I am trying to estimate a panel regression measuring the effect of different events on different stocks. My data set contains a list of 1384 listed companies with daily data of different controls between jun 2018 and dec 2019. I already created a dummy variable which is 1 if the date is an event date and 0 otherwise. My concern now is that the effect might fully unfold only at a later day, so I would like to measure the effect one day, two and three day after the event.
My question would be how do I create dummies for those additional days to also capture that effect?
A secondary question would be as the days are trading days so some dates are missing can I just estimate a panel regression from a dataset containing all the dates with the event dummies or do I need to tell stata first that those are trading dates, or do I even need to include all the other dates, meaning could I just use the event dates instead and delete the other dates?
I would appreciate some guidance, thanks a lot
Here are some information about my data and the panel estimation I have done so far
Code:
obs: 546,285
vars: 13 16 Mar 2020 19:17
size: 41,517,660
---------------------------------------------------------------------------------------------------------------------------------------------------------
storage display value
variable name type format label variable label
---------------------------------------------------------------------------------------------------------------------------------------------------------
c_id int %10.0g c_id
day float %td
grp_id int %10.0g grp_id
size double %10.0g Total Asset Size
oroa double %10.0g Operating ROA
abr double %10.0g Daily Adjusted Abnormal Return
pb double %10.0g Price to Book
roa double %10.0g ROA
sa double %10.0g Sales to Total Assets
lev double %10.0g Total Debt to Total Assets
vol double %10.0g daily Stock Volatility
typeofevent str3 %9s Type of Event
dum_1 byte %8.0g typeofevent==EVTCode:
eststo:xtreg abr dum_1 size oroa pb roa sa lev vol, fe
Fixed-effects (within) regression Number of obs = 530,136
Group variable: c_id Number of groups = 1,379
R-sq: Obs per group:
within = 0.0001 min = 111
between = 0.0069 avg = 384.4
overall = 0.0001 max = 390
F(8,528749) = 7.52
corr(u_i, Xb) = -0.3115 Prob > F = 0.0000
------------------------------------------------------------------------------
abr | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
dum_1 | -.0003579 .000113 -3.17 0.002 -.0005794 -.0001364
size | -1.78e-10 5.27e-10 -0.34 0.736 -1.21e-09 8.55e-10
oroa | -.0000411 .0000117 -3.51 0.000 -.0000641 -.0000181
pb | -9.58e-08 6.85e-07 -0.14 0.889 -1.44e-06 1.25e-06
roa | .0000283 .000011 2.56 0.010 6.66e-06 .00005
sa | .0017127 .000429 3.99 0.000 .0008718 .0025536
lev | -7.32e-08 4.84e-06 -0.02 0.988 -9.56e-06 9.42e-06
vol | .0000124 2.20e-06 5.61 0.000 8.04e-06 .0000167
_cons | -.0005924 .0001779 -3.33 0.001 -.0009411 -.0002437
-------------+----------------------------------------------------------------
sigma_u | .00101958
sigma_e | .02081602
rho | .00239334 (fraction of variance due to u_i)
------------------------------------------------------------------------------
F test that all u_i=0: F(1378, 528749) = 0.74 Prob > F = 1.0000Code:
+-----------------------------------------------------------------------------------------------------------------------------------------+
| c_id day grp_id size oroa abr pb roa sa lev vol typeof~t dum_1 |
|-----------------------------------------------------------------------------------------------------------------------------------------|
1. | 1 25jun2018 4010 . . -.01309818 .7045 . . . 17.445 no 0 |
2. | 1 26jun2018 4010 . . -.00472636 .6986 . . . 17.432 no 0 |
3. | 1 27jun2018 4010 . . -.0038682 .6905 . . . 17.455 no 0 |
4. | 1 28jun2018 4010 . . .02043473 .7001 . . . 18.349 no 0 |
5. | 1 29jun2018 4010 . . -.00893511 .6828 . . . 18.341 no 0 |
|-----------------------------------------------------------------------------------------------------------------------------------------|
6. | 1 02jul2018 4010 6091759 1.1272 -.01051936 .6636 .9105 .014 28.6393 19.516 EVT 1 |
7. | 1 03jul2018 4010 6091759 1.1272 .0035916 .6678 .9105 .014 28.6393 19.848 No 0 |
8. | 1 04jul2018 4010 6091759 1.1272 .00271767 .665 .9105 .014 28.6393 19.839 No 0 |
9. | 1 05jul2018 4010 6091759 1.1272 .00105467 .6614 .9105 .014 28.6393 19.796 No 0 |
10. | 1 06jul2018 4010 6091759 1.1272 .00845053 .6693 .9105 .014 28.6393 20.429 No 0 |
|-----------------------------------------------------------------------------------------------------------------------------------------|
11. | 1 09jul2018 4010 6091759 1.1272 .00719415 .6857 .9105 .014 28.6393 22.034 No 0 |
12. | 1 10jul2018 4010 6091759 1.1272 -.00624704 .6836 .9105 .014 28.6393 21.918 No 0 |
13. | 1 11jul2018 4010 6091759 1.1272 -.00747198 .67 .9105 .014 28.6393 21.99 No 0 |
14. | 1 12jul2018 4010 6091759 1.1272 .00511515 .6836 .9105 .014 28.6393 22.737 No 0 |
15. | 1 13jul2018 4010 6091759 1.1272 -.00675701 .6778 .9105 .014 28.6393 22.257 No 0 |
|-----------------------------------------------------------------------------------------------------------------------------------------|
16. | 1 16jul2018 4010 6091759 1.1272 -.0041779 .6721 .9105 .014 28.6393 22.045 No 0 |
17. | 1 17jul2018 4010 6091759 1.1272 .00716007 .6743 .9105 .014 28.6393 22.132 No 0 |
18. | 1 18jul2018 4010 6091759 1.1272 .01014313 .6793 .9105 .014 28.6393 22.309 No 0 |
19. | 1 19jul2018 4010 6091759 1.1272 .00896441 .6828 .9105 .014 28.6393 22.255 No 0 |
20. | 1 20jul2018 4010 6091759 1.1272 .01805844 .705 .9105 .014 28.6393 24.13 No 0 |
|-----------------------------------------------------------------------------------------------------------------------------------------|
0 Response to Lagged time dummies, panel regression
Post a Comment