Intercorporate default contagion from industry failures: Stress testing on creditee linkage networks of China

This paper explores a simulation framework to examine the financial contagion on the creditee linkage networks, owing to a comprehensive loan-level database of China Banking Regulatory Commission (CBRC). We merge two kinds of connections together, i.e., loan guarantee and shareholding relationships, when constructing our underlying network. Default transmission probabilities are characterized by the two-parameter logistic function with a variant of leverage. Our main finding is that, under the above channel, contagion spreads in a linear form. To be precise, as the scale of the initial shock grows, the total number of default firms and the amount of bad loans incurred both increase linearly. The main reason for this is that the underlying network is rather sparse. Within the same ratio of initially default companies, the Manufacturing Industry is likely to cause the largest number of companies to fail and the largest amount of bad loans. We also find that the default contagion mostly spreads within the same industries, explaining again why the domino effect is very limited. We investigate the impacts on different regions from industry failures, and get many interesting findings, e.g., most industries may influence the Yangtze Delta Economic Circle significantly except for Agriculture and the Mining Industry.