Network Theory and Systemic Risk Propagation

Network theory enables the quantitative representation of complex interdependencies that define systemic risk in sustainability contexts. It provides a formal structure to map the relationships between actors, resources, and institutions, each of which can transmit or absorb shocks.

Network Structures in Sustainability

  • Nodes: Firms, countries, infrastructure assets, resources, or institutions
  • Edges: Financial exposure, trade, supply obligations, emissions, or ecological links
  • Directed networks: Asymmetrical flows (e.g., supply chains, creditor-debtor)
  • Undirected networks: Reciprocal links (e.g., shared infrastructure)
  • Weighted edges: Strength of connection (trade volume, credit exposure, etc.)
Network Centrality Metrics (Sample Node)
Radar chart of key topological metrics for a critical node

Key Topological Metrics

MetricWhat It MeasuresSystemic Risk Insight
Degree CentralityNumber of direct connectionsExposure hubs
Betweenness CentralityShortest paths passing through nodeGatekeeper/transmission risk
Eigenvector CentralityImportance of node’s connectionsEmbeddedness in influential subnetworks
Clustering CoefficientHow interconnected node’s neighbors areLocal collapse risk
ModularityPartitioning into communitiesFirebreaks, containment

Interconnectivity Modeling

  • Bayesian Belief Networks (BBNs): Directed graphs with conditional probability distributions; simulate cascading effects where data is sparse but causal knowledge exists.
  • Directed Acyclic Graphs (DAGs): Causal chains with no cycles; used for policy pathways, environmental impact chains, and identifying confounders/mediators.
Both BBNs and DAGs allow scenario generation and explainable risk propagation, moving beyond simple correlations.

Risk Propagation Channels

  • Input-Output (I-O) Modeling: Tracks how sectoral outputs become inputs elsewhere; extended with environmental flows (EEIO).
  • Contagion Channels: Financial (asset correlations), ecological (trophic cascades), infrastructure (grid failures), sociopolitical (policy diffusion).
  • Systemic Feedback Loops: Reinforcing (positive) or stabilizing (negative) feedbacks amplify or dampen shocks.

Applications in Sustainability

  • Financial ecosystems: Exposure mapping, SIFI identification, ESG-adjusted capital flows
  • Critical infrastructure: Interdependence of electricity, water, transport, digital; failure propagation modeling
  • Global supply chains: Supplier mapping, shock transmission, dual-risk scenario overlay

Spillover & Vulnerability Metrics

MetricDescriptionUse Case
Diebold-Yilmaz IndexShare of variance in one variable from shocks in anotherDirectional spillover in finance, climate, supply chains
Transfer EntropyNonlinear, model-agnostic directional dependencyDetecting spillovers in high-frequency time-series
Granger Causality NetworksStatistical causation among variablesLagged transmission of price, volatility, stress
Vulnerability ConcentrationNodes whose failure triggers large effectsIdentifying bottlenecks, critical suppliers, substations
Systemic Importance ScoreThresholds for “systemically significant” statusMacro-prudential regulation, ESG systemic risk
Sample Spillover Indexes
Directional spillover strengths between sectors (sample data)

References and Further Reading

  • Barabási, A.-L. (2016). Network Science
  • Diebold, F. X., & Yilmaz, K. (2014). On the network topology of variance decompositions: Measuring the connectedness of financial firms. Journal of Econometrics
  • Leontief, W. (1986). Input-Output Economics
  • OECD. (2022). Network Analysis for Sustainability
  • Basel Committee on Banking Supervision. (2022). Principles for the Effective Management and Supervision of Climate-Related Financial Risks