The Insurance and Reinsurance Law Review: Covid-19 and Non-Damage Business Interruption Insurance: The UK Response
Insurers in the United Kingdom have cooperated with the regulator to obtain declaratory judgments on significant coverage issues under 'non-damage' business interruption policies. In January 2021, the Supreme Court gave judgment on these issues on the basis of sample wordings agreed upon by the parties in advance. The parties have agreed to be bound by the outcome of the case and it is hoped that this cooperation will significantly accelerate and simplify the claims settlement process. Nonetheless, important issues as to evidence and proof of loss remain.
Business interruption (BI) insurance is often purchased as an extension to a property damage policy. These policies usually only provide cover for BI loss that results from damage to the insured property. There has been a good deal of debate about whether or not contamination with covid-19 causes 'damage to property' within the meaning of these policies. In the United Kingdom, however, the Financial Conduct Authority (FCA), which regulates the insurance industry, has not challenged insurers' position that the presence of covid-19 alone does not amount to property damage.
By contrast, the regulator has been concerned at insurers' response to claims under two classes of non-damage BI insurance: disease covers and prevention of access covers.
In general terms, disease covers provide protection for business interruption losses flowing from an outbreak of disease at, or within a certain distance of, the insured premises. Prevention-of-access cover protects the insured in respect of the BI loss flowing from the prevention or hindrance of access to the insured premises as the result of the actions of, or restrictions imposed by, the authorities in response to the outbreak of a notifiable disease.2
Insurers' response to claims under these policies raised a number of common coverage issues. For example, insurers commonly argued that disease policies should only respond to a local outbreak of disease and not a national pandemic. In relation to denial-of-access policies, insurers argued that access to premises was only 'denied' for coverage purposes if premises were ordered to close by government action that had the force of law. Access to businesses that closed voluntarily, for example because of government advice, was not, in insurers' view, 'prevented' within the terms of the insurance. There were also significant differences with regard to the methodology for the calculation of loss and the interpretation of 'trends clauses'.
II Declaratory proceedings
To resolve these and other related issues the FCA initiated declaratory judgment proceedings with the cooperation of eight leading insurers,3 who were defendants in the action. Two policyholder groups also joined the action as interveners4 and it was agreed that the proceedings would take place on an expedited basis.
The parties selected 21 sample policies that contained language typical of the non-damage BI cover at issue and asked the court to make a series of rulings on the proper construction of the relevant clauses and the manner in which they respond to the covid-19 pandemic in the United Kingdom. The proceedings took place on the basis of an agreed set of facts relating to the timing and nature of governmental interventions and the court was not asked to make any findings of fact.
The Commercial Court, part of the High Court, Queens Bench Division, handed down its judgment on 15 September 2020: Financial Conduct Authority v. Arch Insurance (UK) Limited and Others.5 The judgment was subject to a 'leapfrog' appeal to the Supreme Court, which gave judgment on 15 January 2021.6
IV Disease policies
Addressing the nature of the risk insured under the disease policies, the Supreme Court held that the peril insured against was loss caused by an occurrence of illness resulting from covid-19 that occurred at or within a specified radius of the insured premises.
The Supreme Court found that an 'occurrence' in this context meant something that happened at a particular time, in a particular place and in a particular way. Accordingly, a disease was not an occurrence and the covid-19 pandemic was not an occurrence. By contrast, an individual case of disease could be an occurrence.
The Supreme Court therefore concluded that the typical disease clause provided cover only in respect of loss resulting from a case of illness at the premises or within a specified radius of the premises. There was no cover under a typical disease clause for loss caused by illness outside the specified radius.
It was accepted by the parties that the disease will have occurred in the relevant area when at least one person is infected. It is not necessary, however, that the case has been diagnosed. Clearly, this raises issues of evidence and burden of proof, which are discussed later in this article.
V Prevention of access and hybrid clauses
The Supreme Court accepted the lower court's analysis that the sample prevention-of-access polices with which it was concerned provided cover in respect of a 'composite' insured peril. While there were differences in detail in relation to the various policies at issue, the key elements comprising the peril were (1) prevention of access to the insured premises, (2) as a result of government action or advice (3) taken in response to an emergency likely to endanger life. It is this composite peril that must cause the interruption to the business for the cover to be triggered.
The insurers accepted that it was not necessary for access to be physically impossible for the clause to be triggered. The Court found, however, that 'prevention of access' was not that same as 'hindrance of access' or 'restriction of access'. Thus, access to premises was only prevented when the business was directed by the authorities to close because it fell within a particular category, such as restaurants. Access to businesses that were not within a class that was directed to close was not prevented, even if those businesses chose to shut in accordance with government advice, to protect staff or because they had no customers in light of the impact of governmental advice or regulations more generally. For example, where law firms instructed staff to stay away from the office in accordance with government advice to work from home where possible, that did not amount to prevention of access to the business.
Consistent with this conclusion, the Court also held that, in the context of the prevention-of-access policies, 'government action' meant something that was mandatory (although not necessarily with the force of law) and was distinct from advice or recommendations, however strongly worded these may have been.
In an important departure from the lower court's decision, however, the Supreme Court held that the prevention-of-access clause could be triggered if only a discrete part of the premises or a discrete part of the business was directed to close. The clause did not require closure of the whole business or of the entire premises.
Finally, the Supreme Court was required to consider the nature of the risk insured under 'hybrid clauses', which respond to loss resulting from an inability to use the insured premises because of restrictions imposed by government following an outbreak of disease at the premises or within a specified radius of the premises.
The Court found that on its proper construction, an 'inability to use' premises meant a complete inability to use: hindrance or restriction of use was insufficient to trigger the clause. As with the prevention-of-access clauses, however, the cover could be triggered by an inability to use a discrete part of the premises or business.
As with the other clauses under consideration, mandatory governmental measures were required – advice was insufficient.
VI Causation and quantum
The Supreme Court's findings with regard to the nature of the peril insured under the sample clauses meant that the issue of causation was crucial. Could it be said that the insured's entire covid-19 business interruption loss was 'caused by' a single occurrence of covid-19 at or near its premises?
Insurers argued that since it could not be said that 'but for' an outbreak of the disease at or near the premises, the insured would not have suffered its covid-19 BI loss, then there was no cover.
The Supreme Court accepted the first part of this analysis agreeing that it could not be shown that but for a single case of covid-19 at the premises the losses would not have occurred. They went on to emphasise, however, that the but-for test is not always appropriate. In particular, it is not appropriate when there are multiple proximate causes of loss. This may occur where there are two distinct causes, either one of which is sufficient to bring about the loss or where there are a number of distinct causes that combine together to bring about the loss. It was this latter analysis that was applicable to these cases. In the Court's view, each individual case of covid-19 was a separate and equally effective cause of the government restrictions, which were a response to all the cases of covid-19 throughout the country. Accordingly, the test of causation could be satisfied if there was a single outbreak of covid-19 at or within the specified radius of the premises.
VII Trends clauses
One of the most contentious areas of dispute arose in connection with the quantification of damages. Most BI policies contain a trends clause. Typically, these require circumstances that would have affected the insured's business had the loss event not occurred to be taken into account as part of the counterfactual for the purposes of calculating the BI loss. Insurers argued that if the outbreak of disease at the insured premises (or within the relevant area) was the loss event, then that was all that should be stripped out to arrive at the counterfactual. Consequently, the insured's BI loss would have to reflect the fact that the rest of the country was subject to lockdown and other restrictions as a result of the nationwide pandemic. Clearly, in the majority of cases, this would substantially reduce the insured's recovery.
In the view of the Court, however, it was necessary to remove from the counterfactual all the elements of the composite peril for the different policies. Consequently, the actions and advice of the government, in so far as they affect the insured premises, were inseparable from the nationwide action and advice. Accordingly, the only way of establishing what the insured business would have achieved if the covered event had not occurred was to assume that there had been no covid-19.
In reaching this decision, the Court considered the controversial case of Orient Express Hotels v. Assicurazioni Generali Spa (UK) (t/a Generali Global Risk),7 which concerned the calculation of BI loss to a hotel in New Orleans following Hurricanes Katrina and Rita. The insured hotel, in the city's French Quarter, was completely devastated by the hurricanes, as were the surrounding districts. On appeal from arbitrators, however, the Court held that the proper interpretation of the trends clause in that policy required the only counterfactual to be the damage to the hotel, with the result that the lost profit should be what an undamaged hotel in a damaged city would have earned.
The Supreme Court in FCA v. Arch held that Orient Express had been wrongly decided and overruled the case. In reaching this decision, the Court found that the principal error in the Orient Express judgment was to apply the but-for test to the alternative causes of loss: damage to the hotel and damage to the surrounding area. The correct approach would have been to consider the damage to the hotel and the damage to the surrounding areas as concurrent causes of loss. If that approach had been taken, it would have been possible for the hotel to establish that its entire loss was caused by one of those concurrent causes (here the damage to the hotel itself) provided that the other cause was not excluded.
During the High Court proceedings, the parties made lengthy submissions on the type of evidence that might be used to prove the prevalence of covid-19 within a relevant area to trigger coverage. The parties were able to agree that official statistics of various kinds could be used in principle, but there were differences as to the inferences that could be drawn from material of this kind, particularly as to the timing of an outbreak of the disease and the consequence of the fact that the statistics were produced by reference to geographical areas that were not the same as the relevant areas in the policies concerned. In the absence of any factual or expert evidence, however, the High Court was reluctant to go much further than recording the areas of agreement and disagreement between the parties.
The High Court took a similar position with regard to what would be required to discharge the burden of proof in any given case using the above statistical sources as the best available evidence. Again, there was significant agreement between the parties, but the point of contention was in relation to the reliability of proposed methodologies put forward by the FCA. In the absence of any evidence on the point, the High Court was unable to offer any guidance other than encouraging the parties to continue to seek agreement on these issues.
This aspect of the case was not subject to appeal and was not, therefore, considered by the Supreme Court in its judgment.
The Supreme Court's ruling in FCA v. Arch has provided answers to a number of pressing coverage issues in respect of both disease and prevention-of-access policies, but it is far from the end of the story. The wordings considered in this case were sample policies and, while they were intended to be reflective of commonly used wordings and formulations, there will inevitably be many wordings in use that contain different language and that will require careful analysis in their own right. Of equal significance, however, is the Court's reluctance to make any declarations with regard to a number of evidential issues, such as the potential sources of evidence available to establish the occurrence of an outbreak of disease within the relevant area, or what would be required of policyholders to discharge the burden of proof in that regard. This is crucial because without this evidence policyholders whose wordings contain a relevant area qualification cannot show if or when their policy was triggered. It is to be anticipated that the FCA will be anxious to reach agreement on these issues, but, in the absence of such agreement, we may well face a further round of litigation to address this uncertainty.
In the meantime, the FCA has written to all insurers requiring them to address claims on the basis of the Supreme Court's judgment (and those elements of the High Court judgment that were not appealed). Under a recent amendment to Insurance Act 2015, it is now implied into all UK insurance policies agreed after 4 May 2017 that claims will be settled within a reasonable period. A failure by the insurer to comply with this provision will enable the insured to claim damages for breach of contract. The reasonable pursuit of coverage litigation does not amount to delay in this context, but it is clear that insurers will be under significant pressure from both the regulator and insureds to settle claims as quickly as possible now that these common coverage issues have finally been resolved by the Supreme Court.
1 Simon Cooper is a consultant at Ince.
2 A notifiable disease is one in respect of which medical professionals and others have a statutory obligation to notify the authorities. Covid-19 became a notifiable disease in England on 5 March 2020 and in Wales on 6 March 2020.
3 Arch Insurance (UK) Limited, Argenta Syndicate Management, Ecclesiastical Insurance Office, Hiscox Insurance Company, MS Amlin Underwriting Limited, QBE Limited, Royal & Sun Alliance Insurance Plc and Zurich Insurance PLC.
4 Hospitality Insurance Group Action and Hiscox Action Group.
5 Financial Conduct Authority v. Arch Insurance (UK) Limited and Others  EWHC 2448 (Comm). The consequent declarations were made available on 20 October 2020.
6  UKSC 1.
7 Orient Express Hotels v. Assicurazioni Generali Spa (UK) (t/a Generali Global Risk)  EWHC 1186 (Comm).