Profit versus cash

Why can a firm reporting its best profits ever fail to pay its suppliers next month?

A company announces record annual profits. Six weeks later it cannot meet its payroll and calls in administrators. The natural reaction is that somebody lied. Sometimes they did. But the collapse happens often enough with entirely honest accounts that fraud cannot be the general explanation.

So take the harder possibility seriously: the profit was real, correctly measured, audited without objection — and the firm still ran out of money. What kind of measurement can be truthful and yet say nothing about whether the wages will be paid?

Ask first what profit is a measure of. Suppose you sell a machine in December, deliver it in December, and the customer pays in March. When did you earn the money? Accrual accounting answers: in December, when you did the thing you were paid for. Revenue is recognised when the obligation is performed, and the costs of performing it are matched against it in the same period, regardless of when cash moved.

Why build it that way? Because the alternative is worse. Measure by cash alone and December collapses to nothing while March looks miraculous, and a firm can flatter any period by paying its suppliers late. Accrual accounting answers "did this year's activity create value?" — and for that question it is the right tool.

But notice what it therefore cannot answer. It is silent on when. And a business does not fail because value was not created; it fails because on a particular Tuesday a payment falls due and the account is empty. Profit measures the economics of the year. Solvency is a question about a date.

Now watch the gap open. Follow one unit through a manufacturer. Money leaves to buy materials; the materials sit as stock for two months; they sell, and profit is booked at that instant; the customer takes ninety days to pay. Cash returns five months after it left, and the supplier wanted paying at day thirty. Every one of those gaps is normal trading, and together they mean the firm funds a queue of unfinished business out of its own pocket, permanently.

Here is the part that surprises people: growth makes this worse, not better. Take a firm with a 10% net margin and a cash cycle of about a third of a year. Add 1,000 of annual sales. The profit on it is 100. The extra working capital needed to carry that business is roughly the fraction of a year it stays tied up multiplied by the sales it supports — about 0.33 x 1,000, call it 330. So the extra business consumes 330 of cash to generate 100 of profit, and the first-year effect is minus 230. (That is a deliberately crude estimate: stock and payables are properly measured on cost rather than on sales price, so the true figure is somewhat lower. The sign is what matters, and the sign does not change.) Double the growth and you double the hole. A firm can therefore be killed by success, and the trade has a name for it: overtrading.

The mismatch runs the other way too, which is a good check that the mechanism is timing rather than dishonesty. A business that bought its assets years ago charges depreciation today — a real cost, but no cash leaves — so it can report losses while generating cash comfortably. Conversely a firm that capitalises development spending records an asset instead of an expense, so profit stays high while the money is gone.

That gives a falsification test. If the mechanism really is timing, failures-while-profitable should cluster in businesses with long cycles and heavy working capital — construction, contract manufacturing, anything billed on milestones — and should be rare in businesses that collect before they pay, like supermarkets and subscription services, whose cash cycle is negative and who are funded by their own suppliers. If profitable firms failed at the same rate in a cash-on-delivery grocery chain as in a civil contractor, the timing account would be wrong and something else would have to explain it.

The tempting conclusion is "profit lies, cash tells the truth". That is too strong both ways. Over the whole life of a business, total profit and total net cash flow converge, because every accrual eventually settles. The divergence is timing within that life — which is precisely why it can be fatal in a single year and yet not be a falsehood.

And cash is not incorruptible. Delay supplier payments across the year-end, factor the receivables, defer maintenance, and the closing balance improves without anything real changing. What makes cash harder to fake is that these moves reverse — stretch your creditors this December and next December is worse. Profit is more susceptible to judgement and cash to short-term arrangement, and reading either alone is how people get surprised.

Who bears the cost is not evenly shared. Suppliers who extended trade credit are the ones left unpaid, and they usually had nothing but the published profit figure to go on when they granted the terms. Managers are typically paid on reported earnings, which points their effort at the accrual number. Banks lend against covenants written mostly on profit-based ratios. Almost every party with power is watching the measure that goes silent about the date — which is why the cash flow statement was eventually made compulsory rather than optional.

• How the cash conversion cycle is computed from published accounts, and what a negative one implies about who finances whom.
• Why revenue recognition on long-term contracts is where honest accounts and optimistic ones diverge most.

Every term the collection defines is gathered in the glossary.