Dear Kalpataru, Thank you. I would like to invite you to join Hossain Academy Facebook at below link to discuss about economics, econometrics and statistical models using EVIEWS, STATA, R, SPSS, Minitab, Microfit, Lingo, and Excel. Thank you, Dr. Sayed Hossain, Adjunct Economic Faculty, Collin College, Texas. facebook.com/groups/hossainacademy/
Dear MsEriss, Thank you. I would like to invite you to join Hossain Academy Facebook at below link to discuss about economics, econometrics and statistical models using EVIEWS, STATA, R, SPSS, Minitab, Microfit, Lingo, and Excel. Thank you, Dr. Sayed Hossain, Adjunct Economic Faculty, Collin College, Texas. facebook.com/groups/hossainacademy/
Dear Anita, Thank you. I would like to invite you to join Hossain Academy Facebook at below link to discuss about economics, econometrics and statistical models using EVIEWS, STATA, R, SPSS, Minitab, Microfit, Lingo, and Excel. Thank you, Dr. Sayed Hossain, Adjunct Economic Faculty, Collin College, Texas. facebook.com/groups/hossainacademy/
Dear Susan, Thank you. I would like to invite you to join Hossain Academy Facebook at below link to discuss about economics, econometrics and statistical models using EVIEWS, STATA, R, SPSS, Minitab, Microfit, Lingo, and Excel. Thank you, Dr. Sayed Hossain, Adjunct Economic Faculty, Collin College, Texas. facebook.com/groups/hossainacademy/
please professor if I use tydl approach, and I use one dummy variable , the estimation of var in this case in eviews, in the exogenous variables I put all the variables with lag, how about dummy variable I use lag for it, I mean dummy (-3) for example or I put just dummy
Hi Mr. Sayed ! How will i know if i will use stationary or nonstationary data in VAR considering that my data are cointegrated? Pls help ASAP thanks ! :)
dear sir i am testing for granger causality between two variables. One of my variable is I(0) and the other is I(2). can i proceed within a VAR framework or do i need to use Toda Yama. Thanks
Hi! Does anyone know the command reference to get the Granger Causality test on a VAR in EViews (instead of going to "View/Lag Structure/Granger Causality" on the VAR ecuation, i'd like to set all the causality tests for different VARs in a program... thank you!
Dear Fernando, I would like to invite you to join Hossain Academy Facebook at below link to discuss about economics, econometrics and statistical models using EVIEWS, STATA, R, SPSS, Minitab, Microfit, Lingo, and Excel. Thank you, Dr. Sayed Hossain, Adjunct Economic Faculty, Collin College, Texas. facebook.com/groups/hossainacademy/
Hi Sir, I'm using EViews version 8, running a unrestricted VAR model. I am checking the impact of oil price shocks on South Africa. I'm using three variables government revenues, money supply and real exchange rate. When I run Granger Causality test, the result shows no causality running from oil price shocks to other variables. However, when I run bi-variate model (in pairs) e.g. Oil price shocks cause government revenues and highly significant. I don't know how to solve this problem!! Which model should I use (multivariate VAR model or univariate model) to explain the real effect.
Thank you so much for your vid. I have a question that I have to check the unit roots test before regressing the VARs or not ? and if YES and then I find that the variable is unit root I need to do first difference before estimating VARs right ?
Hi, If i only have 2 variables which are not cointegrated, do i perform Granger causality in VAR model or Pairwise Granger Causality test? Does VAR model represent short run? How about if 2 variables are cointegrated, can i perform VECM to check long run and short run relationship?
+sin yi Thank you. I would like to invite you to join “Analysis” in Facebook below to discuss about econometrics, statistics and data analysis using EVIEWS, STATA, R, SPSS, Minitab, Microfit, Lingo, Excel etc. Thank you, Sayed from Analysis Group. facebook.com/groups/449520105253087/
Thanks for your usful guidelines! However, I have a question about, what the difference between Granger causality test and Granger causality test in VAR Model. Which one can be used to test the correlation between two variables more clear?
Sayed Hossain hi sir how can we control for variables in causality test if we want to find the causality of independent var. on dependent variable and like to control for the effect of some other variables.
Sayed Hossain Hi Sir. Thanks for the video but I got a question. By applying Granger causality test in VAR Model, you already control the effect of "Money" on GDP when you are testing if "Export" cause "GDP", is that correct?
haifa fahad You have to choose optimum lag number of lags to be used. Hossain Academy has some videos on lag selection. See those and also join Hossain Academy Facebook for discussion below link. facebook.com/groups/hossainacademy/
Thank u for ur videos I have greatly benefited from them. Please, I want to perform diagnostic test (J-B, B-G, ARCH LM, white test and Ramsey test) to check for the robustness of the VAR model I estimate how do I do that? I will appreciate ur view on this.
You just use the t-statistic given in Var and whether or not t s significant compared to T-statistic of adjusted number of observation which in the video is 34
hi, sir, thanks for your video, it is a good guideline. I have a question for my research. after i run unit root test using ADF, my data are integrated at level, I（0），mean i not need differentiate it in 1st order in order make the data stationary , so can i still need run Johansen co integration test ?
dear sayed hossein, thanks so much for your useful guidelines. After watching your demonstration of "Granger causality test in a VAR model", I have a question. Correct me if I am wrong but in this particular model we are assuming that all of my variables are 1) first difference stationary, 2) *not* co-integrated (after Johansen test of co-integration). Therefore, I have to do unit root and co-integration testing before proceeding with "granger causality test in a VAR model". I appreciate your thoughts on this. Thanks.
Granger causality should run using stationary data. Unit root testing will allow you to know the status of data, whether it is stationary or non stationary. If data not stationary, then convert it into stationary, and then run granger causality model.. We run Johansen test to know whether we should run VAR or VECM.
thank you Sayed Hossain for your explanation. If my data (variables) are non stationary at level, then I check if they return stationary at first difference or not. For VAR models, all variables should be at the same order of integration.. Here is my question: Do I run VAR models and co-integration tests with LEVEL data or with first-differenced data?. That is my confusion indeed when you say "convert data into stationary and then run granger causality model"...I really appreciate your thoughts on this.
In other words, let's assume that my variables become stationary at first difference according to unit root testing. Do I have to convert my variables into "first difference stationary" before I run VAR models?
You can not assume it, you have to prove it using unit root testing that data are stationary after first differenced. Then you have to use this first differenced data for granger causality as this data is in stationary state.
dear sir, if i have one of the variables with structural break can i apply co-integrations tests or not, i applied Toda and Yamamoto test and there is one way causality
+Lotfi Ait Hssaine Yes can do. However, I would like to invite you to join Hossain Academy Facebook for greater interaction about economics, finance and econometrics with me. Thank you Sayed Hossain from Hossain Academy. Please join below and post your question.facebook.com/groups/hossainacademy/
The three time series you are using for this exercise are not stationary!! You are using the non stationary GDP,Money and Export to define the VAR. Hence there is no validity for the results you show in this video.
I agree. We only can use non-stationary variables in VAR in case these variables have cointegration relationship. If the cointegration can not be withdrawn, the stationarity of variables is required then.
that's not true." VAR models are usually guided by theory and are dynamic in nature, so you do not have to worry about the nonstationarity. And Prof. Lutkepohl's studies are always good references."
Thanks for explications, nice and clear . Best wishes from Romania. God be with you, dear precious professor !
You are welcome
I am highly grateful for your effort....which is tremendous learning experience for a person who is a non-econometrican
You are welcome
Dear Kalpataru, Thank you. I would like to invite you to join Hossain Academy Facebook at below link to discuss about economics, econometrics and statistical models using EVIEWS, STATA, R, SPSS, Minitab, Microfit, Lingo, and Excel. Thank you, Dr. Sayed Hossain, Adjunct Economic Faculty, Collin College, Texas.
facebook.com/groups/hossainacademy/
Thank you so much, your videos are really helpful! Everything is explained so clearly that there's no need to look it up in a book.
You are welcome MsEriss
If possible leave a comment in Hossain Academy guest book
Dear MsEriss, Thank you. I would like to invite you to join Hossain Academy Facebook at below link to discuss about economics, econometrics and statistical models using EVIEWS, STATA, R, SPSS, Minitab, Microfit, Lingo, and Excel. Thank you, Dr. Sayed Hossain, Adjunct Economic Faculty, Collin College, Texas.
facebook.com/groups/hossainacademy/
Thank you so much sir,it is a very informative and helpful lecture. Thanks once again :)
Dear Anita, Thank you. I would like to invite you to join Hossain Academy Facebook at below link to discuss about economics, econometrics and statistical models using EVIEWS, STATA, R, SPSS, Minitab, Microfit, Lingo, and Excel. Thank you, Dr. Sayed Hossain, Adjunct Economic Faculty, Collin College, Texas.
facebook.com/groups/hossainacademy/
Thanks so much professor. It is very clear and helpful!!
You are welcome
Dear Susan, Thank you. I would like to invite you to join Hossain Academy Facebook at below link to discuss about economics, econometrics and statistical models using EVIEWS, STATA, R, SPSS, Minitab, Microfit, Lingo, and Excel. Thank you, Dr. Sayed Hossain, Adjunct Economic Faculty, Collin College, Texas.
facebook.com/groups/hossainacademy/
Very helpful sir
Please how do the following diagnostic test (J-B, B-G, ARCH LM, white test and Ramsey test) in VAR model estimated
Sayed Hossain, is there a Non Linear causality example? tks
please professor if I use tydl approach, and I use one dummy variable , the estimation of var in this case in eviews, in the exogenous variables I put all the variables with lag, how about dummy variable I use lag for it, I mean dummy (-3) for example or I put just dummy
Dear Sayed,
is it possibile to run a nonlinear granger causality test in eviews? If yes, how?
Thanks
In this case, when no causality exists, can a VAR estimate be applied, and how can it be interpreted?
Yes I have uploaded data
Hi Mr. Sayed ! How will i know if i will use stationary or nonstationary data in VAR considering that my data are cointegrated? Pls help ASAP thanks ! :)
dear sir i am testing for granger causality between two variables. One of my variable is I(0) and the other is I(2). can i proceed within a VAR framework or do i need to use Toda Yama. Thanks
Hi! Does anyone know the command reference to get the Granger Causality test on a VAR in EViews (instead of going to "View/Lag Structure/Granger Causality" on the VAR ecuation, i'd like to set all the causality tests for different VARs in a program... thank you!
Dear Fernando, I would like to invite you to join Hossain Academy Facebook at below link to discuss about economics, econometrics and statistical models using EVIEWS, STATA, R, SPSS, Minitab, Microfit, Lingo, and Excel. Thank you, Dr. Sayed Hossain, Adjunct Economic Faculty, Collin College, Texas.
facebook.com/groups/hossainacademy/
Hi Sir, I'm using EViews version 8, running a unrestricted VAR model. I am checking the impact of oil price shocks on South Africa. I'm using three variables government revenues, money supply and real exchange rate. When I run Granger Causality test, the result shows no causality running from oil price shocks to other variables. However, when I run bi-variate model (in pairs) e.g. Oil price shocks cause government revenues and highly significant. I don't know how to solve this problem!! Which model should I use (multivariate VAR model or univariate model) to explain the real effect.
elsiddigobaid
hi please is 1lag a good lag selection for granger causality test
Thank you so much for your vid. I have a question that I have to check the unit roots test before regressing the VARs or not ? and if YES and then I find that the variable is unit root I need to do first difference before estimating VARs right ?
Unit root testing need to check first whether the variable is stationary or not as you need stationary data to run VAR.
Hi, If i only have 2 variables which are not cointegrated, do i perform Granger causality in VAR model or Pairwise Granger Causality test? Does VAR model represent short run? How about if 2 variables are cointegrated, can i perform VECM to check long run and short run relationship?
+sin yi
Thank you. I would like to invite you to join “Analysis” in Facebook below to discuss about econometrics, statistics and data analysis using EVIEWS, STATA, R, SPSS, Minitab, Microfit, Lingo, Excel etc. Thank you, Sayed from
Analysis Group. facebook.com/groups/449520105253087/
Thanks for your usful guidelines! However, I have a question about, what the difference between Granger causality test and Granger causality test in VAR Model. Which one can be used to test the correlation between two variables more clear?
qj x Granger causality in VAR is in overall VAR model while other one is not.
Sayed Hossain hi sir how can we control for variables in causality test if we want to find the causality of independent var. on dependent variable and like to control for the effect of some other variables.
Control may be possible but I have not done it yet.
Sayed Hossain Hi Sir. Thanks for the video but I got a question. By applying Granger causality test in VAR Model, you already control the effect of "Money" on GDP when you are testing if "Export" cause "GDP", is that correct?
One is as a whole of the model and the second one between two variable.
hi .. why you enter 2 lag and what is this mean ? and how i choose 1 or 2 o 3 lags ?
haifa fahad You have to choose optimum lag number of lags to be used. Hossain Academy has some videos on lag selection. See those and also join Hossain Academy Facebook for discussion below link.
facebook.com/groups/hossainacademy/
Thank u for ur videos I have greatly benefited from them. Please, I want to perform diagnostic test (J-B, B-G, ARCH LM, white test and Ramsey test) to check for the robustness of the VAR model I estimate how do I do that? I will appreciate ur view on this.
Chinasa Onyenekwe Thank you. Please join Hossain Academy Facebook below so that we can communicate better.facebook.com/groups/hossainacademy/
how to do diagnostic tests in a Granger causality in VAR ?
Any news?
You just use the t-statistic given in Var and whether or not t s significant compared to T-statistic of adjusted number of observation which in the video is 34
hi, sir, thanks for your video, it is a good guideline. I have a question for my research. after i run unit root test using ADF, my data are integrated at level, I（0），mean i not need differentiate it in 1st order in order make the data stationary , so can i still need run Johansen co integration test ?
Carina Tiong Normally you can run Johansen when variables are I(1) or integrated of same order.
Carina Tiong No you can not run Johansen test as variables are I(0)
dear sayed hossein, thanks so much for your useful guidelines. After watching your demonstration of "Granger causality test in a VAR model", I have a question. Correct me if I am wrong but in this particular model we are assuming that all of my variables are 1) first difference stationary, 2) *not* co-integrated (after Johansen test of co-integration). Therefore, I have to do unit root and co-integration testing before proceeding with "granger causality test in a VAR model". I appreciate your thoughts on this. Thanks.
Granger causality should run using stationary data. Unit root testing will allow you to know the status of data, whether it is stationary or non stationary. If data not stationary, then convert it into stationary, and then run granger causality model.. We run Johansen test to know whether we should run VAR or VECM.
thank you Sayed Hossain for your explanation. If my data (variables) are non stationary at level, then I check if they return stationary at first difference or not. For VAR models, all variables should be at the same order of integration.. Here is my question: Do I run VAR models and co-integration tests with LEVEL data or with first-differenced data?. That is my confusion indeed when you say "convert data into stationary and then run granger causality model"...I really appreciate your thoughts on this.
In other words, let's assume that my variables become stationary at first difference according to unit root testing. Do I have to convert my variables into "first difference stationary" before I run VAR models?
You can not assume it, you have to prove it using unit root testing that data are stationary after first differenced. Then you have to use this first differenced data for granger causality as this data is in stationary state.
OK, I got it this time..
thank you.
dear sir,
if i have one of the variables with structural break can i apply co-integrations tests or not, i applied Toda and Yamamoto test and there is one way causality
I am not familiar about Toda test yet so unable to comment.
because when i choose lag 1 my variables granger together
Can we use more than 3 variables for this test
+Lotfi Ait Hssaine Yes can do. However, I would like to invite you to join Hossain Academy Facebook for greater interaction about economics, finance and econometrics with me. Thank you Sayed Hossain from Hossain Academy. Please join below and post your question.facebook.com/groups/hossainacademy/
sir hossain, where is the data for this model? i humbly request for the upload of the data for this model. please. more power to you sir.
Hi Sir. Thanks for your interesting lessons. I have one question. Can we use var model if we have only one independent variable?
T Loum Possible
The three time series you are using for this exercise are not stationary!! You are using the non stationary GDP,Money and Export to define the VAR. Hence there is no validity for the results you show in this video.
I agree with you. We need to test the stationary of data series first before putting them in VAR model.
Trọng Phụng Vũ sufficient condition is cointegration for using VAR or stationarity
I agree. We only can use non-stationary variables in VAR in case these variables have cointegration relationship. If the cointegration can not be withdrawn, the stationarity of variables is required then.
What about my series are I(1)? Can I run Granger test within VAR?
that's not true." VAR models are usually guided by theory and are dynamic in nature, so you do not have to worry about the nonstationarity. And Prof. Lutkepohl's studies are always good references."
Thanks so much
You are welcome
in 2x speed this is still slow
Yes I have uploaded data