Measuring and forecasting financial system resilience under multiple shocks: A survival analysis approach

Abstract

This paper addresses the problem of measuring and forecasting financial system resilience. We propose two key extensions to the prevailing shock-based methodologies. First, we incorporate multiple risk sources using generalized impulse response functions, enabling the assessment of general financial resilience rather than specific measure under a single shock. Second, we apply survival analysis – a novel and highly suitable modeling framework for shock methodologies – to integrating financial resilience measurement, key influencing factors identification, and resilience forecasting within a framework. Empirical analysis utilizing US financial system data demonstrates that the proposed measure achieves a 23.81% improvement in accuracy compared to the existing absorption duration measure. Furthermore, the method is forward-looking and proves effective and convenient for identifying critical influencing factors and forecasting financial resilience.