How to spot fake bank statements: 12 checks for lenders
For underwriters, fraud analysts and onboarding teams at lenders, landlords and fintechs: the red flags that give away an edited or fabricated bank statement, how to confirm it, and where automation saves time.
Key takeaways
- Fake bank statements are often edited (a real statement with changed numbers or names) or fabricated (made from a template or with AI). Inscribe flagged 1 in 16 documents across its network as fraudulent in 2025, and 1 in 5 of those showed template-based manipulation.
- The fastest check is the math. Opening balance plus deposits minus withdrawals must equal the closing balance, and every running balance must follow from the line above.
- Visual and file checks find many edits. Look for mismatched fonts or spacing, misaligned columns, and PDF metadata showing an editing tool or a file changed after it was created.
- Cross-document checks find what the page alone doesn't show. Payroll deposits should match pay stubs, the name and address should match the application, and income should match tax returns.
- Automating the extraction and the balance math lets lenders check every statement, not a sample, and spend reviewer time on the files that don't add up.
On this page
To spot a fake bank statement, start with the math. The opening balance plus deposits minus withdrawals must equal the closing balance, and every running balance must follow from the line before it. Then check how the page looks, the PDF's metadata, and whether the statement agrees with the applicant's pay stubs, tax returns and application. If anything doesn't fit, get the statement again through a channel you can verify.
Why fake bank statements are a growing problem#
Bank statements prove income, assets and cash flow, so they're a target for anyone trying to qualify for a loan, a lease or an account. Editing a PDF takes minutes, template sites sell realistic statements, and AI tools can now generate them. Inscribe's 2026 Document Fraud Report found that 1 in 16 documents processed across its network in 2025 was flagged as fraudulent, and 1 in 5 flagged documents showed signs of template-based manipulation, up from 1 in 14 in 2024. Of the 90 fraud and risk professionals it surveyed, 85.6% named bank statements the document type most vulnerable to manipulation.
Types of bank statement fraud#
Fake statements come in four kinds, and each fails a different set of the checks below.
Edited statements
A real statement with changed balances, deposits, names or dates. Often the hardest to see by eye.Fabricated statements
Built from a template or generated with AI for a bank the applicant may or may not use.Borrowed statements
Someone else's real statement, passed off as the applicant's, often with only the name changed.Inflated activity
Real statements showing money moved in temporarily ("round-tripping") to inflate balances before an application.
12 checks to spot a fake bank statement#
The checks fall into four groups. The math (1 to 4), how the page looks (5 to 7), the file itself (8 and 9), and whether the statement agrees with the rest of the application (10 to 12).
- 1. Balance reconciliationOpening balance + total deposits − total withdrawals = closing balance.
- 2. Running balancesEvery line's balance equals the previous balance plus or minus that transaction.
- 3. Dates and continuityPeriods run on without gaps, each opening balance equals the last closing balance, and the dates make sense. ACH doesn't settle on weekends or federal holidays, so an ACH deposit dated on one deserves a question.
- 4. Summary totalsDeposit and withdrawal counts and totals in the summary box match the transaction list.
- 5. Fonts and spacingA different font, size, weight or kerning in names, amounts or dates. Zoom to 400% and edited digits often sit slightly off the baseline.
- 6. AlignmentColumns of numbers line up exactly; edited values often don't.
- 7. Bank detailsLogo quality, address, phone number, routing number and layout match the bank's real format for that period.
- 8. PDF metadataThe creator and producer fields show the bank's system, not an image editor or office tool, and the file wasn't modified after it was created. Metadata can be stripped, so a clean result isn't proof.
- 9. Image vs textA statement downloaded from online banking is usually text-based. A flattened image, or text and image layers mixed on one page, deserves a closer look.
- 10. Deposits vs income documentsPayroll deposits match net pay on the pay stubs, with the same employer name and pay frequency.
- 11. Identity matchName, address and account number match the application, ID and other statements.
- 12. Transaction patternsToo many round numbers, no everyday spending, large unexplained deposits just before the application, and the same amounts moved in and out.
How to confirm a suspicion#
A red flag is a reason to verify, not proof of fraud. What you do next depends on your process and the applicant's consent.
Get the data directly
Ask the applicant to connect their bank through an account verification service, or to download statements through a secure link you control. See instant bank account verification.Verification of deposit
With the applicant's written authorization, request a verification of deposit form from the bank.Ask for more
Additional months, a different account, or documents that source large deposits. Mortgage lenders already do this for large deposits on purchase loans.Document and escalate
Record what you found and follow your fraud policy and any reporting obligations.
Where automation helps#
Manual review works on a sample. Automated review runs the math on every statement and sends people only the files that fail.
Manual review
- Samples or high-risk files only
- Math checks are slow and easy to skip on long statements
- Visual and file checks depend on the reviewer's eye and time
- Pay stubs and the application are compared by hand
- The output is a reviewer's notes
Automated review
- Every statement, every page
- Every balance and every line reconciled
- Automatic comparison across the whole file
- Mismatches shown with the page, sent to a reviewer
| Field | Extracted value | Confidence |
|---|---|---|
| Account holder | Harbor Street Bakery LLC | |
| Address | 118 Harbor St, Portland, ME | |
| Bank name | First Midwest Bank | |
| Account number | •••• 4821 | |
| Account type | Business checking | |
| Statement period | 2026-07-01 → 2026-07-31 |
| Field | Extracted value | Confidence |
|---|---|---|
| Opening balance | 42,180.55 | |
| Total deposits | 88,412.10 | |
| Total withdrawals | 91,686.44 | |
| Closing balance | 38,906.21 | |
| Transaction count | 142 |
| Field | Extracted value | Confidence |
|---|---|---|
| Date | 2026-07-14 | |
| Description | ACH DEPOSIT STRIPE PAYOUT | |
| Amount | +3,284.10 | |
| Debit / credit | Credit | |
| Running balance | 51,902.44 | |
| Category | Card processor payout |
Docsumo is an intelligent document processing (IDP) platform. Its bank statement extraction turns PDFs and scans into a transaction table, joins tables that run across pages and checks running balances; fields it's unsure about go to a reviewer with the source line highlighted. On the Enterprise plan, cross-document validation compares the statement with pay stubs and the application. File and formatting checks stay with your reviewers.
Where we land. The read and the math above are software your own team runs, not a review staff you pay a vendor for. That keeps the call on a flagged statement with the reviewer who knows your credit policy, source line in front of them, instead of a note handed back from someone else's queue.
For lending workflows, see IDP for lending and bank statement verification for business loans. Statements carry account numbers, so use a vendor with SOC 2 Type 2 compliance. Docsumo shows where the numbers don't match; deciding whether a statement is genuine, and any reporting, stays with your team.
What a red flag doesn't tell you#
Most fake bank statements fail on the math, the formatting, the file or the story they tell next to other documents. Check all four on every statement, confirm suspicions through a channel you can verify, and automate the extraction and math so your reviewers spend their time on the files that deserve it.
Book a demo with a few of your own statements, or start a free trial.
Frequently asked questions#
How can you tell if a bank statement is fake?
Check that the balances add up, look for inconsistent fonts, spacing and alignment, inspect the PDF's metadata for editing software, compare deposits with pay stubs and the application, and, if still in doubt, verify directly with the bank or through a bank data connection.
Can a bank verify a bank statement?
Banks generally won't discuss a customer's account without the customer's authorization. With consent, you can use a verification of deposit form, a bank data connection, or have the applicant download statements directly while you watch or through a secure link.
Are PDF bank statements easy to fake?
PDFs are easy to edit with common tools, and online template sites and AI make fabricated statements look convincing. That's why checking the math, the file metadata and other documents matters more than how the statement looks.
What should a lender do if a bank statement looks fake?
Don't accuse the applicant. Ask for statements through a verifiable channel, such as a direct bank connection or a verification of deposit, document what you found, and follow your fraud policy and any reporting obligations.
Is submitting a fake bank statement a crime?
In the US, knowingly making a false statement to influence a loan decision by a federally insured bank or credit union, or by a mortgage lending business, is a federal crime under 18 U.S.C. § 1014, punishable by a fine of up to $1,000,000, up to 30 years in prison, or both.
Can software help check bank statements?
Software can extract every transaction, reconcile the balances and compare deposits with pay stubs and the application, so reviewers only look at statements that don't add up. Deciding whether a statement is genuine, and what to do about it, stays with your team.