Weathering the Storm (April 01, 2006)

Katrina. Rita. Wilma. These names are now associated with some of the worst – and most costly – natural disasters in history. On a global scale, natural catastrophes and man-made disasters set records in 2005. Hurricanes might have captured headlines, but each geographic area had its own set of issues – from mudslides in Switzerland to the earthquake in Kashmir.

Canada didn’t escape this trend. Tornadoes and floods from rising rivers in Canada resulted in US$342 million in insured losses last year. Insured loss figures for a fire that followed two explosions at an oil sands plant in Fort McMurray, AB are still unknown, but the total economic damage was in the neighbourhood of US$1.2 billion. These events are prominent entries in this year’s sigma report from Swiss Re’s Economic Research & Consulting team. Since the early ’70s, sigma has published a series of reports focused on insured property loss and loss of life due to natural catastrophes and man-made disasters – including windstorms, floods, earthquakes, droughts, forest fires and hail.

The trend for the past 15 years has clearly been upward. Insured property and business interruption loss – on an inflation adjusted basis – has been rising dramatically. Year 2005 was no exception. The numbers of events have also been going up: sigma recorded 149 natural catastrophes and 248 man-made disasters in 2005. A driving force behind the long-term trend of rising numbers and catastrophe costs is the increasing geographic concentration of insured assets, fuelled by population growth and economic activity in high-risk areas.

The loss of life in 2005 was as remarkable as it was heartbreaking. On Oct. 8, 2005, a 7.6-magnitude earthquake hit both the Indian and Pakistani administered portions of Kashmir. More than 73,000 people died. Floods, storms – foremost among them hurricane Katrina – shipping and aviation disasters also claimed large numbers of victims. Sigma data places the death toll due to catastrophes in 2005 at more than 97,000.

PROPERTY AND CASUALTY TOLL

Swiss Re estimates total natural catastrophe economic losses in 2005 at over US$230 billion. A significant part of these losses occurred in industrialised nations, where catastrophes affected a high concentration of property assets. In the United States, for instance, a series of hurricanes caused tremendous economic damage – Katrina alone caused an estimated US$135 billion in damages – followed by Wilma with US$20 billion and Rita with US$15 billion, respectively. However, the earthquake in Kashmir, the floods in India in July and hurricane Stan in Central America in October also caused economic losses in the billions in developing nations.

Of the total damage, insurance covered just one-third of the losses. Damage due to natural catastrophes cost insurers worldwide around US$78 billion – making 2005 the costliest year ever for the industry.

In addition, man-made disasters financially impacted property insurers in the amount of approximately US$5 billion. More than two-thirds of this figure relates to large-scale fires and explosions in the industrial and energy sectors. Aviation and space insurance again experienced a comparatively low-loss year.

As in previous years, industrialised nations dominated the table of insured losses – partly a reflection of their higher insurance density, and partly as a result of four hurricanes making landfall in the United States. The exposure characteristics of many population centres and the lack of engineered safety features in highly exposed areas explain in part the high insured losses in industrialised nations and the large number of fatalities in emerging markets despite low insured losses.

WINDSTORM IMPLICATIONS

In 2004, the insurance industry had recorded high losses due to a series of hurricanes in the Caribbean and the United States (Charley, Frances, Ivan and Jeanne), as well as typhoons in Japan (Songda, Tokage, and Chaba). However, the 2005 hurricane season has exceeded that previous record by far. Insured hurricane damages made 2005 the most expensive year for property insurers since 1906, the year of the San Francisco earthquake.

Of US$73.5 billion in losses worldwide, US$69 billion stemmed from hurricanes in the United States and neighbouring countries. The losses of US$45 billion from Katrina alone are equivalent to 9% of U.S. property and casualty direct premiums written. To put this in perspective, the losses from Katrina were more than double the US$22.3 billion paid out following Hurricane Andrew in 1992.

In the search for explanations, climate science points to several reasons for the distinctive traits of the last two hurricane seasons. In general, the frequency of strong windstorms is related to natural climate fluctuations. These can largely be explained by Atlantic Multidecadal Oscillation (AMO), an ongoing series of long-duration changes in the sea surface temperature of the North Atlantic with a cycle period of 50-70 years. The successive cool and warm phases may last for between 20 and 40 years at a time, with a difference of about 0.6C in annual mean temperatures. Since the intensity of a windstorm depends largely on sea surface temperature, far fewer intense hurricanes occur during an AMO cool phase (such as in the ’70s) than during a warm phase.

The world has been in an AMO warm phase since 1995. This has translated into a significant increase in the number of Category 4 and 5 hurricanes as measured on the Saffir-Simpson scale – that is to say, wind speeds of 210 km/h and more. The “power dissipation index,” a measure of windstorm destructiveness, has more than doubled since the AMO cool phase of the ’70s.

The property and casualty industry is in the process of realizing that it is in the middle of an active period, with an increase in expected hurricane activity. Modelling agencies are revising frequency and severity assumptions upward. Insurers and rating agencies now contemplate total aggregate exposures and scenarios with multiple events. As a result, both commercial property and reinsurance premiums in hurricane-exposed regions have increased significantly.

UNDERESTIMATING FLOOD DAMAGE

The extent of flood damage in 2005 raises some noteworthy insurance issues in both Canada and the United States. In the U.S., for example, private persons and small corporations can insure themselves against flood damage only under the National Flood Insurance Program (NFIP).

Up to now, private demand for flood cover has been weak. This is due to the voluntary nature of the insurance, which leads to adverse selection – concentrations of highly-exposed policies – with correspondingly high premiums. Despite the low penetration of flood insurance, claims totalling US$23 billion were reported to the NFIP for all of 2005 (of which US$20 billion were from Katrina). None of the sigma insured loss figures include NFIP claims.

Private insurers provide flood cover for commercial risks as part of all-risk policies, although these often exclude flooding for risks located in highly exposed flood hazard zones. A large part of the Katrina losses is attributable to commercial multi-peril and business interruption claims relating to flooding. Cat models have not covered these dimensions of exposure, as well as unprecedented offshore energy claims, in the past. Also, some underwriters were not sufficiently aware of the limitations of their tools and their data input. Insurers and modelling agencies are currently working on improving or introducing new flood models.

PREPARED FOR WORST

The U.S. property and casualty industry’s underwriting profitability for the first nine months of 2006 was surprisingly strong at a combined ratio of 99.9%. One reason is that the industry had entered the hurricane season in a very strong underwriting position with 1H 2005 combined ratio at 92.7%. The other main reason for the industry’s resilience was the benefit of reinsurance covers. Reinsurers, mo
st of them international, bore the brunt of the losses caused by Katrina, Rita and Wilma. Only 37% was left with U.S. primary insurers.

The magnitude of hurricane claims in relation to the size of some insurers came as a surprise to many. Going forward, rating agencies will be a driving force behind increased capital needs. Higher uncertainties regarding property risk will require more capital for each premium dollar, particularly for smaller and undiversified insurers and reinsurers. A wave of downgrades and negative outlooks by the major rating agencies indicates a tighter capitalization of the reinsurance industry. The new capital raised at the end of 2005 does not match the increase in capital requirements. Clients will also rethink the stability of relationships and balance sheets, fueling a flight to quality.

Paradoxically, a year of great challenges has demonstrated the strength of the insurance industry. Even with a dramatic new dimension in economic and insured losses, insurance remains a powerful tool for coping with natural catastrophes and man-made disasters.