It Takes Two to Tango, but More to Assess Systemic Risk: Credit Networks Through the Lens of Hypergraphs

This paper introduces a higher-order network framework based on hypergraphs for assessing systemic risk in bank-firm credit relationships. Unlike conventional network approaches, which rely on pairwise connections, hypergraphs explicitly represent groups of financial institutions jointly exposed to the same borrower as a multilateral interaction. We compare traditional centrality metrics with hypergraph-specific measures to identify systemically important financial institutions, applying these techniques empirically to credit registry data from the Central Bank of Argentina. An adjusted version of the H-eigenvector centrality measure is also proposed, which nonlinearly combines each creditor’s lending amount with the centrality of its co-lenders participating in the same higher-order interactions. We then assess the systemic impact associated with distress shocks to the top-ranked entities selected by each metric. The proposed hypergraph-based measure consistently identifies sets of institutions whose distress generates greater amplified systemic impact than those selected by traditional centrality measures. To the best of our knowledge, this paper provides the first application of hypergraphs to modern bank-firm credit networks for systemic risk assessment. The results show that explicitly accounting for higher-order interactions reveals dimensions of systemic relevance not fully captured by dyadic metrics, providing financial supervisors with a complementary tool for identifying systemically important institutions. This framework can also be readily applied to different countries and financial systems.