Gyroscopic Investing

just this week i changed my strategy as we retire in 5 weeks.

for decades i followed the fidelity insight newsletter . i used the growth model .  100k in 1987 is 2.1 million today.

about 7 years ago i moved from the growth model to a 50/50 split between the income model and the growth and income model as i toned things down  for retirement.

well i wasn't happy with the bond heaviness of the income model recently so i  switched this week to a 50/50 split between the growth and income model and went back to my roots of 35 years ago and dumped 7  figures into TLT ,VTI ,SAV , AND GLD .

i used to follow the PP decades ago when i first started out.  now that i am retiring i am back

mathjak107 wrote: i used to follow the PP decades ago when i first started out.  now that i am retiring i am back

Did you do better or worse than the PP in the interim?

a lot better

mathjak107 wrote: a lot better

Well that makes sense. 1987-2007 was awesome for stock-heavy portfolios. Really great timing.

MediumTex wrote: It's weird to hear people talk about stocks being overvalued when the major indices are just now breaking through levels first reached 15 years ago.

They were overvalued 15 years ago until this new high in overvaluation eclipsed the previous high set in a new overvaluation 15 years ago.

It's all the Baby Boomers fault!

Pointedstick wrote:

mathjak107 wrote: a lot better

Well that makes sense. 1987-2007 was awesome for stock-heavy portfolios. Really great timing.

1987 to 2003 was an incredible run . 17 years averaging almost 14% cagr.

only 1 problem.  the years leading up  had years of stagnant markets and high inflation . it was sooooo hard to save a dime .  so it was all well and good when the good times came but few regular people had any money , there weren't even 401k's yet in most work places .

mathjak107 wrote:

Pointedstick wrote:

mathjak107 wrote: a lot better

Well that makes sense. 1987-2007 was awesome for stock-heavy portfolios. Really great timing.

1987 to 2003 was an incredible run . 17 years averaging almost 14% cagr.

only 1 problem.  the years leading up  had years of stagnant markets and high inflation . it was sooooo hard to save a dime .  so it was all well and good when the good times came but few regular people had any money , there weren't even 401k's yet in most work places .

Yeah, but you were ready when the opportunity came along. 15 or 20 years of respectable earnings and decent returns are enough for most people. Can't believe I am going to reference Harry Browne twice in one day but his rule #2 about not assuming that your wealth is replaceable comes to mind. 17 years at 14% is amazing. Kudos.

barrett wrote:

mathjak107 wrote:

Pointedstick wrote: Well that makes sense. 1987-2007 was awesome for stock-heavy portfolios. Really great timing.

1987 to 2003 was an incredible run . 17 years averaging almost 14% cagr.

only 1 problem.  the years leading up  had years of stagnant markets and high inflation . it was sooooo hard to save a dime .  so it was all well and good when the good times came but few regular people had any money , there weren't even 401k's yet in most work places .

Yeah, but you were ready when the opportunity came along. 15 or 20 years of respectable earnings and decent returns are enough for most people. Can't believe I am going to reference Harry Browne twice in one day but his rule #2 about not assuming that your wealth is replaceable comes to mind. 17 years at 14% is amazing. Kudos.

I would take 1 year at 14% 

buddtholomew wrote:

barrett wrote:

mathjak107 wrote: 1987 to 2003 was an incredible run . 17 years averaging almost 14% cagr.

only 1 problem.  the years leading up  had years of stagnant markets and high inflation . it was sooooo hard to save a dime .  so it was all well and good when the good times came but few regular people had any money , there weren't even 401k's yet in most work places .

Yeah, but you were ready when the opportunity came along. 15 or 20 years of respectable earnings and decent returns are enough for most people. Can't believe I am going to reference Harry Browne twice in one day but his rule #2 about not assuming that your wealth is replaceable comes to mind. 17 years at 14% is amazing. Kudos.

I would take 1 year at 14% 

Easy, see my VP PP. :-)

[img width=600]http://i.imgur.com/Oz9crTM.jpg[/img]

Here's a full chart of the real CAGRs of PPs started on the first of the month of every month since January 1972.  These PPs are rebalanced annually, with dividends reinvested.  Again, I took the PP data from PeakToTrough.com and CPI-U data from bls.gov/cpi/.

Some interesting features:

There have been several periods of longer than three years with negative real returns.

PPs started in June and July 1973 had negative real CAGRs after nine years!  The worst was -1.1% real CAGR for the June start.  These portfolios got off to bad starts and then rebalanced at terrible times in 1974, 1975, 1980, and 1981.  Inflation was high, too, at 9.13% annualized.

If we ignore everything from pre-1975, then the worst starting dates were February 1980 (-0.1% real CAGR after five years), July-October 1980 (between -0.3 and -1.3% real CAGR after four years), and several months between July 1998 and May 1999 (the worst being December 1998: -0.3% real CAGR after four years).

The average real CAGR has been remarkably consistent at 4.7%.  I take it to mean that if you are in the accumulation phase, adding money monthly and splitting the contributions evenly would have smoothed out the ride.

These returns are for PPs rebalanced every 12 months on the anniversary date.  A PP using 35/15 bands would probably have done better.  For example, starting on June 1 1973 and investing for nine years gave a nominal CAGR for a 35/15 PP (dividends reinvested) of 9.06%, with inflation of 9.13% annually, so a real CAGR of -0.07%.  Still poor, but better than the -1.1% real CAGR for the annual-rebalance strategy.

You can see a higher-resolution version of the chart here.

This was all in the past. What if the future calls for higher interest rates and lower gold prices? What then?

Nice job, Pet Hog.  That's a -lot- of data!

Just as a trivia note, this makes me appreciate doing my rebalancing in December.    Seems to be less red around then.

There are no guarantees in investing -- returns may be lower and may be higher.  I still have yet to find a portfolio I think is better for me.  Having made the best decision I feel I can with the available data, my view is that one who does not over-extend themselves on debt and spending is in the best position to handle any outcome at either extreme.  Control what you can so you don't have to lose sleep over what you cannot. 

Nine years?!!  Woah, fuil stop!!!

We now need to know which month is the absolute best to start and annually rebalance a PP.

Pet Hog wrote:

MachineGhost wrote: Nine years?!!  Woah, fuil stop!!!

We now need to know which month is the absolute best to start and annually rebalance a PP.

Jan 19724.712% 4.838%
4.739% 4.921%
4.719% 4.783%
4.645% 4.732%
4.401% 4.416%
4.306%
4.412%
4.363%
4.456%
4.472%
4.363%
4.583%

Actually, I meant that from the viewpoint of the shortest duration of negative real returns.

MachineGhost wrote: We now need to know which month is the absolute best to start and annually rebalance a PP.

That's easy -- always start your portfolio one month before a big gain!  

I have no data to support it, but it wouldn't surprise me if large negative market events tend to take a break during the holidays along with investors. 

In any case, don't forget that most people do not invest their life savings all at once and walk away.  They save a little bit at a time.  That changes the impact of that month #1 quite a bit.  A short term loss may be quickly mitigated by buying more investments at a discount.

I always rebalance at the end of the year.  Mostly because it's a lot easier to plan for taxes at the end of the year when income is known. 

Sorry about the earlier post (I posted instead of previewed).  These are the real CAGRs from my chart for years 42 and 43 after starting PPs in 1972.  December to April has been a better period than May to November for starting a PP and rebalancing.  February has been the best overall.  The difference between February and November is quite significant at 0.4 or 0.5%.  That makes a big difference when compounding for 40+ years.

Pet Hog wrote: [img width=600]http://i.imgur.com/Oz9crTM.jpg[/img]

Taking the final values from the right-hand side of each line in the chart above gives the following chart of real CAGRs over the years for annual-rebalance-on-anniversary PPs:

[img width=600]http://i.imgur.com/417ep7P.jpg[/img]

That's a pretty consistent portfolio. 

(The numbers assume cashing out on the final anniversary; for example, the CAGRs for all the January portfolios are as of January 1, 2015, not today.)

Beautiful visualizations.

Maybe my eyes are playing tricks on me but you can really see the January/February effect on that line chart. It's small but pretty consistent. Thanks for the charts/graphs.

iwealth wrote: Maybe my eyes are playing tricks on me but you can really see the January/February effect on that line chart. It's small but pretty consistent. Thanks for the charts/graphs.

Iwealth, I think it's an artifact arising from the way I calculated the returns.  The January and February data include the returns up to Jan 1 and Feb 1 2015, respectively, whereas the June to December data only go up to those same months in 2014.  The end of last year and January of this year were good months for the PP -- I'm guessing perhaps 8% real from July 1 to Feb 1? -- so the overall returns for each February will be that much higher than those for each July.  That 8% compounded over 40+ years would be perhaps 0.2% annually.  I think that's what we are seeing.

You can see the effect in the small table I posted earlier today:  The difference in CAGRs from February 1972 to February 2014 (42 years) and from February 1972 to February 2015 (43 years) is almost 0.2%:

Pet Hog wrote:

Code: Select all

1972  2014   2015
Jan  4.712%  4.838%
Feb  4.739%  4.921%
Mar  4.719%  4.783%
Apr  4.645%  4.732%
May  4.401%  4.416%
Jun  4.306%	
Jul  4.412%	
Aug  4.363%	
Sep  4.456%	
Oct  4.472%	
Nov  4.363%	
Dec  4.583%	

Having said that, the value for February 2014 is higher than that for June 2014, so perhaps there is something to it after all!

Last post for a while, I promise.

[img width=600]http://i.imgur.com/qqOtJ6I.jpg[/img]

This chart is similar to the one I posted yesterday, but this one was calculated for a PP with biennial (every two years) rebalancing on the anniversary.  So, starting January 1 1972 and then rebalancing on Jan 1 1974 and Jan 1 1976, etc., up to January 1, 2014, and then cashing out.  So, again, the numbers don't go up to the present, except for the portfolios started on each May 1 of an odd year (e.g., May 1, 1987 was rebalanced every two years and cashed out on May 1, 2015).  It's inconvenient, but easier to calculate with my limited Excel skills.

In theory, rebalancing biennially allows you to let your winners win and your losers lose; you don't catch a falling knife and whatever the opposite of that cliche is: you don't catch a rising firecracker in the crotch?  So generally it has a slightly higher CAGR, but not by much (see the difference below, plotted in green).  There's an obvious effect of the poor return from January 2012 to January 2014 (biennial, two years total, one rebalance) compared with the better return from January 2012 to January 2015 (annual, three years total, two rebalances), rippling all the way back to the 1980s.  I should really take the time to plot each portfolio up to the current day.

Anyway, this is another version of the PP that has had amazingly consistent real growth of 40+ years.

[img width=600]http://i.imgur.com/R9mJtfR.jpg[/img]

Pet Hog wrote: Last post for a while, I promise.

[img width=600]http://i.imgur.com/qqOtJ6I.jpg[/img]

This chart is similar to the one I posted yesterday, but this one was calculated for a PP with biennial (every two years) rebalancing on the anniversary.  So, starting January 1 1972 and then rebalancing on Jan 1 1974 and Jan 1 1976, etc., up to January 1, 2014, and then cashing out.  So, again, the numbers don't go up to the present, except for the portfolios started on each May 1 of an odd year (e.g., May 1, 1987 was rebalanced every two years and cashed out on May 1, 2015).  It's inconvenient, but easier to calculate with my limited Excel skills.

In theory, rebalancing biennially allows you to let your winners win and your losers lose; you don't catch a falling knife and whatever the opposite of that cliche is: you don't catch a rising firecracker in the crotch?  So generally it has a slightly higher CAGR, but not by much (see the difference below, plotted in green).  There's an obvious effect of the poor return from January 2012 to January 2014 (biennial, two years total, one rebalance) compared with the better return from January 2012 to January 2015 (annual, three years total, two rebalances), rippling all the way back to the 1980s.  I should really take the time to plot each portfolio up to the current day.

Anyway, this is another version of the PP that has had amazingly consistent real growth of 40+ years.

[img width=600]http://i.imgur.com/R9mJtfR.jpg[/img]

Thank you for this timely post and don't apologize for producing such meaningful content. I have been debating whether to contribute additional cash to the lagging assets but may postpone purchases until a lower band is breached. I always buy prematurely, but it appears patience and the PP go hand in hand.

Love that last line chart, Pet Hog.  Here's another way to look at it.  (Basically graphing the bottom bar on your main chart, but using my annual data)

[img width=500]http://i62.tinypic.com/307noxx.jpg[/img]

[img width=500]http://i61.tinypic.com/2h72f0k.jpg[/img]

Note the median hill in the middle of the 60/40 chart.  20-year runs get fat on the 80's and 90's but come back to earth the longer you look.  The PP is really consistent in comparison.

As a general note to anyone reading these, please keep in mind that while holding an investment over a long time decreases uncertainty of the average return, the uncertainty in any given year remains the same.  Even in year 30, performance next year could be anywhere in (or out of!) the range at the left side of the chart. 

(And apologies for posting these yesterday but deleting them.  I wanted to take some time to improve the charts to avoid confusion.)

Kudos and many thanks for the terrific charts!...  They really give a uniquely different (and comforting) perspective!

Pet Hog/Tyler, thanks again to both of you for all of the amazing visualizations.

Pet Hog, not sure if you'd be willing to make a modification, but it strikes me there's a little room for more granularity in the color coding. Since the 3-6% average return range makes up such a vast majority of the data, it would be interesting to see that broken down perhaps into two subsets like 3-4.5% and 4.5%-6% or even a 3-4, 4-5, 5-6%. After 30 years, it looks like the max return was 5.8% vs. a minimum of 3.6%. That's a pretty tight range, but it's still fairly large when one starts to consider safe withdrawal rates and such.

Anyways, just a thought..