The following letters are in response to our article Luck versus Skill in Active Mutual Funds, which appeared on August 5, 2008.

Response to:
Luck versus Skill in Active Mutual Funds
Advisor Perspectives, August 5, 2008

C. Thomas Howard, PhD
Professor, Reiman School of Finance
University of Denver
and
CEO and Director of Research
AthenaInvest, Inc.

The Barras, Scaillet, and Wermers (BSW) paper featured in this AP article builds on a long line of studies that conclude active equity managers add no value net of management fees. It also provides additional depressing evidence that manager skill (as defined by a positive alpha) seems to have all but vanished in recent years.

What is surprising to me is that fund performance results point in exactly the opposite direction. Using a more inclusive 1980-2008 active US equity open end mutual fund data base than did BSW, I find that:

• The average fund alpha is characterized by an upward sloping time trend.
• Currently, my best estimate is that the average active US equity fund outperforms the S&P500 by roughly 100bp after fees. So, unlike times past, it is now a rational decision to invest in the average active manager rather than in an index.
• As the average fund alpha has increased over time, the average fund standard deviation has decreased. Thus alphas are improving while risk is declining.
• Over the first six months of 2008, a difficult market environment indeed, the average after fees fund alpha was a whopping 463bp (annualized).

These observations point to increasing managerial skill among active US equity managers rather than decreasing skill as claimed by BSW. It also argues for investing in a portfolio of active managers rather than in an index.

Evidence and Data

My data is all active US equity open end mutual funds that existed during any month over the 28 1/2 year period January 1980 through June 2008 as reported by Thomson Financial. I exclude index, mixed asset, target date, and allocation funds. Monthly fund returns, net of automatically deducted management and other fees, are calculated by averaging the returns across all share classes existing in that month. The final sample is comprised of 4,207 funds and 482,443 fund/month observations. At the end of 1982 there were 300 funds, with the number increasing to a peak of 2,617 at the end of 2003, and then declining to 2,117 as of June 2008. Thus, there has been a roughly 20% drop in the number of funds since 2003.

Annual alphas, based on S&P500 benchmark returns, are calculated by summing the average monthly fund alphas within each year. The resulting annual alphas are shown below, along with the alpha time trend.

* For 2008, annualized sum over first 6 months

Average annual alphas are highly volatile, with some years very negative and others very positive. The average alpha over all fund/month observations is -82bp, similar to what others have found. However, the time trend is positive, with a slope of 14bp annually. The implication is that managers are becoming more skilled over time.

Display article as PDF for printing.

Would you like to send this article to a friend?