China's dependency on foreign oil will exceed 80% by 2030: Developing a novel NMGM-ARIMA to forecast China's foreign oil dependence from two dimensions.

Abstract China is the world's largest net importer of oil and the second largest oil consumer; consequently, changes of China's foreign oil dependence significantly impact both the Chinese and the international oil market. To enhance the forecasting ability of China's foreign oil dependence, this study combines the nonlinear metabolic grey model (NMGM) with the linear autoregressive integrated moving average model (ARIMA), thus obtaining the combined NMGM-ARIMA model. The proposed technique uses the linear ARIMA to correct NMGM forecasting residuals, thus improving forecasting accuracy. The proposed technique achieves a mean absolute error of 2.1–2.3%, reflecting its high reliability. The proposed NMGM-ARIMA was used to forecast China's foreign oil dependence for the period of 2017–2030 from two dimensions. For the first dimension, the gap between China's oil demand and supply was forecast. To fill this gap, China has to import oil; therefore, this gap is responsible for China's foreign oil dependence. For the second dimension, the change of China's foreign oil dependence level was directly forecast. Both dimensions indicate a similar conclusion, namely that the Chinese foreign oil dependence level will increase from 65% in 2016 to over 80% in 2030. A high level of 80% dependence on foreign oil would bring major concern to China. The policy recommendations given at the end of the paper will help China's decision makers respond appropriately. Highlights • The proposed forecasting technique uses the linear ARIMA to correct NMGM residuals. • The mean absolute percent error of the proposed technique is 2.1–2.3%. • Forecasting China's oil gap between production and consumption for 2017–2030. • Forecasting China's foreign oil dependence for 2017–2030. • Two-dimensional forecasting results show China's foreign oil dependence will exceed 80% by 2030. [ABSTRACT FROM AUTHOR]