Limitations Of Active Share As A Measure Of ‘Activeness’

As investors’ low cost options continue to expand and get more sophisticated (okay, sometimes to a fault) active managers have to work harder to justify their fees. One way that managers can’t prove that they’re not just closet indexers is by having a high active share – a measure of how much a portfolio differs from the chosen benchmark index. But investors should be careful not to rely too much on this single measure because it’s not perfectly comparable between different portfolios.

“While active share can give some indication of the level of a portfolio’s activeness, it should be used as a complement to other measures, such as tracking error and specific risk,” writes Axioma senior director of research Dieter Vandenbussche. “Too many factors have an impact on active share to allow it to be the single measure of whether managers are earning their fees.”

Active share doesn’t tell you how much alpha a manager generates

One (fairly technical) example that can influence active share is the kurtosis of a portfolio’s alpha. A portfolio with high kurtosis is more likely to perform in line with the index than one with low kurtosis, but it’s also more likely to have extreme events that vary a lot, which feeds into the portfolio’s active share. The reason this is a problem is that it’s the alpha that you are paying for when you hire an active manager, and in the three scenarios below each distribution has the same alpha.

Active share is also affected by the small cap bias that is common to long-only active managers. The reason for the bias isn’t that active managers love to invest in small caps (although some probably do), it’s just that you can’t be very underweight small-caps – the market weight itself is already tiny (long/short managers can get around this with short positions). Vandenbussche found that this translates into lower active share for investors who specialize in large caps.

Concentrated benchmarks with targeted tracking error force a lower active share

The choice of benchmark obviously affects active share since it’s literally what we measure a manager’s portfolio against, but when you throw in a target tracking error things get even more complicated. For a concentrated index like the S&P 500 active managers won’t be able to over- or underweight very much without going past the target tracking error. But it’s possible to get a much higher active share with the same tracking error against a less concentrated index (eg Russell 3000) basically because there are more levers for the manager to work with.

Of course a high active share doesn’t even attempt to tell you whether managers are any good, but it’s not a perfect measure of how active/passive they are either.