Statistics Seminar（2014-08）

Topic:Inference on Moment Inequalities with Unknown Functions

Speaker:Shengjie Hong, Tsinghua University

Time:Thursday, 3 April, 14:00-15:00

Location:Room 217, Guanghua Building 2

Abstract:In this paper, we consider inference on conditional moment inequalities, in which the unknown parameter may contain infinite-dimensional components. We propose using the sieve minimum of a Kolmogorov-Smirnov type statistic as the test statistic, and derive its asymptotic distribution. A penalized bootstrap procedure, which incorporates a moment selection mechanism, is developed to get critical values. Our methods are robust to partial identification, and allow for the moment functions to be nonsmooth. We extend Hong’s (2012) analysis on conditional moment equalities in the following two directions: First, by allowing for moment inequality constraints, our inference method has a lot more applications; Second, the consistency of our test holds uniformly over a broad family of data generating processes.

Keywords:Conditional moment inequalities, identification set, model specification

test, confidence set, bootstrap.

Jel Classification:C12; C14

About the speaker:Shengjie Hong, Assistant Professor of Economics at the Tsinghua University. He received his Ph.D. in Economics from the University of Wisconsin-Madison in 2012. His main fields of research are econometric theory and applied econometrics. His current work is focused on semiparametric identification and moment inequality models.