摘要
本文回顾了检验羊群行为的实证模型,指出已有的模型不能有效侦测投资者对于某一特定资产组合是否存在羊群行为的缺陷,然后根据GCAPM提出了一个新的测度羊群行为的模型,并得出了不同于传统CAPM的结论:如果投资者对某一资产组合存在羊群行为,那么该组合的预期收益率将与市场上的平均预期收益率呈现非线性关系。通过利用国内深沪两市的板块数据,本文发现两大市场中有相当比例的资产组合存在着显著的羊群行为。
This paper concisely reviews the empirical model to test herd behavior and points out that the extant methods are not applicable to test the herd behavior on given asset portfolios. A new testing model is derived from the basic idea of GCAPM. Compared to the classic CAPM, this model implies : that when investors herd on a given portfolio, a non-linear relationship between the expected return rate of that portfolio and the average return rate in the market will hold. Given the data of Shanghai and Shenzhen Stock Markets, empirical test finds that herd behavior exist on quite a number of portfolios in China.
出处
《金融研究》
CSSCI
北大核心
2007年第05A期108-117,共10页
Journal of Financial Research