Uncovering the inventory-business cycle nexus

Despite being the smallest component of GDP, inventories represent the second largest source of GDP fluctuations, with a contribution comparable to that of fixed investment. Over the past decades, research in inventory management has proposed competing theories about the primary drivers prompting firms to accumulate stocks, yet consensus remains elusive on the source of inventory cycles. This paper imposes structure on US macroeconomic data and disentangles four shocks related to current and expected demand and supply conditions within a unified framework. We find that sales forecast errors drive the highest share of inventory investment in the short run, giving support to the buffer-stock motive for holding inventories, whereas shocks to expected costs gain more relevance in the long run and generate the missing positive correlation between cost-driven inventory investment and sales that the literature has struggled to find. Shocks to expected demand - which relate to stockout-avoidance reasons for inventory investment - are also very relevant. We find that forward-looking behaviours are those that lead production to be more variable than sales. Finally, our results offer a sensible narrative around the post-pandemic period when inventories drove a very high share of GDP fluctuations.