Hello Professor, Thank you for this very rich video for us first-time researchers. I am currently performing an analysis and need to use the Generalized Moment-Based Regression (GMM) method. I wanted to know how to determine the number of lags to include in the endogenous variables.
There's no formal procedure for determining the number of lags to include as instruments for endogenous variables. Lag 1 is usually the default for instrumenting the endogenous regressors while lag 2 of dependent variable is used to instrument the predetermined regressor (ie lag 1 of dependent variable).
Thanks for the detailed explanations. At what point do we have descriptive statistics and correlation analysis? What comes first? diagnostic tests for normal distribution, unit root, serial correlation, heteroskedasticity, multicollinearity or all these are catered for by the regression estimation? I observe that differencing/lagging caters for unit root, endogeneity is also managed here. When is the co-integration test done? What if n if less than t? what estimation techniques should be used? How do we establish whether data requires internal instrumental variables?
Very helpful and informative lecture. Thank You so much, sir. Please can You tell us how to find the Wald test statistics and Hansen test for instruments?
Sir i have run three models. In that for two models AR(1) and AR(2) p values are shown as NA. And in the third model pvalue of AR(1) is 0.963. please explain sir most of the AR p values are shown as NA. What does it mean. Does it mean it's a bad model
Please what is the solution to this? Kindly give a video or explanation to demonstrate this issue I had while running the gmm what does this mean kindly explain 'number of instruments greater than number of observations'
how did you work to graduate a model panal generalized method , i did not show me this model just model generalized method and in this model there is on test bond seriel and there a problem with an application @DYN ??
Thanks for your helpful video tutorial. I want to ask you if we should test the unit root in the Panel Generalized Method of Moment (GMM)? If our variabes are of different order of integration (some are stationary at level and some are stationary after first difference), can we use the Panel Generalized Method of Moment (GMM)?
You should first arrange your data in Excel and then copy to the Eviews to set up your workfile or you can set up a workfile in Eviews and then input your data manually. I will make a separate video for that.
Hello Pr. How can we interprete J-statistic in output of panel System GMM ? How can i obtain sargan-Hansen test with eviews? How can we interprete the output of system GMM model? When you have two equations. I work with eviews 10. Thanks !
Hi. thanks for the valuable information, I tried on my model but keeps rejecting it and giving msg that: Number of instruments greater than number of observations. what do in this case? thanks again
Revisit the instrument specification and follow my guidelines in the video. If the problem persists, re-estimate with fewer instruments. Also, be mindful of the size of N and the number of k as you specify the instruments.
Sir , The autocorrelation at AR(1) should be significant whereas AR(2) should be insignificant (Beck et al., 2013 and Roodman, 2009). May i please know what must be the pvalues for AR(1) and AR(2)
The pvalues for AR(1) should be less than 0.05 (significant) while that of AR(2) should be greater than 0.05 (insignificant) for that condition to be met.
Thank you for the video. How to frame an equation for system GMM. In the video you have copy pasted but I want to learn to frame an equation/estimations for system GMM. Can you help me? Or can you upload a video for the same.
Yes, the data is time series. You can arrange in excel and copy to Eviews or set the data up from Eviews. You need to know how to set up the cross-section ID and date ID. It's very simple.
@@ChekwubeMadichie thanks..actually i am totally new in gmm method. And my data is time series..so i face some problems. I wish a vedio based on time series data not panel data in eviews..
I have other videos on time series data analyses. Except you're talking about country-specific time series GMM, there are videos on time series on my channel. I will make a video on country-specific time series GMM soon.
No, they're not the same. However, if the instrument condition is met, the usual system gmm can be estimated with the forward orthogonal deviation transformation.
@@moyosoreadedapo-aisida9802 thanks for your reply. I'm working on a video for the system gmm in Eviews. Keep sharing and recommending my channel to your colleagues. Also turn on the notification bell so that you can get notified when I upload the video.
@@ChekwubeMadichieFirst of all thank you for the videos. You said that "if the instrument condition is met". What is instrument condition? What does it imly? Also, do you have a video about 2 step system gmm ? If you upload it it will be so good for us. Thanks in advance.
Thank you. From zero level I managed to understand. Keep sharing, you are a very good teacher!
Thank you so much Sir for explaining all the information in a detailed manner
Thank you so much for a very informative and helpful lecture.
Hello thanks for sharing the valuable information.
Welcome
Hello Professor, Thank you for this very rich video for us first-time researchers. I am currently performing an analysis and need to use the Generalized Moment-Based Regression (GMM) method. I wanted to know how to determine the number of lags to include in the endogenous variables.
There's no formal procedure for determining the number of lags to include as instruments for endogenous variables. Lag 1 is usually the default for instrumenting the endogenous regressors while lag 2 of dependent variable is used to instrument the predetermined regressor (ie lag 1 of dependent variable).
Thanks for the detailed explanations. At what point do we have descriptive statistics and correlation analysis? What comes first? diagnostic tests for normal distribution, unit root, serial correlation, heteroskedasticity, multicollinearity or all these are catered for by the regression estimation? I observe that differencing/lagging caters for unit root, endogeneity is also managed here. When is the co-integration test done? What if n if less than t? what estimation techniques should be used? How do we establish whether data requires internal instrumental variables?
thanks alot for the efforts. is there any video in which you gave applied GMM for mediation analysis. kindly share the link
Not yet
Very helpful and informative lecture. Thank You so much, sir. Please can You tell us how to find the Wald test statistics and Hansen test for instruments?
Nice video 👍
Sir i have run three models. In that for two models AR(1) and AR(2) p values are shown as NA. And in the third model pvalue of AR(1) is 0.963. please explain sir most of the AR p values are shown as NA. What does it mean. Does it mean it's a bad model
Please what is the solution to this? Kindly give a video or explanation to demonstrate this issue I had while running the gmm what does this mean kindly explain
'number of instruments greater than number of observations'
how did you work to graduate a model panal generalized method , i did not show me this model just model generalized method and in this model there is on test bond seriel and there a problem with an application @DYN ??
Thanks for your helpful video tutorial. I want to ask you if we should test the unit root in the Panel Generalized Method of Moment (GMM)? If our variabes are of different order of integration (some are stationary at level and some are stationary after first difference), can we use the Panel Generalized Method of Moment (GMM)?
dear Prof., how to solve the problem "number of instruments greater than number of observations" appeared when running GMM panel data, pls.
Thanks to y'all for watching. Kindly support my channel by sharing this video to your friends and colleagues. God bless you.
Hello thanks for your video. I dont know how to input data excel to eviews on this process. Is it the same as entering data in panel data?
You should first arrange your data in Excel and then copy to the Eviews to set up your workfile or you can set up a workfile in Eviews and then input your data manually. I will make a separate video for that.
Dear Riza Laela Tsaro, you can now watch how to input time-series and panel data into the Eviews workfile @ th-cam.com/video/4lqAFyQhfTU/w-d-xo.html
Why there is no Intecept in the regression? please help
A constant is specified but it doesn't come visible in the results.
Hi, do we have any GMM version or method to use for the situation which T is bigger than M ?
Hello Pr.
How can we interprete J-statistic in output of panel System GMM ?
How can i obtain sargan-Hansen test with eviews?
How can we interprete the output of system GMM model? When you have two equations.
I work with eviews 10.
Thanks !
Hi. thanks for the valuable information, I tried on my model but keeps rejecting it and giving msg that: Number of instruments greater than number of observations. what do in this case? thanks again
Revisit the instrument specification and follow my guidelines in the video. If the problem persists, re-estimate with fewer instruments. Also, be mindful of the size of N and the number of k as you specify the instruments.
@@ChekwubeMadichie it's not working Teacher! What should i do please? Thanks for your answer. I'm PhD student.
Sir , The autocorrelation at AR(1) should be significant whereas AR(2) should be insignificant (Beck et al., 2013 and Roodman, 2009). May i please know what must be the pvalues for AR(1) and AR(2)
The pvalues for AR(1) should be less than 0.05 (significant) while that of AR(2) should be greater than 0.05 (insignificant) for that condition to be met.
@@ChekwubeMadichie thank you so much for your reply Sir. Means a lot Sir. Thank you so much
Hello and thank you, should I apply diagnostic tests on panel data set before starting GMM? if yes which tests I should apply?
You should start with endogeneity test and sargan test of overidentifying restrictions.
@@ChekwubeMadichie do you have a videos about how to perform them?
@@afnanghanayem7016 not yet. I will work on that.
@@ChekwubeMadichie Where is the option of computing sargan test in eviews.. I am unable to find so in the current version. will request you to guide
Thank you for the video. How to frame an equation for system GMM. In the video you have copy pasted but I want to learn to frame an equation/estimations for system GMM. Can you help me? Or can you upload a video for the same.
Please become a member of my community to get help and support for your research
Can the time series data be used in gmm in eviews? If possibe how the data be arranged for gmm? Please help me
Yes, the data is time series. You can arrange in excel and copy to Eviews or set the data up from Eviews. You need to know how to set up the cross-section ID and date ID. It's very simple.
@@ChekwubeMadichie thanks..actually i am totally new in gmm method. And my data is time series..so i face some problems. I wish a vedio based on time series data not panel data in eviews..
I have other videos on time series data analyses. Except you're talking about country-specific time series GMM, there are videos on time series on my channel. I will make a video on country-specific time series GMM soon.
@@ChekwubeMadichie thanks
Is system GMM the same as orthogonal deviation?
No, they're not the same. However, if the instrument condition is met, the usual system gmm can be estimated with the forward orthogonal deviation transformation.
@@ChekwubeMadichie Alright thanks, would really appreciate a video on system gmm with Eviews.
@@moyosoreadedapo-aisida9802 thanks for your reply. I'm working on a video for the system gmm in Eviews. Keep sharing and recommending my channel to your colleagues. Also turn on the notification bell so that you can get notified when I upload the video.
@@ChekwubeMadichieFirst of all thank you for the videos. You said that "if the instrument condition is met". What is instrument condition? What does it imly? Also, do you have a video about 2 step system gmm ? If you upload it it will be so good for us. Thanks in advance.
The instruments must be strictly exogenous and must have larger impact on the regressors. I will make a video on two-step.
how to interpret the probability of J stat ?
I don't have model 😞 gmm/dpd I have gmm in my EViews
What version of Eviews do you use?
you need to create a panel data workfile on Eviews to unlock it, watch the link in the description is this video
It is Very much informative. But How can I get the Z-statistic value of variable in GMM Model @vidata solutions