Distributionally Robust Optimal Allocation with Costly Verification
研究委托人分配单一物品给多个代理人时，在无货币转移且验证成本高昂的情况下，如何设计分布鲁棒最优机制，适用于风险投资选择初创公司等场景。
Allocation with Costly Verification under Ambiguity In “Distributionally Robust Optimal Allocation with Costly Verification,” Bayrak, Koçyiğit, Kuhn, and Pınar examine the mechanism design problem faced by a principal allocating a single good to one of several agents with costly verification and without monetary transfers, adopting a distributionally robust optimization perspective. Each agent desires the good and generates value for the principal, who does not initially know these values but can verify them at a cost. This problem arises in various scenarios, such as a venture capitalist selecting a start-up for funding or a procurement manager choosing a supplier. Many such allocation problems occur infrequently, leaving the principal with limited knowledge about the distribution of the agents’ values. The authors model these allocation problems as distributionally robust mechanism design problems, explicitly addressing distributional ambiguity. They introduce the first robustly (and Pareto-robustly) optimal mechanisms in the literature. These mechanisms are not only robust but also simple and interpretable.