基于计算实验的协同羊群行为与市场波动研究  

Research on collaborative herding behavior and market volatility: Based on computational experiments

作　　者：陈莹[1] 袁建辉[2] 李心丹[1] 肖斌卿[1] 

机构地区：[1]南京大学工程管理学院,南京210093 [2]南京信息工程大学经济管理学院,南京210044

出　　处：《管理科学学报》2010年第9期119-128,共10页Journal of Management Sciences in China

基　　金：国家自然基金重点资助项目(70932003);国家自然科学基金资助项目(70671053;70701016);国家社会科学基金资助项目(07CJL014);教育部科技创新工程重大项目培育资金资助项目(708044);教育部人文社会科学研究资助项目(09YJCAH061);江苏省教育厅哲学社会科学基金资助项目(06SJD630037);江苏省生产力学会资助项目(2006-JS-093)

摘　　要：相对于短期实际利率、消费、红利的波动而言,理论界称股价波动水平异常偏高的现象为"股市波动之谜".以往研究表明,羊群行为和市场情绪的协同作用会引发股票市场的波动.在计算实验平台上,通过协同模拟agent间的模仿和市场情绪信号,在实验中观察到明显的协同羊群行为所引发的股票价格泡沫或崩溃.对羊群行为的研究既考察了agent的私有信号,又包含了总体的市场影响,发现羊群行为和收益波动存在较强相关性的证据.将计算金融实验方法用于行为金融研究具有较强的理论价值,同时对投资者和监管方来说都有一定的借鉴和参考意义.In comparison with the volatility of short-term real interest rates,consumption and dividend,the stock price volatility is in an abnormally high level.This phenomenon is called ＂the stock market volatility puzzle.＂ Previous studies show that synergies between herding behavior and market sentiment would cause stock market volatility.In this paper,it is observed obviously that stock price bubbles or crashes are caused by synergy herding behavior through imitation between agent and market sentiment signals on computing experimental platform.The paper＇s research on herd behavior contains not only the agent＇s private signal,but also the overall market impact.The evidence is found that there is a strong correlation between herd behavior and earnings volatility.It has a strong theoretical value that agent-based computational finance is used in the research of behavioral finance.At the same time,the results will give some reference to investors and regulators.

关 键 词：羊群行为 市场波动 计算实验 

分 类 号：F830.91[经济管理—金融学]