The forecast answers one question: the peak, and whether it retires
A lender reading a working capital forecast is looking for a shape, not a number. They want to see the borrowing need rise through the busy stretch, hit a defined peak, and come back down as receivables collect, because a line that cleans down is revolving credit doing its job. A need that rises and never falls is not a working capital problem; it is a permanent funding gap that belongs in term debt or equity, and lenders decline lines that are quietly being asked to fill it.
So before you open a spreadsheet, be honest about which shape your business has. A seasonal retailer or landscaper borrows to build inventory and payroll ahead of the season and repays out of the season's collections: a clean revolving story.
A business growing fast has a different shape, because growth consumes cash even when it is profitable; every new dollar of monthly revenue gets carried as receivables and inventory before it is collected. Both are financeable. What is not financeable is a forecast that hides one shape inside the other.
The distinction has a practical test: ask what the borrowed dollar becomes. If it becomes inventory and receivables that convert back to cash within the cycle, it is working capital and belongs on a revolving line. If it becomes losses, leasehold improvements or equipment, it is not, whatever the request is labelled, and it should be funded with term debt matched to the asset's life or with equity. Lenders run this test instinctively, and a forecast that quietly funds the second kind of spending from a line reads as either confusion or concealment, neither of which prices well.
This page covers the working capital piece specifically. The full projection package a financing request rides on, income statement, cash flow, assumptions and the debt service math, is covered in how to prepare financial projections for a business loan, and the two documents should reconcile to each other line by line.
Three drivers set the whole forecast: receivables, inventory, payables
Working capital is timing, and three day-counts carry nearly all of it: how long customers take to pay you, how long cash sits in inventory or work in progress, and how long you take to pay suppliers. Measure all three from your own ledgers before assuming anything, because the gap between the terms you quote and the days customers actually take is where most forecasts go wrong. Thirty-day terms with sixty-day collections is a different business, and the lender will compute your real days from your statements whether you do or not.
| Driver | What actually sets it | What the lender checks |
|---|---|---|
| Receivables days | Real collection behaviour by customer, not invoice terms; holdbacks and slow anchor accounts stretch it | Aged receivables list; concentration; how much of the book is past due or ineligible for margining |
| Inventory and WIP days | How long cash sits in stock or unbilled work before it becomes an invoice | Whether inventory is current and saleable, and whether WIP converts to billings on schedule |
| Payables days | Supplier terms you can actually use without hurting supply or price | Aged payables; whether the business is already leaning on suppliers as unpriced financing |
| Growth and seasonality | Each new dollar of revenue gets funded before it is collected; seasons pull the need into a peak | Whether the requested limit covers the stressed peak, and whether the line cleans down after it |
Put those day-counts together and you get the cash conversion cycle: the number of days a dollar spends inside the business between paying for inputs and collecting the sale. The longer that cycle, the more cash each dollar of growth demands, which is why the same growth plan can be self-funding in one business and lethal in another.
Measure rather than estimate. Receivable days come from your aged listing read against recent sales, inventory days from stock on hand against cost of sales, payable days from the aged payables against purchases, and all three should be computed over several periods rather than one, because a single month flatters or slanders depending on where the billing cycle fell. If the ledgers are too untidy to support the calculation, that is itself the first finding, and cleaning them up is part of getting financed, since the lender's diligence will read the same ledgers you did.
Build it in five steps, from history forward
The mechanics are straightforward once the drivers are measured. What lenders reward is a build they can trace, where every line ties to a ledger, a contract or a stated assumption.
- Start with the last twelve months of actual cash movement. Receipts and disbursements by week or month from the bank, not from the income statement. This is the skeleton everything else hangs on
- Layer on known future items. Signed contracts, confirmed price changes, the new hire's start date, tax instalments, HST remittance dates, existing loan payments. Debt service goes in as a hard line, current facilities and the proposed one
- Apply your measured day-counts to projected sales. Revenue booked in March becomes cash in May if your real collection cycle says so. Resist the urge to assume collections improve just because the forecast needs them to
- Run it weekly for thirteen weeks, monthly for the rest of the year. The thirteen-week view exposes payroll-versus-collections crunches that a monthly view averages away; the twelve-month view shows the seasonal shape and the clean-down
- Read off the peak. The largest cumulative shortfall, plus a working cushion, is the limit you request. A limit set to the average month guarantees a covenant conversation at the worst possible time
Two build mistakes account for most of the forecasts lenders discount. The first is projecting collections from invoices you have not yet earned the right to send, milestone billings assumed early, retainers assumed renewed, so tie every forecast receipt to a billing event that is contractually real. The second is skipping the lumpy statutory outflows: HST remittances, payroll source deduction due dates and corporate tax instalments land on fixed calendars, and a model without them shows a smoothness no Canadian business actually has.
Keep a one-page assumptions sheet beside the model: the day-counts used and where they came from, the growth rate and what supports it, and what was deliberately left out. Reviewers extend trust to forecasts that show their work and withhold it from spreadsheets that present conclusions.
Build the model in whatever tool your bookkeeping actually feeds. A forecast wired to the ledgers rolls forward every month in an hour; a hand-built spreadsheet ages into fiction by the second quarter, usually right when the lender asks how the season is tracking against plan.
Stress it before the lender does
Every credit reviewer will bend your forecast to see where it breaks, so hand them a version where you already did. Three stresses cover most owner-managed businesses: your largest customer pays thirty days slower, one meaningful receivable goes bad, and the growth case where winning more work makes the cash position worse before it gets better. If the line still covers the stressed peak, say so; if it does not, the honest move is to size the request to the stress case and explain why, because the alternative is discovering the gap while you are living it.
Stress testing also forces the margining conversation early. Operating lines are usually margined: the amount you can actually draw is set by a formula against eligible receivables, and sometimes inventory, with each lender defining eligibility its own way. Old accounts drop out of the calculation, and related-party balances are typically excluded, as are construction holdbacks, which is why contractors so often find their real availability far below the limit on paper.
If your receivables book is heavy with slow, related or holdback balances, model availability, not just the limit; a forecast that ignores margining can show a line covering a peak the formula will never let you draw. This is a live issue for the trades especially, and it is a standing part of our CFO work for general contractors.
Present each stress as a row, not an essay: the scenario, the new peak, and the headroom remaining against the requested limit. Three rows and one sentence of interpretation give a reviewer everything they need, and the format itself signals that you run the business by these numbers rather than having produced them for the occasion. That impression is worth more than it sounds, because working capital lending is ultimately a bet on management's grip.
Five facts change what the forecast says
Two businesses with identical revenue can need entirely different limits. When we build a working capital forecast, these are the facts that move the answer:
- The length of your cash conversion cycle: receivable days plus inventory days minus payable days, measured from your own ledgers
- The growth rate, because faster growth pulls more cash into the cycle every month it continues
- Seasonality: how sharp the peak is and how completely the line cleans down after it
- Customer concentration: one anchor customer changing its payment run moves your whole forecast
- Margining eligibility: how much of your receivables book a lender's formula will actually count
Notice what is not on the list: profitability. A profitable business can run out of cash mid-growth, and a modestly profitable one with a short cycle can self-fund for years. Lenders know this, which is why the working capital forecast, not the income statement, decides the size of the line.
The five facts also explain why a forecast should be rebuilt, not merely updated, whenever the business changes shape. A new anchor customer on sixty-day terms, a move into carrying inventory, a decision to extend terms where you once collected on delivery: each rewrites the day-counts the old model was built on. Lenders notice when the forecast stops resembling the ledgers, and so should you, preferably first.
A useful contrast makes the point. A distributor carrying ninety days of inventory on thirty-day supplier terms lives on its line, while a service firm billing weekly with no stock may need almost none, at identical revenue. The forecast exists to show which one you are, in your own numbers, rather than in your industry's reputation.
Presenting it, and living with it afterwards
Deliver the forecast as part of a coherent package: the thirteen-week and twelve-month views, the assumptions page, the stress cases, and a cover note that states the requested limit and points at the peak that justifies it. It should agree with your financial statements and your projections without a reviewer having to force the reconciliation. The broader list of what sits around it, statements, security, guarantees, the lender's due diligence on your tax accounts, is in what lenders need before approving business financing.
Expect the forecast to become the measuring stick after funding, because lenders file it and read your monthly reporting against it for the life of the facility. That is the argument for realism over salesmanship at the application stage: a forecast you quietly beat builds the credibility that makes the next request easy, while one you miss puts every future number under discount. It is also the argument for keeping the model alive, rolled forward monthly against actuals, rather than rebuilt in a hurry the next time you need something.
Then plan for the reporting that follows funding, because a working capital facility is a relationship with a monthly rhythm: margin certificates, aged receivable and payable lists, covenant calculations, annual statements. Lender reporting done late or wrong erodes exactly the confidence the forecast built. We prepare working capital forecasts and full financing packages for owner-managed businesses across Mississauga and the GTA, business financing and projections work by a CPA firm whose founder came out of banking, and we stay on for the reporting cycle afterwards; how the whole engagement runs is in business financing support for owner-managed businesses. Scope and fee arrive in writing after a free 15-minute discovery call.
