A multi-agent simulation of the pharmaceutical supply chain

The pharmaceutical supply chain is composed of multiple firms interacting to produce and distribute drugs in an uncertain environment. In this work, we develop and validate a multi-agent simulation of the supply chains associated with the pharmaceutical industry. We demonstrate that the operating norms of a particular industry can be accurately represented to create an industry-specific model capable of tracing its evolution. Our model is initialized using 1982 financial data with 30 manufacturers, 60 suppliers, and 60 distributors. Three types of drugs, blockbusters, medium and small, with a 12-year lognormal product life cycle are released by manufacturers. Each quarter the distributors bid for future market share of the released products, and the suppliers bid for acceptable margins. Mergers and acquisitions, based on assets and expected profitability, are allowed at each level. One thousand replications, each lasting the equivalent of 39 years, are used to validate the model.