Property owners in hurricane-prone coastal areas know the impact that catastrophe models can have on their insurance programs. But in inland areas where thunderstorms, hail and tornadoes pose a bigger risk, it’s been a different story. While predictive models for severe thunderstorms have been available for years, they haven’t had quite the same impact. Now, convective storm models are roiling the property market as thunderstorm losses routinely reach into the billions of dollars. From the Southeast across Texas and the Great Plains to the Front Range of the Rockies, more insureds are facing higher deductibles, percent deductibles, and potentially higher rates, as carriers make greater use of convective storm models to manage their own risks.

With thunderstorms accounting for about 40 percent of annual catastrophe losses, carriers are asking more insureds to shoulder more of the risk. Properties in areas classified as high hazard by the models should expect greater scrutiny and higher costs. That’s also true for risks with a history of losses from hail, tornadoes or straight-line winds. Instead of more traditional flat deductibles for wind and hail, carriers are looking for percentage or per-roof deductibles in areas at statistically greater risk. That’s particularly true for frame apartments in the Great Plains region. There is some good news for properties in lower hazard zones, which may see more favorable rates and terms. Here’s how the models work.

STORM DAMAGE

Unlike hurricanes, which are far less frequent and provide plenty of warning, thunderstorms pop up on a daily basis. Violent thunderstorms can spawn tornadoes that blow apart buildings and rip off roofs; generate straight-line winds that knock down trees and mobile homes, and drop hail that destroys roofs, windows and autos. The damage may be highly localized. A tornado can obliterate one building and leave an adjacent one standing. Still, the convective storm models offered by risk modeling firms AIR Worldwide and RMS work like those for hurricanes. They take a baseline set of hazards and rank areas by level of risk based on a simulated 10,000-year—or longer—catalog of storms. The calculations factor in historical weather data as well as billions of dollars of actual insurance claims.

The models can separately evaluate the potential damage from tornadoes, hail and straight-line winds, which cause damage in different ways. Damage from tornadoes and straight-line winds is a function of wind speed, while hail damage is linked to the impact, size and density of the hailstones as well as wind speed. The models calculate the overall risk to a property along with a building’s ability to withstand damage. The surrounding terrain is a consideration as tornadoes can rip up trees and debris and fling them into structures. Every combination of occupancy and construction presents different risks—for instance, the type of roof and its composition, how it’s attached and its slope; how much of the exterior is glass and whether there is window protection. The models can forecast losses on an aggregate or an occurrence basis.

BOTTOM LINE

While property owners may face higher insurance costs due to the models, they also gain a better understanding of their risks with a particular structure or a portfolio of properties. That enables them to make more informed decisions on insurance coverage. Whatever the modeling results, better data is an asset in structuring and placing insurance.

ENDNOTES

i Severe Weather 101 – Thunderstorms, National Severe Storms Laboratory, see: https://www.nssl.noaa.gov/ education/svrwx101/thunderstorms/

ii Texas hailstorm industry loss could reach $1 bn: Karen Clark & Company, Artemis, June 12, 2018, See: http:// www.artemis.bm/blog/2018/06/12/texas-hailstorm-industry-loss-could-reach-1bn-karen-clark-co/

iii Quiet after the storm, Munich Re, July 7, 2018, see: https://www.munichre.com/topics-online/en/2018/07/naturalcatastrophes-first-half-of-2018

iv Tornados and thunderstorms, Insurance Information Institute, see: https://www.iii.org/fact-statistic/factsstatistics-tornadoes-and-thunderstorms