Empirical Research On β Coefficient Prediction Method Of Stock Market Of China

William Sharpe put forward the famous capital asset pricing model in 1964.CAPM andβcoefficient has always been an asset or a meas ure of Systemic risk of the portfolio, It is widely used not only in theoretical study but also in the investment practice. Because of this, Financial theory study on theβcoefficients have been the hot spots from the time whenβcoefficients generated. Throughout the research results at home and abroad, Reserches are spoted in the properties ofβcoefficient (such as stability, diversity, time-varying, etc.) and improvement of the forecasting methods ofβeffectiveness. Single-index model method and CAPM model are two estimate method. And there are two categories ofβcoefficient prediction, they are time-series-based forecasting methods and influencing-factors-based forecasting methods.In this paper, We made an empirical study based on the data from 2002 to 2006 to 229 in Shanghai and Shenzhen Stock Exchange listed company. Single index model as theβcoefficient estimation method, and this method over the yearsβcoefficient estimate its true value as an approximation. Then select the time series prediction method based on the history of law and BloomβAdjustment Act, and construct a two historical data using the prediction method to predict theβcoefficients, and factors based on differences in the methods Rosenbergβcoefficient of the system to predict the final sample data using a variety of forecasting methods and forecast a significant effect of test verification comparison.We can see from the empirical results, historicalβis the the worst prediction methods of several of the methods we selected, and its mean square error (MSE) is the Maximum. The regression coefficient of Bloom adjust is significantly non-zero in the test of significance, So Bloom adjustment is effective. Two previous historical value ofβcoefficient had little effect on the prediction coefficients, two historical prior to the value ofβcoefficient has no effect on the prediction coefficient or a very limited role. From the empirical results this article of the selected financial indicatorsβcoefficient has a strong explanatory power, indicating that I had listed the company's financial indicators of the value ofβcoefficient of the interpretation of a certain role.βcoefficientβ= 1 prediction mean square error is when the lower (slightly higher than the historicalβcoefficient) and even in some years is the lowest among all methods. This shows theβcoefficient is the "1"return, that is,βcoefficient of "becoming one of" the other hand, describes theβcoefficient of stock market instability. Rosenberg predicted China's stock market system, theβcoefficient of mean square error (MSE) larger, indicating that the financial data of listed companies in China there is a certain degree of distortion.