Liquidity doesn’t fail all at once. It fails in a sequence that your KRI dashboard should have been tracking weeks earlier.

Post-SVB, 61% of banks surveyed increased the severity of their liquidity stress test scenarios, and 46% rewrote their contingency funding plans to account for faster deposit runoff. That’s a significant institutional response — but the harder question is whether the monitoring that should have preceded those plans actually exists. A rewritten CFP with vague triggers is still a vague CFP.

The eight metrics below are the liquidity risk indicators that connect to real CFP trigger thresholds, examiner expectations, and the funding dynamics that made 2023 a defining year for liquidity risk management. They’re not a comprehensive treasury function — they’re the specific measures that determine whether a funding problem is visible before it becomes a crisis.

TL;DR

• Uninsured deposit concentration is the single most dangerous liquidity KRI — not because it’s hard to measure, but because most institutions don’t know how fast it can move when it starts
• The eight KRIs below map directly to CFP Yellow (elevated monitoring) and Red (activation) tiers — metrics without that mapping are decoration
• OCC Bulletin 2023-15 updated the Liquidity Comptroller’s Handbook; the OCC Spring 2025 Risk Perspective called out deposit competition and unrealized losses as ongoing concerns
• Failed funding test results are a KRI that most programs don’t track — and examiners are starting to ask about them

Why Liquidity KRIs Exist Separately From Credit and Operational KRIs

Liquidity risk has a different failure mode than most risk categories. Credit losses materialize over months. Operational failures are usually contained events. Liquidity can collapse in hours.

The FSB’s research on depositor behavior documented what practitioners already suspected: online banking reduces the friction that once gave institutions time to respond, and social media accelerates information dissemination faster than any formal crisis communication plan can match. A liquidity stress event in 2025 moves at a speed that renders weekly monitoring insufficient for the metrics most sensitive to runoff.

That means the KRI program for liquidity has to be calibrated differently — faster measurement cycles on the most volatile metrics, real-time awareness of contingent funding line status, and explicit CFP tier mapping so escalation happens on a threshold, not a judgment call.

The Federal Reserve’s November 2025 Financial Stability Report characterized bank balance sheets as satisfactory overall, with elevated focus on liquidity. The FDIC’s 2025 Risk Review and OCC Spring 2025 Risk Perspective both flagged deposit competition and unrealized investment portfolio losses as ongoing monitoring priorities. The risk drivers that made 2023 dangerous haven’t disappeared — they require quantitative monitoring, not refreshed policy narrative.

The Eight Liquidity KRIs That Matter

These metrics cover four dimensions of liquidity risk: concentration, velocity, cost, and availability of backup. Together they give Treasury and Risk a working picture of whether the institution’s funding is stable, deteriorating, or accelerating toward a CFP event.

For each, the table below shows the metric, what it measures, a Yellow tier threshold range (elevated monitoring), and a Red tier threshold range (CFP activation). These ranges are illustrative starting points — calibrate against your institution’s specific funding model, historical data, and board-approved risk appetite.

1. Uninsured Deposit Concentration

This is the single most dangerous liquidity KRI, and it’s not because it’s hard to compute. It’s because concentration in uninsured deposits directly predicts runoff velocity when stress begins.

Deposits above the $250,000 FDIC insurance threshold carry higher runoff assumptions in any stress scenario — and require correspondingly larger High Quality Liquid Asset (HQLA) buffers to offset. The reason is straightforward: depositors above the insurance limit have direct economic incentive to exit an institution they’re uncertain about, because their principal is exposed. Wholesale uninsured deposits — institutional cash, corporate operating accounts, crypto-related balances — move faster than retail deposits for the same reason.

After 2023, the regulatory consensus is clear: an institution where more than 40% of deposits are uninsured requires elevated monitoring and a CFP that accounts for the nonlinear relationship between concentration and runoff velocity. Uninsured deposit growth resumed in 2024 for the first time since 2021 — which means concentration risk is actively re-accumulating across the industry.

Owner: Treasury / CFO
Measurement frequency: Monthly; weekly during elevated market stress

2. Deposit Runoff Rate (7/14/30-Day)

Deposit runoff rate measures the percentage of total deposits that actually exited over a given period. The 7- and 14-day windows are early-warning reads; the 30-day window is the standard stress test measurement horizon and connects directly to LCR methodology.

An increase in deposit runoff rate is often the first observable symptom of a stress event — before social media picks it up, before credit spreads widen significantly. Measuring at multiple time horizons lets Treasury distinguish between a transient outflow pattern (seasonal or large-account driven) and a structural acceleration. When the 7-day rate begins approaching the 30-day rate annualized, acceleration is underway.

Owner: Treasury
Measurement frequency: Daily; report weekly, aggregate monthly

3. Funding Cost Spread

Funding cost spread measures the difference between what the institution is currently paying for deposits and wholesale funding relative to its baseline cost structure. A widening spread is a market signal: counterparties and depositors are demanding a premium to maintain their relationship, which reflects perceived credit risk or competitive pressure.

The OCC Spring 2025 Risk Perspective specifically flagged deposit competition as warranting continued monitoring — the dynamic that drives funding cost spread is not purely idiosyncratic. Even institutions with stable fundamentals can face spread widening in a competitive deposit market.

Owner: Treasury / ALM
Measurement frequency: At each funding event; aggregate monthly

4. Wholesale Funding Renewal Rate

Wholesale funding — brokered deposits, FHLB advances, federal funds purchased, repo — carries higher runoff assumptions than retail deposits because the counterparties are sophisticated and the decision cycle is fast. When wholesale counterparties begin declining to renew, the institution loses access to the funding most likely to fill a gap quickly.

Renewal rate below 80% is a yellow flag; below 60% is a red flag and likely triggers CFP escalation. The practical concern is that a declining renewal rate often precedes — not follows — a funding crisis. Monitoring it provides lead time that other metrics don’t.

Owner: Treasury
Measurement frequency: At each maturity event; report monthly

5. Contingent Funding Line Utilization

Most contingency funding plans identify contingent funding sources — FHLB advances, Federal Reserve discount window access, committed credit facilities — as the backup funding stack. But access to those sources is not the same as available liquidity. Lines need to be pre-positioned, collateral needs to be eligible and pledged, and credit facilities can carry utilization limits or conditions.

Contingent funding line utilization tracks what percentage of available contingent capacity is already drawn. An institution that has drawn more than half its contingent capacity to meet normal operating needs has materially less headroom for a stress event than its CFP may assume.

This connects directly to CFP trigger frameworks — the trigger isn’t just “stress exists” but “the backup funding we planned on is already partially consumed.”

Owner: Treasury
Measurement frequency: Weekly

6. Failed Funding Test Results

This is the KRI most programs don’t track, and the one examiners are increasingly probing. A failed funding test occurs during a CFP tabletop exercise or live drill when a funding source the plan relies on cannot actually be activated within the assumed timeframe.

Common failures include: Fed discount window collateral that was identified as eligible but never pre-positioned; FHLB advances that require a member certification step nobody has completed; committed credit lines that lapsed or had conditions the institution can’t currently meet. The point of a CFP exercise is to surface these gaps — but only if the results are tracked as a KRI and failures trigger remediation.

One failed test in a 12-month period is a yellow flag: investigate and document the gap. Two or more is a red flag: the CFP’s stated backup funding capacity is materially overstated.

For a deeper look at how CFP exercises surface these issues, see the CFP fund flow testing post.

Owner: Treasury / CRO
Measurement frequency: Per exercise; track quarterly

7. Collateral Availability Ratio

Eligible HQLA and pledgeable collateral are only useful if they’re actually unencumbered. The collateral availability ratio measures eligible unencumbered collateral against currently pledged collateral — a ratio below 2.0x indicates limited buffer; below 1.25x indicates the institution is close to exhausting available collateral capacity.

Unrealized losses on investment portfolios reduce the effective value of collateral that might otherwise be pledged. The OCC and FDIC both flagged unrealized investment portfolio losses as an ongoing focus in 2025 guidance. An institution with a large unrealized loss position in its AFS portfolio needs to haircut its effective collateral accordingly.

Owner: Treasury / Investment Management
Measurement frequency: Monthly; stress-test quarterly

8. Net Cash Outflow Coverage

Net cash outflow coverage is the ratio of available liquidity (HQLA plus available contingent funding capacity) to projected net cash outflows over 30 days. It answers the most basic liquidity question: if stress begins today, can the institution cover 30 days of outflows without accessing external markets?

Coverage below 130% warrants elevated monitoring. Below 110% is a red-tier threshold for most institutions. This metric is structurally similar to the LCR — which generally applies to advanced approaches banks with $10 billion or more in assets — but the underlying logic applies to any institution monitoring liquidity exposure. Examiners at community banks and fintechs don’t require a formal LCR calculation, but they do expect institutions to have a clear picture of projected outflows against available resources.

This is the metric that early warning indicator frameworks are ultimately designed to protect — the ability to survive a stress window without relying on market access that may not be available.

Owner: Treasury
Measurement frequency: Monthly; weekly during stress

How These KRIs Map to CFP Tier Activation

The eight metrics above are most useful when they’re explicitly mapped to CFP tier thresholds. Without that mapping, a KRI breach produces a report. With that mapping, it produces an escalation.

Yellow tier (elevated monitoring): Any single metric in yellow range triggers weekly reporting to the designated CFP monitor — typically Treasury or the Liquidity Risk Manager — with documentation of the metric value, trend direction, and a preliminary assessment of whether the condition is transient or structural. Yellow events don’t activate the CFP, but they require documented management attention.

Red tier (CFP activation): Any single metric in red range — or two or more metrics simultaneously in yellow — triggers escalation to ALCO within 24 hours, with board notification for sustained or severe events. The CFP must specify which funding actions are authorized at red-tier activation and who holds that authority.

Correlated movement across multiple metrics is a more reliable stress signal than a single breach. Deposit runoff rate, wholesale renewal rate, and contingent line utilization deteriorating in the same week suggests structural stress, not a single-account event.

What Fintechs Get Wrong About Liquidity KRIs

Fintechs without traditional deposit bases often assume these metrics don’t apply — that they’re bank-specific. That assumption breaks quickly.

If a fintech relies on a bank partner, the bank partner’s liquidity condition directly affects the fintech’s operational continuity. A sponsor bank under funding stress may restrict services, change terms, or exit the relationship with short notice. And fintechs have their own funding vulnerabilities: warehouse line utilization is a structural analog to wholesale funding renewal rate; sponsor bank RFI frequency signals deteriorating cost dynamics; settlement concentration in a single payment rail is contingent liquidity risk under a different label.

The underlying dimensions — concentration, cost, availability of backup — are the same regardless of charter type.

For KRI examples that cover financial risk KRIs adapted for fintechs without deposit bases, the structure is identical: verifiable data source, real owner, calibrated thresholds, documented escalation.

The Regulatory Context You Can’t Ignore

OCC Bulletin 2023-15 updated the Liquidity Comptroller’s Handbook in 2023 — the primary examination reference for OCC-supervised institutions. The revision reflected post-SVB lessons: tighter expectations around CFP operational readiness, stress testing scenario severity, and backup funding source activation. An institution that can walk an examiner through each of the eight metrics above — with data, thresholds, owners, and CFP tier mapping — is demonstrating what the Bulletin requires.

For institutions that haven’t had a liquidity-focused examination since 2023: scenario severity expectations have shifted. 61% of banks surveyed increased LST scenario severity between 2023 and 2025. If stress scenarios are still calibrated to pre-SVB deposit runoff assumptions, they’re likely understating the speed at which concentrated uninsured deposits can exit.

So What?

If the eight KRIs above are being tracked, the question becomes whether the tracking is producing decisions or producing dashboards.

A liquidity KRI program that ends at a RAG status report is missing the point. The value is in the CFP tier mapping — each metric’s threshold is directly connected to a defined management action. When deposit runoff rate crosses 8%, specific people are notified and specific reviews are initiated. When contingent line utilization crosses 50%, the funding team is evaluating backup positions. When failed funding tests accumulate, remediation of the identified gaps is documented and tracked.

That’s the difference between liquidity monitoring and liquidity management. The metrics are the same; the difference is whether crossing a threshold produces action.

If building a complete KRI program from scratch — across liquidity and the other risk domains where similar gap exists — the KRI Library includes 132 pre-built KRIs covering financial, operational, compliance, cyber, vendor, and BSA/AML risk, each with calibrated thresholds, data source mapping, and escalation triggers. The 20 financial risk KRIs in the library cover the liquidity metrics above plus credit, capital adequacy, and concentration — structured for both bank and fintech contexts.

The point isn’t to have a KRI library. The point is to know, with specificity and evidence, when a funding problem is developing — before it’s obvious to everyone else.