Analysis of the correlation network in the US stock market during January 2020

Abstract

In January 2020, our study delved into the US stock market's dynamics as COVID-19 began to affect the global economy. We scrutinized the Dow Jones Industrial Average (DJI) stocks, focusing on the correlations of their returns. We discerned patterns and anomalies through a structural and dynamic analysis of the correlation network facilitated by a distance function applied to the correlation coefficients. The study emphasized the significance of the minimum spanning tree (MST) in shaping the network's structure and influencing the expansion of subnetworks. Central nodes with high connectivity in the MST emerged as crucial, particularly when the market exhibited abnormal behavior. These nodes' daily variations and correlation structures provided insights into the market's evolving nature. We observed that the MST's radius was particularly reactive to market abnormalities, serving as a potential crisis indicator. Our analysis connected the alterations in the MST's central nodes and the overall network structure with shifts in the four fundamental statistical moments of the correlation coefficients and distance weights. These elements proved to be instrumental in detecting and analyzing market irregularities.