A novel hybridization of nonlinear grey model and linear ARIMA residual correction for forecasting U.S. shale oil production

Abstract

Given that the booming U.S. shale oil has reshaped the global oil landscape, better forecasting output of U.S. shale oil can facilitate to better understanding the global oil trend. To more accurately forecasting U.S. shale oil production, we hybridize the nonlinear- and linear-forecasting model to a new forecasting technique in two step: (i) combining a nonlinear grey model with the mentalism idea to develop nonlinear metabolism grey model (NMGM), (ii) combining the proposed NMGM with Auto Regressive Integrated Moving Average (ARIMA) to develop NMGM-ARIMA technique. The sliding validation experiments with the previous U.S. shale oil quarterly output dataset of 2003–2008 and 2008–2013 are adopted to verify the feasibility of the proposed NMGM-ARIMA technique. The empirical results show that this novel NMGM-ARIMA technique can significantly improve forecasting effectiveness and outperform other related four forecasting models. We therefore believe that the proposed forecasting technique can service researchers, policymaker, analysists and others stakeholders to more timely and accurately forecasting U.S. shale oil output and thus better understanding global oil market.