Paper: arXiv 2502.17044
Authors: Jan Fialkowski, Christian Diem, András Borsos, Stefan Thurner
Abstract
Supply chain disruptions constitute an often underestimated risk for financial stability. As in financial networks, systemic risks in production networks arises when the local failure of one firm impacts the production of others and might trigger cascading disruptions that affect significant parts of the economy. Here, we study how systemic risk in production networks translates into financial systemic risk through a mechanism where supply chain contagion leads to correlated bank-firm loan defaults. We propose a financial stress-testing framework for micro- and macro-prudential applications that features a national firm level supply chain network in combination with interbank network layers. The model is calibrated by using a unique data set including about 1 million firm-level supply links, practically all bank-firm loans, and all interbank loans in a small European economy. As a showcase we implement a real COVID-19 shock scenario on the firm level. This model allows us to study how the disruption dynamics in the real economy can lead to interbank solvency contagion dynamics. We estimate to what extent this amplifies financial systemic risk. We discuss the relative importance of these contagion channels and find an increase of interbank contagion by 70% when production network contagion is present. We then examine the financial systemic risk firms bring to banks and find an increase of up to 28% in the presence of the interbank contagion channel. This framework is the first financial systemic risk model to take agent-level dynamics of the production network and shocks of the real economy into account which opens a path for directly, and event-driven understanding of the dynamical interaction between the real economy and financial systems.
Complexity vs Empirical Score
- Math Complexity: 4.0/10
- Empirical Rigor: 8.5/10
- Quadrant: Street Traders — empirical and applied, lower math complexity
Why this score: The paper employs established network propagation models (like DebtRank concepts) with moderate mathematical formulation, but its primary strength lies in extensive real-world data application, including 1M+ firm links and full interbank data for stress-testing with a specific shock scenario (COVID-19).
Research Flowchart
flowchart TD
A["Research Goal<br>Estimate how supply chain disruptions<br>translate into financial systemic risk"] --> B{"Methodology"};
B --> C["Data Inputs<br>• National Supply Chain Network (1M links)<br>• Bank-Firm Loans<br>• Interbank Network"];
C --> D["Computational Model<br>Financial Stress-Testing Framework"];
D --> E{"Scenario Analysis"};
E --> F["Real Economy Shock<br>COVID-19 Supply Disruption"];
D --> G["Simulation<br>Cascading Defaults &<br>Solvency Contagion"];
F --> G;
G --> H["Key Findings & Outcomes<br>• 70% increase in interbank contagion<br>• 28% increase in financial systemic risk<br>• Framework enables event-driven<br>analysis of real economy-finance interaction"];
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 financial stress‑testing framework that combines a national firm‑level supply chain network with interbank network layers to assess financial systemic risk.
Source sentence
We propose a financial stress-testing framework for micro- and macro-prudential applications that features a national firm level supply chain network in combination with interbank network layers.
Assumes 6
- Loss‑given‑default (LGD) is assumed to be 100% for all loans. model
Source sentence
For simplicity, we assume that the loss given default (LGD) is 100% for all loans.
- A reduction in revenue and material costs is assumed to be proportional to the reduced production level. model
Source sentence
In particular, we assume a reduction in revenue and material costs being proportional to the reduced production.
- The supply‑chain shock propagation follows the mechanism of Diem et al. (2022). model
Source sentence
We use a supply chain propagation mechanism developed in (Diem et al., 2022) and used in (Reisch et al., 2022; Diem et al., 2024a,b; Stangl et al., 2024; Fessina et al., 2024).
- A firm defaults if its profit loss exceeds either its equity or its short‑term liquidity. model
Source sentence
A firm defaults, χi = 1, if its profit losses are larger than either its equity, zi, the first part of the conditional, or its available short-term liquidity, ai −si, the second part of the conditional.
Show 2 more
- Interbank solvency contagion is modeled using the DebtRank mechanism. model
Source sentence
(8) Note the difference to FSRI as here we use Lk(T, ψi), i.e., banks losses after the initial shock propagates via the supply chain, causes losses in the banking sector and those then propagate through the interbank market – where we use the normal DebtRank mechanism.
- Supply‑chain links are filtered to retain only stable links that appear in at least two quarters of 2019. data
Source sentence
12 where transactions occur in at least two different quarters in 2019.
Data
- VAT transaction data (2019) for Hungarian firms fundamentals proprietary
Source sentence
Firms’ value added tax reports are used to recon-struct the supply chain network, W. We use data from the year 2019 that contains records of all transactions among the 243,339 firms that exceeded a tax content threshold of 100,000 HUF (approximately 250 EUR).
- Corporate tax reports and registry data fundamentals proprietary
Source sentence
Financial statements of firms.
- Credit registry loan exposures fundamentals proprietary
Source sentence
We filter the dataset for loan contracts and remain with data on the loan exposures of 40,043 firms to 269 banks in Hungary.
- Bank balance‑sheet and Tier 1 equity data fundamentals proprietary
Source sentence
For the banks we use the same dataset as in the stress testing framework of the financial stability report of the Central Bank of Hungary (MNB, 2024), which contains information on 19 banks and their Tier 1 Eq-uity.
- Firm‑level employment numbers (Jan–May 2020) macro proprietary
Source sentence
This dataset also contains information on the employment numbers, which allows us to use the observed employment reduction from January to May of 2020 to construct realizations of realistic shock scenarios modeled after the actual COVID-19 crisis.
- Synthetic COVID‑19 shock scenarios synthetic unknown
Source sentence
As a showcase, we use firm-level monthly employ-ment data during the early stage of the COVID-19 epidemic in Hungary to recreate a realistic shock scenario for the firm-level supply chain network, containing 243,339 firms and 1,104,141 supplier-buyer-links.
Costs
Not applicable
Finds 3
- Interbank contagion increases by 70% when production‑network contagion is present. statistical
Source sentence
We discuss the relative importance of these contagion channels and find an increase of interbank contagion by 70% when production network contagion is present.
- Financial systemic risk for banks rises up to 28% in the presence of the interbank contagion channel. statistical
Source sentence
We then examine the financial systemic risk firms bring to banks and find an increase of up to 28% in the presence of the interbank contagion channel.
- Supply‑chain dynamics play an unexpectedly important role for financial systemic risk and can amplify interbank contagion. theoretical
Source sentence
Conclusion. Supply chain dynamics play an unexpectedly important role for financial systemic risk. Supply chain contagion interacts with other finan-cial contagion mechanisms and can amplify interbank contagion.
Checked by others
No other paper in the archive has been found to support or contradict these findings (compared with 9 related papers). Coverage is partial, so this is neither replication nor refutation.
Extracted 2026-11-14 · same card as JSON · A language model chose these items: check the source sentence before relying on one.