Properties Studies Of Complex Networks On China Securities Markets

Complex networks is a way to abstract and describe complex systems, which prominently pay attention to topology characteristics of system structure. In principle, when we abstract cells as vertices and abstract relations of cells as edges, every complex system that has lots of cells can be studied as complex network. Recently, complex networks have seen much interest from all research circles and have found many potential applications in a variety of fields including engineering technology, society, politics, communications, medicine, neural networks, economics, management and so on.Securities market is also a complex system, it is a great significance to study securities market using theories and method of complex networks. This paper study complex networks characteristic of China securities markets, including constructing network based on price relationships of stocks, small world efficient, scale-free properties, community structure and evolution dynamic model of investors. It provides a bran-new direction for theories investigation and practice application of China securities market. The main works of this paper are as follow.First of all, this paper collected price relationship of stocks that are listed on Shanghai and Shenzhen Stock markets and were trading continuingly in 2006. Then we constructed complex networks of stock prices'relationships in two ways. One is complex networks of stock prices'relationships under thresholds appointed; the other is minimum spanning tree network of stock prices'relationships. Secondly, this paper investigated the two types of networks using theories and method of complex networks. We discovered that networks of stock prices' relationships under thresholds appointed in two markets possess typical characteristic of small world networks——average path length is short and clustering coefficient is high, and present properties of scale-free networks. Minimum spanning tree network of both markets present properties of scale-free networks. Then this paper investigated community structure in complex networks. We divided up networks of stock prices'relationships under different thresholds appointed and got communities which were similar with stock boards plotted out by industries and guided separated investments. Finally, this paper investigated investors'behaviors using complex theories. We constructed evolution dynamic model of investors. After simulating, the network presented balance and has short average path length and high clustering coefficient. Contemporary, we discovered the networks having a certain extent herd behavior.