ABSTRACT

We model and forecast the loan collection of Ifedapo Ijebu – Igbo Cooperative Thrift and Credit Union Limited, Ogun State. We used the cooperative monthly loan data that spanned the period January 2007 to August 2021 and provided two years monthly forecast. We adopted the ARIMA model approach. The original series is mean stationary, and hence we differenced the series once. We automatically estimate about 170 ARMA (p, q) and 170 ARIMA (p, d, q) models using the Eviews software. The model with the lowest Alkaike Information Criterion (AIC), Bayesian Information Criterion (BIC) and Hannan-Quinn (HQ) values were selected from each group. We compare the performance of the two models, ARMA (1, 0) and ARIMA (0, 1, 1), using the forecast performance measures and observed that the ARIMA (0, 1, 1) model performs better in terms of forecast accuracy. We diagnose the selected model for residual normality and autocorrelation, and we found that the model is stable and reliable. Finally, the two years monthly forecast was made, which showed that the highest projected loan by Ifedapo Ijebu – Igbo Cooperative Thrift and Credit Union Limited, Ogun State, is N691,500 by November 2021, while the lowest projected loan is N575,500 by September 2021. Although, there is no indication of upward trend in the projected loan for the cooperative society, however, the forecast values could be an indication of a good development for the cooperative society knowing that the loan collected are mean to improve the standard of living of the cooperative members. We recommend that other sources of income for the cooperative should be actualized and maximized to further generate income that would be enough to satisfy the needs of the cooperative members.