Collateralized Reinsurance

In order to understand collateralized reinsurance a brief description of traditional reinsurance model is helpful. Traditional reinsurance is essentially the insurance of insurance companies. An insurance company prices its business based on the law of large numbers. If it writes a broad enough portfolio the losses sustained by the portfolio will tend to the mean losses sustained by the population. Risk is therefore spread over a large number of homogenous risks, each paying its share of the premium. This however fails where a large single correlating event broadly hits the risks within the portfolio at one instant e.g. an earthquake or a hurricane. It is for these broad impacting events where the diversity of the portfolio breaks down that insurers purchase reinsurance.

A reinsurer takes on reinsurance risks from around the world providing protection for many different perils. A reinsurer therefore spreads its risk across geographic scope and varying perils as well as over time; a reinsurer will, to the extent allowed by their accounting regime, build up reserves in good years to offset losses in bad years.

ILS funds have capital to invest in much the same way as a reinsurer, namely to build a diversified portfolio of large in-frequent events in return for earning a premium i.e. an income on the capital allocated to support the portfolio. Whereas this can be done through investing in cat bonds, the diversity of peril-regions that can be assumed can be increased by also investing in collateralized reinsurance. Most of the cat bond paper outstanding today comprises US perils (hurricane and earthquake) and therefore only limited diversification across peril-regions. To access the broader exposures available in the reinsurance market it is possible to offer insurers a reinsurance contract whose limit is fully backed by collateral, as opposed to a reinsurer who offers contracts based on a claims paying rating from one of the rating agencies. This is known as ‘Collateralized Reinsurance’. To complete such a transaction a regulated reinsurance entity is required to issue the reinsurance policy to the reinsured and an investment product allowing (i) the investor to pay in the collateral and, (ii) to return the collateral and premium back out to the investor at the end of the reinsurace term.

Collateralized reinsurance can be provided either as a traditional reinsurance which protects the ‘Ultimate Net Loss’ sustained by a reinsured. This means that they will reimburse the loss amount, subject to the terms and conditions of the contract, to the reinsured. In order to understand and price such a contract the portfolio of the reinsured and the risks therein needs to be understood in depth. One must understand the monetary impact on that portfolio of each possible event. Another way of providing protection, which has proved popular in the collateralized reinsurance market, is through an ‘ILW’ – an Industry Loss Warranty. In its simplest form, an ILW contract will pay a loss based on a total market loss to a given event e.g. a US hurricane generating insured losses of greater than USD 20 billion. There are currently two independent firms which publish market loss size to large events: PCS in the US and Perils in Europe. The advantages of ILW contracts over UNL contracts are that for an ILW the actual portfolio of the reinsured is disregarded thereby simplifying the analysis and providing little scope for dispute given the independence of the agency providing the loss. Another advantage is that they are settled quickly – the market loss is an approximate estimation whereas a UNL contract requires all losses to be settled before the reinsurance loss is known and can be paid which can result in several years delay before the final result is known. From the reinsured’s point of view however, a UNL contract is a ‘perfect hedge’ whereas the ILW can contain significant basis risk. As such the UNL product is worth correspondingly more and therefore commands higher premium for a unit of risk compared to the ILW.