Last year saw a record issuance of catastrophe bonds. Was this just a hard market blip, or are cat bonds here to stay? By Sarah Goddard.

Back in 1997, there was a mixture of excitement and consternation in the re/insurance markets. Among much chatter and speculation about the real viability of catastrophe bonds, Winterthur came up with one of the first catastrophe bond transactions. Although Hannover Re had sponsored the first ever catastrophe bond back in 1994, it was the $6m issuance for hail in Switzerland that effectively gagged the naysayers who had contended it just wasn't possible, and opened up a new channel of risk capital for catastrophic perils.

The rest of the 1990s still had commentators speculating about the feasibility of the capital markets as a competitive source of cover. During that time, the re/insurance sector continued its downwards trend towards the bottom of the cycle, and there is little doubt that traditional property cat cover was being offered at knock-down (and for some reinsurers knock-out) prices. While some buyers were announcing that the risk environment had changed and that the soft market would be the new, constant market conditions, there continued to be interest in the cat bond sector, albeit dulled by the 'all stock must go' sales in the reinsurance community.

But gradually, through offerings such as the Residential Re, SLF Re and Golden Eagle series, the products became more standardised and the investment community increasingly interested in what cat bonds had to offer.

"It was a gradual change," explained Christopher McGhee, Managing Director of MMC Securities, "people got more educated." And they were more educated to the point that last year saw a record issuance of cat bonds. According to figures published earlier this year in an update on the catastrophe bond market produced by MMC Securities and Guy Carpenter & Co, $1.22bn was issued in 2002, up more than 26% on the previous year's total of $967m. That figure did, however, represent a 15% drop on 2000's then record issuance of $1.136bn. From the Winterthur hail bond in 1997 through to the end of 2002, 46 cat bonds were issued, with total risk limits exceeding $6bn. Of these, 16 bonds were issued in 2001 and 2002 (see table 1).

From Mr McGhee's point of view, one of the main changes he has witnessed in recent years is the ever-increasing investor demand for cat bond products. "There is now a well-developed group of investors which understand cat bonds and cat risk well," he said. These investors are committed to the asset class - indeed, there now exist several dedicated cat bond funds, and word on the street is that there are likely to be more popping up in the not too distant future.

John DiCaro, vp at Chicago-based investment bank Cochran, Caronia & Co, saw the biggest driver for the cat bond market over the past 18 months as being the imbalance between the supply of cat bonds, both new issues and the secondary market, and the demand for securities. In a risk-linked securities market review issued earlier this year, Cochran, Caronia estimated that, including $700m of investor capital raised to assume risk through securitised quota share arrangement, more than $1.6bn in risk-linked securities (RLS) were issued during 2002. "Reports of the demise of the RLS market have been greatly exaggerated," stated the report.

Looking at the RLS market performance over the course of 2002, it becomes obvious where the attraction lies. According to Cochran, Caronia figures, overall the market returned 9.07% during 2002 (see table 2), compared to a return on the Merrill Lynch High Yield Master II Index of -1.89%. Since the end of 1999 to the end of 2002, the RLS market managed to outperform the Merrill Lynch High Yield Master to the tune of 36%.

The continued stability of these returns is pushing the interest ever further, and investors such as PIMCO (Pacific Investment Management Co), one of the largest fixed income managers in the world, have become increasingly involved in the business. At the same time, cat bonds, unlike other investments, are only subject to the vagaries of nature; there is no correlation with equity and debt because of the sheer fortuity of the triggering event. What's more, the very absence of a secondary market in cat bonds at the moment would tend to back the observation that the investment community's interest is high.

From the reinsurers' point of view, diversifying the risk capital backing these very low frequency, high loss events has its obvious attractions, particularly when the reinsurance community continues to see the credit ratings slip across the industry. At a time when AAA is identified more with a US motoring organisation than a reinsurer rating, the security offered by cat bonds has a distinct appeal, as they have gradually shifted up from junk status. Of the top ten reinsurers, eight have issued either catastrophe bonds or securitised quota shares. This could mean, however, that the largest issuers may have reached saturation point, though there is a swell of opinion that the middle market may well start issuing risk-linked securities.

It is Swiss Re which recently has been dominating the cat bond market through its Pioneer 2002 issuance. This bond in particular captured the investment community's imagination, with its regular shelf offerings which are taking advantage of the constant demand in the market (see tables 1 and 2). Pioneer is continuing its offerings during the course of this year, and again is proving hard to buy in the secondary market. According to Cochran, Caronia, Swiss Re obtained $386.5m in protection for three and a half years through 2002's Pioneer offerings. This gave Swiss Re a 4.97% risk premium for a 1.10% expected loss, compared to Munich Re's $300m raised in 2000 for a 7.01% risk premium against a 1.31% expected loss (this was through the Prime CalQuake, Eurowind and Prime Hurricane securities).

Other highlights of 2002 included the first ever cat bond issued by a Lloyd's entity, using St Agatha Re (named after the saint invoked to protect against volcanic eruption and the outbreak of fire, and the patron saint of bellringers) as the special purpose vehicle. This particular bond saw Lloyd's syndicate 33, part of the Hiscox group, issue a $33m bond for earthquake in California and the New Madrid fault. Of particular note during the year was Vivendi Universal, the first-ever US corporate to issue a bond, and only the second ever corporate to do so (the first, Oriental Land, was issued in 1999 as a protection for Disneyland Tokyo should an earthquake occur within a certain radius of the theme park).

Reaping benefits
Although figures are hard to come by, it would appear that 2003 is seeing re/insurers continuing to reap the benefits of the investors' continued interest in the class. According to MMC's Mr McGhee, this year should roughly equal last year's results. But with word on the street suggesting that a new bond, Phoenix, has been successfully issued by Japanese agricultural mutual Zenkyoren, and a Taiwanese bond currently under development, there are signs that it could outpace 2002's record.

Should the Taiwan offering successfully reach the market, it will be the first new geographic region addressed by cat bonds for many years. As a general rule, the risk locations continue to be the main concentrations of values and potentially extreme natural catastrophe events, such as California earthquake, US East/Gulf coast windstorm, European windstorm, and Japanese earthquake and typhoon. But the current buoyancy of the market has shifted the focus to new risk locations, and the World Bank is reported to be looking at using cat bonds for developing economies. Where currently there may be a stumbling block for this, however, is the probable lack of accurate data upon which to base the underlying risk modeling. Nevertheless, parts of the risk modeling community are beginning to look at accruing data from countries which may in the not too distant future have a developing re/insurance sector, with low frequency, high exposure natural perils alongside. How soon there will be sufficiently accurate models to enable a bond issuance is, however, uncertain, bearing in mind that European flood, particularly in central Europe, is still seen as being too uncertain to be able to warrant trying to develop a bond structure.

Nevertheless, with the potential expansion into the mid market, and future geographic diversification (always a selling point for investors), it looks like the cat bond market has a long way to go before it reaches saturation.

To download table 1 click here .

To download table 2 click here .

By Sarah Goddard

Sarah Goddard is the editor of Global Reinsurance.