How to Beat an Efficient Market

Ugly Is Beautiful Yes, that's a click-bait headline. The previous time I tackled this subject, my column title was the unappetizing "There's More to Expected Return Than Risk." That article's combined viewers could have gathered in the back of an independent bookstore, for a reading by Maine's leading poet.

But while click-bait, the headline is not inaccurate.

A bit of theory, before getting to the recommendations. The efficient-market hypothesis (EMH) is largely correct. The notion that the collective wisdom of all market participants is very difficult to outsmart cannot be questioned, as evidenced by 40 years' worth of index fund performance.

However, stock-pricing models that accompany the EMH are problematic. The trouble is that people tend to confuse the models with reality. They talk of "anomalies" that the model cannot explain as if these are investor mistakes. (One example of an anomaly: In William Sharpe's capital asset pricing model, where stock returns are attached to the single indicator of beta, lower-beta stocks tend to perform better than the model success, and higher-beta stocks perform worse.)

Such a claim is inconsistent with the spirit of the EMH, which states that investors are collectively rational, not collectively error-prone. Why believe that the people are wrong and the model is right? The truth almost surely is the opposite.

A new, richer model of stock-pricing is needed--one that can incorporate the many aspects that influence investor decisions, not just a single factor (as with CAPM) or four factors (as with the Fama-French-Carhart model). Last year, Zebra Capital's Roger Ibbotson and Morningstar's Tom Idzorek laid out such a path, in "Dimensions of Popularity," published in the Journal of Portfolio Management. The article suggested that the anomalies mindset be reversed. Rather than mine data to find anomalies, and then searching for reasons to explain those results, researchers should be thinking about aspects of popularity.

(This concept, as with most, follows in the footsteps of other works; the ideas are not brand new, but rather reworked and clarified from previous versions. Indeed, Roger Ibbotson along with two co-authors--Jeffrey Diermeier and Laurence Siegel--articulated some of these ideas a full three decades ago.)

A popular stock is a stock that has desirable characteristics. On the whole, investors find such stocks easy to own. As a result, they are willing to accept a lower rate of return on those securities than they are with unpopular stocks, which for various reasons may be unpleasant holdings. The search for higher return thus becomes the search for the unpopular--along with a willingness to accept their warts.

The popularity concept, unlike that of the current framework of expected model returns plus anomalies, does not assume that the market functions in mysterious ways. Nor does it necessarily posit investor irrationality (although it can permit such a thing, by offering a behavioral-finance explanation for a source of unpopularity). By greatly expanding the potential reasons that investors use when valuing stocks, popularity provides a better framework for thinking about ways to achieve higher returns.

Four Findings Here are some examples, from an unpublished draft paper that Ibbotson, Idzorek, and Morningstar's James Xiong are now writing. Some are familiar, others less familiar. At this stage, the list is highly preliminary; many other sources of popularity remain to be catalogued. But it should give a flavor of the endeavor--and perhaps even an idea or two for how you might wish to tilt your portfolio:

1) Value The historic outperformance of value stocks breaks traditional pricing models. As the authors point out in their new paper, "deep value is less risky than deep growth," as measured by volatility, yet deep-value stocks have handily beaten deep-growth companies over time. But a behavioral preference for the apparent safety of growth companies, which tend to be healthier businesses with stronger corporate brands, can explain what the CAPM cannot.

2) Low volatility As mentioned earlier in this column, stocks with relatively low betas (or volatilities, to use the more general term) are often said to be anomalous because their performance, broadly speaking, keeps pace with that of less-predictable stocks, which should not be the case if return falls in line with risk. There is a popularity explanation, though. Because few investors use leverage, and most investors wish to beat the overall market index (particularly active professional managers, who are employed directly to achieve such a feat), there is a crowding effect. Too much money pursues the market's high-beta stocks, thereby pushing down their expected returns. They are too popular. Conversely, low-volatility stocks are relatively underappreciated.

3) Liquidity This topic is easy. Liquidity kills pricing models because often the less liquid securities show up as being less risky. Since they do not trade very often, their prices can be sticky, so that they show less volatility than a security that can be readily traded. That is exactly backward. A lack of liquidity is an unmitigated bad thing; customary pricing models are bamboozled by the issue; and the popularity approach sensibly states that securities that are less liquid should be expected to have higher future returns, to compensate for their trading drawbacks.

4) Severe downside risk This was a new one for me. Although investors aren't particularly afraid of high-volatility stocks in general, as stated in the second point above, they make an exception for securities that have particularly steep downside risk. It's sensible, of course, to dislike securities that might crater! But per behavioral theory, where people feel losses more strongly than gains (as articulated by Larry Bird: "I hate losing more than I like to win"), the dislike extends further than can be explained by pricing models. The models do not incorporate behavioral findings--but people do, as does the popularity concept.

Looking Forward Much work remains. At this stage, the recommendation is the limited one of favoring relatively illiquid value stocks that might get whacked. That doesn't sound particularly palatable, does it? Tastes bad, performs well--such is the idea behind the popularity concept.

(As you may have noticed, I left low volatility out of the recommendation. While low-volatility stocks fare well on a risk/return basis, they do not necessarily outperform on return alone--and the stated purpose of this column is to identify return opportunities, setting risk aside. That said, tilting toward low-volatility securities does make sense, for taking some of the sting out of the portfolio, but that by itself will not goose returns.)

You may, perhaps, be asking if the popularity concept can help to clarify the chaos known as smart-beta funds (which we call strategic-beta funds). Yes, I think it can. The claims of strategic-beta promoters can now be put to the test. Is there a credible, ongoing reason to explain why a strategic-beta fund's holdings are unpopular? Or is the logic wanting, suggesting that in creating the fund, the sponsoring fund company tortured the data until the numbers confessed?

That is a project that I alone cannot tackle, nor the authors either. It will require a concerted Morningstar effort. Happily, I think the will is there. It would be splendid to see each strategic-beta fund classified according to the source--or sources--of unpopularity that it seeks to exploit. Those cats very much need to be herded.

John Rekenthaler has been researching the fund industry since 1988. He is now a columnist for Morningstar.com and a member of Morningstar's investment research department. John is quick to point out that while Morningstar typically agrees with the views of the Rekenthaler Report, his views are his own.