Justin, how do impone restrictions in the lags in the VAR model? I am interested in make restrictions about the relations between variables and its lags
Hi Justin, I must say this is a fantastic series of videos on (S)VAR modeling and I would like to thank you for this as it has taught me a lot. My questions are more on the economics side of things: 1- In your explanation of why inflation has a positive response to a shock in the output gap, you said that this was due to the economy "overheating". You further said that inflation increased as a result of a continuous increase in productivity. Shouldn't be the opposite instead? Productivity is a supply-side variable and we know that an increase in the short-run aggregate supply will lead to a decrease in the price level (hence deflation) ceteris paribus. Is there anything I am missing? 2- Do you know what the definition of "output gap" is in the dataset? Real GDP minus Potential GDP or the opposite? I am well aware that this is not an economics video, but I just wanted some clarifications. Thank you very much.
Hi, thanks for your comment. On 1 - Could very well be the case. The emphasis I wanted to make was on the order that I chose but a different line of intuition as you are saying may also be feasible which would take a new ordering. Those assumptions could also be validated should this be tested using a different order in which case you may want to order productivity first. On 2 - Real minus Potential. Hope this helps. Thanks
Thank you so much for this! It was very helpful! Just one short question. The policy rate doesn't look stationary as a downward trend can be seen there. Isnt that a problem for the results of the VAR model?
Thank you for your feedback. Yes, you are correct! A potential consequence (although not all the time) if a non-stationary variable is used is that the IRFs are permanent. In this case, it doesn't happen to be. However, to be on the safe side, you may want to log, difference or get the log difference of any nonstationary variable. I retained it in levels so as to match it with the other variables.
@@JustinEloriaga Thank you for your fast reply. Okay, I think the intention was also to show the implementation of a SVAR. I just thought that maybe leave certain variables non-stationary (in this case the policy rate) to keep the economic interpretation. Thanks a lot.
I hope you respond to comments on old videos, I have been following your methods to create an SVAR model for a final data project, but I keep getting the error Error in VAR(y = ysampled, p = 1) : NAs in y. Would you know why this is? I looked through the data by hand and there are no NAs in the sample data.
Hi Justin , thank you so much for this video. I am using the SVAR in my thesis and I have followed through your example keenly with my data. My only challenge is that I cannot estimate the impuse response function in my case. I keep getting this error: "Error in VAR(y = ysampled, p = 8, type = "const", exogen = NULL) : NAs in y". Any form of assistance will be appreciated. Counting on your favorable response. Best.
Hi, I think the problem is with names of impulse variable. The name of the variable should not include spaces, try to do it with underscores, or put the name together with Uppercase. It should help.
Hi, thank you for the video , helped quite a bit ! Im just wondering is it possible to only include the last part of the fanchart ? So its a bit zoomed in
Hi, very good content! I am trying to model an SVAR of two variables, log of the real price of commodities (I (1)) and mining investment as a percentage of GDP (I (0)), both series are non-cointegrated. My idea is to be able to see the effects of shocks in the price of commodities in mining investment, and for that I would like to be able to differentiate between permanent and transitory shocks to the price of commodities, but I don't know how to do that R. Can you help me? Thank you!
Thank you, very neat and informative example. Glad to have stumbled upon your video.
Thank You so much for the video on Time Series Analysis. It will be a great if you can make videos on Univariate and Multivariate GARCH.
Thank you so much for this excellent explanation and step by step procedure.
how to impose some sign restriction on Amat or Bmat. I don't think R package currently handle this? do you have any insight on it? thanks
Did you ever figure this out?
@@sidgundapaneni3268 did you ever figure this out?
Please explain the process if I impose restrictions on B matrix. I did but there were some error messages.
what is the vertical axis of IRF plot? how to interpret?
Very interesting and instructive to watch your videos . Thanks 👍
How do we have the confidence intervals to double bootstrapping?
Justin, how do impone restrictions in the lags in the VAR model? I am interested in make restrictions about the relations between variables and its lags
WHAT A NICE AND ILLUSTRATIVE PRESENTATION THANKS MANY
What is the name of this package to see the graphics and zoom in on the cuts?
Does VAR Represent reduced form VAR? If not how do we go from this VAR to structural VAR?
Do you (or community) have any video on ggplot to combine all these irfs in a single page?
Hi Justin, I must say this is a fantastic series of videos on (S)VAR modeling and I would like to thank you for this as it has taught me a lot. My questions are more on the economics side of things:
1- In your explanation of why inflation has a positive response to a shock in the output gap, you said that this was due to the economy "overheating". You further said that inflation increased as a result of a continuous increase in productivity. Shouldn't be the opposite instead? Productivity is a supply-side variable and we know that an increase in the short-run aggregate supply will lead to a decrease in the price level (hence deflation) ceteris paribus. Is there anything I am missing?
2- Do you know what the definition of "output gap" is in the dataset? Real GDP minus Potential GDP or the opposite?
I am well aware that this is not an economics video, but I just wanted some clarifications. Thank you very much.
Hi, thanks for your comment.
On 1 - Could very well be the case. The emphasis I wanted to make was on the order that I chose but a different line of intuition as you are saying may also be feasible which would take a new ordering. Those assumptions could also be validated should this be tested using a different order in which case you may want to order productivity first.
On 2 - Real minus Potential.
Hope this helps. Thanks
@@JustinEloriaga Thank you for the prompt response.
Is there a way to apply this to a panel data frame? Thank you for your videos!
What exactly does the A matrix tell us? Why are we looking to fill it?
Did u find answer?
Hi, thanks for all your amazing lectures. just wondering if you are going to post the related lectures on FAVAR and panel SVAR.
Thank you so much for this! It was very helpful! Just one short question. The policy rate doesn't look stationary as a downward trend can be seen there. Isnt that a problem for the results of the VAR model?
Thank you for your feedback. Yes, you are correct! A potential consequence (although not all the time) if a non-stationary variable is used is that the IRFs are permanent. In this case, it doesn't happen to be. However, to be on the safe side, you may want to log, difference or get the log difference of any nonstationary variable. I retained it in levels so as to match it with the other variables.
@@JustinEloriaga Thank you for your fast reply. Okay, I think the intention was also to show the implementation of a SVAR. I just thought that maybe leave certain variables non-stationary (in this case the policy rate) to keep the economic interpretation. Thanks a lot.
@@gian-lucaomari5823 Right! That is another consideration.
@@JustinEloriaga Hey. Would you mind if I drop you a private message on Twitter or Facebook? I would like to ask you another question. Thank you!
I hope you respond to comments on old videos, I have been following your methods to create an SVAR model for a final data project, but I keep getting the error Error in VAR(y = ysampled, p = 1) : NAs in y. Would you know why this is? I looked through the data by hand and there are no NAs in the sample data.
Hi Justin , thank you so much for this video. I am using the SVAR in my thesis and I have followed through your example keenly with my data. My only challenge is that I cannot estimate the impuse response function in my case. I keep getting this error:
"Error in VAR(y = ysampled, p = 8, type = "const", exogen = NULL) :
NAs in y".
Any form of assistance will be appreciated.
Counting on your favorable response.
Best.
your y variable seems to have NA values in it (lines with no value) which then can't be calculated. You need to get rid of those before
@@Fissqui Thank you.
I have done that.
What diagnostic tests do I need to perform?
Best.
Hi, Thanks for such a great tutorial. I am getting an error while running the code.
SVARrtgs SVARrtgs
Hi, I think the problem is with names of impulse variable. The name of the variable should not include spaces, try to do it with underscores, or put the name together with Uppercase. It should help.
@@georgekokhreidze5982 thank you sir. It worked.
Hi, thank you for the video , helped quite a bit ! Im just wondering is it possible to only include the last part of the fanchart ? So its a bit zoomed in
Amazing.... I hope you make tutorial Threshold VAR... Thank u so much
Amazing stuff man!!!
Could you do a video on how to run the Panel SVAR in Rstudio?
Or is there an email that I can contact you?
really good idea
Yes please! Panel SVAR would be great!
Nice video, thank you!
thanks for explaining the matrix amat
You are so awesome
Thank you Justin. If you can please make some calculations on output gap in R. It is better to make few ways, for example HP filter, Kalman and TFP
God bless you
Thank you !
Nice video, i try to do this on commodities prices, so i had to put my variables in log and the meaning of my graph is shit
thank you so much
Hi, very good content!
I am trying to model an SVAR of two variables, log of the real price of commodities (I (1)) and mining investment as a percentage of GDP (I (0)), both series are non-cointegrated. My idea is to be able to see the effects of shocks in the price of commodities in mining investment, and for that I would like to be able to differentiate between permanent and transitory shocks to the price of commodities, but I don't know how to do that R.
Can you help me? Thank you!
wonderfull
Top! But, this data is real?
If, 10 qtr to inflation accommodation is game over for millennials.