Quantitative assessment of multi-year non-life insurance risk

Abstract

This cumulative dissertation contributes to the field of quantitative multi-year risk assessment for non-life insurance obligations. Quantifying non-life insurance risk, i.e. the uncertainty in pricing and reserving resulting from the stochastic nature in occurrence, reporting, and settlement of claims, over a horizon of several years is necessary for modern risk- and value-based business planning and regulatory requirements for insurance companies. Practitioners are challenged to develop adequate and feasible multi-year risk projection models that are consistent with the company’s best estimate reserving practices. The research papers included in this dissertation contribute to the multi-year risk literature by providing closed-form estimators of second-moment risk measures for the distribution-free chain ladder model, extending analytical risk quantification to the case of multivariate reserving models among several loss portfolios, and developing simulation algorithms for joint multi-year risk evaluation on company level based on full predictive distributions.