Unit root tests show no unit roots whatsoever among any variables. Is cointegration test even still necessary?

0 ⤊

Unit root tests show no unit roots whatsoever among any variables. Is cointegration test even still necessary?

2006-08-27 05:41:53 · 2 answers · asked by carchatmail 1 in Social Science ➔ Economics

2 answers

Probably not. Absense of unit root means there is no stochastic trend in the data...

2006-08-27 11:07:16 · answer #1 · answered by NC 7 · 0⤊ 0⤋

Nope. If your unit root tests reject the unit root, this means that your time series are stationary. Testing for cointegration is (only) useful if you have non-stationary series.

2006-08-31 00:47:55 · answer #2 · answered by jk 1 · 0⤊ 0⤋