The Driving Mechanism Of External Information Force On The Dynamics Of Financial Systems

Financial markets can be regarded as a complex system with many-body interactions by physi-cists.Many concepts and methods of physics,such as statistical physics,theoretical physics,non-linear science,and complex system theory have been applied to the study of the financial markets.Some basic and significant characteristics in the financial dynamic systems,such as leverage effect,price formation mechanism,financial crisis,large volatility,complex network structure,have been widely concerned and studied,and many progress and achievements have been achieved.The analysis of correlations and interactions between different stocks,sector and community structure has attracted much attention.Various researches have shown that well-performance port-folios can be built with the peripheral and central stocks in the network structure constructed by the stock price returns.Internet technology is developing rapidly with the growth of science and technology,our life is connected close to network.More and more behaviors on the Internet are stored as data,forming massive data,which is known as "big data".Big data,as an information resource of data,has potential value worth people to dig.In recent years,to study financial markets with big data has become a new hot spot.Many researchers extract information from Google Trends,twitter,Financial Times,YAHOO and other data sources to analyze and investigate the financial markets.The focus of our article is to introduce the external information,that is the Google search volumes into the study of financial markets,and to study the driving mechanism of external in-formation force on the dynamics of financial systems.The second chapter mainly discusses the influence of the Google search volumes on the evolution of the network centrality.The network structure is constructed by the stock price returns.Further,we use empirical mode decomposi-tion(EMD)to decompose the Google search volumes,for the purpose of studying the influence of external information in different time scales on the evolution of the network centrality.We find that the low mode of Google search volumes,that is,the basic trend of the Google search volumes,has the greatest influence on the network centrality.In the third chapter,we mainly compare and an-alyze the cross correlation matrices constructed by the Google search volumes and the stock price returns respectively.Meanwhile,in order to study the interaction of network communities between the Google search volumes and price returns,we build a bilayer network of them.