Toward Bracketing the Seasonal Export-Import of Bangladesh: A Time Series based Analytical Approach
Abstract
For any nation, to attain a better growth in business and development, prediction and forecasting is must to confront the gap between export and import. Identifying the trend in export-import as well as predicting the future values of these two sectors, is a major challenge for the national trade policy makers. Hence, to assist in decision making process this article is proposing two time series model based on export and successively on import using export-import data from 2000-2006 in context of Bangladesh, to obtain the information about current export-import trend as well as for future prediction, but still can be developed for any nation. To construct the mathematical export - imports models, different tools of Time series analysis, has been used. Autoregressive (AR) process along with moving average (MA) constitutes autoregressive and moving average (ARMA) process, which is used only to model stationary time series, was considered. In order to extend this model into nonstationary time series model, the concept of autoregressive integrated moving average (ARIMA) has been developed. To identify a perfect ARIMA model for a particular data series, Box and Jenkins methodology that consists of three phases, identification, estimation & diagnostic checking and application was applied. The proposed models also managed to identify the influences of export and imports in past years on export-import in current or future years.