In the era of globalization, companies commonly confront supply disruptions triggered by policy shocks, such as Sino-US trade tensions, trade disputes between Japan and South Korea, and the uncertainty surrounding Brexit. To address this issue, we study the manufacturer’s optimal decisions in various disruption risk levels under two scenarios: non-agent and agent sourcing. We use a model of sourcing in a supply chain with three levels: tier 0 manufacturer, tier 1 suppliers and disruption-prone tier 2 suppliers. The manufacturer may influence the sourcing decisions of tier 1 suppliers via contract parameters. We first study each structure of non-agent sourcing in isolation, showing that the manufacturer will make different decisions for various unit revenue and disruption risk levels and the manufacturer has different psychological bottom lines regarding tier 2 supplier’s prices for different decisions, where psychological bottom lines are related to disruption risks. Based on non-agent sourcing, we also study each structure of agent sourcing in isolation, and show that the manufacturer will make different decisions for various unit revenue, tier 2 supplier’s prices and disruption risk levels. Furthermore, both tier 1 suppliers and tier 0 manufacturer have different psychological bottom lines regarding agent sourcing price for different decisions, where different psychological bottom lines are all related to disruption risks, which causes that agent sourcing optimal decision will come into effect only when the agent sourcing price expectations are consistent. Finally, we conduct a comparative analysis of total supply chain profits between non-agent and agent sourcing under the manufacturer's optimal decisions and show that total supply chain profit will reach the maximum when adopting agent sourcing decisions because all disruption risks are avoided.