Demand forecasting and inventory management for an automotive tire distributor in Northeast Thailand

Abstract

The objective of this research is to improve the inventory management for an automotive tire distributor company in Northeast Thailand to mainly reduce the excessive unnecessary stock. According to the initial investigations, the main causes of overstock are unstable customer demands and a lack of an inventory control system based on theory. In order to better address the issues and scope down the research area, three suitable products and sub-dealers with different demand patterns are selected to find the best-fit forecasting model for each of them and be customized the inventory policy based on the demand patterns. The methodology is divided into two main parts which are the demand forecasting model and inventory control policies for both sub-dealers and the company.

Regarding the development of more accurate demand forecasting models, there are six forecasting models which are three months moving averages, single exponential smoothing, double exponential smoothing, Holt’s winter method, linear regression, and ARIMA conducted to select the best model. The study reveals that Holt’s winter and ARIMA model provides the least forecasting errors as measured by Mean Absolute Percentage Error (MAPE).

According to a proper inventory policy and more accurate demand forecast value from the suitable model, sub-dealers and a company can save up to 47%, and inventory turnover increases by up to 80% while inventory holding days decreases by up to 55 days.