test statistics and iterations

[justin_m]

Hi there!

I just started using eviews and I have a query.
I want to perfom a chow breakpoint and a chow forecast test for each one of a series of iterations using the last 100 observations as a prediction set.

wfcreate (wf=chow) u 350

!constant=0
!beta=13
series x=@rnorm
series eps=@rnorm
series y=!constant+!beta*x + eps
!iterations=1000
for !i=1 to !iterations
series eps = @rnorm ´generate new residuals and series y for a constant x
series y = !alpha + !beta*x + eps
equation eq1.ls y c x ´regression
eq1.chow(f) 251 ´ chow forecast test
eq1.chow 251 ´ chow breakpoint test
next

I wish to know how often one rejects the null for each test. I thought of storing the f-statsitics for both tests in two vectors and count from there. But there is probably a better way.... I also tried to use freeze eq1.chow(f) 251 and count but then "infinitely" many windows open and doing this one by one would consumo way too much time.

[EViews Gareth]

You'll have to freeze the tests into a table, then extract the fstats into a vector.

http://forums.eviews.com/viewtopic.php?f=5&t=18

http://forums.eviews.com/viewtopic.php?f=5&t=1638

[justin_m]

Thank you so much Gareth!

One more question: how can one generate power curves for both types of tests? I've been googling and reading the manuals under Help but I am not getting far!

Thanks in advance,

JM