中国股票市场随机贝塔的估计  

Estimation of SBETA Model in Chinese Stock Market

作　　者：罗捷[1] 劳兰珺[1] 

机构地区：[1]上海复旦大学管理学院,上海200433

出　　处：《系统管理学报》2008年第1期47-50,共4页Journal of Systems & Management

基　　金：国家自然科学基金资助项目(70371010)

摘　　要：通过使用贝叶斯方法，在中国股票市场估计了Jostova与Philipov的SBETA模型。该模型假设投资组合的贝塔系数变化满足一个均值回归随机过程。通过使用马尔科夫链蒙特卡罗算法对中国股票市场部分行业及部分股票2000-06～2005—05的贝塔系数进行估计，发现中国股票市场投资组合的贝塔系数存在随机波动的现象，但是贝塔系数在发生改变以后，会立刻回到长期均值的水平。中国股票市场中股票的贝塔系数可以表示成为一个贝塔系数的长期均值加上一个随机“噪声”的形式。贝塔系数的长期均值可以作为对未来贝塔系数的预测。Parameters of the SBETA model proposed by Jostova and Philipov fitted to data of the Chinese stock market have been estimated using Bayesian methods. The model set a mean-reverting stochastic process for the market betas. The betas of several industries and some individual stocks over June 2000- May 2005 have been estimated based on a MCMC algorithm. The beta variation was confirmed in the Chinese stock market. However, when the betas deviated from their long-term means, they would return immediately. The beta can represent a long-term mean with ＂noise＂. The long-term mean can be used as a prediction for beta.

关 键 词：贝塔系数 均值回归过程 贝叶斯方法 马尔科夫链蒙特卡罗方法 

分 类 号：F830.9[经济管理—金融学]