How to Build a 13 Week Cash Flow Forecast from Bank Statements

Last updated July 2026.

A 13 week cash flow forecast is a rolling, week-by-week projection of cash in, cash out, and closing bank balance for the next quarter. You build it from bank statement history rather than from your income statement, because a forecast has to model when money actually moves, not when revenue is earned. Convert twelve months of statements to a spreadsheet, group the transactions into recurring receipt and disbursement categories, average them by week, then overlay what you already know is coming.

Thirteen weeks is the standard window for a reason. It is one quarter, long enough to see a cash crunch forming with time to act, and short enough that the numbers stay honest. Lenders, turnaround advisors, private equity sponsors, and bankruptcy courts all ask for this exact format, so learning it once pays off repeatedly.

Why bank statements, not the P&L

Accrual accounting deliberately separates economic activity from cash movement. That is the right way to measure profit and the wrong way to predict whether payroll clears on the 15th. A profitable month with 60 day receivables can still empty the account.

Your bank statement has none of that ambiguity. Every line is a real movement of real money on a real date, including the items that never show up as expenses at all: loan principal payments, owner draws, sales tax remittances, credit card payoffs, and transfers between accounts. Those are frequently the largest outflows in a small business, and every one of them is invisible on a profit and loss report.

Step 1: Get twelve months of statements into a spreadsheet

You want a full year so seasonality shows up, plus every account cash actually flows through: the operating checking account first, then any savings sweep, and the business credit cards whose balances get paid from checking.

Download the PDF statements from online banking rather than relying on the transaction export screen. Most US banks cap that export at 90 days to 18 months and drop the running balance, while the PDF statements go back much further and carry the opening and closing balances you will use to prove your work. Then convert the PDF statements to Excel so you have one sheet with date, description, debit, credit, and balance columns for the whole period.

Before moving on, tie it out. Sum the debits and credits for each month and confirm that opening balance plus net movement equals the closing balance printed on the statement. If a month does not tie, you have a parsing problem or a missing page, and every number downstream inherits the error.

Step 2: Bucket transactions into forecast lines

A forecast with 40 line items is unusable. Aim for eight to fifteen. The point is to model behavior, not to reproduce your chart of accounts.

On the receipts side, split by how the money arrives, because timing differs: customer ACH and wires, card and merchant processor deposits, checks, and anything irregular like a tax refund or a loan draw. On the disbursements side, group by predictability rather than by expense type. Payroll and payroll taxes go together. Rent, insurance, and software subscriptions form a fixed monthly block. Then variable vendor payments, debt service, owner draws, and taxes each get their own line.

You can categorize transactions from a bank statement in bulk with a lookup table keyed on merchant text, which is far faster than tagging line by line. Keep a catch-all bucket and watch it: if more than five percent of dollars land there, your rules need work.

Step 3: Convert history into weekly run rates

Add an ISO week number column to the converted data, then pivot dollars by category and week. Now read the pattern in each row.

Fixed items are easy. Rent hits once a month on roughly the same date, so place it in the week it lands, not as a weekly average. Payroll follows its own calendar, and this is where a spread-it-evenly forecast breaks: a biweekly payroll produces three payroll weeks twice a year, and those are exactly the weeks that go negative. Map the actual pay dates for all 13 weeks.

For genuinely variable lines, use a median weekly figure rather than a mean so one unusual month does not distort the run rate. For receipts, look at the last 13 weeks more heavily than the full year unless the business is strongly seasonal, in which case use the same quarter last year adjusted for growth.

Step 4: Lay out the grid and overlay what you know

Columns are weeks, always labeled with the week ending date. Rows run in this order: opening cash balance, receipt lines, total receipts, disbursement lines, total disbursements, net cash flow, closing cash balance. Each week's closing balance becomes the next week's opening balance, so the whole model chains off one starting number.

Then replace averages with facts wherever you have them. Known customer payments with committed dates, a signed insurance renewal, the quarterly estimated tax payment, an equipment deposit, the annual software renewal that only appears once in your history. Statistical run rates are a fallback for the unknown, not a substitute for the calendar you already have.

Finish with a minimum cash line, the balance below which you cannot operate, drawn across the grid. The forecast's whole job is to show you which week you cross it.

Step 5: Roll it forward every week

This is the step most people skip, and skipping it is what makes forecasts useless. Every Monday, drop the completed week, add a new week 13, and enter actuals against your forecast for the week just finished.

The variance column is the real deliverable. Chronic optimism on collections shows up within a month, and you can correct the assumption instead of repeating the mistake for a quarter. Pull the prior week's transactions, compare each category to what you projected, and note why anything moved more than ten percent. After six or eight weeks of this, your forecast starts landing within a few percent, which is the point at which a lender or a board will trust it. Teams that run this discipline consistently often build the same weekly loop for their monthly financial statements so the two views never drift apart.

Common mistakes that break the forecast

• Forecasting revenue instead of collections. Invoicing $100,000 in a week means nothing for cash. Model when it clears, using your actual days-to-pay history.
• Double counting transfers. If you include multiple accounts, transfers between them net to zero. Leave them in and you will inflate both receipts and disbursements.
• Averaging payroll. The three-payroll weeks are the entire reason to build this. An average hides them.
• Ignoring credit card payoffs. Card spending hits cash when the statement is paid, not when the charge is made. Model the payoff date.
• Building it once. A static 13 week forecast is a snapshot. The value is in the rolling update and the variance analysis.

Frequently asked questions

What is a 13 week cash flow forecast?

A 13 week cash flow forecast is a rolling weekly projection of cash receipts, cash disbursements, and closing bank balance over the next quarter. It models actual money movement rather than accrual revenue and expense, which makes it the standard tool for liquidity planning, lender reporting, and turnaround work.

Why is a cash flow forecast 13 weeks and not 12?

Thirteen weeks is exactly one quarter of a 52 week year, so the window aligns with quarterly reporting and covers three full monthly cycles of rent, payroll, and debt service. Twelve weeks would clip the third month for many pay calendars.

Can I build a 13 week cash flow forecast in Excel?

Yes, and most are. You need a weekly grid, categorized transaction history from your bank statements to set run rates, and a chained closing balance so each week feeds the next. No specialized software is required, though converting the statement PDFs to a spreadsheet first saves hours of retyping.

How much bank statement history do I need?

Twelve months is the practical standard. That captures annual items like insurance renewals and tax payments, plus any seasonality in receipts. Six months works for a young or stable business, but you will miss once-a-year outflows that can distort a quarter.

How accurate should a 13 week cash flow forecast be?

Weeks one through four should land within roughly five percent once you have a few cycles of variance analysis behind you, because most of that period is known commitments. Accuracy loosens further out, and weeks nine through thirteen are directional. Precision in the near term is what matters for decisions.

What is the difference between a cash flow forecast and a cash flow statement?

A cash flow statement looks backward and reports what happened, usually monthly and split into operating, investing, and financing activities. A forecast looks forward and is organized by week and by category of receipt or payment, so you can act before a shortfall rather than explain it afterward. You can also build a cash flow statement from bank statements using the same converted data.