The invention of the autopilot was a critical advance in aviation history. Why? As flights got longer, pilots suffered from fatigue and made errors. They needed a stabilizer to keep the plane on course and adjust for changing conditions. When 21-year-old inventor Lawrence Sperry demonstrated his gyroscope-equipped autopilot at a 1914 Paris air safety show—both he and the plane’s mechanic actually walked onto the plane’s wings in mid-air, leaving no one in the cockpit—the spectators were stunned. Sperry became an international celebrity overnight.
In our industry, the target-date fund is intended to provide an autopilot for defined contribution participants—most of whom aren’t any more qualified to pilot their investments than Wile E. Coyote is to jump into the cockpit of a 747. Target-date funds are simple investment solutions whose asset mix becomes more conservative and income oriented as the target date approaches. In theory, that ratcheting down of exposure to risky asset classes should be a good thing. In practice, as the last 10 years has illustrated, risk is not linearly rewarded. The “glide paths” of target-date funds assume constant risk premia from equities and other asset classes.
In this issue we explore another path utilizing the Fundamental Index™ approach and find its inherent contra-trading produces a far better result. Furthermore, this approach is simple and inexpensive, two critical components of an effective 401(k) solution. In addition, we review how a simple tactical asset allocation approach would generate an even bigger retirement kitty.
Risk Premiums are not Constant
Considerable evidence exists that 401(k) investors do not make good investment decisions in their retirement plans. One study (by Munnell et al.) found that 401(k) plans experienced a 1% shortfall relative to defined benefit plans from 1988–2004 “due to poor timing and other investment mistakes.” The study further explained that 401(k) participants on average make sensible investment choices, but on an individual basis are poorly diversified.1 Other research found that employees tend to use a small number of investment options—typically three or four—and offering too many choices can cause “information overload” and can reduce participation and contribution rates.2
Target-date funds are designed to address these issues. They offer a pre-diversified, “one-stop shop” that adjusts asset allocation over time as retirement draws nearer. Equity bets—by far the largest risk source—are diversified by style (growth and value), size (large and small), and geography (U.S., developed ex-U.S., and emerging markets). But they miss the boat in one key area—they fail to adjust to changing market conditions. At some points, investors are amply rewarded for placing their chips on stocks (the forward-looking equity risk premium is large); at other times—usually when markets are in a boom phase and stocks are over-valued—investors receive little reward for taking sizeable equity exposure (the equity risk premium is small).
How volatile is the equity risk premium? Figure 1 illustrates how the excess return of stocks over bonds changes during market cycles. The same is true within equity markets: During some periods, value stocks are priced to deliver a substantial premium while other stretches show value offering little or no premium. A similar effect occurs in size and geographic orientation. One of the key benefits of target-date funds—their simple, formulaic approach to asset allocation—has turned out to be the source of their biggest problems. If market conditions are not factored into the decision on asset allocation, investors could own too much of an asset class at peak valuations—a likely precursor to future underperformance. We saw evidence of this in 2007–2008 when some funds held very high equity exposures despite very high valuations and low dividend yields, shrinking the nest eggs of those close to retirement age when stock markets plummeted. These same strategies missed a spectacular rebalancing opportunity in early 2009.3
The Power of Simple Contra-Trading
We strongly advocate adding tactical asset allocation to target-date glide paths (see our discussion below). However, we realize this approach may not be for everyone. One way to cost effectively receive some of the same benefits is to use a global Fundamental Index approach as the sole equity portfolio.
The beauty of the RAFI™ methodology is its ability to implicitly adjust its exposures—by geographic region, style, and size—through anchoring on company fundamentals and automatically rebalancing back to these measures. This embedded contra-trading inevitably sells recent winners and buys recent losers. This process naturally provides a mechanism to reduce exposure to higher priced assets.
We can graphically see how this process works on a country and regional basis in Figure 2. Japan at the top of its bubble in 1989 reached 51% of total world stock market capitalization. Ten years of brutal underperformance later, it was 8% of world stock market capitalization. On the other hand, fundamentally weighting Japanese companies by their share of world sales, cash flow, book values, and dividends, we find that the Japan allocation in a global equity portfolio ranged much more tightly, from 12–22%. Target-date funds would have assumed that Japanese equities were going to deliver the same return when they were both 51% and 8% of world market capitalization!
This approach also works on a style basis. When value is really cheap—following a strong run in growth stocks—RAFI strategies tend to have a deep value exposure. And when value is expensive, RAFI strategies tend to have less exposure to value stocks.4 In fact, RAFI strategies have achieved most of their added value through this dynamic contra-trading process.5
Of particular interest in a target-date fund setting is the ability of the RAFI methodology to implicitly adjust its overall equity beta—the key driver of risk for these funds. Figure 3 shows how the rolling 12-month beta changes dramatically for RAFI US Large portfolios over nearly half a century, based on simulations.6 During the 47-year period, the beta for the index reached a peak of 1.43 and a nadir of 0.59 at the peak of the tech bubble in early 2000. This figure implied that the RAFI strategy would rise three-fifths as much of the equity market or, conversely, fall three-fifths of the overall market’s decline. What happened in the subsequent three years? The S&P 500 Index declined by 41% cumulatively and the RAFI strategy shed only 12%.
RAFI Glide Path’s Upward Trajectory
So how well would a target-date fund using the FTSE RAFI® All World 3000 Index perform? To estimate the weight between stocks and bonds, we used the old guideline of holding the same percentage of bonds as one’s age. Our hypothetical investor is John Saver, a frugal 30-year old who plans to retire in 2020 at age 66. He starts salting away retirement money in January 1984 with a $10,000 portfolio that is invested 70% in stocks and 30% in bonds. Every January 1 he contributes another $1,000 while his stock weighting drops 1%. Thus, at the start of 1985, when John Saver turned 31, his bond weight increased to 31% and stocks were rebalanced to 69%. In comparison, the Dow Jones Target 2020 Index’s equity allocations hover around 90% from 1984 until 1990, when it slowly starts slipping, hitting the 44% mark in 2010. This sloping allocation gave the Dow Jones Index an advantage during the bull market that started in 1982.
The upshot: Currently, John Saver would have $308,000 in his 401(k) account if he had invested in a portfolio mimicking the Dow Jones Target 2020 Index. If he had invested in funds mimicking the FTSE RAFI All World 3000 Index and the Barclays Capital Aggregate Bond Index, his portfolio would be more than $50,000 bigger—an increase of nearly 20%—and with one-third less risk because he would have been invested in a less equity-centric portfolio.
These superior results could have been achieved with remarkable efficiency. The expense ratio for an age re-weighted blend of two broad index funds—the BarCap Aggregate and the FTSE RAFI All World 3000—would be 40 basis points—well below the average target-date fund expense ratio of 1.03%.7 There would be no active managers to hire and fire in each sleeve. Communication to plan participants—most clamoring for a simple explanation of how their nest egg is being invested—would be a breeze. Plan sponsors and their employees would not need to be rocket scientists to find a simple and effective path to retirement security.
An Asset Allocation Alternative
A systematic, disciplined, tactical asset allocation process embedded within today’s glide paths would represent a further step forward. With such an overlay, the glide path can be modified as market conditions (notably risky asset class valuations) change, ideally shifting the portfolio to a more conservative or aggressive bent. A simple exercise of comparing likely and probable long-term asset class returns and then allocating accordingly could have saved considerable pain in previous market storms. One doesn’t have to be clairvoyant—just sensible.
Let’s assume that when stocks beat bonds by a wide margin, things are likely to mean-revert (go back the other way). So John Saver looks back at the return for stocks and the return for bonds over the past five years. Every 1% difference in the annual returns triggers a 1% shift in his year-end rebalance.
When John got started, stocks beat bonds by 5% per year for the previous five years. So, instead of investing 70% in stocks and 30% in bonds, John started with 65% in stocks and 35% in bonds, taking a contrarian bet that markets could go the other way. A year later, stocks were still ahead of bonds, but the five-year equity average outperformance was now only 2% a year. So, instead of rebalancing to 69/31, he now rebalanced to 67/33.
This simple application—contra-trading against whatever has been best in the past five years—leads to 10% more wealth after a quarter-century. This additional $85,000 in returns means that he’s now 28% ahead of where he’d have been with the first generation glide path strategy, offering a far better prospect for a secure retirement.
The autopilot embedded in today’s target-date funds’ glide paths is far too rudimentary in a world of dynamically shifting risk premiums. Would a wise pilot announce, “Based on average weather conditions and airspeed, the flight from LAX to JFK takes four hours and forty five minutes. Rather than check the weather or air traffic, I will simply assume average conditions. See you in five hours.” Of course not! Conditions are rarely average—in flight or in the capital markets.
Asset allocation is a critical step in the asset management process, whose essence Benjamin Graham once described as “…the management of risks, not the management of returns.” No matter how diversified the portfolio, risk and reward aren’t linear. But target-date funds tacitly assume they are! Just because you are willing to take more risk doesn’t preordain higher returns, even over decades-long stretches. Rather, managing risk should be done either explicitly with active asset allocation of the glide path or implicitly through the natural contra-trading embedded in the Fundamental Index approach.
Learn More About the Author
1. “Investment Returns: Defined Benefit vs. 401(k) Plans,” Alicia H. Munnell, Mauricio Soto, Jerilyn Libby, and John Prinzivalli. Issue Brief, Center for Retirement Research at Boston College, September 2006.
2. “Plan Investment Options and Participant Behavior,” TIAA-CREF: Research Summary, June 2006.
3. See “The Great Contra-Trade,” RAFI Fundamentals, May 2009.
4. See “Dynamic Style and Size Exposures,” RAFI Fundamentals, August 2007.
5. Contra-trading led to a major boost to performance in 2009 when RAFI strategies displayed a strong bias toward deep value stocks. For a longer discussion, please see the May 2009 issue ofRAFI Fundamentals.
6. We use the U.S. large company RAFI strategy due to its longer track record. It should be noted that the proposed RAFI All World strategy would have generated similar results.
7. Due to the long simulation period, we used a cap-weighted bond index for the fixed-income portion of the portfolio, thereby understating the full benefit of this approach.