【摘要】 为分析中国股市房地产板块股票的强相关特性,以101只股票为结点,以近17年股票对数回报的相关系数为加权边,建立复杂网络模型,通过对网络拓扑参数计算,发现该网络为无尺度网络,结点度分布P(s)~s-|s|δ,在不同相关系数阈值下,δ值介于0.8~1.6之间.网络平均集聚系数为0.53.文章也对网络中心性进行测量和子群划分,发现代码为000592和601588的结点在网络中具有很高的中介性,网络中大部分结点都受其影响较大.更多还原

【Abstract】 In order to analyse the strong correlations of stocks in Real Estate Sector in China Stock Markets,the article makes complex networks model,which regard nodes as all 101 stocks,and weighted links as correlation coefficient of logarithm returns of stocks price in recently 17 years.By computing parameters of networks topologies,we find that the network scale-free,node degree distrbution P(s)~|s|-δ with δ∈(0.8,1.6),and mean network cluster coefficient 0.53.The article makes also measures for network centralities,seeks for subgroups,and cut-up for partitions.We find that node 000592 and 601588 have high centrality,which influence entirety network.更多还原