Semiparametric estimation of signaling games with equilibrium refinement
研究信号博弈的计量建模，通过均衡精炼解决多重均衡问题，实现模型参数识别，并利用筛条件极大似然估计非参数类型分布和支付结构，得到结构参数的一致估计和渐近正态性。
We study an econometric modeling of a signaling game where one informed player may have multiple types. For this game, the problem of multiple equilibria arises and we achieve the uniqueness of equilibrium using an equilibrium refinement, which enables us to identify the model parameters. We then develop an estimation strategy that identifies the payoffs structure and the distribution of types from the observed actions. In this game, the type distribution is nonparametrically specified and we estimate the model using a sieve conditional MLE. We achieve the consistency and the asymptotic normality for the structural parameter estimates.