VAR. Model Two. Part 2 of 2. EVIEWS

แชร์
ฝัง
  • เผยแพร่เมื่อ 9 พ.ย. 2024

ความคิดเห็น • 25

  • @daiane_2310
    @daiane_2310 2 ปีที่แล้ว

    Always the best! I always learn too much with your videos! God bless your life!

  • @sayedhossain23
    @sayedhossain23  11 ปีที่แล้ว

    Both are right. One variable may be significant individually but may not be significant jointly when other variables comes it.

  • @leptonhadron9912
    @leptonhadron9912 7 ปีที่แล้ว

    Hello...
    I really appreciate your attempts for making this wonderful videos. some thing that I found in your videos was "simplicity" which was really salient among being erudite and professional and this factor has made your teaching unique. please keep doing it.
    thanks a million

    • @sayedhossain23
      @sayedhossain23  7 ปีที่แล้ว

      Thank you. I would like to invite you to join Hossain Academy Facebook at below link and post your question there for feedback. Thank you, Sayed Hossain from Hossain Academy
      facebook.com/groups/hossainacademy/

  • @sayedhossain23
    @sayedhossain23  11 ปีที่แล้ว

    As the variables are stationary at level, in that case you can use VAR (not VECM model) and can get short run coefficients and also can use Wald Test for short run causality.

  • @alfredoxR1
    @alfredoxR1 11 ปีที่แล้ว

    If the p value says to me that the variables are not significant ( individually) but the wald test tells me that they are ( jointly) which criteria should i accept?

  • @sayedhossain23
    @sayedhossain23  11 ปีที่แล้ว

    Only journal or literature related to this study can tell you whether they have relationship or not. If they are stationary at level, I guess you can not run Johansen Test and VECM, but you can run Granger Causlity Pairwise test as the variables are in stationary at level form.

  • @christychan9113
    @christychan9113 6 ปีที่แล้ว

    your video is amazing. Thank you.

  • @sayedhossain23
    @sayedhossain23  11 ปีที่แล้ว

    Now you have to do trial and error. Either increase sample size, increase variables or drop irrevalent variables or run the model without log and so on. Indeed VAR and regression model estimation will provide u same result.

  • @sayedhossain23
    @sayedhossain23  11 ปีที่แล้ว

    You have to drop those variables which are highly correlated. Correlation coefficient can help you to find out those variables need to be dropped.

  • @sayedhossain23
    @sayedhossain23  11 ปีที่แล้ว

    The concept is close to F statistics.

  • @swarnalakshmi5310
    @swarnalakshmi5310 11 ปีที่แล้ว

    My variables are futures and options daily turnover and securities lending and borrowing daily turnover. I wish to know whether these two have relationship or not? They are stationary at level itself.which method should I use?

  • @melissacagle3707
    @melissacagle3707 4 ปีที่แล้ว

    Hi is this the Konya (2006) method?

  • @swarnalakshmi5310
    @swarnalakshmi5310 11 ปีที่แล้ว

    sir, now i have taken two other variables. FII's and stock market daily prices. One of my variables are stationary at level and the other one is non stationary and is becoming stationary after first differencing. In that case which model should I use and how? VAR or VECM?

  • @EllaWess
    @EllaWess 9 ปีที่แล้ว

    Hi!
    Thanks for a great video! :)
    Can you explain about the F-statistic and related p-value in the diagnostics of models? Why is it a bad sign when F-statistics is not significant? P-value is greater than 0.05, so it must be positive for the model. Is there anything with multicollinearity in these models to do?

    • @sayedhossain23
      @sayedhossain23  9 ปีที่แล้ว +1

      +Ella Wess
      When F statistics is significant it means that all the independnet variables jointly can explain dependent variable and it is a good sign

  • @swarnalakshmi5310
    @swarnalakshmi5310 11 ปีที่แล้ว

    Thank you for your reply sir. one more question... I run the VAR model after taking log of my variables. And then I did the wald test and got answers. But when I check the equations in the VAR model separately for diagnostic test, the ARCH LM test reveals that the residuals are having heteroskedasticity even after log transformation. What should I do to remove it? Thank you in advance again for your reply

  • @sidraaslam1663
    @sidraaslam1663 9 ปีที่แล้ว

    sir thank you so much for such a valuable lectures.i am very happy to see the videos because they taught me alot.i want to learn structural var model.kindly teach it or if you have already done then send me its link?plzz

  • @sayedhossain23
    @sayedhossain23  11 ปีที่แล้ว

    Convert all variables into stationary and use the stationary variables to run VAR model.

  • @dr.najibkhan7876
    @dr.najibkhan7876 10 ปีที่แล้ว

    Thank you very much for this incredibly informative video. One suggestion, if I may, you repeat same words several times, and this takes away a lot of your time. If you don't repeat them, you can cut the video time in half easily. Thanks anyway.

    • @sayedhossain23
      @sayedhossain23  10 ปีที่แล้ว

      Thank you for your suggestion. I shall think about it seriously.

  • @dukelambi2454
    @dukelambi2454 11 ปีที่แล้ว

    Hi master, I have one question: If I'm doing Wald test for all coefficient for exemple c(1)=c(2)=c(3)=c(4)=c(5)=c(6)=c(7)=0 is this test is fairly the same with a F-stat ?
    Thanks for your reponse

  • @sayedhossain23
    @sayedhossain23  11 ปีที่แล้ว

    So far I know u can not run Johansen Cointegration Test

  • @swarnalakshmi5310
    @swarnalakshmi5310 11 ปีที่แล้ว

    My variables are stationary at level itself. Can I run cointegration test?

  • @Nagma2908
    @Nagma2908 11 ปีที่แล้ว

    so , how can avoid the multticolinearity