Skip to content

California’s Insurance Market Crisis: Risk-Based Pricing & Data Gaps

    Even as California moves to address regulatory obstacles to fair, actuarially sound insurance underwriting and pricing, the state’s risk profile continues to evolve in ways that underscore the importance of risk-based insurance pricing and investment in mitigation and resilience, according to Insurance Information Institute and HazardHub’s report.

    Much of California’s challenge is related to a 1988 measure – Proposition 103 – that constrains insurers’ ability to profitably insure property in the state.

    In a dynamically evolving risk environment that includes earthquakes, drought, wildfire, landslides, and damaging floods, Proposition 103 has made it hard for some insurers to offer coverage in the state. In some cases, this has led to insurers deciding to limit or reduce their business in the state.

    With fewer private insurance options available, more Californians are resorting to the state’s FAIR Plan, which offers less coverage for a higher premium.

    For many, this “insurer of last resort” has become the insurer of first resort. This isn’t a tenable situation for the state or its policyholders.

    California’s insurance availability/affordability challenges will require a multi-pronged approach, and underlying every component is the need for granular, high-quality, reliable data.

    Insurance Underwriting Profits Can’t Keep up with Losses

    Insurance Underwriting Profits Can’t Keep up with Losses

    Insurers’ underwriting profitability is measured using a “combined ratio” that represents the difference between claims and expenses insurers pay and the premiums they collect. A ratio below 100 represents an underwriting profit, and one above 100 represents a loss.

    Insurers have earned healthy underwriting profits on their homeowners business in all but two of the 10 years between 2013 and 2022.

    However, the insurance claims and expenses paid in 2017 and 2018 – due largely to wildfire-related losses – were so extreme that the average combined ratio for the same period was 108.1 (see Natural Catastrophes Drive Record-High Economic and Insured Losses).

    California Homeowners Direct Combined Ratio

    California Homeowners Direct Combined Ratio
    Source: Triple-I analysis of NAIC Annual Statements available through S&P Global Market Intelligence

    Underwriting profitability remains essential because it supports policyholder surplus—the financial reserve insurers maintain to pay future claims.

    Accurate pricing based on risk ensures this surplus remains stable. Insurers must align pricing and underwriting with actual risk levels, especially in regions exposed to significant hazards.

    Proposition 103’s impact on pricing

    California differs from most states by prohibiting forward-looking catastrophe pricing. Proposition 103 mandates that insurers rely only on historical data, excluding recent risk developments and advanced modeling tools. It also bars insurers from factoring reinsurance costs into their pricing structures.

    Reinsurance allows insurers to expand capacity, and its costs have increased in parallel with primary insurance.

    When insurers cannot incorporate reinsurance expenses, particularly in high-risk zones, they must absorb these losses through their surplus, scale back their operations in the state, or both.

    Insurance Policy Response in California

    Insurance Policy Response in California

    Insurance Commissioner Ricardo Lara introduced the Sustainable Insurance Strategy to stabilize the insurance market while responding to climate risk.

    One directive requires insurers providing homeowners policies to write at least 85% of their statewide market share in regions identified as under-marketed.

    These are areas where coverage remains limited despite potential insurability.

    Guidewire reports that wildfire mitigation and structural reinforcement can cut wildfire-related losses by up to 70%. However, identifying viable properties in these zones remains challenging.

    Regional Variation and Case Studies

    Mapped case studies show how wildfire risk differs by region due to terrain, population density, income levels, local weather patterns, and proximity to nearby towns.

    These differences reinforce the need for more detailed and location-specific risk evaluations in underwriting practices.

    Areas are differentiated by location and topography

    Areas are differentiated by location and topography

    HazardHub’s wildfire risk assessment

    The HazardHub Wildfire Score uses updated local vegetation data, topographical details, distance to emergency services, past wildfire events, and environmental metrics to generate precise wildfire risk ratings. These ratings support insurers in identifying potentially insurable properties in high-exposure regions.

    Triple-I applied the HazardHub Wildfire Score to several under-marketed areas in California, selected for their geographical and demographic traits.

    The analysis demonstrates how localized data can reveal underwriting insights beyond those offered by traditional risk models.

    Every property being assessed for wildfire risk is unique

    Every property being assessed for wildfire risk is unique; therefore, it’s important to subject as many relevant variables as possible to analysis.

    For example, proximity of structures to fuel is important – but to be more predictive it helps to know more: What kind of fuel? Is there potential for a wind-driven event? Is the property on a hill? If so, is it north-facing?

    The model presented here includes standard variables, such as slope, aspect, wildfire history, wind, and the amount of nearby vegetation. It also includes some key differentiators, including type of vegetation and fire-suppression success rate.

    The charts to the right provide HazardHub’s breakdown of the fuel characteristics for the ZIP Codes under consideration.

    Every property being assessed for wildfire risk is unique
    Source: Triple-I analysis of NAIC
    Every property being assessed for wildfire risk is unique
    Source: Triple-I analysis of NAIC
    Every property being assessed for wildfire risk is unique
    Source: Triple-I analysis of NAIC

    Some model providers would report vegetation types for both 95667 and 95490 as the most prominent – mixed / evergreen – and 90210 would be considered as open space with light urban development.

    Others would use the most hazardous fuel type, in which case all of these study areas may be considered shrub / scrub.

    HazardHub determines the weighted average fuel type according to the amount of each fuel type in the area.

    Consequently, we can consider the entire pie chart for each study area as one component of the model, allowing for a more granular analysis.

    Los Angeles County Case Study

    Los Angeles County Case Study

    Beverly Hills and the Hollywood Hills highlight how wildfire risk in California impacts both high- and low-income communities. These areas sit at the wildland-urban interface, where residential development has expanded into fire-prone terrain on the outskirts of Los Angeles. Residents choose these neighborhoods to remain near the city while avoiding dense urban zones.

    The City of Beverly Hills has taken steps to evaluate local wildfire risk and improve evacuation infrastructure. Despite these efforts, insurance challenges persist.

    ZIP code 90210, home to approximately 21,000 people, has a median home value exceeding $2 mn according to U.S. Census data—or $4.9 mn, based on Zillow estimates. Yet many homeowners face difficulty securing wildfire insurance.

    The entire area north of Sunset Boulevard falls under the “Very High” classification within the Fire Hazard Severity Zone (FHSZ), as defined by the California Department of Forestry and Fire Protection.

    Adjacent regions in Los Angeles County—though facing lower direct risk—are bordered by undeveloped, steep terrain, which complicates firefighting operations and heightens the overall danger.

    Mendocino County Case Study

    Mendocino County Case Study

    Willits is a quieter community located among the trees adjacent to the northern expanse of California’s wine country near Ukiah and Hopland.

    Homes are much more affordable here, with a median home value of $360,000 in the 95490 ZIP code, and residents typically work in forestry, vineyards, or local institutions.

    The relative isolation is probably why this region has so little population, being 2.5 hours from San Francisco.

    Willits and Ukiah are on the rain-shadow side of a coastal mountain range, which means that while towns like Fort Bragg and Mendocino are typically too wet to burn, wildfires are common around Willits.

    The 2018 Mendocino Complex Fire, the largest wildfire in California history at the time, and the 2020 August Complex Fire, the current largest, both burned in a similar landscape elsewhere in Mendocino County.

    In Mendocino County near Willits, there are some properties with composite scores as high as 222. However, the average composite score here is half that of El Dorado County’s. The wind risk here is also half of the wind risk in the first case study.

    El Dorado County Case Study

    El Dorado County Case Study

    Placerville in El Dorado County and Auburn in Placer County represent a middle ground between the high-income communities of Beverly Hills and the rural affordability of Willits.

    These towns offer a quieter, forested setting within reach of Sacramento, attracting residents seeking a more relaxed lifestyle outside the capital.

    ZIP code 95667 in Placerville has a population of approximately 37,283 and a median home value of $485,000. The climate in the Sierra Nevada foothills is generally mild, avoiding the extremes of higher elevations, which adds to the area’s appeal.

    While these towns experienced population growth leading up to 2020, numbers have since declined as the benefit of commuting to Sacramento has become less relevant.

    The region was significantly affected by the 2021 Caldor Fire, which destroyed 1,005 structures. Among the areas studied, El Dorado County presents the highest wildfire and wind risk, with a maximum composite risk score of 320 and an average of 204.

    Over the past 25 years, this area has recorded four times more wildfire incidents than the others under review.

    Frequency of High-Severity Natural Catastrophes

    The rising frequency of severe natural disasters—such as wildfires, floods, and hurricanes—has emphasized the importance of risk-based pricing informed by forward-looking, high-quality data. Insurers must rely on accurate assessments to reflect current and projected risk conditions.

    To maintain availability and affordability of insurance for homeowners, insurers require more detailed and adaptable risk models.

    These models should incorporate live environmental conditions, mitigation activity, and specific property features.

    Advanced approaches—such as wildfire hazard scoring based on vegetation type, fire control success rates, and predictive climate models—support better underwriting, enabling more accurate pricing and risk segmentation.

    Incorporating mitigation actions into these models and rewarding them financially helps lower exposure, expand coverage options, and support preventive measures.

    A viable property and casualty insurance market depends on aligning premiums with actual risk. This ensures financial sustainability while keeping coverage accessible in areas where it is needed.

    Supporting this system requires regulatory flexibility that allows insurers to adopt advanced modeling tools and apply real-time data in pricing.

    These capabilities help identify lower-risk locations within high-risk zones, refine pricing, and encourage mitigation rather than market withdrawal.

    FAQ

    Why is Proposition 103 considered a barrier to effective insurance pricing in California?

    Proposition 103 prevents insurers from using forward-looking catastrophe models and current reinsurance costs in pricing decisions. This restricts their ability to reflect the actual level of risk, particularly in areas exposed to natural disasters.

    What is the impact of restricting catastrophe pricing on homeowners?

    Limiting insurers to historical data has led some to reduce or withdraw coverage in high-risk areas. As a result, more homeowners rely on the FAIR Plan, which typically offers narrower coverage at higher premiums.

    How do combined ratios reflect insurer profitability?

    The combined ratio measures the relationship between claims/expenses and premium income. A ratio below 100 indicates a profit; above 100 means a loss. Wildfire losses in 2017–2018 pushed California’s average ratio over 108 for the decade.

    How does California’s Sustainable Insurance Strategy aim to improve market conditions?

    The strategy requires insurers to write at least 85% of their statewide homeowners business in under-marketed areas. This effort is intended to improve coverage access while encouraging investment in risk reduction.

    What role does data play in identifying insurable properties in wildfire-prone areas?

    Tools like the HazardHub Wildfire Score assess properties using environmental, geographic, and infrastructure data. These detailed scores help insurers evaluate property-specific wildfire risk more accurately than traditional models.

    How do wildfire risks differ across regions like Beverly Hills, Willits, and Placerville?

    Beverly Hills faces high risk despite high property values. Willits has moderate scores but frequent fire exposure due to its location. Placerville shows the highest composite and wind risks, with four times more incidents over 25 years than the other case studies.

    What is needed to stabilize California’s insurance market in the face of climate-related risks?

    The market needs updated regulatory frameworks that permit the use of real-time data and advanced modeling. Insurers must also account for mitigation efforts and offer pricing that reflects current risk conditions to remain viable and ensure broad coverage availability.

    ……………………

    AUTHORS: Michel Léonard – PhD, CBE, Chief Economist and Data Scientist at Triple-I, Dale Porfilio – FCAS, MAAA, Chief Insurance Officer of Triple-I