No standard deviations, t statistics or p values given..?

[Barbara Heerkens]

Hi,

I am looking at the cumulative abnormal returns of 78 firms in Malaysia.
I ran multiple regressions and now I want to perform one last one with only the financial firms included.
However, besides coefficients, no standard errors, t statistics or p values are showed...
Does anybody know what could have gone wrong?

. regress car conn top30 size leverage
note: top30 omitted because of collinearity

      Source |       SS           df       MS      Number of obs   =         4
-------------+----------------------------------   F(3, 0)         =         .
       Model |  .000747805         3  .000249268   Prob > F        =         .
    Residual |           0         0           .   R-squared       =    1.0000
-------------+----------------------------------   Adj R-squared   =         .
       Total |  .000747805         3  .000249268   Root MSE        =         0

------------------------------------------------------------------------------
         car |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
   connected |  -.0408793          .        .       .            .           .
       top30 |          0  (omitted)
        size |  -.1383905          .        .       .            .           .
    leverage |   .1919741          .        .       .            .           .
       _cons |   2.462861          .        .       .            .           .
------------------------------------------------------------------------------

Thank you!

[Rich Goldstein]

you have 4 observations and 4 independent variables - this means no SE's or anything that requires SE's - you need additional observations

[Carlo Lazzaro]

Barbara:
as Rich pointed out, your 4-observation regression is technically unfeasible.
You may want to consider creating a categorical predictor (1/0) for financial firms and include it among the independent variables of the 78 firms regression.