Cash Flow Analysis for Lending

This is the analysis lens specifically. For the full document workflow — which periods to collect, how to spot an altered PDF, what to ask the applicant — see the bank statement review for loan applications guide.

Why cash flow, not just the closing balance

A closing balance is a single moment a borrower can stage — money in two days before pulling statements, then back out. Cash flow is harder to fake: it's the whole month's behaviour, and it answers the question a payslip can't — not "what do they earn" but "what's actually left". A strong salary cleared to zero every month leaves no headroom for a new repayment; a smaller income lived well inside does.

The five reads that matter

Average balance

The average daily balance shows the cushion the account carries — enough buffer means a missed invoice or an unexpected bill won't tip the borrower into the red. Read it from the daily balance across the period, not the simple average of opening and closing figures, which can look fine while the account spent half the month near empty.

Regularity of income

Lenders care less about the headline figure than whether it's dependable. Salaried income should land once a month, same employer, roughly the same date. Self-employed deposits are lumpier, which is normal — what matters is the trend across six to twelve months. The trap is counting things that aren't income: internal transfers, refunds, reversed payments and one-off gifts all inflate it if they're not stripped out. The income verification guide goes deeper on separating genuine revenue from noise.

Returned items and NSF charges

Non-sufficient-funds (NSF) fees, bounced direct debits and unpaid-item charges are loud signals. One in a year might be a glitch; three in three months says the budget is stretched to breaking. They're small in value and easy to miss in a skim, but they show the borrower regularly running out of money before the bills clear — exactly the situation a new repayment would worsen.

Debt servicing

Sort the outgoings by description and the existing commitments surface: loan repayments, credit-card minimums, car finance, buy-now-pay-later instalments, payday lenders. Total them against income for the current debt-service load before your loan is added. Anything here that isn't on the application form is the most common reason an otherwise affordable case falls down — it changes the affordability sum and the trust at once. The detect undisclosed debt guide covers how these payments hide.

The end-of-month dip

The read most reviewers skip, and often the most revealing. Plot the running balance and look at the trough, not the close. An account can end higher than it started and still have spent ten days underwater until payday. A borrower who scrapes the floor every cycle has no slack — a new repayment tips them into overdraft, and the headroom you read from the averages was never really there.

Why the statement has to reconcile first

Every one of those five reads inherits whatever errors sit in the data, so this part can't be rushed. Most underwriters work from a PDF — emailed by the applicant, downloaded from online banking, or scanned from paper. Two things go wrong. PDFs can be edited: a figure changed in a free editor looks convincing. And converting to a spreadsheet slips in predictable spots — a row drops in the join between two pages, a 1,290.00 debit reads as 1,920.00 when the digits flip, a credit lands in the wrong column.

The honest defence is arithmetic, not eyesight. A complete statement reconciles down its own column: opening balance, plus every credit, minus every debit, lands exactly on the closing balance the bank printed. If it doesn't, a line is missing, a figure was misread, or the statement was altered — and you know not to score the cash flow yet.

This is the gap Export Bank Statement is built to close. When it converts a statement PDF — or a scanned or photographed one, using OCR — to Excel or CSV, it walks the running balance from opening to closing and flags any statement that doesn't reconcile, instead of handing over tidy-looking numbers that are quietly short. So before you calculate an average balance or a debt-service ratio, you know the figures underneath are consistent. Most converters just extract and hope, which isn't good enough for a lending decision.

How to run a lending cash flow analysis, step by step

The workflow we'd follow for a clean, defensible read. Adjust the months to the lender's policy.

1. Collect the right period. Three months for salaried income, six to twelve for self-employed. Full statements, every page — the running balance has to be unbroken for the check to mean anything.
2. Convert to a workable format. Turn each PDF into a single Excel or CSV at /convert so you can sort, filter and total. Copy-pasting from a PDF reorders columns and drops minus signs; a proper conversion keeps money in, money out and balance aligned.
3. Confirm it reconciles. Check the totals tie out before reading anything into them. A flagged statement means a gap a manual skim would have missed — go back to the source first.
4. Calculate the averages. Average daily balance and the lowest point the account reached. The trough matters as much as the average.
5. Verify the income. Match regular credits to payslips or invoices, strip out internal transfers and refunds, judge the trend over any single month.
6. Total the debt servicing. Sort by description, add up existing repayments, cross-check against declared commitments. Recurring-payment detection surfaces these quickly.
7. Score affordability. Set genuine fixed outgoings against verified income, factor in the end-of-month dip, compare the surplus to the new repayment.

Steps four to seven take minutes once the data sits in a clean spreadsheet — the whole point of step two.

A worked example

A self-employed applicant, three months of statements, around 200 transactions a month. On the face of it the income averaged well above the stated figure and the closing balances looked comfortable. But footing the first month by hand, page four didn't tie out — two lines had collapsed onto one during a copy-paste, hiding a 1,290.00 loan repayment. The reconciliation check caught it in seconds.

Once fixed, the real picture showed. Roughly a third of the "income" was transfers between the applicant's own accounts, so the genuine figure was lower. Two undisclosed finance payments came to 410 a month. And the balance dipped below zero in the last week of every cycle, surviving on an overdraft until the next client paid. None of that showed in the closing balance; all of it was obvious once the statement reconciled and the transactions were grouped. The averages alone would have approved a loan the cash flow couldn't carry.

The built-in analyser, and what the tool doesn't do

Once a statement reconciles, the built-in analyser does the grouping: cash flow across the period, income verification, expenses by category, and recurring or merchant detection. Because it runs on data you've already proved complete, every figure inherits that completeness. The judgement still belongs to the underwriter — the tool gives clean figures and a clearer view; it doesn't decide the case, and it isn't a bookkeeping service.

To be precise: Export Bank Statement can export CSV in the native bank-import format Xero, QuickBooks and Zoho Books expect. That's a file import — you convert, then import the CSV, and it lands as reconcilable statement lines. It does not push transactions into a ledger through a live bank-feed API; that needs partner certification this tool doesn't claim. For most lending reviews a reconciled Excel file is enough.