The short answer
A 13-week cash flow forecast is a week-by-week view of every cash receipt and payment over one quarter. It is built on the direct method: not derived from profit and loss, but from what actually happens on the bank account.
That makes it a steering instrument, not a reporting one. It answers a single question: is there enough money — and if not, in which week does it run out?
Why thirteen weeks
Thirteen weeks is a quarter. Long enough that countermeasures can still work: a renegotiated payment plan, a collections push or a credit line take weeks, not days. Short enough that the numbers hold, because the order book and the open items already determine most of them.
At twelve months a cash plan is an estimate. At four weeks it is a bank statement with a head start. At thirteen weeks it is a basis for decisions.
A note for readers outside Germany
The instrument itself is the same everywhere. What is not the same is the payment calendar it has to be built on, and this is where finance teams run from abroad lose the most money.
- Social security contributions are due before the month ends. German social security contributions (Sozialversicherungsbeiträge) fall due on the third-to-last banking day of the month they relate to — in most companies before the salaries themselves have even been paid, and based on an estimate that is trued up the following month. A model that assumes payroll costs settle in the following month is wrong by a full month on the largest single outflow in the file.
- Wage tax follows on the 10th of the following month, and VAT prepayments (Umsatzsteuer-Voranmeldung) on the 10th as well — unless the company has a filing extension (Dauerfristverlängerung), which shifts it by a month and is worth confirming rather than assuming.
- The priority order is different. Withholding employee social security contributions and not paying them over is a criminal offence for the managing director in Germany, not a late payment. In a squeeze it is therefore not in the same category as a supplier invoice, whatever a group treasury policy says. The same logic sits behind personal liability for payments made once insolvency has set in — see insolvency filing deadlines in Germany.
Get those three dates into the right weeks before anything else. Everything after that is ordinary forecasting.
What goes in
The structure never changes: opening balance, receipts, payments, closing balance — per week, with each week's closing balance opening the next.
Receipts come from three sources: open receivables with an expected payment date, invoices issued but not yet due, and planned new invoicing out of the order book or subscription revenue. Then the one-offs: tax refunds, grants, funding inflows.
Payments split into fixed and discretionary blocks. Fixed: salaries with social security, rent, leases, insurance, interest and amortisation. Discretionary: supplier invoices, capex, variable costs. The split matters because only the discretionary part can be steered at short notice.
The block most often missing, and the one that does the most damage, is taxes and social security. Large amounts, fixed dates, and not deferrable without creating a liability problem.
Building it in five steps
- Capture balances and lines. Every bank account, the available overdraft facility, any blocked amounts. If the starting figure is wrong, nothing after it is worth anything.
- Pull the open items. Export receivables and payables with due dates from the accounting system. For receivables, plan the actual payment behaviour, not the due date — a customer who has historically paid twenty days late will pay twenty days late again.
- Date the fixed blocks. Payroll on the pay date, social security on the third-to-last banking day, taxes on their dates, interest and amortisation to the bank's schedule.
- Add new business. From the order book and pipeline, deliberately conservative. When in doubt, put it a week later.
- Define the floor. The minimum balance the business needs to operate belongs in the chart as a line. The useful statement is not “we reach zero” but “we drop below the floor in week 9”.
The weekly plan-versus-actual
A 13-week forecast built once and rolled forward monthly is an arithmetic exercise. Its value comes from the weekly rhythm: at the start of the week the closed week is set against the plan, every variance above a defined threshold is explained, and the forecast rolls forward by one week.
The variance analysis is the actual return on the work. Three consecutive weeks of receipts ten per cent below plan is not a planning problem, it is a collections problem — and the forecast shows it eight weeks before the P&L does.
The six typical mistakes
- Deriving it from the P&L. Accruing revenue and costs and then adding payment terms produces an earnings plan on a delay. The direct method starts from the cash flow, not from the result.
- Receipts planned too optimistically. The single most common error. The plan uses the due date; the customer uses habit. The buffer belongs on the receipts side, not in the closing balance.
- Forgetting VAT and social security. Both large, both date-fixed, neither negotiable.
- Too much detail. Plan every line item individually and nobody maintains it weekly. Thirty to fifty rows is enough; the file has to be updatable in under an hour.
- No scenario. Alongside the base case the file needs a downside: receipts fifteen per cent later, no new business. That is what answers the real question — how much headroom is left if it goes worse.
- No action column. Next to every tight week belongs what is being done about it, by when and by whom. Without that column the forecast is a diagnosis without a treatment.
When it stops being optional
Three situations. First, the bank meeting: nobody asking to raise a line or suspend amortisation is taken seriously without a week-by-week forecast — more on what lenders expect in covenant reporting. Second, an approaching covenant breach, where the timing has to be pinned down. Third, anywhere near insolvency: both the solvency test and the going-concern forecast presuppose a solid cash forecast. Building one at that moment costs three weeks nobody has.
Common questions
Is a spreadsheet enough, or does it need software?
For a single entity with a handful of accounts, a well-built spreadsheet is enough. Once several entities, currencies or intercompany settlements are involved, maintenance becomes the bottleneck — then a treasury or planning tool with bank connectivity earns its keep.
Who builds and maintains it?
The finance function builds it; one named person maintains it weekly. Management owns it, because that is where the countermeasures are decided.
How accurate is a 13-week forecast?
Done properly, the variance in the first four weeks is in the low single-digit percentages, because most of it comes from open items. From week eight the uncertainty grows noticeably. That is exactly why it is rolled forward weekly.
How is this different from the annual financial plan?
The financial plan covers earnings, balance sheet and cash flow over twelve months or more and answers strategic questions. The 13-week forecast covers cash flows only, over one quarter, and answers the solvency question. Neither replaces the other.
When should a company introduce one?
Once the cash reserve is smaller than two months of costs, or the business is strongly seasonal or project-driven. In a crisis it is mandatory; before that it is cheap insurance.
Which German payment dates are most often missed?
Social security contributions, due on the third-to-last banking day of the month they relate to — before the month is over and usually before salaries are paid. After that, wage tax and the VAT prepayment on the 10th of the following month, subject to any filing extension.
Read on
- Cutting working capital – the levers that reduce the cash requirement permanently.
- The going-concern forecast – when the cash plan becomes a legal question.
- Subordination agreements and comfort letters – what a shareholder can contribute without paying in cash.
- Finance in a restructuring – the first six weeks, where this forecast is step one.
- Insolvency filing deadlines in Germany – the clock the forecast is measured against.
- Finance as a Service – when the build and the weekly maintenance are staffed externally.
Sources and status
Based on nugrow's mandate experience in growth and special situations, and on the German filing and payment dates for wage tax, VAT prepayments and social security contributions. As of September 2026. This article is an overview and does not replace tax or legal advice.
