UA-cam recently changed the way my content will be monetised. My channel now needs 1,000 subscribers. So it would be amazing if you show your support by both watching my videos and subscribing to my channel if you haven’t done so already. Monetising my videos allows me to invest back into the channel with some new equipment so this small gesture from you will be extremely huge for me. Many thanks for your support….CrunchEconometrix loves to teach, support my Channel with your subscription and sharing my videos with your cohorts.
Great Job. Dr. Ngozi. I had like to ask, is it possible to run a VAR and impulse response if I have a mix of I(0) and I(1) variables. My dependent variable is I(0). Thank you very much, Dr.
Good day and thank you for your amazing videos. They have helped me alot in my thesis which I am still writing. Using a VAR model I am looking at the impact of exchange rate fluctuation on the performance of the manufacturing sector using monthly data. At 13 lags my VAR model is stable, it passed all the diagnostic test but some coefficients of the result are not statisticaly significant while some are. Can I go ahead and interpret the impulse response function and variance decomposition even though my coefficients are not all statistically significant? And can I use the result to make any conclusions on the relationship between the variables? That's my question.
Thanks for the positive feedback, Adamu. Deeply appreciated! Yes, go ahead once the diagnostics are ok. IRF is about innovation accounting of the error terms in the VAR system. Please may I know from where (location) you are reaching me?
Thanks a lot once again for the swift feedback, I highly appreciate it. I am reaching you from Kaduna State Nigeria. One more question Please, relating to the diagnostic check. How important is the normality test in a VAR model? there are models I generated using annual data (29 years) with 2 lags that passed all other test but the residuals are reported to be not normally distributed using the Jaque Bera test. Can I still go ahead and interpret IRF and VD. Thank you very much and God bless you.
I wish to know can we get first difference the raw series to get stationarity. it is said that for IRF it should not be used first difference. In order to use VAR we need to keep the data in the stationary form. therefore, I get confused how to do IRF do I use only stationary series or can we use nonstationary series fro IRF
Hi Selliah, I honestly have no idea what you are talking about. My videos on IRF are well explained. But if you need more information, kindly check other online resources.
Good morning Ma'am. Trust you're good. Please why is there no multicolinearity test in the diagnostic test for a VAR model. Could it be because the variables are endogenous and they only have significance on themselves? I'm not sure. Thank you.
can we employ IRF for non stationary series . I understand VAR can be used for stationary series . so to do IRF , we need to use VAR in sttationary form so we need to do first difference. is it correct
Hi. I really like your videos. It has helped me a lot in my dissertation. I have a question on point 3: VAR must be specified in levels. What if after we did the stationarity test we will need to do first difference on the variables to make it stationary? What's the point of doing stationarity test if during VAR we will need to specify the variables in levels after all? I'm just not understanding this. Don't we need to do VAR on the stationary variables? Thanks!
Sarah Paderes Hi Sarah, thanks for the kind words and good to hear that my videos are helpful in some ways. Ok, relax girl and don't be confused about point 3. I'll clear it up by referring you to look up a basic econometrics textbook (check in Gujarati's Basic Econometrics) and see how the VAR equation is specified. YES to use the VAR model, all variables must be stationary at 1st difference. Another important thing to know is that, the VAR model can be estimated either by using the VAR algorithm or the OLS algorithm. If you are using the VAR algorithm, the variables must be listed in their LEVEL forms (which is the approach I used) and if it's the OLS algorithm, the variables must be listed in their FIRST DIFFERENCE (some authors use this). Also, know that VAR estimates should be given the CETERIS PARIBUS interpretation.
UA-cam recently changed the way my content will be monetised. My channel now needs 1,000 subscribers. So it would be amazing if you show your support by both watching my videos and subscribing to my channel if you haven’t done so already. Monetising my videos allows me to invest back into the channel with some new equipment so this small gesture from you will be extremely huge for me. Many thanks for your support….CrunchEconometrix loves to teach, support my Channel with your subscription and sharing my videos with your cohorts.
very interesting. You are unique academician. Thank you so much
Glad you think so...deeply appreciated! 🥰🙏
Excellent video mam!
Thanks, Kristi for the encouraging feedback. Deeply appreciated! Please may I know from where (location) you are reaching me?
@@CrunchEconometrix I am a PhD student in the US
great job as usual...hope in next video's you explain the Toda Yamamoto test
I'll do my best, Jasem...as usual.
Good day ma, please I would like to know the most appropriate model to adopt when you have a time series data of order 1 and 2 integration.
Hi Yusau, with I(2) series, I advise you read up on the Toda-Yamamoto technique and adopt. Thanks.
I'm so grateful. I must also confess that your videos have been of great help to me. Thank you and keep up the good work up.
Great Job. Dr. Ngozi. I had like to ask, is it possible to run a VAR and impulse response if I have a mix of I(0) and I(1) variables. My dependent variable is I(0). Thank you very much, Dr.
Thanks for the encouraging feedback, Abidemi. Not at all. VAR estimation requires all series to be I(1).
@@CrunchEconometrix Thank you so much.
@@CrunchEconometrixDear Dr. Ngozi, is there any method you recommend apart from ARDL?
These techniques go with I(1) depvar and I(0) regressors: Canonical, FMOLS, and DOLS.
@@CrunchEconometrix Thank you so much Dr. Thank you.
Good day and thank you for your amazing videos. They have helped me alot in my thesis which I am still writing. Using a VAR model I am looking at the impact of exchange rate fluctuation on the performance of the manufacturing sector using monthly data. At 13 lags my VAR model is stable, it passed all the diagnostic test but some coefficients of the result are not statisticaly significant while some are. Can I go ahead and interpret the impulse response function and variance decomposition even though my coefficients are not all statistically significant? And can I use the result to make any conclusions on the relationship between the variables? That's my question.
Thanks for the positive feedback, Adamu. Deeply appreciated! Yes, go ahead once the diagnostics are ok. IRF is about innovation accounting of the error terms in the VAR system. Please may I know from where (location) you are reaching me?
Thanks a lot once again for the swift feedback, I highly appreciate it. I am reaching you from Kaduna State Nigeria. One more question Please, relating to the diagnostic check. How important is the normality test in a VAR model? there are models I generated using annual data (29 years) with 2 lags that passed all other test but the residuals are reported to be not normally distributed using the Jaque Bera test. Can I still go ahead and interpret IRF and VD. Thank you very much and God bless you.
Yes, go ahead. Non-normality of errors occurs in VAR. Please do a Google search for papers who assert this.
@@CrunchEconometrix I will do just that.
Tt
I wish to know can we get first difference the raw series to get stationarity. it is said that for IRF it should not be used first difference. In order to use VAR we need to keep the data in the stationary form. therefore, I get confused how to do IRF
do I use only stationary series or can we use nonstationary series fro IRF
Hi Selliah, I honestly have no idea what you are talking about. My videos on IRF are well explained. But if you need more information, kindly check other online resources.
hi... informative video, can we have this applications for Panel Data?
Hi Charith, thanks for the positive feedback...deeply appreciated. As to your query, not to my knowledge.
Good morning Ma'am. Trust you're good. Please why is there no multicolinearity test in the diagnostic test for a VAR model. Could it be because the variables are endogenous and they only have significance on themselves? I'm not sure.
Thank you.
You answered your question correctly, Judith😊❤️
can we employ IRF for non stationary series . I understand VAR can be used for stationary series . so to do IRF , we need to use VAR in sttationary form so we need to do first difference.
is it correct
Hi Selliah, I have laid out all you need about the IRF. But if you need further information you may seek other online resources. Thanks.
Hi. I really like your videos. It has helped me a lot in my dissertation. I have a question on point 3: VAR must be specified in levels. What if after we did the stationarity test we will need to do first difference on the variables to make it stationary? What's the point of doing stationarity test if during VAR we will need to specify the variables in levels after all? I'm just not understanding this. Don't we need to do VAR on the stationary variables? Thanks!
Sarah Paderes Hi Sarah, thanks for the kind words and good to hear that my videos are helpful in some ways. Ok, relax girl and don't be confused about point 3. I'll clear it up by referring you to look up a basic econometrics textbook (check in Gujarati's Basic Econometrics) and see how the VAR equation is specified. YES to use the VAR model, all variables must be stationary at 1st difference. Another important thing to know is that, the VAR model can be estimated either by using the VAR algorithm or the OLS algorithm. If you are using the VAR algorithm, the variables must be listed in their LEVEL forms (which is the approach I used) and if it's the OLS algorithm, the variables must be listed in their FIRST DIFFERENCE (some authors use this). Also, know that VAR estimates should be given the CETERIS PARIBUS interpretation.
Oh I get it now. Yes, your explanation is very helpful. Keep up the good work.
I’m having issues doing that in my eviews, can u help me please?
Hi Daniela, what issues?
@@CrunchEconometrix can I talk with you privately? Please!
Kindly post them here, Daniela. I'm constrained by time and my schedules to offer personalized tutoring. Thanks for your understanding.