Gyroscopic Investing

An interesting article on momentum investing:

http://www.theglobeandmail.com/globe-in ... le2365067/

The six basic asset classes: three-month Canadian treasury bills, long-term Canadian bonds, the S&P/TSX composite, the S&P 500, international stocks (via the EAFE index), and gold, essentially mimic the PP (Canadian, US, and Int'l make up the stock portion).

Looking at the historical returns of the PP (https://web.archive.org/web/20160324133 ... l-returns/) I decided to test the theory (results below). Investing a theoretical $10,000 in gold in 1973 (since based on the 1972 gold would be the place to put your money).

Year  Allocation    Amount        Gain

1973 (100% gold): $17,560  75.6%
1974 (100% gold): $29,939  70.5%
1975 (100% gold): $23,143  -22.7%
1976 (100% TSM): $29,322  26.7%
1977 (100% TSM): $28,091  -4.2%
1978 (100% gold): $38,400  36.7%
1979 (100% gold): $90,740  136.3%
1980 (100% gold): $100,540  10.8%
1981 (100% TSM): $96,820  -3.7%
1982 (100% STB): $115,700  19.5%
1983 (100% LTB):  $116,510  0.7%
1984 (100% TSM): $121,753  4.5%
1985 (100% LTB): $159,497  31.0%
1986 (100% TSM): $185,176  16.1%
1987 (100% LTB): $179,806  -2.9%
1988 (100% gold): $151,576  -15.7%
1989 (100% TSM): $195,382  28.9%
1990 (100% TSM): $183,659  -6.0%
1991 (100% STB): $203,311    10.7%
1992 (100% TSM): $223,235    9.8%
1993 (100% TSM): $246,898    10.6%
1994 (100% gold): $241,466    -2.2%
1995 (100% TSM): $327,910  35.8%
1996 (100% TSM): $396,772  21.0%
1997 (100% TSM): $519,771  31.0%
1998 (100% TSM): $640,878  23.3%
1999 (100% TSM): $793,407  23.8%
2000 (100% TSM): $709,306  -10.6%
2001 (100% LTB):  $739,806  4.3%
2002 (100% STB): $798,990    8.0%
2003 (100% gold): $955,592    19.6%
2004 (100% TSM): $1,075,041  12.5%
2005 (100% TSM): $1,139,544  6.0%
2006 (100% gold): $1,401,639  23.0%
2007 (100% gold): $1,834,746  30.9%
2008 (100% gold): $1,924,639  4.9%

The results look pretty good to me. 8 years of negative return. 28 years of positive return. Definitely more volatility than the PP, but whereas the PP 10K grows to $317,220 from 1973-2008, this Momentum Strategy grows from 10k to $1,924,639, a whopping $1.6M more. I'm not quite savvy enough to figure out what the CAGR works out to be. 

This seems like a very attractive VP option to me.

Thoughts?

Of course the article says that long bonds, as we already know, was the best performing asset on 2011. This year, so far, they are taking a loss. This kind of result using momentum investing doesn't inspire much confidence. 

I can and do invest using a momentum strategy on a monthly or quarterly basis, but a year seems too long to be married to one asset class.  I would not feel good being completely unhedged for an entire year.

This would definitely need to be a long term strategy, but I find it interesting that it would actually avoid most of the big drops in any one asset class.

Based on this strategy, you would have no stock allocation in 1973, 1974, 2001, 2002, or 2008. So you would miss declines of -18.1%, -27.2%, -11.0%, -21.0%, and most of all -36.7%.

The declines you would have in the TSM would be in '77, '90, and '00, the worst one being -10.6%.

Whereas in gold you have the bad drop of -22.7% in '75 and -15.7% in '88, but other than that you miss declines of -32.8% ('81), -20.2% ('84), and -21.5% ('97).       

You only have you money in Short Term Bonds for 3 years but they all do reasonably well ~10%, and the same goes for LTB.

Between 1972-2008, 15 times momentum would have put you in the highest increasing asset, 3 times it would put you in the 2nd highest asset, 12 times it puts you in the 3rd highest, and only 6 times does it put you in the lowest asset.

A few interesting facts.
- Momentum (following PP) seems to put one in the best class (41.6%) or the third (36.1%). Being in the second (8%) or the worst class (16.6%) happens less frequently.
- Of the 12 times that momentum puts you in the 3rd class only twice does it put you in the negative ('81 at -3.7% and '94 at -2.2%)

So, while many people think that LTB are going to go down this year statistically it is still likely to have positive gains.

It seems difficult to argue with the success of momentum investing on a calendar basis based on backtesting the PP, but again it would have to be done with a 5-10+ year timeframe.

Another point to add (from a Canadian perspective):

Although the starting point doesn't go back very far and therefore it is less useful if you look at a Canadian PP using iShares; XIC for stocks, XSB for short bonds (corporate and gov't not perfect), XLB for long bonds (again corporate and gov't) and IGT for gold, you'd find the following:

              XIC      XSB      XLB      IGT          PP Avg    Momentum Avg.  Contrarian
2006      17.0%    3.8%    4.1%*  22.4%      11.8%          -                    -
2007      9.5%    3.8%      2.9%    30.9%      11.8%          30.9%              3.8%
2008      -33.0%  8.0%      2.1%    5.4%        -4.4%          5.4%                2.1%
2009      34.5%    4.3%    5.3%      23.5%      16.9%          4.3%              34.5%
2010      17.3%    3.2%    12.1%    27.9%      15.2%          17.3%              3.2%
2011      -8.9%    4.4%    17.7%    8.7%      5.5%            8.7%              4.4%

Results

So between 2007 and 2011 this is what would have happened in terms of average return

100% Stock - 3.9%
100% STB - 4.7%
100% LTB - 8.0%
100% gold - 19.3%
Permanent - 9%
Momentum - 13.3%
Contrarian - 9.6%

So unless you were going to stick with one asset allocation for 5 years I think we can toss out the results from those results (100% stock is the only likely one asset class strategy). That leaves us with the Permanent, Momentum, and Contrarian results. Similar to the original Globe and Mail article the contrarian seems like the least likely to perform well over the longer term. It's 9.6% average comes almost entirely from 2008. The Permanent Portfolio of course does well and dampens the 2008 losses, but Momentum comes out on top, outperforming the PP in 4 out of 5 years, and in overall average.

Again this is a small sample size, but it seems like a viable strategy to attempt in a VP. Perhaps as one gets more and more comfortable with it you can shift more from the PP to the Momentum VP. In 2012 an investor should be 100% in LTB. We can reassess at the end of the year how momentum would've worked in 2012.


*XLB didn't have annual data until 2007 so I used the Canadian long bond index return (which would have been higher than XLB)

At least in the U.S., the price of LTT was artificially determined by the Fed in 2011 (Operation Twist). I'm not sure you want to use that as a basis for momentum investing in 2012.

kjoh57 wrote: At least in the U.S., the price of LTT was artificially determined by the Fed in 2011 (Operation Twist). I'm not sure you want to use that as a basis for momentum investing in 2012.

But momentum is based purely on relative strength of price, not fundamentals.  There will always be and always has been manipulation by Central Banks across the globe effecting all asset classes. 

I have decided to hold on tight and not sell my LTT (in my PP) in this current market. That's enough risk for me. I won't be taking on any more risk by following a momentum strategy for this year. I'd be very pleased to find out that LTTs do well (not tank) this year.

I am a fan of momentum investing, but I could never allocate 100% of my assets to this style. I currently have about 50% in a standard HBPP and then allocate most of the other half to two different momentum strategies, both of which utilize a monthly and quarterly time frame however.  All of the momentum studies say that momentum persists not just on a yearly basis, but also on a monthly, quarterly and six month time frame. 

most of the other half to two different momentum strategies

Can you explain your strategies?

Ok, so for 50% of my assets I hold in a traditional HBPP, although I rebalance more frequently (compared to a 15/35 band strategy) because I'm not as concerned about capturing momentum in that portion.

With another 25%, I put into the top 2 of the PP classes (assuming they are ahead of SHY)... That works out to an additional 12.5% in each of the top 2 asset classes, in addition to my buy and hold HBPP.  I calculate this momentum quarterly.

With the last 25%, I put it directly into the leading asset class of a much wider selection of asset classes.  Such as individual country ETF's, a couple more commodity ETF's, REIT's, timber, EDV (zero coupon treasury ETF) and such.  This is calculated monthly.

My thought process is that half of my assets are in a buy low, sell high strategy (HBPP).  This will do the best when markets are somewhat rangebound or volatile.

The other half is based on momentum and should capture most of the big trends in various asset classes. 

Keep in mind this is all in a tax privileged account.  I think transaction costs are somewhat overrated as a reason to not actively trade like this.  If you increase your CAGR by 3-5+%, you're easily going to smash that assuming you have more than a few thousand to invest.  Of course you can negate some of that with firms that have free ETF or Mutual Fund trading such as Vanguard, TDameritrade or Scottrade.

If you did this in a taxable account, I'm not sure it would be worth is when taxes are taken into account.

My taxable account, I use purely for speculative purposes (gut feeling and so forth).  But that is a very small amount of money.

Clacy,

Thanks for your responses. It's interesting to hear from someone who actually implements the strategy. The appeal of the HBPP comes from the fact that I agree with the basic underlying logic of the portfolio, and secondly that it is lazy and easy. The article I cited in my original post piqued my interest if the same ease could be applied to a momentum strategy. I'm not sure that I would desire to pursue momentum strategies on a monthly, or quarterly basis, but I think on a yearly basis it could work for me. In Canada we are allowed to put $5000 a year into a Tax Free Saving Account vehicle that can be put into anything (stocks, bonds, savings accounts, etc..). Right now I'm putting the 5K all into a HBPP, but I think I might try half into a VP following the momentum strategy.

Another interest point you raised was that you take the two best performing asset classes. This was something I was going to look at and see how the results would shake out in that scenario. It seems evident to me that the reason the contrarian strategy did so poorly (in the original article) is that many years you'd be putting that into the cash allocation of an HBPP. So again, it's interesting that you do this as long as the two classes (between stocks, LTB, gold) are ahead of STB.

I'm going to keep researching this, but I think it's something I'm going to try in a VP. This year should be interesting to see how long term bonds do. If they do poorly I suspect people will say that that disproves momentum investing. Historically, however, the four times previously that momentum investing would place you in long term bonds it provided an average of 8% return, and only once was negative. 

1973-2011
        HB 4x25  long top asset    long top 2 assets    long best/short worst  long top 2/short bottom 2
cagr    10.52%        15.73%              14.15%                      5.76%                        4.05%
stdev  8.13%          28.35%              15.39%                    17.96%                        9.89%
% up  94.87%        76.92%              82.05%                    66.67%                      74.36%
             

sfreeman wrote:

1973-2011
         HB 4x25   long top asset    long top 2 assets     long best/short worst   long top 2/short bottom 2
cagr    10.52%         15.73%              14.15%                      5.76%                         4.05%
stdev   8.13%          28.35%              15.39%                     17.96%                        9.89%
% up   94.87%         76.92%              82.05%                     66.67%                       74.36%
             

This sort of coincides with my fairly remedial backtesting similar to monthly/quarterly in the fact that the to 2 tends to be a pretty good combination of higher returns, but stdev that you can live with.  It seems to smooth out the volatility, yet capture much of the momentum of the top performing assets.

It lost 22% in 2009

An interesting new Canadian ETF that brings a passive spin to momentum stock investing: http://www.xtf.ca/overview/?etf=WXM

Here is the index it is based upon: http://www.xtf.ca/documents/fundDoc_42.pdf?52844

This seems like an interesting product to try out in a VP, but keep in mind that it is a very new product.

Since its inception it is up 3% while the TSX is down -1%.

NewPPer wrote: Thoughts?

Its 16.21% CAGR.

Before you get too excited, you are ignoring the actual risk by only looking at calendar year closing values.  A momentum portfolio will be more risky than any other investment strategy; whether or not it is better on a risk-adjusted basis can only be known by comparing the actual risk/reward using the maximum drawdowns on a daily basis.

Here are ETF's using momentum (with poor track records): PDP, PIE, PIZ, MOM, SHMO, HMTM, XHMO, SHMO, EEH.

MG

Speaking of momentum investing, here is the relative strength of cash, bond, stock, gold.


relative strength as of Friday, 5/18/2012

TLT = 96
VTI = 70
SHY = 67
GLD = 39



Source
www.ETFscreen.com


Since TLT has the highest relative strength, I bought 100 shares of TMF for ~ $7500.

TMT = Direxion Daily 20+ Yr Trs Bull 3X Shares
which is 3 times TLT


Let's monitor this weekly, monthly, quarterly, etc.

Please do keep us posted!  This is an interesting strategy, and one that may be preferable to, say, buying individual stocks in a VP.  Reassessing monthly might be safer than quarterly, sticking with the past 12 months as the basis for ranking assets.  You'd also have to avoid the temptation to mess with the criteria, or else you're just plain market timing.

It is a bit counterintuitive that momentum can be so effective when at first glance it is the exact opposite of traditional fixed rebalance bands.  I've run simulations of different bands, though, and HB's recommended 15%/35% gets consistently better results than 20%/30%, regardless of the time period or even the funds chosen.  It's likely because you get to capture some momentum while keeping the portfolio from getting too far out of balance.  It makes intuitive sense that over the long term, markets have a degree of inertia, with sudden switches in direction being relatively rare.  For these reasons, I'd guess that momentum strategies with shorter time frames than 6 months or based on something other than an asset class index wouldn't be as effective, and I would never use it for the core money that I want to keep protected.

okay, sophie, I will try but no guarantee due to time constrain... thanks...

http://seekingalpha.com/article/602681- ... -yields-go

Long-Term Treasuries: How Low Can Yields Go?
May 20, 2012  | 6 comments by: Eric Parnell  |  includes: TLT


Long-Term U.S. Treasury markets are back in rally mode. With uncertainty running high given the deteriorating situation in Europe and a stock market that has been in sharp decline throughout the month of May, investors are clamoring for the relative safety of U.S. Treasury debt. And nowhere has the benefit been more pronounced than at the long end of the curve, which has rallied by +6% for the month to date. But with yields already near historical lows, this latest advance raises the following question: how much lower can Treasury yields go? Depending on how events play out in global markets in the coming months, potentially quite a bit.

sophie wrote: Please do keep us posted!  This is an interesting strategy, and one that may be preferable to, say, buying individual stocks in a VP.  Reassessing monthly might be safer than quarterly, sticking with the past 12 months as the basis for ranking assets.  You'd also have to avoid the temptation to mess with the criteria, or else you're just plain market timing.

It is a bit counterintuitive that momentum can be so effective when at first glance it is the exact opposite of traditional fixed rebalance bands.  I've run simulations of different bands, though, and HB's recommended 15%/35% gets consistently better results than 20%/30%, regardless of the time period or even the funds chosen.  It's likely because you get to capture some momentum while keeping the portfolio from getting too far out of balance.  It makes intuitive sense that over the long term, markets have a degree of inertia, with sudden switches in direction being relatively rare.  For these reasons, I'd guess that momentum strategies with shorter time frames than 6 months or based on something other than an asset class index wouldn't be as effective, and I would never use it for the core money that I want to keep protected.

I think it is really interesting how the PP has an internal momentum bias with the 15/35 bands. I like the bands because they involve less trading/taxes, but the momentum effect is very interesting and luckily it appears to positive (I am not counting on it though).

melveyr wrote: I think it is really interesting how the PP has an internal momentum bias with the 15/35 bands. I like the bands because they involve less trading/taxes, but the momentum effect is very interesting and luckily it appears to positive (I am not counting on it though).

I almost wish this wasn't the case.  The PP has turned me from someone who was constantly screwing around with his investments (in and out of various trades) into to someone who would prefer never to have to adjust anything...even to rebalance. 

Sound Money Nancyboy wrote: I almost wish this wasn't the case.  The PP has turned me from someone who was constantly screwing around with his investments (in and out of various trades) into to someone who would prefer never to have to adjust anything...even to rebalance. 

Funny, new name!

I wouldn't worry that much.  The abnormal (less market return) momentum effect is relatively small in the grand scheme of things.

What damages the PP (and anything else) is down trends.

MachineGhost wrote:
Funny, new name!

MT used it in a post I just read.  Made me laugh for about two minutes.  I'll eventually change it back, though. 

MachineGhost wrote: Funny, new name!

Ha!  Love AdamA's new "sound money nancyboy" moniker.

If you really want to earn the name, be like me!  Go into weeping, mascara-streaked hysterics every time Gumby threatens to mint a $50 quadrillion platinum coin in order to pay off the national debt!