Competition can help predict sales

This paper develops linear and nonlinear forecasting models to propose a sophisticated and accurate forecasting method in a fiercely competitive environment, such as the U.S. auto industry. Our results indicate that companies could operate successfully in a highly competitive market by using the competitors' sales to accurately predict their sales and plan for raw material, production, and finished goods inventories. Our suggested methodology is beneficial when the competitors are within similar strategic groups. The data for this study are obtained from the “U.S. Automotive News” data services, which contain time series records for inventory and sales for multiple automakers. To keep the analysis straightforward, we have chosen data for four major automotive companies known for their high‐level competition: the General Motors Company, the Ford Company, the Toyota Corporation, and the Honda Company because of intense rivalry due to competing within the same strategic business units. The results show a benefit is achieved by including the total sales for at least one competitor in the linear or the nonlinear forecasting models to predict domestic sales for the desired company.