Reinsuring pandemics: the role of government and public–private partnerships between reinsurers and governments

Abstract

Pandemic-related business interruption (BI) losses are generally considered ‘uninsurable’ because, in order to pool sufficient premium revenue to meet valid claims, premiums would be unaffordable for the majority of policyholders. This paper explores whether and how such losses might be made insurable in the U.K. The authors consider post-pandemic governmental responses, including the role of the Financial Conduct Authority (FCA) and the meaning and implications of FCA v Arch Insurance (U.K.) Ltd ([2021] UKSC 1). The central premise of the paper is to highlight the importance of reinsurance in increasing an underwriter’s insuring capacity and to illustrate how, with the support of government in the form of a public–private partnership (PPP), ‘uninsurable’ risks of this type may be made insurable. The authors propose a PPP, ‘Pandemic Business Interruption Re’, which provides, in their view, a feasible and defensible solution that would confer the benefit of increasing policyholders' faith in the industry's ability to underwrite pandemic-related BI claims and reduce reliance on ex post government aid.

Introduction

In the wake of the COVID-19 pandemic, the insurance industry considers its preparedness for the next pandemic. This paper focuses on pandemic-related business interruption (BI) losses in the U.K. More specifically, it considers the roles of government and the insurance industry in response to the COVID-19 pandemic and the possibility of public–private partnerships (PPPs) between reinsurers and government as a means of reinsuring future BI losses.

The authors share the view that insurance ought to play a more prominent role in compensating pandemic losses, not least because it is more sustainable than ex post emergency financial aid from governments. Indeed, aid of this nature provided immediate solutions in the context of COVID-19, but these programmes can cause uncertainty as well as substantial delays and unfairness—real and perceived—in aid distribution. Insurers, however, did not shy away from stressing the financial burden that COVID-19-related losses would have imposed on them.

We suggest that a large proportion of BI losses (initially, at least) may be made insurable by means of a PPP that balances policyholders’ expectations with the financial feasibility of the industry to indemnify these losses in a fair and feasible way. PPPs are not a new concept for the re/insurance industry, as examined below.

The industry’s role cannot be considered in isolation from the risks insured against and the economic realities that it works under. The industry, as explained below, has withdrawn its services in circumstances where it regards the risk ‘uninsurable’. However, with respect to a number of these risks, the industry in multiple jurisdictions has become part of the solution through a wider collaborative response such as a PPP between insurers and governments. This emphasises how government involvement is crucial in resolving the so-called ‘insurability’ matter (see Hartwig et al. 2020). In this regard, we do not depart from what has already been established as a workable system for both the assured and the insurer: our proposal is an adaptation in the context of BI losses caused by an infectious disease. We propose in this paper that a ‘Pandemic Business Interruption Re’ (PBIRe) would enable insurers, with encouragement and support from government, to insure at a reasonably affordable premium rate, risks so large that they would be considered otherwise uninsurable.

To this end, this paper considers the insurability of pandemic risks as compared to natural catastrophe risks. It then discusses whether post-pandemic governmental responses can provide an alternative to reinsurance covers; in particular, the role of the Financial Conduct Authority (FCA) and the meaning and implications of FCA v Arch Insurance (U.K.) Ltd ([2021] UKSC 1) are considered in detail. Pandemic partnership bonds (PPBs) are also discussed because they have been suggested by many as a means of insuring ‘uninsurable’ risks. The paper is dedicated then to analysing two existing, albeit rare, examples of U.K. PPPs—Flood Re and Pool Re—which may inform the design of the PBIRe proposed in this paper. A case study of a similar PPP in China is referred to as it strengthens the viability of such a scheme. We then consider the defensibility of a PPP that is exclusive to BI losses only, before setting out our proposals for the operation of our PBIRe.

The insurability of pandemic risks

The special challenge presented by insurance cover against low-probability, high-consequence events is not an alien concept to the insurance industry (Kunreuther 2015, pp. 741–762). Considering the limited experience that decision makers have with such events, uncertainty as to their probability of occurrence and their resulting consequences, it is not surprising that insurance decisions often differ from those recommended by normative models of choice (Kunreuther 2015, p. 742).

Insurance, in the U.K. at least, is a commercial enterprise, meaning the insurer needs to cover their operational costs, pay valid claims and make a profit from their business. For the insurer to operate successfully, it is crucial to accurately estimate the probability of a loss and its severity so that a sufficient premium can be charged. Also critical to this assessment is how much of each risk it will retain and how much it will reinsure. Notwithstanding reinsurance’s ability to increase the capacity of insurers, certain pandemic-related losses have been deemed so large as to be uninsurable.

‘Insurability’ is an elusive concept. Berlinger (1985) provided three criteria comprising actuarial, market and societal considerations (see Fig. 6 in The Geneva Association 2020, p. 18) and it is clear that not every conceivable loss is, or can be, insurable. Insurers pool a large number of risks such that at any point in time, the ongoing losses will not be larger than the current premiums being paid, plus the residual amount of past premiums that the insurer retains and invests. Thus, a risk is described as ‘systemic’ when the losses, if eventuated, would cause the industry to implode. Reportedly, insurers are reluctant to underwrite pandemic risks because the number of independently exposed policyholders is too small. This impedes the building of large enough risk pools since the premiums would be unaffordable for the majority of those wishing to purchase insurance protection (The Geneva Association 2020).

Considering their eventuation on an enormous scale and uncontrollable aggregation, pandemic-related losses can be described as ‘systemic’ risks (The Geneva Association 2020, p. 26). Notably, while natural disaster catastrophe risks are comparable in these respects, the magnitude of COVID-19-related BI losses has been proven to have exceeded that of any single catastrophic event (OECD 2021). To provide context, a hypothetical pandemic insurance contract price mark-up purportedly corresponds to the top 20% of natural catastrophe insurance contract prices (Gründl et al. 2021, p. 866).

Another consequence of the above paradox is that faith in the insurance industry has seemingly diminished considerably in many jurisdictions. A recent report conducted by Deloitte confirms that faith in the industry is lowest, of the countries surveyed, in the U.K., where only 45% of businesses trust insurers (Deloitte 2022, p. 8). Nevertheless, 50% of small and medium-sized enterprises (SMEs) reported that their level of trust in their insurer neither increased nor decreased, indicating that trust was already low before COVID-19 (Deloitte 2022, p. 8). In contrast, assureds in China report the highest levels of faith in their insurance industry. This is significant because while the governments of both countries responded to COVID-19 by mandating business closures, China devised a PPP to support businesses financially upon their reopening and facilitate employees in their return to work.

As such, our overarching aim is to consider how faith in the U.K. insurance industry’s ability to underwrite and pay pandemic-related BI risks may be restored. Clearly, the industry has a significant role to play and is supported by reinsurance, but it can only be part of a wider collaborative response since the support of government, especially in terms of incentivising the market to insure such risks, is an essential practicality. Considering its impact on the national economy as a whole, it is arguable that the insurability of pandemic losses is, at a broader level, a public policy consideration (see Hartwig et al. 2020) and BI pandemic risks present a unique combination of viral and political risks (The Geneva Association 2020, p. 26).

Natural disasters vs. pandemic losses

The nature of pandemic losses may differ from natural disaster catastrophes in four key respects (Kong 2021). Firstly, unlike natural catastrophes such as earthquakes and floods, pandemic losses fall neither within, and cannot be contained within, predictable geographic parameters. Secondly, in terms of risk accumulation, pandemic losses impact multiple lines of business simultaneously, as opposed to predictable lines like property and casualty for natural catastrophe losses. Thirdly, pandemic losses correlate with capital market losses: the COVID-19 pandemic caused the global stock market to plummet with the consequence that insurers experienced a ‘double hit’ to their investment portfolios as well as their underwriting business (Gründl and Regele 2020, p. 2). Fourthly, actuarial pricing of risk forms the basis of underwriting policies, yet pandemic losses are caused, for the most part, by government responses to the pandemic rather than the particular virus itself. Moreover, there is a lack of historical loss exposure for pandemics, which makes estimating these risks challenging. Nevertheless, while the same can be said of terrorism risks (Qui 2020, p. 296), as investigated below, such risks can be made insurable by a PPP.

Three further matters require noting. Firstly, when a natural disaster occurs, the losses are incurred simultaneously or within a short period of time. Building on the fourth of Kong’s points, it is possible to discern patterns of events and losses and, accordingly, intervening periods of reprieve (Kong 2021). In contrast, pandemic losses are incurred ordinarily over an extended period of time, which threatens the sustainability of re/insurers’ risk pools, thereby contributing to the increase of premiums to unaffordable levels. Secondly, ‘external moral hazard’ considerations—these refer to third parties making ‘insurance-induced’ decisions that impact claims (Richter and Wilson 2020, p. 173)—are not a concern for natural disasters (Jaffee and Russell 1997) or pandemic losses. We are of the view that ‘external moral hazard’ considerations are not a concern because there are almost no risk mitigation steps that businesses could deploy to prevent pandemic BI losses, as COVID-19 has demonstrated; hence, arguing that insurance would discourage people from taking risk-mitigating steps is unlikely to be pertinent regarding a pandemic. Thirdly, ‘insurability’ comprises both a supply and a demand component: policyholders generally underestimate the likelihood of a pandemic and presume emergency financial aid from governments will be available should one arise (The Geneva Association 2020; Hartwig et al. 2020). For instance, Marsh, Munich Re and Metabiota introduced ‘Pathogen Rx’ in 2018, a product to transfer pandemic BI losses; however, it received little uptake (Qui 2020, p. 295). Furthermore, the Deloitte report suggests that after COVID-19, only 19% of businesses in the U.K. (that responded to the survey) are more likely to purchase business insurance cover in the future (Deloitte 2022, p. 4).

The nature and function of reinsurance: in brief

The previous section highlights that the need to encourage reinsurers to back insurance cover for pandemic losses is as essential as for any other large losses, if not more so due to their magnitude. It will be useful here to contextualise the significance of reinsurance by outlining how these contracts operate.

Reinsurance is defined as the insurance of insurance companies ((Merkin 2022, [18-001]; Butler and Merkin (2022); Edelman and Burns 2021; O'Neill and Woloniecki 2019; Gürses 2010; Travellers Casualty & Surety Co of Europe Ltd v Commissioners of Customs and Excise [2006] Lloyd’s Rep. I.R. 63). Through reinsurance, insurers maintain and expand their insuring capacity and safeguard their solvency so that when big losses have to be compensated, insurers can pay the claims without encountering a financial disaster themselves. It is possible for any insured risk to be reinsured, irrespective of the insured amount. Reinsurance cover may be provided in a number of different forms depending on the relationship between the assured and the insurer and the type of insurance contract (Merkin 2022, [18-002]–[18-008]).

Firstly, reinsurance contracts may be proportionate or non-proportionate. In the former, a straightforward proportionate share of premium and risk applies to the relationship between the reinsurer and the reinsured. In the latter, the reinsurers reinsure in excess of the amount that the reinsured (the insurer) retains. Thus, if the original assured’s loss does not exceed the reinsured’s retention, the reinsurer’s liability will not be triggered. This is in contrast to the proportionate form of the contract where the reinsurers and the reinsured share the loss, irrespective of the size of the assured’s claim (naturally up to the maximum limit they agreed to cover). Stop-loss reinsurance is particularly worth noting here, whereby the reinsurer agrees to assume, up to a stated amount, all loss amounts of the reinsured, in excess of a stated amount (in Lord Napier and Ettrick v RF Kershaw Ltd Society of Lloyd's v Woodard [1999] Lloyd's Rep. I.R. 329), the Names, as members of the Outhwaite Syndicate, insured themselves against part of any loss they might incur as members of the syndicate. The stop-loss insurers agreed to “Indemnify the Assured for the amount by which the Assured’s overall ascertained nett underwriting loss as hereinafter defined for the Underwriting Year(s) of Account shown in the Schedule exceeds the amount stated as ‘Excess’ in the schedule”. For the purposes of illustration for the 1982 year of account, a particular hypothetical Name suffered a net underwriting loss of USD 160,000—the excess was USD 25,000 and the limit was USD 100,000. On these figures, the stop-loss insurers paid to the Name USD 100,000, this being the fixed amount of the limit (USD 100,000) which exceeded the excess (USD 25,000).

The original assured’s loss determines the insurer’s liability and the insurer then makes a claim against the reinsurer: reinsurance is a contract between the reinsured and the reinsurer and, as such, the original assured is a third party that cannot make a claim against the reinsurer.

Reinsurance’s contribution to the insurability of risks is undeniable, particularly to those which are incurred on such an enormous scale that they would otherwise be considered ‘uninsurable’.

In the U.K., insurance is a private industry. The London Companies Market and Lloyd’s of London are able to insure risks that are particularly high because the availability of reinsurance increases their capacity. Lloyd’s, in particular, is able to underwrite high risks due to its unique method of ‘subscription’ to insurance and reinsurance covers. Nevertheless, there are some risks that represent financial losses on a scale that even Lloyd’s underwriters are reluctant to insure. Floods are one example and terrorism is another. PPPs have been set up, ‘Flood Re’ and ‘Pool Re’, respectively, to facilitate the insurance of flood and terrorism risks. These PPPs evidence that government involvement in reinsurance, while rare in the U.K., is not unprecedented.

In some countries, insurance and reinsurance of such large risks may be a domestic matter, albeit rarely. By way of example, in France, ‘CCR’ is a public-sector reinsurer that provides cover against natural disasters and ‘uninsurable’ risks.¹ CCR Re operates as an open market reinsurer for a broader range of risks.²

However, the norm is reinsuring risks internationally. The largest market is the U.K. (London Market Group 2020) and others include Germany and Switzerland. In fact, some countries, like Sweden, do not facilitate reinsurance and all domestic risks must be reinsured internationally. This is particularly important in countries where a natural catastrophe may have a detrimental impact on the economy’s growth, e.g. in South America.

Reinsuring the uninsurable

In order to illustrate how reinsurance can make ‘uninsurable’ risks insurable, we briefly summarise the operation of The International Group of P&I Clubs.³ In a different manner to traditional insurers, P&I Clubs are independent, not-for-profit, mutual insurance associations whose members are shipowners and charterers.⁴ Each Club insures the other members’ protection and indemnity (P&I) liabilities, namely liabilities a shipowner might incur to cargo owners, seamen and selected third parties in the use of their ships. Individually, Clubs are competitive; however, 13 P&I Clubs comprise the International Group (the Group), which enables the P&I Clubs to share their large loss exposures between them. Nonetheless, as the risks that P&I Clubs are required to insure increase, their insurability is made possible only with reinsurance backing. As seen below, the cover is provided through non-proportionate reinsurance arrangements, which reach as much as billions of dollars.

The primary function of the Group is the coordination and operation of the P&I Clubs’ claims from the pooling arrangements. Liabilities that exceed the individual P&I Club’s retention (which is currently set at USD 10 million) are shared between all 13 Clubs in accordance with the terms of the ‘Pooling Agreement’.

The Pool is structured in three layers from USD 10 million to USD 100 million. Hydra Insurance Company Limited (Hydra) is the Group’s captive reinsurance vehicle and is incorporated in Bermuda. Hydra reinsures each Club’s liabilities in excess of USD 30 million. Through the participation of Hydra, the Group’s Clubs can retain, within their Hydra cells, premiums which would otherwise have been paid to the commercial reinsurance markets.

The annual Group General Excess of Loss (GXL) reinsurance programme attaches at the Pool ceiling of USD 100 million and provides up to USD 2 billion of reinsurance cover in a four-layer structure (Layer 1: USD 450 million excess of USD 100 million; Layer 2: USD 200 million in excess of USD 550 million; Layer 3: USD 750 million excess of USD 750 million; and Layer 4: USD 600 million excess of USD 1.5 billion).

A further USD 1 billion of reinsurance cover (the Collective Overspill) is purchased by the Group to provide protection in respect of claims exceeding the upper GXL cover limit of USD 2.1 billion.

To illustrate by way of example, an oil tanker accident may cause a large-scale loss that impacts not only the immediate environment but extends to the coastline and the businesses and individuals who occupy the areas affected by the accident. It is the shipowner’s liability to ensure that the environmental damage the accident has caused is cleaned up and that the losses that individuals and businesses suffered in the areas affected by the accident and the oil pollution are compensated.⁵ Such a huge financial burden would be impossible for one shipowner, or even one P&I club, to bear on their own account. However, the insurance and reinsurance arrangements make it possible to compensate such losses and, in effect, virtually guarantee as much.

Can post-pandemic governmental responses be an alternative to reinsurance cover?

Two alternatives have been proposed for risk sharing in extreme disastrous events: ex ante through an insurance market in which premiums are paid into a common pool, and ex post relief paid for by a poll tax levied on each citizen (Jaffee and Russell 2013, p. 475). We propose that the former is preferable.

Undeniably, post-disaster government aid around the world will be shaped by a broad range of social, political and economic factors (Jaffee and Russell 2013, p. 488). Before we set out our reasons in detail for our preferred view, it will be useful to examine insurers’ handling of BI insurance cover after the COVID-19 outbreak in the U.K.

The coronavirus pandemic led to widespread disruption and business closures, which resulted in substantial financial loss. A large number of claims were submitted to insurers; however, there was widespread uncertainty in the market as to whether certain BI policies responded to the losses and concern was also expressed as to the basis on which some firms decided claims (FCA 2020a). Several businesses and groups of businesses indicated their intention to challenge the rejection of their claims.

The FCA, as the conduct regulator of insurers in the U.K., had an interest in the resolution of this uncertainty. It brought a test case, acting in a manner compatible with its strategic objective to ensure the relevant markets function well and to advance its operational objectives to ensure appropriate protection for consumers and market integrity (FCA 2020b).

This intervention had procedural advantages: some 370,000 policyholders were identified as potentially affected by the outcome of the test case (FCA 2021a) and consolidated action prevented significant administrative problems for the courts (where litigation, as opposed to arbitration or other dispute resolution avenues, is contemplated, saving both time and policyholder expense in bringing these claims). Moreover, there was also a significant risk that different courts and tribunals might reach inconsistent decisions on materially similar issues, leading to further costs and uncertainty for insurers and policyholders. In particular, a swifter resolution of the uncertainty was considered to enable the FCA to fulfil its regulatory objectives more expediently.

The intervention was possible following a framework agreement, reached between the insurers involved in the case and the FCA, and the latter’s use, for the first time in its history, of the ‘Financial Markets Test Case Scheme’ (Practice Direction 51M). The Scheme allowed (the Scheme no longer exists, see below) issues of ‘particular public importance’ that required ‘immediately relevant authoritative English law guidance’ to be heard without a cause of action between the parties.

In collaboration with insurers and stakeholders, the FCA selected 21 sample policy wordings, the meaning of which was to be determined by the court. All of the 21 clauses were non-damage BI insurance clauses—damage BI clauses require that physical loss of, or damage to, the insured premises must be established and COVID-19 did not cause any material alteration to the properties (Merlin v British Nuclear Fuels Plc [1990] 2 Q.B. 557; Blue Circle Industries Plc v Ministry of Defence [1999] Ch. 289).

In the test case, FCA v Arch Insurance (U.K.) Ltd, the Supreme Court discussed a number of matters including: (1) the meaning of ‘occurrence’, (2) whether, when SARS-V2 was listed as one of the infectious diseases insured against, it was also intended to include COVID-19, which was also highly infectious, and (3) whether the cause of the loss, the occurrence of COVID-19, was indivisible. The Supreme Court’s assessment of how the assured proves that the loss suffered falls under the insuring clause, i.e. proof of the cause of their loss, is especially novel. This is a matter which falls outside the scope of this paper (see Gürses 2021, 2023; Meggitt 2022).

After the Supreme Court judgement was handed down, the FCA wrote to insurers and actively encouraged them to reassess BI claims and pay valid ones as soon as possible. Indeed, the FCA made it clear that if insurers failed to do so, it would rely on its full range of regulatory tools and powers to force them as such (FCA 2021b).

The test case provided much needed certainty for policyholders. However, it was by no means the ideal solution for all parties involved.

Ordinarily, commercial disputes between professional assureds and insurers do not attract this level of attention; however, the scale of COVID-19-related BI losses impacted the country’s economy, thereby elevating their importance above the confines of an individual-assured insurer relationship. The FCA’s role is to ensure that financial markets work well for individuals, businesses and for the economy by providing honest, fair and effective products and services and ensuring that consumers get a fair deal. To this end, the FCA’s operational objectives are: (1) to protect consumers and financial markets and (2) to promote competition. However, the FCA does not have the authority to amend the contractual terms between assureds and insurers, irrespective of its desire to promote the interests of policyholders.

As the FCA test case has illustrated, whilst some businesses benefited from its outcome, its scope was limited. Some policyholders, namely those whose policies required a physical loss of or damage to property, fell outside the scope of the test case and received no compensation from insurers. Moreover, of the 21 policy wordings tested, only 14 were found to have responded to COVID-19-related claims. Additionally, it has been challenging for the FCA to supervise how efficient insurers have been at providing compensation for assureds that made a claim following the test case.

In any event, the FCA test case appears to be the only one to have been heard and indeed the only one that will ever be heard under the FCA test case scheme because the Practice Direction rule 51M has been revoked, with no explanation as to why.⁶ With regard to the question posed—whether post-pandemic governmental responses may be an alternative to reinsurance cover—the answer, in the U.K. at least, is negative.

Existing public–private partnerships

Pandemic partnership bonds

Strictly speaking, bonds are neither PPPs nor insurance products in the traditional sense; however, they involve a collaboration between governments and private investors. Since numerous authors have suggested that they might be utilised to make ‘uninsurable’ losses insurable (including in Spaeter and Schmitt 2022), it is helpful to consider them, in brief, to contextualise why we favour PPPs over bonds for U.K. pandemic-related BI losses.

Two ‘pandemic bonds’ were issued by the World Bank in 2017 to transfer pandemic risks in low-income, developing countries to the financial markets in the event of an outbreak of six specified diseases (Orthomyxoviruses new influenza pandemic virus A), Coronaviridae (SARS, MERS), Filoviridae (Ebola, Marburg), Crimean Congo, Rift Valley and Lassa fever) (World Bank 2017). The bonds aimed to provide funds to the Pandemic Emergency Financing Facility (PEF), a mechanism for processing these funds to countries where an infectious disease was likely to escalate into a pandemic, and received overwhelming support—the transaction was oversubscribed by 200%. The PEF has two ‘windows’. The first is an ‘insurance’ window—funded by Japanese and German premiums and developed in conjunction with Swiss Re and Munich Re, it comprised bonds and swaps. The second is a ‘cash’ window, whereby Germany provided EUR 50 million of initial funding.

These bonds attracted criticism for their complexity and high running costs, which were seemingly disproportionate to their low value of USD 320 million. Gründl and Regele (2020) draw upon these criticisms and propose the issuing of a PPB to fund future pandemic-related losses, which they specifically label an ‘insurance solution’.

Their PPB would require a fixed amount of funds to be raised from private investors—the suggested figure is EUR 20–50 billion to cover ‘significant’ costs related to a pandemic outbreak (Gründl and Regele 2020, p. 2). The monies are put into a trust fund, out of the reach of governments. The funds are invested in capital market instruments for a fixed maturity period, with member state governments paying additional monies into the trust fund to provide a high rate of return for the investors as compensation for their exposure to pandemic risk. Gründl and Regele (2020) refer to this as an ‘insurance premium’ similar to the context of an ordinary CAT bond (Zonggang and Ma 2013).

If a pandemic risk eventuates, the funds are made available to the governments proportionate to the share of their original ‘premium’ contributions. It would be open for governments to pay the ‘premiums’ of other governments, in effect subsidising them.

If a pandemic does not eventuate, the investors receive the returns, and Helmut and Fabian suggest that whether the funds are to be paid back upon maturity depends upon ‘multiple triggers’. The idea is that if an infectious disease outbreak appears to have the potential to escalate into a pandemic, the funds are not returned and multiple triggers are necessary to reduce the moral hazard of the contributing governments from “‘pulling the trigger’, and thus [reducing] the risk premium required by the investors’” (Helmut and Fabian 2020, p. 3).

As Gründl and Regele (2020, p. 4) point out, a PPB is not the appropriate tool to cover all potential pandemic costs because these are very difficult to predict; as such, it would be necessary to outline at its inception the precise risks against which the funds insure.

We perceive two primary difficulties and one more general difficulty with the PPB solution for BI pandemic-related losses in the U.K. Firstly, the scale of BI losses necessarily imposes a natural limit on investor demand, whether collectively or nationally. Therefore, it might prove challenging to raise sufficient capital to make this a viable solution. Secondly, such enormous funds would be too exposed to the volatility of the financial markets, which is compounded by Russia’s invasion of Ukraine. Indeed, BI losses are the pandemic-related losses likely to cause the sharpest decline in the financial markets, particularly in comparison to the 2008 financial crash, after which CAT bonds became significantly correlated with the market, even if the effect was ‘relatively small’ in comparison to alternative asset classes (Carayannopoulos and Fabricio Perez 2015). Thirdly, any decision to select BI pandemic-related losses for protection in collaboration with private investors in preference to other pandemic-related losses may not be popular politically. The electorate is likely to perceive commercial enterprises as having the means to pay higher insurance premiums for the higher level of cover they require. In comparison, other risks may be deemed more worthy of protection by the electorate. The National Health Service’s pandemic-related risks is one example: ringfencing funds to cover staffing (including agency staffing costs) and equipment costs (such as extra ventilators and the like) as well as the price of outsourcing urgent but non-pandemic-related cases to private hospitals for treatment would gain more support, comparatively.

Flood Re

The U.K. government has been working with insurers since 2000 to make home insurance more affordable as the risk of flooding rises in the U.K. In 2016, the Flood Insurance Statement of Principles was replaced with a reinsurance scheme, Flood Re [Flood Reinsurance (Scheme Funding and Administration) Regulations 2015/1902 and Flood Reinsurance (Scheme and Scheme Administrator Designation) Regulations 2015/1875, Water Act 2014].

Flood Re is a joint initiative between the government and insurers. A not-for-profit, publicly accountable scheme, it is run and managed by the insurance industry. Its aim is to promote the availability and affordability of household insurance policies for those who own and live in properties in flood risk areas and manage the transition to policies that fully reflect flood risk (Flood Re 2018). By 2039, it is expected that insurers will have more information on the likelihood of flood risk occurrences for individual properties to price them on the free market; in the interim, the onus is on the government to have invested in flood mitigation measures, rendering the need for government support redundant. However, the feasibility of this objective is uncertain given that flood risk is ‘underestimated and increasing’ due to, inter alia, climate chance.

It is expected that approximately 350,000 properties will meet the eligibility criteria and benefit from Flood Re, although there is no cap.⁷ This figure represents about 2% of eligible U.K. households. Properties built after 1 January 2009 are not covered by the scheme—to disincentivise home building in flood risk areas—and flats in leasehold blocks containing four or more homes and those used for business purposes are also excluded (reg 2015/1902, Reg 5).

Flood Re is funded from two sources—three including the policy deductible.⁸ Firstly, a levy is charged to each policyholder irrespective of their property’s flood risk. In this respect, those occupying lower-risk properties subsidise those in higher-risk properties since high-income areas and high-value properties benefit relatively more (Bank of England 2022). As of April 2022, the levy is stated as providing GBP 135 million per year, down from GBP 180 million⁹; in 2015, the levy was calculated to be just under GBP 11 per policy (Davey 2015, pp. 28, 30). In ordinary circumstances, policyholders are not expected to subsidise each other. However, in a situation where huge losses occur as a result of a natural event (or a terrorist attack, see below), public policy considerations appear to justify the watering down of the traditional assessment of the risk–premium ratio to support the social policies of the government.

Secondly, when the flood risk part of the policy reaches a certain threshold, the insurer can choose to reinsure this risk with Flood Re. If an insurer does so, it pays a premium to Flood Re. Consequently, only when the cost of covering the flood risk becomes more expensive than the cover offered by Flood Re will it make commercial sense to cede the flood risk to Flood Re. While the insurer is at liberty to choose which flood risks to cede to the scheme, the premiums charged by Flood Re are fixed and determined in accordance with the property’s’ Council Tax band (CTB). Importantly, the premium is artificially lower than the market price, which would take account of the flood risk; setting the premium with reference to the property’s CTB takes into account the householder’s ability to pay and promotes affordability.

The question posed here is whether Flood Re offers a viable template to reinsure pandemic losses. It is hard to assess its success in terms of customer satisfaction because policyholders buy their insurance from insurers or brokers in the usual way and make claims directly to the insurer—they do not deal directly with Flood Re. In fact, it is plausible to assume that the majority of policyholders are not aware of Flood Re’s existence [similarly, they will be unaware that Flood Re then retrocedes (reinsures again) its liabilities to ensure it is able to meet a plethora of simultaneous claims]. Consequently, its success is best judged in terms of its ability to make these risks insurable in the first instance.

It is important to note the differences in the nature of both the policyholders and the risks to be made insurable. Namely, policies that can be ceded to Flood Re insure residential dwellings and any claim monies are for the reparation of physical flood damage. As such, mortgage lenders will require a comprehensive buildings policy as part of the terms of the loan and, as valuable assets, those owning their homes outright will also desire comprehensive cover. In comparison, BI cover is taken out to compensate an insured for loss of earnings/revenue when its business has to suspend operations, usually following a physical loss or damage to commercial property; as such, it is usually issued as an adjunct to a commercial property insurance policy. An example is a fire that causes a shop to cease trading while repair works are carried out: the BI cover will ordinarily compensate the shopowner for loss of revenue, allowing it to make its mortgage payments and pay employees a wage, with the intention that once the physical damage is fixed, the business can continue trading. While commercial property buildings insurance will be required by a mortgage lender, the BI component is almost certainly optional and purchased only when it makes financial sense to do so.

COVID-19 demonstrated, harshly for many businesses, that damage BI insurance clauses are not suitable for businesses that suffer loss due to closure of premises or restrictions on their activities because of government orders vis-à-vis an infectious disease: it is not possible to prove physical damage to the property insured as there is no ‘material alteration’ (FCA v Arch). Consequently, it would only be appropriate for non-damage clauses to be ceded to our PBIRe. At present, there is a significant gap in the market in this respect in Europe (FERMA 2022) and insurers need support and encouragement to develop this line of business in response to the needs of their customers.

Due to the enormous capital pools required, it appears financially necessary for every BI policy purchased, irrespective of whether it covers pandemic-related losses or non-damage losses, to charge a levy to help fund Pandemic BI Re. It is also difficult to see why a large IKEA store ought to be charged the same levy as a small, independent restaurant, for example. In reality, however, given the scale of the capital required, it would likely prove a necessity to make larger businesses pay a correspondingly larger premium. Indeed, Gründl et al. (2021, Sect. "Existing public–private partnerships") suggest that it will take a century to accumulate enough funds to pay for the next pandemic, which industry experts have suggested might be only 30–40 years away (The Geneva Association 2020, p. 9).

Utilisation of the insurance framework also necessitates that each insurer that cedes risk to Pandemic BI Re be charged a premium. It is suggested here that ‘business rates’ offer a useful yardstick in this context, as council tax does for Flood Re. Business rates are charged by local councils to non-domestic properties including shops, restaurants and holiday homes let for commercial purposes.¹⁰ In England and Wales, the amounts charged depend on the property’s ‘rateable value’, which is assessed with reference to a tiered structure. Businesses with a rateable value of below GBP 15,000 are eligible for ‘Small Business Rates Relief’ (SBRR)¹¹ and, if the rateable value is GBP 12,000 or less, the relief is 100% (in other words, these properties do not pay business rates). In response to COVID-19, businesses in receipt of SBRR were eligible for government grants to support them through periods of lockdown closures. Accordingly, the business rates scale seems to be a defensible yardstick to help structure the premiums insurers would be charged for ceding pandemic-related BI risks to PBIRe to ensure their affordability for policyholders.

Insurance of terrorism risks: TRIA and Pool Re

In the U.S., after 9/11, insurers reassessed terrorism risks. For a while, terrorism coverage was scarce because primary insurers excluded it from their commercial policies and reinsurers were unwilling to reinsure policies in urban areas that were perceived to be vulnerable to attack (Information Insurance Institute 2021). Both the difficulties of insuring terrorism risk and concerns about the limited availability of terrorism coverage after the 9/11 attacks led to the enactment of the Terrorisms Risk Insurance Act (TRIA) in 2002. The provisions of TRIA appear in a note of the U.S. Code (15 U.S.C. § 6701 note) and, therefore, references to the provisions of TRIA are identified by sections of the law.

The Secretary of the Treasury administers the programme with the assistance of the Federal Insurance Office. TRIA § 101(b) states that the purpose of the TRIA is to establish a temporary federal programme that provides for a system of shared public and private compensation for certain insured losses resulting from certified acts of terrorism, in order to:

• (1) Protect consumers by addressing market disruptions and ensure the continued widespread availability and affordability of property and casualty insurance for terrorism risk; and

• (2) Allow for a transitional period for the private markets to stabilise, resume pricing of such insurance, and build capacity to absorb any future losses, while preserving state insurance regulation and consumer protections.

TRIA recognises that the ability of the insurance industry to cover the unprecedented financial risks presented by potential acts of terrorism in the U.S. can be a major factor in the recovery from terrorist attacks, while maintaining the stability of the economy (TRIA § 101(a)(3)).

The programme was renewed, initially for an additional six years, and then, by the Terrorism Risk Insurance Program Reauthorization Act (TRIPRA) of 2019 (Pub. L. 116-94, 133 Stat. 2534, ‘the Act’), it was extended to 31 December 2027. The programme acts as a type of reinsurance for commercial property and casualty insurance policies. Each insurer that falls under the definition of ‘entity’ under the Act must join the programme, which only comes into effect when an act of terrorism is ‘certified’ by the Secretary of the Treasury as falling within the definition of terrorism and triggers an ‘event dollar threshold’.

The argument that catastrophic terrorism risk is uninsurable typically focuses on lack of public data about both the probability and severity of terrorist acts (Congressional Research service 2019). In the U.S., however, as a replacement for large amounts of historical data, insurers turned to various forms of terrorism models similar to those used to assess future hurricane losses.

In other lines of insurance such as personal auto, insurers know from historical experience roughly how many covered losses to expect (frequency) and what the costs of those losses will be (severity). This data is used to calculate a premium equal to the risk the insurer is assuming in issuing an insurance policy. Sometimes there is even enough data on natural catastrophes to allow for measurements of frequency and severity, often supplemented with catastrophe modelling.

For terrorism risk, on the other hand, frequency and severity data is scarce. There have been relatively few terrorist attacks in the U.S., so there is little data on which to base estimates for future losses. Furthermore, the range of possible severity of terrorism claims is much larger than in other lines of insurance.

Consequently, despite the difficulties of insuring against terrorism risks, acts of terrorism may be covered under various personal insurance policies. It should be noted here that TRIPRA requires insurers to offer commercial insurance for terrorism risks—initially the uptake by consumers was about 27% but this has steadily increased each year and the price for terrorism cover has been lowered.

The triggering event threshold is the required amount of total insurance industry losses from a certified act of terrorism before federal assistance begins. It was initially USD 5 million, which was raised to USD 100 million and then in 2015 to USD 200 million.

The programme operates in the form of excess of loss reinsurance: if the triggering event threshold is met, each individual insurer participating in the programme is subject to a deductible. The deductible for each insurer is calculated as 20% of the insurer’s direct earned premiums for commercial insurance.

Losses in excess of an individual insurer’s deductible are then shared between the federal government and the individual insurer. The individual insurer’s co-pay of the excess losses is 20%, and the other 80% will be paid by the federal government. TRIPRA caps all losses in a programme year for both insurers and the government at USD 100 billion. Insurers cannot be made to pay for losses in excess of USD 100 billion. Losses above this threshold are therefore uninsured.

The Secretary will collect ‘terrorism loss risk-spreading premiums’ from insurers if federal payments are made to insurers. Under this mechanism, known as ‘recoupment’, programme participants may be required to collect funds from policyholders by placing a surcharge on insurance policies written in TRIP-eligible lines. The surcharges will be set based upon the amount that must be recovered by the Treasury and the time within which it must be recovered, as required by TRIA. Insurers must then remit these surcharges to the Secretary (TRIA §§ 103(e)(7)–(8)). The requirement to collect, or recoup, terrorism loss risk-spreading premiums applies not only to insurers that received federal payments under the programme, but also to all insurers writing policies in TRIP-eligible lines of insurance. Surcharges are placed on all TRIP-eligible insurance policies regardless of whether the policyholder purchased terrorism risk insurance (TRIA § 103(e)(8)).

The TRIA programme has been designed to work in the background through the private insurance system. For these TRIA-eligible lines, insurers are required to make coverage available for new and renewal commercial policies “for insured losses that does not differ materially from the terms, amounts, and other coverage limitations applicable to losses arising from events other than acts of terrorism”. For example, if an insurer offers a policy covering a commercial building for fire damage due to some accidental cause, it must also offer a policy covering that building for fire damage due to terrorism.

TRIA requires an offer but does not mandate the purchase of terrorism risk insurance.

The insurer must allow an insured to decide whether they want terrorism coverage on the same terms and conditions as the insurer offers in their non-TRIA coverage. If the assured selects to purchase the cover for terrorism, the insurer must clearly and conspicuously state the premium charges resulting from coverage for certified terrorist acts. If the assured elects not to purchase a cover for terrorism, the insurer may then add terrorism exclusions to the policy.

According to the Federal Insurance Office (2022), the market for commercial property and casualty insurance in general and terrorism risk insurance in particular remains relatively stable, with insurance both affordable and available to U.S. policyholders. Terrorism risk insurance is available in the market for a relatively low cost and is purchased by the majority of commercial policyholders in the U.S., including by the majority of places of worship. Private reinsurance capacity for terrorism risk insurance has increased since the creation of the programme (Federal Insurance Office, U.S. Department of the Treasury 2022).

Many other countries, after 9/11, have also established some sort of terror pool or government-sponsored insurance entity to spread losses across primary insurers and reinsurers.¹² Similarly, in the U.K., following a number of terrorist attacks, reinsurers withdrew from the terrorism reinsurance market because reinsuring these risks was too expensive and too difficult to model to be commercially viable lines of business (HM Treasury 2020).

Subsequently, Pool Re was established by the Reinsurance (Acts of Terrorism) Act 1993 c.18. Pool Re is a mutual company owned by its members—primarily commercial insurers—to which the U.K. government extended an unlimited, but repayable, public financial guarantee. In other words, Pool Re is a reinsurer for terrorism risks which, at first sight, appear to be uninsurable. The government’s guarantee also means that those unaffected by terrorism risks subsidise those who suffer these losses.

Pool Re’s purpose is to enable the U.K. insurance market to underwrite terrorism risk to commercial property at risk-reflective rates. Since its establishment, Pool Re has paid out, without needing to call on the government’s guarantee, GBP 635 million in relation to 13 claims arising from certified terrorism events in the U.K.

Significantly, the Reinsurance (Acts of Terrorism) Act 1993 c. 18, Sect. "The insurability of pandemic risks" was amended by the Counter-Terrorism and Border Security Act 2019 c.3 to include non-damage BI cover in the scope of Pool Re’s liability. Before the amendment, Pool Re could only reinsure losses incurred if a company’s premises had been physically damaged by terrorists.

An existing PPP for pandemic-related BI losses: China

In Hainan province in February 2020, ‘comprehensive epidemic prevention and control insurance’ was launched, encouraged by The China Banking and Insurance Regulatory Commission (CBIRC). It covered ‘key’ businesses for losses including employee remuneration and quarantine and production losses for the period of the government’s mandated lockdown (The Geneva Association 2021, p. 42). Here, ‘key’ refers to those companies most heavily impacted by COVID-19 lockdown measures. The aim was to facilitate employees’ return to work and the resumption of business enterprise by compensating businesses for employee remuneration and production losses during the mandated lockdown period. Later, the scheme was rolled out to a total of 32 provinces.

The cover was limited to a maximum compensation of RMB 2 million (USD 300,000) for a single enterprise. Specifically, regarding salary expenditure, the compensation was based on the monthly salary of quarantined employees and the maximum compensation limit is RMB 6000 (USD 900) per person per month. A further limit on cover was that the period of insurance was limited to six months.

There is uncertainty, to an extent, as to how precisely this PPP was funded in that information regarding levies and premium rates is not readily available. However, The Geneva Association, having been informed by Gong Xinyu (a Task Force Member, PICC), report that the government provided additional subsidies for insured enterprises (RMB 120,000 or USD 18,000). Moreover, at the provincial level, insurers co-insured the risk underwritten; in Hainan, it was reported that the risks were co-insured by 12 P&C insurers’ co-insurers (The Geneva Association 2021, p. 42).

This demonstrates that a PPP is a viable means of insuring risks that were previously considered ‘uninsurable’—providing limits to cover are incorporated into the framework—in a manner that boosts assureds’ faith in the industry. Indeed, 45% of SMEs in China not covered for pandemic-related risk are interested in purchasing cover, compared with an average of 25% in those countries surveyed by Deloitte (2022, p. 6), and SMEs in China expressed the most interest in purchasing business insurance (Deloitte 2022, p. 4). This reflects the high amount of trust Chinese insureds have in their insurers—68% compared to 42%, which has only increased in light of the PPP and stands at 82% of insureds compared with 67% across other countries (Deloitte 2022, p. 4).

The defensibility of a PPP for pandemic-related BI losses

Pandemic-related BI losses were significant financially worldwide. They amounted to trillions of USD and were said to have exceeded the risk-taking capacity of global BI insurers by a factor of more than 100 (The Geneva Association 2020, p. 26). Moreover, there is a protection gap of more than 99% if global insured BI losses from COVID-19 in 2020 alone are estimated at USD 20–40 billion because claims would cover less than 1% of COVID-19 global GDP losses (The Geneva Association 2020, p. 9). Thus, insurance can be described fairly as ‘the economic stabiliser’ (Qui 2020, p. 294, emphasis added) and is crucial to economic resilience.

For this reason, we believe that it is defensible to select BI losses to ringfence for cadence to PBIRe. Moreover, we consider that the wide-reaching ripple effect of BI losses strengthens this view. From a failure to compensate these losses, at least partially, potentially flows a multitude of losses which manifest, for example, in increased costs of consumer goods and unemployment, and the latter can lead to a decline in health, poverty and even homelessness. In this sense, they impact societies, not just individuals, particularly as unemployment and health expenditure will be compensated by the ‘public purse’.

The question that remains, however, is whether it is defensible to use the framework of insurance for the greater social good. The only parties to insurance contracts, which are private contracts, are the insurer and the policyholder: it is therefore difficult to identify any ostensible social purpose. As Davey (2015, p. 26) points out, insurers are quick to allay any assertion that they have an important social function. Moreover, while insurance is a novel way of dealing with uncertainty, it is neither an instrument of social policy to compensate victims nor a tax to redistribute wealth (Walters 1981).

Naturally, private insurers may not share the view that insurance is a union for mutual aid, the initial promise of which is solidarity (Frezal & Barry 2020); however, in the face of catastrophic risks, insurance can operate on the basis of the complementary principles of solidarity and equity (Harper 1993). This is justifiable given that the consequences of these catastrophic risks are shared, undeniably, by the population. Subject to this view, insurance becomes a risk-sharing compensation scheme whereby those who are less fortunate can be subsidised by those who are more fortunate, regardless of the degree and terms of their actuarial risk classification (Liukko 2010).

A balanced and feasible approach appears to be provided through PPPs. For instance, after 9/11, as well the steps referred to above, to ensure insurance of terrorism, many governments provide large and sometimes unlimited guarantees over and above the losses that can be borne by terror pools or government-sponsored insurance entities (European Central Bank 2007). Such arrangements reflect the reality that such large losses potentially have economic domino effects when insurers are faced with insufficient financial resources to cover all claims. Furthermore, as observed during the peak of the COVID-19 outbreak, governments could have no choice but to provide aid to households, companies and insurers who suffer devastating losses from devastating events, even if they are not insured.

In the U.K., it is arguable that the two PPPs analysed above, as joint initiatives between the government and private insurers, use the framework of insurance to promote the social policy objectives of the government. Flood Re and Pool Re demonstrate that those who are less fortunate can be subsidised by those who are more fortunate, thereby replacing individual prudence with mutualisation. By analogy, the implementation of a subsidy into the insurance framework for pandemic-related BI risks is justifiable given how disruptive they are to societies and individuals. Moreover, the business rates tiered structure could be used to ensure that the cost of premiums ceded to the scheme are assessed fairly and at a price artificially lower than the market rate. Indeed, it is a framework that the government used itself when deciding to make grant payments to those businesses in receipt of SBRR in response to COVID-19.

It is therefore evident that our PPP as formulated is not only feasible but, in our view, defensible.

Proposals

It is noteworthy that previously it was argued that subsidised government insurance markets may actively reduce the economic benefits of mitigation and thus reduce the amount of mitigation that is carried out (Jaffee and Russell 2013, p. 490). Such concerns are unlikely to be relevant in the context of a pandemic; as observed with COVID-19, no mitigation measures could have either prevented or minimised the risk. In any event, we do not propose a government subsidy but rather a government incentive.

Furthermore, some authors appear to support the view that a market solution would have been possible without government intervention (Bruggeman et al. 2010, p. 377). However, they do not appear to propose a practical solution to illustrate how that would be possible. We respectfully disagree and we submit that insurance of BI losses suffered because of a pandemic of a similar scale to COVID-19 can be insured and reinsured by utilising a PPP only.

As a consequence, we attempt to provide a solution that addresses the market failure to insure BI losses. We propose the adoption of legislation that establishes an independent, publicly accountable, not-for-profit scheme, PBIRe, similar to those handling the day-to-day operations of Flood Re and Pool Re.

Legislation ought to stipulate that for every commercial property insurance policy purchased, BI cover is mandatory and should include non-damage as well as damage BI cover. Put differently, we do not propose that commercial property insurance be ‘compulsory’ for individuals and businesses; rather, we propose that damage and/or non-damage BI cover (as appropriate to the risk insured) be a mandatory addendum to any commercial property insurance purchased, irrespective of the type of policyholder. It is noteworthy that the words ‘compulsory insurance’ imply a third-party victim whose interest is aimed to be protected through the insurance. In the present context, we do not address a third-party victim as such; however, potentially all members of society will benefit from the survival of a range of businesses post-catastrophe.

Moreover, if cover were optional, it is reasonable to assume that motivating policyholders to purchase BI cover, including non-damage cover, would be challenging to the extent that many would forgo it. Indeed, the COVID-19 pandemic illustrates an overreliance on government aid in the event of such an enormous catastrophe. This would undermine the function of the PPP. Additionally, unless a large enough number of policyholders purchase an addendum of this type, there would be insufficient monies to fund the pools and the sustainability of the PPP would be at risk. For this reason, our proposal includes every business, irrespective of their size and activity (e.g. whether aviation or hospitality, or whether they provide essential goods and services such as food). Some businesses, such as supermarkets, profited during the COVID-19 pandemic, whilst others suffered substantial losses. If the aim is to have sufficient funds in the pool, this can be achieved only if all businesses are included, whether or not they will likely profit or suffer large scale losses during the next pandemic - the word ‘risk’ here implies uncertainty as to how the next pandemic will impact businesses.

Furthermore, to impose a mandatory addendum in a PPP such as we propose is not novel. In Japan, for example, the Japan Earthquake Reinsurance Co. Ltd. (JER) is the sole earthquake reinsurer (Global Facility for Disaster Reduction and Recovery and The World Bank 2012, p. 6). The JER is a catastrophe insurance pooling mechanism: it retains a portion of liability and cedes the rest back to private insurers and the Japanese government through reinsurance treaties in proportion to their respective liability (Global Facility for Disaster Reduction and Recovery and The World Bank 2012, p. 6). The result is that the liability of private insurers and the JER does not exceed the accumulated reserves of earthquake insurance premiums. Earthquakes are a fairly frequent devastating catastrophe in Japan (OECD 2019). Nevertheless, fire insurance indemnifies policyholders neither against damage caused by fire resulting from an earthquake nor fire damage from a fire that spreads due to an earthquake. Earthquake insurance, however, is exclusively available for damage caused by earthquakes or volcanic eruption or from a tsunami following either of these events, including fire, destruction, burying or washing away. Significantly, earthquake insurance is attached to fire insurance: it is essential to hold a fire policy in order to obtain earthquake insurance (Ministry of Finance, Japan, undated). It is significant that this has been well received and remains relatively uncontroversial.

A further example is EQCover in New Zealand. A seismic insurance provided by the Earthquake Commission (Earthquake Commission Act 1993), EQCover is a de facto compulsory addendum to standard fire insurance policies. Thus, policyholders automatically have earthquake cover if they have a household policy that includes fire cover.¹³ In other words, homes without standard fire cover are not covered by EQCover.

How our scheme would be financed is perhaps the most important consideration. Our overriding aim is to make pandemic-related BI losses more affordable for policyholders. We take into account that premiums should reflect risk, otherwise insurers will have no financial incentive to offer coverage (Kunreuther 2015, p. 758). Under ordinary market circumstances, this would result in some policyholders facing much higher premiums than others. It may be necessary, therefore, for the government to legislate a solution, on a temporary basis, by setting minimum and maximum premium levels to be reviewed on a regular basis. The ultimate objective is to accumulate sufficient funds in the pool—it is important to emphasise that our objective is not to fund losses of this nature through taxpayers’ funds. The pool will be funded by the levy paid by the insurers, which will be determined proportionately to the premium charged (see above). Insurers, similar to TRIA, ought to have a statutory retention of the risks they have insured. An insurer’s exposure to the risk may be limited in this way. Insurers, then, may reinsure the amount exceeding the retention with PBIRe on a stop-loss basis. PBIRe could also purchase its own reinsurance, through private reinsurers, if necessary. The result would be to spread thinly the pandemic-related BI losses: the more thinly spread the losses are, the less catastrophic they are financially. TRIA would support our proposal here: in the U.S., to the extent it is available and purchased by insurers, private reinsurance may serve both to increase the availability of terrorism risk insurance and to reduce the TRIA programme exposure. In the early years of TRIA, the Treasury reported that reinsurance capacity for terrorism risk had remained relatively static. Since 2016, however, the Treasury has observed that reinsurance capacity for terrorism risk has increased (Federal Insurance Office, U.S. Department of the Treasury 2022).

Moreover, one of the justifications for the introduction of a government-centred scheme such as TRIA was that the risk of terrorism draws in the entire nation in a way that natural disasters may or may not: after all, natural disasters do not always geographically affect the entire nation equally (Levmore and Logue 2003, p. 279).In this respect, an analogy can be drawn between terrorism and COVID-19; in fact, more strikingly, the latter affected the nation (and indeed the world) more profoundly than any terrorism risk ever has in such a short period of time.

Finally, it is vital to businesses’ liquidity that they receive prompt payment. Our PPP would provide a timely payment of a pre-determined amount without the need to adjudicate individual claims. This would support these businesses financially if there are almost no commercial activities to retain their employees without, for example, recourse to government furlough schemes.¹⁴

Conclusion

Generally, the need for insurance and reinsurance in our daily activities may not be obvious to those who do not deal with such businesses on a day-to-day basis. Whether small or large, losses can only be insured if there is a reinsurance scheme to underpin the original losses by spreading them even more widely. As COVID-19 reemphasised, from travel insurance for individuals to BI insurance for large businesses, insurance and reinsurance arrangements are essential to individual financial resilience as well as the longevity and financial prosperity of businesses, allowing them to continue functioning for the benefit of society.

The foregoing analysis demonstrates that catastrophic events bearing hallmarks of ‘systemic’ risks like floods and terrorism cannot be insured without both the support of reinsurance and government involvement in such arrangements. So far as the U.K. is concerned, Flood Re and Pool Re demonstrate that government financial involvement has been kept to a minimum. However, such arrangements were only possible because of the relevant statutory provisions and the guarantee given by the government to compensate the losses that exceed re/insurers’ retention. Such arrangements are not based on strict risk modelling but on collection of levies from the interested parties and arranging a pool where the fund will be collected and also from which the premium for reinsurance and retrocession covers could be paid. This would inevitably lead to a subsidy of some policyholders by others. However, the unpredictable nature of pandemic risks and the widespread nature of their effects justify such a practice, although it is far from the norm for insurance and reinsurance businesses.

Moreover, as the TRIA example proves, legislators can intervene and require certain types of policies to cover certain types of risks with no separate treatment between the types of risks included in the cover; for instance, commercial property insurance policies covering pandemic risks with no special exclusion for the latter. In other words, the same level of cover for ordinary property risks and pandemic risks is provided. Following such insurance, a body established with either the encouragement or active involvement of the government can manage the reinsurance of pandemic risks. It can be left to insurers to transfer the pandemic risks to reinsurers. We do not propose that the government provides either limited or unlimited reinsurance through taxpayers’ funding. As the TRIA example shows, over a period of time, the uptake of BI insurance may increase and the price of purchasing such insurance may lower as the pool accumulates income ready to be used if a pandemic strikes.

It is preferable to act sooner rather than later to manage the severe pandemic-related BI losses of the next pandemic by implementing a timely solution; indeed, once legislation establishes a scheme like Pandemic BI Re, the next pandemic will have advanced further than the progression of the BI losses scheme in question. The ultimate aim must be to work to pool sufficient funds in time to compensate the enormous losses.