Forecasting of OPEC's Global Crude Oil Demand using Vector Self-Engagement Models, Collective Exploration and Gravitational Search

Subject Areas : Futurology

heshmatolah asgari ¹ (Moderator, Department of Industrial Engineering, Payame Noor University, Tehran, Iran)
mohammadreza omidi ² (Associate Professor, Department of Economics, University of Ilam, Iran (Corresponding Author) [e-mail])
zahra malekinia ³ (Master of Science (MSc), Energy Economics Department, Ilam University)
ALIAKBAR OMIDI ⁴ (Student, PhD, Political Science, Iran, Islamic Azad University, Kermanshah Branch)

Received: 2017-10-30 Accepted : 2018-11-06 Published : 2019-01-21

Keywords: Prediction, OPEC, Global Demand,

Abstract :

Knowledge about future oil demand is essential for OPEC member countries to set priorities and select policies in order to achieve economic growth and development. So in this study, the OPEC oil demand has been predicted using time series models Including Structural Vector Autoregressive model (SVAR), Autoregressive Integrated Moving Average model (ARIMA) and Gravitational Search Algorithm (That is one of the Innovative Search Algorithms) applying demand data from 1970 to 2014. In this regard, three criteria including Mean Sum of Squared Errors (MSSE), Mean Absolute Error (MAE) and Mean Absolute Percentage Error (MAPE) have been used to measure the predictive power of triple models. Results indicate that the SVAR model has the most appropriate prediction of OPEC global demand. According to results of this model, net export variable has a positive and significant impact on oil demand and OPEC petroleum price and non- OPEC production variables have a negative and significant impact on oil demand.