Newsletter Sign-up

By STEVE E. SMITH, D. Phil. (Oxon.), FRMetS, president, property solutions, ReAdvisory/Carvill America

The transfer of risk at a fair price is the fundamental business of the insurance and reinsurance market. Central to this is the choice of a common method for quantifying risk because, unless we as a market agree on how to measure risk, we have no basis for pricing and transferring it. But here is precisely where innovation is most needed, for our traditional measuring method is preventing us from spreading risk as broadly as we now need to.

The Chicago Mercantile Exchange has launched a series of futures and options based on the CHI, which allow users to trade risk for various sections of the U.S. coast (Florida, for example), as well as the offshore Gulf of Mexico oilfield region.

2008 CME TRADING VOLUME

Trading volume of CHI contracts on CME has been impressive for a new product. In the first seven months of 2008, more than 31,000 contracts were traded with a notional amount of risk close to $60 million changing hands.

As Hurricane Gustav approached the United States, $9 million of "live CAT" capacity was made available through the CHI on CME. So far, trades typically have been between reinsurers and uncorrelated providers of capital such as hedge funds and investment banks. We might not know for a while exactly how many contracts were triggered, though there were no named storm contracts that were triggered.

However, we also have seen interest in the CHI products from sources as varied as homeowners in Florida, farm owners along the Gulf of Mexico and energy companies in the Carolinas.

There are multiple reasons for its success. Given the real-time aspect of the CHI, CME is able to settle trades (i.e., make payments) within three business days of a hurricane landfall, a clear cash-flow benefit. Further, given that trading occurs through a margined exchange mechanism, the credit quality of the product is faultless.

It is also a measure that means the same thing to every participant (i.e., it is fungible). It is completely transparent, relying on a very simple, scientifically based, and publicly available formula to calculate index values using data from an impeccable source, the National Hurricane Center. And it provides a fine-tuned, continuous-scaled estimate of risk in real-time at any point in a hurricane's life cycle.

The CHI is a measure of the potential for damage from an Atlantic hurricane. It is a single measure based on both the maximum sustained wind speed of the hurricane and the extent to which hurricane force winds extend from the center of the hurricane. CHI values thus quantify actual risk, they are easy to understand, and they are available each and every time the NHC issues a Public Advisory--at least every six hours for a live storm in the Atlantic.

A final settlement value for futures and options based on the CHI can be produced within hours of a hurricane making landfall. For example, when Hurricane Gustav hit the US Gulf Coast, the final settlement value for this storm (CHI of 7.2) was calculated and sent to the CME within an hour of landfall; any contracts triggered by this event were settled up within 3 days of the landfall. What we have created is a transparent, real-time, unimpeachable index that enables the development of tradable contracts with clear usefulness to the insurance industry and, at the same time, broader appeal to the outside financial community than any previous ILS.

WHY CHI AND NOT ILS OR ILW?

Traditionally, exposure has been the lingua franca of the insurance market. Companies have calculated their maximum potential losses and aimed to lay off pieces of this exposure on reinsurers. The measurement is simple in concept--the monetary value of individual insurance policies can easily be rolled up to the combined exposure of a particular line of insurance or an entire insurance company.

However, it does impose a constraint on the market, because each company's exposure is made up of a unique constellation of policies, and only specialists can assess what risk it carries. To assess the risk on a portfolio, often the only recourse is to complicated and expensive catastrophe models.

Exposure as a measure of risk is not, in other words, fungible, and our reliance on it has limited the extent to which risk from the insurance sector can be distributed to the wider financial market. There is a broad understanding that one share of IBM stock has the same risk as another other share of IBM stock, but $100 million of property exposure from insurer A will have different risk characteristics from $100 million of property exposure from insurer B.

This lack of fungible risk has lead to the insurance/reinsurance industry being the only sophisticated financial industry without a deep secondary market. This situation is unfortunate because secondary markets have proven their worth in many other financial sectors where historically they have contributed to improved price transparency and stability.

Of perhaps greater interest to the insurance industry, a deep secondary market has the potential to provide a wide variety of un-correlated counterparties, improving the creditworthiness of the entire insurance sector.

The last 10 to 15 years have seen several attempts to expand the secondary insurance market by developing products more familiar looking to outside investors. However, these insurance-linked securities have had only limited success. ILS products are now estimated to make up 10 percent to 15 percent of the total property catastrophe market, but they appeal to only a small subset of investors--those who take the time and effort to understand insurance risk.

The crucial and inhibiting problem, once again, has been the choice of a method of risk measurement. ILS products such as CAT bonds, which are typically based on an insurer's actual losses, suffer the same problem that plagues the traditional insurance market: they transfer disparate pieces of exposure and thus lack fungibility.

In contrast, industry loss warranties, with payouts based on estimates of industry-wide losses, would seem to avoid this problem. But these products rely for common measurement on the Property Claims Service estimate of industry loss, a process which is not transparent or, for that matter, entirely credible, because it is based on the industry's self-reporting of a limited number of lines of business affected by a catastrophic event.

In addition, the final PCS estimate of a loss may take upwards of 24 months to determine, tying up investors' capital for longer than those outside the insurance industry find acceptable.

For all these reasons, the current ILS products have not reached into the broader financial community; instead they have essentially created a new class of pseudo-reinsurers.

For a time, attempts to create an ILS product based on the Saffir-Simpson Hurricane Scale looked promising. The SSHS would offer a simple measure of risk, but because the SSHS is not a continuous scale (instead it has just five discrete categories), and because it takes into account only the strength and not the size of a hurricane, it proved too crude and inaccurate a measure to be of much use to insurers.

But now, engaging the broader financial community to take insurance risk should have significant benefits for the insurance market. It is likely to make pricing more transparent and stable and bring in uncorrelated investors who will improve the market's ability to weather the shocks produced by CAT events.

September 1, 2008

Copyright 2008© LRP Publications