China's Securities Market Risk Research

The people mainly pay their attention about the stock market to two big cores questions - risk and the income, studying how to dodge the risk but to cause the income maximization. Our country's stock market is a young market, is being in the explorative and gradually standard stage. In domestic research, we still primarily introduce and draw lessons from overseas risk management theory, then, under the special environment of our country's stock market, how use these theories and models to measure risks effectively, as well as seeking risk management theory that adapts our country stock market become the most important question.Markowitz proposed to weigh the risk with the profit's standard deviation, namely the Mean- Standard Deviation Model, his Investment Portfolio Theory further proposed that through carries on the suitable combination to the securities, may effectively reduce the risk. Through my empirical investigation on our country's stock market, I discovery that along with the portfolios' scale increasing, the standard deviation of portfolios' profit rapidly reduces, but when the portfolios' scale achieved 6-12 stocks, the standard deviation of portfolios' profit tends to stably, therefore the suitable investment scale on our country's stock market is between 6-12. Moreover, the selections of Markowitz Portfolio, as well as the effectiveness of Mean- Standard Deviation Model were made simple discussion.The target to weight system risk is β coefficient, and the Market Model, the Capital Asset Pricing Model and the Arbitrage Pricing Theory have established the relationship between β coefficient and securities profit from different aspect. According to weak efficacious characteristic of our country's stock market, I carried on the estimate to 239 stocks β coefficient of 2000 - 2004 year with the Market Model, described basic characteristic of the stocks β coefficient, and conducted the comparison research to β coefficient forecast model. Those research discovery that because the correlation between two continuous years' β coefficient was small, the Historical β Method and the Vasicek Adjusted Model all appeared big deviation, but Blume Revised Model was most effective to forecast stock β coefficient.