Em 12 de março de 2011 Allin <cottrell@wfu.edu> escreveu:

(...)

> But we have missing observations... How can we handle this?

I have no idea how this could work, in a time-series model with
lags. That's why I went for the combined version.

Dear Allin, I made a deeper analysis in the RATS operationalization and I discovered that what it does is equivalent to put zeros at the "NA" observations: 

<script>
open australia.gdt

list Y = PAU IAU E
series z = normal()
scalar tau = 1.5
series r1 = z(-1) >= tau
series r2 = 1-r1

loop foreach i Y
    series $i_r1 = r1 * $i
    series $i_r2 = r2 * $i
endloop

list regime1 = *_r1
list regime2 = *_r2 

"VAR Regime 1" 
<- 
var
 1 regime1
"VAR Regime 2" 
<- 
var
 1 regime2
</script>

Thanks a lot for your always kindly and valuable help!

Best, 
--
Henrique C. de Andrade
Doutorando em Economia Aplicada
Universidade Federal do Rio Grande do Sul
www.ufrgs.br/ppge