Forecasting Electric Vehicle Sales with ARIMA and Exponential Smoothing Method: The Case of India

The vehicle industry has shown rapid sales growth in recent years due to which the problem of carbon dioxide (CO₂) emissions is increasing continuously, to address this issue, electric vehicles (EVs) have emerged as a technological innovation aimed at reducing greenhouse gas emissions. Therefore, it is imperative to make accurate forecasts regarding future sales trends of Electric Vehicles (EVs). These projections are crucial for the strategic planning of EV manufacturing and for developing supportive legislation and ensuring an adequate energy supply. Therefore, a suitable model is required for EV forecasting. The objective of this study is to predict the sales of EVs to analyze the prospects of Indian EVs by considering three univariate models named ARIMA and exponential Smoothing. The monthly sales of electric vehicles (EVs) in India have been analyzed from January 2018 to August 2023 to predict the sales for the next two years. The ARIMA model and Holt's Exponential smoothing show contradictory results. The initial model forecasted a rise in sales, but the subsequent model projected a decline in sales. Furthermore, a simple exponential model provided the average forecasted figure for the sale of EVs for the next 24 months. The collective impact of ARIMA model and exponential smoothing led to establishing the structure for efficiently managing the inventory of electric vehicles to fulfill the demand seamlessly and provides valuable insights to the manufacturing industry, investors, and government.