Exploring the interaction of inventory policies across the supply chain: An agent-based approach

The Bullwhip Effect, which refers to the increasing variability of orders traveling upstream the supply chain, has shown to be a severe problem for many industries. The inventory policy of the various nodes is an important contributory factor to this phenomenon, and hence it significantly impacts on their financial performance. This fact has led to a large amount of research on replenishment and forecasting methods aimed at exploring their suitability depending on a range of environmental factors, e.g. the demand pattern and the lead time. This research work approaches this issue by seeing the whole picture of the supply chain. We study the interaction between four widely used inventory models in five different contexts depending on the customer demand variability and the safety stock. We show that the concurrence of distinct inventory models in the supply chain, which is a common situation in practice, may alleviate the generation of inefficiencies derived from the Bullwhip Effect. In this sense, we demonstrate that the performance of each policy depends not only upon the external environment but also upon the position within the system and upon the decisions of the other nodes. The experiments have been carried out via an agent-based system whose agents simulate the behavior of the different supply chain actors. This technique proves to offer a powerful and risk-free approach for business exploration and transformation.