Hello Everyone! Thanks for watching! Part 2 of the tutorial is here! I hope you enjoy it. IMPORTANT: To clarify, if your Error Correction Term is positive (instead of negative), it means your model will not converge to an equilibrium. The disequilibrium will be permanent. ✅ You can buy for a small amount the EViews Wokfile complete explained step by step + video slides + dataset at : payhip.com/b/x9N7v ✅ Visit my website for more information about the topic covered in the video: www.jdeconomics.com/cointegration-and-error-correction-model/ ✅Feel free to subscribe for more content! ruclips.net/channel/UC5P21WGFO4WRUlAiGLcwymg ✅Ensure to watch the 1st Video: "Cointegration - Engle and Granger method in EViews". Link: ruclips.net/video/4DBXBLIOHGE/видео.html I wish you good luck on your research and courses! JDEconomics.
I went through over 15 videos on RUclips and read around 20 articles and many other academic resources about this subject, I can confidently say that this video is a life saveeer and much better than all the unnecessary materials I went through .
Thanks a lot for taking the time to provide such a kind feedback! I am glad I was able to help you understand the topic. Feel free to check my website and share my content with your close ones! Good luck with your courses! JD
Hello JD! This is awesome. Just one question, is there supposed to be a constant in a differenced regression? I would have thought that since constants are the same across periods, they would be deleted out.
Hi Sir, thank you so much for your constructive tutorial. Can you make a video for cointegration bound test applying the ARDL model in Error Correction form showing the estimation for both the long run and Short run error correction model using E-views. Thanks
Thanks! Doing so would have no real sense as there are no corrections in the residuals. I hope that helps! Make sure to check my website: juandamico.start.page and feel free to share my channel with your friends! I wish you good luck! JD
Hi! Great video! I´ve got a question, though. What if I get a statistical significant relationship in the long run but not in the short run (ETC is statistical significant and the values are as expected, but for example imports shows to be only significant in the long run not in the short run)
I am studying price dynamics between retail and farmgate prices. What if the sign of the long run and short run adjustments are different? Also, what if the short run adjustment is not significant but the ect is negative and significant? I hope I can show you my data and results cause I badly need help on this one.
As a tentative to answer your very important question, you may want to re estimate your short run model after watching closely heteroskedasticity and serial correction. Which will give you consistent non unbiased estimates. With these estimated obtained in OLS, you can run an ec approach to which will provide results worth advising public policies
Great work and i appreciated you alot, however, plz, further explain my argument in detail: like what would be the intercept value ?i.e, alpha = 0, alpha > 0 or alpha .01, >.05 and >.10. so, if this is the case, then it means, that the model is stable and valid.explain? example CAPM, FF3FM, FF4FM, FF5FM and FF6FM.
How is the written short-term equation in case of more than 2 variables, should the number of ECT estimated coefficients be the same as the number of analyzed variables in the equation?
Hi, I am using Eviews. You can check my website and see in eviews the first tutorial where I show how to download economic data and import it to the software. Good luck! JD sites.google.com/view/jdeconomics-
Sir I am confused about when ECM model got estimated then how to forecast for the specific dependent variable while ECM independent term residuals are not available for the forecast period.
Hello sir, I hope you will answer my question. I found error correction form negative sign (-1.07) and statistical significant. Is there any problem with that? Because it’s not in the range of 0 and -1. There are many articles that similar to my finding.
Hey. If it’s a linear regression, it should he between -1 and 0. If it is more negative, then the results are explosive. Otherwise, you could say the adjustments are immediate. In other words, takes no time to adjust a disequilibrium. But it’s not really a good analysis. Cherrs
Hi, thanks for your message. They are in order. If you play the playlist, or check the list, they are all ordered. Unfortunately I don't put a number because if I add a new video, then I have to change the numbers all the time. The playlist is here: ruclips.net/p/PLsZ8kVwX52ZGXxJ-bqP-WmRBD2HytO4d4 Regards, JD
Thank you for your explanation video, sir. It's really helpful for my thesis. But, I have little troubles here. 1) What can I do, if the model doesn't fulfill normality assumption? And how can I fix this? 2) the ECT value is positive instead of negative, does it mean I have to re-estimate my model? Thank you very much for your answer and help.
Hello, Thanks for your message. If your residuals aren't normal is not the end of the game. Don't worry. Autocorrelation is more important. There can be some part in your residuals that are spiky, and maybe a dummy variable can help for that particular period. In terms of the error correction term, as you may have seen in the explanation I provided, the negative sign allows the model to correct. Two negative signs will make it positive (picture a trend going very very down, the negative sign will correct it and make it positive) and other way: (very positive (+) a negative sign will make it negative (-) and go down). That's the intuition. Remember the term is allowing the discrepancies between the short and long run to be corrected. Regards, JD
@@JDEconomics Good morning Sir, Thank you very much for your videos. I was just asking myself this question (my model doesn’t fulfill normality assumption) because I had the same problem so I thank you for this answer. All the other conditions are met ( no serial corrélation and no hétérostacity) and my two séries are cointegrated. I was just wondering if it is a problem that I have only between 25 and 30 annual data? Thanks in advance
Hello sir please what can cause the THRESHOLD technic to disappear from the available technique in e-vews if one want to run a data using the threshold analysis for a number of selected countries in the ecowas region.
hello i just wanna ask again, will the result differ if i exchange the dependent and independent variables? what does it mean if it shows that it is cointegrated with significant ect when variable a is used as dependent variable but the results differ when i used variable b as the dependent variable?
hello. i have a question. what if both engle and granger and phillips ouilaris shows that the residuals are stationary which makes my variables cointegrated. however when i generate the short run model the ect shows that it is not significant?
@@JDEconomics what if my ect is finally negative and significant but the other variables like the constant and lagged difference of the other variable is not?
Sir, initially your short run model (ECM) had some variables in differenced form but at 5:10 minute in the video you are writing short term model with all variables in levels. Could you please explain what happened there?
Hi, the coefficient values are from the short run model. I also wrote “short run model” , so we know we are talking about the short run model. Seems I missed to type the differential symbol (triangle) in the variables. Regards
Hello Everyone! Thanks for watching! Part 2 of the tutorial is here! I hope you enjoy it.
IMPORTANT: To clarify, if your Error Correction Term is positive (instead of negative), it means your model will not converge to an equilibrium. The disequilibrium will be permanent.
✅ You can buy for a small amount the EViews Wokfile complete explained step by step + video slides + dataset at : payhip.com/b/x9N7v
✅ Visit my website for more information about the topic covered in the video:
www.jdeconomics.com/cointegration-and-error-correction-model/
✅Feel free to subscribe for more content!
ruclips.net/channel/UC5P21WGFO4WRUlAiGLcwymg
✅Ensure to watch the 1st Video: "Cointegration - Engle and Granger method in EViews".
Link: ruclips.net/video/4DBXBLIOHGE/видео.html
I wish you good luck on your research and courses!
JDEconomics.
I went through over 15 videos on RUclips and read around 20 articles and many other academic resources about this subject, I can confidently say that this video is a life saveeer and much better than all the unnecessary materials I went through .
Thanks a lot for taking the time to provide such a kind feedback! I am glad I was able to help you understand the topic. Feel free to check my website and share my content with your close ones! Good luck with your courses! JD
This is a master class. I am looking forward to you doing a video on the ARDL. Am sure it will be superb like the rest of your videos
Many thanks! Will do! Please subscribe to the channel!
Best, JD
@@JDEconomics I am a subscriber already. Thanks for your content
I love you man.. 😅😅
Thanks!!
thank you so much this clip was great and so helpful for students like me who want to handle their thesis!😀
That’s great to hear! Thanks! Please feel free to subscribe to my channel and share it with your friends! Best, JD
Sir. I'm very happy with your teaching.
Thanks! Best Regards, JD
i appreciated again ,but explain forcasting model in technical and theortical aspects in great detail.
Excellent explanation, very well done! Hope will have more videos on econometric models, with more than two variables and in R software.
Sure! Why not? Have a good one!
Hello JD! This is awesome. Just one question, is there supposed to be a constant in a differenced regression? I would have thought that since constants are the same across periods, they would be deleted out.
Thank You Sir. The video is very informative . Can you please make a video explaining Johannsen Co-integration test and VECM
Hi! Yes. I have to do a tutorial on that. Its in my list. Regards, JD
Hi Sir, thank you so much for your constructive tutorial. Can you make a video for cointegration bound test applying the ARDL model in Error Correction form showing the estimation for both the long run and Short run error correction model using E-views. Thanks
very good explanation sir. is it possible to run the ECT when all variables are I(0)?
Thanks! Doing so would have no real sense as there are no corrections in the residuals. I hope that helps! Make sure to check my website: juandamico.start.page and feel free to share my channel with your friends! I wish you good luck! JD
Hi! Great video! I´ve got a question, though. What if I get a statistical significant relationship in the long run but not in the short run (ETC is statistical significant and the values are as expected, but for example imports shows to be only significant in the long run not in the short run)
Thanks. I replied to your email. Good luck!
I am studying price dynamics between retail and farmgate prices. What if the sign of the long run and short run adjustments are different? Also, what if the short run adjustment is not significant but the ect is negative and significant? I hope I can show you my data and results cause I badly need help on this one.
As a tentative to answer your very important question, you may want to re estimate your short run model after watching closely heteroskedasticity and serial correction. Which will give you consistent non unbiased estimates. With these estimated obtained in OLS, you can run an ec approach to which will provide results worth advising public policies
Excelent. Thant you for this video.
Thanks! Please check my website to see all my tutorials. Www.jdeconomics.com
Good luck!
Hello I am just wondering about the variables we just take it in difference or we should determine how many lages we should use? Tankes
Hey! Short run we use one difference. Regards!
Great work and i appreciated you alot, however, plz, further explain my argument in detail: like what would be the intercept value ?i.e, alpha = 0, alpha > 0 or alpha .01, >.05 and >.10. so, if this is the case, then it means, that the model is stable and valid.explain? example CAPM, FF3FM, FF4FM, FF5FM and FF6FM.
How is the written short-term equation in case of more than 2 variables, should the number of ECT estimated coefficients be the same as the number of analyzed variables in the equation?
HI JD, COULD YOU PLEASE ALSO MAKE A VIDEO ABOUT VECM
Yes! That’s still in my plans. I will hopefully get into vec in the coming week or two. Regards! JD
I want to know that how you feed data in specific software? And which software have you used? i m studying on it. Still I have many confusions!!!
Hi, I am using Eviews. You can check my website and see in eviews the first tutorial where I show how to download economic data and import it to the software. Good luck! JD
sites.google.com/view/jdeconomics-
Sir I am confused about when ECM model got estimated then how to forecast for the specific dependent variable while ECM independent term residuals are not available for the forecast period.
Hello sir, I hope you will answer my question. I found error correction form negative sign (-1.07) and statistical significant. Is there any problem with that? Because it’s not in the range of 0 and -1. There are many articles that similar to my finding.
Hey. If it’s a linear regression, it should he between -1 and 0. If it is more negative, then the results are explosive. Otherwise, you could say the adjustments are immediate. In other words, takes no time to adjust a disequilibrium. But it’s not really a good analysis. Cherrs
It will be usefull if you number the videos (reading sequence)
Hi, thanks for your message. They are in order. If you play the playlist, or check the list, they are all ordered. Unfortunately I don't put a number because if I add a new video, then I have to change the numbers all the time. The playlist is here: ruclips.net/p/PLsZ8kVwX52ZGXxJ-bqP-WmRBD2HytO4d4
Regards,
JD
Thank you for your explanation video, sir. It's really helpful for my thesis. But, I have little troubles here.
1) What can I do, if the model doesn't fulfill normality assumption? And how can I fix this?
2) the ECT value is positive instead of negative, does it mean I have to re-estimate my model?
Thank you very much for your answer and help.
Hello, Thanks for your message. If your residuals aren't normal is not the end of the game. Don't worry. Autocorrelation is more important. There can be some part in your residuals that are spiky, and maybe a dummy variable can help for that particular period.
In terms of the error correction term, as you may have seen in the explanation I provided, the negative sign allows the model to correct. Two negative signs will make it positive (picture a trend going very very down, the negative sign will correct it and make it positive) and other way: (very positive (+) a negative sign will make it negative (-) and go down). That's the intuition. Remember the term is allowing the discrepancies between the short and long run to be corrected. Regards, JD
@@JDEconomics
Good morning Sir,
Thank you very much for your videos. I was just asking myself this question (my model doesn’t fulfill normality assumption) because I had the same problem so I thank you for this answer. All the other conditions are met ( no serial corrélation and no hétérostacity) and my two séries are cointegrated. I was just wondering if it is a problem that I have only between 25 and 30 annual data?
Thanks in advance
Hello sir please what can cause the THRESHOLD technic to disappear from the available technique in e-vews if one want to run a data using the threshold analysis for a number of selected countries in the ecowas region.
Hey how can we calculate short run ECM through ARDL model
Hey JD, can you use this model to hedge ratios for futures contracts ?
hello i just wanna ask again, will the result differ if i exchange the dependent and independent variables? what does it mean if it shows that it is cointegrated with significant ect when variable a is used as dependent variable but the results differ when i used variable b as the dependent variable?
Sir, what about the issue in the diagnosis check of serial correlation and normality? how to handle it?
hello. i have a question. what if both engle and granger and phillips ouilaris shows that the residuals are stationary which makes my variables cointegrated. however when i generate the short run model the ect shows that it is not significant?
Can happen. As an economist, you have to analyze the results and see if they are sensible. Good luck!
@@JDEconomics what if my ect is finally negative and significant but the other variables like the constant and lagged difference of the other variable is not?
Sir, initially your short run model (ECM) had some variables in differenced form but at 5:10 minute in the video you are writing short term model with all variables in levels. Could you please explain what happened there?
Hi, the coefficient values are from the short run model. I also wrote “short run model” , so we know we are talking about the short run model. Seems I missed to type the differential symbol (triangle) in the variables. Regards
@@JDEconomics ok sir. Thank You for responding.
Data set please
I still want to learn
Hi, wha would you like to learn? Regards, JD