Institutional Perspectives

Key Takeaways

• The growing use of active share to evaluate active managers, as well as criticism of the measure, has made it important for investors to recognize its strengths and shortfalls.
• Active share on its own is not a predictor of positive performance, and investors should be aware that high active share strategies are likely to produce wide return dispersion.
• Excess returns are difficult to achieve without the distinct performance generated by high active share strategies.

Introduction

Since publishing our original report on active share in 2014, the measure's usage has grown among investment managers and asset owners and drawn questions about its utility as a singular predictor of performance. ClearBridge views active share as a useful, but not exclusive, statistic to help investors understand the sources of active risk taken by active managers and to validate that active managers are providing differentiated portfolio positioning compared to their benchmark. In this update, we address the risks associated with a high active share approach and how to set proper performance expectations for active strategies.

"We believe placing too much emphasis on the predictive ability of this single statistic has caused investors to pursue high active share strategies for the wrong reasons."

Investor preference for active over passive management relies on the ability of active portfolio managers to consistently add value through distinct performance patterns and measurable alpha. Truly active strategies stand out from active strategies that hew closely to their benchmarks as well as strategies that allocate securities based on risk or style factors by generating unique return streams that result from individual stock selection. In 2009 academic research that led to the development of the active share measure¹, finance professors Antti Petajisto and Martijn Cremers found that active managers whose portfolios bear the closest resemblance to a benchmark fail to deliver excess returns net of fees. This inability to deliver differentiated performance for the management fees charged suggests that certain active strategies fail to serve a useful purpose in an allocation.

Active Share Does Not Measure Skill

While the definitive work on active share occurred six years ago, the measure has more recently gained wider acceptance among asset managers and investors. The growing popularity of active share has also generated criticism from some in the institutional community who question its predictive power. Critics point out that just as a high active share approach has the potential for outperformance relative to a benchmark, it also can lead to significant underperformance. They do allow, however, that active share can be useful in helping investors determine whether the fees they are paying for active management are justified.

Active share can be a valuable tool in deconstructing the sources of a portfolio's dispersion and determining the role of stock selection in an investment process. More recently however, active share has been incorrectly positioned as a sole gauge of manager or investment skill, and we believe the asset management industry has underplayed an important additional measure - tracking error - that should be used alongside active share to give clients an appropriate set of return expectations relative to a passive, market cap weighted benchmark. We believe placing too much emphasis on the predictive ability of this single statistic has caused investors to pursue high active share strategies for the wrong reasons. Our reading of the original active share study is that the best chance for excess returns are from "diversified stock picking" approaches which are high in active share with moderate tracking error.

Used in conjunction with other metrics, active share can act as a thermometer to distinguish genuinely active managers from those who more closely resemble passive indexers. The percentage of active managers considered "closet indexers" grew from 1% of the mutual fund study universe in 1980 to nearly 30% by 2009. Closet indexers - funds with active share below 60% - represent one of five categories into which Petajisto segments active managers by active share percentile. The two active investment styles with the highest active share, diversified stock pickers and concentrated managers, produced the best risk-adjusted returns over a 20-year period from 1990-2009 (Exhibit 1). Active share can help reduce the bulging active universe to these two investment styles for further analysis.

Exhibit1: Excess Returns. Alpha prior to fees/transaction costs (%), 1990-2009

Source: Antti Petajisto, Active Share and Mutual Fund Performance, July 2013. Past performance is no guarantee of future results. This information is provided for illustration purposes only and does not reflect the performance of an actual investment. Mutual funds, are sold by prospectus only, and typically contain fees and expenses. If fees and expense were included in this example, performance would have been lower.

Key Definitions

• Active Share: The percentage of stock holdings in a portfolio that differ from its benchmark index. Active share takes into account actual portfolio holdings and their weightings relative to the benchmark and provides a snapshot of stock overlap. Active share of 0 implies that an investment portfolio is identical to its benchmark while active share of 100 indicates that a portfolio and its benchmark have no holdings in common.
• Tracking Error: The variability of a portfolio’s performance compared to its benchmark index. It calibrates the degree of volatility of a portfolio’s monthly returns versus that of an index. If a portfolio outperforms or underperforms its benchmark by the same amount each month, its tracking error would be 0.

Putting More Perspective Around Active Share

Investors must be careful not to be swayed by an implied correlation between active share and excess returns or other common misconceptions surrounding the metric. Within the subset of diversified stock pickers, not all high active share managers are created equal. Investors must ask where a manager's active bets are coming from: are they purely delivered by stock picking and concentration within the universe defined by investment guidelines or through other measures? Moves such as investing outside a strategy's asset class can result in style and market cap drift and artificially inflate active share. As Petajisto noted in his study, "Both the type and degree of active management matter considerably for performance.

Coincident with the potential for alpha generation is potentially higher volatility. By their nature, high active share strategies rely on return sources that may be independent or bear a low correlation to the sources driving benchmark performance. As a result, high active share strategies are susceptible to underperformance of a greater magnitude than active strategies that fall in the categories of moderately active or closet indexers. Investors seeking to benefit from the stock-picking ability of skilled high active share managers should be aware of the volatility profile of these strategies and be willing to accept meaningful quarter-to-quarter deviation in returns and periods of short-term underperformance in pursuit of longer-term outperformance. New data from Morningstar tracking the return patterns of top decile mutual funds from 2005 to 2014 found that the best performing funds (domestic and international) over the 10-year period spent more time in the bottom calendar-year performance quartile than funds with inferior 10-year records.² In other words, excess returns achieved by the highest of high active share strategies over the long term are most often the result of periods of significant outperformance and underperformance, whereas the objective for active share strategies with moderate tracking error is consistent, above-average performance.

Beyond purely looking at active share as an end-all, be-all metric, we believe investors should include the following additional measures when evaluating equity strategies:

• Tracking error
• Concentration level of top 10 holdings 
• Sector deviations from benchmark
• Market cap deviations from benchmark 
• Out of country/region/index exposure 

Further, investors should scrutinize the benchmark used to determine active share as the fewer constituents in a benchmark, the more difficult it is for a manager to achieve high active share. A strategy's active share should also be evaluated relative to its peer group, which can reveal distortions from some of the approaches mentioned above.

 

Exhibit 2: Active Share and Tracking Error (3 years as of Nov. 30, 215) Active share, when combined with other measures including tracking error, can help investors gauge the degree of a strategy's active management

Data as of Sept. 30, 2015. Source: FactSet. Active share measured versus the following benchmarks: Aggressive Growth vs. Russell 3000 Growth, International Small Cap vs. MSCI EAFE Small Cap, International Value vs. MSCI All Country World ex-US, Large Cap Growth vs. Russell 1000 Growth, Mid Cap Core vs. Russell Midcap, Mid Cap Growth vs. Russell Midcap Growth, Small Cap Growth vs. Russell 2000 Growth, Value Equity vs. S&P 500.

High active share and tracking error - aspects viewed negatively by consultants and gatekeepers as recently as five years ago - are now considered desirable attributes of an active manager. Indeed, active share has raised expectations for all active managers. Investment managers with active share measures from 70% to 90% are considered "high active share," according to the Petajisto study. At ClearBridge, by allowing bottom-up stock selection to guide portfolio construction, our high active share strategies provide measurable differentiation from both passive alternatives and active strategies with low active share (Exhibit 2). The differentiation we provide as active managers can be quantified not only by active share but also sector deviation and the degree of conviction in our best ideas (Exhibit 3). We believe this measurable differentiation is necessary to generate excess returns over the long term.

 

Exhibit 3: Equity concentration and sector weightings reflect ClearBridge’s active stock selection and high conviction

Source: FactSet. Data as of Nov. 30, 2015.

Conclusion

Five years ago, few clients used active share as an evaluation tool. As with many new metrics across a host of industries, such as on base percentage becoming a leading statistic in baseball thanks to the book Moneyball, the term, if not given additional context, can then become "overvalued" as a predictor of efficacy. ClearBridge acknowledges the greater performance dispersion that high active share strategies can generate and places significant emphasis on risk management in our investment process to ensure we are taking risks that are well understood. We believe active share remains a sound tool for benchmarking the degree of active management investors receive for fees paid and the ability of active strategies to generate excess returns over longer time periods. Despite the greater acceptance and growing criticism of active share in the marketplace, ClearBridge has never managed our strategies to achieve a baseline measure of active management. Instead, we remain focused on the same, straightforward approach that we have been providing clients for over 50 years: active stock selection that is driven by proprietary fundamental research and results in high conviction, low turnover portfolios.