Energy is one of the most important resources of the national economy, which plays an importantrole in economic production and life more generally. Given its significance, this paper formulates predictionmodels for electricity consumption (EC), peak load (PL) and gross domestic product (GDP) in Saudi Arabiaby employing the Autoregressive Integrated Moving Average (ARIMA) model; using time series data from1990–2015. It also examines the relationships between EC, PL and GDP through a vector auto-regression(VAR) analysis, which includes Granger causality (GC) testing, impulse response, and forecast error variancedecompositions (FEVD). The results show that ARIMA (1, 1, 1), ARIMA (0, 1, 0) and ARIMA (0, 1, 0) werethe most appropriate univariate models of EC, PL and GDP, respectively, based on the Akaike informationcriterion. The results also revealed significant unidirectional granger causality from PL to EC and PL to GDP.The variance decomposition reveals that in the case of EC, the major changes arise from its own innovationand the contribution from GDP at the 1%.
展开▼
