How do banks decide whether a working capital need is temporary or permanent?

By reading the shape of the borrowing rather than the reason given for it. A temporary need rises and falls: the facility is drawn as stock is bought or invoices are raised, and it clears when the cash comes back. A permanent need never clears — the balance has a floor it has not been below in two years, and the floor keeps rising. The clean-down test names that pattern, but it does not explain it. What explains it is the cash conversion cycle, the trend in debtor and stock days, whether reported earnings converted to operating cash, and what was taken out of the business. Those four resolve the question to one of four different answers.

What is a clean-down covenant and why do banks use it?

A clean-down requires the working capital facility to be reduced to zero, or to a nominated low level, for a continuous period each year — often a fortnight or a month. It is the simplest available lie detector for the temporary-or-permanent question: a genuine trading facility empties itself once a cycle, so the covenant costs a healthy borrower nothing and is impossible for a borrower financing something permanent. When a business asks for the clean-down to be waived, the request is usually more informative than the financials attached to it, because it is the borrower telling the bank which kind of need this is.

Is permanent working capital a bad thing?

No. A growing business permanently carries more working capital than a smaller one did, and that is arithmetic rather than mismanagement — every extra dollar of sales carries its share of the cash cycle. The problem is never that the need is permanent; it is that permanent needs financed on a revolving facility never amortise, so the exposure is renewed indefinitely on a product designed to be repaid every cycle. The honest answer is usually to term out the permanent portion and keep the revolver for the part that genuinely fluctuates. That conversation is easier when the borrower raises it.

What happens if a bank finds the facility is funding losses?

It becomes a different conversation, and one that a larger limit does not improve. A facility that has been absorbing operating losses is financing a gap that trading is not going to close, so an increase buys time at the cost of a bigger eventual problem — which is why an experienced desk will decline the increase as framed even when the security looks adequate. What follows is usually a restructure: what the business needs is capital, a change to the cost base, or both. The bank's role at that point is to be clear early rather than accommodating twice.

// FOR BORROWERS AND BANKERSTHE QUESTION UNDER THE REQUEST

Is a Working Capital Need Temporary or Permanent?

A facility that never clears is not a working capital facility. It is term debt that nobody has scheduled.

BY MICHAEL SHANG · SIXTEEN YEARS IN COMMERCIAL BANKING · UPDATED 2026-09-05

// THE SHORT ANSWER

A working capital facility exists to bridge a gap that closes, so the question under every increase request is whether the money comes back on the cycle. The clean-down covenant detects a facility that never clears but cannot explain it. What explains it is the shape of the balance across a year, the cash conversion cycle against sales, whether reported earnings converted to cash, and what capital expenditure or distributions left the business. Those resolve to four findings — a genuine seasonal need, structural growth, fixed assets funded on the revolver, or operating losses — and each leads to a different facility rather than simply a different answer.

BASIS Practitioner judgment from sixteen years in commercial banking, set out in How Bankers Think (Highbank Press, 2026). Every borrower on this site is a composite constructed for teaching, not a client. The market is mid-market commercial lending as practised in New Zealand and Australia; where a convention differs elsewhere, the page says so.

// 01 — WHY THIS IS THE FIRST QUESTION

A working capital facility exists to bridge a gap that closes. Stock is bought, goods are sold, the customer pays, the facility clears, and the cycle begins again. Every feature of the product — the revolving limit, the pricing on the drawn balance, the clean-down — is designed around a balance that returns to zero.

So when a request arrives to increase the limit, the question underneath it is not whether the business is sound. It is whether the money being asked for will come back on the cycle, or whether the bank is being asked to fund something that will still be there in five years wearing a twelve-month facility. Getting that wrong is the most common structural failure in commercial lending, and it is usually made in good faith on both sides.

// 02 — THE CLEAN-DOWN IS ONE TEST, NOT THE TEST

The clean-down covenant is the best single instrument for this question and deserves its reputation: require the facility to sit at zero for a continuous period once a year, and a genuine trading line clears itself while a disguised term loan cannot. It costs a healthy borrower nothing, which is why the request to waive it is often more informative than the accounts attached to it.

But it only detects. It does not explain. A facility that fails its clean-down could be funding seasonal stock that arrived late, growth the business has genuinely achieved, a piece of plant bought without asking, or losses nobody has named. Those four require four different responses, and the covenant cannot tell them apart. What follows is what does.

// 03 — THE EVIDENCE THAT SEPARATES THEM

The shape of the balance, not its level. A year of utilisation says more than any ratio: whether the line breathes or only inhales, whether the low point each year is falling, holding or rising. The floor is the permanent portion, and it is visible without asking anyone.

The cycle in days, against sales. If the floor rose and the cash conversion cycle held while revenue grew, growth absorbed the money and the business is bigger. If the floor rose because the cycle lengthened, the money went into slower collection or heavier stock, and that is a different problem with a different fix.

Whether earnings became cash. Reported profit alongside operating cash flow, over several years. A business that is profitable and whose facility floor keeps rising is telling the lender that the profit is being consumed somewhere between the invoice and the bank account.

What left the business. Capital expenditure in the period, and distributions. A revolver floor that rose by roughly the amount of an unfunded asset purchase has answered the question by arithmetic. So has one that rose by roughly what was drawn out.

// 04 — THE FOUR ANSWERS

The evidence above resolves to one of four findings, and each leads somewhere different.

01
A genuine seasonal need
The balance peaks and clears on the same rhythm every year, and the peak has grown roughly in line with sales.
ANSWER Keep the revolving facility, sized to the peak rather than the average, with the clean-down intact.
02
Structural growth
The floor rises each year, but so does revenue, and the cycle in days has held. Earnings still convert to cash.
ANSWER Resize the facility to the larger business, and decide whether the increment above the old floor belongs on a revolver at all.
03
Fixed assets on the revolver
The floor rose without the cycle or the sales explaining it, and capex in the period is the missing amount.
ANSWER Term out what bought the asset, over a life that matches it, and hand the revolver back its actual job.
04
Operating losses
The floor rises while revenue does not, earnings do not convert, and the accumulated position has been eroding.
ANSWER Not a facility question. Capital, cost base, or both — and an increase as requested funds the gap rather than closing it.

Only the fourth is a decline, and even that is a decline of the request rather than the borrower. The other three are all a facility — just not always the one that was asked for.

// 05 — THE FILE WHERE THE TWO ARE MIXED

The hardest version is not a request that is wholly permanent. It is a seasonal request with a permanent piece hidden inside it. Issue 06 — Kowhai Retail Group is that file: last Christmas did not clear, this year's buy is larger, and the seasonal uplift request contains stock that has already failed to sell once. Part of the ask is genuinely seasonal. Part of it is last season being refinanced under this season's name.

For the two ends of the range, the published pair works the same arithmetic to opposite answers: Issue 01 where the request is catching up with absorption already in the numbers, and Issue 04 where the cycle held and the answer is a fast, full yes.

// 06 — WHY A BORROWER SHOULD WANT THE RIGHT ANSWER

Borrowers often argue for the temporary reading because it is the easier approval. It is also the worse outcome when it is wrong. A permanent need financed on a revolver is reviewed every year, repayable on demand in substance, and priced as though it fluctuates. Termed out, the same money amortises on a schedule the business can plan around and stops being a renewal risk.

The borrower who arrives having already separated the two — this much fluctuates, this much is permanently in the business, here is what happened to it — is asking a question the bank can answer quickly. That is worth more than the extra limit that a vaguer request might have won.

// QUESTIONS PEOPLE ASK

How do banks decide whether a working capital need is temporary or permanent?: By reading the shape of the borrowing rather than the reason given for it. A temporary need rises and falls: the facility is drawn as stock is bought or invoices are raised, and it clears when the cash comes back. A permanent need never clears — the balance has a floor it has not been below in two years, and the floor keeps rising. The clean-down test names that pattern, but it does not explain it. What explains it is the cash conversion cycle, the trend in debtor and stock days, whether reported earnings converted to operating cash, and what was taken out of the business. Those four resolve the question to one of four different answers.
What is a clean-down covenant and why do banks use it?: A clean-down requires the working capital facility to be reduced to zero, or to a nominated low level, for a continuous period each year — often a fortnight or a month. It is the simplest available lie detector for the temporary-or-permanent question: a genuine trading facility empties itself once a cycle, so the covenant costs a healthy borrower nothing and is impossible for a borrower financing something permanent. When a business asks for the clean-down to be waived, the request is usually more informative than the financials attached to it, because it is the borrower telling the bank which kind of need this is.
Is permanent working capital a bad thing?: No. A growing business permanently carries more working capital than a smaller one did, and that is arithmetic rather than mismanagement — every extra dollar of sales carries its share of the cash cycle. The problem is never that the need is permanent; it is that permanent needs financed on a revolving facility never amortise, so the exposure is renewed indefinitely on a product designed to be repaid every cycle. The honest answer is usually to term out the permanent portion and keep the revolver for the part that genuinely fluctuates. That conversation is easier when the borrower raises it.
What happens if a bank finds the facility is funding losses?: It becomes a different conversation, and one that a larger limit does not improve. A facility that has been absorbing operating losses is financing a gap that trading is not going to close, so an increase buys time at the cost of a bigger eventual problem — which is why an experienced desk will decline the increase as framed even when the security looks adequate. What follows is usually a restructure: what the business needs is capital, a change to the cost base, or both. The bank's role at that point is to be clear early rather than accommodating twice.

// RUN IT ON YOUR OWN NUMBERS

Twelve numbers from two years of financials, and the senior banker's written read on whether the growth is real, the cash converts, and the suppliers are carrying the gap — free, no signup, no data stored.

Open the calculator →