Nonparametric Identification and Estimation of a Common Value Auction Model

Structural econometric studies on auctions have mainly focused on the independentprivate value paradigm. In this paper, we are interested in the “opposite” caseknown as the pure common value model. More precisely, we restrict our attentionto a common value model defined by two functions : the density of the true valueof the auctioned good and a unique function that appears in the definition of theconditional densities of the signals. We establish that this common value model isnonparametrically identified without any further restrictions. We then propose aone-step nonparametric estimation method and prove the uniform consistency of ourestimators. We apply our method on simulated data and show that the technique wepropose is adequate to recover the distribution functions of interest.