Cross-asset momentum: applying momentum signals across multi-asset universes

Time series momentum—the tendency of an asset's recent returns to predict its future returns—is one of the most robust empirical findings in systematic investing. The strategy works in a single asset class, and it works better when applied simultaneously across many asset classes. Cross-asset momentum is the formal name for the multi-asset version, and it has been the basis for some of the most consistently profitable systematic strategies of the last forty years.

What cross-asset momentum is

Cross-asset momentum applies the same momentum signal—typically the trailing 12-month return, or a longer/shorter variant—to a universe of assets that spans multiple asset classes: equity indices, fixed income futures, commodity futures, and currency pairs. Each asset is evaluated on its own historical return; the signal is the same across asset classes; the resulting portfolio is the diversified bet on the assets in positive trend, against (or absent) the assets in negative trend.

The approach is distinct from cross-sectional momentum, which ranks assets within a single universe (typically equities) by their relative performance and goes long the winners and short the losers. Cross-sectional momentum produces relative bets within an asset class; cross-asset momentum produces directional bets across asset classes, with each asset's signal relative to its own recent history rather than relative to other assets.

How it works

The standard cross-asset momentum implementation evaluates each asset's trailing return over a defined window (typically 12 months, with shorter windows of 1–6 months sometimes used in combination). If the trailing return is positive, the asset receives a long signal; if negative, a short signal (or a zero allocation in long-only versions). Position sizes are typically scaled by inverse volatility so that each asset contributes equally to portfolio risk regardless of the asset class's underlying volatility level.

For a universe of, say, 60 assets across equity indices, fixed income futures, commodity futures, and currency pairs, the strategy produces a portfolio of 60 risk-balanced positions in the directions indicated by each asset's signal. Rebalancing occurs monthly, with the portfolio updated to reflect each asset's current signal and current volatility.

The diversification benefit comes from the multi-asset structure. Momentum signals across different asset classes have low correlation: equity momentum and currency momentum, for instance, have produced returns that are roughly uncorrelated over long horizons. Combining 60 such signals into a single portfolio produces a smoother return profile than any single signal in isolation, with materially higher risk-adjusted return.

What the evidence shows

Moskowitz, Ooi, and Pedersen (2012) document time series momentum across 58 liquid futures and forward contracts in equity indices, fixed income, currencies, and commodities over 1985–2009. The strategy produced an annualised Sharpe ratio of approximately 1.8—materially above any single asset class's buy-and-hold Sharpe over the same window—with low correlation to traditional asset classes. The result has been replicated in subsequent work on extended samples and different universe definitions.

The mechanism is consistent across asset classes: positive trends tend to continue over horizons of 1–12 months before mean-reverting beyond that, and the same signal extracts positive expected return whether applied to equities, bonds, commodities, or currencies. The behavioural explanations (under-reaction to news, slow diffusion of information across investors) and the rational ones (time-varying risk premia, slow-moving capital) both predict the pattern, and the empirical evidence is consistent with both.

Managed futures funds and CTA strategies, which historically have been the institutional implementation of cross-asset momentum, have produced multi-decade Sharpe ratios in the 0.5–1.0 range with material diversification benefit relative to traditional 60/40 portfolios. The performance has been weaker in the post-2010 period—strong, narrow market leadership in equities reduced the diversification benefit, and the 2010s drawdown for some CTA funds was the deepest in decades—but recovered strongly in the 2022 bond drawdown when most asset classes fell together except cross-asset momentum, which was already short fixed income.

Limitations and trade-offs

Cross-asset momentum's main weakness is its drawdown behaviour during sharp regime changes. The strategy is by construction late to inflexion points: when a long-running trend reverses, the signal continues to favour the old direction until the new trend is established in the trailing window. The 2009 inflexion from the 2008 drawdown produced a meaningful drawdown for cross-asset momentum strategies that were still positioned for the bear market when the rally began.

The strategy is also vulnerable to whipsaws in choppy, trendless regimes. When asset prices oscillate around a flat trajectory, the signal flips frequently and the strategy bears transaction costs without capturing meaningful directional moves. The best regime for cross-asset momentum is one of clear, persistent trends across multiple asset classes—conditions that are not always available.

Implementation requires access to a broad set of liquid futures and forward markets, which is straightforward at institutional scale but more difficult for retail investors. Cross-asset momentum implemented through ETFs alone produces a less diversified version of the strategy and typically lower realised Sharpe ratios than the futures-based version.

Cross-asset momentum in pfolio

pfolio's momentum signal is applied across the full multi-asset universe at each rebalancing—equities, fixed income, commodities, currencies, and alternatives are all evaluated on the same statistical signal. This cross-asset application is the practical implementation of the cross-asset momentum literature. The full methodology is documented at how we build portfolios.

Related articles

• Time series momentum: the academic evidence behind trend-following strategies
• Cross-sectional momentum: how relative performance generates systematic returns
• Momentum investing explained: how price trends generate systematic returns
• Managed futures: how CTA strategies work and their role as a portfolio diversifier