Arbitration and Insolvency: Clash of Cultures?

The legislative provisions in the Insolvency Act 1986 addressing transactions entered into at an undervalue by a debtor form part of the armoury provided to an office holder to adjust/avoid transactions made in the twilight period prior to the debtor’s entry into an insolvency proceeding. The origin of these legislative provisions can be traced back to the Statute of Elizabeth.¹ They are aimed at debtor misbehaviour and reversing any transaction that had the effect of depleting the value of the estate at the expense of the debtor’s general body of creditors.  As noted in the Cork Report: “The justification for setting aside a disposition of the bankrupt’s assets made shortly before his bankruptcy is that, by depleting his estate, it unfairly prejudices his creditor.”²  The Singapore insolvency regime has similar legislative provisions in the Insolvency Restructuring and Dissolution Act 2018 (“IRDA”) (which were previously contained in the Companies Act, before Singapore consolidated all of its personal and corporate insolvency and restructuring laws into the IRDA) to address transactions entered into at an undervalue by a debtor.

There have been a number of cases recently, particularly in the English courts (but also in the ADGM – see in particular NMC Healthcare LTD and associated companies [2021] ADGMCFI 0006), where judges have held that arbitration clauses have force in insolvency. Whether this is right in a particular case will depend upon two questions:

• Does the relevant dispute fall within the wording of the arbitration clause?
• Is the dispute arbitrable?

Question 1 is unlikely to detain a court for very long. It is trite law that arbitration clauses should be widely and generously construed (Fiona Trust v Privalov [2008] 1 Lloyds Rep 254). As Sir Andrew Smith put it in NMC at [79]:

“Since the decision of the House of Lords in Fiona Trust & Holding Corp v Privalov, [2007] UKHL 40, “the starting point for interpreting an arbitration agreement and determining its scope is not to focus on “fussy distinctions” about the exact terms used, but to construe it liberally, recognising that generally rational businessmen entering into an arbitration agreement will intend that any dispute arising out of their relationship should be resolved by the same tribunal: see esp. at para 13 per Lord Hoffmann and at paras 26 and 27 per Lord Hope.”

A similar approach is taken in Singapore. As stated by the Singapore Court of Appeal (“CA”) in the case of Larsen Oil and Gas Pte Ltd v. Petropod Ltd [2011] SGCA 21 (“Larsen”) at [19]:

“There are, all in all, strong reasons for supporting a generous approach towards the construction of the scope of arbitration clauses, given that such an approach has received widespread acceptance among the leading commercial jurisdictions, and is strongly supported by the academic community. Such an approach is also consistent with this court’s philosophy of facilitating arbitration (see, for instance, the case of Tjong Very Sumito v Antig Investments Pte Ltd [2009] 4 SLR(R) 732 where we adopted a generous interpretation of the word “dispute” in an arbitration clause). Accordingly, we agree that the preponderance of authority favours the view that arbitration clauses should be generously construed such that all manner of claims, whether common law or statutory, should be regarded as falling within their scope unless there is good reason to conclude otherwise.”

There is room, however, even here for a question in the context of transaction at an undervalue claims. Such claims are made by an office holder, arising from a statutory cause of action which comes into play for the first time after the insolvency has commenced. It allows for the swelling of the debtor’s assets which is to be utilised by the office holder in accordance with the statutory regime provided for in the Insolvency Act 1986 – this includes payment out of expenses in the relevant insolvency process. Can this really be said to fall into the normal wording of an arbitration clause, a clause that binds the company and the contracting party inter se? There must be room for argument here. Indeed, it is this argument that succeeded in Singapore, in Larsen.

In Larsen, the Singapore CA drew a line between private remedial claims (either common law or statutory) and claims that can only be made by a liquidator or judicial manager of an insolvent company, and held that arbitration clauses should not ordinarily be construed to cover avoidance claims in the absence of express language to the contrary. Since avoidance claims can only be pursued by a liquidator or judicial manager of an insolvent company, the Singapore CA considered that there is no reason objectively to believe that a company’s pre-insolvency management would ordinarily contemplate including avoidance claims within the scope of an arbitration agreement: see Larsen at [20].

Question 2, will always, however, be the real focus of any argument in an insolvency context. Is the relevant issue arbitrable? As noted in Russell on Arbitration (24th Ed, 2015), para 2-080 the concept of arbitrability depends upon whether a matter is “capable of being submitted to arbitration”. So far, so good, but what determines whether or not this is the case? There are many things that will make a dispute non-arbitrable. Should it not be the case that insolvency is, quite simply, a wholly new event that takes the position out of the norm, leaving matters to be dealt with in the insolvency rather than outside. After all, this is the general approach taken to inward claims, which should normally be dealt with in the proof of debt process, rather than through litigation, unless there are particular reasons to the contrary: see for example the statement by Patten J in A.E.S. Barry Ltd. v TXU Europe Energy Trading (In Administration) [2004] EWHC 1757 (Ch) at [24].

However, this insolvency-centred approach has not garnered extensive approval. As noted by Gary B. Born in International Commercial Arbitration at p.1084, para 6.40(F):

“Parties to international arbitration agreements sometimes become subject to some form of bankruptcy or insolvency, either in their home jurisdiction or elsewhere. In most jurisdictions, only national courts (often specialised courts) have authority to commence, administer and wind-up bankruptcy proceedings, including proceedings to liquidate a bankrupt company, reschedule its liabilities, operate it under some form of receivership or administration, or distribute pro rata to designated creditors and owners. Disputes concerning these “core” bankruptcy functions are almost universally considered nonarbitrable, whether in domestic or international arbitrations, under the laws of developed jurisdictions.

It is much more controversial, however, whether and when disputes merely involving a bankrupt entity as a party or raising questions of bankruptcy law (e.g. the continued effect of a contract), may be resolved in arbitration. Different national legislative regimes and judicial decisions have reached different conclusions about these types of disputes. In many such cases, the desirability of a centralised forum for resolving all disputes involving the bankrupt entity is weighed against that entity’s preexisting commitment to resolve disputes with a contractual counterparty by international arbitration, with different legal systems adopting different resolutions of these competing interests. Again, however, the weight of authority, particularly in recent years, supports narrow nonarbitrability rules in this context.”

So, if it is not the case that it can be said that all insolvency matters are not arbitrable, how do we know whether something is arbitrable or not? For examples of such instances, we can again refer to Russell, at para 2-081:

“In particular, a dispute will generally not be arbitrable if it involves an issue of public policy, public rights or the interests of third parties, or where the dispute in question is clearly covered by a statutory provision which provides inalienable access to the courts”.

This suggests that there is potential for excluding transaction at an undervalue claims, which fall into at least three of these categories:

• They bring into play the public policy question of whether matters should be dealt with in an insolvency context.
• They affect more than just 2 parties (which also feeds into the public policy argument).
• They are clearly covered by a statutory provision which provides for a remedy to be given by the “court” (which, by definition, should be the court with governance of the insolvency process).

Taking all three arguments into account, prima facie that would suggest that there are likely to be very real arguments that transaction at an undervalue claims are not arbitrable.

This was in fact the approach taken by the Singapore courts. In Larsen, the Singapore CA considered that a distinction should be drawn between disputes involving an insolvent company that stem from its pre-insolvency rights and obligations, and those that arise only upon the onset of insolvency due to the operation of the insolvency regime. The objective of the avoidance provisions, which are to recoup for the benefit of the company’s creditors losses caused by the misfeasance and/or malfeasance of its former management, could be compromised if a company’s pre-insolvency management had the ability to restrict the avenues by which the company’s creditors could enforce the very statutory remedies which were meant to protect them against the company’s management. The Singapore CA held that such objective of the insolvency regime should thus override the freedom of the company’s pre-insolvency management to choose the forum where such disputes are to be heard, and to treat disputes arising from the operation of the statutory provisions of the insolvency regime as non-arbitrable, even if the parties expressly included them within the scope of the arbitration agreement.

Significantly, the Singapore CA did not say that the mere fact that one party was insolvent would render any claims non-arbitrable. A clear distinction was drawn by the Singapore CA between claims that arose only upon the onset of insolvency, and disputes that stemmed from pre-insolvency rights and obligations. The Singapore CA accepted that allowing a creditor to arbitrate the latter does not undermine the insolvency regime’s underlying policy aims. Indeed, the Singapore CA was careful to note that a claim under section 73B of the Conveyancing and Law of Property Act for a claim of fraudulent conveyance of property (which has since been repealed and appears in a different form in ss 438 and 439 of IRDA) is one that may straddle both a company’s pre-insolvency state of affairs, as well as its descent into the insolvency regime.

In contrast to the clear position in Singapore, decisions in this area in England do not take quite such an insolvency-friendly approach, particularly in the context of Russian bank insolvencies (which involve the appointment of temporary administrators). One such case that has caused quite a bit of recent discussion is the case of Riverrock Securities Limited v. International Bank of St Petersburg (Joint Stock Company) [2020] EWHC 2483 (Comm) (“Riverrock”), heard just over a year ago in the English High Court where the judge (Foxton J) held that principles of insolvency law does not bar the arbitration of an insolvency claim such as transaction avoidance.

In this regard, the underlying international insolvency policy he identified was that of “modified universalism” concerning the effect to be given to a foreign insolvency (at [80]).  He did not consider this policy was infringed by the arbitration of the actions before him.  He said at [81]:

“However, enforcing the LCIA Arbitration Agreements would not in any way conflict with the principle (or frustrate the policy) of modified universalism. Granting an injunction would not involve recognising a second bankruptcy on the part of IBSP, nor prevent there being a single system of distribution. Any recoveries made by the DIA on IBSP’s behalf in an LCIA arbitration would be subject to, and administered in accordance with, the single scheme for distribution constituted by the St Petersburg bankruptcy proceedings” (Emphasis added)

As we will come back to below, Riverrock may be right on its own facts, in the sense that this was a claim brought by the Bank after the end of the insolvency and therefore there are good grounds for the Court to have reached the conclusion that the relevant dispute was arbitrable. Leaving that point to one side, and looking at the principles applied by Foxton J, it seems to us, with the greatest of respect to the Judge, that he has been led into error in understanding the concept of modified universalism. It is not just about the distribution mechanism, but also about the moratorium. In other words, it is about dealing with the disputes arising in relation to claims into and out of the insolvency in the jurisdiction where proceedings are opened. Had the concept of modified universalism been better understood, it might not have led to a different conclusion, but would have led to a less concerning one.

To understand Riverrock, one needs to start two years earlier, with the case of Nori Holding Limited v. Public Joint Stock Co Bank Otkritie Financial Corporation [2018] EWHC 1343 (Comm) (Males J) (“Nori”). In Nori there were two parallel causes of action in respect of a pre-insolvency transaction: a claim under Russian insolvency law to set aside a transaction for unequal consideration, and a claim under the Russian civil code for an abuse of rights (at [19]-[20]). Prior to the hearing of the application in Nori, there had been a temporary bank administration in Russia in relation to the bank. This had ended. Although both actions had been commenced by the temporary administrator, both were continued by the bank.

The bank submitted that no anti-suit injunction should be granted because the insolvency causes of action was not arbitrable (at [30](2) and [43]-[47]). The bank relied on Larsen in the Singapore CA (which distinguished between the rights that affected all creditors, and the rights that were just bilateral). Males J said that it was unclear whether this distinction was a general rule or just a procedural point for Singaporean courts (although readers of this article may take the view that this distinction is actually a fundamental principle of insolvency and, indeed, arbitration).

This submission was rejected by Males J who relied on the reasoning of the Court of Appeal in case of Fulham Football Club (1987) Ltd v Richards [2011] EWCA Civ 855 (“Fulham”), which held that an unfair prejudice claim was arbitrable. Males J summarised Fulham at [57]-[59]:

“Patten LJ’s conclusion was that in a case where the relief sought was for an order which the arbitrators had power to make and the dispute was essentially contractual, there was no reason why the dispute should not be arbitrated, but that even where an order (such as a winding up order) was sought which arbitrators had no power to make, they could legitimately decide whether there was unfair prejudice and winding up proceedings should be brought, which proceedings could then be brought before the court.

58. So far as construction was concerned, Patten LJ held that in the absence of any statutory restriction or rule of public policy preventing the parties from agreeing to submit certain types of claim to arbitration, it was impossible to read into the wide language of the arbitration clause any limitation excluding claims for unfair prejudice from its scope.
59. Longmore LJ dealt with the issues in reverse order. In relation to construction, he held that the wide expressions “all disputes” and “all differences” meant what they said, while there was no express or implied prohibition in the Companies Act 2006 to prevent arbitration of a dispute about unfair prejudice. Nor was there any principle of public policy to such effect. The fact that an arbitrator could not give all the remedies which a court could give did not afford any reason for treating an arbitration agreement as of no effect.” (Emphasis added)

Males J therefore rejected the Larsen “presumption” that insolvency law claims were not within the scope of an arbitration clause (at [60]-[61]).  Of the question of whether the claims before him were arbitrable, he said:

“62. I deal next with whether the parties’ dispute is arbitrable. For this purpose it is irrelevant in my judgment whether the claim is properly characterised as an insolvency claim under Russian law. It is necessary to focus on the nature of the particular claim and to consider whether that claim is capable of being determined in arbitration. In my judgment it plainly is.

63. What matters is the substance rather than the form. In this case the parties’ dispute is a straightforward factual dispute whether the August transactions constitute a fraud carried out on the Bank to replace valuable secured loans with worthless bonds. If so, the Bank will have a claim to avoid those transactions and to require the claimants to reinstate the position in which it was before they were carried out. A variety of legal labels can be and have been attached to that claim, including the labels of transaction with unequal consideration and abuse of rights under Russian law and conspiracy to defraud under Cypriot law. But in each case the essential dispute is the same, regardless of the label. This is a dispute which arbitrators can determine.

64… There is, in this case, no remedy claimed such as a winding up order which would affect the status of the Bank or which would affect the position of third parties in such a manner as to take the case beyond the consensually derived jurisdiction of the arbitrators…”

The conclusion in Nori will ring a false note in insolvency ears. First, it might well be thought that there is a presumption that insolvency law claims are not within the scope of an arbitration clause. (In other words, as suggested above, Larsen is right on this point.) Secondly, recharacterizing the claims as fraud claims does not mean that they are not insolvency law claims, properly so-called. As we will shortly see, in Riverrock, Foxton J thought that these claims were insolvency law claims.

Having said that, if we look at Nori and Larsen in the context of the underlying rights of the parties, the approaches taken can be seen as broadly consistent, because both approaches start with the same fundamental premise of protection of third parties and public policy. Where they diverged is that the Singapore CA in Larsen drew an important distinction between pre-insolvency claims and claims that could only be brought post-insolvency under the insolvency regime, whereas the court in Nori did not.

So much for Nori. What about Riverrock? The fact pattern in Riverrock was materially the same as Nori: see [5], [9] and [12]-[13], as noted by Foxton J at [39].  On the evidence however, Foxton J considered that the claims were those of the bank commenced by its liquidator on its behalf, rather than distinct claims of a liquidator (at [50]-[51]). By the time of this hearing the temporary insolvency proceeding had ended and the claims were being pursued by the bank.  Foxton J did not therefore decide (nor need to decide) whether office-holder actions were arbitrable (at [54] (last sentence). Still less did he decide (or need to decide) whether English transaction at an undervalue claims would be arbitrable – a point that he expressly left open (‘whatever the position might be if [the avoidance claims] were English law insolvency claims’ [87(iii)]).

Like Males J, Foxton J did not consider the fact that certain of the avoidance powers in bankruptcy might not be available to the LCIA tribunal to be relevant to characterisation of the dispute in that case as arbitrable (at [62]-[66]).  The dispute in relation to the transactions was arbitrable, and even if certain causes of action could not be arbitrated that did not mean the dispute was not arbitrable.

Importantly, Foxton J agreed with Patten LJ in the Fulham case at [69]:

“However, it is clear that the issue of arbitrability can involve more than simply ascertaining whether the relief sought engages third party interests in a relevant sense, or seeks an order that “only a court can make”. In Fulham Patten LJ recognised that a claim might be non-arbitrable for a third reason, namely that it “represent[s] an attempt to delegate to the arbitrators what is a matter of public interest which cannot be determined within the limitations of a private contractual process” ([40]). He referred elsewhere in his judgment to relief which seeks a “state intervention in the affairs of a company which only a court can sanction” ([77]. Examples of such intervention were matters which “engaged the rights of creditors” or impinged on a “statutory safeguard imposed for the benefit of third parties”.

Foxton J also, at [71], expressly agreed with the points made by Patten LJ in the Fulham case:

“There is no doubt that many aspects of this regime are immune from interference by the members of the company whether by contract or otherwise. They cannot override the provisions of the 1986 Act which apply on liquidation by agreeing between themselves or with a particular creditor that property which belongs to the company in liquidation should be dealt with other than in accordance with the Act The same must go for the exercise of the liquidator’s powers under sections 238 to 239 of the Insolvency Act 1986 . They involve an exercise of a statutory power to intervene in and set aside transactions with third parties in the context of the insolvency regime. These are rights vested in the liquidator for the benefit of the creditors as a whole and cannot be overridden by a contract entered into by the company prior to its liquidation”

Foxton J’s reasoning moreover differed from that of Males J.  Foxton J was satisfied that the actions were insolvency actions (at [75]).  In his view, the important question he had to consider was whether that characterisation meant he had to give priority to (i) the policy of party autonomy or (ii) the public interest identified by Patten LJ (at [77]). He concluded that party autonomy trumped the public interest. Why? Because the temporary administrations had ended. These were not claims being brought in the insolvency, for the benefit of the estate, but by a single creditor, for its own benefit.

Where are we left by all of this? Is Riverrock wrong? Probably not (although Nori might well be said to have gone too far – perhaps influenced too much by the “strong pro-arbitration policy of English law” – Riverrock at [78]), unless the English courts can be persuaded to go down the more sensible and predictable Larsen route. However, it can convincingly be argued that Riverrock is right, on its facts, namely only where an insolvency process has ended, and the claims that are being brought are not essentially insolvency claims, or else are claims being brought only for the benefit of an individual. In that case the claim should remain arbitrable.

In contrast, where this is a normal transaction at an undervalue claim, brought in extant insolvency proceedings, by an office-holder, and for the benefit of all of the creditors of the insolvent estate, that is something that involves the interests of third parties (and not merely in a distribution sense). It is something that goes to the central question of the nature of the assets of the insolvent estate. It is something that is governed by a separate statutory regime, and there is a public interest in enabling this to be heard in public, in the public interest, rather than in private, as if it were something subject to a bilateral agreement. As a result, either on the basis of Larsen, or on the basis of arbitrability, transaction at an undervalue claims of this sort should be dealt with in insolvency, and not in arbitration.

To sum up, as noted extra-judicially by Quentin Loh J in The Limits of Arbitration (2014) 1 McGill Journal of Dispute Resolution 66 at 74:

“Apart from the inherent difficulties that come with any two-stage or sequential resolution of issues, there are also other concerns that might complicate matters, e.g., the solvency of the company and possible impact on would-be creditors, and the interests of other shareholders who are not party to the arbitration.

…

Instead of trying to stretch arbitration to its breaking point, the more logical step to take might be to accept that arbitration, useful as it may be, has its limitations. … This is not necessarily inconsistent with a pro-arbitration stance; it is merely to acknowledge the consensual nature of arbitration and its consequent inherent limitations.”