An analytic approach To network-based modelling for contagious defaults

This study proposes a contagious default model using a network-based approach. We design a cyclical structure for the liabilities held by financial entities under which an unexpected cash inflow occurs in the system. In this framework, we derive the probability of multiple defaults and expected recovery rate of the system in analytic form. This model allows us to quantify systemic risk - the likelihood of simultaneous defaults occurring and extent of the losses from default. Using a statistical test, we finally verify that the proposed formula provides stable and accurate results as well as performs faster than the existing method.

Automated summary

This study proposes a network-based contagious default model and derives the probability of multiple defaults and expected recovery rate in analytic form, allowing for the quantification of systemic risk.