In 1965, a young finance professor at the University of Chicago published his Ph.D. thesis on the eternally mystifying question in finance: Are equity prices predictable? His conclusion has resonated across the decades. The empirical evidence, he advised, suggests that “chart reading, though perhaps an interesting pastime, is of no real value to the stock market investor.” It’s more reasonable to expect that share prices will fluctuate randomly than to expect much from studying historical price patterns.
The author of that smackdown, of course, is Eugene Fama, who wrote The Behavior of Stock-Market Prices, one of the most famous and influential research papers in modern financial economics. The study laid much of the intellectual foundation for indexing, a strategy that’s incompatible with the notion of forecasting returns, either by evaluating the fundamentals (earnings, sales, etc.) or studying price charts.
Chart analysis, in particular, has come under harsh criticism since Fama’s paper was published. A recent edition of Burton Malkiel’s best-selling book, A Random Walk Down Wall Street, says bluntly that technical analysis is anathema to academics.
Yet many people who analyze price history beg to differ. And if technical analysis in its various forms (moving averages, momentum metrics, etc.) has suffered under indexing’s advance, the tide may be turning. What’s old in technical analysis seems to be in vogue again among financial advisors, especially those who are using this oft-maligned concept to tweak their asset allocations.
One recent study from academia (Out-of-Sample Equity Premium Prediction: Fundamental Versus Technical Analysis, by Christopher Neely, et al.), says that technical analysis is just as useful as fundamental analysis for anticipating returns. And it’s getting easier to find papers that defend technical analysis in asset allocation (articles such as The Trend is Our Friend: Global Asset Allocation Using Trend Following, by Andrew Clare, et al., and Market Timing with Moving Averages, by Paskalis Glabadanidis).
A trio of finance professors wrote in a recent working paper, A New Anomaly: The Cross-Sectional Profitability of Technical Analysis, that a simple moving average, “when applied to portfolios sorted by volatility, can generate investment timing portfolios that outperform the buy-and-hold strategy greatly.”
To call it a “new” anomaly is an exaggeration—there really is nothing new under the sun. Moving averages and the like have been used for decades in the trading community. (Some market historians trace the origins of studying price squiggles on a page to 17th century rice traders in Japan.) What’s new is that after decades of ridicule, these tools have earned new respect among advisors searching for deeper perspective when adjusting asset allocation.
Greg Vigrass, the president of Folio Institutional in McLean, Va., says advisors who custody assets at the company have shown increased interest in tactical strategies as they look for more dynamic approaches to portfolio management. They haven’t abandoned buy-and-hold concepts, but “there’s an understanding that markets are a little bit different these days,” he reports.
Viva la Difference?
After all, following the autumn of infamy in 2008, those with a buy-and-hold bias were backed into a corner. Though the subsequent rebound in asset prices offered some solace, people no longer take it for granted that, when it comes to stock prices, time can really heal all wounds.
Jeff Vollmer, a former rep with Smith Barney, says he counted himself among those “buy-and-hold guys” before 2008. “We all bought into that mentality,” he recalls. When the markets collapsed, he began to rethink his money management process, and he hasn’t looked back since. Now a principal at Cincinnati-based Hyde Park Wealth Management, Vollmer uses a variety of technical indicators to manage asset allocation for clients. He says it’s a way to set some baselines for entry and exit points into the markets.
He now uses two core indicators that the academics a decade before might have looked at distastefully: the exponentially weighted 20-day and 200-day moving averages. The latter is primarily a warning signal. An ETF that slips under its 200-day marker is a candidate for trimming, if not selling outright. By contrast, the 20-day average offers him guidance when he’s deciding whether a fund is a buy, he says. “I don’t know that it’s a sell [when an ETF falls under its 20-day average], but I may not continue to add to that position.”
Vollmer is quick to point out that he’s not “trading,” at least not by the conventional definition, which is to make frequent moves in and out of a security. There may be a fine line between reckless speculative market timing and the more responsible work of dynamically managing asset allocation to blunt risk. But a growing number of advisors say they recognize that there is, in fact, a distinction between the two.
Nick Olesen, a partner with the Philadelphia Group, which works mostly with wealthy families in the Philadelphia region, has also incorporated more technical analysis into his management strategy. He mainly wants to minimize the potential for deep losses that can strike buy-and-hold portfolios.
“I’m not a day trader,” Olesen emphasizes. Instead, he’s watching broad trends, and also looking for entry and exit points. He and his colleagues build portfolios mostly with ETFs. “Preservation of capital is the No. 1 priority for most clients,” he says. Adding technical indicators—including relative strength and moving-average-convergence-divergence (MACD) metrics—to asset management folds into that priority of capital preservation, he insists. “Investors are starting to realize … you’d be better off following even the basic technical indicators.”
Jerry Miccolis, of Brinton Eaton Wealth Advisors, agrees, and he has the model to prove it. In recent years, he’s helped redesign his Madison, N.J., firm’s risk management practices by adding a role for technical analysis. Moving averages, in particular, are among the first tools he uses to monitor price momentum, giving him a deeper perspective on his asset allocation adjustments.
But he stresses that price momentum is only one of several tools in the company’s risk-management process. He says there are three kinds of risks to manage. One is market volatility, which is usually managed effectively with an expansive definition of asset allocation and periodic rebalancing. Another risk is the potential for market crashes—sudden but relatively short-lived meltdowns, such as the selling wave that struck after Lehman Brothers collapsed in September 2008. Miccolis manages this threat with a “tail risk” hedging strategy that’s designed to limit, if not offset, market volatility’s darkest moments (see Can Volatility Be Tamed? in the September 2011 issue of Financial Advisor). The third risk he faces is a sustained price decline over a longer time frame—in a traditional bear market. Miccolis deals with this hazard by looking for warning signs via price momentum signals.
Brinton Eaton currently applies its momentum-based analysis to domestic equity allocations, though the firm has considered expanding the technique to other asset classes. Meanwhile, it divides its U.S. equity portfolios into primary sector components (energy, financials, etc.) via ETFs. Analyzing short- and longer-term moving averages, Miccolis looks for early clues that trouble is brewing. This technical tool kit is built around 50- and 200-day moving averages, although the details vary—“we optimize [the moving averages] for each sector.” The basic idea is that prices that fall below the moving averages send a bearish signal, and vice versa. (Miccolis and colleague Marina Goodman profiled their momentum strategy in detail in the February 2012 issue of the Journal of Financial Planning: a paper called Dynamic Asset Allocation: Using Momentum to Enhance Portfolio Risk Management.)
The Usual Caveats
Advisors who use technical indicators are usually keenly aware of the criticism that surrounds these tools. The skepticism against most, if not all, forms of market timing runs deep in the financial planning community, and for good reason. There are relatively few verifiable, real-world track records that offer clear and decisive evidence to support market-timing strategies after the investor adjusts for risk, trading costs and tax impacts.
But nobody can prove that the tools are worthless either. Several recent studies bluntly conclude that advisors who use moving averages tactically in managing asset allocation improve their performance, reduce risk or both. A market-timing strategy powered by a 200-day moving average improved risk-adjusted returns substantially from 1973 through 2008, Mebane Faber finds in “A Quantitative Approach to Tactical Asset Allocation” (in the Journal of Wealth Management, Spring 2007). A version of the paper at the Social Science Research Network (SSRN.com) is reportedly the site’s No. 1 download (out of 10,000 papers) for the year through this past September and the No. 2 download overall.
Moving averages have captured the imagination (and increasingly the managed money) of advisors these days, and it’s easy to see why, at least through the lens of history. Consider a simple strategy benchmark with an initial weighting of 60% stocks (represented by the S&P 500) and 40% bonds (by the Barclays Aggregate Bond index). Buying and holding this mix earned you an annualized total return of 7.7% for the 20 years through August 2012 while it gave you an annualized volatility (standard deviation) of roughly 10.7. By contrast, your performance would have considerably improved with a market-timing strategy that adjusted the same initially weighted allocation using signals from a simple 10-month moving average (roughly the equivalent of a 200-day average). You would have seen a return of 9.3% a year and volatility of 7.9 (see Figure 1).
Here’s how the moving average strategy in Figure 1 works: When the equity index falls under its 10-month moving average (based on monthly data) at any month’s end, the entire stock allocation is moved to cash (three-month T-bills). There it stays until the equity index closes above its 10-month average, at which point all the cash is shifted back to stocks. The same rule applies to bonds. In short, the equity portion of the portfolio is either in stocks or cash, and the remaining fixed-income allocation is either in bonds or cash. The result is that this moving average strategy would have sidestepped the worst of the corrections and crashes. If that sounds familiar, it’s because similar results have been documented in numerous studies through the years.
Searching For Godot
You can, of course, generate a wide variety of results by changing the rules for each strategy. Instead of a 10-month moving average, you could use a five-month average. Or you could minimize the extreme allocation swings for the moving-average strategy with a middle-of-the-road approach that changes allocations gradually. Alternatively, the rebalancing trigger points can be raised or lowered to make the strategy more or less sensitive to market fluctuations. The sky, in fact, is the limit for variations on the basic themes, reminding us that you can torture the data to say almost anything you want.
What you can’t do is decide, with certainty, how any one strategy will fare in the years ahead. What is clear is that a true buy-and-hold strategy is probably impractical for most individuals, and perhaps for most institutions as well. Even if you have a lengthy time horizon, a set-it-and-forget-it mind-set isn’t realistic in a world that has seen such extreme market volatility. The main question for most investors is how to manage the volatility. More advisors are using technical analysis as part of the answer.
Is this a change for the better? Much depends on how the indicators’ signals are applied. Whatever benefits technical analysis offers, there are risks to consider and manage. That starts with the glut of noise, when investors are whipsawed by both bullish and bearish signals.
Still, some are wary of technical analysis. One skeptic is Bill Bernstein, author of several popular investment books, including The Intelligent Asset Allocator: How To Build Your Portfolio To Maximize Returns and Minimize Risk. He says that managing asset allocation this way is fairly subjective. It requires the investor to choose his or her own start and end dates, portfolio composition, trigger points for rebalancing, assumptions about trading costs and the length of moving averages. The questions keep coming (and the conclusions are murky).
“The best you can do, I think, is to make some broad heuristic observations,” he says.
He notes that asset classes tend to exhibit momentum in the short run, which eventually gives way to mean reversion. Even though the nine-month moving average “has stood the test of time and empirical studies,” he says, he remains a skeptic. “Combining rebalancing with moving averages sounds like the new new thing, but it’s just one more fad,” he says.
Even if advisors are using moving averages as a supplement—not a replacement—to conventional asset allocation, no one can prove it will help. But its track record looks encouraging.
Figure 2 shows the differences in one-year returns for the moving-average strategy minus the returns for the buy-and-hold strategy. The dots above the zero mark indicate that the moving-average strategy outperformed for the trailing-12-month period, and vice versa. For much of the past two decades, annual returns between the two strategies shared relatively similar results. But the differences widened dramatically around and during recessions—overwhelmingly in favor of the moving-average strategy.
For this reason, finance professor Paskalis Glabadanidis calls moving average-based strategies the equivalent of an “at-the-money put option combined with a long position in the underlying risky asset” (a quote from his working paper, Market Timing with Moving Averages.) In other words, the main value of moving averages has kicked in when the market has trended lower for an extended stretch—a bear market.
None of this should be surprising, says Adam Grimes, the chief investment officer of Waverly Advisors and author of the recently published book The Art and Science of Technical Analysis. “The major crashes usually come well after warnings signaled by technical weakness.” The steep sell-off in the stock market in late 2008 and early 2009, for example, started about a year after equities set new highs. Soon after the peak, investors saw a series of warnings in the moving-average signals.
That’s not unusual, notes Grimes. He adds, however, that there’s nothing magical about 50- or 200-day moving averages—or any other rules for calculating average prices. Moving averages, in all their variations, are simply tools that quantify some of the “repeatable patterns that illustrate the psychology of the markets.”
The main advantage of looking at prices through the prism of trailing averages is that it takes a lot of the emotion out of analyzing market trends, he counsels. “You’d be much better off with this than making emotional decisions,” Grimes says. Is it foolproof? No, of course not. “We don’t deal in certainties—we deal in probabilities.”