Testing Dual Momentum with AllocateSmartly

I am frequently asked about performance and backtest results for my Dual Momentum portfolio, which is inspired by Gary Antonnaci at Optimal Momentum.

I recently began using AllocateSmartly to test some of my favorite tactical strategies, including Dual Momentum.  AllocateSmartly tracks some of the most popular tactical asset allocation strategies, with thorough, up-to-date backtests, and users can combine the strategies to create and test their own custom model portfolio.

The first Dual Momentum strategy is very similar to the strategy I update monthly on Scott’s Investments.

The strategy divides the portfolio into four “modules”, each targeted to a different component of financial markets: equities, credit risk, real estate, and stress. From each module, the strategy selects either one of two related asset classes, or cash. The four modules and their associated assets are as follows:

  • Equities: US equities (SPY) or international equities (EFA)
  • Credit risk: US corporate bonds (LQD) or US high yield bonds (HYG)
  • Real estate: equity REITs (VNQ) or mortgage REITs (REM)
  • Economic stress: gold (GLD) or long-term US Treasuries (TLT)

The test below trades on the last trading day of each month.  You will note the backtest extends further than the launch date for any of the ETFs. In order to extend the backtests as far into the past as possible, AllocateSmartly  often makes use of simulated data from an alternative sources. For example, an alternative data source for SPY could be the Vanguard S&P 500 mutual fund (VFINX).


The second dual momentum strategy is the GEM strategy from Antonacci’s book Dual Momentum Investing.

The strategy rules tested are below:

  • At the close on the last trading day of the month, measure the 12-month return of SPY, EFA (international equities) and BIL (short-term Treasuries).
  • If the return of SPY is greater than BIL (absolute momentum), go 100% long either SPY or EFA, depending on which has the highest return (relative momentum).
  • If the return of SPY is less than BIL, go 100% long  AGG (US aggregate bonds).
  • Positions are held until the last trading day of the following month.

The time frame for this test is much longer than the first test, 1971 – present:


Backtest data courtesy of AllocateSmartly – if you are interested in receiving monthly updates for this strategy or many other tactical strategies give them a try!

6 thoughts on “Testing Dual Momentum with AllocateSmartly”

  1. What many don’t realize is Antonacci also has proprietary models he licenses to investment advisors. They earn substantially higher returns than his public models and are more responsive to market conditions.

  2. Thanks for the article! I have a couple of questions.

    For the 4 module selection, do you select the fund with the higher 12-month return in each module? Do the equity/REITs also have to beat BIL over the last 12 months?

    For the second strategy, what’s your benchmark? Is it some combination of the S&P 500 and cash/bonds? I noticed the drawdown is only 29.5%, while the S&P had a 56% drawdown in 2008-2009.


    1. “For the 4 module selection, do you select the fund with the higher 12-month return in each module? Do the equity/REITs also have to beat BIL over the last 12 months?” Yes and Yes

      The benchmark is a 60/40 (stocks/bonds)

  3. Scott,

    Thanks for the ongoing work!

    Could you show the GEM system performance starting from 1990 to compare with 4 module system?

Comments are closed.