Forecasting GDP with energy series: ADL-MIDAS vs. Linear Time Series Models

In this paper, we offer the following contributions to the extant literature on the energy-growth nexus. First, we test the predictability of energy series in the predictive growth model using autoregressive distributed lag mixed data sample (ADL-MIDAS) approach. Second, we compare the in-sample and out-of-sample forecast performance of the ADL-MIDAS model with the linear time series models involving the first order autoregressive [AR(1)] model and the autoregressive distributed lag (ARDL) model. Third, we consider an array of energy proxies ranging from aggregate data to sectoral data of energy consumption (residential, commercial, industrial and transportation) and those defined by energy sources (petroleum, natural gas, coal, electricity, nuclear electricity and renewable energy). Fourth, we test whether accounting for asymmetries matters in the ADL-MIDAS regression model for the energy-growth nexus. The results support the significant predictability of energy for growth regardless of the measures of energy. In addition, the in-sample and out-of-sample forecast results overwhelmingly favour the ADL-MIDAS over the conventional linear time series models including the restrictive AR model. Thus, allowing for high frequency data for energy in the low frequency growth model will enhance the forecast accuracy of the model. However, we find that accounting for asymmetries may not improve the forecast accuracy of the ADL-MIDAS model in the energy-growth nexus since forecasts of the positive and negative asymmetric models do not differ significantly.