Banking Event Modeling and Simulation in Scenario-Oriented Stress Testing

Abstract

The recent 2008 financial tsunami has made the financial regulators realize the importance of stress testing in banking systems. One of the major challenges in stress testing is to model and calibrate “exceptional but plausible” scenarios in which macroeconomic shocks may cause contagious bank failures that may lead to the breakdown of a banking system. Presently, existing stress testing methods mainly focus on modeling single or multiple risk factors through a “static snapshot” of the banking systems. However, real-world bank crisis scenarios are much more dynamic such that different event occurrence sequences may have different impacts on individual banks and banking systems. For purposes of predicting contagious bank failures in stress testing, we propose the use of event-driven process chains in modeling bank failure scenarios. We refer to this approach as Banking Event-driven Scenario-oriented Stress Testing (or simply the BESST approach). We compare the pros and cons of the BESST approach with two existing approaches in an example scenario. In addition, we conducted a financial simulation based on this example scenario to demonstrate the validity of the BESST approach.