Showing 1–1 of 1 results for author: Phillips, R F

Quasi Maximum-Likelihood Estimation of Dynamic Panel Data Models

Authors: Robert F. Phillips

Abstract: This paper establishes the almost sure convergence and asymptotic normality of levels and differenced quasi maximum-likelihood (QML) estimators of dynamic panel data models. The QML estimators are robust with respect to initial conditions, conditional and time-series heteroskedasticity, and misspecification of the log-likelihood. The paper also provides an ECME algorithm for calculating levels QML… ▽ More This paper establishes the almost sure convergence and asymptotic normality of levels and differenced quasi maximum-likelihood (QML) estimators of dynamic panel data models. The QML estimators are robust with respect to initial conditions, conditional and time-series heteroskedasticity, and misspecification of the log-likelihood. The paper also provides an ECME algorithm for calculating levels QML estimates. Finally, it uses Monte Carlo experiments to compare the finite sample performance of levels and differenced QML estimators, the differenced GMM estimator, and the system GMM estimator. In these experiments the QML estimators usually have smaller --- typically substantially smaller --- bias and root mean squared errors than the panel data GMM estimators. △ Less

Submitted 2 February, 2017; originally announced February 2017.