Indirect Inference

Abstract

The purpose of this master's thesis is to present the Indirect Inference principal in the scope of the analysis of three basic Indirect Inference Estimators, as they appear in the relevant bibliografiic references, and some of their asymptotic and finite sample properties. The study illustrates the framework with the example of estimation of the MA(1) parameter when AR(1) is considered as the auxiliary model and least squares forms the auxiliary criterion. In this context, the already defined estimators are derived and some other ones directly connected with bias correction procedures as the bootstrap and the median estimation, are presented. Finally, a MATLAB simulation study iscarried out in order to highlight the approach along with the essential toolbox.