"Profitable on Paper, Broke on Friday" — Why Construction Cashflow Breaks Most £1m+ Firms

A construction Ltd director rang me last spring. £1.2m turnover, six months into the year, P&L showing £180k profit. His bank balance that Friday morning was £6,400. Payroll for two PAYE staff and four subbies was due to leave the account by 5pm. He needed £18,000 he didn’t have.

Nothing was wrong with his accounts. The profit was real. The bank balance was real. They just weren’t talking to each other. This is the call I take most often from £1m+ construction Ltd directors, and it almost always lands in the same week of the month — the week before the VAT bill, the week after a big retention should have been released, the week the main contractor pushed his valuation back a fortnight.

QUICK ANSWER

A 13-week rolling cashflow forecast shows you, week by week, what’s coming into and leaving your bank account for the next quarter. For a construction Ltd, the gap between profitable on paper and broke on Friday is built into the industry — retentions, CIS deductions, the application-versus-valuation gap, the lag between certified work and paid invoice. Forecast it weekly and the Friday-payroll moment stops being a surprise.

Why your annual P&L is lying to your bank account

Construction is the only industry I work in where a profitable year-end set of accounts can sit next to an empty bank account, and both can be accurate at the same time. The P&L records what you’ve earned — invoiced or about to invoice. The bank records what’s cleared. In most trades those two numbers stay roughly in step. In construction they drift apart, and they drift apart in five specific ways.

Application versus valuation. You apply for £80,000 in stage payments. The main contractor’s QS values it at £72,000 and pushes the rest into next month. Your P&L books the work you’ve done. Your bank gets the £72k — minus retention — and the rest sits in an applications-pending column nobody’s funding.

Retentions on someone else’s balance sheet. Most main-contractor terms hold back 5% of every certificate, typically split into 2.5% released on practical completion and 2.5% twelve months later for defects. On a £1.2m turnover business that’s £60,000 of yours, parked in someone else’s account, earning them interest while you’re working-capital poor on the next job. The P&L counted it as revenue the day the certificate was raised.

CIS suffered, sitting at HMRC. Every subcontractor invoice you send a contractor gets 20% deducted at source (30% if you’re not verified, 0% with Gross Payment Status). That cash never reaches your bank. It sits with HMRC as a credit against your tax bill — or it offsets monthly against your PAYE, but only if your accountant is actually running that offset every month.

DRC making VAT cash-neutral but hiding the timing. Since March 2021 the Domestic Reverse Charge has meant CIS-registered B2B subcontract supplies move VAT-free in the invoice itself. Your customer accounts for the VAT. Net cash effect on you: zero. But it also removes the 20% VAT cushion that used to sit in your account between issuing an invoice and paying HMRC. That cushion was working capital. It’s gone.

Subbie terms tighter than client terms. Most subbies want paying within 14 days of the PDS. Most main contractors pay you 30 days from valuation, sometimes 45, sometimes 60. The arithmetic between those two numbers is the size of the overdraft you’re going to need.

The annual P&L can’t see any of this. It tells you whether you made money. It can’t tell you whether you’ll make payroll on Friday.

What goes into a 13-week construction cashflow forecast

Two columns of money, thirteen weeks across the top. Every line item is dated to the week the cash moves, not the week the work happens.

Inflows worth tracking weekly

• Certified applications. Each main-contractor relationship, with the value certified and the date the cash should clear. Probability-weight the ones not yet certified.
• Retention releases. Every retention held against you, with the contract date it’s due back. Most slip — model 1–2 weeks late as the base case.
• CIS refund or monthly PAYE offset. If your accountant is offsetting CIS suffered against your monthly PAYE liability via the EPS submission, that lands as a reduced outflow on the 22nd. If you’re waiting on the full-year surplus, it lands as a HMRC payment in spring after the EPS year-end.
• Plant hire income, R&D credits, insurance claims, other. Anything else committed.

Outflows worth tracking weekly

• Subcontractor payments. Gross of CIS deduction, dated to the week you actually pay each one. CIS sits as a separate line because the gross outflow leaves your account; the deduction goes to HMRC the following month.
• Gross PAYE plus employer NI. Due 22nd of the following month if you pay electronically, 19th if you pay by post. Falls a working day earlier if the 22nd is a weekend.
• VAT. One calendar month and seven days after the end of the period. So a quarter ending 31 March is due 7 May.
• Corporation tax. Nine months and one day after the year-end for companies with taxable profits up to £1.5m. Forecast it; never let it surprise you.
• Materials, plant hire, fuel, finance and HP, rent and admin, dividends, director loan repayments. All dated to the week the standing order or invoice falls.

The 2026/27 HMRC payment dates worth pinning to your forecast:

We file these monthly for our construction clients — see the CIS returns service and the VAT returns service for how that fits with the forecast.

How to build the forecast — a Monday-morning workflow

A 13-week forecast you build once and never touch again is worse than no forecast — it gives you false confidence. The whole point is the rolling update. Half an hour every Monday morning, never more.

Week 1: the starting bank position and the 13-row grid

Open the bank app at 9am Monday. Write down the cleared balance. That’s your opening cell, column 1.

Then list every committed inflow and outflow for the next thirteen weeks. Confirmed valuations, scheduled retention releases, every subbie pay run, every HMRC date from the table above, every standing order, rent, finance payment. Each one dated to the week the money actually moves.

Weeks 2–13: probability-weight what isn’t certain

A confirmed certification goes in at 100% — but ask yourself whether the main contractor pays within terms (some don’t). A new tender you’re 70% sure you’ve won goes in at, say, 50% weighted value. A retention that’s contractually due in week 9 but contractor X always pushes by a fortnight goes in at week 11. You’re not predicting the future perfectly — you’re building a base case you can stress-test.

Update every Monday — 30 minutes, never more

Refresh last week’s actuals. Roll the window forward by one week — week 14 appears, week 0 drops off. Identify the lowest projected closing balance across the next 13 weeks. That’s the trough. If the trough is below the size of your weekly payroll plus your next two HMRC dates, you’ve got an executive-decision week coming, not an admin task. Don’t bury it in a spreadsheet — bring it to the directors’ meeting.

If the trough goes negative inside eight weeks, the forecast did its job: you found out in week 1, not week 7.

Worked example — a £1.4m construction Ltd, 13 weeks from today

This is a composite. Groundworks subcontractor, four main-contractor relationships, six subbies on the books, two PAYE staff (one site foreman, one admin). Opening bank balance Monday morning: £42,000. Standard 5% retentions on all jobs. One quarterly VAT return due in week 6. The full week-by-week version is in the downloadable template below — the abbreviated extract here shows the trough.

The trough lands in week 7 — the week after the VAT bill — at £4,200. By week 8 the forecast is showing negative. That’s the signal: in week 1 we know the business needs to either pull forward a retention conversation, push a subbie pay run back by a week, or arrange a short-term overdraft drawdown to cover weeks 7–9. None of that is a crisis if you spot it in week 1. All of it is a crisis if you spot it the Friday before payroll.

Five cashflow killers I see in construction Ltd books

The patterns repeat across every first review.

1. Retention releases that nobody is chasing

The single biggest hidden bank balance in a construction Ltd. Retentions sit on the balance sheet as a receivable from day one, but the contractual release dates slide past unchecked. Most firms I see have at least two retentions older than they should be. Working capital sat in someone else’s account — see the Barnsley construction-specialist page for the full mechanic.

2. CIS suffered not offset monthly against PAYE

If your Ltd suffers CIS deductions and your accountant isn’t running the monthly Employer Payment Summary offset, that cash sits at HMRC until April. For a Ltd running £8k–£12k of monthly PAYE liability and suffering £4k–£6k of CIS, the monthly offset is real cash in your bank now. The wait-till-spring approach hands HMRC an interest-free loan of your own money.

3. DRC misclassification on inflows or outflows

Either side of a DRC mistake costs cash. Charging 20% VAT on a job that should be reverse-charged means HMRC gets paid money they didn’t need, and your customer can’t recover it. Missing reverse-charge on an inflow means you’ve put VAT in your bank that wasn’t yours. Both unwind eventually, and both unwind through the cashflow.

4. Subbies on 14-day terms while clients pay on 60

You’re funding the gap. On a six-subbie book that’s the size of your overdraft. Either you renegotiate subbie terms (rarely possible — you need them back next week), shift faster-paying clients into the mix, or you size your facility honestly to the working capital the job actually needs.

5. Overdraft as growth funding instead of working-capital buffer

The most expensive money in the business. Overdrafts are for the weekly trough, not the deposit on the next yard. When the overdraft becomes the permanent baseline, the bank stops seeing a healthy business and starts seeing covenant risk. The 13-week forecast is the document that keeps your bank manager calm, well in advance of the conversation you don’t want to have.

Who owns it, and how often it gets done

The director owns the forecast. Not the bookkeeper. The trough call — do we pull retention forward, push payroll back a day, draw down the overdraft, or chase the late certificate — is an executive decision. The bookkeeper updates the actuals; the director makes the call.

Cadence: refresh every Monday in 30 minutes. Review monthly against the management accounts, so the forecast assumptions stay anchored to the real performance of the business. Forecast and management accounts are the same conversation in two timeframes — month and week.

Frequently asked questions

What’s the difference between a cashflow forecast and a cashflow statement?

The statement is historical — it tells you what cash actually moved last month. The forecast is forward-looking — it tells you what cash will move next week. The statement is for HMRC and your accountant. The forecast is for you, every Monday morning. You need both.

Why 13 weeks specifically — not 4 or 26?

Thirteen weeks is one quarter — far enough out to see a VAT bill and a retention release, close enough that the numbers aren’t fantasies. Below 4 weeks you’re reactive; over 26 weeks the probability-weighting becomes guesswork. Thirteen is the construction-industry standard for a reason.

How is construction cashflow different from any other business?

Retentions, CIS deductions, the application-versus-valuation gap and the gap between subbie and client payment terms — these are construction-specific. Most generic accounting advice assumes invoice-in, payment-in, with no holds. Construction punctures that assumption in five different places.

Do I need a 13-week forecast if my accountant already gives me monthly management accounts?

Yes. Management accounts tell you what happened last month. They don’t tell you whether you’ll cover payroll in week 7. The two are complementary — the forecast tells you what’s coming; the management accounts tell you whether it arrived as expected.

How accurate does the forecast need to be?

Inflows within 10% and outflows within 5% is the working target. The point isn’t perfect prediction — it’s catching the troughs early. A rough forecast in week 1 beats a perfect spreadsheet in week 6.

What software do I need — is the Excel template enough?

For most £1m–£5m construction Ltds, Excel is enough — the editability is worth more than the integration with the accounts software. Above £5m turnover or once you’ve got a finance team, a tool like Float, Fluidly or Syft starts to pay for itself.

When does a construction Ltd grow out of weekly cashflow forecasting?

It doesn’t. The largest tier-one main contractors run weekly cash positions for the same reason a £1.2m subbie does — because cash and profit are different numbers and the gap matters at every size. What changes is who builds it and how much of it is automated.

Sources & further reading

• VAT registration threshold — gov.uk
• VAT payment deadlines — gov.uk
• Corporation Tax payment deadline — gov.uk
• CIS — Make deductions and pay subcontractors — gov.uk
• CIS — File your monthly returns — gov.uk
• VAT reverse charge technical guide (DRC) — gov.uk