Forecasting Loan Consented for Vehicle Purchase in Malaysia

Abstract

Banks are among the institutions that contribute significantly to the country's economic development. Thus, banking sectors play a critical role in the development of the country's economy. Providing credit or a loan is one of the most important services that any bank can do. In this study, a hybrid model was developed, and several existing time series models, such as the seasonal auto regressive integrated moving average (SARIMA), multilayer perceptron (MLP) neural network and a hybrid model were used to forecast the loans approved for purchase of vehicle in Malaysia. The hybrid model is a combination of linear and nonlinear model which is combination of Holt-Winter’s and single exponential smoothing models. Mean absolute percentage error (MAPE) and root mean square error (RMSE) are used to assess the accuracy of the forecast. From the findings, the artificial neural network gives the best forecast compared to the other two models. In conclusion, the advanced model such as MLP gives better forecast compared to the SARIMA and hybrid model. This finding could help bank institution make decision for the future pattern of the loan consent for vehicle purchase.