Dear Fellow members of the Forum, I would like first to express how privileged I am for being a member of this astonishing group and I would like to extend my sincere greetings to all of you. I have a problem and I will discuss it briefly hoping to get some answers.
Description of Data: Unbalanced dynamic panel data on Syrian Banking sector which runs from the year 2005 till 2016, the dependent variable is the profitability of a bank i in year t, the independent variables are grouped into three sets of variables which are bank specific (9 variables), industry specific (5 variables) and Macro specific (3 variables).
Objectives: This dynamic panel data in which two types of GMM are utilized (Difference GMM and System GMM) aims to figure out the determinants of the Syrian banks profitability, and how the behavior of these determinants has changed after the Syrian war which erupted in 2011.
Problem: I need to test for Non-stationarity in the dependent and independent variables series within the context of structural breaks since we know that a war has out-broken around 2011 and 2012, some have suggested the use of Philips-Perron test for the presence of unit roots in the context of structural breaks, and to test for the structural breaks, STATA 14.0 has two commands, namely "estat sbknown" and "estat sbsingle" but the problem is that I cannot run the commands "estat sbknown" or "estat sbsingle" in STATA 14.0 since both of these commands were developed for time-series not for Dynamic panel data, so my question are:
Description of Data: Unbalanced dynamic panel data on Syrian Banking sector which runs from the year 2005 till 2016, the dependent variable is the profitability of a bank i in year t, the independent variables are grouped into three sets of variables which are bank specific (9 variables), industry specific (5 variables) and Macro specific (3 variables).
Objectives: This dynamic panel data in which two types of GMM are utilized (Difference GMM and System GMM) aims to figure out the determinants of the Syrian banks profitability, and how the behavior of these determinants has changed after the Syrian war which erupted in 2011.
Problem: I need to test for Non-stationarity in the dependent and independent variables series within the context of structural breaks since we know that a war has out-broken around 2011 and 2012, some have suggested the use of Philips-Perron test for the presence of unit roots in the context of structural breaks, and to test for the structural breaks, STATA 14.0 has two commands, namely "estat sbknown" and "estat sbsingle" but the problem is that I cannot run the commands "estat sbknown" or "estat sbsingle" in STATA 14.0 since both of these commands were developed for time-series not for Dynamic panel data, so my question are:
- Are there any commands in STATA 14.0 to test for structural breaks in Unbalanced dynamic panel data?
- What unit root tests (commands) do you recommend in my case? Augmented Dicky Fuller (ADF) or Philips-Perron? And why is that?
Comment