In my previous post I took a look at why the momentum factor has worked historically in the markets. The past numbers are impressive, but there are some drawbacks, including high turnover, tax inefficiency, large drawdowns and high costs associated with momentum strategies. It’s also a fairly counter-intuitive approach to get comfortable with. Dual Momentum, a book by Gary Antonacci, looks to minimize these problems.
I get a number of reader requests for writing ideas on this site, but by far the most requested topic has been for my take on this book and the strategy Antonacci lays out. The book explains in detail the difference between the two types of momentum-based investing:
Relative Momentum: Also known as relative strength, this form of momentum compares the trend of an investment to another investment in order to choose which asset class or securities to invest in. You choose the asset or group of securities that have been outperforming on a relative basis in the hope that the outperformance will continue. In the book, U.S. stocks are compared with international stocks on a relative basis.
Absolute Momentum: This is when an asset’s trend is compared to its own performance. For example, Antonacci compares the excess of stock returns over cash returns over a 12 month period to determine if the absolute momentum is positive or negative.
Antonacci combines to two approaches into a single system, hence the term dual momentum. The results in the book are quite impressive, with higher returns, lower volatility and lower drawdowns than the overall stock market.
Here’s what I like about the strategy:
1. It’s simple. The system requires just three funds or ETFs to pull off — an S&P 500 fund, a total international stock index fund and a total bond market index fund. This makes the underlying investments both simple and cost-effective.
2. It’s rules-based. There’s no discretionary decisions required. The relative momentum rule requires a comparison of the past 12 month returns for U.S. versus international stocks. The absolute momentum rule compares the higher trending of these two stock markets to the past 12 month returns for t-bills. If the S&P 500 has a higher return than both international stocks and cash, you hold the S&P. If international stocks have a higher return than the S&P and cash, you hold international stocks. If cash has a higher return than stocks, you hold the bond fund.
3. It’s a low activity strategy. The rules are only checked once a month for rebalancing purposes, which leads to a fairly low turnover compared to other momentum strategies. According to the book, going back to 1974, the strategy only made an average of 1.35 changes per year.
4. It’s based on both historical evidence and investor behavior. I ran the numbers for this strategy myself and they are quite impressive. The back-test makes sense as do the behavioral reasons for the momentum factor. All good investment strategies are based on taking advantage of other investor’s mistakes and behavioral biases and this one is no different on that front.
Now for some drawbacks and caveats:
1. It’s a concentrated strategy. You’re only in one asset class at a time based on the signals of the model. Here’s the breakdown since 1974: 41% of the time in U.S. stocks, 29% of the time in international stocks and 30% of the time in bonds. This strategy is diversified across market cycles more than across asset classes in the more traditional sense.
2. The psychology of going all-in or all-out. Dual momentum is a simple strategy, but no position in the markets is ever easy. It can be very difficult to psychologically go from an all stock to an all bond position. Some investors simple cannot stomach following market timing signals, no matter how simple they may be.
3. There’s nothing special about absolute or relative momentum. These aren’t magical indicators. You could always make tweaks and changes to data-mine better results and it’s quite possible momentum won’t work nearly as well going forward as it has in the past. That’s always a risk of any strategy.
4. It’s not always going to work. Nothing works all the time. Since 2009 this strategy has underperformed the S&P 500 by wide margin. But if you go back to 2008, dual momentum has actually done better than the S&P, because the strategy is built to perform better during bear markets. It works best during large drawdowns, but nothing can save you from all losses. Trend following rules are never going to get out right at the top, nor get in right at the bottom. There’s also the potential for whipsaw of false signals that reverse course in short order. There are always going to be trade-offs.
5. Back-tests only tell you what has happened, not what will happen. This is true of all investment strategies, but it’s worth remembering that the future is promised to no one in the markets.
The verdict? You have to know yourself as an investor when considering this type of strategy. It’s more about knowing yourself than understanding the strategy. Here are some questions to think about in terms of dual momentum:
- Momentum may continue to work, but will it work for you personally?
- Do you understand this type of system enough to force yourself to follow it?
- Will this type of strategy allow you to make fewer emotionally charged decisions during market losses?
- Would adding a trend following rule reduce or accentuate your behavioral biases?
- Would it make you more or less emotional about your portfolio decisions?
It’s almost impossible to recommend this or any system to an investor without first understanding where they’re coming from. Even simple strategies are never easy. This system has worked in the past, but investors have to define what “works” means to them. I say an investment strategy “works” if you’re to follow it over many different cycles. It never “works” if you bail out at the first sign of trouble or relative underperformance.
Even if you’re not interested in the Dual Momentum model, I think the book is worth a read because it covers nearly every other strategy you can think of and cites the pros and cons of each along with a long list of research. The sheer number of research papers cited and explained in the book alone makes it worth the price of admission.
As always, read the book and do your own homework before considering this or any type of investment strategy:
Dual Momentum: An Innovative Strategy for Higher Returns and Lower Risk
Also check out Gary’s website for more on the topic:
Dual Momentum
Further Reading:
Why Momentum Investing Works
The Psychology of Sitting in Cash
I don’t think you should consider performance since 2009 without also including 2008. Giving up some returns when a bear markets abruptly ends is the cost of avoiding the preceding extreme bear market drawdown. It isn’t fair to only look at one side.
True, you always have to consider a full market cycle.
My thoughts on the 5 drawbacks listed in this post:
1. It’s a concentrated strategy.
And hence why it works so well. Stocks have the highest risk premium of all asset classes. By concentrating in stocks when it is opportunistic to do so (based on momentum), you do not have the performance drag that bonds give. Furthermore, by concentrating in bonds when stocks are weak, you do not have the major drawdowns that stocks give.
2. The psychology of going all-in or all-out.
Compare this to the psychology of buy-and-hold. You are routinely living in fear of the next major drawdown. By having a mechanism to switch between stocks and bonds, you have essentially free “insurance” on your portfolio. This insurance gives more peace of mind.
3. There’s nothing special about absolute or relative momentum.
Successful investing shouldn’t be exotic and exciting.
4. It’s not always going to work.
Agreed. I did a backtest of Dual Momentum (in Canadian $) that showed from Jan 2007-Jan 2012 (5 years), the strategy returned essentially 0%. It could have been tough to follow the strategy over this time, but keep in mind:
– There was a sharp bear market in 2008 and sharp recovery in 2009. The model avoided this up & down volatility.
– Subsequent years were abnormally strong: 2012: +11%, 2013: +40%, 2014: +24%. Patience pays!
5. Back-tests only tell you what has happened, not what will happen.
This is why there needs to be an underlying reason (eg. human behaviour, information time lags, etc) why momentum has worked and why it’ll continue to work
Also, I can make the argument that even “buy and hold” is relying on faith and is not guaranteed to work.
You can read more about momentum investing on my blog: http://www.sharpereturns.ca
Good stuff. Makes sense. But I think that buy & hold and trend following have more in common than most people think. Both are forced to hold uncomfortable positions at times. Both require patience and discipline. Both never work all the time. And both are constantly questioned when they don’t work (which is inevitable and also one of the reasons they do work over time).
And I agree about successful investing not having to be exotic or exciting. My entire blog is based in this premise 🙂
Thanks for the response Ben. Recently found your blog through Abnormal Returns – very high quality posts. I’ve got you on my feedly now
Thanks. Abnormal Returns is the best way to find new material. It’s a daily stop for me.
I am a buy and hold investor and disagree with your comments about living in fear of a bear market. In fact, that is not a fear at all! I know they are going to happen. That is given. What I don’t know is their duration and magnitude. But the point of buy and hold for me, as Bill Murray might say, is that it just doesn’t matter. When they occur I won’t do anything about it anyway, so I don’t fear it. You could also argue that, for those making regular purchases, as through an employer based retirement plan, a bear market contributes wonderfully to their long term success.
Very true. If you understand this and can live through the losses, buy and hold works beautifully. This is why understanding your own personal risk profile is so important.
2 things:
1. If that matches your personality, then great. But for many people, this is easier said than done. Emotions are a powerful thing that cause us to do irrational things.
2. An adaptive strategy not only reduces drawdowns, it lets you grow your portfolio from a higher level when the bull market resumes. This excess return combined with reduced volatility is why I use dual momentum. Even a 2% extra return over the long term will compound magnificently.
I have nothing against anyone doing buy-and-hold or any other strategy for that matter. This is the beauty of the markets – everyone has a different time frame and approach.
Well said. I agree it’s all about finding what works for you. If you’re able to stick with something, that’s much better than trying another strategy that you’ll end up bailing on at the first sign of trouble.
From Cliff Asness: “The great strategy you can’t stick with is obviously vastly inferior to the very good strategy you can stick with.”
Just picked up your book on Amazon, Ben. Looks like it’ll be a good read
That’s great. Thanks for the support.
The big risk of buy and hold is the bear market and the big risk of dual momentum is the whipsaw market. I rather like the idea of diversifying those risks by taking some percentage of a portfolio (in a tax protected account) and applying dual momentum to that percentage. As you say, back tests don’t tell you what will happen.
Good point. It doesn’t have to be an all or nothing strategy. The biggest benefit through a diversified approach is that it can provide and emotional hedge and hopefully lower the amount of risk you take by spreading things out amonst asse classes & strategies.
It is very interesting and I believe it works but I know in the periods of underperformance I would be nervous and I would try to outwit the strategy and tinker it. So I remain a boring passive value investor (100% value stock ETFs) with relatvie cheap stocks through thick and thin.
Nice, like I said, it’s more about knowing yourself than anything. It does no one any good to try to fit a square peg in a round hole and try to run a strategy that doesn’t mesh with your personality.
“It works best during large drawdowns.”
I would agree with that. But large drawdowns tend to occur in overpriced markets. If you value invest, the risk of large drawdowns decreases.
Are you talking about using value for timing the market? If you stay fully invested with value, the research I’ve seen indicates drawdowns are larger than with a non-value portfolio.
It depends. A value stock can be risky (the risk is maybe overreacted by the people) or just simply boring so the market doesn’t appreciate it highly. Value=cheap that is the only sure thing.
The US stock market has a PE10 of 25.5; the Brazilian stock market has a PE10 of 9.3. When it comes to which has the greater 50% drawndown risk, I know where I’d place my bet
[…] ← Previous Next → […]
Given the rules based nature of this approach and the requirement of ony three ETFs is there a dual momentum etf that will do all the work for a lazy investor;-0 That would eliminate the need for an individual investor to rebalance, which is especially difficult during market turmoil (status quo bais, investors not willing to realize losses, etc).
Good question. As of right now there are no rule-based ETFs like this but I know of at least one or two that will be coming to market in the next few months. Stay tuned and I’ll be sure to highlight them here.
Any update on this (rule based ETFs like this)?
Nothing on this just yet but I would be shocked if there wasn’t a trend-following rule ETF released this year (that’s the rumor anyways). WisdomTree just released a L/S ETF for 48 bps so I’m guessing competition is coming:
http://www.wisdomtree.com/etfs/fund-details-alternative.aspx?etfid=124
Good analysis of Gary Antonacci’s book; thanks.
Buy-and-holders might take a look at the relative momentum component (S&P vs ACWI ex-USA stock indexes) as a standalone, with the absolute momentum component (a bond fund held only when stocks produce a sell signal) excluded.
Instead of diversifying between US and international stocks, this relative momentum approach holds exclusively one or the other. The biggest single driver of the selection is U.S. dollar strength or weakness. When a strong dollar is eroding international stock performance in USD terms (as it has during the past year), the system sticks with the S&P and skips the performance drag of international diversification.
Obviously, sustained currency trends will benefit the system, while currency whipsaws can cause it to underperform both stock indexes. But over the 40-year backtest period, the S&P vs ACWI ex-USA relative momentum model returned about 13.5% annually, well above the roughly 10% obtained from the S&P and EAFE.
That’s a pretty good boost, using only two well-diversified broad indexes. One can obtain similar numbers using small cap vs large cap or growth vs value relative momentum models. But these approaches inherently involve less-diversified sector overweights.
I was hoping you’d weigh in on this one as you’ve mentioned it before. I’d be curious to hear your thoughts on the absolute momentum piece. The results are fairly impressive, although there of course have been false signals and whipsaw action there are well. I ran the numbers back to 1970 and found there were roughly 60 total trades in the system, so less than 1.5/year. That’s pretty good.
When it comes to absolute momentum (a/k/a ‘market timing’), Gary Antonacci advocates using a 12-month rate of change in place of a moving average, in order to reduce trading frequency. The centroid of a 12-month moving average is only 6.5 months earlier than today, so comparing today’s price to the year-ago price does indeed help to damp spurious whipsaws.
Antonacci’s use of the risk-free rate as the sell threshold for the 12-month rate of change (as opposed to zero or a fixed slope) is innovative, and of course derives directly from the capital asset pricing model: if you can’t beat the risk-free rate with a risky asset, then you should sell it.
Instead of retreating to T-bills, though, Antonacci’s absolute momentum system goes to bonds (Barclays Aggregate, with a duration of 5.7 years). Empirical observations show that bonds perform better than average when stocks are on a sell signal, which often coincides with recession.
Big institutions are never going to engage in absolute momentum market timing. Among other things, trading cost escalates in proportion to share volume raised to the 1.5 power, owing to price impact. Absolute momentum is one of the few edges that individual traders have over institutions.
Good summary, thanks. I didn’t think about it from the standpoint of individuals vs. institutional investor either. That’s a great point. I would add that beyond what you stated the career risk would be far too great for the pension & endowment portfolio allocators. Another reason that these rules should continue to work more oftern than not over time.
[…] In Dual Momentum, Gary Antonacci shares a story about legendary hedge fund manager Michael Steinhardt. When asked what the most important thing an average investor could learn from him, Steinhardt replied, “I’m their competition.” […]
[…] of investment strategy imaginable. My conclusion? Value investing works. Momentum investing works. Trend following rules work. Index investing works. Diversified global asset allocation works. And all of them work much better […]
Ben, thanks for the introduction to this simple strategy. After I read this post, I picked up your book and Gary Antonacci’s Dual Momentum Investing. I used to be a dyed-in-the-wool value guy, but it’s very difficult to deny the evidence presented after a thorough deep dive. I also ran the numbers on the strategy and they are quite impressive even if you change the parameters of the underlying indices.
Regarding your caveats and drawbacks listed, I can see the advantage of using a simple GEM strategy as the tactical asset allocation piece in a portfolio. The diversification benefits are real, but just as importantly, I think the “emotional diversification” for clients in bear markets could be substantial. I can see where this type of systematic strategy would short-circuit a client’s strong desire to “just do something… anything!” What are your thoughts on that approach?
Again, thanks for the overview.
Thanks Daniel. I think you hit the nail on the head for how to approach/frame this for clients. In my mind this could act as a low cost/low activity liquid alt portion of a portfolio that acts as a huge emotional hedge against bailing on the rest of the portfolio.
[…] Carlson describes relative momentum and absolute momentum. The former compares an investment against another investment. Whichever one […]
[…] couple months ago I wrote a post on Gary Antonacci’s book, Dual Momentum, and received a ton of questions and reader feedback on his absolute and relative momentum […]
Thank you for this summary. Can you tell whether Antonacci uses total returns or just price action in determining which investment to use? What do you think of a possible simplification of just comparing us vs international vs bond and investing in best performing?