In a classic puzzle, readers are asked to draw four straight lines through a 3 x 3 matrix of nine dots—without letting their pencils leave the paper (see Figure 1). Most readers fail, at least initially, ending up with one dot left over. The solution requires thinking outside the box. The phrase “thinking outside of the box” has become so overused in recent years as to become trite. And yet, how many investors actually deviate from the norm with their equity allocations? Indeed, most investors follow the pack, implementing one of three “standard” strategies.
In this issue we will look at a different way of constructing the equity portfolio. We will use the concept of “active share”—a measure of how much active equity portfolios actually deviate from their benchmark indexes—as well as what active share tells us about the standard equity structure alternatives.
Equity Structure Choices
The success of an investor’s overall portfolio is highly dependent on how well the equity component performs; stocks are the largest allocation in most portfolios, on average half of assets or more.1 Therefore, paying special attention to the equity strategy decision is very important.
The three common ways to structure stock portfolios are:
1. 100% passive management, tracking a cap-weighted index
2. 100% active management, usually fundamental stock-picking strategies
3. Core-satellite—a combination of passive indexing and active management
An equity structure that entirely uses a passive cap-weighted index approach provides the market return less implementation costs. This equity structure gives you low cost and little shortfall risk relative to the benchmark, but no potential for added value, otherwise known as “alpha.” Without any alpha the passive equity structure locks in future stock returns that are unlikely to meet an investor’s return goals.2
To beat the market, investor portfolios must be different than the market. So, an investor who wants or needs to beat the market must populate his or her portfolio with active strategies because of the bets they take away from the benchmark. But exactly how active are active managers? In 2006, Martijn Cremers and Antti Petajisto of Yale University introduced a tool—dubbed “active share”—to measure how much active managers differ from the benchmark index.3 The active share measurement ranges from 0% for an index-tracking portfolio, such as an S&P 500 fund, to 100% for a fund that holds no overlap with the index, such as a concentrated active stock-picking strategy. Most active managers lie somewhere in between.4
Cremers and Petajisto found that funds in the highest active share quintile achieved the largest average alpha—1.1% per year net of fees and transaction costs.5 The average active share of this quintile of active managers is approximately 90%—that is, the typical portfolio had only a 10% overlap with the index. For the active management industry as a whole, the authors found that active share was just 30% and that the average fund returned –0.43% against the market. Net net, investors frequently end up paying high active management fees for index-like returns (or worse)!
Clearly, alpha-seeking investors should invest only in the managers with the highest active share, where they get the most compensation for taking active management risk. But the catch is that top quintile active share managers suffer from large tracking error—a statistical measure of volatility of excess returns versus the benchmark. Although an excess return of 1.1% is attractive, studies show that both institutional and retail investors don’t have the stomach to sit through bouts of underperformance inherent in high tracking error strategies.6
As a result, retail investors (imitating their institutional cousins) are increasingly shifting assets into low-cost passive index funds, which have grown from 5% of equity mutual fund assets in 1996 to 15% in 2010.7 This trend has led to a “core-satellite” approach, an equity structure that blends cap-weighted indexing as the “core” part of the portfolio with highly active managers as the satellites. A hypothetical equity structure of 50% passive cap-weighted index and 50% actively managed satellite strategies produces a total equity portfolio active share of 45% with an expected excess return of 0.5%. That return potential is better than that offered by a purely passive portfolio, but it is still well shy of the return target.
RAFI™ Strategy: The Best of Both Worlds
Clearly, the three equity structure approaches, which are summarized in Table 1, are suboptimal to achieve portfolio return goals. The Fundamental Index™ strategy offers an attractive alternative. The Research Affiliates Fundamental Index (RAFI) strategies are able to earn returns over the cap-weighted benchmarks by their superior design—the approach maintains many of the attractive characteristics of traditional passive investing, and avoids the performance drag associated with linking portfolio weights to security prices. By constructing a portfolio built on fundamental weights rather than stock price—and rebalancing back to the fundamental weight anchor as stock prices mean revert—the RAFI methodology has delivered live excess returns of 2% per year on average in developed markets and more in less efficient markets.8 And the methodology has accomplished this track record across most stock markets, with 87% of FTSE RAFI® indexes outperforming their respective cap-weighted index since their launch dates.9
Because RAFI portfolios are broadly diversified, they tend to have a relatively low active share of 30%. RAFI portfolio excess returns are made more attractive because they are achieved in a smooth fashion, with a low tracking error of 4% (relative to 8% for active managers). Accordingly, the RAFI approach provides the best of both worlds—strong excess returns without the big out-of-index risks active managers take—as Table 1 shows.
The way to solve the puzzle in the introduction is to extend one or more lines outside the matrix (or box). In other words, you must literally think outside the box to find the solution (see Figure 2).10 The puzzle of achieving superior equity performance will not be solved by staying within the box of traditional equity portfolio construction strategies. Investors need to explore newer approaches that might offer superior solutions. A RAFI-centric equity structure improves the risk-return trade-off compared to the three common equity structures implemented today. It’s an out-of-the-box approach that is easy and inexpensive to implement, without the rollercoaster ride offered by high active-share managers.
1. 52% in public DB plans, 54% in corporate DB plans, 60% in 401(k) plans. See 401(k) Plan Asset Allocation, Account Balances, and Loan Activity in 2009.
2. We forecast 10-year forward equity returns of 5–6%. This is based on the building block approach of 2% dividends, 1% long term real earnings growth, and 2.5% inflation. Typically, investors have a 7–8% total portfolio rate of return target.
3. K.J. Martijn Cremers and Antti Petajisto, 2009, “How Active Is Your Fund Manager? A New Measure That Predicts Performance,” International Center for Finance, Yale University Working Paper (March 31, revised).
4. The average asset-weighted fund active share is approximately 65%.
5. For comparison, funds in the bottom three quintiles of active share gave up 1.2% per year on average to the index.
6. Scott Stewart and John Neumann, 2009, “Absence of Value: An Analysis of Investment Allocation Decisions by Institutional Plan Sponsors," Financial Analysts Journal, (November/December).
7. Investment Company Institute, 2011 ICI Factbook, Chapter 2.
8. A two percentage point excess return is the average of the 23 developed market FTSE RAFI indexes from November 2005 through June 2011.
9. Research Affiliates and FTSE with data from Bloomberg. 87% of the primary 44 FTSE RAFI indexes have outperformed their respective index since inception.
10. The Original "Thinking Outside the Box" Puzzle!