Dual Momentum: Relative or Absolute Momentum First?

When Gary Antonacci first shared the Dual Momentum strategy, he outlined a specific process that DIY investors could follow to narrow their investment of choice using absolute momentum (time-series momentum) and relative momentum (cross-sectional momentum).

Both forms of momentum are demonstrated by ample academic research as generating better risk-adjusted returns compared to simply holding the underlying security in all market conditions.

Gary was the first person to publicly share the outsized returns an investor can achieve by combining the two forms of momentum using major asset classes. He called this investment strategy "Dual Momentum" and it has become very popular among self-directed investors.

Dual Momentum is noted for its simplicity. The strategy only uses three broad asset classes: U.S. Stocks, International Stocks, and Bonds. The investor uses a 12-month lookback period to find the recent returns of each asset class and then follows a simple process to get the "Dual Momentum signal" for the next month.

Gary Antonacci's Process

As you can see, the process is simple and covers both forms of momentum: absolute momentum and relative momentum. Gary tests for absolute momentum on U.S. stocks first in his process. This means an investor would neither be invested in U.S. nor International stocks if the U.S. equity market was underperforming Treasury bills.

While this model is simple to follow, my first thought as an investor based outside of the U.S. was: why test absolute momentum on U.S. stocks first? What if the rest of the world is performing well while U.S. stocks are experiencing a correction?

By testing absolute momentum on U.S. stocks first, we open ourselves up to single market risk. There is a real possibility of being invested in U.S. bonds while global stocks and global currencies were doing well. This would compound the downside from a global purchasing power standpoint.

In his book, Gary shares that U.S. stocks lead markets. Based on this thesis, an investor could test for absolute momentum on U.S. stocks first as a type of leading indicator on all equities. While this argument has certainly held true for most of the past century, I'm not so sure it will persist.

In the 1900s, we saw a major shift to a U.S.-based global economy. The U.S. economy dominated the world and global currencies were almost exclusively pegged against the dollar (via gold standard or trust in U.S. institutions). In fact, during the second World War this arrangement was formally adopted via Bretton Woods.

However, in recent decades other economies have increased their influence substantially. The People's Bank of China, the European Central Bank, and to a lesser extent the Bank of Japan carry a lot of weight. We are entering a period where the U.S. is slowly receding in relative economic dominance. This could hurt investors who rely solely on U.S. stocks to drive portfolio returns.

Considering the Dual Momentum model provides global diversification, I could understand a case for testing absolute momentum first if the test was applied to the global stock universe. We could evaluate the absolute momentum of the MSCI ACWI Index to determine if we should invest in stocks or bonds.

If the absolute momentum test determined global stocks were outperforming cash, we could run a relative momentum test to determine if we should invest in U.S. stocks or International stocks. In large part the relative momentum gains realized by the investor are driven by currency changes. U.S. stocks tend to outperform when the U.S. dollar is doing well against other global currencies; International stocks do well for U.S. investors when global currencies are outperforming the U.S. dollar.

That said, there is an easier way to do this without having to track another indicator. It would also keep the currency performance impacts separate.

My Dual Momentum Process

In my process, I test the relative momentum component first.

I begin by identifying which broad stock market class is performing better: U.S. stocks or International stocks. Once we've identified the stronger of the two major equity markets, we ensure we are investing in a rising asset by checking the absolute momentum of that market.

In this slight variation of the process, we can eliminate the risk of holding bonds while foreign stocks and currencies are doing well.

Relative Momentum First or Absolute Momentum First: U.S.A.

Sources: TheRichMoose.com, MSCI Inc., FRED

When comparing the two methods of Dual Momentum using U.S. stocks as the base, there is a slight historical performance advantage for testing absolute momentum first.

In this backtest, both methods are very comparable over the test period with the only substantial deviation occurring in the early-1970s. Even in this period, my suggested method quickly caught up to Gary's method by the end of the decade.

In the end, an investor testing for absolute momentum on U.S. stocks first (Gary's method) would have realized a +16.14 percent compound annual return during the test period.

The same investor testing for relative momentum of equities first (my method) would have realized a +15.93 percent compound annual return. This is effectively an indiscernible difference.

We can also look at a rolling period return to examine the differences in returns between the two methods more closely.

Sources: TheRichMoose.com, MSCI Inc., FRED

Testing absolute momentum first with U.S. stocks does show a general advantage earlier in the test period. The noticeably better performance at the start of the testing period largely stems from a single month in 1973 where the relative momentum first model had the investor in International stocks while the absolute momentum first model had the investor in bonds.

While one month does matter, we can't ignore the snap back in the following years where the relative momentum first model quickly caught back up.

The performance advantage of Gary's model has shrunk to nothing in the past two decades. I suspect this may be due to the corresponding rise of China and an expanding European Union during this period.

Relative Momentum First or Absolute Momentum First: Japan

Sources: TheRichMoose.com, MSCI Inc., FRED

Japan provides us with a stress test example of a highly diversified, high impact market going through a period where local stocks performed poorly while International stocks did extremely well.

To perform this test, I used the MSCI Japan Index, MSCI Kokusai Index, the MSCI ACWI ex-Japan Index, and Japanese CD data from the Federal Reserve Bank of St. Louis. All momentum evaluations used a 12-month lookback period and all data was priced in Japanese yen.

In this simple backtest, we can clearly see the many periods where Japanese stocks were underperforming CDs. This would put the investor into local bonds when absolute momentum was evaluated first (Gary's method). The investor missed years of International stock market exposure.

However, when doing a relative momentum evaluation first (my method), the investor was able to participate in International stock growth while the local market was underperforming.

In the end, an investor testing the absolute momentum of local stocks first would have realized a reasonable +8.03 percent compound annual return. The same investor testing relative momentum first would have realized a +10.46 percent compound annual return.

As seen by the chart below, the performance advantage for testing relative momentum first was noticeable across most time periods.

Sources: TheRichMoose.com, MSCI Inc., FRED

I acknowledge Japan might be a bit of an anomaly as an enormously inflated stock market going into the late-1980s. However, it does show us a recent example using a very diversified market, one that has a meaningful impact on the global economy, and a country with reserve status currency.

Conclusion

As long as the U.S. economy is the world's leading economy and the U.S. dollar is the base currency for global currency valuations and global economic activity, the method of testing absolute momentum first in the Dual Momentum model should work.

While U.S. stocks and the U.S. economy have performed well this past century and have been a leading indicator of the global economy, this phenomenon is likely to subside as the U.S. declines in relative impact. Just as the U.S. rose to prominence in the early 20th century, China and India are rising today.

Gary's method of evaluating Dual Momentum relies on the performance of U.S. stocks to get exposure to any equities. This limitation in the process needlessly exposes investors to single market risk.

Single market risk is very real. We can see the negative effects of a single market on Dual Momentum when we apply Gary Antonacci's process to Japanese stocks. Although Japan is not the U.S., it is also not a fringe economy with a non-influential currency or small global impact.

Given the shifts we are seeing today, it is not out of the realm of possibility that the U.S. experiences a similar decline in market influence, becoming a market laggard instead of a market leader.

We can significantly reduce our single market risk by slightly changing the Dual Momentum process. Instead of testing absolute momentum on U.S. stocks first, we should start with a relative momentum evaluation on our equity assets. We will still always test for absolute momentum on the better performing equity asset to ensure we are not investing into a declining market.

Comments & Questions

All comments are moderated before being posted for public viewing. Please don't send in multiple comments if yours doesn't appear right away. It can take up to 24 hours before comments are posted.

Comments containing links or "trolling" will not be posted. Comments with profane language or those which reveal personal information will be edited by moderator.

4 Replies to “Dual Momentum: Relative or Absolute Momentum First?”

  1. Interesting post Daren. However, I might add some researchers postulate that China might fall behind the US shortly after catching up with it. All those years of the one-child policy will have an impact some day.

    Could China become the next Japan? Maybe not, but surely many saw Japan beating the US as a sure thing… look at us now.

    As Buffet says: “Never bet against America” 🙂

    For reference….“The Long View: Scenarios for the world economy to 2060”.

    1. Daren (Editor) says:

      Good points. I don’t think China or India will become richer than the U.S. on a per capita basis anytime soon. But with 4x as many people in each country they don’t need to in order to have massive impacts on global markets. It is almost inevitable that the U.S. will remain strong for many decades to come, but their relative dominance will wane as countries like China and India rise in wealth.

      Plus other lesser countries such as Indonesia, Nigeria, Eqypt, Brazil, DR Congo, Pakistan, Ethiopia, and The Philippines will become wealthier and carry more influence than they do today. Today many of these countries still have per capita GDP of less than $5,000.

  2. Very good post! US-centrism could indeed perhaps become a blindfold in the future. Do you have any thoughts on whether the same could apply in terms of testing absolute momentum with respect to US T-bills? If the US market loses some of its preponderence, shouldn’t the risk-free rate also be dependent on the rising economies? Would it be worthwhile to develop something like a “global risk-free rate indicator”? Or would it make more sense to base the AM trigger on the local (i.e. Canadian) risk-free rate?

    1. Daren (Editor) says:

      It’s definitely an important point that I’ve thought about and should be monitored.

      Right now we measure the value of equities in USD. For that reason we should also measure the best available risk-free return in USD. In my view we should only change our current measure if we have reason to believe that the U.S. government will default on 3-month T-bills.

Leave a Reply

eighty one + = eighty six