Dynamic Study On Network Models Of Chinese Stock Market

The stock market is an important part of a country’s financial system.It has great significance to study the feature of the stock market formanagers and investors. Complex network is a powerful tool to study thestock markets. Using the theory of the complex network and weightednetwork, we study the dynamic property of network models of Chinesestock market. The main work is as follow:First, the constituent stocks of the HuShen300index are seen as arepresentative of Chinese stock market.With the theory of complexnetwork,we build a series of consecutive complex networks in order tostudy the feature of Chinese stock market. As found, with the increase ofthe total network degree, the power law of the network decays byexponential. The fluctuation of the stock index and the total network degreeare almost consistent. In addition, the average of network coefficientbecome lager when the market rise and fall, especially for the average offitting error of the degree probability distribution.Further more, the changeof the average fitting error and the stock index are also consistent.We findthat the scalefreeness of the degree distribution is disrupted when the market experiences fluctuation.These conclusions are obvious for theclosing price networks. In addition, the change of the stocks’ degree isalmost consistently when the market experiences fluctuation.Second, the stock market is a highly relevant system that the BBVmodel can not simulate. In this paper, we propose a generalized BBVmodel (FBBV model).In our model, when a new node is added to thenetwork, not only the edge between the connected node and its neighborsbut also other edges will change. But this change is different for differentnotes. And this phenomenon is accord with the stock market.We give theevolutionary algorithm of the FBBV model and the theoretical analysis ofthe strength. We also give the numerical simulation of the FBBV model.The simulation results are consistent with theoretical analysis.Third, using a new method, we present a weighted network of thesample stocks of the HuShen300index. The strength distribution of thenotes is a power-law distribution. It suggests that there are a few notes withhigh strength. These notes are very important in the weighted network.