Cross-Asset Momentum
Rank assets across equities, bonds, commodities, FX by 12m momentum; long top quartile, short bottom quartile.
Thesis (edge)
Momentum exists across asset classes. A cross-sectional long/short portfolio based on 12m momentum, vol-scaled, captures the premium while diversifying across assets.
Where it works (regimes)
Works in trending regimes across assets. Can have large drawdowns when many assets reverse together (e.g. 2009, 2020).
Signals
- ( r_{12}(i) ) for each asset ( i ). Rank. Long top quartile, short bottom quartile. Scale by inverse vol.
Portfolio construction
Equal risk contribution or equal dollar within long and short. Rebalance monthly. Max single-asset weight.
Risk model
Tail: correlated reversal. Stress: 2009 reversal. Use vol targeting and possibly trend filter.
Costs & implementation
Moderate turnover. Use ETFs or futures for liquidity. Consider implementation delay (e.g. next-day rebalance).
Failure modes
Overfitting formation period; ignoring transaction costs; illiquid names.
Our Notes & Suggestions
Use liquid, tradeable instruments only. Backtest with 6m and 12m; consider skip-month to reduce crowding. Monitor correlation of longs and shorts.
Our Notes & Suggestions
See the "Our Notes" subsection in the body above for practical guidance, gotchas, and best practices. Always validate regime assumptions and transaction cost assumptions before scaling.
Implementation Checklist
- Define universe and 12m return calculation
- Rank; long top 25%, short bottom 25%
- Vol-scale positions; rebalance monthly
- Transaction cost and capacity check