Forecasting and the Causal Relationship of Sectorial Energy Consumptions and GDP of Pakistan by using AR, ARIMA, and Toda-Yamamoto Wald Models

The objective of this research was to forecast the sectorial energy consumption of Pakistan for five fiscal years, i.e., from FY18 to FY23 using two different time series techniques and explore the causal relationship between total energy consumption and its sectorial components, and Gross Domestic Product (GDP). The study further analyzed the efficiency of two different time series models, such as the Autoregressive model (AR with seasonal dummies) and Autoregressive Integrated Moving Average model (ARIMA/ARMA). In any economy, forecasting energy consumption and its relationship with GDP is paramount to ensure the economic development and fiscal policies. This study used components of total energy consumption (TEC) such as domestic energy consumption (DEC), commercial energy consumption (CEC), industrial energy consumption (IEC), agricultural energy consumption (AEC), transport energy consumption (TrEC) and other government energy consumption (OGEC). The data is taken from FY1977 to FY2017 (41 annual observations) and focused on forecasting for FY18 to FY23. For the forecasting of total energy consumption independently and taking the sum of all sectorial components. The results of this study revealed that among these models, the ARIMA model gives better-forecasted values for Pakistan's total energy consumption. The findings of the Granger causality test shows that there is no causal relationship between CEC, OGEC, and GDP variables and there is a one-way causal relationship between IEC & GDP, the direction is from IEC to GDP. The Toda & Yamamoto Wald model's findings demonstrated similar results, and there is one-way causation from IEC to GDP.