Skip to content

Dual Momentum: Testing the Dual Momentum Strategy and Implications for Lifetime Allocations: summary and key takeaways

The paper tests Antonacci's dual momentum rule on US equity indexes and T-bills, first as an allocation framework and then as an overlay on target date funds. Returns beat a static benchmark in every formation period, but Sharpe ratio gains are not statistically significant, and Monte Carlo results are weaker than historical ones.

Ha, S.; Fabozzi, F.J. · 2022 · The Journal of Portfolio ManagementSummary updated 20 November 2026
Published2022Journal article
Asset focusMulti-assetMomentum / TAA
TypeBothTactical Allocation
Reading time4 minthis summary

Dual Momentum: Testing the Dual Momentum Strategy and Implications for Lifetime Allocations at a glance

The question
Does dual momentum, which combines time-series and cross-sectional momentum, work as an asset allocation framework and as a tactical overlay for lifetime (target date) allocations?
Dataset & period
Monthly Fama-French US Total Market, Small Cap and Value Research indexes plus the one-month US T-bill, June 1926 to September 2016 (Dimensional Fund Advisors and CRSP data). Monte Carlo: 1,000 simulated paths per formation period. Target date fund tests: 584 rolling 40-year windows. Transaction cost of 10 bps per trade.
Method
Each month the rule checks whether any of the three equity indexes beat the T-bill over the formation window. If none does, it holds T-bills. If at least one does, it holds the index with the strongest past return. The benchmark (GAA) is a static mix of US stocks and T-bills. The Monte Carlo run uses geometric Brownian motion for stocks, a CIR model for rates, and Student-t shocks. The target date fund test applies the signal on top of three glide paths with monthly contributions and compares terminal wealth.

Key findings of Dual Momentum: Testing the Dual Momentum Strategy and Implications for Lifetime Allocations

  • In the historical test, the dual momentum return beat the GAA in every formation period tested, from two to twenty months.
  • With a nine-month formation period the geometric return was 12.04% against 7.46% for the GAA.
  • The nine-month Sharpe ratio was 0.62 against 0.55 for the GAA, but Sharpe differences were not statistically significant in any formation period.
  • Maximum drawdown for the nine-month version was 63.78% versus 71.62% for the GAA, and the drawdown difference was significant in all formation periods.
  • In the Monte Carlo test, the strategy return was on average below the simulated GAA at 6.66% per year, with higher volatility and lower Sharpe ratios than in the historical test.
  • Under the 100-Age glide path, mean terminal wealth was $5,695,899 with a seven-month overlay versus $5,180,950 without it.

Robustness, as reported in the paper

  • Monte Carlo simulation with 1,000 paths is used as a robustness check; it gives weaker risk-adjusted results than history.
  • A 10 bps cost per trade is included, and the paper reports turnover of 92% per year for the 16-month version.
  • Three glide paths and nineteen formation periods are tested for the target date fund overlay; the paper says 20- and 30-year vintages gave similar results to the 40-year case.
  • Return differences versus the GAA are statistically significant only for a subset of formation periods (about eight of them).

Caveats and limits

  • The paper says the Sharpe ratio improvement is not statistically significant, so the evidence is mainly for higher return and lower drawdown.
  • The Monte Carlo assumptions (Student-t returns, a fixed correlation matrix) may be unrealistic, and the paper cannot explain why a high share of positive months did not lift the simulated Sharpe ratio.
  • The test universe is only US equity indexes and T-bills, so it is not the global stock/bond setup used in the original Antonacci model (inferred).
  • Higher terminal wealth for the target date overlays comes with a wider spread of outcomes.

Items marked (inferred) are our reading of the design, not statements by the authors.

Why it matters

The paper separates return gains from risk-adjusted gains, which is the point most backtests of momentum overlays blur. It also shows how a trend filter changes retirement-style accumulation paths, including the left tail of terminal wealth. Practitioners can use it as a reference for how fragile the Sharpe edge is.

Questions about this paper

Did dual momentum beat the benchmark in the paper?

On return, yes, in every formation period, though only some of the gaps were statistically significant. The Sharpe ratio gap was not significant in any period.

How often did the strategy have positive months?

Around 70% of monthly returns were positive across formation periods in the historical test, and the simulated average was also 70%.

What did the overlay do for target date funds?

Mean terminal wealth was higher with the overlay in most formation periods, and minimum terminal wealth was also higher in many cases, at the cost of a larger spread of outcomes.

Where can I read the original paper?

The original is at its publisher: https://doi.org/10.3905/jpm.2022.1.336. We link to the paper and do not host the PDF.

Follow new quant research

Download MacroFinance market data as CSV or Excel, or pull it by API, and test the ideas in these papers yourself. New summaries are added as we review them.

Create a free account
Summaries are written by MacroFinance from the published papers, in our own words, for information only. They are not investment advice and may omit detail. Papers are not hosted here: each page links to the original at its publisher.