How do banks conduct liquidity stress tests and what scenarios do they typically model?

Liquidity stress testing simulates severe but plausible funding disruptions to determine whether the bank can survive without external support. It's fundamentally different from solvency stress testing because liquidity crises unfold in days, not quarters.

Three Standard Scenarios:

1. Idiosyncratic stress — The bank experiences a reputation-damaging event (e.g., fraud discovery, earnings miss, credit downgrade) while the broader market remains normal.
2. Market-wide stress — A systemic event (e.g., sovereign crisis, market crash) disrupts funding markets for all participants.
3. Combined stress — Both idiosyncratic and market-wide shocks occur simultaneously (the most severe scenario).

Key Assumptions That Change Under Stress:

Example — Crestline Financial (hypothetical):

Under the combined stress scenario:

• Starting liquid assets: $45B
• 30-day cash outflows under stress: $62B
• 30-day cash inflows (stressed): $22B
• Net outflows: $40B
• Survival period: $45B / ($40B / 30 days) = 33.75 days

Crestline barely survives 30 days. The board requires a minimum 60-day survival horizon, so the treasury must:

• Increase the HQLA buffer by $15B
• Reduce reliance on short-term wholesale funding
• Secure additional committed backup facilities

How Results Feed Into Management:

• LCR and survival horizon calculations under each scenario
• Identification of the primary source of liquidity drain (wholesale runoff? collateral calls?)
• Calibration of the liquidity buffer size
• Trigger points for the contingency funding plan
• Board reporting and limit setting

Exam tip: FRM Part II expects you to distinguish between the three scenario types, explain why combined stress is most conservative, and calculate survival periods.

Practice liquidity stress testing problems in our FRM question bank on AcadiFi.