The Great Bond Bubble?

Discussion of the Bond portion of the Permanent Portfolio

Moderator: Global Moderator

Hal
Executive Member
Executive Member
Posts: 182
Joined: Tue May 03, 2011 1:50 am

Re: The Great Bond Bubble?

Post by Hal » Tue Mar 06, 2018 4:38 am

Kevin K. wrote: Personally I think that reducing fat tails through some iteration of the Larry (Swedroe) approach of ~30% internationally diversified (and value/small tilted) equities and putting everything else in T bills or CDs going out no more than 2 years makes all kinds of sense in this environment.
Kevin, I would agree with your proposal of 30% Shares 70% short term bonds as a good alternative. It is very similar to a suggestion in Nassim Talebs "Black Swan" book. It is also supported by this economists five part post.

http://idiosyncraticwhisk.blogspot.com. ... ation.html

BUT this all depends on the currency not collapsing. And it for that reason I hold gold.

Try backtesting, say, 20% Gold, 20% All world value stocks and 60% Short term treasuries over at the excellent portfolio charts site.

Regards,
Hal

ps: Watch from the 16:30 min mark onwards. https://www.youtube.com/watch?v=VBJX5eH3X6A
Kevin K.
Full Member
Full Member
Posts: 83
Joined: Mon Apr 26, 2010 2:37 pm

Re: The Great Bond Bubble?

Post by Kevin K. » Tue Mar 06, 2018 10:09 am

That's great stuff (both the article and the video link) Hal - thank you!

It is indeed almost impossible, in playing around with the variables on Portfolio Charts, to avoid including at least 10% gold in these kind of defensive allocations.
User avatar
Desert
Executive Member
Executive Member
Posts: 3613
Joined: Sat Jun 26, 2010 2:39 pm

Re: The Great Bond Bubble?

Post by Desert » Tue Mar 06, 2018 10:04 pm

glennds wrote:
Desert wrote:
ochotona wrote:My bond duration is down to 5 years. Taken together with cash, they are 3 years. Definitely not HBPP. Well... there's a small HBPP in there somewhere, surrounded by a huge rind of Otherness.
Historically, it hasn't mattered much. Buy federally-backed fixed income with a composite duration in the 5-10 year range, and you've captured the essence of the HBPP fixed income philosophy.
But doesn't the HBPP look to the long bond for it's maximum volatility, which is exactly what you want in a negative correlation risk parity strategy? I would think shortening up the duration would reduce the bond volatility and interfere with the strategy.

The federal backing is the same in both scenarios, but that's only one piece of the bond equation in the HBPP strategy. If I'm wrong on this, please correct me.
glenn
Sorry, I may not have explained sufficiently. You're correct that the long bond volatility is important to HBPP past performance. But if one looks at the combination of 25% cash & 25% LTT ("barbell" fixed income portfolio), and compares it with a "bullet" fixed income portfolio with a roughly equivalent average duration, the overall portfolio performance has been very similar. In other words, one can allocate the fixed income into 25% cash and 25% LTT or instead 50% 10-year treasuries with little difference in past performance.
User avatar
sophie
Executive Member
Executive Member
Posts: 2889
Joined: Mon Apr 23, 2012 7:15 pm

Re: The Great Bond Bubble?

Post by sophie » Wed Mar 07, 2018 7:20 am

Yes, that's true: you can substitute a 50% allocation of intermediate bonds for the 25% cash and 25% LTT barbell and get about the same result on backtesting. Two caveats:

1. The barbell lets you use the PP as a cash management system as well as a long-term portfolio. That's one of the PP's best features IMHO. Think twice before giving this up.

2. Backtesting only tells you how the portfolio performs in the specific market environments that prevailed during the backtest. If you backtest during a period dominated by prosperity, your conclusion might not be worth much if you're about to enter a period of extended inflation or deflation.

I'd like to see a simulated backtest for the four major financial conditions described by Harry Browne. It would use actual data drawn from the appropriate market periods, but strung together differently so you could explore different scenarios, individually or in sequence.
User avatar
buddtholomew
Executive Member
Executive Member
Posts: 2032
Joined: Fri May 21, 2010 4:16 pm

Re: The Great Bond Bubble?

Post by buddtholomew » Wed Mar 07, 2018 7:55 am

sophie wrote:Yes, that's true: you can substitute a 50% allocation of intermediate bonds for the 25% cash and 25% LTT barbell and get about the same result on backtesting.
This is an accurate statement as I update my results daily and compare the LTT/Cash allocation (5.6 years) to the BND ETF with the same duration.
Psychologically it is more challenging to hold the LTT/Cash position (volatility in LTT's), but in aggregate the two positions perform within 5-15 basis points of each other on a daily basis.
On occasion I have witnessed some divergence, but the gap narrows in a few days with the laggard outperforming to catch-up.
The two approaches are closer than they are different.
Kevin K.
Full Member
Full Member
Posts: 83
Joined: Mon Apr 26, 2010 2:37 pm

Re: The Great Bond Bubble?

Post by Kevin K. » Wed Mar 07, 2018 8:33 am

sophie wrote:Yes, that's true: you can substitute a 50% allocation of intermediate bonds for the 25% cash and 25% LTT barbell and get about the same result on backtesting. Two caveats:

1. The barbell lets you use the PP as a cash management system as well as a long-term portfolio. That's one of the PP's best features IMHO. Think twice before giving this up.

2. Backtesting only tells you how the portfolio performs in the specific market environments that prevailed during the backtest. If you backtest during a period dominated by prosperity, your conclusion might not be worth much if you're about to enter a period of extended inflation or deflation.

I'd like to see a simulated backtest for the four major financial conditions described by Harry Browne. It would use actual data drawn from the appropriate market periods, but strung together differently so you could explore different scenarios, individually or in sequence.
To your first point not only is the 25% cash convenient it (and not gold) is what provides a large part of the PP's inflation protection, as Tyler shows pretty convincingly in this excellent post on his site:

https://portfoliocharts.com/2017/05/12/ ... -investor/

Regarding backtesting and how the 4 assets work during various market conditions, that's pretty much the entire subject of William Bernstein's little book "Deep Risk," which originated in conversations he and our esteemed mentor Craig Rowland had. One of Bernstein's key take-aways is that (a) gold is really more SHTF insurance than anything else and (most important) that the four economic conditions the PP is constructed in light of are nowhere near equally likely to occur or equally important to protect against, meaning that 4 x 25% is overly simplistic. The Golden Butterfly with its slight tilt towards prosperity and slight diversification of equities beyond TSM is one of many intriguing PP iterations that arise when considering Bernstein's points.
Post Reply