Hi,I came across this blog entry:http://autospreader.wordpress.com/2010/ ... ucture/And found it (and the rest of the blog) fairly interesting. I am trying to analyse some other curves in the way the ED is done and I can't understand the 3rd graph, covariance. I have a grid of relative month settles (M1,M2,M3 etc.) by date, I can work out the correlation (M1/M2, M1/M3, M1/M4 etc) and Std dev of returns easily enough, but then he graphs covariance.... Is that the covariance of correlation to s.d.? if so that only yields one number, not a graph of them. Or is it covariance of M1/M2, M1/M3 etc? Any help will be very much appreciated!Cheers

in a comment at bottom of the post autospreader explains the graphs depicts the diagonal of the covariance matrix.

Yeah, I could have been more informative there.....my apologies.

ah ! henderson, you're Autospreader man. i wondered who it is.