The Motley Fool Discussion Boards
Investing/Strategies / Mechanical Investing
|Subject: Re: CAGR/GSD Limits of Blends||Date: 11/16/2007 8:12 PM|
|Author: emintz||Number: 203976 of 256291|
when you say “It is sufficient that the correlation be less than 1.”, I think you have the correlation coefficient in mind. The correlation coefficient is the Cov(X,Y) divided by the product of the standard deviations, stddev(X)*stddev(Y). Again, the correlation coefficient must be negative to achieve the goal of a blend (diversification).
No, it doesn't.
The correlation between screen 1 and screen 2 is 0.33. And yet the blend has a higher CAGR and a lower GSD.
Certainly negative correlation would be better, but it is not necessary for there to be a benefit.
|Copyright 1996-2015 trademark and the "Fool" logo is a trademark of The Motley Fool, Inc. Contact Us|