“Risk Everything” proclaims Nike’s latest ad campaign. The ad depicts a world in which the greatest athletes have been replaced by superhuman clones of themselves. After some moping, the true athletes recognize their only chance to beat the supposedly superior versions of themselves is to capitalize on their creativity to inspire moves unanticipated by their pre-programmed opponents. The approach that may appear risky is, in fact, the only way to win. For the athletes risking extinction, it is the safest play of all. Nike’s tagline is “there is no greater danger than playing it safe.” The same may be true in investing.

Based on recent results, it may appear that the case for active management is getting harder and harder to make. But that doesn’t mean it isn’t worth making. When conventional thinking leads to conventional results, unconventional thinking can be both prudent and rewarding.

The argument for passive or index investing is simple. Most active managers underperform their benchmark and index funds charge much lower fees than active managers. These two facts lead many investors to conclude that passive investing is practical, cost effective, better performing and even “safer” than active management.

The SPIVA scorecard (S&P Indices Versus Active Funds)¹ has served as a report card for the active versus passive debate. Tracking the annual returns of all large cap, mid cap and small cap funds each of the last ten years against their benchmarks, the S&P 500, S&P Mid-Cap 400 and the S&P Small-Cap 600, the average percentage of funds that underperformed their benchmark each year was 59%, 63% and 65%, respectively. Not particularly inspiring.

Another recent study, the S&P Persistence Scorecard², would seem to put an exclamation point on the side of passive investing. The study attempts to answer the question — if some managers outperform, are they able to do so consistently? Out of 687 funds in the top quartile of peer performance for the twelve months ended March 31, 2012, only 4% managed to stay in the top 25% for each of the next two twelve-month periods. Over five years, less than .5% (1 in 200) maintained their top quartile position each year.

But the SPIVA and S&P Persistence scorecards don’t tell the whole story. Active managers vary widely in strategy and tactics. There are identifiable subsets of active managers that have outperformed their benchmarks and have done so by meaningful margins (even after fees). Further, there are risks associated with passive investing that receive little attention. And finally, while the dearth of manager persistence in the S&P Persistence Scorecard may seem alarming, its relevance is questionable. Why does the top quartile of an arbitrary initial period, and a relatively short one at that, define the starting universe of potential top performers? And, why is performance over five consecutive twelve-month periods more important than performance over the entire five-year span? Most investors care more about the magnitude of return over time rather than making sure they are in the top quartile every sub-period along the way.

While indexing, by definition, eliminates benchmark risk and, by default, eliminates manager selection risk, it does nothing to eliminate volatility. Owning an index can actually have more risk than making active investment decisions, particularly in extreme environments when indexes may become distorted. In fact, in negative return years, truly active managers (defined below) tend to outperform by even greater margins through loss mitigation. In other words, when investors need protection the most, the seemingly “safe” choice may actually have more risk and provide less downside protection.

One reason for underperformance in down markets is because indexing does not involve price discovery, the process of evaluating the factors that affect market price. For the most common indices, like the S&P 500, the only thing that matters is the relative size of the component company. The bigger the market capitalization of a company, the more an index investor must own. Thus, indexation is a momentum based strategy. Investment dollars are allocated not on expected return but rather on the current price of a stock relative to others. The higher the proportional price of a stock in the S&P 500, the more index investors must buy with every marginal dollar invested.

A system based on momentum is inherently unstable, ultimately leading to booms, busts and back again. It forces concentration risk in companies or sectors that are in vogue because they have performed well. Index funds buy more as a stock’s price goes up and buy less as the price goes down. This is the exact opposite behavior than that of most fundamental investors and can lead to unintended consequences. For example, almost a third of the S&P 500 was invested in the energy sector in the 1970’s and the technology sector in the late 1990’s, and almost a quarter in the financial sector just before the financial collapse in the mid 2000’s. In each instance, the index became distended in a sector that turned out to be grossly overvalued.

There are many occasions when the safety and diversification sought through an index is an illusion and the accompanying risks can be substantial. Passive investors have a high probability of achieving average performance and virtually no chance of being above average. Sometimes “average” can be a very bad and painful result.

If index funds entail systemic risks and active portfolios underperform indexed portfolios, where should equity investors turn? It should not be surprising that active equity managers, on average, underperform their benchmarks. The aggregation of all active manager holdings into a consolidated portfolio would essentially resemble the benchmark, deliver benchmark-like returns, and underperform by the amount of their higher active management fees.

The average active manager loses to a low cost index fund, net of all fees and expenses. However, active managers are not all equal or average. They differ in how active they are and the type of active management they practice. These differences are measurable and informative.

Martijn Cremers and Antti Petajisto, while at Yale’s International School of Finance, developed a metric called “active share.” ³ Active share is the percent of a portfolio that differs from its benchmark index. Imagine an index with four holdings equally weighted at 25% each. If a managed portfolio is entirely invested in one of the four, its active share would be 75%, because only 25% of the portfolio is an exact replication of the index. A perfectly indexed portfolio would have an active share of 0%; active shares below 60% are generally considered closet or quasi-indexed portfolios; truly active managers have an active share of 80% or higher.

The Cremers and Petajisto research found that portfolios with higher active share significantly outperformed those with lower active share. More important, they found that active share was predictive of subsequent performance. A follow-up study by Petajisto⁴ found that the most active stock pickers have been able to add value by beating their benchmark indices by an average 1.26% per year after fees. In contrast, the average closet indexer underperformed, net of fees, by .91%, annually. The study covered the twenty-year period from January 1990 through December 2009.

These results indicate that stock selection skill as measured by high active share is rewarded in the marketplace and stock pickers, as a group, are able to add value for their clients, despite charging higher fees than index funds.

Prior to the introduction of active share, the industry relied primarily on tracking error (the standard deviation of the difference between a portfolio’s returns and its benchmark’s returns) to ascertain the extent to which a portfolio was actively managed. Essentially, tracking error measures the volatility of a portfolio that is not explained by the change in returns of its benchmark index. These differences in return are usually caused by factor bets. Factor bets can be as varied as overweighting or underweighting sectors, industries or regions, disproportionately favoring either small or large companies, or over or under-emphasizing a particular investment style such as growth or value.

Tracking error by itself has been found to have little, if any, correlation to portfolio returns, positively or negatively. As shown in the following chart, portfolios characterized by factor biases produce no discernible advantage or disadvantage before fees and negative returns relative to the benchmark after fees.

 

Tracking error associated with extreme factor biases actually negatively impacts performance. As the chart below shows, if the portfolios with the highest twenty percent of tracking error are excluded from those in the top twenty percent of active share, performance is thirty percent better than that of all portfolios in the highest quintile of active share managers. The first group is identified as stock pickers because performance is influenced primarily by stock selection. The second group includes portfolios such as sector funds where, despite a high active share, performance is overwhelmingly driven by a factor bias.

 

High active share defined as top quintile in fund universe, medium as middle 3 quintiles, low as lowest quintile; reduced factor exposure defined as not highest quintile tracking error

 

In addition to high active share, more concentrated portfolios have been found to have a positive influence on performance. Concentration is measured simply by the number of positions held in the portfolio. Logically, holding fewer but larger positions implies more conviction in each holding. Additionally, holding fewer positions may focus a manager’s attention, reducing the potential for complacency and enabling a deeper understanding of each security.

A study by Cambridge Associates⁵ shows that in multiple sub-categories and over long periods of time, concentrated portfolios outperform un-concentrated portfolios by a meaningful amount, net of fees (chart below). Concentrated US large cap strategies earned a return premium of 127 basis points annually, but were 12 basis points more expensive, implying an annual net return premium of 1.15%.

 

 

Concentrated portfolios for purposes of the above chart include those with forty or fewer holdings for the US Small-Cap and International categories, and thirty or fewer for US Large-Cap. We believe that for an unconstrained manager across geographies and market capitalizations, a portfolio of forty to fifty stocks would be relatively concentrated without inherently increasing volatility.

Another study by Baks, Busse and Green⁶ analyzed performance of over 2,000 funds from 1979-2003. They found that focused managers outperform their more broadly diversified counterparts by approximately 30 basis points per month or almost 4% per year. The Baks, Busse and Green study expanded the concept of concentrated managers to include those who took big bets in a relatively small number of stocks (regardless of the number of stocks in the total portfolio). They found that the top holdings of concentrated managers outperformed the top holdings of more diversified managers. Additionally, conviction in their largest holdings was rewarded by significantly better performance than lower weighted holdings. Thus, concentrated managers have some ability to correctly identify the relative attractiveness of stocks and correctly pick stocks.

Conclusions

• High active share is linked to benchmark outperformance. This is consistent with the notion that the only way to outperform an index is to be different from it.
• High active share is no guarantee that a manager will outperform, but it is more than likely an important condition for outperformance.
• A multi-manager approach putting together several active managers, each acting independently, will likely result in a low aggregate active share, and as a result, a potentially very expensive index-like fund.
• High active share managers with concentrated portfolios have an even higher probability of outperforming their benchmarks and peers.
• Investors can reduce the risk of severe underperformance by avoiding managers with very high tracking error.
• Combining insights about active share, portfolio concentration and tracking error does not guarantee success, but can help identify managers more likely to succeed.

Amy P. Raskin
Chief Investment Officer
Chevy Chase Trust

 

References

¹SPIVA U.S. Scorecard, S&P Dow Jones Indices, McGraw Hill Financial.
²S&P Persistence Scorecard, S&P Dow Jones Indices, McGraw Hill Financial.
³Cremers, K.J. Martijn and Antti Petajisto, 2009, “How Active Is Your Fund Manager? A New Measure That Predicts Performance”, Review of Financial Studies.
⁴Petajisto, Antti, 2010, Active Share and Mutual Fund Performance, Financial Analysts Journal.
⁵Ely, Kevin, 2014, “Hallmarks of Successful Active Equity Managers”, Cambridge Associates, LLC.
⁶Baks, Klaus P., Jeffrey A. Busse, and T. Clifton Green, 2006, Fund Managers Who Take Big Bets: Skilled or Overconfident”, Emory University.