A new PYMNTS Intelligence report published this week documents one of the most consequential blind spots in U.S. consumer banking: a stable population of 44 million subprime adults who actively seek credit, manage their finances with measurable discipline, and remain largely invisible to the legacy underwriting infrastructure that most banks still operate.
The study draws on 47 consecutive monthly surveys dating back to March 2022 and finds that subprime consumers have consistently represented approximately 17% of the U.S. adult population across that entire period. That persistence across four years of economic volatility distinguishes the segment from a crisis-driven phenomenon. Researchers describe it as a durable, identifiable market. For banks anchored to conventional FICO-based origination workflows, however, that durability represents a gap that is not closing on its own.
The Structural Problem in Bank Underwriting
The core issue is architectural. Traditional underwriting is built around revolving credit history, bureau files and repayment records accumulated through conventional accounts. That approach works for consumers who have participated in standard credit channels. It fails, however, for a significant share of the subprime population that operates outside those channels entirely.
Notably, 35% of subprime consumers hold no credit or store card at all, compared with just 4% of super-prime consumers. Standard origination systems cannot evaluate what they cannot see. As a result, these borrowers are not simply underwritten conservatively. They are not underwritten at all.
At the same time, 55% of subprime consumers report living paycheck to paycheck with difficulty paying bills, more than double the rate for the broader population. However, that statistic does not mean these borrowers are poor credit risks. It means they manage cash flow differently, with greater attention to payment timing, bill prioritization and liquidity preservation. The behavioral evidence exists. The models to read it, in most bank origination platforms, do not.
What Cash-Flow Signals Actually Reveal
PYMNTS Intelligence identified several behavioral patterns that may prove more useful for underwriting than static bureau data alone. Among subprime consumers who received tax refunds, 67% described the funds as critical or very important to maintaining financial stability. Most directed those funds toward recurring bills and essential expenses, not discretionary purchases. That pattern reflects disciplined liquidity management, not financial recklessness.
Healthcare spending also emerged as a meaningful signal. Among subprime consumers aged 18 to 43, 23% delayed a doctor’s visit because of cost, while 14% did not fill a prescription. Consumers who consistently negotiate bills, maintain payment plans and prioritize essential obligations demonstrate a repayment orientation that conventional scores do not capture. For this reason, cash-flow behavior and payment sequencing may offer lenders a clearer picture than bureau files alone.
Fintech Providers Are Already Filling the Gap
The competitive picture makes the cost of inaction concrete. The PYMNTS Intelligence data on installment provider usage shows Klarna, Sezzle, FuturePay and Quadpay/Zip all over-indexing with subprime users. By contrast, PayPal Pay in 4 and Uplift under-index sharply with the same population. The divergence reflects underwriting philosophy. Providers whose models accommodate near-term cash-flow signals and thin-file borrowers are reaching consumers that conventional bank origination screens out.
That market share is not theoretical. The PYMNTS Intelligence research characterizes the subprime segment as economically significant. As fintech providers accumulate origination data on this population, refine their models and deepen repeat-customer relationships, the gap between what they can underwrite and what banks can underwrite compounds over time.
The Revenue Cost of a Structural Blind Spot
For consumer lenders, the implication is clear. A segment representing 17% of the adult population, stable across nearly four years of monthly measurement, is not a marginal tail risk. It is a permanent feature of the credit market. Banks that cannot reach it through their existing origination workflows are not simply passing on higher-risk loans. They are ceding a measurable and durable portion of consumer credit demand to competitors whose infrastructure is designed for it.
The study suggests the underwriting question is evolving. It is no longer whether to serve subprime consumers. It is which behavioral signals best predict repayment resilience in this population. Banks that cannot answer that question with their current infrastructure are already positioned behind lenders that can.
