The Conditional CAPM And Its Empirical Tests On Chinese Stock Market

The capital asset pricing model(CAPM) is one of the classical theoretical models in modern financial economics,but recent empirical evidence of the capital asset pricing model is disappointing,that is,the capital asset pricing model fails to explain the cross-sectional variation of asset expected returns.This may due to many reasons. One crucial reason is that its static specification fails to take into account the effects of time-varying investment opportunity set.Intetemporal asset pricing models(ICAPM) can remedy this theoretical defect of the capital asset pricing model.However, existing empirical tests based on ICAPM diverge on the signal of risk-return relation. What is the intertemporal asset pricing model and what about its application on Chinese stock market? Will the intertemporal asset pricing model reflect risk-return relation and give instruction for Chinese traders?Firstly,by reviewing correlative literature,this paper introduces content and forms of the static asset pricing models and intertemporal asset pricing models as well as empirical tests on them.Based on this,we conclude that divergence on the signal of the risk-return relation may result from the different measures on volatility of returns. Therefore,based on conditional capital asset pricing model of Engle,Lilien & Robins (1987),this paper applies different methods and models to estimate the volatility of stock market returns,and then investigates the intertemporal relation of risk and return on Chinese stock market.Our sample covers stocks that listed on Shanghai A-Share market from 1997 to 2008.Following the traditional literatures,we estimate the risk-return tradeoff using a single series of the SSE(Shanghai Stock Exchange) A-Share Index.Three different approaches are utilized when investigating the conditional CAPM based on the market portfolio:GARCH-in-mean,realized volatility,and range volatility models. Consistent with earlier studies,with the method of realized volatility and range volatility model,the risk aversion estimates provide no evidence for a significant link between the conditional mean and volatility of excess returns on the market portfolio. However,when considering the effect riskless return as one of the explanatory variable in the variance equation,the GARCH-in-mean model provides evidence for a significant positive relationship between risk and return.Different from the traditional literatures,we divide the aggregate stock market portfolio into ten book-to-market portfolios over the sample period of January 1997 to December 2008.We then estimate the dynamic covariance between book-to-market portfolios and market portfolio based on the mean-reverting dynamic conditional correlation(DCC) model of Engle(2002).By pooling the time series and cross section together,we find the intertemporal risk-return tradeoff is significantly positive.There is seldom direct research about the dynamic relation of risk and return on Chinese stock market,and this paper attempt to do that,which is one of the innovations.What's more,we move away from the narrow focus on the single series of a market portfolio return.Instead,we enlarge our sample of observation by constructing ten book-to-market portfolios.And the third innovation is that we estimate the volatility of market excess returns by using three models,which makes our study comprehensive.