The Allin method is perfect, I have only changed when there is the error message. In my application I have use the following commands to obtain the pseudo-data. But the bootstrap results are strange. Is it right my script to generate the pseudo-data? r=154 ## number of observations system method=sur equation Y const Y(-1) Z(-1) equation X const Y(-1) X(-1) Z(-1) equation Z const Y(-1) Z(-1) end system genr residui=$uhat genr M=$coeff ## Procedure to obtain the pseudo-data ysim=Y xsim=X zsim=Z genr A=resample(residui) loop for i=2..154 --quiet ysim[i]=M[1]+M[2]*ysim[i-1]+M[3]*zsim[i-1]+A[i-1,1] xsim[i]=M[4]+M[5]*ysim[i-1]+M[6]*xsim[i-1]+M[7]*zsim[i-1]+A[i-1,2] zsim[i]=M[8]+M[9]*ysim[i-1]+M[10]*zsim[i-1]+A[i-1,3] endloop