It’s one of the most documented findings in finance: the majority of equity funds underperform stock markets. However, defenders of active management say the statistics are distorted by closet index funds containing so many positions as to make their performance almost indistinguishable from benchmark indices. Can investors therefore earn outsized profits by avoiding the closet indexers and embracing concentrated, high-conviction funds?
Neither passive nor genuinely active investors have a good word to say about closet index funds. Such funds are all too common, for two main reasons. Firstly, fund managers run the risk of being fired or losing investors if they suffer short-term underperformance; buying hundreds of stocks means outperformance becomes almost impossible, but it also ensures their performance will not deviate too much relative to their benchmark, thereby minimising career risk.
Secondly, managers want to bag the extra fee that comes with increased assets under management, but it’s difficult to manage large sums of money if your investment universe is limited to 30 or so stocks. Accordingly, managers are naturally incentivised to over-diversify and own more stocks; diversification becomes diworsification.
That’s bad news for investors hoping to outperform markets. According to a study conducted by money manager and author Robert Hagstrom, a randomly generated portfolio of 15 stocks has a 27 per cent chance of outperforming indices, but the odds of outperformance decline to just 2 per cent for portfolios consisting of 250 stocks.
What if managers took a more concentrated approach? Few if any investors can have in-depth knowledge of 200 or 300 stocks, but what if they had only to pick a few stocks – say their best ideas? One 2010 study, ‘Best Ideas’, found returns generated by managers’ “high-conviction investments” – for example, their top-five stocks – comfortably outperformed the market and their other holdings. Investors, the authors concluded, “would benefit if managers held more concentrated portfolios”.
Similar conclusions are reached by the authors of ‘Diversification versus Concentration . . . and the Winner is?’, a 2012 paper that examined 4,700 US equity funds over the 1999-2009 period; the more concentrated the portfolio, the better the performance.
Active share
One way of avoiding closet index funds is to buy funds with a high active share, a concept popularised by Yale researchers Martijn Cremers and Antti Petajisto.
Active share refers to the percentage of a fund’s portfolio that differs from its benchmark. A fund with a score of 0 is identical to its benchmark index, whereas a score of 100 means it holds none of the stocks in the index. A 2009 Cremers and Petajisto paper that examined the 1980-2003 period found US funds with the highest active share “significantly outperform their benchmarks, both before and after expenses”.
A follow-up Petajisto paper in 2013 that analysed the 1990-2009 period found the average actively managed fund underperformed, with one exception – the most active stock pickers, who beat indices by 1.26 per cent per year after fees and by 2.61 per cent annually before fees, indicating a “nontrivial amount of skill”.
Other studies have replicated these findings regarding the value of active share. One recent study found significantly better performance from emerging market funds with a high active share over the 2009-14 period. Similarly, recent research by Premier Multi-Asset Funds’ manager Simon Evans-Cook has found UK funds with the highest active share have easily outperformed the FTSE All-Share index over one-, three-, five- and 10-year periods.
Sceptical
Many researchers are sceptical, however. A recent study of the Norwegian equity market found "no correlation between the active share and performance, persistence, nor skill". Japanese firm Nomura's analysis of the US market found that closet trackers actually outperformed concentrated stock pickers in eight of the 11 years from 2004 to 2014. Active share "may have been useful before 2004 but not since then", cautioned Nomura.
Similarly, Morningstar researcher Russ Kinnel noted last month that funds with high active share have underperformed those with low active share over the 2011-15 period. Low fees are a better guide to fund performance, said Kinnel; 64 per cent of the cheapest funds outperformed, he said, compared to just 29 per cent of the funds with the highest active share.
Advocates of passive investing are also sceptical. A Vanguard report found high-conviction funds were associated with higher costs but not better performance. An S&P Dow Jones Indices report, meanwhile, cautions portfolio volatility “will almost certainly increase” if an active manager reduces his holdings from 100 to 20 or so stocks. The S&P report also suggests portfolio concentration makes management skill “harder to detect, and less likely to matter”; a few losing bets can make a skilled manager look bad just as a few lucky punts will make bad managers look good, resulting in investors’ decisions being “informed by luck rather than skill”.
The most trenchant criticism, however, comes from researchers allied to hedge fund AQR Capital Management. Their paper, ‘Deactivating Active Share’, says Cremers’ and Petajisto’s findings have “attracted considerable attention in the investment community”, with more active managers opting to “tout their active share”. However, AQR researchers argue that while the highest active share funds have outperformed in recent decades, this is because such funds were largely invested in small- and mid-cap companies; this, rather than active share, explains the results.
Patience
Cremers and Petajisto, who have vociferously rejected the AQR thesis, argue that critics are missing the point. Active share is “not a measure of skill”, Cremers said at a Morningstar conference earlier this year. “You do not need skill to have high active share, you just need to buy a bunch of different stocks”.
On its own, a fund’s active share tells you little, he admits. The key is to examine a fund’s active share and its average holding period. Cremers’ latest paper, ‘Patient Capital Outperformance: The Investment Skill of High Active Share Managers Who Trade Infrequently’, finds that managers that trade frequently “systematically underperform”, even if they are funds with a high active share. However, funds characterised both by high active share and patience (defined as holding positions for at least two years) have significantly outperformed over the last 26 years, by over 2 percentage points per year.
Useful
Overall, the concept of active share appears useful on a number of levels. Calls for funds to reveal their active share have grown louder in recent years and that is a good thing – low active share scores expose closet index funds that are taking their investors for a ride. Additionally, if a fund’s active share declines rapidly, it may suggest that asset bloat – a rush of money into a fund after a period of outperformance – is driving a change in strategy.
Nevertheless, it’s clear investors would be naive to dive headlong into a fund purely on the basis of it boasting a high active share or a low number of stock holdings. Deviating from an index is risky. Investors frequently make the mistake of thinking a portfolio consisting of 10 or 20 stocks has a 50:50 chance of outperforming, but the aforementioned Robert Hagstrom statistics show a randomly-generated 15-stock portfolio has only a 27 per cent chance of beating the market.
That’s because most stocks actually underperform the market, with a small handful of so-called super-stocks historically accounting for the bulk of market returns. Doubtless, those odds will plummet further if a fund manager relies on a dumb strategy – managers that frequently trade expensive growth stocks will underperform, whether or not their fund boasts a high active share.
In that sense, active share is, as Prof Cremers puts it, “a very basic tool” in a large toolbox. If combined with other tools, however, the measure would appear to be a relevant one for investors seeking market-beating returns. Cremers’ research indicates the best managers outperform by focusing on stocks often shunned by others: specifically, picking safe, value and high-quality stocks “and then sticking with those over relatively long periods until their apparent undervaluation has been reversed”.
In that sense, high active share doesn’t mean a manager will outperform; it is, however, a necessary condition for outperformance.
