FLAMES: AAIS Data-Driven Fire Risk Model Sparks Approval
Fire Loss and Mitigation Evaluation Score (FLAMES) helps insurers predict expected fire severity more accurately AAIS (American Association of Insurance Services) has received approval from the Florida Office of Insurance Regulation for its new predictive model. The AAIS Fire Loss and Mitigation Evaluation Score (FLAMES) is a significant improvement on existing methods of quantifying expected severity of fire losses.
Current methods of measuring fire risk reflect fire protection capabilities within the geographic boundaries of individual fire departments, but ignore variation in capabilities and claims severity across communities. Using fire department boundaries also makes it difficult for insurers to integrate risk scores into underwriting systems and processes.
“Our Member carriers have expressed frustration with existing methods of assessing fire risk, which haven’t kept pace with changes in public fire protection capabilities and modern mitigation approaches,” said AAIS CEO Ed Kelly. “Our Data Engineering and Actuarial Team took a rigorous, analytical approach to solving this decades-old problem, and now insurance carriers and regulators can have a more accurate picture of fire risk.”
Several key principles guided the development of the FLAMES model, including the need for a more data-driven, analytical solution that is acceptable to insurance carriers, regulators, and other industry professionals. AAIS evaluated a wide range of data including National Fire Incident Reporting System (NFIRS) data, carrier claims data, weather, crime, economic activity, fire station locations, fire hydrant coverage, traffic patterns, population density, building stock characteristics, and vacancy rates, which best reflected the drivers of loss.
Current approaches prioritized the location of insured properties relative to fire protection resources and infrastructure. According to Matt HindsAldrich, PhD, AAIS Senior Risk Strategy Lead, “Legacy models have been overly simplistic given advancements in local fire protection. Rather than a checklist of input measures to grade individual local fire department performance, FLAMES measures outcomes in terms of severity of fire losses by ZIP code.
