May 30th, 2013
by Jim Fink, Investing Daily
A long literature finds that, on average, value stocks outperform growth stocks and stocks with high positive momentum outperform stocks with low positive momentum.
(Clifford Asness, AQR Capital Management)
Every investor wants to “beat the stock market,” but we all know that it is easier said than done. Becoming a superior value investor like Warren Buffett is probably the best and most sustainable way to outperform index investing, but Buffett’s version of value investing requires years of experience and expertise in fundamental analysis that most will never attain.
Outside of Buffett-like prowess, the holy grail of investing is finding a quantitative and mechanical methodology based on a simple set of screening criteria that anybody can follow. In my previous article Mechanical Investing and Fundamental Indexing: Be a Quant!, I provided a few examples of pre-defined stock screens that have good back-tested results. Looking back, I’ve found a few of the most important characteristics of the stocks selected from these successful screens:
- Low valuation (based on ratios of market price to cash flow, earnings, or sales)
- Small company market capitalization
- Price momentum
In Buy Small-Cap Stocks Before They Grow Up, I discussed the Fama-French Three-Factor Model (TFM), which successfully predicted 95% of a stock portfolio’s future return based on the first two factors listed above (value and small size), as well as stock price volatility (i.e., beta).
In The Great Investment Truth Behind Simple Arithmetic, I discussed how downside volatility is much more destructive of wealth than upside volatility is beneficial, so the higher a stock’s beta, the greater the chance that the stock will suffer a large stock decline that destroys wealth. Investors should require a higher return potential from such stocks to compensate for this risk (the fact that they don’t is one of the great anomalies of finance). Similarly, small-cap stocks are arguably more vulnerable to economic recessions (due to fewer financial resources to weather bad times) and low-valuation stocks arguably sport low valuations because their businesses are distressed and thus at risk of never recovering to full health.
Advocates of market efficiency argue that these added risks cause stock prices of these companies to sell at a discount and subsequently generate abnormally-large returns if the companies end up succeeding. On the other hand, advocates of market inefficiency, argue that small-cap and value stocks outperform for irrational behavioral finance reasons rather than the rational discounting of increased risk. Specifically, small-cap stocks are underfollowed and ignored by Wall Street brokers/analysts because they earn their money from advising large institutional investors that focus on large companies. Since many individual investors unfortunately rely on Wall Street broker-salesmen for “advice” (despite the fact that these salesmen are not fiduciaries), individual investors also unjustly ignore small caps.
Value stocks that aren’t troubled but simply slow growers are similarly ignored unjustly because investors emotionally overpay for high growth (the “growth trap”) and underpay for companies with lower growth. It’s the essence of human emotion to engage in overshooting and undershooting behaviors and in the context of the stock market this means selling stocks until their market prices are pushed below intrinsic value and buying stocks until their market prices are pushed above intrinsic value. Shrewd investors then swoop in to take advantage of these stock-price inefficiencies by buying negative-momentum value stocks and selling positive-momentum stocks.
Momentum is a fascinating investment phenomenon because it can be exploited in two completely opposite ways. Value investors profit if they successfully bet that the momentum has gone too far and will reverse direction, whereas momentum investors profit if they successfully bet that the momentum will continue in the same direction. Isaac Newton’s first law of motion is:
An object in motion tends to remain in motion, and an object at rest tends to remain at rest.
Newton’s experience in 1720 investing in the South Sea Company demonstrated both the promises and pitfalls of momentum investing. He initially bought some South Sea stock and then sold it after momentum had earned him a 100 percent profit. Presumably, he sold the stock because he thought the stock’s momentum was about to reverse. But the stock kept going up and Newton decided to jump back in right at the top, expecting the momentum to continue, only to watch the momentum peak and reverse and he ended up losing not only all of his earlier profit but much, much more. Newton famously stated:
“I can calculate the motion of heavenly bodies, but not the madness of people.”
Momentum by its very nature continues in the short term and reverses in the long term. A rubber band can stretch very far, but it eventually snaps back. Academics Fama and French recognize that momentum in stock prices exists, but have argued that the effect only lasts a few months and requires so much trading that the transaction costs eat up all of the abnormal returns. On the other hand, in a 2011 paper, they concluded:
We find strong momentum returns everywhere, except Japan . . . . Last year’s winners show positive momentum returns in all size groups, but persistence is stronger for small stocks, especially microcaps. (pp. 9, 11)
The evidence of short-term momentum is so strong that Fama and French appear ready to update their three-factor model to a four-factor model to incorporate momentum effects.
Combining Value and Momentum is the Holy Grail
The holy grail of investing involves finding investments that individually produce strong returns over time, but that also have a negative correlation with each other that reduces or eliminates downside volatility of the portfolio as a whole and smooth out returns. Benefiting from momentum both ways – as a short-term continuation trade (a.k.a. momentum investing) and as a long-term reversal trade (a.k.a value investing) — may be the holy grail.
In a 2012 paper entitled Value and Momentum Everywhere, hedge-fund manager Clifford Asness of AQR Capital studied value (low price to book value) and momentum (12-month price appreciation) characteristics in the stock markets of eight different countries and found that both significantly outperform everywhere in the world (except that momentum doesn’t work in Japan):
We find consistent and ubiquitous evidence of value and momentum return premia across all the markets we study. We also highlight that studying value and momentum jointly is more powerful than examining each in isolation. The negative correlation between value and momentum strategies and their high positive expected returns implies that a simple combination of the two is much closer to the efficient frontier than either strategy alone, and exhibits less variation across markets and over time.
Asness found that a value portfolio rebalanced annually and a momentum portfolio rebalanced monthly both outperformed the overall stock market (Figure 2, page 42), with momentum’s outperformance almost double value’s outperformance. But what was truly amazing is that a 50/50 combination portfolio of both strategies performed best of all by almost double the momentum strategy’s outperformance! The reason for the combo’s superiority is that the value and momentum strategies sport an amazingly negative (i.e., good) correlation of -0.65 (best possible is -1.00). This negative correlation makes sense because momentum works when price continues in the same direction and value works when price reverses.
The fly in the ointment is the short-term nature of momentum, which requires costly portfolio balancing each month (pp. 29-30). Asness admits that his study’s gross returns of the momentum strategy would be less if transaction costs were taking into account, but argues that “we focus on an extremely large and liquid set of equities in each market (approximately the largest 17% of firms), where trading costs and price impact and capacity constraints are minimized.”
52-Week Highs: A More Sustainable Measure of Positive Momentum
Asness doesn’t claim that measuring momentum by 12-month price appreciation is the best value of momentum (pp. 31-32), but wanted to keep things simple. Fortunately, there is another measure of momentum that offers robust outperformance and yet is much more long-lasting and doesn’t require monthly rebalancing to work!
As I mentioned in January Effect and 52-Week Highs/Lows: How to Beat the S&P 500, measuring a stock’s price momentum by the relationship of its current price to its 52-week high is more powerful in predicting future returns than the magnitude of 12-month price appreciation. In fact, in a 2010 paper, the authors discovered that whereas positive momentum based on 12-month price appreciation peters out and reverses after the following 12 months (interestingly, negative momentum based on 12-month price depreciation is longer lasting), positive momentum based on closeness to the 52-week high does not reverse even after the following 24 months! (pp. 2158-2161).
Two Distinct-Style Portfolios or One Composite Portfolio?
The impressive results from Asness’ value and momentum study involved a 50%-weighting to one group of pure value stocks and a 50%-weighting to another group of pure momentum stocks. But another way to construct a value and momentum stock portfolio to search for a single set of stocks that possess both value and momentum characteristics. As I wrote in "What Works on Wall Street” and Trending Value: Best Stock Screen of All Time!, author James O’Shaughnessy has formulated a stock screen called “Trending Value” that filters stocks in two stages. First, it screens for value stocks based on a composite of six different low-valuation criteria (e.g., price-to-earnings, price-to-sales, price-to-free cash flow). Second, from this list of value stocks, it selects the value stocks with the highest six-month price momentum. Consequently, it is a value screen first and foremost with a secondary sorting by price momentum.
The back-tested performance of the “Trending Value” screen is impressive and the rationale behind the screen’s criteria are persuasive: value stocks with some price momentum are not value traps of deteriorating businesses destined to remain cheap, but are companies investors have started to buy because they recognize that the distressed business is “on the mend.” In other words, the price momentum vindicates the cheap valuation. Similarly, the cheap valuation vindicates the price momentum. Because the stocks are still valued cheaply, they are not growth traps, and the price momentum that has occurred is likely just the beginning and will last much longer than the price momentum on stocks that already sport expensive valuations that reflect unrealistic expectations of future growth.
But none of the performance or rationale behind O’Shaughnessy’s “Trending Value” screen proves that its single, composite portfolio of value stocks sorted by price momentum is better than Asness’ 50%/50% combined portfolio of two pure groups of value and momentum stocks. For one thing, Asness’ methodology benefits from the awesome -0.65 negative correlation between the two separate stock groups, whereas O’Shaughnessy’s single portfolio of stocks enjoys no such benefit. Of course, stocks that exhibit both value and momentum characteristics are probably much more stable and suffer less downside volatility than pure value and pure momentum stocks, so one could argue that the Asness stocks need the negative correlation effect much more than O’Shaughnessy’s stocks would.
A second factor in favor of Asness is that pure-bred value and momentum stocks may simply perform better than composite stocks that are restrained by their mixed nature from exhibiting brilliance. This is especially true with regard to momentum stocks since O’Shaughnessy’s stock screen is really a value screen that only secondarily sorts by price appreciation and thus treats momentum as a second-class citizen. Asness gives momentum the equal credit with value it deserves.
On the other hand, pure value stocks will have a much larger percentage of value traps and pure momentum stocks will have a much larger percentage of growth traps that a composite model would have weeded out. Furthermore, Asness rebalanced the momentum stocks monthly – which likely increases transaction costs substantially – whereas O’Shaughnessy rebalances his single portfolio of stock hybrids annually.
Bottom line: A small-cap equity portfolio composed of 50% value stocks and 50% momentum stocks based on 52-week highs may turn out to be the holy grail of stock investing.
Fama and French’s fourth possible outperformance risk factor — beta — is problematic because of the anomalous conflict between theory and experience, so I’ll have to think more about that factor before including it in any stock-picking methodology.