Title: Bounds testing for panel unit roots using common correlated effects estimators
Authors: Josep Lluis Carrion-i-Silvestre - Universitat de Barcelona (Spain) [presenting]
Anindya Banerjee - University of Birmingham (United Kingdom)
Abstract: A panel data unit root statistic is proposed which has cross-section dependence driven by unobserved common factors that are approximated by means of the common correlated effects estimation method. The null hypothesis of panel data unit root focuses on the idiosyncratic component, although the statistical inference is conducted using a bounds testing strategy. Proceeding in this way the analysis takes into account that the cross-section dependence might be driven by $I(1)$ non-stationary common factors, $I(0)$ common factors, or a mixture of both $I(1)$ and $I(0)$ common factors and, therefore, extending some existing proposals in the literature.