Hello everyone,

I am trying to estimate a system GMM with Windmeijer correction. For doing that I have a panel with 66 countries running from 1989-2012 but will data averaged in five year periods such as 1989-1993 makes observation, 1994-1998 the second observation and so on. In the end there is a total of 5 observations per country. I am looking at the link between the stock market, banks and economic growth. I have tried the following command but i am not entirely sure if it is the correct command for the system GMM which includes Windmeijer correction:

xtabond2 log_rgdpc l.log_rgdpc log_prcreditB log_Tratio log_govsize log_trade log_school log_infl period*, gmm(l.log_rgdpc, collapse sp) gmm( L.( log_prcreditB log_Tratio log_govsize log_trade log_school log_infl ), collapse) iv(period*) twostep robust small

Where: log_rgdpc - log of real gdp per capita /GDP
l.log_rgdpc - lag of real gdp per capita/GDP to capture the covergence effect
log_prcreditB - log of bank credit to private sector /GDP
log_Tratio - log turnover ratio/GDP
log_govsize, log_trade, log_school and log_infl are control variables
period - the data are averaged over 5 year periods

Can you please inform me if this is the correct command or which one should be used instead?

Thank you !