The Desert Portfolio

[barrett]

Hey Edward,

Two questions...

Do you use 63 trading days because there are roughly that many in a quarter?

Also, what is your current asset allocation using the x/y/y formula?

Thanks.

[EdwardjK]

barrett,

Hi.  I use 63-trading days because there is literature that 63-days is a good measure of current trend/momentum.  As of June 30, 2015, my allocation is:

VTI - 26%
TLT - 19%
GLD - 23%
SHY - 32%

If I updated it today, the allocation would be:

VTI - 24%
TLT - 24%
GLD - 21%
SHY - 31%

[ochotona]

You could get over the asset per unit price influencing the weight if you used the ratio of beginning and end prices.

So if the asset is $50 at the beginning, and $55 at the end, you have a ratio 55/50 = 1.1. It would remove the asset price totally from the calculation. If the asset were $25 and ended at $27.50, you'd still get 1.1.

The question is, what kind of behavior do you want that 1.1 to correspond to? Do you want to use 1.1 itself as a relative weight, or square it, or cube it, or multiply it by a constant, or take the "e" exponent? Then after you do those operations on the individual assets, you'd form them into weights adding to 100%.

Those would be cool backtests, actually.

[dualstow]

A potentially novice question: why does the Desert Portfolio have such a long drawdown when it has such a small allocation to stocks?

This morning, I was looking at three portfolios: the pp, Larry's min fat tails, and Desert.
At Tyler's portfolio charts, we see the longest drawdowns for those are 2 years, 5 years, and 10 years respectively.

The worst year for Desert is a nice small -10% and I still find the DeP attractive, but why the protracted drawdown?

[barrett]

A potentially novice question: why does the Desert Portfolio have such a long drawdown when it has such a small allocation to stocks?

This morning, I was looking at three portfolios: the pp, Larry's min fat tails, and Desert.
At Tyler's portfolio charts, we see the longest drawdowns for those are 2 years, 5 years, and 10 years respectively.

The worst year for Desert is a nice small -10% and I still find the DeP attractive, but why the protracted drawdown?

For what time period do you see that long drawdown? When I run it on Peaktotrough for the 1975 to 2015 period, I really only see 1981 and 1994 as negative years. 1977 and 2001 (and 2015 so far) are also in the red but with losses that are virtually zero. For the bond portion I am just using 10-year treasuries, which is just my own simplification but close enough I think.

[Pointedstick]

It looks like 10-year treasuries didn't have a good run in the 70s. Also, TSM underperformed sliced-and-diced stock allocations during that time. Try swapping the 30% TSM out for for 10% each in large, medium, and small value stocks and see the difference. Also try turning the bullet bond allocation into a barbell. That helps a bit too.

[dualstow]

Will do, PS, thank you!

For what time period do you see that long drawdown?

I don't know, but it's in Tyler's chart at the right. Says, "10 yrs, Longest drawdown."
http://portfoliocharts.com/portfolio/desert-portfolio/

[barrett]

Will do, PS, thank you!

For what time period do you see that long drawdown?

I don't know, but it's in Tyler's chart at the right. Says, "10 yrs, Longest drawdown."
http://portfoliocharts.com/portfolio/desert-portfolio/

Yeah, I see it there on the chart for the ten years starting in 1972. It's due to the crappy performance of stocks and bonds in the 1970s (big generalization, I realize). The lower gold allocation wasn't enough to pull them up into positive territory.

[dualstow]

Yeah, I see it there on the chart for the ten years starting in 1972. It's due to the crappy performance of stocks and bonds in the 1970s (big generalization, I realize).

Yeah, that's it. Top line.

The lower gold allocation wasn't enough to pull them up into positive territory.

But the other portfolio mentioned in my question is Larry Swedroe's, and it has no gold. Maybe TIPS {hypothetically would have} worked back then?

[barrett]

But the other portfolio mentioned in my question is Larry Swedroe's, and it has no gold. Maybe TIPS {hypothetically would have} worked back then?

I don't know Swedroe's stuff but a quick glance shows 40% short-term bonds. That would be more or less like holding a lot of cash and cash had huge nominal returns in some of those years. In the summer of 1981 I was briefly getting 16% or so on a money market fund. A strong return in 40% of a portfolio is usually enough to avoid a losing streak. Hope that helps.

[barrett]

Dualstow, Hope I am not getting in the way in your search for truth! I just wanted to suggest having a look at the return of a portfolio that was 100% in one-year T-Bills from 1/1/72 to 1/1/82. That "portfolio" had a a CAGR of 7.78 and was especially juicy the last three years. This is all in nominal terms, of course, but with the choppy stock market of the 1970s (some years were terrific, others horrible), it would have been very helpful to be able to park some of your stock winners in a cash fund that was cranking out solid gains.

I will now crawl back into my hole.

[dualstow]

Not at all; you're very helpful.

[dualstow]

But, are you using a different tool than Tyler's? Whether I focus on short-term treasuries or the treasury money market (there is no T-Bill choice) I still get plenty of red in the 1970s.

[dualstow]

Thank you for that comprehensive reply! I never messed much with backtesting until Tyler's site came along. Before that, I mainly just looked at snapshots of what other people did with Simba's work at Bogleheads. Now, I put on ambient music and can't stop messing with these calculators.

With regard to what you said about gold, yes, the days of Harry's 'Coming Devaluation' book are long gone. But since 'Best Laid Plans' and 'Fail-Safe' came so much later, and he still recommended about 25% in gold in those, he must have still seen it as playing an important role  . Assets and portfolios continue to surprise, and gold may become The Great Multiplier Again. It was simply easier to think and talk about it when it was at 1700/oz. In the meantime, I like your idea of starting in '75 for simulations.

Ah, it seems like just yesterday many people were talking about how the miners were "on sale" and how it would be a great time to jump in. These days, nearly everything seems like it's on sale, and at the same time it doesn't feel like much of a bargain.

[ochotona]

I am thinking in particular about Newmont Mining (NEM), it has never been lower in inflation-adjusted price than right now, but my quotes only go back to the late 1970s. Schwab Equity Ratings has it as an "A" rated stock in an "F" rated industry. I think I'll buy some this week. It has been beat down -77% since its 2011 peak.

[dualstow]

Be careful. It can always be beaten down further.

[ochotona]

Be careful. It can always be beaten down further.

Sounds like many people here have had that experience!