Empirical Research Of VaR Method Based On GARCH Model

Affected by the globalization and financial innovation, huge changes in the financial market are taking place, how to control the risk has become a focus of every financial institution. VaR is a new risk management instrument after1990s. As a model of risk measurement and control, it is easy to operate. Compared with the traditional risk management,VaR model is more practical and referential significance. At present, VaR technique has become one of the main methods to measure market risk、assess achievement and expose information in the word.The paper, to begin with,introduces basic methods to calculate VaR, then makes a comparison in face of the advantages and disadvantages of them. Finally Shanghai stock index and Shenzhen element index are studied and the profitability of them is tested through normality, stability test and GARCH LM test. The outcome of the tests indicates the earning ratio of those markets presents skew and distributed stably with the effect of ARCH. Hereby, GARCH model is suitable to be applied to calculating VaR of China’s Market. Also, the paper fits GARCH、TARCH、EGARCH model and the GARCH model under the hypothesis normal distribution, t distribution and GED distribution to calcullate VaR of the Shanghai stock index and Shenzhen element index. Kupiec test is adopted to test the accuracy of the model to find good model for each securities market.