Research On Stock Asset Pricing In China

Asset pricing theory aims to analyze the uncertain payoff of an asset, where portfolio selection theory, CAPM and option pricing theory have won Nobel Prizes in economic sciences. As new winner of Nobel Prize, Fama proposes efficient market hypothesis. However, there exist a lot of anomalies in developed financial market, including calendar effect, historical performance, trading activity and fundamental information. Empirical models such as Fama-French three-factor model and Carhart four-factor model are popular in business community. In contrast to developed stock market, the features of institutional background and investors are quite different, and classic asset pricing models cannot always work well in China. In all, we should develop new pricing models which capture Chinese characteristics.This doctoral thesis focuses on several stock pricing issues in China. China’s stock market is known as "policy market" and "retail investor market". Frequent policy interventions, absence of individual charecteristics, blind investment following the trend, China’ stock returns may exhibit many anomalies which are different from developed countries. Based on the above consideration, this thesis attempts to investigate asset pricing issues from the aspects of market risk, fundamental information, historical performance and trading volume.The classic Fama-French three-factor model is based on the American stock markets, but it is also widely used by the investment professionals in China. Examining the portfolio returns during the same period of the Chinese and American stock markets, we find that the market risks in China are dominant. We also show that the size effects exist in China, but the book-to-market effects are not significant. Therefore, the two-factor model with the market factor and the size factor works better in the Chinese stock market. Comparing with the United States, the government policies related to the stock market change frequently, the returns of individual shares are highly sensitive to the market changes. As the factor of firm growths is insignificant in share pricing in China, the Chinese stock market has not been able to optimize the distribution of capital resources.The stock returns experience both under-reaction and over-reaction in China. Examining the data of share prices from1992to2012, we find that, different from New York Stock Exchange and other developed stock markets, there are reversal effects in the super-short term and the long term, but there are merely occasions for the inertia to take effect. We further show that the reversal premium cannot be captured or explained by Fama-French three-factor model, but the Carhart (1997) and Novy-Marx (2012) help. In this paper, we study the features of the Chinese stock returns, structure the reversal factor and present the new four-factor model of asset pricing which is suitable for Chinese stock market. Moreover, the long term reversal effect can be captured by the short term reversal factor. We further find that in the subperiod of2005-2012and in bull market state, the reversal effect is more pronounced. Also, the stocks with small size, high betas and high turnover rate exhibit more significant short term reversal effect.From the aspect of RMB trading volume and turnover rate, this paper investigates how trading volume impact expected stock returns. Employing the data of weekly returns of A-share stocks listed in Shanghai Stock Exchange and Shenzhen Stock Exchange from1995to2012, we construct a modified asset pricing model. Fama-MacBeth cross-sectional regression indicates that RMB trading volume and turnover rate have significantly positive correlation to expected stock returns, which is particularly pronounced in the subsample of2004-2012. Time series regression results show that volume premium cannot be explained by market risk, size, book-to-market and short-term reversal effect. Trading volume factors have a3%marginal contribution to the explanatory power of Fama-French three-factor models. Turnover rate has a more powerful impact in small-caps, high β stocks, and well-performed stocks.Further, we examine how the trading volume shock impacts the expected stock return. Testing the intertemporal role of trading volume in predicting price changes, we find that stocks encounted by volume shock perform worse than those ignored by a large fraction of the investors. This reversal phenomenon is particularly significant in small caps and in bear market. In early subperiod, the trading volume shock has little impact on expected stock return, expect big caps. While in late subperiod, the impact on expect stock return is significantly negative.