For Risk Analysts ·
What you'll accomplish
By the end of this guide, you'll have a comprehensive scenario library for your risk models — 20–40 stress scenarios covering your key risk domains, ready to use in your next stress testing cycle, risk committee discussion, or contingency planning session. A scenario library that would normally take days to compile will be ready in under an hour.
What you'll need
I'm a risk analyst at a [regional bank / mid-size insurance company / corporate treasury function]. Key characteristics: [size, geography, business lines, key exposures]. I need to build a scenario library for [stress testing / risk committee / business continuity planning].
Press Enter to set the context, then proceed to the scenario generation.
For each risk domain, use this prompt structure:
Generate 8 stress scenarios for [risk domain, e.g. credit risk / operational risk / liquidity risk] for our organization. For each scenario include:
- Scenario name (2-4 words)
- Narrative description (3-4 sentences)
- Severity: Moderate / Severe / Extreme
- Speed of onset: Sudden / Gradual
- Key transmission channels (how does this affect us specifically?)
- Early warning indicators to monitor
- Primary business lines / portfolios affected
Ask for scenarios in order from most likely to least likely — this is useful for prioritization.
These are required by many regulators and often overlooked:
Now generate 3 reverse stress scenarios — scenarios that would threaten the viability of our organization. Work backward from organizational failure to identify what combination of events could cause it.
After generating all domains:
Summarize all the scenarios we've discussed in a table: Scenario Name | Risk Domain | Severity | Speed of Onset | Primary Affected Areas | Key Monitoring Indicator
Copy this table into Excel — it becomes your scenario register.
Select 3–5 of the most relevant scenarios and ask for expanded narratives:
Expand scenario [name] with a full narrative suitable for a stress testing document: quarter-by-quarter timeline of events, quantitative assumptions I should consider, regulatory precedents for this type of scenario, and assumptions I'd need to make to model it.
What you should see: Detailed scenario narratives with timeline, assumptions, and data points you can use to parameterize your models
Troubleshooting: If scenarios feel generic, add more organizational specifics: your top 3 industry concentrations, geographic markets, key revenue streams, or dependencies on specific vendors or counterparties
1. Single risk domain library:
Generate 8 scenarios for [risk domain] at a [org type]. Include name, narrative, severity, speed of onset, transmission channels, and early warning indicators.
2. Cyber/operational risk scenarios:
Generate 6 cyber risk scenarios for [org type], ranging from data breach to ransomware to third-party system failure. Include estimated recovery time and financial impact range.
3. Reverse stress scenarios:
Generate 3 reverse stress scenarios that would threaten our viability. Work backward from organizational failure.
4. Regulatory scenario alignment:
Map our existing scenarios to [DFAST / CCAR / ICAAP / ORSA] requirements. Are we missing any required scenario types?
5. Narrative expansion:
Expand [scenario name] with a quarter-by-quarter timeline, quantitative assumptions, and data points needed to model it.