ZIP Code Precision, Eight Policy Years: What the NAIC's Homeowners Data Call Puts in Regulators' Hands

Over 330 insurers, 246 million policies, five policy years, and every ZIP code in the country: that was the Federal Insurance Office's January 2025 report, the most granular homeowners insurance analysis the federal government had ever assembled. It found average premiums growing 8.7% faster than general inflation between 2018 and 2022. It found Florida's non-renewal rate at 3.35% of policies in force, the highest of any state. It found Louisiana posting the sharpest five-year jump. What it could not show was what happened next. The NAIC's June 2026 homeowners data call extended the same ZIP code-level dataset through 2025, adding three more years, the three years when non-renewal rates reached records in California and Florida, when State Farm and Allstate ceased writing new California homeowners business, and when Florida's Citizens Property Insurance grew to 1.4 million policies, making it the state's largest homeowners insurer by count. That data now sits with 50 state insurance regulators. The public report they are building for early 2027 will become the definitive regulatory benchmark for homeowners availability and pricing across the country, and carriers that model its likely findings before it publishes will enter the next state legislative cycle in a fundamentally different position than those who wait.

Jurisdictions participating in the NAIC's 2026 homeowners data call, covering every carrier writing at least $50,000 in homeowners premium in any year from 2018 to 2025

Policy years captured at ZIP code granularity, spanning the pre-crisis baseline (2018-2020) through the peak non-renewal cycle (2021-2025)

Early 2027

Planned public release of the NAIC's homeowners market report, with a public comment period before finalization, creating the regulatory benchmark for state DOI rate review proceedings

What Carriers Were Required to Submit

Florida Insurance Commissioner Mike Yaworsky, who chairs the NAIC's Homeowners Market Data Call Task Force, announced the data call at the NAIC's Spring 2026 National Meeting on March 26. All 50 states and territories agreed to participate, a unanimity that distinguishes this collection from prior efforts where coverage gaps let carriers argue that cross-state comparisons were not meaningful. The applicability threshold was $50,000 in homeowners premium in any participating state in any year from 2018 to 2025, a figure low enough to capture regional writers and specialty carriers as well as the national standard markets.

The data elements run eight categories deep at ZIP code granularity: policy type, covering home, renter, condo, and mobile home forms; annual premiums written and earned; peril-specific claims counts and losses; deductible structures by type; cancellations and non-renewals with cause coding; coverage limits by tier; replacement cost versus actual cash value elections; and mitigation discounts applied at the policy level. The combination of peril-level loss data with ZIP code assignment across eight years is the structural difference between this collection and any prior regulatory homeowners dataset. Individual state market surveys have produced some of these data elements at the state level; the NAIC data call produces all of them at ZIP code resolution across every state simultaneously.

The initial submission deadline was June 15, 2026, with a one-month extension to July 15 granted in recognition of the expanded scope. NAIC staff were explicit that the extension was a one-time accommodation, not a precedent, and that submitted data should reflect carrier quality-assurance review rather than preliminary estimates. The data call letter identified the NAIC's online regulatory data collection portal as the submission mechanism, with validation rules posted in advance and a carrier webinar series walking through the template. As of the original June 15 deadline, the majority of carriers had filed, with the remainder using the July 15 extension window.

The eight-year lookback was designed to capture two distinct market regimes in the same dataset: the competitive period of 2018 through 2020, when non-renewal rates were low and admitted markets were broadly available in most geographies; and the crisis period of 2021 through 2025, when catastrophe losses, reinsurance cost increases, and replacement cost inflation drove the largest homeowners non-renewal cycle in the modern record. Having both regimes in the same ZIP code-level dataset is what makes the 2027 report analytically tractable for regulatory purposes. A regulator reviewing a carrier's 2024 non-renewal rate in coastal Mississippi can compare it against that same carrier's 2019 non-renewal rate in the same ZIP code, against the all-industry 2024 non-renewal rate in that ZIP code, and against the all-industry 2019 baseline. That four-cell comparison is new to state-level rate review proceedings.

The FIO Partnership and What the 2022 Dataset Already Showed

The 2026 data call builds on a coordinated agreement between the NAIC and the U.S. Department of the Treasury's Federal Insurance Office, signed in 2024. Under that arrangement, the NAIC collects ZIP code-level homeowners data on behalf of participating states and shares a subset with FIO, which uses the aggregated data to conduct what it describes as a nationwide assessment of climate-related financial risks to insurance consumers. The NAIC began transmitting data to FIO in June 2024. Treasury published the analysis in January 2025, releasing a large subset of the aggregated ZIP code-level data alongside the report.

The January 2025 FIO findings covered 330-plus insurers on more than 246 million homeowners policies aggregated to the ZIP code level for policy years 2018 through 2022, an annual average of 49.3 million policies per year. Average homeowners premiums grew 8.7% faster than general inflation across the five-year period. Non-renewal rates were materially higher in ZIP codes with elevated expected losses from climate-related perils. The FIO published geographic climate-risk scores keyed to ZIP codes alongside the dataset, creating a public mapping between non-renewal concentration and modeled hazard exposure that was immediately cited in state legislative proceedings in California, Florida, and Louisiana.

The insurance industry pushed back on the January 2025 report, arguing that the FIO analysis overweighted climate factors relative to inflation, litigation costs, construction labor shortages, and population migration patterns. The critique has technical merit. A cross-sectional analysis correlating non-renewal rates with climate-risk scores does not isolate climate exposure from the social inflation and replacement cost pressures that concentrated in the same coastal and wildfire-exposed ZIP codes over the same period. High non-renewal rates in South Florida coastal ZIP codes in 2022 reflect inadequate rate levels under Florida's prior ratemaking environment as much as they reflect hazard exposure, and the FIO report's methodology does not fully untangle the two.

The 2026 data call is designed to be thick enough to support multivariate analysis. Eight years of peril-specific loss data at ZIP code resolution gives analysts the ability to separate wind-driven loss experience from hail, flood, and fire loss at a spatial granularity that no prior dataset supported. Whether NAIC staff and state DOI analysts use that analytical capability for nuanced peril attribution or for straightforward non-renewal mapping will be the first signal worth watching when the 2027 report appears.

Three Missing Years: 2023-2025 and the Non-Renewal Surge

The FIO dataset's 2022 cutoff is where the policy gap is sharpest. The non-renewal and market-exit events that define the current homeowners crisis accelerated significantly after 2022. State Farm announced in May 2023 that it would stop writing new California homeowners policies, citing catastrophe exposure and reinsurance costs. Allstate had made the same decision earlier that year. The FAIR Plan's California exposure grew from approximately 50,000 surplus lines homeowners policies in 2023 to more than 320,000 by the end of 2025, a 540% increase driven almost entirely by admitted market withdrawals. Louisiana's non-renewal acceleration, already the steepest five-year rate in the FIO dataset, continued through 2023 and 2024 as Hurricane Ida loss development ran above initial reserve estimates and reinsurance attachment points reset.

Florida's Citizens Property Insurance reached 1.4 million policyholders, making it the largest homeowners insurer in the state and an explicit measure of the admitted market's retreat from coastal exposure. The Florida legislature's 2022 and 2023 tort reform packages were designed to reverse that pattern, and the depopulation program that extracted 184,000 Citizens policies in May 2026 alone suggests the reform cycle is working. But the non-renewal data from 2021 through 2024 will show the full magnitude of the retreat before the recovery. That record is now inside the NAIC dataset.

Policy year 2023 is particularly important for the 2027 report's analytical structure. It was the year market exits shifted from incremental to explicit, the year Citizens reached its policyholders peak, and the year when the gap between admitted and surplus lines pricing in coastal Florida and California wildfire-urban interface geographies became wide enough to be visible in premium data at the ZIP code level. Adding 2023, 2024, and 2025 to the dataset creates a continuous ZIP code-level record spanning the run-up to the crisis, the peak of the non-renewal surge, and the early recovery phase. A carrier whose non-renewal rate in a given ZIP code was elevated in 2022 and has returned to baseline by 2025 tells a different story than a carrier whose rate stayed elevated through the full period. Regulators will have both curves, in the same ZIP code, across the full carrier market.

Building Anticipatory Models Before the Report Publishes

The NAIC dataset is not public and will not be before the 2027 report. Carriers and pricing actuaries who want to model what the report is likely to surface do not have to wait passively; the combination of currently available public datasets is richer than most pricing teams are using systematically.

The FIO-NAIC 2022 dataset is the starting point. Treasury released a large subset of the aggregated ZIP code-level data alongside the January 2025 report, and the file is directly accessible from the Treasury website. It includes ZIP code-level non-renewal rates, average premiums, and claim frequencies by peril for 2018 through 2022. A carrier that loads this dataset and matches it against its own historical experience in the same ZIP codes can identify geographies where its non-renewal rate deviated materially from the all-industry average, and in which direction. A carrier that non-renewed substantially more than the industry average in a given ZIP code has an actuarial story to tell about why; a carrier that non-renewed substantially less may have a different exposure profile or pricing advantage in that geography, which is also information worth understanding before regulators start asking. Building ZIP code-level homeowners exposure models against this public benchmark is straightforward with the FIO data in hand, and the combination of the public non-renewal rates with peril-specific loss cost factors creates a precision instrument that the 2027 report will validate or complicate.

State rate filing loss cost factors are the second data layer. NAIC annual statement data and state-specific market share filings are publicly available through the NAIC's market data tools and individual state DOI websites. A carrier that knows its market share and written premium in each ZIP code can construct an approximation of what its submitted data will look like relative to the all-industry aggregate the NAIC report will publish. The geographies where the approximation shows material deviation from the expected industry benchmark are the geographies worth examining actuarially before a regulator raises them in a rate proceeding.

The FIO climate-risk scores published alongside the 2022 dataset offer a third analytical layer. Treasury keyed the hazard scores to ZIP codes and organized them by peril. Overlaying a carrier's current book of business against the FIO risk scores identifies the ZIP codes where exposure concentrations are highest relative to observed non-renewal patterns in the public dataset. That overlay is not predictive in the modeling sense, but it identifies the geographies where the 2027 report is most likely to generate regulatory attention, which is a different and more actionable question for a pricing actuary preparing rate support files for the next cycle.

NAIC market conduct examination data is the fourth source, and the one most commonly overlooked in anticipatory analysis. State market conduct reports document carriers' non-renewal rationale coding, cancellation notice practices, and mitigation discount application rates at the company level. A carrier whose non-renewal cause coding in high-hazard ZIP codes defaults to broad underwriting judgment language rather than specific peril-related rationale is creating a record that a regulator reviewing the 2027 non-renewal maps may choose to connect to the ZIP code-level patterns. Carriers that have not recently audited their own non-renewal cause coding for consistency with their actuarial loss-cost analysis of high-hazard geographies have work to do before the report establishes a public benchmark they cannot control.

What the 2027 Report Will Almost Certainly Show

NAIC staff have described the planned report as organized around four analytical dimensions: how coverage options and deductibles affect cost and access; whether mitigation discounts are producing measurable loss reductions; insurer financial health monitoring at the state level; and consumer awareness of insurance options in high-risk markets. Each dimension produces a specific finding type that actuaries should anticipate.

The geographic non-renewal map will be the report's most politically visible output. ZIP code-level non-renewal rates from 2018 to 2025 across all 50 states will give state DOIs, legislators, and consumer groups the ability to publish a precise picture of where carriers concentrated their market exits. The FIO 2022 data already identified Florida and California as national leaders and Louisiana as the fastest-growing non-renewal market; the 2025-vintage extension will show whether the acceleration continued, plateaued, or reversed in each state, and will do so at a ZIP code resolution fine enough to connect specific carrier non-renewal patterns to specific communities. That connection is what legislators and advocacy groups have been waiting for. The data will be in the public record.

The peril-specific loss trend analysis will be the most actuarially consequential output. Eight years of loss data broken out by wind, hail, fire, flood, and other perils at ZIP code level gives regulators a dataset that no individual carrier has for the full market. Carriers know their own loss experience in a given ZIP code. Regulators will know the all-industry aggregate loss experience in that ZIP code, by peril, across eight policy years. The comparison becomes a reference point in rate hearings. A carrier with loss experience materially above the industry aggregate in a high-hazard ZIP code without a clear actuarial explanation for the divergence is in a weaker position in a rate proceeding than it was before the benchmark existed.

The mitigation discount analysis will be the most technically novel finding. Requiring carriers to report the actual mitigation discounts applied at the policy level is information regulators have not had in aggregate form before. The 2027 report will be the first time anyone can examine whether mitigation discounts, applied across the full carrier market, are associated with lower observed loss frequencies in the ZIP codes where they are most heavily used. Florida's wind mitigation credit program, which has been active since 2002 and generates roughly $1.5 billion in annual premium credits, is the natural test case. If ZIP codes with high wind mitigation credit penetration show meaningfully lower wind loss frequency per policy than comparable ZIP codes without comparable credit penetration, that correlation strengthens the actuarial support for the credits. If it does not, the conversation becomes considerably more complicated at the next Florida DOI rate hearing.

The replacement cost versus actual cash value breakdown will show the geographic pattern of coverage form elections across the carrier market. In high non-renewal markets, the documented carrier strategy of moving policyholders from RC to ACV to reduce claims exposure while lowering premium appears in aggregate data as a shift in the coverage form mix within a ZIP code. The 2027 report will identify ZIP codes where that shift was concentrated, enabling regulators to examine whether coverage contraction was a systematic response to high-hazard exposure or a selective application in specific communities.

How the 2027 Report Reshapes State DOI Rate Reviews

The actuarial consequence of the 2027 report is not the report itself. It is what happens to the evidentiary standard in state DOI rate review proceedings once regulators have a public ZIP code-level benchmark for the full carrier market.

Before the data call, a carrier appearing in a state rate hearing brought its own data: its own loss experience, its own non-renewal history, its own peril-level loss costs. Regulators had general market share filings and actuarial publications but not the ZIP code-level all-industry dataset that would let them compare a specific carrier's non-renewal pattern against the market aggregate in a specific ZIP code in a specific year. After the 2027 report, that comparison is available, and it is the regulator who will bring it to the proceeding.

A carrier whose non-renewal rate in coastal Louisiana ZIP codes ran materially above the all-industry average during 2022-2024 faces a new category of question. The carrier may have had sound actuarial reasons: inadequate rate levels confirmed by subsequent loss development, concentrations of aging housing stock in FEMA-designated flood zones, or class-of-business mix that skewed toward the highest-hazard coverage tier in those geographies. But the regulator reviewing that carrier's next rate filing will have the all-industry comparison available, and the carrier's actuary will need a disaggregated geographically grounded explanation for why its non-renewal pattern diverged from the market. That explanation should already exist in the carrier's actuarial work product. The 2027 report makes it necessary to have that documentation in the rate filing support file rather than in an internal actuarial memorandum that a regulator would have to request.

Florida, California, Louisiana, and Colorado are among the states most likely to deploy the 2027 dataset early in rate proceedings. Florida's DOI has been tracking Citizens depopulation and admitted market non-renewal patterns as part of its post-2022 legislative reform framework. California's CDI is managing the Sustainable Insurance Strategy rate filing pipeline simultaneously with the NAIC data submission process, and CDI examiners will have both the 2027 all-industry non-renewal map and the first three years of SIS model-based rate filings to cross-reference. Louisiana's regulator has watched the steepest non-renewal acceleration in the country through the FIO five-year dataset period; the 2025-vintage extension will confirm whether that acceleration continued or whether the market correction that higher rates should eventually produce has begun. Colorado has been developing its own cat-model ratemaking framework for wildfire, and the NAIC non-renewal map for Colorado ZIP codes will inform that effort directly.

What to Watch When the Report Drops

Three methodological choices will determine how the 2027 report enters the policy conversation and how durable its findings prove once carriers and their actuaries engage in the comment process.

The first is whether the NAIC defines "inadequate availability" with a hazard-normalized benchmark or as a raw non-renewal rate. A 3.35% non-renewal rate in coastal Broward County is not the same actuarial event as a 3.35% non-renewal rate in suburban Kansas City. If the report presents non-renewal rates without adjustment for underlying hazard exposure, the geographic map will generate political pressure on carriers in high-risk markets regardless of whether their withdrawal decisions were actuarially consistent with hazard levels and rate adequacy. The report that does this analysis well will publish non-renewal rates alongside a modeled loss cost benchmark, allowing readers to separate carriers that were acting rationally given their rate inadequacy from carriers that were withdrawing selectively beyond what the actuarial evidence supports. The 2022 FIO report did not fully achieve this normalization; the 2027 NAIC report has the more complete dataset to do it.

The second is whether the peril-specific loss cost figures become the new de facto actuarial standard in state rate hearings. If the NAIC publishes ZIP code-level wind, hail, fire, and flood loss cost factors derived from the full carrier dataset, those figures will function as reference points in rate proceedings whether or not the NAIC intends them to. State DOIs in prior-approval states will cite them in rate hearings. Carriers whose indicated peril-level loss costs in a given ZIP code deviate materially from the all-industry figure will need actuarial documentation of why their experience justifies the deviation, which is standard credibility analysis but is new work in the specific context of an all-industry ZIP code benchmark. Actuaries who have not yet thought through how their companies' peril-split methodologies would compare against such a benchmark should do that analysis now, before the question arrives in a rate hearing with a short response deadline.

The third is whether non-renewal patterns become the trigger for state market conduct reviews rather than merely legislative attention. The NAIC report will be a public document. Consumer advocacy organizations, state attorneys general, and legislative staff with the public non-renewal maps and access to state market share filings can construct a reasonably detailed picture of which carriers withdrew from which communities. If those patterns correlate with demographic or income variables in a way the carrier cannot explain with actuarial data, the market conduct examination risk becomes real independent of whether the carrier's rate filings were otherwise in order. The carriers best positioned when the 2027 report publishes are those that have already constructed a geographically disaggregated actuarial account of their non-renewal decisions, not because they expect to defend it in a market conduct proceeding, but because having it means they can.

The NAIC has committed to a public comment period before the final 2027 report is released. That process is the actuarial profession's clearest leverage point for shaping how the report's metrics are defined and presented. Comments that engage the measurement methodology, propose hazard-normalization approaches, or document the analytical distinction between actuarially justified non-renewal and selective community withdrawal will be in the public record and will shape how state DOIs interpret the report's findings in subsequent proceedings. Carriers and their actuaries who wait until the report publishes to engage its methodology will find the interpretive frame already established.