Valuing Momentum: Part 2
In part two of this series, Cameron Hight references Cliff Asness’ paper on momentum and what it means for Alpha Theory and value investing.
I’ll highlight one major article written by Cliff Asness and his team at AQR, published in May 2014 (it’s also worth checking out “What Works on Wall Street” by O’Shaunassy and their fund strategies which combine value and momentum and have solid long-term track records). The AQR piece titled “Fact, Fiction and Momentum Investing” evaluates some of the most prominent myths regarding Momentum and uses empirical research to refute those myths. In doing so, it gives a compelling account showing why the marriage of value and momentum are potent partners. Here are a few excerpts to give a sense of their conclusions:
As we’ll show in this essay, value and momentum work better when used as complements, and it is the combination of the two we stress and most-strongly recommend. We are fans of both momentum and value but bigger fans of their combination (and not fans of myths at all).
Evidence for Momentum
The (momentum) return premium is evident in 212 years (yes, this is not a typo, two hundred and twelve years of data from 1801 to 2012) of U.S. equity data,3 dating back to the Victorian age in U.K equity data,4 in over 20 years of out-of-sample evidence from its original discovery, in 40 other countries, and in more than a dozen other asset classes.
88% of returns positive for momentum and 89% for value.
Israel and Moskowitz (2013) show that the long and short side of momentum is equally profitable using 86 years of U.S. data as well as 40 years of international equity data, and another 40 years of data from five other asset classes outside of equities. Everywhere they looked and in every way, they could not find any evidence that the short side profits were systematically larger or more important than the long side.
Benefits of Momentum and Value Combined
Sharpe ratio and percent of years with positive returns increase with a 60% value / 40% momentum strategy.
Suppose, despite all of the evidence to the contrary and our strong belief it’s positive, momentum had a zero expected return going forward. Would it still be a valuable investment tool? The answer is clearly, though perhaps surprisingly, yes. The reason is because of momentum’s tremendous diversification benefits when combined with value.
The diversification benefits are so great that even a zero expected return would be valuable to your portfolio! The logic is simple. Since value is a good strategy and momentum is -0.4 correlated with it, one should expect momentum to lose money based only on that information. Yet, the fact that it does not lose but in this assumed case breaks even makes it a valuable hedge. (We note that using the definition of value in Asness and Frazzini (2013) dramatically increases the magnitude of this negative correlation (to -0.7) and the power of combining value and momentum. Following their methodology, the results of this section would be far stronger.)
But, there’s an even simpler and equally effective way to mitigate these crashes, as we mention repeatedly: combining momentum with value. This combination has effectively eliminated these crashes in our long-term sample evidence — and not just those for momentum but also the crashes that can occur for value investing. In other words, the diversification benefits of combining momentum with value don’t just appear during normal times, but also during these extreme times, which makes their combination even more valuable. For example, Asness and Frazzini (2013) show that the combination of value and momentum did not suffer as badly in 2009. Going the other way, in 1999 momentum helped ameliorate value’s pain. Both factors have worked well over the long-term, but neither has a Sharpe ratio of 10, meaning that both will have hard times occasionally, but when combined together they will have fewer hard times. Using Kenneth French’s data, we can show similarly that these very poor episodes for momentum and value are ameliorated. The diversification benefits between momentum and value are evident, even during these extreme times. For example, the worst drawdown over the full sample is -43% for value, -77% for momentum, but only -30% for a 60/40 combination of value and momentum.
By the way, we fully recognize and acknowledge that the past ten years have not been great for momentum, with the 10-year return for UMD (Momentum) falling in the 7th percentile of rolling 10-year returns (going back to 1927). At the same time, the past ten years have not been great for value, either, with the 10-year return for HML (Value) falling in the 5th percentile of rolling 10-year returns. That, of course, makes the prior 10-year return of the 60/40 combination of the two low (2nd percentile), but still positive (12%). You know a strategy has a pretty great history when the 2nd percentile return is still positive.
Summing up the points from the AQR paper:
- Momentum works better with value (negatively correlated with each other)
- The better the value mechanism the better the whole portfolio performs (see the bolded section in the excerpt above)
This is where our clients shine. They are great value estimators and their research is not easily systematized. What should be systematized is the translation of that research into a portfolio and a new push for Alpha Theory will be to give our clients tools to incorporate momentum.
Active manager's search for alpha is more difficult today than it has ever been. There is an existential requirement for active managers to leverage the tools and evidence around them and maximize the return they get from their research. To that end, over the coming months, you will see Alpha Theory develop new functionality to better account for momentum in position sizing. We welcome your input as we embark upon this journey.