Modern day coastal development is making U.S. hurricanes more potent, but catastrophe modeling firms may be applying “incomplete” historic loss information for insurance and reinsurance industry clients, according to a report issued Tuesday by catastrophe consulting firm Karen Clark & Co.
According to Clark’s assumptions, losses from the largest historic hurricanes would be much greater than previously estimated. The report states that 13 out of 28 of the largest hurricanes since 1900 would have caused insured losses of $10 billion or more, given today’s values.
Florida would have largest loss at $125 billion if the intensity and track of the 1926 Great Miami Hurricane was repeated, Clark adds.
The report argues that catastrophe modelers method of providing loss estimates for historical hurricanes — by simulating storms and superimposing them on estimates of current property values — is often flawed because each firm uses different damage functions, industry databases and meteorological parameters.
Clark argues that the meteorological data on historical hurricanes is particular troublesome since it is “sparse, subject to error, and particularly for storms occurring early in the 20 the century, it is often incomplete. “
“Different meteorological data sources can provide different information for the same key hurricane parameters, and this is what drives the variability in the simulated losses,” the report states.
In addition, modelers also rely on a single vender — Property Claims Service — as the primary source for insured losses in the US, even though its data “is also imperfect and incomplete.”
“PCS figures are only available going back to 1950, and the PCS loss estimates may not account for all sources of loss, depending on the companies surveyed and the nature of their property business,” the report states. “PCS does not adjust their loss estimates on an ongoing basis to reflect what size losses the historical hurricanes
would cause today. Adjustments that have been made underestimate how much larger the losses would be today.”
“We conducted this study because our clients asked us to provide historical event information as another tool they can use to assess their catastrophe loss potential,” said Glen Daraskevich, senior vice president of the Karen Clark & Co. in a statement. “Researching the various data sources we found incomplete and conflicting information, which led us to compile and thoroughly evaluate all of the sources.”
