Spatial Statistical Modeling of Insurance Risk An epidemiologist approach to car insurance

Spatial models, such as the Besag, York and Mollie (BYM)model, have long been used in epidemiology and disease mapping. A commonproblem in these subjects is the modelling of number of disease eventsper region; here the BYM models provides a holistic framework for bothcovariates and dependencies between regions.We use these tools to assess the relative insurance risk associated withthe policyholders geographical location. The models are placed in a Bayesianframework, and inference is made using Integrated Nested Laplace Approximation(INLA).The model is applied to car insurance data from If P&C Insurance togetherwith spatial referenced covariate data of high resolution, provided byInsightone. Including spatially dependence in the modelling of number ofclaims signicantly improves on the result obtained using ordinary generalisedlinear models. However, the support for adding a spatial componentto the model for claims cost is weaker