An Application of Credit Scoring Models to Microloans in the Us

Nancy A. White, Hofstra University

ABSTRACT
A substantial body of research has demonstrated that the credit risk determinants for large corporations differ from those for SMEs (small and medium enterprises). As a result, credit models developed to assess default probabilities for large corporations have been modified to incorporate these differences. One change that has improved model performance for SME's is the addition of qualitative variables, such as firm structure and industry type. The objective of this paper is to extend earlier work on the measurement of credit risk in SME's to the analysis of microloans granted in the United States, specifically those granted by Community Development Financial Institutions (CDFI's). Our data set includes qualitative and quantitative information about borrowers that will be used to (1) develop a credit risk profile of this CDFI's clientele (2) develop a credit risk model that includes quantitative and qualitative variables to assess borrowers' default probability. Results will enable CDFI's to identify borrowers that need additional monitoring and assistance and will also be used to compare credit risk of CDFI issued loans to those issued to SME's.

(Return to Program Resources)

Updated 03/19/2014