Hello,
I'm trying to compute a correlation matrix for a number of series, and then take, say, 100 at least correlated series. Is there any way to achieve this? In other words, how can I take names of those 100 series which are at least correlated among the entire sample?
Best
Correlation
Moderators: EViews Gareth, EViews Moderator, EViews Jason, EViews Matt
-
EViews Gareth
- Fe ddaethom, fe welon, fe amcangyfrifon
- Posts: 13600
- Joined: Tue Sep 16, 2008 5:38 pm
Re: Correlation
what does "and then take, say, 100 at least correlated series" mean?
-
mr.interested
- Posts: 16
- Joined: Wed Jul 14, 2010 5:57 am
Re: Correlation
OK, so let's say I have 10,000 series + 1 index, and I want to create a portfolio of 100 series which have the lowest correlation with the index. As I understand, I would have to create a correlation matrix, and then take names of 100 series that have the lowest correlation with the index. How can I do that?what does "and then take, say, 100 at least correlated series" mean?
-
EViews Gareth
- Fe ddaethom, fe welon, fe amcangyfrifon
- Posts: 13600
- Joined: Tue Sep 16, 2008 5:38 pm
Re: Correlation
I think I would loop through them calculating the covariance once at a time, recording the covariances in to a column of a matrix. Then you could sort the matrix and take the top (or bottom) 100.
-
mr.interested
- Posts: 16
- Joined: Wed Jul 14, 2010 5:57 am
Re: Correlation
Thank you Gareth.I think I would loop through them calculating the covariance once at a time, recording the covariances in to a column of a matrix. Then you could sort the matrix and take the top (or bottom) 100.
Who is online
Users browsing this forum: No registered users and 2 guests
