The correct allocation of EL mesothelioma losses to excess of loss reinsurances has been an issue in the market since at least as far back as the 1990s. Mesothelioma differs from other asbestos diseases in that it is “indivisible”, namely injury can be caused by any material exposure rather than proportionately by reference to exposure.

Reinsurances written post 1984 often contain an ACOD/B clause, which was introduced in that year and which allocates loss on a “pro rata / pro rata” basis, namely the loss allocated by reference to the period of reinsurance versus exposure and with limits and retentions proportionately reduced. However, what is the correct legal position prior to 1984 or where ACOD/B was not included in the reinsurance?

Arguments historically from reinsureds have been that loss could be “spiked” (a word generated by one reinsured running this argument) 100% in any relevant reinsurance year with the application of one retention only. Reinsurers have argued for pro rata allocation across reinsurance years with a full retention each year. Most of the major EL players have settled with their principal reinsurers over the years on an ACOB/B basis but with some discount on their claim.

MMI has not so settled. Rather, it has sought to pursue “spiking” against a reinsurance year written by Lloyd’s (now Equitas) and with the effective application of one retention post contribution/recoupment. The Court of Appeal, in a decision handed down in April 2019, has found it is not able to do so, but has to spread its loss over years of exposure and with full reinsurance retentions each year.


Before turning to the Court of Appeal decision, it is helpful to summarise briefly the mesothelioma position at employer and insurer level:

  1. As a result of the House of Lords’ decision in Fairchild v Glenhaven2, a mesothelioma victim tortiously exposed by two or more employers, need only prove material contribution to the risk to be treated as having caused the disease – a far weaker test than the usual balance of probabilities.
  2. Following the House of Lords’ decision in Barker v Corus3, the employer is only liable at common law on a several basis for the extent of its relative degree of contribution to the risk. This finding was expressed to be due to fairness to employers, given the significant relaxation of the law on causation in Fairchild.
  3. The Compensation Act 2006 reversed Barker, providing that employers were to be liable for 100% of the loss.
  4. In Zurich v IEG4, the Supreme Court decided by a majority that an EL insurer in turn was liable to indemnify the employer for 100% of the loss each year (i.e. the employer could “spike”), but would gain contribution / recoupment rights as against insurers on different years and employers to the extent uninsured on other years. Such decision ensured the credit risk in respect of other insurers was borne by the insurer rather than the employer, thus protecting the victim.

Equitas v MMI

The central submission of Equitas was that the Fairchild and subsequent jurisprudence was innovative in order to ensure compensation for the victims of mesothelioma. However, that objective had been achieved and the same policy considerations did not apply at reinsurance level. Accordingly, the law should return to basic insurance principles of non-selection and coverage by risk, such that a Barker approach should be applied to ensure that liability was apportioned by reference to reinsurance time on risk.

Equitas put forward three ways in which this objective could be achieved: (1) by way of deemed allocation or implied term requiring MMI to allocate proportionately (it being accepted that MMI had not allocated itself due to its 100% liability each year to its insured employer); (2) by way of the duty of good faith or good faith principles generally; and / or (3) by way of contribution / recoupment where the loss was treated as proportionately allocated to the reinsurance programme and from the ground up.

The Court of Appeal accepted that the law should “return in a principled way to a more orthodox approach” given victims had now been protected. As to the legal manner of producing such result:

1. Deemed Allocation / Implied Term

The Court of Appeal accepted in principle that as the employer was 100% liable, so was the reinsurer as reinsurance was a type of insurance of the underlying risk (a result which would still have followed on the alternative analysis of reinsurance as a type of liability insurance of the reinsured) – and reflecting the IEG finding that the insurer was 100% liable each year. Accordingly, any deemed allocation or implied term would be inappropriate as inconsistent with the legal effect of the reinsurance terms.

2. Duty of Good Faith

The Court of Appeal noted that the duty of good faith was limited in principle to a duty of disclosure on inception and not to act fraudulently on a claim, and thus it had no part to play in the current context.

However, the Court considered a line of cases involving the implication of a term to the effect that a party to a contract shall act in good faith i.e. not irrationally in exercising a contractual power – most notably Gan v Tai Ping5 where such term was imposed to fetter a reinsurer’s ability to decline to consent to a settlement.

The Court of Appeal found the authorities concerned the proper construction of the contract. Here, there were “powerful reasons” to imply such a term to the effect that MMI could only present reinsurance claims by reference to each year’s contribution to the risk. This was because the concept of spiking was inconsistent with the presumed intentions and reasonable expectations of the parties at the time of conclusion of the reinsurance contracts. Fairchild and its progeny could not then have been predicted and produced a result inconsistent with fundamental principles of liability insurance law – such principles being an inability to select cover against insurers, coverage by reference to risk during the policy period, and loss falling into one rather than into a series of separate periods. The position was materially different at insurance level due to the need to ensure victim compensation.

3. Contribution and Recoupment

Assuming 100% liability each year and an ability to spike, the Court of appeal found the “just” solution to eliminate anomalies flowing from the Fairchild jurisprudence was to adopt the approach proposed by Equitas, namely to treat the loss as spread across the years with a full retention in each, and then to operate contribution / recoupment accordingly. The full retention each year reflected the fact that MMI agreed to such in years victims were exposed, and such approach ensured higher layer reinsurers were not exposed until full exhaustion of retention and underlying reinsurances – all no doubt, it was said, reflecting the lower premium charged by such higher layer reinsurer.

The net effect of the above, given the level of retentions and increasing over the years, will be very significantly to reduce any MMI reinsurance recovery.


One of the results of reinsurance disputes usually being resolved by way of arbitration is that there is a dearth of authority on allocation. Indeed, prior to Equitas there were only two cases directly addressing the issue – MMI v Sea Insurance6 and IRB v CX Re7. The latter is a difficult decision from which to draw any general principles. The position generally on reinsurance allocation seems to be as follows:

  1. Where the reinsured has allocated to particular years by settlement, the reinsured cannot recover more than such allocation due to the operation of the indemnity principle.
  2. The reinsurer is able to open up the settlement to ascertain the “real basis” of the settlement to ensure no breach of the indemnity principle. The Court of Appeal appears to have proceeded on this basis in Equitas by analogy with authorities addressing the ability of an insured to demonstrate quantum of a loss when claiming under a liability insurance and a reinsured’s ability to establish loss fell within the terms of a reinsurance when relying on a follow clause.
  3. Leaving to one side the indemnity principle, the reinsured’s own allocation is legally irrelevant to reinsurance coverage. That is dictated by the terms of the reinsurance itself and the underlying factual position (MMI v Sea).
  4. The reinsured is obliged to allocate and claim EL mesothelioma losses on a proportionate allocation basis. In Equitas, the Court of Appeal was at pains to stress this position is applicable only to such losses due to the manner in which the law on mesothelioma has developed.

As regards the Equitas decision itself, a number of questions arise:

  1. Is it right that “spiking” offends basic principles of English insurance law? The reinsurance was written on the basis of loss due to occurrences during the reinsurance period, not relative risk of indivisible loss. It might be said that all that has happened is that the causal test for loss has been weakened to both insurers’ and reinsurers’ detriment given indivisible loss but which still falls within the reinsuring clause.
  2. Is it right in any event that parties contracting, in say, the 1960s would objectively have decided that if the causal test for the insured’s liability was weakened, the insurer alone must bear the consequences of such decision? The general principle is that parties to a contract must bear the consequences of a change of law.
  3. One can see in principle that allocation spread across the years for the purposes of contribution / recoupment should be effected from the ground up to reflect the reinsurance market’s relative participation in the risk. Can it be said the risk has been realised due to 100% liability in principle in each year?

The Court of Appeal has granted leave to appeal to the Supreme Court, a Court which presently contains no Justices experienced in insurance or reinsurance law. The market will await with interest the final say on this subject.