What are insurance-linked securities and how do they work?
Events in the natural, rather than the corporate world, drive the performance of insurance-linked securities, known as ILS. This means performance is not correlated with traditional asset classes, whose returns are more closely linked to factors such as economic strength or weakness, a company’s good or bad performance, or geopolitical concerns.
This diversification benefit is a key reason investors are increasingly interested in the area. But for those not familiar with them, what are ILS and how can they fit into a portfolio?
The market for ILS
The best-known part of the ILS market is probably catastrophe bonds or “cat” bonds. These instruments are more conventionally tradeable and normally have a life span of between three and five years. This is the area for which most data is available. However, of the overall ILS market, which is estimated at around $103 billion, only around one third is comprised by cat bonds.
The other two-thirds consist of non-tradable, “over-the-counter” contracts, mostly with a 12-month lifespan. This market gives investors access to a wider range of insurance perils than those available in the cat bond market, including marine, aviation and specialty risk and a broader range of investment structures.
These instruments also require managers to model the risks themselves rather than using a third party, and there is no secondary market. Their non-tradeable nature means there is an additional “illiquidity premium” for managing this segment of the ILS market.
How does ILS differ from corporate or sovereign bonds?
Essentially, ILS is a way for companies to buy protection against the risk of incurring a loss as a result of an event. These companies are therefore often referred to as “protection buyers”. The protection buyers of ILS are generally insurance or reinsurance companies (and also corporations or public entities) looking to reduce or remove the risk of paying out on an insured event. An investor in ILS will receive interest payments, paid out of the insurance risk premium plus a money market return. As such the return is mainly determined by the insurance risk assumed.
Unlike corporate or sovereign bonds, cat bonds and other insurance-linked instruments are not directly exposed to the credit risk of the issuer. This is because of the specific structure of ILS. The insurance risk (i.e. the risk of paying a claim) is made available to investors via a special purpose vehicle (SPV) known as a “transformer”. The buyer of the protection (which could be an insurer or reinsurer) passes a risk, or a proportion of it, to the protection seller (such as an investment fund). The risk moves from the seller to the buyer through a reinsurance contract in the form of a bond (an ILS), issued by the SPV.
This reinsurance contract is backed by collateral which is paid to the SPV by investors when the transaction starts. The SPV then issues securities – such as cat bonds or preferred shares in the case of a private transaction – against this collateral. The result is an insurance risk that has been transformed into an investible instrument. The financial health of the protection buyer is therefore not a concern in assessing the riskiness of the bond. To help visualise the structure, it is outlined below.
Structure of an insurance-linked instrument
Source: Schroders, for illustration only
The structure means that the cash paid for an ILS is not directly exposed to the credit risk of the issuer, as it is held separately in a trust account and invested in money market funds or instruments. As a result, the insurance-linked instrument is not exposed to the issuer’s ability to pay claims. The likelihood that the pre-defined risks occur, which could impact the investment, is the metric used instead to assess the riskiness of the instrument. For this reason, most of the universe is not rated by an agency like corporate or government bonds.
In the unlikely event that the issuer of an ILS defaults (for example, by not paying the agreed risk premium), the notional capital remains unaffected. The insurance-linked instrument would cease to exist and the collateral in the separate trust account would then be paid back to the investors. What could result in a loss of notional capital, of course is the “insured event”. If the tropical cyclone, flooding or earthquake the protection buyer has sought cover for were to happen, this would cause losses to the instrument. This is the insurance risk embedded in the instrument, rather than credit risk. It is also a significant reason for the asset’s lack of correlation to traditional markets.
ILS performance versus traditional assets
Source: Schroders, Bloomberg
Past performance is not a reliable indicator of future returns, prices of shares or bonds, and the income from them, may fall as well as rise and investors may not get the amount originally invested.
Duration (interest rate) risk
Duration (the sensitivity to changes in interest rates) for ILS instruments is also generally negligible. ILS are notionally floating rate instruments, in so far as part of the coupon paid is based on money-market return. However, the coupon is reset quarterly, meaning shifts in monetary policy make little difference to the instrument’s value. Furthermore, the bulk of the coupon is based upon the risk premium. This is a combination of the modelled likelihood of the insured event occurring and – particularly in the case of privately negotiated ILS instruments – an illiquidity premium.
Where are the opportunities in ILS just now?
A fairly valued investment opportunity for ILS is one that compensates investors sufficiently for taking the insurance risk embedded in it (both on a stand-alone basis and within a portfolio).
This means understanding and rigorously testing modelled losses relative to potential return. Third party vendors will model losses relating to specific events, and these tools are available to buy “off the shelf”. Schroders will overlay our research on top of these models to determine where we believe they might be wrong or if risk is insufficiently priced in. If this is the case, we would adjust the model to reflect our view of the underlying risk.
As an example, wildfires were a dominant aspect of insurance event activity in 2017 and 2018, and our own views – which differed from third party models – prompted us to reduce exposure to this area in 2018 and not buy a bond that covered this peril as a stand-alone risk.
2017 and 2018 were respectively the largest and fourth-largest insured loss years ever recorded (at $140 billion and $80 billion globally). The combined impact of natural catastrophe events over these years has resulted in more compelling pricing in 2019. With premiums adjusting to offer better compensation, especially on loss-affected deals, we are currently seeing the highest spread levels in six years.