Paraguay-based Financiera Familiar is in the risky business of loans. Established in 1967 as a micro-credit institution for consumer goods, today it specializes in micro-credit loans to low-income persons.
Paraguay's gross domestic product (GDP) per capita has been declining over the past 10 years. The bleak economy, combined with a hyper-competitive financial market, prompted Financiera Familiar to look into solutions that would improve the profitability of its loans without sacrificing the speed in which loans were qualified and approved.
Financiera Familiar developed a credit scoring system to better predict loan risk and more profitably manage its assets. The institution's risk manager used SPSS for Windows and SPSS Regression Models to develop the scoring system, which is capable of predicting the risk of $9.3 million in loans spread over 45,000 customers.
Financiera Familiar, a Paraguay-based financial institution founded in 1967, is in the high-risk business of providing loans. Originally established as a micro-credit institution for consumer goods, today it focuses on offering micro-credit loans to low-income individuals. The micro-loan segment in Paraguay is comprised of credit cards (23 percent), loans targeted at micro-enterprises (35 percent), and loans aimed at customers with fixed salaries (42 percent).
Due to Paraguay's economic instability, the risk accompanying low-income loans has become more acute. To better understand potential loan problems, Financiera Familiar's risk manager, Diego Balanovsky, used PASW Statistics and PASW Regression to develop a credit scoring system capable of accurately forecasting the risk associated with its customers—in sum, $9.3 million in loans spread over 45,000 individuals.
This process would have been impossible without SPSS Inc. and the people at SPSS Inc. Argentina. Not only were the applications very user-friendly, we appreciate the high quality of SPSS Inc. Argentina's post-sales support. We feel these factors have greatly helped turn the company around, and have helped our department grow professionally by learning to deploy new techniques.Diego Balanovsky
Balanovsky has scored 58,000 loans over the past two years with the help of SPSS Inc. software. He was able to derive a credit scoring model that allowed the bank to accurately predict which loans would be paid on time versus which loans would be paid late. The system also gave Financiera Familiar an accurate forecast of the percentage of loans approved by risk segment.
The credit scoring system provided the basis for Financiera Familiar to implement new procedures that segmented loans by predicted risk, giving smaller loans to high-risk customers and larger loans to lower-risk customers. The bank was able to adopt organizational policies, allowing them to increase the number of loan application approvals while keeping the default rate under control.
As a result of the new credit scoring system and subsequent procedural and policy changes, the bank has been able to decrease late payments from 14 to nine percent. Additionally, the bank can more efficiently use its staff during the application assessment process, resulting in better-served and happier customers.
One of the most significant benefits of installing the scoring system was the elimination of approximately $350,000 in unrecoverable payments per year. By anticipating the risk level for loan applicants, Financiera Familiar was able to confidently loan more money to low-risk customers and less money to high-risk customers. As expected, errors in projected loan estimations decreased, resulting in a reduction of annual loan default loss by 35 percent.
"The added value we appreciate the most is being able to predict the risk involved with a specific loan request," said Balanovsky. "The accuracy of these predictions has created greater confidence in the loans we give and insight on potential loan recipients."
Financiera Familiar is not content to rest on its success. In the near future, it is seeking to develop a model based on its loan qualification process for credit cards and loans for micro-enterprises. Given the company's success working with SPSS Inc., the bank has also decided to acquire additional software such as AnswerTree® and PASW Advanced Models™ to work on other phases of the credit cycle, such as retention and collection.
"This process would have been impossible without SPSS Inc. and the people at SPSS Inc. Argentina. Not only were the applications very user-friendly, we appreciate the high quality of SPSS Inc. Argentina's post-sales support," said Balanovsky. "We feel these factors have greatly helped turn the company around, and have helped our department grow professionally by learning to deploy new techniques."
can make your organization