Title: Predicting loan repayment default among second tier borrowers in Ghana

Authors: Kwami Adanu; Emma Boateng

Addresses: Business School, Ghana Institute of Management and Public Administration (GIMPA), Achimota, Accra, Ghana ' Ghana Commercial Bank, Kisseman, Accra, Ghana

Abstract: This paper evaluates factors that affect the probability of loan repayment defaults using repayment data on loans granted by Snapi Aba Savings and Loans Ltd, Ghana. Results from logistic regression analysis of the data indicate that, contrary to findings elsewhere, age of borrowers is irrelevant in predicting probability of defaults in this study. Rather, credit screenings should focus on the gender of borrowers, payment frequency, loan cycle, and loan size. The probability of repayment defaults increases by about 1.3% for male borrowers as compared to female borrowers. Further, clients with longer credit history with the lending agency are more likely to default, and default probability is declining in loan size. Also, one-time lump-sum loan repayments reduce repayment defaults by about 2.2% as compared to regular monthly repayment options. Finally, educational level of borrowers and interest rates are irrelevant in predicting defaults.

Keywords: microfinance; logit; loans; small and medium-sized enterprises; SMEs; default prediction; loan repayments; repayment default; loan defaults; second tier borrowers; Ghana; credit screening; gender; payment frequency; loan cycle; loan size.

DOI: 10.1504/IJESB.2015.067466

International Journal of Entrepreneurship and Small Business, 2015 Vol.24 No.3, pp.417 - 432

Received: 25 Jun 2014
Accepted: 11 Sep 2014
Published online: 13 Feb 2015 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article