 -- load the data into sca. Name the data ir1 & ir3.

input ir1,ir3. file 'm-gs1n3.dat'

 -- take log transformation. Call the transformed series r1t & r3t.

r1t=ln(ir1)

r3t=ln(ir3)

 -- identify VAR model via M-stat & AIC

miden r1t,r3t. no ccm. arfits 1 to 8.

 -- specify a VARMA(2,1) model. Name the model m1.

 -- Denote the coef-mtx by p1, p2, and t1.

mtsmodel m1. series r1t,r3t. @

model (i-p1*b-p2*b**2)series=p0+(i-t1*b)noise.

 -- estimation

mestim m1. method exact. hold resi(res1,res2)

 -- put zero parameter constraints

p1(2,1)=0

cp1(2,1)=1

p2(2,1)=0

cp2(2,1)=1

p2(2,2)=0

cp2(2,2)=1

t1(2,1)=0

ct1(2,1)=1

 -- redo estimation

mestim m1. method exact. hold resi(res1,res3).

 -- check residuals cross-correlation matrices

miden res1,res3.

stop

