Forecasting Lending Interest Rate and Deposit Interest Rate of Bangladesh Using the Autoregressive Integrated Moving Average Model

Abstract

The purpose of this paper is to predict the lending interest rate and deposit interest rate of Bangladesh using the Autoregressive Integrated Moving Average (ARIMA) model. The dataset collected from World Bank Open Data consists of 46 years of secondary data from 1976 to 2021. Box-Jenkins (BJ) model has been adopted to prepare the appropriate ARIMA model based on three parameters (p,d and q). Six ARIMA (p,d,q) models have been estimated and to check the goodness of fit among the estimated ARIMA models such as ARIMA (1,0,1), ARIMA (1,1,1), ARIMA (1,1,2), ARIMA (1,2,1), ARIMA (2,1,2) and ARIMA (2,2,1), AIC (Akaike Information Criterion) and BIC (Bayesian Information Criterion) index values have been calculated. ARIMA (1,0,1) model has been found appropriate in predicting the lending and deposit interest rates from 2022 to 2026. Diagnostic tests such as Dickey-Fuller unit root test and Correlogram have been conducted in order to make the dataset and residual of the selected model constant across time or stationary. Finally, based on the ARIMA (1,0,1) model, lending interest rate and deposit interest rate are predicted for the next couple of years revealing the increasing trend of the predicted values which are subject to the adjustment due to macroeconomic conditions and policy implications. This paper shall provide important insight of the lending and deposit interest rates in Bangladesh to the managements, regulators and other concerned parties.