What a bank looks for in a cash-flow forecast

Start with how a credit analyst reads

I spent years reading forecasts that companies sent to investors, first as an equity analyst and then as a fund manager. A bank's credit team reads in much the same way, with one difference: they are not looking for upside. They want to know whether the money comes back. They read in a set order and stop the moment something does not tie up. This is what they check.

Monthly columns, not annual totals

An annual figure tells a lender almost nothing. A business can make £40,000 of cash in a year and still run out of it in March. The credit team wants monthly columns for at least twelve months, and for a term loan the full first year of repayments. For an overdraft, run it far enough to show the balance recovering after the low point.

The opening balance ties to the bank statement

The first number I ever checked on any forecast was the opening one. If it did not agree with the last reported figure, I stopped and asked why. Your opening cash should match the closing balance on your most recent business bank statement, and the statement should be in the pack. If there is a gap between the statement date and the forecast start, show the bridge. A forecast that opens with a round number and no statement behind it has told the reader how the rest was built.

The low point, and when it falls

Next they find the lowest closing balance and the month it lands in. That is the peak funding need, and it is the number the facility has to cover. They also want to know why it falls there. A dip in month four because a van is bought and a VAT bill lands is one thing. A balance that gets lower every month is a loss-making business asking the bank to fund the losses. Make the cause obvious.

Headroom against the facility

If you are asking for a £30,000 overdraft and the forecast touches minus £29,800 in month seven, you have a problem, and so does the person approving it. Lenders want margin between the worst month and the limit, because forecasts are optimistic by nature. Show the limit as a line and the headroom in each month. If it is thin, say what you would do about it rather than hoping nobody notices.

The two numbers a credit analyst reads first: the lowest closing balance, and the month it falls in.

Debtor days and retentions

Sales are not cash. If your terms are 30 days, do not have every customer paying in 30 days, because in my experience they do not. Use your actual debtor days from the last year; if you are new, use the sector norm and say so. Construction and fit-out businesses need to show retentions: the slice of each certified payment the client holds back, part of which may not arrive until the end of the defects period, often a year after the job is finished. Book retention money when it arrives, not when it is invoiced.

VAT quarters and tax dates

If you are VAT-registered, the return and payment are due one calendar month and seven days after the end of each quarter, so a quarter ending 31 March means money leaving on 7 May. A limited company pays corporation tax nine months and one day after its year end. If you employ staff, PAYE and National Insurance go to HMRC by the 22nd of the following month. Sole traders have self-assessment payments on 31 January and 31 July. Put each one in the month it actually leaves the account.

Loan repayments included

This sounds obvious. It is missed constantly. The forecast must include the repayments on the loan you are applying for, starting the month after drawdown, plus every existing commitment: hire purchase, asset finance, director loans being repaid. A forecast that shows the loan arriving and never shows it going back is not a forecast of the business with the loan. It is a forecast of the business with a gift.

A downside case

A single set of numbers invites the question "and what if you are wrong?" One extra sheet with sales 15 to 20 per cent lower, or customers paying a month later, and a line showing whether the facility still covers it. If it does not, show the action you would take. Credit teams do not expect the downside to look comfortable. They expect you to have looked at it.

Assumptions written down

One page. Every line should trace back to a stated assumption: price, volume, debtor days, creditor days, wage bill, VAT status, loan terms. The analyst will test two or three of them. If they can find them quickly and the logic holds, they trust the rest. If they cannot, they assume the worst.

What gets a forecast binned

• An opening balance that does not match the bank statement.
• No repayments on the loan being requested.
• Sales growth with nothing behind it, such as a 40 per cent jump in month three because "marketing kicks in".
• No VAT, PAYE or corporation tax lines.
• Totals that do not add up. Yes, still.
• Annual columns only, or a forecast that stops at the low point.
• An owner who cannot explain their own numbers on the phone.

None of this guarantees a yes. Nobody can promise a lending decision, and anyone who does is selling something. What a well-built forecast does is get you a fair hearing from someone whose job is to find reasons to say no.

Common questions