(EViews10): VECM and 3-Ways Causality Checks (3)

[@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, support my Channel with your subscription and sharing my videos with your cohorts

[@ogamartinz2757]

You 've made my day, very sound and precise to point. Bless you.

[@CrunchEconometrix]

Glad it was helpful!

[@Mmagnoliaa_]

Thank you so much for the video 💝

[@CrunchEconometrix]

U're welcome, Belles. Please may I know from where (location) you are reaching me?

[@Mmagnoliaa_]

CrunchEconometrix I’m from Singapore. I want to ask you, if my variable is significant at level (intercept), not significant at (intercept & trend) however significant at 1st difference (both intercept ; intercept & trend) can i proceed to var ? Btw I’ve been reading your blog too. Because it’s very hard for me to find a reliable sources to analyse my data. Lucky I found your video and writing 👍🏻❤️

[@victoriaisere109]

great video. keep up the good work.

[@CrunchEconometrix]

Thanks Vic, for the positive feedback and remarks on my video. Deeply appreciated. May I know from where (location) you are reaching me?

[@victoriaisere109]

@@CrunchEconometrix auchi, edo state

[@CrunchEconometrix]

@@victoriaisere109 Awesome! I'll appreciate it if you can share the link to my KZbin Channel with your friends and academic on social media for awareness. They'll learn some useful tips and hints too. Thanks 😊

[@yamiletumba345]

Thanks for the video. One question, Can i use these tests for panel data? Or I have to use Dumitrescu Hurlin test?

[@CrunchEconometrix]

Not sure if you can...but no harm in trying.

[@MShaf-ic9td]

Dear professor, Thank you so much for such an excellent explanation. I was curious to know how the researchers report the results of VAR, VECM, particularly the short and long run Causal relationship. Do you know any journal article who has done so? That will help a lot. Thanks

[@CrunchEconometrix]

Hi Shaf, a Google search will yield several papers. Give it a try. Thanks.

[@MShaf-ic9td]

OK

[@surojitdey574]

Madam I'm from University of calctta, India. Your video really help me a lot in my research work.

[@CrunchEconometrix]

Thanks for the encouraging feedback, Surojit! I'm glad you find my videos helpful. Wish you God's greatness🙏

[@ammarali4420]

If we take log of all variables and diagnostic test is still showing the problem of heteroscedasticity normality and auto correlation then what should be the next step ?

[@CrunchEconometrix]

Hi Ammar, you may want to modify the lag structure and re-estimate the model.

[@ammarali4420]

@@CrunchEconometrix Thank you Ma'am ...I am highly obliged

[@anshulagrawal8529]

Hello ma'am, in my data there is a cointegration between all the variables, but there is no unidirectional and bidirectional causality between them.. Is it possible if there is a cointegration between variables but there is no unidirectional and bidirectional causality exist between them.

[@CrunchEconometrix]

Hi Anshul, that is possible since cointegration and causality address DIFFERENT issues in a model/study.

[@chrisalwis1626]

Dear Dr. your tutorials are extremely helpful. Therefore, I would like to suggest to start a basic statistics series for those who have no basic knowledge on the subject. This is lacking in the subject. Nobody has started such a basic statistics and econometric you tube tutorials. pls. think of this suggestion. Thanks

[@CrunchEconometrix]

Hi Chris, I humbled and encouraged by your comments about my KZbin videos. Deeply appreciated! As per your suggestion, statistics is not my area, unfortunately. Thanks!

[@praisedee7696]

Thank you ma for the presentation. In a case where, the Jarque-Bera test reveals a result that is not normally distributed, can we go ahead with the result? Given that there is no heteroskedasticity, no autocorrelation and no multicollinearity.

[@CrunchEconometrix]

Yes, go ahead.

[@mickeykozzi]

Firstly, thank you for your videos. They are excellent. Just a question on the pairwise granger causality test. I can see that you did not difference the series. Is this an error? Thank you :)

[@CrunchEconometrix]

Hi Michael, thanks for the positive feedback and kind remarks on my KZbin videos. Deeply appreciated! As to your query, causality should be performed on the raw series or their log transformation not on the differenced series. Please may I know from where (location) you are reaching me?

[@mickeykozzi]

@@CrunchEconometrix Hello and thank you for your reply. I am a PhD student from Australia :). Because you are running a vecm and know the series are non stationary, I thought that we have to difference the series when running a pairwise granger causality test? I remember my supervisor saying this. Also, I have found the below youtube video where they difference the series as well. kzbin.info/www/bejne/hpOnZmp3g7h5j7c Thank you Dr.

[@samueljlaltlanzaua363]

Professor, may I humbly submit one request? I have learned that for working with non-stationary time series there is an approach by the name of Cochrane-Orcutt iteration method. Can you make a detailed video on the topic while showing how it is done in eviews? How the results are to be interpreted etc?

[@CrunchEconometrix]

Hi Maps, I will have to do some findings on this. Thanks.

[@sharon246787]

How do you deal with multiple co-integrating equations? Also, multicollinearity, heteroskedasticity?

[@CrunchEconometrix]

Hi Sharon, please watch my clip on multicollinearity. It will guide you on what to do. For heteroscedasticity, there are battery of measures: using the robust option, dropping variable, logs etc. You may need to do some of these.

[@sharon246787]

@@CrunchEconometrix Thanks for responding so quickly. These videos really helped me. Very few go into details and take such care in explaining VECM.

[@CrunchEconometrix]

@@sharon246787 Thanks for being appreciative Sharon it's a great pleasure helping students and fellow researchers. Keep watching, keep sharing!😊

[@m.dawoodasri2755]

Hi teacher can we apply Causality through fixed and random effect models in Panel data, where N

[@CrunchEconometrix]

When N

[@ronnyjimenez22]

Seems to me that every variable is differenced by the time you run Granger's test. Why is that? I have both variables on Logs and DIFF(LOGS)) and as soon as I perform a granger test on the differenced data, Grander will perform a second difference.

[@CrunchEconometrix]

Hi Ronny, kindly follow the steps as shown in the clip. Anything contrary will yield a contrary outcome. Please may I know from where (location) you are reaching me?

[@ronnyjimenez22]

CrunchEconometrix Maryland, USA

[@CrunchEconometrix]

@@ronnyjimenez22 Awesome! Kindly share the link to my KZbin Channel with your students and academic community in the US 🇺🇸...and stay safe 🙏

[@pedromrfernandes]

in my model with two variables, i have one ECT with a negative but not significant, and the other ECT is positive and significant. i switched the independent variable for the dependent (target) variable in the model and now the negative ECT is significant and the positive is not significant. is that ok?

[@CrunchEconometrix]

Hi Pedro, you switched a regressor to a regressand all because you need a negaTive and significant ECT? I don't understand this.

[@pedromrfernandes]

@@CrunchEconometrix i tried to assess whether the causality was from one thing to another, thus, i put that another thing as the target variable. however, no ECT from that vector was negative and significant. one was positive and significant and the other was negative and not significant. thus (to find a more normal result, is guess), i switched that another thing to a regressor and now i have one negative and significant ECT and one positive and not significant ECT. what does this mean? afterwards, i used the IRF and the pairwise granger causality test. indeed, the long-run relationship is the reserve of what i initially tested.