Author Topic: Dual Momentum Investing  (Read 1943561 times)

milesdividendmd

  • Handlebar Stache
  • *****
  • Posts: 1913
  • Location: Portlandia
    • Miles Dividend MD
Re: Dual Momentum Investing
« Reply #100 on: April 18, 2015, 08:34:07 AM »
This has been looked at extensively by Antonacci, CXO advisory, gestaltU, to name a few.  Read up!

FIPurpose

  • Handlebar Stache
  • *****
  • Posts: 2059
  • Location: ME
    • FI With Purpose
Re: Dual Momentum Investing
« Reply #101 on: April 18, 2015, 09:50:10 AM »
Here is a paper that I found extremely enlightening.

http://dorseywrightmm.com/downloads/hrs_research/SSRN-id1585517.pdf

Runs a number of comparisons back to 1928 comparing different lookback periods and number of assets held at any one time.

hodedofome

  • Handlebar Stache
  • *****
  • Posts: 1463
  • Age: 44
  • Location: Texas
Re: Dual Momentum Investing
« Reply #102 on: April 18, 2015, 12:28:58 PM »

The lookback period is the heart of the model; it is the main variable that is used to optimize performance.  Too long, you get creamed; too short, you get whipsawed.  It is also the variable that is most vulnerable to over-fitting (i.e., constructing a model that predicts the past instead of the future).  Many posts have already alluded to the number one rule of modeling anything: construct your model using one set of data and validate it using another set of data.  So, the logical thing to do would be to divide the historical market data into two sets.  Use, say, 1900 - 1960 to optimize the lookback period, then use 1961 - present to test the model's performance.  Surely Antonacci (or someone) has already done this?  I'm too lazy to read all the background material - I'm hoping one of you already has and can enlighten me.
The look back period is much less important than the choice of asset classes. Asset choice is by far the most important.
« Last Edit: April 20, 2015, 10:37:29 AM by hodedofome »

mtnrider

  • Pencil Stache
  • ****
  • Posts: 518
  • Location: Frozen tundra in the Northeast
Re: Dual Momentum Investing
« Reply #103 on: April 19, 2015, 11:30:46 AM »

I just read up a little about it now.  Color me suspicious, especially since it's partially sold by the backtesting.  Of course it backtests well!  It grew out of the backtests themselves, if it didn't fit them, he would have changed the algorithm.  This seems like survivorship bias to me.  (That doesn't mean it won't continue to work for a while longer, of course.)

I also can't find anything about a monte carlo simulation, which would be much better than backtesting.

How could you perform a monte carlo? Don't you lose the whole momentum thing (which is critical to the strategy) by randomizing it like that?

That's the point.  It's really impossible to predict the future, but these backtests rely on something the authors saw by looking at the data in hindsight.  It's actually not hard to find patterns in the past.  You can mine PI for patterns.  You can read "The Bible Code" for people who've found patterns in the bible.  Etc...

To be clear though, momentum itself is a real thing.  It's small and unpredictable, but real.  The question is - can you really play it to your advantage.  There's tons of other strategies you can follow (eg Dogs of the Dow).  They all look good in backtesting, then they fail.  Then someone tweaks the backtests, they look good again, repeat.  I suspect this falls in that bucket.

Now... I will be impressed if they can keep their algorithm set in stone and report in 10 years that they made a better return than VTI.

That said - it's not a horrible strategy - you're not trading naked options or something where you can blow up.  Worst case, you backtested to a fund that drops in half in a month, or are invested in treasuries during a bull run.  (This is essentially a market timing strategy - and Bogle has laid out the pitfalls of that.)


As a big proviso: I haven't read the book.  I've only read blog posts like this.  Maybe if I read it, I'd be a convert.  :)

mtnrider

  • Pencil Stache
  • ****
  • Posts: 518
  • Location: Frozen tundra in the Northeast
Re: Dual Momentum Investing
« Reply #104 on: April 19, 2015, 11:53:05 AM »


I just read up a little about it now.  Color me suspicious, especially since it's partially sold by the backtesting.  Of course it backtests well!  It grew out of the backtests themselves, if it didn't fit them, he would have changed the algorithm.  This seems like survivorship bias to me.  (That doesn't mean it won't continue to work for a while longer, of course.)

I also can't find anything about a monte carlo simulation, which would be much better than backtesting.

Please justify your use of the term survivorship bias.



More specifically, see the entry on data mining bias here:

http://www.investopedia.com/exam-guide/cfa-level-1/quantitative-methods/sampling-bias.asp


Crushtheturtle

  • 5 O'Clock Shadow
  • *
  • Posts: 51
Re: Dual Momentum Investing
« Reply #105 on: April 19, 2015, 12:20:45 PM »


I just read up a little about it now.  Color me suspicious, especially since it's partially sold by the backtesting.  Of course it backtests well!  It grew out of the backtests themselves, if it didn't fit them, he would have changed the algorithm.  This seems like survivorship bias to me.  (That doesn't mean it won't continue to work for a while longer, of course.)

I also can't find anything about a monte carlo simulation, which would be much better than backtesting.

Please justify your use of the term survivorship bias.



More specifically, see the entry on data mining bias here:

http://www.investopedia.com/exam-guide/cfa-level-1/quantitative-methods/sampling-bias.asp

Is there an appreciable difference between the historical data used to support momentum and that used to support long term buy and hold?

With momentum, I don't see any overly complex "algorithm" being fitted to cherry picked backtests. The strategy consists entirely of concentrating into recently outperforming global asset classes. There's no guarantee such a ploy will generate results in the future, but betting on buy and hold ("stocks will generally do better than bonds in the long term") seems no less uncertain.

milesdividendmd

  • Handlebar Stache
  • *****
  • Posts: 1913
  • Location: Portlandia
    • Miles Dividend MD
Dual Momentum Investing
« Reply #106 on: April 19, 2015, 12:28:06 PM »


I just read up a little about it now.  Color me suspicious, especially since it's partially sold by the backtesting.  Of course it backtests well!  It grew out of the backtests themselves, if it didn't fit them, he would have changed the algorithm.  This seems like survivorship bias to me.  (That doesn't mean it won't continue to work for a while longer, of course.)

I also can't find anything about a monte carlo simulation, which would be much better than backtesting.

Please justify your use of the term survivorship bias.



More specifically, see the entry on data mining bias here:

http://www.investopedia.com/exam-guide/cfa-level-1/quantitative-methods/sampling-bias.asp

Oh so you were confusing survivorship bias with data mining.

One is relevant to the discussion and one is not, hence the confusion.
« Last Edit: April 19, 2015, 01:02:19 PM by milesdividendmd »

GGNoob

  • Pencil Stache
  • ****
  • Posts: 726
  • Age: 37
  • Location: Colorado
Re: Dual Momentum Investing
« Reply #107 on: April 19, 2015, 06:07:54 PM »
First time reading this thread and looking into Dual Momentum Investing. It's very intriguing. I have a few questions that pertain mainly to milesdividendmd and anyone else who is currently doing this. I would do a lot more reading before implementing something like this. But it could be fun to test out with my 457 and 401k since they both have the same funds available through the same company. One could be a 3-Fund account (100% stock), the other DMI.

1. How did you choose your loopback period?

From what I've read, 6 months sounds about right. I'd hate to be in a bear market too long and not get back into the bull soon enough if I chose 12 months, and 3 months seems like too little data to go with.

2. How did you choose your asset classes?

With my 457 and 401k, it would be easy as I'd probably just go with 4 funds...Capital Preservation/Fixed Income, US Large-Cap, US Mid/Small-Cap, International.

But in my IRA, I also invest in US REITs, Emerging Markets, and International Small-Cap. I'd be tempted to include those in the mix.

 


milesdividendmd

  • Handlebar Stache
  • *****
  • Posts: 1913
  • Location: Portlandia
    • Miles Dividend MD
Dual Momentum Investing
« Reply #108 on: April 19, 2015, 07:11:56 PM »
Due to the limitations of my 403B I toggle between an S&p fund VIIIX, an foreign developed fund, FSPNX, and a short term treasury fund.

This is almost identical to Antonacci's GEM, with short term treasuries in place of total bond.

If I had a low cost EM  fund I would include that too.

I chose 6 months for behavioral reasons, Ie not wanting to get out of a bear too slowly. The downside is more trades.

In my last blog posts I posit that it might be optimal to diversify Lookback periods. I.e. 1/3 of the portfolio to three months Lookback, 1/3 on 6 month, 1/3 12 month to mimimize the risk of period specific whipsaws.

The downside would be more trading and more complexity.
« Last Edit: April 19, 2015, 08:10:25 PM by milesdividendmd »

mtnrider

  • Pencil Stache
  • ****
  • Posts: 518
  • Location: Frozen tundra in the Northeast
Re: Dual Momentum Investing
« Reply #109 on: April 19, 2015, 08:46:57 PM »


I just read up a little about it now.  Color me suspicious, especially since it's partially sold by the backtesting.  Of course it backtests well!  It grew out of the backtests themselves, if it didn't fit them, he would have changed the algorithm.  This seems like survivorship bias to me.  (That doesn't mean it won't continue to work for a while longer, of course.)

I also can't find anything about a monte carlo simulation, which would be much better than backtesting.

Please justify your use of the term survivorship bias.



More specifically, see the entry on data mining bias here:

http://www.investopedia.com/exam-guide/cfa-level-1/quantitative-methods/sampling-bias.asp

Oh so you were confusing survivorship bias with data mining.

One is relevant to the discussion and one is not, hence the confusion.

Yeah.  My bad.  I tend to lump data mining bias in as a subset of survivorship bias (because one throws away all the strategies that fail), but I can see how not everyone would see it that way.


GGNoob

  • Pencil Stache
  • ****
  • Posts: 726
  • Age: 37
  • Location: Colorado
Re: Dual Momentum Investing
« Reply #110 on: April 19, 2015, 08:57:56 PM »
Due to the limitations of my 403B I toggle between an S&p fund VIIIX, an foreign developed fund, FSPNX, and a short term treasury fund.

This is almost identical to Antonacci's GEM, with short term treasuries in place of total bond.

If I had a low cost EM  fund I would include that too.

I chose 6 months for behavioral reasons, Ie not wanting to get out of a bear too slowly. The downside is more trades.

In my last blog posts I posit that it might be optimal to diversify Lookback periods. I.e. 1/3 of the portfolio to three months Lookback, 1/3 on 6 month, 1/3 12 month to mimimize the risk of period specific whipsaws.

The downside would be more trading and more complexity.

Interesting blog post. Using my Roth IRA and my 401k as an example...my 401k would be split between 2 funds (2/3 US Small/Mid Cap and 1/3 US Large Cap). My Roth IRA on the other hand would be split between 3 funds (1/3 International Small-Cap, 1/3 Extended Market, and 1/3 US REIT). Using an approach like that would make me feel safer since I'd be invested in more than 1 fund and I would think it could reduce volatility a bit.

Leisured

  • Pencil Stache
  • ****
  • Posts: 696
  • Age: 79
  • Location: South east Australia, in country
  • Retired, and loving it.
Re: Dual Momentum Investing
« Reply #111 on: April 20, 2015, 05:38:03 AM »
Dual Momentum Investing is an extraordinary method, but a Dual Momentum ETF is impractical. As has been pointed out, a DM ETF will hold just one or two ETFs, and when the time comes to sell the ETF, the entire holding has to be sold. This might be, say, $100 million over the space of a few days, thus flooding the market and forcing down the price. When the DM ETF buys one or two bond funds, it floods the market with buy orders, so forcing up the price.

I commend milesdividendmd and sirdoug for their links.

milesdividendmd

  • Handlebar Stache
  • *****
  • Posts: 1913
  • Location: Portlandia
    • Miles Dividend MD
Dual Momentum Investing
« Reply #112 on: April 20, 2015, 08:30:15 AM »
Due to the limitations of my 403B I toggle between an S&p fund VIIIX, an foreign developed fund, FSPNX, and a short term treasury fund.

This is almost identical to Antonacci's GEM, with short term treasuries in place of total bond.

If I had a low cost EM  fund I would include that too.

I chose 6 months for behavioral reasons, Ie not wanting to get out of a bear too slowly. The downside is more trades.

In my last blog posts I posit that it might be optimal to diversify Lookback periods. I.e. 1/3 of the portfolio to three months Lookback, 1/3 on 6 month, 1/3 12 month to mimimize the risk of period specific whipsaws.

The downside would be more trading and more complexity.

Interesting blog post. Using my Roth IRA and my 401k as an example...my 401k would be split between 2 funds (2/3 US Small/Mid Cap and 1/3 US Large Cap). My Roth IRA on the other hand would be split between 3 funds (1/3 International Small-Cap, 1/3 Extended Market, and 1/3 US REIT). Using an approach like that would make me feel safer since I'd be invested in more than 1 fund and I would think it could reduce volatility a bit.

You seem to be describing more of an absolute momentum strategy. Also an excellent strategy. For it to be dual momentum you must toggle between two imperfectly correlated assets with cash as your backstop.

Here is an excellent paper on absolute momentum to help you with your implementation.

http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2244633
« Last Edit: April 20, 2015, 10:13:07 AM by milesdividendmd »

GGNoob

  • Pencil Stache
  • ****
  • Posts: 726
  • Age: 37
  • Location: Colorado
Re: Dual Momentum Investing
« Reply #113 on: April 20, 2015, 08:54:31 AM »

Due to the limitations of my 403B I toggle between an S&p fund VIIIX, an foreign developed fund, FSPNX, and a short term treasury fund.

This is almost identical to Antonacci's GEM, with short term treasuries in place of total bond.

If I had a low cost EM  fund I would include that too.

I chose 6 months for behavioral reasons, Ie not wanting to get out of a bear too slowly. The downside is more trades.

In my last blog posts I posit that it might be optimal to diversify Lookback periods. I.e. 1/3 of the portfolio to three months Lookback, 1/3 on 6 month, 1/3 12 month to mimimize the risk of period specific whipsaws.

The downside would be more trading and more complexity.

Interesting blog post. Using my Roth IRA and my 401k as an example...my 401k would be split between 2 funds (2/3 US Small/Mid Cap and 1/3 US Large Cap). My Roth IRA on the other hand would be split between 3 funds (1/3 International Small-Cap, 1/3 Extended Market, and 1/3 US REIT). Using an approach like that would make me feel safer since I'd be invested in more than 1 fund and I would think it could reduce volatility a bit.

You seem to be describing more of an absolute momentum strategy. Also an excellent strategy. For it to be dual momentum you must toggle between two uncorrelated assets with cash as your backstop.

Here is an excellent paper on absolute momentum to help you with your implementation.

http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2244633

Still too new to this idea so I didn't know exactly how it worked! I actually bought the book and will begin reading it tonight. Been playing around with backtesting on PortfolioVisualizer.com and there's some pretty impressive results. Since I wasn't sure how it worked, I did some tests with lots of funds and some with just US stock, International Stock, and Total Bond. The simple ones seemed to give better results.

hodedofome

  • Handlebar Stache
  • *****
  • Posts: 1463
  • Age: 44
  • Location: Texas
Re: Dual Momentum Investing
« Reply #114 on: April 20, 2015, 01:21:53 PM »
Dual Momentum Investing is an extraordinary method, but a Dual Momentum ETF is impractical. As has been pointed out, a DM ETF will hold just one or two ETFs, and when the time comes to sell the ETF, the entire holding has to be sold. This might be, say, $100 million over the space of a few days, thus flooding the market and forcing down the price. When the DM ETF buys one or two bond funds, it floods the market with buy orders, so forcing up the price.

I commend milesdividendmd and sirdoug for their links.

This is why the very large managed futures funds (like Winton Capital and others with billions under management) are pretty much stuck with trading interest rates and currencies. They are literally the only markets that are liquid enough for them to get in and out of their positions within a few days.

hodedofome

  • Handlebar Stache
  • *****
  • Posts: 1463
  • Age: 44
  • Location: Texas
Re: Dual Momentum Investing
« Reply #115 on: April 20, 2015, 02:16:26 PM »
Just throwing up some momentum research here:

http://www.dualmomentum.net/2013/09/momentum-back-testing.html
http://www.aqrindex.com/AQR_Momentum_Indices/Momentum_Research/Content/default.fs
http://www.dualmomentum.net/2011/03/history-of-momentum-research.html
http://www.dualmomentum.net/2011/05/efficient-marketsnot.html
http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2435323
https://www.aqr.com/~/media/files/papers/aqr-a-century-of-trend-following-investing.pdf
https://drive.google.com/file/d/0BzyyTlvGE-T2TFdZSG1rVmZYLVE/view?usp=sharing
https://drive.google.com/file/d/0BzyyTlvGE-T2R2pjMWhSbjVSSGc/view?usp=sharing
https://drive.google.com/file/d/0BzyyTlvGE-T2RWNINEpzc25Ma1U/view?usp=sharing
https://drive.google.com/file/d/0BzyyTlvGE-T2MURxZGtqNnYxMDA/view?usp=sharing
https://drive.google.com/file/d/0BzyyTlvGE-T2RG5zYkstZURrdDA/view?usp=sharing
https://drive.google.com/file/d/0BzyyTlvGE-T2c1ZFQllrMl92eFU/view?usp=sharing
https://drive.google.com/file/d/0BzyyTlvGE-T2eVFVZEF2ZHdkNlE/view?usp=sharing
https://drive.google.com/file/d/0BzyyTlvGE-T2QW1jM2M1ejdLSkU/view?usp=sharing
https://drive.google.com/file/d/0BzyyTlvGE-T2MHlLbS1tbzNVVDA/view?usp=sharing
https://drive.google.com/file/d/0BzyyTlvGE-T2S19LSVF5UlUtS0k/view?usp=sharing
https://drive.google.com/file/d/0BzyyTlvGE-T2eEswVmdGcjE2Q28/view?usp=sharing
https://drive.google.com/file/d/0BzyyTlvGE-T2UXV6Q3ZrYlkzYjg/view?usp=sharing


As for the comments about the lookback periods, the previous momentum research going back 80 years has consistently looked at previous returns for the past 6-12 months. Those lookback periods still work today so that's a pretty decent amount of out of sample evidence. Find me another strategy with that much out of sample evidence.
« Last Edit: April 20, 2015, 02:19:22 PM by hodedofome »

smilla

  • Stubble
  • **
  • Posts: 145
  • Location: Canada
Re: Dual Momentum Investing
« Reply #116 on: April 20, 2015, 03:33:22 PM »
This is very interesting and I plan to try it in one very small (<5% of portfolio) account of mine.  I do have a question though.  It seems to me that choosing the top performer over a specified period shows which asset had relative momentum to that point, not which asset is currently enjoying momentum

Wouldn't a more accurate but still simple assessment be achieved by looking at the 12 mo, 6 mo & 3 month performance and choosing the investment that is increasing momentum at the greatest rate across the periods (or at least losing momentum the slowest)?   

I.e.     (periods ending Mar 31/15)
Stock symbol    12 mo    6 mo    3 mo
VUN  (US)         28.27    20.92    10.9
XEF  (dev)         16.31    18.24    14.53
XEC  (em)         14.85    10.56    11.79

At a glance this suggests that the US market is actually slowing down and that foreign markets have the momentum.  Even if you delete the 3 month column to lower risk and reduce trading, it seems reasonable to think that the comparison between 12 month and 6 month rates of growth would be a better way to capitalize on relative momentum than simply choosing the top 6 month performer.

Since it isn't done that way I expect I am missing something.  What makes this wrong or dangerous? 

(These are Canadian stock symbols but they are basically broad index ETFs for each market.)
« Last Edit: April 20, 2015, 03:38:43 PM by smilla »

milesdividendmd

  • Handlebar Stache
  • *****
  • Posts: 1913
  • Location: Portlandia
    • Miles Dividend MD
Re: Dual Momentum Investing
« Reply #117 on: April 20, 2015, 04:59:35 PM »
The question you're asking  is what's more predictive, the absolute return over the look back period or the acceleration of the price (dp/dt)? 

Your basis for this question seems to be gestalt feeling that  price acceleration jibes with your idea of "momentum" linguistically.

I don't know the answer to that question specifically, but I do know that simple price momentum (ie total returns over the Lookback period) is very predictive of future short term returns and it is simple to implement.

A literature search or your own backtest may answer your question better.

sirdoug007

  • Pencil Stache
  • ****
  • Posts: 585
  • Age: 43
  • Location: Houston, TX
Re: Dual Momentum Investing
« Reply #118 on: April 21, 2015, 10:03:38 AM »
I've been doing more and more reading and some more backtesting, including with the funds I have access to in my 401(k).

I was actually relieved to discover some Dual Momentum under-performance.  Otherwise this thing would be a free lunch and my econ professor drilled into me there ain't no such thing!

DM is really a medium to long term investing approach.  The key is that absolute momentum will occasionally pull you out of equities on loss of momentum just at the time equities push to new highs.  However, where DM really shines is over time periods that include a major drawdown which we really haven't had since 2009, summer 2011 wasn't bad and quickly bounced back.

This spreadsheet uses the mutual funds available to me in my T. Rowe Price 401(k) but you can pull your own data from the historical data on yahoo finance.

I compared the performance of DM starting in January of each year with the performance of a 100% S&P500 fund (close to what I am invested in now).  The results show significant outperformance for periods starting in 2005-2008.  However, after the great recession DM has underperformed the S&P500 (like pretty much everything else out there).  Note the outperformance of the longer periods in in the +70% to +140% range while underperformance is in the -0 to -32% range (i.e., a bit lopsided with advantage to long term outperformance).  You'll see the same thing in my earlier spreadsheet in the 1995-1999 period.

My conclusion is that DM can under-perform in bull markets as it moves between funds and occasionally jumps into bonds on small hiccups.  However, over long periods that include a bear market, DM does very, very well.

This makes some sense intuitively if you think of equities markets as cyclical with significant up and significant down periods.  I think the cyclical nature of equities markets has been pretty well established.  If you are in a bull/up market, the relative momentum may help a small amount but you can also underperform, especially when the S&P500 is the best performing equity asset sector.  It's when the bears come out of the woods that the absolute momentum part of DM saves your ass.  This is why you see great performance over 10+ years but so-so performance over 0-5 year periods.

« Last Edit: April 21, 2015, 11:17:05 AM by sirdoug007 »

sirdoug007

  • Pencil Stache
  • ****
  • Posts: 585
  • Age: 43
  • Location: Houston, TX
Re: Dual Momentum Investing
« Reply #119 on: April 21, 2015, 10:38:13 AM »
Here is the same analysis with the other spreadsheet that goes back to 1995.

The results since 2009 are barely perceptible while years that have endured the 2000s bust and 2008 recession have done incredibly well compared to the S&P500.  I had to double check but yes that is 26x the performance of the S&P500 since 1995.  However, since 2009 it's been less than 50% and only 2% over the last year.

This may be why Wall Street has not embraced this.  It doesn't really work on the time-frames they live in.  Monthy, quarterly and yearly it's not impressive.  Over 20 years with some big equity draw-downs it can really do some awesome stuff.

milesdividendmd

  • Handlebar Stache
  • *****
  • Posts: 1913
  • Location: Portlandia
    • Miles Dividend MD
Re: Dual Momentum Investing
« Reply #120 on: April 21, 2015, 02:01:15 PM »

I've been doing more and more reading and some more backtesting, including with the funds I have access to in my 401(k).

I was actually relieved to discover some Dual Momentum under-performance.  Otherwise this thing would be a free lunch and my econ professor drilled into me there ain't no such thing!

DM is really a medium to long term investing approach.  The key is that absolute momentum will occasionally pull you out of equities on loss of momentum just at the time equities push to new highs.  However, where DM really shines is over time periods that include a major drawdown which we really haven't had since 2009, summer 2011 wasn't bad and quickly bounced back.

This spreadsheet uses the mutual funds available to me in my T. Rowe Price 401(k) but you can pull your own data from the historical data on yahoo finance.

I compared the performance of DM starting in January of each year with the performance of a 100% S&P500 fund (close to what I am invested in now).  The results show significant outperformance for periods starting in 2005-2008.  However, after the great recession DM has underperformed the S&P500 (like pretty much everything else out there).  Note the outperformance of the longer periods in in the +70% to +140% range while underperformance is in the -0 to -32% range (i.e., a bit lopsided with advantage to long term outperformance).  You'll see the same thing in my earlier spreadsheet in the 1995-1999 period.

My conclusion is that DM can under-perform in bull markets as it moves between funds and occasionally jumps into bonds on small hiccups.  However, over long periods that include a bear market, DM does very, very well.

This makes some sense intuitively if you think of equities markets as cyclical with significant up and significant down periods.  I think the cyclical nature of equities markets has been pretty well established.  If you are in a bull/up market, the relative momentum may help a small amount but you can also underperform, especially when the S&P500 is the best performing equity asset sector.  It's when the bears come out of the woods that the absolute momentum part of DM saves your ass.  This is why you see great performance over 10+ years but so-so performance over 0-5 year periods.

I completely agree with your analysis.

The power of Dual momentum lies mostly in absolute momentum which allows you to limit drawdowns in all market conditions except for flash crashes. 

One point to make about dual momentum however, is that its ability to limit drawdowns allows you to take on more risk.

So comparing to a 100% S&P 500 fund might not be the best comparison on a personal level.

The ability to limit drawdowns allows you to up the risk of your holdings during bull markets.

As an example, prior to switching over my retirement accounts I was in a 75/25, stock bond passive global portfolio. I am nowholding 100% S&P 500.

Under most (almost all) market scenarios my drawdowns should be far less in bull markets then with my prior 75/25 portfolio, and there are no 10 year periods were dual momentum has not outperformed such a portfolio.

This ability to take on more risk significantly increases upside, without increasing downside. The one exception to this observation would be a flash crash such as what occurred on Black Monday 1987.

frugalnacho

  • Walrus Stache
  • *******
  • Posts: 5055
  • Age: 41
  • Location: Metro Detroit
Re: Dual Momentum Investing
« Reply #121 on: April 21, 2015, 02:32:09 PM »
I am still trying to wrap my head around this.  I understand how and why indexing works.  It's not a free lunch, you just cut out all the useless middlemen that add no value, and get average market returns.  Sprinkle in a dash of compounding interest, and bake for 10-15 years, and retire.

Maybe my perception of the market is wrong, but I see it as a positive sum game, where the positive sum is the total market capitalization.  For example is the total market is $1M, and then in the future it is $2M, then $1M in real (at least on paper) wealth has been created, but no more.  So buy and hold index investors would have realized average growth during that period, dual momentum investors would have realized above average growth, and because of the math some other group of rubes has achieved below average market returns for that period. 

So who is consistently making below average market returns? My first thought is that it is the unsophisticated investor that doesn't know anything.  They just toss money at whatever is hot, and panic and sell when it's not hot, they end up buying high and selling low.  They help amplify the momentum that dual momentum investors hope to capitalize on, but because they are unsophisticated they don't know when to get out and end up bearing the brunt of the losses when the tides change.

milesdividendmd

  • Handlebar Stache
  • *****
  • Posts: 1913
  • Location: Portlandia
    • Miles Dividend MD
Re: Dual Momentum Investing
« Reply #122 on: April 21, 2015, 02:53:12 PM »
Not hard to find losers in the stock market.  The answer to your question is: almost everyone including indexers.

milesdividendmd

  • Handlebar Stache
  • *****
  • Posts: 1913
  • Location: Portlandia
    • Miles Dividend MD
Re: Dual Momentum Investing
« Reply #123 on: April 21, 2015, 03:01:14 PM »
Not hard to find losers in the stock market.  The answer to your question is: almost everyone including indexers.

But to answer your question in a more genuine manner,  the power of trend following approaches is really limiting drawdowns.  Trendfollowing approaches generally provide reliable signals as to when to get the hell out of dodge, and when to get back in.

Dual momentum is just one such trendfollowing approach (thanks to absolute momentum) , with a little juice added to the upside courtesy of relative momentum.

And in my way of seeing the market, not losing is the name of the game and is far more powerful than outperforming to the upside.  I wrote a whole "cowards" investing series based on that premise, and dual momentum was merely the last chapter.

The first installment was here:

http://www.milesdividendmd.com/on-the-wisdom-of-cowardice/


frugalnacho

  • Walrus Stache
  • *******
  • Posts: 5055
  • Age: 41
  • Location: Metro Detroit
Re: Dual Momentum Investing
« Reply #124 on: April 21, 2015, 03:03:42 PM »
Not hard to find losers in the stock market.  The answer to your question is: almost everyone including indexers.

I'm not really satisfied with that answer.  Of course there are losers/winners depending on which sector you went into.  Also i'm sure part of that can be explained by fees (which I already addressed).  Pay someone 2% to invest in index funds for you, and it should be no surprise that you consistently under perform the market by 2% each year. 

Chuck

  • Bristles
  • ***
  • Posts: 407
  • Age: 35
  • Location: Northern VA
Re: Dual Momentum Investing
« Reply #125 on: April 21, 2015, 03:24:41 PM »
I have been reading furiously about this for hours now. It appears to work. It appears to work in multiple time frames.

I just can't shake this question: Why isn't everyone doing this? Why is such a simple method not in widespread application? Indexing has an entire religion devoted to it for fucks sake, why doesn't this? I can't shake the feeling that all I'm looking at is the best, most delicious bait... on a nasty hook.
« Last Edit: April 21, 2015, 03:27:48 PM by Chuck »

sirdoug007

  • Pencil Stache
  • ****
  • Posts: 585
  • Age: 43
  • Location: Houston, TX
Re: Dual Momentum Investing
« Reply #126 on: April 21, 2015, 03:56:48 PM »
Maybe my perception of the market is wrong, but I see it as a positive sum game, where the positive sum is the total market capitalization.  For example is the total market is $1M, and then in the future it is $2M, then $1M in real (at least on paper) wealth has been created, but no more.  So buy and hold index investors would have realized average growth during that period, dual momentum investors would have realized above average growth, and because of the math some other group of rubes has achieved below average market returns for that period. 

For each transaction there is a buyer and a seller.  So if you are correctly timing the market, you are selling high to someone who may get crushed in the near future and then buying low from someone who may have just been crushed and is missing out on the subsequent gains.  Because of the emotions involved with money and investing people do this all the time.  Implementing DM may be stomach churning in real life because it doesn't feel good to go against the good times and then buy back in when it seems like the world is falling apart.

DM is a rule based approach that takes the emotions of buying and selling out of it and gives a clear signal when the ride to the top of the roller coaster is over and it's time to get off.

I'm just as amazed by everyone else at how well this backtests.  I even did a backtest with Robert Shiller's S&P500 dataset back to 1871 with getting out to cash to test absolute momentum and it does amazingly well.

This only works because of human's herding tendencies which result in the cyclical nature of markets.  It seems to only be useful over 10+ year timeframes which may be why traders haven't adopted it and it isn't more well known.

milesdividendmd

  • Handlebar Stache
  • *****
  • Posts: 1913
  • Location: Portlandia
    • Miles Dividend MD
Re: Dual Momentum Investing
« Reply #127 on: April 21, 2015, 04:45:27 PM »
Maybe my perception of the market is wrong, but I see it as a positive sum game, where the positive sum is the total market capitalization.  For example is the total market is $1M, and then in the future it is $2M, then $1M in real (at least on paper) wealth has been created, but no more.  So buy and hold index investors would have realized average growth during that period, dual momentum investors would have realized above average growth, and because of the math some other group of rubes has achieved below average market returns for that period. 

For each transaction there is a buyer and a seller.  So if you are correctly timing the market, you are selling high to someone who may get crushed in the near future and then buying low from someone who may have just been crushed and is missing out on the subsequent gains.  Because of the emotions involved with money and investing people do this all the time.  Implementing DM may be stomach churning in real life because it doesn't feel good to go against the good times and then buy back in when it seems like the world is falling apart.

DM is a rule based approach that takes the emotions of buying and selling out of it and gives a clear signal when the ride to the top of the roller coaster is over and it's time to get off.

I'm just as amazed by everyone else at how well this backtests.  I even did a backtest with Robert Shiller's S&P500 dataset back to 1871 with getting out to cash to test absolute momentum and it does amazingly well.

This only works because of human's herding tendencies which result in the cyclical nature of markets.  It seems to only be useful over 10+ year timeframes which may be why traders haven't adopted it and it isn't more well known.

Actually, In practice I think that Dual momentum is behaviorally much easier than buy and hold since it merely asks me to follow my own base instincts and chase performance.  No need to buy recent losers or sell recent winners as in rebalancing a buy and hold.

I buy assets that are doing well, and sell ones that are doing less well.  I will re-enter the market after the recovery has started and exit when the sh*t is hitting the fan.  (But I haven't employed this technique through a bear market yet, of course.)

The main discomfort is in tracking error and the increased volatility that comes with holding only one asset at a time as opposed to holding a diversified portfolio (which smoothes the ride considerably.)

frugalnacho

  • Walrus Stache
  • *******
  • Posts: 5055
  • Age: 41
  • Location: Metro Detroit
Re: Dual Momentum Investing
« Reply #128 on: April 21, 2015, 06:36:54 PM »
I have been reading furiously about this for hours now. It appears to work. It appears to work in multiple time frames.

I just can't shake this question: Why isn't everyone doing this? Why is such a simple method not in widespread application? Indexing has an entire religion devoted to it for fucks sake, why doesn't this? I can't shake the feeling that all I'm looking at is the best, most delicious bait... on a nasty hook.

Yeah pretty much what i've been thinking. But to add to that, if it's such a simple method and it's so easy to tell when the roller coaster ride is over...why does every book i've read state emphatically that market timing is impossible?  It seems it's ridiculously easy using this method.  Why has every author dismissed this approach that has been known about since before I was even born?

milesdividendmd

  • Handlebar Stache
  • *****
  • Posts: 1913
  • Location: Portlandia
    • Miles Dividend MD
Re: Dual Momentum Investing
« Reply #129 on: April 21, 2015, 06:42:33 PM »
Dual Momentum was only recently described.  You may be confusing it with relative price momentum which was first described in academia in the 80s.

You could make a similar argument about using the 200 day moving average approach, which has also outperformed since first described.

Trend following works at decreasing drawdowns.  It empirically does.

arebelspy

  • Administrator
  • Senior Mustachian
  • *****
  • Posts: 28444
  • Age: -997
  • Location: Seattle, WA
Re: Dual Momentum Investing
« Reply #130 on: April 21, 2015, 07:04:02 PM »
Dual Momentum was only recently described.

How do we know it will continue to work?

Trend following works at decreasing drawdowns.  It empirically does.

A posteriori, apparently.  But a priori?

My question is: it has worked.  Apparently it's new though.  Logically, what is the reason it works, and will continue to work?  All market timing back tested works empirically.  Until it doesn't.  So I want something that rationally works (like index funds).  Does this?
I am a former teacher who accumulated a bunch of real estate, retired at 29, spent some time traveling the world full time and am now settled with three kids.
If you want to know more about me, this Business Insider profile tells the story pretty well.
I (rarely) blog at AdventuringAlong.com. Check out the Now page to see what I'm up to currently.

hodedofome

  • Handlebar Stache
  • *****
  • Posts: 1463
  • Age: 44
  • Location: Texas
Re: Dual Momentum Investing
« Reply #131 on: April 21, 2015, 07:35:20 PM »

I have been reading furiously about this for hours now. It appears to work. It appears to work in multiple time frames.

I just can't shake this question: Why isn't everyone doing this? Why is such a simple method not in widespread application? Indexing has an entire religion devoted to it for fucks sake, why doesn't this? I can't shake the feeling that all I'm looking at is the best, most delicious bait... on a nasty hook.

Yeah pretty much what i've been thinking. But to add to that, if it's such a simple method and it's so easy to tell when the roller coaster ride is over...why does every book i've read state emphatically that market timing is impossible?  It seems it's ridiculously easy using this method.  Why has every author dismissed this approach that has been known about since before I was even born?

Because you've only been reading the authors that don't believe in market timing, and missing the ones that do.


Sent from my iPhone using Tapatalk

frugalnacho

  • Walrus Stache
  • *******
  • Posts: 5055
  • Age: 41
  • Location: Metro Detroit
Re: Dual Momentum Investing
« Reply #132 on: April 21, 2015, 07:38:45 PM »

I have been reading furiously about this for hours now. It appears to work. It appears to work in multiple time frames.

I just can't shake this question: Why isn't everyone doing this? Why is such a simple method not in widespread application? Indexing has an entire religion devoted to it for fucks sake, why doesn't this? I can't shake the feeling that all I'm looking at is the best, most delicious bait... on a nasty hook.

Yeah pretty much what i've been thinking. But to add to that, if it's such a simple method and it's so easy to tell when the roller coaster ride is over...why does every book i've read state emphatically that market timing is impossible?  It seems it's ridiculously easy using this method.  Why has every author dismissed this approach that has been known about since before I was even born?

Because you've only been reading the authors that don't believe in market timing, and missing the ones that do.


Sent from my iPhone using Tapatalk

My follow up is: Why do those authors not believe in market timing if it apparently works so well?  How did they all miss the memo and not discover it during their own research?  And why does everyone recommend them if they are all so wrong?  It just doesn't add up to me.

frugalnacho

  • Walrus Stache
  • *******
  • Posts: 5055
  • Age: 41
  • Location: Metro Detroit
Re: Dual Momentum Investing
« Reply #133 on: April 21, 2015, 07:50:14 PM »
Dual Momentum was only recently described.  You may be confusing it with relative price momentum which was first described in academia in the 80s.

You could make a similar argument about using the 200 day moving average approach, which has also outperformed since first described.

Trend following works at decreasing drawdowns.  It empirically does.

This paper seems to describe momentum and was published in 1937:

http://www.e-m-h.org/CoJo37.pdf

So it was pointed out and published in a paper nearly 100 years ago and has been completely ignored since?

milesdividendmd

  • Handlebar Stache
  • *****
  • Posts: 1913
  • Location: Portlandia
    • Miles Dividend MD
Re: Dual Momentum Investing
« Reply #134 on: April 21, 2015, 08:39:16 PM »

Dual Momentum was only recently described.  You may be confusing it with relative price momentum which was first described in academia in the 80s.

You could make a similar argument about using the 200 day moving average approach, which has also outperformed since first described.

Trend following works at decreasing drawdowns.  It empirically does.

This paper seems to describe momentum and was published in 1937:

http://www.e-m-h.org/CoJo37.pdf

So it was pointed out and published in a paper nearly 100 years ago and has been completely ignored since?

It hasn't been. Momentum is used by almost every trader out there, consciously or subconsciously.

The better question is: if it has persisted for 85 years, why do you expect it to disappear? 

arebelspy

  • Administrator
  • Senior Mustachian
  • *****
  • Posts: 28444
  • Age: -997
  • Location: Seattle, WA
Re: Dual Momentum Investing
« Reply #135 on: April 21, 2015, 08:41:39 PM »
It hasn't been. Momentum is used by almost every trader out there, consciously or subconsciously.

The better question is: if it has persisted for 85 years, why do you expect it to disappear?

I am a former teacher who accumulated a bunch of real estate, retired at 29, spent some time traveling the world full time and am now settled with three kids.
If you want to know more about me, this Business Insider profile tells the story pretty well.
I (rarely) blog at AdventuringAlong.com. Check out the Now page to see what I'm up to currently.

milesdividendmd

  • Handlebar Stache
  • *****
  • Posts: 1913
  • Location: Portlandia
    • Miles Dividend MD
Dual Momentum Investing
« Reply #136 on: April 21, 2015, 08:45:00 PM »
Dual Momentum was only recently described.

How do we know it will continue to work?

Trend following works at decreasing drawdowns.  It empirically does.

A posteriori, apparently.  But a priori?

My question is: it has worked.  Apparently it's new though.  Logically, what is the reason it works, and will continue to work?  All market timing back tested works empirically.  Until it doesn't.  So I want something that rationally works (like index funds).  Does this?

A priori.

Trendfollowing was described in the 1930s at the latest, and it still has the same effect 80 years later:  decreased drawdowns.

Momentum was described as the 4th factor in the efficient market model (after beta, size, and value) in the 1980s, and its effect persists undiluted.

Dual momentum simply combines the 2 ; trend following (absolute momentum) and price momentum (relative momentum). It's future success rests solely on the persistence of those 2 observations.

Which do you doubt?
« Last Edit: April 21, 2015, 08:47:03 PM by milesdividendmd »

arebelspy

  • Administrator
  • Senior Mustachian
  • *****
  • Posts: 28444
  • Age: -997
  • Location: Seattle, WA
Re: Dual Momentum Investing
« Reply #137 on: April 21, 2015, 08:47:25 PM »
Something being described and observed doesn't provide logical reasoning for why it works, or if it will continue to.

I don't understand why you're citing dates for an argument of reason.

https://en.wikipedia.org/wiki/A_priori_and_a_posteriori
Quote
A priori knowledge or justification is independent of experience

I am a former teacher who accumulated a bunch of real estate, retired at 29, spent some time traveling the world full time and am now settled with three kids.
If you want to know more about me, this Business Insider profile tells the story pretty well.
I (rarely) blog at AdventuringAlong.com. Check out the Now page to see what I'm up to currently.

Chuck

  • Bristles
  • ***
  • Posts: 407
  • Age: 35
  • Location: Northern VA
Re: Dual Momentum Investing
« Reply #138 on: April 21, 2015, 08:51:22 PM »
Why do we have faith that the market will go up (the underlying assumption underpinning buy and hold index investing)?

Because it has in the past? I think your questions are fair arebelspy, but I also think they could be asked of bogglehead investing as well. Ultimately, all we have to presume future returns is prior performance.

arebelspy

  • Administrator
  • Senior Mustachian
  • *****
  • Posts: 28444
  • Age: -997
  • Location: Seattle, WA
Re: Dual Momentum Investing
« Reply #139 on: April 21, 2015, 08:53:46 PM »
Why do we have faith that the market will go up (the underlying assumption underpinning buy and hold index investing)?

Because it has in the past? I think your questions are fair arebelspy, but I also think they could be asked of bogglehead investing as well. Ultimately, all we have to presume future returns is prior performance.

No, I think there are other valid, logical reasons as to why it will go up long term that have nothing to do with "because it has in the past."  If that's your only reason for thinking it will, I'd suggest reading a lot more before investing.  :)
I am a former teacher who accumulated a bunch of real estate, retired at 29, spent some time traveling the world full time and am now settled with three kids.
If you want to know more about me, this Business Insider profile tells the story pretty well.
I (rarely) blog at AdventuringAlong.com. Check out the Now page to see what I'm up to currently.

Chuck

  • Bristles
  • ***
  • Posts: 407
  • Age: 35
  • Location: Northern VA
Re: Dual Momentum Investing
« Reply #140 on: April 21, 2015, 08:59:01 PM »
Why do we have faith that the market will go up (the underlying assumption underpinning buy and hold index investing)?

Because it has in the past? I think your questions are fair arebelspy, but I also think they could be asked of bogglehead investing as well. Ultimately, all we have to presume future returns is prior performance.

No, I think there are other valid, logical reasons as to why it will go up long term that have nothing to do with "because it has in the past."  If that's your only reason for thinking it will, I'd suggest reading a lot more before investing.  :)
I think the primary reason you are referring to is the ever expanding US economy. What guarantee do we have this trend of perpetual expansion will continue? It didn't work that way for Japan, and others.

The logic behind momentum, from what I have read over the past six or so hours, is that humans are fearful and irrational. They buy high and sell low. So long as this remains true, I think momentum investing as described by MDMD has a solid chance of outperforming the market, and it certainly has up until this point.

Like I said, I understand your questions, but if you aren't satisified with the answers I don't understand why you are satisified with the leap of faith involved in index investing either...

hodedofome

  • Handlebar Stache
  • *****
  • Posts: 1463
  • Age: 44
  • Location: Texas
Dual Momentum Investing
« Reply #141 on: April 21, 2015, 09:20:28 PM »

My follow up is: Why do those authors not believe in market timing if it apparently works so well?  How did they all miss the memo and not discover it during their own research?  And why does everyone recommend them if they are all so wrong?  It just doesn't add up to me.

Ben Graham didn't believe in it And he's pretty much responsible for most modern investors' beliefs. So everyone totes the party line without realizing its original origins and whether or not it's even true.

Rice traders were using Ichimoku trend following several hundred years ago. Maybe even 1000 years ago I can't remember how long. Successful speculators in the 1700 and 1800s mentioned 'hang onto your winners and cut your losers short.' This is trend following and momentum at its core.

A few here have wondered whether momentum will continue in the future. I personally believe that it will continue as long as irrational humans run the markets. Momentum and value are both strategies that prey on human cognitive biases and so I don't see any reason why that would go away. However, here's a market structure explanation of the source of trend following returns (at least in the futures markets):

http://www.michaelcovel.com/2013/01/18/eric-crittenden-on-the-podcast-now/

Start listening at 23:00 into the podcast. According to Eric, as long as we have commercial hedgers we'll have sustained trends. It may or may not be true but it's something to think about.


Sent from my iPhone using Tapatalk
« Last Edit: April 21, 2015, 09:41:19 PM by hodedofome »

dragoncar

  • Walrus Stache
  • *******
  • Posts: 9918
  • Registered member
Re: Dual Momentum Investing
« Reply #142 on: April 21, 2015, 09:23:56 PM »
Double momentum is fine for consumer suckAs but triple moment investing is streets ahead.

milesdividendmd

  • Handlebar Stache
  • *****
  • Posts: 1913
  • Location: Portlandia
    • Miles Dividend MD
Re: Dual Momentum Investing
« Reply #143 on: April 21, 2015, 09:35:49 PM »

Something being described and observed doesn't provide logical reasoning for why it works, or if it will continue to.

I don't understand why you're citing dates for an argument of reason.

https://en.wikipedia.org/wiki/A_priori_and_a_posteriori
Quote
A priori knowledge or justification is independent of experience

I must've misunderstood your point ARS.

My point was that if the effect described was based on data mining (what I thought you meant by a posteriori knowledge), then why has it persisted long after its initial description?

If a posteriori knowledge is merely that which based on experience, (ie it is empirical) then that is the best sort of knowledge on which to make decisions about the future IMO.




brooklynguy

  • Handlebar Stache
  • *****
  • Posts: 2204
  • Age: 43
Re: Dual Momentum Investing
« Reply #144 on: April 21, 2015, 10:17:58 PM »
If a posteriori knowledge is merely that which based on experience, (ie it is empirical) then that is the best sort of knowledge on which to make decisions about the future IMO.

No, a priori knowledge is the best sort of knowledge on which to make decisions about the future.  But since there are no purely a priori truths about the market (in the sense that it is impossible for them to be wrong, because they are inherently true), the best sort of knowledge on which to make decisions about the future is knowledge that both has an underlying logical explanation and can be demonstrated empirically, which I think is what arebelspy is getting at.  As he pointed out earlier, if an investing strategy based on making trades every other Tuesday in months starting with the letter J or M, except in years ending in double-digit primary numbers, happened to backtest perfectly, would you trust it to continue to work into the future?

The logic behind momentum, from what I have read over the past six or so hours, is that humans are fearful and irrational. They buy high and sell low.

Miles, do you agree with this?  Is there any other logic behind this strategy?  If not, it's interesting to note that the logic behind this strategy is in some sense fundamentally pessimistic (it relies on an assumption of human irrationality/stupidity) while the logic behind buy and hold indexing is fundamentally optimistic (it relies on an assumption of human productivity/ingenuity).

milesdividendmd

  • Handlebar Stache
  • *****
  • Posts: 1913
  • Location: Portlandia
    • Miles Dividend MD
Dual Momentum Investing
« Reply #145 on: April 21, 2015, 11:41:18 PM »
If a posteriori knowledge is merely that which based on experience, (ie it is empirical) then that is the best sort of knowledge on which to make decisions about the future IMO.

No, a priori knowledge is the best sort of knowledge on which to make decisions about the future.  But since there are no purely a priori truths about the market (in the sense that it is impossible for them to be wrong, because they are inherently true), the best sort of knowledge on which to make decisions about the future is knowledge that both has an underlying logical explanation and can be demonstrated empirically, which I think is what arebelspy is getting at.  As he pointed out earlier, if an investing strategy based on making trades every other Tuesday in months starting with the letter J or M, except in years ending in double-digit primary numbers, happened to backtest perfectly, would you trust it to continue to work into the future?

The logic behind momentum, from what I have read over the past six or so hours, is that humans are fearful and irrational. They buy high and sell low.

Miles, do you agree with this?  Is there any other logic behind this strategy?  If not, it's interesting to note that the logic behind this strategy is in some sense fundamentally pessimistic (it relies on an assumption of human irrationality/stupidity) while the logic behind buy and hold indexing is fundamentally optimistic (it relies on an assumption of human productivity/ingenuity).

There are multiple explanations for momentum. Some efficient market types even try to explain it as a risk story which has never made a bit of sense to me and just ends up seeming Panglossian.

I don't think that momentum has to do with human irrationality or stupidity, in a pejorative way. I think it has to do with the heuristics that govern human decision-making.  i'm talking recency, loss aversion, representativeness, etc. i.e. everything in Daniel Kahneman's thinking fast and slow.  These are useful techniques that we all use to process information quickly, but they are nonstatistical ways of thinking.

There is also some good data that momentum is perpetuated by the flow of capital in and out of funds. So as one strategy becomes successful based on regime change, money flows toward it in the short-term from other funds creating price momentum (and negative momentum from the donor funds!). And because money cannot flow instantaneously because of liquidity issues, this momentum lasts for a significant amount of time (anywhere from 3 to 12 months).

To me momentum is an expression of humanity as it exists, not as we think it exists.

We all believe ourselves to be rational, but we have irrational reactions to loss, and to prospective gain. We all chase performance (which is probably why there is such a strong predilection here for passive low cost investment. It is the SmartMoney bet!)

But to me passive investment has always been about one thing and one thing only. Low costs.

Passive investment wins because it is cheaper. Period. Full stop.

That's it. All of the arguments about winners and losers, zero sum games, and market efficiency strike me as post Facto rationalization. What matters to me is getting exposure to the market at the lowest price possible.

And one of the best things about the dual momentum strategy is that it trades so incredibly infrequently and can be implemented with only low-cost Cap weighted index funds.

To be sure there is some friction in the form
of bid ask spreads, so it is a little bit more expensive than buy-and-hold. But you trade about as often as you rebalance in a yearly rebalance.

But if you believe that the momentum anomaly is real, and is likely to persist, then it is easy to get exposure to it for cheaper than it is to get exposure to other factors like size, value, quality etc.

So to me the momentum story could not be more persuasive. But admittedly that's probably because its The fastest horse out there as far as I can see right now.
« Last Edit: April 21, 2015, 11:44:18 PM by milesdividendmd »

milesdividendmd

  • Handlebar Stache
  • *****
  • Posts: 1913
  • Location: Portlandia
    • Miles Dividend MD
Dual Momentum Investing
« Reply #146 on: April 21, 2015, 11:57:25 PM »
All of which relates to relative momentum. By far the less important half of dual momentum.

The real bang for the buck comes from sitting out of bear markets based on absolute momentum which is just another form of trend following not to dissimilar from moving average rules.

Trend following does not always beat buy-and-hold, but it almost always limits draw downs. (the only exception I can think of being flash crashes.)

So if you are considering the strategy or are interested in it then that is really the most important question to answer for yourself. Does it make sense to you that absolute momentum (or 200 day moving average rules) in other words trend following, will reproducibly truncate drawdowns in the future?

That is the single most important a priori question that you must answer.

My answer is (obviously) "almost always yes."

michaelrecycles

  • Stubble
  • **
  • Posts: 187
  • Location: Monterey Bay
Re: Dual Momentum Investing
« Reply #147 on: April 22, 2015, 12:40:24 AM »
Very interesting discussion. Following.

I understand you can backtest, but before going whole hog, I'd probably conduct a real test with a two-horse race between DM and buy-and-hold over a number of years. Then I'd trade a winning horse for a whole hog. I still wouldn't have enough animals for a farm, though.

halfshellmeijin

  • 5 O'Clock Shadow
  • *
  • Posts: 35
Re: Dual Momentum Investing
« Reply #148 on: April 22, 2015, 07:21:38 AM »
So here is another way of thinking about this. If you use the Dual Momentum strategy but instead of back testing 6-12 months, you back test as far as you can go. You reach the result that the US stock market has produced the best returns over the period of time and should be invested in 100% in US stocks. This is the argument that MMM puts forth in his blog. So the question is do you think back testing for the whole historical period is a better choice then back testing 6-12 months? Back testing the whole historical period gives more data to work with, but could also be argued that much of the data is outdated. Back testing such a short period of time reduces the sample size but makes the data more relevant.

frugalnacho

  • Walrus Stache
  • *******
  • Posts: 5055
  • Age: 41
  • Location: Metro Detroit
Re: Dual Momentum Investing
« Reply #149 on: April 22, 2015, 07:46:49 AM »
Very interesting discussion. Following.

I understand you can backtest, but before going whole hog, I'd probably conduct a real test with a two-horse race between DM and buy-and-hold over a number of years. Then I'd trade a winning horse for a whole hog. I still wouldn't have enough animals for a farm, though.

The issue with that is that it might take a long time before you are able to confidently reach a conclusion on which is the better strategy.  As in it might take my entire working career before one strategy shows clear superiority.  What I need is a strategy I can implement right now, that not only back tests well, but has sufficient evidence to lead me to believe it will continue to perform.  Passive investing has that evidence.  Not only that, it makes perfectly logical sense to me - cut out middlemen and any unnecessary costs that don't add value, and minimize the costs that are absolutely necessary - that means more money goes into my account and I can calculate the effect of compounding that excess money.  It's not an anomaly, it's just costs reduction and math, so I have every reason to believe that advantage will persist.

Dual momentum however, doesn't make logical sense to me.  I don't understand how it works.  It seems to decouple risk from reward.  You bear essentially no risk when the market is on the way up yet you reap all the rewards, and when the market is going to go down you get out of the market so you essentially bear no risk.  All the reward (more than your fair share), and none of the risk.  A giant free lunch.  The phenomena has apparently existed for hundreds of years and pre dates the USA according to some, and yet still persists as an easily exploitable anomaly. 

Why don't you just take the bullet proof dual momentum strategy, and implement it one day earlier?  Just get onto that roller coaster one day earlier, and get off one day earlier.  Wouldn't that give you slightly more gains, and slightly less losses?

I also don't believe that ben graham just decided timing the market is impossible, presented no data, and then everyone just parroted him ad infinitum.  I've heard the argument from more than just ben, and i've seen it supported with evidence and data, over and over.  Whenever I have heard claims of being able to time the market and been presented with evidence of how it can be done, when I dig into the data I find it's either false or can be explained by some other logical fallacy, 100% of the time.  I've yet to see any strategy (excluding dual momentum - still undecided as of yet) that cuts the mustard.  Furthermore I fully expect any strategy that actually does work will be exploited to the point it is no longer possible.  The idea that an easily exploitable anomaly exists, is widely known, and yet remains unimplemented is troubling to me.