The Study Of Risk Prediction Based On Improved GARCH Model

With the development of economic globalization, information technology, financial innovation and a great number of most financial institutions trading portfolio, the financial market is unprecedented volatility and vulnerability, Some financial crisis happened frequently. Market risk and credit risk as the main form of financial risk, making the financial regulatory authorities and financial institutions continue to strengthen market risk and credit risk monitoring. China's securities market as developing markets. Establishment of a suitable risk of stock market risk management, more accurate estimate of financial risk, there will be of great theoretical significance and practical value.Volatility model for researching and development of increasingly complex financial issues provide a solid foundation. With the rapid development of economic globalization, the links between the financial markets are becoming increasingly close, the correlation between financial markets analysis, asset pricing, portfolio strategy, and many issues need to find a suitable financial model to describe the case of multiple variables. If you use an existing multi-GARCH and multivariate SV model, because of more variables, resulting in parameter estimation becomes more hardly. However, if introduce the multivariate Copula theory into the financial model of the case, you can make research easier. We can study the marginal distribution of the variable and the dependencies between the variables separately. Copula theory for us to build a new multivariate financial time series model provides a new function.This paper mainly base of financial markets, domestic and international risk management background, focuses on the characteristics of GARCH model of risk management, application and calculation methods, On the GARCH model of risk management study has been made of two risk management model and empirical study, the one introduce high-frequency financial data into the GARCH model and the other is based on Copula theory. Risk management model based on Copula take full account of the nonlinear relationship of financial markets. Through theorical and empirical research of these two kinds of risk management model, the results show that the two methods more accurate.