Stress Testing

Stress testing estimates how adverse conditions could affect a bank, company, portfolio, or financial system. A stress test applies specified shocks or scenarios to exposures, income, cash flows, and capital to reveal vulnerabilities and support decisions.

A stress scenario is hypothetical. It is not a forecast of what will happen or a probability-weighted prediction.

Key Takeaways

• Stress tests are forward-looking complements to current ratios and historical-loss measures.
• The scenario, horizon, balance-sheet assumptions, models, and management actions determine the result.
• Capital stress testing and liquidity stress testing answer different questions.
• A passing result does not prove an institution is safe under every shock.
• The test is useful only when results affect limits, capital, funding, hedging, contingency plans, or other decisions.

Main Types of Stress Tests

A bank usually needs a portfolio of tests rather than one enterprise-wide scenario.

Stress-Testing Workflow

1. Define the objective. Specify the decision, legal entity, portfolio, risk, horizon, and metric.
2. Design the scenario. Make shocks severe enough to reveal vulnerability while preserving internal coherence.
3. Map risk factors to exposures. Connect unemployment, rates, spreads, prices, defaults, deposit flows, or operational events to the balance sheet.
4. Project financial effects. Estimate losses, provisions, revenue, expenses, RWA, capital, cash flows, and collateral.
5. Apply constraints and actions. State whether balance sheets are fixed or dynamic and which management actions are permitted.
6. Review and challenge. Validate data, models, assumptions, overlays, and aggregation.
7. Act on the result. Change capital, funding, limits, pricing, hedges, underwriting, recovery plans, or risk appetite where warranted.

Simplified Capital Stress Example

Assume a bank begins a two-year scenario with $12 billion of CET1 capital and $100 billion of RWA.

If stressed RWA rise to $110 billion, the ending CET1 ratio is about 6.8%.

The result does not answer whether the bank passes an actual supervisory test. The applicable minimums, buffers, scenario instructions, model rules, and permitted capital actions must also be applied.

Capital vs. Liquidity Stress Testing

A bank can remain solvent in a capital projection yet run out of cash under a rapid deposit run. It can also have ample liquidity initially but become insolvent after severe credit or market losses.

Stress Test vs. Nearby Methods

Stress testing should not be reduced to a VaR calculation. Severe scenarios often examine conditions outside or poorly represented by recent statistical history.

Scenario Design

A useful scenario specifies:

• severity and narrative
• time path, not only endpoint shocks
• macroeconomic and market variables
• sector, geographic, and counterparty effects
• deposit and funding behavior
• interaction between credit, market, liquidity, and operational risks
• second-order effects such as rating downgrades, margin calls, or fire sales

The Federal Reserve’s supervisory scenarios explicitly state that they are hypothetical and should not be interpreted as forecasts. Internal scenarios should maintain the same distinction.

How to Evaluate a Stress Test

1. Check relevance. Does the scenario target the institution’s material concentrations and vulnerabilities?
2. Inspect data lineage. Can exposures be traced from source systems through models and aggregation?
3. Challenge models. Review calibration, nonlinear effects, missing risks, overlays, and uncertainty.
4. Test balance-sheet assumptions. Fixed and dynamic balance sheets can produce materially different outcomes.
5. Scrutinize management actions. Capital raising, asset sales, hedging, or funding actions may be unavailable during system-wide stress.
6. Review legal entities. Capital and liquidity may be trapped and not freely transferable.
7. Compare multiple scenarios. One severe scenario cannot span every failure path.
8. Link to decisions. Record the limit, plan, buffer, or control changed by the result.

Common Mistakes and Limitations

• Treating the scenario as a forecast or assigning an unsupported probability.
• Designing a severe headline shock that misses the institution’s actual concentrations.
• Assuming historical correlations remain stable under stress.
• Allowing unrealistic management actions or unlimited market liquidity.
• Ignoring feedback loops, contagion, margin calls, and fire sales.
• Combining capital and liquidity effects without preserving their different time horizons.
• Reporting one point estimate without uncertainty or sensitivity ranges.
• Treating a pass as proof against untested scenarios.
• Running the model without changing a decision or control.

Authoritative Sources

• Basel Committee stress-testing principles
• Federal Reserve 2026 stress-test scenarios
• Federal Reserve stress-testing guidance

Related Terms

• Solvency: Longer-term ability to absorb losses and support liabilities.
• Liquidity Risk: Risk that cash or funding is unavailable when obligations fall due.
• Common Equity Tier 1: A central loss-absorbing capital measure in bank stress tests.
• Bank Ratings: Current assessments that stress testing can challenge with forward-looking evidence.
• Texas Ratio: A point-in-time asset-quality screen rather than a forward-looking scenario model.

Educational Use

This page provides general financial education, not a supervisory stress test, capital plan, risk model validation, or personalized investment, banking, legal, or regulatory advice.