Cat bond fever has continued unabated in 2007. Willis Capital Markets looks at how their pricing has evolved and asks if they can weather the subprime storm currently brewing.

Recent years have seen a continuous and steady growth in the catastrophe bond market. From its earliest days, an annual compound growth rate in excess of 33% in the volume of securities outstanding has been the norm (see figure 1).

As the catastrophe bond market expands and liquidity increases, the long-term drivers of the pricing of cat bonds becomes easier to understand. This is particularly relevant in the current market environment, where the subprime mortgage "implosion" and its impact on the stock markets may limit the willingness of capital market investors to offer reinsurance capacity. There are several factors likely to affect the price of capacity offered by the cat bond market and these include:

• Credit spreads in the debt capital markets;

• Traditional reinsurance pricing and capacity availability; and

• Ratings migration.

The ten-year history of the cat bond market encompasses a period of dramatic extremes within the reinsurance industry and provides us with an ideal opportunity to observe the influences of these factors.

Pricing dynamics in the cat bond market are particularly open to comparative analysis. Most cat bonds are rated by a rating agency such as Standards and Poor's. The rating, which reflects the perceived degree of risk associated with holding an investment in the security, is comparable to the ratings of other types of security. It is therefore possible to compare the pricing of cat bonds with the pricing of other forms of investments that display similar rating profiles.

Up until mid-2003 it would appear that the dominant driver of cat bond prices was the credit spread in the debt capital markets. The price of reinsurance in the catastrophe bond market closely followed the credit market over this time, albeit cat bond spreads were a little above corporate spreads. This suggests that investors were prepared to put faith in the risk analysis that underpins the ratings and only sought a slight premium over the price of other, more familiar, investments.

From mid-2003 until mid-2005 there was divergence as corporate spreads continued to tighten while cat bond spreads held firm. This was largely due to a restricted investor market, limited supply and a consequent lack of liquidity.

“The hedge fund and insurance-linked funds are prepared to accept greater risk for higher returns

The Katrina effect

The real shift in the market arises from late 2005, following Hurricane Katrina. In effect, it represents a shift in the dominant driver of cat bond pricing from credit spreads to reinsurance market conditions and pricing. In an efficient market one would not have expected to see such a significant shift, since the effective arbitrage opportunity would have been in part eliminated by the free movement of capital.

The post-Katrina surge in demand for new sources of risk transfer forced up cat bond market yields with the side-effect of opening the market up to a whole new class of investor. The hedge fund and insurance-linked funds are prepared to accept greater risk for higher returns. Attracted by the relative returns on offer, the supply of investor capacity has continued to grow leading to a surplus of supply over demand. As a result there has been a steady re-convergence of spreads during the last 12 months.

This cause and effect is more clearly demonstrated when we look at the perils to which catastrophe bonds are exposed (see figure 2). Unsurprisingly, given traditional reinsurance capacity constraints, US hurricane is by some way the peak exposure in the catastrophe bond market. By comparing cat bond pricing for US hurricane risk to other similarly rated perils it quickly becomes clear that demand for US hurricane capacity is directly influencing cat bond pricing.

This interaction between the effects of credit spreads and reinsurance pricing has also been influenced by what we might term "ratings migration". The early attempts by the rating agencies to assign credit ratings to cat bonds were overshadowed by an aura of extreme caution. The consequence of this was the assignment of ratings relative to the expected loss for the bond. These were considerably more conservative than the equivalent debt security with the same expected loss.

However, as the rating agencies have become more familiar with the technical issues surrounding catastrophe risk modelling, and the models themselves have become more robust, the ratings relative to the expected loss have improved. Following the hurricane season of 2005 however, the risk models were recalibrated and ratings have since been based on more conservative near-term views.

Credit crunch challenge

Having experienced such a good run, will current stresses in the debt capital markets influence the future prices of catastrophe bonds? The fact that the debt capital markets have come under stress in the last few weeks cannot be doubted. The fallout of the collapse of the subprime mortgage market in the US continues to be felt across the financial markets.

There have been no new catastrophe bond issues since the credit crisis began, so forming a definitive view of how the cat bond market will respond is difficult. Anecdotal evidence however would suggest that the market is actually responding positively. In a recent interview, John Brynjolfsson, who manages $2bn of catastrophe bonds at Pacific Investment Management, said: "Cat bonds are kind of a hybrid between the capital markets and the reinsurance markets. Although credit-market spreads have been blowing out, reinsurance pricing has been softening. The premiums that clients have been paying to insurance companies have been shrinking... That seems to be consistent in the cat bond market." In summary we are seeing the spreads (equivalent to the cost of reinsurance) tightening in the cat bond market.

“Experienced investors recognise and value the differences between credit risk and event risk, not least of which is the lack of correlation between the two

While such an observation might be intuitively obvious, caution should be exercised. It is important to remember that stresses in the debt capital markets have come at a time when the spreads on cat bonds are still relatively high historically. Furthermore, there has never been a period in the past when the spreads on cat bonds have traded below the equivalent debt capital market spreads.

If credit market spreads continue to widen, the critical issue to be resolved is whether they will pass through cat bond spreads or simply push them wider? The latter could be expected to lead to a reduction in the supply of business to the catastrophe bond market. In fact, if conditions in the debt capital markets suddenly forced spreads wider, there is a material risk that cat bond issuance might evaporate altogether.

This is unlikely to occur in our opinion. A characteristic of the current credit crunch is not just the widening of credit spreads but also the bid-offer spread, the latter being associated with the greatly reduced liquidity in the market. Where liquidity is at a premium the alternative catastrophe bond market will be an attractive alternative for buyers of high yield debt. Furthermore, as has been observed, reinsurance pricing is falling and, barring significant losses in the reinsurance markets, this is likely to continue to exert pressure on cat bond spreads.

Protective bubble

Many factors have influenced the pricing of catastrophe bonds. Over time there has been a growing divergence between the broad debt capital markets and the cat bond markets, a divergence that has been driven by extreme circumstances in the reinsurance markets. The question now is whether the current extreme conditions in the credit markets could have a similar bearing?

The credit crunch may well slow down the supply of capital to the cat bond market and the insurance-linked sector as a whole, but is unlikely to stall the market entirely. There are now many experienced investors who recognise and value the differences between credit risk and event risk, not least of which is the lack of correlation between the two.

In essence, we see that the insurance-linked sector is a complex amalgam of two distinct markets and is, in effect, a conduit that permits the free flow of risk and capital between the two. The larger the sector becomes, the more effective the conduit, and the closer the two markets can be expected to track each other. In the meantime, there exists a degree of elasticity and we expect to see the insurance-linked sector behaving, in part, independently of the credit markets.

This feature was researched and written by the team at Willis Capital Markets.