Severe convective storms have become the costliest peril in US property insurance, yet the way most carriers measure and price that risk has not kept pace with how quickly it has grown. The result is a widening distance between the risk a book actually carries and the price set against it. Anvil exists to close that distance, and to show its working while doing so.
We measure risk where it concentrates. Severe storm exposure sits at the county level, not across whole states, so a state-wide average tends to hide as much as it reveals. Our analysis is built county by county, so the figures reflect the specific areas a carrier writes in rather than a regional approximation. Every figure we publish is tested against events the model was not built on, and we do not make a claim we cannot demonstrate.
We work independently. Anvil is not tied to a vendor model or a capacity provider, and the analysis answers to the data rather than to anyone's existing position. That work begins from public severe-storm records, which means our findings can be traced back to their source, checked, and relied upon.
Anvil is the climate-risk arm of Scrape Group, a data house built around finding signal that others overlook. Anvil applies that same discipline to a single problem: severe convective storm risk in US property insurance.