Measuring fund manager performance is not as easy as it sounds. Sure you know the basic measures: Sharpe Ratio, Treynor measure, and Jensen's alpha.
Sharpe: (Return-Risk Free)/ Standard DeviationBut each measure has its problems. For instance as far back as the early 1980s, financial economists (including Henricksson and Merton 1981) showed that the Sharpe Ratio could be a poor measure in the presence of "non-linear payoffs".
Treynor: ( Return-Risk Free)/ Beta)
Jensen's Alpha: Return = RF + Beta (Market Risk Premium) + Alpha
There really has been no good solution to this. Indeed in Robert Strong's Portfolio Construction, Management, and Construction Text, he concludes a discussion on the topic with a true, but unsatisfying statement:
"We have numerous analytical tools and we have a brain; we should use both in evaluating the performance of a portfolio."In Sharpening Sharpe Ratios, William Goetzmann, Jonathan Ingersoll, Matthew Spiegel, and Ivo Welch discuss the problems of the traditional performance appraisal methods and show that this is particulary troubling with the widespread use of derivatives (because their payoffs are inherently non-linear).
The paper then shows how these biases can be "gamed" by fund managers using options to take advantage of the limitations of the Sharpe Ratio. Once this is established, the authors develop a "manipulation-free" measure.
While on simplicity grounds alone, I doubt the new measure will be a big hit in undergrad classes, it definitely addresses the key attributes that a new measure should have. To understand the derivation, I recommend you read the paper. :) I tried to copy the equation in (equation 30) but I could not paste it and have top get to class now...sorry...
Suggested Citation
Goetzmann, William N., Ingersoll, Jonathan E., Spiegel, Matthew I. and Welch, Ivo, "Sharpening Sharpe Ratios" (November 2004). Yale ICF Working Paper No. 02-08; AFA 2003 Washington, DC Meetings. http://ssrn.com/abstract=302815
No comments:
Post a Comment