Liquidity stress testing in a bank

A liquidity stress test answers one question: how many days could this bank meet its obligations if funding stopped and depositors left. The answer is a number the management board owns, and the MaRisk module BTR 3 requires a German bank to produce it.

Two standing ratios sit underneath the test. The liquidity coverage ratio and the net stable funding ratio in Part Six of Regulation (EU) 575/2013 are prescribed calculations with prescribed assumptions. The internal stress test is where the bank states what it believes about its own depositors.

A bank liquidity stress scenario shown through an outflow chart and survival horizon on a tablet.

LCR and NSFR, and why they are not the stress test

The LCR measures high quality liquid assets against net outflows over a 30-calendar-day stress window, with the regulation fixing the runoff rates: a stable retail deposit covered by a deposit guarantee scheme runs off at 5 percent, a less stable one higher, an operational corporate deposit at 25 percent, non-operational wholesale funding far higher. The ratio has to be at least 100 percent. The NSFR looks out a year and compares available stable funding against required stable funding, also with prescribed factors.

Both are useful and neither is the bank's own view. The runoff rates are a supervisory average applied to every institution, so a bank whose deposit base is more concentrated, more digital or more rate-sensitive than the average is under-measured by the LCR and knows it. The internal stress test exists to correct for that, which is why a supervisor compares the two and asks about the gap.

The three scenarios and the minimum periods

The EBA guidelines on institution-specific stress testing and MaRisk BTR 3 both ask for three scenario families. An idiosyncratic scenario assumes the bank alone is in trouble: a rating downgrade, a negative press report, the loss of a funding counterparty, while markets function. A market-wide scenario assumes the opposite: markets freeze, asset prices fall, but the bank's name is not the problem. A combined scenario runs both at once, which is the one that binds.

Minimum horizons apply. Supervisory practice sets at least five working days for the idiosyncratic scenario and at least one calendar month for the market-wide and combined ones, with granular time buckets inside: overnight, to seven days, to one month, and onward. The buckets are what reveal a bank that survives thirty days on paper and fails on day four, which an aggregate figure hides.

The survival period and what sets it

The survival period is the length of time the counterbalancing capacity covers the cumulative net outflow of a scenario. It is a calculation and not a target: take the available liquidity after haircuts, divide by the outflow the scenario produces per day, and read the day on which the cumulative gap turns negative. The ECB guide to the ILAAP expects a bank to define the survival period it requires per scenario and to show the calculation behind the one it achieves.

Counterbalancing capacity means what the bank could actually monetize in the scenario, which is narrower than its liquid assets. A government bond repoed at a stress haircut yields less cash than its market value. A corporate bond may not be repoable at all in a market-wide scenario. A central bank facility has eligibility criteria and an operational process that takes time. A plan that assumes a bond portfolio sells at mid-market in week one has assumed the stress away.

The deposit runoff assumption, after 2023

The deposit runoff is the single assumption that decides most of the result, and the bank failures of 2023 changed how supervisors read it. Deposits left faster than any prescribed rate contemplated, because the customers were concentrated, informed, and able to move money from a phone within minutes of reading the same news. The Bundesbank Financial Stability Review discussed the speed of that outflow and the role of digital access and social media in it.

What follows for a stress test is specificity. A runoff assumption derived from the bank's own data, by customer segment, channel, balance size and coverage under the deposit guarantee scheme, is defensible. A single rate applied to all retail deposits is not, and supervisors now expect that finding to be addressed, not repeated.

Intraday liquidity, which most frameworks keep separate

Intraday liquidity risk is the risk of failing to meet a payment at the moment it falls due, even with ample liquidity measured at the end of the day. The Basel Committee's monitoring tools for intraday liquidity management set the measures: the daily maximum intraday liquidity usage, the available intraday liquidity at the start of the day, total payments, and the timing of time-specific obligations.

A bank in a real-time gross settlement system or an instant payments scheme carries this risk continuously, and instant payments removed the cutoff that used to let a treasurer net before the end of the day. Our page on instant payments in Europe covers the payments side of that change.

How does the ILAAP package the result?

As a document set with a statement at the front. The internal liquidity adequacy assessment process is the bank's own case that its liquidity suffices, and the ECB guide to the ILAAP organizes it into seven principles covering governance, the framework, the funding strategy, the buffer and collateral management, the stress testing and the contingency funding plan. The management body signs a liquidity adequacy statement, which is the sentence a supervisor holds the board to.

The submission also carries a reader's manual listing what was sent and where each topic sits, because the authority reads a package assembled across several teams. In Germany the ECB receives the ILAAP of a significant institution, and BaFin with the Bundesbank for the rest. The result feeds the SREP, which our ICAAP and SREP page covers.

What is the contingency funding plan for?

For the decisions nobody wants to take while the phones are ringing. The plan names the early warning indicators that trigger it, the escalation path and who may invoke it, the funding actions available in order of preference and cost, the collateral that would be pledged, and the communication to supervisors, rating agencies, depositors and staff.

The stress test and the plan connect through the trigger. A scenario that shortens the survival period below the board's tolerance is what activates the plan, so a stress test without a defined tolerance produces a number with no consequence. The plan also needs to be tested, because a funding action that requires a signed agreement nobody has signed is not an action. MaRisk requires the plan and its regular review, and the recovery plan under recovery and resolution planning carries liquidity indicators that point at the same thresholds.

Liquidity stress testing and Finance Loop

Finance Loop is the meeting place for the treasury, risk and data teams in German institutions who build these models and defend their assumptions. Liquidity stress testing is where a bank's data quality becomes visible, because a runoff assumption is only as good as the customer segmentation underneath it.

Finance Loop is a professional network and has the goal of driving the adoption of emerging technologies in finance, such as AI, tokenization, stablecoins, and DeFi. Finance Loop helps its members build skills and personal networks in these fields: Investment & Digital Assets, Payments & Digital Money, Digital Infrastructure & Sovereignty, and Risk & Compliance.

Let's stay in touch

4,000+ members in finance and tech. Become a Network Member for free.

Get updates for free!

Exclusive event invitations, member perks and news from the network. Unsubscribe at any time.

By submitting you agree to the terms.