Novel approaches to banking sector stress testing methodology development and validation

Posted: Sep 05, 2020

• Dr. Prof. Emma Peterson
• Dr. Prof. Ethan Nasser
• Dr. Prof. Harper Romano

Abstract

The global financial crisis of 2008 exposed critical limitations in traditional banking sector stress testing methodologies. Conventional approaches, predominantly based on historical data analysis and linear regression models, have proven inadequate in capturing the complex, interconnected nature of modern financial systems. This paper addresses these challenges by proposing a fundamentally new framework for stress testing that integrates insights from quantum computing, behavioral economics, and network theory. The research develops a novel stress testing methodology that incorporates quantum-inspired optimization algorithms to identify the most relevant stress scenarios, integrates behavioral economics principles through agent-based modeling to capture the impact of cognitive biases on financial decision-making during stress periods, and employs dynamic network analysis to model the propagation of shocks through the financial system, providing a more comprehensive assessment of systemic risk.