| Abstract [eng] |
This thesis analyzes the forecasting accuracy of Value at Risk (VaR) and Expected Shortfall (ES) using conditional heteroscedasticity models. The theoretical analysis includes risk measures, three innovation distributions – normal, Student’s t and skewed Student’s t – and conditional heteroskedasticity models – GARCH and APARCH. The accuracy of forecasting Value at Risk is assessed using unconditional coverage, independence and conditional coverage tests, while the accuracy of forecasting Expected Shortfall is assessed using a regression-based ES test. |
