Logistic regression model of default risk in micro-finance a case study of financial services associatio (FSA) Mulot

Abstract

Default among loan borrowers is a common vice which has affected both the MFls and banks. The banks and MFIs offer loans to different categories of people and as a result different people usually perform differently in terms of repayment of loans. Among the challenges the banks and MFls faces in offering loans are default to loan repayment and prepayment of loans. There are different reasons that bring about the variation in people's repayment rate for example the financial condition, age, gender, level of education, occupation, demographic area marital status, loan period. For our research we looked at gender, marital status, demographic area and the occupation as factors that we thought they could affect the repayment of microcredit in MFls using data of 107 customer from Mulot FSA, a Micro-Finance institution based in Bomet County. The data was analyzed using STAT A software, the coefficient results of the four variables was used to come up with a logistic regression model. Based on our analysis we found out that among the variables that was studied only one out of the four variables significantly influence probability of default. According to our analysis gender significantly influences the probability of default while marital status, occupation and demographic area have no effect on the rate of default.