Empirical Study Of Industry Portfolio's Time-varying Betas In China Stock Market

In this paper, the constant beta model, GARCH(1,1) model, the SCHWERT and SEGUIN model, an asymmetric beta model are used to estimated time-varying betas of 55 industry portfolio in China stock market over the period January 22th 2002 to October 23th 2009. Then this paper uses these time-varying betas to predict the forecasted excess return of every industry portfolio, and compares the forecasted excess return with the real excess return to get the minimum mean squared error criterion(MSE criterion), and then use the MSE criterion to test the accuracy of the four different model. The results show that none of the four model can be applied to 55 industry portfolio in Chinese stock market. This means that different industry portfolio should be used different models to estimate the mean of betas. On the basis of this result, this paper suggests the institutional investors, especially those Open-end funds who has the limited minimum position, that when the stock market is in a bull market, they should buy more coal, non-ferrous metals, real estate, comprehensive public and finance or other industry portfolio whose mean of time-varying beta is bigger than 1 to get more return; when the stock market is in a bear market, they should buy more road transport, electricity, agriculture, medicine, regular consumption or other industry portfolio whose mean of time-varying beta is smaller than 1 to against a bigger market risk.