Impact of Variety and Distribution System Characteristics on Inventory Levels at U.S. Retailers

Over the past six decades, numerous analytical models have been developed to determine optimal inventory levels. These models predict that inventories carried by a retailer should be a function of the product variety carried by the retailer, distribution system characteristics, economies of scale, etc. A few recent empirical studies have explored the impact of some of these factors on aggregate inventories at U.S. retailers. Building on these works, this study empirically explores the role of key factors such as product variety, number of stores, and number of warehouses in explaining inventory levels at U.S. retailers using data obtained from both primary and secondary sources. We find that variety as measured by the number of stock-keeping units carried and number of stores is associated with higher inventories, whereas scale economies are associated with lower inventories. We do not find the number of warehouses to be significant in explaining inventory levels. Increased demand fluctuations are also associated with higher inventories, although the effects are less robust. The significant variables together with retailer segment identifiers explain a substantial fraction of the variance in inventory levels and can be potentially useful to managers in benchmarking their inventory levels.