针对目前国外文献中提出的时间变动Beta估计方法不适用于实施了涨跌停制度的股票市场,本文依据涨跌停制度下收益的审查特性提出了一个审查——SS模型,并用该模型对中国股市的时间变动系统风险的特性进行了实证研究,认为市场的波动对大公司股票和小公司股票的影响方向是不一致的,二者风险的差异将随市场波动的加剧而加大。通过样本内预测误差的比较证明,本文所提出的审查——SS模型可以较SS市场模型和市场模型提供更为精确的Beta估计,并且忽略了系统风险的时间变动特性和涨跌停的制度特征将会导致对中国股市大公司股票系统风险的高估和小公司股票系统风险的低估;当只考虑了时间变动特性而未考虑涨跌停制度的影响时,则会导致对股票系统风险的普遍低估。
We proposed a censored-SS model for estimating time-varying betas under price limits which incorporates the censored effect into the SS model by taking into account the effects of price limits. Using the daily data of Chinese Stock Exchange in 1997 to 2001, the time-varying betas yielded by the censored-SS model show that in China the systematic risks of large firms( small firms) tends to increase(decrease) with an increase in aggregate market volatility. To demonstrate that the censored-SS model yields more accurately betas, we compare the RMSE in-sample of the market model, the SS model and the censored-SS model. The results support the censored-SS model.