Providing sustainable solutions to wildfire perils

Kevin Stein, founder and CEO of Delos Insurance Solutions, explains how applying AI and human expertise to risk models can provide sustainable solutions to wildfire-exposed homes.

Many people in California are now living in a wildfire zone. Often they did not move into a disaster zone; rather, the disaster zone moved into their backyard. Climate is changing the paradigm of the catastrophe peril itself and traditional models, which analyse weather-related risk, are no longer able to keep up with the evolving risk of wildfire.

Flood, fire and wind events are increasing in intensity and starting to strike in areas with no prior history of occurrence. In addition, California has experienced several new types of wildfire behaviour in recent years: big wind-driven urban conflagration first struck in 2017 before expanding into new geographies in 2018, followed by dry lightning ignition events in 2020 and fires that created their own fire weather in 2021.

Environmental observations indicate that burned areas in northern and central California during summer increased fivefold from 1996 to 2021, compared with 1971 to 1995. Further, 10 of the largest California wildfires have occurred in the last 20 years – five of which occurred in 2020 alone.

While the wildfire season was once contained to four months a year, it can now last for around six to eight months, according to the US Department of Agriculture. Western fire chiefs have gone further, suggesting California is approaching a year-round fire season.

The constantly changing wildfire pattern means that the third-party data models that insurers rely on can often be out of date on the true exposure – leaving residents and insurers unprepared.

To add more complexity and challenges to the wildfire insurance landscape, various initiatives from state governments have been implemented to ensure people have access to insurance at affordable rates. For example, the California Senate Committee on Insurance has just passed the Fire Insurance Risk Evaluation Act, designed to address the cost and availability of insurance.

For insurers trying to provide wildfire coverage, climate change and various legislative requirements have made it hard to remain profitable. Between 2017 and 2022, the California home insurance industry had a 124 percent combined ratio. Indeed, seven of the 12 largest homeowners insurance companies have withdrawn, or reduced, their insurance capacity.

Traditional insurers tend to avoid wildfire-stressed areas, since they are not wildfire experts. Rising loss ratios for incurring wildfire claims are not sustainable for insurers, but neither is the dwindling protection available to property owners, who face a Hobson’s choice of staying exposed or relocating.

Answers to a burning question

Fortunately, there is a third option for insurers and homeowners: wildfire insurance specialists focusing on wildfire-exposed regions. By better collecting and integrating a wide variety of data, specialists can identify homes that are lower risk, despite being located within these wildfire-exposed areas.

Utilising and effectively integrating all available data, and applying both human and technological expertise, allows specialty insurance providers to adapt their pricing and risk appetite to respond to changing climate conditions far more quickly than traditional approaches.

Using inputs such as slope, peak temperature, and amount and type of vegetation together determine a property’s true risk profile. An AI-driven wildfire model that understands the risk spatially, as well as the data for a specific location and how the geography influences the risk, is a solution that can make underwriting sustainable in wildfire-stressed areas.

For example, an AI algorithm that understands each location’s geography will know that the same peak temperature in Lake Tahoe means something different from that same temperature in Beverly Hills, along with the location-specific meaning of each data point.

Delos’ research has determined that out of the approximately six million homes across 13 states that struggle to find insurance because of wildfire risk, only around half are actually exposed to significant hazard.

This represents a great opportunity for the insurance industry to improve the granularity of its underwriting approach and profitably insure these homes. Being able to do this benefits everyone in the value chain and, in particular, provides homeowners in stressed areas peace of mind that they can now obtain and afford wildfire insurance.

Former California Insurance Commissioner Dave Jones was quoted in the New York Times as saying, “I believe we’re marching toward an uninsurable future” in many places. Encouragingly, the advances in AI and other new technology, coupled with focus and expertise, can assist insurance companies in providing more stable, sustainable solutions in climate change-impacted communities.