The notion that truly active portfolio managers make meaningful bets against their market-based benchmarks is far from new. This concept has been discussed for at least two decades – or probably longer.
Martijn Cremers and Antti Petajisto popularized this concept in their 2009 paper, “How Active is your Fund Manager? A new measure that predicts performance,” which was published in the Review of Financial Studies.
That paper introduced the investment industry to Active Share (AS), which measures the extent of a fund’s active bets based on differences in stock weightings vs a benchmark. The paper concluded:
– AS is unique, in that it measures the extent of active management based on the fund’s stock positions (whereas “tracking error” – the volatility in monthly differences between fund and benchmark returns – aggregates many risk exposures).
– AS is better than tracking error in distinguishing between active funds and closet indexers.
– High AS funds tend to outperform their respective benchmarks.
These conclusions were unchanged in Petajisto’s updated research in 2013. But both his papers have been challenged.
In a 2012 examination, research from Vanguard Group Inc. found no clear evidence of AS’ predictive power. This research found that most of the high AS funds are small-cap equity funds and found a wider dispersion of returns for higher AS funds. In other words, higher AS funds had a greater range of best-to-worst performers – making the tool more likely to pick a disappointing fund.
Finally, Vanguard’s research found that AS is a good measure of active management, but the tool must be combined with careful qualitative analysis for a complete assessment.
In “Deactivating Active Share” in the April 2015 issue of the Financial Analysts Journal, Andrea Frazzini, Jacques Friedman and Lukasz Pomorski used the same data and methodology that Cremers and Petajisto did in their original 2009 paper. Frazzini et al calculated the same basic results as Cremers and Petajisto: that high AS funds outperform low AS funds. But the co-authors of the 2015 paper found that after accounting for differences in benchmark returns, the return differences between low and high AS funds disappeared. In other words, Frazzini et al claimed that AS is useless as a predictive tool. Petajisto then wrote a spirited defence – mostly in response to the Frazzini et al paper.
Having read the views of all sides, here is some food for thought:
AS is overused as a stand-alone active-management selector, but still is a good way to distinguish between closet indexers and truly active managers. But to find future outperformers, AS must be combined with a thorough assessment of a money manager’s structure, philosophy, processes, costs, policies, performance and fit with clients’ needs. There’s no secret sauce. There’s only smart use of data, a focus on characteristics linked to future success and a diligent analytical process.
Dan Hallett, CFA, CFP, is vice president and principal with Oakville, Ont.-based HighView Financial Group, which designs portfolio solutions for affluent families and institutions.
© 2016 Investment Executive. All rights reserved.