Hi STATA community
I have a panel data for period 2012-2020, with #obs 1988 and #groups 223, I run the xtreg, fe and the results as shown below, after that I run xtserial and I have serial autocorrelation issue! How can I deal with this problem please?
xtreg WROAI FEMproportion ln_FS1 LEV_PERCENT BS BIND CEO_DUA firmage i.year i.INDUSTRY*, fe
note: 2020.year omitted because of collinearity
note: 1.INDUSTRY1 omitted because of collinearity
note: 1.INDUSTRY2 omitted because of collinearity
note: 1.INDUSTRY3 omitted because of collinearity
note: 1.INDUSTRY4 omitted because of collinearity
note: 1.INDUSTRY5 omitted because of collinearity
note: 1.INDUSTRY6 omitted because of collinearity
note: 1.INDUSTRY7 omitted because of collinearity
note: 1.INDUSTRY8 omitted because of collinearity
note: 1.INDUSTRY9 omitted because of collinearity
note: 1.INDUSTRY10 omitted because of collinearity
Number of obs = 1,988
Group variable: id Number of groups = 223
R-sq: Obs per group:
within = 0.1469 min = 6
between = 0.0006 avg = 8.9
overall = 0.0001 max = 9
F(14,1751) = 21.54
corr(u_i, Xb) = -0.9574 Prob > F = 0.0000
WROAI Coef. Std. Err. t P>t [95% Conf. Interval]
FEMproportion 2.219943 2.045081 1.09 0.278 -1.791114 6.231
ln_FS1 .9539394 .5602033 1.70 0.089 -.1447984 2.052677
LEV_PERCENT -.1649817 .0163612 -10.08 0.000 -.1970713 -.1328922
BS .3332249 .1401214 2.38 0.018 .0584019 .6080478
BIND -.8364824 2.397737 -0.35 0.727 -5.539211 3.866246
CEO_DUA -10.46369 4.086443 -2.56 0.011 -18.47852 -2.448872
firmage -.7101108 .0989218 -7.18 0.000 -.9041281 -.5160936
year
2013 -.0385384 .5529471 -0.07 0.944 -1.123044 1.045968
2014 1.556931 .5258847 2.96 0.003 .5255028 2.588359
2015 1.800889 .508461 3.54 0.000 .8036348 2.798144
2016 1.858767 .5045516 3.68 0.000 .8691804 2.848354
2017 3.16199 .5125383 6.17 0.000 2.156739 4.167242
2018 3.066505 .5300157 5.79 0.000 2.026975 4.106035
2019 2.695152 .5482957 4.92 0.000 1.619768 3.770535
2020 0 (omitted)
1.INDUSTRY1 0 (omitted)
1.INDUSTRY2 0 (omitted)
1.INDUSTRY3 0 (omitted)
1.INDUSTRY4 0 (omitted)
1.INDUSTRY5 0 (omitted)
1.INDUSTRY6 0 (omitted)
1.INDUSTRY7 0 (omitted)
1.INDUSTRY8 0 (omitted)
1.INDUSTRY9 0 (omitted)
1.INDUSTRY10 0 (omitted)
_cons 25.41833 7.117782 3.57 0.000 11.45809 39.37858
sigma_u 25.773241
sigma_e 6.0717567
rho .94741855 (fraction of variance due to u_i)
F test that all u_i=0: F(222, 1751) = 11.62 Prob > F = 0.0000
xtserial WROAI FEMproportion ln_FS1 LEV_PERCENT BS BIND CEO_DUA firmage
Wooldridge test for autocorrelation in panel data
H0: no first order autocorrelation
F( 1, 222) = 12.314
Prob > F = 0.0005
How can I deal with this problem? I tried robust and lagging 1, 2 and the same results. How many lagging can I generate to fix this issue?
0 Response to serial autocorrelation in panel data
Post a Comment