13-Week Cash Flow Forecast: How to Build One, with a Worked Example

What a 13-week cash flow forecast is, and why it is weekly

Profit and cash are different numbers. Profit counts a sale when it is made and a cost when it is incurred; cash counts money when it arrives in or leaves the bank. A company can report a profit every quarter and still run out of cash, because the cash is sitting in unsold stock or in invoices customers have not yet paid.

A 13-week forecast shows cash on the dates it actually moves. Payroll, rent, supplier runs and loan repayments each fall on a specific day, and a monthly forecast averages them away. The lowest point of the month, which is the one that matters, disappears into the monthly total.

Thirteen weeks is one quarter. That is long enough to show the next payroll cycles, loan payments and quarterly tax, and short enough to forecast from invoices, orders and contracts that already exist rather than from estimates. Lenders ask for one when a company is close to its covenant, a condition in a loan agreement such as a minimum cash balance, and turnaround advisers build one in the first week of any engagement.

The rows of the forecast

Each column is a week. Each row is a type of cash movement, with the source the number should come from.

This layout is the direct method: it lists actual cash in and out. The indirect method, used in the cash flow statement of annual accounts, starts from profit and adjusts for non-cash items. The indirect method explains the past well; the direct method is the one that can show which week cash runs short.

How to build it in seven steps

1. Start from the bank, not the books

Opening cash is the total of today’s bank balances. The accounting ledger can lag by days, and payments that have been sent but not yet cleared make it look higher than it is.

2. Forecast receipts from collections, not sales

Customers pay on their own schedule. Use the receivables ageing, the list of unpaid invoices by how overdue they are, and each large customer’s actual payment history. Days sales outstanding (DSO), the average number of days customers take to pay, sets the timing for everything else.

3. Put every payment on its date

Payroll on payroll day, rent on the first, suppliers on their due dates. Days payable outstanding (DPO), the average days the company takes to pay suppliers, sets the timing for the supplier rows.

4. Add financing, tax and capital items

Loan interest and repayments, tax instalments and approved capital spending are large, dated and easy to forget because they are not in the monthly operating budget.

5. Find the low point and the headroom

The single most important output is the week with the lowest closing cash and how far it sits above or below the minimum balance. Everything else in the forecast exists to make that one number reliable.

6. Find the binding constraint

If the low point breaches the minimum, test the possible causes in turn: is the business unprofitable, is growth consuming cash, or is cash tied up in stock and receivables? The binding constraint is the one cause that, if fixed, moves the low point. Fixing the others does not.

7. Roll it forward every week

Each week, drop the week just finished, add a new week 13, and compare last week’s forecast with the actual bank movements line by line. The variance, the gap between forecast and actual, shows which assumptions to correct.

Worked example: Kestrel Outdoor

Kestrel’s profit and loss statement did not show the problem. Reported EBITDA, earnings before interest, tax, depreciation and amortisation, was $3.1 million. After removing a one-off $0.9 million settlement booked as revenue and $1.2 million of marketing spend recorded as an asset, and adding back $0.4 million of founder pay above market rate, normalised EBITDA was $1.4 million. Only 31 cents of cash arrived for every dollar of reported profit.

Testing the causes in turn found the binding constraint:

Working capital is the cash tied up in running the business: stock, plus money owed by customers, minus money owed to suppliers. The measure that connects it to the forecast is the cash conversion cycle, the number of days between paying for stock and collecting the cash from selling it. Kestrel holds stock for 118 days, collects from customers in 4 and pays suppliers in 38: 118 + 4 − 38 = 84 days.

The size of the lever follows from the same figures. At $48 million of revenue and a 41% gross margin, cost of goods is 59% of revenue, $28.3 million a year, or about $78,000 a day. Each day of stock Kestrel holds ties up about $78,000 of cash at cost. The $190,000 gap in week 9 is the cash held in about two and a half days of stock.

The conclusion changes the board conversation. Kestrel does not have a profit problem or a growth problem to fix before week 9; it has a stock problem. Reducing inventory days is the one action that moves the low point.

What a one-line prompt gets wrong

Ask an AI tool to “build me a 13-week cash flow forecast” with a spreadsheet attached, and it will usually produce a tidy table. The errors are in how it fills it.

Try it: a shortened Prompt 07

This is a shortened version of Prompt 07 from the Finance & Analysis pack, set up to run on its own. Paste it into ChatGPT, Claude or Gemini and fill in the brackets.

You are a restructuring CFO who has built weekly cash forecasts for lenders during covenant reviews. A 13-week cash flow forecast uses the direct method: cash in and cash out on the dates they actually happen, never the budget divided into weeks. Receipts come from how customers actually pay, not from sales, and a forecast is only useful if it names the week cash falls below the minimum and the cause.

THE POSITION:
- Cash in the bank today, by account: [amounts]
- Minimum cash you must hold: [loan covenant, facility terms or internal floor]
- Payroll dates and amounts: [dates, net pay, payroll taxes]

THE FLOWS (paste what you have):
- Receivables ageing and how your main customers actually pay: [paste]
- Payables and supplier terms: [paste, with due dates]
- Fixed payments - rent, loan interest and repayments, tax: [list with dates]
- Inventory days, receivable days and payable days, if known: [figures]

Produce:

1. THE 13-WEEK TABLE. One column per week: opening cash, receipts by type, payments by type, net cash flow, closing cash, the minimum balance and the headroom. Show the arithmetic for week 1.

2. THE LOW POINT. The week with the lowest closing cash, its amount, and whether it breaches the minimum. If it does, the size of the gap.

3. THE CASH CONVERSION CYCLE. Inventory days plus receivable days minus payable days, and the cash tied up by one day of inventory.

4. THE BINDING CONSTRAINT. Test profitability, growth and working capital in turn, and name the one that causes the low point, with the figure that proves it.

Mark every figure you estimated rather than took from my data. If a payment date or amount is missing, list it as a gap; do not fill it with an average.

The full Prompt 07 has eight parts rather than four. It also reads the Finance Brief built by prompts 01 to 06, so the forecast starts from normalised earnings and from inventory days checked against actual results, not the ones assumed in the budget.

Mistakes that make the forecast wrong

• Starting from the ledger. Uncleared payments make the opening cash look higher than the bank will allow you to spend.
• Treating sales as receipts. A sale on 60-day terms is cash two months later, not this week.
• Averaging dated payments. Payroll, rent and loan repayments land on specific days; spreading them evenly hides the low point.
• Forgetting non-operating items. Quarterly tax, loan repayments and capital spending are the payments most often missing.
• Adjusting a figure until it looks right. If the forecast does not reconcile with the bank, find the wrong assumption; do not change a number to close the gap.
• Never checking it against actuals. A forecast that is not compared with what happened each week gets less accurate every week.

Questions

Written by a former Gartner Managing Partner and investment banking SVP

The guides and the prompt templates on this site come from a career spent building these documents: board decks, forecasts, business cases and hiring decisions, as Managing Partner at Gartner, SVP in investment banking and Country Manager at international subsidiaries. The worked examples are published in full on each product page. Browse the Prompt Library.