Announcement

Collapse
No announcement yet.
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • repeated time values in sample problem

    Hello, I have panel data on cryptocurrencies.
    Uniquely identified by coin's rank and date. I used tsset ranknow date. Panel data is unbalanced and time variable is with gaps.
    I tried basic VAR regressing price and it says repeated time values in sample.
    Any idea where is the problem? I have data uniquely identified by coin's rank and date.
    Thank you for your help.

  • #2
    EDIT: I think I chose a wrong model, no need to reply, thanks

    Comment


    • #3
      EDIT: My model was right, there must be something else. Please help

      Comment


      • #4
        I used tsset ranknow date. Panel data is unbalanced and time variable is with gaps.
        I tried basic VAR regressing price and it says repeated time values in sample.
        Any idea where is the problem?
        The problem is that you are using a time-series command on panel data. Unless you want to analyze each time-series separately, whatever you want to achieve will not be possible with this command.

        Comment


        • #5
          Thank you very much. What model would you recommend if I am interested in the effect of certain event on cryptocurrency price?

          Comment


          • #6
            If you have a wide panel (N > T), the standard panel data models such as fixed effects and random effects work well.

            Comment


            • #7
              I was thinking linear regression with AR(1) disturbance .

              Comment


              • #8
                With linear regression, you will be ignoring the panel structure of your data. Even if your residuals are AR(1), there are panel data methods that accommodate this. Specifically, for unbalanced panels, you should take a look at Baltagi and Wu (1999) - Unequally spaced panel data with AR(1) disturbances.

                Comment


                • #9
                  thank you very much

                  Comment

                  Working...
                  X