Dear Statalisters,
I study the long run relationship between real exchange rates and underlying fundamentals based on panel data (N=17; T= 36). To this end, I am trying to estimate a cointegration relationship with unobserved common factors using GMM, as described by TimothyNeal (2015) in this article. I rely on the xtabond2 command from ssc in Stata 13 SE. Here is one of my syntaxes:
Please, can anyone help me to know if I am right by first-differencing my dependent variable and if I can expect to obtain stationary residuals - I(0) - with zero mean for every country after a system-GMM like this. To the best of my knowledge, residuals normality is not required after a GMM. If, I am right, is it reasonable to standardized my residuals after the above command?
Thanks a lot in advance.
I study the long run relationship between real exchange rates and underlying fundamentals based on panel data (N=17; T= 36). To this end, I am trying to estimate a cointegration relationship with unobserved common factors using GMM, as described by TimothyNeal (2015) in this article. I rely on the xtabond2 command from ssc in Stata 13 SE. Here is one of my syntaxes:
Code:
xtabond2 d.lnreer L1.(lnreer) lntot lnopen lngovexp lnprod nfa_gdp L(1).D.( lnreer ) L(0).D.( lngovexp ) L(0/1).D.( lnopen ) L(0).( l1lnreer_cs lngovexp_cs lnprod_cs ), gmm(lnreer lntot lnopen lngovexp , lag(2 2) eq(diff) orthog ) iv( growth_ocde growth_rgdppc l(1/3).lnprod l2.nfa_gdp , eq(level) mz ) orthog pca h(2)
Thanks a lot in advance.
