ARBITRATION AGREEMENTS AND THE WINDING-UP PROCESS: RECONCILING COMPETING VALUES

Abstract

Courts in a number of jurisdictions have attempted to resolve the relationship between winding-up proceedings and arbitration clauses, but a unified approach is yet to appear. A fundamental disagreement exists between courts which believe that the approach of insolvency law should be applied, and those which prefer to prioritise arbitration law. This article argues that a more principled solution emerges if the problem is understood as one of competing values in which the process of characterisation can offer guidance. This would allow both a more principled approach in individual cases, and a more coherent dialogue between courts which take different approaches to the issue.

I. INTRODUCTION

The promotion and growth of commercial arbitration around the world has inevitably led to tension between arbitration and other areas of law. An important example of this is the relationship between arbitration and the winding-up process in insolvency law. An aspect of this relationship which is increasing in prominence is what courts should do when a defendant in the winding-up process disputes the existence of a debt which is subject to an arbitration clause.¹ Despite courts in a number of jurisdictions attempting to deal with this problem, a unified approach is yet to emerge and it has received little academic attention. Some jurisdictions have followed the English decision in Salford Estates v Altomart (Salford Estates) and have held that in such a situation a winding-up petition should be dismissed if there is a dispute as to whether the debt is owed. Others, however, have found that allowing an arbitration clause to restrict a creditor's right to bring a winding-up petition is contrary to public policy. This article has three aims. First, the article reviews the decisions which have dealt with this issue, and shows that a preference between the values of arbitration law and insolvency law plays a decisive, but to date unrecognised, role in how courts have resolved these cases. Secondly, the article advances an understanding of these cases which emphasises the importance of how courts have framed the problem, and explains the relationship between the framing decision, the values of arbitration and insolvency law, and the approach ultimately adopted. Thirdly, the article contends that fresh insights on the problem can be gained by viewing it as akin to the question of characterisation in private international law. This perspective clarifies that when deciding how to deal with the clash between arbitration law and insolvency law, courts are making a more complex choice between the policies and values which underpin each regime.

Section II explains the problem that courts face with reference to the facts of one of the leading cases on the issue—Salford Estates. Section III then compares the approaches that have been taken by courts which have dealt with this issue, identifying a ‘pro-arbitration’ approach and a ‘pro-insolvency’ approach. In Sections IV and V the article examines the policy values which underpin arbitration and insolvency law and identifies how these values have shaped the manner in which courts have analysed and framed the relationship between arbitration and insolvency law. Section VI then proposes tackling this relationship by viewing the problem as one of characterisation, and advances a new case-by-case approach to resolving the conflict. Section VII concludes.

II. THE PROBLEM STATED

The dispute in Salford Estates concerned an underlease between the claimant and defendant. Under the arrangement, the defendant was to pay a service charge and a share of the cost of the insurance premium on the property. The lease also contained a provision which referred to arbitration ‘[a]ny dispute arising between the Lessor and the Lessee as to their respective rights, duties or obligations or as to any other matter arising out of or in connection with this Underlease’.² The parties referred to arbitration a number of disputes concerning the defendant's obligation to pay the service charge and the insurance premium, resulting in an award which found that the defendant owed the claimant £64,431.79 plus costs. However, the defendant failed to pay immediately the sums due under the award, leading the claimant to present a winding-up petition on the basis that the defendant was unable to pay the debt under the award, as well as other alleged debts which had not been referred to arbitration.³

The issue for the Court was whether, in light of the existence of the arbitration clause in the contract between the parties, the winding-up petition should be stayed so that the disputed debts could be dealt with by arbitration. Doing so would have the effect of giving the tribunal the power to determine the claimant-creditor's claim, rather than requiring the creditor to submit a proof of debt to the liquidator as would be the norm in the insolvency regime.⁴ The first question was whether the mandatory stay provision in section 9 of the Arbitration Act applied. Although the Court at first instance found that section 9 did apply, the Court of Appeal disagreed. The Court of Appeal noted that section 9 referred only to claims and counterclaims, and held that winding-up petitions fell outside the scope of the provision.⁵ The main reason given for this was that, unlike a claim for a debt, the presentation of a winding-up petition does not necessarily mean that the petitioner will be able to recover the value of the debt; the amount recoverable depends both on the debtor's assets and on the claims and ranking of other creditors.⁶

However, this was not the end of the matter. As the Court's power to wind up a company under section 122(1) of the UK Insolvency Act is discretionary, it did not necessarily follow from the Court's finding on section 9 that the defendant had to be wound up.⁷ Ordinarily, a winding-up petition will only be stayed or dismissed if the debtor establishes that the debt is bona fide disputed on substantial grounds and that the petition is therefore an abuse of process.⁸ However, imposing this standard where there is an arbitration agreement risks undermining the bargain struck by the parties as it would mean that a liquidator rather than a tribunal would determine the dispute. The Court of Appeal in Salford Estates therefore had to choose between the solution provided by the insolvency regime, and the solution provided by the arbitration regime.

Underpinning this conflict is a difference in view as to which of the parties’ rights is most relevant to the dispute. The insolvency regime focuses on the right to the payment of a debt, and adopts an expansive understanding of this right. Section 123(1)(a) of the Insolvency Act therefore allows a creditor to bring a winding-up petition in respect of a debt exceeding £750 if a written demand has been served on the creditor and the creditor has failed to pay the sum or secure or compound for it within three weeks. A similar rule can be found in the legislation of the other jurisdictions examined in this article.⁹ At this stage, there is no consideration of the debtor's actual financial position, or the existence or interests of other creditors. The insolvency regime therefore allows a single creditor to put a debtor under significant pressure with relative ease, even where there is no evidence of the debtor actually being in financial distress. The burden is then on the creditor to show that the debt is bona fide disputed on substantive grounds and that the petition is therefore an abuse of process. Put differently, it is for the debtor to justify the departure from a default rule which itself takes an expansive view of the creditor's rights.

In contrast, the focus of the arbitration regime is the prior issue of how the parties have agreed to settle their disputes. When dealing with the arbitration regime, courts are therefore generally unconcerned with the substantive rights of the parties, and only with ensuring that the dispute is heard and resolved in the appropriate forum. To ensure this, arbitration law only requires a debtor to show the existence of an arbitration agreement and a genuine dispute; assessing whether the dispute was ‘substantive’ would itself undermine the arbitration agreement. The court's task in cases like Salford Estates is therefore to determine which rights are most relevant to the dispute.

Although Salford Estates concerned a winding-up petition, the existence of a discretion in other aspects of insolvency law has resulted in the same tension between insolvency and arbitration law being found in other forms of proceedings, such as the setting aside of statutory demands (see further below). Courts have taken two conflicting ways to resolving this conflict.

III. RESOLVING THE CONFLICT

A. The Pro-Arbitration Approach

The first, and dominant, approach is to favour the arbitration regime. This approach was followed in Salford Estates itself. The Court of Appeal expressed concern that allowing the petition in such circumstances would risk encouraging parties to an arbitration agreement to present winding-up petitions as a way of by-passing the agreement and pressurising the alleged debtor to pay the disputed sum immediately.¹⁰ The Court therefore concluded that except in ‘wholly exceptional circumstances’ it would exercise its discretion ‘consistently with the legislative policy embodied in the 1996 Act’ and would not consider whether a debt was bona fide disputed on substantial grounds.¹¹ That the defendant disputed the debt was itself enough. The decision has since been consistently followed in the English courts, with later judgments emphasising the narrowness of the ‘exceptional circumstances’ exception. In Fieldfisher LLP v Pennyfeathers Ltd the Court went as far as to say that even previous admissions of the disputed debt did not constitute ‘wholly exceptional circumstances’, and that the language of the judgment in Salford Estates indicates that the Court of Appeal in that case did not envisage that any circumstances would be sufficiently ‘exceptional’.¹² Unsurprisingly, no English court has yet found circumstances so ‘exceptional’ so as to justify departing from the general approach in Salford Estates.

A similar approach has been taken by the Singaporean courts. The dispute in BDG v BDH concerned a winding-up petition presented in relation to a debt due under a contract for the supply of drilling units for fossil fuel production. The contract contained a tiered dispute resolution clause culminating in arbitration. Noting the similarity between section 254 of the Singapore Companies Act 2006 and section 122 of the UK Insolvency Act, the Singapore High Court held that Salford Estates should generally be followed and so that the debtor only had to show the existence of a prima facie dispute and did not have to meet the higher ‘triable issues’ standard.¹³ Although the language of ‘bona fide’ was used, this was subsequently interpreted by the Singapore Court of Appeal in AnAn Group v VTB Bank to be a reference to the ‘genuineness’ of the debtor's dispute and not as a suggestion that the court should consider the merits of the argument.¹⁴

The pro-arbitration approach has also been followed in some decisions of the courts of the British Virgin Islands (BVI). Rangecroft v Lenox International concerned an application to set aside a statutory demand. Following the debtor's failure to repay a loan, the creditor issued a statutory demand and sought the appointment of a liquidator. The debtor sought to have the statutory demand set aside due to the existence of an arbitration agreement in the loan contract. Under section 157 of the BVI Insolvency Act a statutory demand can be set aside if, inter alia, there is a ‘substantial dispute’ as to whether the debt is owing or is due or if substantial injustice would otherwise be caused due to a defect in the demand or for some other reason. The BVI High Court acknowledged that in other judgments the BVI courts have declined to follow Salford Estates (see further below), but nevertheless held that it was appropriate to do so as it was ‘obliged to apply’ the mandatory stay provision of the Arbitration Act.¹⁵ The result was that the Court exercised its discretion under the Insolvency Act in a way that gave priority to arbitration law. The High Court in A Creditor v Anonymous Company Ltd took the same approach in dealing with the appointment of a liquidator. The Court again acknowledged that the BVI has not generally followed Salford Estates, but nevertheless held that asking whether the debt was honestly disputed on substantial or reasonable grounds (the ‘Sparkasse Bregenze test’) would undermine the policy of the Arbitration Act where there were no other creditors.¹⁶

A slightly different approach has become dominant in Hong Kong. The issue was first considered by the Hong Kong Court of First Instance in Lasmos Ltd v Southwest Pacific Bauxite (Lasmos). The underlying dispute in Lasmos concerned an alleged debt under a management services agreement. Following non-payment of the alleged debt, the creditor issued a statutory demand and later a winding-up petition. The debtor disputed the existence of the debt and sought to have the winding-up petition struck out, pointing to the existence of an arbitration clause in the contract between the parties. After reviewing the existing Hong Kong authorities, which followed the traditional requirement of a bona fide dispute on substantial grounds,¹⁷ the Court considered decisions in other jurisdictions such as that in Salford Estates, BDG v BDH and Jinpeng Group v Peak Hotels and Resorts (see further below). The Court ultimately held that where a winding-up petition is based on a debt which is subject to an arbitration clause, the petition should generally be dismissed if the debtor disputes the debt and the debtor has taken ‘the steps required under the arbitration clause to commence the contractually mandated dispute resolution process’.¹⁸ In other words, the Court in Lasmos followed the approach in Salford Estates but added the condition that the debtor must have taken the steps necessary to commence the arbitration process. The effect of the decision is that a different standard will apply depending on whether or not the debtor has taken steps to commence arbitration proceedings. This approach still fits within the broad ‘pro-arbitration’ approach discussed above, but the existence of arbitration proceedings rather than merely an arbitration agreement is taken to be the normatively significant factor. The Lasmos criteria have been applied subsequently by the Hong Kong courts,¹⁹ albeit not always enthusiastically as will be seen below.

More recently, the Hong Kong Court of Appeal has shown a willingness to move even further towards the Salford Estates approach. Although dealing with the relationship between a winding-up petition and an exclusive jurisdiction clause, the Court of Appeal in Guy Kwok-Hung Lam v Tor Asia Credit Master Fund considered cases involving arbitration clauses at great length, and relied on such cases to support its conclusion that a winding-up petition should be dismissed if the dispute is the subject of an exclusive jurisdiction clause and there are no ‘strong reasons’ to the contrary.²⁰ Notably, G Lam JA (with whom Barma JA agreed) expressly disagreed with the reasoning of some of the Hong Kong cases which favour a ‘pro-insolvency’ approach.²¹ Whilst Chow JA in his concurring judgment sought to argue that the treatment of arbitration clauses and exclusive jurisdiction agreements should be considered separately,²² the majority's judgment seems to indicate that the two forms of dispute resolution agreements should be considered together (and therefore that the Court would have taken the same approach had the case concerned an arbitration agreement).

It is notable that, in sharp contrast to the clear pro-arbitration approach in these cases, courts in some of these jurisdictions have often taken the opposite approach when considering the interaction between a winding-up petition and an exclusive jurisdiction clause. Despite the similarity between arbitration clauses and exclusive jurisdiction agreements, courts in these jurisdictions have tended to find that the existence of an exclusive jurisdiction clause does not affect the applicable standard when determining whether to stay or dismiss an insolvency application. In Citigate Dewe Rogerson Ltd v Artaban the English High Court stated that any argument that the presentation of a statutory demand was an abuse of process would only succeed if it could satisfy ‘the tradition, and well-established, ground that there is a bona fide and substantial dispute’.²³ Despite being decided before Salford Estates, the Court in Salford Estates did not refer to this decision. Likewise, the BVI High Court in Integrated Whale Media Investments Inc v Highlander Management LLC held that a debtor seeking to have a statutory demand set aside must show that it is disputing the debt on ‘genuine and substantial grounds’ despite the existence of an exclusive jurisdiction clause in the share purchase agreement in favour of the courts of Delaware.²⁴

The reasons for this differential treatment have not received any meaningful judicial or academic attention, and may be the subject of productive future discussion. Such discussion is, however, beyond the scope of this article, as the differences in the treatment of arbitration agreements and exclusive jurisdiction agreements go well beyond the winding-up process. For present purposes, it suffices to say that judicial consideration of any fundamental differences between these two types of agreement would help to rationalise both their inconsistent treatment in some jurisdictions and provide for a more informed debate as to how each interacts with winding-up petitions.

B. The Pro-Insolvency Approach

In contrast, in a number of other decisions courts have rejected the argument that the existence of an arbitration agreement or the commencement of arbitral proceedings should affect the relevant standard when determining to stay or dismiss an insolvency application, and have applied the traditional standard of a ‘bona fide dispute on substantial grounds’. This approach is perhaps most clearly seen in the courts of the BVI, notwithstanding the cases discussed above. The first case to consider the issue after the decision in Salford Estates was C-Mobile Services Ltd v Huawei Technologies Co. Ltd (C-Mobile) which concerned the appointment of liquidators over an alleged debtor. In a relatively brief judgment, the Court agreed with the reasoning of the English Court of Appeal in Salford Estates insofar as it related to a mandatory stay, but declined to follow the Salford Estates approach regarding the exercise of discretion. However the Court's reasoning is primarily based on the fact that the debtor had previously only argued that the there was a bona fide dispute on substantial grounds and had not asked the Court to exercise its discretion under section 157(2) of the Insolvency Act.²⁵ The Court did not disapprove of the Salford Estates approach on the exercise of discretion; in fact, the Court said that it was ‘in full agreement with the sentiments’ expressed in Salford Estates on the exercise of discretion.²⁶

The decision in Salford Estates was explored in more depth in Jinpeng Group v Peak Hotels and Resorts Ltd (Jinpeng), which concerned the alleged non-repayment of a loan and the creditor's application to have liquidators appointed over the alleged debtor. The loan contract contained an arbitration agreement. Although the Court in Jinpeng recognised the decision in Salford Estates, it relied on the earlier decision in C-Mobile as support for the proposition that the traditional standard was ‘too firmly a part of BVI law’ to adopt the Salford Estates approach and require a creditor to show ‘exceptional circumstances’.²⁷ The Court therefore maintained that to have an order appointing liquidators set aside it was necessary to show that the debt was bona fide disputed on genuine and substantial grounds.²⁸ The Court went on to find that, in any event, ‘exceptional circumstances’ did exist on the facts of the case due to allegations of fraud against individuals associated with the debtor and missing assets.²⁹

Some support for the pro-insolvency approach can also be seen in some recent decisions of the Hong Kong courts. After finding that the Lasmos requirements were not met on the particular facts before it, the Hong Kong Court of Appeal in But Ka Chon v Interactive Brokers LLC entered into an obiter discussion, stating that it had ‘reservations if the discretion under the insolvency legislation should be exercised only one way to substantially curtail the right of a creditor to present a petition’.³⁰ In a dispute concerning a sum allegedly due under a charterparty agreement, the Court of First Instance in Dayang Marine Shipping Co., Ltd v Asia Master Logistics Ltd went further. After applying the Lasmos requirements and finding that they had not been satisfied, the Court entered into a discussion of the merits of the ‘pro-arbitration’ approach, arguing that the approach in Salford Estates was ‘conceptually muddled’,³¹ and was ‘antithetical to the nature of a discretion and represents an unprecedented fetter on the Court's discretion’.³² The Court ultimately took the view that the existence of an arbitration agreement should be irrelevant to the exercise of discretion.³³ In Re Hongkong Bai Yuan International Business Co Ltd, Linda Chan J took the view that regardless of whether the prima facie or bona fide standard is applied, the debtor had to show that there was ‘a genuine dispute on the debt which requires the determination of a tribunal’.³⁴ Although the judge was of the view that there was no real difference between the various approaches, the suggestion that it is appropriate for the Court to consider the merits of the dispute and ‘review the evidence and arguments adduced by the parties’ means that this approach closely resembles the standard of the ‘pro-insolvency approach’.³⁵ Notably, in later decisions, Linda Chan J reverted to the requirement of a ‘bona fide dispute on substantial grounds’, and appeared to equate that standard with the existence of a ‘genuine dispute’.³⁶

A brief mention should also be made of the position in the Cayman Islands, where the issue has been considered only at first instance since the Salford Estates decision. In Re Grand State Investments, the Cayman Grand Court was faced with a dispute arising out of a failure to pay a share redemption price which had allegedly fallen due. The claimant brought a petition to wind up the company on the basis of an inability to pay its debts, but the company pointed to an arbitration agreement in favour of Hong Kong contained in the shareholders agreement. After finding that the petition should be dismissed on the basis of there being a bona fide and substantial dispute, the Court entered into an obiter discussion of the significance of the arbitration agreement. The Court began by acknowledging the decision of the Court of Appeal in Salford Estates and ‘the need to ensure that the parties’ freedom to choose a dispute resolution mechanism is respected’.³⁷ However, the Court went on to point to a concession from counsel for the petitioner that ‘the mere assertion of a dispute which does not pass the bona fide dispute on substantial grounds test’ will not be sufficient.³⁸ It seems likely that this concession was made due to the existence of Cayman Court of Appeal authorities pre-dating Salford Estates,³⁹ making it difficult to predict how the Cayman appellate courts may deal with the issue if faced with it again.

IV. IDENTIFYING THE ROLE OF POLICY

The question that arises from these cases is why different courts have reached such different decisions on the same issue, and what this tells us about the law in each jurisdiction. The question is particularly pressing in light of the disagreements between courts in the same jurisdiction, and the striking similarities between the legislative frameworks in each jurisdiction. This section of the article argues that the only way to explain the emergence of these divergent approaches is by recognising that the courts in each jurisdiction have made clear policy choices. This runs contrary to the argument advanced by some courts and commentators, which is explored further below, that the nature of the problem means that one approach is correct and the other incorrect. The first subsection considers the policy values emphasised by arbitration law, before exploring how these values have been reflected in the cases following the pro-arbitration approach. The second subsection then follows the same structure in connection with insolvency law and the cases which follow the pro-insolvency approach.

A. Arbitration Policy

Modern arbitration law is dominated by three major policy values: party autonomy, pacta sunt servanda and efficiency.

The significance of party autonomy can be seen most clearly in provisions such as section 1(b) of the UK Arbitration Act, which states ‘the parties should be free to agree how their disputes are resolved, subject only to such safeguards as are necessary in the public interest’.⁴⁰ The arbitration legislation of all jurisdictions examined in this article gives the parties a wide scope to shape the arbitral proceedings according to their own preferences. This practical significance of this principle should not be underestimated. In addition to the power to choose the number and identity of arbitrators,⁴¹ parties who elect to resolve their disputes through arbitration are able to control almost every aspect of arbitration proceedings including the procedural rules⁴² and the law applicable to the merits.⁴³ Certain jurisdictions even allow the parties to exclude the right to challenge an award on a point of law, to the discomfort of some judges.⁴⁴ Mandatory provisions are limited to dealing with establishing a basic framework in which tribunals can operate, such as establishing the immunity of arbitrators, and allowing a challenge to an award in extreme situations such as where there is evidence of serious procedural irregularities.⁴⁵ These permissive provisions are often accompanied by others which give the courts the ability to support ongoing arbitral proceedings actively through, for example, the power to preserve evidence and secure the attendance of witnesses.⁴⁶

Given that arbitration agreements are a contractual clause, it is unsurprising that the second major value in modern arbitration law is pacta sunt servanda. The role of pacta sunt servanda is perhaps most clearly seen in mandatory stay provisions,⁴⁷ but is also evident in the ability and willingness of courts to make orders in support of arbitration proceedings. In addition to issuing anti-suit injunctions,⁴⁸ anti-anti-suit injunctions⁴⁹ and anti-enforcement injunctions,⁵⁰ courts will give damages for a breach of an arbitration agreement.⁵¹ Indeed, it is arguable that the importance of pacta sunt servanda is heightened in relation to arbitration agreements through the doctrine of separability. To show that an arbitration agreement is void or voidable it is not enough to show that the contract as a whole is void; rather, it is necessary to show that there are circumstances relating specifically to the arbitration agreement which make that agreement void or voidable.⁵² This is indicative of a particular concern with ensuring that arbitration agreements are not unnecessarily found to be invalid, which is not extended to contractual provisions more generally.

Finally, modern arbitration law shows a clear concern with dispute settlement efficiency. Although not explicitly mentioned in arbitration legislation, this is apparent in the way that arbitration agreements have been interpreted. The leading English case on this issue is the House of Lords decision in Fiona Trust & Holding Corp v Privalov (Fiona Trust). The underlying dispute in Fiona Trust was whether a ship owner (the Appellant) had validly rescinded charter agreements entered into with the Respondent on the basis that they had been obtained by bribery. The issue before the House of Lords was whether an arbitration clause which stated that the parties could refer to arbitration ‘any dispute arising under [the] charter’ was broad enough to cover questions of bribery.

In answering the question positively, the House of Lords took the opportunity to review how courts should approach the interpretation of arbitration agreements. The Court took the view that the tendency of earlier courts to pay close attention to the language of the particular arbitration agreement before them, which resulted in the drawing of fine linguistic distinctions, should be rejected.⁵³ Instead, Lord Hoffmann explained that, in the absence of clear language to the contrary, courts should start from the presumption that ‘rational businessmen’ were likely to have intended that all their legal disputes should be heard in the same forum.⁵⁴ A similar approach has been followed by courts elsewhere. In Larson Oil and Gas Pte Ltd v Petropod Ltd, for example, the Singapore Court of Appeal stated ‘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’.⁵⁵

The Fiona Trust approach is often understood as being based on a concern for party autonomy, probably due to Lord Hoffmann's statement that ‘[a]rbitration is consensual. It depends on the intention of the parties expressed in their agreement. Only the agreement can tell you what kind of disputes they intended to submit to arbitration.’⁵⁶ Whilst party autonomy is clearly a concern, the Fiona Trust approach cannot be rationalised on the basis of party autonomy alone. Party autonomy could be served equally well, if not better, by adopting a granular analysis to the language of arbitration clauses. Further, no empirical evidence was presented by the House of Lords to demonstrate that business people actually prefer that their disputes are all settled in a single forum.⁵⁷ This presumption of ‘dispute resolution efficiency’ has resulted in arbitration agreements being interpreted broadly.

A concern to protect these three values drives the reasoning in the cases following the ‘pro-arbitration’ approach. In Salford Estates the Court of Appeal expressed concern at the possibility of a creditor using a winding-up petition to ‘by-pass the arbitration agreement and the 1996 Act’ and breaching the ‘parties’ agreement as to the proper forum’ for the resolution of disputes.⁵⁸ The Court also noted the ‘very wide’ terms of the arbitration agreement on the facts before it.⁵⁹ Reference to the need to hold a creditor ‘to the bargain it struck’ with the debtor and not to ‘undermine the policy of the Arbitration Act’ can also be seen in the decisions of the BVI courts which have followed the pro-arbitration approach.⁶⁰

Perhaps the most detailed treatment of these values can be seen in the Singaporean cases and the decision of the Hong Kong Court of First Instance in Lasmos. In BDG the Singapore High Court relied heavily on the principle of pacta sunt servanda when explaining why it was not appropriate to require alleged debtors to satisfy the traditional standard.⁶¹ The Court in AnAn expanded on these concerns, arguing that a court following the traditional approach risks undermining the parties’ agreement as it would result in either the court or the liquidator assuming the role of the arbitral tribunal and determining the disputed debt.⁶² The Hong Kong courts following the Lasmos approach have taken the view that, in light of the fact that creditors bring winding-up petitions to recover a debt, courts should be resistant to the circumvention of the contractually agreed forum.⁶³ When making similar remarks in the context of the relationship between winding-up petitions and exclusive jurisdiction agreements, the Court of Appeal in Guy Kwok-Hung Lam relied on a number of cases concerning arbitration clauses including Salford Estates.⁶⁴ In China Europe International Business School v Chengwei Evergreen Capital the Court, applying the Lasmos approach, also relied heavily on the ‘one-stop method of adjudication’ when interpreting the relevant arbitration agreement.⁶⁵

B. Insolvency Policy

The values underpinning modern insolvency law and the winding-up or liquidation of debtors differ significantly from those discussed above. Although varied, these values can be grouped under three broad headings: ensuring the fair distribution of an insolvent entity's assets between creditors; maximising the return to creditors; and protecting the public interest.⁶⁶

The concern of insolvency law with ensuring the fair distribution of an insolvent entity's assets amongst creditors can be seen primarily through the winding-up process itself. Although creditors have the right to bring a winding-up petition, doing so rarely results in a creditor recovering the full sum it is owed. Instead, a petition triggers a collective enforcement process in which the claims of all the insolvent entity's creditors are considered and ranked. As explained by Buckley J in In Re Crigglestone Coal, a creditor seeking to wind up a company is not seeking ‘an order for his benefit, but an order for the benefit of a class of which he is a member’.⁶⁷ Any hint that the creditor is using the winding-up process to pressure the debtor into payment of a debt which is disputed or not yet due is generally frowned upon and considered an abuse of process.⁶⁸ This collective enforcement mechanism eliminates the prospect of a race to judgment or enforcement, ensuring that assets are distributed according to principled and objective criteria.

Closely related to this is the second aim of maximising the return to creditors. This is achieved by a number of features of insolvency law, the most relevant in the current context being the collective enforcement mechanism triggered by a winding-up petition. The race to judgment which would occur in the absence of a collective enforcement procedure would be ‘inimical to the efficient organisation of the company's affairs for the benefit of the creditors as a whole’ as it would multiply the costs of debt collection.⁶⁹ Managing the process through a collective enforcement mechanism avoids the need for the insolvent company or its liquidators to pay the cost of dealing with several individual claims, thereby preserving more of the assets for distribution to individual creditors. Moreover, individual creditors are spared from having to bear the cost of bringing their own claim.

The public interest values of insolvency law can be seen in a number of places. First, such concerns are clear when considering who is treated as a ‘creditor’ in the winding-up process. The process is not limited to creditors whose claims are already in existence and have matured. Rather, tort claimants, employees and persons with claims for unliquidated damages are all treated as having provable debts which they can seek to recover through the winding-up process. Indeed, some of these creditors may be given priority over unsecured creditors whose claims are in existence and have matured. The English Insolvency Act, for example, gives priority to, inter alia, ‘preferential creditors’ over unsecured provable debts which includes employees (in relation to claims for pension contributions and remuneration) and certain tax liabilities (known as Crown Preference).⁷⁰ Consequently, in the majority of cases where a company is wound up, the private interests of unsecured creditors are affected by, if not subordinated to, broader public interest concerns.

Secondly, the concern for the broader public interest can be seen in the role of the liquidator.⁷¹ The liquidator plays an important role in helping creditors to protect their private rights, as the liquidator's primary function is the collection and distribution of the insolvent company's assets.⁷² Nevertheless, the liquidator also has a number of other important roles. The liquidator is responsible for investigating the reason for the insolvent company's failure, and for taking appropriate steps to punish any directors who were at fault.⁷³ To this end, the liquidator is typically able to make applications to the court to summon any persons who the liquidator believes may have information regarding the dealings or affairs of the company and require such persons to produce any relevant documents.⁷⁴ Such a role demonstrates that the insolvency regime has a clear concern with the healthy operation of the commercial market and not merely the private interests of creditors. The significance of the ‘public’ role of the liquidator was well explained by Lord Millett in Re Pantmaenog Timber:

[…] I reject the unspoken assumption that the functions of a liquidator are limited to the administration of the insolvent estate. This is only one aspect of an insolvency proceeding; the investigation of the causes of the company's failure and the conduct of those concerned in its management are another. Furthermore such an investigation is not undertaken as an end in itself, but in the wider public interest with a view to enabling the authorities to take appropriate action against those who are found to be guilty of misconduct in relation to the company.⁷⁵

Perhaps unsurprisingly, these concerns and values are reflected in the cases following or supporting the ‘pro-insolvency approach’. This is clearest in some of the decisions of the BVI. Although the Court in Jinpeng took the view that the dispute over the debt fell within the scope of the arbitration agreement, it emphasised that the nature of winding-up petitions meant that ‘the applying creditor [was] seeking a collective remedy on behalf of itself and all the creditors of the respondent […] [i]t is not a claim by the [creditor] to recover its debt from the respondent company’.⁷⁶ Although the public aspect of the insolvency process was not explicitly mentioned in the judgment in Jinpeng, the need to investigate the allegations that the director of the company had acted fraudulently appears to have weighed heavily on the Court's mind.⁷⁷ The Court in L Capital KDT also paid close attention to the collective nature of winding-up proceedings when following the decision in Jinpeng.⁷⁸

Such reasoning can also be seen in the decisions of the Hong Kong courts expressing concern about the approach in Lasmos. In But Ka Chon the Hong Kong Court of Appeal stated that it was ‘contrary to public policy to preclude or fetter’ a creditor's right to petition a court for the winding up of a company.⁷⁹ Similar concerns were expressed by the High Court in Dayang, albeit expressed in terms of the court's own power rather than the rights of creditors.⁸⁰

Some, such as the Singapore Court of Appeal in AnAn, have suggested that despite the clear division in values there is no actual conflict between the arbitration and insolvency regimes. The Court suggested that ‘the view that adopting the prima facie standard of review deals a blow to the insolvency regime begs the question, as it assumed that the company is in fact a debtor, when that question is exactly what the company and creditor have agreed to refer to arbitration’.⁸¹ Similarly, in Dayang the Hong Kong High Court argued that the presentation of a winding-up petition would not breach an arbitration agreement on the basis that the role of the court when hearing a petition is only to consider whether the debtor has shown a triable case on the defence, not to make any final determination.⁸² As will be seen further below, this reasoning does not demonstrate the lack of a conflict; it merely avoids confronting its existence by adopting a particular view of the problem, and of the arbitration and insolvency regimes.

V. EXPLAINING THE POLICY CHOICES: THE IMPORTANCE OF FRAMING

Two intertwined issues arise from the discussion so far. The first is whether it is possible for courts following either the pro-arbitration or pro-insolvency approach to recognise the importance of, and give weight to, the values of the non-preferred regime where appropriate. The second is how best to explain how and why courts have reached different conclusions as to the preferred regime and set of values. It is important to recognise that judgments following both approaches pay at least some attention to the concerns and values of the other approach. This is particularly true of the judgment in AnAn in which the Singapore Court of Appeal carried out a detailed review of the case law in Singapore and other jurisdictions, and considered arguments regarding the impact of its approach on values underpinning the insolvency regime.⁸³

However, there is an important distinction between merely laying out competing values and giving effect to them, and the discussion above shows that the courts have so far been unable to achieve the latter. Courts have been able to list the policy and values undermining both arbitration and insolvency law, but have allowed only one set of values to inform the rule ultimately adopted. The reason for this inability appears to be a failure to acknowledge the impact that formulating the problem has on a court's analysis. Put differently, by choosing to frame the problem in certain terms, courts make an implicit decision as to the lens through which the problem should be viewed and bind their own hands.

The pro-arbitration approach of the decision in AnAn is a good example of this. In the very first paragraph of its judgment, the Court explained that its starting point was that where there is a dispute falling within the scope of an arbitration agreement any court proceedings should be stayed in favour of arbitration.⁸⁴ The Court went on to compare the presentation of winding-up petitions to ordinary claims for disputed debts, and to suggest that disputed debts which form the basis for a winding-up petition concern ‘pre-insolvency rights and obligations’.⁸⁵ This naturally led to a concern to protect party autonomy rather than the public policy values which underpin insolvency law.

Notably, the Court did not explicitly consider the proper interpretation of the arbitration agreement. Instead, the Court seems to have followed the broad approach taken in Fiona Trust and assumed that the dispute falls within the scope of the arbitration agreement, when it formulated the issue before it as being about the appropriate ‘standard of review when a dispute that is subject to an arbitration agreement arises in relation to a debt which forms the basis of a winding-up application’.⁸⁶ In light of the Court's framing of the issue, it is unsurprising that it took the approach that it did. The framing of the issue itself has the effect of deciding which values are important when answering the question. Once the problem is understood as being concerned solely with the ‘pre-insolvency’ rights and obligations of the parties, it is difficult to see how there could be any place for the values which underpin insolvency law.

The same point can be made by considering cases which follow the ‘pro-insolvency’ approach. In Dayang the Hong Kong Court of First Instance began its analysis by taking as its starting point the rule that:

[…] if a debtor-company fails to satisfy a statutory demand for payment of a debt, the creditor-petitioner may petition to wind-up the debtor-company. The debtor-company may then apply to dismiss or stay the petition on the ground that there is a bona fide dispute on substantial grounds. It is not permissible for the debtor-company to merely deny, without more, the existence of the debt. One cannot just hold up his or her hand and simply say there is a dispute.⁸⁷

This passage shows that unlike the Singapore Court of Appeal in AnAn, the Court in Dayang took the position in insolvency law as its starting point. Further, the Court in Dayang did not recognise the distinction between pre- and post-insolvency rights and obligations which played an important role in the decision in AnAn. Instead, the Court emphasised the unique features of the winding-up process such as the collective enforcement mechanism and the role of the liquidator.⁸⁸ As a result, the Court framed the issue before it as about the exclusion of the winding-up process which does not, in itself, purport to determine disputes between the parties. This focus has the effect of elevating the values inherent in insolvency law at the expense of those which shape arbitration law. As above, framing the problem in this way makes it impossible for the court to give any real weight to arbitration values. If the winding-up procedure is understood not as a dispute resolution mechanism but as the commencement of a process which aims to achieve various public interest goals, it is difficult to see why an arbitration agreement and values such as party autonomy should have an impact on that process.

Some courts have recognised the possibility of exceptions, but this does not in itself do much to weaken the link between the framing of the issue and the approach ultimately adopted by the court. In the first place, such exceptions remain narrow and poorly defined. More important, however, is the fact that exceptions are, by their very nature, exceptional. Having a general rule accompanied by an exception means that the values of the non-preferred regime are considered only as an afterthought rather than at the outset. This creates a clear bias in favour of the default rule which is automatically followed in subsequent cases; in contrast, a departure from the default rule needs to be justified clearly. In most cases the default rule will be followed with little or no consideration of whether the values of the non-selected regime should be given weight.

Once the significance of framing is properly understood, it is easier to make sense of the claims from courts and commentators that there is, in fact, no clash between insolvency and arbitration. Such arguments are made possible because the conflict has been defined away. By understanding the problem as one of ‘pre-insolvency’ rights and obligations, the Court in AnAn is able to argue that its approach does not dismiss the aims and values of insolvency law as, in its view, those aims and values have simply not yet become relevant. Similarly, focusing on the distinct public interest elements of the winding-up process rather than the commercial aims of the petitioner allows the Court in Dayang to claim that it takes party autonomy seriously but simply does not see it as relevant to the problem it is dealing with. Although superficially attractive, such reasoning is illusory and does little more than shift the locus of the conflict to the level of framing the issue before the court. When courts decide this framing decision, they implicitly make a choice as to which side of the conflict they are choosing. As has been seen above, however, courts have tended not to justify their framing of the issue. The only potential exception to this is the decision in Lasmos where the Court drew a distinction based on whether the debtor had taken steps to commence arbitration.

VI. A QUESTION OF CHARACTERISATION

The final question is how to make sense of the framing decisions of courts in juridical terms, and how to provide courts with the tools to engage in more principled reasoning when determining whether to follow the ‘pro-arbitration’ or ‘pro-insolvency’ approach. Aspects of the process of ‘characterisation’ in private international law may offer some useful guidance. The authors of Dicey, Morris & Collins explain that ‘[t]he problem of characterisation consists in determining which juridical concept or category is appropriate in a given case’.⁸⁹ The question of characterisation is often significant in determining which national court should have jurisdiction over a dispute, or which national legal system should be applied to the merits of the dispute.⁹⁰

As a result, characterisation is often seen as a multistage process. In Macmillan v Bishopsgate Investment Trust (No 3) Staughton LJ gave what is now a classic statement of the characterisation process:

First, it is necessary to characterize the issue that is before the court. Is it for example about the formal validity of a marriage? Or intestate succession to moveable property? Or interpretation of a contract?

The second stage is to select the rule of the Conflict of Laws which lays down a connecting factor for the issue in question. […]

Thirdly, it is necessary to identify the system of law which is tied to the connecting factor in stage 2 to the issue characterised in stage 1.⁹¹

The Salford Estates problem does not necessarily involve a cross-border element, making Staugton LJ's second and third stages irrelevant for present purposes. There is, however, a close analogy between Staughton LJ's first stage and the ‘framing’ decision made by courts dealing with the Salford Estates problem as both involve ‘the allocation of the question raised by the factual situation before the court to its correct legal category’.⁹² Whereas a court dealing with a cross-border dispute may have to select between a number of legal categories, a court dealing with the Salford Estates problem will only ever have to choose between arbitration and insolvency law. Notably, the characterisation of the Salford Estates problem has jurisdictional consequences that are similar to cross-border disputes, only with the relevant jurisdictional choice being between a court and an arbitral tribunal rather than between two national legal systems. In light of this, the concept of characterisation can help to explain the decisions being made by courts when deciding how to frame the Salford Estates issue.

In addition to allowing courts to face up to the conflict and the nature of the decision that they are making, viewing the framing decision as one of characterisation provides guidance as to how courts could reach more reasoned answers. First, viewing the problem as one of characterisation opens the way to a more transparent and principled approach to deciding how the issue should be framed. Characterisation typically occurs in the context of a cross-border legal dispute, but it has also been used as a technique for resolving other legal and non-legal problems.⁹³ This is in large part due to the ability of the characterisation process, and of private international law more generally, to acknowledge that legal categories often exist in horizontal rather than hierarchical relationships.⁹⁴ Recognising this dynamic allows courts to find solutions to conflict problems which are sensitive to the aims and values of different legal systems, something that would not be possible by viewing the problem from the perspective of one of the conflicting systems. In particular, the characterisation process could be used to help courts to distinguish between cases in which there are real concerns about the debtor's financial position affecting the interests of other creditors and the wider public, and cases in which there are no such concerns and the case is best described as a bilateral dispute.

Secondly, viewing the problem as one of characterisation removes the need to apply a hard-and-fast rule. The first stage of characterisation is an inherently fact-sensitive exercise, which means that the problem will be viewed differently in each case. Interestingly, it is evident from many of the judgments considered above that courts see the need for a degree of flexibility within an otherwise strict approach. In Salford Estates this took the form of an, admittedly narrow, ‘exceptional circumstances’ exception.⁹⁵ The Court in AnAn adopted a slightly broader test, and held that the creditor would be permitted to bring a winding-up petition if it could show that the debtor's actions in disputing the debt amount to an abuse of process.⁹⁶ Examples given by the Court include the debt being admitted on both liability and quantum, and the dissipation of the debtor's assets.⁹⁷ However the existence of a default rule means that such values are not given weight at the outset and are only considered in exceptional situations.

Thirdly, viewing the framing decision as a characterisation exercise offers guidance on how courts might make the decision in individual cases. Much of the limited judicial and academic discussion of the first stage of characterisation is unhelpful for present purposes as it focuses on whether courts should take into account how other legal systems categorise obligations or simply rely on how they are categorised by the lex fori. However some guidance can be gleaned from the English cases on the issue. The practice of the English courts has been to take a multi-factor approach when attempting to determine the relevant legal category.

The issue arose in Macmillan v Bishopsgate itself, as the parties disagreed on whether a claim for the return of shares was restitutionary or proprietary in nature. The Court did not ultimately need to decide the issue as it held that the real issue was whether the defendants were bona fide purchasers for value which was considered to be an issue of property law. However the Court's discussion does provide support for a factor-based approach. Staughton LJ noted the existence of a number of factors for and against each categorisation, including the claimant's decision to plead that it had an equitable interest in the shares and the fact that equity acts in personam.⁹⁸

A similar approach can be found in the decision in Raiffeisen Zentralbank Österreich v Five Star General Trading which concerned the characterisation of a dispute relating obligations under an assigned insurance contract. Although the Court found that the text of the Rome Convention led to the conclusion that the issue was contractual, it also considered and weighed a number of factors. These included the significance of party autonomy in contractual assignments, the fact that the claim was for damages and potential impact of the assignment on third-party creditors.⁹⁹

An attempt to carry out a characterisation exercise can arguably be seen in the Hong Kong Court of Appeal decision in Lasmos. Rather than adopting a fixed approach, the Court in Lasmos argued that it is appropriate to take into account wider factors such as whether steps to commence arbitration proceedings have been taken and other, as yet undefined ‘exceptional circumstances’. In reaching this view, the Court considered both the need to hold parties to their contractual agreements, and the need to take into account the collective nature of the winding-up process and the public-interest elements of the insolvency process.¹⁰⁰

On closer inspection, however, the approach in Lasmos is unsatisfactory. First, although the Court in Lasmos mentioned the competing policy concerns, its approach ultimately did not balance those concerns. Unless and until the alleged debtor takes the steps required to commence arbitration, a court following Lasmos will require alleged debtors to satisfy the traditional requirement of showing that the debt is bona fide disputed on substantial grounds. In such a situation the values of arbitration receive no recognition, and it is as though the arbitration clause did not exist. In contrast, where an alleged debtor has taken steps to commence arbitration the values animating the insolvency regime are given no weight (unless the creditor can show the existence of exceptional circumstances).

Secondly, it is not clear why one party taking steps to commence arbitral proceedings should be a consideration in deciding which standard applies, let alone the main consideration. This aspect of the Lasmos approach appears to be based on deference to the tribunal rather than to the arbitral agreement, but the Court in Lasmos advanced no explanation on why such deference should attach to the taking of steps to commence arbitration and not to the arbitration agreement. Indeed, given the relationship between the arbitration agreement and the values of the arbitration regime discussed above, it is difficult to envisage why taking steps to commence arbitral proceedings should be considered a significant factor.

The factors which could be relied on under this characterisation approach should help a court to identify whether the case is one in which there are real concerns about the debtor's financial position, or whether it is closer to a standard disputed debt in which there are no indications that the debtor is in financial distress. Relevant factors will vary from case to case, and it is not possible to give a comprehensive list here; however, some factors which may be relevant to multiple-fact patterns can be identified. Many of these factors have been considered or relied on by courts dealing with the Salford Estates problem when arguing for either the ‘pro-arbitration’ or ‘pro-insolvency’ approach, but courts have not considered such factors as part of a holistic characterisation exercise. Factors which may be relevant include: the existence of other creditors;¹⁰¹ the nature of the application being made by the creditor;¹⁰² the existence of suspicions justifying the investigation of the debtor;¹⁰³ the debtor having admitted the debt;¹⁰⁴ and the creditor's subjective reason for making its application.¹⁰⁵ Although, as noted, the decision in Lasmos pays particular attention to whether the debtor had taken steps to commence arbitration, it is not clear why this is significant or affects the characterisation of the claim.

The natural objection to this approach is one of certainty: a multi-factoral approach would not have the advantages of predictability and convenience associated with a fixed rule, at least initially. This approach would require courts to consider the relevance and weight of various factors over a number of cases, finetuning their reasoning over time. Parties and their legal advisors would have to wait until the jurisprudence was relatively well developed before being able to predict with any certainty how a court would deal with their particular case. A multi-factoral approach akin to the process of characterisation would nevertheless be preferable. The discomfort expressed by some courts with the current arrangement and the uncertain scope of exceptions leaves the door open for considerable change and uncertainty under the current fixed-rule approach. In any case, parties and their legal advisors are well used to dealing with the gradual development of law in other areas, and there is no reason why this issue should be any different. The potential of establishing a more nuanced approach which is better able to weigh the competing values of the two regimes and is sensitive to the facts of particular cases is surely worth any initial uncertainty.

Even in the absence of a wholesale adoption of a multi-factoral approach, understanding the problem as one akin to characterisation may allow courts to engage more coherently with courts in other jurisdictions which have reached a different conclusion as to the correct test to be applied. Viewing the problem as a matter of characterisation and recognising the role of policy means that courts can avoid criticising judgments which take a contrary view as illogical or the result of muddled decision-making. Instead, the difference in opinion can be explained on the basis of legitimate differences in how the courts in a particular jurisdiction wish to balance the competing policy aims of arbitration and insolvency law.

VII. CONCLUSION

This article has shown that despite similarities between the legal systems, the courts which have examined the relationship between arbitration and insolvency law have produced radically different solutions. Whilst some courts have referred to both arbitration and insolvency values, they have tended to define away the tension rather than acknowledge and deal with it. Recognising that the issue the courts face is akin to that of characterisation in the conflict of laws provides the tools to help determine which regime is best suited to any given case. This allows the tension between the two regimes to be resolved in a way which gives appropriate weight to the values of both.