Graveyard / Momentum / US stocks
Dual momentum (Global Equities Momentum)
Does dual momentum beat buy and hold?
The idea
Dual momentum picks the stronger of US or international stocks, and moves to bonds when stocks lose to cash. The book version is Gary Antonacci's Global Equities Momentum, with a public out-of-sample record from 2015.
How we tested it
| Data | SPY, EFA, AGG and a bill proxy (SHY). Month-end, at the closing auction: if SPY's 12-month return beats SHY's, hold the better of SPY and EFA over 12 months, else hold AGG. A single variant, defined before results. A volatility filter variant was also tested. |
|---|---|
| Period | About 22 years (39 switches). Sub-periods 2002-2014 and 2015 onward. |
| Costs | 1 bp per side; stress 6.5 bp, immaterial. |
| Holdout | Full, in-sample (80%) and out-of-sample (20%) gate, plus a volatility filter. |
| Control | None stated. |
Result
| Trades | 39 |
|---|---|
| Sharpe full / in-sample / out-of-sample (11 trades) | 0.63 / 0.60 / 0.73 |
| Max drawdown / profit factor | −35.9% / 5.03 |
| Sharpe 2002-14 / 2015 onward | 0.82 / 0.47 |
| Sharpe in bear / bull / chop markets | −0.96 / 2.53 / 0.07 |
| With volatility filter, full Sharpe | 0.45 |
Why it failed
It is slow trend following on equity beta. It was whipsawed in 2018, 2020 and 2022, with a −36% drawdown and only 39 switches in 22 years. The public record since 2015 (Sharpe 0.47) is weaker than before. The volatility filter made it worse.
What would make it worth retesting
No reopen condition is stated. It belongs to the same family as the earlier time-series momentum test.
Past results of a backtest say nothing certain about the future.