题目：Hybrid Generalized Empirical Likelihood Estimators: Instrument Selection with Adaptive Lasso

主讲人：Qingliang Fan, Assistant Professor, Xiamen University

Abstract

In this paper we use adaptive lasso estimator to select between relevant and irrelevant instruments in heteroskedastic and non Gaussian data. To do so limit theory of Zou (2006) is extended from univariate iid case. Then we use the selected instruments in generalized empirical likelihood esti-mators (GEL). In this sense, these are called hybrid GEL. It is also shown in the paper that Lasso estimators are not model selection consistent whereas adaptive lasso can select the correct model in xed number of instruments case. It is also shown that adaptive lasso estimator can achieve near minimax risk bound even in the case of heteroskedastic Gaussian data. This is a new result and extends the standard normal iid data in Zou (2006). In simulations we show that hybrid GEL estimators have very good bias and mean squared error compared with other estimators.