APLIKASI MODEL KREDIBILITAS HIRARKI TIGA LEVEL PADA INDUSTRI ASURANSI UMUM

In statistical inference, credibility theory achieves the compromise between the notions of stability, precision, and responsiveness to the most recent events. Credibility theory studies how to calculate premium based on past experiences and combine between individual experiences and collective experiences. In practice, we often find hierarchical structure in many line business, such as in the industrial, business, insurance, etc. Hierarchical credibility model is seen as a tool to distribute premium fairly among a heterogeneous portfolio classified in a hierarchical structure. Fundamental idea from this theory is divide a given population into homogeneous subportfolio based on certain citeria. This thesis will analyzed the way to get hierarchical credibility estimator and structural parameter that involve mathematical properties of conditional expectations and of conditional covariances. Case study in this thesis analyze loss ratio credibility estimation among 12 countries that joined legal protection insurance RIAD that classified based on World Bank�s income group