Sompo Japan Insurance Inc

On 8^th February 2011 I made an order under Part 7 of the Financial Services & Markets Act 2000 (“FSMA”) sanctioning a scheme (“the Scheme”) for the transfer of certain insurance business from the UK branch of Sompo Japan Insurance Inc (“Sompo”) to Transfercom Ltd (“Transfercom”). I did so at the conclusion of a hearing attended by the applicant Sompo and by the Financial Services Authority (“FSA”) which, after due consideration, did not object to the Scheme in any way.

When the Scheme first came before the court for approval in March 2010, objections had been made by four policyholders with policies constituting in the aggregate more than 40% by value of those affected by the transfer contemplated by the Scheme. Directions were given by Floyd J on 26^th March 2010 for written statements of objection, further evidence, including expert evidence from the independent expert, a further report by the FSA, and other case management directions sufficient to ensure an effective determination of Sompo’s application. That process led to three of the four objectors withdrawing their objections, leaving only Axa Corporate Solutions Assurance (“AXA”) which, while not withdrawing its objections, decided not to appear or be represented at the hearing on 8^th February 2011.

Being satisfied by the end of the hearing, after reading the evidence and hearing submissions from Mr Martin Moore QC for Sompo and Mr Jonathan Nash QC for the FSA, that I should sanction the Scheme, and in order to save those parties and their substantial legal teams from attendance on a second day, I made the order sanctioning the Scheme there and then, but stated that I would provide written reasons for doing so thereafter. My reasons now follow.

Section 111 of, and Schedule 12 to, FSMA require that, before sanctioning an insurance business transfer scheme, the court must be satisfied as to a number of jurisdictional and procedural requirements. Section 111(3) provides that:

The principles to be applied by the court in considering whether to exercise its discretion by sanctioning such a scheme are well settled: see Re The London Life Association Ltd (unrep) 21^st February 1989 per Hoffmann J, Re AXA Equity and Law Life Assurance Society plc & anr [2001] 2 BCLC 447, per Evans-Lombe J, in particular at paragraph 6, and Re Hill Samuel Life Assurance Ltd [1998] 3 All ER 176, at 177, in the following summary by Rimer J:

In the present case (as in most others) the main classes of persons affected by the Scheme are:

Sompo and its shareholders;

Transfercom and its shareholders;

policyholders whose re-insurer will change from Sompo

to Transfercom under the Scheme (“the transferring policyholders”);

the remaining policyholders of Sompo;

the existing policyholders of Transfercom.

The requirement that the Scheme should operate fairly as between different classes of affected persons may not be met if, for example, its effect is materially to benefit one class at the expense of another. Generally, and in the present case, objections are commonly made by transferring policyholders that the commercial interests of the proponents of the scheme (here Sompo and Transfercom) are being unfairly advanced at the expense of the transferring policyholders by substituting a less well resourced insurer for the insurer of their choice. An objection of this type calls for close examination, both by the independent expert and by the court, of the question whether the security level for the transferred policies is likely to be reduced by the implementation of the Scheme.

The court has nonetheless recognised that such a reduction in the security level of transferred policies is not automatically and in every case unfair to the transferring policyholders, either viewed separately, or in the context of the realisation of some commercial objective of the proponents. The strict regulation of the conduct of insurance business in the UK, supervised by the FSA, requires specified levels of security to be maintained by insurers but, provided that those levels are not adversely impacted, an insurer is in principle at liberty (subject to any other constraints imposed by company law) to reduce its resources, for example by distribution to its shareholders, or by deployment in other business activities which may adversely impact upon its credit rating: see in particular Re Norwich Union Linked Life Assurance Ltd [2004] EWHC 2802 Ch per Lindsay J at paragraphs 14-15.

In most cases, including the present, the court is not concerned with any question of unfairness to classes (i) or (ii). The Scheme is the product of a commercial bargain between Sompo and Transfercom, which it may be assumed is perceived by their respective boards of directors to be beneficial to each of them. All the objectors come from class (iii), namely the transferring policyholders and, before the withdrawal of three out of four of them, they represented a large proportion of that class by value. Even the remaining objector AXA represents, so I was told, 14.9% of that class by value. As for class (iv), the business being transferred out of Sompo by the Scheme is such a small proportion of its total insurance business that the effect of the Scheme on Sompo’s continuing policyholders will, in the unchallenged opinion of the independent expert, be de minimis. As for class (v), the expert’s opinion (again unchallenged save perhaps as to quantum) is that Transfercom’s existing policyholders will if anything obtain a marginal benefit in terms of increased security if the Scheme is implemented. It follows that the court’s present concern is only with the question whether the Scheme is fair to the transferring policyholders, and whether any benefits of the Scheme to its proponents and to Transfercom’s existing policyholders are being unfairly conferred at the transferring policyholders’ expense.

The business to be transferred by the Scheme mainly comprises reinsurance contracts that were underwritten on behalf of Nissan Fire & Marine Insurance Co Ltd (“Nissan”) by Fortress Re Inc between 1981 and 2003. Sompo was formed upon the merger in July 2002 of Nissan and Yasuda Fire & Marine Insurance Co and inherited Nissan’s insurance business upon the merger. The book written by Fortress Re consisted mainly of aviation reinsurance and property catastrophe business, together with some general marine and non-marine reinsurance and retrocessional business. Nissan’s share varied from year to year between 7% and 26%.

Due mainly to the events of 11^th September 2001, and to the air crash over Queens in New York later that year, Nissan’s book proved to be seriously loss-making. It has been in run-off since 2003. By April 2010 Sompo had paid claims in excess of US$1.353 billion against an estimated gross liability of US$1.573 billion. As at 31^st March 2009 the transferring policies had claims reserves of approximately US$237.1 million.

Sompo’s principal place of business is in Tokyo. It has a market share of about 20% of the Japanese domestic market for non-life business, and credit ratings of AA3 (Moody’s) and AA− (Standard & Poor’s). The business to be transferred has been carried on from Sompo’s UK branch in London.

As is inherent in aviation and catastrophe reinsurance, the transferring business is relatively short tail. Sompo’s participation consisted mainly of excess of loss reinsurance within the bracket between US$50 million and US$500 million. Sompo’s enjoys substantial protection from claims for latent losses from its US$50 million lower limit of liability and, where aggregation was permitted (mainly before 1985), the exhaustion of its liability for large claims, many of which have exceeded the upper limit of US$500 million. Thorough research by the independent expert Mr Graham Fulcher of Watson Wyatt Ltd has satisfied himself that exposure to asbestosis claims is minimal, limited to a few hundred thousand dollars, that the book is stable, conservatively reserved, and expected to run off mainly within to two to three years.

Due to the time which has elapsed between March 2010 and the present date, Mr Fulcher has had, and has used, the opportunity to review Sompo’s reserves and expectations as to the continuing run-off of the book over time. Those expectations have been fortified by that review. Further, the conservative nature of Sompo’s reserves has been demonstrated by the fact that on no occasion has Sompo’s liability to any transferring policyholder been determined or settled in an amount greater than that reserved in respect of it.

The 10,374 policies which, as at March 2009, comprised the transferring business are governed by a number of different governing laws. Sampling exercises undertaken on behalf of Sompo, and described in the evidence, suggest that some 47% of those contracts, measured by value of reserves, were governed by English law, 7% by US Law and 39% by the laws of EEA member states other than England.

Transfercom is a recently incorporated wholly-owned subsidiary of National Indemnity Company (“NICO”) a US corporation of which the ultimate parent company is Berkshire Hathaway Inc. NICO has a credit rating of AA+ (Standard and Poor’s) and, as at December 2009, total assets in excess of US$79 billion, with free capital after liabilities as at 30^th June 2010 of US$61 billion.

Transfercom was incorporated in England in April 2006 as NICO’s intended vehicle for the acquisition of another insurance book from Sompo, the transfer of which was approved by David Richards J in March 2007: see Re Sompo Japan Insurance Inc [2007] EWHC 146 (Ch). It was a more varied book than the subject matter of the present Scheme, with a significantly longer tail, and included substantial asbestosis exposures. It has gross technical provisions of US$346 million, with an anticipated run-off period substantially longer than the transferring business.

Leaving aside what is proposed in relation to the Scheme, Transfercom’s capital resources consist of the following:

The reinsurance policy is written on terms which, in effect, transfer from Transfercom to NICO the whole of Transfercom’s risk under its existing business, subject to the upper limit which I have described. It is therefore the main asset out of which Transfercom expects to discharge its ongoing liabilities to its existing policyholders, while retaining sufficient additional liquidity with which to ensure that, (as has been the case thus far), all claims by its policyholders are duly paid on time.

Transfercom has agreed with Sompo that, if the Scheme proceeds, it will fund the run-off of the transferring business in the following manner:

Transfercom will enter into a new reinsurance treaty with NICO (“the new reinsurance”) with an upper limit of US$277.1 million together with US$25 million for unallocated loss adjustment expenses, written on terms which transfer the whole of Transfercom’s risk in relation to the transferred business to NICO subject to that limit, and which prevent NICO from avoiding or cancelling on grounds of non-disclosure or misrepresentation, falling short of fraud or deceit, by Transfercom.

By way of endorsement, and to provide additional security and liquidity to Transfercom (“the funds withheld endorsement”) the whole of the premium payable by Transfercom for that reinsurance (derived from the amount paid by Sompo to Transfercom for taking on the transferring business) will be retained in a special account by Transfercom (or at NICO’s election in a trust account) for use in paying claims. That premium is in excess of 90% of the current outstanding losses and IBNR attributable to the transferring policies. The funds withheld endorsement permits NICO to withdraw amounts in excess of 102% of the aggregate of Transfercom’s attributable loss reserves and expected operating expenses in relation to the transferred business from time to time. In return Transfercom will pay NICO interest on the funds retained at a rate of 1% over the US dollar Treasury Bill three month rate from time to time.

NICO will raise the limit in its existing reinsurance of Transfercom’s existing business (“the existing reinsurance”) by US$100 million (from US$482 million to US$582 million). The amount of US$100 million has increased during the formulation of the Scheme from an originally proposed US$75 million.

By way of endorsement to the existing reinsurance, NICO undertakes to place in trust for Transfercom funds or securities sufficient to enable Transfercom to discharge its liabilities to its existing policyholders, in the event that NICO’s credit rating falls below BBB+ (“the ratings trigger endorsement”).

Sompo’s purpose in promoting the Scheme is to divest itself of the continuing management and residual uncertainties of a loss-making book, having no synergies with the rest of its much larger insurance business, where it regards the continued running-off of that book as being outwith its general business strategy going forward. Transfercom’s and NICO’s commercial purpose in acquiring the transferring business has not been specified in the evidence. It has agreed to acquire it for a reverse premium which, for reasons of commercial sensitivity has not been disclosed either to objectors or to the court, but which has been disclosed to Mr Fulcher. It is a fair inference that NICO and Transfercom anticipate that the acquisition and completion of the run-off of that book will be a profitable business activity, and that the amount of the premium payable by Sompo has been negotiated by reference to the two companies’ probably different perceptions as to the economics of completing that run-off.

Mr Fulcher’s primary focus, both in his research and in his series of four successive reports, made respectively in January, March and November 2010, and finally in January 2011, has been upon the likely effect of the Scheme upon the security or confidence levels affecting each of the three classes of policyholder which I have already described, namely the transferring policyholders, Sompo’s continuing policyholders, and Transfercom’s existing policyholders. He has taken as his benchmark the FSA’s Individual Capital Assessment (“ICA”) solvency criterion of 99.5% value-at-risk over a one year time horizon which, in layman’s terms, implies a 99.5% probability that the insurer will be able to meet its liabilities in full during the stated period. Mr Fulcher’s evidence (which the FSA does not challenge) is that non-life insurance firms in run-off may properly use, in the alternative, a time horizon constituted by the total outstanding duration of the business until run-off is completed, with a correspondingly reducing series of required confidence levels, reducing by 0.5% per annum until year five, and remaining constant thereafter. Thus for a time horizon of five years or greater, such a firm may satisfy the FSA’s regulatory requirements by demonstrating a 97.5% confidence level.

Against that yardstick, Mr Fulcher’s opinion in his first report (in January 2010) may be summarised as follows:

The transferring policyholders’ confidence level while remaining with Sompo was substantially in excess of the 99.5% one year level.

If the Scheme was implemented, then there would be a 97.5% confidence level in relation to all Transfercom’s policyholders (i.e. both its existing policyholders and the transferring policyholders) being paid in full over the whole period of the combined run-off.

Mr Fulcher was taken to task by the objectors for having addressed the security level of the transferring policyholders after implementation of the Scheme by an aggregate approach to all Transfercom’s policyholders, rather than to the transferring policyholders viewed separately. In his second and third supplementary reports (in November 2010 and January 2011) Mr Fulcher has (in part with the benefit of further information and research) addressed that issue directly. His current opinion may be summarised as follows:

Over the two year average anticipated run-off period for the transferring policyholders, their level of confidence while customers of Sompo is 99.88%.

On the implementation of the Scheme, the level of confidence attributable over the same two year mean run-off period for the transferring policyholders as customers of Transfercom is 99.6%.

His opinion as to the level of confidence, if the Scheme is implemented, of Transfercom paying all its policyholders (existing and transferring) remains unchanged, at 97.5%.

Mr Fulcher very properly acknowledges that in the event of an insolvency of Transfercom, then the pooling of its resources to meet the claims of its unsecured creditors would mean that the transferring policyholders could not have any assurance that their prospects of payment in full would be unaffected by any adverse experience in the run-off of the business relating to Transfercom’s existing policyholders. Nonetheless his opinion that it is proper to attribute a significantly higher level of confidence to the transferring policyholders is based upon his (unchallenged) perception that, since their policies will run off substantially earlier on average than those of Transfercom’s existing policyholders, it is the transferring policyholders which will have first access to Transfercom’s free capital, to the extent necessary to make good any shortfall in the new reinsurance from NICO as against liabilities to transferring policyholders.

None of the objectors, let alone AXA as the only objector which has not withdrawn, have provided any evidence or argument which calls into doubt Mr Fulcher’s opinion as to the marginal reduction in security (now only 0.28%) which the implementation of the Scheme would impose upon transferring policyholders. The consequential confidence level of 99.6% over the full expected mean run-off period for the transferring business is itself significantly in excess of the FSA benchmark. There is in my judgment no basis upon which the court could, let alone should, second-guess that unchallenged expert assessment.

Viewed separately, I do not regard a reduction in the confidence level attributable to the transferring policyholders from 99.88% to 99.6% as material, and I accept Mr Fulcher’s opinion that, having regard in particular to the FSA benchmark, the 99.6% chance of being paid in full remains a satisfactory level of security for the transferring policyholders.

I must however consider the objectors’ separate case that the Scheme is unfair to them because, in addition to reducing their security, it creates a corresponding increase in the security available for Transfercom’s existing policyholders. In that context, Mr Fulcher’s opinion is that the implementation of the Scheme would increase the confidence level for Transfercom’s existing policyholders from 95.2% to 97.5%. That might be a consideration of significant weight if the existing policyholders’ benefit was being obtained at the expense of the transferring policyholders. But the evidence shows that it is not. The improvement in the security of the existing policyholders is attributable, so far as I can ascertain, to the combination of the US$100 million increase in the limit of the existing reinsurance from NICO and the addition of the rating trigger endorsement to the existing reinsurance. Those are contributions to Transfercom’s resources made by its parent company NICO which, although conditional upon the implementation of the Scheme, are not in any sense taken from, or at the expense of, resources otherwise available to meet the claims of the transferring policyholders. Since their business is shorter-tailed than that of the existing policyholders, any increase in the prospect of the existing policyholders being paid in full will in practice be attributable to resources available to Transfercom after the transferring policyholders have all been paid in full.

I have dealt at length with the objection based upon a reduction in the transferring policyholders’ security because I consider that it constituted what was potentially the most significant of the list of objections. I can therefore deal with the remaining objections more briefly.

The starting point is that all the objections originally raised in and after March 2010 have been addressed by Mr Fulcher in comprehensive detail in his second and third supplementary reports. This process has apparently led three out of the original four objectors to withdraw their objections. I have already described how Mr Fulcher has addressed the criticism that the post-Scheme security position of transferring policyholders was not originally analysed separately from that of Transfercom’s existing policyholders. He has, similarly, addressed criticisms about the quality of the new reinsurance from NICO, in particular by obtaining independent legal advice as to the extremely limited circumstances in which it might be avoided.

He has dealt to my satisfaction (and without subsequent challenge) with suggested doubts about his independence, and criticism of his reliance upon other research and review carried out on behalf both of Sompo and Transfercom. In the context of large scale and complex insurance business, it is in my judgment both inevitable and acceptable that the independent expert should place appropriate reliance, after sufficient review, upon the analytical work of other qualified professionals, so as to avoid what would otherwise be the unacceptable consequence that his own work would become disproportionately time consuming and expensive. I am satisfied that, to the extent that Mr Fulcher has placed reliance on the work of other experts, that he has done so only after his own sufficient review of their work.

Criticism was made that Sompo either had no rational commercial purpose for the transfer, or that its purpose was to achieve finality at the transferring policyholders’ expense. I consider that a desire to achieve finality is itself a rational commercial purpose, and I have already concluded that Sompo would not by the Scheme achieve it to any material extent at the transferring policyholders’ expense.

Criticism was made that the Scheme would expose the transferring policyholders to risks arising from future underwriting business undertaken by Transfercom. As to that, there appears to be no reason to suppose that Transfercom will endeavour to establish new insurance business of its own (rather than acquire run-off business from other insurers). To the extent that it seeks to acquire further books in run-off by transfer either from Sompo or from other insurers, such transfers will themselves be subject to scrutiny and sanction under FSMA. The FSA itself, the independent expert and the court will all be obliged to consider whether such transfers are unfair to Transfercom’s existing policyholders, which would in such an event include the transferring policyholders under the present Scheme.

Administration of the transferring business is not proposed to be changed by the Scheme, since it is already administered by Resolute, a member of the Berkshire Hathaway group, and this will continue after the Scheme is implemented. It is suggested that transferring policyholders are thereby exposed to a potential conflict of interest between Resolute and Transfercom. In my judgment that objection is without substance.

Complaint is made that by the enforced substitution of Transfercom for Sompo as reinsurer, some policyholders may incur an obligation to make bad debt provisions due to a concentration of their reinsurance credit risk with the Berkshire Hathaway group. The only objector to provide any specific particulars of that otherwise very general concern has since withdrawn its objections.

Criticism was made of Mr Fulcher’s original report, in relation to a supposed failure to ascertain whether Transfercom was taking sufficient steps in anticipation of the introduction by the FSA of what is known generally as Solvency II, as a new security benchmark for insurers. Mr Fulcher’s later reports satisfied me that the Berkshire Hathaway group was taking sufficient steps in that regard, both generally and in relation to the implications for NICO and Transfercom.

One objector (which has since withdrawn) complained that it had exposures on certain treaties with persons in countries currently on the list maintained by the US Office of Foreign Assets Control (“OFAC”) or countries which may in the future be included on that list, such that they might suffer prejudice if their reinsurance in respect of such exposures was transferred to a group domiciled in the USA, and therefore liable to comply with the OFAC Rules. Mr Fulcher took independent legal advice on that question, which demonstrates to my satisfaction that this is an essentially theoretical concern. None of the objectors have adduced evidence to show that they have exposures in the only country in relation to which there could be an OFAC problem which affected Transfercom, namely Cuba. To the extent that NICO might be more widely affected, the funds withheld endorsement in the new reinsurance seems to me to provide sufficient protection.

Finally, objection was made that, since probably a majority in value of the transferring policies were governed by laws other than that of the UK, there was a serious risk that the Scheme would be legally ineffective to bring about a substitution of Transfercom for Sompo as reinsurer. This is a concern which was ventilated before David Richards J in connection with the 2007 scheme between Sompo and Transfercom. His view was that, provided the evidence showed that a sufficient proportion of transferring policies were governed by English law (and therefore by FSMA), so that the court’s sanction of the scheme could not be characterised as acting in vain, then the risk that transferring policies governed by non-English law might not be validly altered as to reinsurer by the scheme, was not a persuasive reason to refuse sanction.

In principle, I agree, provided that a risk as to the partial ineffectiveness of a given scheme does not undermine the independent expert’s analysis in relation to issues such as security for policyholders, so as to introduce some distinct element of unfairness or unacceptable level of uncertainty. In the present case Mr Fulcher has responded to this criticism by pointing out that, since the upper limit of the new reinsurance by NICO is not to be attenuated in the event that, in relation to one or more policies, the Scheme proves ineffective to substitute Transfercom for Sompo, the practical consequence of any such legal ineffectiveness will be to increase rather than reduce the security of the transferring policyholders whose policies are validly affected. To the extent that the Scheme is ineffective in relation to any particular policy, then the policyholder thereunder will remain reinsured by Sompo. Since the Scheme affects such a small part of Sompo’s overall business, the security thus available for that policyholder will itself not be adversely effected. Similarly, the enhancements to the security for Transfercom’s existing policyholders effected by the Scheme is not in any way conditional upon the 100% legal effectiveness of the transfer in relation to policies not governed by UK law.

I should note that a criticism as to the adequacy of the sample on the basis of which the court was originally invited to address this question was in my view unfounded. Furthermore, it was subsequently supplemented by a second sampling exercise which produced a similar result.

I have dealt, albeit briefly, with what appear to me to have been all the objections of any substance persisted in by AXA. In doing so I have covered most, but not necessarily quite all, of the objections of those transferring policyholders which have been withdrawn. More generally, I am satisfied that Mr Fulcher’s response to all the objections (including those now withdrawn) has demonstrated a sufficient depth of experience, research and analysis that his continuing favourable opinion as to the Scheme may properly be relied upon by the court.