Dear all
I have three variables : average of global crude oil, real activity index , real oil price.I want to decompose the real oil prices into three shocks as kilian did in paper 2009.Kilian use cholesky factorization in his restrictions. SO my quetsions: First, Could review my codes if it is in the right way? Second , the most important question : I want to help me to calculate impulse response function for the three shocks (eta) on the three variables using Bootstrap 2000:

matrix A1 = (1,0,0 \ .,1,0 \ .,.,1)
. matrix B1 = (.,0,0 \ 0,.,0 \ 0,0,.)
. svar TProduction Global_Index RealOilPrice , lags(1/24) aeq(A1) beq(B1)
predict e1,res eq(TProduction )
predict e2,res eq(Global_Index )
predict e3,res eq( RealOilPrice)

mkmat e1 e2 e3, matrix(e)
mat list e
matrix sigma = e(Sigma)
mat list sigma
matrix B = inv(cholesky(sigma)')
mat list B
matrix eta = e * B
mat list eta

Matrix eta represents the reduced error term
If my coding right : how I can calculate the impulse response function of my three variables with the three shocks using bootstrap 2000?