Credilogros Cía Financiera S.A. is a credit company based in Buenos Aires, Argentina, and owned by the gST Group (Grupo de Servicios y Transacciones). Its assets are valued at $95.7 million and it is the fifth largest credit company in Argentina.
After acquiring Credilogros in 2006, gST began a search for a new system to enable the company to better manage customer credit risk. It was important for Credilogros to identify the risk potential associated with prospective customers up front, so that risk exposure could be minimized.
The company's first objective was to implement a decision engine to process credit applications to interact with the core system of the company, as well as with the systems of two credit report companies. Credilogros was also looking for customized risk scores for the low-income customer segment it serves. Additional requirements included that the solution operate in real time from any of its 35 branches, as well as from the more than the 200 associated points of sale, including retail appliance chains and a cell phone company.
After evaluating several competitors, Credilogros chose PASW Modeler, SPSS Inc.'s data mining software, due to its flexibility and ease-of-use in integrating within Credilogros' core information system.
Two scoring models were developed—one for clients with a credit history and the other for new clients. Both could be combined with scores provided by a credit rating company in order to increase predictive power in classifying a potential client. The system also could determine the maximum credit limit given to a customer, based on paying ability and risk level predicted by the models.
Prospect data was then collected in a Web application and sent to the decision engine designed within the SPSS Inc. data mining workbench. In the first stage, basic requirements defined by market rules are checked. Then the demographic data provided by prospects is validated against an external database, and the bureau score is requested online.
All of this information is fed to the software's model to obtain the final score for each prospect. The system can also manage business strategies to determine the appropriate credit limit for each individual. Depending on the limit assigned by the system and the level of risk, Credilogros can also handle cross-selling opportunities by offering a second product, such as a credit card, at the same time.
By determining credit risk in advance, we are better positioned to make intelligent credit-related decisions, protecting our bottom line.— Carlos Risetti
Credilogros Cía Financiera S.A.
By implementing PASW Modeler, Credilogros reduced the time it takes to process credit data and deliver a final credit score to less than eight seconds. This enables the organization to approve or reject the credit request with no delay.
This decision engine also enables Credilogros to minimize the number of identification documents each customer has to present, as a large amount of the data is validated online. In some special cases, credit can be approved just by presenting one piece of identification.
Credilogros is currently processing an average of 35,000 applications per month using PASW Modeler. The implementation was simple and was completed in 11 weeks.
Additionally, the system provides monitoring capabilities that have helped Credilogros reduce delinquency on loan payments by 20 percent, after only three months of implementation.
"PASW Modeler allows us to accurately assess the credit worthiness of potential customers and gain real-time results to determine whether to extend credit and, if so, how best to minimize the risk level and maximize the potential return to the organization," said Carlos Risetti, risk manager at Credilogros. "By determining credit risk in advance, we are better positioned to make intelligent credit-related decisions, protecting our bottom line."
can make your organization