An Improved GMM Estimation of Panel VAR Models with Applications to Granger Causality Test and Impulse Response Analysis

In this paper, we propose improved IV/GMM estimators for panel vector autoregressive models by extending Hayakawa (2009) where an alternative form of instruments is suggested. We show that the proposed IV estimator has the same asymptotic distribution as the bias corrected maximum likelihood estimator by Hahn and Kuersteiner (2002). Since the proposed estimator is simply change the form of instruments, it is very easy to implement. As applications of the proposed estimators, we consider a panel Granger causality test and panel impulse response analysis where the asymptotic distribution of generalized impulse response functions of Pesaran and Shin (1998) is newly derived. Monte Carlo simulation results show that the proposed estimator outperforms the conventional IV/GMM estimators using instruments in levels in many cases. ∗Part of this paper was written while the author is visiting the Faculty of Economics, University of Cambridge as a JSPS Postdoctoral Fellow for Research Abroad. Financial support from the JSPS Fellowship and KAKENHI(22730178) is also gratefully acknowledged. All remaining errors are my own.