Title: Higher order properties of estimators in nonparametric moment conditions models with weakly dependent data
Authors: Francesco Bravo - University of York (United Kingdom) [presenting]
Abstract: Localized versions of Generalized Method of Moments (GMM) and Generalized Empirical Likelihood (GEL) are considered that can be used for estimation in nonparametric moment conditions models with weakly dependent data. The local linear estimator is used to estimate the unknown parameters and shows that the local empirical likelihood estimator has the smallest second order bias within the class of local efficient GMM and GEL estimators. Bias corrected local GMM/GEL estimators are proposed that are second order accurate. The results are illustrated analytically and numerically with two examples: a nonparametric generalized linear model and a smooth coefficient model where some covariates can be correlated with the unobservable errors.