Matthew X. Hanauer and Steffen Windmueller, two scholars affiliated with the Technical University of Munich, compare the performance of three risk management approaches applicable to the momentum strategy. Their new paper also explores the risk management techniques available for hedge fund managers and others who pursue a momentum strategy.
A “momentum” strategy sometimes compared to the strategy of the gambler who puts his money on red at the roulette wheel because it has been stopping a lot on red lately. It is typically contrasted with contrarianism, the strategy of the gambler who puts his money on black when the wheel has been stopping on red lately.
Since streaks do end, momentum strategies have their drawdowns. Sometimes very markedly so, as in 2009, when the momentum factor for both US and non-US equity markets experienced great losses.
The existing literature proposes the following three approaches: idiosyncratic momentum, constant volatility scaling, and dynamic scaling.
Defining Risk Management Approaches
“Idiosyncratic momentum” simply means that a momentum strategy can be adjusted to draw only on the idiosyncratic returns from market regressions. This removes the return component from market beta exposure, and that means lessens the vulnerability what there is a market-wide crash (or a reversal to the upside).
“Constant volatility scaling” is an approach akin to the maintenance of a risk parity portfolio. It involves weighing different instruments depending on the ratio between a constant target volatility and realized volatility.
“Dynamic scaling” weighs instruments depending on the ratio between expected market returns and realized volatility.
Hanauer and Windmueller find, first, that combining the momentum strategy with any of these approaches works. Each style decreases momentum crashes producing an increase in risk-adjusted return. Further, both within the US specifically and on a global scale, each is implementable.
The Simplest Approach is the Best
Which is best? Perhaps not surprisingly, the simplest is the best, “in a multiple model comparison test that also controls for other factors, idiosyncratic momentum emerges as the best momentum strategy.”
The improvements in results that come from using only idiosyncratic momentum are “more than twice that of the improvements of volatility-scaling strategies [either constant or dynamic] and the reduction in maximum drawdowns is the highest.”
This study did not explicitly measure the after-trading cost performance associated with the three different approaches, or the effectiveness of any transaction cost mitigation techniques. This might be material for follow-up studies. In a footnote the authors observe that the idiosyncratic approach may lead on average to a portfolio of larger stocks, with “higher liquidity and lower transaction costs” than the other approaches.
But these authors did look more deeply into a related question: how high would transaction costs have to be in order to render the strategies’ returns statistically insignificant? Looked at from this perspective, too, it is the idiosyncratic momentum approach that excels. “[F]or the global sample, idiosyncratic momentum clearly gives the highest bounds for all types of round-trip costs.”
As for the distinction between the two different methods of scaling volatility: the result there depends on the metric.
The January Effect
Momentum strategies in general are subject to a January effect. That is, at the end of a new year portfolio managers make adjustments, often going short on prior losers, betting on downward momentum. Some of those prior losers, especially the small-cap stocks among them, reverse their fortunes in January: enough so that January is a negative month—the only consistently negative month—for the momentum strategy in general.
Again, the idiosyncratic approach is the winner. It outperformers the other approaches tested in protecting the strategy against the chill winds of January.
Looking to Japan
There has been a great deal of discussion in the literature about why the momentum strategies works in some countries better than others. According to a 2010 paper by Chui, et al., momentum is a persistent factor around the world outside of Asia. Other scholars have said that Japan is immune from momentum.
Hanauer and Widmueller seek to contribute to this line of investigation as well. They agree that momentum as a strategy has weak (through it still has positive) results in Japan. But they indicate that in Japan, too, “idiosyncratic momentum emerges as the best momentum strategy.”