Document fraud in lending: how to catch fake statements and payslips

Last updated: 2026-07-09

Every lender or MFI has written off at least one loan that should never have been disbursed. The application looked clean, the documents checked out at a glance, and only later — when the repayment schedule went silent — did someone pull the originals and find the salary figure had been edited, the bank balance inflated, or the statement generated wholesale from a template.

Document fraud in lending is not exotic. It is one of the most common forms of application fraud, and the cost is direct: the principal you disburse rarely comes back. Understanding how it enters your funnel — and where your current checks fall short — is the first step to stopping it before disbursement.

How document fraud enters the lending funnel

Most fraudulent applications do not arrive with obviously forged documents. They arrive looking like normal applications — because the applicant has changed only the details they need to change. A monthly net salary bumped from 3,400 to 5,200. An end-of-month balance shifted upward. One transaction removed that would have shown a gambling site or a missed payment.

The forger does not need to reproduce a perfect bank template from scratch. They open the original PDF, edit a number in a tool that re-saves the file, and resubmit. Done in minutes. The result passes a visual scan because everything except the edited value is genuine.

The documents most often forged

Two document types account for the large majority of cases: bank statements and payslips or income certificates. Bank statements are targeted because they show the full picture — balance, regular income, outgoings — in a single file. Forging a statement is straightforward with any PDF editor. Payslips and income certificates are targeted because they are the simplest way to overstate income: change gross, net, or employer contribution, and the document supports a larger loan.

Payment confirmations are less commonly altered to obtain credit, but they do appear in instalment-fraud schemes where the applicant claims a deposit was already paid. Verifying a payment confirmation uses the same metadata approach.

Manual checks your team can do today

Before any forensic tool is involved, trained underwriters can catch a meaningful proportion of forgeries through disciplined manual review. These checks cost nothing except time:

PDF metadata: open Properties in a PDF viewer and check Author, Producer, Creator. A genuine bank-generated statement names the bank or its core-banking software. Adobe Acrobat, Word, or Canva as the Producer is a red flag.
Modification date vs statement date: if the file was last modified after the statement period ended, it was re-saved — and re-saving after the bank issued it usually means editing.
Balance arithmetic: opening balance plus all credits minus all debits must equal the closing balance. A forger who edits one transaction often forgets to adjust the running total.
Payslip maths: gross minus tax and social contributions should equal net. Benefit figures should be consistent with gross. Errors here show the figure was altered without recalculating.
Employer verification: search the company registry for the employer name on the payslip, check the registered address and status, call the number on the document rather than the one supplied by the applicant.

Where manual checks break down at volume

Individual underwriters running the checks above will catch careless forgeries. What they struggle with at scale is the careful one: a file whose metadata has been cleaned, whose numbers add up, and whose visual appearance matches a genuine statement closely enough. The edit is there — in the PDF's internal object tree, its incremental-update history, or the font fingerprints of a substituted glyph — but none of those are visible in a PDF viewer.

Volume compounds the problem. A reviewer processing thirty applications in a day will spend less time on each one. Forgeries are designed to slip through under time pressure: numbers plausible enough to not trigger a second look, a document that is 95% genuine and 5% edited.

Verifying documents before disbursement

Forensic document verification reads what a PDF viewer does not show: the internal object tree, incremental-update records, edit layers, and font fingerprints. When those structures are inconsistent with how the bank's genuine files are built, the discrepancy is flagged — regardless of how the document looks on screen. See the full verification capabilities for what is checked, and the bank statement verification page for the end-to-end process.

Adding verification to your disbursement workflow means every document submitted receives the same level of scrutiny, regardless of how busy the underwriter is or how convincing the application looks. The check produces a structured report you can attach to the loan file — useful both for your internal process and as an audit record if the loan does later default.

For teams with higher volume, per-document pricing scales with submission rate. There is no minimum commitment; each document is priced and reported individually.

Security and data handling

Loan applications contain sensitive personal data. Any verification workflow needs to handle that data carefully. Documents submitted for forensic review are stored only for the duration of the review, then deleted. Access is limited, transfers are encrypted, and processing is conducted under GDPR. Full details are in the security and data handling page. We do not use submitted documents to train models or share them with third parties.

Frequently asked questions.

How do lenders detect forged bank statements?

By checking what a PDF viewer cannot show: metadata and edit traces, balance maths across every transaction, font and layout consistency, and whether the file's structure matches the software a bank would have produced it with. At volume these checks are automated.

Which documents do loan applicants forge most often?

Bank statements and proof-of-income documents — pay stubs and employment letters — edited to inflate balances or salary lines. Payment confirmations are also faked to simulate repayments or deposits.

Can document verification be automated in a lending funnel?

Yes. Forensic PDF analysis runs in seconds per file and can sit before disbursement, clearing genuine documents automatically and flagging suspicious ones for manual review instead of hand-checking every application.

Checking these documents inside your product?

The same check runs as a REST API for software that receives documents from its users:

Lending, BNPL and credit organizations