基于计算实验的证券市场羊群行为研究

发布时间：2018-07-26 15:30

【摘要】：本文研究了证券市场上的羊群行为现象,证券市场中的羊群行为是指投资者不是基于自己分析挖掘的信息进行投资决策,而是受到其他投资者投资行为的影响,模仿其他投资者的投资行为,导致这些投资者在某段时间内做出相同投资策略的一种从众行为。目前关于羊群行为的理论研究侧重研究羊群行为的成因,实证研究主要是用统计学的方法检验投资者行为的相似性,该方法的主要不足在于无法区分真实羊群行为和伪羊群行为。本文采用基于Agent的计算实验的方法,通过基于计算机技术的可控制、可重复的实验,从微观主体的视角来研究证券市场上投资者的羊群行为,克服了传统数学统计方法不能区分“真实羊群行为”和“虚假羊群行为”的缺陷,为研究投资者的羊群行为提供了一个新思路。研究发现人工股票市场上的个体投资者存在明显的羊群行为,羊群行为水平的上升会导致市场收益率和波动性的同时增加；市场的整体特征以及投资者类型结构会影响羊群行为水平,羊群行为水平只是影响市场收益率和波动性的一个中间变量,真正影响投资者收益和波动性的是市场的流动性深度、投资者对私人信息的发掘程度以及投资者类型的结构；在人工股票市场中的各种策略中,仅有类“追涨”策略能战胜市场。
[Abstract]:This paper studies the herding behavior in the securities market. The herding behavior in the securities market means that the investors do not make investment decisions based on the information they analyze and excavate, but are influenced by the investment behavior of other investors. A herd behavior that mimics the investment behavior of other investors and leads them to make the same investment strategy for a certain period of time. At present, the theoretical research on herding behavior focuses on the causes of herding behavior. The empirical research mainly uses the statistical method to test the similarity of investor behavior. The main drawback of this method is that it can not distinguish real herding behavior from pseudo herding behavior. In this paper, we adopt the method of computing experiment based on Agent, through the controllable and repeatable experiment based on computer technology, to study the herding behavior of investors in the stock market from the perspective of microcosmic subject. It overcomes the defect that the traditional mathematical statistical method can not distinguish between "real herding behavior" and "false herding behavior", and provides a new way to study herding behavior of investors. It is found that the herding behavior of individual investors in artificial stock market is obvious, and the increase of herding behavior level will lead to the increase of market return and volatility simultaneously. The overall characteristics of the market and the structure of the investor type will affect the herding behavior level. The herding behavior level is only an intermediate variable that affects the return and volatility of the market. What really affects the return and volatility of the investors is the depth of liquidity of the market. The degree to which investors discover private information and the structure of investors' types; among the various strategies in the artificial stock market, only the "chasing up" strategy can defeat the market.
【学位授予单位】：南京大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：F830.91;F224

【参考文献】