SVAR model - Error

[mj13]

Hello!

I´m trying to estimate a four-variable SVAR of the Sims-Bernanke type – ATS/Dollar and Yen/Dollar exchange rates, and the Austria/US and Japanese/US industrial production ratios, from Q1:1196 to Q3:2013

I already estimated the VAR model, and now I selected Proc <- Estimate Structural Factorization

Now I´m writing the following restrictions (which I now it has to be 6 to be an exactly identified system).
@e1 = C(1)*@u1 + C(2)*@u3 + C(3)*@u4
@e2 = C(4)*@u2 + C(2)*@u3 + C(5)*@u4
@e3 = C(6)*@u1 + C(7)*@u3 + C(8)*@u4
@e4 = C(9)*@u2 + C(7)*@u3 + C(10)*@u4

which corresponds to the following restrictions:

e1t ... a11 0 a13 a14 ... u1
e2t ... 0 a22 a13 a24 ... u2
e3t ... a31 0 a33 a34 ... u3
e4t ... 0 a42 a33 a44 ... u4

The problem is that I get the following error: Structural VAR matrix is singular at starting values. Reset starting values or respecify restrictions to ensure that B is non-singular

How can I solve this? Please help! It is urgent