SOS! When I run Johansen cointegration, I got rank deficiency...

losemind · Post by **losemind** » Wed Jul 20, 2011 4:14 am

SOS! When I run Johansen cointegration, I got rank deficiency...

I have three time series.

I ran johansen([x y z], 0, 1) in Matlab using the Spatial-Econometrics toolbox.

I got rank deficiency.

What shall I do?

Thanks a lot!

Baltazar · Post by **Baltazar** » Wed Jul 20, 2011 8:46 am

please stop creating several threads on the same subject.

please stop doing 2 sentences posts that indicate that you didn't first try to find an answer by using google or the search function.

please post in basic when it is basic.

now to that specific subject:

did you read on cointegration?

did you test your series were I(1)?

did you test the series weren't already pairwise correlated?

I suspect while you search the answer to these question, you'll understand that rank deficiency

CokeHead · Post by **CokeHead** » Wed Jul 20, 2011 9:00 am

Baltazar I do appreciate the fact that you are really trying to help and rescuing this train wreck. Do you think it makes sense for admin to delete these threads as we already have a few cointegration threads going and this is just taking away from the dialogue.

Losemind, I also understand that you are trying to learn, but when you create these threads without doing appropriate research you just degrade the signal to noise ratio on this forum. Use the search button, check google, check other finmath forums and if you think your question is still interesting, please do air your views. Currently these are just homework questions. Thanks. Blendertime

Baltazar · Post by **Baltazar** » Wed Jul 20, 2011 9:06 am

I fully agree with CokeHead.

Cheng · Post by **Cheng** » Wed Jul 20, 2011 9:06 am

Blendertime

Reactor Core · Post by **Reactor Core** » Sat Jul 23, 2011 2:26 pm

This thread was moved from [url=/Show%20Forum.aspx?ForumIDKey=2]Trading[/url] to [url=/Show%20Forum.aspx?ForumIDKey=12]Trashcan[/url].

Reason: Thread elicited too many negative ratings.