Who this entry is for — Anyone who needs to understand how an economic, financial or operational story propagates through factors, positions and resources without turning it into a certain forecast.
Scenario analysis describes a coherent set of conditions and, where needed, their evolution over time. It connects a narrative to paths for risk factors — prices, rates, volatility, spreads, currencies, liquidity or operational variables — and estimates effects on the selected scope.
In the Basel Committee framework it is one methodology in the stress-testing family, alongside sensitivity analysis and reverse stress testing. A scenario may be historical, hypothetical or hybrid. It is not necessarily the most likely forecast and does not always require a numerical probability.
In plain terms — A sound analysis does not say “this will happen”. It says “if these coherent conditions occurred, this is how they would propagate and where the system would be vulnerable”.
Boundaries with adjacent methods
| Method | Structure | Main use |
|---|---|---|
| Sensitivity analysis | Changes one or a few inputs, even without a full narrative | Identify exposures and non-linearities |
| Scenario analysis | Combines coherent factors and, when needed, paths | Explore joint transmission and impacts |
| Reverse stress testing | Starts from a breaking outcome and seeks conditions that could cause it | Find vulnerabilities and viability thresholds |
| Post-entry plan | Conditional rules for managing one trade | Prepare operational trade decisions |
The final row is covered separately in Scenario. It may use an “if X, then Y” structure, but it does not replace risk scenario analysis.
Structure of a traceable analysis
- Objective and scope — Define the decision, entity, portfolio, positions, date, currency and resources under review.
- Horizon and dynamics — Distinguish an instantaneous shock, a multi-period path and when losses or constraints are recognised.
- Narrative — Explain the mechanism linking events and why it is relevant to vulnerabilities in the scope.
- Factors and paths — Specify levels, changes and relationships among rates, prices, volatility, spreads, currencies, liquidity and other drivers.
- Consistency — Check that factors are not an arbitrary collection of incompatible shocks, and document any exceptions.
- Transmission — Revalue linear and non-linear payoffs, margin, collateral, funding, counterparties and potential second-round effects.
- Results and limitations — Separate impacts, contributions, excluded risks, sensitivity to assumptions and model uncertainty.
- Use and challenge — Connect findings to possible decisions, checking that actions remain feasible under the stated conditions.
There is no universal number of scenarios. One focused story can answer a specific question; a broader set may be needed to cover different vulnerabilities. The choice follows the purpose, not a fixed “three to five” rule.
Severity, plausibility and probability
Basel principles call for scenarios that are sufficiently severe and varied for the purpose and, for adverse scenarios, severe but plausible. “Plausible” does not mean “forecast” or “frequent”. Historical events can inform calibration, while hypothetical scenarios represent emerging risks or combinations absent from the data.
Assigning a probability is useful only where the method supports it. A number without an empirical or model basis creates false precision. The Federal Reserve, for example, states that its stress scenarios are hypothetical paths, not forecasts.
Interpretation errors
- turning the scenario into a forecast or price target;
- selecting dramatic but mutually inconsistent shocks;
- ignoring liquidity, funding, non-linearity or counterparty responses;
- assuming historical correlations remain unchanged under stress;
- prescribing generic actions — such as “limit orders only” — without testing execution, costs and effects;
- reporting precise outputs without exposing model risk.
Review frequency depends on purpose, portfolio change, environment and risk materiality. No monthly or quarterly schedule is valid for every analysis.
Typical mistake — Treating the narrative as certainty and the modelled number as observable precision. Both are conditional inputs that require challenge.
Sources
- Basel consolidated guidelines — RMA30, Stress testing
- Basel Committee — Stress testing principles
- Basel Committee — Supervisory and bank stress testing: range of practices
- Federal Reserve — Policy Statement on the Scenario Design Framework for Stress Testing