A cashflow forecast shows a cash shortfall months before it happens. To read one, you need to know which four or five figures to look at.
What a cashflow forecast shows
A forecast isn't trying to predict your profit. Profit and cash are different things, and it's entirely possible to be profitable on paper while running out of money in the bank — because cash gets tied up in stock, in unpaid invoices, or in a tax bill landing at an awkward moment. A cashflow forecast is about timing: when money is expected to move, so you can see the gaps before they open up.
That distinction has a practical consequence. A profitable month can be a terrible cash month, and a loss-making month can look fine in the bank. Neither statement is wrong; they are measuring different things.
The four numbers to look at first
Ignore the detail for a moment and find these:
- Opening balance — what's in the bank at the start of the month.
- Closing balance — where the month is expected to end.
- The lowest point within the month: whether the bank balance will cover every payment that month.
- Your facility headroom — the overdraft or funding line you could draw on, and how much of it is already used.
A worked example
Take an agency with a 30 September VAT quarter end, opening October with £42,000 in the bank. October's expected receipts are £58,000, of which £36,000 is a single large client who pays on the 28th. Payments due are rent of £3,500 on the 1st, the VAT bill of £11,400 on 7 November, supplier payments of £18,000 between the 10th and the 20th, and payroll of £31,000 on the 28th.
A quarter ending 30 September is due one calendar month and seven days later, so it leaves the bank on 7 November, more than a month after the quarter closed. Get that wrong in the forecast and October looks £11,400 worse than it is while November looks fine. Timing errors like this are the single most common fault in a homemade forecast.
With the VAT correctly placed in November, October's closing balance is £42,000 plus £58,000 in, less £3,500 rent, £18,000 suppliers and £31,000 payroll — a closing balance of £47,500. That reads as a comfortable month.
Now look at the low point. On the morning of the 28th, before the big client pays, the position is £42,000 plus the other £22,000 of receipts, less £3,500 and £18,000 — that's £42,500. Payroll of £31,000 goes out the same day, leaving £11,500. Still fine, but a long way from £47,500. And if that client pays eight days late, as large clients routinely do, you are running the last week of October on £11,500 with the next payroll five weeks out. The £11,400 VAT payment on 7 November would then fall due before that money arrives.
Reading the trend as well as the total
A single month tells you very little on its own. What matters is direction over six to twelve months. Is the closing balance broadly stable, climbing, or falling? A forecast showing cash shrinking by £3,000 a month is telling you something important even though no individual month looks alarming. Seeing it months ahead leaves time to chase customers, delay a purchase or arrange finance.
It also helps to separate the forecast into three parts: income already invoiced, income you're reasonably confident of but haven't billed, and fixed outgoings that happen regardless of trading. That split shows how much of the forecast is certain and how much is assumption.
Tax payment dates
The lumps that break forecasts are nearly always tax, and the dates are entirely knowable in advance. Set them out on the actual days they leave the bank:
- VAT — one calendar month and seven days after the quarter end, so a 31 March quarter is paid by 7 May.
- PAYE and National Insurance — the 22nd of the following tax month if you pay electronically, the 19th by post.
- Corporation Tax — nine months and one day after the accounting period end, so a 31 March year end is paid by 1 January.
- Self Assessment — 31 January for the balancing payment and the first payment on account, and 31 July for the second. If you draw dividends, January carries your Self Assessment balancing payment and first payment on account, so put both in the forecast.
If you came into Making Tax Digital for Income Tax on 6 April 2026, with qualifying income above £50,000, the quarterly updates do not move cash. They do keep your figures current four times a year. That makes an accurate rolling forecast considerably easier to maintain.
Where forecasts usually go wrong
Update the forecast every month after you build it. One built in January is only as accurate as January's assumptions, which are usually wrong within a few weeks. Compare the forecast against what happened every month.
The second failure is optimism on timing. If your terms are 30 days and your customers pay in 47, forecast 47. Take your last six months of invoices, work out the real average gap between invoice date and payment date, and use that number in the forecast.
The third is forgetting that VAT you've collected isn't yours. On a £58,000 month of standard-rated sales, roughly £9,667 of that is VAT you are holding on HMRC's behalf.
Who should be looking at it
The forecast should be read by whoever makes the day-to-day calls, such as whether to place a large order, whether to hire, or whether a supplier payment can wait a week. You do not need to build it yourself, but you need to understand it well enough to act on it.
Set a standing half hour each month: open the forecast, compare last month's prediction to what happened, ask why the difference arose, and roll the window forward.
Our management accounts include rolling cashflow reporting alongside profit and loss each month. Our cashflow and budgeting page covers how we set it up, and the cashflow forecasting guide walks through building one.
