Systemic and nexus risk
Systemic Risk: Definition, Cascading Risks and the Polycrisis
Systemic risk is the risk that a disruption in one part of a system spreads through its connections and impairs the functioning of the whole system, with serious consequences beyond the point where it started.
Published Topic: Systemic and nexus risk
Systemic risk is the risk that a disruption in one part of a system spreads through its connections and impairs the functioning of the whole system. The term began in finance, where it means a failure that disrupts financial services and harms the real economy, and it now also describes hazards and failures that cascade across power, water, food, health and ecosystems.
Key takeaways
- The financial definition comes from the FSB, IMF and BIS (2009): disruption to financial services caused by impairment of the system, with serious consequences for the real economy [1].
- Disaster science, led by UNDRR's Global Assessment Reports since 2019, defines systemic risk as risk embedded in a system that is not tracked or managed but can impair the whole system when conditions change [3].
- Systemic risk works through interconnection, non-linearity and feedback; it cannot be found by adding up single-risk assessments.
- Disaster losses illustrate the scale: about USD 202 billion a year in direct losses, more than USD 2.3 trillion once cascading and ecosystem costs are counted [5].
- Governments recognised in 2023 that disaster risk is "increasingly complex and systemic" and that conventional disaster management is "no longer adequate" [12].
What is systemic risk?
All definitions share one idea: the harm comes from the connections between parts of a system, not only from the parts themselves. A single bank failure, a single power cut or a single crop failure is an ordinary risk. It becomes systemic when the structure of the system carries the shock onward, amplifies it, and degrades functions that many others rely on.
Three properties recur across the literature. Systemic risks are interconnected, so impacts move across sectors and borders; non-linear, so small triggers can have disproportionate effects and thresholds can be crossed abruptly; and emergent, so they appear only at the level of the system and are invisible in any one component's risk register [6][7]. The International Science Council and UNDRR briefing note of 2022 deliberately sets out these attributes "without prescribing a single definition", and groups them under the scale of the system, the level of understanding, relationships between elements, transboundary effects and outcomes [6].
Systemic risk in finance: the FSB, IMF and BIS definition
After the 2008 crisis, the G20 asked the Financial Stability Board (FSB), the International Monetary Fund (IMF) and the Bank for International Settlements (BIS) to define and assess systemic importance. Their 2009 report defines systemic risk as "a risk of disruption to financial services that is (i) caused by an impairment of all or parts of the financial system and (ii) has the potential to have serious negative consequences for the real economy" [1]. It places "negative externalities" at the heart of the definition and proposes three criteria for systemic importance: size, substitutability and interconnectedness [1].
That work underpins the annual identification of systemically important institutions, which are allocated to buckets carrying higher capital requirements. The FSB's 2025 list of global systemically important banks (G-SIBs), published on 27 November 2025 using end-2024 data, contains 29 banks, with no additions or removals compared with 2024 [2]. Financial systemic risk is therefore measured and regulated, even if imperfectly, through capital buffers, resolution planning and macroprudential policy.
Systemic risk in disaster and climate science
The disaster-risk community adopted the concept later and broadened it. UNDRR's Global Assessment Report (GAR) 2019 proposed a working definition: systemic risk is "risk that is endogenous to, or embedded in, a system that is not itself considered to be a risk and is therefore not generally tracked or managed", but which systems analysis shows to have a latent or cumulative potential to impair overall system performance when the system changes [3]. The same chapter introduced the "femtorisk", a seemingly small event that triggers consequences at a much higher level of organisation [3].
GAR 2022, Our World at Risk, asked how governance can evolve to address systemic risk, and found that between 350 and 500 medium- to large-scale disasters occurred each year over the previous two decades, with the number projected to reach 560 a year, about 1.5 a day, by 2030 [4]. GAR 2025, Resilience Pays, quantified what interconnection costs: direct disaster losses of about USD 202 billion a year rise to more than USD 2.3 trillion a year once cascading and ecosystem costs are included [5].
The Intergovernmental Panel on Climate Change (IPCC) reached the same conclusion from the climate side, with high confidence: "climatic and non-climatic risks will increasingly interact, creating compound and cascading risks that are more complex and difficult to manage" [13].
Cascading, compound and interconnected risks
Systemic risk is often confused with its mechanisms. The distinctions matter because each calls for a different analysis.
| Concept | What it describes | Typical example | Analytical approach |
|---|---|---|---|
| Compound risk | Multiple drivers or hazards combining in time or space [9] | Heatwave during drought; storm surge with river flood | Joint probability, multivariate hazard models |
| Cascading risk | Impact propagating through dependencies [8] | Flood cuts a substation, which stops water pumping, which closes a clinic | Dependency mapping, network and cascade models |
| Interconnected and interacting risk | Risks linked through shared drivers or feedbacks [8] | Food price spike, unrest and migration | Systems mapping, causal loop analysis |
| مەترسیی سیستەمی | Possibility that the system itself loses function [1][3] | Financial crisis; regional blackout that halts multiple services | Network analysis, stress tests, scenarios with stated uncertainty |
| Polycrisis | Several crises causally entangled across global systems [10] | Pandemic, energy shock and debt distress together | Cross-system causal analysis and foresight |
Pescaroli and Alexander's 2018 framework for compound, interconnected, interacting and cascading risks was written to reduce confusion between these overlapping terms and to support implementation of the Sendai Framework [8]. Zscheischler and colleagues showed that compound events, where floods, wildfires, heatwaves and droughts arise from interacting physical processes, are underestimated when risk is assessed one driver at a time [9].
Polycrisis: when crises become entangled
The World Economic Forum's 18th Global Risks Report, published in 2023, "considered the possibility of a 'polycrisis'", with risks from multiple domains unfolding at the same time, as the 2026 edition recalls [11]. Lawrence, Homer-Dixon, Janzwood, Rockström, Renn and Donges gave the term an analytical definition in 2024: "the causal entanglement of crises in multiple global systems in ways that significantly degrade humanity's prospects" [10].
Current perceptions point the same way. In the Global Risks Report 2026, half of the more than 1,300 surveyed leaders and experts anticipated a turbulent or stormy outlook over two years and 57% over ten years, and inequality was again identified as the most interconnected global risk over the coming decade [11]. Over ten years, the most severe risks ranked were extreme weather events, biodiversity loss and ecosystem collapse, and critical change to Earth systems [11].
How to assess systemic risk, step by step
No single model captures systemic risk. A defensible assessment combines several methods and is explicit about what it cannot see.
- Define the system and its boundary. The briefing note warns that boundaries drawn too narrowly can miss important feedback loops or cascading effects [6].
- Map functions and dependencies. List the services the system provides and what each depends on: power, water, transport, data, labour, finance and ecosystems.
- Characterise coupling. For each dependency, record the mechanism, its strength, buffers (storage, redundancy) and restoration times.
- Identify triggers and thresholds. Include small but well-connected triggers (femtorisks) and points where failure becomes non-linear [3].
- Simulate propagation. Use network, input-output, agent-based or cascade models to trace how shocks spread, and test compound scenarios, not single hazards [9].
- State uncertainty. Report ranges and confidence, and separate observed, modelled and expert-judged results.
- Identify leverage points. Look for nodes where redundancy, decoupling, early warning or governance changes reduce propagation most.
- Govern and revisit. Assign shared ownership and repeat the assessment as the system changes; the IRGC notes that "traditional risk management practices are often not adequate in the face of high uncertainty" [7].
Governing systemic risk
Because no single owner holds a systemic risk, governance is as important as analysis. The IRGC's 2018 guidelines call for a shift towards understanding collective behaviour and emergent properties of systems rather than individual components [7]. In the Sendai Framework's midterm review, governments recognised that "hazards can trigger each other with cascading and compounding impacts within and across sectors and geographies", and called on states to strengthen the monitoring of "systemic risk, cascading effects, compounding hazards and multiple risk drivers" [12]. They also expressed concern that institutional silos limit coordination between disaster risk reduction, climate, biodiversity and economic policy [12].
In practice, this means shared data between agencies and operators, joint scenarios and exercises, clear decision rights for cross-sector incidents, and finance that rewards prevention.
Common mistakes in systemic risk analysis
- Summing single risks. Adding sector risk scores ignores the interactions that make risk systemic.
- Assuming independence. Treating hazards or failures as independent understates joint and tail risk [9].
- Stopping at the organisational boundary. The most important dependencies are often on public systems and nature.
- Using only historical data. Systemic events are rare and the system changes; scenarios are needed alongside statistics.
- Presenting model outputs as forecasts. Cascade models illustrate mechanisms; they rarely support point predictions.
How GCRI works on systemic risk
GCRI's programme starts from the observation that risks are coupled: a flood that closes a substation also stops the pumps that treat drinking water and the road to the clinic. Its model places six sectors (water, energy, infrastructure, food, health and biodiversity) on one hexagon with thirty directed relations, each with a named mechanism such as irrigation or hospital power. A simple teaching cascade shows how a first hit spreads and how much coupling enlarges it. GCRI labels these strengths as teaching values for a typical city-region, not measurements for any place, and its method pages list where a real analysis must do more: place-specific strengths, thresholds and non-linear failure, restoration times, geography, behaviour and stated uncertainty.
Its systemic risk management and systems modelling pages cover governance, cascades, stress tests, coupled-system models and validation. A participating country replaces teaching values with its own evidence through a National Model that GCRI describes as never presented as a validated causal model or a forecast. GCRI informs and does not decide: it is not a regulator, certifier, lender or insurer, and the programme is at service level R0, as recorded on the programme status page.
Further reading
- FSB, IMF and BIS (2009), guidance on systemic importance [1]
- UNDRR, GAR 2019 Chapter 2, GAR 2022 and GAR 2025 [3][4][5]
- ISC, UNDRR and Risk KAN, Briefing Note on Systemic Risk (2022) [6]
- IRGC, Guidelines for the Governance of Systemic Risks (2018) [7]
- Lawrence et al. (2024), Global polycrisis [10]
Written by the GCRI knowledge team. Last reviewed 2026-10-30. Corrections are welcome through the knowledge base report link on this page.
زاراوە سەرەکییەکان
- Systemic risk (finance)
- A risk of disruption to financial services caused by an impairment of all or parts of the financial system, with the potential for serious negative consequences for the real economy (FSB, IMF and BIS, 2009). [1]
- Systemic risk (disaster risk science)
- Risk embedded in a system that is not itself considered a risk and so is not generally tracked or managed, but which systems analysis shows can impair overall system performance when the system changes (UNDRR GAR 2019). [3]
- Cascading risk
- Risk that an initial impact propagates through dependencies, such as power to water to health services, producing secondary impacts larger than or distant from the first event. [8]
- Compound event
- An event with significant impact produced by a combination of multiple climate drivers and/or hazards, such as heat and drought together, whose joint risk exceeds the sum of single-hazard estimates (Zscheischler et al., 2018). [9]
- Polycrisis
- The causal entanglement of crises in multiple global systems in ways that significantly degrade humanity's prospects (Lawrence et al., 2024). [10]
- Femtorisk
- A seemingly small-scale event that can trigger consequences at a much higher level of organisation through complex chains of events (UNDRR GAR 2019, after Levin). [3]
- Interconnectedness
- Linkages between components of a system through which distress in one component raises the likelihood of distress in others; one of three criteria for systemic importance with size and substitutability (FSB, IMF and BIS, 2009). [1]
Frequently asked questions
What is systemic risk in simple terms?
Systemic risk is the chance that a problem in one part of a connected system spreads and damages the whole system. In finance it means a failure that disrupts financial services and harms the real economy. In disaster and climate science it means hazards and failures that cascade across sectors such as power, water, food and health.
What is an example of systemic risk?
The 2008 global financial crisis is the classic financial example: distress at a few large, interconnected institutions spread through funding markets to the real economy. The COVID-19 pandemic is a cross-sector example, spreading from public health to supply chains, finance and education. A regional flood that cuts power, which stops water treatment and closes hospitals, is a local example.
What is the difference between systemic risk and systematic risk?
Systematic risk is market-wide risk that cannot be diversified away, such as interest-rate or recession risk affecting all assets. Systemic risk is the risk that the system itself breaks down because failures propagate through its connections. A systematic shock can trigger systemic risk, but the two concepts answer different questions.
What is the difference between cascading and compound risk?
Compound risk concerns several drivers or hazards occurring together or in sequence, such as heat and drought, that jointly cause more harm than each alone. Cascading risk concerns an impact that propagates through dependencies, such as a power failure that stops water treatment. Real disasters often involve both.
What is the polycrisis?
Polycrisis describes several crises in different global systems that are causally entangled, so that each worsens the others. Lawrence and colleagues defined it in 2024 as the causal entanglement of crises in multiple global systems in ways that significantly degrade humanity's prospects. The World Economic Forum's Global Risks Report used the term in its 2023 edition.
How is systemic risk measured?
In finance, supervisors assess systemic importance with indicators built on criteria such as size, interconnectedness and substitutability, alongside network models and stress tests. Outside finance, analysts use dependency mapping, network models, cascade simulations and scenario analysis. No single number captures systemic risk, so assessments combine models with stated uncertainty.
سەرچاوەکان
- Financial Stability Board, International Monetary Fund and Bank for International Settlements. (2009). Guidance to Assess the Systemic Importance of Financial Institutions, Markets and Instruments: Initial Considerations.
- Financial Stability Board. (2025). FSB publishes 2025 G-SIB list.
- United Nations Office for Disaster Risk Reduction (UNDRR). (2019). Global Assessment Report on Disaster Risk Reduction 2019, Chapter 2: Systemic risks, the Sendai Framework and the 2030 Agenda.
- United Nations Office for Disaster Risk Reduction (UNDRR). (2022). GAR2022: Our World at Risk — Transforming Governance for a Resilient Future.
- United Nations Office for Disaster Risk Reduction (UNDRR). (2025). From billions to trillions: Flagship UN report reveals true cost of disasters and how to reduce them (GAR 2025: Resilience Pays).
- International Science Council, UNDRR and Risk KAN. (2022). Briefing Note on Systemic Risk (Sillmann et al.).
- International Risk Governance Center (IRGC), EPFL. (2018). Guidelines for the Governance of Systemic Risks.
- Society for Risk Analysis / Wiley. (2018). Understanding Compound, Interconnected, Interacting, and Cascading Risks: A Holistic Framework (Pescaroli and Alexander), Risk Analysis 38(11): 2245-2257.
- Nature Portfolio. (2018). Future climate risk from compound events (Zscheischler et al.), Nature Climate Change 8: 469-477.
- Cambridge University Press. (2024). Global polycrisis: the causal mechanisms of crisis entanglement (Lawrence et al.), Global Sustainability 7, e6.
- World Economic Forum. (2026). The Global Risks Report 2026, 21st edition.
- United Nations General Assembly. (2023). Political declaration of the high-level meeting on the midterm review of the Sendai Framework (A/RES/77/289).
- Intergovernmental Panel on Climate Change (IPCC). (2023). Climate Change 2023: Synthesis Report, Summary for Policymakers.