There is a new post-Katrina understanding of catastrophe exposure and retro prices are bound to seem severe as a result, explains Andrew Elliott

One of hurricane Katrina's most significant impacts is a quantum shift in the perception of the potential scale of hurricane and other disaster losses. With market estimates of the insured loss from Katrina alone in excess of $40bn this is an order of magnitude above the previously generally-accepted exposures. The result is a rapid rethink of the risks by retrocessional underwriters, with reduced capacity being offered and significantly higher prices required.

So, does this mean the retrocession market is in a crisis? The answer is potentially yes - there is little point in an underwriter offering retro covers at rates that are uneconomic to purchase or fail to give adequate return; the capital would be more wisely deployed elsewhere, hence the significant reduction in capacity. The hopes and aims of retro underwriters are that they can write a diversified book, with geographic and perils spread covering earthquake, wind etc, without undue concentration in single zones.

Serious questions are now being asked of exposure modelling techniques.

Whilst some programmes gave reasonable estimates of hurricane losses, many did not, and the fallout is that retro underwriters are taking a very cautious view of modelled numbers and greater interest in total exposed aggregates.

The new capital that came into the reinsurance market in 2001 has been seriously impacted by windstorm losses. They are rapidly recapitalising, with the aim of gaining payback in an environment of increased rates.

But is the market willing to pay the right price? Reinsurers will have to charge for the catastrophe risk if they are to be able to purchase worthwhile retro, and pass these costs on to the direct market, who in turn must of course charge a realistic rate for the catastrophe exposure.

Potential retro prices may look unpalatable at present but they reflect the current market perception of a catastrophe impacting this book in the next 12 months and ultimately may look more attractive once the cedants have re-engineered their catastrophe exposures during 2006.