Abstract

In the forecasting literature, it has been observed that simple techniques can be employed unless there is a strong case for complexity in the historical data. This paper studies the annual trends and patterns in the domestic sales of the four main segments of the Indian automobile industry namely the passenger vehicles, commercial vehicles, three-wheelers and two wheelers in the last two decades from 2001-02 to 2020-21 in which the benefits of opening the economy after the reforms of 1991 seemed to be realised with an investment in equity, technology transfer and managerial skills that would reflect the reliable prediction for the year of 2021-22. The data was obtained from the secondary source mainly from the Society of Indian Automobile Manufacturers’ (SIAM) database. The forecasting methods comprising of regression linear trend, Simple Moving Averages (SMA), Weighted Moving Averages (WMA), Brown’s Exponential Smoothing (ES), and Holt’s Double Exponential Smoothing (DES) have been used for the time series based on annual data and these techniques are evaluated on the basis of Mean Absolute Deviation (MAD), Mean Absolute Percentage Deviation (MAPD),Root Mean Square Error (RMSE) from Mean Square Error (MSE) .It was observed that out of all the methods Holt’s DES was carrying the least MAPD for the passenger vehicles, for the commercial vehicles and three wheelers and for two-wheelers , MAD was minimum with DES in the passenger vehicles, for the commercial vehicles and for two-wheelers but not in three wheelers segment, and RMSE value was minimum for ES method in commercial vehicle, three wheelers and two wheelers’ forecast estimates for the passenger vehicles, for the commercial vehicles and three wheelers and for two-wheelers.

Keywords: Indian Automobile Industry, Forecasting, Domestic Sales