Comparing sourcing strategies in two-echelon supply chains

In this paper we investigate four sourcing models with respect to both cooperative or noncooperative planning strategies, and sole or dual sourcing. A two-stage supply chain is considered, which includes a single buyer and either a single or dual vendor(s). At the buyer the product is consumed at a constant rate, and an (r, Q) policy is used to generate replenishment orders. The vendors have a finite production rate and ship to the buyer in equal-sized batches. The delivery lead time is stochastic. Shortages are allowed at the buyer’s side and are fully backordered. The cost function comprises five elements: inventory holding costs at the buyer and the vendors, backorder costs and ordering costs at the buyer, and setup costs at the vendors. The objective is to minimize the total system costs incurred by the buyer and the vendors. As there is no dominating combined strategy, a major finding is that determining the best strategy requires a detailed analysis. Furthermore, when total system costs are taken into account, dual sourcing does not appear as beneficial as sometimes claimed in the literature on lot splitting.