Credit Scoring Model for Payroll Issuers: A Real Case

In this paper we present the development of a credit score model for payroll issuers based on a credit scoring methodology. Typically, in the Mexican banking system, it is common to provide and administer payroll service for companies via third parties (outsourcing). This service allows employees to get payroll loans of which periodic payment is retained automatically by the creditor. However, if their relationship with the company is lost, the payment is omitted incresing the risk of default. Addressing the problem described, a statistical model was built to predict whether a payroll issuer will churn in the next six months, this allows the decision maker to determine the appropriate business retention actions in order to avoid future payment loan losses. Results showed that the developed model facilitates a practical interpretation based on scoring system and showed stability when it was implemented.