LenderAnalyzer rebuilds a borrower's real cash flow from the raw transactions, so a lender sees when deposits are inflated by transfers, when the running balance does not tie to the printed ending balance, and when the account contradicts the tax return, before the money goes out. Self-serve from $99 a month.
Upload a document to extract
Drop files here or click to upload
Up to 50 files
Uploading...
Upload a bank statement and watch the analysis run live, free, no signup required.
Bank statement fraud in lending comes in two shapes, and they call for two different tools. The first is document authenticity: a statement that was altered in an editor, with inconsistent fonts, off formatting, or figures pasted over the original. The second is content and consistency fraud: a statement that may be perfectly genuine as a file but tells a false story, where deposits are inflated by transfers between the applicant's own accounts, where debt has been left off, or where the income claimed cannot be supported by the money that actually moved. LenderAnalyzer is built for the second problem, and it is honest about the first. It reconstructs the account from the transactions: it recomputes the running balance and checks whether the arithmetic ties to the ending balance the statement prints, because when the transactions do not sum to the stated balance the document has almost certainly been edited. It nets deposits down to true revenue by identifying transfers and owner injections, so a business that appears to collect $80,000 a month but really collects $40,000 once internal transfers are removed cannot pass on gross deposits alone. It compares bank activity to the tax return, since a systematic gap between deposits and reported income is one of the most reliable signs that one of the two documents is wrong. And it finds the debt an applicant did not disclose, including merchant cash advances collecting daily under an ACH descriptor. Scope, stated plainly: LenderAnalyzer is not a pixel-level document-forensics or metadata-tamper model. It catches the arithmetic and the cash-flow inconsistencies that survive a clean-looking forgery, and it flags them with the transactions behind them. A lender whose primary threat is professionally forged PDFs and synthetic identities should pair it with a dedicated forensic fraud model; the two check different things, and the strongest fraud programs run both.
A skilled forger can produce a PDF that looks right. What is far harder to fake is a coherent set of numbers: transactions that sum to the printed balance, deposits that hold up once transfers are stripped out, and an account that agrees with the tax return. These are the checks LenderAnalyzer runs on every file.
The single most reliable authenticity check is arithmetic. Take the opening balance, apply every credit and debit in order, and see whether the result matches the ending balance the statement prints. When it does not, the document has been altered, because a genuine statement always reconciles. Fabricated deposits and deleted debits are the usual cause, and they are invisible to a reader skimming for a total but obvious to a tool that recomputes the running balance line by line. LenderAnalyzer does that reconciliation automatically and flags the file when the transactions do not sum to the stated balance.
The most common way a genuine statement misleads is not forgery at all, it is presentation. An applicant moves money between their own accounts, or injects personal funds, and the gross deposit total swells to look like strong revenue. Underwriting on gross deposits rewards exactly this. The fix is to identify internal transfers and owner deposits and net them out, so the number driving the decision is what the business actually collected from customers. A merchant showing $80,000 in monthly deposits that nets to $40,000 of real revenue is not necessarily a fraud, but a lender who never separates the two will be repeatedly surprised.
When a borrower submits both bank statements and a tax return, the two should broadly agree on the scale of the business. A systematic mismatch, especially bank deposits far above reported income, means one document is unreliable, and it is a signal that shows up only when you actually reconcile the two rather than reading each in isolation. Sometimes the explanation is innocent, a fiscal-year difference or non-taxable transfers, and asking the borrower resolves it. Sometimes it is the whole case. Either way the gap has to be measured before approval, not discovered in a post-mortem.
Omitted debt is a quieter fraud than a forged balance, and often more damaging. Fixed daily or weekly ACH debits under a funder descriptor mean a merchant cash advance is already collecting, and applicants routinely leave these off because they think of them as a sale of receivables rather than a loan. An income figure that looks serviceable against the disclosed obligations may be underwater against the real ones. Finding undisclosed debt in the statements, and measuring the true debt load against true revenue, is as much a fraud control as it is a credit one.
What each approach catches and what it misses. The strongest programs combine an authenticity check with a cash-flow consistency check. Last updated July 2026.
| Approach | What it catches | What it misses | Typical cost |
|---|---|---|---|
| LenderAnalyzer This page | Math that does not tie to the printed balance, deposits inflated by transfers, bank-to-tax-return mismatches, and undisclosed debt including stacked advances, all traceable | Pixel-level image tampering and forged metadata, which need a dedicated forensic model | Transparent, $99 to $399/mo |
| Manual underwriter review | Obvious inconsistencies a careful reader notices, and font or formatting that looks off | Reconciliation errors and inflated-deposit math across hundreds of transactions, which the eye cannot tally reliably | Free, but slow and inconsistent |
| Forensic tamper model | Altered pixels, inconsistent fonts, edited metadata and known forged templates | A genuine file that tells a false story: inflated real deposits, undisclosed debt, income the account cannot support | Quote-based, often per document |
| Bureau and identity check | Identity mismatches and reported term debt on file | Statement-level fraud entirely, and the advances that never reach a bureau | Per-pull, varies |
Comparison compiled by LenderAnalyzer from public vendor materials, June 2026. Competitor names are trademarks of their respective owners; figures may change, so verify current details with each vendor.
Computed deterministically from every extracted transaction, every figure traceable to its source line.
Computed across the full statement period, carried forward day by day.
Deposits vs withdrawals and net flow, broken down month by month.
Every insufficient-funds and overdraft incident counted, with fees totaled.
Recurring deposits grouped into income streams with estimated monthly amounts.
Debits to other lenders and funders detected and totaled per month.
Days below zero across the period, a direct stress signal.
The biggest credits with dates and sources, concentration flagged.
Automatic red and yellow flags your analysts can review in seconds.
Drop in PDFs, scans or photos, one statement or a multi-month package, from any bank.
Every transaction is extracted, then cash flow, balances, income streams, NSF activity and debt payments are computed.
Read the underwriting snapshot, download the Excel report, or pull structured JSON into your LOS via API.
28 lending document types extracted out of the box, build the complete picture of an applicant's financial situation.
Common questions from lending and credit teams.
The most reliable check is arithmetic: add every credit and debit to the opening balance and see whether the result matches the ending balance the statement prints. If it does not reconcile, the document was almost certainly altered. Beyond the math, watch for deposits inflated by internal transfers, a mismatch between bank deposits and tax-reported income, and undisclosed recurring debits. Formatting and font inconsistencies point to editing too, though catching those reliably is the job of a forensic tool rather than a cash-flow analysis.
The common ones are a bank statement whose transactions do not sum to its printed balance, deposits that look strong only because personal transfers were counted as revenue, reported income that the account cannot support, debt left off the application that is visibly collecting in the statements, and documents that contradict each other, such as a tax return well below the deposit history. A single flag is a question to ask the borrower. Several together, or a math error in the statement itself, is a decline.
Yes, and the arithmetic catches most of them. A genuine statement always reconciles: the opening balance plus every transaction equals the ending balance. When fabricated deposits are added or debits deleted, that reconciliation breaks, and recomputing the running balance line by line exposes it even when the file looks clean. What arithmetic does not catch is a professionally forged image whose numbers were made internally consistent, which is why a full fraud program pairs a reconciliation check with a forensic authenticity model.
Synthetic identity fraud is when a fraudster builds a fake but credible profile by combining real and fabricated information, often nurturing it over time, then applies for credit and defaults with no intent to repay. It is an identity problem more than a document one, and it is best countered at the identity-verification stage. Where statement analysis helps is downstream: a synthetic business often cannot produce a coherent, reconciling account history, so inconsistencies in the cash flow can corroborate what an identity check flags.
Not at the pixel or metadata level, and it is important to be direct about that. LenderAnalyzer detects fraud through the numbers: it reconciles the transactions against the printed balance, nets out inflated deposits, reconciles bank activity to the tax return, and surfaces undisclosed debt. That catches most altered statements, because forgers rarely make the math tie. For detecting a professionally forged image or manipulated metadata, use a dedicated forensic model alongside it. The two check different things and work best together.
In the United States, knowingly submitting falsified financial documents to obtain a loan can be prosecuted as bank fraud under federal law, which carries penalties of up to 30 years in prison and fines up to $1 million. It can also support charges for wire fraud and making false statements to a financial institution. The severity is why documented, traceable underwriting matters: a lender that can show exactly which figures did not reconcile has both a cleaner decline and a stronger record if the matter is ever pursued.
Through a combination of checks. The arithmetic reconciliation confirms the transactions tie to the printed balance. Comparing the statement to the tax return and to any other financials confirms the story is internally consistent. Direct verification, such as a bank-linked feed or a lender contacting the institution, confirms the account exists and the balances are current. Automated analysis handles the first two at speed and flags the files that need the third, so verification effort goes where the risk actually is.
Analyze your first statements free, plans from $99/month, 50% off billed annually.
From the same family of tools
