Bankruptcy Propagation on a Customer-supplier Network: An empirical analysis in Japan

Author Name Yoshiyuki ARATA (Fellow, RIETI)
Creation Date/NO. June 2018 18-E-040
Research Project Large-scale Simulation and Analysis of Economic Network for Macro Prudential Policy
Download / Links

First Draft: June 2018
Revised: February 2019


Since firms are interrelated via customer-supplier relationships, the bankruptcy of a firm may lead to the bankruptcy of its suppliers. Due to this contagion effect, one bankruptcy may trigger many subsequent bankruptcies of direct and indirect and have a nonnegligible impact on an aggregate economy. This paper empirically analyzes this bankruptcy propagation on a customer-supplier network by using a comprehensive dataset consisting of more than one million firms and their customer-supplier relationships, and bankruptcy records over April 2013 to February 2017 in Japan. We find that the contagion effect is significant at the firm-level; for example, if 50% customers of a firm go bankrupt, the firm's bankruptcy probability approximately triples. However, it does not immediately imply that there is a substantial risk at the aggregate level that a nonnegligible fraction of firms are forced into bankruptcies by the contagion effect. In fact, by simulating our model, we find that the reach of bankruptcy propagation is very limited in most cases and it is highly unlikely that bankruptcy spread extensively on the network. This is because of the structure of the customer-supplier network. The network structure contributes to absorbing bankruptcy shocks by an aggregate economy rather than spreads bankruptcy on the network.