Lessons from the “Naughties”

By Rob Arnott

FEBRUARY 2010 Read Time: 10 min

Last month, we examined the Lost Decade and learned that much of the pain of the past 10 years was caused by an overreliance on the equity risk premium and the corrosive effect of capitalization weighting our equity holdings. Simply bypassing these two practices would have delivered respectable 7–8% annual returns. But past is not prologue. History is littered with the folly of building yesterday’s army to fight tomorrow’s war. The capital markets are especially unkind to those whose success is dependent on extrapolations of recent successes.

In this issue we apply the lessons of the recent Lost Decade to current market conditions. From an asset allocation perspective, the outlook for the ubiquitous 60/40 blend remains bleak. Unfortunately, moving away from this standard mix to a broader toolkit of risk exposures is likely to be less profitable than it was in the past decade as yields from diversifiers like REITs, TIPS, and emerging market bonds are well below the levels of 10 years ago. The key to better returns will be to respond tactically to the shifting spectrum of opportunity, especially expanding and contracting one’s overall risk budget. This approach, combined with “better beta” choices like the Fundamental Index™ concept (which currently sports an unusually deep discount, relative to capitalization weighting), should help us to achieve our targeted returns in what—we shudder to suggest—is likely to be another tough slog for investors.

Busting Out the Crystal Ball
Naïve mean reversion would indicate that 10 lean years for the 60/40 blend (60% S&P 500/40% BarCap Aggregate) ought to be followed by a decade of relatively strong results, especially when the recent lean years delivered the first ever decade of negative real returns! Of course, this assertion can only be verified with a perfectly tuned crystal ball.

While we take great pride in our asset class forecasting, we unfortunately don't have such a device buried in our research department.But we can reasonably project likely future asset class returns by starting with their key Building Blocks. The long-term return on any investment can be broken down into income, growth in income, and changes in valuation levels. Table 1 illustrates these components, save for changes in valuations levels (more on that later), for the S&P 500 and BarCap Aggregate Bond Index as of December 31, 1999, and December 31, 2009.

Let’s start with equities because we spent most of last month’s issue of Fundamentals on their Lost Decade. The dividend yield on the S&P 500 was 2.1% as of December 31, 2009. True, that’s almost double the rate at the end of the 1990s, but it’s still puny relative to a long-term average of 4.5% since 1900. If we add a historic growth rate to those dividends, we arrive at an annualized real long-term expected return of 3.3% for stocks, assuming no change in valuations. Clearly, 10 years of poor returns hasn’t materially impacted expected future returns. As some wags have suggested, the Tech bubble discounted not only future growth but also growth in the hereafter.

On the bond side, the current yield to maturity is an excellent predictor of future long-term returns. Accordingly, bonds helped the 60/40 portfolio in the Lost Decade as they started with a yield of over 7%. Today the yield is about half as large. Backing out today’s break even rate,we see a core bond portfolio can be reasonably expected to achieve only 1.8% real return.

So, a reasonable expectation for a standard 60% stock and 40% bond mix over the next 10 years is a real return of 2–3% per year, again assuming no change in valuations. Yikes! The Lost Decade has most assuredly not paved the way for easy times in the years ahead. We’re still in a low return environment. This is a commonplace observation but most observers refer to low returns relative to the 1980s and 1990s, not the last decade. How many observers expect mid-single-digit returns for the foreseeable future (with the hopeful assumption of no changes in valuations)?

The Impact of Valuations—The BIG Wildcard
But valuations do change and have large multiplier effects on 10- year returns from asset classes, especially stocks. Consider that during the naughties a rise in dividend yields from 1.1% to 2.1% implies a 48% drop in the value that the market was willing to pay for each dollar of dividends. That works out to a 6.5% annualized drop in valuation multiples. If we examine Table 1, we find that a valuation change of 6.5% pulls our annualized real return down from 2.3% to –4.2%. What was the actual result? A whole lot closer to the latter: –3.6%!

The annualized contribution of changing valuations to equity returns has ranged from +10.9% to –6.8% over the past six decades. So where are we today in the stock market? Figure 1 details the Shiller P/E ratio over time, which uses 10-year smoothed earnings in the denominator. The Mother of All Recoveries has pushed equity valuations, marginally cheap in a historical context back in February 2009, back into the low 20s, a 25% premium to the long-term average.

As you can see in Figure 1, equities traded at the same P/E ratios as they did in early January 2010 in four distinct time periods (highlighted): 1928–1930, 1936–1937, much of the 1960s, and 1992–1995. Table 2 shows the subsequent average 10-year equity returns, inflation, and ending P/E ratios from each of these periods. Not surprisingly, the subsequent 10 years after 1928–1930 (even to those who slept through American history classes) showed negative nominal returns and deflation due to the Great Depression. The 10-year periods following 1936–1937 and the 1960s showed average annual inflation in the 4% range, well higher than we’ve seen in the past 25 years. With equity P/E ratios contracting into the 14–15 range, against a headwind of inflation, stock investors suffered skinny real returns of 1.5–2.5%. Only following the early 1990s did 10-year returns bump into double digits, as low inflation and rising valuation multiples allowed the S&P 500 to average 10.6% per annum, gains that were subsequently lost.

If we believe in higher long-term inflation over the next decade,then equity valuations are likely to contract, meaning our Building Blocks return forecast for stocks and bonds may be too high. Stocks will produce less due to the downward pressure on valuation multiples, while higher inflation eats into today’s skinny nominal bond yields. So, one lesson of the Lost Decade is likely to hold true—an equity-centric mix of mainstream stocks and bonds is likely to disappoint. Again. Net of inflation, it could even be worse than the past 10 years.

Diversification and Alternative Assets—With No Fat Pitch, Think Tactical
A key tonic to the past 10 years was a more diversified, less equity-centric approach. A risk premium  government bond isn’t restricted to equities; plenty of assets offer premiums in line with stocks and occasionally higher. In the last issue we used the 16-asset class portfolioto illustrate the benefits of diversifying across a wider spectrum of asset classes. For the decade 2000–2009, this more-diversified approach achieved an annualized return of 6.8%, a 450 bps premium over 60/40. Abandon cap weight for stocks and the return jumps to 8.5%, nearly matching most investors’ targeted returns.

Looking forward, the outlook is not as “attractive” as it was in 2000. Today, yields on most of these diversifying assets are well off the rich premium levels at the turn of the century. Back then, NASDAQ-induced neglect led to a whole spectrum of alternative asset classes, favorably priced for attractive long-term returns. Today, we aren’t so lucky as many off-the-beaten path categories sport rock bottom yields (and, therefore, low forward-looking returns). Figure 2 provides a quick snapshot of “Then Versus Now” in four asset classes: REITs, TIPS, emerging market bonds, and high-yield bonds.

Emerging markets bonds, REITs, and TIPS offer half of their Y2K yields. Even high-yield bonds, whose 1999 yields were pushed down due to heavy issuance by adored tech and telecom players, show significantly lower yields today. The fat pitch of diversification into risk premiums beyond mainstream stocks and bonds is largely gone.

So what to do? Manage the asset mix! Vitally important in this exercise is to shift risk postures. Too often asset allocation programs are governed by a relatively constant risk tolerance, say on par with a 60/40 stock/bond mix. This approach encourages swapping one risky asset class out for another (e.g., non-U.S. developed stocks for emerging markets stocks, REITs for U.S. stocks, etc.). But in the current environment, when all asset classes are rich, shouldn’t we consider a more conservative posture? This approach isn’t market timing but risk budgeting. We choose to take long-term risk when risk-bearing is likely to be rewarded, and a conservative, well-diversified posture when it is not. Rich forward-looking risk premiums typically prevail when investors are terrified, as they were in early 2009. As Warren Buffett suggests, we should be “greedy when others are fearful and fearful when others are greedy.”

Out-of-mainstream markets can still add value if we use them tactically and opportunistically. Inevitably, investors sell the assets they least understand when times get rocky and buy them when conditions are calm. Thus, diversification can still be powerful, but only if we practice diligent tactical asset allocation.

Outlook for Equities? Depends on your Index!
Stocks were terribly disappointing during the Naughties, but the results of the Fundamental Index approach (and, for that matter, equal weighting) illustrate that the shortfall was largely attributable to the cap-weighted construction of traditional indexes. Never rebalancing and always chasing winners, cap-weighted indexes held far too much in new age tech before the 2000–2002 bear market and far too little in the left-for-dead financials, industrials, and consumer sectors ahead of the Mother of All Recoveries that closed out the decade.

But the destruction of capitalization weighting wasn’t restricted to these two bookend years as evidenced in Table 3. A global, all country Fundamental Index (FTSE RAFI® All World 3000) portfolio beat the representative global, all country cap-weighted portfolio (MSCI ACWI) 9 years out of 10, falling short by a scant 30 bps in 2008. Even equal weighting, the most naïve of all price-indifferent approaches, managed to win by 600 bps and did so consistently (8 years out of 10).


Of course, yesterday’s winners typically become tomorrow’s laggards. However, a comparison of current valuation discounts for Fundamental Index strategies versus cap-weighted ones indicates that avoiding the negative alpha of capitalization weighting is likely to still be profitable at today’s valuation levels. Previously, we noted that when RAFI US Large trades at a price/book ratio 27% or more “cheaper” to the S&P 500, the odds are good for subsequent outperformance—in the United States, the RAFI portfolio beats the S&P 500 in over 80% of subsequent three-year periods, with an average of 3.6% of additional return.5

So where does this discount stand today? Despite achieving its second best year ever of relative outperformance in 2009, the FTSE RAFI US 1000 still trades at a discount of 48% to the S&P 500. Similar discounts can be had elsewhere, including 38% for a Global All Country application as evidenced in Table 4. These approach the historical peak discounts seen at the top of the Tech bubble in early 2000.

It’s not realistic to expect another 10 years of 600–700 bps per annum return drag from capitalization weighting. Nonetheless, given today’s discount levels, we expect continued sizeable gains from non-cap-weighted indexes and, therefore, continued benefits from using a Fundamental Index approach.

“Lost and Found” will not describe investment results for the first two decades of this millennium, as sizeable real returns will prove to be difficult for the second 10-year stretch in a row. Most investors will fall short of their goals, as almost all asset classes—whether mainstream or alternatives—are priced richly relative to historical norms. But odds can be tilted back in our favor by tactically altering our portfolio risk based on measures as simple as yields and yield spreads. The most successful investors are those with the discipline to shun risk when the markets seem tranquil, and the fortitude to seek risk when others are terrified. The best path to future success marries risk management—tactical asset allocation—with a more efficient beta like the Fundamental Index methodology and a full toolkit of alternative markets.

Featured Tags

Learn More About the Author

Partner, Chairman of the Board


1. Ironically, many of our turn-of-the-century predictions proved remarkably—and sadly—prescient. Early drafts of “The Death of the Risk Premium” (published in early 2001) were circulated as early as February 2000. Before the top! But, even a good crystal ball doesn’t assure success with clients. The mid-decade bull market caused some shorter term investors to bail out of asset allocation programs, despite their eventual reliability over the full decade. Many paths can be taken to achieve a spot-on 10-year forecast. Successfully managing expectations is often harder than successfully managing assets!
2. Admittedly, breakeven rates are a poor predictor of future inflation as they can be influenced by many things. In 2000, the relative newness of TIPS and the tech bubble allowed TIPS yields to briefly cross 4%. On the flip side, the liquidity based sell-off in the fall of 2008 disproportionately hurt TIPS versus nominal Treasuries.
3. See “3-D Hurricane Force Headwind,” Fundamentals, November 2009. Incidentially, this longer term near inevitability of inflation probably isn’t going to be an issue shorter term—next 12–24 months—as a weak recovery and falling rents will put pressure on CPI figures. But on a 10-year outlook (the minimum planning horizon for institutional investors and most retirement programs), our bet is on higher inflation. Perhaps even far higher.
4. The equally–weighted portfolio comprises the following 16 indexes, rebalanced monthly: ML US Corporate & Government 1–3 Year; LB US Aggregate Bond TR; LB US Treasury Long TR; LB US Long Credit TR; LB US Corporate High Yield TR; Credit Suisse Leveraged Loan; JPM EMBI + Composite TR; JPM ELMI Composite; ML Convertible Bonds All Qualities; LB Global Inflation Linked US TIPS TR; FTSE NAREIT All REITs TR; DJ AIG Commodity TR; S&P 500 TR; MSCI Emerging Markets TR; MSCI EAFE TR; Russell 2000 TR.
5. “Discounts and Relative Performance,” Fundamentals, February 2009.