Working Capital: The Money You Have That Isn't Really Yours
The median small business could survive 27 days if the money stopped arriving. That number is not a measure of wealth — it is a measure of how long you get to keep being wrong before it matters.
There is a specific flavour of vertigo that comes from looking at a bank balance you cannot spend.
Eleven thousand, four hundred and something. A good number, on a Tuesday, in an account belonging to a business that had just had its best quarter. And I remember doing the arithmetic on the back of a delivery note — rent on the 1st, two invoices at thirty days that were now at fifty-one, payroll on the 28th, a VAT bill I had been carefully not thinking about — and arriving at the conclusion that of eleven thousand four hundred pounds, roughly none of it was mine.
Not spoken for. Not earmarked. Not mine. The money was already someone else's; it was simply still, temporarily, in my account, the way a train is briefly in a station it has no intention of staying in.
That is the entire idea of working capital, and I had been running a business for two years without understanding it, because every account of it I had read began with a formula.
The definition that actually helps
Textbook first, because it's short and then we can leave it: working capital = current assets − current liabilities. Everything you'll turn into cash within a year, minus everything you must pay within a year.
The formula is correct and nearly useless, because it produces a single number at a single moment and tells you nothing about timing — which is the only thing that ever kills anyone.
The JPMorgan Chase Institute produced a far better instrument in 2016. Diana Farrell and Chris Wheat analysed 470 million transactions conducted by 597,000 small businesses from February to October 2015, across 12 industries and 367 metropolitan areas, and built their analysis around one idea:
“Cash buffer days are the number of days of cash outflows a business could pay out of its cash balance were its inflows to stop.
”
That reframing does something the formula can't. It converts a quantity into a duration — and duration is the unit your actual anxiety is denominated in. Nobody lies awake worrying about a current ratio of 1.4. They lie awake wondering how long they've got.
The headline finding:
“The median small business holds 27 cash buffer days in reserve.
”
Twenty-seven days. Half of all small businesses have less than a month.
The spread matters more than the median
Medians hide things. Here's the distribution, which is where the argument lives:
A 4.8× spread between the bottom quartile and the top. Two businesses can post identical revenue and identical profit and be in completely different states of danger, and nothing on the profit-and-loss statement will tell you which is which.
By industry, the pattern is structural rather than about competence:
Restaurants at 16 days. Real estate at 47. The restaurant owner is not more reckless than the estate agent — they are running a business where money leaves daily (staff, stock, spoilage) and arrives daily in small amounts, with almost no float. The structure of the industry sets the buffer before anyone makes a single decision, in the same way that structure rather than character explains most of what looks like individual failure.
Why 'we're profitable' is not an answer
Profit is a claim about a period. Cash is a claim about a Tuesday.
You can invoice £80,000 in March, book every penny as March revenue, be entirely correct under accounting rules, and still be unable to pay a £4,000 wage bill on the 28th because the money arrives in May.
Nothing has gone wrong. Nobody has lied. The business is profitable and insolvent at the same time, and those are not contradictory states — they're answers to two different questions that everyone has been trained to treat as one.
Where the money actually goes while you're not looking
The mechanism has a name — the cash conversion cycle — and three moving parts, each of which is a decision someone else made about your money.
Receivables. You did the work. The invoice says thirty days. It gets paid at fifty-one, because your customer is also managing their buffer days, and the cheapest credit available to any business is simply paying its suppliers late. Every day past terms is an interest-free loan you did not agree to make.
Inventory. Cash converted into things, sitting on a shelf, doing nothing until sold. It appears on the balance sheet as an asset, which is technically true and psychologically catastrophic — it feels like wealth and behaves like a hole.
Payables. The one lever pointing your way, and the reason the whole system persists: you're also paying your suppliers late. Everyone in the chain is financing themselves with everyone else's patience.
The Federal Reserve's 2026 Small Business Credit Survey — 6,525 responses, fielded 3 September to 14 November 2025 — found that rising costs of goods, services and wages remained the most common financial challenge, with 77% of firms reporting rising costs, tariff-related cost increases, or both. Rising costs hit the timing problem directly: the outflow side of the buffer calculation grows while the inflow side keeps its own schedule.
A statistic this piece deliberately does not use
Search for anything in this territory and you'll be told that "82% of small businesses fail because of cash flow problems." It's on hundreds of pages. It is usually attributed to a US Bank study that nobody ever links, and I could not locate a primary source for it.
It may well be roughly true. But an unsourced number repeated ten thousand times is still an unsourced number, and this piece would rather give you 27 buffer days from 470 million verifiable transactions than a rounder figure with no paper behind it.
Treat the 82% the way you'd treat any statistic whose citation chain terminates in another blog post.
The turn
Everything above is business writing, and this is The Ledger, so here's the part I actually think matters.
Your personal account balance is not your money either, and you already know this, and you have almost certainly never computed it.
You have current liabilities. Rent or mortgage on the 1st. The subscriptions that leave on dates you no longer remember agreeing to. The tax you owe but haven't been asked for yet — the single most dangerous item on any freelancer's balance sheet, because it doesn't feel like debt, it feels like money you already earned. The card balance. The thing you promised someone.
Subtract all of it from what's in the account. That remainder is your working capital, and it is almost never the number you think of as "what I have."
Then do the buffer-days version, which is the one that changes behaviour: divide what's actually available by your daily outflow. You'll get a number between about four and ninety. That number is not a measure of wealth.
It's a measure of how long you get to be wrong before it costs you something.
That's the whole thing. Buffer days are the distance between a bad month and a decision you can't take back — the reason one person can turn down bad work and another can't, why one can wait out a slow quarter and another takes the first offer. Neither is more disciplined. One has more days.
And this is why the number deserves a place in the personal ledger alongside everything else that quietly governs your options. It's also why buffer days behave so much like the debts nobody sends you a statement for: both are real, both compound, and neither appears anywhere you'd naturally look.
Computing yours
- 1
Write down every committed outflow for the next 30 days
Not what you expect to spend — what is already committed. Rent, loans, subscriptions, standing orders, the tax accruing on income you've already received. This list is always longer than people expect, and the first draft is always wrong in the same direction.
- 2
Subtract it from what's actually available
Available means today, not once an invoice lands. What's left is your working capital. If it's negative, that isn't a moral failure — it's the ordinary condition of a great many solvent people, and it's precisely the information the balance was hiding.
- 3
Divide by your daily outflow to get buffer days
Total 30-day committed outflow ÷ 30 = daily burn. Available cash ÷ daily burn = your buffer days. Compare against the median small business's 27. This single number will tell you more about your real position than any budget you have ever abandoned in February.
- 4
Attack the denominator, not just the numerator
Buffer days is a ratio, and almost everyone only ever tries to raise the top of it. Work an example, because the honest answer is more interesting than the tidy one.
Say you hold £3,000 against a £3,000/month outflow — £100/day, so 30 buffer days.
A one-off £2,000 takes you to £5,000 against the same £100/day: 50 days. A gain of twenty, immediately.
Cutting £200/month takes daily burn to £93.33. Your £3,000 now covers 32.1 days — a gain of only 2.1. On day one the windfall wins, and wins by a lot.
But the cut also retains £200 every month. By month ten it has returned £2,000 in cash and it is still lowering the denominator, permanently, every month after. The windfall is a step; the cut is a slope. Take the windfall if it's offered — just don't mistake it for the thing that changes your position.
- 5
If you invoice, treat payment terms as pricing
Thirty-day terms paid at fifty-one is a real cost — you financed a customer's operations for three weeks for free. Either price it in, or shorten terms, or charge for lateness. Basic bookkeeping makes the pattern visible; until you can see which customers habitually pay late, you can't charge the ones who do.
Isn't 'buffer days' just an emergency fund with a fancier name?
Closely related, and the objection is fair, but the framing does one thing an emergency fund doesn't: it's a ratio, so it responds to both sides. "Three months of expenses" is a target you either hit or don't, and it treats your outflow as fixed. Buffer days makes visible that reducing what leaves each month raises your resilience permanently, whereas saving raises it once. Same underlying money, materially different behaviour — and the JPMorgan data shows buffer days varying by industry structure, which an emergency-fund framing can't explain at all.
Is 2016 data still relevant a decade later?
Partly, and it's worth being precise about which parts. The specific medians are a snapshot of 2015 transactions and will have moved. What ages well is the structure: the 4.8× spread between quartiles, the industry pattern where restaurants sit near the bottom and real estate near the top, and the definition itself. Those follow from how the industries move money, not from 2015 conditions. Treat 27 as an anchor rather than a current reading, and compute your own.
If most businesses run on under a month of cash, isn't that just normal and fine?
It's certainly normal — that's what a median means. Whether it's fine depends on volatility, and this is where the Institute's own framing is more cautious than the cheerful reading: their conclusion was that most small businesses hold reserves that would be an insufficient cushion in a significant downturn. Twenty-seven days is survivable in a stable month and not survivable in a bad one, which is the same as saying most small businesses are fine until they aren't.
Doesn't this just make people anxious about a number they can't change?
Some of it isn't changeable — if you run a restaurant, the structure of the industry sets much of your buffer regardless of how well you operate, and pretending otherwise would be dishonest. But the piece points at the one lever most people never pull, which is the denominator. You may not be able to raise cash on demand; you can almost always reduce a committed monthly outflow, and that moves the ratio permanently. Anxiety about a number you can't compute is worse than knowing.
Eleven thousand, four hundred
I worked it out eventually, on the back of that delivery note, and the answer was somewhere around nineteen days.
Nineteen days of being wrong. That was the actual size of the business — not the revenue line I'd have quoted at a party, not the eleven-four in the account, but nineteen days of runway between an ordinary bad week and having to accept terms I'd otherwise have refused.
What changed wasn't the balance. I cancelled about £180 a month of things I had genuinely forgotten I was paying for, moved two customers to fourteen-day terms, and stopped treating the account as a scoreboard. Buffer days went to somewhere near forty over the following quarter — and forty days is not security, it's just enough room to say no to one bad contract, which turned out to be the only thing I actually needed it for.
The money in your account is a train in a station. The only question worth asking is how long before it leaves, and whether you'll still be standing there when it does.
It's just business.
Sources
- Diana Farrell & Chris Wheat, JPMorgan Chase Institute — 'Cash is King: Flows, Balances, and Buffer Days' (September 2016)
How this was checked
470 million transactions from 597,000 small businesses, February–October 2015, across 12 industries and 367 metro areas. Median 27 cash buffer days; 25th percentile 13, 75th percentile 62; restaurants 16, real estate 47; labor-intensive 23 vs capital-intensive 38.
- Federal Reserve Banks — 2026 Report on Employer Firms: Findings from the 2025 Small Business Credit Survey
How this was checked
6,525 responses, fielded 3 September – 14 November 2025. Rising costs of goods, services and/or wages the most common financial challenge; 77% of firms reported rising costs, tariff-related increases, or both.
Keep reading
This article is educational and satirical content from Business Dog. It is not financial, legal, or tax advice. It's just business.