An Ontology-Based Approach for Bank Stress Testing

The 2008 banking crisis has demonstrated that there is the lack of effective methods for modeling and analyzing """"exceptional but plausible"""" risk scenarios in bank stress testing. However, existing bank stress testing practices mainly focus on modeling probability-based risk factors and events in a """"static snapshot"""" of the banking systems, but largely ignore the dynamic processes in which financial crisis events and their interactions creates various complex risk scenarios. In addition, the rare (low probability) risk events such as the bankruptcy of Lehman Brothers that can cause """"exceptional but plausible"""" crisis scenarios are largely ignored due to the lack of appropriate modeling and analysis methods. To address this problem, we developed an approach called Banking Event-driven Scenario-oriented Stress Testing (or simply the BESST) which mainly includes three components: 1) a set of stress testing ontologies, 2) an event-driven scenario model (OESM), and 3) a scenario recommendation component. In addition, we show how to use BESST to model and examine """"exceptional but plausible"""" stress testing scenarios in an example process of crisis events. In general, this research has provided the bank stress testing stakeholders a novel approach for modeling and analyzing the rare risk events and their dynamic processes in various financial crisis scenarios.