Every third person in the US wants to live in the path of a tropical storm or beside an earthquake fault - or so it seems. According to the US Census Bureau, the largest net population growth over the next three decades is expected in California, Texas and Florida.

It is almost exactly ten years since Hurricane Andrew struck Florida. In inflation-adjusted terms, it cost the insurance and reinsurance market about $20bn. It remains the largest ever insured loss from a natural catastrophe and is second only to the losses of September 11 as the largest ever from any type of catastrophe. Hurricane Andrew was a category 4 storm on the Saffir-Simpson scale of intensity and it avoided Miami. Two previous hurricanes that made landfall in Florida were more intense - category 5 - but they were in 1935 and 1969 and affected far less property than they would today.

The second most costly insured lost from a natural catastrophe was the Northridge, California, earthquake in January 1994, just 18 months after Hurricane Andrew. Claims have mounted to about $15bn in today's terms. Tropical storm Allison inundated parts of Texas in early June 2001 with more than 1,000mm of rain in four days, causing over $5bn in economic loss in the state and $2.7bn insured loss.

These are not the only catastrophic exposures in the US. Three intense, magnitude-8 earthquakes struck the central Mississippi Valley during the winter of 1811-1812. At the time, the area was sparsely populated. Even so, by winter's end, few houses within 250 miles of the town of New Madrid, Missouri remained undamaged.

High risk
According to the United States Geological Survey (USGS), scientists estimate that the probability of a magnitude 6 to 7 earthquake occurring in this seismic zone within the next 50 years is higher than 90%. It could be any time. But now millions of people live there, many of them in the cities of

St Louis, Missouri, and Memphis, Tennessee, but most structures have not been built to withstand earthquake shaking.

Increasing concentrations of population and insured assets in exposed areas are increasing the ultimate cost of earthquakes, storms, floods and other natural perils. Preventive measures and higher deductibles are curbing the rise in insured losses, said the Swiss Re sigma Report on Natural Catastrophes and Man-Made Disasters 2001, but the statistics show that these loss-increasing factors are still out-stripping the mitigating measures.

The realistic disaster scenarios that Lloyd's syndicates use to stress test their risk management plans work on the basis of industry losses of $50bn for a windstorm affecting Florida and the Gulf coast or $44bn for a Florida windstorm. A Los Angeles earthquake could cost $53bn in insured losses; one on the New Madrid fault in the Mississippi

Valley $29bn. Munich Re believes probable maximum losses are in the order of $20bn for a windstorm loss event in Europe, $50bn for a hurricane in the US and $80bn for an earthquake in California.

High cost
Natural catastrophes are going to cost more in real terms. Although in 2001 man-made catastrophes, inevitably, claimed public attention and natural catastrophe losses were comparatively low, there were events that indicate likely future trends. In Kansas City, the largest hail loss of all time occurred in April. It cost the insurance industry almost $2bn. The wind and hail storm Hartmut in Germany in August caused insured losses of $600m.

An intense earthquake in January in Gujarat, India, killed about 15,000 people. It destroyed infrastructure and made important transport routes impassable over several thousand square kilometres, according to Munich Re in its topics report on catastrophes in 2001. There was damage to property in a city 400km away from the epicentre. With insured damages of $100m, it must be one of the most costly ever events in India. Typhoon Nari in Taiwan in September caused insured damages of $500m-600m. Parts of the country's underground railway system were under water for weeks. The stock market and important trading centres were closed for some days.

"The most extensive possible identification of unknown loss potentials calls for a rigorous interdisciplinary application and collation of scientific, engineering and underwriting specialist knowledge," said Munich Re in topics.

In this issue of Global Reinsurance, we have tried to show the inter-relationship of these disciplines in mitigating the effect of natural disasters as well as compensating businesses and people who suffer when they occur. The events in India and Taiwan, for example, demonstrate that natural catastrophes in many parts of the world today have the potential to cause major losses to the re/insurance industry. It is also hard to imagine that that the effect of damaged roads and railways in India or trading market closures in Taiwan will not in some way feed back into the global business network.

Lee Coppack is an insurance and risk management writer with a particular interest in catastrophic losses. She is a former editor of Global Reinsurance and guest editor of this edition.