The RATS Software Forum

Hi All,

I wanted to get some suggestions on how to do a "ROLLING CHOW TEST" in RATS when the date of structural break is not known. In the final results, we should get the test statistic for all the observations (am I right?). I am new to RATS coding. I tried to go through the manual but it has examples only for known structural dates.
Your help will be greatly appreciated.

Use the @APBREAKTEST procedure or the @REGHBREAK procedure. I've attached a program (and its data file) which use @APBREAKTEST and also do the calculation of break test statistics using HAC standard errors.