A novel quantitative forecasting framework in energy with applications in designing energy-intelligent tax policies

Energy, monetary and fiscal policies are going to play a crucial role towards the broad transformation of global energy. The design and optimization of these policies as well as the efficacy in achieving the desirable results require quantitative approaches. In this study, we extend the methodology of our novel quantitative framework, the Energy Price Index (EPIC) that represents the average price of energy in the US, to the design, optimization and assessment of energy-intelligent tax policies. The non-availability of actual energy demand and price data for the recent months is addressed with the introduction of a rolling horizon methodology that demonstrates excellent forecasting ability over a testing period of 181 months. The mean absolute percentage error of the initial and the 2nd adjusted EPIC are just 2.71% and 1.02% respectively over this period, with the forecasting framework demonstrating excellent robustness even during the unprecedented pandemic of COVID-19. The same framework can be used to forecast the energy demands for the next four years. Two case studies for energy-intelligent tax policies are presented, investigating parametrically the effects that a gasoline tax hike and a carbon tax could have retrospectively and prospectively. A hypothetical gasoline tax hike of 15 cents per gallon would increase EPIC by 1.6%, with the greatest increase of 3.67% being associated with the transportation sector (TEPIC). This policy would result in an estimated annual average revenue of $20.253 billion. Likewise, a hypothetical incremental carbon tax hike over the next 10 years would lead in substantial reductions in CO2 emissions, while the average price of energy would increase by just $1.5/MMBtu, generating though more than $110 billion in annual revenue.