ARIMA and VAR Modeling to Forecast Malaysian Economic Growth

Nur Azreen Abdul Razak, Azme Khamis, Mohd Asrul Affendi Abdullah


This study presents a comparative study on univariate time series via Autoregressive Integrated Moving Average (ARIMA) model and multivariate time series via Vector Autoregressive (VAR) model in forecasting economic growth in Malaysia. This study used monthly economic indicators price from January 1998 to January 2016 and the economic indicators used to measure the economic growth are Currency in Circulation, Exchange Rate, External Reserve and Reserve Money. The aim is to evaluate a VAR and ARIMA model to forecast economic growth and to suggest the best time series model from existing model for forecasting economic growth in Malaysia. The forecast performances of these models were evaluated based on out-of-sample forecast procedure using an error measurement, Mean Absolute Percentage Error (MAPE). Results revealed that VAR model outperform ARIMA model in predicting the economic growth in term of lowest forecasting accuracy measurement.


univariate; multivariate; growth; forecast; ARIMA; VAR; MAPE

Full Text:


Copyright (c) 2017 Journal of Science and Technology

ISSN : 2229-8460

e-ISSN : 2600-7924

Creative Commons License
This OJS site and its metadata are licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.