Customized DFS UPM’s increase dealership profit by substantially reducing quick repossessions and “pushing out” the timing of the later repossessions to where their profit contribution exceeds the repossession loss. The UPM’s are based on artificial intelligence, specifically behavioral modeling; the result is a prediction, not the traditional score produced in the consumer credit industry.

Customized DFS formulates the UPM’s by analyzing the historic loan performance, application and credit bureau data for each particular dealer’s separate BHPH locations. Customized DFS never uses generic data.

Each dealership has the choice of whether each of its locations utilize the models specific to each particular location, or whether they utilize the same set of models for all of the specific dealership’s locations. Following the dealership’s enrollment, the UPM’s are refined from time to time based on the additional data generated.

Dealership employees enter data via the web into the UPM and receive an instantaneous decision: “Approved”, “Approved-Marginal”, “Modify” (deal structure), “Rejected-Marginal’, or “Rejected”.

“Rejected” and “Rejected-Marginal” decisions state the basis as credit, stability or income – or a combination thereof – with detailed reasons within each rejection category. As Customized DFS does not obtain addresses for applicants and the dealership might have compelling reasons to override the UPM decision, the dealership has the responsibility for providing turndown letters to applicants. Further, the UPM data entry does not include social security numbers, as the dealership employee is entering data from the credit bureau report that the dealership has pulled.