Identifying Key Predictors of Loan Default in Kenyan Microfinance Institutions: A Backward Logistic Regression Analysis in Makueni County, Kenya

Microfinance institutions (MFIs) play a critical role in improving financial inclusion in Kenya; however, high loan default rates continue to threaten their financial sustainability. This study aimed to determine the predictors of loan default using logistic regression among microfinance institution clients in Makueni County, Kenya. The study adopted a quantitative research design and used secondary data comprising 4,592 borrower records obtained from selected MFIs operating within the county. Data analysis was done using python programing. Borrower socio-economic and financial attributes were extracted from loan records and preprocessed through data cleaning, normalization and encoding procedures. Prior to model fitting, logistic regression assumptions, including multicollinearity, independence of observations and model goodness-of-fit, were assessed and found to be satisfactory. The dataset was split into training and testing sets using an 80:20 ratio. A binary logistic model was fitted with all predictor variables and was found to be significant (χ2 (19)= 2695.70, p < 0.001) at α=0.05. Backward logistic regression was then run to identify the key predictors of loan default. The findings revealed that key predictors of loan default included loan amount, interest amount, outstanding loan balance, loan term, number of serviced loans, employment status, loan purpose and borrower type (new or repeat clients). In conclusion, these results indicate that loan default is primarily driven by financial exposure (loan amount, outstanding balance, repayment term) and borrower characteristics (employment status and repayment history). These findings demonstrate the effectiveness of data-driven credit risk modelling in improving loan approval decisions and enhancing financial sustainability of MFIs in rural Kenyan contexts.