On returns and network configuration in supply chain dynamics

This research focuses on how two common modeling assumptions in the Bullwhip Effect (BWE) literature (i.e., assuming the return of the excess of goods and assuming a serial network) may distort the results obtained. We perform a robust design of experiments where the return condition (return vs. no return) and the configuration of the Supply Chain Network (SCN) (serial vs. divergent) are systematically analyzed. We find an important interaction between these assumptions: the impact of returns on the BWE strongly depends on the SCN configuration. This study highlights the importance of accurately modeling SCNs to properly assess SCNs managers.