Assessing the systemic risk in Rwandan banking sector using clustering approach

Research Article

Assessing the systemic risk in Rwandan banking sector using clustering approach


Abstract

This article assessed the dynamics of homogeneity (similarities) among banks in Rwanda from 2017 to 2021 by utilizing agglomerative clustering techniques on a detailed breakdown of the banking system’s balance sheet structure. Specifically, we analyzed homogeneity in the banks’ portfolios, namely, the investment, lending, and funding portfolios, as they represent the most crucial areas of fundamental banking activities. The results indicated that the Rwandan banking sector showed high similarity between the lending and funding portfolios of the banks, with the investment portfolio showing the least similarity. In particular, the largest banks exhibited high homogeneity in their investment and lending portfolios. Most medium-sized banks showed strong similarities in the funding and lending portfolios, while all small banks demonstrated a more identical structure across all three portfolios (investment, lending, and funding). These findings have important implications for both policymakers and regulators. High homogeneity between banks can amplify systemic risk, particularly during periods of economic stress when institutions are similarly exposed to shocks. As such, regulatory efforts should aim to encourage diversification in banks’ portfolio structures and monitor herd behaviour that could threaten financial stability. Furthermore, macroprudential tools could be tailored to address the concentrated risks arising from such uniformity in portfolio composition.

Get new issue alerts for African Journal of Science, Technology, Innovation and Development