深圳股市时变Beta、条件CAPM实证研究

建立Engle(2002)提出的动态条件相关多元GARCH模型计算深圳股市诸行业指数2001/07/02～2005/07/15期间的时变Beta系数,进而对系统风险Beta系数与收益的关系进行传统的检验和由Pettengill et al.(1995)提出的条件检验,并且探讨了非系统风险、总风险在资产定价中的作用.研究结果表明,Beta与收益间不存在传统的无条件相关关系;部分行业指数的Beta系数与收益符合条件相关关系:当超额市场收益大于0(上市场)时,Beta和收益正相关;当超额市场收益小于0(下市场)时,Beta与收益负相关.但对大多数指数而言,Beta与收益仅在下市场时呈显著的负相关关系.同时非系统风险以及总风险均得到了补偿,表明深圳股市的投资者并没有充分分散化其投资,政府应大力发展机构投资者.

Empirical Research on Time-varying Betas and Conditional CAPM of Shenzhen Stock Market

In this paper,time-varying betas of industry indices in Shenzhen stock market over the period July 2~(nd) 2001 to July 15~(th) 2005 are estimated by using the dynamic conditional correlation multivariate GARCH(DCC-MVGARCH) proposed by Engle(2002),and then the risk-return relations are examined by using traditional approach and conditional approach proposed by Pettengill et al.(1995).In the same time,whether unsystematic risk or total risk is priced in Shenzhen stock market or not is tested.The findings suggest that there is no evidence of a significant positive risk premium on betas when the traditional unconditional relation between beta and return is considered.And there is a significant conditional relation between beta and return for some industry indices,namely a significant positive(negative) relation between beta and return when the market excess return is positive(negative).But only in the down market(the market excess return is negative) is the negative relation between beta and return significant.Also our results indicate that unsystematic risk as well as total risk is priced in Shenzhen stock market and investors diversify their investment only to a limited extent not fully,and the institutional investors should be greatly developed to make the stock market developing well.