Structural changes, common stochastic trends and unit roots in panel data

In this paper we propose a new test statistic that considers multiple structural breaks to analyse the non-stationarity of a panel data set. The methodology is based on the common factor analysis in an attempt to allow for some sort of dependence across the individuals. Thus allowing for multiple structural breaks in the â€Panel Analysis of Non-stationarity in Idiosyncratic and Common components†(PANIC) methodology increases the degree of heterogeneity when assessing the stochastic properties of the panel data set