(EViews10): ARDL and 3-Ways Causality Checks (1)

[@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.

[@tomaseduardochungaseminari5605]

What can I do if I have a variable that is stationary of order 2, which could I resort to

[@CrunchEconometrix]

Use the Toda-Yamamoto technique.

[@daciovillarreal4742]

I'm glad I suscribed to this channel. Thanks for your useful videos. Keep on the good work!!

[@CrunchEconometrix]

Hi Dacio, I am honoured by your subscription...and thanks for the encouragement. Grateful!

[@mmworks.6767]

Cant stop appreciating you my Prof

[@CrunchEconometrix]

Thanks for the encouraging words and feedback, Elijah. Deeply appreciated!

[@KeziaSpeaksHerMind]

Good day prof, thanks for all you do. How can I run the diagonistics tests on a Panel ARDL using eviews. I tried tofollow the steps you outlined but it seems to not be working

[@CrunchEconometrix]

As indicated in my panel ARDL videos, I use Stata not EViews for PMG analysis. Thanks.

[@nikosstromplos4677]

what can i conclude regarding long-run causality when the coefficient of error correction term is significant while the individual coefficients of long-run relationship are not significant?

[@CrunchEconometrix]

Hi Nikos, there is JOINT causal relation. Watch the videos again. Regards.

[@masterakande5313]

Good day ma, I am one of your subscriber on your channel. What is the implications of having ECT of (-1.29) in an ARDL model? What can be done to have the value within accepted region of 1? Please I am currently doing a research and my ECT is greater than 1.. Thank you ma

[@CrunchEconometrix]

Hi Akande, nobody will take your results serious. Rework your analysis doing a combination of several measures: adjust lags, drop regressors, change the depvar to a closer proxy etc

[@apica1234]

Dear Mam You mentioned somewhere that you interpret the coefficient values of ARDL short run model as reverse from the signs in results. can you clarify that part please.

[@CrunchEconometrix]

Dear Apica, I never said this.

[@immaculatelum5102]

Thanks ma

[@CrunchEconometrix]

You are welcome, Immaculate 😊

[@patrickting4445]

Hi @CrunchEconometrix lets say I have a combination of data sets of stationary and non stationary data. how do i carry out the causality test using ardl?

[@CrunchEconometrix]

Hi Patrick, I showed how in the ARDL-Causality videos.

[@patrickting4445]

@@CrunchEconometrix thanks !

[@nurehh2271]

Thank you

[@CrunchEconometrix]

You are welcome, Sir!

[@larsahnland3411]

So how do we perform Granger and Wald tests from an ARDL model?

[@CrunchEconometrix]

Hi Lars, that is what these causal videos are about. Regards.

[@arhamibrahim8619]

Really a good attempt.

[@CrunchEconometrix]

Thanks for the kind comments, Arham. Deeply appreciated!

[@fatimazahramoussaid350]

hello, can you please make a video on augmented ARDL?, my dependent variable is I(0) and as i know to run ARDL one of the main conditions is that the dependent variable should be I(1), so by reading articles to surpass this problem they use Augmented ARDL!

[@CrunchEconometrix]

Hi Fatima, suggestion is noted. Thanks!

[@fatimazahramoussaid350]

@@CrunchEconometrix Thank you so much

[@RN-jr2qe]

Hello Prof. I want to ask the book references in this video because i'm working with my thesis

[@CrunchEconometrix]

Hitar, there are references listed at the end of my ARDL videos. You can also check out other online resources.

[@RN-jr2qe]

@@CrunchEconometrix Thank you Prof🙏 I have some problem when running the Short-Run ECM form. The selected ARDL model for my thesis was ARDL(3,1,3,0,1,0) and there is cointegration among variables through bound testing. But my independent variables which has 0 lag didn't appear in ECM form. Is there any way to make these independent variables keep appear in ECM form? Because when i read some journal, the variables with 0 lag keep appears in the Short-Run ECM form (Ashfaq & Padda, 2019; Bhatta & Mishra, 2020)

[@CrunchEconometrix]

Hi Hitar, 0 lag implies that the explvar impacts the depvar contemporaneously but NOT with a lag. Nothing wrong with that. You may initiate an email to the authors of those papers for clarity on what they did.

[@belfqihhamza1632]

(from Morocco) hi professor, i used the ARDL bounds test with a mixed order of variables and found evidence of cointegration. now i want to look for causality. can i only infer that causality from ARDL bounds test without resorting to granger or Toda Yamamoto tests?

[@CrunchEconometrix]

Hi Belfqih, my causality videos are quite clear. Please follow through and align your research accordingly. Thanks.

[@michaels.1799]

Hello. I would like to ask. For the case of short run causal effect, do we interpret the coefficients of the significant regressors (based on the t-statistics) from the long run ARDL model or the ECM model if it is proven that there's cointegration within the model and the error correction term can be calculated? Thank you, hope to hear your reply as soon as possible.

[@CrunchEconometrix]

If you estimate ARDL, you get only short-run (SR) coefficients, but with ECM you obtain both LR and SR estimates.

[@michaels.1799]

Okay. But for the short-run coefficients needed for short-run causality interpretation, do I use the coefficients from the short-run ARDL model where the variables have been differenced or the original ARDL model used for detecting cointegration via Bounds Test?

[@CrunchEconometrix]

Hi Michael, this is quite simple. I answered this in the last query even after explaining in detail with the "soft" notes in the video. If there is cointegration, you run the ECM and obtain both LR and SR causal effects. If there's no cointegration, estimate only the ARDL model and obtain SR causal effects. Nothing complicated in this, at all.

[@grizzlybear4807]

Ma'am. I have two I(1) variables, A and B. I used ARDL bound test and found a cointegration when I used B as the dependent variable. However when I applied Granger Causality on VECM and VAR framework, it shows that B granger cause A. Is this even possible or am i doing anything wrong?

[@CrunchEconometrix]

Grizzly, you are mixing 2 different techniques. Makes your work very confusing. Depending on your research objective, it is advisable you stick to ONE technique. Tx

[@danielkayembe6655]

bonjour je suis Daniel KAYEMBE s'il vous plaît veuillez nous poster aussi une vidéo sur le test de causalité de breitung et candelon(2006).

[@CrunchEconometrix]

Hi Dan, I have since responded to this. Thanks

[@goldenresearch88]

CrunchEconometrics: "Subscribing is too small to your Videos seriously am saying" !! ,I have mate recently your Videos, what I did already is subscribe only wich is easy too when compared with my gain from... Thanks and Long live in advance is my wish (GelanaD from Ethio...)

[@CrunchEconometrix]

Hi Gelana, thanks for the positive feedback and remarks on my KZbin videos. Deeply appreciated! I'm glad you find my content helpful. I hope to improve in the next year. Encourage me by SHARING the link to my KZbin Channel with your friends and academic community in Ethiopia 🇪🇹 for awareness. Thanks! ❤️ 🙏

[@danielkayembe6655]

bonjour je suis Daniel KAYEMBE s'il vous plaît veuillez nous poster aussi une vidéo sur le test de causalité de breitung et candelon(2006).

[@CrunchEconometrix]

Hi Daniel, I have responded to this on another thread. Thanks.

[@danielkayembe6655]

bonjour je suis Daniel KAYEMBE s'il vous plaît veuillez nous poster aussi une vidéo sur le test de causalité de breitung et candelon(2006).

[@CrunchEconometrix]

Hi Daniel, thanks for the suggestion. It is duly. However, all new videos are henceforth published on my Teachable paid platform. You may want to enroll. Here's the link cruncheconometrix.teachable.com