FRM Exam Part II · Liquidity Stress Testing

Designing Liquidity Stress Test Scenarios for FRM Part II

Q: What is the difference between idiosyncratic and market-wide liquidity stress scenarios?

An idiosyncratic scenario hits one bank, for example after a downgrade, so markets stay open. A market-wide scenario hits many institutions at once, so asset sales are harder and funding markets can freeze. A combined scenario has both together.

Q: How do I choose the time horizon in liquidity stress testing?

Match it to how fast the shock works. Use days or one week for runs, about 30 days for tests aligned with the Basel LCR, and longer periods to test survival. Banks usually run several horizons.

Q: Why is the combined scenario usually the most severe?

The bank loses funding because of its own weakness, while the value and saleability of its buffer assets are also impaired because markets cannot absorb its asset sales. Outflows rise and counterbalancing capacity falls at the same time. This is a typical pattern, though it depends on the bank.

Q: What is reverse stress testing in this context?

It starts from a defined failure, such as the buffer running out, and asks which events could cause it. It helps find vulnerabilities that forward scenarios might miss.

Updated 11 October 2026 · Fact-checked

Liquidity stress test scenario design means building plausible but severe shocks to a bank's cash inflows, outflows and buffers. You pick the scenario type (idiosyncratic, market-wide or combined), set severity and time horizon, tie shocks to the bank's business model, then project cash flows and compare them with available liquidity.

Understand Designing Liquidity Stress Test Scenarios

A liquidity stress test asks one question: can the bank meet its payments as they fall due under adverse conditions? The scenario is the story that drives the answer. It defines which shocks hit, how hard, and for how long.

There are three standard scenario types. An idiosyncratic scenario is specific to the bank: a rating downgrade, a fraud loss, a cyber event or a rumour. Only this bank is hit, so markets generally keep functioning for others and asset sales are typically easier than in a market-wide scenario. Even so, the bank's own access to secured funding and its collateral terms may worsen, and counterparties may pull funding from it. A market-wide scenario hits all institutions: a funding market freeze, a sharp rise in haircuts, a fall in asset prices. Here many banks sell at once, so asset sales are harder and central bank support may matter. A combined scenario joins both. It is the most severe, because the bank's own weakness appears while the market is closed. Supervisors usually expect a bank to run all three.

Severity must be severe but plausible. Too mild and the test tells you nothing. Too extreme and management ignores it. Use history (2007-2009 crisis, the failures of Bear Stearns and Lehman Brothers) and hypothetical events. Reverse stress testing works the other way: start from the point of failure and ask what would cause it.

The time horizon matters because liquidity dries up in stages. Overnight and one-week horizons capture the first run on wholesale and uninsured funding. One-month horizons align with the Basel LCR, which uses a 30-day stress. Three-month or longer horizons test whether the bank can survive until it restructures or receives support. Shorter horizons typically assume the most abrupt outflows with little management action, while longer horizons allow for management actions and partial market recovery.

Finally, link the scenario to the business model. A bank funded by retail deposits is exposed to deposit runs. A dealer bank funded by repo is exposed to haircut increases and loss of rollover. A bank with large derivatives books is exposed to margin calls and collateral downgrade triggers. A bank with big foreign currency positions is exposed to FX swap market stress. Good design shocks the vulnerabilities that this bank actually has.

Key formulas to remember

Survival horizon: Survival horizon = the first day on which cumulative net cash flow plus starting buffer falls below zero; Compare it with the minimum the bank's risk appetite requires.
Net stressed cash flow: Net stressed outflow = Stressed outflows − Stressed inflows; Apply run-off rates to liabilities and haircuts or delayed receipt to assets.
Liquidity buffer after stress: Buffer after stress = Σ (market value of asset × (1 − stressed haircut)); Include only assets that are unencumbered and can truly be monetised in that scenario.
Stressed run-off outflow: Outflow = Balance × run-off rate; Rates are higher for unsecured wholesale and uninsured funding than for stable retail deposits.
Scenario severity ordering: Idiosyncratic < Market-wide < Combined (typical severity for a given bank); A rule of thumb, not a law. Market-wide can be milder for a bank with little exposure to it.

How to solve Designing Liquidity Stress Test Scenarios questions

Use this sequence for any scenario design or interpretation question.

1Identify the bank's business model and funding profile: retail, wholesale, repo, derivatives, FX.
2Identify its main liquidity vulnerabilities, such as concentrated funding, collateral triggers or currency mismatch.
3Choose the scenario type: idiosyncratic, market-wide or combined, and say whether markets stay open for asset sales.
4Set severity and plausibility, using historical events, hypotheticals or a reverse stress test.
5Choose the time horizon to match the speed of the shock: days for runs, 30 days for LCR-style tests, longer for survival.
6Apply shocks: run-off rates, haircuts, drawdowns of credit lines, margin calls, and limits on rollover.
7Project cumulative net cash flow, compare it with the buffer, and state the survival horizon.
8Interpret the result and link it to actions: buffer size, limits, contingency funding plan and governance.

Quickest way: Match the scenario to the funding model

When to use it: Use when a question gives a short case and asks which scenario or horizon is most appropriate.

Ask: is the shock specific to the bank, or to markets? That settles the type.
If both, choose combined and expect asset sales to be impaired.
Pick the funding source most exposed: repo means haircuts, deposits means runs, derivatives means margin calls.
Pick the horizon: very short for run risk, about 30 days for LCR-type, longer for survival.
Reject options that assume unaffected asset markets in a market-wide scenario.

Common mistakes in Designing Liquidity Stress Test Scenarios

Assuming assets can be sold at normal prices in a market-wide scenario.
Students carry over the idiosyncratic logic, where markets stay open.
Fix: In market-wide and combined scenarios, apply higher haircuts and limit sales to the truly liquid assets.
Calling a combined scenario simply the sum of two mild shocks.
It sounds like addition.
Fix: A combined scenario is a bank-specific shock occurring during market stress, with interactions such as loss of access to wholesale funding when the bank cannot sell assets.
Using one horizon for every scenario.
Students memorise the 30-day LCR period.
Fix: Choose horizons from the speed of the shock. Use short ones for runs and longer ones to test survival and management actions.
Ignoring the business model.
Students apply generic run-off rates.
Fix: Shock the vulnerabilities that apply: haircuts for repo funded dealers, margin calls for derivatives books, deposit runs for retail banks.
Counting encumbered or non-transferable assets in the buffer.
Total assets look large.
Fix: Include only unencumbered assets available in the relevant legal entity and currency, after stressed haircuts.
Making scenarios so severe they are implausible, or so mild they are useless.
Severity is treated as a number to maximise.
Fix: Aim for severe but plausible, and use reverse stress testing for extreme tail events.

Worked examples

Example 1

A bank holds USD 2,000 million of unsecured wholesale funding maturing within 30 days. In a combined stress it expects 80% not to roll over. Its liquidity buffer is cash of USD 200 million plus government bonds with a pre-stress market value of USD 700 million. The bonds face a 10% stressed haircut. Assume no other cash flows. What is the surplus or shortfall?

Show the solution

Outflow = 2,000 × 80% = USD 1,600 million.
Stressed bond value = 700 × (1 − 0.10) = USD 630 million.
Buffer = 200 + 630 = USD 830 million.
Net position = 830 − 1,600 = −USD 770 million.

Answer: A shortfall of USD 770 million over 30 days. The bank would not survive the full horizon without management actions or outside support.

Example 2

A dealer bank funds most of its inventory in overnight repo with lenders that can demand higher haircuts. Which scenario and horizon best test it, and why?

Show the solution

Funding model: short term secured repo, so the main vulnerability is loss of rollover and rising haircuts.
Scenario: a combined one, as a bank-specific loss of confidence during market stress makes lenders raise haircuts and refuse rollover, while its collateral is hard to sell.
Horizon: very short, such as overnight to one week, because repo runs happen within days. Add a longer horizon to test survival.
Link: shocks should include haircut increases by collateral type and loss of the least liquid repo lines.

Answer: Use a combined scenario with a very short primary horizon (days to one week), plus a longer survival horizon. The design targets repo haircuts and rollover, which is the bank's key weakness.

Exam tips

Read the case for the funding model first. The right scenario usually follows from it.
If markets are described as frozen, assume asset sales are limited and haircuts are higher.
Check whether an option confuses LCR's 30-day stress with a required horizon for all stress tests.
Check arithmetic: apply haircuts to assets and run-off rates to liabilities, never the reverse.
Choose answers that link results to action: buffer, limits or contingency funding plan.

Practice questions from Liquidity Stress Testing

Designing Liquidity Stress Test Scenarios in other exams

The same ground in other exams, if you are preparing for more than one or want another angle on it.

FRM Part IStress Testing Fundamentals and Scenario Design

Designing Liquidity Stress Test Scenarios: frequently asked questions

What is the difference between idiosyncratic and market-wide liquidity stress scenarios?

An idiosyncratic scenario hits one bank, for example after a downgrade, so markets stay open. A market-wide scenario hits many institutions at once, so asset sales are harder and funding markets can freeze. A combined scenario has both together.

How do I choose the time horizon in liquidity stress testing?

Match it to how fast the shock works. Use days or one week for runs, about 30 days for tests aligned with the Basel LCR, and longer periods to test survival. Banks usually run several horizons.

Why is the combined scenario usually the most severe?

The bank loses funding because of its own weakness, while the value and saleability of its buffer assets are also impaired because markets cannot absorb its asset sales. Outflows rise and counterbalancing capacity falls at the same time. This is a typical pattern, though it depends on the bank.

What is reverse stress testing in this context?

It starts from a defined failure, such as the buffer running out, and asks which events could cause it. It helps find vulnerabilities that forward scenarios might miss.