摘要
本文根据随机折现因子方法的基本理论,结合广义矩阵法和蒙特卡罗模拟,对随机折现因子方法和传统的CAPM对风险溢价的计算进行实证比较研究。实证结果表明,对于中小样本,随机折现因子方法比传统的CAPM方法优越。估计量较为精确,误差小;对于大容量样本,这两种方法性能接近。另外,随机折现因子方法得到Jensen’sα均值比CAPM方法得到Jensen’sα均值小,而且标准偏差明显较小,也从另一角度说明了随机折现因子方法的优越性。
The risk premium, as the core of asset pricing theory, is generally estimated by beta method. Recently, the stochastic discount factor (SDF) is widely used to appraise the risk premium. In this paper, we compare the variance of the estimates of the risk premium under both methods based on asset pricing theories, the generalized method of moments (GMM) and Monte Carlo Simulation with Chinese stock market data, and show that to some extent the SDF method is more efficient than the Beta method for estimating risk oremiums.
出处
《数量经济技术经济研究》
CSSCI
北大核心
2006年第5期131-139,共9页
Journal of Quantitative & Technological Economics
基金
对外经贸大学"十五""211""中国金融市场的发展与风险防范"课题(项目号:e12003)资助。