il sent you my data into your mail to try. thanks
-
Login or Register
- Log in with
- You are not logged in. You can browse but not post. Login or Register by clicking 'Login or Register' at the top-right of this page. For more information on Statalist, see the FAQ.
-
-
Please professor I sent you my data into your mailComment
-
Professor I have a same problem I don't why? please help meComment
-
-
I will change my stata version if I have the same i will come back to youComment
-
Dear prof. Kripfganz,
Hi, I have a question if there's any way to implement Difference GMM than System GMM in the first stage. I've read your coauthored article and found that there's no restriction for first-stage estimator and I'd like to employ Difference GMM in my case.
I read 'help xtseqreg' and thought the option for gmmiv(.. difference model(difference)) was for Difference GMM but when I implemented 'xtabond2 ...' that you specified in your post as an equivalent way to mimic the first stage results, I found that it was System GMM. So I don't know how I could do Difference GMM with xtseqreg.Last edited by Jieun Lee; 25 Mar 2019, 04:09.Comment
-
Yes, you can use the difference GMM estimator in the first stage. Such an example is given in my [URL=https://www.statalist.org/forums/forum/general-stata-discussion/general/1374005-xtseqreg-new-stata-command-for-sequential-two-stage-gmm-estimation-of-linear-panel-models#post1374005]post #1[7URL] of this Statalist topic.
Why do you believe it was system GMM?Comment
-
Thank you for your information! I ran the code of 'xtabond2 ...' that you said in the post and Stata reported that it is 'Dynamic panel estimation - System GMM' and so I thought it was about System GMM. (Please refer to the attached captured files.)
So the 'xtseqreg' example in your link (or the very first post in this thread) is for Difference GMM then?Attached FilesLast edited by Jieun Lee; 25 Mar 2019, 06:46.Comment
-
-
To be precise, you obtain the difference GMM estimator by adding the option noconstant to the xtseqreg command, and equivalently the option noleveleq to the xtabond2 command. As you will notice, the constant term disappears from the regression output but all the other coefficients remain the same.Comment
-
Oh, I see. I really appreciate your kind and quick feedback!
Comment
-
Hi, I am estimating a dynamic panel model for trade. The model is:
lnexports l.lnexports lngdpi lngdpj lndistance inflationi inflationj hombias
Where I am treating l.lnexports as endogenous and for the moment assuming other regressors are exogenous. The coefficients of interest are the homebias dummies which are equal to 1 if shipments travel within a country. I have data for T=3 and N=6912 and it is a balanced panel.
I had two questions. First I have just read this thread, your paper with Schwartz and wanted to quickly check whether I have successfully implemented this above model for XTSEQREG as I want to implement a system GMM estimation (Blundell and Bond 1998) to address endogeneity in l.lnexports.
Code:xtseqreg lnexports l.lnexports lngdpi lngdpj lndistance neighbour homebias, /// gmm(l.lnexports, lag(1 2) collapse model(difference)) /// iv(l.lnexports, difference model(level))/// iv(lngdpi lngdpj lndistance neighbour homebias priceincreasei priceincreasej , difference model(level)) twostep vce(robust) estimate store gmm1 xtseqreg lnexports (l.lnexports lngdpi lngdpj lndistance neighbour homebias) lndistance neighbour homebias, /// vce(robust) first(gmm1, nocons) iv(lngdpi lngdpj inflationi inflationj)
Secondly following the literature for the pooled OLS I have computed time varying origin and destination fixed effects as well as split my homebias dummy by year, thus forming: hombias1 homebias2 and homebias3. Since neither are time invariant would you suggest implementing system GMM using xtabond2?
Many Thanks,
SamuelComment
-
It is not entirely clear from your question but I assume that lndistance neighbour homebias are time-invariant variables. In any case, you should not specify the same variables in the first and second stage. Your first-stage model should not include these three variables in this case (neither in the regressor list nor in the list of instruments). Other than that, the specification looks alright, assuming that inflation and GDP are relevant and valid instruments, i.e. correlated with the time-invariant variables and uncorrelated with the unobserved effects.
When you interact the the time-invariant variables with time dummies (to obtain hombias1 homebias2 homebias3), the two-stage approach might be harder to justify. A (system) GMM approach is applicable here. The following Statalist topic provides more information on GMM estimation of dynamic panel models in Stata:
XTDPDGMM: new Stata command for efficient GMM estimation of linear (dynamic) panel models with nonlinear moment conditionsComment
-
Dear Sebastian,
I have a question about the consistency of results between the xtabond2 and the xtseqreg commands. I am estimating a production function using a panel dataset of the following form.
Here is the output when I run the xtabond2 command:Code:xtset q1 time panel variable: q1 (unbalanced) time variable: time, 1 to 3, but with gaps delta: 1 unit
There are 50 instruments that have been used above. Now, I run a similar command using xtsreqreg:Code:.xtabond2 lnlabprod_v3 capprod rawprod npprod dummy maxprod i.Year if (q13a ==1 & q11a == 0 & industry_same ==1), gmm > (lnlabprod_v3 capprod rawprod npprod dummy maxprod, lag(2 2) eq(level)) gmm (lnlabprod_v3 capprod rawprod npprod du > mmy maxprod, lag(2 2) eq(diff)) iv(i.Year i.q17a, eq(level)) iv(i.q17a, eq(diff)) robust twostep Favoring speed over space. To switch, type or click on mata: mata set matafavor space, perm. Warning: Two-step estimated covariance matrix of moments is singular. Using a generalized inverse to calculate optimal weighting matrix for two-step estimation. Difference-in-Sargan/Hansen statistics may be negative. Dynamic panel-data estimation, two-step system GMM ------------------------------------------------------------------------------ Group variable: q1 Number of obs = 5612 Time variable : time Number of groups = 2982 Number of instruments = 44 Obs per group: min = 1 Wald chi2(8) = 16799.40 avg = 1.88 Prob > chi2 = 0.000 max = 3 ------------------------------------------------------------------------------ | Corrected lnlabprod_v3 | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- capprod | .1737902 .068021 2.55 0.011 .0404714 .307109 rawprod | .0695543 .0560353 1.24 0.215 -.0402729 .1793815 npprod | .2612364 .1383583 1.89 0.059 -.0099408 .5324136 dummy | 3.12106 1.320894 2.36 0.018 .5321542 5.709965 maxprod | .4540816 .180837 2.51 0.012 .0996476 .8085157 | Year | 2011 | 0 (empty) 2013 | -4.680937 .047691 -98.15 0.000 -4.774409 -4.587464 2015 | -4.759981 .054865 -86.76 0.000 -4.867515 -4.652448 | _cons | 10.94667 1.147688 9.54 0.000 8.697245 13.1961 ------------------------------------------------------------------------------ Instruments for first differences equation Standard D.(0b.q17a 1.q17a 2.q17a 3.q17a 4.q17a 5.q17a 6.q17a 7.q17a 8.q17a 9.q17a 10.q17a 11.q17a 12.q17a 13.q17a 14.q17a 15.q17a 16.q17a 17.q17a 18.q17a 19.q17a 20.q17a) GMM-type (missing=0, separate instruments for each period unless collapsed) L2.(lnlabprod_v3 capprod rawprod npprod dummy maxprod) Instruments for levels equation Standard 2011b.Year 2013.Year 2015.Year 0b.q17a 1.q17a 2.q17a 3.q17a 4.q17a 5.q17a 6.q17a 7.q17a 8.q17a 9.q17a 10.q17a 11.q17a 12.q17a 13.q17a 14.q17a 15.q17a 16.q17a 17.q17a 18.q17a 19.q17a 20.q17a _cons GMM-type (missing=0, separate instruments for each period unless collapsed) DL2.(lnlabprod_v3 capprod rawprod npprod dummy maxprod) ------------------------------------------------------------------------------ Arellano-Bond test for AR(1) in first differences: z = -5.90 Pr > z = 0.000 Arellano-Bond test for AR(2) in first differences: z = . Pr > z = . ------------------------------------------------------------------------------ Sargan test of overid. restrictions: chi2(35) = 75.26 Prob > chi2 = 0.000 (Not robust, but not weakened by many instruments.) Hansen test of overid. restrictions: chi2(35) = 38.54 Prob > chi2 = 0.312 (Robust, but weakened by many instruments.) Difference-in-Hansen tests of exogeneity of instrument subsets: gmm(lnlabprod_v3 capprod rawprod npprod dummy maxprod, eq(diff) lag(2 2)) Hansen test excluding group: chi2(29) = 30.33 Prob > chi2 = 0.397 Difference (null H = exogenous): chi2(6) = 8.21 Prob > chi2 = 0.223 iv(2011b.Year 2013.Year 2015.Year 0b.q17a 1.q17a 2.q17a 3.q17a 4.q17a 5.q17a 6.q17a 7.q17a 8.q17a 9.q17a 10.q17a 11. > q17a 12.q17a 13.q17a 14.q17a 15.q17a 16.q17a 17.q17a 18.q17a 19.q17a 20.q17a, eq(level)) Hansen test excluding group: chi2(16) = 10.85 Prob > chi2 = 0.818 Difference (null H = exogenous): chi2(19) = 27.69 Prob > chi2 = 0.090 iv(0b.q17a 1.q17a 2.q17a 3.q17a 4.q17a 5.q17a 6.q17a 7.q17a 8.q17a 9.q17a 10.q17a 11.q17a 12.q17a 13.q17a 14.q17a 15 > .q17a 16.q17a 17.q17a 18.q17a 19.q17a 20.q17a, eq(diff)) Hansen test excluding group: chi2(17) = 23.73 Prob > chi2 = 0.127 Difference (null H = exogenous): chi2(18) = 14.81 Prob > chi2 = 0.675
As you can see, some of the coefficients are completely changing signs and significance. Also, the instrument count has reduced to 44. Am I not replicating the code properly?Code:. xi: xtseqreg lnlabprod_v3 capprod rawprod npprod dummy maxprod if (q13a == 1 & q11a == 0 & industry_same ==1), gmmiv > (lnlabprod_v3 capprod rawprod npprod dummy maxprod, lag (2 2) model(level)) gmmiv (lnlabprod_v3 capprod rawprod npp > rod dummy maxprod, lag (2 2) model(diff)) iv(i.Year i.q17a, model(level)) iv(i.q17a, model(diff)) teffects vce(robus > t) twostep i.Year _IYear_2011-2015 (naturally coded; _IYear_2011 omitted) i.q17a _Iq17a_0-20 (naturally coded; _Iq17a_0 omitted) Group variable: q1 Number of obs = 5612 Time variable: time Number of groups = 2982 Obs per group: min = 1 avg = 1.881958 max = 3 Number of instruments = 50 (Std. Err. adjusted for clustering on q1) ------------------------------------------------------------------------------ | WC-Robust lnlabprod_v3 | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- capprod | .217903 .0325344 6.70 0.000 .1541367 .2816693 rawprod | .065866 .035395 1.86 0.063 -.003507 .135239 npprod | .4307212 .0701663 6.14 0.000 .2931978 .5682446 dummy | -.5710566 .2646682 -2.16 0.031 -1.089797 -.0523164 maxprod | -.0702817 .0373012 -1.88 0.060 -.1433908 .0028274 | time | 2 | -4.613127 .0382481 -120.61 0.000 -4.688092 -4.538162 3 | -4.65214 .036684 -126.82 0.000 -4.72404 -4.580241 | _cons | 13.80343 .5484781 25.17 0.000 12.72843 14.87843 ------------------------------------------------------------------------------
Thanks a lot for your help, in advance.
Suchita
Comment
-
Sorry, just a correction, the instrument count is 44 for the GMM and 50 for the xtseqreg and not the reverse, as I previously stated.Comment
-
By default, xtabond2 creates first differences of the GMM-type instruments for the level equation and of the standard instruments for the first-differenced equation. With xtseqreg, you need to explicitly specify the difference suboption because by default no such transformation is applied, i.e.
Code:gmmiv(..., difference lag (2 2) model(level)) iv(..., difference model(diff))
Comment
Comment