Home ListingApril 14, 2015

The Problem with Active Share

Two academics have hit back after their popular measurement for active management came under scrutiny.

Reported by Featured Author

AQR has challenged the recent trend towards investors prioritizing “active share” as a measure of how active an equity manager is.

A white paper published this month by the fund manager cast doubt on the popular measurement, which is designed to highlight by how much a manager’s portfolio deviates from its benchmark.

“If investors care about how their mutual fund performs relative to the benchmark, active share matters.”—Martijn Cremers, University of Notre DameBut the authors of the original research—Martijn Cremers, of the University of Notre Dame, and Antti Petajisto, formerly of New York and Yale universities and now a vice president at BlackRock—have argued in turn that AQR actually reinforced some of their findings.

“Just because a portfolio claims to have high conviction and thus looks a lot different than its benchmark doesn’t mean it should perform better than its benchmark,” said AQR co-founder Cliff Asness, commenting on his colleagues’ research in his blog.

The company’s report used the same data as the original research that made active share popular. Cremers and Petajisto in a 2009 paper reported a positive correlation between their new measure and outperformance.

However, AQR’s Andrea Frazzini, Jacques Friedman, and Lukasz Pomorski wrote that the original research was “subject to misinterpretation”.

“Just because a portfolio claims to have high conviction and different to its benchmark doesn’t mean it should perform better.”—Cliff Asness, AQRThe authors argued in their paper “Deactivating Active Share” that funds with high or low active share “systemically have different benchmarks”—in other words, funds are more likely to measure highly on the scale if they invest in small caps, and low on the scale if they buy large caps.

“Controlling for benchmarks, active share has no predictive power for fund returns,” the authors wrote, “predicting higher performance within half of the benchmark indexes and lower fund performance within the other half.”

But Cremers and Petajisto both told CIO AQR’s paper had it wrong—and actually backed up their link to performance.

“They were able to replicate the main results in our paper,” Cremers said, “showing strong evidence that active share has been strongly predictive of mutual fund performance once you adjust for the benchmark performance.”

Adjusting for benchmark performance was “valid and important,” he said, and was acknowledged “clearly” in his work with Petajisto.

“As a result, if investors care about how their mutual fund performs relative to the benchmark, active share matters, according to both our as well as their results,” Cremers said.

Petajisto referred to a 2013 paper in which he had tested the active share measure by adjusting for style. While the AQR authors stated that “performance predictability can be explained by a bias towards the small-cap sector,” Petajisto said his previosu research had found that “the active stock picker group outperforms the closet index group by a significant margin,” more or less regardless of style.

There was one area all sides agreed on, however: fees. Frazzini, Friedman, and Pomorski said active share “could be useful for evaluating costs” when used alongside other measures, such as tracking error. The indicators can show how much true active management investors getting for their fees.

AQR’s paper “Deactivating Active Share” can be found on the company’s website.
Both Martijn Cremers and Antti Petajisto have published several papers in support of active share, including “How Active is Your Fund Manager? A New Measure that Predicts Performance” (Cremers & Petajisto, 2009), and “Active Share and Mutual Fund Performance” (Petajisto, 2013).