Questions to Consider, Interested in Thoughts

[buddtholomew]

Should it matter which asset/s carry the portfolio to earn a positive inflation adjusted return? Also, what is the thought process behind having one or more assets outpacing the other assets to achieve this gain?

[dualstow]

Should it matter which asset/s carry the portfolio to earn a positive inflation adjusted return? Also, what is the thought process behind having one or more assets outpacing the other assets to achieve this gain?

a. No.
b. Not knowing?

[buddtholomew]

Assets have a floor (0) but no upper ceiling. Does this explain why we should expect to earn a positive, inflation adjusted return? Is there anything more to it than that?

[MachineGhost]

No, but why would you want to carry an asset that doesn't have a real return going forward unless it helped other assets achieve a real return? 

You get a real return for risking your money and taking your chances.  Otherwise inflation eats your capital away like a rat to Swiss cheese.  It's not really more complicated than that.

[rickb]

Assets have a floor (0) but no upper ceiling. Does this explain why we should expect to earn a positive, inflation adjusted return?

No.

Is there anything more to it than that?

Yes.

50% of the PP is in short term T-Bills and gold which have an expected long term return roughly equal to inflation, so the rest must have an expected return of more than inflation.  Long term bonds return about inflation plus 2% and stocks return about inflation plus 3%.  You might think the PP should therefor return about inflation plus 1.25%.  You're wrong because 3 of the assets in the PP are extremely volatile and essentially uncorrelated with each other.  By rebalancing, you're selling these assets when they're high and buying them when they're low.  Definitely no guarantees - but over the last 4 decades this has consistently added about another 3% (!) to the return. 

The PP not only has you own assets with 0 floor and unlimited upside and each with an expected upside greater than or equal to inflation, but forces you to own extremely volatile assets and rebalance between them.

Genius is perhaps too weak of a word.

[Mark Leavy]

Should it matter which asset/s carry the portfolio to earn a positive inflation adjusted return? Also, what is the thought process behind having one or more assets outpacing the other assets to achieve this gain?

Some very good comments so far.

I've always liked the 4 bucket analogy for the PP.

The US's gross domestic product is growing at some rate.  Someone is creating value and that value is going somewhere.  Harry surmised that holding bonds, equities, gold and cash would be a sufficient portfolio such that the value created by the US economy would somehow slosh into one of those buckets.

Maybe he was right.  Maybe he wasn't.  But I like the concept of holding all of the sinks that all of the country's added value has to slop into.

[EdwardjK]

Also, what is the thought process behind having one or more assets outpacing the other assets to achieve this gain?

I do not think that overweighting an outperforming asset(s) is contrary to the PP principles.  As long as you remain invested in the four PP components, overweighting can lead to higher portfolio returns in exchange for very modest higher volatility.

My backtest using 63-day day returns of the four assets resulted in an average 2% higher return per year during the backtest period.  The backtest assumed I rebalanced on January, April 1, July 1, October 1.

Although I have not tested the idea, NOT investing in a PP asset that has a negative return over the same 63-day period may result in WORSE returns, as you may miss out on rebounds while you sit on the sidelines.

[FarmerD]

Some very good comments so far.

I've always liked the 4 bucket analogy for the PP.

The US's gross domestic product is growing at some rate.  Someone is creating value and that value is going somewhere.  Harry surmised that holding bonds, equities, gold and cash would be a sufficient portfolio such that the value created by the US economy would somehow slosh into one of those buckets.

Maybe he was right.  Maybe he wasn't.  But I like the concept of holding all of the sinks that all of the country's added value has to slop into.

I think CraigR mentioned that idea briefly somewhere but I've never seen where Harry wrote that.  I'd be extremely interested to see some more on that - do you have any references where Harry talks about that?  Perhaps one of his books?

[barrett]

I do not think that overweighting an outperforming asset(s) is contrary to the PP principles.  As long as you remain invested in the four PP components, overweighting can lead to higher portfolio returns in exchange for very modest higher volatility.

My backtest using 63-day day returns of the four assets resulted in an average 2% higher return per year during the backtest period.  The backtest assumed I rebalanced on January, April 1, July 1, October 1.

How are you going about overweighting an outperforming asset? And how are you rebalancing in your backtest? Thanks.

[EdwardjK]

[quote author=barrett link=topic=7022.msg116225#msg116225 date=1426513616
How are you going about overweighting an outperforming asset? And how are you rebalancing in your backtest? Thanks.
[/quote]

On the first of January, April, July and October, I calculate the 63-day return of the four assets using the formula (Price today / Price 63-days ago) / Price 63-days ago.  Since we want to remain in the four assets regardless of a gain or loss, this formula results in a positive number for every calculation.  The assets are then allocated based upon the formula result for that asset divided by the sum of all four formula results. 

For example, on March 10, 2015, the Yahoo Finance adjusted values were as follows:

VTI  $106.22
TLT  $126.28
GLD  $111.42
SHY  $84.60

The results of the calculations using the December 8, 2014 values (63 days from March 10, 2015), results in the following values:

VTI    0.95%
TLT    0.85%
GLD  0.83%
SHY  1.19%

Sum  3.82%

You then allocate based upon the asset's result divided by the total:

VTI  25%  (0.95% / 3.82%) = 25%
TLT  22%
GLD  22%
SHY  31%

Sum  100%

My backtest from January 1, 2006 through mid-December, 2014 resulted in a 9.4% CAGR and standard deviation of 5.2%, compared to the HBPP of 7.2% CAGR and 5.1% standard deviation (using ETF Replay).  FYI, the backtest was limited to January 1, 2006 because some of the ETFs history does not go back further.

I also backtested using monthly rebalancing , 21-day returns and relative strength; and combinations of them; but using the 63-day return and quarterly rebalancing yielded the best results.

If you want to see my Excel file, send me a private message with your e-mail address.

Thanks.

[Mark Leavy]

Some very good comments so far.

I've always liked the 4 bucket analogy for the PP.

The US's gross domestic product is growing at some rate.  Someone is creating value and that value is going somewhere.  Harry surmised that holding bonds, equities, gold and cash would be a sufficient portfolio such that the value created by the US economy would somehow slosh into one of those buckets.

Maybe he was right.  Maybe he wasn't.  But I like the concept of holding all of the sinks that all of the country's added value has to slop into.

I think CraigR mentioned that idea briefly somewhere but I've never seen where Harry wrote that.  I'd be extremely interested to see some more on that - do you have any references where Harry talks about that?  Perhaps one of his books?

Farmer, I may be making a wrong attribution.  I had thought it was a topic I had heard him discuss on his radio shows, but now that you ask, I could easily have picked it up here on the boards somewhere.

[buddtholomew]

The PP not only has you own assets with 0 floor and unlimited upside and each with an expected upside greater than or equal to inflation, but forces you to own extremely volatile assets and rebalance between them.

Genius is perhaps too weak of a word.

The above comment is clear to me. How much weight do you assign to this argument - rising assets will advance higher in percentage terms than falling assets will drop lower to produce a positive return? I heard HB provide this argument when discussing gold returns and the 70's although he did stress the increase was unlikely to repeat in our lifetimes.

[sophie]

50% of the PP is in short term T-Bills and gold which have an expected long term return roughly equal to inflation, so the rest must have an expected return of more than inflation.  Long term bonds return about inflation plus 2% and stocks return about inflation plus 3%.  You might think the PP should therefor return about inflation plus 1.25%.  You're wrong because 3 of the assets in the PP are extremely volatile and essentially uncorrelated with each other.  By rebalancing, you're selling these assets when they're high and buying them when they're low.  Definitely no guarantees - but over the last 4 decades this has consistently added about another 3% (!) to the return. 

....

Genius is perhaps too weak of a word.

+1!!

Save that quote next time someone on the OTHER investing forum complains about how useless it is to invest in gold.

Incidentally, this is why I believe people who tinker with the PP balances may be shooting themselves in the foot.  During any rebalance event, there is always one asset that will be sold off and you don't know in advance which one that will be.  If you deliberately underweighted that asset, you've lost out on the rebalance boost in addition to the gains enjoyed by the asset on its own.  Backtesting doesn't count for much, since new situations will always be...new.

[Kbg]

EdwardjK,

Very cool. Could you post some allocation extremes of this?

[barrett]

Some very good comments so far.

I've always liked the 4 bucket analogy for the PP.

The US's gross domestic product is growing at some rate.  Someone is creating value and that value is going somewhere.  Harry surmised that holding bonds, equities, gold and cash would be a sufficient portfolio such that the value created by the US economy would somehow slosh into one of those buckets.

Maybe he was right.  Maybe he wasn't.  But I like the concept of holding all of the sinks that all of the country's added value has to slop into.

I think CraigR mentioned that idea briefly somewhere but I've never seen where Harry wrote that.  I'd be extremely interested to see some more on that - do you have any references where Harry talks about that?  Perhaps one of his books?

Farmer, I may be making a wrong attribution.  I had thought it was a topic I had heard him discuss on his radio shows, but now that you ask, I could easily have picked it up here on the boards somewhere.

This is Sophie's summary of the PP exactly... if you substitute 'feeding troughs' for 'buckets' and throw in some hungry pigs.

[EdwardjK]

TennPaGa,

Thank you from our reply.  I agree with your math and had similar observations.  Please keep in mind though, that the PP strategy requires you to be invested in all four assets, even if one or more of those assets has lost value.  In fact, the strategy requires you to increase investment in losing assets in order to maintain values within the 15%-35% range.

With that said, your examples show that an investor would allocate a higher percentage allocation to that asset(s) with the highest growth, and a lower allocation to that asset(s) with the lowest growth or a loss.  So why is that a bad thing?

My backtest shows that favoring growing assets over either slower growing or declining assets results in a better overall return compared to the traditional PP strategy.  And quarterly rebalancing helps capture any change in asset momentum.

For the curious, following is the historical allocation from my backtest file:

Date         TLT VTI GLD SHY
12/19/14 25% 27% 18% 30%
10/1/14 25% 27% 18% 30%
7/1/14 24% 27% 20% 29%
4/1/14 25% 26% 21% 28%
1/2/14 23% 30% 18% 29%
10/1/13 22% 30% 20% 28%
7/1/13 21% 35% 14% 30%
4/1/13 20% 37% 14% 29%
1/2/13 21% 36% 14% 29%
10/1/12 20% 37% 16% 27%
7/2/12 25% 33% 14% 28%
4/2/12 19% 41% 15% 25%
1/3/12 19% 40% 14% 27%
10/3/11 32% 29% 15% 24%
7/1/11 26% 33% 15% 26%
4/1/11 24% 37% 15% 24%
1/3/11 19% 42% 16% 23%
10/1/10 22% 36% 18% 24%
7/1/10 27% 34% 17% 22%
4/1/10 23% 37% 17% 23%
1/4/10 19% 40% 19% 22%
10/1/09 21% 42% 18% 19%
7/1/09 15% 49% 17% 19%
4/1/09 16% 41% 21% 22%
1/2/09 29% 24% 24% 23%
10/1/08 27% 29% 19% 25%
7/1/08 24% 32% 21% 23%
4/1/08 26% 26% 23% 25%
1/2/08 25% 26% 27% 22%
10/1/07 26% 24% 28% 22%
7/2/07 24% 28% 24% 24%
4/2/07 24% 27% 26% 23%
1/3/07 22% 29% 28% 21%
10/2/06 27% 28% 23% 22%
7/3/06 23% 27% 27% 23%
4/3/06 21% 29% 29% 21%
1/3/06 21% 27% 32% 20%

[EdwardjK]

TennPaGa,

You are correct.  Fortunately, the asset prices are not so dissimilar as to cause the distortion you have displayed. 

There may be some way to standardize or normalize the asset prices and still retain the overall relationships, but I have to research that to be sure.

In any event, I believe my backtest does trigger the imagination to investigate alternative approaches to the PP strategy, yet still comply with the four asset requirement.

If one wanted to step outside the formal strategy, consider the same approach using the actual 63-day returns ((Price today - Price 63-days ago) / Price 63-days ago) for the four assets and then invest only in those assets with a positive return.  Determining the allocation using the actual returns as the weightings results in a much superior result compared to the traditional HBPP and the method I described earlier.

In this case, the CAGR is 14.1% with a STD-P of 4.9%, compared to a HBPP CAGR of 7.2% with a STD-P of 5.1%.  If you can stomach not investing in all four assets, the allocations would be as follows:

Date       TLT VTI GLD SHY
12/19/14 62% 36% 0% 2%
10/1/14 62% 36% 0% 2%
7/1/14 47% 28% 22% 3%
4/1/14 43% 20% 36% 1%
1/2/14 0% 100% 0% 0%
10/1/13 0% 42% 56% 2%
7/1/13 0% 100% 0% 0%
4/1/13 0% 99% 0% 1%
1/2/13 0% 100% 0% 0%
10/1/12 0% 38% 61% 1%
7/2/12 98% 0% 0% 2%
4/2/12 0% 74% 26% 0%
1/3/12 0% 96% 0% 4%
10/3/11 82% 0% 17% 1%
7/1/11 30% 13% 48% 9%
4/1/11 0% 86% 14% 0%
1/3/11 0% 66% 34% 0%
10/1/10 30% 0% 66% 4%
7/1/10 59% 3% 33% 5%
4/1/10 0% 63% 26% 11%
1/4/10 0% 49% 51% 0%
10/1/09 19% 47% 31% 3%
7/1/09 0% 86% 14% 0%
4/1/09 0% 0% 100% 0%
1/2/09 78% 0% 16% 6%
10/1/08 71% 0% 0% 29%
7/1/08 0% 63% 37% 0%
4/1/08 23% 0% 59% 18%
1/2/08 25% 3% 62% 10%
10/1/07 26% 0% 63% 11%
7/2/07 0% 90% 0% 10%
4/2/07 7% 33% 45% 15%
1/3/07 1% 41% 53% 5%
10/2/06 71% 12% 0% 17%
7/3/06 0% 0% 89% 11%
4/3/06 0% 32% 66% 2%
12/30/05 15% 11% 69% 5%

[MachineGhost]

The problem is when you need the asset after a fat tail, your exposure will be non-existent to minimal.  All momentum strategies increase both gain and maximum drawdown without exception.

[EdwardjK]

All momentum strategies increase both gain and maximum drawdown without exception.

MachineGhost,

This sounds like an overly broad statement.  Would you please elaborate?

[MachineGhost]

This sounds like an overly broad statement.  Would you please elaborate?

Nature of the beast in my experience.  It amps up both risk and reward, sometimes the risk a lot for very little additional reward.  To get a better ratio of risk:reward you have to use trendfollowing not momentum.