Statistical Methods in Inventory Effect and Analysis

Abstract

Abstract Inventory effect suggests that the correlation between spot and futures prices is positively related to the inventory level. It was proposed to be a possible explanation for backwardation equilibrium in futures markets. A test for the inventory effect requires the specification of the joint distribution between spot and futures prices. This chapter summarizes four approaches currently available in the literature: bivariate normal distributions, bivariate log‐normal distributions, ordered bivariate normal distributions, and ordered bivariate log‐normal distributions. These approaches generate different conclusions, indicating the result is not robust to the distribution assumption. As a consequence, the inventory effect may prevail in some commodity markets but not in others.