Project contact is Étienne Marceau at Université Laval
Building insurance accounts for a significant portion of income for general insurance companies. Actuaries in insurance companies use predictive models based on history for pricing and underwriting purposes. As explained in Frees et al. (2011), a home insurance contract usually includes protections for several "perils": fire, wind, hail, theft and vandalism, water and weather, non-wetted water, lightning, responsibility, and other. Predictive models based on explanatory variables are used to model the losses associated with a home insurance contract. Two families of approaches can be used to model the losses of a home insurance contract. One approach ignores perils and models total losses, while in the other approach perils are taken into account. In this project, we aim to analyze the losses for eco-responsible wood constructions and to compare the results obtained for other types of construction.