the outcome because we would nonetheless have found that Mr Yong could not rely on the defence of limitation here.

[56] Dr Tang’s argument on the doctrine of limitation by analogy in Singapore is premised on the claim being characterised as one for equitable compensation instead of a constructive trust. Since, as we have earlier explained, it would be inappropriate to characterise the claim purely as one for equitable compensation instead of as a Class 1 constructive trust, the rest of Dr Tang’s argument cannot succeed. We nonetheless make some further observations about the applicability of the doctrine of limitation by analogy in Singapore as this was challenged by Mr Jeyaretnam.

the arguments raised

[57] Dr Tang relied principally on Katherine Tang in support of his contention that the doctrine of limitation by analogy is applicable in Singapore. In Katherine Tang, G P Selvam J held:

[58] As noted above at [41], the nub of Mr Jeyaretnam’s challenge to the applicability of the doctrine of limitation by analogy was that Selvam J in Katherine Tang had failed to appreciate a material difference between our legislation and the corresponding English legislation. To assess this argument, it is necessary to trace the history of our legislation.

[59] When the relevant statutory provision was originally enacted as s 6(6) of the Limitation Ordinance 1959 (No 57 of 1959) (“the 1959 Limitation Ordinance”), it provided as follows:

[60] Subsequently, when the Limitation Act (Cap 10, 1970 Rev Ed) (“the 1970 Limitation Act”) was reviewed by the Law Revision Commissioners, the words “(if necessary by analogy)” were omitted from the equivalent provision. Section 6(8) of the 1970 Limitation Act provided as follows:

[61] The omission of these words cannot be traced to any legislative amendment made between 1959 and 1970 (or subsequently, for that matter). Section 6(8) of the 1970 Limitation Act is in pari materia with s 6(7) of the Limitation Act that is currently in force in Singapore today. Section 6(7) of the Limitation Act provides:

[62] The equivalent of s 6(7) of our Limitation Act is s 36 of the English 1980 Limitation Act which provides as follows:

[63] Mr Jeyaretnam submitted that: (a) the omission of the words “(if necessary by analogy)” in the 1970 Limitation Act and subsequent Limitation Acts leaves no room for application of the doctrine in Singapore; and (b) Selvam J in Katherine Tang wrongly relied on the English position (in particular, in Knox v Gye (1872) LR 5 HL 565 (“Knox v Gye”)) as he had failed to appreciate that with the omission of those words, our statutory provision was materially different from the corresponding English provision.

our views

[64] To the extent that Mr Jeyaretnam’s argument was put on the footing that the words in the statute which expressly referred to the applicability of the limitation periods by analogy were omitted from the 1970 Limitation Act and subsequent editions, the point can be answered quite shortly. We do not think that the omission could possibly have been intended to signify that the doctrine of limitation by analogy would no longer apply in Singapore. Under the Revised Edition of Laws Act (No 16 of 1966) (the “1966 Revised Edition of Laws Act”), which was the governing law at the time of the preparation of the 1970 Limitation Act, the powers of the Law Revision Commissioners to make omissions was circumscribed by the provisions of s 4, which provided as follows:

[65] It is evident from this that the Law Revision Commissioners only had the power to make omissions which, broadly speaking, either: (a) did not affect the meaning of the affected statute; or (b) had been repealed either expressly, specifically or by necessary implication. Any other omission, particularly one that had the effect of changing the scope or application of the statute in question, would have had to be presented to Parliament and passed as law as provided under s 7 of the 1966 Revised Edition of Laws Act, which reads:

[66] Since: (a) no Act was passed to give effect to the deletion of the words “(if necessary by analogy)” before the 1970 Limitation Act came into operation; and (b) no Act was passed to amend s 6(8) (and its subsequent equivalent provisions) after the 1970 Limitation Act came into operation, it must follow that the omission was neither intended nor contemplated to make any difference to the meaning of s 6(6) of the 1959 Limitation Ordinance. In short, the Law Revision Commissioners did not have the power to effect any substantive change to this (or, for that matter, any other statute) and the omission of these words could not validly have affected or altered the scope and meaning of s 6(6) of the 1959 Limitation Ordinance or its successor provisions.

[67] This is sufficient to dispose of the argument but for completeness, we also observe that there is a yet more fundamental difference in the wording and structure of s 6(7) of the Limitation Act and the corresponding English provision, which strengthens our view that the doctrine of limitation by analogy does apply in Singapore.

[68] This can be seen first by examining s 6 of the Limitation Act which provides:

[69] The effect of s 6(7) is that “this section”, ie the entire s 6, applies to all claims for equitable relief, whether these be founded upon contract, tort, a trust or other ground in equity. This, however, is expressly subject to s 22 (limitation of actions in respect of trust property) and s 32 (the equitable jurisdiction to refuse relief on the ground of acquiescence, laches or otherwise). In other words, outside the operation of the two exceptions in ss 22 and 32, s 6(7) contemplates that the relevant limitation period for a particular cause of action in law (eg, six years for claims for damages for breach of contract under s 6(1)(a)) will also apply when the claim is for equitable relief instead (eg, specific performance of that contract).

[70] Section 6(6) of the 1959 Limitation Ordinance and its English correspondent, s 36(1) of the English Limitation Act 1980, are juxtaposed for comparison:

[71] The language of the English provision indicates that the statutory limitation periods would not apply to equitable reliefs, save to the extent that the doctrine of limitation by analogy made them applicable. In contrast, the language of the Singapore provision indicates that the statutory limitation periods would apply in effect to all claims for equitable relief. This can only be given effect to by finding the particular provision elsewhere in s 6 by which an express limitation period has been prescribed for an action at law, which bears the closest correspondence to the relevant claim for equitable relief. Seen in this light, it becomes evident why the Law Revision Commissioners in 1970 considered that s 6(6) of the 1959 Limitation Ordinance was wide enough to render the words “(if necessary by analogy)” superfluous.

[72] The way in which our provision has been drafted in fact enables us to avoid much of the difficulty evident in the bedevilled history of the development of English law in this area (see for instance Tito and Others v Waddell and Others (No 2) [1977] 1 Ch 106 (see especially 249–252); Coulthard v Disco Mix Club Ltd and another [2000] 1 WLR 707 (“Coulthard”) and Cia de Seguros Imperio v Heath (REBX) Ltd and others [2001] 1 WLR 112 (“Cia de Seguros”).

[73] Though the streams of equity and the law sprang out of different sources, it became increasingly untenable to keep them flowing entirely separately as litigants often looked to equity to overcome the harsh rigidity and formalism of the law. It was out of this that the doctrine of limitation by analogy developed so that limitation periods applicable to claims in law could not be side-stepped by the simple device of seeking equitable relief. The point is succinctly and clearly explained by Jules Sher QC sitting as a deputy High Court Judge in Coulthard where, after citing a passage from the speech of Lord Westbury in Knox v Gye (at 674–675) which he described as “the best description of the circumstances in which the court of equity acted by analogy to the statute,” he continued (Coulthard at 730):

[74] The approach in Coulthard was endorsed by the English Court of Appeal in Cia de Seguros. In our judgment, essentially the same result was achieved by s 6(7) of the 1959 Limitation Ordinance and its successor provisions and this is unaffected by the omission of the words “(if necessary by analogy)”.

[75] It will be recalled that s 6(7) of the 1959 Limitation Ordinance and its successor provisions make this regime subject to ss 22 and 32 of the same Ordinance and its successor legislation. This is the consequence of two things. First, not every claim to equitable relief will have a corresponding claim in the law such that the relevant limitation period specified for the latter can be readily applied by analogy. There is a historical rationale for this. The confluence of the streams of law and equity was such that there nonetheless remained areas in which only equity could intervene. This was referred to as the exclusive jurisdiction of equity as opposed to its concurrent or auxiliary jurisdictions.

[76] The concurrent, auxiliary and exclusive jurisdictions of equity are explained as follows by William Swadling in his chapter on “Limitation” in Breach of Trust (Hart Publishing, 2002) (Peter Birks and Adrianna Pretto eds) at p 323:

[77] For claims that fell within the exclusive jurisdiction of equity, since the plaintiff had no legal claim at all, there was no basis for invoking a statutory limitation. Instead, the equitable doctrine of laches applied, although the courts, in determining the time limit for laches, would usually follow the lead given by the Legislature and adopt the statutory period of limitation (see Smith v Clay (1767) 3 Bro CC 639).

[78] Under our framework, s 22 statutorily deals with the limitation period that is specifically applicable (or not, as the case may be) to claims for breach of trust or for the recovery of trust property, which claims would ordinarily have fallen within the exclusive jurisdiction of equity. We have dealt with this above (at [49] to [54]). Second, equity developed its own rules that allowed for equitable defences to be raised against equitable claims. Section 32 preserves the applicability of those equitable defences (including laches and acquiescence) which may arise according to the facts of the case in any setting where equitable remedies are sought. Accordingly, we agree with Dr Tang that the doctrine of limitation by analogy applies in Singapore.

[79] We turn to consider the appeal against Mdm Lim’s liability. The Judge found that Mdm Lim was liable for knowing receipt since there was: (a) a disposal of Panweld's assets in breach of Mr Yong's fiduciary duty; (b) beneficial receipt by Mdm Lim of these assets; and (c) knowledge on Mdm Lim's part that she received these assets as a result of Mr Yong’s breach of his fiduciary duty. Mdm Lim’s liability for dishonest assistance was also made out given that: (a) Mr Yong, as Panweld's director, was in the position of a trustee in relation to Panweld’s assets; (b) Mr Yong breached that trust when he misapplied monies belonging to Panweld; (c) Mdm Lim facilitated Mr Yong's breach by allowing her bank account to be used for such purposes; and (d) Mdm Lim was dishonest in that she knew full well that she was not entitled to the salaries paid into her account over the course of those 17 years.

[80] It was submitted on Mdm Lim’s behalf that she should not be held liable for knowing receipt because she had no knowledge that the payments had been made in breach of Mr Yong’s fiduciary duty when she received the funds. It was also argued that Mdm Lim should not be held liable for dishonest assistance because Panweld had not discharged its burden of proving that her conduct was dishonest. It was submitted that Mdm Lim is not a sophisticated or highly educated business person; rather, it was said, she is a traditional wife who trusted her husband.

[81] We see no basis for disturbing the Judge’s findings on this issue. The test for liability for knowing receipt is whether the recipient possessed the “state of knowledge ... such as to make it unconscionable for him to retain the benefit of the receipt” (see George Raymond Zage III and another v Ho Chi Kwong and another [2010] 2 SLR 589 at [23]). This can either be actual knowledge or the wilful avoidance of knowledge (see Comboni Vincenzo and another v Shankar's Emporium (Pte) Ltd [2007] 2 SLR(R) 1020 at [64]). In a similar vein, for liability for dishonest assistance to attach, the assistor does not need to know exactly what is going on so long as he suspects that something dishonest might be going on (see Banque Nationale de Paris v Hew Keong Chan Gary and others [2000] 3 SLR(R) 686 (“Gary Hew”) at [147] and [148]).

[82] Mdm Lim knew that she was not a genuine employee of Panweld and she therefore either knew, or wilfully avoided knowing, that the only reason the payments were made into her bank account was because her husband was channelling funds from Panweld to her in breach of his fiduciary duty. This was not a one-off transaction such that it might be said Mdm Lim had just been careless or was misled as to the reason she was paid. On the contrary, she received substantial salary payments, filed returns, and paid income tax on the same, for an extended period of 17 years, knowing all the while that she was not a genuine employee of Panweld. Significantly, she stood to gain, and did in fact gain, substantial benefits from the arrangement (see Gary Hew at [166]).

[83] Accordingly, we uphold the Judge’s findings of liability on the part of Mdm Lim. There was no appeal against the finding that the six-year statutory limitation period applied to the claim against Mdm Lim; and as there is no merit in the appeal against the Judge’s finding as to the quantum of monies misappropriated, we affirm the Judge’s finding that Mdm Lim is liable for the sum of $338,410, being the sum that the parties agreed was paid out in the last six years immediately preceding the commencement of this action.

[84] We dismiss this aspect of the appeal because we agree with the Judge that the third party claim against Mr Loh had absolutely no merit. If Mr Loh had agreed to the salary payments, then Mr Yong would not be liable in the first place. On the other hand, if Mr Loh had not agreed to this, then there would be no basis for seeking any indemnity or contribution from Mr Loh.

[86] In the interests of securing finality in this litigation, to the extent that payments were made to Mrs Loh and Sook Min, we order that an account be taken of all such payments and for those monies to be reimbursed to Panweld. The parties are directed to exercise their best efforts to come to an agreement as to: (a) the period over which such payments were made; (b) the quantum of these payments; (c) the terms of reference for the account to be taken; and (d) all other matters relevant to put an end to this litigation. Should the parties be unable to reach an agreement on these issues, they have liberty to apply to this Court for further directions.