Momentum Crashes
Read the paperopens doi.org in a new tab
What they found
Momentum's strong average returns hide rare but devastating crashes. The authors document that in U.S. stocks since 1927 the momentum strategy lost the bulk of its value in a few episodes, most famously 1932 and 2009, when a market rebound after a crash caused the 'loser' portfolio (the short side, full of beaten-down high-beta stocks) to rocket. These crashes are somewhat predictable: they occur after bear markets when volatility is high. A version of the strategy that scales exposure by forecast volatility roughly doubles the Sharpe ratio and dramatically reduces the crashes, and the same pattern appears in international stocks, currencies, commodities, and bonds.
What you can use
- Momentum is left-skewed: it earns steadily and then loses years of gains in months, especially in market rebounds.
- The danger is on the short side; after a bear market the 'losers' are the most sensitive to a recovery.
- Cutting momentum exposure when volatility is high and the market has just fallen historically avoided the worst episodes.
- Never size a momentum strategy off its average return alone; size it off its worst months.
Caveats
Volatility-scaling improvements are in-sample by construction and depend on volatility forecasts; the effect is smaller in some replications. Crash timing is identifiable only in broad regimes, not to the day.
Tags: momentum, tail-risk, volatility-scaling, crashes
Summaries are our own reading of the paper, not the authors' words. Educational only, not advice. Discuss it in Book Club.