Abstract Insurance risk arising from natural catastrophes such as earthquakes is a key component of the minimum capital test for federally regulated property and casualty insurance companies. This paper proposes an integrated, open-source, simulation-based actuarial framework for the assessment of earthquake insurance risk and solvency capital requirements. The framework combines spatio-temporal earthquake occurrence modeling, physics-informed ground-shaking estimation based on Canadian seismic hazard maps, building exposure and vulnerability modeling, and detailed insurance loss and claim calculations within a unified pipeline. Spatial heterogeneity in seismic risk is captured through kernel-based spatio-temporal point process modeling, while Voronoi-based deviance residuals are employed as localized diagnostic tools to validate model adequacy. Simulated insured losses are used to estimate regional and country-wide probable maximum losses (PMLs), and a new capital aggregation formula is proposed that explicitly incorporates cross-provincial dependence in earthquake losses, in contrast to the current region-based regulatory aggregation. The proposed framework enables spatially resolved loss and capital assessment at a fine geographic scale and is implemented in a fully reproducible open-source environment. An interactive web application is also provided to allow users to simulate earthquake damage and the resulting financial losses and insurance claims at user-specified epicenter locations.
Bairakdar et al. (Thu,) studied this question.