Forecasting the Growth Rate of GDP and its Indicators Remittance and Labor Force Participation of Bangladesh using VECM and ARIMA Models

Abstract

The present study used VECM model and ARIMA model for modeling the growth rate of GDP, Remittances and total Labor Force Participation. All of the data of this study are collected from the World Bank Database with indicators as the time series variable from 1999 to 2020. The stationarity of the time series variables has been tested by the Augmented Dickey-Fuller test and investigated that they become stationary at first difference. The lag selection criteria has been applied to find out the no. of lags needed to be included in the model. Lag 2 has been selected for this study. VEC model has been applied in the time series data. Also the ARIMA model imposed on these time series variables. These two models have been used for forecasting. Among the predictions obtained by these two models the better one has been selected by using RMSE, MAPE and MAE. It has been found that for growth rate of GDP and total Labor Force Participation, the ARIMA model is more preferable but for Remittance, VECM is desirable. It has been found that the Remittance and Labor Force have significant positive impact on the growth of GDP. Also, a long run relationship has been found among the growth rate of GDP, Remittances and total Labor Force Participation.

Dhaka Univ. J. Sci. 70(2): 42-47, 2022 (July)