CA FAIR

California moved $750B of property risk onto a backstop built for a fraction of it.

The FAIR Plan is the state's property insurer of last resort — a private pool every licensed carrier must backstop, with no capital requirements and no published audited financials. As regulation held private rates below wildfire risk and carriers withdrew, the backstop became one of California's largest property insurers. The January 2025 fires were its first modern stress test, absorbed only by assessing every insurer in the state. The analysis below addresses what a subsequent event would cost, and who would pay it.

In fiscal 2021 the FAIR Plan held 242,440 policies and $161B of exposure. Four years later: 642,010 policies and $694B — growth of 165% and 332% respectively, with the quarterly book reaching $750B by March 2026. This is not organic demand; it is the voluntary market's risk, transferred.

Insured value is compounding faster than policy count — 332% vs 165% over the window. The backstop isn't just adding homes — it is adding ever more expensive homes, $1M of exposure per policy on average by FY2025.

The retreat concentrates precisely where fires do. In the mountain resort communities it is no longer a marginal phenomenon — in Lake Arrowhead (92352), 89% of the housing stock is on the backstop. The top 10 ZIP codes alone hold 9% ($63.06B) of statewide FAIR exposure; the top 100 hold 43% ($295B).

ZIPs with ≥100 housing units; shares capped at 100% (seasonal-home ZIPs can exceed ACS unit counts — see methodology). Full list on the rankings page; every lens on the map.

The chart below illustrates the growth in FAIR policies after private insurers reduced their exposure in the burn corridor. Pacific Palisades went from 527 to 1,557 residential FAIR policies between FY2021 and FY2025, with an 85% single-year increase in FY2024 — the year before the fire:

California's backstop classifies its own book into wildfire-risk bands; FEMA's National Risk Index scores every community's wildfire hazard independently. Classification is a rate driver: statewide, the average High-band FAIR policy costs $4,008 per year against $1,661 in the Low band — a 2.4× difference. Where the two models disagree, that difference is the measure of what the disagreement is worth.

The disagreement runs in both directions, with different consequences. The 30 ZIPs where FEMA rates hazard highest relative to FAIR's classification hold $7.98B of FAIR exposure across 8,227 policies, priced at an average of $1,890 per policy — if FEMA's assessment is closer to the truth, the plan is collecting roughly $17M per year less premium on these ZIPs than its own High-band pricing implies, a shortfall that converts to assessment-and-surcharge liability when they burn. The 30 ZIPs on the other side pay an average of $4,140 per policy against a $1,661 Low-band benchmark — roughly $64M per year in premium above what federal hazard modeling would support, paid by homeowners who had no alternative carrier.

Divergence (“Gap”) = FAIR high-risk-band percentile − FEMA NRI wildfire percentile, in points; ZIPs with ≥50 policies; 30 ZIPs per direction. NRI measures community hazard, not the insured book — a flag for examination, not proof of mispricing (see methodology). Explore the full pattern with the map's FAIR vs FEMA lens.

The FAIR Plan has no state guarantee behind it. In the event of a shortfall, the cost is borne directly by policyholders across the state: losses cascade through unlimited member-insurer assessments — CDI approved $1.00B in February 2025, the first in roughly thirty years — and insurers may recoup half of the first $1B, and all of any amount above it, through approved surcharges on California policies.

The January 2025 fires are the calibration point: $4.05B of FAIR gross loss against $4.78B of exposure in the perimeters — 85 cents of loss per dollar of exposure, 2.7× the plan's own historical average — and against the pre-fire book of the affected corridor, 37% of corridor-wide exposure. We apply that same corridor-wide ratio to four documented high-risk clusters, run the plan's actual reinsurance structure ($900M retention; tower to $5.78B with FAIR retaining ~53% inside it), and translate the retained loss into the surcharge across ~9M residential policies. The historical-average severity (~14% of footprint TIV) is derived, not assumed: FAIR's own ~31% historical claims-to-exposure average, scaled by the same perimeter-to-corridor geometry as the calibration event (31 ÷ 85 × 37%). These are illustrations of documented fire-prone clusters, not predictions.

A historical-average event in any of the four clusters is sufficient to re-run the February 2025 assessment. A Palisades-severity event in any of them exhausts the reinsurance tower entirely — the repeat-Palisades case leaves $5.60B uncovered and implies a surcharge near $955 per policyholder. Because the book continues to grow, these figures increase with each quarterly release.

Holding a policy did not mean being made whole. A year after the fires, 69% of total-loss survivors were underinsured, short by an average of $247 per square foot (United Policyholders year-one survey) — for a 2,000-square-foot home, an unfunded rebuilding gap of roughly $494,000. 61% of respondents reported that their insurer's own replacement estimate was inadequate. Independent evidence points the same direction: peer-reviewed analysis of Colorado's 2021 Marshall Fire found 74% of destroyed-home owners underinsured, 36% severely. Roughly seven in ten is the consistent finding for a total-loss event.

For FAIR policyholders the gap is structural: FAIR dwelling policies pay depreciated actual-cash-value by default (replacement cost is an add-on), and coverage is capped at $3M — a limit set in April 2020 and never indexed. In 31 California ZIPs the typical home value now exceeds that cap, by a median of $564k. Statewide, 13% of ZIPs show average FAIR coverage below 75% of typical home value — a median dollar gap of $549k per home — concentrated in the high-value, high-risk corridors. The pattern is mapped in the coverage-vs-home-value lens, where the darkest ZIPs insure the smallest fraction of their housing stock's value.

Income data indicates that few households could absorb such a gap out of pocket. Of the 200 ZIPs with coverage below 75% of home value, 141 of 200 (71%) have typical home values exceeding eight years of local average income — an uninsured shortfall at that scale is not recoverable from savings or borrowing. Self-funding is realistic only at the highest incomes: Atherton's average filer ($1.6M AGI) could bridge its cap gap in roughly 3.5 years, while the same gap in Newport Coast represents more than four years of a $620k average income, and most cap-binding ZIPs sit well below that:

Income = IRS SOI mean AGI per return (2022) — a mean, skewed high by top filers; the affordability picture for the median household is worse. “Years to cover gap” = (typical home value − $3M cap) ÷ average income.

Payment speed compounds the gap. Market-wide, $23.70B had been paid by March 2026 against estimated insured losses of $40.00B–$45.00B; 69% of surveyed survivors reported payment delays.

The retreat was regulatory before it was climatic. Every event below is verified against the primary source it cites.