Here's How To Take Advantage Of Momentum Investing
Tim Melvin: We're on today with Gary Antonacci with Portfolio Management Consultants and the author of an award-winning book called "Dual Momentum." Thanks for coming on with us today. I really appreciate it.
Gary Antonacci: My pleasure.
TM: You've done most of your research and work on the momentum concept of investing. Now, a lot of folks aren't familiar with that. They have kind of a distorted image from the media of exactly what momentum investing is. Can you basically described what momentum investing is for us?
GA: Momentum means persistence of performance. It's akin to Newton's First Law that a body in motion tends to stay in motion. The same thing applies in the financial arena, as well as in the physical universe. Some people think of momentum in the old sense of those who just go after high-risk stocks, like the .com era people.
But the way I talk about momentum, and the way that it's used in the academic world, is a systematic, rules-based approach, in which you look at performance during a certain specified look back period. Whatever performs well during that period, it's expected that asset will continue to perform well. That's been shown to exist, the momentum effect, across a great number of assets – stocks, stock indexes, fixed income, commodities, currencies, real estate – pretty much everything.
TM: So it's not just chasing the popular stocks of the day, which I think, for the average investor out there, that's what they think of when they hear the phrase momentum, that you're trading whatever is popular. That's really not necessarily the case.
GA: No, it's not the case at all. Many momentum stocks are popular, of course, because they've been doing well. But you can't just pick something out of thin air because you have no way of knowing when to get in and went to get out. You can't really back test something that way. Momentum is a systematic approach which was first written about in 1937 by Cowels and Jones, noted economists, who went back painstakingly over the New York Stock Exchange stocks by hand. They discovered stocks that had done well over the preceding 12 months tended to outperform during the coming year.
That was forgotten about until the 1990s when it was picked up again in the academic world by Jegadeesh and Titman who did a very rigorous study basically validating the same thing. Because it flew in the face of efficient market ideas, momentum was challenged by many people. There were probably more momentum studies done than anything else in financial economics during the 20 years since then. Momentum was looked at as going forward in time, going backward in time, across different asset classes. And it's held up remarkably well. In fact, studies of momentum go back now to the year 1801. It's been shown to perform not only well, but very consistently over that time.
TM: In your book and in a lot of your research you talk about the two different types of momentum. There is absolute momentum and relative momentum. Can you walk us through what each of those is and the difference between them?
GA: Most people are more familiar with relative momentum, or what's sometimes called relative strength. That's where you're comparing the performance of one asset to its peers. It's usually done with individual stocks in the research literature. So, you'll take cross sections of stocks, those having the strongest performance, and you segment the market into deciles or quintiles, of the top 10 or the top 20 percent, and you compare that to the lesser quintiles. That's what is meant by relative momentum. Absolute momentum is where you compare performance to itself over time. It is sometimes called time-series momentum because, instead of comparing your performance over, say, the last year, to other assets that are similar, you just look at how that asset has done over that period of time.
For instance, if you were looking at 12 months' absolute momentum, you would say that absolute momentum is positive if the excess return of that asset has been positive over the preceding 12 months. Now, one thing to keep in mind is that when you're looking at absolute momentum, you're looking at excess return because you're comparing the performance of that asset to the performance of the risk-free asset. Usually, Treasury bills are used for that. Some people don't give this proper consideration, and right now it doesn't matter very much, because Treasury bill rates are close to zero.
But in the past, short-term interest rates have been up in the double digits. Let's say stocks have returned 5 percent over the past year and Treasury bills returned 6 percent. There's no reason to assume the risk of stocks when you can go into Treasury bills risk free, and you would have earned a higher return. That's why it's important not just to look at the total returns of the asset that you're considering, but to compare it to some type of low risk alternative.
TM: I'm a geek, I read a lot of academic papers and theories. It's something I like to do when I'm sitting around on the weekends. I've never seen anybody put the two concepts together the way you have. In the book and your other papers, you call it dual momentum. Can you explain that for us?
GA: Both types of the momentum will enhance performance. You'll get a boost in returns and a boost in risk-adjusted returns over whatever benchmark you're comparing yourself to. The problem with relative momentum is that it does little or nothing to attenuate downside volatility. You'll still have the same downside risk and downside bear market exposure that you would have from just being in the stock market, assuming we're talking about stocks here. What absolute momentum does, and why I actually prefer it to relative momentum, is that because it is trend following, it will get you out when the market is weak.
That way, you can avoid the severity of these really bad bear markets. If I had to choose between the two, I would choose absolute momentum over relative momentum. But there's no reason why you have to choose between the two. You can put them both together and get the return enhancements benefits of both, and also the potential drawdown and bear market exposure reduction that comes from absolute momentum. You get the best of both worlds.
TM: In your book, you put this together into what you call the global equity momentum strategy. The results are pretty impressive, and it's a unique idea. Can you describe how that works for us?
GA: Sure. The premise behind it, and people lose sight of this fact, is you're trying to capture the most risk premium that you can, because that's how you get rewarded. Historically, the highest risk-adjusted returns, the highest risk premium, has come from equities, particularly U.S. stocks.
So, we want to have that as our core holding. For that, we use the S&P 500. Now, in order to have relative strength momentum, you have to compare that to something else. What I do is compare U.S. stocks to the rest of the world. And we use the MSCI All World Index – ex US – for that purpose. We'll go into whichever of the two has been stronger according to relative strength momentum. However, we also use absolute momentum as a filter. Not only does our selection need to be stronger than the other, but it has to exhibit positive absolute momentum during that time.
For example, if U.S. stocks are stronger than non-U.S. stocks at the time, but U.S. stocks have been down on an excess return basis over the past year, we would stand aside of stocks entirely and go into some type of short-term or intermediate-term fixed income asset until such time as absolute momentum turns positive for stocks.
TM: What are you using for the risk-free rate for absolute momentum in the model?
GA: I use the total return for 90 day T-Bills. I'm comparing the total return of U.S. stocks to the total return of 90 day T-Bills over the past 12 months.
TM: That's not a very high bar these days, is it?
GA: No, it's not. I think the T-Bill ETF return over the past 12 months is slightly negative. In that case, you could just use zero, or you can use a one-year type Treasury bill ETF such as SHV as an alternative.
TM: And how often to you make this comparison to re-jigger the portfolio?
GA: We look at it monthly. When you look at momentum, your returns are best if you do monthly type reassessments and rebalancing. That creates a problem when you're dealing with individual stocks, because the turnover can be substantial and transaction costs can take away much of the potential profit that you would get from relative strength momentum with individual stocks. With asset classes, that's not such a problem. We have only about 1.3 trades per year on average while switching based on both relative and absolute momentum.
TM: I was going to touch on that, because initially, it sounded like, "Monthly? My goodness, we're going be trading all the time." But, that really has not been the case, as one asset class or another has pretty much sustained momentum for a lengthy period of time.
GA: That's correct. Absolute momentum is not a short-term, in-and-out way of trying to time the markets. It's a longer-term, trend-following approach that's meant to pick up on the major trends. Because of that, we retain much of the short-term volatility of stocks. People should be aware of that. There will still be day-to-day, week-to-week and month-to-month volatility. You don't get rid of that by using dual momentum, but you can attenuate much of the severe bear market downside that comes from being in stocks continually.
TM: When I look at the performance posted on your website, it actually does avoid this – and by the way, that's www.optimalmomentum.com – and the results of everything in the model are up there. But, about 30 percent of the time, during the study period, this model's actually been in cash or cash equivalents. The timing has been kind of remarkable.
GA: It is pretty impressive, and the further back you go, it continues to hold up.
TM: Going back, you started in 1971, just before we go into the nasty bear market of the seventies. You spend from May 1973 to May 1975 in cash equivalents, and then again from ‘77 all the way up through the end of the summer of 1978, which was a nasty bear market. The model actually just sidesteps it.
GA: Yes, it did, and it avoided much of the 2000 and the 2008 bear market as well.
TM: What's really impressive is that it stays in either U.S. or international stocks almost all the way through the 1990s. There's only just a very small period where you come out of equity markets, which of course, as you got into ‘97, ‘98, ‘99, people were getting a little twitchy about the equity markets, because the movie was so powerful. In the recent unpleasantness that may be more familiar to you, the model got out in February of ‘08, and didn't get back in until November of ‘09. So, you sidestep a lot of the disaster of 2008 and 2009. I've never seen a model that actually achieved this as precisely as this one did. So, as I told you earlier, I was very impressed. Can dual momentum be used with sectors? Could you use sector ETFs to take advantage of the momentum changes among the sectors?
GA: It can. I have a dual momentum sector rotation model. However, it doesn't hold up as well as using broader market indexes. Initially, I only had data back to the early 1990s for sectors. More recently, I've acquired data back to the 1970s. When you look further back, the broader based models at the time, like my global equity momentum model, actually out-performed sector rotation models.
TM: Okay, so, that sounds like it's best to stick with just the major market asset classes and avoid trying to fine-tune it too much.
GA: Well, there's a place for both when you look at portfolio theory. You can always benefit from diversification. So having sectors as a minor portion of your portfolio makes sense. I think people who are "gung ho" on sectors may end up being disappointed because they don't have as much data as I have to go back and see how it's done all the way back to the 1970s. Most people who put together sector rotation programs don't have nearly that much data. And going back just the last 20 years, it looks very impressive. But further back, I think a broader-based approach has much more going for it in terms of risk-adjusted returns.
TM: When you're looking at the sectors, would you just take a certain number of sectors and buy the top couple that were also showing positive excess returns? Or, would you just do the one sector that was the leader?
GA: I wouldn't do just one. The idea behind sector rotation is to be in the top-performing sectors so that you gain some benefits of diversification. But with one, you're going to have quite a bit of volatility. You will have, potentially, a higher return. It's a tradeoff, the number of sectors you use is really an individual thing that's based on your own risk preferences. The fewer sectors you have, the higher the potential returns, and the higher volatility and drawdown.
TM: It really does sound like, for the average investor out there, they're better off sticking with the three-part global equity model and just rotating between sectors using that.
GA: Right, very much so.
TM: With individual stocks, with most of your writings that I've read, it doesn't really work with individual stocks, because of turnover, does it?
GA: Most of the momentum products out there use individual stocks. That may be because the research on momentum, most of it doesn't incorporate transaction costs, or there's differences of opinion of what those transaction costs might be. I've seen estimates of anywhere from 2.5 percent per year up to 7.5 percent per year. That can take away quite a bit of momentum profits.
TM: Yeah, at 5.5 percent, just the bid-ask spread would be a killer.
GA: Momentum stocks have wider bid-ask spreads anyway because they tend to be more volatile, and sometimes that can affect performance also. And that's not accounted for in much research that's been done. There is one research paper by Lesmond, and his conclusion is that transaction costs take away all of the potential profits that we see from momentum.
TM: Using individual stocks?
TM: That's crazy. I don't know what else to say with that one, all the profits, remarkable. Now, you mentioned momentum products a minute ago. Then, when you look around the investment world, there's not really a ton of momentum products available to the individual investor. Why is that? So, you just have to keep the big picture in mind, and realize that this is a long-term approach, and very carefully work what the benefits and costs might be to you.
GA: There's becoming more and more, but they tend to come as part of multifactor portfolios, and multifactor ETFs. They also tend to be oriented toward individual stocks. I think some of this has to do with the fact that momentum is not as intuitively appealing as something like a value. Everybody recognizes that you want to buy something that's cheap. But to go into momentum, you're buying something that isn't cheap; you're buying something that's been going up. And when things go up, people sometimes will back away from it instead of wanting to jump in and buy it.
The other thing that's related to that is that both types of momentum, both absolute and relative, are forms of trend following. With absolute momentum, it's obvious that you're following the major market trends. But even with relative momentum, you're dealing with strength of one trend versus another when you compare different assets. And people have been brought up in the efficient market world thinking that trend following doesn't work, that you should just buy and hold until you grow old.
TM: That brings me to an interesting side question. There's tons of research by people that have achieved outstanding results using both of value and momentum strategies, these two huge anomalies on opposite ends of the market. But a great deal of the academic world still clings rather tightly to the efficient market hypothesis. Why is that?
GA: Well, they're not clinging as tightly anymore as they used to. But, I think the general principle behind efficient market is still a good one, that markets are generally efficient, and it's really hard to make money when you're trying to do something different than just owning the market. However, the momentum and value anomalies have been well established in the academic literature now for the last 20-30 years. I don't think you'll find an academic around nowadays who would say that momentum doesn't work.
TM: What's the major risks to a dual momentum strategy? Because it sounds like it pretty much does keep you in the market when it's going up and out of it for the bulk of a horrible crash. What are the major risks for this?
GA: One risk, of course, is the future may not be like the past. The equity premium in stocks may go down. Momentum returns are based on what the stock market does. So momentum returns may suffer as well. There's also some risk in terms of tracking errors. Momentum doesn't always outperform, and it can be difficult for people to stay the course when they see the benchmarks outperforming momentum, which does happen occasionally. There's a risk that you get in and out at inopportune times, just like people do with stocks in general. Whipsaws happen occasionally. When you get out of the market, the market may just reverse on you quickly and go back up.
Those kinds of issues create psychological problems that people need to be prepared to deal with. Re-entry lags also are a thing that people should recognize. Even though the trend-following component of absolute momentum may get you out of the worst part of the bear market, it's not going to get you back into the coming bull market immediately. There's going to be a lag there. Some people aren't going to like that, either.
TM: To jump back to the academic part of the discussion, in the last year or two, Cliff Assess and some others have done a lot of work on combining value and momentum in different ways. Have you looked at that at all?
GA: Of course. What they've done is taken long/short value and compared it to long/short momentum. As factors, yes, they're not correlated. If you're going to have a long/short portfolio, then yes, there's some good value to be had by combining the two. However, most people don't invest in long/short portfolios.
They invest in long only. You're not going to gain a lot of diversification value just from being in both long-only momentum and long only value. You'll get some diversification benefits, but it won't be any type of great benefit as you might think from reading some of the papers that look at long/short momentum combined with long/short value.
TM: That makes a lot of sense. If you're just doing the long only, they're matching the market when you put them together?
GA: Exactly. Both of them tend to exhibit the same type of volatility and drawdown. In fact, value shows considerably more drawdown than the market. You're not going to get rid of that just by buying value and momentum together.
TM: Could you use a dual momentum type strategy to switch between value and momentum?
GA: You could.
TM: Do you think that would work?
GA: I've looked at that, and there are better ways to do it. You're usually talking about individual stock momentum and individual stock value. You have potential problems there that make it so that I don't care to do things that way. Even talking of indices, I think there are better indices that can be used rather than value and momentum ones. In fact, my back testing has shown that's the case.
TM: So, what really works the best is this global equity momentum strategy. It's simple, it's easy, using very liquid easily traded indexes, and you only really have to look at it once a month. That seems like it might fit the average investor a little bit better than trying to be too fancy.
GA: It does. It's fully disclosed in my book. People can easily do it using a free charting package. It takes one or two minutes per month. It's easy to implement and easy to understand.
TM: What do you use for a look back period with the two?
GA: With my Global Equity Momentum model, I use 12 months. That's the one that's most commonly used in the academic literature. It's performed very well over time. And it gives fewer trades and transaction costs than going with shorter-term lookbacks.
TM: So, you use 12 months for both absolute and relative?
TM: Would there be any merit to a longer-term relative and a shorter absolute?
GA: There might be. You can play around with these if you want, and you might find something that, based on whatever particular data that you're looking at, fits the data a little better. What I like about 12 months is, it's been used across many different assets, and many different time periods, and it's held up very well. And also, you're basically applying the same type of rationale to both relative and absolute. So, in a sense, they give some confirmation of each other. This look back was first used in a 1937 research paper, and it has held up well ever since then.
TM: Does Fed policy keeping rates so low for so long affect the way the dual momentum model is going to work?
GA: Yes. It has in several ways. First of all, quantitative easing and low interest rates have made stocks appreciate quite a bit over the last 6 years. I believe the stock market is up 160 percent from 2009 through 2014. So, when you have a very powerful stock market, it's hard for any other approach, one that includes elements of trend following especially, to keep up with it.
I think, on a relative basis, that explains why we have lagged behind somewhat during the last six plus years. I don't think that kind of strong performance is likely to continue indefinitely into the future. The second aspect is that low interest rates have meant that when we get out of stocks temporarily, we've earned very little during that time. Again, I think when more normal conditions prevail, we'll find a better return on the fixed income portion of our portfolios.
TM: Back in the nineties, when you were out of the market, you could still be earning 5 or 6 percent on that portion of the money. That's just not the case the last six years, and probably for another year or two, I guess. Would that sound about accurate?
GA: That sounds accurate to me. I think the future bodes well for dual momentum on a relative basis compared to what's been going on the past six or so years.
TM: Before we wrap up, do you have any thoughts on dual momentum and individual investors using it?
GA: I don't see any reason why they shouldn't use it. If they'll go to my website, they'll see some model performance figures and other information there. My blog goes into even more detail. If you compare it to anything else out there, I can't see any reason why many people wouldn't want to use dual momentum.
TM: The website is www.optimalmomentum.com. And your book, "Dual Momentum," has been a top seller, and is easily available in bookstores and Amazon. Is that correct?
GA: Yes. It's a 5-star rated book on Amazon with over 200 reviews, which is unusual for an investment book. If you go to Amazon, you can read some of the reviews and endorsements there. Also, you can take advantage of the look inside feature to get a better feel for what dual momentum is all about.
TM: If this becomes really popular and everyone starts doing it, does the anomaly of dual momentum go away?
GA: If it becomes really popular, the anomaly effect can become reduced. However, I don't think it's ever going to become really popular. Based on all of the push back I've been getting since talking about dual momentum, I don't see that happening at all. There's an army of buy and hold type people out there who refuse to consider anything like it. The same applies to value investors. You have so many people being brought up with the fact that you just buy and hold, so there's considerable career risk, especially for investment professionals who won't do any type of market timing or anything that isn't along traditional lines. I don't think they're going to jump on the momentum bandwagon, certainly not on the dual momentum.
TM: Right. It's okay to lose money as long as everybody else is losing money.
GA: That's correct.
TM: I always remind people, when they're dealing with Wall Street, that the fund manager managing their funds, his biggest fault when he came to work to this morning is not how much money is he going to make you, but how does he keep his mid six-figure job.
GA: Unfortunately, that's true. That's why I wanted to put something together that anyone can go out and do themselves very easily and quickly, and where it makes logical sense. In my book, I explain all the reasons why momentum works and why I think it should continue to work. They won't have need for these high priced advisors.
Image Credit: Public Domain
© 2017 Benzinga.com. Benzinga does not provide investment advice. All rights reserved.
Tags: Dual Momentum