Announcement

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

  • Reporting Newey-West autocorrelation corrected t-statstics after test of differences in mean variables (ttest)

    Hello everyone,

    I want to test whether two mean variables are significant different from each other. In general, I would do a simple ttest. However, as I am conducting an asset pricing model and considering cross-sectional variation, I need/want to report Newey-West autocorrelation corrected t-statstics for this ttest. My question is hence: how can I combine a simple ttest and Newey-West t-statistics?
    I there such a thing as as a Newey Ttest?

    Thank you so much in advance for your help!

  • #2
    Lisa:
    welcome to this forum.
    Short answer: no, there's no such a thing as a -Newey-West t-test-.
    But you can test the hypotehesis you're interested via -newey-, though.
    Kind regards,
    Carlo
    (Stata 19.0)

    Comment


    • #3
      Thanks for your answer!
      My problem is, that I dont unterstand how I can use the newey command when I am not doing a regression (reg / newey depvar indepvar) ?
      Do you have an idea?

      Comment


      • #4
        in the -newey- command, depvar will be your quantitative variable and indepvar will be your group variable; try the following:
        Code:
        sysuse auto, clear
        ttest gear, by(foreign)
        regress gear i.foreign
        and, by comparing the results, you will see that they are exactly the same

        Comment


        • #5
          Hi, I want to check if a hedge position, one column of numbers is significantly different from 0. However, as the numbers are derived through overlapping periods, I would need to correct for newey and west. Could anyone please tell me what command to run, I tried ttest==0,
          newey variable, lag(#) but it is invalid. Please help me. How do I correct for these numbers through newey and west and then tests if they are significantly different from 0 using stata

          Comment


          • #6
            Hi, I would like to raise the above topic since I am interested in similar. In all articles, they are referring to t-statistics of differences in average returns using the Newey-West test, so I am interested in command in STATA (it is not regression, for which code is clear). Many thanks in advance.

            Comment


            • #7
              Zrinka:
              not my research field, but I'm under the impression that the t-statistics were obtained from Newey-West regression outcome.
              That said, you may want to take a look at https://stats.stackexchange.com/ques...t-t-statistics
              Kind regards,
              Carlo
              (Stata 19.0)

              Comment


              • #8
                meldoy:
                hi, if you want to get the newey-west t-statistics for the means of series X, you can try this:
                gen t = _n
                tsset t
                newey X, lag(6) // you can set the lags according to the characteristics of your data

                Comment


                • #9
                  Originally posted by Zrinka Lovretin Golubic View Post
                  Hi, I would like to raise the above topic since I am interested in similar. In all articles, they are referring to t-statistics of differences in average returns using the Newey-West test, so I am interested in command in STATA (it is not regression, for which code is clear). Many thanks in advance.
                  hi, if you want to get the newey-west t-statistics for the means of series X, you can try this:
                  gen t = _n
                  tsset t
                  newey X, lag(6) // you can set the lags according to the characteristics of your data

                  Comment


                  • #10
                    Thank you Wuqi Song, I try your syntax but it doesn't work because my data include repeated values for date and stock_id. Do you have any suggestions? Many thanks.

                    Comment

                    Working...
                    X