Style Analysis: A Holding-Based Microscope or Return-Based Telescope?
May 22nd, 2007 | Filed under: Performance, Analytics & MetricsWith the release of Andrew Lo’s recent paper on a new way to calculate alpha, this article by Edhec on style analysis seems apropos. As you may recall, Lo argues that the true measure of active management is a manager’s ability to correctly forecast security price movements and to express those forecasts in the portfolio. The key metric, therefore, is the correlation between portfolio positions and price movements.
This is clearly a “holding-based” approach to analyzing manager value-add that differs from the traditional returns-based approach which infers the make-up of a portfolio from the way it responds to exogenous variables. The return-based approach was, of course, first popularized by William Sharpe late last century (for modern examples of such detective work, see this Amaranth case study or this Fidelity Magellan case study).
While return-based analysis is computationally easier, it has come under attack recently as a somewhat blunt instrument. In its place, the infinitely more granular holding-based approach has captured the limelight. To be sure, holding-based analysis is elegant and is being adopted by legions of funds of funds, and other institutional investors who have position-level transparency into the funds they own. But is it infallible?
To continue reading this article please login (at the right) or click here to learn more about accessing our archives.
I think part of the problem is a need to do “style” analysis. Style boxes are a bad simplification and have no true economic significance. I do agree that managed account platforms provide a lot of benefit for investors where it really matters.