Plaid has launched the Cash Advance Index, a risk-scoring solution built for cash advance and earned wage access providers.
The solution addresses a structural gap in how cash advance providers evaluate borrowers. Most existing decision-making relies on cash flow data and point-in-time signals drawn from one bank account, which produces an incomplete view of user behaviour.
In addition, as consumers increasingly use multiple income sources and several lending, BNPL, and cash advance applications simultaneously, that gap has widened. With this in mind, Plaid has positioned the CAI as a response to that growing blind spot.
How the scoring model works
The CAI generates scores ranging from one to 99, predicting the likelihood of repayment within 30 days for advance amounts between USD 25 and USD 500. A higher score indicates a greater likelihood of repayment. The model draws on both cash flow data and proprietary signals derived from account connection activity across the Plaid network, and has been trained on actual cash advance repayment outcomes.
Two distinct models underpin the product. An onboarding model was designed for first-time users, assessing repayment likelihood to determine appropriate advance sizing from the outset. In addition, an ongoing model applies to repeat users, incorporating repayment history and usage patterns to refine risk signals over time. The CAI can also be deployed at the repayment stage, allowing providers to identify users whose risk profiles have shifted since the original advance was issued.
A key differentiator is the network-level visibility that underpins the scoring. Because Plaid's infrastructure spans thousands of applications and services, it can surface behavioural patterns that single-account data cannot, such as a user establishing new connections to multiple cash advance providers within a short window or drawing advances from several bank accounts simultaneously. These signals indicate concurrent borrowing against the same paycheck, a pattern the company identifies as a marker of elevated repayment risk.
Performance and industry context
According to the official press release, in a live production test, the CAI produced an 8% relative reduction in delinquency with no change in approval rates. The model was reported to be 17% more predictive in distinguishing between low- and high-risk users, and twice as performant as alternative models used by the same provider.
The launch is relevant against the backdrop of the US consumer credit environment, where, according to figures cited by Plaid, approximately 67% of consumers live paycheck to paycheck. Cash advance and earned wage access products have expanded significantly in recent years as an alternative to traditional short-term credit, but risk assessment in the segment has remained constrained by limited data sharing across providers.
Moreover, through the process of leveraging network-wide data, the CAI offers providers a way to extend access to higher-risk consumers while managing delinquency rates, a balance that has been difficult to achieve using conventional cash flow analysis alone. The product was designed to integrate across multiple decision points in a provider's workflow, from initial qualification through to repayment management.
