Developing Credit Scoring Models When Small Sample Sizes Are Available

Abstract

Making lending decision is an important process for financial institutions because it has a direct impact on the profits and losses of financial institutions. Therefore, financial Institutions try to develop good credit scoring models to make lending decisions. The purpose of this research is to compare the performance of the credit scoring models between multiple linear regression and logistic regression. The comparison of the credit scoring models is done through using three sets of population data generated through simulation. The odds ratio is adopted in this research as an evaluation tool. The findings of this research are useful for financial institutions especially commercial banks because they present the evidence of how well each credit scoring model can predict the credit score of the loan applicants.