Investigating Interbank Contagion with Agent-Based Modeling and Functional Dependency Network Analysis (FDNA)

Abstract

Functional Dependency Network Analysis (FDNA) is a tool developed by engineer management researchers for developing systems that are more resilient and less prone to catastrophic failure. Considering financial networks as a system-of-systems, we introduce FDNA for analyzing non-engineering systems. This paper applies FDNA to the financial system to investigate not only what causes contagion but also how it spreads through a financial network, specifically to understand how the failure of one bank affects others in the system. A simple agent-based model of interbank lending will be presented and FDNA will be applied to it. A main goal of the paper is to promote application of a new tool from systems engineering to financial networks.