Clustering and Banking Interconnectivity: Identification of Contagion Channels in ECOWAS


  •  Abdoulaye Soumaila    

Abstract

This paper analyzes the interconnectivity of banks in the Economic Community of West African States (ECOWAS) and identifies potential channels of contagion. From a sample of 120 banks covering the period 2001-2023, we use the K-means clustering algorithm to form homogeneous groups of banks with similar risk profiles. The results reveal the existence of ten distinct clusters, with banks in Ghana and Nigeria appearing as interconnected central nodes with several clusters. WAEMU banks have an average probability of default (PoD) of 9.9%, significantly lower than that of banks in the West African Monetary Zone (WAMZ) which reaches 20.7%. Moroccan banks, although foreign to the region, contribute less to systemic risk (PoD of 6.5%) than local banks. The study highlights that interbank debt acts as a moderator of systemic risk, with a significant negative relationship to the lower quantiles of the distribution. These results argue for enhanced cooperation between banking supervisors in the region.



This work is licensed under a Creative Commons Attribution 4.0 License.