16 July 2019
Fermat Capital’s John Seo, manager of GAM Investments’ ILS strategies, discusses exotic ILS perils, such as terrorism, and why exposure to these risks can be a useful addition to a portfolio, albeit with relatively small weightings at present.
Insurance-linked securities (ILS) are most commonly known for their exposure to natural catastrophe risks such as earthquakes or hurricanes, however there is a small and lesser known category of ILS exposed to what might be termed exotic or man-made risks, for example, pandemic, terrorism or cyber risk. These tend not to be core weightings in a typical ILS portfolio and some investors are understandably uncomfortable with exotic ILS risks. Therefore we thought it pertinent to consider the question: why bother with exotic risks at all?
The answer first requires a review of the basic principles, as we see them, of ILS investing. In our view, a good ILS investment has three general features.
The first feature is that ILS should relate to important risks. In this context, important typically means risks related to a highly concentrated economic activity, but it can also mean risks of national importance or of strategic importance to a corporation. In the field of insurance, highly concentrated economic activity is easiest to recognise by a dense concentration of housing and buildings and ILS typically cover regions with trillions of US dollars in building values located in relatively small areas. However ILS can also cover things like national emergency aid entities or rail networks, even where they pass through sparsely populated areas, if they are strategically important to the issuer. This first feature gives the ILS a chance to offer good “reward”.
The second feature is that the ILS should have risks that are reasonably quantifiable. This should allow an ILS investor to set a reasonable upper bound on the estimate of risk to the security. Put together with the first feature, a good “risk-reward” might be achieved. Whether a risk is “modellable” is often misunderstood. For example, some will say that the risk of natural catastrophes can be modelled, but so-called man-made risks cannot. We do not believe this is true. The misconception likely arises from the notion that complex market risks can be difficult to model. This, we acknowledge, is true, but the catalogue of quantifiable man-made risks is large. The main problem is that they are often not well compensated for in the market. In practice, “modellability” is a function of how much forgiveness is allowed, i.e. what degree of uncertainty is acceptable in the modelling. When a risk is not well compensated, the risk estimate often needs to be more perfect than is practical to achieve. When a risk is extremely well compensated, an intelligent non-specialist is able to set a reasonable upper bound on the risk and find it adequate to drive a rational decision to invest.
The third feature of a good ILS investment is that it should not be a plain vanilla market investment in disguise. Some call this last feature non-correlation, which is a hallmark of alternative investments, but the term as it pertains to ILS is often misunderstood and misapplied. Put simply, ILS non-correlation should mean a crash in the stock markets will not cause an ILS investment to lose principal. It would be incorrect, however, to say that the alternatives investment thesis is possibly violated because, for example, a large enough earthquake might cause a loss to both an ILS investment and a temporary drop in a stock market index. The earthquake risk is already borne by a stock market investor with no compensation. By shifting a unit of exposure from the stock market to an earthquake ILS, an investor is moving from an uncompensated to a compensated component of risk. That has immense benefit to portfolio-level risk-reward, similar to that achieved through classic diversification. By explicitly targeting risks therefore investors stand a chance of being adequately compensated for them in the context of an ILS.
None of the above provides an obvious bar to engaging in ILS focused on man-made risks. It is important to note that man-made risk is a natural part of insurance markets in the sense that insurance claims themselves are subject to significant behavioural elements. Human behaviour may not cause an earthquake to occur, but the scale of individual policyholder claims will affect the final totals paid on earthquake insurance claims. When it comes to ILS risks we believe the uncertainties around policyholder behaviour are relatively well-bounded – certainly relative to some of the significant uncertainties inherent to global macroeconomic risks – but the fact remains an element of an insurance claim’s outcome is behavioural.
The prevalence of natural catastrophe risks in the ILS market is primarily due to the first principle of importance - not the second or third principles (modellability and non-correlation), as many believe. There is a saying in the insurance market that even if only one home is destroyed by a storm, it is a catastrophe for the homeowner, even if it is not a catastrophe for the insurer. This is true. All losses to human property are important to the owners of that property, but in the grand scheme of things, the actions of people, no matter how devastating, are dwarfed by the destructive actions of Mother Nature. An earthquake or hurricane routinely delivers the energy of a hundred nuclear bombs. Therefore, insured losses from acts of nature, not man-made causes, often outstrip the capacity of insurance and reinsurance markets. This tends to bring the more natural catastrophe risk into the ILS market rather than man-made risk. While this is a strong tendency, it is not a rule against man-made ILS risks. In some cases, such risks can create attractive intra-portfolio diversification opportunities for ILS funds.
An example of a recent exotic ILS is the standalone terrorism risk catastrophe bond sponsored by Pool Re, the public-private terrorism risk pool which reinsures commercial property damage and business interruption against terrorism in Great Britain. This bond, called Baltic PCC Ltd, was a GBP 75 million issue in the first quarter of 2019. With the aim of further distancing Her Majesty’s Treasury and the UK taxpayer from any liability in the event of a major claim due to a large terrorism attack (or attacks), this deal was noteworthy for two reasons. First, it was brought to market to ostensibly address an issue of national importance. Second, the risk disclosures were sufficient to enable the replication of the risk modelling of the bond by investors to allow them to perform their own sensitivity and model stress tests. Such exemplary risk disclosures should not be a surprise; some of the world’s leading experts in terrorism risk are based in the UK. Pool Re itself is arguably the most advanced and knowledgeable terrorism risk pool in the world. Two key ILS investments elements came together in this proposition.
The only surprise is that this transaction came to the ILS market at all. As off-putting as terrorism risk may be, the traditional reinsurance market covers this risk well, and Pool Re enjoys significant private market reinsurance support – deservedly so. The motivation of the deal sponsor appears to be a desire to develop a good relationship with the ILS market. The ILS market is not only a different, diversifying pocket of capital from the traditional reinsurance market, but by virtue of being part of the larger capital markets, we believe it has the potential to be the deepest of pocket of them all, making it capable of withstanding a loss better than any other insurance mechanism. Baltic PCC Ltd, though modest in size, incrementally increases the security and stability of the broader UK insurance market. In our assessment, the pricing of the bond reflects its important contribution to the security of the national UK insurance market and therefore to the UK economy.
So where does the market go from here? At ILS conferences we often hear people asking whether we will see more terrorism risk bonds, or even cyber risk bonds. We do not know where the market will go, but we believe that exotic perils will, in general, remain a minor part of the ILS market in the near term – at least in terms of notional risk exposure outstanding. Again, this comes back to the observation that natural catastrophe risk is so much bigger than exotic ILS risks. In the case of terrorism bonds, the question is not whether or not the traditional reinsurance market is getting “full on the risk”. The traditional reinsurance markets seem to have the risk well in hand, both from a risk capacity and a modelling point of view.
However as people and productivity continue to progress, both physically and virtually, new risks will increasingly emerge – perhaps even some with the potential to create insured losses that equal or exceed the current risks to property from natural perils. In the long run therefore, in the decade and more ahead, as demand for risk capacity begins to exceed the available reinsurance supply, what we currently call ILS exotic risks may not be so exotic after all. All the more reason, we believe, for us to pay attention to these investment opportunities as they emerge today.
Source: GAM unless otherwise stated. The information in this document is given for information purposes only and does not qualify as investment advice. Opinions and assessments contained in this document may change and reflect the point of view of GAM in the current economic environment. No liability shall be accepted for the accuracy and completeness of the information. The mentioned financial instruments are provided for illustrative purposes only and shall not be considered as a direct offering, investment recommendation or investment advice. Reference to a security is not a recommendation to buy or sell that security. Past performance is not an indicator of future performance and current or future trends.