This study aims to develop a two-stage stochastic framework that guides governments and firms on where and how much semiconductor production capacity to add, and how to procure chips when large-scale supply disruptions are possible.
The authors model countries as players in a two-stage stochastic logistics network game: in Stage 1, they choose capacity investments under budget limits; and in Stage 2, once a disruption unfolds, they optimize procurement under costs, export quotas and shortage penalties. The problem is formulated as a stochastic generalized Nash equilibrium and numerical examples are solved with a variational inequality projection algorithm.
Investing in domestic production and neutral manufacturing hubs can cut expected shortage penalties significantly. Limited export windows ease system-wide shortages. Technological capability, rather than budget size, is the primary driver of where new capacity locates. When shortage penalties are high, domestic production expansion comes first; after that, access to neutral hubs deliver the best penalty relief.
The model operates as a “what-if” stress-test, enabling policymakers, fab operators and downstream manufacturers to compare investment strategies and export control rules before committing capital or negotiating trade terms.
To the best of the authors’ knowledge, this is the first study to integrate stochastic logistics network design with game-theoretic competition in the semiconductor sector. This work provides a tractable equilibrium algorithm and actionable insights for building resilient semiconductor supply chain networks.
