Xtabond2 struggling with syntax for sysGMM

[Ralph Ross Camelo]

Hello,

I am trying to run system-GMM on the equation: CAGDP = CAGDP_t-1+quinn+logGDPPC+quinn*logGDPPC with the instruments as such: CAGDP_t-1 and quinn lagged by 2 time periods and logGDPPC, quinn*logGDPPC lagged by 1 time period.

The code beneath is my attempt to achieve this. Is this correct? Thanks in advance.

code: xtabond2 CAGDP L1.CAGDP quinn logGDPPC c.quinn#c.logGDPPC, gmmstyle(L3.(CAGDP) L2.(quinn) L1.(c.quinn#c.logGDPPC logGDPPC), laglimits(1 3) collapse) ortho

[Ralph Ross Camelo]

Sorry if I'm not following the protocol for askin questions as I am new to this. I've changed my choice of code to: xtabond2 CAGDP L1.CAGDP quinn logGDPPC c.quinn#c.logGDPPC dependency NFA popgrowth fiscalbalancetoGDP tradeopenness TOT, gmmstyle((logGDPPC c.quinn#c.logGDPPC L2.CAGDP L.quinn), collapse) ortho twostep robust

Which seems a bit more like what I want. Any advice on this? CAGDP and quinn are lagged once more as they are assumed to be endogenous whilst logGDPPC and by extension quinn#logGDPPC should be predetermined. This gives me a hansen value of 0.169 which indicates that my results may be weakened by many instruments. However, my number of instruments is only 48 compared to 85 groups so do you think I should be in the clear?


Instruments for orthogonal deviations equation
GMM-type (missing=0, separate instruments for each period unless collapsed)
L(1/14).(logGDPPC c.quinn#c.logGDPPC L2.CAGDP L.quinn) collapsed
Instruments for levels equation
Standard
_cons
GMM-type (missing=0, separate instruments for each period unless collapsed)
D.(logGDPPC c.quinn#c.logGDPPC L2.CAGDP L.quinn) collapsed
------------------------------------------------------------------------------
Arellano-Bond test for AR(1) in first differences: z = -2.68 Pr > z = 0.007
Arellano-Bond test for AR(2) in first differences: z = -0.80 Pr > z = 0.423
------------------------------------------------------------------------------
Sargan test of overid. restrictions: chi2(37) = 69.76 Prob > chi2 = 0.001
(Not robust, but not weakened by many instruments.)
Hansen test of overid. restrictions: chi2(37) = 45.11 Prob > chi2 = 0.169
(Robust, but weakened by many instruments.)

Difference-in-Hansen tests of exogeneity of instrument subsets:
GMM instruments for levels
Hansen test excluding group: chi2(33) = 43.97 Prob > chi2 = 0.096
Difference (null H = exogenous): chi2(4) = 1.13 Prob > chi2 = 0.889

[David Roodman]

Sorry, I'm not taking the time to figure out exactly which you'd want to lag once and which twice, etc., but it looks like you understand how to use the command correctly.