A recent deal between the New York state attorney general’s office and major investment firms highlighted the role of “active share” in evaluating mutual funds or other pooled investment vehicles. While the New York deal, in which 13 firms representing roughly 400-plus funds agreed to publish active share data for their funds, was aimed at benefiting retail investors, the announcement is a timely reminder to institutional investors about evaluating the role active share plays in their portfolios.
Active share is a holdings-based measure of how different a portfolio is from its benchmark index. The origins of the metric stemmed from research by Martijn Cremers and Antti Petajisto at the Yale School of Management, who introduced the metric in 2006. In 2009, the duo published their seminal article, “How Active Is Your Fund Manager? A New Measure That Predicts Performance.” Cremers and Petajisto concluded that funds with the highest active share outperformed their benchmarks after fees, while funds with the lowest active share underperformed their benchmarks. Furthermore, Cremers and Petajisto point to the persistence of outperformance among high active share managers. In a bold statement, the authors argued that active share should be used as a litmus test to predict future performance.
Callan began incorporating active share into our analysis of investment managers in 2011. As with any judgment that relies on quantitative and qualitative information, active share needs to be assessed in a broader context. At Callan, active share primarily serves as a risk assessment tool (i.e., could be an indicator of expected tracking error). The more different a portfolio is from its benchmark, the more different its return may be. We also utilize the metric as a “sanity check” when building multi-manager portfolios to protect against inadvertently over-engineering what ends up being an expensive index fund.
Unlike retail investors, institutional investors have the resources to rely on investment staff or investment consultants to identify best-in-class active managers. But it is easy to have high conviction in a strategy before money is invested; successfully using active share in building a portfolio requires patience. Performance rough patches occur. A high active share portfolio, particularly one with high security concentration, may be hit with idiosyncratic risks (e.g., Valeant, Volkswagen), testing the conviction of even the heartiest investors.
Defined contribution (DC) plan providers should also closely consider whether to use active share in building investment menus. Empirical evidence points to performance-chasing behavior among DC participants (i.e., buy high and sell low). In a DC investment menu, a high active share (and high tracking error) portfolio may not be suitable as patience, conviction, and a strong stomach may be required. Instead, a fund with a more moderate level of active share may be a better strategy.
One final word of caution: beware the reference benchmark. Active share does not “beta adjust.” High active share may come from non-index securities and reflect systematic factor tilts such as a capitalization, style, or regional bias. To the extent these biases dramatically influence the risk profile of the fund, active share may mask the source of return (i.e., factor bets as opposed to security selection).
Since 2009, even Cremers and Petajisto appear to have capitulated on their position, publishing in the Financial Analysts Journal in 2017 and 2013, respectively, more nuanced takes on using active share.
The bottom line here: the disclosure of active share has the potential for misuse if it leads investors to invest in strategies for which they do not have the wherewithal to endure. Successfully using active share requires:
- A full understanding of what active share does—and does not—indicate about a strategy
- A holistic assessment of the total portfolio to avoid inadvertent indexing
- Patience and a long-term investment focus
