Testing for near I (2) trends when the signal to noise ratio is small
Publikation: Working paper › Forskning
Standard
Testing for near I (2) trends when the signal to noise ratio is small. / Juselius, Katarina.
Kbh. : Økonomisk institut, Københavns Universitet, 2014.Publikation: Working paper › Forskning
Harvard
APA
Vancouver
Author
Bibtex
}
RIS
TY - UNPB
T1 - Testing for near I (2) trends when the signal to noise ratio is small
AU - Juselius, Katarina
PY - 2014
Y1 - 2014
N2 - Researchers seldom find evidence of I(2) in exchange rates, prices, and other macroeconomics time series when they test the order of integration using univariate Dickey-Fuller tests. In contrast, when using the multivariate ML trace test we frequently find double unit roots in the data. Our paper demonstrates bysimulations that this often happens when the signal-to-noise-ratio is small.
AB - Researchers seldom find evidence of I(2) in exchange rates, prices, and other macroeconomics time series when they test the order of integration using univariate Dickey-Fuller tests. In contrast, when using the multivariate ML trace test we frequently find double unit roots in the data. Our paper demonstrates bysimulations that this often happens when the signal-to-noise-ratio is small.
M3 - Working paper
T3 - University of Copenhagen. Institute of Economics. Discussion Papers (Online)
BT - Testing for near I (2) trends when the signal to noise ratio is small
PB - Økonomisk institut, Københavns Universitet
CY - Kbh.
ER -
ID: 101017294