What are your thoughts on the "Golden Butterfly" portfolio?

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
User avatar
Forester
Posts: 2400
Joined: Sat Jan 19, 2019 1:50 pm
Location: UK

Re: What are your thoughts on the "Golden Butterfly" portfolio?

Post by Forester »

CULater wrote: Mon Jun 15, 2020 12:03 pm Here's the thing about Gold. The longer you have an allocation to gold, the less benefit it provides. Historically, once you go beyond 3-5 years holding period you might as well have replaced your Gold allocation with an allocation to safe Treasuries because your portfolio returns would have been better.
I don't follow that... gold did great in the late 1970s, then required great patience in the 80s & 90s, then it paid off again. Two decades out of the last five, gold had a huge positive impact. Granted one of those decades involved the USA severing the link to convertible gold.

If not gold then a portfolio needs some exposure to hard assets (REITs, commodities, CTA commodity trend), something which is not government bonds. The 60/40 portfolio lost 4.5% a year , in real terms, 1973 to 1980. Permanent Portfolio, bolstered by gold did 3.2%. 60/40 was basically flat in the 2000s while the Permanent Portfolio achieved 3.91% (this time, helped by bonds as well as gold).
Amateur Self-Taught Senior Macro Strategist
User avatar
Lee_WSP
Posts: 10401
Joined: Fri Apr 19, 2019 5:15 pm
Location: Arizona

Re: What are your thoughts on the "Golden Butterfly" portfolio?

Post by Lee_WSP »

Lately gold seems to be highly correlated to the stock market. Just saying.
User avatar
CULater
Posts: 2832
Joined: Sun Nov 13, 2016 9:59 am
Location: Hic sunt dracones

Re: What are your thoughts on the "Golden Butterfly" portfolio?

Post by CULater »

For the longest period for which data is available using Portfolio Visualizer 1978-2020:

Portfolio 1: Golden Butterfly (20% TSM, 20% SCV, 20% Long Treasuries, 20% Short Treasuries, 20% Gold)
Portfolio 2: Golden Butterfly - No Gold (20% TSM, 20% SCV, 30% Long Treasuries, 30% Short Treasuries)

Portfolio, CAGR, SD, Largest Drawdown, Sharpe
Portfolio 1: 9.91%, 8.29%, -16.64%, 0.65
Portfolio 2: 9.96%, 7.69%, -16.49%, 0.70

As you can see, there was virtually no difference in compound return, annualized standard deviation, largest drawdown, and Sharpe Ratio if you had not held any Gold, substituting Treasuries for the 20% Gold allocation.

Once again, the historical data for long term portfolio returns shows that Gold added nothing to portfolio performance if you had allocated that portion instead to U.S. Treasuries with an intermediate term duration.

It's pretty hard to show that for long term buy-and-hold investors that Gold has provided any unique long term benefit.
On the internet, nobody knows you're a dog.
Horton
Posts: 1480
Joined: Mon Jan 21, 2008 2:53 pm
Contact:

Re: What are your thoughts on the "Golden Butterfly" portfolio?

Post by Horton »

CULater wrote: Mon Jun 15, 2020 5:16 pm For the longest period for which data is available using Portfolio Visualizer 1978-2020:

Portfolio 1: Golden Butterfly (20% TSM, 20% SCV, 20% Long Treasuries, 20% Short Treasuries, 20% Gold)
Portfolio 2: Golden Butterfly - No Gold (20% TSM, 20% SCV, 30% Long Treasuries, 30% Short Treasuries)

Portfolio, CAGR, SD, Largest Drawdown, Sharpe
Portfolio 1: 9.91%, 8.29%, -16.64%, 0.65
Portfolio 2: 9.96%, 7.69%, -16.49%, 0.70

As you can see, there was virtually no difference in compound return, annualized standard deviation, largest drawdown, and Sharpe Ratio if you had not held any Gold, substituting Treasuries for the 20% Gold allocation.

Once again, the historical data for long term portfolio returns shows that Gold added nothing to portfolio performance if you had allocated that portion instead to U.S. Treasuries with an intermediate term duration.

It's pretty hard to show that for long term buy-and-hold investors that Gold has provided any unique long term benefit.
Assuming the numbers are correct, your Portfolio 2 might be a new portfolio to track. The Nonmetallic Butterfly? The Achromatic Butterfly? Doesn’t quite have the same ring to it, though. :beer

PS - plan for lower returns in the future.
80% global equities (faith-based tilt) + 20% TIPS (LDI)
User avatar
CULater
Posts: 2832
Joined: Sun Nov 13, 2016 9:59 am
Location: Hic sunt dracones

Re: What are your thoughts on the "Golden Butterfly" portfolio?

Post by CULater »

Once again, we see that over short rolling time periods (1 - 3 years) the Golden Butterfly did produce larger maximum returns in one or more of those time periods compared to the Golden Butterfly without Gold. But as the rolling time period lengthens, it did not. Also, the average returns for rolling time periods of 1,3,5,7,10 and 15 years were all lower.

Golden Butterfly, Golden Butterfly - No Gold
Period, Avg Return, Largest Return, Avg Return, Largest Return
1 year: 10.31%, 38.92%, 10.33%, 32.24%
3 years: 9.76%, 22.08%, 10.17%, 19.00%
5 years: 9.61%, 18.18%, 10.23%, 20.72%
7 years: 9.53%, 16.63%, 10.18%, 17.50%
10 years: 9.46%, 15.20%, 9.99%, 15.62%
15 years: 9.33%, 13.41%, 9.87%, 14.32%

As in my previous post, historically the longer the time period the less likely that the allocation to Gold in the Golden Butterfly provided any meaningful benefit beyond simply investing the Gold portion in U.S. Treasuries instead.
On the internet, nobody knows you're a dog.
stuper1
Posts: 508
Joined: Tue Apr 03, 2018 9:30 am

Re: What are your thoughts on the "Golden Butterfly" portfolio?

Post by stuper1 »

CULater wrote: Mon Jun 15, 2020 5:16 pm For the longest period for which data is available using Portfolio Visualizer 1978-2020:

Portfolio 1: Golden Butterfly (20% TSM, 20% SCV, 20% Long Treasuries, 20% Short Treasuries, 20% Gold)
Portfolio 2: Golden Butterfly - No Gold (20% TSM, 20% SCV, 30% Long Treasuries, 30% Short Treasuries)

Portfolio, CAGR, SD, Largest Drawdown, Sharpe
Portfolio 1: 9.91%, 8.29%, -16.64%, 0.65
Portfolio 2: 9.96%, 7.69%, -16.49%, 0.70

As you can see, there was virtually no difference in compound return, annualized standard deviation, largest drawdown, and Sharpe Ratio if you had not held any Gold, substituting Treasuries for the 20% Gold allocation.

Once again, the historical data for long term portfolio returns shows that Gold added nothing to portfolio performance if you had allocated that portion instead to U.S. Treasuries with an intermediate term duration.

It's pretty hard to show that for long term buy-and-hold investors that Gold has provided any unique long term benefit.
Some of the charts at portfoliocharts.com suggest a different picture. The Rolling Returns chart shows that the Golden Butterfly has been much more consistent than your suggested alternative. This is because when stocks/bonds zig, gold often zags. The Withdrawal Rates chart calculates a perpetual withdrawal rate for the GB of 5.1%, while for your alternative it is 3.9%, again for the same reason.
A 10-20% allocation to gold has helped with the sequence of returns problem. Some gold held physically is also good insurance against the all-digital-assets problem.
Horton
Posts: 1480
Joined: Mon Jan 21, 2008 2:53 pm
Contact:

Re: What are your thoughts on the "Golden Butterfly" portfolio?

Post by Horton »

There's something different between Portfolio Visualizer and Portfolio Charts.

Portfolio Visualizer for Golden Butterfly vs CULater's AA.
80% global equities (faith-based tilt) + 20% TIPS (LDI)
User avatar
CULater
Posts: 2832
Joined: Sun Nov 13, 2016 9:59 am
Location: Hic sunt dracones

Re: What are your thoughts on the "Golden Butterfly" portfolio?

Post by CULater »

stuper1 wrote: Mon Jun 15, 2020 5:52 pm
CULater wrote: Mon Jun 15, 2020 5:16 pm For the longest period for which data is available using Portfolio Visualizer 1978-2020:

Portfolio 1: Golden Butterfly (20% TSM, 20% SCV, 20% Long Treasuries, 20% Short Treasuries, 20% Gold)
Portfolio 2: Golden Butterfly - No Gold (20% TSM, 20% SCV, 30% Long Treasuries, 30% Short Treasuries)

Portfolio, CAGR, SD, Largest Drawdown, Sharpe
Portfolio 1: 9.91%, 8.29%, -16.64%, 0.65
Portfolio 2: 9.96%, 7.69%, -16.49%, 0.70

As you can see, there was virtually no difference in compound return, annualized standard deviation, largest drawdown, and Sharpe Ratio if you had not held any Gold, substituting Treasuries for the 20% Gold allocation.

Once again, the historical data for long term portfolio returns shows that Gold added nothing to portfolio performance if you had allocated that portion instead to U.S. Treasuries with an intermediate term duration.

It's pretty hard to show that for long term buy-and-hold investors that Gold has provided any unique long term benefit.
Some of the charts at portfoliocharts.com suggest a different picture. The Rolling Returns chart shows that the Golden Butterfly has been much more consistent than your suggested alternative. This is because when stocks/bonds zig, gold often zags. The Withdrawal Rates chart calculates a perpetual withdrawal rate for the GB of 5.1%, while for your alternative it is 3.9%, again for the same reason.
I think the difference is attributable to the fact that Portfolio Charts data starts in 1970, while Portfolio Visualizer data starts in 1978. The entire argument for Gold depends on the high returns for Gold from 1972 - 1976, which is arguably a "one-off" event due to the unpegging of the US$ to Gold in 1972. Other factors to consider are that very few investors held gold back then because there were very few ways to own it and the costs associated with owning it and trading it were quite high and never accounted for in historical return data. I suggested to the author of Portfolio Charts for these reasons that an option to select a starting year other than 1970 be provided on the website but he disagreed. But, if you think that 1972-76 Gold returns will be repeated anytime soon, then that would be a reason to hold it. I guess it's not impossible, but I don't think I'd bet 20% of my portfolio on it. Like many assets that were previously illiquid and not widely held, the current price of Gold might not be likely to behave as it once did.
On the internet, nobody knows you're a dog.
User avatar
Tyler9000
Posts: 740
Joined: Fri Aug 21, 2015 11:57 am

Re: What are your thoughts on the "Golden Butterfly" portfolio?

Post by Tyler9000 »

CULater wrote: Mon Jun 15, 2020 7:44 pm I suggested to the author of Portfolio Charts for these reasons that an option to select a starting year other than 1970 be provided on the website but he disagreed.
I disagree because Portfolio Charts already studies every start date simultaneously rather than just the period starting in 1970. :wink: And for a full explanation, I address your concern about the gold data here: How to study portfolios when the data is full of bubbles.

But all data aside, I would never recommend that someone invest in something they simply don't trust. There are lots of good portfolios for all types of people, and I very much appreciate your perspective.
User avatar
Forester
Posts: 2400
Joined: Sat Jan 19, 2019 1:50 pm
Location: UK

Re: What are your thoughts on the "Golden Butterfly" portfolio?

Post by Forester »

Real rates presently negative and the next bond bear market might be worse than the last one; in that context making room for gold in that 50% to 20% "risk off" portion might at least be harmless.
Amateur Self-Taught Senior Macro Strategist
tigerdoc93
Posts: 255
Joined: Wed Oct 07, 2015 8:50 pm

Re: What are your thoughts on the "Golden Butterfly" portfolio?

Post by tigerdoc93 »

I’m certainly not an economist but I predict gold is in the early phases of a bull run. Only time will tell.
CDub
Posts: 79
Joined: Thu Jan 18, 2018 7:27 am

Re: What are your thoughts on the "Golden Butterfly" portfolio?

Post by CDub »

CULater wrote: Mon Jun 15, 2020 5:16 pm I think the difference is attributable to the fact that Portfolio Charts data starts in 1970, while Portfolio Visualizer data starts in 1978. The entire argument for Gold depends on the high returns for Gold from 1972 - 1976, which is arguably a "one-off" event due to the unpegging of the US$ to Gold in 1972. Other factors to consider are that very few investors held gold back then because there were very few ways to own it and the costs associated with owning it and trading it were quite high and never accounted for in historical return data. I suggested to the author of Portfolio Charts for these reasons that an option to select a starting year other than 1970 be provided on the website but he disagreed. But, if you think that 1972-76 Gold returns will be repeated anytime soon, then that would be a reason to hold it. I guess it's not impossible, but I don't think I'd bet 20% of my portfolio on it. Like many assets that were previously illiquid and not widely held, the current price of Gold might not be likely to behave as it once did.
If you compare them side by side - the results do look quite similar with the exception of the 1970s. This seems to prove both CULater and Tyler's points.

Since the early 1980s, they've performed very similarly. However, excluding the 70s when inflation was rampant poses it's own issues.

With interest rates having very little room to fall further, I have a suspicion the allocation to gold may be more useful than a higher allocation to treasuries (i.e. see start-date-sensitivity)

Image
Image
Image
Image
Image
User avatar
CULater
Posts: 2832
Joined: Sun Nov 13, 2016 9:59 am
Location: Hic sunt dracones

Re: What are your thoughts on the "Golden Butterfly" portfolio?

Post by CULater »

Tyler9000 wrote: Mon Jun 15, 2020 9:08 pm
CULater wrote: Mon Jun 15, 2020 7:44 pm I suggested to the author of Portfolio Charts for these reasons that an option to select a starting year other than 1970 be provided on the website but he disagreed.
I disagree because Portfolio Charts already studies every start date simultaneously rather than just the period starting in 1970. :wink: And for a full explanation, I address your concern about the gold data here: How to study portfolios when the data is full of bubbles.

But all data aside, I would never recommend that someone invest in something they simply don't trust. There are lots of good portfolios for all types of people, and I very much appreciate your perspective.
Thoughtful piece. I do admire and appreciate the great work you have done with Portfolio Charts and your insights. To make the contrary argument about Gold, I feel the concept of "bubble" is a little slippery. I think you can make a case for a "bubble" in gold price in the early 70s based on a very pronounced and relatively short-term spike in the price of gold that seems largely attributable to a specific "one-off" event: the decoupling of gold from the US$. I don't know that I'd label the run-up in bond prices since 1982 as a "bubble" as much as a secular trend; the same for the runup of stock prices from 1982-2000.

Also, we have the practical matter that gold was not a very tradeable asset back in the 1970s and it has become one since the early 2000s due to gold ETFs. Stocks and U.S. treasuries were liquid and tradeable assets. The impact of that fundamental change in the asset characteristics of gold is an important factor. In addition, nobody includes the costs of owning and trading Gold back in the 1970s in calculations of portfolio returns. The spreads, commissions, cost of storage were very significant. Another factor is that it was actually illegal for individual U.S. investors to own gold bullion until 1975. All in all, I feel that the price of gold during the early 1970s is anomalous and misleading because of these several factors, and I never include that period in my backtesting calculations.

And if you exclude just the years 1972-1976 then Gold becomes an unremarkable investment asset. This disregards the fact that it's hard to find a living human who dedicated 20% - 25% of their portfolio to an asset that lost about 4% annually for two decades through the 1980s and 1990s, while experiencing a drawdown at one point during that period of 57%. If you invest in Gold then "stay the course" could take on a whole new definition.

Even though I'm a skeptic about the long-term benefits of Gold, I'm not an atheist. I do have a small allocation. Hard to break the grip of the Yellow Dog based on logic alone. Can't explain fully why I don't take my own advice, other than the fact that I'm subject to the same short-term thinking that everybody else is.
On the internet, nobody knows you're a dog.
User avatar
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 2:17 pm
Location: USA
Contact:

Re: What are your thoughts on the "Golden Butterfly" portfolio?

Post by willthrill81 »

CDub wrote: Tue Jun 16, 2020 8:24 am
CULater wrote: Mon Jun 15, 2020 5:16 pm I think the difference is attributable to the fact that Portfolio Charts data starts in 1970, while Portfolio Visualizer data starts in 1978. The entire argument for Gold depends on the high returns for Gold from 1972 - 1976, which is arguably a "one-off" event due to the unpegging of the US$ to Gold in 1972. Other factors to consider are that very few investors held gold back then because there were very few ways to own it and the costs associated with owning it and trading it were quite high and never accounted for in historical return data. I suggested to the author of Portfolio Charts for these reasons that an option to select a starting year other than 1970 be provided on the website but he disagreed. But, if you think that 1972-76 Gold returns will be repeated anytime soon, then that would be a reason to hold it. I guess it's not impossible, but I don't think I'd bet 20% of my portfolio on it. Like many assets that were previously illiquid and not widely held, the current price of Gold might not be likely to behave as it once did.
If you compare them side by side - the results do look quite similar with the exception of the 1970s. This seems to prove both CULater and Tyler's points.

Since the early 1980s, they've performed very similarly. However, excluding the 70s when inflation was rampant poses it's own issues.

With interest rates having very little room to fall further, I have a suspicion the allocation to gold may be more useful than a higher allocation to treasuries (i.e. see start-date-sensitivity)

Image
Image
Image
Image
Image
It's clear to me at least that the GB had much more consistent performance, even excluding the 1970s, than the portfolio with only Treasuries.

Heat maps are very good for examining the impact of starting dates. Examining the impact of even a relatively small allocation to gold in portfolios otherwise comprised of stocks and bonds has convinced me that gold's stabilizing effect was not at all limited to the 1970s.

That said, there's no denying that the portfolio with only Treasuries did pretty well outside of the 1970s, but the argument that bonds are very unlikely to repeat their last 40 years' of performance over the next 40 years is more persuasive to me than arguments that gold is 'just a shiny metal that's of no value to long-term investors'.
The Sensible Steward
User avatar
CULater
Posts: 2832
Joined: Sun Nov 13, 2016 9:59 am
Location: Hic sunt dracones

Re: What are your thoughts on the "Golden Butterfly" portfolio?

Post by CULater »

Once again, it would be instructive to see the Portfolio Charts data run from a starting point of 1976 to 1978. Clearly, the inflationary 1970s were damaging to both stocks and nominal bonds, while Gold price did well. As I pointed out, when you look at returns from 1978, Gold has been unremarkable. If you look at the period 1972-77, it would have helped a lot - but we don't know if that was just hypothetical because very few investors could have owned or traded it prior to 1975 in sufficient amounts to have benefitted and backtesting results are based on price, not net price taking into consideration costs. And it was illegal for U.S. citizens to even own Gold until 1975, had they the means and determination to invest in it.

One other factor to consider is that most investors are not "lump sum" investors who can put all of their capital to work immediately. Using Portfolio Visualizer, I was able to compare an approximation of the Golden Butterfly (GB) and Non-Golden Golden Butterfly (NGGB) for the period 1972-Present. This was done by using Intermediate Treasuries instead of Short and Long Treasuries, as follows:

GB: 20% TSM, 20% SCV, 40% ITT, 20% Gold
NGGB: 20% TSM, 20% SCV, 60% ITT

If we lump-sum at the beginning of 1972, it's clear that the GB does better than the NGGB over the entire period in compound annual return, but not in risk-adjusted return or drawdowns:

CAGR = 10.08 vs. 9.40
SD = 8.35 vs. 7.57
Max Drawdown = 16.9% vs. 16.3%
Sharpe = 0.64 vs. 0.62

But if we are not able to lump-sum, and make periodic annual contributions, there is no difference in compound annual return over the entire period, while the NGGB portfolio has lower annualized volatility:

MWRR = 9.04% vs. 9.04%
SD = 8.35 vs. 7.57
Max DD = 16.09 vs. 16.3
Sharpe = 0.64 vs. 0.62

*MWRR = internal rate of return accounting for cashflows.

In other words, even if the Gold return data were actually obtainable by investors in the early 1970s and they had invested in the Golden Butterfly since then, 99% of retail investors (who must invest over time) would not have realized any benefit at all from investing in the GB vs. the Non-golden GB. So, even including the most favorable period for Gold price (early 1970s) the data show no practical advantage for real-world investors in the GB for the last nearly 5 decades.
On the internet, nobody knows you're a dog.
User avatar
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 2:17 pm
Location: USA
Contact:

Re: What are your thoughts on the "Golden Butterfly" portfolio?

Post by willthrill81 »

CULater wrote: Tue Jun 16, 2020 10:24 am Once again, it would be instructive to see the Portfolio Charts data run from a starting point of 1976 to 1978.
The heat map on that site shows all starting years since 1970.
The Sensible Steward
User avatar
CULater
Posts: 2832
Joined: Sun Nov 13, 2016 9:59 am
Location: Hic sunt dracones

Re: What are your thoughts on the "Golden Butterfly" portfolio?

Post by CULater »

willthrill81 wrote: Tue Jun 16, 2020 10:44 am
CULater wrote: Tue Jun 16, 2020 10:24 am Once again, it would be instructive to see the Portfolio Charts data run from a starting point of 1976 to 1978.
The heat map on that site shows all starting years since 1970.
Like to see the complete suite of backtesting results.
On the internet, nobody knows you're a dog.
stuper1
Posts: 508
Joined: Tue Apr 03, 2018 9:30 am

Re: What are your thoughts on the "Golden Butterfly" portfolio?

Post by stuper1 »

Also use the Rolling Returns chart with it set to 5 or more years of average returns. The GB was much better during the 2000s than the non-gold alternative. That was 30 years past the 1970s.
A 10-20% allocation to gold has helped with the sequence of returns problem. Some gold held physically is also good insurance against the all-digital-assets problem.
User avatar
Lee_WSP
Posts: 10401
Joined: Fri Apr 19, 2019 5:15 pm
Location: Arizona

Re: What are your thoughts on the "Golden Butterfly" portfolio?

Post by Lee_WSP »

Wouldn't the non preferential tax treatment make gold significantly less worthwhile in a taxable account?
User avatar
CULater
Posts: 2832
Joined: Sun Nov 13, 2016 9:59 am
Location: Hic sunt dracones

Re: What are your thoughts on the "Golden Butterfly" portfolio?

Post by CULater »

stuper1 wrote: Tue Jun 16, 2020 11:23 am Also use the Rolling Returns chart with it set to 5 or more years of average returns. The GB was much better during the 2000s than the non-gold alternative. That was 30 years past the 1970s.
Yes, that's true. If I start in 2000 then the rolling returns for 1,3,5,7,10 and 15 years, as well as the highest period returns, favor the GB over the Non-Gold GB. Gold had a run from 1972-1981 and another run from 2002-2011, both periods about a decade long. It was dead money for two decades from 1981-2001. If you were fortunate enough to invest a lump sum in a portfolio that includes a significant allocation to Gold at the start of a run, you were better off than if you had avoided Gold. But isn't this cherry-picking the periods when it would have been beneficial to own gold?

If that's allowed, I think I can find some pretty good portfolios that didn't include gold. What I need to know is how gold will do in the future. Is now the start of a run and the right time to have a 20% - 25% gold allocation, or should I wait until the next decade?
On the internet, nobody knows you're a dog.
stuper1
Posts: 508
Joined: Tue Apr 03, 2018 9:30 am

Re: What are your thoughts on the "Golden Butterfly" portfolio?

Post by stuper1 »

How do the rolling returns look if you start in 1981?

If you use the Rolling Returns chart on portfoliocharts.com, you can see the X-year rolling returns starting in each year since 1970 on a single chart, where you get to pick X. To me the GB looks much more consistent than the non-gold GB. I guess we each get to make our guesses about what the future will look like. It makes sense to me that when stocks/bonds go out of favor at the same time, which does happen on occasion, that then some of that money will go into gold.
A 10-20% allocation to gold has helped with the sequence of returns problem. Some gold held physically is also good insurance against the all-digital-assets problem.
User avatar
Forester
Posts: 2400
Joined: Sat Jan 19, 2019 1:50 pm
Location: UK

Re: What are your thoughts on the "Golden Butterfly" portfolio?

Post by Forester »

A two-part article on the Permanent Portfolio worth reading... the classic boring Harry Browne version holds up versus the "improvements" IMHO;

https://investresolve.com/blog/permanen ... wn-part-1/
https://investresolve.com/blog/permanen ... wn-part-2/
https://investresolve.com/blog/the-perm ... -japanese/
Amateur Self-Taught Senior Macro Strategist
JBTX
Posts: 11227
Joined: Wed Jul 26, 2017 12:46 pm

Re: What are your thoughts on the "Golden Butterfly" portfolio?

Post by JBTX »

I really don't like gold as an investment, for many of the reasons Buffet articulates. But I held my nose and bought some GLD recently, and a less quantity of SLV. I'm having a hard time justifying a robust weighting in long term bonds when they are earning practically zero in nominal terms. There's a lot of money being injected into the system, and assets like gold could benefit.
NoRegret
Posts: 505
Joined: Sat Dec 16, 2017 1:00 am
Location: California

Re: What are your thoughts on the "Golden Butterfly" portfolio?

Post by NoRegret »

CULater wrote: Mon Jun 15, 2020 5:16 pm For the longest period for which data is available using Portfolio Visualizer 1978-2020:

Portfolio 1: Golden Butterfly (20% TSM, 20% SCV, 20% Long Treasuries, 20% Short Treasuries, 20% Gold)
Portfolio 2: Golden Butterfly - No Gold (20% TSM, 20% SCV, 30% Long Treasuries, 30% Short Treasuries)

Portfolio, CAGR, SD, Largest Drawdown, Sharpe
Portfolio 1: 9.91%, 8.29%, -16.64%, 0.65
Portfolio 2: 9.96%, 7.69%, -16.49%, 0.70

As you can see, there was virtually no difference in compound return, annualized standard deviation, largest drawdown, and Sharpe Ratio if you had not held any Gold, substituting Treasuries for the 20% Gold allocation.

Once again, the historical data for long term portfolio returns shows that Gold added nothing to portfolio performance if you had allocated that portion instead to U.S. Treasuries with an intermediate term duration.

It's pretty hard to show that for long term buy-and-hold investors that Gold has provided any unique long term benefit.
If the predicate is similar past performance and by implication an analogous future then I choose the portfolio with the greater number of uncorrelated assets, since it will be more robust going forward. The simpler portfolio — the one with fewer components, will be at a greater risk from a correlation regime change. The simpler portfolio is also at greater danger from one component unexpectedly underperforming. I can make this choice from first principles — without knowing what the actual components are.

Now knowing what they are, in terms of correlations I have reasonable confidence gold will continue to have low correlation to equities. As for equities/bonds we’ll have to see if or when inflation comes. The historical precedent is not necessarily favorable.

The above should be sufficient reason from a portfolio construction point of view. I also have my own subjective return expectations that others may or may not share.
Market timer targeting long term cycles -- aiming for several key decisions per asset class per decade
thermo
Posts: 41
Joined: Sat May 30, 2020 10:32 pm

Re: What are your thoughts on the "Golden Butterfly" portfolio?

Post by thermo »

Tyler9000 wrote: Mon Jun 15, 2020 9:08 pm
CULater wrote: Mon Jun 15, 2020 7:44 pm I suggested to the author of Portfolio Charts for these reasons that an option to select a starting year other than 1970 be provided on the website but he disagreed.
I disagree because Portfolio Charts already studies every start date simultaneously rather than just the period starting in 1970. :wink: And for a full explanation, I address your concern about the gold data here: How to study portfolios when the data is full of bubbles.

But all data aside, I would never recommend that someone invest in something they simply don't trust. There are lots of good portfolios for all types of people, and I very much appreciate your perspective.
That's a great article. Arguably every asset class is currently in a bubble, stocks, bonds, gold and real estate. Is there any way to model or visualise what returns would be like if all the bubbles burst simultaneously?
User avatar
CULater
Posts: 2832
Joined: Sun Nov 13, 2016 9:59 am
Location: Hic sunt dracones

Re: What are your thoughts on the "Golden Butterfly" portfolio?

Post by CULater »

NoRegret wrote: Tue Jun 16, 2020 11:44 pm
CULater wrote: Mon Jun 15, 2020 5:16 pm For the longest period for which data is available using Portfolio Visualizer 1978-2020:

Portfolio 1: Golden Butterfly (20% TSM, 20% SCV, 20% Long Treasuries, 20% Short Treasuries, 20% Gold)
Portfolio 2: Golden Butterfly - No Gold (20% TSM, 20% SCV, 30% Long Treasuries, 30% Short Treasuries)

Portfolio, CAGR, SD, Largest Drawdown, Sharpe
Portfolio 1: 9.91%, 8.29%, -16.64%, 0.65
Portfolio 2: 9.96%, 7.69%, -16.49%, 0.70

As you can see, there was virtually no difference in compound return, annualized standard deviation, largest drawdown, and Sharpe Ratio if you had not held any Gold, substituting Treasuries for the 20% Gold allocation.

Once again, the historical data for long term portfolio returns shows that Gold added nothing to portfolio performance if you had allocated that portion instead to U.S. Treasuries with an intermediate term duration.

It's pretty hard to show that for long term buy-and-hold investors that Gold has provided any unique long term benefit.
If the predicate is similar past performance and by implication an analogous future then I choose the portfolio with the greater number of uncorrelated assets, since it will be more robust going forward. The simpler portfolio — the one with fewer components, will be at a greater risk from a correlation regime change. The simpler portfolio is also at greater danger from one component unexpectedly underperforming. I can make this choice from first principles — without knowing what the actual components are.

Now knowing what they are, in terms of correlations I have reasonable confidence gold will continue to have low correlation to equities. As for equities/bonds we’ll have to see if or when inflation comes. The historical precedent is not necessarily favorable.

The above should be sufficient reason from a portfolio construction point of view. I also have my own subjective return expectations that others may or may not share.
If you haven't read it, Dr. Bernstein's article on the Permanent Portfolio is worth a look - "Wild About Harry". He points out that, while the correlations look inviting, investors holding portfolios with a large Gold allocation are subject to the behavioral risk of abandoning the allocation at some point because of it's historically large tracking error and few have the right stuff to stick with it over the long run. The most important thing in investing is Investor Know Thyself.

http://www.efficientfrontier.com/ef/0adhoc/harry.htm
In many respects, this allocation is a thing of beauty. Not only does it provide some protection against all but the most dire of scenarios, but its correlation grid is one rarely seen in finance: four non-derivative assets populated entirely by near-zeros:

And therein lies the real problem with the TPP: because of its huge tracking error relative to more conventional portfolios, it attracts assets and adherents during crises, then sheds them in better times. Theres nothing wrong with Harrys portfolio nothing at all but theres everything wrong with his followers, who seem, on average, to chase performance the way dogs chase cars.

Investment success accrues not so much to the brilliant as to the disciplined, and the nature of the chosen strategy contributes mightily to this calculus. The very worst place an investor can find herself is, in the words of Mark Kritzman, "wrong and alone"; this is a near certainty at some point given the TPPs huge tracking error relative to that of the overall market portfolio, approximated by a 60/40 mix of stocks and bonds. Thus, it will be nigh-impossible for even the most disciplined investors to adhere to the TPP in the long run. (And lord knows, most investors are unable to stick to even a 60/40 portfolio.)
On the internet, nobody knows you're a dog.
ignition
Posts: 366
Joined: Sun Dec 11, 2016 10:28 am

Re: What are your thoughts on the "Golden Butterfly" portfolio?

Post by ignition »

Tyler9000 wrote: Mon Jun 15, 2020 9:08 pm I disagree because Portfolio Charts already studies every start date simultaneously rather than just the period starting in 1970. :wink: And for a full explanation, I address your concern about the gold data here: How to study portfolios when the data is full of bubbles.
It seems a bit strange to call 1982 to 2000 a stock market bubble. Maybe 1996-2000 was a bubble but prices before 1996 didn't seem very "bubbly".
CDub
Posts: 79
Joined: Thu Jan 18, 2018 7:27 am

Re: What are your thoughts on the "Golden Butterfly" portfolio?

Post by CDub »

Any benefit to splitting the STT allocation between STT and TIPS? Results in Portfoliovisualizer seem about the same, but this is only since 2001. I'm assuming results won't change much but just curious.

https://www.portfoliovisualizer.com/bac ... GlobalBond
User avatar
Spinola
Posts: 328
Joined: Fri Nov 16, 2018 12:32 pm

Re: What are your thoughts on the "Golden Butterfly" portfolio?

Post by Spinola »

tigerdoc93 wrote: Tue Jun 16, 2020 5:28 am I’m certainly not an economist but I predict gold is in the early phases of a bull run. Only time will tell.
Seems likely. Especially after the FOMA/Crypto crowd gets burned when that vaporware-ponzi scheme collapses or gets regulated into oblivion or if rampant inflation becomes a reality.. :shock:
"If more of us valued food and cheer and song above hoarded gold, it would be a merrier world." | Thorin Oakenshield
User avatar
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 2:17 pm
Location: USA
Contact:

Re: What are your thoughts on the "Golden Butterfly" portfolio?

Post by willthrill81 »

The GB is up 3.7% YTD.

Steady as she goes, which is what this portfolio seems to do pretty darn well.
The Sensible Steward
GaryA505
Posts: 2907
Joined: Wed Feb 08, 2017 1:59 pm
Location: New Mexico

Re: What are your thoughts on the "Golden Butterfly" portfolio?

Post by GaryA505 »

Has anyone considered replacing SCV with ex-US international stock. I'm going to call it the "World Golden Butterfly" or WGB.

20% TSM
20% TISM
20% cash/STT
20% LTT
20% gold

This is basically 40/40/20 stock/bonds/gold (as is Tyler's GB).
The 40% stock could be divided up however you like, such as 80% TSM and 20% TISM like we see preferred by many on this forum.
The 40% bonds could also be divided up as well, using cash/STT, ITT and LTT to fit your preferred duration.

You may have noticed that both the GB and WGB are really just a 50/50 split between stocks and bonds, with a considerable chunk of gold. So, maybe we should call it the "Golden Couch Potato" instead.
Get most of it right and don't make any big mistakes. All else being equal, simpler is better. Simple is as simple does.
User avatar
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 2:17 pm
Location: USA
Contact:

Re: What are your thoughts on the "Golden Butterfly" portfolio?

Post by willthrill81 »

GaryA505 wrote: Sun Aug 02, 2020 2:54 pm Has anyone considered replacing SCV with ex-US international stock.
Giving that the purpose of SCV is to add exposure to small and value companies, ex-U.S. stock is a poor substitute since it is dominated by large-cap blend companies. Ex-U.S. SCV would be a much better option, and a few funds (beyond DFA's) are now available that provide access to that asset class, such as VFSVX.
The Sensible Steward
seajay
Posts: 1656
Joined: Sat May 01, 2021 3:26 pm
Contact:

Re: What are your thoughts on the "Golden Butterfly" portfolio?

Post by seajay »

Tyler9000 wrote: Mon Dec 11, 2017 4:54 pm
EarlTheGreySnapper wrote: Mon Dec 11, 2017 8:41 am Living in the UK (as a British citizen with no plans to move abroad any time soon), what are the downsides (if any) to investing in a US portfolio like this?
If you're building a Golden Butterfly for the UK or any other smaller market, splitting the stocks between half local and half All-World might be a reasonable compromise. I could also see the argument for diversifying into a good unhedged international bond fund of the appropriate average maturity if one is available, although in the UK I'd probably just stick with gilts. But I would generally steer clear of investing all of your money in a foreign market, as completely detaching the engine of your returns from the inflation and currency of your home country may have side effects that are difficult to anticipate.
UK FTSE 250 midcaps (iShares MIDD) is small cap in US scale, and somewhat value'ish. It's also more reflective of the UK, around 50% of earnings sourced from foreign, whilst the FTSE 100 (largest 100 stocks by market cap) have around 70% of earnings sourced from foreign.

Combined with US S&P500 (CSP1) as the non-PP stock holding and since 2016 (Brexit referendum) ...

Code: Select all

+─────────────────+─────────+─────────+─────────+─────────+─────────+─────────+
| Calendar years  | 2015    | 2016    | 2017    | 2018    | 2019    | 2020    |
+─────────────────+─────────+─────────+─────────+─────────+─────────+─────────+
| IGLT TR         |         | 9.97    | 1.7     | 0.4     | 6.8     | 8.2     |
| IGLS TR         |         | 2.44    | -0.39   | 0.18    | 1.01    | 1.48    |
| MIDD TR         |         | 6.16    | 17.34   | -13.62  | 28.36   | -4.99   |
| SGLN $          |         | 8.85    | 11.58   | -1.39   | 18.53   | 23.95   |
| CSP1 $          |         | 11.54   | 21.4    | -4.72   | 31.02   | 18.02   |
| Year end £$     | 1.4746  | 1.2337  | 1.3529  | 1.2763  | 1.3269  | 1.3662  |
| SGLN £          |         | 30.10   | 1.75    | 4.53    | 14.01   | 20.38   |
| CSP1 £          |         | 33.32   | 10.70   | 1.00    | 26.02   | 14.63   |
|                 |         |         |         |         |         |         |
| GB (UK £)       |         | 16.40   | 6.22    | -1.50   | 15.24   | 7.94    |
+─────────────────+─────────+─────────+─────────+─────────+─────────+─────────+
Sourced from iShares https://www.ishares.com/uk/individual/e ... ckerSearch except Pound/Dollar rates that were sourced from FRED

Many have been saying that long dated Treasury (Gilts) have only one way to go for years now, despite which IGLT has done OK. Short dated Treasury (IGLS) - low rewards has been the norm for years now. Gold (SGLN) has had a few good years in the last 5 years. 51.5% 5 year nominal gain, 8.7% annualised nominal gain. All values are total return (TR).

Physical gold/Gold fund wise, using data from 1939 with a 4% SWR applied to 30 year periods calendar year granularity and the worst run had 48% of the inflation adjusted start date value still remaining at the end of the 30 years (median run had 97%). That 48% was also the lowest low (yearly granularity). So if gold is 1/5th then 10% physical gold would likely have remained untouched, leaving 10% allocated to more liquid/tighter spread gold fund (ETF).

For a 3% SWR the lowest 30 year ended with 98% of the inflation adjusted start date amount available at the end of 30 years, lowest dip was down to 73%, suggestive of 15% physical gold unlikely being touched (so 5% gold fund/ETF). Median case ended with 164% of the inflation adjusted start date portfolio value at the end of the 30 years.

Personally given relatively wide physical gold spreads, it can feel more comfortable to defer purchasing physical and initially just hold gold funds until such times that good portfolio gains occur in a year and then use some of that 'other peoples money' to fund swapping some of gold funds over to physical gold. 5% physical gold premium above spot, 10% of total portfolio to be moved into physical gold, 0.5% cost relative to the total portfolio value ... so count 2016 as a 15.9% gain year instead of 16.4%. Whilst that's just mental trickery it's a trick that made things feel for comfortable for me personally.

Presently with ii brokerage, but contemplating opening several Freetrade accounts (multiple family members) and all of the above iShares funds are listed as being available in their free/general accounts https://freetrade.io/etfs. Since ii took over from TD Waterhouse multiple monthly account fees whilst relatively small in the scale of things just doesn't feel right when it was free before.
seajay
Posts: 1656
Joined: Sat May 01, 2021 3:26 pm
Contact:

Re: What are your thoughts on the "Golden Butterfly" portfolio?

Post by seajay »

CULater wrote: Wed Jun 17, 2020 7:04 am If you haven't read it, Dr. Bernstein's article on the Permanent Portfolio is worth a look - "Wild About Harry". He points out that, while the correlations look inviting, investors holding portfolios with a large Gold allocation are subject to the behavioral risk of abandoning the allocation at some point because of ...
Recently its much the same for long dated treasury holdings. Trouble is that's then market timing ... when do you re-enter holding gold or long dated treasury bonds. If you just persist come what may then at least rebalancing will have you partake of the good years as/when they might occur for those assets rather than having missed those good years because you were out of the asset altogether. The PP design in part factors in that losses in one asset are inclined to be offset by gains in other(s). Nature of the beast.

With gold across the 1980's/1990's whilst it was pretty much a case of repeated reducing stock shares to add to gold only to see stocks continue up and gold continue down, you did see the number of ounces of gold being held expanded by multiples, so somewhat felt like gold was earning dividends when you opened a safe to see several more stacks of coins having been accumulated.
Post Reply