Hi, Thanks for explaining so well. I have been struggling to find articles that support using lags as instrument variable if you have would you mind sharing. Also how did you determine the number of lags for each instrument?
Stationary series are naturally long run series. By stationary, I mean stationarity at levels I(0). There will be no need to conduct stationarity test when all your series have this attribute. You can go ahead with OLS
Hi, Thanks for explaining so well. I have been struggling to find articles that support using lags as instrument variable if you have would you mind sharing. Also how did you determine the number of lags for each instrument?
Thank you.
MANY THANKS
Sir in 2SLS what must be the value of j statistics and in Least Squares what must be the value of F statistics for model fit?
What do we do when we noticed that variable are stationery but no long run relationship, can we just go ahead to use OLS
Stationary series are naturally long run series. By stationary, I mean stationarity at levels I(0). There will be no need to conduct stationarity test when all your series have this attribute. You can go ahead with OLS
share the link of data
Why tbr is not log
Because tbr is in percentage(already transformed). The essence of logging variables is to transform them for better distribution.