Sovereign Marine & General Insurance Company Ltd & Ors, Re

I have before me applications by 16 insurance companies (“the Scheme Companies”) for, in the case of each Scheme Company, an order convening a meeting of creditors to consider, and if thought fit, to approve a scheme of arrangement pursuant to section 425 Companies Act 1985 (“section 425” and “CA 1985”).

The first applicant, Sovereign Marine & General Insurance Company Limited (“Sovereign”) is insolvent and already subject to a scheme (“the Original Sovereign Scheme”) under section 425. All the other Scheme Companies are solvent. Two of them, Greyfriars Insurance Company Limited and Sovereign Insurance (UK) Limited (“the Sovereign Subsidiaries”), are wholly owned subsidiaries of Sovereign. Two of them, Allianz Marine & Aviation (France) (“Allianz Marine”) and Hibernian General Insurance Company Limited (“Hibernian”), are incorporated and regulated in, respectively, France and Ireland. Two others, Atlantic Mutual Insurance Company (“Atlantic”) and Continental Reinsurance Corporation International Limited (“Continental”), are incorporated and regulated in, respectively, New York and Bermuda. Richard Snowden QC appears for Sovereign and the Sovereign Subsidiaries. Gabriel Moss QC, Jeremy Goldring and Daniel Bayfield appear for the other Scheme Companies (“the Solvent Companies”). Richard Sheldon QC and Hilary Stonefrost appear for a number of objectors (“the Opposing Creditors”) which I will identify later.

Sovereign is a wholly owned subsidiary of the Willis group. Until 1991, all of the Scheme Companies underwrote insurance and reinsurance business in pooling arrangements (“the WFUM Pools”) through underwriting agents in the Willis group called Wills Faber (Underwriting Management) Limited (“WFUM”), Devonport Underwriting Agency Limited (“DUAL”) and Willis Faber & Dumas Limited (“WF&D”). In 1991, they ceased accepting new WFUM Pools business and went into run-off. Since 1998, the run-off has been administered by PRO Insurance Solutions Limited (“PRO”).

In 1997, Sovereign was placed into provisional liquidation on the ground of insolvency. The Original Sovereign Scheme, which is a “run-off” scheme, was sanctioned in 2000. The run-off of the WFUM Pools has continued to be conducted on a unified basis. Sovereign’s Scheme Administrators now wish to close its run-off by entering into a “cut-off” scheme. The terms of the schemes for each Scheme Company are found in the current draft of a comprehensive scheme document (“the Scheme Document”). The Scheme contained in Part II of the Scheme Document encompasses the separate schemes (a “Scheme”) for each of the Scheme Companies. The Scheme for each Scheme Company is, technically, a separate scheme and is capable of separate approval and sanction: the Schemes are not inter-dependent although they are conceived as a whole.

The proposed “cut-off” Scheme for Sovereign (“the Amended Sovereign Scheme”) is an estimation scheme dealing with all of Sovereign’s liabilities. The Schemes for the Sovereign Subsidiaries are estimation schemes which will deal with all of their insurance (and other) liabilities. The Schemes for the Solvent Companies are estimation schemes for those companies’ liabilities which arise out of the WFUM Pools business other than any UK compulsory liability insurance.

The Scheme Document contains in Schedule 1 a history of the WFUM Pools and a description of the business included in the schemes (“the Explanatory Statement”) which explains that approximately 50% of the liabilities of the WFUM Pools are liabilities of Sovereign or the Sovereign Subsidiaries and approximately 95% of Sovereign’s liabilities and those of the Sovereign Subsidiaries arise out of WFUM Pools business. The Scheme Companies contend that, if there were to be a fragmentation in the administration of the WFUM Pools, there would be many difficulties for policyholders, reinsurers and the Scheme Companies, including the duplication of effort and increased costs to all parties. This is one important factor on which the Solvent Companies will rely in promoting their own Schemes given the promotion by Sovereign and the Sovereign Subsidiaries of the Amended Sovereign Scheme and their own Schemes.

Sovereign’s application to convene a single meeting of its scheme creditors is not opposed. Similar applications by the Sovereign Subsidiaries were initially opposed but that opposition was withdrawn in the course of the hearing. However, the Solvent Companies, in seeking the summoning of a single class meeting of the scheme creditors of each of the Solvent Companies, are opposed by the Opposing Creditors which are a number of insureds in the WFUM Pools. There are 13 Opposing Creditors: Goodrich Corporation (“Goodrich”, a leading manufacturer of tyre and rubber products for most of the last century which divested itself of this business to become one of the world’s largest aviation component companies) and three of its corporate affiliates, Textron Inc (“Textron”) and six of its corporate affiliates (US industrial manufacturers), Sears Holdings Corporation (“Sears”, a leading US retailer) and Exxon Mobil Corporation (a US oil company) each of which has purchased policies known as “occurrence” policies. These policies provide broad prospective coverage against claims relating to the risks covered by the policies (which include asbestos, pollution and health hazard liabilities) if the underlying act or omission that gives rise to the claim happened during the relevant policy period. These claims (known as long tail claims) typically arise and are asserted years or even decades after the claimant’s latent exposure to the allegedly hazardous substance. There are other creditors who have communicated their objections to the Scheme Companies or their solicitors but which do not appear before me. Their objections do not raise any issues which have not been fully aired before me.

There are two main areas of objection. The first objection is jurisdictional and relates to the position of Allianz Marine and Hibernian. Mr Sheldon submits that the English court has no jurisdiction to sanction a scheme in relation to them and that class meetings should not, therefore, be convened at all. The second is that a single class meeting for each of the Solvent Companies would not be proper; Mr Sheldon submits that there are at least two, and possibly more, classes which are entitled to separate class meetings. Before dealing, in that order, with the objections, I need to deal with the nature of the WFUM Pools business and with some, at least, of the provisions of the proposed Schemes.

The WFUM Pools give rise, as Mr Moss says, to complex inter-relationships between the participants in the underwriting arrangements, policyholders and reinsurers. The history is set out at some length in the Explanatory Statement. The important points are as follows (which I gratefully adopt from the opening written submissions on behalf of the Scheme Companies).

The Opposing Creditors’ case is that they purchased the occurrence policies (referred to in paragraph 7 above) issued by or subscribed to by the Scheme Companies for substantial premiums in order to secure the unlimited prospective cover which those policies provided. The policies were intended to protect them against the risks that the latent liability exposures referred to in paragraph 7 could materialise years or decades later. These policies, they say, are a crucial component of the policyholders’ ability to manage their liability risks and without these policies each policyholder would have had to set aside a proportion of its capital each year. Replacement coverage cannot be purchased at any price in the current insurance market as companies operating today no longer provide this type of cover and the Scheme Companies do not dispute that.

The Opposing Creditors say that the effect of the proposed Scheme is to extinguish a valuable asset of the policyholders, namely the insurance cover, by terminating their existing contractual rights and relieving the Solvent Companies of executory obligations that they freely assumed for substantial premiums over the course of six decades. In effect the policyholders, if the Schemes are eventually sanctioned against their wishes, are being forced to commute their claims on terms that they say are more favourable to the Solvent Companies than could be achieved by negotiated agreement if policyholders were in the present circumstances prepared to negotiate commutation.

The evidence shows the extent of the insurance cover provided to the Opposing Creditors in the WFUM Pools:

Six other US companies, KeySpan Corporation, General Motors Corporation, National Passenger Railroad Corporation, Norfolk Southern Corporation, The Boeing Company and United Technologies Corporation and one company incorporated in this jurisdiction, National Grid plc, also object to the Schemes proceeding on the basis now proposed.

The principal objection at this stage on the part of the Opposing Creditors and the other objectors is that the Solvent Companies propose only one class meeting with the result that future claims that have been incurred but not yet reported, typically because no claim has been made by the person ultimately suffering damage eg an individual injured as a result of exposure to asbestos who does not yet know that he or she has suffered injury, will fall, for voting purposes, into the same class as claims which have been notified and whose value can, according to the Opposing Creditors, more easily be estimated.

The Scheme Document is long and the Schemes are not straightforward. It is not necessary, for the purposes of this application, to go into the details of them. The following is a summary of the salient points which I again take from the written opening submissions of the Scheme Companies. A longer overview can be found in the evidence of Mr Hunt in paragraphs 27 to 53 of his first witness statement.

The Scheme contains many definitions some of which I should mention in order to make sense of what follows:

The Schemes are “cut-off” schemes as set out briefly in paragraph 5 above. The purpose is to terminate the run-off of Scheme Claims by estimating the value of all Scheme Claims as at the Ascertainment Date (31 December 2005) of Scheme Creditors, which must be submitted before the Bar Date (now projected to be 28 February 2007). Claims against the Solvent Companies will be paid in full although a time value discount will be applied. A dividend will be paid on claims against Sovereign.

Under the scheme, Liabilities will be dealt with, in summary, in the following way:

The Financial Services Authority (“FSA”) has been consulted in relation to the Schemes and, in relation to those of the Scheme Companies over which it has direct regulatory control, ie all of the companies other than Allianz Marine, Atlantic, Continental and Hibernian, has confirmed that it has no objection to them.

The FSA has no direct regulatory control over Allianz Marine, Atlantic or Continental. However, those companies have liaised with their respective regulators with regard to the Scheme. The FSA has no direct regulatory control over Hibernian either. The Regulation and Compliance Manager of Hibernian decided not to consult the Irish insurance regulator “due to the extremely low and immaterial scale of the WFUM Pools Business in [Hibernian’s] books”.

Mr Hunt explains that Clause 2.8.4 of the Scheme provides that where (as will generally be the case) a Scheme Creditor is a creditor of more than one Scheme Company, it must also abide by the terms of the other Schemes which become effective. He says that this clause has been included to address the position of Atlantic which, for the reasons outlined in Schedule V to the Explanatory Statement, is not eligible under US law for Chapter 15 (of the US Bankruptcy Code) protection of its Scheme in the United States. Provided that the English Court sanctions the schemes, the Scheme Companies other than Sovereign and Atlantic will apply for permanent injunctive relief from the United States Bankruptcy Court under Chapter 15 to ensure that the Schemes will be given full force and effect and be binding on and enforceable against all Scheme Creditors in the United States. Sovereign currently has protection in respect of the Original Sovereign Scheme under section 304 of the US Bankruptcy Code and will request that this is modified to grant recognition of the Amended Sovereign Scheme. A summary of the terms of the Chapter 15 and section 304 permanent injunction orders are contained at Schedule V of the Explanatory Statement.

Once the Schemes are effective, clause 4.1.1 provides that Scheme Creditors will not be permitted, save as provided for in the Schemes, to commence or continue any Proceedings against the Scheme Companies to establish the existence or amount of any Scheme Claim. However, the Schemes do not prevent a Scheme Creditor from taking proceedings where the relevant Scheme Company has failed to perform its obligations to make a payment to the Scheme Creditor under its Scheme. The stay is modified in relation to Protected Policyholders of Sovereign.

The position of Sovereign is different from those of the other Companies in that: (a) it is insolvent; (b) it is already subject to the Original Sovereign Scheme (which extends to all its liabilities and is a single class scheme, notwithstanding that it applies to both protected policyholders and non-protected policyholders); and (c) the Amended Sovereign Scheme will likewise extend to all its liabilities.

Clauses 9 to 17 of the Scheme apply only to the Sovereign Amended Scheme. There are certain provisions to which I should make specific reference:

The Estimation Methodology is the means by which Liabilities are to be valued when the disagreement about value is referred to the Scheme Actuary rather than the Scheme Adjudicator. It uses a number of definitions many of which I need to refer. The primary division for the purposes of the Estimation Methodology is into “Unpaid Agreed Claims”, “Outstanding Claims” and “IBNR” (“incurred but not reported”). They correspond to the categories in BAIC of “unsettled paid claims”, “outstanding losses” and IBNR. The definitions in the Estimation Methodology are, paraphrasing, as follows:

There is one other aspect which it is convenient to mention here. Apart from Outstanding Agreed Claims, there are some other claims not requiring valuation or estimation. These are:

The Opposing Creditors’ principal objection is that the Schemes in relation to the Solvent Companies fail to constitute a separate class of Scheme Creditors in respect of IBNR for the purposes of section 425. Logically prior, however, is the issue of jurisdiction in relation to Allianz Marine and Hibernian. There is no jurisdictional issue in relation to Atlantic and Continental in the light of the decision in In re Drax Holdings Ltd [2004] 1 WLR 1049 (“Drax”) to which I will come in a moment.

Section 425 makes provision in relation to schemes of arrangement in respect of a “company”. For the purposes of the section, “company” is defined as “any company liable to be wound up under this Act”. In relation to Allianz Marine and Hibernian, Mr Sheldon submits that the court has no jurisdiction to sanction the Scheme under section 425 CA 1985 on the basis that neither is a company which is “liable to be wound up under this Act”. The words “this Act” in section 425(6)(a) include, by virtue of section 735A(1) CA 1985, Parts I to VII IA 1986 and thus incorporate sections 220 and 221 IA 1986. It should be remembered that the ordinary definition of “company” in section 735(1) CA 1985 does not include unregistered companies. Accordingly, schemes for such companies, whether solvent or insolvent, would not be possible at all without widening the definition of “company” for the purposes of section 425. A similar distinction was found in predecessor legislation. Thus, section 206 Companies Act 1948 contained the provision for the making of schemes of arrangement but “company” was defined, for the purposes of that section, as “any company liable to be wound up under this Act”. A company registered under that Act was, of course, one to which the winding up provisions applied; but those provisions were capable of application, at least in certain circumstances, to certain other companies too ie other companies “liable to be wound up under this Act”.

Under section 221(1) IA 1986, an unregistered company may be wound up under that Act. The definition of “unregistered company” in section 220 IA 1986 is wide enough to include a company incorporated in a foreign state. The jurisdictional requirements of what is now section 221(1) were examined by Lawrence Collins J in Drax in the course of considering the meaning of “liable to be wound up” in section 425(6)(a) CA 1985. It is to be noted that the Judge was there dealing with companies incorporated in the Cayman Islands and in Jersey; the case did not, therefore, raise any questions under the Community instruments which I will come to in due course. After referring to, but not citing from, a number of cases, he said this at paragraph 22:

“22 In Stocznia Gdanska SA v Latreefers Inc (No 2) [2001] 2 BCLC 116, 140, the Court of Appeal confirmed that the presence of assets of the company in England was not a precondition to the exercise of jurisdiction (although the presence of assets would constitute good reason in the normal case), and that (approving the formulation of Knox J in In re Real Estate Development Co [1991] BCLC 210, 217) before a foreign company could be wound up in England, three core requirements had to be fulfilled: (1) there must be a sufficient connection with England which may, but does not necessarily have to, consist of assets within the jurisdiction; (2) there must be a reasonable possibility, if a winding up order is made, of benefit to those applying for the winding up order; and (3) one or more persons interested in the distribution of assets of the company must be persons over whom the court can exercise jurisdiction.”

The Judge then went on to consider the nature of those three requirements, that is to say whether they go to the court’s jurisdiction to wind up an unregistered company or whether they are simply judicially imposed restrictions on the way in which the court will exercise its discretion. As to that, the Judge said this at paragraphs 23 to 27:

“23 It was not necessary in that case (or in the other cases which led to, or followed, the formulation of the three preconditions) to decide whether these requirements were preconditions for the existence of the statutory jurisdiction of the court, or principles to be observed in considering the discretion to exercise the jurisdiction. The latter approach was favoured (obiter) by Sir Donald Nicholls V-C in In re Paramount Airways Ltd [1993] Ch 223, 240.

24 In most cases the distinction will not matter. The English court will not wind up a foreign company where it has no legitimate interest to do so, for that would be to exercise an exorbitant jurisdiction contrary to international comity, and for that purpose it does not matter whether the preconditions are couched in terms of the existence of jurisdiction or the exercise of jurisdiction.

25 But in the present case it may make a difference, because the question is one of the jurisdiction to approve a scheme of arrangement, and the second and third conditions may not be relevant because they were formulated in the context of winding up. If they go to the jurisdiction to order a winding up, the words "any company liable to be wound up" in section 425(6) may require those conditions to be fulfilled even in the case of schemes of arrangement. If they go to the discretion to wind up, then they do not have to be fulfilled in the case of a scheme of arrangement, although the first condition would plainly be relevant in any event.

26 The question therefore is whether (as was assumed in the present matter by the companies) the combined effect of section 425(6) of the 1985 Act and of section 221(1) of the Insolvency Act 1986, and the cases on the winding up of foreign companies, is that the three conditions must be satisfied before the court can exercise its powers under section 425. In my judgment the three conditions go to the discretion of the court, and not to the existence of its jurisdiction. If that it is right, then the conditions do not have to be satisfied for the purposes of section 425, because they do not go to the question whether a company is "liable" to be wound up under the Insolvency Act 1986. So also it is not necessary for the purposes of section 425 that the grounds for winding up in section 221(5) exist.

27 Any other result would led to very odd and artificial consequences, since schemes of arrangement are used in many circumstances having nothing to do with insolvency.”

It can be seen that the Judge therefore concluded that the three conditions went to discretion to wind up and not to jurisdiction, observing that the first condition would plainly be relevant in any event, that is to say as a matter of discretion in relation to the sanction of a scheme of arrangement. He concluded that the Court had jurisdiction to sanction a scheme of arrangement in relation to the Cayman Islands and Jersey companies with which he was concerned provided that there was a sufficient connection with England.

It is to be noted that the Judge says that, for the purposes of section 425, the grounds for winding up in section 221(5) do not need to exist. Those grounds, it seems to me, go to jurisdiction; so that the Judge is effectively answering the question which he left open in paragraph 25 (viz that, if the conditions he was considering went to jurisdiction, it may be a requirement that they are fulfilled even in the case of a scheme of arrangement). In fact, it seems that both companies were insolvent (see the end of each of paragraphs 5 and 7 of the judgment) so that the condition of section 221(5)(b) was fulfilled and jurisdiction was founded in any case. Drax is not, therefore, authority for the proposition that a solvent unregistered company is “liable to be wound up”.

It was not, however, suggested to me that the Court has no jurisdiction to sanction a scheme in relation to a solvent foreign company outside the EU/EEA unless one of the conditions in section 221(5) is in fact fulfilled. That would be a surprising conclusion and one which I consider to be incorrect for the reasons which follow. The question “Is this company liable be wound up under IA 1986?” could be taken in two senses: first, in the sense whether it could be subject to a winding up process under IA 1986 on the facts as they stand at present; secondly, in the sense whether the company is the sort of company which is capable of being wound up under IA 1986. In my judgment the latter sense affords the correct approach to the meaning of “liable to be wound up” in section 425(6)(a): it is not necessary to show that any of the conditions of section 221(5) is in fact fulfilled. Thus a foreign (non-EU/EEA) company is a company which is capable of being wound up in the sense that, if any of the circumstances set out in section 221(5) arises, then the court has power, subject to its discretion and thus, in particular, to the three conditions considered in Drax, to wind it up.

The relevance of this analysis is this: it is, I consider, proper to take account of each of the circumstances set out in section 221(5) under which a company could be wound up in considering whether it is “liable to be wound up”. It is not necessary to focus, in the case of a solvent company, only on those circumstances in which a solvent company could be wound up. This is because, in order actually to fall within any of the circumstances set out in section 221(5), a change in the actual factual position must be posited: there is no warrant for saying that one sort of change (eg to bring the case within section 221(5)(c)) is to be taken into account but that another (eg to bring the case within section 221(5)(b)) is not. The importance of this point will become apparent when I turn to Mr Sheldon’s argument in relation to the effect of Regulation 5(1) of the Insurers (Reorganisation and Winding Up) Regulations 2004 (“the Insurers Regulations”). It follows from this that the Court has jurisdiction to sanction the Schemes for Atlantic and Continental.

Mr Moss goes further. He submits, relying on Drax, that the position has now been reached where it is sufficient, in order to be able to say that an unregistered company is “liable to be wound up under this Act” for the purposes of section 425(6)(a), that there is a sufficient connection between the relevant scheme and England. It is then common ground that, if that is correct, there is on the facts of the present case such a sufficient connection in the cases of both Allianz Marine and Hibernian.

Mr Moss submits that it is unnecessary to look further to Community instruments concerning jurisdiction when deciding whether a company incorporated in another Member State is “liable to be wound up under this Act”. He says that none of the Community instruments which might conceivably be of relevance (the Jurisdiction Regulation, the Insolvency Regulation and Directive 2001/24/EC (“the Directive”) referred to below) nor the Brussels and Lugano Conventions as implemented by English statute, has anything express to say about the jurisdiction of the English court in respect of schemes of arrangement. It could only be because Community instruments are concerned, in various situations, with the jurisdiction to open the winding up of a company incorporated in a Member State that there could be any effect on the jurisdiction in relation to schemes; and that could only be because a change in the situations in which such a company could in fact be wound up by the English court, has incidentally and unintentionally brought about an indirect change in the class of companies “liable to be wound up under this Act”.

As to that, he correctly points out that the “liable to be wound up” criterion was inserted in the English legislation long before there were international rules of allocation of jurisdiction between Member States in the EU or under the Brussels and Lugano Conventions: one need go back no further than section 206(6) Companies Act 1948 to illustrate that point. He submits that the criterion was inserted simply in order to ensure an adequate connection with the English jurisdiction. (I am not sure that that point is correct, but nothing turns on it: I would have thought that the provision was inserted simply to provide a definition of “company” for the purpose of schemes which went beyond the ordinary meaning of “company” as defined in the legislation and did so in a short-hand, referential, way.) There is no reason, he says, to suppose that, when rules of allocation of international jurisdiction were introduced between the relevant states (first under the Brussels and Lugano Conventions and then under the Jurisdiction Regulations) they had any effect one way or the other on the “adequate connection” requirement: there is, he says, no logical basis for importing any rules of allocation of international jurisdiction which may exist in relation to solvent winding up into schemes which, according to him, have no such rules. He points out that, in the analogous case of the importation of rules of allocation of international jurisdiction into domestic law criteria for the voluntary winding up of a foreign unregistered company, this was done expressly by section 221(4) IA 1986; he says that the same would have been done in relation to schemes if that is what had been intended.

Let me put what I see as the thrust of that argument (I shall refer to it from now on as “the Primary Argument”) in my own words. First, absent the Conventions (and their implementation into English law) and Community instruments, the English court would have jurisdiction to sanction a scheme in relation to a company incorporated in an EU Member State if there is a sufficient connection with England. Secondly, there is nothing in the Conventions or Community instruments which purports expressly to deal with jurisdiction in relation to schemes of arrangement and nothing to suggest that the framers of those documents had schemes in mind. Accordingly, as a matter of Community law there is nothing which requires a Member State to apply to schemes the same rules of jurisdiction as it requires to be applied in respect of winding up. Thirdly, the criterion “liable to be wound up under this Act” is relevant only in the context of schemes and its meaning should not be taken as having been affected, as it were by a side wind, as the result of Community instruments (or their implementation in English law); the meaning of “liable to be wound up under this Act” remains the same today as it was in the past, for instance immediately after the commencement of the Companies Act 1948. It makes no difference to this third proposition, according to Mr Moss, that the earlier legislation has been replaced by CA 1985 and IA 1986: CA 1985 is a consolidation Act and section 425 simply re-enacted without amendment longstanding provisions concerning schemes.

I have already dealt with the first of those propositions in considering Drax. I consider that it is correct. Mr Sheldon would take issue with both the second and third of those propositions. I will come to the second of them in due course but mention here that he submits that the Directive does indeed deal with creditors’ schemes which, he says, are “reorganisation measures” as defined in the Directive.

As to the third of those propositions, even if Mr Moss is right in saying that the “liable to be wound up” criterion was inserted by Parliament to ensure an adequate connection with England, the fact is that Parliament chose to define the class of unregistered companies which could be made subject to a scheme by reference to companies which were “liable to be wound up”. What is more, it did so by reference to the criteria relevant to compulsory winding up, there originally being no power at all to wind up an unregistered company voluntarily, the most important class of company subject to compulsory winding up being insolvent companies. And this was done even though many schemes (ie members’, rather than creditors’, schemes) may have nothing to do with creditors or insolvency at all.

One might therefore think that if some law – be it an Act of Parliament or an overriding piece of EU legislation – were passed which provided expressly that the English court should not have jurisdiction to wind up a particular class of unregistered company, it could no longer be said that a company within that class was “liable to be wound up under this Act”. That, I do not doubt, is the literal meaning of the words; and if schemes of arrangement had been invented and introduced only after such a jurisdictional law as I have just mentioned had been made, it is, I venture to suggest, inconceivable that any judge would say that the company was “liable to be wound up under this Act”. It is only the history of the legislation which opens up the possibility of the argument which Mr Moss puts forward. It is necessary, perhaps, to put one slight gloss on what I have just said: it may be that the same, or another piece of legislation, also provided that a voluntary winding up of a foreign registered company could in some circumstances take place in England (as section 221(4) now appears to suggest in relation to companies within the scope of the Insolvency Regulation) and it might then be said that the company “is liable to be wound up under this Act” if the conditions for a voluntary winding up of such a company could be fulfilled.

Thus, suppose that an Act of Parliament were passed which stated that a Jersey company should be wound up only under the Jersey Companies Acts and that the English court should have no jurisdiction to wind up such a company. It could not, applying a literal meaning of the words in section 425(6)(a), then be said that Jersey companies nonetheless remained “liable to be wound up under this Act” provided that there was a sufficient connection with England. Accordingly, there would, applying the literal meaning, be no jurisdiction to sanction a scheme in relation to a Jersey company under section 425. The position would be the same in the case of a company incorporated in an EU Member State if a piece of EU legislation having direct effect in England provided for an allocation of international jurisdiction which precluded the English court from winding up such a company and which allocated winding up to the foreign state.

In the light of considerations such as that, Mr Sheldon submits that, if EU legislation provides that exclusive jurisdiction to wind up a company is allocated to a particular Member State other than the UK, then it cannot be said that that company is “liable to be wound up under this Act”. In effect, he says that an allocation of international jurisdiction in relation to winding up would flow through to the meaning of “liable to be wound up under this Act” and does so as a matter of construction of section 425(6)(a) itself.

In making that submission, Mr Sheldon has to meet what Mr Moss says is the effect of the decision in Drax. In my judgment, that decision is not an authority contrary to the result for which Mr Sheldon contends as the analysis which I have already carried out shows. The cases which Lawrence Collins J considered in Drax were all concerned with the limits on the exercise of an apparently wide jurisdiction to wind up a company: absent those conditions, there would have been jurisdiction to wind up the companies concerned. In Drax, Lawrence Collins J, having held that all three of the conditions went to discretion to wind up and not to the jurisdiction to do so, did not discuss any other restrictions on jurisdiction (other than the special restrictions under EU Law) to wind up but simply assumed that there would be jurisdiction to do so. Drax has nothing to say about the effect, if any, of a separate jurisdictional bar on the English court winding up a foreign company.

That is not to say that the Primary Argument is wrong. However, I do not, at least at this stage, decide whether Mr Moss or Mr Sheldon is correct on the Primary Argument. Instead, I turn to look beyond CA 1985, IA 1986 and the existing case law to see whether there is anything in the case of a company with its seat in an EU Member State or an insurance company within the scope of the Directive and the Insurers Regulations which either (i) robs the English court of any power which it would otherwise have to wind up such an unregistered company or, to put it another way, which restricts the jurisdiction apparently conferred by section 221; or (ii) allocates international jurisdiction in relation to schemes of arrangement in relation to such companies.

The EU legislation and other material which has featured in argument includes the Jurisdiction Regulation and the Jenard and Schlosser Reports which preceded them, the Insolvency Regulation and the Directive on the reorganisation and winding up of insurance companies. The last of those has been implemented in the UK by the Insurers Regulations although Mr Sheldon submits that it has not been properly implemented. Mr Moss submits that jurisdiction in relation to schemes is, in the case of insurance undertakings, preserved or conferred by the Insurers Regulations. Even though the Directive (being a directive rather than legislation) is not of direct effect in Member States, it is common ground that the Insurers Regulations should be construed in the light of, and so as to give effect so far as possible, to the Directive.

It is helpful, I think, to start with the Directive. This applies to insurance undertakings regulated by a contracting State of the European Economic Area (“EEA”): it is common ground that Allianz Marine and Hibernian are both such insurance undertakings. The Directive applies in relation to both solvent and insolvent insurance undertakings and contemplates that matters relating to the winding up and reorganisation (so far as covered by the Directive) should be dealt with exclusively by the home State, that is to say the regulating State.

The Directive was, I have mentioned, implemented by the Insurers Regulations. That implementation appears, rightly or wrongly, to have proceeded on the basis that schemes of arrangement were neither winding up proceedings nor reorganisation measures within the definitions in the Directive, Regulation 4(1) dealing only with winding up, provisional liquidators and administration orders. Further, one sees, in Regulation 5(1), schemes expressly being excluded from the allocation of international jurisdiction in the Directive.

Regulation 4(1) Insurers Regulations provides that

“…..a court in the United Kingdom may not, in relation to an EEA insurer or any branch of an EEA insurer –

(a)

make a winding up order pursuant to section 221 [IA 1986]…;

(b)

appoint a provisional liquidator;

(c)

make an administration order”.

For the purpose of the Insurers Regulations, an EEA insurer is an undertaking, other than a UK insurer, authorised to write direct insurance. The UK retains jurisdiction over UK insurers, that is to say insurers regulated under the Financial Services and Markets Act 2000 even if incorporated in another EEA State . Although section 221(4) IA 1986 appears to contemplate voluntary winding up of certain companies taking place under the Act, this is so only in accordance with the Insolvency Regulation. That Regulation does not, however, apply to an insurance undertaking (eg to Allianz Marine and Hibernian): see Article 1(2) Insolvency Regulation. Accordingly, the English court has no jurisdiction to wind up an insurance undertaking regulated in another EEA State, and voluntary winding up is also excluded.

If the Insurers Regulations said nothing more then, on the literal reading of section 425(6)(a), it would follow that no scheme of arrangement in relation to an insurance undertaking regulated by an EEA State other than the UK could be sanctioned by the English court. However, consistently with the view that schemes of arrangement were neither winding up proceedings nor reorganisation measures within the meaning of the Directive, Regulation 5(1) provides as follows:

“For the purposes of section 425(6)(a) of the [CA 1985]…..an EEA insurer or a branch of an EEA insurer is to be treated as a company liable to be wound up under that Act….if it would be liable to be wound up under that Act…..but for the prohibition in regulation 4(1)(a)”.

Mr Moss submits that this is a clear provision the result of which is that a scheme of arrangement in relation to an EEA insurer can be sanctioned by the English court provided that there is a sufficient connection with England in accordance with the decision of Lawrence Collins J in Drax.

Mr Sheldon, however, submits that that is to misunderstand what Regulation 5(1) really says. Regulation 5(1) does not, he says, prescribe the jurisdictional basis for winding up an EEA insurer: that basis has to be found elsewhere. He says that “but for” the prohibition in Regulation 4(1)(a) (ie as he would say, in the absence of that prohibition) an EEA insurance company would not be “liable to be wound up under this Act”.

According to this argument, because Allianz Marine and Hibernian are clearly solvent, any liquidation would be a solvent liquidation. The Insolvency Regulation would not apply, it being common ground that they apply only to insolvent liquidations, and in any event do not apply to insurance undertakings. Instead, Mr Sheldon submits that Article 22(2) of the Jurisdiction Regulation would apply to the winding up to assign exclusive jurisdiction to the place where the company has its seat (in relation to Allianz Marine and Hibernian, France and Ireland respectively). On that basis, Allianz and Hibernian are not “liable to be wound up under this Act” even if one ignores the prohibition in Regulation 4(1)(a) as one is enjoined to do by Regulation 5(1).

Without embarking on a consideration of Mr Sheldon’s argument as to the scope of Article 22(2) of the Jurisdiction Regulation, I reject his submissions on this aspect of the case for the following reasons.

There can be no doubt that the Directive is intended to be the exclusive instrument for the allocation of jurisdiction in relation to the winding up of insurance undertakings whether or not the winding up proceedings are based on insolvency or not. That is reflected in the express exclusion in the Insolvency Regulation of insurance undertakings.

It is also reflected in recital (24) of the Jurisdiction Regulation which provides that that Regulation “should not affect rules governing jurisdiction and the recognition of judgments contained in specific Community instruments”. The Directive is just such a Community instrument. Mr Sheldon, however, submits that if one is to ignore Regulation 4(1)(a) of the Insurers Regulations as directed by Regulation 5(1), then that must be done also for the purposes of recital (24). I disagree. The Directive is a specific Community instrument within the meaning of recital (24) which governs, to the exclusion of the Jurisdiction Regulation, the allocation of jurisdiction in relation to the winding up of insurance undertakings. There is no scope for the application of Article 22(2) in relation to such companies. The Directive, so far as concerns actual winding up, is implemented by the Insurers Regulations so that there is (subject to one argument to which I will come concerning the meaning of “reorganisation measures” in the Directive) no scope for the application of the Jurisdiction Regulation. The statutory hypothesis which has to be applied under Regulation 5(1) in relation to the meaning of “liable to be wound up” does not require one to assume, contrary to the reality, that the Jurisdiction Regulation is to be treated as applying to a situation to which it could in fact never apply.

A related point (or perhaps the same point put in another way) is this: it is to be remembered, that so far as concerns Community law, the Directive is a special Community instrument within recital (24) of the Jurisdiction Regulation; Article 22(4) does not therefore apply to a solvent EEA insurer. It does not, in my judgment, have any impact on that proposition whether a contracting State (in the present case the UK) has or has not implemented the Directive or, where it has done so, that it might have done so inadequately. Even if Mr Sheldon is correct in saying that the “but for” proviso in Regulation 5(1) means that Regulation 4(1)(a) must be ignored for all purposes, Regulation 5(1) does not provide that the Directive itself is to be ignored in determining the scope of the application of the Jurisdiction Regulation. Accordingly, the “but for” proviso does not result in the Jurisdiction Regulation applying in determining whether an EEA insurer is “liable to be wound up”.

The conclusion, therefore, is that Regulation 5(1) disapplies Regulation 4(1)(a) but does not thereby result in the application of Article 22(4). The application of the “but for” proviso therefore results in a notional application of the same test in determining whether an EEA insurer is “liable to be wound up” as apply to a non-EU/EEA company. I have addressed that test at length already and concluded that the English court does have jurisdiction to sanction schemes in relation to such companies. Accordingly, in my judgment the English court has jurisdiction to sanction a scheme of arrangement in relation to an EEA insurer provided that it has a sufficient connection with England.

Moreover, Mr Sheldon’s argument in relation to Regulation 5(1) of the Insurers Regulations leads to the conclusion that an insolvent EEA insurer is “liable to be wound up”. Under that argument, the “but for” proviso in Regulation 5(1) throws one back to the position which would obtain ignoring, for all purposes, Regulation 4(1) and the Directive. (Mr Sheldon has to say that the Directive, too, is to be ignored since otherwise it would clearly remain a relevant special Community instrument precluding the application of the Jurisdiction Regulation, the application of which is essential to his argument in relation to solvent companies.) But in the case of an insolvent EEA insurer, the Jurisdiction Regulation does not apply (see Article 2(b)) so that there is nothing to deprive the English court of jurisdiction (the Insolvency Regulation not applying to an insurance undertaking: see Article 1(2)).

In considering, in the course of my analysis of Drax, why it is that a solvent unregistered company (having nothing to do with the EU or EEA) is one which is “liable to be wound up”, I concluded that it was appropriate to consider each of the circumstances in which such a company could be wound up including those applicable on insolvency. The same applies equally, in my judgment, in the case of a solvent EEA insurer. Since, as is common ground, the English court would, on Mr Sheldon’s argument, have jurisdiction to wind up an insolvent EEA insurer it follows that a solvent EEA insurer would be “liable to be wound up” since insolvency (or “unable to pay its debts”: section 221(5)(b) is one of the relevant circumstances in which winding up can take place. On that basis as well, Allianz Marine and Hibernian are, in my judgment, “liable to be wound up” even though they are solvent and even if Mr Sheldon is correct in all of his arguments in relation to the effect of Regulation 5(1).

However, that is not quite the end of the jurisdiction argument. I have so far dealt only with aspects of Mr Sheldon’s argument which are based on winding up. A separate question is whether there are any other Community rules applying to schemes of arrangement which might have a bearing on the jurisdiction of the English court to sanction them.

The Insolvency Regulation is not of relevance: it contains nothing which could apply to schemes. Nor do the jurisdictional provisions in Chapter II (Articles 2 to 31) of the Jurisdiction Regulation extend to schemes. Although that Regulation applies “in civil and commercial matters” (see Article 1) and although there is a dispute between the parties about whether the exclusion of “judicial arrangements, compositions and analogous proceedings” in Article 2(b) extends to schemes of arrangement in relation to solvent companies, I do not need to resolve that difference because, as I have just said, none of the jurisdictional rules in Chapter II is wide enough to encompass schemes of arrangement.

That leaves the Directive. The Directive extends to “reorganisation measures” as well as to “winding up proceedings”. Under Article 2, exclusive jurisdiction is given to the home Member State (ie the regulating State) “to decide on the reorganisation measures with respect to an insurance undertaking….”. The definition of “reorganisation measures” is as follows:

“measures involving any intervention by administrative bodies or judicial authorities which are intended to preserve or restore the financial situation of an insurance undertaking and which affect pre-existing rights of parties other than the insurance undertaking itself, including but not limited to measures involving the possibility of a suspension of payments, suspension of enforcement measures or reduction of claims”.

Mr Moss submits that schemes of arrangement in relation to solvent companies are not within that definition: they do not involve an intervention nor are they carried out with the intention of preserving or restoring the financial situation of an insurance undertaking. He says nothing, and needs to say nothing, about schemes in relation to insolvent companies. Mr Sheldon submits (i) that a creditors’ scheme of arrangement is an intervention (because it requires the sanction of the Court) and (ii) that such a scheme is carried out with the intention of preserving the financial situation.

I do not need to resolve those differences (which might, in any case, be matters for reference to the ECJ) since the Directive is not of direct effect: if it has not been properly implemented, that is a matter to be resolved between the UK and the Commission. What Mr Sheldon can rely on, of course, is the implementation of the Directive by the Insurers Regulations; but, unfortunately for him, that part of the Regulations which relate to reorganisation measures (see Regulation 4(1)(c) in particular) does not use any language which is wide enough to cover schemes of arrangement.

However, Mr Sheldon reminds me, correctly, that the Insurers Regulation must be interpreted in accordance with the Directive and, so far as possible, to give effect to it. On that basis, he submits that Regulation 5(1) should be given the wide interpretation for which he contends ie so that the “but for” proviso in Regulation 5(1) takes one into the Jurisdiction Regulation and, on his submission, into Article 22(2) of the Jurisdiction Regulation. In this way, jurisdiction in relation to the proposed schemes in relation to Allianz Marine and Hibernian would be allocated to France and Ireland respectively. That would be in accordance with the allocation required by the Directive assuming that a scheme is indeed a reorganisation measure as Mr Sheldon submits.

Although this results in the “correct” allocation of jurisdiction in relation to Allianz Marine and Hibernian, that is only because each of them is regulated in the State where it has its seat. Under the Directive, jurisdiction in relation to “reorganisation measures” is allocated to the regulating State whereas under the Jurisdiction Regulation (which, according to Mr Sheldon’s argument would apply) it is allocated to the “seat”: those may be different. Mr Sheldon’s argument will not always give the “correct” result. That is not, perhaps, of much significance; what is more important is what follows.

Let me accept for the purposes of argument that solvent schemes are within the definition of “reorganisation measures”. It is difficult to conceive of any construction of “reorganisation measures” which includes schemes for solvent companies but not for insolvent companies. I could see the force of the argument that one should construe the Insurers Regulations consistently with the Directive if it achieved the result that the English court was deprived of jurisdiction in relation to schemes of arrangement in relation to both solvent and insolvent EEA insurers as the Directive would, on his argument, require. But the result of Mr Sheldon’s argument is to exclude jurisdiction in relation to solvent insurers but not insolvent insurers. It makes little sense, I think, to adopt a strained construction of Regulation 5(1) to bring schemes for solvent EEA insurers in line with the Directive when to do so would leave schemes in relation to insolvent EEA insurers out of line with the Directive. And from what I have already said, it will be apparent that I consider the construction for which Mr Sheldon contends to be, indeed, a strained construction.

In these circumstances, I do not propose to address the long and difficult arguments concerning the scope of Article 22(2) of the Judgment Regulation, including reference to the Jenard and Schlosser Reports. Those arguments have consequences for companies other than insurance companies and arise, in any case, only if I am wrong on the conclusions which I have reached so far.

In conclusion on the argument about jurisdiction, the English court does, in my judgment, have jurisdiction to sanction a scheme of arrangement in relation to both Allianz Marine and Hibernian as well as in relation to Atlantic and Continental.

The law on the proper constitution of classes for the purposes of section 425 is not a matter of significant dispute. Its application to the facts of a particular case can, however, give rise to considerable difficulty; in the present case, the class issues are hotly contested. Argument has focused to a large extent on the decision of Lewison J in BAIC. The parties dispute precisely what it is that he is to be taken as having decided as a matter of law and what merely turns on an application of the law to the facts with which he was concerned. They also dispute whether the facts are materially distinguishable from those in BAIC. Although it is certainly the case that I have different policies to consider and that the expert evidence is more extensive than that which Lewison J heard, Mr Sheldon notes that some of the policies with which I am concerned are the very same policies with which Lewison J was concerned and it is also the case that some of the expert evidence in that case is also before me in a report from one expert, Dr Francine F Rabinovitz, who provided evidence in BAIC and has provided evidence in the present case.

There is a three stage process in obtaining sanction under section 425: first, an application to the Court under section 425(1) for an order that a meeting or meetings of creditors be convened. It is this stage alone with which I am concerned in the present applications and it is at this stage that the proper constitution of the voting classes must be decided upon: see Chadwick LJ in Re Hawk Insurance Co Ltd [2002] BCC 300 at [11], p 510. Secondly, the scheme proposals are put to the meeting or meetings convened following the first stage. If the requisite majorities are obtained, the matter proceeds to a further court hearing to sanction the scheme. Chadwick LJ explains that each of the three stages has its own purpose. At the first stage:

“…the court directs how the meeting or meetings are to be summoned. It is concerned, at that stage, to ensure that those who are to be affected by the compromise or arrangement proposed have a proper opportunity of being present (in person or by proxy) at the meeting or meetings at which the proposals are to be considered and voted upon.”

I would only add that, at that stage, the proper opportunity to be present requires that the classes are properly constituted and votes properly allocated.

In BAIC, after setting out part of the Practice Statement issued by Morritt V-C on 15 April 2002 concerning the convening of class meetings (Practice Statement: Schemes of Arrangement with Creditors [2002] 1 WLR 1345) Lewison J said this at paragraph 56:

“The function of the court at the first stage is “emphatically not” to consider the merits or fairness of the proposed scheme: Re Telewest Communications plc [2004] BCC 342 at p 348.”

It is no doubt correct that the issues of fairness are not, of themselves, ones which should dictate the composition of classes, such issues being best left to the sanction stage. But given the comparatively narrow scope for assessing what is and is not open to challenge at the sanction stage when the hypothetical reasonable and honest shareholder of the class concerned comes on the scene, one must be careful not to fall into the logical error of thinking that a factor which does, properly, go to class issues is to be left out of account because it is also relevant to fairness. Some of the same factors which result in a scheme being “unfair,” so that sanction should be refused, may also be highly relevant to deciding whether separate classes are appropriate in the first place.

A small number of authorities in this area are routinely referred to and passages from them routinely cited. I cannot, unfortunately, avoid yet further repetition in the light of the arguments which I have heard since it is important to identify precisely what they decided.

I can, however, take as the starting point the decision of the Court of Appeal in Re Hawk Insurance Co Ltd [2002] BCC 300 where the court considered in detail the earlier decision in Sovereign Life Assurance Co (in liquidation) v Dodd [1892] 2 QB 573. In Hawk, provisional liquidators were appointed in relation to an insolvent insurance company; they presented a petition under section 425 for approval of a scheme between the company and its creditors. The scheme was a “cut-off” scheme requiring insurance creditors to submit claims by a specified date. Claims were weighted for dividend purposes with creditors whose claims were admitted receiving a distribution of 100% of unsettled paid claims (ie Unpaid Agreed Claims in the language of the Scheme in the present case) 75% of outstanding losses (ie Outstanding Claims) and 50% of claims incurred but not yet reported to a creditor (ie IBNR claims). The scheme was unanimously approved at a meeting of creditors. The meeting was not attended by any ordinary unsecured creditors who were not insurance creditors. Arden J, of her own motion, took the point that a single class meeting might not have been appropriate and held that, since the creditors were to be treated differently for distribution purposes with different weightings attaching to different classes of claim, separate class meetings should have been held. Her decision was reversed on appeal.

Chadwick LJ, in Hawk, dealt with the decision whether to summon more than one meeting in the section of his judgment starting at paragraph 13. He said this:

“13.

The decision whether to summon more than one meeting - and, if so, who should be summoned to which meeting - has to be made at the first stage. If the matter were free from authority, I would have regarded the basis upon which that decision has to be taken as self-evident. The relevant question is: between whom is the proposed compromise or arrangement to be made?”

He went on to consider three possible answers to that question. Mr Moss suggests that the present case falls within the first answer ie it is plain that there is a compromise or arrangement between each Company and all its (insurance) creditors. I do not agree. The case is more difficult than that and it may be – this is essentially what I have to decide – that the case falls within the third answer ie where what appears at first sight to be a single compromise or arrangement between the Company and all its insurance creditors can be seen, on a true analysis, to be two or more linked compromises or arrangements with creditors whose rights put them in several and distinct classes. As Chadwick LJ put it at paragraph 23:

“…..the relevant question at the outset is: between whom is it proposed that a compromise or arrangement is to be made? Are the rights of those who are to be affected by the scheme proposed such that the scheme can be seen as a single arrangement; or ought the scheme to be regarded, on a true analysis, as a number of linked arrangements? The question may be easy to state; but, as the cases show, it is not always easy to answer. Nor can it be said that, hitherto, the courts have posed the question in quite those terms.”

But one must note the warning in paragraph 33:

“33……it is necessary to ensure not only that those whose rights really are so dissimilar that they cannot consult together with a view to a common interest should be treated as parties to distinct arrangements - so that they should have their own separate meetings - but also that those whose rights are sufficiently similar to the rights of others that they can properly consult together should be required to do so; lest by ordering separate meetings the court gives a veto to a minority group. The safeguard against majority oppression is that the court is not bound by the decision of the meeting. It is important that the test should not be applied in such a way that it becomes an instrument of oppression by a minority.”

Going back in his judgment, Chadwick LJ undertook a detailed analysis of the only relevant authority, Sovereign Life Assurance Co v Dodd [1892] 2QB 573 including citation of the well-known passage from the judgment of Bowen LJ at p 583:

“The word “class” is vague, and to find out what is meant by it we must look at the scope of the section, which is a section enabling the Court to order a meeting of a class of creditors to be called. It seems plain that we must give such meaning to the term "class" as will prevent the section being so worked as to result in confiscation and injustice, and that it must be confined to those persons whose rights are not so dissimilar as to make it impossible for them to consult together with a view to their common interest. If that be so, in considering the deed of arrangement made with the company which took over the business of the Sovereign, we must so construe it as not to include in one class those whose policies had already ripened into debts, and those whose policies might not ripen into debts for years to come; for the position of a person like the defendant, who had an ascertained sum of 2,000l. due to him from the company was entirely different from that of those policy-holders whose future was entirely uncertain.”

In relation to that, Chadwick LJ went on to say this:

“The answer, therefore, which Lord Justice Bowen may be taken to give to the question ... "are the rights of those who are to be affected by the scheme proposed such that the scheme can be seen as a single arrangement; or ought it to be regarded, on a true analysis, as a number of linked arrangements?" is clear enough. The scheme proposed may be regarded as a single arrangement with those creditors whom it is intended to bind if, but only if, the rights of those creditors are not so dissimilar as to make it impossible for them to consult together with a view to their common interest. If the rights of those creditors whom the scheme is intended to bind are such as to make it impossible for them to consult together with a view to their common interest, then the scheme must be regarded as a number of linked arrangements. In the latter case it will be necessary to have a separate meeting of each class of creditors; a class being identified by the test that the rights of those creditors within it are not so dissimilar that as to make it impossible for them to consult together with a view to their common interest.”

I comment at this stage that this test, whilst easy like many other tests to state, is not always easy to apply. The test is formulated by reference to rights rather than interests but, confusingly, the word interest (perhaps with a different meaning) creeps in at the end of the test. Take for instance a class of UK and US policyholders of an insurance company whose rights are identical but whose interests (eg because of entirely different tax consequences in the UK and in the US) are widely divergent so that one group would clearly vote in favour and one group would clearly vote against a proposed scheme. The two groups clearly form one class. But equally clearly, consultation between the two groups, if not impossible, would be pointless given their divergent interests; and yet, because it is not different rights but only divergent interests which make consultation fruitless, the test is fulfilled. That illustrates the difficulty which leads to this problem: where the rights of two opposing groups are not identical, it can sometimes be very difficult to determine whether it is the difference between their rights which leads to opposition between the groups or whether it is the divergence in their interests which does so.

Chadwick LJ has some important observations to make about Sovereign Life Assurance and what it does, and does not, decide. Thus in paragraphs 29 and 30 he says this:

“29.

……[Sovereign Life Assurance]……has been relied upon from time to time in later cases for the proposition that creditors whose rights have vested must, necessarily, be regarded as a different class from creditors whose rights are contingent. It has long been cited, in successive editions of Buckley on the Companies Acts, as authority for the proposition that "in the case of a life assurance company holders of matured policies are a different class from holders of current policies…….In my view, the Sovereign Life case is authority for neither of those propositions. On its facts the case is authority for the proposition that, in relation to the terms of the scheme in that case, a person with an existing right to set off moneys due to him under a policy which had matured against moneys owed by him to the company was not in the same class of creditors as those who had no such right. It may well be said, also, that this Court would have found, had it been necessary for it to do so, that, the terms of the scheme in that case did lead to the conclusion that those whose policies had matured constituted a different class of creditors from those whose policies had not matured; but that is because the terms of the scheme substituted for rights under policies which had matured during the life of the policy holder the rights which those policy holders would have had on death if the policies had not matured.

30.

But it will not necessarily follow, in every case, that the treatment under the scheme of vested and contingent rights, or the rights under matured and current policies, will be so dissimilar that the holders of those rights must be regarded as persons in different classes in the context of the question "with whom is the compromise or arrangement made". In each case the answer to that question will depend upon analysis (i) of the rights which are to be released or varied under the scheme and (ii) of the new rights (if any) which the scheme gives, by way of compromise or arrangement, to those whose rights are to be released or varied. It is in the light of that analysis that the test formulated by Lord Justice Bowen in order to determine which creditors fall into a separate class - that is to say, that a class "must be confined to those persons whose rights are not so dissimilar as to make it impossible for them to consult together with a view to their common interest" - has to be applied.”

That the test formulated by Bowen LJ is the test to be applied in determining which creditors fall into the same class was a proposition which Chadwick LJ considered must be regarded as settled law. But it must, he reminds us, be kept in mind that the underlying question, to which that test must be directed, is that posed by the statutory language: with whom is the compromise or arrangement to be made? Or, as he put it earlier in his judgment: "are the rights of those who are to be affected by the scheme proposed such that the scheme can be seen as a single arrangement; or ought it to be regarded, on a true analysis, as a number of linked arrangements?"

Paragraph 30 is of great importance. Chadwick LJ states clearly that the answer to the question “With whom is the compromise or arrangement made?” will depend upon an analysis (i) of the rights which are to be released or varied under the scheme and (ii) of the new rights (if any) which the scheme gives, by way of compromise or arrangement, to those whose rights are to be released or varied. As Chadwick LJ says later (at paragraph 39) it is necessary to analyse first whether the rights which are to be released or varied under the scheme are so distinct that the scheme must be treated as a compromise or arrangement with more than one class of creditor; and, second, whether the new rights which the scheme gives, by way of compromise or arrangement, to those whose rights are to be released or varied lead to that conclusion. And one sees the same approach again reflected in paragraph 51 where, after an examination of the terms of the scheme, he concluded that the provisions of the scheme did not reflect any difference in the rights to be released or varied, but also that the new rights did not fall into distinct classes and said that, applying the Sovereign Life test “neither the rights released or varied, nor the new rights given under the scheme, are so dissimilar as to make it impossible for the persons entitled to those rights to consult together with a view to their common interest”.

In Hawk, the company was insolvent. That was a very important factor. Thus, at paragraph 42, Chadwick LJ said :

“It is, to my mind, essential to have regard to the fact that the scheme is proposed as an alternative to a winding-up. There is no doubt that the company is insolvent. It has presented a petition for winding up and the court has appointed provisional liquidators. The right approach in those circumstances, as it seems to me, is to consider the position on the basis that the relevant rights are those which creditors would have in a winding up.”

So, in addressing whether creditors’ rights before the scheme are similar, one is to look at the position as it is on the facts of the particular case and to assess their rights accordingly. This is not to say that a comparison between the rights before the scheme and the rights under the scheme is not required. One sees this being done in Telewest. In that case, the dispute was whether there should be different classes for dollar and sterling bondholders. Since the reality was that the company would, absent a scheme, go into liquidation, the present rights of all the bondholders would be assessed as sterling amounts as at the date of the commencement of the liquidation. It was against that starting point that a comparison had to be made with the terms of the scheme to see whether the rights given under the scheme were such that different classes should be constituted. Further, a comparison may be necessary to see whether creditors’ rights absent the scheme are so dissimilar as to make consultation impossible, something which creditors could do only in the context of the scheme about which they are to consult. Thus, the rights which creditors are to receive under the scheme inform the answer to the question whether their rights absent the scheme are so dissimilar to make consultation, that is to say consultation on the scheme itself, impossible.

Then, after referring to certain provisions of the insolvency legislation Chadwick LJ concludes in paragraph 44:

“It follows that (but for any special rules applicable to the valuation of claims under insurance policies) the rights of non-insurance creditors, insurance creditors with unsettled paid claims, insurance creditors with outstanding losses and insurance creditors with IBNR losses are the same in this respect: that in the context of a winding up of the company they will all be entitled to submit claims in the winding up and to have those claims admitted or rejected. The difference between the position of non-insurance creditors and insurance creditors with unsettled paid claims (on the one hand) and insurance creditors with outstanding losses or IBNR losses (on the other hand) is that, in the case of the latter, their claims are in respect of debts which by reason their "being subject to any contingency or for any other reason" do not bear a certain value and so must be the subject of an estimate. But that does not lead to the conclusion that the rights of, say, non-insurance creditors and insurance creditors with IBNR losses are different. They have the same rights in a winding up. It is simply that, in order to give effect to the rights of the creditors with IBNR losses, it is necessary to estimate their value.”

I have already referred to paragraph 51 of the judgment. It is, I consider, interesting to note something else said in that paragraph: Chadwick LJ identified the common interest as being “in achieving a relatively simple, inexpensive and expeditious winding up of the company’s affairs outside a formal liquidation. It is a striking feature of this case that the creditors have, in fact, found it possible to consult together with a view to that common interest: there have been no dissentient voices”. Accordingly, the fact that the alternative to a scheme was an insolvent winding up was of significance not only because it provided a direct comparator for financial outcome, but also because it enabled a common interest to be identified.

It should not be forgotten that, in Hawk, Pill LJ also gave a reasoned judgment (the third judge, Wright J being content simply to agree with both judgments). The significance of Pill LJ’s judgment appears from what he says in paragraphs 60 and 61 since it underlines how fact-dependent each case is and gives some indication of how the court might approach the question whether creditors’ rights do in fact differ in a relevant way. Thus:

“[60]. This is not a case in which, in the event, there are creditors whose potential claims upon the fund are heavily weighted for example, towards IBNR and in whose interest it may be to seek a deferment of any proposed scheme. The overwhelming majority in value of creditors, Mr Moss tells the Court, have claims in all three categories spelt out in Clause 19.3.1 [ie Unsettled Paid Claims, Outstanding Losses and IBNR]. Mr Moss submits that the Court would have jurisdiction even if that were not the case. Mr Philip Jones, as amicus, submits that different considerations might apply if that factor had been absent. Mr Jones' caution is in my view justified. In a case where some creditors have only unsettled paid claims and others only potential claims which are incurred but not reported, different considerations might apply, especially if the state of scientific knowledge were to be such that the IBNR claims are likely to be numerous, valuable and long deferred.

[61]. I hope it is not unfair to Mr Moss to say that he did not encourage detailed analysis by the Court of the facts of the particular case, submitting as he does, virtually as a matter of principle, that the rights of the creditors are sufficiently similar to constitute them a class within the meaning of section 425(1). In my view, scrutiny of the facts is essential to the decision on jurisdiction.”

I am quite sure that in using the word “interest” in paragraph [60], Pill LJ was not confusing “rights” and “interests”. Rather, it reflects the fact that the relevant “rights” under consideration all sprang from the contracts of insurance concerned. It is not because of some special position of an individual policy that the policyholder has interests which diverge from the generality of the class. Rather, it may be inherent in the very nature of a particular right that it indicates a particular approach on the part of holders of that right to a proposed scheme (as it might be said, it is in their interests to adopt that approach) which is at variance with the approach which would be adopted by holders of a different right. Thus, it seems that Pill LJ at least (and certainly Arden J at first instance) would, on the facts of Hawk, have seen those policyholders with exclusively or mainly IBNR claims as having a different interest (in that sense) from those of those with exclusively or mainly Unsettled Paid Claims or Outstanding Losses. And a similar view must have been taken by Lewison J in BAIC. Those differences in interest derive, however, from the very nature of the rights concerned and not from some special position of the particular policyholder.

Mr Sheldon relies on the judgment of Pill LJ, suggesting that it lends support to an argument that, even in the case of an insolvent company, a creditor whose claims were limited to IBNR claims might be seen as falling into a separate class from creditors whose only claims were accrued claims. This reasoning, Mr Sheldon said repeating his argument as recorded in paragraph 84 of Lewison J’s judgment in BAIC, applied with greater force to the case of a solvent insurer.

In carrying out an analysis of a creditor’s existing rights, it is necessary to adopt a realistic assessment of those rights. Accordingly, if the relevant company is, although not actually in winding up, insolvent with no prospect of recovery, a creditor’s rights are realistically to be assessed by reference to what he would receive in a winding up and not what he is contractually entitled to receive. Thus a contingent creditor’s rights would be assessed as the dividend which he would be likely to receive as a result of proof for the value of his claim. As I have already mentioned, one sees this approach being adopted by David Richards J in Telewest (supra) at paragraph 29: current rights were to be assessed by reference to an insolvent liquidation rather than the theoretical possibility that the company might remain solvent.

I have thought it right to consider the decision in Hawk at some length not only because it is the leading Court of Appeal authority in this area, but because it is necessary to understand what was said in it in order to understand the decision in BAIC itself on which so much of the argument has focused. Before going to that case, there is one further authority which I would mention since it contains another helpful summary of the law which, coming as it does from Lord Millett (albeit as a judge of the Court of Final Appeal of Hong Kong and thus not binding on me) is deserving of the greatest respect. It is Re UDL Holdings Ltd [2002] 1 HKC 172 where, at p 184, he set out six principles which could be derived from the consistent line of authority which he had examined. Included in those were the following:

“(2)

Persons whose rights are so dissimilar that they cannot sensibly consult together with a view to their common interest must be given separate meetings. Persons whose rights are sufficiently similar that they can consult together with a view to their common interest should be summoned to a single meeting.

(3)

The test is based on similarity or dissimilarity of legal rights against the company, not on similarity or dissimilarity of interests not derived from such legal rights. The fact that individuals may hold divergent views based on their private interests not derived from their legal rights against the company is not a ground for calling separate meetings.

(4)

The question is whether the rights which are to be released or varied under the Scheme or the new rights which the Scheme gives in their place are so different that the Scheme must be treated as a compromise or arrangement with more than one class.”

There is a slight difference of emphasis in the first of those (principle (2)) from the test laid down in Sovereign Life Assurance as confirmed in Hawk. The traditional expression of the test is based on such a dissimilarity of rights “as to make it impossible” rather than “that they cannot sensibly” consult together. I do not perceive any substantive difference between the two formulations. But if there is any difference, I follow, as I must, the traditional formulation.

I come, now, to BAIC. That case has many similarities with the present case; indeed, as I have already mentioned, some of the policies concerned in that case are also concerned in the present case. However, the evidence before me – and in particular the expert evidence – is different and I must determine the class issue by reference to the evidence in this case, especially bearing in mind the observations Pill LJ in Hawk about the need for a careful scrutiny of the facts.

BAIC wrote mainly aviation direct business or facultative reinsurance and was a recognised lead underwriter in the aviation risks in which it participated. The categories of risk underwritten by it (both directly and as reinsurer) included the following: hull; war risks; liability for passengers; third parties and baggage/cargo; airport liabilities; product liability cover; personal accident cover; loss of licence and cargo all risks insurance. BAIC’s main potential liabilities affected by the scheme were claims and potential claims by policyholders in the USA under product liability and general liability insurance covering potential exposure to claims arising out of exposure to asbestos, pollution and other health hazards.

BAIC’s evidence was that about 92 per cent of its asbestos liabilities were owed to direct insureds, all of whom were resident in the USA. The remaining 8 per cent of these liabilities arose under reinsurance policies issued by it. BAIC had "current claims" arising under Scheme Business from 459 policyholders, the majority of which were in the USA and the UK, most of whom would have more than one claim.

BAIC was solvent.

The insurance and reinsurance policies underwritten by or on behalf of BAIC for the United States and Canadian aviation industry that comprise the Scheme Business were primarily "occurrence" policies, providing unlimited prospective coverage against claims relating to the risks covered by the policy (for example, asbestos, chemical and product liability) as long as the underlying act or omission happened during the relevant policy period. These long-tail claims typically arise and are asserted years or even decades after the claimant's latent exposure to the allegedly hazardous substance.

Occurrence policies (at least for these kinds of claims) were no longer available in the market at any price.

Claims under insurance or reinsurance contracts of this kind could fall into one of the three classes which I have already mentioned and which correspond to the three classes in the present case.

As in the present case in relation to the Schemes for the Solvent Companies, the scheme in BAIC did not include the whole of the business written by it.

I should set out what Lewison J included in his judgment in relation to the expert evidence. I need to do so in the light of the submissions in the present case. Mr Sheldon says that I am, in effect, bound by BAIC to decide that there must be more than one class meeting for each of the Solvent Companies. Mr Moss, in contrast, says that the facts are materially distinguishable, that I am not bound and that I should hold there to be only one class.

Under the heading “Pollution and asbestos claims” in paragraphs 14 to 19 appears the following:

“14 It was common ground that asbestos related injuries are often latent for long periods before their existence becomes known; and the most serious asbestos related diseases, mesothelioma and lung cancer, usually do not appear until 30 or 40 years after exposure. In the aviation sector, the products which are typically alleged to have caused asbestos related diseases are brakes and other friction products, together with gaskets. Exposure to asbestos in aircraft hangars can also give rise to claims. The majority of claims relevant to BAIC are made by aviation mechanics who have been exposed to these products; although family members, who have been exposed to contact with a mechanic's clothing, also make claims.

15 According to Dr Rabinovitz, the expert for the opposing creditors, aviation equipment manufacturers were not made defendants in the first or even the second wave of asbestos litigation that dates back to the 1970s. New asbestos claims, caused by exposures from the 1940s to the 1970s, are expected to continue to arise until about 2049; and aviation asbestos claims are expected to extend yet further into the future because asbestos continued to be used in aviation products into the 1980s, long after asbestos had been banned in other products. While the manufacturers of aviation products were not among the "traditional" primary asbestos defendants (most of whom are now bankrupt), the first non-traditional defendants were sued in the 1980s; and companies in the aviation industry are now being named in potentially very large and expensive claims and lawsuits. She characterises these claims as "an immature mass tort". Mr Sanders, the Company's expert, disagrees. He considers that there is a limited pool of potential claimants; and that it is obvious that, if a claim is brought, it will be brought against a well-known manufacturer of aviation products, such as Boeing, McDonnell Douglas, Lockheed, Honeywell or Goodyear. It is common ground that there was a surge of such claims in 2003. Dr Rabinovitz considers that claims of this nature are likely to increase in frequency, whereas Mr Sanders thinks that the surge was "one-off" and that an actuary, measuring probabilities, can take account of this in a number of different technical ways.

16 Dr Rabinovitz points to widely varying expressions of expert opinion presented to the US courts in asbestos cases. Whereas Dr Rabinovitz considers that the claims history in the aviation sector is too short and too small to allow reliable estimates of future liability to be made, Mr Sanders disagrees.

17 It is, however, important to note the extent of his disagreement. Mr Sanders accepts that the valuation of future claims involves a degree of uncertainty; and that the uncertainty will be greater in the case of claims that do not have a lengthy claims history. Claims of this kind go back to 1998. However, he says that the level of uncertainty relating to asbestos claims is decreasing, because of a lengthening claims history. He also says that, because of the limited pool of claimants in the aviation sector, the uncertainties are reduced on that account too. His conclusion is that the uncertainty may be estimated by applying stochastic or simulation techniques. By this approach a range of outcomes can be estimated. He considers that the range of outcomes for asbestos claims arising out of product liability in the aviation sector is unlikely to be significantly greater than the range of estimates for other asbestos liabilities.

18 Mr Powell, the actuary who was the "principal architect" of the Estimation Methodology for which the scheme provides, said:

"Estimation of asbestos related IBNR claims involves a valuation of future contingent liabilities and is, therefore, inherently uncertain. The degree of uncertainty will depend upon the impact of various external factors and I accept that wide variations might be experienced. However, the existence of such uncertainty does not render any estimation unusable, unreasonable or unfair. Where the estimation is based on legitimate assumptions which are fair and reasonable, and can be supported by evidence, fair and reasonable conclusions can be drawn. It is accepted that, with the benefit of hindsight, an estimation may be shown to undervalue or overvalue a liability."

19 None of these witnesses was cross-examined. I cannot choose between them. It seems to me that I should approach the decision that I have to make on the basis that each of them has expressed a reasonable and tenable view.”

Lewison J dealt with the class issue on the facts of BAIC at paragraphs 82 to 97 of his judgment under the heading “Were the classes of creditors correctly identified?”.

At paragraph 83, Lewison J considered the contrasting positions of policyholders with accrued claims and those with IBNR claims (although it is not entirely clear to me whether “accrued claims” is to be read as including only “unsettled paid claims” and not “outstanding losses”). In relation to accrued claims, he says (a) that under the scheme, a policyholder with an accrued claim will have that claim paid in full and (b) if the scheme is not approved, he will still have his claim paid in full; so that the measure of the claim would be the amount for which he is entitled to indemnity under the policy in respect of the known claim. The position of a policyholder with an IBNR claim would, he said, be different:

“Under the scheme he will be entitled to have his contingent claim valued. He will then be entitled to be paid the full amount of the valuation (less a discount for the time cost of money). Although a valuation of a future (and contingent claim) can be made, and may even be described as a fair valuation, it is only a valuation. It is not an indemnity. Indeed, whatever else one may be able to say about a valuation of a future contingent claim the one thing that one can say with near certainty is that, barring a miracle, the valuation will not be the same amount as the indemnity. If, on the other hand, the scheme is not approved, the Company will remain in run-off. It will pay claims as and when they arise; and the measure of the payment will be the full indemnity to which the policyholder is entitled. It may be that anticipated claims by some policyholders will never arise; in which case the Company will not have to pay. But that is what insurance is about. The policyholder bargains for the insurer to bear the risk of a contingency materialising. The insurer is in the risk business; and the policyholder is not. Unlike the policyholder with an accrued claim, who knows the extent of his exposure to that claim, the policyholder with an IBNR claim does not. The essence of the scheme is that it retransfers the risk from the insurer (who had contracted to bear it) to the policyholder (who did not). Thus the rights of a policyholder with an IBNR claim are fundamentally different under the scheme from the rights that he would have in the absence of the scheme [my emphasis].”

At paragraphs 85 and 86, Lewison J records (i) Mr Moss’s submission that the appropriate comparator was the rights that the policyholders would enjoy in a solvent liquidation and (ii) Mr Moss’s reasons for saying that the Scheme Creditors were correctly grouped into a single class. In particular, he says that Mr Moss accepted that if the scheme were sanctioned, Scheme Creditors would have different rights under the scheme from the rights that they would have had in a "no scheme world". But although their rights would be different, it would be the same difference for all of them. Accordingly, there was no need to separate them into different classes.

The judge appears to focus on the issue of the comparison between the rights of the creditors under the scheme and in the absence of the scheme. However, as I have sought to point out earlier, it is not only, or even primarily, to carry out such a comparison that it is necessary to identify what would happen in the absence of a scheme. Rather, it is necessary to do so in order to see what the existing “rights” of creditors are and thus to see whether their rights are the same or different, and if different whether they are so dissimilar that the creditors must be put into separate classes for voting purposes because they cannot consult together on the scheme. I note in this context that Mr Moss’s submissions, recorded in paragraph 86, that there should be only one class follow on from his submissions that the appropriate comparator was a solvent liquidation. In such a liquidation, all creditors would be entitled to have their contingent claims valued and be paid accordingly. There would have been no class issue arising as a result of different types of claim giving rise to different “rights” for voting purposes since all creditors would have similar rights in a solvent winding up.

Lewison J (see paragraph 88) nonetheless regarded the comparator as the starting point. This he viewed as critical to deciding whether all the policyholders form a single class. He observed that in Hawk the appropriate comparator was an insolvent liquidation because the company was insolvent, although I would prefer to say that, because the company was insolvent, the creditors’ rights were to be established by reference to an insolvent winding up and that the comparison was then necessary to see whether creditors with those rights could, in the context of the proposed scheme, consult together in their common interest. In Telewest a comparison was appropriate and the comparator was again an insolvent liquidation because that was the real alternative to the scheme. In contrast, in BAIC itself, Lewison J held that the only realistic alternative to the scheme, as things stood, was a continuing solvent run-off and that was the appropriate comparator. I agree entirely that the starting point is to establish what rights the creditors would have absent the scheme and, with respect, would agree also entirely with his conclusion that, on the facts of BAIC, it was a solvent run-off. That is also clearly the starting point for making a comparison with the rights of creditors under the scheme. One should not lose sight, however, of the question at issue “is there one compromise or arrangement between the company and the creditors, or a series of linked compromises and arrangements?” in relation to which it is necessary to address whether the rights (ie in a solvent run-off in the case of BAIC) of creditors are so dissimilar as to make it impossible for them to consult together on the scheme in question.

Lewison J then comes to consider (see paragraphs 89 and 90) what he considered that a Scheme Creditor would, in the light of the evidence before him, have been entitled to in a solvent run-off:

In paragraph 91, Lewison J said that he did not consider that the fact (if it was a fact) that policyholders may have both accrued claims and IBNR claims was of any great moment. It appears that he took that view because at most the evidence only established that some policyholders might have had both accrued and IBNR claims and that at least one of the direct insureds who voted in favour of the scheme had substantial accrued claims, but no IBNR claims. The evidence was, in any case, directed only at those who actually voted, as to which the judge observed that an analysis of the votes cast at the meeting by some 15 per cent of policyholders with known claims was not an adequate basis for safely concluding that there are no creditors who only have IBNR claims. He concluded that the “fact that a creditor may fall into more than one than one class does not, in my judgment, mean that separate classes are inappropriate”.

Then, at paragraph 92, comes a most important statement:

“In my judgment in the particular circumstances of a solvent scheme, where a solvent liquidation is not a realistic alternative, those with accrued claims and those with IBNR claims have interests which are sufficiently different as not to make it possible for them sensibly to consult together "in their common interest". In truth, they do not have a common interest at all.”

Mr Sheldon submits that this is the real ratio of the decision and that, if not technically binding on me, is one which I should follow unless I think that it is clearly wrong. He submits that Lewison J is laying down a perfectly general statement of principle and that, accordingly, whenever there is a solvent scheme where the comparator is a solvent run-off, it is never possible for those with accrued claims and those with IBNR claims to consult together in their common interest with the result that there must always be separate class meetings.

I reject that submission. The apparently general statement made by Lewison J must be read in the light of the facts before him and in particular in the light of his analysis of the rights of policyholders with different types of claim in a solvent run-off and the uncertainties to which they give rise. It was largely because of his perception, on the evidence before him, of those uncertainties that he reached the conclusions which he did. And whilst his approach will no doubt be very influential in the approach which another judge will take to cases with some similarities, each case will, as Pill LJ observes, be heavily fact-dependent. That Lewison J cannot be taken as laying down a general rule can be illustrated, I think, by two examples:

Examples such as those show that there are some limits on the principle for which Mr Sheldon contends. But once one accepts that that is the case, there is no clear test to apply in deciding what does and does not fall within the principle. The correct approach, in my judgment, is to apply the well-established principles as explained in Hawk to the facts of the case in hand, bearing in mind, of course, the approach which Lewison J took in BAIC but without being bound by a statement which, in context, I do not think is one which was intended to be of universal application.

Lindsay J, albeit on an unopposed application for sanction of a scheme, certainly did not regard Lewison J as laying down a general rule as wide as that for which Mr Sheldon contends. In NRG Victory Reinsurance Ltd (27 March 2006) at paragraph 17 he said this:

“….BAIC is not, of course, a decision that where there are both “accrued” and “IBNR” claims they invariably have to vote in separate classes. Whether separate classes for them are truly necessary will depend on a long list of variables such that what is right for one company and one scheme will not necessarily be right for another.”

He then lists a number of factors and goes on:

“Lewison J, I would think, would be surprised and even perturbed were he to find that BAIC was being treated as if it had laid down that invariably and without more a meeting mixing accrueds and IBNRs would fail the Sovereign Life test.”

Mr Sheldon points out that Lindsay J emphasised that he was not disagreeing with Lewison J but was merely dealing with what was to be done at the convening stage on the evidence before him and in the absence of any representation to the contrary. In the light of that, I do not suggest that Lindsay J’s expression of view is conclusive since he accepts that, at the sanction stage, it would still be open for any objector to take the point. Be that as it may, if he had not held the view which he did, he could not have made the order which he in fact made but would have had to order separate class meetings. Accordingly, whilst observing that I would not share Lindsay J’s view about Lewison J’s suggested state of perturbation, I do agree with his view about the effect of the decision.

Mr Sheldon complains that Lindsay J does not explain why the factors which he lists detract from the conclusion that the existence of both accrued claims and IBNR claims inevitably results in separate classes. I would have thought, however, that the reason is clear, namely that it is only following an assessment of all the facts (including the factors identified by Lindsay J) that one can decide whether or not identified groups of creditors can consult together in their common interest: it is necessary to see what divides them and what unites them which is something that can be done only by reference to that factual enquiry.

Mr Moss, however, has what is really a prior point on paragraph 92 of Lewison J’s judgment. He submits that the judge is there, incorrectly, focusing on “interests” rather than “rights”: this reference to “interests” rather than “rights” puts Re BAIC, in relation to the classes issue, in direct conflict with the Court of Appeal in Re BTR plc [2000] 1 BCLC 740 (CA). In that case, parties in a solvent scheme (in that case shareholders) with “quite different interests” but the same legal “rights” were held to have properly been placed in the same class. Mr Moss submits that the use of “interests” rather than “rights” is plainly not a slip in Re BAIC, given the repeated and obviously deliberate references to “interests”. It is, however, clear that Lewison J was well aware of the distinction, referring as he did to the correct approach when considering the class issue in relation to reinsurers at paragraph 95.

I reject Mr Moss’s approach to what Lewison J was saying. I think that the use of the word “interests” rather than “rights” is a mere slip of the judicial keyboard. I have no doubt that, had it been suggested to Lewison J that his decision was flawed because he had wrongly considered “interests” rather than “rights” he would have said that he has simply used the wrong word. Even if that is not so, I feel confident that his decision would have been exactly the same if, indeed, he had made the error which Mr Moss says he made and had been given the opportunity to correct it.

I now turn to consider the existing rights of Scheme Creditors of the Solvent Companies, which are clearly their rights in a solvent run-off, and their rights under the Scheme. The starting point in relation to the Sovereign Subsidiaries was accepted to be their rights in a solvent liquidation, on the basis that liquidation was the realistic alternative to the Scheme for those companies. A given policyholder may have claims which fall into one or more of the claims categories which have been adopted for the purposes of the Scheme in applying the Estimation Methodology. These categories are the ones which the Scheme Companies have chosen for the purposes of the Scheme; and they are the categories on which the parties have focused in the argument. Each claim, whichever category it falls into, represents, in whole or in part, the rights which the policyholder has against the relevant Scheme Company under the contract of insurance and each category is a claim which has reached a particular state of maturity.

There is a considerable amount of expert evidence which has been filed in relation to the issue of classes and which Mr Moss submits distinguishes the present case from Re BAIC, including evidence filed by the Solvent Companies which is similar to the evidence which Lindsay J. referred to in Re NRG Victory. The evidence filed in support of their applications by the Solvent Companies includes actuarial evidence from Mr David Hindley. His evidence is not independent expert evidence; he is the architect of the Estimation Methodology and the proposed Scheme Actuary.

The Opposing Creditors responded to the evidence from Mr Hindley by producing expert evidence from Ms Mary Miller, an actuary, and from Dr Rabinovitz based on her experience and work in the field of mass tort and asbestos litigation and policy. Dr Rabinovitz refers to her extensive experience in “the disciplines of economics, econometric modelling, statistics, finite mathematics and computer programming, and in the application of those disciplines to the evaluation and quantification of mass tort insurance claims”. She provided expert evidence in BAIC and, indeed, her report in that case is annexed to her report in the present case.

Further rounds of evidence produced more expert testimony from Mr Hindley, Ms Miller and Dr Rabinovitz. There was also expert evidence from Mr Stuart Shepley another actuary and from Ms Lucy Allen whose area of expertise is similar to that of Dr Rabinovitz. Mr Moss remarks that a great deal of the evidence contained in the reports of Ms Miller and Dr Rabinovitz appears to be the groundwork which has been laid for the future “fairness” arguments to be run on behalf of the Opposing Creditors at the sanction hearing stage: I have some sympathy with Mr Moss’s comment and will, myself, do my best not to be seduced by the forensic attractiveness of “fairness” points in determining the class issues.

The expert evidence is directed principally at demonstrating the level of uncertainty which exists in relation to the estimation and valuation of claims within each of the three categories. The Opposing Creditors seek to show that the difficulties and uncertainties of estimation and evaluation of IBNR claims are significantly greater than, and even of a different type from, those of Outstanding Claims. The Solvent Companies’ evidence seeks to show that in relation to both types of claim, there are difficulties and uncertainties of estimation and valuation, that those difficulties are common and that there are cases where it is more difficult and more uncertain to estimate and value an Outstanding Claim than an IBNR claim. The evidence goes far beyond that in BAIC to which I have already referred.

The importance of the issue of difficulty and uncertainty is this: the central planks in the argument of the Opposing Creditors are that Outstanding Claims and IBNR claims are so different that it is impossible for creditors of each class to consult together in their common interest and that those differences arise out of the difficulties and uncertainties in estimation and valuation. In contrast, the Solvent Companies say that there are difficulties and uncertainties in relation to both types of claim and such differences of type or degree in those difficulties and uncertainties are not so significant as to make it impossible for the creditors to consult together in their common interest.

Accordingly, the dispute between the parties depends, to a significant extent, on the extent to which the estimation and valuation difficulties differ between the different types of claim. Mr Moss, however, makes a point which is logically prior to examination of those differences. He submits that all policyholders in fact have the same (or at least very similar) rights under a solvent run-off of the WFUM Pools and that they all have the same (or at least very similar rights) under the Scheme and can, accordingly, form one class for the purposes of voting on the Scheme. In a solvent run-off, each policyholder has the same right namely a right to an indemnity in respect of any insured loss in relation, it should be remembered, to events which have already occurred (ie prior to 1993 or 1991 as the case may be): the right in respect of any particular claim will have reached a particular stage of maturity depending on whether it is an Unpaid Agreed Claim, and Additional Unpaid Claim, an Outstanding Claim or an IBNR claim; and a policyholder who has a number of claims (whether in one or more of those three categories) has a bundle of rights which make up the rights which his policy gives him. Similarly, under the Scheme, each policyholder has the same rights namely a right to have the value of each of his claims under the policy estimated and valued according to a common estimation and valuation process; again, there is a bundle or rights where an insured has a number of claims (whether in one or more of the three categories).

It is no doubt true that each policyholder, in a run-off, has a right of indemnity; and it is also no doubt true that it is possible to describe the totality of his rights of indemnity as a bundle of rights. That, however, does not lead to the conclusion that the “rights” – as that word is used in the context of the authorities dealing with class issues under section 425 – are the same rights. This can most easily be shown by taking an example:

In my judgment, it cannot be said that P1’s rights against the insurer, for the purposes of ascertaining the appropriate voting classes under section 425, are the same rights both before any claim is made and after both of Claim 1 and Claim 2 are made. Although Claim 1 and Claim 2 arise as a result of the policy and its terms, the bundle of rights which P1 possesses has changed. The bundle has certainly matured because P1, instead of having a right to an indemnity on certain future events which may never occur (eg the past exposure to asbestos resulting in a future manifestation of personal injury giving rise to a cause of action by the employee against P1), has a right to an immediate payment in respect of Claim 1 and a right to a payment once quantum is agreed in respect of Claim 2.

To say that the right remains the same at all times because it was and is a right to an indemnity is to beg the question. P1’s right to an indemnity comprises the bundle of rights which P1 possesses to receive payment from the insurer sufficient to meet P1’s own liability in respect of insured losses suffered by P1 (in the example, the employees’ claims giving rise to Claim 1 and Claim 2). It does not seem to me that the right of P1 to receive payment in relation to an unknown future claim is the same right as the right to receive actual payment of Claim 1 or, once quantified, of Claim 2. The fact that both rights can be described as an indemnity and the fact that an actual claim derives from a right of indemnity does not make them the same right.

The same point can be made in a slightly different way. Consider the position of the injured employee under Claim 2. He has a right against P1 to damages for personal injury. Before the time when he first knew he had suffered any injury – before it first manifested itself – it would be difficult to describe him as having any right against P1: the most he could say is that he had been exposed to asbestos and might develop some unpleasant disease as a result. But even if that is not correct, he certainly has no subsisting claim and his right, once the injury has manifested itself cannot, I think, be described as the same right as the one which he (might have) had before it did so. But if the injured employee’s right is not the same, I find it difficult to see how P1’s right against the insurer can be said to be the same both before and after Claim 2 is made; and that is so even though P1’s rights can at all times be described as an indemnity.

It follows that the rights of different policyholders with different mixes of claim have different rights – they do not each have the same right, namely a right of indemnity. Accordingly, I reject Mr Moss’s submission that all policyholders have, for the purposes of section 425, the same rights under run-off namely the right to be indemnified. It is therefore necessary to see whether the rights which they do have are so different as to make consultation in their common interest impossible.

So far as concerns rights under the Scheme, Liabilities which are Unpaid Agreed Claims are paid in full. In relation to Additional Unpaid Claims, Outstanding Claims and IBNR claims, a right to indemnity is replaced by a right to receive payment based on estimation and valuation, the amount being ascertained in cases of dispute by the Scheme Adjudicator or the Scheme Actuary, cases referred to the Scheme Actuary being dealt with in accordance with the Estimation Methodology. In the case of Outstanding Claims and IBNR claims, the rights given by the Scheme are thus entirely different from the rights provided in a solvent run-off whereas in the case of Unpaid Agreed Claims there is no material difference and possibly little in relation to Additional Unpaid Claims.

The question which then arises is whether the rights under the Scheme are sufficiently dissimilar to make it impossible for the Scheme Creditors to consult together in some common interest.

Mr Sheldon’s principal case is that a separate class is necessary for policyholders having IBNR claims. He does not seek to put Scheme Creditors with Unpaid Agreed Claims (or other creditors where no estimation of claims is necessary) in a separate class from those with Outstanding Claims, although I will in due course consider whether this is in fact necessary. He has a separate class issue (to which I will come towards the end of this judgment) in relation to reinsurers and possibly a further issue in relation to foreign creditors.

So far as concerns IBNR claims and Outstanding Claims he submits that the present case is on all fours with the decision and reasoning in BAIC. As in BAIC, he makes the submission that, unlike Outstanding Claims against a policyholder (where the policyholder will, by definition, have received notice of a claim against him), IBNR claims involve exposures that by definition have already occurred during the coverage periods (such as third party exposure to harmful agents eg asbestos, silica or pollutants that give rise to environmental liabilities) but which claims have not been notified to the policyholder. The common feature of these “long-tail” claims is, he says, that they typically arise and are asserted years or even decades after the claimant’s initial latent exposure to an allegedly hazardous substance.

According to the evidence of Mr Timothy Carter, Director of Risk Management at Goodrich, in some cases the claims arise very suddenly with little or no claims history. He says this:

“The estimation and valuation of IBNR asbestos, silica and other toxic tort claims in the U.S. is fraught with difficulty. The experience of the “Big 3” U.S. automobile manufacturers - which had small volumes of asbestos claims until the past few years and now have many thousands of such claims - illustrates the extent to which particular sectors of the U.S. economy have been targeted by asbestos plaintiffs at different periods of time, often in response to bankruptcy filings by other asbestos defendants…….”

This is a point which Mr Sheldon also makes, having in mind changes in the state of knowledge, by reference to what Pill LJ said in Hawk:

“…… In a case where some creditors have only unsettled paid claims and others only potential claims which are incurred but not reported, different considerations might apply, especially if the state of scientific knowledge were to be such that the IBNR claims are likely to be numerous, valuable and long deferred.”

The point is of even more force in the present case which involves solvent run-off for the Solvent Companies, in contrast with Hawk where the company was insolvent. It is a point which is also illustrated by recent developments in the US, where legal claims in relation to lead paint toxicity have, for the first time, been allowed in the Rhode Island court against manufacturers and distributors of lead paint.

Mr Sheldon points out (by reference to paragraph 9 of the judgment in BAIC and the evidence in the present case) that the policies that were acquired from the Scheme Companies by the Opposing Creditors are in all material respects the same as those in issue in BAIC. As in BAIC, the liabilities covered by the Schemes include claims from US policyholders under product liability and general liability insurance in respect of exposures to hazardous substances such as asbestos and pollution. Moreover, the policies issued by some of the Scheme Companies that form part of the business that is to be included in the Schemes are the same policies as those that were in issue in BAIC itself. And as can be seen from the description which I have already given of the Scheme in the present case, there is the same division of claims as in BAIC albeit under different names.

Mr Sheldon says that an analysis of the Scheme in the present case and an examination of the scheme in BAIC shows that everything which Lewison J said in that case is applicable in the present case. There is, on that basis, a real difference between an Outstanding Claim and an IBNR claim such that policyholders with different claims have different “rights”. He accuses the Solvent Companies of an attempt to blur the distinction. He points out that even the proposed Schemes treat IBNR Claims differently from other Scheme Claims. IBNR Claims are required to be supported by substantially more supporting evidence than all other claims and to be calculated by reference to a complex Estimation Methodology which is principally directed to IBNR Claims. This plainly reflects, he says, the substantially greater difficulties of estimating IBNR claims as compared with other claims. He relies on these factors in particular:

I now turn to the expert and other evidence. Since, like Lewison J, I have heard no cross-examination of the experts, and there is nothing which any of the experts say which I can say is definitely wrong, it is right that I should like him, proceed on the basis that each of them has expressed a reasonable and tenable view. There is however, one aspect which I would like to highlight at once: it is common ground between Mr. Hindley and Mr. Shepley, on the one hand and the experts who have produced reports for the Opposing Creditors on the other hand that, in the normal course, the uncertainties in estimating IBNR claims exceed and are materially different from those arising in the valuation of other claims including Outstanding Claims.

The first evidence in time of an expert nature came from Mr Hindley: Mr Moss, in his written submissions, has drawn my attention to a number of paragraphs of Mr Hindley’s first witness statement. I mention in particular the following:

The next evidence in time was contained in the first reports of Ms Miller and Dr Rabinovitz on behalf of the Opposing Creditors. Ms. Miller, an actuary, made a number of relevant points including these:

Dr Rabinovitz expresses similar views on some of these points:

Mr. Shepley agrees with Mr. Hindley on the relative valuation difficulties saying that “….there is general agreement amongst claims estimators that it is usually more difficult to estimate IBNR Reserves than outstanding claims reserves.”. It is to be noticed that he, too, makes this comment in the context of reserving.

The only dissenting voice is that of Ms Allen, an economist who has produced an expert report for the Scheme Companies which states that “there is not much distinction in practice between valuing Outstanding and IBNR Claims.” This is contrary to the evidence of the Scheme Actuary and the report of the Scheme Companies’ actuarial expert. Moreover, her evidence is based on the existence of common factors in the estimation of both types of claim, but even where some of the sources of uncertainty listed in the Scheme Companies’ evidence do apply to some Outstanding Claims as well as IBNR Claims, Mr Sheldon submits that the nature and degree of their applicability is substantially greater for IBNR Claims than for Outstanding Claims. I am satisfied that, as a general rule, that will be so.

The annual reports and financial statements of insurance companies of two of the Scheme Companies make it clear that the estimation of IBNR Claims are inherently substantially more uncertain as compared with estimation of other claims. For example:

As to the four passages quoted from Ms Miller, Mr Moss makes these submissions:

I shall comment on a. and b. later. In relation to c., it does not appear to me that Ms Miller is relying on the paragraphs of the Scheme Document which Mr Moss identifies in saying what she does. A reading of the preceding paragraphs of her report show that Ms Miller, as part of her analysis, considered a number of factors showing why the valuation of IBNR claims is so difficult and (making a fairness point rather than a class issue point) why the Scheme is disadvantageous to Scheme Creditors in respect of their IBNR claims. What I think she is doing, in referring to a zero valuation in some cases is simply putting the limiting point of her argument. That can be justified, I think, when it is realised that experience shows that changes in scientific or legal knowledge and principles can result in a class of claim when before those developments, such a claim was unknown, for example, the Rhode Island lead paint litigation. Mr Moss does not, in any case, make any submission at this stage of his case that Ms Miller is wrong when she says that: “For Scheme Creditors with outstanding claims… the proposed scheme merely accelerates the settlement process with Scheme Companies”.

As to d., Ms Miller does respond to the point (albeit in relation to Mr Shepley’s evidence which was to the same effect as the submission made by Mr Moss) in her second report. She says this:

“Mr. Shepley admits that there may legitimately be very different views among claim estimators on the valuation of IBNR claims, yet under the proposed Scheme, in the long run, only the views of the Scheme-appointed actuary matter. The policyholder’s principal leverage for negotiating a fair value settlement does not exist under a Scheme like the proposed Scheme, but instead is replaced by a forced commutation of extremely valuable insurance rights with no recourse to meaningful judicial review.

In this regard, Mr. Shepley’s disagreement with my prior statement that IBNR claims may not be quantifiable, and insistence that there could never be a situation of an “unquantifiable” IBNR claim, simply mean that the Scheme Actuary can put some value on IBNR claims (including a zero value) and the Scheme Creditor has to live with it.”

Mr Sheldon suggests that, notwithstanding their evidence that in the normal course it is more difficult to estimate IBNR claims as compared with Outstanding Claims, Mr Hindley and Mr Shepley have attempted to blur the distinction between these two types of claims by providing hypothetical and atypical examples of circumstances in which an estimation of an IBNR Claim could be less uncertain than the estimation of an Outstanding Claim. Moreover, he asserts, correctly in my view, that neither Mr. Hindley nor Mr. Shipley has adduced evidence of “real world” examples that are relevant to the business underwritten by the WFUM Pools, notwithstanding their long experience of estimating insurance claims and, in Mr. Hindley’s case his experience in dealing with the WFUM Pools. Mr. Hindley’s own evidence is that the examples he actually gives are of the types of claim that “might be made under the WFUM Pools”, similarly, Mr. Shepley describes Mr. Hindley’s examples as of the “type that might well occur in respect of the business underwritten by the WFUM Pools…” Neither Mr. Hindley nor Mr. Shepley has produced a single example where the Estimation Methodology, that is directed principally at IBNR Claims, would apply to an Outstanding Claim. In response to these criticisms, Mr Hunt has given some further examples which, he says are based on real cases within the WFUM Pool and which he asserts do not suffer from the defects which Mr Sheldon alleges. I set out these examples later.

Mr Sheldon says that the Scheme Companies also attempt to minimise the uncertainty attendant on estimating IBNR Claims by suggesting that the uncertainty inherent in estimating IBNR Claims is reduced by the passage of time since the Scheme Companies wrote their last “occurrence” policies in that the last live policies issued to a US Creditor expired in 1993. Mr. Hindley states that it is his understanding that the last direct policy written in the US Casualty account expired in early 1988, which was 18 years ago and the last direct policy covering a US Creditor expired at the end of 1993. He concludes that:

“…the chance of a currently universally unknown claim emerging in future which would affect the relevant WFUM Pools is remote. This is because the very first notification of such a latent claim would need to occur at least 12 years after the period of exposure…This would require a minimum latency period much longer than that for asbestos-related diseases, which are well known as having a long latency period. Therefore I regard the likelihood of any such unknown latent claims arising in the future and affecting WFUM Pools’ Scheme policies and remote and the likelihood of the emergence of any material such latent claims affecting the WFUM Pools’ Scheme policies as very remote.”

Mr. Hindley’s statement is, according to Mr Sheldon, contrary to the facts for the reasons he gives and set out below, some of which are based on the evidence before the Court in BAIC which has been filed in this case. He observes that, in this context, it is, therefore, relevant to note that BAIC ceased to write the business that it proposed to cover by its scheme in 1990. Thus no new “occurrence” policies had been written by BAIC for 15 years prior to the scheme proposed in that case.

First, it is not right that the latency period would need to be much longer than that for asbestos related diseases. According to Dr Rabinovitz, claims for injuries from asbestos exposures from the 1940s to the 1970s when asbestos products were widely used are still emerging in the US and UK. The general view is that new asbestos claims will continue to arise from these exposures until about 2049 because the most serious asbestos related diseases (mesothelioma and lung cancer) do not appear for 30 to 40 years after exposure. There are, according to Ms Miller, numerous examples (she lists asbestos, DES, Perchlorate, dioxin, silicosis, lead paint, benzene in paint, construction defects) of long-tail toxic claims that have not materialised for periods far greater than 12 to 18 years. I accept that evidence.

Secondly, in the context of US tort and environmental pollution claims it is now common, according to Ms Miller and Dr Rabinovitz, rather than remote for latent injury claims to take several decades to emerge, in some cases because of medical and scientific progress. On the evidence before me I am satisfied that:

Accordingly, Mr Sheldon submits that, if Scheme Creditors with exposures of the type described above had been required to value their IBNR Claims before the recent developments that made these claims known, their claims would have been highly likely to have been valued at zero under the Estimation Methodology. This is particularly so in the light of Mr. Hindley’s opinion that the prospect of emergence of such new claims is remote. By contrast, in a solvent run-off, these creditors would be entitled to payment in full of claims of many millions of dollars.

In addition, now that nearly all the traditional asbestos defendants are in bankruptcy, Dr Rabinovitz evidence shows that the plaintiffs’ bar in the US has started to target previously peripheral defendants. Scheme Creditors, such as Textron and Sears, are among the peripheral defendants who may face asbestos claims. It is too early to predict the likely trajectory of such claims.

Mr. Hindley continues his evidence with the statement that:

“In any case, should such an unknown latent claim emerge in future then I believe that it would be equally likely to affect all creditors regardless of whether they currently have paid, outstanding or IBNR claims, or any combination of them”

That produces criticisms from Mr Sheldon that Mr. Hindley has not adduced any empirical support for this view and that this point appears to be contrary to Mr. Shepley’s evidence which is that the WFUM Pools business covers a wide variety of types of business including marine, property, non-marine, aviation, personal and employer’s liability, which includes cover for types of claim that cannot be characterised as unknown latent claims.

He also says that Mr. Shepley also appears to overlook the fact that claims can suddenly emerge. Thus Mr Shepley says:

“My interpretation of the Scheme …is that the Scheme Actuary is likely to consider creditors’ methodologies reasonable if they satisfy 7.3 a) to e). Specifically this section implies that creditors may be able to support claims where they have minimal or no historical claims data. As long as there is a demonstrable link between the claim type(s) being claimed and the activities of the creditor, it will still be able to make and validly support a claim, normally through an exposure based approach. With the addition of the requirement that claims should be from recognised sources, this appears reasonable and would be a minimum that I would require to be able to decide a particular methodology was appropriate.”

This approach, Mr Sheldon says, fails to take account of the fact that the point of the “occurrence” policies is to protect the policyholders from claims that arise from all sources – whether recognised or unrecognised. For example, had the Schemes been promoted in 1996, Goodrich would surely have received zero value for liabilities associated with perchlorate contamination of the Rialto Site.

Further, Mr. Shepley acknowledges that there may be very different views among claims estimators on the range of eventual outcomes and the appropriate weighting to apply to these outcomes.

Mr. Hindley states that the general approach outlined in the Estimation Methodology for estimating asbestos and other mass torts can be characterised as “exposure-based” and says:

“In addition, the Estimation Methodology explicitly recognises that some creditors might have limited claims history, and outlines the approach that will be adopted to value such claims. Such an approach would not merely rely on the projection of the prior claims history…”

However, whatever method, Mr Sheldon’s submission runs, an actuary uses to value a claim, the smaller the body of known claims the more difficult it is reasonably to estimate IBNR claims. For Scheme Creditors with very little information about known claims, the valuations of IBNR Claims will vary widely. The scope for substantial variations in estimates is clear: for example the Estimation Methodology does not provide any guidance as to the exposure base for a policyholder such as Sears facing allegations that products sold by them contain asbestos or how to approach an exposure-based analysis for lead paint liability when the most substantial uncertainty is whether there will be any claims at all. Moreover, that Scheme Creditors may propose their own methodologies further emphasises the potential for substantial variations in IBNR Claims.

The estimates of IBNR Claims for the purposes of the Schemes are virtually certain not to be the same as the amount that the Scheme Creditor would be entitled to receive in a solvent run-off. In these Schemes, because of the difficulties in valuing IBNR, such Scheme Creditors with such claims would be very likely, according to Mr Sheldon, to receive less than the amount to which they would otherwise be entitled. Thus, as in BAIC, the substantial inherent uncertainty in estimating the likelihood of there being an IBNR Claim and the risk of the Scheme Creditors having inadequate resources to meet such claims is transferred from the Scheme Companies to Scheme Creditors with IBNR Claims. There is no similar transfer of risk, in the normal course, for other claims whose rights against the Scheme Companies would not be materially changed by the Schemes.

Mr Sheldon therefore concludes that there is no material distinction that can legitimately be made between the facts in BAIC and the facts in this case. The IBNR claims should be a separate class for voting purposes from Unpaid Agreed Claims and from Outstanding Claims.

Mr Moss, unsurprisingly, focuses on different passages from the second round of evidence from Ms Miller and Dr Rabinovitz (see a. to d. below), saying that they in fact add very little; and he draws my attention as well to passages in the evidence from his own side (see e.and f below) in the form of (i) further statements from Mr Hindley aimed at answering (a) the criticisms levelled at his original evidence and (b) in detail, the points made by Ms Miller and Dr Rabinovitz (ii) further expert actuarial evidence from Mr Shepley and (iii) the expert evidence of Ms Allen.

I have set out the essential points of the rival evidence at considerable length to show where the differences are and how difficult they would be to resolve. I do not think that I can, or need, to resolve those differences. Indeed, it would be undesirable to set a precedent which would encourage proponents of and objectors to schemes such as the present Scheme to adduce ever more detailed evidence on a convening application.

What the evidence relevantly establishes, in my judgment, includes a small number of propositions which are of central importance to the case:

I conclude from the evidence, in particular attaching significance to those few central points, that there is a huge range of uncertainty in estimation and valuation of claims. It is reasonably clear that Unpaid Agreed Claims are, if not completely, then very nearly completely certain. There is very considerable certainty in the estimation and valuation of the most certain of the Outstanding Claims; and there is a large uncertainly in the least certain of IBNR claims. Normally, Outstanding Claims will be more certain than IBNR claims, but there can be cases where that is not so.

If relative uncertainty is to be the touchstone of separate classes, and if the uncertainty in all claims (both Outstanding Claims and IBNR claims, and perhaps even including Unpaid Agreed Claims) forms a continuum, then there is some force in the argument that there should be only one class.

However, I have already rejected Mr Moss’s submission that all Scheme Creditors have the same right, namely the right to be indemnified and have concluded that policyholders with different mixes of claim have different rights. Accordingly, I consider that an Outstanding Claim gives the Scheme Creditor a different right from an IBNR claim (which is not to say that the former is necessarily more uncertain than the latter, but it is different). A fortiori, an Unpaid Agreed Claim gives different rights from an IBNR claim.

Further, it also seems to me that the rights of a Scheme Creditor in respect of an Outstanding Claim (and even more so in respect of an Unpaid Agreed Claim or other claim which does not require estimation) at the certain end of the range of uncertainty are so different from those of a Scheme Creditor in respect of an IBNR claims at the uncertain end of the range that it is impossible for them to consult together (in respect of those divergent rights) for the purposes of voting on the Scheme (just as it was impossible for the different classes of claimant to consult together in BAIC). It is not just that the level of uncertainty is different, but also that the uncertainty is qualitatively different. Contrasting the positions of those two policyholders at opposite ends of the range, one will know he has a claim against one or more Scheme Companies and will have a reasonable idea of its value. He has a right to recover a sum of money which simply requires quantification. In contrast, the IBNR claim is uncertain even as to its occurrence.

It might be said that the only significance of that contrast is the difference in the level of uncertainty which the uncertainty of occurrence engenders so that where the uncertainty is of the same magnitude in relation to an Outstanding Claim and an IBNR claim, they can nonetheless be put in the same class for voting purposes. That may be so, but it does not detract from the very real difference in rights where the claims are at the opposite end of the uncertainty range.

Thus, assume for the moment that there is, as Mr Moss suggests, a continuum of uncertainty and picture it with the most certain of Outstanding Claims at the low end and the most uncertain of IBNR claims at the other. One could imagine the Outstanding Scheme claims of each Scheme Creditor and the IBNR claims of each Scheme Creditor being set out in order on that continuum, each Scheme Creditor having a different mix of the two types of claim. It may be that some Outstanding Claims appear higher up the continuum than some IBNR claims but that, according to the evidence, will not be the normal case. The evidence is that, for the normal case, Outstanding Claims are more certain in estimation and valuation than IBNR claims so that, accordingly, the continuum will appear with most Outstanding Claims lower down than most IBNR claims.

Now, it may also be that the rights of a particular Scheme Creditor are not so different from the rights of his immediate (and perhaps more distant) neighbours on each side as to make it impossible for them to consult together on the Scheme. On that basis it can be said that there should be only one voting class. Any class separation based on uncertainty – the only criterion by which, according to Mr Moss, it is sought to divide the Outstanding Claims from the IBNR claims - would have to be based on something different from a division into Outstanding Claims and IBNR claims, otherwise claims of each type would fall into the wrong classes based on uncertainty. In any case, it is clearly practically impossible to rank the Scheme Claims of all Scheme Creditors in order of uncertainty. Moreover, if uncertainty is the criterion, then a division based on different types of policy (ie a difference in the type of cover provided) might be more appropriate: but nobody before me has suggested a division on that basis.

However, if there is a single class for voting purposes, then that will result in Scheme Claimants having to vote together in respect of widely divergent claims at opposite ends of the continuum where, for reasons already given, it is not to be supposed that it is possible for them to consult together. Even if it were possible to put a gloss on the Hawk test so as to be able to ignore a de minimis group (either in terms of number or value) the present case cannot be assumed to be such a case. Whilst it might not be correct to place most Outstanding Claims at the low extreme of the continuum or most IBNR claims at the high extreme, the evidence is, in my judgment, sufficient to establish that many Outstanding Claims will be of a sufficiently different degree of uncertainty from many IBNR claims as to make consultation between the relevant Scheme Creditors in respect of those claims impossible.

Accordingly, it seems to me that the appropriate course is to place Outstanding Claims and IBNR claims into different categories and for separate classes to be constituted in respect of those different categories. It may be that this will result in some, atypical, claims being placed in the “wrong” category if the only criterion were uncertainty. However, there is nothing in the evidence before me which would establish that, even given the different levels of uncertainty which exist, it would be impossible for all Scheme Claimants to consult together in respect of their Outstanding Claims or, separately, in respect of their IBNR claims. What the evidence does suggest is that it would be impossible for many of the Scheme Claimants in respect of their Outstanding Claims, to consult together with many other Scheme Claimants in respect of their IBNR claims thus making a single class inappropriate.

Whether Unpaid Agreed Claims (and other claims which do not require estimation) should fall into a separate category and give rise to a separate class from Outstanding Claims is another question. These claims are (a) not significant in the overall picture in terms of number or value and (b) are intended so far as possible to be paid off before the meeting or meetings of creditors are held. Mr Moss says, in reliance on the decision of Lloyd J in Equitable Life Assurance Society [2002] BCC 319 that the court can refuse to carve out some little group of people and give them a veto. This should be particularly so where that group has no real interest one way or the other and is going to be paid in full whatever happens. Mr Sheldon has not pressed for a separate class. I am not going to require one; but instead will direct that these Scheme Creditors vote in the same class as Scheme Creditors in respect of their Outstanding Claims. I see no reason why, although their rights are different, they should not be able to consult with Scheme Creditors with Outstanding Claims in respect of this Scheme in their common interest. Similarly, other claims which do not require estimation should fall into the same category. Further, I see no reason why Scheme Creditors with Additional Unpaid Claims should not be included in this class also in respect of those Claims.

In my judgment, therefore, two separate classes should be constituted in relation to Solvent Companies to enable Scheme Creditors to vote (i) in relation to their Unpaid Agreed Claims, other claims not requiring estimation, Unpaid Additional Claims and Outstanding Claims and (ii) in relation to their IBNR claims.

Some Scheme Creditors of the Scheme Companies are also reinsurers of one or more Scheme Companies. Under the Scheme (other than in relation to Sovereign itself), reinsurers remain entitled to vote with a time discount (as with other Scheme Creditors) to reflect the time value of money. However, there is to be no set-off in respect of amounts which may become due from contingent Scheme Creditors as reinsurers to the Scheme Creditors. The evidence shows that Scheme Creditors who are also reinsurers are significant in terms of number and value. In contrast with this absence of set-off for voting purposes, claims under the Scheme will be admitted only after taking account of set-off.

The issue whether reinsurers should constitute a separate class was addressed by Lewison J in BAIC; he found it to be a difficult question and decided, accepting Mr Moss’s submissions, that a separate class was not required:

“Mr Moss submitted that, legally, both direct insureds and reinsureds constituted a single class of unsecured creditors, each of whom had the same right of indemnity against the Company in respect of insured risks. It is true that set off will be applied to those reinsured who are also reinsurers; but it is only those reinsureds who, after the application of set-off, were adjudged to be net creditors of the Company at the time of the scheme meeting who were admitted to vote. As net creditors of the Company, it did not matter how the net position arose. They also had the same economic balance to make as the direct insureds. In each case it was a question of choice between certain cash now and uncertain cash later. If the reinsureds choose cash now (as overwhelmingly they did), they will recognise that they are losing their reinsurance cover.”

Mr Sheldon does not seek to say that Lewison J was wrong. But what he does say is that the present case is different. In BAIC, reinsurers were entitled to vote only on the net value of their claims ie after set off; in contrast, in the present case, Scheme Creditors who are also reinsurers will be permitted to vote without set-off. Mr Sheldon says that this will have a grossly distorting effect on the vote at the meetings. The vote will not reflect the economic interests of creditors in the companies. Nor will it reflect their entitlements under the schemes if they become effective. By proposing to exclude the application of set-off for the purpose of voting, the votes cast by creditors who are also reinsurers will, notwithstanding their different rights and interests, count disproportionately towards the final result, unless they are constituted as a separate class.

Mr Moss says that, whilst the approach in BAIC was not wrong, it was not the only one; the approach in the present case is also a permissible one. He says that the question for me is not whether I prefer the BAIC approach but whether it is proper for a Scheme Company to adopt that approach. The suggested justification for the Scheme Companies’ approach is this. Reinsurers have a right to payment of sums due to them in a solvent situation without a set-off. As a matter of general principle set-off is only for allowed for liquidated and ascertained claims in a solvent situation and that is what the proposed voting arrangements reflect. It is said that this is a perfectly proper way to proceed and that at the voting stage that the reinsurers are in the same position as they are under the ordinary law. Mr Moss points out that the reinsurers are not before the court and that they might say that, under the BAIC formula, they are being disadvantaged because their votes are not being given the same weight as they are entitled to under the general law. If that is wrong, his clients are willing to amend the Scheme so that voting is based on the value of claims after set-off.

Like Lewison J, I have found this a difficult issue. I accept Mr Moss’s submission that the question is not which approach I would adopt if the decision were for me. I should allow the proposals of the Scheme Companies to take effect unless I am satisfied that their approach is not a proper one. On that question, I think that Mr Moss is right in saying that the Scheme Companies’ approach is a proper one. It seems to me that the rights of a Scheme Creditor do not depend on whether he is also a reinsurer where his liability as such reinsurer remains contingent. His concerns, as a reinsurer, may affect his voting decision but that reflects his interests (as reinsurer) rather than his rights (as an insured). The position is different where his liability is not contingent and, in those circumstances, the Scheme provides for set-off for voting as well as for payment. Accordingly, I do not consider that the proposed vote allocation is such as to make it impossible for the reinsurers within a particular class of Scheme Creditor to consult with other members of that class. In my judgment, therefore, it is not necessary to constitute a separate class for reinsurers.

Mr Sheldon submits that certain foreign creditors should constitute a separate class. His argument is succinctly recorded in his Outline Opening Submissions:

“The Schemes, were they to be approved, would only be binding in this jurisdiction. Any creditors that are subject to the jurisdiction of this Court can be prevented from enforcing their policies other than in accordance with the terms of the Schemes. Foreign creditors, to the extent that they have contracts governed by the law of a foreign jurisdiction would not have had their rights varied or discharged by the Schemes and may take proceedings against the assets of the Scheme Companies that are not in this jurisdiction as the policies would be enforceable outside this jurisdiction. The fact that creditors have contracts governed by different laws is relevant when determining whether these creditors should be treated separately from those whose who are bound by the Schemes. In the present case it would appear (a) that many of the policies written by the Solvent Companies are governed by the laws of foreign jurisdictions; and, (b) many of those companies have assets abroad. In circumstances where certain creditors have the right to recover payment in full abroad, even if the proposed schemes were to become effective, it is submitted that such creditors should form a separate class from those who would be bound by those schemes.”

I reject that submission at least insofar as it concerns creditors who can be bound as a result of an order of the foreign court. In the present case, a large number of Scheme Creditors are United States creditors and if and to the extent that their contracts are governed by a law of a State (some are I understand governed by the law of Ohio) it is hoped and expected that an order will be obtained, save in the case of Atlantic, under Chapter 15 of the United States Bankruptcy Code with the result that the creditors will be bound. Given that expectation, there are no sufficient grounds for excluding those creditors from the class in which they would otherwise vote. So far as concerns Atlantic, in relation to which Chapter 15 protection cannot be obtained, the Scheme makes special provision (see paragraph 21 above) designed to ensure that, in practical terms, the same economic consequences will follow for Atlantic and its Scheme Creditors as for other Scheme Companies. Apart from the US creditors, there is no evidence of any other foreign contract that might be subject to Mr Sheldon’s objection; nor have the Scheme Companies been given the opportunity – this point being raised at a late stage – of dealing with any relevant ripostes on the evidence. Mr Sheldon has not persuaded me that the rights of any foreign Scheme Creditor in this category are so different from the rights of other Scheme Creditors that it is impossible for them to consult together in their common interests as creditors in each class.

There is one discrete issue which arises in relation to Sovereign. Some of Sovereign’s insurance and reinsurance business was assumed and ceded through numbered underwriting syndicates at Lloyd's of London (‘Lloyd's’). On the Original Sovereign Scheme, Scheme Claims against Sovereign by each Lloyd's syndicate had one vote in number and a vote value as allocated, subject to the Chairman's discretion. Sovereign seeks the direction of the Court to treat votes from Lloyd's syndicates in a similar manner on the proposed Scheme. This would result in Equitas Limited having approximately 230 votes and being valued in accordance with their claim with reference to the Vote Adjudication Protocol, subject to the Chairman's overriding discretion. I am willing to make that direction.

Having dealt with the disputed issues of jurisdiction and class meetings, I do not understand there to be any dispute or difficulty over the notice, timing and conduct of the Scheme Meetings which will now need to be convened.

I will make appropriate directions, as well as initialling documentation, following the handing down of this judgment.