金融市场的无标度特征研究  

Scale-free properties of financial markets

作　　者：周艳波[1] 蔡世民[1] 周佩玲[1] 

机构地区：[1]中国科学技术大学电子科学与技术系,安徽合肥230027

出　　处：《中国科学技术大学学报》2009年第8期880-884,共5页JUSTC

基　　金：国家自然科学基金(70571075;60874090)资助

摘　　要：基于股票价格波动序列的相关特性,提出一种金融市场的复杂网络建模机制.通过研究基于股票价格波动序列建立的复杂网络模型,发现金融市场的网络节点度分布具有无标度特征.它说明少数"中心"节点的股票对金融市场整体价格波动影响力比较大,甚至可以影响全局,其他大多数股票影响力相对较小.进一步研究网络聚类系数与最近邻平均度,发现金融市场具有分层结构和非相关联匹配的特征.这些结论对于从复杂网络的角度理解金融市场相互作用机制可能有重要的启示和作用.Based on the correlations of stock price fluctuation, a mechanism was proposed to extract complex networks from financial markets. The results suggest that the stock correlation networks of financial markets have a scale-free property of degree distribution, which indicates that the few hub stocks have a great influence on the Whole fluctuation of financial markets while the others have a weak influence on it. The hierarchical structure and disassortative mixing properties of stock correlation networks are investigated according to the clustering coefficient and degree correlation. The conclusions in complex networks aspect may provide some insights into the understanding of the intrinsic mechanism of financial markets.

关 键 词：复杂网络 无标度 金融市场 关联矩阵 

分 类 号：N94[自然科学总论—系统科学] O59[理学—应用物理]