(EViews10)Interpret VAR, Forecast Error Variance Decomposition

[@CrunchEconometrix]

KZbin 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, help me stay online.

[@achudakhinkudachin2048]

That was very helpful how you interrelated the coefficients to fevds in the var model, other teachers don't do that. It is a great video!

[@CrunchEconometrix]

Thanks for your encouraging feedback, deeply appreciated! 🙏🥰

[@Econotitans]

your work is commendable. u explain the way a student really want to be explained

[@CrunchEconometrix]

Thanks Zunerah, for the encouraging feedback. Deeply appreciated! 🙏☺️

[@Kotletka131]

I'm now writing an essay for an econometrics course, and your videos rescued me a lot! Box-Jenkins videos were the most understandable among all I've seen, and VAR ones are also great. Thank you very much!

[@CrunchEconometrix]

Thanks for the encouraging feedback, Kot. Deeply appreciated! Please may I know from where (location) you are reaching me?

[@Kotletka131]

@@CrunchEconometrix from Russia 😄

[@marnix122]

Your tutorials are super helpful madam, thank you so much for all your hard work and explanation!

[@CrunchEconometrix]

Thanks Mamix, for the encouraging feedback! Deeply appreciated! May God bless you, amen 🙏

[@antonkajor1218]

amazing job.... I learn alot from your tutorial.. God bless you

[@CrunchEconometrix]

Hi Anton, thanks for the kind words and positive feedback. Deeply appreciated! Please may I know from where (location) you are reaching me?

[@antonkajor1218]

@@CrunchEconometrix I'm from yogyakarta Indonesia. I'm waiting for your next videos :))

[@panggaulibau9849]

thnk you i love you. hopefully can graduated next year

[@CrunchEconometrix]

That's gr8, Pang...glad to be of help! 💕 😊

[@kanya1998]

Dear Ngozi, You are doing a great job, i like the way you put your explanations, easy to understand, you are a great my great tutor! i do have one suggestion though, could you please number your tutorials? sometimes it is not easy to tell whether there is a previous tutorial or next. Once again, keep up the good work, spread the knowledge!

[@CrunchEconometrix]

Hi Anthony, thanks for the positive feedback on my videos. Deeply appreciated! 💕 Yes, I number my videos and I always mention the prerequisite videos to watch in all my presentations. Kindly share my Channel link with your students and academic networks. May I know where you are reaching me from?

[@daiane_2310]

Always amazing videos! God bless you!

[@CrunchEconometrix]

Thanks for the kind words, Daiane... deeply appreciated! ❤️

[@mohammedarmah9035]

You are doing a great job. Keep it up.

[@CrunchEconometrix]

Hi Mohammed, thanks for the encouraging feedback. Deeply appreciated! Please may I know from where (location) you are reaching me?

[@florencesitima2697]

You are doing an immense job,keep it up.But iwould like to hear much on step by step procedure in performing bivariate VAR model.

[@CrunchEconometrix]

Hi Florence, thanks for the encouraging remark. Deeply appreciated! Procedure for a bivariate VAR is the same as in any multivariate VAR. May I know from where (location) you are reaching me?

[@genesisb.kollie753]

Thank you very much for your tutorial. It is helpful

[@CrunchEconometrix]

Hi Genesis, I'm really glad that you got value while watching and I hope that you have subscribed so that you get my videos the moment I upload them. Kindly help to share my link kzbin.info to your friends and social media community. A lot of beginners need to know that econometrics is not difficult at all.

[@saada.izeddin7712]

Hello CrunchEconometrix team, thank you for the great efforts. My question is about the S.E. column, how to interpret it and its role in the model?

[@CrunchEconometrix]

Hi Saad, I restrict myself to the relevant areas of FEVD. You may want to check out other online resources on the SE column. Thanks

[@bazezewchekol6458]

Thank you very much and wish long live

[@CrunchEconometrix]

You too, Sir😊

[@ucheosemenam4101]

Nice videos. Please Ma, can you make out time to do a video on SVAR and how to impose restrictions on Eviews?

[@CrunchEconometrix]

Great suggestion Uche. I will but it will be posted on my paid Teachable platform. Here is the link cruncheconometrix.teachable.com/p/practical-econometrics-for-researchers-beginners-and-advanced-level-users-perba/

[@abusuhayb6791]

Thanks for your generosity for providing such useful tutorial, if I understand you well, is this what you mean The T-stat LNRGDP(-1) showed highly significant 7.68 which means a 0.99% increase on LNGDP for a 1% increase in LNGDP(-1). Similarly, The T-stat LNEDS(-1) showed highly significant 2.04 which means a 0.31% increase on LNGDP for a 1% increase in LNEDS(-1).

[@CrunchEconometrix]

You got it right, Abu!👏🏽👏🏽👏🏽 Please may I know from where (location) you are reaching me?

[@abusuhayb6791]

@@CrunchEconometrix Qatar

[@moisedjepangkouamo469]

very helpfull tutorial. i appreciate. but my principal preoccupation is how to run GMM model througt eviews?

[@CrunchEconometrix]

You may need to check other online resources.

[@phillipsturm5701]

You are doing a great job, thank you. By which percentage is said that the exogenous variable has a great influence on the endogenous variable. ?

[@CrunchEconometrix]

Hi Phillip, thanks for the positive feedback on my videos. Deeply appreciated! 💕 Strong exogenous influence is when other variables in the VAR model do not influence movement in the outcome variable. No industry guidelines as to the percentage. But any figure below 50% should do. May I know from where (location) you are reaching me?

[@willwu5366]

Hi, I got a question that what if I obtained a large S.E in the variance decomposition analysis? Thank you!

[@CrunchEconometrix]

Hi Will, I suggest that you use the p-value to take your decision.

[@sarahben1]

Hi, this video was very informative! I have a question regarding the p values and coefficients I have obtained after running a VAR - some of my variables show a statistically significant p-value (less than 0.05) however they have a negative coefficient and negative t-statistics. What is your opinion on this? How shall I go about interpreting these results? Or is my model wrong? Fortunately this is only occurring to my control variables and not my variable of interest.

[@CrunchEconometrix]

Thanks for the positive feedback, Sarah. Negative coefficients imply that when the regressors change the dependent variable falls.

[@dimasmukhlas3952]

Thanks! Its helpful!

[@CrunchEconometrix]

Thanks, Dimas for the encouraging feedback. Deeply appreciated! 🙏💗

[@sbodavtians5591]

Hi Ngozi thank you so much for your videos, i just have a question, how can you tell by the t-statisic what a variable is significant or not? What do you compare the t-statisic to?

[@CrunchEconometrix]

Hi Davtians, thanks for the positive feedback on my videos. Deeply appreciated. The regression output is based on a 2-tailed test using the 95% confidence interval at which a t-stat of 1.96 is considered to be significant at the 5% level. However, because many students often get confused about the t-stat, I encourage the use of the p-values which gives the exact level of significance at which the H0 can be rejected...may I know from where (location) you are reaching me?

[@BrocoliEnjoyer1]

Thank you for the explanation ma. For the VAR estimates, how do you know when the results are significant?

[@CrunchEconometrix]

Use the t-stat like I mentioned in the clip. I'll refer you to any basic econometrics textbooks for more info on hypotheses testing and inferences.

[@TheAndreapop22]

hi!! we do not have to invert the signs??

[@CrunchEconometrix]

Hi Andrea, not sure what you mean.

[@gobindaroy6616]

Excellent !

[@CrunchEconometrix]

Thanks Gobinda, for the positive feedback. Deeply appreciated! Kindly share my Channel link with your students and academic community...thanks!

[@dorstellappiah3173]

Hello Dr, please in presenting your VAR results is there a need to presents all the results of the other dependent variables. in my case i have one dep. variable and three indep. variables. Do i need to presents the other independent variables being used as dependent variables?

[@CrunchEconometrix]

The choice is yours really. You can report on all or only the main variable of interest. Also, check papers that used VAR and adapt their style.

[@okanaybar]

I like your video. Thank you. My question is in regards to the endogenity-exogenity terms. Regarding the terms under "From other variables", should the first one, "Strongly Exogenous" not be "implies strong influence on dependend variable"?. Or is it true to say "implies weak influence on dependend variable"?

[@CrunchEconometrix]

Hi Okan, thanks for complimenting my video, deeply appreciated. The interpretation is correct. Remember that VAR is an endogenous model, so if other variables have weak influence on the outcome variable, it implies that those other variables exhibit "strong exogeneity"....may I know where you are reaching me from?

[@okanaybar]

@@CrunchEconometrix That's right. I am from Turkey. I am also Phd studen in France. I taught Finance at Rennes School of Busines and at Rennes Campus of Coventry University. I am also finance practicioner. I am commodity trading expert.

[@CrunchEconometrix]

@@okanaybar Amazing profile...impressive!!! Kindly share my videos with your students and academic networks. Thanks!

[@okanaybar]

When you do Var decomposition (multivariable process) with 2 lags, would it be meaningless to do bi-variate VAR and interpret results? Or the VAR model should also be multivariable?

[@okanaybar]

@@CrunchEconometrix I sure will

[@goherfatima9791]

Hello. thank yo for uploading these good and very informative videos. I want to make you request that can you please make videos for estimating structural VAR (SVAR) and specifically how to impose restrictions. I am learning it these days but your help in this regard would be highly appreciated. Many thanks

[@CrunchEconometrix]

Alright Goher, I'll have to learn SVAR and once I fully understand, then I will do videos on it. thanks!

[@saada.izeddin7712]

Hi Fatima any helping results do you have for your question? I have to understand the SVAR model and how to apply restrictions to finish my Dessertaion. Best Regards

[@CrunchEconometrix]

Saad, there are several KZbin videos on SVAR. Do a thorough search on KZbin.

[@axelndjore1896]

When do we use first difference to run VAR estimation

[@CrunchEconometrix]

Axel, I have responded to you on a different thread.

[@paleesemendy7497]

thanks for the head up😍

[@CrunchEconometrix]

Ure so welcome 🥰🙏

[@milesjd3021]

Hello. Thank you so much for your tutorial! It is helping me a lot right now. :) But I am just wondering, in writing the VAR model, do I have to write the full model (even some of the coefficients are not significant) or I just need to present the model with only the significant coefficients included? Thank you so much!

[@CrunchEconometrix]

I think you mixed up your queries. You specify a model BEFORE estimation and coefficients (significant and not significant) are obtained AFTER estimation

[@milesjd3021]

@@CrunchEconometrix what if I already specified the model, but after estimation, some of the obtained coefficients are not significant. Can I still use the nonsignificant coefficients in substituting to the specified model? P. S. Sorry if I have too much question.

[@CrunchEconometrix]

I don't understand what you imply by "Can I still use the nonsignificant coefficients in substituting to the specified model?"

[@shubhamgarg9540]

Mam for running VAR model, our variable should be stationery at the same level. Is it necessary only for endogenous variables or also for exogenous variable. If all our variable are stationery at same level, but one exogenous variable are not, then we can run Var model or not. Please clarify

[@CrunchEconometrix]

Hi Shubham, ALL the variables MUST be integrated of order one. I will advise you watch my VAR videos. Thanks.

[@pedromrfernandes]

Thank you for the video. A 93,2% increase on real GDP? GDP almost doubles after a 1% increase em GDP(-1)?! does this make sense? or am I understanding this wrong? or is the variation of GDP in period t-1 responsible for 92,3% of the variation in period t?

[@CrunchEconometrix]

You scientifically interpret based on the outcome of your estimation. You can then find the intuition for it.

[@pjmccloskey225]

I agree with Pedro. Bosede means a 0.932% increase for a 1% increase.

[@CrunchEconometrix]

Thanks for the clarity PJ, that was definitely a slip from me. Gracias!

[@immaculatelum5102]

Thanks ma

[@CrunchEconometrix]

You are very welcome, Immaculate.

[@30Oliviaa]

Is it possible to do a Generalised Forecast Error Variance Decomposition in Eviews? How?

[@CrunchEconometrix]

Hi Priya, apologies for the late response. Could be, not sure.

[@ብሬአለክስ]

Thanks alot

[@CrunchEconometrix]

Glad you liked the video! 🥰🙏

[@mehdifarouki330]

WHAT IS SE PLEASE

[@CrunchEconometrix]

Hi Mehdi, SE means Standard Error.