Pars Ram Brothers (Pte) Ltd (in creditors’ voluntary liquidation) v Australian & New Zealand Banking Group Ltd and others [2018] SGHC 60

The Liquidators of the plaintiff Pars Ram Brothers (Pte) Ltd (“the Company”) applied under s 310(1)(a) of the Companies Act (Cap 50, 2006 Rev Ed) (“the Companies Act”) for the court to determine the method of distribution with regard to the sale proceeds of four categories of pepper stock held on trust by the Liquidators for the Company’s creditors, the defendants. This application raises the question of the appropriate method of distribution when assets have been commingled into a mixed bulk and are insufficient to satisfy all the claims made in respect of it by claimants (none of whom is a wrongdoer towards another) who have a security interest in the mixed bulk.

Background

Prior to its liquidation, the Company was in the spice business, trading primarily in pepper and cashew nuts. It financed its import of stock mostly through trade financing facilities granted by banks. Typically, lending banks would disburse funds directly to the relevant supplier of stock upon being furnished with proof of the Company’s purchase. As security for the loan, the Company would pledge the shipping documents (eg, the bill of lading) for the financed stock to the relevant bank. Thereafter, to enable the Company to sell the stock to its end-customers, the bank would release the relevant shipping documents to the Company. In consideration for this release, the Company would execute a trust receipt on terms that the Company held the financed stock or its proceeds of sale on trust for the bank. The trust receipts typically identified the financed stock with reference to their corresponding bill of lading number, invoice number and/or a description of the financed stock.

After the Company became insolvent, the Liquidators were informed of stock in the Company’s possession which was held in a warehouse in Singapore, which included 17 different categories of pepper. As they were perishable, the plaintiff proposed to sell the stock and hold the proceeds on trust for the creditors pending the determination of their claims. None of the creditors objected to this proposal.

On 7 February 2017, Steven Chong J (as he then was) heard the plaintiff’s application under s 310(1)(a) of the Companies Act for the court to determine whether the gross sale proceeds of 12 categories of pepper stock should be (a) held for the benefit of the general pool of the Company’s creditors; or (b) paid to the defendant lenders who could assert a security interest in the pepper stock which they financed. Chong J found that the gross sale proceeds of the pepper stock should be paid (in proportions to be resolved separately) to those defendants who could assert a security interest in the underlying stock which they financed, save for the proceeds for four categories of pepper stock (“the Disputed Categories”) in which the general creditors were also entitled to assert an interest in certain quantities of the stock. The grounds for his decision can be found in Pars Ram Brothers (Pte) Ltd (in creditors’ voluntary liquidation) v Australian & New Zealand Banking Group Ltd and others [2017] 4 SLR 264 (“Pars Ram”).

All of the Company’s stock, save for the Disputed Categories, has been accounted for or distributed. The Disputed Categories consist of:

The claims in respect of the stock under the Disputed Categories exceed their respective sale proceeds. It is also undisputed that stock for the Disputed Categories have been commingled into a mixed bulk. This brings me to the present application, where the parties sought a determination on the method of distribution to be adopted for the sales proceeds of the Disputed Categories of pepper. The underlying legal question is this: when assets commingled into a mixed bulk are insufficient to satisfy all claims made in respect of the said assets by claimants (none of whom is a wrongdoer towards another) who have a security interest in them, what method of distribution should be used?

The various methods of distribution

The foreign case authorities have identified a number of possible approaches to the distribution of mixed funds where such funds are insufficient to satisfy every claim. This includes the “first in, first out” approach, the pari passu approach (or what the Liquidators termed as the single-block method) and the rolling charge approach (or what the Liquidators termed as the multi-block method). The parties had confirmed that they were not advocating for the “first in, first out” method. The second defendant, Bank of Baroda, submitted that the rolling charge method should be used whilst the fifth and sixth defendants, DBS Bank Ltd and Indian Bank, argued for the pari passu method to be applied. The rest of the defendants, whilst they had previously informed the Liquidators of their preferred approach, did not attend and make submissions in court. The Liquidators maintained a neutral position although they supported the rolling charge method. In the Annex to this Grounds of Decision, I have inserted a table (produced by the Liquidators and not disputed by the defendants) illustrating how the selection of the method of distribution affects the defendants’ share of the sale proceeds.

Having heard the submissions, I determined that the rolling charge method, rather than the pari passu method, should be adopted. Given the paucity in local case authority on this issue, it would be useful to discuss the various methods that the courts have considered before coming to the reasons for my decision.

The “first in, first out” method

The “first in, first out” approach was set out in Devaynes v Noble [1814-23] All ER Rep 1, more commonly known as Clayton’s Case. Essentially, “when sums are mixed in a bank account as a result of a series of deposits, withdrawals are treated as withdrawing the money in the same order as the money was deposited” (Barlow Clowes International Ltd (in liq) and others v Vaughan and others [1992] 4 All ER 22 (“Barlow Clowes”) at 35).

The scope of application for the rule in Clayton’s Case is rather limited. For one, it is questionable whether it has any application out of a banker-customer relationship (Barlow Clowes at 44; Re Diplock’s Estate, Diplock v Wintle [1948] 1 Ch 465 (“Re Diplock”) at 555; Re Ontario Securities Commission and Greymac Credit Corp (1986) 55 OR (2d) 673 (“Greymac”) at 688). In Q & M Enterprises Sdn Bhd v Poh Kiat [2005] 4 SLR(R) 494 (“Q & M Enterprises”) at [56], Andrew Phang JC (as he then was) observed that the rule was applied almost invariably in the context of running accounts as such accounts are, ceteris paribus, ideal situations for the application of the rule. While Phang JC posits that there was no reason to treat any debtor-creditor relationship involving a running or current account differently, he also notes the judicial limitation of the applicability of the rule in Clayton’s Case, which can be attributed to the rule’s perceived arbitrariness. The effect of the rule is that it favours later contributors over earlier contributors. Citing the dicta of Learned Hand J in Re Walter J Schmidt & Co, ex p Feuerbach (1923) 298 F 314 at 316, Woolf LJ in Barlow Clowes agreed that the fiction of the “first in, first out” method apportioned a “common misfortune through a test which has no relation whatever to the justice of the case”, and that it led to “capricious consequences” (at 35).

It is observed that the courts have in various subsequent cases distinguished or disapplied the rule in Clayton’s Case in favour of a method that would produce a more just and equitable outcome. In Re Diplock, Lord Greene MR characterised the rule as “really a rule of convenience based on so-called presumed intention” (at 554). In Barlow Clowes, Woolf LJ held (at [42]) that if the application of the rule would be impracticable or result in injustice between the investors it will not be applied if there is a preferable alternative or if the application of the rule would be contrary to the express, inferred or presumed intention of the investors. In Russell-Cooke Trust Co v Prentis [2003] 2 All ER 478 at [55], Lindsay J, in rejecting the use of the rule in Clayton’s Case and adopting the pari passu method instead, held that the modern approach in England is not “to challenge the binding nature of the rule but rather to permit it to be distinguished by the reference to the facts of the particular case”, and for the rule to be “displaced by even a slight counterweight” such that it would be more accurate to call it the exception (rather than the rule) in Clayton’s Case. Phang JC suggested at [56] in Q & M Enterprises that the rule could be an “evidential presumption and no more”.

It also bears mention that the rule in Clayton’s Case would be displaced when there are mixed funds made up of contributions from a beneficiary and a fiduciary, as the operative presumption would be the one that works best in favour of the claimant and against the fiduciary (Graham Virgo, Principles of the Law of Restitution (Oxford University Press, 3rd Ed, 2015) at 622, citing Re Hallett’s Estate (1880) 13 Ch D 696 and Re Oatway [1903] 2 Ch 356). Finally, whether Clayton’s Case is even applicable in the context of tracing remains unresolved (for differing views, see Barlow Clowes at 44; Q & M Enterprises at [56]; Greymac at 684; O’Connor Rosamund Monica v Potter Derek John [2011] 3 SLR 294 at [47]).

The cases therefore suggest that save for exceptional cases, the rule in Clayton’s Case is not the appropriate method to be applied to the resolution of claims made by multiple claimants with security interests in a commingled fund, and that an alternative method which produces a more just outcome should be applied.

The pari passu and rolling charge methods

The second and third methods – ie, the pari passu method (or the pari passu ex post facto method) and the rolling charge method (also known as the “North American” method) – are related, and it would be convenient to discuss them together. Although the parties have used the expressions “single-block approach” and “multi-block approach” respectively, I will nevertheless refer to the “pari passu method” and “rolling charge method” in line with the body of decided cases.

The pari passu method involves the pari passu sharing of the total pool of assets according to what each of the claimants are owed, ignoring the dates on which they have made their respective investment or contribution (Barlow Clowes at 36). The rolling charge method treats the commingled fund as a blend or cocktail of credits made at different times and from different sources with the result that “when a withdrawal is made from the account it is treated as a withdrawal in the same proportions as the different interests in the account (here of the investors) bear to each other at the moment before the withdrawal is made” (Barlow Clowes at 35). Both methods are similar in approach in that calculations are done on a pari passu basis. However, the rolling charge method requires that the contributor’s rateable interest in the mixed fund vis-à-vis the other contributors be recalculated at every instance of withdrawal, whereas the pari passu method only requires that all available assets be divided on a pari passu basis at the point of distribution.

Given that the rolling charge method takes into account the rateable interests of each contributor to the mixed fund immediately before any withdrawal, it is more precise and is deemed to produce “the most just result” (Barlow Clowes at 35). An application of the pari passu method may potentially be unfair to the most recent contributors as they may have their interests in the fund diminished by withdrawals prior to their contribution (Boughner v Greyhawk Equity Partners Limited Partnership (Millenium) 111 OR (3d) 700 (“Boughner v Greyhawk”) at [92]; see generally Shalson v Russo [2005] Ch 281 at [149]–[150]). The rolling charge obviates this potential unfairness as it favours neither the earlier nor the more recent contributors (whereas the “first in, first out” method favours the latter) at the expense of the other group (Lynton Tucker et al, Lewin on Trusts (Sweet & Maxwell, 19th Ed, 2015) at para 41-071; Underhill and Hayton: Law Relating to Trusts and Trustees (David Hayton gen ed) (LexisNexis, 19th Ed, 2016) at paras 90.34–90.35).

The choice between the pari passu and the rolling charge methods

In my view, the rolling charge method should be preferred unless it is impracticable or unworkable (Barlow Clowes at 42; Boughner v Greyhawk at [84]). In Boughner v Greyhawk, the Ontario Supreme Court of Justice was concerned with the distribution of commingled funds to 24 victims of a fraudulent investment scheme. On the facts, the receiver had determined that it was able to make the accurate calculations to distribute the funds by using the rolling charge method. The court ordered the funds to be distributed on that basis as “justice dictates that the funds should be distributed proportionately based on the interests of the parties at the time of commingling” (at [91]–[93]). In rejecting the pro rata ex post facto distribution method (or what is essentially the pari passu method), the court commented (at [92]) that this method would not have resulted in a fair outcome as the earlier investors had already lost over 88% of their investment value by the time the more recent investors came on board.

Even in respect of cases where the pari passu method is applied, it has been acknowledged that the rolling charge method would have been adopted had it been practicable to do so. In Barlow Clowes, following the collapse of a Gibraltar deposit-taking company (“BCI”) and related entities, 11,000 investors were owed monies in excess of £100m, and the English Court of Appeal was confronted with the task of distributing monies and assets that were available for distribution. In deciding on the pari passu method, Woolf LJ observed (at 42) that “[i]f the North American solution is practical this would probably have advantages over the pari passu solution”; but on the facts, the complications of applying the rolling charge method meant that the pari passu method had to be used. Leggatt LJ also accepted that the fairness of the rolling charge method was “obvious” but the required calculations in Barlow Clowes proved too difficult and expensive (at 44).

In Charity Commission, an unincorporated charitable trust operated a website that facilitated charitable giving by members of the public by inviting them to make donations to the trust for the benefit of charities of their choice. An inquiry into the trust revealed a substantial shortfall between the funds held by the trust and the amounts due unpaid to the designated charities. One of the determinations that the Charity Commission sought from the court was a direction as to the most appropriate method of distributing the funds. Henderson J applied the pari passu method even though he had earlier commented, citing Barlow Clowes, that the rolling charge method was likeliest to produce a fair and just result and he was prepared to proceed on the assumption that it was open to him to adopt the rolling charge method if he were satisfied that it provided a fairer result than either “first in, first out” or pari passu distribution (at [54]). In rejecting the rolling charge method, Henderson J recognised that the problem faced was that of impracticability in the reconstruction of the raw data which would be needed in the absence of any adequate computerised record keeping by the unincorporated charitable trust (at [55]).

In my judgment, the rolling charge method is the method that would likely produce a fairer and more equitable result compared to the pari passu method, all other things equal. It should therefore be preferred, unless it is impracticable for whatever reason to adopt and subject to considerations such as the intention of the parties as regards methods of distribution (including an agreement to a particular method of distribution). I turn to discuss the significance of parties’ intentions.

The parties’ intentions

A key consideration relevant for the court’s determination on the appropriate method to adopt is the parties’ intentions, whether express, inferred or presumed. Where there is an express agreement as to the method of distribution, such agreement would be given effect to unless it was unworkable or impracticable to do so. In the absence of such agreement, the court will look at the parties’ presumed intentions, which are to be gleaned from the facts of the case.

In Barlow Clowes, Woolf LJ relied on the parties’ presumed intentions to reject the “first in, first out” method, and to adopt the pari passu method instead. He found that contributions were intended to be made in respect of a “collective investment scheme”; and even if BCI (the deposit-taking company) was under an obligation to create separate funds for each investor, the presumed intentions of the investors would have been that what could be salvaged as a result of the “common misfortune” should be dealt with in accordance with the pari passu approach, instead of being subjected to the vagaries of chance that follow from applying a “first in, first out” principle (at 41). Leggatt LJ echoed this analysis in his judgment, emphasising that “[t]he court goes by what must be presumed to have been the intention of the investors” (Barlow Clowes at 46).

The principle to be gleaned from Re IIUT and Barlow Clowes is that the parties’ intentions are a key consideration to bear in mind when selecting the appropriate method of distribution. To ascertain what the parties’ intentions are, the terms and structure of the contribution or investment will be indicative, though all the circumstances of the case should be taken into account.

Summary of the applicable principles

To summarise, the principles to be applied in a situation where there are multiple claimants to a pool of commingled assets that is insufficient to satisfy every claim are as follows:

(a) The scope of application for the “first in, first out” rule in Clayton’s Case is limited and can be displaced by a slight counterweight. The rule in Clayton’s Case should not be applied if such application would be impracticable or unjust and there is a preferable alternative or if the application of the rule would be contrary to the intention of the claimants.

(b) In most cases, a fairer and more equitable method can be employed, being the pari passu or rolling charge method. Where practicable and subject to considerations such as the intention of the parties or any agreement to a particular method of distribution, the rolling charge method is to be preferred to the pari passu method as the former more accurately reflects the parties’ interests in the mixed fund. In the usual case, the rolling charge method produces a result that is more just as the pari passu method might occasion unfairness to the most recent contributors to the mixed fund. There may, however, be situations in which the pari passu approach is more suitable – for instance, where it would be too complicated or costly to apply a rolling charge approach because of a prohibitively large number of claimants or transactions.

(c) The parties’ intentions are an important overarching consideration. Where there is an express agreement as to the method of distribution, such agreement would be given effect to unless it was unworkable or impracticable to do so. In the absence of such agreement, the court will look at the parties’ presumed intentions. To ascertain what the parties’ intentions are, the terms and structure of the contribution or investment will be indicative, though all the circumstances of the case must be looked at in the round.

Application to the facts

The parties confirmed that they were not advocating for the “first in, first out” method in Clayton’s Case to be applied, so the only pertinent question is whether it is practicable to apply the rolling charge method or the pari passu method.

In my view, the arguments raised by the fifth and sixth defendants against adopting the rolling charge method can be condensed into three points:

(a) First, the rolling charge method does not take into account the fact that new incoming stock would likely be packed for outgoing shipments first since they are stacked on top of existing stock.

(b) Second, the rolling charge method cannot be used because the chronological order of some same-day entries has not been established, and it is therefore impossible to establish an accurate sequence of transfers for three out of four of the Disputed Categories (Categories 11, 13 and 17).

(c) Third, the pari passu method should be used as it is a just and equitable method of distributing funds between victims of a common misfortune.

The first point requires me to elaborate briefly on the Company’s warehousing system as the stock of pepper was stored in its warehouse. This warehouse was managed by an external company, Sea-shore Transportation Pte Ltd (“Sea-shore”), which provided warehousing, storage, packing and handling services to the Company. According to the evidence of a Sea-shore warehouseman working in that warehouse, one Mr Shatik, there was no system as to how incoming stock was stored. Stock of the same description was stored together in various unmarked locations within the warehouse (hence, the bags of pepper under the four Disputed Categories should be handled as four distinct mixed bulks). At each location, there would be one or multiple stacks of stock of similar description. New stacks were created whenever existing stacks became unstable, and stacks were rearranged whenever the space became too cluttered. There was also no system as to how outgoing shipments were processed. Upon receiving an order, the personnel manning the warehouse would fulfil the order by retrieving the required stock from the relevant location, and then arranging for delivery out of the Company’s warehouse. There was no “first in, first out” or “last in, first out” method used to retrieve stock for outgoing shipments. All said, arrangements for incoming and outgoing shipments were “completely random” and how the bags were stacked was always changing as Mr Shatik had to figure out how to make better use of the space in the Company’s warehouse.

The fifth defendant challenged Mr Shatik’s evidence. One Mr Sharaf, the former Managing Director of Sea-shore, deposed in his affidavit that incoming stock will always be stacked on top of existing stock and stock will always be removed from the top of the stack to meet outgoing orders (suggesting that a “last in, first out” method is always applied). Mr Sharaf added that there was no need to rearrange stacks in the warehouse as that would be a waste of time and manpower. The fifth defendant therefore submitted that the rolling charge method ignores practical reality as a “later Lender who had financed stock at a later date would have a greater risk of having its stock taken out from the Warehouse...”

I reject the factual premise of the fifth defendant’s submission as I prefer Mr Shatik’s evidence to Mr Sharaf’s. Mr Shatik was the warehouseman on the ground and he would have a better understanding and recollection of how the warehouse operated. Mr Shatik’s evidence also seemed to be the likelier version because pictures of the warehouse showed numerous stacks – many of which were not very tall and suggested that new stacks had to be made continuously and existing stacks had to be rearranged. In any case, the only point of contention is whether the pari passu or rolling charge method should be applied. The fifth defendant’s submission is neither here nor there because any suggestion that the warehouse dealt with shipments in and out of the warehouse on a “last in, first out” basis not only militates against the adoption of the rolling charge method, but the pari passu method as well.

I turn to the fifth and sixth defendants’ second point: the evidential uncertainty regarding the sequence of same-day entries in the warehouse ledger. On a preliminary note, the parties agreed that there was no objection in principle to the adoption of different methods of distribution for each of the Disputed Categories, each category being a distinct mixed bulk on its own. This is relevant here because Category 15 did not face the same problem as regards the sequence of transfers (although ultimately, this did not matter). I will elaborate.

To track incoming and outgoing shipments of stock to and from the Company’s warehouse, handwritten warehouse ledgers are kept. Each ledger contains details relating to the date of transfer, description of stock, quantity of stock transferred (for pepper, this is expressed in number of bags), as well as the amount of balance stock left. The financing bank for each incoming shipment can also be identified by looking at the relevant trust receipts.

For Categories 11, 13 and 17, there were same-day entries detailing transfers of stock in and out of the warehouse without any evidence or indication that they were recorded in the right chronological sequence. To illustrate using an excerpt of the entries recorded between 25 September 2015 and 28 September 2015 for pepper bags under Category 11:

In respect of the three entries on 26 September 2015, the fifth and sixth defendants argued that the exact chronology of transfers in and out of the Company’s warehouse had not been established. If the incoming shipment of 1,000 bags on 26 September 2015 came after the two outgoing shipments of 500 bags each, the results obtained by using the rolling charge method would differ. This problem arose for multiple days across the three categories mentioned. Doubts about the sequence of the transfers seemed especially serious since Mr Shatik had deposed in his affidavit that he does not immediately record a delivery every time the stock arrives or leaves the warehouse because of time constraints, and it was not uncommon to update the ledger only after a few days.

After I heard the parties on 27 October 2017, I directed that Mr Shatik file a further affidavit to explain whether records for transactions within the same day accurately reflected the chronological order in which the transaction took place. All of these difficulties were then addressed in Mr Shatik’s second affidavit dated 9 November 2017, wherein he confirmed that the order in which he has recorded the entries was the order in which the incoming and outgoing shipments were made. He was confident about the order of the transactions after reviewing the ledgers, and also deposed in his affidavit that he was able to distinctly remember the movement of stock in and out of the Company’s warehouse as he was constantly being asked to check on the status of the deliveries at or around the time when the entries were recorded. Given the foregoing, I did not see the second point as an insuperable difficulty making it impracticable or unrealistic to apply the rolling charge method.

Indeed, in Re IIUT, Williams J observed at [25] that there would be difficulties and inequities regardless of the method chosen. Hence, even if there were minor inaccuracies in the ledger entries in the present case, such inaccuracies would not have tilted the balance in favour of the pari passu method. Unlike Barlow Clowes where there were 11,000 investors to distribute to, and Charity Commission where the required raw data could not even be reconstructed, the parties in the present case shared considerable common ground and were even able to agree on a set of figures to be paid out depending on the method used. It could scarcely be said that it would be impracticable to apply the rolling block method in the present case.

I turn now to the third and final point raised by the fifth and sixth defendants: that the pari passu method would be the most just and equitable way to distribute the sale proceeds amongst the financing banks which were victims of a common misfortune. The fifth defendant submits that this is borne out by the fact that the terms of the trust receipts did not clearly require segregation of stock and that the defendants’ ignorance about the warehousing system meant that they were “indifferent to whether the self-same bags the subject of the original bills of lading they received were delivered out of the warehouse against their own trust receipt”, ie, the bags of pepper were treated as fungible.

In my view, the choice between the pari passu and rolling charge methods cannot simply be resolved by referencing Woolf LJ’s holding that the presumed intentions of victims to a “common misfortune” would be to deal with what is salvaged on a pari passu basis. The competing method he considered was the “first in, first out” principle, and his holding was directed at the “vagaries of chance” that the “first in, first out” principle subject claimants to. It should be highlighted that the court in Barlow Clowes seemed ready to implement the rolling charge method, had it been practicable for them to do so. In the present context, where the two competing methods are the pari passu and rolling charge methods, Re IIUT provides a more helpful comparison. The present case is very different from a situation where the claimants are investors to a joint investment fund like the IIUT. Contrary to the submissions of the fifth defendant, the financing banks expressly tried to set themselves apart from each other. In Pars Ram, Chong J noted that the second defendant’s and the fifth defendant’s trust receipts contained an obligation to hold and store the goods in the bank’s names or to pay the proceeds of sale of the underlying goods into a designated account. Moreover, the sixth defendant’s trust receipt contained an express term stating that the Company should hold or store the goods in a manner capable of separate identification (Pars Ram at [2(b)]). Even though these terms may not have been complied with, the Company’s breach is irrelevant for ascertaining the financing banks’ intention, and the terms used in the trust receipts point towards a finding that the creditors did not intend to weather this “common misfortune” on a pari passu basis.

Conclusion

Audrey Lim

Judicial Commissioner

A.1 The following table sets out the four Disputed Categories and how the sale proceeds would be split between the defendants and the general creditors under the single-block (pari passu) and multi-block (rolling charge) approaches (to the nearest dollar):

Category 11

Category 13

Category 15

Category 17