Econometrics of Asymmetric Information Models with Applications to Nonlinear Pricing, Insurance and Compensation

自九十年代以来，拍卖的实证分析一直是一个活跃的研究领域，ppi在以前的NSF资助的支持下做出了贡献。拍卖是一种机制，在这种机制中，买家出于个人意愿为一件物品付费。卖家的问题是设计拍卖规则，确保获得最大回报。拍卖是不对称信息模型。pi建议对类似建模的新数据进行分析。正如他们之前的工作一样，他们从一个经济模型开始，深入研究它与数据之间的映射关系，并从经验过程、核估计和特征函数中开发定量方法。首先，研究了对大买家的价格折扣（非线性定价）。pi收集黄页广告和移动电话服务的新数据。对于后者，他们考虑一家公司提出两种产品，即语音和短信。他们还通过推导可观察对象的模型限制来解决测试逆向选择的问题。其次，研究项目分析了保险公司在风险和风险态度上的异质性。此外，pi考虑有限数量的合同。两者都会导致均衡的汇集，这代表了新的挑战。他们评估偏离合同最优和由此造成的保险公司损失。政策问题用汽车保险的新数据来解决。第三，本项目研究高管薪酬，并开发了包含道德风险和(i)公司之间吸引管理人员的竞争以及（ii）风险态度的异质性管理人员的模型；利用标准普尔执行公司的数据。本项目在解决基本问题的同时，进一步推进了非对称信息模型的结构分析。收集了关于黄页广告、移动电话服务和汽车保险的若干新数据集，并将提供给研究界。pi开发的方法将被来自产业组织的实证研究人员使用，也将被来自其他领域的研究人员使用，如卫生、劳工和企业融资。该项目的成果还将有助于公共机构更好地了解监管政策对保险业的影响。

Since the nineties, the empirical analysis of auctions has been an active research area to which the PIs contributed with the support of previous NSF grants. Auctions are mechanisms in which buyers compete given their private willingness to pay for an object. The seller's problem is to design auction rules that secure the maximum return. Auctions are asymmetric information models. The PIs propose to analyze new data modeled similarly. As in their previous work, they begin with an economic model, study thoroughly the mapping between it and the data as well as develop quantitative methods drawing from empirical processes, kernel estimation and characteristic functions.First, price discounts (nonlinear pricing) to big buyers are studied. The PIs collect new data on yellow page advertising and cellular phone services. For the latter, they consider a firm proposing two products, namely, voice and SMS. They also address the problem of testing adverse selection by deriving the model restrictions on observables. Second, the research project analyzes insurance in which insurers are heterogeneous in both risk and in attitudes toward risk. Moreover, the PIs consider a finite number of contracts. Both lead to pooling at equilibrium, which represents new challenges. They assess deviations from contract optimality and resulting losses for the insurer. Policy questions are addressed with new data on automobile insurance. Third, this project studies executive compensation and develops models incorporating moral hazard and (i) competition among firms to attract managers and (ii) heterogeneous managers in risk attitudes; utilizing data from S&P ExecuComp.This project further advances the structural analysis of asymmetric information models while addressing fundamental questions. Several new data sets on yellow page advertising, mobile phone service and automobile insurance are collected and will be available to the research community. The methods developed by the PIs will be used by empirical researchers from industrial organization but also from other fields such as health, labor and corporate finance. The results of this project will also assist public institutions in better understanding the impact of regulatory policies on the insurance industry.