Stationarity test - Is this correct?

SleepingGiant · June 9th, 2004, 9:33 pm

After searching the archives for discussions on stationarity, I found a post from "Student" regarding the implementation of the Dickey-Fuller test. He wrote the following:QuoteSat May 24, 03 01:30 PM If you have to stay with excel, I reckon it won't be too difficult either. At the end of the day, it's just a t-test on a linear regression, isn't it? Is "Student" correct? Is this an acceptable (sufficient) method of testing for stationarity? My gut tells me that it isn't, but since I'm not sure I thought I'd ask.Thanks!

Garywind · June 9th, 2004, 10:30 pm

no, that is incorrect. Dickey-fuller test does looks like a t-test(testing whether P=0) but it follows different distribution. I think you can perform ADF calucation in Excel, but you have to check the DF distribution table supplied by authors to perform hypothesis test. hope this help. ^_^cheers,Gary