GRA · Financial Stability
Spatial financeWhere capital meets hazard, exposure and transition
Stress testing methods
Stress tests ask how institutions and the system would fare under severe but plausible scenarios. Supervisors use them to set capital, central banks to study system dynamics, and boards to test their plans. Methods range from bank-run models to system-wide exercises that include nonbanks. This page sets out how stress tests are built and where they mislead, the records they produce, and who decides how they are used.
Stress testing methods on the record.
Stress testing methods: how the risk develops.
Scenario, sensitivity and model limits. The dossier below sets out how the risk develops, what is measured, where the data come from, how it is modelled and who decides.
Scenarios set the answer
A stress test shows losses under the scenario chosen. Scenarios that are too mild, or that miss the next shock's shape, give false comfort.
Static balance sheets miss reactions
Many tests hold balance sheets fixed. Real firms deleverage, sell assets and cut lending, which can amplify stress across the system.
System-wide exercises add feedback
Exercises that model banks, nonbanks and central counterparties together capture second-round effects that firm-level tests miss.
Results drive capital and conduct
When results set capital buffers, firms have incentives to optimise models to the test. Varying scenarios and transparency limit gaming.
What stress testing methods means for exposure, loss and capital.
The exposure, loss, scenario and capital views a risk officer, supervisor or ministry of finance would ask for, with the frameworks they report against. Structures follow supervisory and actuarial practice. Values are illustrative: never prices, ratings or advice.
Transmission channels
How a shock in stress testing methods reaches credit, market, underwriting, liquidity and operational risk, and feeds back.
What is measured, and how well.
Each indicator carries its unit, its cadence, the kind of source behind it and the evidence label a value would carry. No values are shown: a programme supplies them under its own authority.
| Indicator | Unit | Cadence | Source type | Label |
|---|---|---|---|---|
| Capital depletion under the adverse scenario | percentage points of CET1 ratio | annual | Supervisory stress test results | MODELLED |
| Stress capital buffer | percent of risk-weighted assets | annual | Supervisory determinations | Published statistics |
| Liquidity survival horizon under stress | days | periodic | Liquidity stress tests | MODELLED |
| Second-round losses in system-wide exercises | currency | per exercise | System-wide exercise reports | MODELLED |
| Scenario severity | peak-to-trough change in key variables | per exercise | Scenario documents | Published statistics |
| Climate scenario losses in exploratory exercises | currency or percent | per exercise | Climate exercise reports | MODELLED |
Where the evidence comes from.
Named public datasets and institutional records, each with its licence class. Sources by agreement are used only under their terms and never republished.
| Source | What it gives | Licence class |
|---|---|---|
| Federal Reserve stress test scenarios and results | Supervisory scenarios and bank results | Public domain |
| EBA EU-wide stress test results | Scenarios and bank-level results | Free, EU reuse terms |
| Bank of England system-wide exploratory scenario report | System-wide exercise covering banks and nonbanks | Free, Bank of England terms |
| NGFS climate scenarios | Reference climate scenarios for exploratory exercises | Free, NGFS terms |
| IMF FSAP stress test technical notes | Stress test methods and results by country | Free, IMF terms |
Who decides, and what the evidence serves.
Financial Stability prepares evidence. The decisions below are taken by the authorities named, under their own rules, and never by the platform.
| Who decides | The question | Evidence this page serves |
|---|---|---|
| Prudential supervisors | How stress test results set capital buffers | Scenario design, model review and results |
| Central banks and macroprudential authorities | Whether system-wide dynamics need policy response | System-wide exercise findings |
| Bank and insurer boards | Whether capital and liquidity plans survive severe scenarios | Internal and supervisory test results |
| Scenario design teams | Which risks and severity to include | Historical episodes and emerging risk evidence |
Outputs this topic produces
- Guide to stress test types and what each answers
- Method note on scenario design and severity
- Brief on second-round effects in system-wide exercises
- Questions for boards on reverse stress testing
Shared questions the industry asks.
- 01
What scenario is severe but plausible?
How to design scenarios that challenge institutions without being arbitrary.
- 02
What do tests capture?
Which channels, such as credit losses, market moves and funding outflows, are modelled and which are left out.
- 03
Do tests include feedback?
Whether institutions' reactions and system-wide effects are represented.
- 04
How should results be read?
What a stress test shows and what it cannot show.
How the risk arises and transmits.
Sensitivity analysis changes one factor at a time; scenario analysis changes many together in a coherent narrative. Reverse stress testing starts from failure and asks what would cause it.
Most traditional tests are static and firm-by-firm: they apply a shock and measure losses at each institution separately. They miss how institutions' responses, such as selling assets or cutting lending, affect each other and the economy.
System-wide and macroprudential tests add feedback, contagion and nonbank sectors. They are more realistic in structure but rely on more uncertain behavioural assumptions.
New risks challenge methods. Climate, cyber and pandemic scenarios involve long horizons, deep uncertainty and non-financial transmission paths that models were not built for.
Model risk is inherent. Results depend on data, assumptions and model choices; communicating uncertainty is as important as the headline results.
How evidence is gathered, and where it stops.
| Method or data | What it shows | Its limit |
|---|---|---|
| Sensitivity analysis | Exposure to single factors | Ignores interactions |
| Macro scenarios | Coherent multi-factor stress | Scenario design is judgemental |
| System-wide models | Feedback and contagion | Behavioural assumptions uncertain |
| Reverse stress tests | Paths to failure | Can identify implausible paths |
How it is studied, and what the methods cannot say.
- 01Scenario design and stress testingMacro-financial scenario design · System-wide exercises · Reverse stress testsApplied here
- 02EconometricsEarly warning indicators · Quantile regression for growth-at-risk · Spillover estimationApplied here
- 03Complex systems and agent-based modellingAgent-based market models · Early warning of regime shifts · Feedback loop mappingApplied here
- 04Network science and contagionInterbank contagion simulation · Common exposure networks · Multilayer networksApplied here
Questions and answers.
Is a stress test a forecast?
What is reverse stress testing?
What is a system-wide stress exercise?
Are GRA scenarios supervisory stress tests?
Can a stress test prove an institution is safe?
Where are scenarios published?
Contribute evidence from this page.
Reports is the Consortium's quarterly publication. What you know about this subject can become evidence, a case or a correction in the next edition, credited to you and published under CC BY 4.0 where it is your own work.
- EmailGateway GW-GRASend it from this pageEvidence, a correction, a case, a method, a review, a translation or profile input. It arrives tagged with this page, so the editor who owns it receives it.
- ZenodoOpen depositDeposit it and submit to the communityPublish your own record on Zenodo with your ORCID iD, then submit it to the Reports community there. Curators review it; the DOI stays yours.
- ReportsGlobal Risks AllianceSee every way to contributeThe contribution types, how each is reviewed, what credit you receive, the licence and what is never accepted.
Never send personal data, sensitive infrastructure detail or anything you have no right to share. A contribution is reviewed before use; sending it confers no authorship or editorial rights, and sponsorship never buys a finding.
Bring an exposure or capital question to the Alliance.
Information and enquiries are open today. An enquiry receives a reference and names the next step; applications, payments, compute allocations and registrations open only when their team, terms and review capacity exist.
Legal documents.
Maps, data and figures on this site are shown for context and general information. They are not advice, forecasts or warnings; the limits of the information are set out in the Disclaimers and Data Notices.