Despite the unprecedented nature of the four hurricanes making landfall in Florida in 2004 and the huge disasters spawned in the Gulf of Mexico in 2005, panelists at the recent annual meeting of the Casualty Actuarial Society believed these types of anomalies are inevitable over the long term. In fact, actuarial models already included even worse storms in their ultimate scenarios. That the U.S. windstorm-exposed coastal and inland properties clearly need to become better prepared for these events was one of several observations offered at the meeting.
Michael A. Walters, Consulting Actuary, Tillinghast – Towers Perrin, said that while the insurance industry will ultimately pay out record claims of some $50 billion from Hurricane Katrina alone, the solidity of the industry remains strong because of worldwide risk-transfer and risk-spreading measures put in place after the last two mega-catastrophes: Hurricane Andrew and the 9/11 terrorist attack. However, new capital clearly is needed to replenish the capacity lost worldwide.
Future challenges, said Walters, include deciding how best to apportion the pricing so that those most at risk are not overly subsidized by those who are not. Much more needs to be done to mitigate the hurricane risk through construction improvements and retrofitting, but without recognition of the true expected cost of hurricanes and a regulatory willingness to include this cost in the charged rates, insureds may find themselves without adequate incentive to invest in such upgrades.
Actuaries can’t anticipate the ‘next surprise,’ said Sean R. Devlin, Chief Actuary – Direct Reinsurance, GE Insurance Solutions. “And nine out of ten surprises are bad.”
For example, he said, potential loss costs from natural disasters have been frequently underestimated in the past. In reviewing the 2004 and 2005 hurricane seasons, the insurers and reinsurers need to “alter the way they think of loss costs.” It is important for insurers and reinsurers to “know what you are covering, what your models are covering and what your risks are,” said Devlin.
Laurie Johnson, Vice President, RMS, observed that catastrophe modelers have recognized changes in weather patterns that have helped produce record losses. She also said that the possibility of hurricane clusters, such as the four storms that struck Florida in 2004, must be accounted for in the catastrophe models although clusters of intense landfalling storms are relatively rare.
“There are a number of phenomena that relate to each other” and impact the likelihood of multiple landfalls, said Johnson, including multiple storms taking similar paths over the same pockets of very warm water, which help them intensify. “Modeling research has shown that the risk posed by clustered storms does not significantly influence annualized losses,” she said.
Johnson also noted that the 2004 and 2005 storms also validated many modeling concepts and damage factors. She pointed out that the high activity in 2004 and 2005 marked a tipping point in the multi-decade cycle of tropical storms, heralding more frequent and severe hurricanes. “We have hit a point where we are seeing conditions right for more storms with greater potential for severe damage.”
Alice Gannon, Senior Vice President, USAA, said that, overall, the Florida Hurricane Catastrophe Fund (FHCF) has worked well in providing insurers with a form of financial backstop, noting that the FHCF retention was lowered for 3rd and subsequent events in one hurricane season beginning with 2005. “In Florida, the biggest issue is how to pay for the cost of adequate capital,” she said.
Gannon observed that if the hurricane risk was priced accurately the resulting underwriting profits would be large enough to attract additional investment capital. Ironically, the increase in capital could create marketplace competition which, in turn, might drive down prices for some individuals.
“There may never be enough discipline to assure you keep adequate rates in the highest risk places,” she said.
Source: Casualty Actuarial Society