Page 1 of 1

Estimate a cointegration model with a deterministic variable

Posted: Thu Jul 02, 2015 2:06 pm
by xin_zh
Hi Tom,

I know I can use "dettrend=" option in the procedure @cats(options) to include a constant or a time trend in a cointegration model. But what should I do if I need to include a deterministic dummy variable with "1" for some specific days and "0" for others? I searched the user guide and other manuals but did not find the answer. please help!

Thank you very much!
Xin

Re: Estimate a cointegration model with a deterministic vari

Posted: Fri Jul 03, 2015 2:47 pm
by TomDoan
Isn't that what the "DUM" option is for? Section 3.5 (in the CATS manual) covers an example of its use.

Re: Estimate a cointegration model with a deterministic vari

Posted: Wed Jul 08, 2015 6:46 pm
by xin_zh
Exactly! Thank you very much Tom!