We’re excited to announce our new Small Dollar Loan Score. This new score offers a risk assessment of a user’s ability to repay a small dollar loan.
The Small Dollar Loan Score evaluates the probability of a user making their first four small dollar loan payments. It provides a series of four probability scores, one for each of the initial four installments.
Small dollar loans, typically capped at $5,000, are unsecured, short-term installment loans. Repayments are structured in installments for up to two years. They provide quick financial relief to borrowers who may not have access to other sources of credit (e.g., high credit card limits or larger and longer term personal loans).
The Small Dollar Loan Score is a predictive machine-learning model that leverages Cashflow Attributes calculated from the user’s real-time transaction history and balance data, updated with each new data upload.
The Small Dollar Loan Score is trained on an expansive dataset of loan delinquencies and defaults and hundreds of Attributes such as overdraft fees, changes in income, balance trends, and more.
The Small Dollar Loan Score is used to:
Read more about Small Dollar Loan Score use cases here.
By implementing Pave’s Small Dollar Loan Score, lenders can drive improvements across the lending experience. This ensures lenders can safely extend credit, manage risks, and support financial health.
Book a demo to see how the Small Dollar Loan Score can be applied to optimize your risk decisions.
I understand that the probabilities decrease from the 1st to the 4th payment. Is this always the case?
This won’t always be the case. There are 4 separate models trained to predict the first 4 installments. This means in some cases, a user may end up with higher probabilities for later payments. However, we’ll analyze this and put in place mitigating factors. For example, passing the probabilities of payments for the previous installments to the model.
What happens if I rescore the same person’s loan application after several installments? They might have missed a payment or two, or they might not have. Do you account for these changes?
The user’s scores may change with new transactions and missed/paid installments. For example, if a user missed a payment, became unemployed, has seen soaring debts or other expenses vs their income, etc., they’ll likely have lower scores.