I estimated sytem consisting of two equations and obtained below output.
In this regards, my questions are
(1) What are general diagnostic tests (for GMM estimation) provided by Eviews? I am using Eviews 7.1.
(2) J statistics is very low, almost 0 as seen in the output. This means that this model is something wrong. To my understanding, as the value of J is larger, the model is better. Am I right?
Please give me your short reply?
Thank you in advance.
-------------estimation output -------------------------------------------
System: COEFF_RATIO
Estimation Method: Generalized Method of Moments
Date: 06/29/11 Time: 11:42
Sample: 1999Q1 2009Q3
Included observations: 43
Total system (balanced) observations 86
Kernel: Bartlett, Bandwidth: Fixed (4), No prewhitening
Linear estimation after one-step weighting matrix
Coefficient Std. Error t-Statistic Prob.
C(1) 0.005819 0.004665 1.247377 0.2164
C(2) -1.028896 0.042897 -23.98519 0.0000
C(4) 0.228810 0.105806 2.162547 0.0340
C(6) 0.036068 0.031669 1.138908 0.2586
C(8) -0.017446 0.044273 -0.394069 0.6947
C(10) -0.336310 0.432287 -0.777978 0.4392
C(12) -0.047116 0.105523 -0.446505 0.6566
C(14) -2.629151 1.317966 -1.994855 0.0500
C(16) 0.007033 0.003321 2.117552 0.0378
C(17) -0.820989 0.062046 -13.23202 0.0000
C(19) 0.250807 0.097160 2.581390 0.0119
C(21) -0.029336 0.043103 -0.680604 0.4984
C(23) 0.010782 0.029999 0.359393 0.7204
C(25) -0.359385 0.363671 -0.988214 0.3265
C(27) -0.011578 0.089454 -0.129430 0.8974
C(29) -0.350884 0.209483 -1.675001 0.0984
Determinant residual covariance 8.42E-09
J-statistic -5.61E-45
Equation: DNDA/GDP_NO_SA = C(1) + C(2)*DNFA/GDP_NO_SA +C(4)*CA
/GDP_NO_SA +C(6)*DRUS_ADJ +C(8)*FXDIF +C(10)*INF +C(12)*YC
+C(14)*D1SDRKO
Instruments: DNFA/GDP_NO_SA CA/GDP_NO_SA DRUS_ADJ FXDIF INF
YC D1SDRKO C
Observations: 43
R-squared 0.972699 Mean dependent var -0.026997
Adjusted R-squared 0.967239 S.D. dependent var 0.083736
S.E. of regression 0.015156 Sum squared resid 0.008040
Durbin-Watson stat 2.753145
Equation: DNFA/GDP_NO_SA = C(16) + C(17)*DNDA/GDP_NO_SA + C(19)
*CA/GDP_NO_SA + C(21)*DRUS_ADJ + C(23)*FXDIF + C(25)*INF +
C(27)*YC+ C(29)*D2SDFX
Instruments: DNDA/GDP_NO_SA CA/GDP_NO_SA DRUS_ADJ FXDIF INF
YC D2SDFX C
Observations: 43
R-squared 0.973912 Mean dependent var 0.032511
Adjusted R-squared 0.968695 S.D. dependent var 0.077905
S.E. of regression 0.013784 Sum squared resid 0.006650
Durbin-Watson stat 2.859623
Diagnostic test for GMM estimation(need quick reply-please)
Moderators: EViews Gareth, EViews Moderator
Return to “Econometric Discussions”
Who is online
Users browsing this forum: No registered users and 2 guests
