When Hurricane Katrina hit New Orleans in August 2005, it caused the biggest insurance losses ever seen from a natural disaster. The city’s flood defenses – walls, levees, and pumps – failed to hold back the water. But it wasn’t just the infrastructure that let people down. The models insurers used to predict and price risk also missed the mark.
Back then, these catastrophe models overestimated how strong the levees really were. They didn’t fully understand the value and exposure of many commercial buildings. Most importantly, they underestimated the power of the storm surge. The tide of water pushed by huge waves is what ultimately flooded much of New Orleans. As Karen Clark, a pioneer in catastrophe modeling, explained, the storm surge was far worse than anyone thought for a Category 3 hurricane.
Since Katrina, a lot has changed. More people have moved to coastal areas, building bigger and fancier homes. These newer homes cost much more to fix or replace, and inflation has only added to the bills. Clark points out that houses today aren’t just simple structures. They come with three-car garages and marble bathrooms, pushing insured values higher.
Climate change adds another layer to the problem. According to the Swiss Re Institute, warmer temperatures could make hurricanes stronger, while rising sea levels may cause worse storm surges. This means future storm damage might be even more expensive. In fact, some hurricanes from the early 1900s would likely cause over $100 billion in insured losses if they struck today.
Today’s catastrophe models are better than ever. More computing power, detailed property data, and artificial intelligence have improved their accuracy. They also now include storm surge in their calculations. Still, there are gaps. Models often struggle with risks like tornadoes, hailstorms, and floods—events that don’t happen right on the coast but cause major damage farther inland.
For example, last year’s Hurricane Helene caused heavy flooding and landslides in places like Asheville, North Carolina, well away from the coast. Many models focused mostly on coastal damage and missed the inland flooding risk. This caught people by surprise and left many without flood insurance, since they weren’t officially seen as living in high-risk areas.
Insurance companies find themselves stuck. They don’t want to offer flood coverage if they can’t price the risk accurately, but people in vulnerable spots often can’t afford expensive premiums. Meanwhile, models rely on good data to work well. When Katrina struck, data on residential homes was decent, but information on commercial properties was poor, with many values too low and locations mixed up.
This gap in commercial property data still exists today. If another Katrina-like storm hit, insurers and businesses might face huge unexpected losses. Clark warns that this low quality of data is a serious issue. If property values rise, insurance payouts and costs will follow.
Hurricane Katrina showed that both cities’ defenses and insurers’ models can fail in big ways. Almost 20 years later, the stakes are even higher. With more expensive homes, a changing climate, and tricky risks farther inland, understanding and preparing for the next big storm has never been more important.