ÈȵãºÚÁÏ

Delos to Insurers Shedding Wildfire Risk: Send the Homes to Us

By | July 12, 2023

About 6 million homes in the U.S. are considered uninsurable due to wildfire risk. The models at Delos ÈȵãºÚÁÏ say about half of those homes are absolutely insurable.

“Delos sees a lot more insurable homes in California than other insurance companies,” Delos CEO Kevin Stein told ÈȵãºÚÁÏ Journal on the heels of some of the biggest names in insurance – State Farm, Allstate, Nationwide, Chubb, AIG, and most recently Farmers – making dramatic changes to their personal insurance books of business in states like California.

Wildfire-stressed areas of California. Courtesy of Delos.

Stein said , founded in 2017, is committed to filling the homeowners insurance gaps for those with wildfire risk by using a different approach than the large, generalist insurers. With an entire team from California, Delos is currently focused on homeowners in the Golden State but is in development of a nationwide wildfire model and has plans to expand with products for commercial property and vacant homes.

Schooled in aerospace engineering and satellite imagery, Stein teamed with Shanna McIntyre – herself an aerospace engineer and predictive modeler – to look at wildfire risk using imagery, data analytics, and a deep roster of scientists, professors, and researchers within a think tank to power its model and inform the insurance industry in a novel way.

“The structure of a third-party data model selling a single risk to a generalist carrier is something that worked for a very long time,” said Stein. “But now we find ourselves in a more dynamic, climate-affected environment. That structure no longer works.”

Eight out of the 10 largest wildfires in California have occurred within the last six years. That’s a huge risk to a profitable portfolio, but even more troublesome is the fact wildfire characteristics and behavior has changed, Stein said. The wind is different; traditional “seasons” are changing; temperatures are fluctuating and drying out vegetation, for example. Given the time it takes for an insurer to acquire and put a new model into production, Stein said, “It’s already out of date because the peril itself has changed.”

“Being a 3rd-party vendor comes along with many stakeholders, including many carriers with different objectives and 50 state regulators. This adds up to a lot of restrictive requirements,” Stein continued. “Vendors can’t use artificial intelligence; they can’t use predictive models. They’re sort of stuck in a world where they have a ton of constraints. Therefore, we believe strongly that only the specialist MGA structure can solve this coverage gap problem.”

Stein said that six weeks after a new wildfire behavior is observed, Delos has the data to take action and know whether its portfolio is profitable. Working in the nonadmitted market with a focus on one complicated peril allows Delos to “actually have an economically sustainable solution,” Stein added.

“The E&S world allows us to be able to do what we believe is necessary to accurately model,” he said, adding that Delos has never nonrenewed a home for wildfire exposure and has only charged more at renewal in order to account for inflation. “We have an [artificial intelligence] model. You have to be in the E&S world for that. But our product looks like an admitted product, and we want that for our brokers.”

Stein said he and McIntyre became licensed brokers in order to understand the struggles and even operated a small brokerage for several years before selling it when Delos became an MGA. Delos now also sells through a couple of carriers and is onboarding others.

“We’re going to the insurers that are pulling out and saying, ‘Hey, let us solve this problem for you,'” Stein said. “Send [the risk] to us and let us offer a policy. That way agents stay happy and they can sell other lines of business in these areas.”

Photo: (AP Photo/Jae C. Hong)

Topics Catastrophe Natural Disasters California Carriers Wildfire

Was this article valuable?

Here are more articles you may enjoy.