Hi,I obtained the ect term from cajorls in urca. I found a result that would
like to obtain some explanation here. My setting is that I have 2 variate time
series. I use ca.jo to perform Johansen test. 1. I found that sometimes, in the
case where the ca.jo test statistics suggest that I have 1 relationship. After
I investigate the ect obtained from cajorls (with r=1), the ect series fails
unit root test like adf.test completely. I still see high degree of
non-stationarity. Would you please provide some insight?2. The opposite of
above also occurs. I found that the statistics suggests that I have no
cointegrated relationship. But when I tried to perform adf.test on the ect, it
gracefully rejects the null hypothesis of I(1). 3. Sometimes I found test
statistics for both r <= 1 and r <= 0 are statistically enough. How should I
interpret this result? Do I have 1 cointegrated relationship or not at all?Any
explanation would be really appreciated. Thank you.- adschai
[[alternative HTML version deleted]]
______________________________________________
[email protected] mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
