You need to enable JavaScript to run this app.
By Travis Yule — CEO & Founder, Full Send Funding
Published 2023-01-09 · Updated 2026-09-23
The textbook definition is current assets minus current liabilities. The useful version is a number you can compute: daily operating outflow times your cash conversion cycle.
In one sentence: Working capital is mostly locked in the cash conversion cycle and cannot be spent; how much cash to keep is a separate, far smaller number — a floor in days of operating outflow plus the largest thing that can break, read at each month's lowest balance — and confusing the two is how a solvent business misses payroll.
Working capital is current assets minus current liabilities — the cushion that carries payroll, rent and inventory across the gap between money going out and money coming in. That is correct, and on its own it will not tell you one thing you can act on.
Two questions follow it and they have different answers. How much working capital does the business require? is about the operating cycle, and the answer is a large number that is not cash, cannot be spent, and is set by how long you wait to be paid rather than by how much you earn. How much cash should you keep? is about the balance in the account on the worst morning of the month, and the answer is a much smaller number you can read off four bank statements in ten minutes. Most owners run the two together, which is how a business carrying half a million dollars of working capital ends up one delayed cheque from missing payroll. This guide works both, with the arithmetic shown.
Working capital is current assets less current liabilities: cash, receivables and inventory, minus everything owed within twelve months. Net working capital is the same calculation under a second name, so ask which one somebody means. The refinements below are not academic — one balance sheet run through working capital, operating working capital, operating cash flow and free cash flow yields figures from $93,000 to $410,000.
Operating working capital strips out cash and short-term debt: receivables plus inventory, minus payables. It describes the operating cycle, because it measures what is tied up in doing the work rather than what happens to be in the bank on the day. A company can improve its working capital by borrowing long — a term loan adds cash and a non-current liability, so the headline number rises — while its operating working capital, and its real position, are unchanged. Drawing a revolving line does not even do that: the drawn balance is a current liability, so cash and current liabilities rise together and working capital does not move.
Surety-adjusted working capital is a third version worth knowing if you are bonded, and it is usually well below the book figure: retainage on distant jobs, aged receivables and related-party balances are discounted or excluded. Bonding capacity is sized off that adjusted number, typically at ten times for a single job, so a dollar of working capital is worth ten dollars of bidding room and a dollar lost costs the same.
Which one you are using matters because each answers to a different reader: a surety reads the third, a bank with a covenant reads the first, and a revenue-based funder reads none of them — a section of its own below.
The most expensive misreading here is treating that figure as money.
Take an ordinary distributor — Bellwether Supply, an illustrative business rather than a client: a plumbing and heating wholesaler, eleven employees, five years trading, depositing $315,000 a month. Its current assets are $46,000 of cash, $420,000 of receivables and $340,000 of inventory, which is $806,000. Its current liabilities are $250,000 of trade payables and $86,000 of accruals and card balances, which is $336,000. Working capital is $806,000 − $336,000 = $470,000.
Now sort that figure into the three pockets it is actually sitting in.
Locked. Receivables plus inventory less payables — $420,000 + $340,000 − $250,000 = $510,000 — is operating working capital, and none of it is available. It is stock on a rack and invoices in somebody else's accounts payable queue. It does not shrink while the business operates at this size, because each dollar collected is replaced by a dollar of new receivable the same week. This is a permanent commitment, not a balance.
Owed. The $86,000 of accruals is real, dated and mostly not negotiable — payroll tax, the insurance instalment, the card.
Free. The cash, $46,000 — the entire amount this business can spend on Tuesday.
So a business showing $470,000 of working capital has $46,000, and its operating working capital of $510,000 is larger than the headline figure rather than smaller, because the headline deducts $86,000 of accruals and adds back only $46,000 of cash. Both numbers are correct and neither is money. An owner who plans a purchase off the working capital figure is planning to spend stock.
A business needs working capital because of a sequence with a gap in it: cash goes out to buy material or run payroll, the cost sits as inventory or unbilled work, you deliver and invoice, and the customer pays some number of days later. The distance from the first step to the last, less however long your own suppliers let you wait, is the cash conversion cycle:
CCC = days inventory outstanding + days sales outstanding − days payables outstanding
Each term is a balance divided by a daily rate, and each is a lever worth more than most cost-cutting: a day removed from collection is a day of your own capital returned permanently, and it does not have to be re-earned next month. Which of the three dominates is business-specific rather than industry-specific — two contractors doing identical work can sit thirty days apart on collection alone, depending entirely on how fast they close out and invoice. The cycle guide derives all three components, prices what a single day is worth and works four businesses end to end; the cash conversion cycle calculator runs the arithmetic on your own balances.
The cycle is not the same fact as the cash position, and Bellwether shows why. On $3,780,000 of annual revenue — $10,356 a day — and 78% cost of sales — $8,078 a day — its $420,000 of receivables is 40.6 days, its $340,000 of inventory is 42.1 and its $250,000 of payables is 31.0, so the cycle is 40.6 + 42.1 − 31.0 = 51.7 days. The account still bottoms out below one day of outflow, and the next two sections are why those facts do not contradict each other.
Skip the rules of thumb expressed as a percentage of revenue. The requirement is close to:
Average daily operating cash outflow × cash conversion cycle in days
A business spending $10,000 a day on payroll, materials and overhead, with a 45-day cycle, needs roughly $450,000 in the business permanently. Not once — permanently, because as one cycle's cash comes back the next cycle's has already gone out.
The working capital a business must hold permanently is its daily operating outflow multiplied by the length of its cash conversion cycle. Every fifteen days of cycle costs this business $150,000 of committed capital.
Arithmetic, not a benchmark: daily operating outflow × cycle length. Substitute your own daily outflow to get your own number.
The chart is that multiplication and nothing else. Every fifteen days of cycle costs this business $150,000 of permanently committed capital — the real price of a slow close-out process, and the reason shortening the cycle is usually a better investment than financing it. The working capital requirement calculator runs the same arithmetic on your own outflow and cycle.
The approximation is good enough to act on. Bellwether's operating outflow is $3,528,000 a year, or $9,666 a day on a 365-day basis, and at the 51.7-day cycle that gives $499,700 against an exact operating working capital of $510,000 — within about 2%, on two numbers you can say out loud instead of three off a balance sheet.
Two properties of that number catch good operators. It does not depend on profitability — a business with excellent margins and a 45-day cycle needs the same working capital as a thin-margin business with the same cycle, because profit determines whether you can build that capital over time rather than how much of it you need. And it scales with revenue: double the volume and you double the daily outflow, so you double the requirement.
The requirement above is not an answer to this question; it describes what is already tied up. The question a bank balance answers is narrower and more urgent: how far can this account fall before something bounces, and how often does it get there?
A thirteen-week cash flow forecast is the right instrument when you are about to do something unusual — a large job, a season, a purchase — and building one properly takes a template and a week's discipline. For the standing question of how much to keep, you need no forecast: you need the four months of statements you would hand a funder anyway, and four readings from each.
Divide the second by the fourth and you have the number that matters: how many days of outflow the account was holding at its worst point. No accounting package prints it, and it is the one reading that describes the morning something goes wrong.
Bellwether's four statements show lowest end-of-day balances of $24,700, $31,200, $11,900 and $28,400, no negative days, and an average daily balance across the window of $46,000. Its operating outflow is $294,000 a month across 21 banking days, or $14,000 a banking day.
On the average balance it is holding $46,000 ÷ $14,000 = 3.3 days of outflow. At the worst point of the worst month it held $11,900 ÷ $14,000 = 0.85 of one day. Those two readings describe the same business four weeks apart, and only one of them is true when a customer pays late.
The target is a decision rather than a formula, and it has three parts.
The floor you will not go below. Name it in days of outflow, because dollars go stale and days do not. Bellwether's owner picks one week — five banking days, or 5 × $14,000 = $70,000. The worst month bottomed at $11,900, so the gap is $58,100, and at $21,000 a month of retained margin — deposits of $315,000 less outflow of $294,000, before owner distributions and tax — it takes 2.8 months to build if every dollar stays in.
The largest single thing that can break. Not a formula either. For Bellwether it is a delivery truck engine, about $14,000, one more day. For a restaurant it is a compressor; for a fabricator, a machine with a six-week lead time. Price yours and add it.
Nothing at all for the cycle. The $510,000 the cycle has locked up does not belong in this calculation and cannot be funded out of a buffer. Running the two together is how an owner concludes the target is impossible and stops.
That gives Bellwether $84,000 at the bottom of the month — $70,000 plus $14,000 — which is six banking days of outflow, and $72,100 above its $11,900 low, or under four months of retained margin. It is not a rule of thumb, and it is not a textbook answer. It is a number this business can state, measure monthly and actually reach. Floor and allowance together are what the thirteen-week forecast guide calls the minimum operating balance — the number its funding requirement is measured against — so if you build that forecast, the figure to use is $84,000, not $70,000.
For many businesses, a buffer sized to the worst normal month and an unbudgeted breakage and a season is not reachable out of retained margin in any useful timeframe. That is the honest position, and it has two answers.
The first is to shorten the cycle, which reduces the requirement rather than the buffer and costs nothing but process. The second is to hold a smaller buffer against a facility that is committed — the argument for a revolving line rather than a term product, and the reason an undrawn line is worth a non-use fee in a year you never draw it. A committed facility is a reserve somebody else holds for you; an offer you have not accepted is not, because the file behind it is re-underwritten on the day you need it, which is the worst day your statements will ever show. What does not work is holding the smaller buffer against nothing and calling it discipline.
Everything above changes the moment a fixed debit leaves the account every banking day, and not in the way most owners brace for.
Suppose Bellwether takes $420,000 — 133% of monthly revenue, inside the published 80% to 150% sizing range — on an eighteen-month term at a 1.35 factor, the pricing how much can I borrow works through for a file of this shape. Payback is $567,000 across 378 banking days, or $1,500 a banking day.
Run the two tests a funder runs. The remittance is $1,500 against average daily deposits of $15,000 — that is $315,000 ÷ 21 — for a remittance share of 10%, a light share by the measure that most often declines a file. Days of cover is the $46,000 average daily balance divided by $1,500, or 30.7 banking days, a deep cushion. On both capacity tests this file is nowhere near the line.
Now run the test neither of them performs. Across a 21-day month the remittance removes $31,500, and the business currently keeps $21,000 a month. So the $420,000 has to produce at least $10,500 a month of additional retained margin from the day it lands, or the difference comes out of the balance — and at the worst month's $11,900 low that is $11,900 ÷ $10,500, about five weeks, before a debit comes back. The $46,000 average would say nearly four and a half months, which is the average-versus-low misreading the section above warns against.
Whether it does is a question about the deployment, not the funding. If the $420,000 buys stock at the business's own 78% cost of sales, it sells for about $538,500 — $118,500 of gross profit each time it turns — and at four turns a year that is $473,800, or $39,500 a month, nearly four times the $10,500 the payment needs. If the same stock turns once a year it is $9,900 a month, and the answer is no. Both are margin, not cash, and the first half of this article is why that matters: stock sold on Bellwether's 40.6-day collection period becomes a receivable before it becomes money. Four turns add about $2,154,000 of annual sales, which carry about $239,000 of new receivables against about $142,000 of new supplier credit, so the deployment ties up roughly $97,000 more than the advance supplies — twice the $46,000 average balance — while the remittance starts the next banking day. Handling, storage and the extra hands take a further slice. Treat $39,500 as a steady-state ceiling, reached only once the receivables the growth creates have been funded, not as a number to bank from the first month.
That is worth saying plainly about our own product: the remittance-share test is a test of the deposits, not of the margin. It establishes that the account can carry the debit and does that well, but it cannot see what the money is going to do, because nothing in four months of bank statements can. The owner is the only party to the transaction who knows the turn rate.
So size the advance to cover the stock and the receivables it creates, against the incremental margin the deployment produces rather than the maximum the file supports, and rebuild the buffer out of that margin first — a daily remittance and a thin balance in the same account is what turns one slow week into a missed payment, cheap to handle early and expensive to handle late.
A profitable, growing business can run out of money, and the mechanism is the one above run forward. Every additional unit of work has its costs funded before its revenue arrives, so grow 40% and you need roughly 40% more working capital deployed, before the growth pays for itself, out of a balance sheet sized for last year. There is a rate above which a business cannot fund its own growth out of retained profit at all, and the sustainable growth rate, worked across a full year on a service contractor, is where that arithmetic is done properly.
The tell is specific: profit rising while cash falls, quarter after quarter. That is not an accounting error. It is working capital being absorbed by growth, and it is the most common way a healthy business gets into trouble.
Two show up constantly and they measure different things. Current ratio = current assets ÷ current liabilities; Bellwether's is $806,000 ÷ $336,000 = 2.40. Above 1.0 means current assets cover current obligations on paper — the crude version, which flatters a business holding slow inventory. Quick ratio, or acid test = (current assets − inventory) ÷ current liabilities; Bellwether's is ($806,000 − $340,000) ÷ $336,000 = 1.39. The gap between the two is the stock, $340,000 of it, and for an inventory-heavy business that gap is the informative part.
Both are snapshots, which is their limitation and the reason Bellwether can post a current ratio of 2.40 and hold less than one day of outflow at the bottom of a month. Both are also movable without changing anything real — a ratio above 1.0 rises when both sides shrink by the same amount, so paying down payables the week before a reporting date improves it. What a lender actually reads when the covenant certificate arrives is that problem.
The more useful fact is who still uses them. A bank lending against a balance sheet uses both and writes one into the agreement as a covenant; a surety uses the adjusted figure; a revenue-based funder sizes on neither: on most files it never sees a balance sheet, and above about $250,000, where it asks for year-to-date financials, the balance sheet is read for what is already owed, not for a ratio. What an underwriter here actually looks at is four months of business bank statements — deposit consistency, average daily balance, negative days and the debits already running — the cash conversion cycle showing up in the account rather than on a statement. The ratios are not wrong. A snapshot taken on the last day of a quarter simply cannot answer the question a daily remittance asks, and a statement window can.
Borrowing to bridge a timing gap is what working capital funding is for. A signed contract that pays in sixty days, a seasonal ramp, an inventory buy ahead of proven demand — the revenue exists and simply arrives later than the expenses. That is a solvable problem and capital solves it.
Borrowing to cover a structural loss — a business that spends more than it makes in a normal month — fixes nothing and adds a payment. The next shortfall is larger by exactly the new debt service, which is how a solvable operating problem becomes an insolvency over two or three rounds. Five situations where borrowing is the wrong answer, with the three-blank test that separates them is the long version; the short one is the test from the remittance section run backwards. Name what the money buys, name the margin it produces, and name the month that margin arrives. If all three fill in, it is a timing gap. If the first is "the shortfall", it is not.
One thing to be precise about, because both of these sentences are true and they are not the same sentence. A longer term costs more in total dollars is a fact about how an offer is priced: the factor rises with the number of banking days the money is out, and the eighteen-month 1.35 above is dearer in total than a nine-month 1.28 on the same amount. Stretching the remittance does not increase what you owe is a fact about the deal after signing, because the payback was fixed at signing and a reconciliation moves the daily figure rather than the total. Confusing them costs money in both directions — choosing a term by total dollars instead of by whether the daily figure clears your worst week, or refusing a reconciliation in the belief it will cost more.
Diagnose which one you have before you apply. An honest funder will ask, and if the answer is a structural loss they should say so out loud.
Working capital drains in four places, three of them internal. Receivables drift — customers stretching net-30 to net-60 is an interest-free loan you are making them, usually without a conversation. Unbilled work is the same money held up by your own close-out process: every day between the period ending and the invoice going out is a day of your own capital. Stock is cash you cannot spend, and slow stock consumes capital and then gets discounted. The fourth is contractual: retainage is far larger relative to profit than the percentage suggests, and gross charges overstate what a payer will actually pay, often by most of it.
Four moves come before financing, and none of them needs a lender. Measure the cycle in days per customer, because the aggregate hides the one account doing the damage and most businesses find one running two to three weeks longer than they believed. Invoice the day the work is billable. Attack the largest of the three terms — collection for a service business, usually inventory for a distributor, and a day of collection is worth more than a day of inventory by exactly the gross margin, because receivables are carried at the selling price and inventory at cost. And stop treating deposits and progress billing as optional. Each of those levers has a price — in margin, in a supplier relationship, or in nothing but process — and the cycle guide ranks them by what turning each dial costs. Only then finance the residual.
There is no benchmark here, deliberately. No public dataset reports receivable, inventory and payable days for businesses of the size we fund, and a number borrowed from a company a hundred times your size is worse than no number. The cycle guide sets out that reasoning with its sources.
Bellwether is arithmetic, not evidence. Every figure attached to it was chosen so that the results could be checked, and they follow from those inputs and nothing else. Your balances will differ; the method is what transfers.
The reserve target is a judgement, not an output. Five banking days of outflow is a number Bellwether's owner picked. No arithmetic produces the right floor: it depends on how lumpy your receipts are, how many customers you have and what breaks. What the method does is turn the question into one you can answer and re-measure every month, which is the difference between a policy and a hope.
Working capital is current assets minus current liabilities, and almost none of it is money. Sort it into what the cycle has locked up, what is owed on a date and what is free; only the third is spendable. The requirement is roughly daily operating outflow times the cycle in days, held permanently, independent of profit and scaling directly with revenue — which is why growth consumes cash and why profit rising while cash falls is the classic warning sign.
How much to keep is a smaller, different question, and four bank statements answer it: take the lowest end-of-day balance of each month and divide by your daily operating outflow. That ratio, not the average, is what you are running on. Pick a floor in days, price the largest single thing that can break, add them, and build to it out of retained margin. If the number is out of reach, shorten the cycle or hold the buffer against a committed facility — not against an offer you have not taken. And before financing a gap, establish that it is a timing gap and not a structural loss: a capacity test reads the deposits, and only you can read the margin.
We fund $5,000 to $10 million against monthly revenue, on terms from four months to three years, with decisions within 24 business hours. Estimate your range with the calculator, read how offers get sized, or check what you would qualify for in about a minute — no credit pull.
The cycle-cost chart multiplies $10,000 of daily operating outflow by the cycle length in days at each point shown, and nothing else; the $450,000 in the prose is the same product at 45 days, and the site's arithmetic audit recomputes every value before publication. The formulas are the standard definitions.
The article deliberately publishes no industry benchmark for the cycle. The public datasets that report receivables and sales by industry cover corporations far larger than the businesses this site serves and exclude several of the sectors it serves most, and the figures that circulate for smaller companies contradict one another by weeks. The ranges of collection days quoted are from the firm's underwriting of bank activity and are stated as practice.
Bellwether Supply is an illustrative business rather than a client, and every figure attached to it is a stated input: $315,000 of monthly deposits across 21 banking days ($15,000 a day), $294,000 of monthly operating outflow ($14,000 a banking day), $46,000 of cash, $420,000 of receivables, $340,000 of inventory, $250,000 of trade payables, $86,000 of accruals, 78% cost of sales, and month-low balances of $24,700, $31,200, $11,900 and $28,400 against an average daily balance of $46,000. Every other number follows from those — $470,000 of working capital, $510,000 of operating working capital, a 51.7-day cycle, a current ratio of 2.40 and a quick ratio of 1.39, 3.3 days of outflow on the average balance and 0.85 at the low, an $84,000 reserve target, and a $1,500-a-day remittance on a $420,000 advance at a 1.35 factor over eighteen months (378 banking days, $567,000 of payback). The firm publishes no cut-off for remittance share or days of cover — each is judged against the whole file, as the article on why applications are declined explains; this example's remittance share is 10% of average daily deposits and its days of cover is 30.7, both comfortable readings on any file. The five-day reserve floor, the $14,000 breakage allowance and the deployment's turn rate — four turns a year, or one — are choices the example makes, not rules; the deployment is priced at the business's own 78% cost of sales and its new receivables and supplier credit at the business's own 40.6 and 31.0 days.
Current assets minus current liabilities — cash, receivables and inventory, less everything owed within twelve months. It is the cushion that carries payroll, rent and materials across the gap between money going out and money coming in. The definition is easy; the useful question is how large that gap is in your business and how much cash it takes to survive it.
Approximately your average daily operating cash outflow multiplied by your cash conversion cycle in days. A business spending $10,000 a day with a 45-day cycle needs roughly $450,000 in the business permanently — not once, but continuously, because the next cycle’s cash goes out before the last one’s comes back. Rules of thumb expressed as a percentage of revenue are a poor substitute, because the cycle varies enormously between businesses in the same industry.
Days inventory outstanding plus days sales outstanding, minus days payables outstanding. It measures how long your own cash is tied up between paying for something and being paid for it. Each of the three terms is a lever, and a day removed from collection is a day of your own capital returned permanently — it does not have to be re-earned next month.
Working capital includes cash and short-term debt; operating working capital strips both out, leaving receivables plus inventory minus payables. The second describes the operating cycle more honestly, because a company can improve its headline working capital simply by borrowing long — a term loan adds cash and a non-current liability — while nothing about its actual position has changed; drawing a revolving line does not even do that, because the drawn balance is a current liability.
Because every additional unit of work has to be funded before it is paid for. Grow 40% and you need roughly 40% more working capital deployed, and you need it before the growth pays for itself, out of a balance sheet sized for last year. Profit determines whether you can build that capital over time; it does not reduce the requirement. The tell is profit rising while cash falls, quarter after quarter.
Above 1.0 means current assets cover current obligations on paper, and lenders generally want comfortable headroom above that. But both the current ratio and the quick ratio are snapshots, so a company can look sound on the last day of a quarter and be unable to make payroll on the fifteenth. That is why revenue-based underwriting reads bank activity across months — deposit consistency, average daily balance and negative days — rather than a balance sheet at a point in time.
In order of effect: measure your actual cycle in days, per customer where you can; invoice the day the work is billable, since internal delay is the cheapest capital available; attack whichever of the three terms is largest, which is collection for most service businesses and inventory for most distributors; negotiate supplier terms, the one term that works in your favour; and take deposits or progress billings rather than treating them as optional.
To bridge a timing gap — a signed contract that pays in sixty days, a seasonal ramp, an inventory buy ahead of proven demand. The revenue exists and simply arrives later than the expenses, and capital solves that. Borrowing to cover a structural loss, where a normal month spends more than it earns, fixes nothing and adds a payment, so the next shortfall is larger by exactly the new debt service. Diagnose which one you have before you apply.
Set a floor in days of operating outflow rather than in dollars, then add the price of the largest single thing that can break. Take total debits for a month, less financing payments and owner draws, divided by the banking days in it — that is your daily outflow. A business spending $14,000 a banking day that decides never to drop below one week is holding $70,000, plus perhaps $14,000 for one unbudgeted repair, so $84,000 at the bottom of the month. Two things make that number honest: it is measured at the month's low point rather than on average, and the working capital tied up in receivables and inventory is excluded from it, because that money cannot be held as cash.
Routinely, because working capital is not cash. A distributor with $46,000 of cash, $420,000 of receivables and $340,000 of inventory against $336,000 of current liabilities shows $470,000 of working capital and can spend $46,000. Receivables plus inventory less payables, $510,000, is operating working capital: stock on a rack and invoices sitting in somebody else's accounts payable queue. It is larger than the headline because the $86,000 of accruals exceeds the cash, and it does not shrink while the business keeps trading, because each dollar collected is replaced by a new receivable the same week. The balance sheet is correct. It is simply not a statement about what is in the account on the fifteenth.
The average is the mean end-of-day balance across the statement window; the low is the worst single end-of-day balance in it, and the two can be far apart. A business averaging $46,000 against $14,000 a day of operating outflow is holding 3.3 days on the average and, at a low of $11,900, less than one day at the bottom of the month. A funder's days-of-cover test reads the average, because that is the cushion a remittance is actually paid out of over time; a line of credit is read at the low point as well. An owner sizing a cash reserve should read the low, because that is the morning a late customer payment turns into a returned item.
Yes, until the money produces margin — and margin reaches the account only after what the money bought has been sold and collected. A remittance of $1,500 a banking day removes $31,500 across a 21-day month. If the business was keeping $21,000 a month before the advance, the funding has to add at least $10,500 a month of retained margin from the day it lands, or the difference comes out of the balance. Capacity tests do not see this: on that file a remittance share of 10% of deposits and roughly 30 banking days of cover both pass comfortably, because they measure whether the account can carry the debit rather than whether the deployment pays for it. Run the margin arithmetic yourself before signing.
Travis Yule worked in business funding before starting Full Send Funding in 2021, and has more than five years in business funding in all. He leads the company from Middle Grove, New York, and writes about working capital from the underwriting side of the table — what the numbers actually have to say before a business gets funded.