A Comparison of the VAR Model, ARIMA Model and Holt’s Exponential Smoothing Approach in Forecasting Remittance and GDP for Bangladesh

Abstract

As remittance is one of the most crucial factors driving our economy, this study employs various statistical techniques to analyze its impact on the Gross Domestic Product (GDP). First of all, Vector Autoregressive (VAR) model and Granger causality test were applied to assess the influence of remittances on GDP. Subsequently, the study utilized Autoregressive Integrated Moving Average (ARIMA), VAR and Holt’s exponential smoothing technique to analyze and forecast the selected variables: Remittance and GDP. The yearly time series data on GDP and remittances were sourced from the Journals of Bangladesh Bank covering the period from 1990 to 2022. The VAR model analysis revealed a significant relationship between remittances and GDP, indicating that remittances play a vital role in driving economic growth. Forecasting remittances and GDP is crucial because these variables are essential indicators of economic stability and growth. By accurately forecasting remittance inflows and GDP trends, policymakers and businesses can better plan for future economic conditions, manage foreign exchange reserves and make informed decisions about resource allocation. Additionally, ARIMA and Holt’s exponential smoothing technique were used for forecasting purposes and a comparison of these methods indicated that the VAR model is more efficient, particularly when applied to remittance and GDP data.

Dhaka Univ. J. Sci. 74(2): 243–256, 2026 (July)