Panel index var models: specification, estimation, testing and leading indicators

Abstract

This paper integrates panel VARs and the index models into a unique framework where cross unit interdependencies and time variations in the coefficients are allowed for. The setup used is Bayesian and MCMC methods are used to estimate the posterior distribution of the features of interest and to verify hypothesis concerning the model specification. The approach reduces substantially the dimensionality of the problem, can be used to construct multiunit forecasts, leading indicators and to conduct policy analysis in a multiunit setups. The methodology is employed to construct leading indicators for inflation and GDP growth in the Euro area.