Multi-fractal Characteristics And Herd Behavior In Chinese Stock Market

The stock market is a very complex nonlinear dynamic system. The existing empirical studies have shown that the stock market has shown complex multi-fractal characteristics. For the stock market, the causes of multi-fractal characteristics of the study will help us reveal the complexity of the stock market to grasp the dynamic changes of the stock market.This article select the sample data from CSI300, the term is from July 11th, 2006 to March 3th, 2009, there is 886 Closing prices. We use Multi-fractal method and Econometrics method in this article. The main content is below:(1) Using the MF-DFA method research Multi-fractal characteristic of the CSI300. The study found that China's stock market return series does not meet the normal distribution; there are long-range correlation characteristics in our country's stock market. Generalized Hurst index depend on the q, China's stock market has clear multi-fractal characteristics.(2) using the CCK model research the herd behavior of CSI300, the study have shown that, there is obvious phenomenon of herd behavior in both"bull"market and"bear"market, and the herd behavior of"bull"market is more obvious. Further research found that Herding behavior is closely related to the market complexity.(3) Based on sliding window technique, we introduce the multi-fractal degree and herd behavior degree. Using the econometric Granger causality test study multi-fractal characteristics and herding behavior of China's stock market. This article shows that the herd behavior in China's stock market is the Granger cause of Multi-fractal. Further research found that the herd behavior have greater impact on the multi-fractal in the rising phase.