Liquidity is a timing problem, not a ratio
A liquidity ratio tells you how much. Institutions fail on when. The gap that matters is on a specific day, against obligations that do not reschedule themselves.
Liquidity gets managed as a ratio in most institutions because a ratio is easy to compute, easy to report monthly, and easy to compare. It is also structurally incapable of describing the thing that actually goes wrong.
A ratio is a stock measure. It answers how much. Liquidity failure is a flow event: obligations landing on a particular day, against sources that either arrive by that day or do not. A bank can hold a comfortable ratio and still be unable to meet a Tuesday, because the assets are real but slow and the outflow is real and fast.
Build the ladder before the ratio
The right primitive is a maturity ladder: sources and uses placed into time buckets, so that what you read is a gap per bucket rather than a single number. Overnight, one week, thirty days, ninety days, one year is a reasonable set for most community institutions.
| Bucket | What lands here | What actually counts as a source |
|---|---|---|
| Overnight | Settlement, same-day outflows | Cash above the operating floor, and nothing else |
| 1 week | Maturing wholesale funding, large depositor moves | Unpledged securities that can settle inside the window |
| 30 days | Deposit attrition, draws on committed lines | Contingent borrowing capacity you have actually tested |
| 90 days | Contractual maturities | Asset sales at a price you would accept |
| 1 year | The structural funding gap | Anything needing a market that may not be open |
Uses are the contractual and behavioural outflows in each bucket: maturing wholesale funding, expected deposit attrition, draws on committed lines, operating expense. Sources are what you can actually raise inside the bucket. The discipline is entirely in that word actually.
What is not liquidity
Three things routinely get counted that should not be.
Pledged collateral. A securities portfolio that is already pledged against public deposits or borrowings is not a source. This sounds obvious and is one of the most common errors in practice, because the securities schedule reports the portfolio and the pledging is tracked somewhere else, and the two do not get netted before the number goes into the board packet.
Operating cash. The cash on the balance sheet is not all available. A bank has to open in the morning: vault, settlement, clearing, payroll. There is a floor beneath which the balance cannot go without the institution ceasing to function, and only the amount above that floor is a liquidity source. Institutions that skip this are overstating their most immediate bucket, which is the bucket that matters most.
The same asset twice. An unpledged security cannot simultaneously be the collateral behind your contingent borrowing capacity and a separate line of available securities. It is one asset. Double-counting here is easy, because the borrowing capacity is computed by one process and the securities inventory by another.
Uninsured deposits are the fast money
On the outflow side, the assumption that decides the answer is deposit behaviour, and the useful split is insured versus uninsured. Insured deposits are historically sticky through stress, which is the entire point of the insurance. Uninsured balances are held by people and businesses with a live incentive to move and, increasingly, the means to move them in an afternoon.
The estimate of uninsured deposits sits in the Call Report, in the schedule built for deposit insurance assessment, which is why it is rarely read outside the assessment process. It belongs in the liquidity ladder as the driver of the near buckets, with an attrition assumption applied to it that is stated, defended, and considerably more severe than the insured assumption.
The question is never how much liquidity you have. It is how much of it arrives before the obligation does.
The contingency funding plan is an ordering, not a document
Most contingency funding plans list the available sources. The useful version puts them in order and attaches two facts to each: how long it takes to draw, and what it signals.
Speed is obvious once written down: some sources are same-day and some take a week you may not have. The signalling cost is the part that gets omitted and it can dominate. Sources drawn late in the sequence are frequently the ones that tell counterparties, correspondents and eventually depositors that the institution is under pressure. Using them can accelerate the outflow they were meant to cover.
An ordered plan, with a trigger for moving to each next rung and an honest note on what that rung costs in confidence, is a usable document. An inventory of sources is a compliance artifact.
Stock and flow
The two measures answer different questions and the difference is not subtle. A ratio answers how much. A ladder answers when.
Institutions are examined, reported and compared largely on the first, which is a stock measure, computed monthly, easy to benchmark. Failure happens on the second, which is a flow, arrives on a particular morning, and does not benchmark against anything.
That mismatch is not an oversight by anyone in particular. It is what happens when the thing that is easy to measure and the thing that actually breaks are not the same object, and it is worth noticing how often that shape recurs elsewhere.
← Back to all insights