Accelerating Dual Momentum (ADM)
Developed by Engineered Portfolio · Dual Momentum · High Risk
Accelerating Dual Momentum was developed by the research team at Engineered Portfolio and published in their 2018 analysis of momentum lookback optimization. The strategy addresses a fundamental tension in Antonacci's original GEM framework: the twelve-month lookback that makes GEM so stable also makes it slow to detect trend reversals. ADM's contribution is demonstrating that a blended multi-period momentum score — averaging returns across one, three, and six months — can detect trend changes substantially earlier without abandoning the dual momentum architecture's protective properties.
The term "accelerating" refers to the shortened composite lookback, which gives greater implicit weight to recent price action compared to GEM's uniform twelve-month window. While GEM's long lookback captures the full momentum effect documented in academic research, ADM argues that the early months of a trend reversal carry disproportionate information value — and that a faster signal can meaningfully reduce the depth of drawdowns at the cost of modestly higher turnover and occasional whipsaws during trendless periods.
During the 2008 financial crisis, GEM's twelve-month absolute momentum signal did not turn negative until several months into the decline, by which point the portfolio had already absorbed meaningful losses. Similarly, the signal was slow to turn positive during the subsequent recovery, causing the strategy to miss the powerful initial surge off the March 2009 bottom. ADM was designed to compress this lag by replacing the single twelve-month lookback with a composite that blends shorter timeframes. The one-month component captures immediate market direction, the three-month component identifies emerging trends, and the six-month component provides intermediate-term context.
The strategy maintains a focused two-asset equity universe — US large-cap stocks versus international small-cap stocks — with long-term Treasuries as the sole defensive position. The choice to compare against international small caps rather than large caps (as in GEM) reflects a deliberate effort to maximize the momentum differential between options. International small caps are more volatile and more cyclically sensitive, creating larger momentum spreads that produce clearer relative signals and more decisive positioning.
How It Works
The Accelerated Momentum Score
Each month, both US large-cap stocks (SPY) and international small-cap stocks (SCZ) are scored by averaging their one-month, three-month, and six-month total returns. This composite gives equal weight to each timeframe, creating a score that responds to recent moves more quickly than a standard twelve-month return while maintaining awareness of the medium-term trend.
The equal averaging across three periods means that a sharp one-month decline has three times the impact on the score relative to its contribution to a twelve-month return, allowing the strategy to react within one to two months of a genuine trend reversal rather than the three to four months typical of twelve-month momentum systems. This faster detection compresses the depth of drawdowns during bear market entries, which is the primary performance advantage over GEM.
Relative and Absolute Momentum
The asset with the higher accelerated score wins the relative comparison. Before allocation, the strategy applies an absolute momentum check: the winner's composite score must be positive. If the best equity option has a negative accelerated score — meaning its short-to-medium-term returns are collectively negative — the absolute momentum filter blocks allocation and the portfolio shifts entirely to long-term Treasuries (TLT).
The choice of TLT as the sole defensive asset reflects a specific design decision: long-term Treasuries tend to rally strongly during equity bear markets as investors seek safety and central banks ease monetary policy. This flight-to-quality effect historically amplifies the strategy's protective benefit during the exact periods when defense is most valuable. However, this concentrated duration exposure also means performance is highly sensitive to interest rate cycles — during rising rate environments like 2022, long-term Treasuries suffered historic losses.
The US vs. International Small-Cap Universe
The choice to compare US large caps against international small caps — rather than international large caps as in GEM — reflects a deliberate effort to maximize the momentum differential between the two options. International small caps (SCZ) are more volatile and more cyclically sensitive than international large caps (EFA), creating larger momentum spreads that produce clearer relative signals. When international markets are leading, small caps tend to lead by more; when lagging, they lag by more.
This amplified differential reduces the ambiguity zone where the two assets have similar momentum, resulting in more decisive and longer-lasting positioning. The trade-off is higher volatility in both directions — the strategy captures stronger trends but also experiences sharper reversals when the signal changes.
Turnover and Practical Considerations
The shorter lookback windows increase ADM's turnover relative to GEM. While GEM typically generates one to three trades per year, ADM may produce six to ten as the faster signal responds to shorter-term market movements. Each trade involves a complete portfolio transition — from 100% equities to 100% long-term bonds or vice versa — creating large discrete jumps in risk exposure, duration, and asset class.
The binary all-or-nothing allocation means the portfolio transitions instantaneously between full equity exposure and full bond exposure. During volatile but trendless periods, the strategy can generate multiple round-trip trades in quick succession — each incurring transaction costs and potential tax consequences — without producing net positive returns. This whipsaw cost is the insurance premium paid for the faster protective response during genuine bear markets.
Explore Accelerating Dual Momentum (ADM)
See the full backtest across 18 years of market data, or run your own what-if scenarios by adjusting all research parameters.