Zpětné testování modelů pro odhad rizika na bázi Value at Risk

Abstract

This thesis is focused on market risk of commercial banks. In the theoretical part, the VaR methodology is presented with emphasis on three basic methods of VaR calculation: historical simulation, variance-covariance method and Monte Carlo simulation, which is used in the practical part. Another section is devoted to individual methods of backtesting, such as Kupiec unconditional test, Christofferson test or mixed Kupiec test. In the final part the theoretical knowledge are applied in order to analyze VaR for selected financial instruments and verify the estimates using selected methods. The primary aim of this thesis is the backtesting model for estimating risk, VaR and comparing the two distributions, normal and variance gamma distributions.