Accelerating Dual Momentum (ADM): Faster Signals, Better Protection
Accelerating Dual Momentum (ADM) takes the dual momentum framework pioneered by Gary Antonacci and makes it faster. Where GEM relies on a single 12-month return to judge momentum, ADM blends three shorter lookback periods into a composite score that detects trend changes earlier. The result is a strategy that enters uptrends sooner, exits downtrends faster, and provides stronger protection during market dislocations.
This article examines ADM's exact mechanics, explains the multi-period scoring innovation, covers the ADM_IP variant with inflation-protected bonds, and explores when faster signals help and when they create friction.
ADM Core Strategy
Strategy Parameters
| Parameter | Value |
|---|---|
| Risk-On Assets | SPY, SCZ (two assets only) |
| Risk-Off Asset | TLT (single defensive) |
| Momentum Method | ADM = avg(R1M, R3M, R6M) |
| Lookback | 6 months (longest component) |
| Top-N | 1 |
| Protection | ABSOLUTE (score must be > 0) |
The Two-Asset Race
ADM sets up a head-to-head competition between two deliberately different equity exposures: SPY (US large-cap stocks) and SCZ (international small-cap stocks). Each month, both assets are scored using the ADM composite formula — the simple average of their 1-month, 3-month, and 6-month returns. The asset with the higher composite score wins the relative test.
Then comes the absolute test: if the winner's composite score is positive (greater than zero), hold it at 100% of the portfolio. If the winner's score is negative — meaning that even the best option is losing momentum across all three timeframes — the evidence is clear that risk conditions have deteriorated. The entire portfolio moves to TLT.
The ADM Scoring Formula
The composite score is calculated as:
ADM Score = (R1M + R3M + R6M) / 3
Where R1M is the trailing 1-month return, R3M is the trailing 3-month return, and R6M is the trailing 6-month return. Note the absence of R12 — ADM deliberately excludes the 12-month lookback used by GEM. This is a design choice: the 12-month return is slow to respond and can keep the strategy invested in a deteriorating trend long after shorter signals have turned negative.
The three-period composite creates a natural consensus mechanism. For the score to be strongly positive, the asset must be performing well across short, medium, and intermediate horizons — all three timeframes must agree. For the score to turn negative, deterioration must be sufficiently broad that the average across all three periods goes below zero. This averaging smooths out noise while maintaining sensitivity to genuine trend changes.
Why SPY vs SCZ
Maximum Behavioral Divergence
The pairing of SPY and SCZ is deliberate. International small-cap stocks (SCZ) behave very differently from US large-caps in several important ways:
- Currency sensitivity: SCZ is denominated in foreign currencies, meaning dollar weakness benefits SCZ while dollar strength benefits SPY. This creates a natural rotation opportunity when the dollar trend changes.
- Economic cycle positioning: International small caps tend to lead during global growth accelerations — periods when capital flows toward smaller, cyclical companies in emerging industrial economies. US large caps tend to lead during quality-driven environments and late-cycle technology booms.
- Correlation: The correlation between SPY and SCZ is meaningfully lower than between SPY and broad international large-cap indices like VEU or EFA. Lower correlation means more distinct momentum signals and fewer ambiguous relative rankings.
This behavioral divergence is essential for a two-asset momentum strategy. If both assets moved in lockstep, the relative momentum signal would be noise. The wider the behavioral gap, the more meaningful the relative ranking becomes.
ADM vs GEM: Speed vs Stability
| Dimension | ADM | GEM |
|---|---|---|
| Scoring | avg(R1, R3, R6) — no R12 | R12 only |
| Risk-On | SPY, SCZ | SPY, VEU |
| Risk-Off | TLT | AGG |
| Absolute Test | Winner score > 0 | SPY R12 > T-bill rate |
| Signal Speed | Fast (1-month component) | Slow (12-month lag) |
| Whipsaw Risk | Higher | Lower |
ADM's faster signal is its primary advantage during crisis periods. During the 2020 COVID crash, ADM's 1-month component detected the deterioration immediately — the R1M alone would have turned sharply negative in March. Combined with the already-weakening R3M, the composite score went negative quickly, triggering a move to TLT. GEM, relying on R12, was still reflecting the strong 2019 returns and took longer to exit.
The tradeoff is increased whipsaw during sideways markets. When SPY and SCZ are both drifting with modest positive or negative returns, the composite score oscillates near zero, and the strategy can flip between risk-on and TLT monthly. This whipsaw generates unnecessary trades and friction — the cost of ADM's faster responsiveness.
The ADM_IP Variant
Adding Inflation Protection to Defense
ADM_IP modifies the defensive mechanism by expanding the risk-off universe from TLT alone to TLT and TIP (Treasury Inflation-Protected Securities).
| Parameter | ADM | ADM_IP |
|---|---|---|
| Risk-On | SPY, SCZ | SPY, SCZ |
| Risk-Off | TLT | TLT, TIP |
| Defensive Selection | Always TLT | Best of TLT or TIP by R1M |
When ADM_IP moves defensive, it selects the better-performing of TLT or TIP based on their trailing 1-month return. This adaptive defense addresses a specific weakness of pure TLT defense: during inflationary downturns (like 2022), TLT can decline significantly as rates rise. TIP, with its inflation adjustment mechanism, tends to hold up better during these periods. By selecting the stronger of the two, ADM_IP adapts its defense to the prevailing regime — TLT during deflationary crises, TIP during inflationary ones.
When ADM Works Best
Sharp Market Turns
ADM's greatest advantage appears during sharp V-shaped market turns. The 1-month component of the composite score reacts immediately to sudden changes, and the 3-month component confirms within one quarter. This speed allows ADM to exit falling markets faster and re-enter recovering markets sooner than strategies using longer lookbacks.
Trending Markets
During sustained trends — either up or down — all three components of the ADM score align. The strategy holds its position with high conviction and makes few trades. These are the environments where momentum strategies of all types perform best, and ADM is no exception.
When ADM Struggles
Range-bound markets with frequent reversals are ADM's Achilles' heel. When the market moves 5% up, then 5% down, then 5% up over a six-month period, the composite score oscillates near zero. The 1-month component flips positive and negative, dragging the average back and forth across the zero threshold. Each flip triggers a trade, and the cumulative friction of these whipsaws erodes returns.
Practical Implementation
Monthly Calculation
At each month-end, the investor needs to compute three returns for each of SPY and SCZ: the trailing 1-month, 3-month, and 6-month total returns. Average these three numbers for each asset. Compare the two averages: the higher wins the relative test. Check if the winner's average is above zero: if yes, hold it; if no, hold TLT (or the better of TLT/TIP for ADM_IP).
Blending with Slower Strategies
ADM's fast signals complement slower strategies well in blended portfolios. Pairing ADM with a GEM-type strategy creates a composite that responds quickly during crises (via ADM) while maintaining stability during trending markets (via GEM). This kind of signal-speed diversification is a powerful portfolio construction technique used by many defensive allocation approaches.