Contrarianism has a long history in investing, and quite a bit of success as well. Graham and Dodd’s Security Analysis was an early investigation of the power of thinking independently as an investment framework. If you had to boil it down, Baron Rothschild’s maxim to buy when there’s blood in the streets sums it up nicely. Kipling’s poem “If…” does the trick too: “If you can keep your head when all about you are losing theirs…”
Jumping back into finance proper, contrarianism draws support from the idea of mean reversion, which is another way of saying that what goes up comes down, and vice versa. Not immediately, of course. Between the medium-to-long-term reversions is momentum, or the tendency of asset prices to move up or down for periods of roughly 12 to 24 months. A fair amount of the art/science of managing money is deciding when momentum gives way to mean reversion, and then back to momentum in the other direction. Of course, if you were truly a long-term investor, you could ignore momentum. Or if your focus is on the short run, mean reversion could be and arguably should be overlooked. But for everyone else, a blending of the two may be practical.
In any case, it’d be foolish to mindlessly buy fallen angels, and sell the winners, but that’s more or less the idea in mean reversion. But the details matter.
Analyzing the phenomenon in stock prices was launched as a formal inquiry in a pair of 1988 studies—one from Fama and French and another written by Porterba and Summers. Fast forward to 2010, and you could spend a few years reviewing the finer points in the literature that supports the idea that betting against the crowd is a winning principle.
There’s lots of debate about why mean reversion exists. Some claim it’s evidence of market inefficiency and irrational investors. A competing line of analysis is that mean reversion tells us that expected returns fluctuate and that risk premiums, like everything else, vary, largely in line with changing expectations about the economy. The crowd demands higher—perhaps much higher—compensation during times of economic stress vs. those periods when all seems right with the world. Regardless of where you come down in this debate, the larger point is that mean reversion exists, or at least there’s a mountain of empirical studies telling us so.
And the research keeps on giving. A study from earlier this year, for instance, analyzed 17 stock markets in the developed world since 1900 and found that, yes, reversion to the mean is alive and kicking around the globe. But as the paper reminds, the potency of this factor varies, depending on the prevailing conditions. The authors report that the deeper the shock that knocked equities off the pedestal, the quicker the subsequent mean reversion.
Over the full period of study (1900-2008), the paper notes that “it takes stock prices on average 13.8 years to absorb half of a shock.” But the results vary if you look at specific time periods. As the authors explain,
…using a rolling-window approach we establish large fluctuations in the speed of mean reversion over time. The highest speed of mean reversion is found for the period including the Great Depression and the start of World War II. Similarly, the early years of the Cold War and the period covering the Oil Crisis of 1973, the Energy Crisis of 1979 and Black Monday in 1987 also show relatively fast mean reversion. Overall, we document half-lives ranging from a minimum of 2.1 years to a maximum of 23.8 years. In a substantial number of time intervals no significant mean reversion is found at all, which underlines the fact that the choice of the data sample contributes substantially to the evidence in favor mean reversion. Our results suggest that stocks revert more rapidly to their fundamental value in periods of high economic uncertainty, caused by major economic and political events.
Mebane Faber, author of The Ivy Portfolio has crunched the numbers too and found some interesting results using asset classes. As he wrote earlier this week, “It is pretty well established that markets mean revert from returns of 2-4 years prior.” Turning to the data, he analyzed five asset classes (U.S. stocks, foreign stocks, Treasuries, commodities and REITs) and reported that a simple portfolio that owns the worst performer over the past three years, and re-evaluates once a year, does quite well vs. buying and holding since the mid-1970s.
So, how does the current round-up of asset class returns look these days? The big losers for trailing 3-year annualized total return through August 31, 2010 are equities in the U.S. and foreign developed markets, commodities overall and REITs, according to data from Morningstar Principia:
Foreign stocks (MSCI EAFE): -10.8%
US Stocks (Russell 3000): -8.3%
Commodities (DJ-UBS): -6.6%
REITs (MSCI REIT): -6.1%
On the flip side, the leading winners for the major asset classes over the past three years comes down to bonds. Again using annualized 3-year total return:
Emerging Market Bonds (Citigroup ESBI-C): +10.7%
Foreign Developed Market Bonds (Citigroup WGBI ex-US): +8.3%
U.S. Bonds (Barclays US Aggregate Bond): +7.7%
TIPS (Barclays Treasury TIPS): +7.2%
Does this mean we should simply buy the losers and sell the winners? No, at least not without considering how our own portfolios are currently structured. Other factors to weigh include the details of our investment objective, risk tolerance, time horizon, etc. But the above list of winners and losers is a good place to start for evaluating how to rebalance an asset allocation that’s wandered away from its strategic targets of late.
Yes, rebalancing has a history of modestly enhancing return, controlling risk, and perhaps both. True for broadly defined asset classes as well as within single-asset class portfolios. Why? Reversion to the mean offers some perspective. Indeed, once we consider reversion to the mean through the prism of rebalancing, it’s hardly surprising to find a track of return-boosting results. Is it alpha or beta? That’s another topic. But rebalancing is certainly part of the reason why the equal weighted S&P 500 has beaten its cap-weighted counterpart in recent years. It also helps explain why the fundamental indices designed by Research Affiliates have an encouraging track record against their conventional value-weighted equilvalents. Indeed, as Research Affiliates founder Rob Arnott and his co-authors discuss in The Fundamental Index, cap-weighted indices don’t rebalance whereas fundamental weighted benchmarks do. That’s not the entire story of why cap-weighted indices lag (the value and small-cap premia are relevant too, perhaps even more so). But rebalancing is certainly a critical factor.
Is using rebalancing as a tool for exploiting mean reversion a free lunch? No, absolutely not. When you rebalance—if you’re doing so in a timely manner—you’re embracing a particular type of risk at a time when the crowd generally seeks safety. A recent study makes a persuasive case that rebalancing boosts returns because relatively few investors do so in a timely manner, which in turn creates conditions for volatile swings in expected return. That opens the door for boosting performance with opportunistic rebalancing.
That reluctance to rebalance, or at least engage in timely rebalancing isn’t surprising. How many of us were buying stocks in late 2008/early 2009 and selling bonds? Not many. For those who did, recent trailing returns today probably look quite handsome vs. those who waited for the all-clear sign. But to reap the higher return, you had to endure some sleepless nights. For most folks, it was much easier to sit in cash. But that comes at a price, even if that’s a weak argument when the world seems to be coming apart. In fact, there are studies, such as this one, that document that rebalancing is relatively rare among individuals. That’s one reason–maybe a big reason–why there’s a risk premium connected with rebalancing for the few who exploit this opportunity.
There’s no guarantee that mean reversion will work, although almost everything’s suspect in finance in real time. But history suggests we should be cautious before dismissing the idea entirely.