What’s the penalty for using the Square Root Law of inventory centralisation?

Abstract

Purpose The square root law (SRL) is a popular model for assessing inventory levels when changing the number of warehouses. Previous empirical research, however, has shown that mostly its underlying assumptions do not hold in practice. This sparks the question how inaccurate the SRL’s results are. The paper aims to discuss this issue. Design/methodology/approach In 26 company cases of reducing the number of warehouses, the estimation error of the SRL is analysed irrespective of its underlying assumptions. Findings The analysis reveals an average estimation error for total inventory of 27.85 per cent (median=27.84 per cent), but a high variability across the cases. The SRL seems to mostly overestimate inventory savings from centralisation and inventory increases from decentralisation. Managers should only use the SRL if inventory depends on the number of warehouses in their situation, i.e. if they use the economic order or production quantity policy and safety factor approach. Suggestions for coping with the SRL’s estimation error are given. Research limitations/implications This paper is based on the 26 cases that could be found in a thorough literature review in the ten most widely spoken languages and that contained or allowed to deduce the necessary information. In order to enable wider generalisations, this sample could be extended. Originality/value Most past research has been more theoretical in nature. This research is the first to investigate the SRL’s estimation error using a variety of company cases and how to cope with this error.