Go Active or Go Home

In 1993, the late political scientist and U.S. Senator Daniel Patrick Moynihan wrote an influential essay on the dangers of becoming accustomed to increasing levels of crime and social deviance. He called it "Defining Deviancy Down."

In the mutual fund world, an opposite situation has been unfolding--there's now greater tolerance, perhaps even a desire, for less deviant behavior or active management on the part of fund managers. As the number of funds has expanded, more and more of them do little but conform tightly to an index: 41% of all mutual fund assets in 2003, as opposed to 2% in 1980, according to the study highlighted below. The typical "actively managed" fund today is still technically actively managed, but it no longer deviates enough from the index to justify its fees.

Happily, recent academic studies, including one devising a metric called "active share" that compares a fund's holdings with those of its best-fit index, are making it more difficult for those who sell and promote funds to essentially define deviancy up by calling index-huggers actively managed. Of course, continuing with the comparison to Moynihan's essay, becoming increasingly invested in index-hugging funds is not as serious as becoming more accepting of crime. But it's oddly similar in that it's a little bit like accepting having your pocket picked occasionally. Deeming index-hugging acceptable can end up costing many investors substantial sums in unnecessary fees. (Morningstar's director of mutual fund research, Russel Kinnel, also completed an active share study of his own recently.)

In this piece, we'll review some recent academic research on active share and highlight how Morningstar's nonindex domestic large-cap Analyst Picks independently corroborate the findings of recent active share studies. Our picks, derived from a variety of fundamental factors, confirm the findings of active share studies, that high deviation from the index wins over the long run. We chose to examine our large-cap Analyst Picks because large-cap stocks may be reasonably compared to the S&P 500 Index, even if that's not precisely the best-fit index in each case, and because most investors have a large-cap fund as a core holding.

The Research Behind Active Share
K.J. Martijn Cremers and Antti Petajisto of the Yale School of Management have conducted primary research into active management and defined active share. Cremers and Petajisto studied portfolios from 1980 through 2003, and discovered that those displaying the greatest active share tended to outperform their relevant indexes both before and after expenses. In other words, a high active share was a good predictor of strong performance regardless of fees. This is impressive given how Morningstar's own expense studies have shown the predictive quality of fees.

Of course, by definition, not all funds that deviate from the index can outperform it. However, Cremers and Petajisto discovered that funds that deviated only mildly from the index--those we might call index-huggers, or closet-indexers--tended to underperform the index. So, more deviation was predictive of outperformance, while slight deviation was predictive of underperformance.

Similarly, a 2008 study titled "Best Ideas" by Randolph Cohen at the Harvard Business School, Christopher Polk at the London School of Economics, and Bernhard Silli at Goldman Sachs defends high-conviction portfolios. The authors studied managers' favorite stocks and showed that, in fact, those stocks outperformed funds' appropriate indexes. The study concluded that U.S. stocks are not efficiently priced and that many managers are able to identify stocks that will outperform--but that their funds' returns don't fully reflect that ability because managers are hampered from emphasizing their best ideas in meaningful ways.

Indeed, for a professional money manager, underperforming over the short term while looking mostly like your peers is tolerable to your employer; underperforming while deviating significantly from the index isn't. Maintaining job security and attendant institutional factors, including pressure from consultants and advisors, often push active managers to mediocrity.

Where Morningstar and Active Share Meet
Although Morningstar analysts don't rely heavily on active share when we debate and choose our Analyst Picks, our picks confirm the thesis that high active share is highly correlated with good performance. When we consider funds to be Analyst Picks, we look for strategies that have a high likelihood of beating their peers and index by an appreciable amount over at least a full market cycle. These strategies must have a durable competitive advantage, which often means a penchant or willingness to hold stocks when they're generally out of favor or hold cash when bargains are unavailable. Many of our Analyst Picks have spells of short-term underperformance while posting superior longer term results, the latter of which hold great weight in our deliberations.

Not coincidentally, these funds rarely look or behave like their benchmark indexes--which is one reason we think they have a good chance of succeeding over the long haul. In other words, our Analyst Picks have high active share numbers, even though, depending on the fund, that specific figure played little or no role in the selection process. Collectively, our nonindex domestic large-cap equity Analyst Picks have an average active share of 76%. The median active share of these 31 funds is 80%. (Both numbers would be even higher if we excluded  Vanguard Tax-Managed Growth & Income  and  Vanguard Tax-Managed Capital Appreciation , which, though not technically index funds, essentially try to match index performance and thus have active shares of less than 25%.) By contrast, the median active share scores for Morningstar's domestic large-cap categories range from 66% to 70%, and Cremers and Petajisto consider index-huggers to be funds with active shares of 20% to 60%.

 Click here for the active share data for all our nonindex large cap domestic Analyst Picks.

Beyond these two Vanguard Funds, none of the domestic large-cap Analyst Picks have markedly low active share scores. In fact,  American Funds Fundamental Investors (ANCFX) and  American Funds Washington Mutual (AWSHX) clocked in with the next-lowest Active Share metrics at 60% and 59%, respectively.

Meanwhile,  Longleaf Partners (LLPFX) has the highest Active Share at 98%, with  Fairholme (FAIRX) and  Sequoia (SEQUX) not far behind, both with 96%. Although these three funds have different investment processes that lead them to unique holdings and portfolios, they all strive to achieve positive absolute returns and not suffer permanent capital impairment instead of participating in what hedge fund manager Seth Klarman calls "the short-term relative-performance derby." They're not afraid to look different from their peers and the index even if it means short-term periods of underperformance, which they've all experienced or will almost certainly experience.

Conclusion (and One Caveat)
Funds with high active share scores won't always outperform, especially over short periods of time, and those with lower-than-average active share scores can sometimes succeed, as shown by the strong long-term performance of the two American funds mentioned above. But funds with high active share scores stand a better chance of outperforming over longer periods of time than funds with only a modicum of active share. Funds with high active share scores outperform by so much that they even tend to overcome the bad effects of high expenses in the cases of funds that are burdened by them.

Of course, funds with high active share scores aren't for everyone, and high active share isn't a blanket guarantee of good absolute performance. The risk of underperforming the index may be too great for some investors to bear, even over a short period of time, and we continue to think indexing is a perfectly reasonable investment strategy. Investors should also remember the painful 25% 10-year cumulative loss through June 30, 2010, of  Janus Twenty  (which is not an Analyst Pick but has an active share of 87%), when they think about the gaudy 226% 10-year cumulative return of Fairholme over the same period. (The S&P 500 Index dropped 15% over that time.)

Also, investors should keep in mind the caveat that the major components of the index haven't looked as cheap on a price/cyclically adjusted earnings basis as they do now in a decade or more. In other words, now may not be the worst time for active managers to own some of the index's top holdings, even if, for the time being, that makes their portfolios look more like the index.

Ultimately, though, if you choose the active route, it's worth considering the studies that show greater deviation from the index is better than only a modicum of it. And investors need to be more vigilant, given the massive proliferation of index-hugging funds since 1980.

In the end, if you're going to choose to pay up for active management, you may as well get it.

John Coumarianos co-authored this article.