The eye of the storm

The eye of the storm

A new model predicts that inland damage could be on the rise with future hurricanes, leading to higher rates for people in places such as Baton Rouge.


Click a photo to enlarge

The threat of major hurricanes making landfall along Louisiana’s fragile coast has always been a huge concern for residents as well as insurance companies and brokers, all of whom must deal with the aftermath.



At least one hurricane computer model predicts that the damage inland could be worse in future storms, creating worries for areas beyond the coast. But some people in the state’s insurance industry are skeptical of RMS-11, the latest model created, proposed and distributed by California-based Risk Management Solutions.



Jeff Nielsen, RMS’s senior manager for natural catastrophe and portfolio solutions, recently spoke to the Independent Insurance Agents and Brokers of Louisiana about the model.



“This was the first major update since 2003, where we actually included data from the 2004-05 and ’08 hurricane seasons,” Nielsen says, referring to Katrina and Rita in 2005, and Gustav and Ike in 2008.



“We found that there’s more information to support what happens with hurricanes, especially in regard to what happens after they move inland. We have been able to more accurately represent those risks inland: How likely is it that there will be damaging winds and structural damage once the storm moves away from coastal areas?”




One finding predicts a 66% overall probable loss under the worst of circumstances, with a large percentage of that amount tied into what happens in inland cities like Baton Rouge, which three years ago took the brunt of Hurricane Gustav, with a peak wind gust of 91 mph and fallen trees that damaged homes, pulled down power lines and left much of the city in the dark for one week.



Insurers and brokers who choose to follow the RMS-11 or combine those findings with other models are feeling the impact. Such companies are being pressed by ratings agencies to increase rates paid by customers because capital reserves will have to be increased and because many will have to invest more with reinsurance companies to support their potential loss.



“This is meaningful for customers, which means it’s meaningful for us,” says David Daniel, president of the Daniel and Eustis Insurance Agency. “This new modeling concludes that the exposure for inland properties is greater than the insurance industry expected it to be. Not everybody believes in the model, but it does appear that it’s affecting the marketplace, and we have to be ready to adjust to that and how it will affect our customers.”



Adds Jeffrey Mohr of Lewis Mohr Real Estate and Insurance: “It could certainly lead to rate increases. Some companies will adopt RMS-11, and that could initially lead to a tightening of the market, especially in states like ours, and that will affect everybody, including customers in the coastal areas.”



The projected impact has given rise to some skepticism.




Before insurance companies and brokers start increasing rates, in the event that RMS-11 is widely adopted, David Albright, CEO of the Independent Insurance Agents and Brokers of Louisiana, would like to see more data.



“It raises more questions than provides answers,” he says. “It’s too early to tell what the impact is going to be. I’m skeptical of the model, given the dramatic change from their last model. It raises a question of, was the last model that far wrong or is this model that far wrong. I’m not sure how it can change so dramatically.”



An argument has been offered that the potential damage hasn’t changed, just the estimates of what might happen. The only way to know for sure, unfortunately, is for another major hurricane to make landfall in Louisiana and establish a baseline for the residual impact on inland locations.



Karen Clark is one of the forerunners of the risk management assessment field, having developed one of the first hurricane catastrophe models in 1987 for Applied Insurance Research.



In a year that has seen such disasters as the tornadoes in Joplin, Mo., Raleigh, N.C., and Tuscaloosa, Ala., and Hurricane Irene affecting the East Coast from the Carolinas to New England, Clark has been critical of the latest wave of computer models.



She refutes the notion that models can precisely forecast losses, though she advocates their use, along with human judgment, for prediction purposes.



“These models cannot pinpoint average annual loss by location or ZIP code at nearly the level of certainty they are promising,” Clark says in a recent interview with the Birmingham (Ala.) News. “It’s a false precision. We are never going to have an accurate model. These models are not, and will not be, accurate in my lifetime. We don’t have the data.”



Nielsen disagrees. He says RMS-11 is reliable.



“We’ve got technology and data to help us generate twice the amount of information than we’ve ever used before,” he says.



Part of Nielsen’s presentation to IIABL touched on how and why the projections for inland damage spiked from the previous model, done eight years earlier.



Included in the criteria was a cross-section of the physical characteristics of insured buildings inland, many of which have not historically carried high levels of insurance because the threat of a hurricane has not been perceived to be as catastrophic as the threat in the coastal areas.



Similarly to codes used for buildings and homes in areas susceptible to earthquakes, the criteria includes the construction of those structures, occupancy numbers, the year they were built and the number of stories.



While consumers and business owners in coastal areas have always paid higher premiums because of the threat of damage from wind and tidal surge, new information from RMS-11 could potentially increase premiums for people in places such as Baton Rouge.



Nielsen says the model isn’t intended to be the sole impetus for widespread price increases, but rather a component of an evaluation to see if that’s necessary. He says insurance companies and brokers can use the model for underwriting purposes, but it remains their choice as to how much they want to incorporate the information into policies and to how it will affect the amount of capital needed for reinsurance.



There might be a need to step back even further than that.



“I do not believe the property insurance industry has been charging rates that are too low,” Albright says. “Although weather patterns may be changing, it’s hard to predict or conclude what might happen inland the next time a storm hits. We need to look at each building and region separately.”



comments powered by Disqus