How to do a Chow Test with a moving window.

[Fernando Canales]

Dear Statalist,

I want to do a chow test but I have not found information with moving window in Stata, only in R.

I am testing data of financial frauds, in specific I want know what happen with the coefficients when occur a fraud but using a moving window to know when the coefficient returns to be the same than before the fraud. I need to found significance between the coefficient before and after and know the date when the significance occurs. I already know the date of the break, but I need to compute this in Stata and I am lost.

In R exist a package called strucchange but I want to use Stata.
This is a paper where that package is used:

Link

I want do the same in Stata.

Can anybody help me?
Thanks you.

[Phil Bromiley]

You might try estat sbknown . Alternatively, make a dummy equal 0 before and 1 after, then include i.dummy#x with your x variables. Then you can use any of a variety of test procedures. Finally, you can run the two regressions (one before, one after), and use sureg to combine them and do then do the tests.
See also http://www.stata.com/support/faqs/st...cs/chow-tests/

[Fernando Canales]

You might try estat sbknown . Alternatively, make a dummy equal 0 before and 1 after, then include i.dummy#x with your x variables. Then you can use any of a variety of test procedures. Finally, you can run the two regressions (one before, one after), and use sureg to combine them and do then do the tests.
See also http://www.stata.com/support/faqs/st...cs/chow-tests/

Thanks Phil, but I still don't know how to include the moving window; stat sbknown only works for to know when a break happen.
I thought that could do regressions with a loop and stop when found a p-value, but I am a noob in Stata.