Based On The Garch Model Var Calculation

With the development of economy-globalization and finance-integration, the global financial market has changed basically and structurally; therefore many banks and investment institutions have attached importance to measure and predict financial risk. The method of Value-at-Risk (VaR) based on statistics is known as the mainstream in this field nowadays. The key step of computing the VaR is to estimate the fluctuation of prices. In the article we discussed the methods of estimate the parameters of GARCH model that is the mainstream method of estimating the fluctuation of prices. Aiming at the BHHH algorithm which estimates the parameter of GARCH model, we adopt the PHR algorithm and simulated annealing algorithm to improve accuracy of the parameters of GARCH model ,and the result are applied to calculating the Value-at-Risk. The computing examples of Microsoft Corp stock prices are presented , and the computation results indicate that VaR-GARCH model based on simulated annealing algorithm outperformed the conventional numerical method on the aspects of computational accuracy of VaR.