NAIC Homeowners Data Call Sets Nationwide Peril Baseline for the First Time

The NAIC sent letters to every homeowners carrier in the country on March 25, 2026, opening the Homeowners Market Data Call. Any carrier writing at least $50,000 in direct written premium in a participating jurisdiction has until June 15, 2026, with no extensions, to file 113 fields covering 2018 through 2025 at the zip-code level. Fifty jurisdictions are participating. The change that matters is one column: wildfire, reported separately from fire.

Key Takeaways

• 113 fields across four parts, filed at zip-code level by policy form, by year and separately for new and renewed business, on an individual company basis rather than at group level.
• Fifty participating jurisdictions, being 48 states plus the District of Columbia and Puerto Rico. Alabama and Tennessee are the only absences.
• Wildfire is broken out from fire for the first time on a standardized basis. On the Annual Statement both collapse into a single fire loss figure.
• $88,170 average claim severity for fire and lightning against $14,747 for wind and hail in 2023, while wind and hail was 42.5% of national losses and fire and lightning 21.6%.
• An 82-day window with no extensions, against eight years of history that most carriers do not hold in the reported format.

What Is Being Demanded, and From Whom

The Homeowners Market Data Call (C) Task Force finalized requirements at the March 24, 2026 Spring National Meeting session and the letters went out the next day. This is not a survey; it runs on each participating jurisdiction's statutory examination authority.

Section	Fields	Key Data Elements
Part I: Premium, Coverage, Deductibles	12-70	Written premium in-force, policies in force, Coverage A/B/C/D limits, wind/wildfire/earthquake exclusion counts, RC vs. ACV breakdowns, deductible distributions by peril
Part II: Claims and Losses	71-82	Paid claims and losses by peril: fire (excluding wildfire), wind/hail, water damage/freezing, wildfire, all other
Part III: Cancellations and Non-Renewals	83-90	Non-payment cancellations, company-initiated cancellations by timing window, non-renewal counts, premium for cancelled policies
Part IV: Mitigation Discounts	91-112	State-required and voluntary discounts for fortified standards, wind, wildfire, impact/hail, and water mitigation

The reporting grain is the burden. Data comes at zip-code level, by policy form across DP-1 through DP-3 and HO-1 through HO-8, by year for 2018 through 2025, and split between new and renewed business. Each response is filed on an individual company basis, not consolidated to the group. A carrier writing in all fifty jurisdictions across eight years, several forms and thousands of zip codes produces a very large submission.

The $50,000 threshold is deliberately low. It triggers on homeowners line 4 of the Annual Statement in any year from 2018 through 2025, or on dwelling and homeowners policies written outside line 4, which pulls in small domestic carriers and surplus lines writers that targeted requests have always missed.

The Peril Split Is What Makes a Loss Ratio Computable

Part II requires paid claim counts and losses paid decomposed into five perils: fire excluding wildfire, wind and hail, water damage and freezing, wildfire, and all other. Peril counts must sum exactly to total claims and peril losses exactly to total losses.

Separating wildfire from fire is the substantive change. The Annual Statement collapses a kitchen fire and a wildfire into one number. For California, Colorado, Oregon and New Mexico this produces the first clean wildfire-specific loss data a regulator can hold against a rate filing across every jurisdiction at once.

The reason it matters is that the frequency mix and the severity mix point in different directions. In 2023, wind and hail was 42.5% of national homeowners losses, water damage and freezing 22.6%, and fire and lightning 21.6%. But fire and lightning carried $88,170 in average claim severity against $14,747 for wind and hail, a factor of six. A peril distribution read off claim counts describes a different book than one read off severity, and until now no standardized dataset let a regulator hold both at once.

The availability side gets the same treatment. Part III separates company-initiated cancellations into windows of the first 59 days, 60 to 90 days, and beyond 90 days, which distinguishes post-binding underwriting from mid-term reassessment driven by a model update or an inspection, and counts non-renewals separately. Against residual markets that now hold 1.34 million Citizens policies at $4.6 billion in premium, 431,300 California FAIR Plan policies at $1.42 billion and 161,647 Louisiana FAIR Plan policies at $518 million, that will map carrier retreat to zip codes rather than to press reports.

One convention limits what the recent years can carry. Claims are reported in the year they closed, on a paid basis net of salvage and subrogation, excluding case reserves, unpaid amounts and loss adjustment expenses. For a long-developing peril that is a real understatement: the 2024 and 2025 wildfire figures will sit below ultimate, and a loss ratio built off them without a development adjustment reads better than the exposure is.

Premium Can Hold Flat While the Product Shrinks

Part I carries the fields that make the premium series interpretable, and they are not premium fields.

Carriers must report policies in force with replacement cost coverage on dwelling separately from actual cash value coverage, with the same split applied to roof and to siding. That tracks a shift that has run since 2022: ACV roof endorsements that depreciate roof value by age, moving the replacement gap to the policyholder. If 30% of policies in a state now carry ACV roofs against 10% five years earlier, coverage has been withdrawn without any rate filing recording it.

The deductible fields do the same work. For all perils, hurricane or named storm, and wind and hail, carriers report policies across six buckets from $500 or lower up through 5% and above, with a separate seven-bucket distribution for earthquake. A percentage deductible is a different instrument from a flat one: a percentage deductible on a $500,000 home can translate to $10,000 or more, and a state where 40% of policies sit above 5% has a materially different consumer profile from one where most carry $1,000 flat.

Put those two series against the premium series and the interpretation changes. A state can show stable average premium across eight years while replacement cost migrates to actual cash value and flat deductibles migrate to percentages. Rate adequacy measured on premium alone would read that book as unchanged. The HMDC is the first collection that makes the offsetting movement visible, which is also why the first submission is the hardest.

That is the near-term constraint. Eighty-two days, no extensions, eight years of history at zip-code and form grain, reported per company rather than per group, in a format most carriers do not currently hold. The dataset will be the baseline for peril-level regulation for years. The first cut of it will carry whatever each carrier's reconstruction of 2018 looked like in the spring of 2026.