An Empirical Study On Time Variability Of Beta In Shanghai Stock Market

In the theory and practice of capital market, the measurement of investment risk has always been the focus of theorists and practitioners, and Beta coefficient which is an important parameter in CAPM, is a widely used risk measurement indicators. With the gradual deepening of globalization and financial integration, China's stock markets are gradually integrated into the global financial system. Factors that affect the stock market will be more complex, and investors will face greater systemic risk. Moreover, the Chinese stock market is an emerging and immature market, so it has its own characteristics. Therefore, the study of systematic risk and its characteristics in China's stock market, especially the prediction of risk is of great significance for investors and regulators to implement risk prevention and control.First, a large number of domestic and foreign research results have shown that coefficient is unstable. Therefore, the thesis starts with the theoretical definition of Beta coefficient, and use dynamic analysis method—DCC-MVGARCH model to analyze the conditional variance and dynamic correlation coefficient, and then estimate the time-varying Beta coefficient. Studies show that Beta coefficient is very unstable and fluctuating. Further studies show that the systematic risk of Shanghai stock market is high, but the systemic risk level of listed companies with larger share is lower than that of listed companies with smaller share.Secondly, in this paper, we use classic R/S analysis to study the long-term characteristics of the time-varying Beta coefficient for the first time. Studies found Beta coefficient series has long memory characteristic.Finally, we estabish the ARIMA and ARFIMA forecasting model, and the results show that ARFIMA model can describe the beta coefficient more better and have more precise predictive value.