This final project is about computing market risk (in this case is stock risk) VaR using volatility measured by ARCH/GARCH model. The ARCH/GARCH model are used due to the data of stock market return which show a relatively heterocedasticity nature. In this research, GARCH (1,1) model is the chosen model to forecast the volatility. The result from this research shows that, practically, ARCH/GARCH model is a valid model to measure the volatility that will be used in calculating Value at Risk. Almost GARCH model in measure the Value at Risk is the appropriate in accordance to Kupiec Criteria for 99% and 95% of confidence level.
Key words:
VaR, ARCH/GARCH