Value And Momentum: Everywhere, But Not All the Time

“Momentum” and “Value” strategies have had well-documented return premia in multiple geographies and asset classes (Asness, Moskowitz, & Pedersen 2013). Average monthly returns to momentum are larger than average returns to value, caveated by large pullbacks (“crashes”) in the momentum portfolio. Practitioners often include both approaches in their investment strategy.

In this report, we present a dynamic risk-weighting scheme. Historically, this scheme outperforms both value and momentum strategies, as well as a naïve equal-weighting of the two, by capturing the upside of momentum while avoiding large drawdowns.

• Dynamically weighting value and momentum strategies by a function of the trailing volatility in the momentum portfolio produces a superior information ratio (IR), total return, and lower maximum drawdown compared to a naïve equal weighting.
• Results are consistent in six regions (U.S., Europe, Asia Ex-Japan, Japan, Latin America, and Emerging Markets) and in multiple robustness checks. We maintain dollar neutrality and persistent leverage of 1.0 in all specifications.
• Monte Carlo simulation supports the conclusion that the shift of tail density from left- to right-tail drives the performance improvements. That is, large drawdowns are avoided.