Paper: arXiv 2502.15611

Authors: Ilias Aarab, Thomas Gottron

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.

Complexity vs Empirical Score

  • Math Complexity: 3.5/10
  • Empirical Rigor: 7.0/10
  • Quadrant: Street Traders — empirical and applied, lower math complexity

Why this score: The paper uses established network theory concepts (e.g., degree distribution, clustering coefficient) and graph metrics without deriving new complex mathematical models, resulting in low to moderate math complexity. However, it heavily relies on real-world, granular datasets (RIAD, AnaCredit, SFTDS, etc.), details data integration and preprocessing steps, and applies these to actual Euro Area banking data for systemic risk analysis, indicating high empirical rigor.

Research Flowchart

  flowchart TD
  A["Research Goal:<br>Network Topology of Euro Area Interbank Market"] --> B["Data Source:<br>Granular EA Banking Data Collections"]
  B --> C["Methodology:<br>Multi-layer Network Approach"]
  C --> D["Computational Process:<br>Distill & Separate Financial Relationship Layers"]
  D --> E["Analysis:<br>Describe, Analyze & Compare Topology"]
  E --> F["Key Outcome:<br>Complete Picture of Financial Market"]
  F --> G["Applications:<br>Financial Stability & Risk Monitoring"]

Evidence Card

What the authors claim, on what data, and who has checked it. Read from the paper's own text by a language model: every quote is verbatim, nothing was re-run. How to read this card

Proposes

The paper proposes a multi‑layer network approach to represent the Euro Area banking system in distinct layers, each capturing a specific type of financial relationship between banks.

Source sentence
In this paper we propose a multi-layer network approach to distill the Euro Area (EA) banking system in different distinct layers.

Assumes 1

  • Some network assumptions such as small‑world properties or power‑law behaviour are valid only for certain types of financial relationships between banking groups. other
    Source sentence
    That is, some assumptions (e.g., small-world properties or power-law behaviour) are valid for only a certain type of financial relationship between banking groups. 16 That is, some assumptions (e.g., small-world properties or power-law behaviour) are valid for only a certain type of financial relationship between banking groups.

Data

  • Register of Institutions and Affiliates Data (RIAD) · European Central Bank / ESCB other unknown
    Legal entities of Euro Area banks
    Source sentence
    In the subsequent sections, we explain their distinct characteristics and provide a rationale for their selection in 2 our analysis: Register of Institutions and Affiliates Data (RIAD): RIAD is the shared master dataset serving several European System of Central Banks (ESCB) and Single Supervisory Mechanism (SSM) business processes and statistical data collections. In our application, RIAD served to identify
  • Centralised Securities Database (CSDB) · European Central Bank / ESCB other unknown
    Securities issued by banking groups
    Source sentence
    CSDB and SHSG data: The Centralised Securities Database (CSDB) is a reference database for all securities relevant for statistical purposes of the ESCB. It comprises information on debt securities, equity instruments and investment fund shares which are stored on a security-by-security basis.
  • Securities Holdings Statistics Database by Group (SHSG) · European Central Bank / ESCB other unknown
    Securities holdings of banks
    Source sentence
    While CSDB covers the issuance of securities by banking groups, the Securities Holdings Statistics Database by Group (SHSG) dataset contains the holdings of securities by banks. Thus, using both databases we gain insights into the securities being issued by banking groups as well as the securities they are currently holding.
  • AnaCredit · European Central Bank / ESCB other unknown
    starting September 2018, monthly · Loan‑by‑loan data from Euro Area banks
    Source sentence
    AnaCredit data: The AnaCredit dataset contains loan-by-loan information on loans extended to corporations collected from EA banks. This data is reported at monthly frequency starting from September 2018.
  • Securities Financing Transactions Data Store (SFTDS) · European Central Bank / ESCB other unknown
    SFT reporting data
    Source sentence
    SFT – Securities Financing Transactions data: The Securities Financing Transactions Data Store (SFTDS) collects and processes data reported under the Securities Financing Transactions Regulation (EU) 2015/23652. SFTs and the scope of the SFTR include3: (a) repurchase/ reverse repurchase transactions, (b) buy-sell back or sell-buy back transactions, (c) securities or commodities lending and securities or commodities
  • FINREP/COREP · European Central Bank / ESCB other unknown
    Financial and regulatory reporting data
    Source sentence
    FINREP/COREP: FINREP (Financial Reporting Standards) and COREP (Common Reporting Stan-2Regulation (EU) No. 2015/2365 on transparency of securities financing transactions and of reuse and amending Regulation (EU) No. 648/2012 Securities Financing Transactions Regulation – SFTR).

Costs

Not applicable The paper reports no trading or execution result.

Finds 5

  • The multi‑layer network constructed from granular data allows analysis of topology and structure of Euro Area banks from multiple perspectives, providing a more complete picture of the financial market. theoretical
    Source sentence
    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. We demonstrate the construction of a multi-layer network covering the broad scope of the EA financial market using data from different granular collections, detailing the preprocessing, integration, and
  • Weighted degree distributions in all layers are consistent with a truncated power‑law, as the null hypothesis of a truncated power‑law is not rejected for any layer. statistical
    Source sentence
    As shown in Figure 4, for all layers we fail to reject H0, and thus find evidence that the (weighted) degree distributions are consistent with a truncated power law.
  • Clustering coefficients and low diameters suggest potential small‑world properties in the layers, though these properties vary across relationship types. statistical
    Source sentence
    In conjunction with the low diameters, all layers demonstrate potential small-world properties as defined by [30]. To explore this further, we follow the approach of [13] and analyze the relationship between degree distributions and clustering coefficients across the layers.
  • PageRank centrality reveals heterogeneity across layers: nodes highly central in one layer may be absent from the top‑10 in another, indicating layer‑specific importance. statistical
    Source sentence
    The results reveal clear heterogeneity between layers; certain nodes are highly central in one layer but not nearly as significant in another. For instance, the largest banking 15 group FR0 (depicted in green) ranks as the most central node in the short-term credit market but doesn’t even make it to the top 10 central nodes in the cross-securities market.
  • Benchmarking real‑world network against common theoretical assumptions highlights the need for caution when simulating interbank networks, as assumptions like small‑world or power‑law hold only for certain relationship types. theoretical
    Source sentence
    By benchmarking the real-world data-derived network against common theoretical assumptions, our findings underscore the necessity for researchers and practitioners to exercise caution when simulating networks to target specific segments of the interbank network. It is important to acknowledge and incorporate the intricacies inherent in the financial markets being modeled.

Checked by others

No other paper in the archive has been found to support or contradict these findings (compared with 14 related papers). Coverage is partial, so this is neither replication nor refutation.

Cited by 1 paper in the archive

Code

No repository by the authors found in the paper.

Extracted 2026-11-14 · same card as JSON · A language model chose these items: check the source sentence before relying on one.

Cited References (62)

  1. powerlaw: A Python Package for Analysis of Heavy-Tailed Distributions (2014) — link
  2. SciPy 1.0: Fundamental Algorithms for Scientific Computing in Python (2020) — link
  3. Centrality measures in networks (2023) — link
  4. Network Analysis for Economics and Finance: An Application to Firm Ownership (2021) — link
  5. Systemic risk in financial systems (2001)
  6. Network topology of the interbank market (2004)
  7. Multi-layered interbank model for assessing systemic risk (2016)
  8. Too interconnected to fail: A survey of the Interbank Networks literature (2016)
and 54 more
  1. Interbank exposure networks (2016)
  2. Networks in finance (2009)
  3. Systemic Risk in Banking (2012)
  4. A survey of systemic risk indicators (2018) — link
  5. The multiplex structure of interbank networks (2015)
  6. Multiplex interbank networks and systemic importance: An application to European data (2018)
  7. Financial contagion (2000)
  8. Network models and financial stability (2007)
  9. Liaisons dangereuses: Increasing connectivity, risk sharing, and systemic risk (2012)
  10. Financial networks and contagion (2014)
  11. How likely is contagion in financial networks? (2015)
  12. Liquidity risk and contagion (2005)
  13. Modelling the emergence of the interbank networks (2015)
  14. Network formation in the presence of contagious risk (2013)
  15. Finding the core: Network structure in interbank markets (2014)
  16. Emergence of scaling in random networks (1999)
  17. Interbank tiering and money center banks (2014)
  18. The structure and resilience of the European interbank market (2013)
  19. Using granular data to facilitate advanced analytics (2022)
  20. The Analytical Credit Dataset. A magnifying glass for analysing credit in the euro area (2017)
  21. DebtRank : Too central to fail? financial networks, the FED and systemic risk (2012)
  22. Epidemics of liquidity shortages in interbank markets (2018)
  23. Structural measures for multiplex networks (2014)
  24. On the origin of systemic risk (2020)
  25. Financial integration in Europe (2012)
  26. The topology of interbank payment flows (2007) — link
  27. Collective dynamics of `small-world’ networks (1998) — link
  28. Network structure and systemic risk in banking systems (2013) — link
  29. Network topology and interbank credit risk (2016) — link
  30. Contagion risk in the interbank market: a probabilistic approach to cope with incomplete structural information (2016) — link
  31. SinkRank : An Algorithm for Identifying Systemically Important Banks in Payment Systems (2013) — link
  32. Light- and Heavy-Tailed Density Estimation by Gamma-Weibull Kernel (2018) — link
  33. Power-Law Distributions in Empirical Data (2009) — link
  34. Beyond the power law: Uncovering stylized facts in interbank networks (2015) — link
  35. A Graph-theoretic perspective on centrality (2006) — link
  36. The systemic importance of financial institutions (2009) — link
  37. Power law distributions in information science: Making the case for logarithmic binning (2010) — link
  38. Likelihood Ratio Tests for Model Selection and Non-Nested Hypotheses (1989) — link
  39. Calibration of optimal execution of financial transactions in the presence of transient market impact (2012) — link
  40. The index of general nonlinear D A E S (1995)
  41. Clinical implications of dysregulated cytokine production (2000) — link
  42. Quasimonotonicity, regularity and duality for nonlinear systems of partial differential equations (1995)
  43. Algorithms for C omputer A lgebra (1992)
  44. Software engineering—from auxiliary to key technologies (1992)
  45. Conductive P olymers (1981)
  46. Neuromuscular blocking drugs in man (1976)
  47. Isolation and characterization of plasmid deoxyribonucleic acid from Streptomyces fradiae (1978)
  48. Global integrated drought monitoring and prediction system (GIDMaPS) data sets (2014)
  49. Quantum scissors: teleportation of single-mode optical states by means of a nonlocal single photon (2002)
  50. Mixing induced CP asymmetries in inclusive B decays (1997) — link
  51. deep SIP : deep learning of S upernova I a P arameters (2020)
  52. Rank Correlation Methods (1970)
  53. Network Topology of the Euro Area Interbank Market (2024)
  54. Multivariate Density Estimation: Theory, Practice, and Visualization (1992)

0 of 62 cited works are in the corpus; resolved via arXiv source bundles, Semantic Scholar and OpenAlex. Corpus-wide: 2596 linked papers.