Risky Science Podcast and Model Q&As · · 2 min read

Why Flood Modeling and Insurance Still Falls Short — And How It Can Be Fixed

Why Flood Modeling and Insurance Still Falls Short — And How It Can Be Fixed

Today for RMN paid subscribers:

TWIA Actuarial Committee Spikes Rate Hike: “Not Worth The Hassle”

Despite years of losses and new legislation that broke in their favor, leaders shot down a minimum increase citing “PR problems.”


The Risky Science Podcast

a yellow flood sign sitting on the side of a road
Photo by Brian Wangenheim / Unsplash
Listen and follow the ‘Risky Science Podcast’
Apple Podcasts | Spotify | Overcast

In the latest Risky Science Podcast, we sit down with Cory Isaacson of reThought Flood Insurance Corporation, who’s spent years tackling flood risk modeling’s blind spots.

"The reason there was an opportunity here is because the flood zones were so inaccurate and crude," Isaacson explains. "In those days, the NFIP was pricing only based on flood zone... What’s interesting is they don’t price based on flood zone anymore."

Isaacson, a former CTO at RMS (now Moody's), shared why flood risk demands a fundamentally different approach than traditional catastrophe models:

Flood is so variable. You could have one building right next to another building where one building is great and the other one is very, very bad for flood risk.

ReThought’s answer: blending multiple cat models using what Isaacson calls "mod fusion."

Another key breakthrough came from integrating an flood resilience score:

It took us two and a half years to develop the flood resilience score. We failed for 18 months badly. It was a very, very difficult thing to learn and then prove was it right.

That blend of MGA and cat modeling, Isaacson argues, builds a product people actually want to buy and Ito expand access to private flood coverage beyond what’s mandated.

There isn’t a product with good coverage that will cover what happens. That’s exactly the problem.

🎧 Listen to the full discussion

📝 Episode Show Notes:

Read next