7 min read
The data an underwriter actually needs on a property risk
A working inventory of the datasets that inform a homeowners or small commercial property decision, what each one answers, who publishes it, and where the gaps are.
The short answer
A property risk decision draws on five distinct families of data: what the structure is, what it is worth, what has been done to it, what natural hazards it sits in, and how quickly someone can respond when something goes wrong. No single vendor covers all five, which is why the work is usually done across a dozen browser tabs.
Structure and parcel
Year built, construction type, roof cover, living area, lot size and the presence of a pool or attached garage drive both the exposure and most of the eligibility rules in a homeowners program. This data originates with county assessors and is aggregated commercially by providers such as ATTOM, CoreLogic and Black Knight.
The failure mode is staleness rather than absence. An assessor record can lag a renovation by a full assessment cycle, which is why permit history matters as a corroborating signal.
Valuation and transaction history
An automated valuation, the last sale price and date, and the assessed value and tax amount give a reconstruction-cost sanity check and flag properties where the requested coverage is out of line with anything the market has paid.
Permit history
Permits are the closest thing to a maintenance log. A re-roof permit five years ago answers the roof-age question that drives so much of the loss picture in wind and hail states. An absent permit history is not proof that no work was done, but a present one is strong evidence of what was.
Natural hazard layers
Flood comes from FEMA's National Flood Hazard Layer, which publishes the zone, the Special Flood Hazard Area flag, base flood elevation and the effective date of the governing map panel. Wildfire hazard potential is published by the USDA Forest Service as the WHP raster, graded one to five, alongside annual burn probability.
Both are public datasets, and both are queryable per coordinate. The reason underwriters still do this by hand is not access, it is that each one lives behind its own map viewer with its own interface.
Response and protection
Distance and drive time to the nearest responding fire station is a genuine input to loss severity. It is worth being precise about what it is not: an ISO Public Protection Classification is graded by Verisk on department staffing, water supply and communications, and cannot be derived from distance alone. Treating a distance band as a PPC is a common and costly shortcut.
Where the gaps are
Crime and neighbourhood safety data is the hardest of the set to source consistently. Municipal open-data portals cover a minority of the country, the FBI's aggregate data lags by a year or more, and the commercial grading products are proprietary and expensive. Any tool that shows you a neighbourhood crime grade should be able to tell you exactly which feed produced it.