Investigating the Determinants of Inventory Holding and Ranking them Using Decision Tree and Neural Network Algorithms

Inventories constitute the main part of firms’ working capital. Maximizing the return from inventory holding is one of the financial managers’ goals and to achieve this purpose, an optimal level of inventory holding must be identified by manager and this requires the identification of factors affecting firms’ inventory holding. In this research, first the determinants of inventory holding are identified using multivariate regression analysis in 158 firms from 2002 to the end of 2013 and then the importance of significant factors is ranked using decision tree and neural network algorithms. The results of regression analysis show that except firms’ leverage and age, other research variables show a significant relationship with the level of inventory holding. Also, the ranking process of variables indicate that capital expenditure and firm size posit the most importance and cash from operations and firm growth have the least importance in explaining the level of firms’ inventory holding.