›› 2017, Vol. 29 ›› Issue (8): 43-52.

• 经济与金融管理 • 上一篇    下一篇

我国股票市场流动性的非线性动力学特征研究:基于分形理论的检验

尹海员, 华亦朴   

  1. 陕西师范大学国际商学院, 西安 710119
  • 收稿日期:2016-07-05 出版日期:2017-08-28 发布日期:2017-09-26
  • 通讯作者: 尹海员,陕西师范大学国际商学院副教授,硕士生导师,博士
  • 作者简介:华亦朴,陕西师范大学国际商学院硕士研究生。
  • 基金资助:

    教育部人文社科基金项目(16YJA790061);陕西省自然科学基础研究项目(2015JM7367)。

Study on the Nonlinear Dynamic Characteristics of Stock Market Liquidity in China——A Test Based on Fractal Theory

Yin Haiyuan, Hua Yipu   

  1. International Business School of Shaanxi Normal University, Xi'an 710119
  • Received:2016-07-05 Online:2017-08-28 Published:2017-09-26

摘要:

利用我国股市日度交易数据构建流动性指标,通过BDS检验、赫斯特指数、关联维检验及李雅普诺夫指数等方法,从微观视角分析了股票市场流动性的非线性动力学特征。实证结果显示:我国股票市场流动性具有一定程度上的非线性结构;进一步的赫斯特指数计算结果证明市场流动性呈现出均值回复的特点,具有反持续性和不可预测性;相应的关联维估计值说明我国股票市场流动性属于一个具有分形分维特征的低维混沌系统;同时市场流动性具有相当的脆弱性,对微小的因素变化非常敏感,极易发生极端性集聚。这些研究结果为股票市场流动性的日常监管、危机爆发时干预政策的实施提供了依据。

关键词: 市场流动性, 非线性动力学, 混沌系统, 关联维, 李雅普诺夫指数

Abstract:

Based on China's stock market data, the paper analyzes the nonlinear dynamics characteristics of stock market liquidity through the BDS test, Hurst exponent, correlation dimension inspection and Lyapunov exponent method from the micro perspective. Empirical test shows that the liquidity of both Shanghai and Shenzhen Stock Markets has a nonlinear structure to a certain extent. Further Hurst index calculation results prove that the market liquidity is characterized by mean reversion, anti-persistency and unpredictability. The corresponding correlation dimension estimation values confirm that the liquidity of China's stock market is typical of a low dimensional chaotic system with fractal characteristics, indicating that market liquidity is considerably vulnerable, highly sensitive to minor changes of relevant factors and very likely to incur extreme agglomeration. These results provide empirical evidence for the normal supervision of liquidity in the stock market and the implementation of intervention policy against crisis.

Key words: market liquidity, nonlinear dynamics, chaos, correlation dimension, Lyapunov index