I’m a big fan of Professor Ken French’s data library. I’ve used data from his library at this site many times. If you’re not familiar with Dr. French, he’s a well-known finance professor at the Tuck School at Dartmouth. He’s also known for his long-time association with Eugene Fama at the University of Chicago.
I was digging around some of the files in the library and I was completely stunned by the incredible outperformance of stocks with high momentum, meaning stocks that are surging have a tendency to keep on surging. I was aware of some of the academic literature on this subject, but I have to confess that I was completely dumbfounded by the results.
I know that stocks with favorable valuation characteristics do better than the rest of the market. For example, stock in the lowest decile (or 10%) of price/earnings ratio have historically beaten the market. The same is true for stocks with higher dividend yields or low price/book ratios. Also, small-caps do better than large-caps (although I’m not particularly impressed by the small-cap premium). These phenomena are very well-known and have been documented countless times. But simply put—high momentum creams them all.
At the data library, French has ten portfolios listed by momentum (see “10 Portfolios Formed on Momentum”). He gets his data from the Center for Research in Security Prices at the University of Chicago. I looked at the long-term returns of stocks with the greatest momentum.
From the beginning of 1927 through August of 2007, the overall market has returned an average of 10.10% a year. The highest momentum stocks returned an average of 17.76% a year. What’s more, that’s just the value-weighted portfolio. By looking at the equal-weighted portfolio, which gives more say to smaller-cap stocks, the results are even more impressive. The equal-weighted high-momentum portfolio returned an average of 21.94% a year. Here's the chart:
Also, while the momentum portfolios are more volatile, they don’t strike me as being usually high. The monthly standard deviation for the value-weighted momentum stocks is about 20% greater than the rest of the market. The equal-weighted stocks are 37% more volatile.
The problem I have with many small-cap or value-related models is that the results are highly cyclical. It’s true that small-caps do well after several decades, but it’s not unusual to see underperformance for five years of more. That happened to small-caps in the 1990s and I think value is entering a down phase right now. With momentum, the results are much more consistent. Heck, just look at the red and blue lines.
There’s also the question of what we mean by stocks with high momentum. I called Dr. French just to make sure I had it right and he was very helpful in explaining it to me. By momentum stocks, he ranks every stock by how it did over an 11-month period, then skips a month and then tracks them for one month. At the end of the month, the whole thing is repeated.
Confusing? Here’s an example. On January 1, we take the top 10% of stocks by their performance for the previous January 1 through November 30. The stocks are held for exactly one month and the process is repeated again on February 1. This system in completely mechanistic and all emotions are banished. I’ve known lots of people who are momentum investors but they rarely have the discipline to act by strict rules.
Another interesting aspect of a momentum strategy is the turnover probably isn’t that high. Since it encompasses the best returns for 11 months, many stocks will remain each month. Dr. French said that he thinks the turnover is 91%. That’s high, but not as high as many mutual funds.
There are lots of historically interesting strategies but many are very impractical. For example, the stock market has been net down on Monday, Tuesday and Thursday combined. But it’s highly impractical to sell all your stocks and buy them a few times each week. But I don't think that's the case with a momentum strategy. Also, I’m sure you could even use ETFs to mimic the high-momentum portfolios.
There’s also the question of why. Why do stocks with high momentum continue to outperform for a bit more? Is there something inefficient in their...frothiness? Can a person really play the height of their frothiness all the time by constantly shifting?
According to the Efficient Market Hypothesis, this outperformance would rationally be exploited away. I’m not a believer in efficient Market Theory (ironically, it was developed by Eugene Fama) but I do think stocks show a bias towards efficiency. Perhaps there’s something in a surging stock that causes the pre-requisites for an efficient market to break down (flow of information?).
Dr. French was careful to say that he’s not the discoverer of momentum premium. That award goes to Jegadeesh and Titman who in 1993 found that the best-performing stocks of the last six months outperform the worst over the six to twelve month. I'm very impressed. I’m planning to look into more of the academic literature.