Rethinking release rules for international grain reserves under production risk

This study develops a global stochastic computable general equilibrium (CGE) model to evaluate trigger mechanisms and coverage rates for an international joint grain reserve among China, Japan, and South Korea. Framing the reserve as a cooperative risk-sharing institution with public-good characteristics, it links release-rule design to incentive compatibility and institutional sustainability. Unlike prior studies emphasizing reserve size and contribution rules, we examine how release triggers interact with probabilistic shock distributions. We compare VaR-based triggers at 70–95% confidence levels with uniform thresholds of 2.5–20% under 500 correlated productivity-shock paths. Results show that VaR-based triggers generate consistent welfare gains, risk reduction, and complete Pareto improvements. Uniform triggers produce more asymmetric outcomes; at the 10% threshold, Japan and South Korea benefit while China incurs welfare losses. For VaR designs, the optimal coverage rate is approximately 75%, since increasing coverage to 100% brings only marginal additional gains. Excessively aggressive trigger–coverage combinations can weaken performance by accelerating reserve depletion under extreme shocks. Overall, an 80% VaR trigger with 75–100% coverage provides the strongest stabilization effect and more balanced welfare distribution, supporting voluntary participation and long-term cooperation