How to Do Bank Statement Analysis: A Practical Method

Last updated July 2026.

Bank statement analysis means working out what a business or a person actually earns, spends, and holds, using the bank record rather than the accounting record. It matters because the bank record is the one nobody can edit. A lender does it before funding, a buyer does it before acquiring, a forensic accountant does it when the books are in question, and a bookkeeper does it when there are no books at all. The method below is the same in every case, only the emphasis changes.

Step 1: Get every page of every month

Partial statements are the most common reason an analysis goes wrong, and the failure is silent. Missing pages remove transactions, which quietly improves the picture. Insist on complete statements, all pages, consecutive months, every account the entity uses. If a month is missing, the analysis does not begin until it turns up.

Check the sequence as you go. Each statement's closing balance must equal the next statement's opening balance. If it does not, either a statement is missing or something has been altered, and both are worth knowing about before you spend an hour on the numbers.

Step 2: Get the data into columns

You cannot analyze a PDF. Everything below assumes date, description, amount, and running balance sitting in four columns, sorted by date, with all accounts either in one sheet with an account column or in clearly labeled tabs.

This step is why most people stop. A year of statements across three accounts runs to a few thousand lines, and retyping is not a plan. Bank portals will usually export recent activity as CSV, but the window is typically limited to somewhere between 90 days and 18 months while the PDF e-statements go back years, so for a twelve or twenty four month review the PDF is often the only complete source. Running the statements through a bank statement analyzer gets you the columns directly, with the running balance preserved, which matters for the next step.

Step 3: Foot the balances before you trust anything

For each statement period, take the opening balance, add the credits, subtract the debits, and confirm you land on the closing balance. Do this per account, per month, before drawing a single conclusion.

This check does two things. It confirms your extracted data is complete, which is the practical reason. It also catches altered documents, which is the reason lenders insist on it. A statement that has been edited to add a deposit or remove an overdraft usually fails to foot, because whoever changed the transaction rarely recalculates every subsequent running balance correctly.

Step 4: Separate real revenue from money that only looks like revenue

This is where the analysis is actually done, and it is where inexperienced reviewers produce numbers that are far too high. Total deposits is not revenue. Strip out:

• Transfers between the entity's own accounts. Money moved from savings to checking is a deposit in the bank record and is not income. On a multi account review this is the single largest distortion.
• Loan proceeds and advances. A merchant cash advance or a line of credit draw lands as a deposit and is a liability.
• Owner contributions. Cash the owner put in is not the business earning anything.
• Refunds and reversals. A returned payment coming back in is not new revenue.
• Credit card and processor settlements already counted. Watch for the same sales arriving twice through two paths.

What remains is eligible deposits, which is roughly what a lender means by revenue from the bank record. Do the mirror exercise on the outflow side: separate operating expenses from debt service, owner draws, and transfers out, because those three tell you very different things about the business.

Step 5: Run the measures that actually get used

Average monthly deposits. Eligible deposits divided by the number of months. The headline figure in almost every credit decision.

Average daily balance. Sum the daily balances and divide by the number of days. This is the honest liquidity measure. A business can show strong deposits and still run at near zero most of the month, and only this figure reveals it.

Negative days and NSF count. Count days with a negative balance and count non sufficient funds or overdraft fees. This is read as cash flow stress and it carries real weight, particularly with FHA and SBA lenders where a pattern can trigger manual review.

Deposit concentration. What share of revenue comes from the largest customer. Heavy concentration is a risk finding regardless of how healthy the totals look.

Debt service coverage from the bank record. Identify the recurring loan and advance payments leaving the account and compare them to the cash actually generated. Undisclosed debt shows up here more reliably than anywhere else, because a payment schedule is hard to hide when it leaves the account every week.

Month to month volatility. Plot monthly eligible deposits. A steady line and a spiky line with the same average are not the same business, and most credit models take the lower months more seriously than the average.

What are the red flags in bank statement analysis?

Six recur often enough to be a checklist. Large unsourced deposits, particularly any single deposit worth more than half of monthly income, which a mortgage underwriter will ask you to source and season. Round number deposits that do not match any invoice. Money shuffling, where funds cycle between related accounts to inflate apparent activity. Recurring payments to lenders that were never disclosed. NSF fees clustering at the same point in each month, which usually means payroll is landing before receivables. And statements that do not foot, which is the one that ends the conversation rather than continuing it.

Step 6: Tie it back to something

An analysis nobody can check is not worth much. Reconcile your eligible deposits against reported revenue, whether that is a tax return, a profit and loss statement, or a seller's claim. Differences are not automatically a problem, since timing, cash sales, and accrual accounting all create legitimate gaps. But every material difference should have an explanation you can write down in one sentence.

Keep the converted spreadsheet as your working paper, with your categorization visible. When someone disagrees with a number, and on any consequential review someone will, being able to point at the specific rows behind it settles the question in minutes. Where the analysis needs to go further into what each transaction was for, transaction categorization assigns categories across the whole file at once, and converting the statements into a profit and loss gives you a statement to compare against the books. Where a large deposit or an expense needs documenting, the underlying paperwork can be pulled into the same sheet with receipt and invoice extraction rather than sitting in a separate folder.

How many months should you analyze?

It depends on who is asking. Mortgage underwriting typically wants the two most recent statements, all pages. Short term business lenders and merchant cash advance funders usually look at three to six months and often use a rolling deposit average, taking the lowest period rather than the mean. Bank statement loan programs for self employed borrowers run on twelve or twenty four months. SBA and acquisition diligence commonly reach twenty four months, because a full cycle plus a comparison period is the only way to separate growth from seasonality. If you are doing the analysis for yourself rather than for a lender, twelve months is the useful minimum, since anything shorter cannot show you a seasonal pattern.