SVAR

[learner]

Dear All

I am estimating Structural VAR (SVAR), I face the following problems

(1) The upper and lower lines of Confidence Interval are mirror image of each other. They do not move in tandem (when I use VAR, they behave well)
(2) The values of coefficients are .1000 for all coefficients (for A matrix)
(3) Although EViews (version 6) gives results but also gives message Failure to Improve after 1 iteration (see below)

　Structural VAR Estimates
Date: 12/29/11 Time: 14:52
　Sample (adjusted): 1991:3 2009:3
　Included observations: 73 after adjustments
　Estimation method: method of scoring (analytic derivatives)
　Failure to improve after 1 iterations
　Structural VAR is over-identified (3 degrees of freedom)

What could be the reason and possible solution.

Thanks
Learner