Coronavirus and its impact on the world will remain in our thoughts for many years to come…Not all of my blogs will start in this way, but the next few might. For the Insurance industry it marked the beginning of a new chapter. Not just in terms of the use of technology in operations but also as it highlights the uncertainty that remains in some policies in relation to cover. Some of you will know that I have been on the campaign trail for a decade or more, raising the profile of Business Interruption (BI) insurance, and finding solutions to issues that still haunt us. In more recent years, it was non damage BI (NDBI) projects at the forefront, so infectious disease is no stranger to me. But I must admit, of all the NDBI risks I had thought that cyber would be the first to bring the world to its knees, not necessarily infectious disease.
As the more specialist non-damage BI policies appeared, we found they were excluding pandemic due to the vast potential cumulative exposure that would attract a similarly vast premium, but NDBI cover was creeping more and more into Property damage (PD) BI policies. Risk managers weren’t necessarily ready to dive in and buy a whole new policy to cover NDBI risks, but were happy to dip their toe in the water through extensions to existing ones.
In the US, infectious disease is often an included cause in Time Element policies through an endorsement, therefore in those cases sometimes cover is offered at full value. In the rest of the world, including the UK, it is mainly dealt with through an extension clause with restricted limits and limited cover. Although, there are some progressive policies that attempt a wholistic approach, and include risks such as infectious disease as a main cause under the policy.
Where the extension clause approach is adopted, it tends to be offered as either:
Coronavirus and ‘civil authority’ orders were given to restrict movement and interaction of people, and lockdowns to try to halt social activity all together. We are in unprecedented times and there is no blueprint for everyone to follow. The uncertainty in cover from PD/BI policies in the UK resulted in the UK regulator for insurance and financial services – the Financial Conduct Authority (FCA) – bringing a ‘class action’ style case on behalf of Insureds against Insurers, using a selection of policies. The aim was to establish clarity of cover and avoid costly, lengthy, individual legal disputes. A difficult task considering each contract is slightly different and each Insured definitely is. According to the FCA, there are over 60 Insurers involved, with 700 different types of policy and over 370,000 policyholders, it has therefore been a major case of incredible significance. 21 policies were chosen to be representative.
The FCA trial was virtual, and streamed live for anyone to view. The FCA website has extensive information available from the full transcripts to useful summary tables. The wordings were published, the Insurers were named, transparency has been paramount throughout. There are many useful summaries available, including one from Herbert Smith Freehills who represent the FCA, but I wanted to share my take on the initial judgement from a claim’s perspective.
To begin with, the case gives us some fabulous phrases that I’m using more and more, including the ‘counter factual’ and ‘quantification machinery’....and who would have thought that the many cases referred to would include the ‘rainy skies’ case, alongside the ‘silver cloud’. In our panel discussion (highlighted in my previous blog) we contemplated what the silver lining to this Covid cloud might be, but had no idea that the cloud itself might actually be silver.
Overall, the Court initially found mainly for Insureds other than where clear exclusions or restrictions were apparent, giving 12 out of the 21 policies the ability to claim under the various extension clauses. The High Court then gave permission for any Insurer appeals to be expedited by an immediate hearing into the Supreme Court (well, the virtual equivalent of), and the judgement of that appeal is due soon. I include a brief summary of where we are to date focusing on key claims issues:
An age-old interpretation issue for PD/BI policies, raised by many over the years (including the Chartered Institute of Loss Adjusters in its wording review from 2010 updated in 2019). In this case, where it was used without a specified/restricted radius, ‘vicinity’ was interpreted very broadly in line with the nature of the cause. As the cause was a highly infectious disease, it was found to mean the whole of the UK. This allowed both a claim under infectious disease as it was deemed to be within the ‘vicinity’, but also a calculation of expected revenue/gross profit ‘but for’ the overall infectious disease.
It is therefore the application of the trends clause which is also vital. In calculating loss of revenue/gross profit in BI, you would compare expected to actual, the difference being the loss. The calculation of expected revenue/gross profit considers standard (previous 12 months), but also applies trends in the business and the industry. This is where, as Forensic Accountants, although we don’t have a crystal ball to see the future, we must establish the expected position using the information available.
This is where the counterfactual comes in. Previously referred to as the scenario ‘But for’ for the event, I do like the use of the term ‘counterfactual’ to explain this calculation. The wording is usually ‘But for’ the damage, and where there is no damage, it must be ‘but for’ an incident. The judgement focuses on ‘causation’ rather than damage or incident, and ‘causation in the counterfactual’, matching proximate cause and counterfactual cause.
The findings were that if the claim was for infectious disease (being the full pandemic in the case of ‘vicinity’) then it is the effect of that full pandemic that must be removed in the calculation of expected revenue/gross profit. This attempts to establish the position the business would have been in if the Coronavirus had not occurred at all. This could greatly increase the expected position, and therefore the loss. In the appeal, it was further argued by Gavin Kealey QC among others, that it was not cases of Coronavirus in the vicinity (in this case within a specified radius) that caused the interruption, but it was the lockdown itself that caused it, and that would have happened regardless of localised occurrences. He identified that there was some confusion over 'what was covered but not causative with what was causative but not covered'.
For many this is a familiar case, so in a nutshell, in the major hurricane season of 2005, Rita and Katrina took its toll on New Orleans and many businesses. A hotel in the area, within the Orient Express hotels group, suffered some damage and therefore claimed BI. But was it the damage to the hotel, or the wide area damage that caused the loss? Did the counterfactual need to reflect the position had the hurricane not occurred, or just the position if the damage to the hotel had not occurred? The case found for Insurers who argued that the hotel should be treated as though it were an undamaged hotel in an otherwise damaged city. The damage to the hotel was the covered cause, and it was the damage that should be removed in the counterfactual. Expected revenue was greatly reduced as a result, and therefore so was any loss. It was found that it was the wide area damage that caused the overall loss, and although the denial of access/loss of attraction clause could be relied upon, those lower limits applied.
The case was referred to in the FCA test case judgement, and was dismissed. It was argued that if the object of the quantification machinery is to put the Insured in the same position if the Insured peril had not occurred, the insured peril must be the causation. The counterfactual is therefore the position if the Insured peril (Infectious disease) had not occurred. The cases are in fact different though, there is no damage to property in the FCA case, therefore it must be an Insured peril, rather than damage, that is the focus. Interestingly, two of the 5 judges in the FCA test case appeal were directly involved in the Orient Express hotels case. We await the judgement with much anticipation…
There are many cases whereby policies have grown over time and developed some inherent contradictions. One example from the past where this was highlighted was the yellow shirt/red shirt civil commotion in Thailand back in 2010, which saw the Government declare it as a terrorism event. With non-damage denial of access extensions (where bomb threat was used as an example) pitched up against a general exclusion for terrorism, it was the exclusion that generally took precedence. However, the FCA test case found that you can’t expressly grant cover in an extension clause and then eliminate it with a general exclusion (although guidance on the scope of the specific cover provided is allowed). Many underwriters are very much aware of the need to have direct exclusions attached to extensions to ensure clarity, but others will no doubt take note of these findings and review wordings to correct any such contradictions.
We are not alone in our contemplation of issues of course, there are many individual cases across the world, and a similar test case in Australia with the Australian Financial Complaints Authority (AFCA) and the Insurance Council of Australia (ICA). After a similar initial ruling mainly in favour of Insureds, and an appeal the ICA requested an appeal to the High Court. In Australia, following SARS scenario testing, general exclusions were introduced for quarantinable or infectious disease. A Quarantine Act ‘carve-out’ was established whereby there would not be cover, and policies would have been priced on this basis. If it is found that the carve out does not apply to COVID-19, cover will be available. However, the CEO of the ICA Andrew Hall warned of the harsh reality faced:
“…If the industry is forced to pay out for risks it has not collected premiums for, or sought reinsurance for, it would compromise our ability to provide the Australian business market with protection against other risks.”
Other countries that have faced strict lockdowns, and the details of each case will be different, as will the policy wordings (after all, a policy is a contract like any other). In France for example, in the case of Axa v Maison rostang, the result came in for the Insured, but the French Insurance Federation estimated BI losses at 60 billion euros – which was estimated as 110 years of premiums. Not a sustainable model. Whereas in the case of Gavrilides Management Company et al. vs. Michigan Insurance Co. in the US, the view of Insurers was upheld.
The FCA case will resolve some of the immediate uncertainty over coverage and policy interpretation, but that’s just the beginning. The normal issues remain in relation to the loss investigation and quantification. We await the results of the appeal, which include vital arguments about Covered vs causative impact, so that if necessary, we can start the next stage in the process.
With regard to furlough funds or government grants received, it’s not just the potential deduction as a saving in relation to claims that is important (see ‘Dear CEO letter’ and FCA statement August 2020), but the impact of any related decisions on loss mitigation. There is an overall duty for an Insured to mitigate any loss, in the case of retailers it might be online alternatives, in the case of restaurants it would be takeaway sales. If a business has decided to take the money from the Government and furlough workers, what impact does this have from a loss mitigation perspective? Self-preservation usually prevails, and Insureds would act as if they were uninsured in the case of a loss, but decisions made early on were without the benefit of hindsight or knowledge of how long Government restrictions and/or this disease might continue. The duty to mitigate was up against a duty to employees.
So, there are many other issues to be dealt with, even after the appeal result comes in. The reinsurance position is not always clear, and quantum issues in determining the counterfactual remain paramount as they become even more remote from reality.
In the past I have commented that some seem to wear the ‘property damage blinkers’ when thinking about BI risk, and here I am, wearing them (for the purpose of this blog at least). Even remaining in our world of insurance I haven’t even touched the liability impact. But to avoid a concise (ish) blog becoming a slog, we must stay within the boundaries set.
In the short term, if Insurers are found liable, we move onto those quantum issues and the analysis of individual cases and circumstances. Independent Forensic Accountants/quantum experts such as ourselves will be vital to that process. In the long term (and irrespective of the outcome), the aim will be to create more certainty in policies to avoid surprises and such deep-rooted disputes. Forensic accountants can again assist, as this can be partly achieved through pre-loss reviews and testing. I’m well known for saying that I would much rather be sitting round having a cup of tea, playing ‘let’s pretend you have a loss’ and discover issues then, rather than wait until after the loss. BI reviews, if combined with Business Continuity testing programmes can be of great benefit to all. Insured values will be more accurate, expectations of cover are managed on all sides, any mitigating actions advised in business continuity plans can be aligned to the BI policies to confirm cover in advance and if lawyers are also engaged upfront to look at wording, some disputes can be avoided. Utopia beckons, but remains just out of reach for the moment…