Good evening. I am running the command xtdpdsys ETR ROA SIZE LEV PPE NOLs filing_timely_interaction, twostep in stata in order to estimate the effect of the dependent variables on the etr index, but the sargan test and Arellano–Bond test give me p=0.0000. My professor suggested that I try manual instruments. How is it applied in stata and what does it mean? Thank you very much.
xtdpdsys ETR ROA SIZE LEV PPE NOLs filing_timely_interaction ,twostep,
System dynamic panel-data estimation Number of obs = 438,301
Group variable: firm_id Number of groups = 64,921
Time variable: econ_year
Obs per group:
min = 1
avg = 6.751298
max = 19
Number of instruments = 196 Wald chi2(7) = 24902.98
Prob > chi2 = 0.0000
Two-step results
-------------------------------------------------------------------------------------------
ETR | Coefficient Std. err. z P>|z| [95% conf. interval]
--------------------------+----------------------------------------------------------------
ETR |
L1. | .6362364 .0050627 125.67 0.000 .6263137 .6461592
|
ROA | .0848115 .0018484 45.88 0.000 .0811887 .0884342
SIZE | .0069132 .000754 9.17 0.000 .0054354 .0083909
LEV | .0101254 .0011278 8.98 0.000 .0079149 .0123359
PPE | -.0040782 .0009223 -4.42 0.000 -.0058859 -.0022705
NOLs | .073814 .0013805 53.47 0.000 .0711083 .0765197
filing_timely_interaction | .0592145 .0050197 11.80 0.000 .0493761 .0690529
_cons | -.0803129 .0109657 -7.32 0.000 -.1018052 -.0588206
-------------------------------------------------------------------------------------------
Warning: gmm two-step standard errors are biased; robust standard
errors are recommended.
Instruments for differenced equation
GMM-type: L(2/.).ETR
Standard: D.ROA D.SIZE D.LEV D.PPE D.NOLs D.filing_timely_interaction
Instruments for level equation
GMM-type: LD.ETR
Standard: _cons
. estat abond
Arellano–Bond test for zero autocorrelation in first-differenced errors
H0: No autocorrelation
Order z Prob > z
--------------------------
1 -92.721 0.0000
2 12.119 0.0000
--------------------------
. estat sargan
Sargan test of overidentifying restrictions
H0: Overidentifying restrictions are valid
chi2(188) = 3587.72
Prob > chi2 = 0.0000
.
xtdpdsys ETR ROA SIZE LEV PPE NOLs filing_timely_interaction ,twostep,
System dynamic panel-data estimation Number of obs = 438,301
Group variable: firm_id Number of groups = 64,921
Time variable: econ_year
Obs per group:
min = 1
avg = 6.751298
max = 19
Number of instruments = 196 Wald chi2(7) = 24902.98
Prob > chi2 = 0.0000
Two-step results
-------------------------------------------------------------------------------------------
ETR | Coefficient Std. err. z P>|z| [95% conf. interval]
--------------------------+----------------------------------------------------------------
ETR |
L1. | .6362364 .0050627 125.67 0.000 .6263137 .6461592
|
ROA | .0848115 .0018484 45.88 0.000 .0811887 .0884342
SIZE | .0069132 .000754 9.17 0.000 .0054354 .0083909
LEV | .0101254 .0011278 8.98 0.000 .0079149 .0123359
PPE | -.0040782 .0009223 -4.42 0.000 -.0058859 -.0022705
NOLs | .073814 .0013805 53.47 0.000 .0711083 .0765197
filing_timely_interaction | .0592145 .0050197 11.80 0.000 .0493761 .0690529
_cons | -.0803129 .0109657 -7.32 0.000 -.1018052 -.0588206
-------------------------------------------------------------------------------------------
Warning: gmm two-step standard errors are biased; robust standard
errors are recommended.
Instruments for differenced equation
GMM-type: L(2/.).ETR
Standard: D.ROA D.SIZE D.LEV D.PPE D.NOLs D.filing_timely_interaction
Instruments for level equation
GMM-type: LD.ETR
Standard: _cons
. estat abond
Arellano–Bond test for zero autocorrelation in first-differenced errors
H0: No autocorrelation
Order z Prob > z
--------------------------
1 -92.721 0.0000
2 12.119 0.0000
--------------------------
. estat sargan
Sargan test of overidentifying restrictions
H0: Overidentifying restrictions are valid
chi2(188) = 3587.72
Prob > chi2 = 0.0000
.
