Consolidation effects: assessing the impact of tail dependence on inventory pooling using copulas

This paper assesses the impact of tail dependence on the consolidation effect. Based on different R packages, it was not only possible to use copulas for modelling the dependence structure between decentralised demands, but also to randomly generate and optimise different scenarios for various product characteristics, thus finding the best pooling policy: inventory centralisation, regular transshipment, or independent systems. Results suggest a different pattern for the decision-making rationale involving correlation and inventory holding costs. While independent systems and regular transshipments seem to compensate for strong upper tail dependence structures, centralisation seems to compensate for strong lower tail ones.