Network topology of the Euro Area interbank market

Ilias Aarab, Thomas GottronarXiv 2025 · DOI · arXiv

Abstract

This study utilizes granular regulatory data from the European Central Bank to construct a multi-layer network model of the Euro Area interbank market. By analyzing five distinct layers of financial exposure—including credit, securities, and repos—the research identifies a universal "truncated power law" structure across markets. However, the analysis also reveals that bank centrality is highly context-dependent, suggesting that systemic risk assessments must account for the multiplex nature of interbank relationships to accurately capture potential contagion channels.

Introduction

Since the 2008 financial crisis, the banking sector has been increasingly recognized as a deeply interwoven system rather than a collection of isolated firms. Contagion can spread through multiple channels, such as unpaid loans, devalued securities, or fire sales of shared assets. Traditional models often oversimplify these relationships by collapsing them into a single interbank network. This paper addresses this gap by employing a multi-layer (multiplex) network approach to map distinct financial layers and understand their unique topological properties and associated risks.

Methodology and Data

The study identifies 114 significant banking groups in the Euro Area using the ECB's RIAD and ROSSI databases. Five distinct layers of exposure are defined:

  1. Long-term credit: Loans with maturity > 3 months (AnaCredit).
  2. Short-term credit: Loans with maturity < 3 months (AnaCredit).
  3. Cross-securities: Interbank equity and debt holdings (CSDB & SHSG).
  4. Short-term funding (Repo): Securities financing transactions (SFTDS).
  5. Overlapping portfolios: Shared investments in non-bank entities (FINREP/COREP).

Network properties were analyzed using graph statistics (density, clustering, diameter) and centrality measures (PageRank, Hubs, and Authorities). Maximum Likelihood Estimation (MLE) was used to test for "scale-free" power law distributions in bank exposures.

Network Characteristics

The analysis reveals significant variance in connectivity across the different layers. As shown in Table 1, the "Overlapping Portfolio" layer is the most densely connected, facilitating rapid asset-side contagion, whereas the "Short-term Credit" layer is the sparsest.

Table 1: Summary statistics for the different layers of the Euro Area interbank network.

Despite these differences in density, all layers exhibit "small-world" properties, characterized by low diameters and high clustering coefficients relative to random networks. This implies that most banks are only a few steps away from one another, which can accelerate the propagation of shocks.

Topological Insights

A key finding is the universality of the "truncated power law" for weighted in-degrees (total exposure) across all layers, with power-law exponents ($\alpha$) typically falling between 1.75 and 2.25. This suggests a fundamental regularity in market organization.

Figure 1: Clustering coefficient (cc) vs. degree (k) across different network layers, illustrating the small-world and hierarchical characteristics of the market.

However, bank centrality is not persistent across layers. A bank dominant in the repo market may be relatively insignificant in the securities market. Kendall Rank correlations between layers are only moderate (30-60%), confirming that "flattened" single-layer models lose critical information about which nodes are pivotal in specific contexts.

Conclusion

The multi-layer approach provides a more nuanced view of systemic risk than traditional single-layer models. While the market exhibits stable topological regularities, the shifting importance of banks across different layers necessitates a granular, multiplex perspective for effective financial supervision. Future research should aim to move beyond this June 2021 snapshot to explore how these network structures evolve dynamically during periods of market stress.

Summary updated

Abstract

The rapidly increasing availability of large amounts of granular financial data, paired with the advances of big data related technologies induces the need of suitable analytics that can represent and extract meaningful information from such data. In this paper we propose a multi-layer network approach to distill the Euro Area (EA) banking system in different distinct layers. Each layer of the network represents a specific type of financial relationship between banks, based on various sources of EA granular data collections. The resulting multi-layer network allows one to describe, analyze and compare the topology and structure of EA banks from different perspectives, eventually yielding a more complete picture of the financial market. This granular information representation has the potential to enable researchers and practitioners to better apprehend financial system dynamics as well as to support financial policies to manage and monitor financial risk from a more holistic point of view.
BibTeX
@misc{aarab2025network,
  author = {Ilias Aarab and Thomas Gottron},
  title = {{Network topology of the Euro Area interbank market}},
  year = {2025},
  eprint = {2502.15611},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2502.15611},
}
Figure