Contagion in Banking Networks: The Role of Uncertainty

We study the role of information and confidence in the spread of financial shocks through interbank markets. Confidence in financial institutions has only recently been introduced in computational models studying the stability of financial networks (Arinaminpathy, Kapadia, & May, 2012). However, so far it has been assumed that all agents have complete information about the system. Here we add realism to a model of interbank markets by introducing uncertainty into what banks know about other banks. In our model, information spreads through the lending network and the quality of information depends on the proximity of the information source. Instead of having complete information, banks receive information that is delayed, noisy, or local. This affects their confidence and the resulting lending decisions. We show that introducing uncertainty leads to a substantial increase in the probability of whole-system collapse after an idiosyncratic bank failure. In contrast, when the same shock is distributed among multiple smaller banks, uncertainty mitigates the impact of the shock. The consequences of a large bank’s failure are the most difficult to predict. Our study demonstrates the need for a better understanding of the role of information asymmetries in systemic risk in financial networks.