Predicting Credit Default Among Microfinance Customers at Eclof Kenya Limited

Abstract

Microfinance Institutions (MFIs) carry out a crucial role in the financial sector for economic development of many countries. The MFIs advance credit to the active poor who mostly lack formal collaterals to secure the loans. The advancement of loans can lead to the MFIs to suffer financial crisis/distress called credit default causing financial losses. Therefore, all MFIs are concerned whether the applicant of the loan will become a good or bad payer in order to minimize probability of loan repayment default. The Financial Reports from Central Bank of Kenya show that MFIs continue to incur losses emanating from Non-Performing loans due to loan defaults. The study sought to predict the credit default in MFIs customers. The research objectives entailed finding the factors that are important in establishing credit customers' default risk and to evaluate the relative degree of the importance of each of the factors that affect credit default in MFI customers using Altman Model. The study applied a descriptive research design. The population target was the 2000 EKL credit customers of Kisumu branch. Stratified Random sampling was used among the target population to get defaulted and non-defaulted loans for analysis. Secondary Data was obtained from loan applications advanced during the year 2018. The data extracted included the borrower and the loan factors such as Gender, Age, Marital Status, Guarantor, Loan Amount, Loan Term, Weekly Installments, LSF Contribution, Loan Purpose, and Loan Cycles. The sample size constituted 35 defaulted loans and 35 non-defaulted loans that were randomly selected from each category of defaulted and non-defaulted loans that formed 70 cases. The SPSS (IBM SPSS statistics 23) software and Discriminant Analysis were employed for data analysis. Loan Repayment status formed the dependent variable, while Loan and Borrower’s characteristics were the independent variables. The findings of the research showed a statistical significant relationship between Borrower characteristics and Loan characteristics and Repayment status. The study indicated Loan Cycles, LSF Contribution, Weekly Installments and Loan Amount were more important in discriminating default and non-default categories hence determining the credit default of the borrower. The study also highlighted LSF Contribution, Number of Loan Cycles, Weekly Installments and Loan Amount as the highest predicting factors in the model. The study established that the Discriminant Analysis Model was able to predict default cases by 82.9% and non-default cases by 88.6% indicating that Altman MDA model is a strong model with a high prediction rate that can be used to predict the default status of credit clients. The study recommended that MFIs should consider xi weekly LGF Contributions by all clients, do thorough loan appraisals to determine the client’s potential to pay the loan. It is important for the MFIs to fund the right loan size that leads to good loan repayments while avoiding default. Again, the study recommended use of Altman Model-Discriminant Analysis by MFIs to discriminate good and bad borrowers in order to minimize the default risk in lending