Dynamic analysis of influential stocks based on conserved networks

Abstract

Characterizing temporal evolution of stock markets is a fundamental and challenging problem. The literature on analyzing the dynamics of the markets has focused so far on macro measures with less predictive power. This paper addresses this issue from a micro point of view. Given an investigating period, a series of stock networks are constructed first by the moving-window method and the significance test of stock correlations. Then, several conserved networks are generated to extract different backbones of the market under different states. Finally, influential stocks and corresponding sectors are identified from each conserved network, based on which the longitudinal analysis is performed to describe the evolution of the market. The application of the above procedure to stocks belonging to Standard \& Pool's 500 Index from January 2006 to April 2010 recovers the 2008 financial crisis from the evolutionary perspective.

arXiv commentlatex, 16 pages, 1 figure, 14 tables

References in corpus (18)

• The structure and function of complex networks
• Noise Dressing of Financial Correlation Matrices
• Universal and non-universal properties of cross-correlations in financial time series
• A tool for filtering information in complex systems
• Dynamics of market correlations: Taxonomy and portfolio analysis
• Topology of correlation based minimal spanning trees in real and model markets
• Clustering and information in correlation based financial networks
• Evolution of worldwide stock markets, correlation structure and correlation based graphs
• Correlation based networks of equity returns sampled at different time horizons
• Reconstruction methods for networks: the case of economic and financial systems
• Structure of local interactions in complex financial dynamics
• Structural and topological phase transitions on the German Stock Exchange
• Dissecting cross-impact on stock markets: An empirical analysis
• Cross-response in correlated financial markets: individual stocks
• Efficient construction of threshold networks of stock markets
• Nonparametric sign prediction of high-dimensional correlation matrix coefficients
• Financial instability from local market measures
• Complex Correlation Approach for High Frequency Financial Data