Underwriting risk control in non-life insurance via generalized linear models and stochastic programming

Abstract

We focus on rating of non-life insurance contracts. We employ multiplicative models with basic premium levels and specic surcharge coefficients for various levels of selected segmentation criteria (rating factors).

We use generalized linear models to describe the probability distribution of total losses for a contract during one year. We propose stochastic programming problems with reliability type constraints for the surcharge coecients estimation which take into account riskiness of each rate cell, prescribed loss ratio and other business requirements.

We apply the approach to Motor Third Party Liability (MTPL) policies.

Keywords

• non-life insurance
• rate making
• generalized linear models
• stochastic programming
• MTPL