Collaborative Research: New Applications of Dynamic Games with Asymmetric Information: Mergers, Capacity Discipline and Price Rigidity

这项研究调查了企业之间的信息不对称如何影响价格和消费者福利，以及如何改变反垄断政策来考虑这些影响。目前，几乎所有的市场分析--无论是量化市场力量还是进行反事实分析--都假设了完整的信息，在这些信息中，所有公司都准确地知道竞争对手的需求、边际成本和目标。然而，如果公司更多地了解自己的头寸，那么他们可能会试图利用他们的定价或产能选择来影响竞争对手的信念，以影响这些竞争对手未来可能的行为。一些简单的例子表明，即使是少量的不对称信息也可能严重扭曲价格，并可能解释为什么合并后价格大幅上涨；为什么即使菜单成本应该很低，价格也显得僵化；以及为什么航空等一些行业似乎经历了运力纪律时期。本研究将在理论、实证和方法论上做出贡献。研究人员首先刻画了信息不对称的多期寡头定价、投标和容量选择博弈中的非合谋马尔可夫完美贝叶斯均衡。该项目进一步开发了解决这种平衡的计算方法，这通常需要求解非常大量的微分方程式。相关代码将提供给其他研究人员。调查人员还使用消费者包装商品的扫描仪数据和航空市场运力投资的数据来测试和校准这些模型。此外，该项目还分析了围绕实际合并的价格变化和公司行为，以了解不对称信息是否能够解释为什么结果与反垄断机构在合并获得批准时的预期有很大不同。最后，这项研究培训研究生，使他们能够在自己的研究中使用这些和其他技术。

This research investigates how asymmetric information between firms can affect prices and consumer welfare, and how antitrust policies might be changed to account for these effects. Currently almost all analyses of markets -- whether to quantify market power or perform counterfactuals -- assume complete information, in which all firms know the demand, the marginal costs and the objectives of their rivals exactly. However, if firms know more about their own positions, then they might try to use their pricing or capacity choices to influence what their rivals believe in order to affect how those rivals might behave in the future. Some simple examples suggest that even small amounts of asymmetric information can distort prices significantly, and potentially explain why prices rise significantly after mergers; why prices appear rigid even when menu costs should be small; and, why some industries, such as airlines, appear to go through periods of capacity discipline. This research will make theoretical, empirical and methodological contributions. The investigators first characterize non-collusive Markov Perfect Bayesian equilibria in multi-period oligopoly pricing, bidding and capacity choice games with asymmetric information. The project further develops computational methods to solve such equilibria, which often require solving very large numbers of differential equations. The associated code will be made available to other researchers. The investigators also test and calibrate the models using scanner data on consumer packaged goods and data on capacity investments in airline markets. In addition, the project analyzes changes in prices and firm behavior around actual mergers to understand whether asymmetric information can explain why outcomes were significantly different from what the antitrust authorities expected when the mergers were approved. Lastly, this research trains graduate students so that they can use these and other techniques in their own research.