The structural role of weak and strong links in a financial market network

Abstract

We investigate the properties of correlation based networks originating from economic complex systems, such as the network of stocks traded at the New York Stock Exchange (NYSE). The weaker links (low correlation) of the system are found to contribute to the overall connectivity of the network significantly more than the strong links (high correlation). We find that nodes connected through strong links form well defined communities. These communities are clustered together in more complex ways compared to the widely used classification according to the economic activity. We find that some companies, such as General Electric (GE), Coca Cola (KO), and others, can be involved in different communities. The communities are found to be quite stable over time. Similar results were obtained by investigating markets completely different in size and properties, such as the Athens Stock Exchange (ASE). The present method may be also useful for other networks generated through correlations.

To appear in EPJ-B

References in corpus (5)

• Uncovering the overlapping community structure of complex networks in nature and society
• A tool for filtering information in complex systems
• New Model of Internet Topology Using k-shell Decomposition
• Optimal Paths in Disordered Complex Networks
• Spanning Trees and bootstrap reliability estimation in correlation based networks

Cited by in corpus (9)

• A k-shell decomposition method for weighted networks
• Worldwide spreading of economic crisis
• Impact of the topology of global macroeconomic network on the spreading of economic crises
• Structure of local interactions in complex financial dynamics
• A network approach for the scientific collaboration in the European Framework Programs
• Filtering of complex systems using overlapping tree networks
• Distance Distribution in Extreme Modular Networks
• Coalition Formation Algorithm of Prosumers in a Smart Grid Environment
• Abrupt efficiency collapse in real-world complex weighted networks: robustness decrease with link weights heterogeneity