Abstract (eng)
Suppliers as a central factor of successful supply chain management face the risk of several disruptions, e.g. natural disasters, bankruptcy, labor strikes, machine breakdowns, and business failures. Buyers that rely on a single supplier, therefore, expose themselves to great risk of disruptions. Dual sourcing, i.e. sourcing from two suppliers, is a potential countermeasure.
This thesis studies dual sourcing models in the presence of supply disruption risks. Thereby, the optimal sourcing and allocation policy, i.e. how much to order from which supplier, depends on various factors, such as diversity in reliabilities, prices, and geographical location of the suppliers. To examine the value of the buyer’s optimal dual sourcing policy it is compared to single sourcing policies and simple (heuristic) dual sourcing policies.
We study different model variants that include supply disruptions from the perspective of the buying firm. First, we analyze the optimal inventory and allocation policy under the risk of supply disruptions, stochastic lead times, and stochastic demand. Second, we study the optimal allocation policy under supply disruptions and learning, i.e. suppliers reduce production cost over time through learning. By considering supply disruption risk and learning, we analyze the trade-off between risk reduction via dual sourcing and learning benefits on supplier sourcing costs induced by long-term relationships with a single supplier. Finally, the sensitivity of this model to unknown supplier reliabilities, a stochastic demand, and risk averse decision makers is investigated. Our results illustrate the advantage of a flexible dual sourcing strategy and provide important insights into effective supply disruption risk management.