Error correction model - part 1

In this video I introduce the concept of an Error Correction Model, and explain its importance in econometrics.
Check out www.oxbridge-tutor.co.uk/under... for course materials, and information regarding updates on each of the courses. Check out ben-lambert.com/econometrics-... for course materials, and information regarding updates on each of the courses. Quite excitingly (for me at least), I am about to publish a whole series of new videos on Bayesian statistics on youtube. See here for information: ben-lambert.com/bayesian/ Accompanying this series, there will be a book: www.amazon.co.uk/gp/product/1...

[@repuglia]

You're a real life saver, and you explain a lot better than my current teacher. Keep up the good work, cheers

[@1nSense]

Ben the Oxford undergraduates taking the Econometrics FHS exam on Saturday salute you!

[@SpartacanUsuals]

Ha! Good luck to you all. Enjoy the Turf/KA/Lamb-and-flag/Bear/Cape afterwards. (By “/“ here, I mean “and” not “or”.) If you are in a particularly felicitous mood afterwards, then ask your college library to buy my Bayesian book. Cheers, Ben

[@ebateru]

Dude, Im falling in love with you! thanks a bunch!

[@shoaibkahut]

Greetings from *China* sir ! YOU are the real LIFE saver . Thank you

[@TheDiegoAdrenalina]

Thank you so much for this helpful video! Really appreciate your work.

[@mihaililiev5932]

Splendid! I finally get to understand the error correction model!

[@Mrnmc4911]

This is so good. Very clear explanation.

[@perrygogas]

Great Ben I am gonna send this to my students.

[@szpacur]

Absolutely brilliant job buddy

[@mauroacunacaliso2358]

Fantastic video, very good explained, congrats for your work!

[@igorstanisic5380]

Ben, both videos (part 1 and 2) explain very well the EC model. Thank you for sharing it. I would like to run Engle-Granger methodology in R. Could you suggest any material which shows how Engle-Granger methodology is applied on data in R? Thank you for any assistance.

[@krabbypatty6896]

Hi, Ben. Thanks so much.

[@tedlu2502]

Thanks, mate. You are a legend.

[@tolulopeobele9886]

Interesting, very well understood.

[@momcilomracajac2242]

THE GREATEST!

[@l8524]

1:11 , I think an important correction is that it's not simply more powerful to run ECM, but actually running the VAR in first differences when there is cointegration is wrong: there does not exist a VAR specification in differences that allows to achieve the structual shocks of the WOLD, since the matrix of the Wold will not be invertible. VAR exist only in levels but not in first differences under cointegration.

[@leventis1991]

Greetings dear Ben, I was wandering if you could answer me the following two questions first of all i would like you to clearify where do i use the VECM equation...i still haven't figure out where it should be used (for prediction or analysis) and what to we mean when say "short run" what is considered short run to an estimation and secondly which model is better suited for a prediction the normal var or the VECM? Thank you in advance G.K.

[@olivier306]

Beyond grateful still

[@Internal0str]

you legend mate. thanks a lot

[@syedtabrez]

Ben, I have seen that after generating Long Run and Short Run equations, the error correction coefficient is multiplied with the Long Run coefficients and then deduced a final regression equation based on new Long Run and Short Run, can you explain this logic?

[@Azam_Pakistan]

Please add a video on ARDL model

[@user-ov1to6cs7i]

thank you very much, you are very great

[@RayDV88]

what is alpha supposed to be? I Can't see why you add ir it and why did you change the constant?

[@OgaMariOga]

Thank you so much, your videos are extremely helpful!

[@quinnschelske4903]

Thank you!

[@gulzameenbaloch9339]

Thank you so much

[@hassanaber392]

Goooood teacher merci

[@jackiecomendador9428]

just wanted to say thank you for all your videos and being one of the reasons why I'll graduate college

[@lastua8562]

You do this in college? I thought this is university only stuff...

[@jackiecomendador9428]

@@lastua8562 college...like university...where i came from, that's interchangeable college/university = tertiary education. But I did study in a University. Also, this stuff is done no matter if you're in a "college" or uni -- it's more on your degree program! We also had this in high school actually but not discussed in depth (hope this clarifies the confusion!)

[@lastua8562]

@@jackiecomendador9428 Then that is a very good high school!

[@endrity]

Silly question but what kind of software do you use to produce these videos? Great work btw.

[@lastua8562]

check his website

[@khanhlinhpham558]

Why does (S1+S2) equal to (1- mu) at 6:42s? please!

[@DuyNguyen-lo8xx]

No, it's not equal. If (S1+S2) = (1-mu) then Beta would be fixed to 1 and make no sense. Beta is obtained by regressing Yt and Xt.

[@samiasamssoume4752]

useful video! One question; I do not understand where come 1-m parameter for y (t-1). thanks a lot.

[@mihaililiev5932]

because Ben takes away y(t-1) from both sides. (1-m)*y(t-1) = y(t-1) - m*y(t-1)

[@samiasamssoume4752]

@@mihaililiev5932 Thanks a lot.

[@mohammadpoltaksimbolon8019]

What do the I(1) and I(0) mean?

[@lastua8562]

order of integration. Check the I in ARIMA

[@3000MEGRA]

Can you show me the derivation of the alpha?

[@juliusvinson5004]

yes, am wondering too where this alpha comes from

[@henrikhansen123]

​@@juliusvinson5004 I think he basically defines c' such that c' + (1-mu)alpha = c => alpha = (c - c')/(1-mu). So, he basically split the constant into two parts: LR and SR. In this case, the SR-part and the LR-part of the ECM both includes a constant. This makes it easier to interpret the model. Alternatively, alpha can be defined to be alpha = c/(1-mu) , but then the SR-part doesn't have a constant. The intuition can be found in his video "Lagged dependent variable ARMA".

[@Azam_Pakistan]

All well till step 4.After that couldn't follow. Please add a few more steps to clarify.

[@taylorhamlin6949]

hey does anyone know how to do this on stata