Stress testing has become one of the most important tools in financial risk management, particularly after the 2008 crisis highlighted the limitations of statistical models in extreme conditions.
What Is Stress Testing?
Stress testing evaluates the impact of severe but plausible adverse scenarios on an institution's financial condition, including capital adequacy, liquidity, and profitability.
Types of Stress Tests
Sensitivity Analysis
- Changes a single risk factor while holding others constant
- Quick and easy but may miss interaction effects
- Example: "What happens if interest rates rise 300bp?"
Scenario Analysis
- Applies multiple simultaneous changes across risk factors
- Historical scenarios: Replaying actual crises (2008, COVID, 2023)
- Hypothetical scenarios: Designing custom severe events
Reverse Stress Testing
- Starts with a defined adverse outcome (e.g., breach of capital minimum)
- Works backward to find scenarios that would cause that outcome
- Identifies hidden vulnerabilities
Regulatory Stress Testing
Key Regulatory Programs
- US CCAR/DFAST: Federal Reserve stress tests for large banks
- EU EBA Stress Tests: Biennial exercises for European banks
- UK PRA Stress Tests: Bank of England annual stress testing
- Climate Stress Tests: Emerging regulatory exercises for climate risk
Regulatory Expectations (Basel)
- Regular stress testing of all material risks
- Range of scenarios including severe but plausible events
- Board and senior management involvement
- Integration into risk management and capital planning
Practical Implementation
Designing Scenarios
- Identify key risk factors for the institution
- Develop narratives linking macroeconomic/market conditions
- Translate narratives into specific parameter shocks
- Ensure internal consistency of the scenario
Running Stress Tests
- Apply shocks to risk models
- Calculate impact on P&L, capital, and liquidity
- Consider second-round effects (behavioral responses)
- Document assumptions and limitations
Using Results
- Capital planning and adequacy assessment
- Setting risk appetite and limits
- Identifying concentrations and vulnerabilities
- Contingency planning and preparedness
Limitations
- Results are only as good as the scenarios chosen
- Models may not capture all second-round effects
- "Known unknowns" vs "unknown unknowns"
- Scenario selection can be biased
Practice stress testing concepts with our FRM question bank!