A question on Unit root

Sun, Jan 11, 2009 5:12 PM

In my textbook, I read that for a Integrated with order 1 process, there
would be atleast one unit root. But if I take following DGP : y[t] =
1.001y[t-1] + epsilon then what I found was, this is a Integrated with order
1 process, because, data generated from this process becomes stationary
after 1 difference. However there is, strictly speaking, no unit root. The
Ch. polynomial has root as 0.999001, this is not exactly one.Or I should
consider 0.999001 as 1?

Can you give a better clarification pls?

View this message in context: http://www.nabble.com/A-question-on-Unit-root-tp21406847p21406847.html
Sent from the Rmetrics mailing list archive at Nabble.com.