MIMIC model estimation in state space
Posted: Mon Apr 26, 2010 10:57 am
Hello. I am new to this forum. I try to estimate a Multiple indicators multiple causes model (MIMIC), it's written in eviews guide that it can be estimated as a state space model. So I specify the model:
@state ue_sub=c(1)*itax_sa+c(2)*scontrib_sa+c(3)*dtax_sa+c(4)*govsal_sa+c(5)*subsidies_sa+c(6)*d_socbenef_sa+c(7)*openess+c(8)*un+c(9)*infl_sa+e1
@signal d_curr_ratio_sa=c(10)*ue_sub+e2
@signal gdp_g=c(11)*ue_sub+e3
@signal lforce_g=c(12)*ue_sub+e4
@ename e1
@ename e2
@ename e3
@ename e4
@evar var(e1)=1
@evar var(e2)=exp(c(13))
@evar var(e3)=exp(c(14))
@evar var(e4)=exp(c(15))
@evar cov(e1,e2)=0
@evar cov(e1,e3)=0
@evar cov(e1,e4)=0
@evar cov(e2,e3)=exp(c(16))
@evar cov(e2,e4)=exp(c(17))
@evar cov(e3,e4)=exp(c(18))
But when I try to estimate it I get "Missing value found in state innovation vector", I don't get what is wrong with the model??? What is the problem with the specification???
@state ue_sub=c(1)*itax_sa+c(2)*scontrib_sa+c(3)*dtax_sa+c(4)*govsal_sa+c(5)*subsidies_sa+c(6)*d_socbenef_sa+c(7)*openess+c(8)*un+c(9)*infl_sa+e1
@signal d_curr_ratio_sa=c(10)*ue_sub+e2
@signal gdp_g=c(11)*ue_sub+e3
@signal lforce_g=c(12)*ue_sub+e4
@ename e1
@ename e2
@ename e3
@ename e4
@evar var(e1)=1
@evar var(e2)=exp(c(13))
@evar var(e3)=exp(c(14))
@evar var(e4)=exp(c(15))
@evar cov(e1,e2)=0
@evar cov(e1,e3)=0
@evar cov(e1,e4)=0
@evar cov(e2,e3)=exp(c(16))
@evar cov(e2,e4)=exp(c(17))
@evar cov(e3,e4)=exp(c(18))
But when I try to estimate it I get "Missing value found in state innovation vector", I don't get what is wrong with the model??? What is the problem with the specification???