Statistical model for personal loan prediction in bhutan

Abstract

© 2019, Institute of Advanced Scientific Research, Inc.. All rights reserved. Banking plays a vital role in functioning the economy of a country. All over the world banks are one of the institutions which significantly contributes towards economic growth of the country. One of the major role played by every bank is providing credit facilities. Time series models help to predict and forecast the number of credit borrowers in future, which would help the concern authority to plan and work accordingly. In this study Box-Jenkins approach of time series was used to model and forecast the number of personal loan consumers at Bhutan Development Bank in Bhutan. The monthly number of personal loan consumers from January 2011 to June 2016 is modelled by ARIMA model of Box-Jenkins approach. A comprehensive study shows that ARIMA(2,1,2) works efficiently in forecasting future number of personal loan borrowers. The best fitted models were tested based on forecast accuracy test such as Root Mean Square Error (RMSE), Mean Absolute Error (MAE) and Bayesian Information Criterion (BIC).