Momentum factor and trend following

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
Topic Author
konik
Posts: 33
Joined: Mon Aug 17, 2020 2:44 pm

Momentum factor and trend following

Post by konik »

Hi all, I have a theoretical question. Please explain me something, links to papers are welcome.

There is a momentum factor explaining some part of stock returns. Basically if I screen great performers over last 12 months and buy them, while screening worst performers and selling them, I get better than broad market returns (although with greater risks).

On the other hand there is trend following strategy for asset classes, where you buy or sell them depending on the last say 200 days performance. As I understand it is based on the very similar phenomenon. But the result of trend following is reduced returns with reduced risk of drawdowns.

Why two similar effects leads to two opposite results?
HappyPappy
Posts: 162
Joined: Sun Jan 02, 2022 4:05 pm

Re: Momentum factor and trend following

Post by HappyPappy »

konik wrote: Sun Jul 31, 2022 12:21 pm Hi all, I have a theoretical question. Please explain me something, links to papers are welcome.

There is a momentum factor explaining some part of stock returns. Basically if I screen great performers over last 12 months and buy them, while screening worst performers and selling them, I get better than broad market returns (although with greater risks).

On the other hand there is trend following strategy for asset classes, where you buy or sell them depending on the last say 200 days performance. As I understand it is based on the very similar phenomenon. But the result of trend following is reduced returns with reduced risk of drawdowns.

Why two similar effects leads to two opposite results?
To quote an astute poster on these boards: "links to papers are welcome." :) Is what you posted been seriously found to be true?
Topic Author
konik
Posts: 33
Joined: Mon Aug 17, 2020 2:44 pm

Re: Momentum factor and trend following

Post by konik »

HappyPappy wrote: Sun Jul 31, 2022 12:26 pm
konik wrote: Sun Jul 31, 2022 12:21 pm Hi all, I have a theoretical question. Please explain me something, links to papers are welcome.

There is a momentum factor explaining some part of stock returns. Basically if I screen great performers over last 12 months and buy them, while screening worst performers and selling them, I get better than broad market returns (although with greater risks).

On the other hand there is trend following strategy for asset classes, where you buy or sell them depending on the last say 200 days performance. As I understand it is based on the very similar phenomenon. But the result of trend following is reduced returns with reduced risk of drawdowns.

Why two similar effects leads to two opposite results?
To quote an astute poster on these boards: "links to papers are welcome." :) Is what you posted been seriously found to be true?
Regarding momentum factor there is a huge pile of historic evidences, which you can easily find anywhere on this forum as well as a lot of criticism for this strategy. Regarding trend following - I don't know, I have started reading on that only recently. See for example Larry Swedroe https://www.etf.com/sections/index-inve ... -insurance
Logan Roy
Posts: 264
Joined: Sun May 29, 2022 10:15 am

Re: Momentum factor and trend following

Post by Logan Roy »

konik wrote: Sun Jul 31, 2022 12:21 pm Hi all, I have a theoretical question. Please explain me something, links to papers are welcome.

There is a momentum factor explaining some part of stock returns. Basically if I screen great performers over last 12 months and buy them, while screening worst performers and selling them, I get better than broad market returns (although with greater risks).

On the other hand there is trend following strategy for asset classes, where you buy or sell them depending on the last say 200 days performance. As I understand it is based on the very similar phenomenon. But the result of trend following is reduced returns with reduced risk of drawdowns.

Why two similar effects leads to two opposite results?
Well with asset classes, a simple explanation could be that your average portfolio holdings in a multi-asset trend strategy (over a year) contain more bonds, gold, etc. So an appropriate benchmark might not be particularly stock-heavy. In my experience, if one benchmarks correctly, trend following backtests tend to perform similarly enough to buy-and-hold strategies, but with smaller drawdowns.

The problem with momentum is probably best elucidated if one imagines the whole market follows a 12 month buy and switch strategy. Who wins? What happens to the effect? Of course, it's still zero-sum. But the winners are those front-running the losers on the trades. An old trading joke was that 'edge' was using the 199 day moving average (because everyone else was using the 200 day). But the principle is sound. You have to be ahead of the curve. So if the market's switching on 12 months, the profitable momentum effect shifts to 11.

So the reason everyone can't profit from momentum (aside markets being zero-sum) is because the profitable traders still have to be ahead of the average traders, and that process of people trying to get in front changes the timing you'd have to use. So the principle of momentum is pretty mathematically solid, but the details (e.g. 12 months, 200 SMA, etc.) are completely unstable, and very prone to back-fitting. And the details dictate whether you're winning or losing – the equivalent of a bear-market for a momentum traders is 'whipsawing' (which is when the timing signals consistently put you in the wrong assets – which, in a random walk, is just as likely as the contrary). The idea momentum 'exists' (like other factors) in some pure form is (imo) very tempting and very dangerous.
Hydromod
Posts: 848
Joined: Tue Mar 26, 2019 10:21 pm

Re: Momentum factor and trend following

Post by Hydromod »

Logan Roy wrote: Sun Jul 31, 2022 2:08 pm
konik wrote: Sun Jul 31, 2022 12:21 pm Hi all, I have a theoretical question. Please explain me something, links to papers are welcome.

There is a momentum factor explaining some part of stock returns. Basically if I screen great performers over last 12 months and buy them, while screening worst performers and selling them, I get better than broad market returns (although with greater risks).

On the other hand there is trend following strategy for asset classes, where you buy or sell them depending on the last say 200 days performance. As I understand it is based on the very similar phenomenon. But the result of trend following is reduced returns with reduced risk of drawdowns.

Why two similar effects leads to two opposite results?
Well with asset classes, a simple explanation could be that your average portfolio holdings in a multi-asset trend strategy (over a year) contain more bonds, gold, etc. So an appropriate benchmark might not be particularly stock-heavy. In my experience, if one benchmarks correctly, trend following backtests tend to perform similarly enough to buy-and-hold strategies, but with smaller drawdowns.

The problem with momentum is probably best elucidated if one imagines the whole market follows a 12 month buy and switch strategy. Who wins? What happens to the effect? Of course, it's still zero-sum. But the winners are those front-running the losers on the trades. An old trading joke was that 'edge' was using the 199 day moving average (because everyone else was using the 200 day). But the principle is sound. You have to be ahead of the curve. So if the market's switching on 12 months, the profitable momentum effect shifts to 11.

So the reason everyone can't profit from momentum (aside markets being zero-sum) is because the profitable traders still have to be ahead of the average traders, and that process of people trying to get in front changes the timing you'd have to use. So the principle of momentum is pretty mathematically solid, but the details (e.g. 12 months, 200 SMA, etc.) are completely unstable, and very prone to back-fitting. And the details dictate whether you're winning or losing – the equivalent of a bear-market for a momentum traders is 'whipsawing' (which is when the timing signals consistently put you in the wrong assets – which, in a random walk, is just as likely as the contrary). The idea momentum 'exists' (like other factors) in some pure form is (imo) very tempting and very dangerous.
You might check out Allocate Smartly for more insight. This is a site with lots of tests of various strategies.

I agree with much of this explanation.

It seems to me that momentum strategies are a bit hit and miss. For long periods, you probably don't gain much return with momentum and just see greater volatility. Then there are those crisis times when things are making big moves, where timing is very critical to success. Momentum strategies react differently than the market, so there's potential for big undershoots and big overshoots relative to the market. Momentum traders hope that there are more or bigger hits than misses, but it can be very stressful to be out of synch with the market, and there really aren't a lot of data points to go off of.

Historically the momentum trends were easier to spot and take advantage of, IMO. I am convinced that trends are moving faster and faster as the market becomes more influenced by algorithmic trading and less influenced by trading frictions.
Logan Roy
Posts: 264
Joined: Sun May 29, 2022 10:15 am

Re: Momentum factor and trend following

Post by Logan Roy »

Hydromod wrote: Sun Jul 31, 2022 2:32 pm
Logan Roy wrote: Sun Jul 31, 2022 2:08 pm
konik wrote: Sun Jul 31, 2022 12:21 pm Hi all, I have a theoretical question. Please explain me something, links to papers are welcome.

There is a momentum factor explaining some part of stock returns. Basically if I screen great performers over last 12 months and buy them, while screening worst performers and selling them, I get better than broad market returns (although with greater risks).

On the other hand there is trend following strategy for asset classes, where you buy or sell them depending on the last say 200 days performance. As I understand it is based on the very similar phenomenon. But the result of trend following is reduced returns with reduced risk of drawdowns.

Why two similar effects leads to two opposite results?
Well with asset classes, a simple explanation could be that your average portfolio holdings in a multi-asset trend strategy (over a year) contain more bonds, gold, etc. So an appropriate benchmark might not be particularly stock-heavy. In my experience, if one benchmarks correctly, trend following backtests tend to perform similarly enough to buy-and-hold strategies, but with smaller drawdowns.

The problem with momentum is probably best elucidated if one imagines the whole market follows a 12 month buy and switch strategy. Who wins? What happens to the effect? Of course, it's still zero-sum. But the winners are those front-running the losers on the trades. An old trading joke was that 'edge' was using the 199 day moving average (because everyone else was using the 200 day). But the principle is sound. You have to be ahead of the curve. So if the market's switching on 12 months, the profitable momentum effect shifts to 11.

So the reason everyone can't profit from momentum (aside markets being zero-sum) is because the profitable traders still have to be ahead of the average traders, and that process of people trying to get in front changes the timing you'd have to use. So the principle of momentum is pretty mathematically solid, but the details (e.g. 12 months, 200 SMA, etc.) are completely unstable, and very prone to back-fitting. And the details dictate whether you're winning or losing – the equivalent of a bear-market for a momentum traders is 'whipsawing' (which is when the timing signals consistently put you in the wrong assets – which, in a random walk, is just as likely as the contrary). The idea momentum 'exists' (like other factors) in some pure form is (imo) very tempting and very dangerous.
You might check out Allocate Smartly for more insight. This is a site with lots of tests of various strategies.

I agree with much of this explanation.

It seems to me that momentum strategies are a bit hit and miss. For long periods, you probably don't gain much return with momentum and just see greater volatility. Then there are those crisis times when things are making big moves, where timing is very critical to success. Momentum strategies react differently than the market, so there's potential for big undershoots and big overshoots relative to the market. Momentum traders hope that there are more or bigger hits than misses, but it can be very stressful to be out of synch with the market, and there really aren't a lot of data points to go off of.

Historically the momentum trends were easier to spot and take advantage of, IMO. I am convinced that trends are moving faster and faster as the market becomes more influenced by algorithmic trading and less influenced by trading frictions.
Absolutely – and trend following is often a huge part of what hedge funds, and hedge fund traders (Soros, Druckenmiller, etc.), do, and the game of trying to play it better than the next guy has been raging since certainly the days of Jesse Livermore. But of course, the problem with published strategies is no different from the problem with publishing a winning poker strategy - as soon as you know what the other guy's doing, you can front run them on every trade.

I think it's a very interesting engineering problem. Modern trend following algorithms will often do things like slash position sizes as volatility picks up – because of course: volatility is noise, and in many ways, all successful trend following is is delineating trend from noise. So the more noise, the less you can trust your ears. And then avoiding whipsawing involves generalising how you measure trends (so short-term volatility doesn't trigger false signals) and how you trade (so you don't want to be jumping in and out of things, in a way that can have you jumping out when the market's about to rise, and in, when it's about to fall).

I've got a concept (very presumptuous) called decision-compounding, or signal-compounding, which is similar in practice to the trading concept of 'laddering', and means if you get whipsawed, it's on these fractional positions. And when you're right about a trend, it's because there's been a succession of right decisions. Another thing I always tell people is if you didn't design the system yourself, you'll never know how to deal with things when it goes wrong (which they all will). But it's a very interesting problem, based on a very simple observation. And I think the standard of academic work on momentum is absolutely terrible (which always gives me some hope that there are people doing it worse than I am).
Topic Author
konik
Posts: 33
Joined: Mon Aug 17, 2020 2:44 pm

Re: Momentum factor and trend following

Post by konik »

Hydromod wrote: Sun Jul 31, 2022 2:32 pm
Logan Roy wrote: Sun Jul 31, 2022 2:08 pm In my experience, if one benchmarks correctly, trend following backtests tend to perform similarly enough to buy-and-hold strategies, but with smaller drawdowns.
It seems to me that momentum strategies are a bit hit and miss. For long periods, you probably don't gain much return with momentum and just see greater volatility. Then there are those crisis times when things are making big moves, where timing is very critical to success. Momentum strategies react differently than the market, so there's potential for big undershoots and big overshoots relative to the market.
Doesn't those two statements contradict each other? That exactly was my initial question. Trend following and momentum chasing is basically the same. But in one case we get smaller drawdowns, while in another - greater volatility. What do I miss?
Topic Author
konik
Posts: 33
Joined: Mon Aug 17, 2020 2:44 pm

Re: Momentum factor and trend following

Post by konik »

I think I got it.

In trend following we spend some amount of time in cash (in case of just 1 risk asset), reducing the average exposure to the named asset, reducing both risk and return.

With momentum factor we just rotate between two equally risky assets, while additionally taking load of factor risk or undiversified risk (depending on your ideological stance). So higher risk with maybe higher return.

Am I correct with this explanation?
User avatar
hiddenpower
Posts: 247
Joined: Wed Nov 18, 2020 12:24 am

Re: Momentum factor and trend following

Post by hiddenpower »

Relative momentum keeps you switching between equities, absolute momentum determines if you’re in equities at all
User avatar
whodidntante
Posts: 11013
Joined: Thu Jan 21, 2016 11:11 pm
Location: outside the echo chamber

Re: Momentum factor and trend following

Post by whodidntante »

The problem is that the trading costs make off with most of your returns. I do like trading time series momentum of an asset class, which if done at the fund level is generally referred to as managed futures. And I also like momentum screens like those those employed by DFA and Avantis.
User avatar
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Momentum factor and trend following

Post by willthrill81 »

konik wrote: Sun Jul 31, 2022 12:21 pm On the other hand there is trend following strategy for asset classes, where you buy or sell them depending on the last say 200 days performance. As I understand it is based on the very similar phenomenon. But the result of trend following is reduced returns with reduced risk of drawdowns.
Maybe, maybe not. It depends on what the mode of comparison is. If you're comparing absolute returns, it's somewhat dependent on the starting date. If you're taking the reduced volatility and maximum drawdowns into effect, the 200 DMA has provided much better returns than portfolios that used bonds for the purpose of reducing volatility and maximum drawdown. Here is a link to the data going back to 1985 in PV.
I have left the forum but occasionally check PMs.
User avatar
hiddenpower
Posts: 247
Joined: Wed Nov 18, 2020 12:24 am

Re: Momentum factor and trend following

Post by hiddenpower »

whodidntante wrote: Sun Jul 31, 2022 4:15 pm The problem is that the trading costs make off with most of your returns. I do like trading time series momentum of an asset class, which if done at the fund level is generally referred to as managed futures. And I also like momentum screens like those those employed by DFA and Avantis.
Dual momentum author handwaves that the tax drag isn't that bad :confused, but manual momentum strategies seem fine in tax sheltered accounts.
Logan Roy
Posts: 264
Joined: Sun May 29, 2022 10:15 am

Re: Momentum factor and trend following

Post by Logan Roy »

konik wrote: Sun Jul 31, 2022 3:19 pm
Hydromod wrote: Sun Jul 31, 2022 2:32 pm
Logan Roy wrote: Sun Jul 31, 2022 2:08 pm In my experience, if one benchmarks correctly, trend following backtests tend to perform similarly enough to buy-and-hold strategies, but with smaller drawdowns.
It seems to me that momentum strategies are a bit hit and miss. For long periods, you probably don't gain much return with momentum and just see greater volatility. Then there are those crisis times when things are making big moves, where timing is very critical to success. Momentum strategies react differently than the market, so there's potential for big undershoots and big overshoots relative to the market.
Doesn't those two statements contradict each other? That exactly was my initial question. Trend following and momentum chasing is basically the same. But in one case we get smaller drawdowns, while in another - greater volatility. What do I miss?
It's all about implementation. If you're just holding stocks with the highest 6-12m returns, you could be very concentrated in certain sectors and styles, and if something changes – like an earnings report falling short – your whole portfolio could very quickly be on the wrong side of the market.

With multi-asset strategies, you're probably trading the market as a whole (rather than individual stocks), and switching into things like bonds. So you're always more diversified, and you're usually being pulled out of those markets at early signs of trouble (whereas, again, in a stock market momentum strategy, you're always in stocks).

But there are much more sophisticated approaches to risk management – so trend following with individual stocks doesn't have to be volatile (as above, it could involve much more sitting in cash). In fact, the Tom Basso coin flip demonstrated that, with decent risk management, you could enter positions at random and still do very well.
Laurizas
Posts: 418
Joined: Mon Dec 31, 2018 4:44 am
Location: Lithuania

Re: Momentum factor and trend following

Post by Laurizas »

willthrill81 wrote: Sun Jul 31, 2022 4:18 pm [Here is a link to the data going back to 1985 in PV.
Why 1985?
User avatar
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Momentum factor and trend following

Post by willthrill81 »

Laurizas wrote: Tue Aug 02, 2022 12:49 pm
willthrill81 wrote: Sun Jul 31, 2022 4:18 pm [Here is a link to the data going back to 1985 in PV.
Why 1985?
That's the latest that Portfolio Visualizer goes back when doing this type of analysis. Others have done similar analyses in peer-reviewed journal articles going much further back and found similar results. Clare et al. (2013) said the following about the 200 DMA strategy using data going back to 1952.
We find that a range of fairly simple rules, including the popular 200-day moving average trading rule, dominate the long only, passive investment in the index.
https://openaccess.city.ac.uk/id/eprint ... 126476.pdf
I have left the forum but occasionally check PMs.
Laurizas
Posts: 418
Joined: Mon Dec 31, 2018 4:44 am
Location: Lithuania

Re: Momentum factor and trend following

Post by Laurizas »

Strange, I can go to 1978
User avatar
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Momentum factor and trend following

Post by willthrill81 »

Laurizas wrote: Tue Aug 02, 2022 12:54 pm Strange, I can go to 1978
Ah, they must have changed it.

With VFINX and cash, the 200 DMA returned 11.21% nominal vs. 11.61% for buy-and-hold, but the 200 DMA had a max. monthly drawdown of -23.5% compared to -51% for buy-and-hold. And this is start date sensitive since the 200 DMA was ahead in returns back in early 2020.
I have left the forum but occasionally check PMs.
Beliavsky
Posts: 1071
Joined: Sun Jun 29, 2014 10:21 am

Re: Momentum factor and trend following

Post by Beliavsky »

A long-only strategy that has you invested in the quintile of stocks with the highest momentum has full exposure to the stock market and may outperform the market in CAGR if the momentum effect is real.

For an asset such as SPY that outperformed cash by a wide margin, a trend-following strategy that has you in cash some of the time will have a tough time generating a higher CAGR in backtests, but the trend-following strategy may have a higher Sharpe ratio. I have written about this in previous posts, which can be searched.
User avatar
nisiprius
Advisory Board
Posts: 47244
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: Momentum factor and trend following

Post by nisiprius »

Discussions of momentum are incomplete without discussions of "momentum crashes." A 2014 article in Chicago Booth Review, Understanding momentum crashes, Chicago Booth Review says that
Curiously enough, this strategy often works. Yet while momentum can produce consistent returns, it also is prone to rare, sudden crashes.

Those crashes predominantly occur at pivotal points in the economy, when a bear market suddenly ends and stocks dramatically rebound. In such environments, last year’s losers perform magnificently, while winners lag dramatically—the opposite of what a momentum trader would be betting on....
The article is referring to a method that tries to address this issue, "dynamic momentum trading." Instead of simply allocating to the momentum factor, it layers on a mechanism for predicting momentum crashes and dodging them:
Moskowitz and Daniel propose a formula for “dynamic momentum trading,” which allows investors to increase their momentum trading when volatility is low and decrease it when volatility is high.
That sounds like factor timing to me.

The Chicago Booth article concludes ambiguously:
Defenders of the efficient-markets hypothesis—which suggests in its strongest form that past price movements should not predict future results—speculate that investors can reap consistent gains from momentum trading for a time because they accept the risk of a calamitous reversal. Yet there also are signs that investors are slow to incorporate new information, or that there is a bandwagon effect for recently fortunate trading strategies. The explanation for the mechanics underlying momentum trading, like the success of the phenomenon itself, remains elusive.
It should also be noted that, quoting a summary by Jared Kizer,
Some factor premia have negative correlation with each other, namely value and momentum. Practically, it is impossible, therefore, to own a stock portfolio that is both deeply tilted toward value and deeply tilted toward momentum. At best, stock portfolios can be moderately tilted toward both.
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.
User avatar
nedsaid
Posts: 16451
Joined: Fri Nov 23, 2012 12:33 pm

Re: Momentum factor and trend following

Post by nedsaid »

I think there is something to trend following. I had a now deceased friend who used Dick Fabian's moving average strategy. He advocated keeping a moving a 39 week moving average. When the price of your fund was above the average, you would buy and stay in. When the price of your fund moved below the moving average, you would sell. So this got you in somewhere above the bottom and got you out somewhere below the top. On the surface, this idea makes a lot of sense but I never would do this myself.

My understanding was that Dick Fabian's method worked fairly well and I remember that he had a newsletter. What I read was that when the son took things over, he tweaked his father's simple method and it hasn't worked very well since.

Mostly I am a buy and hold kind of guy. My belief is that if you hold good assets for long periods of time, riding out the volatility, that you will be richly rewarded. The old saying that it is time IN the market and not the timing OF the market that makes you money. You need to be on the train when it leaves the station.
A fool and his money are good for business.
Gaston
Posts: 604
Joined: Wed Aug 21, 2013 7:12 pm

Re: Momentum factor and trend following

Post by Gaston »

Not a helpful comment, I know, but I find momentum investing kind of odd. My guess is that the #1 practice that costs investors the most money is buying the hottest stocks or sectors on the assumption they will get hotter still. You see it all the time, with money flowing into funds whose managers allegedly have a hot hand. That’s why dollar-weighted returns always fall below time-weighted returns.

I hope the academics have a better handle than we laymen on how to profit from momentum.
User avatar
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Momentum factor and trend following

Post by willthrill81 »

nedsaid wrote: Tue Aug 02, 2022 8:10 pmMy belief is that if you hold good assets for long periods of time, riding out the volatility, that you will be richly rewarded. The old saying that it is time IN the market and not the timing OF the market that makes you money. You need to be on the train when it leaves the station.
Some of the factors (no pun intended) that moved me to switch to trend following were that assets like U.S. or ex-U.S. stock can and have gone nowhere for decades, and you have no assurance whatsoever that such assets will recover in time to 'save your day'. That the U.S. stock market, for instance, has always recovered eventually does not mean that it will do so in time to keep you from going broke. And bonds aren't a panacea these days due to their yields being around 0% real.

I don't view trend following as a good means of generating above market returns. Rather, I have always viewed it primarily as a risk mitigation tool. Getting below market returns during the good times (usually true of most trend following strategies) is worth it to me if it means that I'm likely, to the extent that the future resembles the past, to get above market returns during the bad times. I know that I might still underperform even during the bad times, and I accept that risk. After all, it's undeniable that all paths carry risk, and we must all live with the consequences of our chosen path.
I have left the forum but occasionally check PMs.
User avatar
Forester
Posts: 2346
Joined: Sat Jan 19, 2019 2:50 pm
Location: UK

Re: Momentum factor and trend following

Post by Forester »

One can probably achieve similar results, but remove the risk of bad timing luck, by pairing low volatility stocks with high quality bonds.

Even a multitude of lookback periods can incur whipsaw: https://www.portfoliovisualizer.com/bac ... ion4_3=100
Amateur Self-Taught Senior Macro Strategist
sixtyforty
Posts: 626
Joined: Tue Nov 25, 2014 12:22 pm
Location: USA

Re: Momentum factor and trend following

Post by sixtyforty »

Regardless of what momentum or trend following strategy is used, the most difficult part is following it consistently and long enough through difficult times for it work (if it works). It's also very difficult to follow a strategy without tweaking it. Once you start tweaking it your off on another strategy.
"Simplicity is the ultimate sophistication" - Leonardo Da Vinci
bog007
Posts: 352
Joined: Fri Feb 28, 2020 3:27 am

Re: Momentum factor and trend following

Post by bog007 »

The 200 day (red line) as I noted then that broke above 200 day in 2020. need to get above then test as happened in spring of 2020. spring it did til drop below 200day. Soup for u.


https://schrts.co/ZcAGRKrW
If I had a dollar for every time I heard something was on sale I'd be rich.
User avatar
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Momentum factor and trend following

Post by willthrill81 »

sixtyforty wrote: Wed Aug 03, 2022 8:47 am Regardless of what momentum or trend following strategy is used, the most difficult part is following it consistently and long enough through difficult times for it work (if it works). It's also very difficult to follow a strategy without tweaking it. Once you start tweaking it your off on another strategy.
This is true, but it's true of all investment strategies. We see 'buy-and-hold' investors here 'tweaking' their strategies all the time. But I don't see this as necessarily a bad thing either.
Last edited by willthrill81 on Wed Aug 03, 2022 11:59 am, edited 1 time in total.
I have left the forum but occasionally check PMs.
Logan Roy
Posts: 264
Joined: Sun May 29, 2022 10:15 am

Re: Momentum factor and trend following

Post by Logan Roy »

Gaston wrote: Tue Aug 02, 2022 9:06 pm Not a helpful comment, I know, but I find momentum investing kind of odd. My guess is that the #1 practice that costs investors the most money is buying the hottest stocks or sectors on the assumption they will get hotter still. You see it all the time, with money flowing into funds whose managers allegedly have a hot hand. That’s why dollar-weighted returns always fall below time-weighted returns.

I hope the academics have a better handle than we laymen on how to profit from momentum.
Personally, I'd trust the monkeys and random number generators with my money long before I'd trust the academics.

Momentum's just the raw game of trading. If you're ahead of the curve: you generate alpha; if you're behind it: negative alpha. One of the main reasons I use forums and social media is because if I spot a trend, and no one's talking about it, it's a clue that I may be ahead of the curve. If I'm thinking of adding to a momentum trade, and other investors are looking for reassurance that it's a good time to buy, it's a flashing amber light. "Whenever you find yourself on the side of the majority, it is time to pause and reflect." – Mark Twain.

I hate the way momentum's been turned into a 'factor'. You play chess for 30 years, and then a bunch of humanities grads start discovering chess 'factors', or poker factors – and it's just a load of nonsense to me. Momentum's just a way to measure the latency of information as it moves through markets, and to some extent, the time it takes for large investors to build and exit positions. Which is why it seems very persistent. To erase momentum, markets would need the ability to price in sudden step changes in price, which isn't common.
User avatar
BlueEars
Posts: 3930
Joined: Sat Mar 10, 2007 12:15 am
Location: West Coast

Re: Momentum factor and trend following

Post by BlueEars »

Logan Roy wrote: Wed Aug 03, 2022 10:15 am ... Momentum's just a way to measure the latency of information as it moves through markets, and to some extent, the time it takes for large investors to build and exit positions. Which is why it seems very persistent. To erase momentum, markets would need the ability to price in sudden step changes in price, which isn't common.
I like that "latency of information" way of describing it.
Logan Roy
Posts: 264
Joined: Sun May 29, 2022 10:15 am

Re: Momentum factor and trend following

Post by Logan Roy »

BlueEars wrote: Wed Aug 03, 2022 10:55 am
Logan Roy wrote: Wed Aug 03, 2022 10:15 am ... Momentum's just a way to measure the latency of information as it moves through markets, and to some extent, the time it takes for large investors to build and exit positions. Which is why it seems very persistent. To erase momentum, markets would need the ability to price in sudden step changes in price, which isn't common.
I like that "latency of information" way of describing it.
Thank you. I should say I missed another interesting take on momentum, which is the reflexive. Which is the idea that, e.g., if large amounts of capital flow into a business (assuming the business is made of the right stuff, and there's a place for it), that has an effect on the business' ability to build a market-leading position. Tesla might be an example of reflexive momentum. The idea that price action isn't just an indicator, but can have a material effect on fundamentals.
Morik
Posts: 1038
Joined: Tue Nov 25, 2014 12:26 pm

Re: Momentum factor and trend following

Post by Morik »

Logan Roy wrote: Wed Aug 03, 2022 12:16 pm
BlueEars wrote: Wed Aug 03, 2022 10:55 am
Logan Roy wrote: Wed Aug 03, 2022 10:15 am ... Momentum's just a way to measure the latency of information as it moves through markets, and to some extent, the time it takes for large investors to build and exit positions. Which is why it seems very persistent. To erase momentum, markets would need the ability to price in sudden step changes in price, which isn't common.
I like that "latency of information" way of describing it.
Thank you. I should say I missed another interesting take on momentum, which is the reflexive. Which is the idea that, e.g., if large amounts of capital flow into a business (assuming the business is made of the right stuff, and there's a place for it), that has an effect on the business' ability to build a market-leading position. Tesla might be an example of reflexive momentum. The idea that price action isn't just an indicator, but can have a material effect on fundamentals.
But doesn't trading stocks on the market not put money into the company? They only get money when they issue new shares?
Or is the idea that a run-up on the stock allows them to raise more capital per share they issue going forward?
Logan Roy
Posts: 264
Joined: Sun May 29, 2022 10:15 am

Re: Momentum factor and trend following

Post by Logan Roy »

Morik wrote: Wed Aug 03, 2022 12:24 pm
Logan Roy wrote: Wed Aug 03, 2022 12:16 pm
BlueEars wrote: Wed Aug 03, 2022 10:55 am
Logan Roy wrote: Wed Aug 03, 2022 10:15 am ... Momentum's just a way to measure the latency of information as it moves through markets, and to some extent, the time it takes for large investors to build and exit positions. Which is why it seems very persistent. To erase momentum, markets would need the ability to price in sudden step changes in price, which isn't common.
I like that "latency of information" way of describing it.
Thank you. I should say I missed another interesting take on momentum, which is the reflexive. Which is the idea that, e.g., if large amounts of capital flow into a business (assuming the business is made of the right stuff, and there's a place for it), that has an effect on the business' ability to build a market-leading position. Tesla might be an example of reflexive momentum. The idea that price action isn't just an indicator, but can have a material effect on fundamentals.
But doesn't trading stocks on the market not put money into the company? They only get money when they issue new shares?
Or is the idea that a run-up on the stock allows them to raise more capital per share they issue going forward?
Yes, so Tesla's done these periodic share issuances since becoming public. You also look at the amount of debt they've been able to take on (which you might view as relative to their market cap or cash balance).
User avatar
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Momentum factor and trend following

Post by willthrill81 »

Morik wrote: Wed Aug 03, 2022 12:24 pm But doesn't trading stocks on the market not put money into the company?
Not at all. If I buy a stock from you, the company whose stock is traded doesn't get a penny. To do that, they would need to issue new stock, which dilutes the value of the existing stock.
I have left the forum but occasionally check PMs.
Morik
Posts: 1038
Joined: Tue Nov 25, 2014 12:26 pm

Re: Momentum factor and trend following

Post by Morik »

willthrill81 wrote: Wed Aug 03, 2022 12:50 pm
Morik wrote: Wed Aug 03, 2022 12:24 pm But doesn't trading stocks on the market not put money into the company?
Not at all. If I buy a stock from you, the company whose stock is traded doesn't get a penny. To do that, they would need to issue new stock, which dilutes the value of the existing stock.
I didn't write that very clearly--I meant "trading stocks on the market doesn't put money into the company".
User avatar
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Momentum factor and trend following

Post by willthrill81 »

Morik wrote: Wed Aug 03, 2022 1:09 pm
willthrill81 wrote: Wed Aug 03, 2022 12:50 pm
Morik wrote: Wed Aug 03, 2022 12:24 pm But doesn't trading stocks on the market not put money into the company?
Not at all. If I buy a stock from you, the company whose stock is traded doesn't get a penny. To do that, they would need to issue new stock, which dilutes the value of the existing stock.
I didn't write that very clearly--I meant "trading stocks on the market doesn't put money into the company".
I see. The answer is still no.
I have left the forum but occasionally check PMs.
Morik
Posts: 1038
Joined: Tue Nov 25, 2014 12:26 pm

Re: Momentum factor and trend following

Post by Morik »

willthrill81 wrote: Wed Aug 03, 2022 1:18 pm
Morik wrote: Wed Aug 03, 2022 1:09 pm
willthrill81 wrote: Wed Aug 03, 2022 12:50 pm
Morik wrote: Wed Aug 03, 2022 12:24 pm But doesn't trading stocks on the market not put money into the company?
Not at all. If I buy a stock from you, the company whose stock is traded doesn't get a penny. To do that, they would need to issue new stock, which dilutes the value of the existing stock.
I didn't write that very clearly--I meant "trading stocks on the market doesn't put money into the company".
I see. The answer is still no.
Regarding diluting the stock--I used to think that but I read some article that explained it in a way that changed my view. I don't recall where the article was or I'd link it.

Yes the # of outstanding shares went up so your ownership % in the company decreased, but the company also got cash in exchange for that, which should increase the book value per share such that it ends up roughly being a wash for existing equity holders in terms of their overall holding value, right?
If a company pays its employees fully in cash, vs partly in cash and partly in newly issued stock, either way is roughly equivalent in terms of book value per share. If they hadn't issued the new stock they'd have had to spend cash instead.
User avatar
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Momentum factor and trend following

Post by willthrill81 »

Morik wrote: Wed Aug 03, 2022 1:25 pm
willthrill81 wrote: Wed Aug 03, 2022 1:18 pm
Morik wrote: Wed Aug 03, 2022 1:09 pm
willthrill81 wrote: Wed Aug 03, 2022 12:50 pm
Morik wrote: Wed Aug 03, 2022 12:24 pm But doesn't trading stocks on the market not put money into the company?
Not at all. If I buy a stock from you, the company whose stock is traded doesn't get a penny. To do that, they would need to issue new stock, which dilutes the value of the existing stock.
I didn't write that very clearly--I meant "trading stocks on the market doesn't put money into the company".
I see. The answer is still no.
Regarding diluting the stock--I used to think that but I read some article that explained it in a way that changed my view. I don't recall where the article was or I'd link it.

Yes the # of outstanding shares went up so your ownership % in the company decreased, but the company also got cash in exchange for that, which should increase the book value per share such that it ends up roughly being a wash for existing equity holders in terms of their overall holding value, right?
If a company pays its employees fully in cash, vs partly in cash and partly in newly issued stock, either way is roughly equivalent in terms of book value per share. If they hadn't issued the new stock they'd have had to spend cash instead.
A company issuing more stock does not create more money and/or value out of thin air, and accurate accounting must reflect that.
I have left the forum but occasionally check PMs.
Morik
Posts: 1038
Joined: Tue Nov 25, 2014 12:26 pm

Re: Momentum factor and trend following

Post by Morik »

willthrill81 wrote: Wed Aug 03, 2022 1:27 pm
Morik wrote: Wed Aug 03, 2022 1:25 pm
willthrill81 wrote: Wed Aug 03, 2022 1:18 pm
Morik wrote: Wed Aug 03, 2022 1:09 pm
willthrill81 wrote: Wed Aug 03, 2022 12:50 pm

Not at all. If I buy a stock from you, the company whose stock is traded doesn't get a penny. To do that, they would need to issue new stock, which dilutes the value of the existing stock.
I didn't write that very clearly--I meant "trading stocks on the market doesn't put money into the company".
I see. The answer is still no.
Regarding diluting the stock--I used to think that but I read some article that explained it in a way that changed my view. I don't recall where the article was or I'd link it.

Yes the # of outstanding shares went up so your ownership % in the company decreased, but the company also got cash in exchange for that, which should increase the book value per share such that it ends up roughly being a wash for existing equity holders in terms of their overall holding value, right?
If a company pays its employees fully in cash, vs partly in cash and partly in newly issued stock, either way is roughly equivalent in terms of book value per share. If they hadn't issued the new stock they'd have had to spend cash instead.
A company issuing more stock does not create more money and/or value out of thin air, and accurate accounting must reflect that.
I'm not saying it creates more money or value out of thin air--I'm saying it is (roughly) a wash. They are reducing the % ownership of all existing shareholders, but they are also increasing their total assets by the amount of cash they get from that issuance. Of course the price per share isn't strictly linked to the company's book value, so it isn't a complete wash... but it also isn't a straightforward reduction of existing shareholder's total holding value by the same percentage that their ownership % decreased.

I think of it kind of like how turning an unrealized gain into a realized gain is just shifting money around. It isn't quite that, since a X% increase in company book value doesn't translate to a X% increase in their share price, but it is more like that I think than it is just straight up reducing existing shareholders' total holding value.
Logan Roy
Posts: 264
Joined: Sun May 29, 2022 10:15 am

Re: Momentum factor and trend following

Post by Logan Roy »

Yes, so issuing shares doesn't raise the market cap of Tesla, but it provides them with utilisable capital for growth. In 2020, Tesla issued $5bn in shares. This was a year before Tesla made their first full-year profit. So the value the market assigns a business affects their access to capital, which is why businesses go public. And reflexivity is the idea that the market puts capital where it thinks there's potential for growth; but there's also more potential for growth where there's more capital.
User avatar
nedsaid
Posts: 16451
Joined: Fri Nov 23, 2012 12:33 pm

Re: Momentum factor and trend following

Post by nedsaid »

willthrill81 wrote: Tue Aug 02, 2022 9:13 pm
nedsaid wrote: Tue Aug 02, 2022 8:10 pmMy belief is that if you hold good assets for long periods of time, riding out the volatility, that you will be richly rewarded. The old saying that it is time IN the market and not the timing OF the market that makes you money. You need to be on the train when it leaves the station.
Some of the factors (no pun intended) that moved me to switch to trend following were that assets like U.S. or ex-U.S. stock can and have gone nowhere for decades, and you have no assurance whatsoever that such assets will recover in time to 'save your day'. That the U.S. stock market, for instance, has always recovered eventually does not mean that it will do so in time to keep you from going broke. And bonds aren't a panacea these days due to their yields being around 0% real.

I don't view trend following as a good means of generating above market returns. Rather, I have always viewed it primarily as a risk mitigation tool. Getting below market returns during the good times (usually true of most trend following strategies) is worth it to me if it means that I'm likely, to the extent that the future resembles the past, to get above market returns during the bad times. I know that I might still underperform even during the bad times, and I accept that risk. After all, it's undeniable that all paths carry risk, and we must all live with the consequences of our chosen path.
Trend following is a great concept. I know Larry Swedroe has articles on it and on the Momentum factor in general, I just haven't studied the topic in depth and I never have tried it myself with my own investing. Mostly a buy and hold guy myself but I have done mild tactical asset allocation. So tactical asset allocation in my experience was more about mitigating risk and not so much about enhancing return. Sounds like trend following might be much the same.

A Bill Bernstein comment in one of his books stood out. He said that over long periods of time, and he means centuries, that the returns of stocks and bonds are identical. He also said that as societies get wealthier and wealthier that investment returns will decrease. So there might be a day when the Equity Risk Premium disappears, I don't think it will but it absolutely could happen. This is the thought experiment of thinking the unthinkable. We all need to do that.
A fool and his money are good for business.
Logan Roy
Posts: 264
Joined: Sun May 29, 2022 10:15 am

Re: Momentum factor and trend following

Post by Logan Roy »

nedsaid wrote: Wed Aug 03, 2022 7:37 pm
willthrill81 wrote: Tue Aug 02, 2022 9:13 pm
nedsaid wrote: Tue Aug 02, 2022 8:10 pmMy belief is that if you hold good assets for long periods of time, riding out the volatility, that you will be richly rewarded. The old saying that it is time IN the market and not the timing OF the market that makes you money. You need to be on the train when it leaves the station.
Some of the factors (no pun intended) that moved me to switch to trend following were that assets like U.S. or ex-U.S. stock can and have gone nowhere for decades, and you have no assurance whatsoever that such assets will recover in time to 'save your day'. That the U.S. stock market, for instance, has always recovered eventually does not mean that it will do so in time to keep you from going broke. And bonds aren't a panacea these days due to their yields being around 0% real.

I don't view trend following as a good means of generating above market returns. Rather, I have always viewed it primarily as a risk mitigation tool. Getting below market returns during the good times (usually true of most trend following strategies) is worth it to me if it means that I'm likely, to the extent that the future resembles the past, to get above market returns during the bad times. I know that I might still underperform even during the bad times, and I accept that risk. After all, it's undeniable that all paths carry risk, and we must all live with the consequences of our chosen path.
Trend following is a great concept. I know Larry Swedroe has articles on it and on the Momentum factor in general, I just haven't studied the topic in depth and I never have tried it myself with my own investing. Mostly a buy and hold guy myself but I have done mild tactical asset allocation. So tactical asset allocation in my experience was more about mitigating risk and not so much about enhancing return. Sounds like trend following might be much the same.

A Bill Bernstein comment in one of his books stood out. He said that over long periods of time, and he means centuries, that the returns of stocks and bonds are identical. He also said that as societies get wealthier and wealthier that investment returns will decrease. So there might be a day when the Equity Risk Premium disappears, I don't think it will but it absolutely could happen. This is the thought experiment of thinking the unthinkable. We all need to do that.
A lot of us thought we might already be there (and we still might). There's this idea of the "1% everything" economy. With persistently low inflation, we keep easing (as Japan has been for decades) and borrowing, and we want to maintain negative or neutral real yields (so we can keep borrowing), but certainly avoid deflation, so inflation sticks around 1%, bond yields stick around 1%, and as such, stocks are the only game in town, so markets climb to PE100 and stay there.

I think that's what's happening – and it's driven by automation – but that the inflation we've got now is like throwing a boulder into a calm pond, and that it's a heavily engineered global effort.
moneyflowin
Posts: 171
Joined: Sat Oct 16, 2021 6:14 am

Re: Momentum factor and trend following

Post by moneyflowin »

There are many trend-following strategies, and they've back-tested well the last 30 years. The reason is that trend following and TAA greatly outperforms buy and hold in deep bear markets, and we happened to have two of them in since 2000. TF can outperform by >30% in that type of market, which shows up as a significant CAGR advantage.

However, these strategies under-performed SPY since 2010 because we've been in a strong bull market with a number of sharp pullbacks and quick reversals (2010, 2011, 2015, 2018, 2020). TF does poorly in that type of market because they switch out of the asset after most of the loss has already happened, and they get you back into the asset after the sharp rally. I.e., they sell low and buy high.

In short, whether TF will outperform B&H in the future depends on the type of market we get from here.
Logan Roy
Posts: 264
Joined: Sun May 29, 2022 10:15 am

Re: Momentum factor and trend following

Post by Logan Roy »

moneyflowin wrote: Thu Aug 04, 2022 10:41 pm There are many trend-following strategies, and they've back-tested well the last 30 years. The reason is that trend following and TAA greatly outperforms buy and hold in deep bear markets, and we happened to have two of them in since 2000. TF can outperform by >30% in that type of market, which shows up as a significant CAGR advantage.

However, these strategies under-performed SPY since 2010 because we've been in a strong bull market with a number of sharp pullbacks and quick reversals (2010, 2011, 2015, 2018, 2020). TF does poorly in that type of market because they switch out of the asset after most of the loss has already happened, and they get you back into the asset after the sharp rally. I.e., they sell low and buy high.

In short, whether TF will outperform B&H in the future depends on the type of market we get from here.
Just my 2c ... People don't publish profitable strategies (whether they're exploiting real market anomalies, or getting lucky). A huge part of trend following is knowing when other traders are going to buy and sell – so your competition can be shaken out of trades, and you can buy back at better prices.

My experience as a trend follower is that 2009-10 onwards has been optimal. I've not had an underperforming year, and in 2020, there was an easy 140% to be made on the Cathy Wood/Chase Coleman type stocks, just following price trends that were in place at the start of the year. The problem with reductive academic models is they tend to look at assets in isolation. Very few articles talk about relative strength. But my experience is it's a lot better to exit trades for better trends, than play this optimal stopping distance game. Soros' Quantum fund operated this way – whenever they wanted to enter a new trade, they had to find something to cut. So there's a constant culling of things that look less certain than other things.

My message would be that trending is a very simple mathematical idea. None of the details or best ways to implement it are fixed (or exist as esoteric 'factors'). It's an engineering problem: take this very simple idea, that the price direction over one period has some statistical relationship to direction over a future period, and work out how to measure that. It's the measurement that generates alpha, as all alpha is is adding information to the market that correctly estimates something.
Beliavsky
Posts: 1071
Joined: Sun Jun 29, 2014 10:21 am

Re: Momentum factor and trend following

Post by Beliavsky »

nisiprius wrote: Tue Aug 02, 2022 6:02 pm Discussions of momentum are incomplete without discussions of "momentum crashes."
A current article by Larry Swedroe is Avoiding Momentum Crashes.
Post Reply