A regional lender came to us with a problem:
They were approving borrowers with high credit scores… who kept defaulting on mid-size personal loans.

Their model leaned heavily on score bands and stated income. But it wasn’t catching the deeper signal: a mismatch between what people said they could handle, and what they’d actually handled before.

What We Did

We built them a lightweight decision layer that calculated max historical exposure per borrower, using just the tradeline data in their credit pulls (TransUnion and Experian). No external data sources. No scoring black box.

Then we introduced a simple rule:

Don’t approve more than 75% of a borrower’s historical max exposure—unless they’ve held that level for at least 12 months without any lates.

We layered it into their pre-approval logic alongside score and DTI.

What Changed

In the first 90 days:

  • Early delinquencies dropped 23% in the 680–720 score band
  • Average loan size stayed the same—because high-capacity borrowers still passed
  • Manual review time went down, since edge cases were flagged more precisely

The kicker?
They didn’t lower their approval rate. They just stopped approving the wrong people.

Why It Worked

Max exposure created a floor for financial experience. It let them ask:

  • Has this borrower ever handled this much money before?
  • Did they make it through that period without slipping?

That was the real test—not the score on paper.

Credit reports tell more than just scores. If you know where to look, you can make better decisions with the data you already have.

Want a quick way to surface max exposure on your reports? We’ve got tools for that—drop us a line.