Quantifying supply-chain vulnerability and expected systemic risk during economic crises.
A firm's Economic Systemic Risk Index (ESRI) measures the economy-wide output lost if that firm were suddenly removed from the network. The shock doesn't stop at its immediate trading partners — it propagates upstream to suppliers losing a buyer, and downstream to buyers losing a supplier, until it dissipates.
To know whether this risk comes from genuine economic relationships — and not just from firm size — we compare the real network against a null model: a randomized version that preserves each firm's total trade volume by sector, but scrambles who actually trades with whom.
Firms are connected in a supply-chain network of directed transactions (Supplier → Buyer). Shocks propagate upstream (demand/financial shocks) and downstream (supply bottlenecks).
Click any firm node to trace upstream & downstream propagation.
Firms are connected in a supply-chain network of directed transactions (Supplier → Buyer). Shocks propagate upstream (demand/financial shocks) and downstream (supply bottlenecks).
Firm Details
-
To evaluate systemic risk and the influence of international trade, we compare empirical networks with null models representing a local economy (Budapest area) before and during the COVID-19 pandemic.
We build our baseline using temporal snapshots of the empirical transaction network. In the structural analysis, we color nodes based on their involvement in global trade: green nodes are firms engaged in international export/import, whereas gray nodes represent firms operating solely in local trade.
Systemic risk values fluctuate differently across distinct economic sectors. We compare the empirical (Real) ESRI trajectories with reconstructed null models to identify structural vulnerabilities.
During the COVID-19 crisis (2020–2021), a distinct divergence between empirical and reconstructed values occurs across almost all sectors. The sole exception is the postal and courier sector, which tracks the null model closely, suggesting its risk profile remained structurally linked to baseline expectations.
The postal and courier sector (Sector 53) underwent significant structural changes between 2015 and 2022. Driven by the digitization of commerce and accelerated by pandemic lockdowns, the postal hub (Firm F06, highlighted in crimson) became increasingly connected.
Use the timeline controller to trace how this node increases its connection density, expanding in degree and link strength to become a central hub in the overall economic transaction structure.
Adaptive supply chains don't just survive shocks — when firms are free to rewire, the economy as a whole ends up more resilient than a static model would ever predict.
Adaptation beats rigidity
Post-COVID, real systemic risk fell below what the null model expected — firms actively rewiring their supply links made the network more resilient than its static structure alone would suggest.
Risk is not static
The pandemic reshuffled which firms carry the highest systemic risk: trade-enabling firms — like the postal & courier sector — became structurally central almost overnight.
Volume matters, links don't (alone)
A firm's international trade volume predicts its systemic risk, but severing specific international links isn't the mechanism — it's the broader adaptive rewiring of the whole network.
Monitor risk continuously
Because the highest-risk firms change identity during a crisis, static pre-crisis risk assessments can miss exactly the firms that matter most once a shock actually hits.
Data Statement
This research relies on aggregated transaction records of the domestic Hungarian supply-chain network. Node IDs are completely anonymized to protect business confidentiality. Through collaborative agreements, network datasets and reconstruction models are integrated to analyze systemic risks and supply-chain shocks. The network models compare empirical trade dynamics against bipartite null models to assess expected propagation pathways and isolate structural systemic anomalies during economic crises.
Publication
- Mancini, A., Lengyel, B., Di Clemente, R. & Cimini, G., Evolution and determinants of firm-level systemic risk in local production networks Pnas Nexus, pgag233 (2026). doi:10.1093/pnasnexus/pgag233 PDF
Code & Resources
Partners
