Fortunately, every year is not as cataclysmic as 1999, officially the second most catastrophic 12 months on record. Many of the tempests and maelstroms of the year, despite being relatively modest in ferocity, hit built-up areas in developed countries. Thus they delivered to many reinsurers their biggest catastrophe claims experience ever, triple or quadruple what they are used to.

A similar and successive pounding could bring the reinsurance industry to its knees. The chief executive of the London operation of one of the world's few billion-dollar reinsurers put it this way: “Most reinsurance senior managements failed to recognise that '99 was going to be bad, even before the year-end storms, due to inadequate pricing in all classes. Then, in January and February of this year, they failed to comprehend just how bad the winter storms would be. Hence you will see continuing deterioration throughout 2000.” The red ink will flow, he warned.

Late in July, Bermuda's remaining monoline cat reinsurer, IPC Re, announced its results for the second quarter of 2000, lending support to the London chief executive's observation about underestimation of losses. Jim Bryce, placed in the hot seat as the new ceo of IPC, said his company's second quarter results reflected “further increases in the market loss estimate for the storms of late December 1999. In addition to the impact previously advised for companies in France, we have seen an increase in the reported losses emanating from companies in Switzerland and Germany.” He blamed the revisions on “the magnitude of the events”.

Meanwhile, those that must, including some of the biggest, have drawn heavily on reserves. Swiss Re, for example, delivered a technical result on its non-life business of SFr132 million on premiums of SFr13.2 billion, but only after a reserve release of SFr696 million. In addition, the Swiss took a SFr450 million hit as an extraordinary charge direct to the 1999 profit and loss (P&L) account. The technical result also includes a portion of an allocated investment return of SFr5,932 million, but the amount attributable to non-life business remains unstated, making the company's true underwriting performance rather difficult to distil.

Others, in the face of unsettling speculation, were defiant, if remaining non-committal. A statement issued by GE, parent of big-four reinsurer ERC Group, illustrates this: “In response to media speculation, GE states unequivocally that losses in its ERC unit, due to the windstorms in Europe in late December, were less than $75 million – and were fully reserved for in GE's record fourth quarter results. [sic]” Since ERC is privately held by GE, one can only guess exactly how to interpret the last part of the statement, but it is widely believed that ERC Frankona was hammered particularly severely by Lothar and Martin, thus the “unequivocal” announcement cited above.

Warren Buffet, has always been less reticent about saying exactly what is happening, financially, in his insurance concerns. His website reveals the beating taken by his reinsurance business. “General Re's international property/casualty underwriting results for 1999 were poor.” The Berkshire Hathaway website says the year's loss ratio of 87.1%, up 14.8 points, was brought on by “inadequate” rates, higher catastrophe losses and deteriorating results in the excess liability, motor, Australian professional indemnity and film finance. But it is cat losses that are singled out: “Losses from catastrophic events, including fourth quarter 1999 European winter storms, earthquakes in Taiwan and Turkey, and an Australian hailstorm, aggregated $126 million in 1999, or 5.4 loss ratio points.”

The cost of these events is news to no one. Armed with such figures, wily reinsurance underwriters are angling for price rises in most markets. But how much do rates need to go up? To say for sure, one would have to know how windy, shaky and wet the future is likely to be.

Munich Re says that, as we begin a new millennium, there is no sign of any slowing of the long-term trend of rising losses. “If we compare the last ten years of the 20th century with the 1960s, we will see that the number of great natural catastrophes increased by a factor of three, with economic losses – taking into account the effects of inflation – increasing by a factor of more than eight, and insured losses by a factor of no less than 16,” the geoscience team concludes.

The group of 76 windstorms in the United States on 2 May, which were particularly hard on Oklahoma but made groundfalls in more than a dozen states, is expected to be the second most costly series of tornadoes in US history, and may be its tenth most expensive weather-related catastrophe. According to Swiss Re, January 1999 was a record-breaker in terms of the number of tornadoes experienced, with more than three times the number of storms than the previous record year.

If things are not getting stormier, it is clear that storms are getting costlier. In practice, however, many reinsurers seem to feel that storms of all sizes are becoming more frequent, or at least more frequent that they thought. SCOR had rated events with the loss intensity of Lothar and Martin as 100- and 30-year events respectively, but the frequency estimate changed after the two storms cut a swath through SCOR's P&L account. “We now believe Lothar-type events are 30-year events, and Martin-type events seven-year events,” admitted Jacques Blondeau while announcing the company's result for the year. He said SCOR took the brunt of Lothar and Martin on its 1999 books, which ate up its catastrophe reserves, but added that the 2000 accounts will see more impact of the two events.

“We believe the price of windstorm rates in France should double,” Mr Blondeau said. A doubling of rates has not happened, but there has been widespread realisation that PMLs have been massively underestimated, and that windstorm coverages need to be much broader as well. Anatol, the storm that crashed into Denmark in early December 1999, is said to have blown through the local industry's reinsurance programmes completely. “Virtually every insurer went through its reinsurance layers last December, even relatively large players. One company had a loss experience of DKr1 billion more than its reinsurance cover. Small, mutual and local insurers were as badly affected,” said Susan Lund, regional general manager of Scandinavia at Copenhagen Re

Overall in 1999, 138 events met the qualifications set out by Swiss Re's research group Sigma to count as catastrophes. They delivered an average loss of $177 million, compared to 125 events with an average loss of $77 million over each of the preceding four years. It is not yet clear whether, in the long term, catastrophes are becoming more numerous and more intense, or simply more expensive due to increased exposures. It is clear, however, that reinsurers, particularly those with big European exposures, took a powerful thumping last year. Most have, in their annual reports, blamed their poor performance on the “spate” of cat losses in 1999. Equalisation reserves have been drawn down, and optimism about rising rates has been preached. Yet these comments make it seem that, after the low catastrophe years of the late 1990s, some reinsurers (and their shareholders) began to forget that major loss events do happen, and their business is paying for them. One hopes that those four years of calm are not followed by four of tempest.

Adrian Leonard is a freelance insurance journalist.