Active Management Really Works
June 23, 2009
“Active Share”
In 2007, the Yale University School of Management published an important paper called “How Active Is Your Fund Manager? A New Measure That Predicts Performance,”1 by Martijn Cremers and Antti Petajisto, which has taken the air out of previous studies that say active management doesn’t work. As the authors explain, the other studies comparing fund managers to benchmarks rely on the traditional method of evaluating how funds perform compared to benchmark returns, which is to use tracking error (the amount of the volatility between a fund’s return and its passive benchmark return) as a guide. But tracking error does not explain whether fund performance is impacted by stock selection or by sector or industry selection.
Instead of merely analyzing tracking error, the authors of the Yale study did something entirely different: They compared the active fund portfolio holdings to its benchmark index holdings. Based on this comparison, they concluded that if a fund overweights a fund holding relative to the benchmark holding, it is considered to have an active long position in the stock. If a fund underweights an index stock or doesn’t own it at all it implicitly has a short position in that stock. With this methodology they then show that funds own 100% of the benchmark index stocks, plus or minus the implied long and short positions. They call the net total of the implied long and short positions the “Active Share” of a fund. Since mutual funds almost never take actual short positions, they find that the Active Share of funds is always between 0% and 100%. Funds with Active Share greater than 80% are considered “stock pickers.” Funds that are non-index funds with an Active Share less than 60% are considered to be closet indexers. Funds with Active Shares less than 20% are Index funds.
Cremers and Petajisto came to these conclusions after analyzing massive amounts of data. They analyzed individual stock holdings and monthly and daily returns for three different families of index funds, including the S&P/Barra, Russell, and Wilshire Indexes. They then collected an exhaustive amount of data on the universe of mutual funds, including the individual holdings of each fund and the monthly and daily returns of the funds from a variety of databases. In the end, after thoroughly scrubbing the fund data to make sure that they had a reliable sample of funds, they studied the Active Share of 2,650 funds in the period 1980-2003. Because they compared the individual holdings of the funds to the individual holdings of the benchmark indexes, they were able to choose the best possible benchmark for each fund by coming up with the closest match between fund holdings and benchmark holdings. This resolved the problems with similar studies where critics argued that active managers were not being compared to an accurate benchmark, thereby invalidating the results of the study.
The Yale study conclusions about active management are startling. They show that active management should be measured in two dimensions: tracking error and Active Share. Tracking error measures the volatility of portfolio return around a benchmark index, whereas Active Share measures the deviation of portfolio holdings from the holding of the benchmark index. The study’s methodology identifies different types of active management: diversified stock picks, concentrated stock picks, sector rotators (factor bets), closet indexing, and pure indexing.
Figure 7.1
2 Note: Figure 7.1 shows different types of active and passive management, as revealed by a two-dimensional picture. Active Share represents the fraction of portfolio holdings that differ from the benchmark index, thus emphasizing stock selection. Tracking error is the volatility of fund return in excess of the benchmark, so it emphasizes bets on systematic risk. Active funds with concentrated portfolios of individual stocks have the highest exposure to Active Share and tracking error. Factor bets represent active managers who make bets on systematic risk relative to the index. They could be overweighting value versus growth, large versus small-cap stocks, or have an index beta different from 1.
Display article as PDF for printing.
Would you like to send this article to a friend?
Remember, if you have a question or comment, send it to .
