Pay active management fees only for truly active management.
“If you are a professional and have confidence, then I would advocate lots of concentration. If it’s your game, diversification doesn’t make sense. It’s crazy to put money into your 20th choice rather than your first choice.”
Buffett makes sense. But most active U.S. equity managers own diffuse portfolios. Of the nonindex, nonsector U.S. equity funds for which Morningstar has portfolio data, the average fund holds 134 stocks and keeps only 27% of its assets in its 10 largest holdings. The popularity of benchmark hugging is a consequence of the incentives managers face. Closet indexing won’t win you much in the way of assets, but you can keep your job for many years before asset outflows become a big problem--for example, Robert Stansky ran Fidelity Magellan FMAGX as a closet S&P 500 fund for nine years before retiring. On the other hand, truly active investing will either make you look brilliant or get you fired in a hurry.
Investors suffer by paying steep active-manager fees on the indexlike portion of their funds. If you're going the active route, why not just own truly active, benchmark-bucking funds? Aside from offering more bang per buck, maverick managers may offer outperformance and valuable diversification. And many of their perceived flaws can be easily addressed by combining them with index funds.
Truly Active Funds Outperform
Researchers have found that maverick managers tend to outperform their benchmarks and peers. Antti Petajisto invented “active share,” a measure that expresses as a percentage a fund’s deviation from its stated benchmark. He found that stock-picking funds with high active share on average beat their benchmarks by 1.26% per year. Yakov Amihud and Ruslan Goyenko found that funds with low correlation to their benchmarks tended to have more persistent performance.
This makes sense. Truly good managers are likely more confident in their ability to discern undervalued securities, so they are willing to make big bets. Buffett famously bet 40% of his partnership’s capital on American Express AXP during the “Salad Oil Scandal.” He has said he was willing to devote up to 75% of his capital on a high-confidence, low-risk trade. Like skilled managers, truly good investment ideas are also rare. When a manager invests in her best ideas, she has a better chance of overcoming her costs. A study by Randolph Cohen, Christopher Polk, and Bernhard Silli found that the highest-conviction bets the average manager makes beats the market. This suggests skill is somewhat more common than thought but is masked by benchmark-hugging and high fees.
As a bonus, it’s also easier to be confident in benchmark-bucking fund managers: Their performances are more persistent than that of indexlike managers, according both Petajisto’s and Amihud and Goyenko’s studies. Petajisto speculates that big betting appeals to managers on the extreme ends of the skill spectrum, the gifted and the incompetent. That may be one reason, but I think a more important one is the fact that maverick managers make more truly independent “skill,” or “alpha,” bets. A market-beating performance produced by many bets is more indicative of skill than one that reflects a single monster bet that paid off in spades. One might object that a manager with lots of bets will have a diversified, benchmarklike portfolio. On the contrary, the overdiversified manager has actually increased his market exposure (because idiosyncratic stock movements cancel each other out) at the expense of skill bets, reducing his effective number of independent bets.
Low Returns, Low Correlation? No Problem!
The modest outperformance of maverick managers is, in aggregate, not game-changing, and in fact may not be their most compelling trait. Their relatively low correlation to traditional benchmarks makes them potent portfolio diversifiers. Portfolio theory emphasizes that one should judge funds by how they change your overall portfolio’s risk-reward character, not whether they simply beat their benchmarks. In some cases, a fund with low risk-adjusted returns and low correlation to your portfolio can be better than one with high returns and high correlation.
Consider the quintessential maverick fund, Fairholme FAIRX: Morningstar crowned its manager, Bruce Berkowitz, Manager of the Decade in 2010, right before the fund fell off a cliff. From the beginning of 2008 to the end of June 2012, it returned only 0.73% annualized, barely topping the S&P 500, but with over a third more volatility. The fund’s risk-adjusted performance was a bust. Manager Bruce Berkowitz concentrates his positions, and his fund’s active share is among the highest of all U.S. equity mutual funds. Unsurprisingly, Fairholme’s correlation to its benchmark, the S&P 500, has been among the lowest in its category. Despite its miserable risk-adjusted performance, replacing half of the equity portion of a 60/40 portfolio with Fairholme improved the portfolio’s absolute and risk-adjusted returns. This is diversification in action.