Study On Risk Premiums Of Correlation And Idiosyncratic Volatility Using DCC

In the field of CAPM, except the well-known three factors in Fama-French framework, researchers are always searching other systematic factors to reinforce the explaining ability of CAPM theory. Recently, some scholars began to pay attention to the correlation risk between individual stocks and the idiosyncratic volatilities of individual stocks. However, few researchers examined both two factors' influence to cross-section returns in one pricing model. Moreover, considering the time varying attribution of the second moments of stock return, how to measure these two factors accurately is also a meaningful problem.This article uses data from Shanghai A type stock market and originally measures the conditional expectation of correlation risk and idiosyncratic volatility by DCC-MV GARCH model. At the same time, considering these two factors correlated with other factors that influence market return, there may be co-linearity in our regression equations if we ignore this phenomena. Thus, we select some variables, such as macro-economic variables and variables about market index volatility and eliminate these variables' effect from the correlation risk between individual stocks and the idiosyncratic volatilities of individual stocks. Then we use innovations of them to construct pricing models.This article adopts two-stage regression and Shanken adjustment to estimate the prices of correlation and idiosyncrasy and draws two main conclusions: (1) Correlation does not lead to a statistically significant risk premium. (2) Idiosyncrasy has a negative and statistically significant risk premium, which should be considered as an anomaly pricing phenomena.At last, we infer two reasons of these phenomena by referring to some relevant studies: (1) That there is no adequate diversification action in China stock market can explain that correlation has no risk premium; (2) Idiosyncrasy can reflect the divergence of investors' expectation to a company's stock price. In China, the information discourse of listed companies is not adequate. Thus, it is possible for investors to have large divergence on a company's stock price. So far there is no mechanic for short trade in type A stock market. Thus the price cannot sufficiently include the view point of pessimistic investors. This could lead to an overestimated price and thus a lower return. Also, this article gives some pertinent policy advice to administration of China stock market to solve the two problems mentioned above.