Why high-risk applicants slip through and how custom transaction insights can stop them

Bank statement fraud is not an edge case. The FBI reported nearly $12.5 billion in fraud losses from U.S. consumers in just the first three quarters of 2025, a 25% year-over-year increase. Our 2026 Document Fraud Report found roughly 1 in 16 documents showing signs of manipulation, with bank statements flagged as the top concern by 85.6% of fraud and risk leaders.

The problem isn't just fraud volume. It's that most detection tools weren't built with your risk policies in mind. What's a red flag at one institution is a non-issue at another. And when the tools don't reflect that, high-risk applicants slip through.

That's what custom transaction insights are designed to fix.

‍Deep bank statement analysis, built around your risk criteria

A bank statement tells you a lot more than a balance. It shows you how someone behaves with money over time, and that's usually where the real picture emerges. It's worth noting that over 90% of flagged documents contain altered financial details, meaning the risk is almost always hiding in the numbers.

With custom transaction insights, you define what you're looking for.

Build rules based on transaction category, merchant, payment processor, method, amount, or timeframe. Fire them based on raw transaction count or percentage of total activity. Set balance-based thresholds, track average daily balance, or flag volatility against a rolling average.

Once your signals are live, Inscribe runs the analysis automatically. No manual cross-referencing. No relying on reviewers to remember your institution's specific criteria. The flags are there when they open the file.

You're not looking at a generic risk score. You're looking at a bank statement through the lens of your own policies.

How fraud and risk teams are putting it to work

A few examples based on how Inscribe customers are already using this:

Risky spending behavior

Flag if more than 20% of transactions in a 60-day period fall under gambling, or more than 20% are tied to crypto exchanges.

Suspicious money movement

Flag if average daily balance over 60 days falls below a threshold your team sets.

Low average balance

Flag if daily balance swings more than 40% of the average daily balance over 30 days.

High balance volatility

Flag if daily balance swings more than 40% of the average daily balance over 30 days.

What's more, before anything goes live, you can backtest against your existing applicant population to see exactly how many would be flagged and what your flag rate looks like.

You control how these insights affect the risk rating

Inscribe produces a customer-level risk rating across all submitted documents. With custom transaction insights, you choose whether each signal rolls into that rating or surfaces as additional context.

This also means you can test new signals without affecting decisions, phase in stricter rules gradually, or keep certain flags as FYIs before they influence outcomes automatically.

Built for the way fraud actually works

Off-the-shelf document fraud solutions tell you what they were trained to flag. But fraud has gotten more layered. 60% of cases now involve both identity and document falsification, up nearly 20% year over year.

Catching it requires detection that reflects how risk actually shows up at your institution, not someone else's.

Interested in learning more? Book time with us.