[Lecture Overview]

   Many semiparametric models in economics are increasingly more flexible and yet are more difficult to ensure point-identification of structural parameters of interest. In this talk, Professor Chen will describe some published and unpublished papers on penalized sieve likelihood ratio (LR) inference on nonstandard semiparametric likelihood models, and penalized sieve optimally weighted Quasi LR inference on nonstandard semiparametric conditional moment restriction models.  Professor Chen will first review some published papers on sieve (Q)LR inferences for possibly irregularly identified semiparametric models, and then discuss some working papers on sieve (Q)LR inference for partially identified semiparametric likelihood models or conditional moment restriction models. Empirical applications of models with latent heterogeneity, shape invariant semi/nonparametric quantile instrumental variables Engel curve regression, nonparametric endogenous demand in industry organization, semi/nonparametric asset pricing models are presented.

 

Finished