Coronavirus (COVID-19)—some reinsurance perspectives

Coronavirus (COVID-19)—some reinsurance perspectives


Insurance & Reinsurance analysis: Stephen Carter, partner at Carter Perry Bailey LLP, considers the implications of coronavirus on reinsurance contracts.

One hundred years after the Spanish flu pandemic, Marsh and Munich Re, in May 2018, launched an innovative parametric insurance product, PathogenRX. This provided insurance against business interruption caused by a pandemic. In October 2018, Marsh presciently observed in a paper entitled ‘Pandemic Readiness: Risk Finance and Mitigation Strategies’:

‘Although recent pandemics and epidemics have been deadly, the mortality rates from these outbreaks are generally far lower than health crises of the past, owing in large part to advances in medicine and infrastructure. Yet the potential economic impacts of today’s health crisis can be far greater in scope than earlier ones. The increasing reliance of businesses on technology, frequent and unrestricted travel, and far-reaching supply chains means that an outbreak in a single country can have global repercussions. The World Bank estimates that the cost of a severe flu pandemic could total as much as 5% of global GDP’.

Nevertheless, not a single policy was sold prior to the coronavirus pandemic. There is now considerable interest being shown, but it will be too late to assist with the current pandemic.

Those suffering losses from coronavirus therefore have to look at their traditional insurance policies for indemnity, if and to the extent that they provide cover. However, insurers have been wary of providing such cover, largely because of the scale of losses that might arise over multiple classes of business. Hence, most existing policies have not been intended to cover pandemic losses and, further, many contain exclusions intended to apply to pandemics. Consequently, most reinsurances are not designed and rated to cover them either. However, inevitably those suffering loss are looking for ways in which to claim under their insurances, and political pressure is being brought to bear on insurers to agree claims, even though there may be no legal liability to pay them.

In USA, it is being taken a step further. Legislation is being proposed in a number of states that would require insurers to cover business interruption losses even when the policy only provides such cover (as do many UK policies) consequent upon physical damage. It is also reported that both the New York Mayor and the New Orleans Mayor have inserted language in their civil shutdown orders, stating that the coronavirus outbreak is causing property damage (presumably in the full knowledge that under many business interruption wordings, indemnity will only be triggered if there is damage to property). A wave of litigation is expected.

Insurers’ accumulation of losses is, of course, one of the driving forces behind reinsurance. Inevitably, insurers who agree claims will look to their reinsurance programme for recovery. The question therefore arises as to whether reinsurance will provide cover for coronavirus claims paid by insurers, and, if so, on what basis the cover will apply. Problems with which reinsurers will be familiar will arise, but applied to these new circumstances; issues of whether reinsurers have to follow their reinsured’s settlement and of aggregation of claims. Of course, as ever, the application of each reinsurance contract to the circumstances of a claim will depend upon the particular wording of the contract and the particular facts involved. Without that information, it is not possible to be definitive. Even with it, that may be difficult!

However, there are broad categories of common provisions, which constitute a good starting point for guidance.


Aggregation provisions in reinsurance contracts broadly fall into two categories; event based and cause based.

Event based clauses

These provide for separately covered losses to be treated as a single loss by reference to the event, or occurrence, from which they arise. Typically, the clause will define a loss as ‘each and every loss and/or occurrence and/or series of occurrences arising out of one event’ or similar wording. A clause may, for example, refer to ‘each and every loss or series of losses arising out of one occurrence’. In these examples, the unifying factor is the ‘event’ or ‘occurrence’. Both words appear commonly, and the courts have concluded that, used as a unifying factor, they should be treated as synonymous, unless it is clear from the context that they are not intended to be.

In the absence of a specific ‘event’ definition in the reinsurance contract wording, it is axiomatic that the starting point must be that the unifying factor must be something that can properly be called an event. Lord Mustill stated in AXA Re v Field (1996) 2 Lloyd’s Rep 233 (HL) (not reported by LexisNexis) that:

‘In ordinary speech, an event is something which happens at a particular time, at a particular place and in a particular way.’

An event is therefore what has happened—not the reason it happened, which is the cause (as to which, more below). Lord Mustill’s above comments were amplified in what has become known as the ‘unities test’, first put forward in the ‘Dawsons Field’ arbitration award, which the parties agreed to release into the public domain. This award has been referred to in a number of Court judgments. The following words of the arbitrator, Mr Justice Kerr (who became Lord Justice Kerr), have been referred to in most aggregation cases since, so I make no apologies for the length of the quotation:

‘… [B]oth sides gave numerous examples which would or would not in their view be regarded as loss or damage resulting from a single occurrence, such as damage resulting from an air raid, the losses of several ships in an attack on a convoy by a single submarine, a ship breaking loose from her moorings and colliding with a number of other ships, damage from an earthquake, or from the Fire of London, etc etc (the same examples were also used in the context of “arising out of one event”). On which side of the line each of these is to be placed depends in my view on the position in which the person who has to make the determination is placed and on the way in which he will therefore approach the question. The crews of a submarine and of ships which are attacked or sunk in a convoy would no doubt regard each attack and sinking as a separate occurrence. An admiral at a naval headquarters might regard the whole attack and its results as one occurrence; an historian almost certainly would. An earthquake may have a number of tremors producing different damage at different times and in different places; the victims would no doubt regard each tremor as a separate occurrence, but others might not. Whether or not something which produces a plurality of loss or damage can properly be described as one occurrence therefore depends on the position and viewpoint of the observer and involves the question of degree of unity in relation to cause, locality, time and, if initiated by human action, the circumstances and purposes of the persons responsible’.

Analysed strictly on the basis of the unities of cause, location, time and purpose (the latter only being applicable to the extent a loss is initiated by human action), it is problematic for coronavirus or COVID-19 (the former being the virus that causes the latter disease) themselves to be unifying factors under an event based clause.

However, one does not approach the issue exclusively by strict scientific analysis. In Scott v Copenhagen Re, it was argued that the unities test was inappropriate to its particular circumstances (which related to aviation losses following the Iraq invasion of Kuwait). Rix J disagreed and applied the unities test, but made clear that the unities are aids to, rather than the sole analytical criteria of, construction:

‘That question can only be answered by finding and considering all the relevant facts carefully, and then conducting an exercise of judgement. That exercise can be assisted by considering those facts not only globally and intuitively and by reference to the purpose of the clause, but also more analytically, or rather by reference to the various constituent elements of what makes up one single unifying event. It remains an exercise of judgement, not a reformulation of the clause to be construed and applied’.

Cause Based Clauses

The House of Lords in AXA Re v Field considered that whereas an event happens at a particular time, at a particular place and in a particular way, a ‘cause’ is ‘altogether less constricted’. Hence, when a unifying factor is stated to be a ‘common cause’ or ‘the same cause’ or ‘a single source’, or similar wording, the search for a unifying factor can be much wider, including, for example, a continuing state of affairs. Where the words ‘originating cause’ were used, their Lordships considered it to open up the widest possible search.

In all cases, whether the unifying factor is event based or cause based, there has to be a significant causal connection between the event or the cause and the loss suffered.

Finding a single event or occurrence which happened at a particular time, at a particular place and in a particular way, which gave rise to all cases of coronavirus is likely to be very challenging (even bearing in mind Rix J’s bringing intuition into the equation). If the virus itself is not a unifying factor, identifiable events giving rise to multiple individuals contracting coronavirus, may be. Of course, identifying such events is itself likely to be problematic. However, losses caused by identifiable decisions made to contain the spread of virus (such as cancellation of an entertainment event) may more readily be classed as events or occurrences. Questions of degree may still arise. Is it the decision to cancel the event at a specific venue, or the decision to cancel a number of events over a period of time at the same venue, or at a series of venues in different places but in the same ownership/management? In each respective case, the limits of the unities are stretched a little more. The wider the spread, probably in particular of time and place, the less likely a potential unifying factor actually is to be one. There have been a number of decisions, which I will not delve into in detail here, in the past as to whether different circumstances constitute an event or occurrence. For example, coordinated riots throughout Indonesia and the 9/11 terrorist attacks have both been held by the English Courts not to constitute single events. However, depending on the tribunal, the jurisdiction, the nature of the policy and the class of business, the terrorist attacks on the Twin Towers in New York have been held, in different awards/judgments, both to be one and two events.

There are many potential arguments that may be put by insurers, if significantly impacted by coronavirus related losses, as to potential aggregation of losses. Each will depend on the policy wording, the reinsurance contract wording and the circumstances of the loss concerned. However, we hope that the above will at least provide a useful starting point.

Follow the settlement provisions

There are likely to be considerable differences of opinion as to whether inwards risks are covered under the wordings—for example, whether the presence, or potential presence, of coronavirus constitutes property damage. Some claims may be paid by insurers as a result of political pressure, rather than on the basis of legal liability; and I have referred above to moves in the USA to legislate for cover where none would otherwise have existed.

At one extreme, if there are no follow the settlements provisions, the reinsured must prove its loss under both the original policy and the reinsurance contract. At the other extreme, some reinsurance wordings provide cover for ex gratia payments, for example by providing that all settlements ‘including ex gratia and compromise settlements, provided the same are within the terms of this agreement, shall be unconditionally binding upon the reinsurer’. This is very unusual. Most contracts will be subject to some form of follow the settlements provision which falls between these two extremes, the wording of which is key to establishing the extent to which the reinsurer must follow its reinsured’s settlements.

Wordings of follow the settlements clauses may vary, but can be viewed as falling into two broad categories: full follow clauses and qualified follow clauses.

The full follow clause is essentially a simple proviso that the reinsurer should follow the settlements of the reinsured. The Courts have found that the effect of this is that the reinsurer must satisfy two criteria. The first is a matter of fact—it must have acted honestly and in a reasonable businesslike manner. The second is a question of law—the claim, as recognised by the reinsured, must fall within the terms of the reinsurance contract. Hence, the reinsurer cannot go behind the settlement, absent fraud, but it can argue that the reinsurance contract itself does not actually cover the underlying loss as settled.

The qualified follow clause introduces another requirement. Typically, it may provide:

‘All loss settlements by the reinsured shall be binding upon reinsurers provided that such settlements are within the terms and conditions of the original policies and within the terms and conditions of this policy …’

This wording was considered in the case of Hill v Mercantile & General Reinsurance Co Plc. Its effect is that the reinsurer cannot be held liable unless the loss, as a matter of law, falls within both the original contract of insurance and the cover provided by the reinsurance. Each clause and original policy must be considered on its own merits against the circumstances of the claim or claims.

In the case of a retrocession, the ‘loss settlements shall be binding’ provision refers to the immediate underlying loss settlement of the retrocedant. It is not necessary for the retrocedant to go into all of the underlying loss settlements, which could constitute a long chain.

The liability has to be shown on the balance of probabilities. When certain claims in the LMX spiral had been dealt with by a chain of underlying insurers and reinsurers, on an incorrectly aggregated basis, it was held in Equitas v R&Q Reinsurance Co, that in the absence of the ability to rework all of the underlying figures, coverage of the claims could be proven on the balance of probabilities using an actuarial model. Once liability was established in that way, the claimant syndicates did not have to prove correctly aggregated losses by precise calculation. They could recover what was the minimum amount that would have been paid after removing improperly aggregated claims, again using actuarial models to establish the figures on the balance of probabilities. Might similar methodologies be utilised in life reinsurance, if it is necessary to distinguish deaths that were deaths from pandemic and those that were, for example, death by natural causes, in circumstances where it is not practically possible to do so?

Many London market reinsurance contracts cover inwards policies issued overseas and subject to a foreign law. In those circumstances, the reinsurer under English law will have to follow the decision of the foreign Court or tribunal as to insurers’ liability for the underlying loss, even if that decision would have been different had it been considered by the English Courts.

Nevertheless, the reinsurer does not have to follow an underlying Court decision based on overseas law, if to do so offends a fundamental provision of the reinsurance contract itself. In Wasa v Lexington, AGF v Same, the underlying policy was subject to a foreign law. It contained a period clause which covered losses occurring during the period 1/7/77 to 1/7/80. There was a facultative reinsurance contract, subject to English law. It contained a period clause in substantially the same terms as the underlying policy, hence also covering losses occurring during the period 1/7/77 to 1/7/80. The applicable foreign Court broadly treated the period clause in the underlying policy as covering damage whenever it occurred, if some damage existed during the policy period. The cover provided by the reinsurance contract came before the English court. The House of Lords applied the English law interpretation of the period clause, with the result that even though the reinsured had been held liable under the underlying policy, for losses that occurred outside its policy period, the reinsurance only covered those of the losses which actually occurred (as interpreted by English law) during the period of the reinsurance contract.

Lord Phillips explained:

‘… the ‘full reinsurance’ clause in this case, and follow the settlements clauses in general did not and do not have the effect of bringing within the cover of a policy of reinsurance risks that, on the true interpretation of the policy, would not otherwise be covered by it’.

Coronavirus and reinsurance—final observations

Finally, it should be borne in mind that many US policies, in particular, include ‘follow the fortunes’ wording, rather than ‘follow the settlements’ wording. These can impose a broader liability on reinsurers to follow underlying settlements, than an English law ‘follow the settlements’ clause generally would. If a retrocedant is obliged under the local law to pay a claim under the ‘follow the fortunes’ clause, then a ‘follow the settlements’ clause in an English law retrocession will likely oblige the retrocessionaire to follow that settlement.

The above are some of the factors that may feed into the recovery of claims relating to coronavirus under reinsurance contracts. I would again emphasise that there are a variety of clauses, a wide range of classes of business, original wordings and circumstances potentially involved in such claims. While it is not unduly helpful to repeat the old mantra that each case will be considered on its merits, there is, unfortunately, no escaping it.

Nevertheless, it is fair to say that insurers proposing to pay claims that do not in law fall within the insurance cover of their inwards policy, need to be wary of their ability to recover on reinsurance. It may well be wise, if practicable to do so, for reinsureds to liaise with their reinsurers before entering into such settlements. It may be the case that the reinsurance contracts contain claims cooperation clauses or even claims control clauses, giving reinsurers rights to be involved in, or to control, the settlements of their reinsured. Where this is the case, it should go without saying that it is particularly important for the reinsured to work with its reinsurers before entering into settlements.

Of course, other potential issues may also emerge. For example, the use of the words ‘other perils’, especially in cat bond arrangements, has increased over recent years, expanding coverage to natural perils beyond those actually named. Questions have arisen as to whether this may include a pandemic.

From whatever perspective one looks at it, there is clearly the potential for both misunderstanding and dispute between reinsurers and insurers as to the payment of original claims, their coverage under reinsurance and, if covered, the way in which they may be aggregated. As specific scenarios play out, these areas of dispute will become clearer and no doubt, there will be significant decisions either in the Courts or in arbitration. Many reinsurance contracts are subject to arbitration provisions, but it is distinctly possible that the legal issues involved and their importance are such that the Courts may be called upon by way of appeal from arbitral awards which deal with key issues of law (a process which is available in a very limited class of cases, under section 69 of the Arbitration Act 1996, AA 1996).

This month, the Reinsurance News and Artemis COVID-19 Coronavirus Survey revealed a high level of concern in the insurance marketplace over business interruption litigation, aggregation and solvency pressure. Moreover, the Director General of the Association of British Insurers (ABI) is reported as saying that insuring businesses for pandemics would be impossible using the normal model. He observed that the UK insurance industry is ‘up for’ debate with the government and business about the potential creation of a state-backed insurance scheme for pandemic risk. Last week it was reported that in USA that a legislative proposal has been drafted whereby a government backstop, capped at US$500b annually, would be triggered in the event of a future pandemic declared as a public health emergency, when aggregate losses to participating insurers exceed US$250m.

However, I leave the last words with Huw Evans, Director General of the ABI:

‘… Even in the UK, providing widespread insurance cover against pandemics will be virtually impossible without state support, because the amount of capital insurers would have to hold against the risk would result in completely unaffordable prices for customers. Last year, UK companies turned over £4.1trn and employed 27 million people. Insuring these businesses for pandemics is impossible using the normal model, given UK insurers hold total assets of £2.2trn … That is why we need to start thinking about new solutions. Partnerships between governments and insurance markets to help solve big problems are nothing new; so called ‘protection entity’ schemes exist round the world, most commonly for flooding, terrorism and earthquakes. Here in the UK, we have Flood Re, which I helped set up, as well as Pool Re, while other examples include the California Earthquake Authority, the CRC in France and the Earthquake Commission in New Zealand. Each is structured with different levels of state involvement but all seek to enable insurance protection for risks that would otherwise be uninsurable’

This content is based on an article first published by Stephen Carter of Carter, Perry Bailey LLP and is published with permission.

Related Articles:
Latest Articles: