Metody pro zrychlení odhadu závazků životních pojišťoven

The thesis deals with methods for estimation of a life insurance company’s liabilities and it aims to find a faster alternative to the traditionally used cash flow analysis. It studies two approximation methods, cluster analysis and stratified random sampling. The thesis is structured into four parts. First, it briefly characterizes life insurance, presents common types of life insurance contracts and shows a typical structure of a life insurance company’s liabilities. Then, it describes the process of liabilities estimation using cash flow analysis and characterizes the properties of cluster analysis and stratified random sampling. Next, all the studied methods are applied to a modified portfolio of homogeneous life insurance policies and to a real portfolio of heterogeneous policies. Finally, the thesis compares the methods and defines the contexts in which an insurance company would prefer each of the tested methods.