The I Hate "This" Asset PP

[Kbg]

I find this board pretty interesting in terms of its personalities. Some I have no idea why they even post as they obviously can't stand the PP or one of its assets and then there are around 10 who make some awesome contributions...anyway, I think the HBPP/Ray Dalio stuff has merit as an investing concept, but I agree with the thought that one is insuring against 4 scenarios that aren't going to all happen at the same time or with the same likelihood. However, there is a very simple way around this. Just use momentum. Momentum predicts nothing it just goes with the flow of the market, so it is true to HB's premise that no one can predict the market. I personally believe that the market is relatively good at figuring out what the investing/economic climate is and adjusting/voting with asset flows. Not immediately, but most of the time it is right. So here is my solution to A) those who can't stand a particular asset (thought at times it really should be the one they are in) and B) I believe will always outperform the normal PP due to the "over insurance" factor. The cost may be more volatility, but I'm not so sure it will be any more volatile.

Test Dates 1/1/05-11/20/15, Monthly Rotation, Absolute Momentum, SHY, SPY, TLT, GLD, Dividend adjusted returns

Hold top ranked based on X momentum (CAR/MDD)

1mo 3.43/25.61
3mo 13.23/25.24
12mo 9.2/28.53

Hold top 2 ranked based on X momentum (CAR/MDD)

1mo 7.74/16.7
3mo 11.88/13.79
12mo 10.07/14.98

Hold top 3 ranked based on X momentum (CAR/MDD)

1mo 6.52/11.39
3mo 8.14/18.12
12mo 7.69/10.08

By comparison

PP Annual Rebalance - 7.25/13.96
50/50 SPY/AGG no rebalance - 5.87/27.27
50/50 SPY/AGG ann rebalance - 6.26/27.89

60/40 SPY/AGG no rebalance - 6.17/33.31
60/40 SPY/AGG ann rebalance - 6.58/33.96

Buy and Hold
SHY - 2.27/2.23
SPY - 7.33/55.17
TLT - 6.74/26.58
GLD - 8.35/44.56

Observations

For all the gold haters, the facts are what the facts are. Over this almost 11 year period GLD was in fact the best raw performer. And I repeat, these are dividend adjusted (total return) figures.

A straight PP has outperformed  both 50/50 and 60/40 portfolios with FAR less MaxDD

The momentum versions provide better performance than a normal PP at about the same volatility. Momentum is also an undisputed performance factor backtested over centuries of data and it is also the strongest performance factor of all factors identified academically. (If you begin seriously thinking about doing this version of a PP, please read up on "momentum crash." It will happen to you when using momentum and is unavoidable using momentum only techniques. You should also consider tax consequences which will be intrinsically tied to whatever length of momentum you use. Shorter = more turnover). Please note, these figures also include rebalancing on a monthly basis as well.

If one thinks about the four economic quadrants the two holdings version makes a lot of intrinsic sense to me...and I think it is probably the best way to implement the theory in a practical application. You pretty much get the same DD with better performance and you don't have to rue the fact you are holding the "I hate this asset." If you want to dial down risk, you could use one of the Top X holds variants and add a permanent cash element which would reduce performance and DD by whatever % allocated to cash. If one was to do this I'd recommend 3+ month momentum and the 2 or 3 holdings versions.

Maybe MG can take this one back for us further in time which would be cool.

[mathjak107]

tell me what asset you want me to make "the best " and i will gladly find the time frame to make it shine .

the real deal is what it means to you and you own balance and time frame .

how about the last 5 years ?    the last 7 years ?

randomly picking time frames means very little to anyone . most have balances that are far bigger the last 5 years then 11 years ago carrying far more weight .  what happened to my money even 15 years ago is a drop in the bucket compared to the last 5 years when investment dollars are at max .

just the fact the pp is down 2% ytd and my 50/50 model of funds is up about 2% represents a difference of over 110k at the moment .  so the weighting of your balance is going to count  very heavily  in the equation as opposed to throwing out random  feel good numbers that you mold to fit . .

great if they apply to you but very different results if they don't . 

i was an investor since 1987 and i can recite the fact that the most amazing bull market in history took place in my time frame . 17 years of almost 14% returns cagr from 1987 to 2003 .

but i had a fraction of the money i have today  invested back then  so the time frame meant very little in my scheme of things .  looks great in back testing and pretty charts but that is the extent of the impact on me .



.

[ochotona]

Nice work kbg, what is your opinion on using a 200 day moving average kill switch to avoid bear markets?

[Kbg]

just the fact the pp is down 2% ytd and my 50/50 model of funds is up about 2% represents a difference of over 110k at the moment .  so the weighting of your balance is going to count  very heavily  in the equation as opposed to throwing out random  feel good numbers that you mold to fit . .

Kettle, black.

I picked 2005 because that is how far back I could do the backtest due to ETF history limitations. But ok, let's take your "not random feel good numbers" against a 50/50 model and run the numbers.

11/20/14 -11/20/15
MoPP (Monthly Check)
              Top1        Top2          Top3
63          -3.74        -4.27        -4.99
252        -6.84        -5.47        -3.73

One Full Market Cycle 10/1/07 -11/20/15
              Top1                    Top2                    Top3
63          11.05/-26.89        7.51/-15.43        5.63/-12.52
252          6.22/-32.57        6.63/-21.41        6.71/-8.69

Bull Market Cycle 3/1/09 -5/31/15
              Top1                    Top2                      Top3
63          10.28/-26.88        8.14/-15.43        6.40/-11.46
252          5.48/-32.57        7.56/-12.30        6.76/-8.36

PP Annual Rebalance, Dates Per Above
-0.11/-6.97
6.24/13.96
8.91/-7.87

50/50 Annual Rebalance, Dates Per Above
2.80/-6.10
6.06/-27.17
13.21/-7.80

So what can we conclude from facts alone without cherry picking dates from very recent history? (And, yes I am going to ignore a single year's results because it is utterly pointless to conclude anything from a financial data series using a single year's results. But just so you could do a little victory lap, and to avoid being labeled a cherry picker, I included the past year's performance.)

- 50/50 BH performs better in an up cycle than HBPP or MoPP

- HBPP and MoPP performed better over a full market cycle

- When using mo it is probably better to use a shorter look back length

My take: If you are good at predicting the future, then you are just plain stupid not to take advantage of that ability. If you are not, I don't think there is any question that something that diversifies beyond just stocks and bonds is a safer and more consistent bet.

The numbers, they don't lie...and there is absolutely nothing here that should surprise anyone.  So beware a couple of posters on the board. Facts can be so inconvenient to recency biased beliefs.

[mathjak107]

if you want , compare a typical rolling  25 or 30 year accumulation time frame  between the two . want to bet it is no contest , not even close  over most if not all of them  ?

[Pointedstick]

Just put mathjak on ignore, you'll never budge him.

[mathjak107]

facts are facts .  you can bend , twist and mold them any way you like but when you look at a typical savings time frame  over 25-30 years there is no comparison .  odds are the 50/50 mix blew the pp away .

it should , the voltility is higher and it would be expected to . but to pull out various chunks of time and make that claim is more mental masturbation then actual long term results will demonstrate  .

even a 2% a year difference on average  is 50 or 60% difference in the balance at a min and i bet it is greater than that over most of them.

[Kbg]

Nice work kbg, what is your opinion on using a 200 day moving average kill switch to avoid bear markets?

As a general rule, absolute momentum works better if one backtests/compares the two. If you do use a 200MA, it is important you put a volatility band around it or whipsaw will eat your lunch. But regardless of one's preference, the impact is generally the same. You get near the same returns with about half the DD. On an absolute basis this technically shouldn't matter. But if time matters in your investing horizon, then it matters a great deal. This has been amply demonstrated elsewhere on this board via retirement withdrawal rate analysis.

And of course, this all assumes one can avoid tinkering. There is something to be said about a portfolio that you should just ignore for long periods of time.

However, at the end of the day...over long periods of time most of these fixed asset percentage ports end up in pretty much the same place performance wise. Thus, from a performance perspective I'm not impressed with the PP at all as compared to any of the rest.  What I find quite interesting/useful in my own investing is the phenomenal reduction in volatility. This is a quality that can be exploited with more advanced investing methods.

[dutchtraffic]

Mathjak107 talks a whole lot of nonsense.

But he is totally right in this case.

[mathjak107]

the only one around here my friend with zero facts to back up what they spew as well as verbally attack posters is YOU!

what i forgot is likely more than what you know based on what you demonstrate here . .

don't like the posts , use ignore or just don't read them . obviously you keep reading them .

[Kbg]

facts are facts .  you can bend , twist and mold them any way you like but when you look at a typical savings time frame  over 25-30 years there is no comparison .  odds are the 50/50 mix blew the pp away .

MJ,

My last post on this (and for the record I'm not arguing with you, but rebutting your B.S. so that folks who would like facts can make an intelligent decision).

The correct answer is it all depends on the starting and end points of the time period examined. Your statement is utterly and demonstrably false depending on the start date. It is also utterly and demonstrably true depending on the start date. What is absolutely unarguable is that across all starting points the HBPP has delivered comparable results with far less volatility than a traditional 50/50.

So I've laid out the numbers, please demonstrate the truthfulness of your statement with facts and math.

[Kbg]

Mathjak107 talks a whole lot of nonsense.

But he is totally right in this case.

Numbers please and describe your methodology when you post them.

Thanks.

[dutchtraffic]

There are plenty of portfolios that have absolutely destroyed the PP, it is obvious that a 1% difference is going to be a big deal in the longer term. I don't think we can deny that.

random example:

[mathjak107]

facts are facts .  you can bend , twist and mold them any way you like but when you look at a typical savings time frame  over 25-30 years there is no comparison .  odds are the 50/50 mix blew the pp away .

MJ,

My last post on this (and for the record I'm not arguing with you, but rebutting your B.S. so that folks who would like facts can make an intelligent decision).

The correct answer is it all depends on the starting and end points of the time period examined. Your statement is utterly and demonstrably false depending on the start date. It is also utterly and demonstrably true depending on the start date. What is absolutely unarguable is that across all starting points the HBPP has delivered comparable results with far less volatility than a traditional 50/50.

So I've laid out the numbers, please demonstrate the truthfulness of your statement with facts and math.

so you are trying to tell us you find 25-30 year period's  from 1975 on that  the pp beat a 50/50  with dividends reinvested ?  show me the money

[Kbg]

There are plenty of portfolios that have absolutely destroyed the PP, it is obvious that a 1% difference is going to be a big deal in the longer term. I don't think we can deny that.

random example:

This is your rebuttal? Laughable. So how about a Sharpe or Sortino ratio? What this chart tells me is that I can leverage a HBPP to 146% and achieve an 13.46% return with the exact same risk profile (at least if your snapshot of risk is annual/the duration of this chart...who knows what the daily risk would be without running the numbers on a daily basis). I'm pretty sure at risk equivalent the blue squiggly goes above the orange squiggly now.

Thanks...next attempt?

[mathjak107]

a 50/50 mix is hardly something that is typically used for growth for typical long time frames . that is the most popular retirement model .  that is what i use being retired  and i can tell you ytd  there  is a 110k difference between the model and the pp  just ytd  on the amount i have invested  . i post the results weekly in my thread .

if you come up with more than 1 rolling 25-30 year typical accumulation period where the balance in dollars was more than even a 50/50 portfolio , which is not a growth portfolio , i would be quite surprised . i am not even sure there is one that was comparable .

but that is the expected results when you pay for the insurance of having lower volatility .

the pp is good at what it does but it is going to be very rare it is going to match a more aggressive conventional model long term  and rightfully it shouldn't .

but to try to find smaller chunks of time and proclaim victory is just silly .

[ochotona]

As a general rule, absolute momentum works better if one backtests/compares the two. If you do use a 200MA, it is important you put a volatility band around it or whipsaw will eat your lunch. But regardless of one's preference, the impact is generally the same. You get near the same returns with about half the DD. On an absolute basis this technically shouldn't matter. But if time matters in your investing horizon, then it matters a great deal. This has been amply demonstrated elsewhere on this board via retirement withdrawal rate analysis.

I got whipsawed into cash during September-October. Fortunately, it's commission free ETFs in my IRA & 401(k). Being in my mid-50s, the MaxDD starts to matter a great deal. I'll take some lashes and bruises to avoid severing limbs.

I like this experiment done with four basic PP components. It's a non-discretionary trading algorithm based on the data. People forget that 15/35 rebalancing bands are also a chosen trading algorithm. None of these are ordained by the Almightly.

[ochotona]

Indulge us once more, please... run the simulation again with 10% buy-and-hold gold (Desert allocation), and 15% allocated by momentum.

Sorry, I figured out how to run this on Portfolio Visualizer.

[Kbg]

I'd love to run the rolling 25-30 year time frame analysis I just don't have the data. The gold standard messes up the US calculations, but my sense is that it really does depend on the start point.  Essentially anything that had stocks in it/weighted stocks for the last half of the 20th century did great due to the epic 1980-2000 run. However, the first half of the century was pretty much the opposite due to the 1907 and great depression bears.

The fact is no one knows what is going to work best going forward and while I'm not a big gold fan of the three non-cash PP components only gold has been thrashed so my unscientific gut says over the next 10 years it may well be the leader of the performance pack. (This would be known as the wild **s guess...but mean reversion is an investing fact). It isn't going anywhere though until the dollar turns south which given the rest of the world isn't likely to happen any time soon. Once the dollar turns gold will do well again but the best place to be will probably be foreign stocks which are in a sustained bear market starting back in mid 2014. Valuation measures are beginning to look interesting.

[mathjak107]

my feeling is holding so much gold is like waiting so long for your ship to come in the pier collapses .

so much money ends up not being made with the money dead in gold and the fall so great that even if it doubled or tripled and the other assets fell by a lot you still missed the boat and are behind  .

gold in my opinion is not like equity's in the sense that time in the market makes them successful  here in this country , gold is about timing the market .

betting so much on calamity and disaster happening on a full time basis  has always left you poorer over the long term .even if you bought your gold back in 1975 at 175 an ounce you lagged the lowly treasury bill  today .

i can see a tactically driven pp  possibly doing well  long term  where instead of the shotgun approach  you cut back those assets classes where the cost of insurance is way to high .

you don't want to work at the umbrellas stand in a drought that lasts years .  at least wait for signs of approaching rain .

you can give up the first 10% and bottom 10% by timing and still capture 80% of the middle according to all those who believe in the IVY  or hedgable  concept . . 

[Kbg]

my feeling is holding so much gold is like waiting so long for your ship to come in the pier collapses .

so much money ends up not being made with the money dead in gold and the fall so great that even if it doubled or tripled and the other assets fell by a lot you still missed the boat and are behind  .

gold in my opinion is not like equity's in the sense that time in the market makes them successful  here in this country , gold is about timing the market .

betting so much on calamity and disaster happening on a full time basis  has always left you poorer over the long term .even if you bought your gold back in 1975 at 175 an ounce you lagged the lowly treasury bill  today .

i can see a tactically driven pp  possibly doing well  long term  where instead of the shotgun approach  you cut back those assets classes where the cost of insurance is way to high .

you don't want to work at the umbrellas stand in a drought that lasts years .  at least wait for signs of approaching rain .

you can give up the first 10% and bottom 10% by timing and still capture 80% of the middle according to all those who believe in the IVY  or hedgable  concept . .

I don't actually disagree with any of this. I am neither a gold lover nor hater. More recent US history suggests gold does spectacularly well and spectacularly poorly. HB was quite accurate in it's purpose and under what conditions it moves though. I thought what he wrote was the clearest explanation of gold I've ever read. And my read was while inflation is usually the condition under which it moves, the reason it really moves is people are looking for a USD replacement for whatever reasons. However, I do think it is legitimate to question gold allocation, should I hold it, should I hold this much of it. Historically it provides zero return but it holds value over time as well. Kind of like holding a dollar that doesn't deflate and likes to ride roller coasters.

[ochotona]

If you add 10%, 20% gold to a portfolio, and historically backtest it, there is no doubt it shows significantly improved safe withdrawal rates for retirement... based on 1975 - 2014.

But what is missing is a SWR analysis based on gold entry price. Buying gold in 1980 or 2011 was like buying tech stocks in 1999... the following years are likely filled with tears. We need to qualify what this looks like.

[Kbg]

If you add 10%, 20% gold to a portfolio, and historically backtest it, there is no doubt it shows significantly improved safe withdrawal rates for retirement... based on 1975 - 2014.

But what is missing is a SWR analysis based on gold entry price. Buying gold in 1980 or 2011 was like buying tech stocks in 1999... the following years are likely filled with tears. We need to qualify what this looks like.

I personally believe it is folly to look at a portfolio's constituents in isolation. Now you absolutely want to look at them in isolation when constructing the portfolio so as to pick constituents that do whatever it is you are trying to achieve in the basket of items you have chosen. But once in the basket, the basket is the only thing that matters.  Once in the basket you monitor the individual constituents to see if their behavior has fundamentally changed...but that takes several years before you can make a legitimate call, not 1-3 years.  If you are going to try and time a particular asset, then use momentum. It is time tested and works...so long as you can stomach transition periods which are very frustrating.

[mathjak107]

i am in 100% agreement . i always say you have to look at how the pieces play together and do they hurt each other or support each other as th big picture changes . .

as an example if we have stock returns that are going to be weak , then at this point in time my feeling is you do not want anything in your portfolio grabbing equity's by the collar  and trying to hold their feeble attempts at gains back .

in a strong bull market that may not matter and you want the protection . but now that can hurt more then help .

that is why every fidelity insight model i track is positive for the year and the pp can't gain traction .  every time what has been the most likely  horse  makes a run for the finish line either the weight of long term treasury's or gold stick their foot out and trip it .

these are not the times for equal weighting , these are times for the other pieces to help  the player that seems to have traction get to the finish line as best as they can without getting in the way ..

[sophie]

Probably I should stay out of this, but for the benefit of the people watching on the sidelines I'll just inject a few common sense points:

- There is always going to be a portfolio that does better than the one you're holding.  Like, say, 100% gold during the 1970s. The problem is that you don't know which one it's going to be.  Also of course if you're not actually holding it during the entire period of your backtest, it does you no good, and only makes you frustrated with the "I could have done better" feeling.

- CAGR is not the whole picture - portfolio management and usability are easy to overlook.  A steep drawdown during the withdrawal phase (i.e. retirement) can hurt significantly, as you may be forced to sell assets that have dropped in value, thus locking in losses.  This problem is exacerbated by portfolios that do not include a cash allocation.  Even when you're in the accumulation phase, you may suffer a catastrophic event (job loss, medical condition) that requires that you dip into your retirement funds.  And, there is a better than even chance that a job loss will coincide with a bad market condition, as Libertarian666 pointed out.

- The above portfolio comparison completely ignores taxes and trading fees.  A finely sliced portfolio like #2 would have incurred a lot of steep fees over the backtest period, and probably would have to be rebalanced annually.  In comparison, the PP does not require much trading as rebalances occur only about every 2-3 years, and it's highly tax efficient.