Correlation structure networks of stock market during terrorism: evidence from Pakistan

Abstract

We study topology of correlation structures and construct Pearson correlation-based networks (MST-Pearson) and partial correlation-based networks (MST-Partial) during two time periods of extreme terrorist activities (High civilian and security forces fatalities) and relaxed period (low civilian and security forces fatalities). Our results find that probability density function of Pearson correlation coefficients for relaxed period and partial correlation coefficients for both periods slightly deviates from the gaussian function. MST-Pearson during the period of extreme terrorist activities is a crisis like less stable market structure in comparison with the meta-stable market state structure of relaxed period. Our studies also find presence of two most prominent clusters belonging to cement, and chemical and pharmaceutical sectors among four MSTs. In addition, we find important role of few sectors during the period of extreme terrorist activities in comparison with a more diversified sectoral role in the relaxed period. Furthermore, time varying topological properties indicate an expansion in both MST-Pearson and MST-Partial length in the relaxed period due to counter terrorism strategies. Thus, the study reveals interesting findings and implications for the policymakers and investors of Pakistan stock market during the event of terrorism.

Keywords

• correlation structure
• network topology
• partial correlation
• minimum spanning tree
• terrorism
• emerging market
• stock market