Ordered Probit Estimation

[Artur Patricio]

Hi I am trying to make forecast for an ordered probit model. I have already realized that in order to do this I have to make a model firts, bit after I do this and try to forecast my endogenous variables by some reason there's an error message that does not allow me to do the forecast.

This is what I see on my estimation output

Dependent Variable: RTG_MDY
Method: ML - Ordered Probit (Quadratic hill climbing)
Date: 05/21/15 Time: 16:20
Sample (adjusted): 1981 2013
Included observations: 793 after adjustments
Number of ordered indicator values: 21
Convergence achieved after 11 iterations
Covariance matrix computed using second derivatives

Variable Coefficient Std. Error z-Statistic Prob.

ASS_1/NE_EXP_GNFS_CD -0.025025 0.005371 -4.659421 0.0000
GGSB_NPGDP(-1) -0.042150 0.012522 -3.366092 0.0008
BCA_NGDPD -0.045744 0.007019 -6.517404 0.0000
PCPIPCH 0.054769 0.008019 6.830014 0.0000
NGDPPC(-1) -8.29E-09 9.54E-09 -0.869345 0.3847
NGDP_RPCH -0.017589 0.012497 -1.407425 0.1593
DEV_IND -2.244267 0.106431 -21.08666 0.0000

Limit Points

LIMIT_2:C(8) -2.137352 0.127925 -16.70787 0.0000
LIMIT_3:C(9) -1.757372 0.123725 -14.20386 0.0000
LIMIT_4:C(10) -1.457038 0.120146 -12.12725 0.0000
LIMIT_5:C(11) -1.233189 0.117314 -10.51189 0.0000
LIMIT_6:C(12) -0.917200 0.112517 -8.151650 0.0000
LIMIT_7:C(13) -0.567269 0.107344 -5.284613 0.0000
LIMIT_8:C(14) -0.382248 0.105306 -3.629875 0.0003
LIMIT_9:C(15) -0.087150 0.103576 -0.841411 0.4001
LIMIT_10:C(16) 0.127114 0.103672 1.226126 0.2202
LIMIT_11:C(17) 0.461855 0.105533 4.376401 0.0000
LIMIT_12:C(18) 0.852110 0.111143 7.666788 0.0000
LIMIT_13:C(19) 1.003352 0.114428 8.768437 0.0000
LIMIT_14:C(20) 1.267768 0.122105 10.38257 0.0000
LIMIT_15:C(21) 1.564110 0.133329 11.73122 0.0000
LIMIT_16:C(22) 1.663284 0.137626 12.08552 0.0000
LIMIT_17:C(23) 1.963694 0.153749 12.77209 0.0000
LIMIT_18:C(24) 2.149239 0.168163 12.78071 0.0000
LIMIT_19:C(25) 2.251483 0.179098 12.57123 0.0000
LIMIT_20:C(26) 2.378641 0.195049 12.19509 0.0000
LIMIT_21:C(27) 2.834274 0.289753 9.781686 0.0000

Pseudo R-squared 0.175262 Akaike info criterion 4.120366
Schwarz criterion 4.279568 Log likelihood -1606.725
Hannan-Quinn criter. 4.181550 Restr. log likelihood -1948.164
LR statistic 682.8784 Avg. log likelihood -2.026135
Prob(LR statistic) 0.000000



Error message

Model: M_RATING
Date: 05/21/15 Time: 16:27
Sample (adjusted): 1981 2020
Solve Options:
Dynamic-Deterministic Simulation
Solver: Broyden
Max iterations = 5000, Convergence = 1e-08

Scenario: Baseline
Solve begin 16:27:16
1981 NA generated for I_RTG_MDY
Solve terminated - Unable to compute due to missing data in "I_RTG_MDY = @COEF(1) *


Does any one knows how to solve this ?