Dual Momentum Explained: GEM, ADM, and Beyond
Few concepts in systematic investing have proven as durable and well-documented as momentum. The idea that recent performance predicts near-term future performance — that winning assets tend to keep winning and losing assets tend to keep losing — sounds almost too simple. Yet decades of academic research across dozens of markets and asset classes confirm it as one of the most robust return factors in finance.
Gary Antonacci's Dual Momentum framework synthesizes two distinct dimensions of momentum into a single, elegant strategy that has attracted a devoted following among systematic investors. This article explains the foundational concepts, walks through how GEM (Global Equity Momentum) and its variants work, and covers the key considerations any investor should understand before implementing a momentum-based approach.
What Is Momentum in Investing?
The Academic Foundation
In financial research, momentum refers to the empirical tendency for assets to continue moving in the same direction they have recently been moving. An asset that has outperformed over the past several months tends, on average, to continue outperforming over the next several months. An asset that has underperformed tends to continue lagging.
This phenomenon was formally documented at the individual stock level by Jegadeesh and Titman in their landmark 1993 paper, which showed that buying recent winners and selling recent losers produced significant risk-adjusted returns. Subsequent research by Carhart (1997) established momentum as a standalone factor alongside value, size, and market beta in asset pricing models.
Critically, momentum has been found to persist across asset classes — not just individual stocks, but equity indices, bonds, commodities, currencies, and real estate investment trusts. Research by Asness, Moskowitz, and Pedersen demonstrated that momentum works in a broadly consistent way across these asset classes and across international markets. This cross-asset persistence is what makes momentum-based tactical asset allocation compelling: it appears to be a feature of markets rather than a statistical artifact of any particular dataset.
Why Does Momentum Exist?
Several behavioral and structural explanations have been proposed. Behavioral finance researchers point to anchoring, herding, and the slow diffusion of information — investors underreact to new information initially, leading to continued price trends as sentiment catches up. Institutional factors such as performance chasing and career risk among professional investors may reinforce trends. Structural features like index rebalancing and liquidity constraints may also contribute.
The academic debate about the precise source of momentum is unresolved, but the empirical fact of momentum's persistence is not seriously contested in the literature.
Relative Momentum vs Absolute Momentum
Before examining the Dual Momentum framework specifically, it is important to understand the distinction between its two components.
Relative Momentum
Relative momentum (also called cross-sectional momentum) measures how one asset has performed relative to other assets in a comparable universe. If there are five candidate assets, the one with the strongest trailing return over a defined lookback period has the highest relative momentum rank. The strategy selects whichever assets rank at the top.
Relative momentum is useful for rotating among a set of alternatives — for example, preferring U.S. equities over international equities when the former has been stronger, or rotating from equities into bonds when bonds have outperformed. It answers the question: "Of my available choices, which has performed best?"
Absolute Momentum
Absolute momentum (also called time-series momentum) measures an asset's performance against a fixed benchmark — typically its own past performance or a cash return. If an asset's trailing return over the lookback period is positive, it has positive absolute momentum; if negative, it has negative absolute momentum.
Absolute momentum answers a different question: "Has this asset actually made money in absolute terms over my lookback window?" An asset can have the highest relative momentum in a universe of poor performers and still have negative absolute momentum if everything has declined.
This distinction is practically important. During broad market downturns, every equity asset class may have negative trailing returns. A pure relative momentum strategy would still pick the "best" option, which might still be a declining asset. An absolute momentum filter catches this scenario: when the top-ranked asset has negative absolute momentum, the strategy shifts to a safe haven asset rather than investing in the least-bad option.
Gary Antonacci's Dual Momentum Framework
Gary Antonacci, a portfolio manager with decades of experience in quantitative strategies, synthesized relative and absolute momentum into the Dual Momentum framework, most comprehensively described in his 2014 book "Dual Momentum Investing."
The core insight is simple: use relative momentum to select among risk assets, and absolute momentum to decide whether to hold a risk asset at all or retreat to safety. The two filters work together and address each other's blind spots.
Antonacci applied this framework to a streamlined three-asset universe, keeping the strategy simple enough to be reliably implemented and robust enough to avoid overfitting. The result is GEM — Global Equity Momentum.
GEM: Global Equity Momentum
The Three-Asset Universe
GEM operates with three instruments:
- 1. U.S. equities — typically represented by a broad U.S. stock market index ETF
- 2. International equities — typically represented by a developed-market ex-U.S. equity ETF
- 3. Aggregate bonds — typically represented by a U.S. aggregate bond index ETF (the safe haven asset)
The simplicity is intentional. Antonacci's research showed that most of the benefit from cross-asset momentum is captured with just these three instruments. Adding more assets increases complexity without proportional improvement in outcomes, and risks overfitting to historical data.
The Monthly Decision Process
GEM evaluates its universe at the end of each month and follows a two-step decision rule:
Step 1 — Absolute momentum check. Compare the trailing 12-month return of U.S. equities against the return on short-term Treasury bills (cash). If U.S. equities have outperformed T-bills over the past 12 months, the environment is favorable for risk assets and the strategy proceeds to Step 2. If U.S. equities have underperformed T-bills (negative absolute momentum), the strategy allocates 100% to bonds and stops.
Step 2 — Relative momentum check. Compare the trailing 12-month return of U.S. equities against international equities. Whichever has the stronger return receives 100% of the portfolio allocation.
The result is a strategy that holds U.S. equities when they lead globally, international equities when they lead globally, and bonds as a defensive position when the entire equity complex is in a downtrend.
Why 12 Months?
The 12-month lookback is the most commonly used window in momentum research, and Antonacci's work confirmed it as effective for GEM. It is short enough to respond to changing market conditions within a few months, but long enough to avoid overreacting to short-term noise. Shorter lookbacks (1 to 3 months) capture recent reversals as well as trends. Longer lookbacks (beyond 12 months) can be slow to respond to regime changes.
That said, the 12-month window is not magic. Research has shown that momentum signals are generally robust across lookbacks ranging from roughly 3 to 15 months. Sensitivity testing around the core lookback is useful, but investors should be cautious about over-optimizing to historical data.
ADM and Other Dual Momentum Variants
The elegance of the Dual Momentum framework is that it is a template, not just a fixed recipe. Researchers and practitioners have extended it in numerous directions.
Accelerating Dual Momentum (ADM)
ADM modifies GEM by blending momentum signals across multiple lookback windows rather than relying solely on a 12-month trailing return. The motivation is that a single lookback can be slow to detect regime changes at the turning points — a strategy might have strong 12-month momentum even as conditions have deteriorated sharply over the past 1 to 3 months.
ADM typically averages the trailing 1-month, 3-month, and 6-month returns (and sometimes 12-month) into a composite signal. The blended signal responds faster to recent changes while retaining the noise-filtering benefit of medium-term lookbacks. This can improve how quickly the strategy shifts to defensive positioning when market conditions deteriorate, at the cost of somewhat more frequent rebalancing.
Expanded Universe Variants
Some extensions of Dual Momentum expand beyond the three-asset GEM universe to include additional asset classes such as emerging market equities, real estate investment trusts, commodities, or sector-specific ETFs. The same dual-filter logic applies — relative momentum to rank among candidates, absolute momentum to gate risk-taking — but the broader universe provides more opportunities for rotation.
Adding assets to the universe must be approached carefully. Each additional asset introduces more degrees of freedom and increases the risk that any observed improvement in historical performance reflects overfitting rather than genuine robustness.
Multiple Protective Assets
Classic GEM uses only one safe-haven asset (aggregate bonds). Some variants replace aggregate bonds with a choice between short-term Treasury bonds, gold, or cash depending on which has the stronger momentum at the time of the defensive shift. This adds a layer of selection within the defensive allocation and can improve outcomes when aggregate bonds themselves are in a downtrend — as they were, notably, in 2022.
Historical Performance Context
Antonacci's original research and the peer-reviewed academic literature supporting cross-asset momentum provide strong evidence that Dual Momentum strategies have historically produced better risk-adjusted returns than static buy-and-hold allocations over long periods.
The distinctive characteristic that explains most of this outperformance is not spectacular upside capture — it is significantly reduced exposure during sustained market downturns. By shifting to bonds when equity momentum turns negative, GEM and its variants have historically avoided the worst phases of bear markets. This drawdown reduction is the engine of long-term compounding improvement: avoiding catastrophic losses allows capital to compound continuously rather than spending years recovering from deep declines.
It is important to note that momentum strategies, like all rules-based approaches, go through extended periods of underperformance relative to a simple buy-and-hold strategy. In strong, sustained bull markets with few corrections, a momentum strategy that occasionally shifts to bonds will lag an equity index that never leaves the market. Investors must have the discipline to stay with a systematic approach through these periods, which can last for years.
Limitations and Important Considerations
Whipsaw Risk
No momentum signal perfectly separates uptrends from downtrends at their turning points. Signals will occasionally shift to defensive positioning just before a market recovery, resulting in a "whipsaw" — the strategy sells near the bottom and then buys back higher after the market has already rebounded. These events are uncomfortable but are an unavoidable cost of the systematic approach. They do not invalidate the strategy's long-term logic.
One-Month Lag
Momentum strategies that rebalance monthly can experience one-month lags when markets reverse sharply within a calendar month. A strategy holding equities at the end of October might experience a significant decline before the November rebalance occurs.
Transaction Costs and Taxes
Momentum strategies typically produce more turnover than buy-and-hold portfolios. In taxable accounts, frequent rotation can generate short-term capital gains taxed at ordinary income rates. Investors implementing momentum strategies in taxable accounts should account for this, or consider implementing in tax-advantaged accounts where possible.
The Look-Back Bias Problem
When evaluating any backtested strategy, investors should be alert to the possibility that the parameters — particularly the lookback window and asset universe — have been chosen because they performed well historically rather than because of a principled reason. The most robust momentum-based strategies are those where the core logic is justified by economic reasoning and where performance is shown to be consistent across a range of reasonable parameter choices.
How Portfoliowiser Implements Dual Momentum Strategies
Portfoliowiser includes both GEM and ADM as pre-built strategies in its library, along with several additional dual momentum variants that extend the framework in different directions. Each strategy is fully transparent: investors can see exactly what the rules are, how the signal is calculated, and what the historical allocation history looks like month by month.
The platform's Strategy Builder allows full customization of the underlying parameters — investors can adjust the lookback window, change the safe-haven asset, modify the universe, or blend multiple momentum signals — without any coding required. This is particularly useful for investors who want to understand how sensitive any given strategy is to its parameter choices.
The AI Assistant is available to explain any aspect of the strategy mechanics, clarify how signals are calculated, and provide context for understanding historical performance across different market regimes.
For investors who are new to momentum strategies, the Portfoliowiser Strategy Library provides a straightforward starting point: view the historical performance of GEM, understand when and why it shifted to bonds, and compare its behavior against static alternatives. The platform's backtesting engine provides full analytics including monthly return heatmaps, drawdown history, and annualized performance metrics.
Next Steps
Dual Momentum is one of the most thoroughly researched approaches in systematic investing, with a compelling logic and a strong empirical track record across multiple decades and market environments. It is also one of the simpler systematic strategies to understand and implement, making it an excellent entry point for investors exploring tactical asset allocation for the first time.
The key to success with any momentum strategy is understanding what it does and does not do — and having the patience to stay systematic through the inevitable periods of underperformance that are part of any disciplined, rules-based approach.
Ready to explore GEM, ADM, and other momentum strategies? Browse the full strategy library at app.portfoliowiser.com and build a backtested portfolio in minutes.