Autocorrelation test (i need he command) when dealing with Panel Data FE

[Matteo Torchio]

can someone help me finding the right way (the also woks, because most of the attempts that I performed were not working) to check for autocorrelation when dealing with panel data FE? this is my work so far: //----------------------------*4. Original Model-----------------------------
*Preparation of the Work Environment
clear all
set more off
cd "/Users/catz/Desktop/Master in Finance/FEP- MÉTODOS ECONOMÉTRICOS/Assignment"
use "/Users/catz/Desktop/Master in Finance/FEP- MÉTODOS ECONOMÉTRICOS/Assignment/dataset.dta"
log using assigecon.log, replace

*Program Installation
ssc install asdoc
ssc install coefplot
net install dca, from("https://raw.github.com/ddsjoberg/dca.stata/master/")replace

*Database
use "dataset.dta", clear

global ylist firmgrowth
global xlist female_director directors_number board_size ceo_duality board_independence firmsize

describe c_id yr $ylist $xlist
summarize c_id yr $ylist $xlist
* Set data as panel data
sort c_id yr
xtset c_id yr
xtdescribe
xtsum c_id yr $ylist $xlist

///////////////////////////////*4.1.ORIGINAL MODEL//////////////////////////////////////
*Descriptive Statistics
describe c_id yr $ylist $xlist
summarize c_id yr $ylist $xlist

* Pooled OLS estimator
reg $ylist $xlist

* Population-averaged estimator
xtreg $ylist $xlist, pa

* Between estimator
xtreg $ylist $xlist, be

*Estimation of the Fixed Effects Model
xtreg $ylist $xlist, fe
estimates store fixed

*Estimation of the Random Effects Model
xtreg $ylist $xlist, re
estimates store random

* Breusch-Pagan LM test for random effects versus OLS
quietly xtreg $ylist $xlist, re
xttest0

*Hausman Test to Compare FE and RE
hausman fixed random, sigmamore

* The Hausman test shows significant differences between the coefficients for the fixed effects and random effects model. Therefore, we need to use the fixed effects model. *

Specification
xtreg $ylist $xlist, fe
predict yhat_fe, xb
gen yhat_fe2 = yhat_fe^2
gen yhat_fe3 = yhat_fe^3
xtreg $ylist $xlist yhat_fe2 yhat_fe3, fe
test yhat_fe2 yhat_fe3


////////////MULTICOLINEARITY/////////////////////////
global ylist firmgrowth
global xlist female_director board_size ceo_duality board_independence firmsize
//////////Correlation Matrix///////
pwcorr $xlist
///////////VIF////////////
regress $ylist $xlist
vif

////////////AUTOCORRELATION/////////////////////////
xtregar firmgrowth female_director directors_number board_size ceo_duality board_independence firmsize, fe

///////////////////////////////////////////////////////
* 5. HETEROSKEDASTICITY
///////////////////////////////////////////////////////

* Baseline pooled OLS
regress $ylist $xlist i.yr

* 5.1 Breusch–Pagan / Cook–Weisberg test
estat hettest

* 5.2 White's general test
estat imtest, white

* FIXED EFFECTS MODEL
xtreg $ylist $xlist i.yr, fe

* MODIFIED WALD TEST FOR GROUPWISE HETEROSKEDASTICITY
ssc install xttest3
xttest3

* MAIN FE MODEL – firm-clustered robust SE
xtreg $ylist $xlist i.yr, fe vce(cluster c_id)

[Carlo Lazzaro]

Matteo:
welcome to this forum.
The policy is not to consider posts related to home or class assignments (please see Help - Statalist, #4).
That said, you may want to take a look at the community-contribuyed module -xttest2-.
I am also not clear with your switching from -xtreg,fe- to -xtregar,fe-.

[Matteo Torchio]

Thank you very much for your prompt response. First of all I was not aware about the policies of thegroup, I am extremely sorry. However, my main goal is to understand what I am doing and see why when it comes to conduct a test for Autocorrelation it does not work on my STATA. Regarding the xttest2, could you please be more kind as to give me more information so that I can check this area on my own? On the other hand, which one do you think is more appropriate to use between xtreg and xtregar? Thank you in advance.

[Carlo Lazzaro]

Matteo:
1) new posters are kindly requested to read the FAQ before their first message. Skipping this step is not a waiver.
2) just type -search xttest2-;
3) your last question has no "one size fits all" answer. -xtreg- works for N>T pabel datasets.

[Matteo Torchio]

Thanks! However, I do have a question... isn't the xttest2 the breusch–pagan lm test? If so, why I will need to use it in my FE Panel Data if I am trying to check for autocorrelation? I tried to used the xtserial option but it is not working.

[Carlo Lazzaro]

Matteo:
-search xtcsd-.

[Matteo Torchio]

what about this option:

////////////AUTOCORRELATION/////////////////////////

xtreg firmgrowth female_director directors_number board_size ceo_duality board_independence firmsize, fe

///////////////////////////////////////////////////////
///////////// MANUAL WOOLDRIDGE TEST //////////////////
///////////////////////////////////////////////////////

reg D.firmgrowth D.female_director D.directors_number D.board_size D.ceo_duality D.board_independence D.firmsize, noconst vce(cluster c_id)

predict ehat, resid

reg ehat L.ehat, noconst vce(cluster c_id)

test L.ehat = -0.5