Submitted to Management Science manuscript MS-xxxx-xxxx.xx Disruption Risk and Optimal Sourcing in Multi-tier Supply Networks

We study sourcing decisions in a supply chain with three levels: Tier 2 suppliers that sell to Tier 1 suppliers, Tier 1 suppliers that sell to a manufacturer, and a manufacturer that sells finished goods to the market. The suppliers in Tier 2 are prone to independent disruption risk due to, e.g., natural disasters such as earthquakes or floods. The manufacturer may not directly dictate which Tier 2 suppliers are used, but may influence the sourcing decisions of Tier 1 suppliers via the contract parameters it offers to Tier 1 suppliers. We show that the manufacturer’s optimal sourcing strategy depends critically on the degree of overlap in the upper levels of the supply chain: generally speaking, as overlap in Tier 2 increases (i.e., as Tier 1 suppliers have more Tier 2 suppliers in common), the manufacturer should rely less on direct mitigation (procuring excess inventory and multi-sourcing in Tier 1) and rely more on indirect mitigation (increasing the price offered to Tier 1 suppliers to induce them to procure excess inventory and multi-source). We demonstrate that, predictably, a supply chain with no overlap (Tier 1 suppliers share no Tier 2 suppliers) yields the greatest profit for the manufacturer; surprisingly, however, a supply chain in which Tier 1 suppliers share their most reliable (but still prone to disruption) Tier 2 suppliers performs equally well in many cases, implying that it is most important for a manufacturer to ensure its Tier 1 suppliers are diversified amongst their riskiest Tier 2 suppliers. Lastly, we consider the manufacturer’s sourcing problem if it does not know the structure o