Real-Time Data Helps Subprime Lenders Reduce Risk Without Cutting Approvals

Plaid has released its LendScore system that tries to compensate for shortcomings in the conventional credit scores often dependent on limited or stale information. 

The new system should favorably impact subprime borrowers’ access to credit.

LendScore provides time-sensitive information regarding income and cash flow patterns, together with the type of financial apps that applicants use. The new system is designed to reduce lender risk and support cheaper credit.

Plaid already has an open banking network that connects financial accounts with thousands of apps and services. LendScore builds on this foundation by allowing lenders to see customer-permitted transaction data. The data shows cash flows — income deposits, bills, and spending.

LendScore isn’t wedded to the same old credit bureau reports. It instead uses current data to assign a customer score from 1 to 99. Lenders can access the score and associated reason codes through the system API. 

The system offers various “network insights.” For example, it explains that borrowers who use at least four earned wage access apps are 22% more likely to default on a loan. This risk indicator is generally not available elsewhere.

Tests proved that combining LendScore with traditional credit data improved default predictions by 25%. The same combo helped lenders reduce non-prime defaults by 10% to 20%. All this without tightening loan approvals.

LendScore can provide lenders with a more accurate picture of risk to improve their assessments.

These are valuable improvements for subprime lenders who operate on tight margins, so every default is a big ouch. With LendScore, lenders should be able to land more borrowers who used to be routinely denied credit. More loans and less risk equals more profits — what’s not to love?

It’s a boon for borrowers who want to rebuild their credit after financial setbacks such as missed payments or charge-offs. LendScore reveals how these borrowers currently manage their finances without harping on previous screwups.

“Traditional credit scores still matter, but they rely on historical data and miss what’s happening right now in a borrower’s financial life,” said Plaid Product Lead Michelle Young.

The math for lenders is simple — more accurate data yields fewer defaults and stronger loan performance. Consumers pay less and get better terms.

Old-style scoring models often overlook the financial realities of many Americans, including gig workers, freelancers, and part-timer employees. Lenders who rely solely on credit reports often reject subprime applicants unnecessarily.

Plaid is part of a growing shift toward cash-flow based underwriting. Fintechs and subprime lenders are getting new tools from Experian, Equifax, Prism Data, Nova Credit, and others.

The trend is toward data-rich scoring that makes it easier for subprime borrowers to get approved. Using LendScore and other models helps lenders make smarter decisions, and consumers benefit from greater access.

Plaid’s focus is on domestic data. By way of contrast, Nova Credit helps immigrants use foreign credit histories to establish U.S. credit. The growing number of alternative models might use different technologies, but all try to increase access to credit.

Plaid claims this tech has the potential to reduce personal lenders’ credit losses by more than $1 billion and lower borrowers’ average APR by around three points. Those are cost-effective moves that get LendScore closer to fair lending for everyone.

Plaid’s LendScore offers a forward-looking way to understand credit risk.

With the integration of real-time financial information, signals such as app utilization, and fairness verification by third-party audits, it may enable smarter, quicker lender decisions — and afford greater access to credit for subprime and near-prime consumers.