Unit Root Testing. Model One. Part 2 of 2. EVIEWS

Поділитися
Вставка
  • Опубліковано 10 січ 2025

КОМЕНТАРІ • 87

  • @kabinehkpukumu6182
    @kabinehkpukumu6182 5 років тому

    Fantastic video sir! God bless you

  • @marwanassar5028
    @marwanassar5028 11 років тому +1

    really thank you sir, you have helped me alot

  • @vaishalidhingra279
    @vaishalidhingra279 12 років тому

    That's fine sir, but i read from damodar gujarati that if trend is present then we have to regress it against trend variable for removing trend. and i did the same thing and solved the problem. Thank you so much for your reply.

  • @sayedhossain23
    @sayedhossain23  12 років тому

    Sometime it is difficult to make the variable stationary. In that case, I would advise, either convert variables into log or increase the data size and try. In future I have plan to upload video on ARMA. but normally we insert AR(1) to capture serial correlation problem

  • @elninou3174
    @elninou3174 10 років тому

    Fantastic video dear Hussein. Thank you.

    • @sayedhossain23
      @sayedhossain23  10 років тому

      Marc Jack You are welcome Jack. Thank you Sayed Hossain from Hossain Academy at www.sayedhossain.com

  • @michelangelolandgraf
    @michelangelolandgraf 9 років тому

    Thank you for the video. It was very enlightening.

    • @sayedhossain23
      @sayedhossain23  9 років тому

      michelangelolandgraf You are very much welcome. Thank you Sayed Hossain from Hossain Academy at www.sayedhossain.com

  • @mrdubbledee6227
    @mrdubbledee6227 6 років тому +1

    Thanks a lot sayed Hossain

    • @sayedhossain23
      @sayedhossain23  5 років тому

      Thank you. I would like to invite you to join Hossain Academy Facebook Group at below link and join our group discussion. Thank you. Sayed Hossain from Hossain Academy. facebook.com/groups/hossainacademy/

  • @sayedhossain23
    @sayedhossain23  12 років тому

    There are three ADF equations to test whether a particular time series data is stationary or not. All these three must give the same decision to come to a decision whether a particular time series is stationary or not.

  • @sayedhossain23
    @sayedhossain23  11 років тому

    You see all my videos properly. I guess you will understand finally and also read literature on it.

  • @0451452
    @0451452 12 років тому

    Hi Sayed, do you know if it is possible to carry out the Dickey Pantula test for multiplie unit roots on eviews?

  • @vaishalidhingra279
    @vaishalidhingra279 12 років тому +1

    Dear Sir,
    I saw your videos for unit root test but I'm not getting the video which shows how to detrend the series if trend is present. Please revert back as soon as possible.

  • @JML335
    @JML335 12 років тому

    When I did "level" and "intercept" unit root was present but the coefficient was positive. For every other combination the coefficient was negative. Does it matter that we couldn't accept the result for level and intercept, as the P and T values showed it to have unit roots anyway?

  • @NMAUTUBE
    @NMAUTUBE 13 років тому

    Sir, In Eviews manual, there is no metion of obsoulte value The manual says " Notice here that the t statistic value is greater than the critical values so that we do not reject the null at conventional test sizes. " the t value shown in manual is -1.41 which is mentioned as GREATER than at 1% level -3.45 Are you correct or is the manual correct? Pl clarify

  • @ashbritish7331
    @ashbritish7331 12 років тому

    Dear sayed,
    can you post anything to show us how to use DOLS on a money demand function?
    Thank you....

  • @DhavalSaifaleeAaryash
    @DhavalSaifaleeAaryash 12 років тому

    Sir, In one of my results on GDP, I am trying to find out stationarity. ADF Statistics = 5.4255 and critical values at 1%, 5% and 10% are -3.83, -3.02, -2.65 respectively and p-value is 1.00. Please tell me how to interpret such result. Thank You.

  • @yasinhared1283
    @yasinhared1283 3 роки тому

    Hi sir.
    Its must to use intercept, trend and none, or i can use only one that i get stationary by using level first and second difference

  • @MIbtehajKhan
    @MIbtehajKhan 5 років тому

    In my model, I have 5 variables, 3 of them are stationary at level but the other two are stationary at first difference. So is it okay or I have to make them stationary at level? If YES, then please tell me the technique.

  • @JML335
    @JML335 12 років тому

    What do you do if the coefficient doesn't have a negative sign?

  • @MrJoshdur
    @MrJoshdur 12 років тому

    Hi, the question is what do i include my model? none?, drift?, or time-trend and drift?..

  • @abdixaliim
    @abdixaliim 11 років тому

    thank you Mr sayed, but still im confused, how to make cointegration and the propose of it

  • @lazregsaid3830
    @lazregsaid3830 12 років тому

    what is the best criterion to choose the right model for example the first différence with
    1: none
    2: intercept
    3: trend and intercept
    thank you in advance

  • @vaishalidhingra279
    @vaishalidhingra279 12 років тому

    Sir, for one of my series i went for adf test, in which the p- value for test statistics is 0.0110 and the trend co-efficient is significant as the p- value is 0.0082.
    This is the result of E-view
    Augmented Dickey-Fuller test statistic -3.934060 0.0110
    @TREND(1) 0.213549 0.0082
    Should I consider this series stationary? If yes then there is presence of deterministic tred, isn't it? N if yes then how do i make the series trend stationary?

  • @sayedhossain23
    @sayedhossain23  12 років тому

    You are welcome

  • @choimaabolor3527
    @choimaabolor3527 7 років тому

    Hello, what if ADF results are not consistent across 3 models(intercept, intercept and trend, none)? One of my variables is exhibiting significance at intercept and has unit root on other 2. Can i take it as no unit root or should i go to 2nd difference?

    • @sayedhossain23
      @sayedhossain23  7 років тому

      Dear Bolor, Thank you. I would like to invite you to join Hossain Academy Facebook at below link and post your question there. Actually I am in that group and may help you. Thank you once again, Sayed Hossain from Hossain Academy. facebook.com/groups/hossainacademy

  • @sayedhossain23
    @sayedhossain23  12 років тому

    Yes.

  • @sayedhossain23
    @sayedhossain23  12 років тому

    Better test all

  • @sayedhossain23
    @sayedhossain23  12 років тому

    May be but I have never used Dickey Pantula Test

  • @aditkasim
    @aditkasim 9 років тому

    nice video! i have a question.. if the data is stationary at 1st difference with "none" and "intercept", but when include "trend and intercept" it becomes not stationary. so, can we call it stationary or should we continue checking it at 2nd difference?
    thanks before.

    • @sayedhossain23
      @sayedhossain23  9 років тому

      Adiatma Kasim All three must come to same decision to come to a single decision.

    • @aditkasim
      @aditkasim 9 років тому

      but my proff said that the data would not be that reliable when it comes to the 2nd difference. so that the result not so accurate. what do you think sir?
      anyway, thanks so much for the reply and keep up the good work!

  • @firdausibibinu2821
    @firdausibibinu2821 11 років тому

    what if the result only shows two give the same result while the othe gives a different result. what does that imply?

  • @hoseahandan1538
    @hoseahandan1538 9 років тому

    God bless you sir, I most confess to you that videos on running regression has been great. But sir, my question is, what would you do when you discover that a variable is not stationary?

    • @sayedhossain23
      @sayedhossain23  9 років тому

      hosea handan It can always happen but you can solve it

    • @sayedhossain23
      @sayedhossain23  9 років тому

      Sayed Hossain I would like to invite you to join Hossain Academy Facebook for greater interaction about economics, finance and econometrics. Thank you Sayed Hossain from Hossain Academyfacebook.com/groups/hossainacademy/

    • @hoseahandan1538
      @hoseahandan1538 9 років тому

      Cool

  • @renn9236
    @renn9236 9 років тому

    Hi Sir, may i knw..did u have the video show the 2nd different model?

    • @sayedhossain23
      @sayedhossain23  9 років тому

      puong renn Lau No I do not have but the procedure is same like first difference.

  • @sayedhossain23
    @sayedhossain23  12 років тому

    It would be better if all three give the same result. Try to get same result from three.

  • @nicob744
    @nicob744 11 років тому

    Dear Sayed,
    You're absolutely helping me out with your videos, i've seen quite some already!
    Just one question: you happen to use only one variable for unit root testing. I would like to test it for multiple variables, (4,5,6,) but it does not work the same. Should I test it for all variables individually or as a group?
    Thank you

    • @sayedhossain23
      @sayedhossain23  11 років тому

      You have to use each and every variable individually..

    • @nicob744
      @nicob744 11 років тому

      Sayed Hossain Alright thank you, and then what should one do if not all three (intercept, trend & intercept, none) follow the same line? If one (none) is significant while the others are not? Convert to first difference?

    • @sayedhossain23
      @sayedhossain23  11 років тому

      If all three give the same decision only then accept it. Otherwise do not take any decision.

    • @nicob744
      @nicob744 11 років тому

      Sayed Hossain But what should I do then? If I cant take a decision? In the next step (Johansen test) I need to have all variables in first difference anyway..

    • @sayedhossain23
      @sayedhossain23  11 років тому

      In that case, you run only VAR, not VECM model but make the data stationary before running it.

  • @AdilKhashtamov
    @AdilKhashtamov 12 років тому

    @NMAUTUBE the idea is to use absolute value of t-statistic

  • @flower-py4jw
    @flower-py4jw 11 років тому

    thanks!!!!your video have help me a lot!
    but i still have 1 question.
    why the coefficient value of the variable must be negative sign ?
    if I get a positive sign?what its mean?

    • @sayedhossain23
      @sayedhossain23  11 років тому

      It should be negative but if not, meaning that there is no long run causality...

    • @flower-py4jw
      @flower-py4jw 11 років тому

      Prob*:1.000 (do not reject H0)
      Augmented Dickey-Fuller test statistic:-4.852017
      Test critical values:
      1% level -3.482035
      5% level -2.884109
      10% level -2.578884 (more than critical value=reject H0)
      GE(-1): 0.082220 (positive sign)
      this is the result i get, i should conclude it have unit root or not ?

    • @sayedhossain23
      @sayedhossain23  11 років тому

      I have never seen it to be 1...There is problem somewhere either in data or data setting...

    • @sayedhossain23
      @sayedhossain23  11 років тому

      Test statistics is more than critical value...you can reject null hypothesis...what about probablity value?

    • @flower-py4jw
      @flower-py4jw 11 років тому

      you mean the prob*?
      but after i proceed to 1st differencing test,
      prob*=0.0000
      ADF test statistic more than critical value
      coefficient is negative sign.
      so can i conclude that the variable is stationary?

  • @zulqarnainmushtaq9972
    @zulqarnainmushtaq9972 6 років тому

    ASALAM ul KUM Sir; I had taken log of all five variables and now all five variables are stationary at second difference (NONE). Is it possible to proceed to next test with variables stationary at second difference or need to transform the variables to make them stationary at first difference then can proceed to next test? I'll be using ARDL model to test the long run relationship then to VECM followed by Granger Causality test. I am really struck jazakallah if you can help me to proceed.

    • @sayedhossain23
      @sayedhossain23  6 років тому

      Thank you. I would like to invite you to join Hossain Academy Facebook Group at below link and join our group discussion. Thank you. Sayed Hossain from Hossain Academy.

    • @jasveerkaur4182
      @jasveerkaur4182 2 роки тому

      I am also facing same situation..can I use ardl if some variables stationary at 2nd difference????
      Pls reply ASAP

  • @weeminalbert
    @weeminalbert 8 років тому

    hello, sir
    what if the variable is stationary with tested on level form, do we need to test it on 1st difference?

  • @mkjoshi21
    @mkjoshi21 6 років тому

    Sir, If the series is stationary at I(1) and also stationary at I(2) then what is the final conclusion? Is it I(1) or I(2)?

    • @sayedhossain23
      @sayedhossain23  6 років тому

      Thank you. I would like to invite you to join Hossain Academy Facebook Group at below link and join our group discussion. Thank you. Sayed Hossain from Hossain Academy.

    • @sayedhossain23
      @sayedhossain23  6 років тому

      Thank you. I would like to invite you to join Hossain Academy Facebook Group at below link and join our group discussion. Thank you. Sayed Hossain from Hossain Academy. facebook.com/groups/hossainacademy/

  • @nathankigosa1640
    @nathankigosa1640 11 років тому

    Hi Sir, ADF test shows that the variable is nonstationary at level but PP shows that it is stationary. Of course both ADF and PP show that the variable is stationary after first difference. Which decision should be taken?
    Can running the correlogam and the LB test help to clear this uncertainty?

    • @sayedhossain23
      @sayedhossain23  11 років тому

      You can choose any one but if both gives the same decision, that is better always.

    • @nathankigosa1640
      @nathankigosa1640 11 років тому

      Sir, is it a common scenario of having ADF and PP showing different results and can running a correlogram and Q statistics be of any help?

    • @sayedhossain23
      @sayedhossain23  11 років тому

      Yes ...it is common scenario results may vary test to test...

    • @nathankigosa1640
      @nathankigosa1640 11 років тому

      Thank you Mr. Hossain.

  • @sayedhossain23
    @sayedhossain23  12 років тому

    Plan in future

  • @sayedhossain23
    @sayedhossain23  11 років тому

    Normally when three equation gives the same result, we accept it.

  • @yoshuanugraha
    @yoshuanugraha 9 років тому

    Mr Sayed, what is the meaning of p-value equal to 0.0000 after 1st difference?

    • @sayedhossain23
      @sayedhossain23  9 років тому

      yoshua nugraha it means that you can reject null hypothesis. What is your null?

    • @sayedhossain23
      @sayedhossain23  9 років тому

      yoshua nugraha Dear Youshua, Please join Hossain Academy Facebook below for greater interaction with me regarding data analysis. Thank you Sayed Hossain from Hossain Academy
      facebook.com/groups/hossainacademy/

  • @vaishalidhingra279
    @vaishalidhingra279 12 років тому

    Sir, I'm a beginner in time series analysis. So, if you find my question silly, I am sorry for that. But please guide me properly.

  • @sayedhossain23
    @sayedhossain23  12 років тому

    Out of three options, at least two should be negative

  • @sabihabbasrazvi
    @sabihabbasrazvi 9 років тому

    Sir, what if my probability (p-value) for all three conditions is 0.00000?

    • @sayedhossain23
      @sayedhossain23  9 років тому

      +dada It means that reject null hypothesis. What is your null hypothesis?

  • @sayedhossain23
    @sayedhossain23  12 років тому

    In that case we cannot accept the result

  • @sayedhossain23
    @sayedhossain23  12 років тому

    I am not understanding your question

  • @sayedhossain23
    @sayedhossain23  12 років тому

    You are welcome

  • @sayedhossain23
    @sayedhossain23  12 років тому

    Better test all

  • @sayedhossain23
    @sayedhossain23  11 років тому

    You are welcome