The loan officer's subjective judgment and its role in microfinance institutions

Abstract

Microfinance institutions can classify their customers into different risk classes either empirically, by means of credit scoring models, or manually by having human judges recollect their professional experience. Regarding the applicability of credit scoring models in microfinance institutions, recent studies have concluded that credit scoring is still unable to fully replace the traditional human loan assessment procedure, though it can stand as a complementary decision-support tool that helps increase operational efficiency and reduce cost. In fact, while the eligibility of loan officers’ subjective judgments is widely accepted as input for final scoring model, this variable real contribution has not been documented above. In fact, on applying data collected from Tunisian microfinance bank, we have found some evidence that the combined use of hard as well as soft information leads to a higher prediction accuracy of potentially default events than does the separate use of single hard information. On average, loan officers’ subjective judgments have helped reduce the bad loans’ proportion classified as good loans by 1.31%, and that of good loans classified as bad ones by 1.59%. The loan-officers’ judgment role varies with respect to loan types. As a matter of fact, it turns out to vary according to loan purpose and firm activity sector.