An Empirical Analysis Of The Interdependency And Topology Of Shanghai Stock Network Under The Global Financial Crisis Based On Complex Network

Since the 90 s of the 20 th century, with the development of graph theory, big data and computer science, complex network theory has attracted people’s attention and developed rapidly,it is used as a tool to find the structures and properties of the networks. As a place for the issuance and trading of listed stocks, stock market has many functions and is influenced by many factors.Stock network is a typical complex system, we can study the correlation between members, the stability and the topology of the network by complex network theory.In this paper, we take the daily closing price of the constituent stocks of Shanghai Stock 180 Index as the research object, select the period from 2007 to 2010 including the special time of global financial crisis, and use each stock as a node to model the stock network. Firstly, we establish the stock correlation network, through analysis stability, modularity, and the average clustering coefficient as a function of time and the network determined by a threshold that was decided through these three parameters, we find that the correlation between stock nodes is inversely proportional to the fluctuation of stock market, the linkage between stocks is very large especially under the financial crisis. Moreover, we find that Shanghai stock network is quite unstable by the reason that the communities of network changes a lot and division is unclear under the financial crisis. Secondly, through analysis the node weight distribution and the ranking of influence factor before and after the financial crisis, we find that the interdependency between stocks is larger and the industry that hub nodes belonged to changes a lot under the financial crisis.Finally, we analysis the network using the K-means clustering algorithm, proving the conclusion that the division of community is basically based on industry or similar industry, and the number of club members varied considerably under the financial crisis, nodes of large influence factors are concentrated in several communities and combined to have a greater impact on the whole network.This paper analyzes the Shanghai stock network both on macroscopic and microcosmic and the method we used to construct network is not only applicable to the network we selected, but also can be applied to other models.