- In the recent past, higher risks have not led to higher returns and cat bonds outperformed private transactions
- Successful implementation depends on the ability of the selected asset managers to invest in ‘good’ risks that are adequately compensated
Hurricane Ida in late August and early September caused great damage to the southern coast of the US. Fortunately, for people in this area, insurance policies often cover destructions to their properties. Since covering such damage can lead to severe losses for insurance companies, they are keen to reinsure themselves.
This is the traditional business of reinsurance companies. For some time, a portion of these policies is pooled and transferred to institutional investors in the form of insurance-linked securities (ILS). ILS offer interests (insurance premiums) for bearing risks for losses that occur in the case of an insured event.
An increasing number of catastrophic events has recently led to intense examination of the rationale for ILS as a unique asset class, as well as performance dispersion among asset managers. This article addresses strategic considerations, how implementations differ in terms of risk-return characteristics, which managers were successful risk allocators, and what this implies for investing in ILS.
A unique selling point of ILS is that the asset class represents a truly independent alternative risk exposure, as there is hardly any relation-ship between natural disasters and financial crises. Insurance-related assets therefore tend to have low correlations with ‘traditional’ financial market risks and returns (for example, interest rates, equities or credit). Investors in ILS provide protection against an insured event to the insurance buyer and continuously collect premiums, but face significant downside risks (events with low probability and high severity, also known as tail risk) in case of occurrence.
Owing to their short duration and mostly variable interest rates, ILS are not suitable for liability-hedging purposes and cannot be viewed as a substitute for ‘traditional’ fixed income. In addition, investors must be able to cope with the low liquidity and limited transparency. There are also potential conflicts of interest between the protection taker and provider. For example, there is the inherent risk that insurers pass on ‘bad’ risks to investors (adverse selection) and that insurers no longer have any interest in minimising losses after the risk is transferred (moral hazard).
In theory, ILS could improve portfolio diversification. However, implementation is essential. Owing to the challenges mentioned above, specialised asset managers are usually entrusted with the management of ILS.
Small market, big differences in implementation
The following overview shows the implementation of ILS strategies in practice. The market overview is based on data collected from 43 different ILS products of 22 leading asset managers that manage ILS amounting to $43bn (€37bn). This represents almost half of the entire ILS market that is estimated at about $100bn, of which investments of Swiss pension funds alone account for nearly 10% ($9bn).
In terms of risks, US wind (38%) and US earthquakes (14%) dominate the portfolios, while European wind, flood and Japanese wind each account for roughly 5%. Other risks include US wildfire, mortality or longevity, or specialty risks such as terrorism or marine risks that all contribute less than 5% to the total risk.
The products in the market overview on average have an allocation of 45% to catastrophe (cat) bonds (securitised, tradable instruments) and 37% to illiquid private transactions (bilateral contracts between reinsurers and investors). The remainder is split between cash and other instruments. However, the allocations of the individual products in terms of instruments vary greatly, from products that invest exclusively in cat bonds, to products that contain only private transactions.
Risk and return expectations of asset managers are in line with theory – higher expected losses are related to higher return targets (see figure 1). The average no-loss return among the products is 7.8% a year (in dollars) and the average expected loss is 3.3% a year, resulting in an average expected after-loss return of 4.5% a year. The data also shows that a higher allocation to cat bonds tends to be associated with lower expected losses.
Pricing insurance risks – a retrospective view
Contrary to the expectations of asset managers and economic theory, over the past five years the relationship between risk and return was inverse – products with higher expected losses had lower realised returns (see figure 2).
At first glance, this suggests that risks are not accurately priced. However, this conclusion would be premature because of the long-term nature of ILS. Higher premiums accumulated over years without major events should be sufficient to cover large losses. In the short term, average returns of insurance-linked products can vary and events like hurricane Irma can lead to more market-wide negative returns in individual years (see figure 3). Over the last 10 years, realised performance (4.5% a year), on average, has been in line with managers’ expectations.
Historically successful risk allocators
Although Hurricane Irma led to low average returns in 2017, there is also a significant difference between the best and worst-performing managers in that year. Dispersions in returns among asset managers can be explained by different allocations to risk perils and instruments.
Products with a relatively high allocation to US wind risks experienced higher losses in 2017 owing to Hurricane Irma and Harvey. In 2020, primarily US winds and other insurance risks (especially Covid-19) had a negative impact on returns.
Another factor that explains part of the difference in returns is the allocation across instruments – that is, cat bonds versus private transactions. Products with a focus on cat bonds had higher returns (+4.3% a year), on average, over the last five years compared with products with a focus on private transactions (+1.1% a year).
One rationale for this performance difference is that cat bonds typically insure large catastrophes with severe loss potential (for example, hurricanes, earthquakes), while private transactions also capture small to medium catastrophes (for example, floods, wildfire) or catastrophes that arise as secondary consequences of large events (for example, hurricane-related rainfall, fire after earthquake). In the last five years, damages from small to medium events, respectively secondary events of large catastrophes (secondary perils) were more frequent compared to large-scale catastrophes (primary perils). Asset managers with higher allocations towards private transactions therefore, on average, had higher losses compared to cat bonds focused investors.
Implications for the future
In efficient markets, if risks of losses from catastrophic events increase, the price of insurance against these risks, respectively the premiums, should increase. This is exactly what is currently happening in the market. Spreads of cat bonds, which in the recent past have offered adequate compensation for the risks assumed, have fallen. On the other hand, premiums paid for private transactions have increased. In the end, it is crucial for investors to mandate asset managers who can assess which market segments currently offer the best compensation for the risk budget (for example, select the ‘good’ risks).
When investing in ILS there are a number of considerations to bear in mind. ILS could improve portfolio diversification but are not suitable as a ‘bond substitute’ and substantial losses can occur. Investment products are complex and there is great heterogeneity between asset managers. Diversification among risk perils and instruments is essential. When investing and selecting suitable asset managers, it is important to know the portfolio characteristics. When invested, concentration risks, performance and fees of the products must be continuously reviewed for market conformity to ensure that positive aspects of the asset class outweigh the costs.
Romano Gruber is senior investment consultant and Marc Staub is senior investment consultant at PPCmetrics