Coronavirus (COVID-19) business interruption test case—pre-judgment analysis and the future of pandemic claims

Coronavirus (COVID-19) business interruption test case—pre-judgment analysis and the future of pandemic claims


The High Court’s decision in the Financial Conduct Authority’s (FCA) coronavirus (COVID-19) business interruption (BI) test case will be published on 15 September 2020. The decision is widely expected to be appealed, whatever the High Court decides. Accordingly, the wait for the legal certainty needed to adjust and settle valid coronavirus BI claims is set to continue for several months yet. Pamela Freeland of Weightmans and Edward Rushton of LexisNexis consider the implications of this impasse for coronavirus BI claims advanced in other forums, such as the Financial Ombudsman Service (FOS), arbitration or separate actions in the county Court and the High Court—all the time while the coronavirus saga continues to unfold and produce new conundrums, such as the gradual withdrawal of government support through the coronavirus job retention scheme. We also take a step back and consider the essence of the parties’ respective positions on the critical issue of causation, and the implications that the courts’ final reckoning on the issue might have for insurance law.

The impasse

The delay in resolving coronavirus BI claims has been deeply frustrating for policyholders, many of which are small businesses struggling to survive the economic impact of lockdown and the coronavirus pandemic. In some quarters this frustration gave way to anger as policyholders compared notes and came to the conclusion that insurers were trying to have it both ways.

Some policyholders with BI cover responsive to closure by order of a public authority state that they were informed that their losses were not covered because they would have been suffered anyway, because of the pandemic, and all the fear, lockdown and social distancing that it entails.

On the other hand, other policyholders with BI cover responsive to the outbreak of an infectious disease state that they were informed that their losses were not covered because they would have been suffered anyway, because of the government’s instruction to business to close their doors.

One can readily understand policyholders’ anger and frustration, especially in light of the Prime Minister and Chancellor’s assurances on 17 March 2020 that insurers would make payments to companies forced to close as a result of the coronavirus pandemic, if they had policies to that effect. They said an agreement had been reached with insurance companies to ensure businesses received payouts where appropriate. These assurances, while caveated, may unhelpfully have raised many policyholders’ expectations of having claims paid when they did not in fact have a valid claim on any view, irrespective of the issues in the test case. Policyholders with BI insurance without ‘non-damage’ extensions may, in particular, have taken false hope from the government’s assurances.

The present impasse has, however, also been frustrating for brokers and insurers, who have been widely and often unfairly maligned in the mainstream press and who have had to deal with high volumes of distressed customers in circumstances where, in truth, there was little that they could do to assist pending legal clarity from the courts.

Insurers seldom elicit sympathy, especially when declining claims. The fact is, however, that they found themselves in an invidious situation. However sympathetic an insurer (or claims handler) may be to the plight of an unfortunate policyholder, placating them by making an ‘ex gratia’ payment is very often a terribly bad idea. Unfortunately for the reputation of the insurance industry, not all of the reasons for this are widely understood.

Some reasons are fairly obvious—it would be unfair for insurers to pick and choose who gets paid and who does not on an arbitrary basis. Most people would also understand that an insurer will not stay in business very long if it depletes its reserves paying invalid claims. Fewer, however, consider the implications of making ex gratia payments for other stakeholders, including shareholders and other policyholders.

The interests of these other stakeholders are diminished when invalid claims are paid. Policyholders’ interests are diminished if their insurer pays invalid claims to other policyholders because the insurer will have to martial its remaining funds more cautiously in order to maintain its regulatory solvency margin. This may manifest in many ways that do not favour policyholders, such as narrower cover, higher premiums and more stringent controls on paying claims in future.

Restrictions on the availability and scope of insurance cover also diminishes the role that the insurance industry plays in de-risking the global economy and international commerce. This has negative consequences for the national and global economy at large.

Shareholders’ interests are of course harmed if the insurer pays substantial amounts in respect of invalid claims. Since institutional investors such as pension funds account for a sizable proportion of the ownership of insurers, the impact on shareholders will have an indirect but negative effect on fund performance.

The rule of law and contractual certainty

The legal uncertainty had to be resolved urgently, for the sake of policyholders and to preserve the reputation of the insurance industry. This is of course why the FCA and the insurance industry acted swiftly in getting the test case afoot and expediting its progress. The resulting impasse, pending judicial clarity, is frustrating, but in the long run it is far preferable to forcing insurers to pay ex gratia claims by means of political pressure or, worse, retroactive legislation.

The delay will mean that some cash-strapped businesses will fail before the test case has concluded. This is deeply regrettable. It is, however, better in the long run for claims to be paid in accordance with the law and the terms on which policyholders contracted. This does of course mean that policyholders whose losses do not fall within the scope of their policy will have to bear their own losses. However, only by settling claims in accordance with the terms of the contract, properly construed, can the rule of law be upheld. The alternatives, however appealing they may seem in the throes of the pandemic emergency, undermine our societal interest in striving for contractual certainty and predictable legal outcomes.

The necessity of the test case does demonstrate that the subject policy wordings were not clear enough. That is not to say that insurers were remiss for not having foreseen the construction issues to which the parties were joined in the test case—the insurance industry has a long history of responding to crises by adapting policy wordings to cater for unprecedented developments. If, as is hoped, insurers are willing and able to offer insurance against future pandemics, they will doubtless do so using new clauses, crafted in the light of knowledge of how the coronavirus pandemic unfolded and the legal clarity that it is hoped the courts will soon deliver.


There were many issues in the test case but those relating to causation are fundamental. Both sides averred that their respective positions proceed from the proper construction of the relevant policies, but they diverged significantly regarding the appropriate legal test for causation. Insurers contended that the correct test is the classic ‘but for’ test, familiar to most lawyers from tort law and criminal law. The FCA, on the other hand, argued that for the purposes of insurance, it is sufficient that a loss was proximately caused by an insured peril so long as it was not also caused by an excluded peril.

It is on this legal basis that the FCA submitted that policyholders’ BI losses were proximately caused either by a single peril (being the national coronavirus outbreak), or a multitude of concurrent perils of approximately equally effective causes (ie the aggregate of all known and presumed coronavirus infections in the country). The FCA cited the Silver Cloud as authority, in which case the insured loss was found to have been caused equally by advice to US travellers regarding the dangers of cruise holidays in the wake of the 2001 terrorist attacks on the United States (an insured peril), and also by the fact of the attack itself causing worried US citizens to avoid going on cruises (not insured against, but not excluded either). Flaux LJ, (presiding over the FCA test case alongside Butcher J) was leading counsel in the Silver Cloud case for the (largely) successful party.

In a sense, the approaches of both sides beg the question of what the relevant insured perils are and indeed it may be instructive to ask oneself that question when considering a particular BI policy.

The FCA submitted that Orient Express Hotels, which insurers rely on in support of the ‘but for’ test, was wrongly decided because the tribunal had incorrectly identified ‘Damage’ (as defined) as the relevant peril, not hurricane, and that this error led the tribunal to a misconceived application of the ‘but for’ test. The FCA characterised the resulting counterfactual analysis as being unreal—whereby the insured hotel’s insured BI loss was found to be that amount it would have earned had it not sustained ‘Damage’ but where the surrounding area has nevertheless been heavily damaged such that guests would not be able to get to it.

Insurers’ riposte was that counterfactual analyses are unreal by their very nature. This is of course true, but a scenario in which a hotel is left unscathed while everything around it is destroyed seems more ‘unreal’ than a more typical BI loss scenario, such as a kitchen fire at a hotel. The ‘but for’ counterfactual analysis would simply remove the kitchen fire to assess what the policyholder’s BI loss was. It would seem, therefore, that a straightforward ‘but for’ causation test produces particularly strange counterfactuals where more than one concurrently operative peril (some insured, some not) is considered to have caused the loss.

Once again, this line of reasoning tends to come back to the question of what, exactly, the relevant insured peril is.

This was among the most fascinating parts of the test case. Counsel for both sides debated colourful hypothetical scenarios with the judges. Both sides also provided vivid analogies to illustrate the logic of their competing cases on causation and, in so doing, their competing analyses of the insured perils under the relevant policies.

Counsel for Hiscox, Jonathan Gaisman QC, imagined a system of pipes as a metaphor to describe the relevant peril for the purposes of the non-damage businesses interruption section of a given policy. The ‘pipework’ obviously depends on the wording in any relevant policy but the analogy is apt to describe insurers’ analysis of the relevant insured peril. Counsel put it as follows:

‘Imagine that the pandemic is a large bore pipeline through which liquid flows. The liquid is the loss which the pandemic causes people, businesses. We will call that pipeline A. At a certain point in the pipeline there is a flange at which several smaller bore pipelines lead off the main pipeline, each of them taking a small proportion of the liquid flowing through pipeline A. One of them, one of those smaller pipelines, is called public authority action, and we will call it pipeline B. In due course it leads to pipeline C and to pipeline D but we will leave those out of the equation for simplicity. Let us assume that pipeline B receives 25% of the total liquid. Other pipelines, which we will call X, Y and Z, take the other 75% of the liquid from A.
The insurance here is against the loss caused by public authority action following or caused by… among other things an occurrence of disease. It therefore only insures against the loss coming out of pipeline B.’

In other words, insurers’ case is that the insured peril for non-damage BI is exactingly circumscribed by the requirements of the non-damage BI insuring clause, such that the insured peril is loss caused by B. Accordingly, the ‘but for’ counterfactual becomes A minus B, (or, to put it another way, ‘X, Y and Z’), if pipelines C and D are again left out of the equation for simplicity.

An attraction of this approach is that it very clearly gives full effect to all the conditions of the insuring clause. It may well be legally correct. If it is, there may be severe practical difficulties producing evidence to establish what element of a policyholder’s coronavirus loss was caused by B and what was caused by ‘X, Y and Z’. Such difficulties may, perhaps in most cases, render it impossible or impractical for policyholders to discharge their initial burden of proving their loss. This may produce situations where a policyholder’s coronavirus loss probably was caused by both A and B but it is unable to recover anything because it cannot ascertain what part of its loss was caused by B rather than A (or the other way around, depending on the ‘pipework’ in the insuring clause of the relevant policy).

This result would do little to allay policyholders’ perception that insurers are trying to have it both ways when denying claims on the grounds that the loss would have happened anyway, whether because of mandated closures or because of the coronavirus outbreak, depending on the insuring clause in the relevant policy. This perception is unfair because each insurance policy is a separate contract which falls to be construed in accordance with its terms, and there can be no reasonable suggestion that insurers somehow acted in concert to frustrate potential pandemic BI claims. Nevertheless, if the test case results in policyholders being unable to recover simply because of the practical evidential difficulty of distinguishing loss caused by A from that caused by B, insurers stand to make what might be regarded as a windfall saving at the expense of policyholders.

Colin Edelman QC, for the FCA, vividly described the rapid emergence of the nationwide pandemic and showed the judges a map of the UK showing clusters of coronavirus outbreaks emerging at the same time across virtually all of England. By reference to contemporaneous Scientific Advisory Group for Emergencies minutes, he argued that nationwide business closures, lockdown and social distancing were imposed by the government in response to a single cause, being the nationwide coronavirus outbreak, or, alternatively, that all of the outbreaks comprising the nationwide outbreak comprise a multitude of concurrent causes of the same, or similar efficacy.

Accordingly, the FCA’s case is that policyholders’ losses were proximately caused either by a single insured peril (being the national coronavirus outbreak resulting in closures), or a multitude of concurrent causes of approximately equally effective causes, some insured and some not, resulting in government measures (ie the aggregate of all known and presumed coronavirus infections in the country). In support of the latter position, the FCA relied on the Silver Cloud as authority that a policyholder is entitled to an indemnity where its loss was caused by an insured peril and a concurrent uninsured peril of equal efficacy. On this basis, it was argued, policyholders should be entitled to recover their loss in full, irrespective of whether the non-damage BI insuring clause requires the outbreak of an infectious disease or closure ordered by a competent authority.

Implicit in this argument is a different analysis of what the relevant insured peril is. The FCA’s approach has the attraction of simplicity—the insured peril is simply the outbreak of an infectious disease, or mandatory closure of premises because of the outbreak of an infectious disease, depending on the particular policy. On this view, the further requirements of the BI insuring clauses, such as a requirement for there to have been an outbreak within 25 miles of the premises, are not part of the insured peril but merely pre-requisites for cover. This may be the correct analysis, but as ever the devil is in the detail. Simplicity has its attractions, but it should never be a substitute to giving effect to what the parties actually agreed.

Taking a step back (putting the law and policy wording aside for a moment), purchasers of BI insurance probably would, rightly or wrongly, have conceived of their insurance policy as protecting them against perils of disease outbreak or closure by public authority. They might have been quite confused if their broker had explained to them that the actual peril insured against was loss caused by the outbreak of infectious disease within 25 miles minus any part of its loss attributable to qualitatively indistinguishable outbreaks of the same disease at the same time as the disease spread in virtually all parts of the country.

In reality there were few, if any, conversations of this nature because nobody was very concerned about the possibility of an outbreak of this scale. We are learning the hard way. Had such a conversation arisen, however, is it more likely that the broker would have explained that the insurance was against outbreak or closure, subject to further requirements such as proximity of the virus? Or is it more likely that the broker would have explained that such further requirements were inextricable elements of the insured peril itself?

Lord Justice Flaux and Mr Justice Butcher have a challenging task indeed and there are many issues to be decided, but it is hoped that the result of the test case will not only facilitate the proper adjustment and settlement of valid claims, quickly, but perhaps also provide guidance on the nature of insured perils and causation. A more precise understanding of what the insured peril is might make future questions of what loss was caused by an insured peril much easier to answer. Of particular interest would be guidance on how to discern the insured peril when construing a complex insuring clause. Does the inclusion of further requirements or limitations further delineate the insured peril itself, or is it the nature of an insured peril that it is apt to cause loss in a particular way, with any further requirements or limitations being simply conditions rather than part of the peril?

Complaints to the FOS and parallel proceedings

While the result of the FCA test case is awaited, complaints that have been made by policyholders to the FOS regarding BI policies are being held in abeyance so as to ensure that decisions are consistent and are not at odds with judicial authority. It also indicates that the FCA and the FOS are adopting a joined up approach, albeit that they have very different roles to play. How long however will this moratorium last? Will the FOS resume its assessment of BI related complaints when the first instance decision in the FCA test case is published or will it wait until the case has run its course, including any and all appeals that may be pursued? Having waited for the first instance decision, it would seem unlikely that the FOS would resume its assessment of complaints in circumstances where any central issues that would feature in its decision-making process were the subject of an appeal. But from a practical perspective, how long can the FOS put off its substantive review of such cases?

No matter the timing, the FOS will have to deal with the logistical challenges caused by the backlog of complaints and the competing demands of policyholders, who having waited (at least) several months, and will be likely to expect rapid decisions when the moratorium is eventually lifted. As at the date of publishing, the FOS states that it takes around five months to allocate a complaint relating to other types of insurance products to a claims handler; a clear sign of the burden faced by the FOS and the uphill struggle it will have processing claims relating to BI policies.

Resourcing will clearly be a key concern. Perhaps the FOS can address this with a combination of new hires (there are some indications that recruitment is underway) and the retention and retraining of case handlers who had been processing PPI claims (the deadline for making such claims now having passed). No doubt that dealing with the backlog of cases, while continuing to address complaints unrelated to BI policies presents the FOS with an unprecedented challenge. The spotlight will be on the FOS and its handling of complaints when the first instance decision in the FCA test case is published.

Prior to the commencement of the FCA test case, the Hiscox Action Group (HAG), which comprises around 400 Hiscox policyholders, embarked on an expedited arbitration process, seeking indemnities for BI claims relating to coronavirus. Arbitration is a private process and it is considered likely that the parties in this case have agreed terms relating to confidentiality meaning that the outcome may not become public knowledge. That privacy and/or confidentiality and the window of opportunity between the hearing and the decision in the FCA test case may have given the parties to the arbitration an opportunity to ‘de-risk’ these claims. Outsiders can only speculate, highlighting the stark contrast between litigation and arbitration as forums for dispute resolution.

As the arbitration commenced by the HAG demonstrates, the FCA test case does not preclude policyholders from pursuing a claim independently. Claims could therefore make their way into county Courts and High Courts up and down the country. If a significant number of such claims are made, the courts may also face resourcing challenges as they too begin to deal with the backlog of certain types of cases (such as possession and eviction hearings) which have been paused as a result of the pandemic. Will the courts use their general case management powers to adopt the same approach as the FOS and stay such claims until the ultimate outcome of the FCA test case is known? Again if so, how long is too long to hold such cases in abeyance?

Coronavirus job retention scheme

In March 2020 the government launched the Coronavirus Job Retention Scheme with the aim of saving jobs that might otherwise have been made redundant as a result of the lockdown and social distancing measures introduced to reduce the spread of the virus. The scheme is now winding down and HMRC has indicated that it will now start taking a closer look at certain ‘high risk’ cases where payments may have been made in error and/or fraudulent applications may have been made (such cases may, HMRC states, amount to £3.5bn). Litigation on this issue will be likely to follow, with HMRC seeking to claw back payments that should not have been made.

However, furlough payments could also feature in claims relating to BI policies. There has been some criticism in the press about the categorisation of furlough payments as a ‘saving’ to be deducted from any indemnity paid as part of the loss adjusting process pursuant to the terms and conditions of BI policies. Some say that insurers are using the public purse to reduce indemnities paid under policies that do respond to claims relating to coronavirus. That criticism seems unfair because insurance is by its very nature designed to be a back-stop or final safety net when other avenues have proven unsuccessful and when other savings made when a business is unable to operate have been calculated and included in the loss adjustment process to date without attracting criticism. It remains to be seen whether HMRC and/or policyholders will have the appetite to pursue this issue with insurers now that the immediate issues presented by coronavirus have been addressed and the focus shifts to the more medium to long term challenges in its aftermath.

Other issues, such as the imposition of local lockdowns, complicate the landscape still further. See: Local lockdowns—complex implications for coronavirus (COVID-19) business interruption insurance cover.

Future pandemic cover

Pandemic cover was incredibly scarce and largely unavailable outside of the US prior to the outbreak of coronavirus and only a small minority of businesses had obtained such cover, most probably due to pricing and/or the perception that the risk of a pandemic was remote in modern times. Now, with the risk of a second wave of coronavirus and with the knowledge of the significant impact a pandemic can have on many sectors of commerce, the question of how this issue can be addressed in the future is a top priority for the government, businesses and insurers.

For many, the obvious answer is an insurance system backed by government. This is not a new idea: Pool Re was set up in 1993 and provides insurance backed by government, for losses relating to / arising from acts of terrorism. In the event of an insured loss resulting from an act of terrorism, insurer members of Pool Re pay out claims up to certain thresholds. If losses surpass those thresholds, Pool Re’s reserves will be deployed to pay the remainder. In the event that Pool Re’s reserves are depleted, the government will step in and provide an unlimited loan to pay the remainder.

Given the financial costs associated with a pandemic, it is difficult to see how any scheme would work without the state being the ultimate backstop. However, the costs of the Pool Re scheme would be dwarfed by the potential costs of a scheme providing cover for a pandemic. Policyholders may need to share the burden by way of higher policy excesses for damages relating to or arising from a pandemic and/or may have to pay more for other insurance products so that funds can be diverted to assist with the financing of pandemic risk. An industry levy (like that applied to property insurance, which helps finance Flood Re to make property insurance in high risk areas affordable) may also be required. Further still, an international effort might be required; after all pandemics do not stop at borders and cannot be eradicated without worldwide effort.

Concluding comments

It is hoped that the widely anticipated decision in the FCA test case will provide some clarity to policyholders and insurers as to the extent to which certain types of non-damage BI policies will respond to the losses caused by the closure of businesses to tackle the spread of coronavirus. However, given the importance of the case, an appeal on at least some of the issues in dispute is likely meaning that, for many, uncertainties will remain. For many, who have made complaints to the FOS or who have, or are considering litigating or arbitrating, the decision in the FCA test case will not be the final word on the subject. The impact that the decision will have on policy coverage questions more generally, such as the meaning of an insured peril and the causation test to be applied when considering concurrent causes of loss remains to be seen.

The focus has to some extent shifted to the question of how the government, businesses and insurers can be better prepared for the next pandemic (having considered it unlikely or not even having contemplated at all that a pandemic would occur most people now believe that more will follow). Discussions with government and insurers are afoot but for any scheme to succeed it will need the full support of government, insurers and businesses alike.

Related Articles:
Latest Articles:
About the author:
Edward leads the Lexis®PSL Insurance and Reinsurance team, which is dedicated to producing market-leading resources that reflect and are responsive to the needs of legal and market professionals in the re/insurance industry and which leverage the technological innovations of LexisNexis.