Network structures and credit risk in cross-shareholdings among listed Japanese companies

Abstract

This study assesses the network structures of cross-shareholdings among listed Japanese companies using a dataset of 2936 companies for the fiscal years 2008–2015. First, I analyze the network structure of cross-shareholdings on the Japanese stock market using centrality measures. Financial institutions and companies in the motor manufacturing industry are central in the network in terms of degree centrality. By contrast, companies in various industries play a central role in terms of betweenness centrality. Second, I use credit risk analysis to investigate the effects of cross-shareholdings using a panel regression. The results show that an increase in the share of foreign companies does not increase the counterparty credit risk in the network. In addition, an increase of a direct centrality measure (i.e., degree or eigenvector centrality) indicates that the connections among cross-shareholdings become denser in a neighboring area and increase credit contagion risk. By contrast, an increase of an indirect centrality measure (i.e., eccentricity or farness centrality) shows that the connections among cross-shareholdings become dispersed at the global level and decrease credit contagion risk. The results could provide senior managers with an important signal for credit risk management related to cross-shareholdings.