Semiparametric estimation of firstprice auctions with risk-averse bidders

This paper proposes a semiparametric estimation procedure of the first-price auc-tion model with risk averse bidders within the independent private value paradigm. We show that the model is nonidentified in general from observed bids. We then exploit heterogeneity across auctioned objects to establish semiparametric identi-fication under a conditional quantile restriction and parameterization of the bid-ders ’ von Neuman Morgenstern utility function. Next we propose a semiparametric method for estimating the corresponding auction model. This method involves sev-eral steps and allows to recover the parameters of the utility function as well as the bidders ’ private values and their density. We show that our semiparametric estimator of the utility function parameters converges at the optimal rate, which is slower than the parametric one. An illustration of the method on U.S. Fores