Identification and estimation by penalization in nonparametric instrumental regression

The nonparametric estimation of a regression function x from conditional moment restrictions involving instrumental variables is considered. The rate of convergence of penalized estimators is studied in the case where x is not identified from the conditional moment restriction. We also study the gain of modifying the penalty in the estimation, considering for instance a Sobolev-type of penalty. We analyze the effect of this modification on the rate of convergence of the estimator and on the identification of the regression function x.