Modeling Uncertainty in Banking Networks

Abstract

Recent evidence from simulations of banking networks suggests that properties of the network design such as connectivity, bank size, or concentration affect networks’ ability to withstand stress (Arinaminpathy et al. 2012; Gai at al. 2011). However, those studies typically assume that all banks have complete knowledge about the whole system. Here we introduce uncertainty into what banks know about other banks. We model uncertainty scenarios in which uncertainty is translated into asymmetric distribution of information among banks relative to the proximity of information source. Instead of knowing everything agents are faced with information delay and limited information availability. We show that when uncertainty is introduced, the system becomes more fragile to sudden shocks and a relatively small distress can push the system over the tipping point in which the whole banking network collapses.