This thesis examines the application of credibility theory in determining insurance premiums, with a particular focus on comparing hierarchical and non-hierarchical models. The importance of this research lies in its focus on the potential of using hierarchical models, rather than common non-hierarchical models, to enhance the accuracy of premium estimates and achieve lower error rates. The thesis employs a blend of theoretical exploration and empirical analysis, drawing from probability theory, Bayesian statistics, and linear estimators to establish a robust framework. Key findings from a case study using data provided by Allianz Versicherungs-AG indicate that hierarchical models, despite their complexity, produce the lowest error rates, outperforming both general and Poisson-based non-hierarchical models. The research concludes that the hierarchical method is justified for this dataset, o↵ering a superior alternative to simpler models and paving the way for further exploration in diverse insurance branches.     «

This thesis examines the application of credibility theory in determining insurance premiums, with a particular focus on comparing hierarchical and non-hierarchical models. The importance of this research lies in its focus on the potential of using hierarchical models, rather than common non-hierarchical models, to enhance the accuracy of premium estimates and achieve lower error rates. The thesis employs a blend of theoretical exploration and empirical analysis, drawing from prob...     »