Catastrophe insurance has typically been covered through insurance and reinsurance markets. Due to the capital structures of reinsurance companies, they have struggled to provide sufficient coverage of tail risks at prices which are fair for the insurers and subsequently for households. This leaves many companies and households retaining risks beyond what is preferable. The impacts of climate change on the propensity and severity of extreme weather events threatens to endanger more even more people. Additionally, economic development also acts as another driver for increased risk exposure.
Therefore, additional methods of financing are sought after to provide catastrophe risk coverage at fair prices, determined with respect to the scientific modelling. Insurance Linked Securities (ILS) and collateralized reinsurance have been designed to provide a solution to these issues. CAT bonds are a popular example of an ILS and have increased in popularity over the past 20 years. Nevertheless, CAT bonds still hold other problems and new financial products are being considered for catastrophe risk insurance.
2. Traditional reinsurance approach to catastrophic risk and its inefficiencies
The need for reinsurance
Tail risk poses a significant threat to society. The increase in the severity and intensity of extreme weather events, coupled with economic development leaves an increasing amount of capital exposed to these risks. The role of reinsurers looks to disperse these risks taken on by insurers.
Traditionally, insurers and reinsurers raise capital through debt and equity markets. Diversification of catastrophe risk occurs through the sale of premiums across unrelated geographic locations. Therefore, if an event were to occur, payouts would be guaranteed to not affect the whole portfolio.
Despite insurance companies being able to spread risk among their portfolios by diversifying, the magnitude of a single catastrophe can be too high to bear by a single insurance company. The estimated potential insured loss for a single peril in the US can be up to $100 billion; whereas the premium of all US’s insurers is about $239 billion (Froot, 1999). Therefore, insurance companies are eager to distribute risks with other parties by purchasing reinsurance.
The benefits of such approach to insurers is obvious. First, the reinsurance contracts can transfer the individual insurer’s accumulated commitment to a larger number of parties. Second, as the reinsurance market expands insurance companies’ capability to cover greater number of losses without increases in administrative costs that could potentially affect their solvency abilities. Third, reinsurance market can provide liquidity and capital support to insurers under the occurrence of contingent cat events.
The theoretical model for reinsurance
From the the theoretical model for reinsurance, two major assumptions can be derived.
The first assumption presumes that households and (insurance) companies are risk averse, and therefore regard severe losses as more costly. Hence, assuming that the household or company is budget constrained, they have the incentive to reinsure low-probability but high-severity outcomes, as hedging against such catastrophic outcomes has higher benefits than hedging against high-probability outcomes with low impacts. In other words, the need to hedge is the greatest for the most severe risks.
The second assumption presumes that catastrophic event losses are uncorrelated with financial wealth, and therefore the price for reinsurance against catastrophic events should equal expected losses. This assumption is derived from the Capital Asset Pricing Model (CAPM). The probability and expected losses of catastrophic events can be measured using scientific models. According to the CAPM, expected returns for assets equal the risk free rate plus the return of the market in excess of the risk-free rate, which depends on beta. Beta is a measure of systematic risk. Since the risks of catastrophic events are independent from systemic risks in the financial markets, the beta of reinsurance assets is zero, and thus the expected return should be equal to the risk-free rate. In other words, the expected return is equal to the rate of expected loss. Therefore, the fair reinsurance price should equal expected losses (Froot, 2001).
Reinsurance in reality
When these two particular assumptions, derived from the theory, are tested against the observed markets, the results hint towards a poorly functioning market for catastrophe risk. Derived from the assumptions discussed above, the following two properties should be observed in the markets for catastrophic risks :
- The optimal reinsurance profile should protect against the largest losses first.
- Fair premiums for cat risk should approximate expected losses
However, data from the US market shows that, contrary to the first assumption, insurers cede the most risk to reinsurers for small to medium sized losses. The largest losses, those that could be associated with catastrophic risk, are the least ceded by insurers (Figure 1). Secondly, analysis of the same data shows that the second assumption is violated. Industry wide prices for reinsurance of catastrophic events is greater than the assumed fair value of the expected loss and the price excess is greatest for the lowest probability risks (Figure 2) (Froot, 2001).
Therefore, in the following paragraphs, the practical inefficiency of reinsurance for cat market from both demand and supply side will be discussed.
Potential supply and demand issues
Over the past three decades, the greatest contribution to investigating the efficiency of the catastrophe risk insurance market has been from three papers; Froot (2001), Froot & Cornell (2008) and Froot (2008). The following is a summary of these findings from these papers which cover evidence for inefficiencies in the market as well as some supply and demand issues that could potentially explain them.
Both supply and demand issues could potentially be contributing to the failure of catastrophic risks to be purchased by reinsurers and for the prices to consistently exceed the theoretical fair value. First, we will deal with potential supply issues that could explain these market failures which can be summarised as:
- Capital shortages for large cat events
- Inefficiency of the corporate form of ownership in reinsurance
- Rating agencies encourage too much capital to be allocated to diversifying risks
Evidence exists of insufficient capital in the reinsurance market. Following a serious season of catastrophic storm events that included hurricanes Katrina, Rita and Wilma in 2005, the markup for cat risks increased three to four times. It is possible that this may be explained by forces other than capital shortage, such as new information for hurricane risk in the US. However, prices also doubled for earthquake insurance in the same period, suggesting that capital depletion was at least in part causing the price fluctuations. The reasons behind the lack of capital could be due to the high costs of raising capital from the capital markets, the difficulty of finding investors willing to accept relatively low returns for holding catastrophe risk or simply the costs of hoarding capital for a reinsurer.
The corporate form of ownership may also introduce inefficiencies into the market. Corporate insurance owners are encouraged to hoard capital by customers seeking certainty of payment and capital markets that are biased towards companies with high liquidity and low levels of debt. The hoarding of capital causes agency problems and increases the risk of mismanagement by executives.
Finally, rating agencies encourage insurers to focus overly on risk than on returns. As a result, intermediaries over supply capital to cover diversifying risks and create a capital shortage for the largest catastrophe risks. For example, too much capital is allocated by US reinsurers to cover wind and earthquake exposures in Australia, Europe and Japan which are used to diversify their exposure to hurricane and earthquake exposure in the US.
On the demand side, a couple of issues could be contributing to the market inefficiency. These include:
- Ex-post financing by third parties
- Managers’ actions distorted by agency issues
- People struggle to internalise the risk from very low probability events
The demand for catastrophic risk insurance is particularly susceptible to distortions created by third party substitutes for insurance. The most common form of substitution (also known as ex-post financing) occurs when governments step in to provide emergency finance to those affected by a large event. For example, the US government has given more finance since 1970 than the average annual loss borne by reinsurers from catastrophe coverage. While ex-post financing has a large impact on insurers’ demand for the highest reinsurance layers, it can not account for capital shortages and so can not entirely explain the low levels of risk transfer.
As well as affecting the supply side, agency costs can also negatively impact the demand side of a market. In the insurance industry, managers affected by agency issues tend to transfer value from policyholders to shareholders. In the context of reinsurance of catastrophe risks, agency issues can reduce demand for the highest layers of reinsurance.
Finally, an inherent failure in the human condition to properly assess, internalise and act on the lowest probability risks, even if these risks pose the highest losses, may be another factor that explains low demand for catastrophe insurance. Equally, because of the ambiguous nature of catastrophe risk, insurers and reinsurers may set premiums irrationally high.
3. Capital market solutions
As the previous section discussed, there are some major problems with using traditional reinsurance to hedge against climatic risks. To tackle these issues, alternative mechanisms have been created to transfer risk to other parties and capital markets, known collectively as Alternative Risk Transfer (Artemis, 2017a). Insurance-linked securities, cat bonds and collateralized reinsurance provide new methods to insure against cat events, linking insurance products to capital markets.
Insurance-linked securities emerged as a solution for the problems of traditional reinsurance in the late ‘90s, after hurricane Andrew struck the United States in 1992. Andrew is regarded as one of the most devastating hurricanes in the history of the US, the total damage was approximately $26.5 billion in 2016 and led to the bankruptcy of 11 insurance firms (Lewis, 2007). A single cat event such as Hurricane Andrew can create losses that are too large to be diversified among reinsurers. Thus, the insurance industry created alternative insurance-linked securities to hedge the risk of such cat events through collaboration with capital markets. Destructive cat events happening in mid 1990s, such as Hurricane Andrew, which highlighted serious issues with the market were regarded as the main drivers of such action. (Luciano, text is great, can you include a reference or two)
Catastrophe (cat) bonds help to reduce the contingent impacts of future natural disasters on reinsurers. Since their introduction in the late 1990s, the volume of risk capital issued through ILS and cat bonds has increased by a factor of ten (Figure 3), resulting in $26 billion outstanding in 2017 (reference).
Today cat bonds are a commonly used insurance-linked securities, deployed by insurers to hedge themselves against low-probability climatic risks such as earthquakes, windstorms, hurricanes, typhoons, and temperature risks.
Structure & Functioning
A counterparty or sponsor, looking to hedge a certain catastrophic event, signs a contract with a special purpose vehicle (SPV), who acts as the insurer. The SPV agrees to cover a part of the risk in return for premiums received from the sponsors. To cover the risk, the SPV issues bonds to investors, and receives a principal in return. The principal serves as collateral for the underlying limit that the SPV agreed to cover, and is invested in safe securities such as US treasuries. In this way, credit risk can be virtually eliminated for investors (Artemis, 2017b; Edesses, 2014). A diagram illustrating the structure of cat bonds can be seen in Figure 5 in the appendix.
If a catastrophic events occurs before the bond reaches its maturity, investors may not recover the full principal at the end of maturity as the principal will be used to cover insured claims. Different categories of bond notes may be issued in tranches with different risk profiles, eg. A1 bondholders do not have the full principal at risk, while A2 bondholder bear the risk to lose the full principal. In this way, the risk investors bear depends on the category of their notes
Advantages compared to traditional reinsurance
There are several reasons why sponsors and investors prefer cat bonds over traditional reinsurance to hedge against cat events.
From the sponsor’s perspective, a major advantage is that credit risk is virtually eliminated, meaning that the counterparty will pay out in the event that a catastrophe occurs. Moreover, where traditional reinsurance contracts had to be renewed every year, cat bonds usually have a maturity of 3-5 years, providing multi-year pricing stability for the sponsor. Cat bonds form a cheaper alternative to traditional reinsurance contracts, and more contracts are available as the problem of capital shortages in the traditional reinsurance market is eliminated. Cat bonds enable access to diverse sources of funding from the capital markets.
Various categories of investors (hedge funds, pension funds, investment managers etc.) participate in the cat bond market for two major reasons. First, cat bonds generate high risk adjusted returns, with low volatility compared to other asset classes. Investors receive the base interest on the treasury fund in which the principal is deposited, plus premiums paid by the sponsor for the insurance coverage. Since 2002, annual rates of return ranged from 7 to 9%. Second, returns are uncorrelated with stock markets as the risk of cat bonds (i.e. the occurrence of a natural catastrophes) is uncorrelated with the risk of equity market fluctuation, credit risk, and interest rate risk. This allows for a high degree of portfolio diversification (Artemis, 2017b; Edesses, 2014).
Even though the cat bond market has expanded quickly during the last decade, questions can be asked about whether this growth is viable in the long term. Until 2013, only three out of 200 bonds issued were triggered and resulted in a loss to investors. However, if the incidence of climatic risks were to increase dramatically in a given year, losses could be potentially increase quickly and cat bonds could suddenly become a less attractive asset to investors (The Economist, 2013).
According to the chairman of Lloyd’s of London, there is the danger that the recent influx of capital into the insurance industry could lead to systemic problems caused by a lack of understanding from investors of the inherent risks of cat bonds. Cat bonds are esoteric, and in the case that the industry is not pricing risks correctly, inexperienced investors may encounter big losses (Financial Times, 2017).
Another reason for concern is that the bank crisis of 2008 showed that cat bonds are not completely uncorrelated with market risk. Four catastrophe bonds were downgraded, and investors experienced losses. Their principal was deposited into CDOs with an AAA rating, so they were supposed to be safe securities. However the bank crisis showed that rating agencies failed to properly assess risk and the price of CDO’s crashed. As a result, investors experienced losses, not due to catastrophic events but due to credit risk, despite the cat bonds supposedly being free of credit risk. However, following the crisis, cat bonds are now invested primarily in treasuries so it can be assumed that these cat bonds truly do not bear any credit risk (Towers Perrin, 2009).
Industry Loss Warranties (ILW)
An industry loss warranty has similar characteristics to an insurance contract. The warranty is triggered when the losses experienced by the entire industry exceeds a certain amount, previously agreed within the contract (Benfield, 2015). In this case, the reinsurer may not bear as much risk from individual entities as it is the cumulative loss that will determine whether the ceding insurer is returned capital. The introduction of these, and other conditions, allows the price of the security to be reduced in comparison to traditional reinsurance contracts (Artemis, 2017).
Other tools exist in the alternative risk transfer market such as catastrophe swaps and derivatives. These again seek to diversify the risk geographically, and reduce exposure to particular catastrophic events.
Collateralised reinsurance does not differ greatly from the traditional reinsurance market, with the collateralized reinsurer underwriting the insurer. The collateralised reinsurer then transfers this risk to one or more risk bearing entities, usually special purpose insurance companies. As with cat bonds, this function distributes the risk away from the reinsurance company, transferring it into the capital markets. The risk bearing entities are then able to receive a premium for bearing the risk.
Collateralised reinsurance is not an insurance linked security, and is therefore more flexible, not being bound by the same regulatory limitations that may be found with instruments such as the cat bond (Benfield, 2015). These smaller transactions, usually in the range of $1m to $100m, are not rated or traded and so do not suffer from the same risks around mispricing as catastrophe bonds. Whilst this avoids agency fees related to rating, the inability for the product to be traded negatively impacts the interest of investors (Solidum Partners, 2017).
4. Alternative risk transfer in the context of climate change
4.1 Extreme Climate Facility
Climate change is one of the main risks for the future of African agriculture. It will affect the economy of hundreds of million of Africans. As a result the African Risk Capacity (ARC) was founded to assist the Member States in improving their capabilities to reduce the impact of natural disaster by using modern finance mechanisms such as risk pooling and risk transfer (African Risk Capacity, 2017).
The ARC has created the Extreme Climate Facility (XCF), funded by cat bonds, using climate indices as a parameter to determine when the principal of the cat bond will be triggered. The Extreme Climate Index (ECI) measures any increase in magnitude or frequency of extreme climate events. If the ECI reaches a certain value compared with the baseline, the cat bond is triggered. Then, the affected country will receive the funds from the XCF so it can activate its climate adaptation plan. These payments could increase if the ECI deviates even more from the baseline climatology. This will secure direct access to capital for African governments to reduce the impacts of climate change. XCF is expected to rise more than $1 billion over the next 30 years from public and private funds. It also represents an excellent opportunity for investors as the bond’s coupon will be financed by recognised donors such as the Rockefeller foundation.
4.2 Carbon Cat bond
According to the IEA, one of the main uncertainties for climate change investors is whether governments are engaged to support the decarbonisation of the economy. Cleantech projects provide better returns when government policies lead to higher carbon prices. Therefore, uncertainty about future government policies is one of the main risks that investors want to avoid (International Energy Agency, 2007).
Carbon cat bonds, described for first time by Professor Michael Mainelli, are a potential solution that can help governments to make cleantech projects more attractive to investors. This sort of index-linked bond can provide natural insurance against climate policy failures. If investors were able to invest in a cleantech project and simultaneously buy a proportion of carbon cat bonds, they would reduce their exposure to the risk of climate policy failure (Knight & Howard, 2011).
If governments strive to move towards a low carbon future, cleantech projects will have greater benefits. If governments are not committed, then they will be forced to pay a higher coupon. This eliminates the climate policy risk of cleantech projects, the uncertainty of government policy being directed at a low carbon future, attracting more investors and making capital cheaper.