PSA: Read Tyler of Portfolio Charts' recent blog post on portfolio construction?

[SeattleCPA]

Wow, @Tyler , great job. I say that as both a writer and a finance guy. Super insightful. Well worth the time to read and then reread a time or two.

TLDR summary (and quoting Tyler), "Combinations of volatile complementary assets can have much more desirable performance than any one asset by itself."

https://portfoliocharts.com/2021/12/16/three-secret-ingredients-of-the-most-efficient-portfolios/

P.S. The blog post particularly valuable for folks thinking 100% stocks is optimal.

[Imanuels]

I'm wondering about two points:
1. There's historical evidence that momentum provides better returns. Yet, in this case SCV stands out.
2. The importance of gold in most efficient portfolios. By looking at how poorly gold itself has performed for several periods, I'm wondering if this could be further adjusted by some other metric that avoids buying gold at the peak and thus sitting on an asset with negative return. See the attached image, at least visually the similarity to the previous peak looks scary.

[vand]

I'm wondering about two points:
1. There's historical evidence that momentum provides better returns. Yet, in this case SCV stands out.
2. The importance of gold in most efficient portfolios. By looking at how poorly gold itself has performed for several periods, I'm wondering if this could be further adjusted by some other metric that avoids buying gold at the peak and thus sitting on an asset with negative return. See the attached image, at least visually the similarity to the previous peak looks scary.

There are plenty of backtested portfolios that introduce trend following rules, eg only hold something that is above its 200 day moving average.

One of the best examples is Meb Faber's Trinity portfolio, all the backtests suggested that it would produce superior long term results with less volatility than a buy and hold multi-asset approach. However since he launched it (TRTY) the performance has been pretty lacklustre. Maybe that's unfair because it hasn't been launched for that long, but it just shows that backtested approaches that introduce too many rules tend to perform worse going forward than they have done so historically.

agree that it's another excellent article from PortfolioCharts. Tyler really deserves a commendation for his services to the DIY investor.

[SeattleCPA]

...agree that it's another excellent article from PortfolioCharts. Tyler really deserves a commendation for his services to the DIY investor...

Yup. It's really well-crafted work... Again, thank you @Tyler ... :-)

P.S. Been thinking ever since I reread it that I'd like to think seriously about tweaking the Swensen formula I've been using for insights Tyler highlights... It really would not be too hard. E.g., use LT treasuries in place of IT treasuries... Peel off 5% from the REIT percentage and TIPS percentage, reallocate to Gold... split the 30% in the US Stocks allocation to large cap blend and small cap value... voila...

[Tyler]

I'm glad you found the data as interesting as I did. :) This is a topic I've been stewing on for a while but kept on the back burner because of the work involved. So it feels pretty good to finally be able to put some solid numbers and words to it.

[Wintergreen78]

Tyler,

I really appreciate the work you put into these topics and how well you present everything. I’ve found much of your writing thought-provoking and helpful as I make decisions.

Personally, my equity funds are a mix of total market US and total market worldwide. One thing that I’ve always gotten hung up on when looking at small cap value is the turnover rate and therefore the tax drag for funds held in taxable accounts. I just looked and saw that the fidelity SCV index fund had a 60% turnover last year, compared to 2% for their SP 500 fund. You do have me thinking about this more, and whether it makes sense to move some of my equity funds in my Roth IRA to SCV, as well as re-thinking my stock/bond balance and my bond choices.

What do you think about tax impacts on performance, and how that would impact either fund choices, or which accounts are better suited to which investments?

[SeattleCPA]

What do you think about tax impacts on performance, and how that would impact either fund choices, or which accounts are better suited to which investments?

You're not asking me... but I'm going to jump in here and give an answer anyway since I'm a tax guy.

I think you try to put bonds into tax deferred because the interest is taxed an ordinary income if in taxable.

You try to put the US total market stocks or if you go with Tyler's formulation the large cap blend into your taxable because there the qualified dividends and LT capital gains get treated preferentially (maybe even at a zero rate when you're retired?)

I have not thought a ton about this... but I think you put gold into a taxable account (which means a safe deposit box)...

And ideally everything else (so small cap and REITs and International) you put into a tax deferred if you can... If you can't make that work maybe you put some international into the taxable space and get the foreign tax credit and then deal with the Form 1116s that'll end up on your 1040 return.

The other thing to think about is how you rebalance if you're holding lots of money in taxable accounts because you want to do that only in tax-deferred. As much as possible.

BTW, the Swensen asset allocation (updated for his bump up in emerging markets which he did because they become bigger chunk of market after his book was published) is 30% US stocks... 15% IT treasuries, 15% TIPS, 15% developed market, 15% REITs... and 10% emerging.

What I'm considering a Tyler-ized Swensen might look something like this: 20% Large Cap US stocks, 15% LT treasuries, 15% developed markets, 10% TIPS, 10% REITs, 10% small cap value, and 10% gold. This formulation retains 30% in real assets and close to the Swensen's percentages in riskless assets (25% instead of 30%) and equity (65% instead of 70%).

[Tyler]

What do you think about tax impacts on performance, and how that would impact either fund choices, or which accounts are better suited to which investments?

SeattleCPA is far more qualified than I am to talk about tax optimization. But here's my 2-cents.

I generally don't worry so much about turnover in passive index funds because it's the index I care about most. A small cap fund will always have more turnover than a large cap fund because they have to do more work maintaining the right cross-section of companies (like when a company out-grows the index, which never happens in a large cap fund). Some indices have different trading band rules to reduce turnover, but it will never be zero.

That said, I do generally prefer ETFs over index mutual funds because the ETFs are structured a little differently and almost never realize internal capital gains. So they're usually a little more tax efficient.

When it comes to where to store each asset, there are two schools of thought. One (as SeattleCPA describes) is to find the most tax efficient account type for each fund based on the tax burden of each asset. That certainly has a lot of merit, and his recommendations are solid. The other (which I personally follow) is to maximize tax flexibility by holding the same full portfolio in both taxable and tax-deferred accounts. I like it because:

1) It's way easier to rebalance
2) It opens up more opportunities for things like tax-loss harvesting
3) I want the taxable money that I have easy access to in ER to have the full portfolio characteristics I care about. And it's hard to do that with only a partial recipe.

But the best choice depends on your personal situation, so obviously YMMV.

And regardless of where you keep assets, the simplest and most effective tax minimization strategy during accumulation is to always top off the lagging asset with new money. Keep that up, and you may go many years without having to realize any capital gains to rebalance.

BTW, the Swensen asset allocation (updated for his bump up in emerging markets which he did because they become bigger chunk of market after his book was published) is 30% US stocks... 15% IT treasuries, 15% TIPS, 15% developed market, 15% REITs... and 10% emerging.

It's funny -- I added a note about that on the Swensen Portfolio page just yesterday after someone else brought it to my attention. Do you have a good reference that explains Swensen's thought process in the change? I'd love to read it.

[SeattleCPA]

Do you have a good reference that explains Swensen's thought process in the change? I'd love to read it.

See the bottom of the blue sidebar on this article in Yale Magazine: https://yalealumnimagazine.com/articles/2398-david-swensen-s-guide-to-sleeping-soundly

"Today, Swensen says, economic conditions might call for a modest revision. He now recommends that investors have 15 percent of their assets in real estate investment trusts, and raise their investment in emerging-market stock funds to 10 percent."

[Tyler]

See the bottom of the blue sidebar on this article in Yale Magazine: https://yalealumnimagazine.com/articles/2398-david-swensen-s-guide-to-sleeping-soundly

Thanks. I've added a link on the Swensen page for posterity.

[Chris Pascale]

Posting just to bump it up for you.

Great job.

[uniwelder]

Also posting because I like Tyler’s site and analysis. Thank you!

[vand]

I'm wondering about two points:
1. There's historical evidence that momentum provides better returns. Yet, in this case SCV stands out.
2. The importance of gold in most efficient portfolios. By looking at how poorly gold itself has performed for several periods, I'm wondering if this could be further adjusted by some other metric that avoids buying gold at the peak and thus sitting on an asset with negative return. See the attached image, at least visually the similarity to the previous peak looks scary.

There are plenty of backtested portfolios that introduce trend following rules, eg only hold something that is above its 200 day moving average.

One of the best examples is Meb Faber's Trinity portfolio, all the backtests suggested that it would produce superior long term results with less volatility than a buy and hold multi-asset approach. However since he launched it (TRTY) the performance has been pretty lacklustre. Maybe that's unfair because it hasn't been launched for that long, but it just shows that backtested approaches that introduce too many rules tend to perform worse going forward than they have done so historically.

It's also a bit unfair to single out gold as an asset class you particularly want to avoid when it's going down.  All asset classes have suffered catastrophic losses at some point in the last century. 
(from https://mebfaber.com/2020/03/05/the-stay-rich-portfolio/)

Largest real drawdown since 1926:

US Stocks:  -79%
Foreign Stocks:  -78%
10 Year US Bonds: -61%
Foreign Bonds:  -78%
Gold:  -85%

What is more likely to cripple your portfolio: having 60-70% of your portfolio in an asset class that has gone down -79%, or 10% of your portfolio in an asset class that has gone down -85%?

[Wintergreen78]

What do you think about tax impacts on performance, and how that would impact either fund choices, or which accounts are better suited to which investments?

SeattleCPA is far more qualified than I am to talk about tax optimization. But here's my 2-cents.

I generally don't worry so much about turnover in passive index funds because it's the index I care about most. A small cap fund will always have more turnover than a large cap fund because they have to do more work maintaining the right cross-section of companies (like when a company out-grows the index, which never happens in a large cap fund). Some indices have different trading band rules to reduce turnover, but it will never be zero.

That said, I do generally prefer ETFs over index mutual funds because the ETFs are structured a little differently and almost never realize internal capital gains. So they're usually a little more tax efficient.

When it comes to where to store each asset, there are two schools of thought. One (as SeattleCPA describes) is to find the most tax efficient account type for each fund based on the tax burden of each asset. That certainly has a lot of merit, and his recommendations are solid. The other (which I personally follow) is to maximize tax flexibility by holding the same full portfolio in both taxable and tax-deferred accounts. I like it because:

1) It's way easier to rebalance
2) It opens up more opportunities for things like tax-loss harvesting
3) I want the taxable money that I have easy access to in ER to have the full portfolio characteristics I care about. And it's hard to do that with only a partial recipe.

But the best choice depends on your personal situation, so obviously YMMV.

And regardless of where you keep assets, the simplest and most effective tax minimization strategy during accumulation is to always top off the lagging asset with new money. Keep that up, and you may go many years without having to realize any capital gains to rebalance.

BTW, the Swensen asset allocation (updated for his bump up in emerging markets which he did because they become bigger chunk of market after his book was published) is 30% US stocks... 15% IT treasuries, 15% TIPS, 15% developed market, 15% REITs... and 10% emerging.

It's funny -- I added a note about that on the Swensen Portfolio page just yesterday after someone else brought it to my attention. Do you have a good reference that explains Swensen's thought process in the change? I'd love to read it.

Just looking at this a little more:

I’m seeing that Fidelity’s SCV fund is at $25/share right now, it paid out $0.41/share in dividends and $2.06/share in capital gains this year. So someone in the 24% bracket would pay 0.4% of their invested balance in taxes on dividends and 1.2% in taxes on long term capital gains at 15%. That seems like a pretty big tax drag to me.

You’ve got me intrigued enough to think about this some more, but it seems like you would need to hold this particular asset in a tax-advantaged account.

[Tyler]

Just looking at this a little more:

I’m seeing that Fidelity’s SCV fund is at $25/share right now, it paid out $0.41/share in dividends and $2.06/share in capital gains this year. So someone in the 24% bracket would pay 0.4% of their invested balance in taxes on dividends and 1.2% in taxes on long term capital gains at 15%. That seems like a pretty big tax drag to me.

You’ve got me intrigued enough to think about this some more, but it seems like you would need to hold this particular asset in a tax-advantaged account.

That's a good example of my point about ETFs vs mutual funds. The Fidelity SCV fund FISVX has a mutual fund structure (you can tell by the 5 letters in the ticker). It tracks the Russell 2000 Value index. Now look at the distributions for IWN, an ETF that tracks the same index. It hasn't had a capital gains distribution since 2000, right after its inception. It's not a Fidelity vs iShares thing, but a fundamental difference between mutual funds and ETFs. ETFs are just structured differently and capital gains distributions are very rare.

Of course, IWN also has a higher expense ratio than FISVX so you have to find the right balance. Personally I use VBR. For me, the slightly higher ER is worth it for the better tax treatment.

[Wintergreen78]

Just looking at this a little more:

I’m seeing that Fidelity’s SCV fund is at $25/share right now, it paid out $0.41/share in dividends and $2.06/share in capital gains this year. So someone in the 24% bracket would pay 0.4% of their invested balance in taxes on dividends and 1.2% in taxes on long term capital gains at 15%. That seems like a pretty big tax drag to me.

You’ve got me intrigued enough to think about this some more, but it seems like you would need to hold this particular asset in a tax-advantaged account.

That's a good example of my point about ETFs vs mutual funds. The Fidelity SCV fund FISVX has a mutual fund structure (you can tell by the 5 letters in the ticker). It tracks the Russell 2000 Value index. Now look at the distributions for IWN, an ETF that tracks the same index. It hasn't had a capital gains distribution since 2000, right after its inception. It's not a Fidelity vs iShares thing, but a fundamental difference between mutual funds and ETFs. ETFs are just structured differently and capital gains distributions are very rare.

Of course, IWN also has a higher expense ratio than FISVX so you have to find the right balance. Personally I use VBR. For me, the slightly higher ER is worth it for the better tax treatment.

That is a good distinction to keep in mind. When i started years ago putting money into the Fidelity SP500 index fund, I wasn’t really aware of the distinctions between mutual funds and ETFs. The SP500 mutual funds have relatively low distributions, so I’ve never bothered switching to ETFs.

For SCV funds in tax-advantaged accounts you probably don’t need to worry about mutual funds vs. ETFs, but it seems like an important consideration if you are holding them in taxable accounts.

[SeattleCPA]

Just looking at this a little more:

I’m seeing that Fidelity’s SCV fund is at $25/share right now, it paid out $0.41/share in dividends and $2.06/share in capital gains this year. So someone in the 24% bracket would pay 0.4% of their invested balance in taxes on dividends and 1.2% in taxes on long term capital gains at 15%. That seems like a pretty big tax drag to me.

You’ve got me intrigued enough to think about this some more, but it seems like you would need to hold this particular asset in a tax-advantaged account.

That's a good example of my point about ETFs vs mutual funds. The Fidelity SCV fund FISVX has a mutual fund structure (you can tell by the 5 letters in the ticker). It tracks the Russell 2000 Value index. Now look at the distributions for IWN, an ETF that tracks the same index. It hasn't had a capital gains distribution since 2000, right after its inception. It's not a Fidelity vs iShares thing, but a fundamental difference between mutual funds and ETFs. ETFs are just structured differently and capital gains distributions are very rare.

Of course, IWN also has a higher expense ratio than FISVX so you have to find the right balance. Personally I use VBR. For me, the slightly higher ER is worth it for the better tax treatment.

That is a good distinction to keep in mind. When i started years ago putting money into the Fidelity SP500 index fund, I wasn’t really aware of the distinctions between mutual funds and ETFs. The SP500 mutual funds have relatively low distributions, so I’ve never bothered switching to ETFs.

For SCV funds in tax-advantaged accounts you probably don’t need to worry about mutual funds vs. ETFs, but it seems like an important consideration if you are holding them in taxable accounts.

Not to muddy the waters here, but to go "big picture" on this...

My thought is if you can use the taxable space for something like your US stocks allocation which should mean mostly qualified dividends and LT capital gains, during retirement, that's pretty efficient. E.g., if you have $1M in your nest and $100K of that is taxable all in US stocks, your taxable account's income probably isn't triggering federal income taxes. Because the LT capital gains and qualified dividends tax rate is probably 0%.

BTW if double your portfolio to $2M and hold $200K in US stocks in taxable account, would guess that "taxable income" also doesn't trigger any federal income taxes.

To go off topic, it's interesting to think about a gold EFT with a .4% expense ratio versus coins. (I've worked with clients who can pretty easily put their gold allocation (in coins) into a safe deposit box which is a pretty low annual expense ratio.) Obviously, the gold eft easier to rebalance. But have gotten used to paying .04% expense ratios... Paying .4% would be sobering.

[Imanuels]

In principle yes. The difference I see is that gold is not productive - is doesn't generate earnings, pay dividends or interest. Thus, I'm not sure how to assess its 'real value' and if I see a reason why it actually should necessarily always recover after losses. Moreover, what happens when interest rates rise?
But I apologize, this is a specific discussion about gold then. And Tyler has a great article on topic: https://portfoliocharts.com/2020/08/21/metal-money-and-the-measurable-value-of-gold/
There's a graph relevant to the current times showing relationship between the gold price and real interest rates.

I'm wondering about two points:
1. There's historical evidence that momentum provides better returns. Yet, in this case SCV stands out.
2. The importance of gold in most efficient portfolios. By looking at how poorly gold itself has performed for several periods, I'm wondering if this could be further adjusted by some other metric that avoids buying gold at the peak and thus sitting on an asset with negative return. See the attached image, at least visually the similarity to the previous peak looks scary.

There are plenty of backtested portfolios that introduce trend following rules, eg only hold something that is above its 200 day moving average.

One of the best examples is Meb Faber's Trinity portfolio, all the backtests suggested that it would produce superior long term results with less volatility than a buy and hold multi-asset approach. However since he launched it (TRTY) the performance has been pretty lacklustre. Maybe that's unfair because it hasn't been launched for that long, but it just shows that backtested approaches that introduce too many rules tend to perform worse going forward than they have done so historically.

It's also a bit unfair to single out gold as an asset class you particularly want to avoid when it's going down.  All asset classes have suffered catastrophic losses at some point in the last century. 
(from https://mebfaber.com/2020/03/05/the-stay-rich-portfolio/)

Largest real drawdown since 1926:

US Stocks:  -79%
Foreign Stocks:  -78%
10 Year US Bonds: -61%
Foreign Bonds:  -78%
Gold:  -85%

What is more likely to cripple your portfolio: having 60-70% of your portfolio in an asset class that has gone down -79%, or 10% of your portfolio in an asset class that has gone down -85%?

[vand]

In principle yes. The difference I see is that gold is not productive - is doesn't generate earnings, pay dividends or interest. Thus, I'm not sure how to assess its 'real value' and if I see a reason why it actually should necessarily always recover after losses. Moreover, what happens when interest rates rise?
But I apologize, this is a specific discussion about gold then. And Tyler has a great article on topic: https://portfoliocharts.com/2020/08/21/metal-money-and-the-measurable-value-of-gold/
There's a graph relevant to the current times showing relationship between the gold price and real interest rates.

I'm wondering about two points:
1. There's historical evidence that momentum provides better returns. Yet, in this case SCV stands out.
2. The importance of gold in most efficient portfolios. By looking at how poorly gold itself has performed for several periods, I'm wondering if this could be further adjusted by some other metric that avoids buying gold at the peak and thus sitting on an asset with negative return. See the attached image, at least visually the similarity to the previous peak looks scary.

There are plenty of backtested portfolios that introduce trend following rules, eg only hold something that is above its 200 day moving average.

One of the best examples is Meb Faber's Trinity portfolio, all the backtests suggested that it would produce superior long term results with less volatility than a buy and hold multi-asset approach. However since he launched it (TRTY) the performance has been pretty lacklustre. Maybe that's unfair because it hasn't been launched for that long, but it just shows that backtested approaches that introduce too many rules tend to perform worse going forward than they have done so historically.

It's also a bit unfair to single out gold as an asset class you particularly want to avoid when it's going down.  All asset classes have suffered catastrophic losses at some point in the last century. 
(from https://mebfaber.com/2020/03/05/the-stay-rich-portfolio/)

Largest real drawdown since 1926:

US Stocks:  -79%
Foreign Stocks:  -78%
10 Year US Bonds: -61%
Foreign Bonds:  -78%
Gold:  -85%

What is more likely to cripple your portfolio: having 60-70% of your portfolio in an asset class that has gone down -79%, or 10% of your portfolio in an asset class that has gone down -85%?

The answer that I like to give to this is that gold is good money.  One of the functions of money is a long term store of wealth.  Good money absolutely should increase in purchasing power as society gets richer - it doesn't *need* to be put to work or do anything special or have an income stream. Over the long term as society becomes richer holder of real money benefits from the work that productive assets do to raise the standard of living.  You don't need to exclusively hold all you wealth in the wealth producing asset (companies) to benefit from the work that they do, you can hold some of it in real money.

I have posted this video before and will continue to refer to it when people ask why should gold go up?
Good money appreciates: https://www.youtube.com/watch?v=GnCHmLsV6Ro

[JJ-]

Thanks @Tyler for this research and @SeattleCPA for posting it. My plan was to start diversifying into EM next year before doing gold/REIT but this has me reconsidering and probably pulling the plug that approach to start with gold.

[SeattleCPA]

Thanks @Tyler for this research and @SeattleCPA for posting it. My plan was to start diversifying into EM next year before doing gold/REIT but this has me reconsidering and probably pulling the plug that approach to start with gold.

So, I think Tyler's post is excellent. And (perhaps indoctrinated on modern portfolio theory from my MBA in finance) agree with the portfolio construction insights Tyler provides.

Those points stipulated, however, I think I'm going to just stick with my Swensen asset allocation formula which suggests following:
30% US stocks
15% Developed markets
15% REITs
15% Intermediate term treasuries
15% TIPS
10% emerging market

A couple three reasons for this which I'll share because someone ( @JJ- ?) may find parallels with their own situation.

First, in digging around to see what Swensen had to say about gold, I found this article where Swensen explains why he thinks individuals should avoid gold... the logic basically that gold only tracks inflation. Here's the link:

https://www.barrons.com/articles/SB113140725152890541?tesla=y

Second, and related to above, I already (via Swensen and then alternative investments) have a bunch of inflation protection built into my asset allocation formula.

Third, I'm at stage where portfolio simplicity matters. And Swensen formula already pushes the boundaries there.

For what that's worth.

Again, though, I do think the overall logic of Tyler's post? Totally agree. And in fact as he notes in his article, Swensen's formula works well... and I would say because Swensen looks not at individual assets but at the portfolio construction as a whole.

[vand]

My current AA is:

General Equity   39.03%
REIT   6.57%
Government Bonds   4.99%
Precious Metals   23.49%
Precious Metal Equity   8.14%
Defensive Investment Trusts   15.11%
Commodities   1.23%
Cash   1.43%

It's a work in progress and I've been reducing my Bond allocation throughout the year.  Overall I've been really pleased with the steady performance. I think the max drawdown throughout the year has been in the region of about -4%, while the internal rate of return has been around 11-12%.  It's not a passive approach though, I have about 15% in specifically selected stocks, and I hold quite a few actively managed funds.

[JJ-]

Thanks @Tyler for this research and @SeattleCPA for posting it. My plan was to start diversifying into EM next year before doing gold/REIT but this has me reconsidering and probably pulling the plug that approach to start with gold.

So, I think Tyler's post is excellent. And (perhaps indoctrinated on modern portfolio theory from my MBA in finance) agree with the portfolio construction insights Tyler provides.

Those points stipulated, however, I think I'm going to just stick with my Swensen asset allocation formula which suggests following:
30% US stocks
15% Developed markets
15% REITs
15% Intermediate term treasuries
15% TIPS
10% emerging market

A couple three reasons for this which I'll share because someone ( @JJ- ?) may find parallels with their own situation.

First, in digging around to see what Swensen had to say about gold, I found this article where Swensen explains why he thinks individuals should avoid gold... the logic basically that gold only tracks inflation. Here's the link:

https://www.barrons.com/articles/SB113140725152890541?tesla=y

Second, and related to above, I already (via Swensen and then alternative investments) have a bunch of inflation protection built into my asset allocation formula.

Third, I'm at stage where portfolio simplicity matters. And Swensen formula already pushes the boundaries there.

For what that's worth.

Again, though, I do think the overall logic of Tyler's post? Totally agree. And in fact as he notes in his article, Swensen's formula works well... and I would say because Swensen looks not at individual assets but at the portfolio construction as a whole.

Thanks for this breakdown. Currently our portfolio is relatively simple, 80/20 equities to bonds, but it gets a little more complicated than that.

The equity component is 35% US total market, 30% US SCV, 20% EX US total market,  15% EX US SCV.

The bond component is 65% TSP G fund 35% LTT.

Given the sizable chunk in the g fund whose downside is inflation risk, it seems the easiest/simplest step for me to shore up the portfolio is by adding gold.

Before pulling the trigger I'll read more into the Swenson portfolio logic.

[SeattleCPA]

Before pulling the trigger I'll read more into the Swenson portfolio logic.

Swensen's book for individuals, Unconventional Success, is really well worth the read. Also, usefully, he wrote the book like 15 years ago so you can kinda "back test" his suggestions...

His basic logic, in a nutshell...

Improve the classic 60%/40% stock and bonds portfolio by just going with US treasuries but changing stock/bond percentages to 70%/30%...

He wants to have 30% of your portfolio in riskless assets (thus the 15% to intermediate term treasuries and 15% to TIPS)...

And he want to have 30% of your portfolio in real assets (thus 15% to REITs and 15% to TIPS)...

He noted in his book that if you already have a house (and so some inflation protection) that that dials down your need for 15% in REITs...

He also notes that if in your "job" you already already own part of a small business, that maybe suggests you dial down your equity percentage.

It's imminently readable. And should be valuable even if you use some other portfolio construction philosophy.

[JJ-]

Before pulling the trigger I'll read more into the Swenson portfolio logic.

Swensen's book for individuals, Unconventional Success, is really well worth the read. Also, usefully, he wrote the book like 15 years ago so you can kinda "back test" his suggestions...

His basic logic, in a nutshell...

Improve the classic 60%/40% stock and bonds portfolio by just going with US treasuries but changing stock/bond percentages to 70%/30%...

He wants to have 30% of your portfolio in riskless assets (thus the 15% to intermediate term treasuries and 15% to TIPS)...

And he want to have 30% of your portfolio in real assets (thus 15% to REITs and 15% to TIPS)...

He noted in his book that if you already have a house (and so some inflation protection) that that dials down your need for 15% in REITs...

He also notes that if in your "job" you already already own part of a small business, that maybe suggests you dial down your equity percentage.

It's imminently readable. And should be valuable even if you use some other portfolio construction philosophy.

I have borrowed it from the library

[index]

To go off topic, it's interesting to think about a gold EFT with a .4% expense ratio versus coins. (I've worked with clients who can pretty easily put their gold allocation (in coins) into a safe deposit box which is a pretty low annual expense ratio.) Obviously, the gold eft easier to rebalance. But have gotten used to paying .04% expense ratios... Paying .4% would be sobering.

There are many gold ETFs with expense ratios lower than GLD's 0.4%. GLDM 0.18% (still issued by State Street and same setup GLD), BAR 0.175%, SGOL 0.17%.

The expense of holding physical gold are not at the safety deposit box level, but transaction costs of buying a selling gold coins. You are going to pay 5-10% to buy and the same to sell a gold coin. If you are trading very high dollar amounts, you are still going to pay about 3% (6% round trip). 

[ChpBstrd]

@Tyler a few dumb newbie questions you've no doubt answered 100x before:

1) Are returns for the indices or are they adjusted for expense ratios to simulate the existence of ETFs in the distant past? We're talking about the cumulative effects of very small long-term performance differences, and we're also talking about various asset classes where the differences in ER's could be a half-percent or more. Do gold and SCV still shine when we subtract a typical mutual fund ER/load every year?

2) If gold was illegal for individual investors to own prior to 1975, and it had already multiplied by that point, is it accurate to say portfolios containing gold "would have" done well since 1970 if they literally "could not" have existed? Why not use silver, mining companies, or a commodities index, to avoid this problem, or (the easy solution) start the gold series when it became investable?

3) Obviously the efficient frontier is approached via rebalancing. What's the rebalancing frequency implicit in this study?

4) At this point, it is highly improbable that a LT Treasury allocation would deliver a double-digit positive annual returns. The experience of the last 40 years of falling interest rates is not going to repeat - there's not another 13% for long duration interest rates to fall. Are long duration bonds still a secret ingredient if there is no plausible way for them to deliver the returns required to be a secret ingredient - or even positive real returns?

[joe189man]

PTF - and learn/reread

[SeattleCPA]

@Tyler a few dumb newbie questions you've no doubt answered 100x before:

1) Are returns for the indices or are they adjusted for expense ratios to simulate the existence of ETFs in the distant past? We're talking about the cumulative effects of very small long-term performance differences, and we're also talking about various asset classes where the differences in ER's could be a half-percent or more. Do gold and SCV still shine when we subtract a typical mutual fund ER/load every year?

2) If gold was illegal for individual investors to own prior to 1975, and it had already multiplied by that point, is it accurate to say portfolios containing gold "would have" done well since 1970 if they literally "could not" have existed? Why not use silver, mining companies, or a commodities index, to avoid this problem, or (the easy solution) start the gold series when it became investable?

3) Obviously the efficient frontier is approached via rebalancing. What's the rebalancing frequency implicit in this study?

4) At this point, it is highly improbable that a LT Treasury allocation would deliver a double-digit positive annual returns. The experience of the last 40 years of falling interest rates is not going to repeat - there's not another 13% for long duration interest rates to fall. Are long duration bonds still a secret ingredient if there is no plausible way for them to deliver the returns required to be a secret ingredient - or even positive real returns?

To help Tyler, he does provide an expense ratio input on the portfolio finder portfoliochart: https://portfoliocharts.com/portfolio/risk-and-return/

And then, here's something that helps me understand this. You can look at gold just for last 20 years at portfoliovisualizer. And you can see the effect it has on your portfolio. E.g., if I take Swensen asset allocation formula and put 10% into gold taking the 10% from the other real asset categories, TIPS and REITs, it nudges things in the right direction. A little higher return... a little less variability.

[Tyler]

1) Are returns for the indices or are they adjusted for expense ratios to simulate the existence of ETFs in the distant past? We're talking about the cumulative effects of very small long-term performance differences, and we're also talking about various asset classes where the differences in ER's could be a half-percent or more. Do gold and SCV still shine when we subtract a typical mutual fund ER/load every year?

Yes, everything is adjusted for realistic expense ratios for every individual asset type. You can read more about that here.

2) If gold was illegal for individual investors to own prior to 1975, and it had already multiplied by that point, is it accurate to say portfolios containing gold "would have" done well since 1970 if they literally "could not" have existed? Why not use silver, mining companies, or a commodities index, to avoid this problem, or (the easy solution) start the gold series when it became investable?

Gold is a global asset and was not illegal for private investors outside of the US. That's how we have records for the price. So just like how large cap blend stock data can be tracked before index funds made it investable, the gold data is still valid.

3) Obviously the efficient frontier is approached via rebalancing. What's the rebalancing frequency implicit in this study?

It assumes all portfolios are rebalanced once a year.

4) At this point, it is highly improbable that a LT Treasury allocation would deliver a double-digit positive annual returns. The experience of the last 40 years of falling interest rates is not going to repeat - there's not another 13% for long duration interest rates to fall. Are long duration bonds still a secret ingredient if there is no plausible way for them to deliver the returns required to be a secret ingredient - or even positive real returns?

There actually is a plausible way for long term bonds to deliver significant real returns today, and it doesn't require that rates fall another 13%. I suggest reading this article on bond convexity.

[mntnmn117]

4) At this point, it is highly improbable that a LT Treasury allocation would deliver a double-digit positive annual returns. The experience of the last 40 years of falling interest rates is not going to repeat - there's not another 13% for long duration interest rates to fall. Are long duration bonds still a secret ingredient if there is no plausible way for them to deliver the returns required to be a secret ingredient - or even positive real returns?

There actually is a plausible way for long term bonds to deliver significant real returns today, and it doesn't require that rates fall another 13%. I suggest reading this article on bond convexity.

I was just discovering the 90 SCV/ 10 TLT portfolio and do you think that it's performance is a fluke of past conditions/bias? It's peaks and valleys just happen to line up better than anything else or does it have a fundamental relationship causing the inverse correlation? 

Punching in 90/10 SCV/TLT gives an average return of 10.1 and Perpetual WR of 6.2. Same Ulcer# as 60/40, No-brainer, and Core 4 but with almost twice the Average, Baseline LT, and PerpWR.

[Tyler]

I was just discovering the 90 SCV/ 10 TLT portfolio and do you think that it's performance is a fluke of past conditions/bias? It's peaks and valleys just happen to line up better than anything else or does it have a fundamental relationship causing the inverse correlation? 

Punching in 90/10 SCV/TLT gives an average return of 10.1 and Perpetual WR of 6.2. Same Ulcer# as 60/40, No-brainer, and Core 4 but with almost twice the Average, Baseline LT, and PerpWR.

I'm not going to pretend to be some macroeconomic expert, and I'm definitely NOT a financial adviser. But if you want to read the reasoning behind SCV, I recommend this book. It covers several of your questions and may help you decide whether that much SCV makes sense to you or not.

[VanillaGorilla]

I adore portfoliocharts.com and have spent way too much time on it over the years. Thank you @Tyler!

This is an interesting analysis and I have a few amateurish thoughts:

1.) I think the SCV tilt is too well documented and has been arbitraged out. Using a SCV tilt in your portfolio and assuming that future results will resemble backtesting strikes me as unwise.

2.) Optimizing my portfolio for the 15th percentile of poor economic conditions is not something I'm interested in doing. I save 70% of my income, I don't need to minimize risk, I want to maximize returns. Hence I'm 100% equities (well, and real estate....in that my house is worth a lot and I owe very little).

3.) Overfitting to historical niches is tempting but unwise. SCV is very impressive historically, but in modern timeframes it has not overperformed. In my mind it's too well known and cannot return similar figures as it once did. Simply backtesting various asset classes makes it tempting to go 100% SCV or REIT and enjoying a 5% safe withdrawal rate, but then you're overfitting to very particular bits of history that will by definition never repeat. To make backtesting useful for predicting future performance I only feel comfortable painting with very broad brushes. Saying a 60/40 portfolio is quite safe at 3.3% is fine. Arguing that, say, pork futures over the next 50 years will behave like the previous 50 is risky.

I'm at 30x expenses at 100% large cap equities and I will consider shifting to 20% bonds and 10% gold when I no longer have an income, but while I have a job and am saving 70% of my income I want to optimize for returns, not for low risk, so I find this analysis interesting but not compelling.

I would love to see it repeated for maybe a median and a 75th percentile of market return!

[Tyler]

I appreciate the different perspectives. As I mentioned in the article, different people have a variety of mindsets and needs that can't be solved with a single chart or portfolio. But if nothing else, I hope the data is thought provoking enough to encourage people to step back and think about their own assumptions.

[vand]

@Tyler , regarding bond convexity, I've been trying to understand the implications of this for a while... I understand that when rates are low then all bonds become more senstive to changes in interest rates. When interest rates are 1% then a +/- 1% move in rates is equivilent to a +/- 100% change in the interest rate, compared to when the rates are 10% then a 1% move is only equivilent to a +/- 10% change.

But isn't the downside to this is that they become prone to large drop off a relatively small increase in interest rate? For example, all long duration bond funds such as TLT are well down from their peaks of summer 2020 when long term rate were only a percent or so lower than they are today.

I don't see a magic get-out-of-jail card for the Bond market here. The only way for them to continue delivering strong returns is if rates go even more negative, but that increases the risk of capital losses when interest rates move the other way.

From an real return point of view, bonds will still return their coupon rate with ~90% accuracy by maturity minus whatever the prevailing rate of inflation is in the meantime. The convexity may, however, mean that their capital value is more sensitive to rate changes and therefore are "better" crash protection than when rates are higher - so handily you may not need to hold as many of them to have the same portfolio dampening effects. 70/30 or 75/25 may be the new "efficient" portfolio with rates where they are today.

[SeattleCPA]

...but then you're overfitting to very particular bits of history that will by definition never repeat. To make backtesting useful for predicting future performance I only feel comfortable painting with very broad brushes.

I agree with above, for what it's worth...

And then what I think is tricky is separating some effect that's basically "noise" or what Daniel Kahneman called the "law of small numbers" or what you're articulately labeling "very particular bits of history" with something more, hmmm, statistically sound.

Not to change the topic, but based on Swensen's concern about gold (referenced in an earlier comment in this thread), I tried to see (using portfoliovisualizer) whether gold helps a portfolio if all it does is dampen the volatility. I.e., if all gold is doing is maintaining its value and not correlating.

And Swensen's comment is right. E.g., you look at ten years starting 2010 when gold shows lots of volatility, nearly zero correlation with stocks and generates a very small annual return... and yes a 10% allocation to gold dampens the variability. But you're better off leaving gold out of your portfolio. And for this particular bit of history, the theoretical benefit I'd love to enjoy doesn't show up.

Summing up and returning to @VanillaGorilla 's point, I kind of worry I'm at risk when I look at something like this of assuming the positive portfolio effect I see comes from clever portfolio construction. When actually, ugh, the positive portfolio effect stems from stumbling on a "particular bit of history."

The very clever Bill Bernstein in one of his books says something like, and I'm wildly paraphrasing probably, "sure, you can improve your investment results by using modern portfolio theory thinking... but the benefit is often disappointingly modest and often not worth the cost."

Here's portfolio visualizer chart for 100% Gold (portfolio 1), 60/40 stocks and bonds (portfolio 2) and 55/35/10 stocks, bonds and gold (portfolio 3). The inputs are start with a $1 million and draw $40K a year. Gold over this ten year time frame earns about a 1% real rate of return. Yes, it nicely dampens variability. But it really doesn't do what you'd hope.

[index]

I adore portfoliocharts.com and have spent way too much time on it over the years. Thank you @Tyler!

This is an interesting analysis and I have a few amateurish thoughts:

1.) I think the SCV tilt is too well documented and has been arbitraged out. Using a SCV tilt in your portfolio and assuming that future results will resemble backtesting strikes me as unwise.

2.) Optimizing my portfolio for the 15th percentile of poor economic conditions is not something I'm interested in doing. I save 70% of my income, I don't need to minimize risk, I want to maximize returns. Hence I'm 100% equities (well, and real estate....in that my house is worth a lot and I owe very little).

3.) Overfitting to historical niches is tempting but unwise. SCV is very impressive historically, but in modern timeframes it has not overperformed. In my mind it's too well known and cannot return similar figures as it once did. Simply backtesting various asset classes makes it tempting to go 100% SCV or REIT and enjoying a 5% safe withdrawal rate, but then you're overfitting to very particular bits of history that will by definition never repeat. To make backtesting useful for predicting future performance I only feel comfortable painting with very broad brushes. Saying a 60/40 portfolio is quite safe at 3.3% is fine. Arguing that, say, pork futures over the next 50 years will behave like the previous 50 is risky.

I'm at 30x expenses at 100% large cap equities and I will consider shifting to 20% bonds and 10% gold when I no longer have an income, but while I have a job and am saving 70% of my income I want to optimize for returns, not for low risk, so I find this analysis interesting but not compelling.

I would love to see it repeated for maybe a median and a 75th percentile of market return!

There are many 15 year periods where 80/20 outperforms 100% equities. 2005 to 2020 being the last. (blue)



@Tyler, I am interested in factors other than growth and value - specifically sector concentration. I have used a risk parity strategy for over a decade now that relies on leverage. Have you investigated the effect of leveraging the golden butterfly?

[index]

...but then you're overfitting to very particular bits of history that will by definition never repeat. To make backtesting useful for predicting future performance I only feel comfortable painting with very broad brushes.

I agree with above, for what it's worth...

And then what I think is tricky is separating some effect that's basically "noise" or what Daniel Kahneman called the "law of small numbers" or what you're articulately labeling "very particular bits of history" with something more, hmmm, statistically sound.

Not to change the topic, but based on Swensen's concern about gold (referenced in an earlier comment in this thread), I tried to see (using portfoliovisualizer) whether gold helps a portfolio if all it does is dampen the volatility. I.e., if all gold is doing is maintaining its value and not correlating.

And Swensen's comment is right. E.g., you look at ten years starting 2010 when gold shows lots of volatility, nearly zero correlation with stocks and generates a very small annual return... and yes a 10% allocation to gold dampens the variability. But you're better off leaving gold out of your portfolio. And for this particular bit of history, the theoretical benefit I'd love to enjoy doesn't show up.

Summing up and returning to @VanillaGorilla 's point, I kind of worry I'm at risk when I look at something like this of assuming the positive portfolio effect I see comes from clever portfolio construction. When actually, ugh, the positive portfolio effect stems from stumbling on a "particular bit of history."

The very clever Bill Bernstein in one of his books says something like, and I'm wildly paraphrasing probably, "sure, you can improve your investment results by using modern portfolio theory thinking... but the benefit is often disappointingly modest and often not worth the cost."

Here's portfolio visualizer chart for 100% Gold (portfolio 1), 60/40 stocks and bonds (portfolio 2) and 55/35/10 stocks, bonds and gold (portfolio 3). The inputs are start with a $1 million and draw $40K a year. Gold over this ten year time frame earns about a 1% real rate of return. Yes, it nicely dampens variability. But it really doesn't do what you'd hope.

Your back test is showing lower performance due to the 60% vs. 55% equity component. Here is a back test with your same parameters - start with $1M and draw 40k/year with 60/40 stocks/bonds, 60/20/20 stocks/bonds/gold, and 60/40 stocks/gold. Also, you should really be using treasuries if you are trying to reduce volatility - not total bond market.

[index]

Gold helps out because not only is it not correlated to equities, but it is not correlation to treasuries either (first attachment). The second attachment shows the same back test extended 43 years (with LT treasuries)

[Blender Bender]

Good data to chew on.

However the future very likely (almost certain) will not look like the past:
1. Bonds are subject to face raising interest rates.
2. Gold in many minds got replaced by "crypto".

[SeattleCPA]

Your back test is showing lower performance due to the 60% vs. 55% equity component. Here is a back test with your same parameters - start with $1M and draw 40k/year with 60/40 stocks/bonds, 60/20/20 stocks/bonds/gold, and 60/40 stocks/gold.

So I don't think so... I think maybe the change in the date range produces the effect you're pointing to. But check my accuracy...

E.g., if I go 70% stocks, 10% gold and 20% intermediate treasuries, and look at just ten years (from 1/1/2010 to 12/31/2019), I get an almost unmeasurable bump in real return (6 basis points) but also a slightly noticeable bump in my variability.

BTW, I would have guessed if we swap in LT treasuries, that'll help. But it doesn't. In that MPT-y portfolio, the math doesn't work. At least over 10 years when looking at the average. BTW I would not be surprised if the baseline return Tyler likes to use shows differently.

My point in all this isn't to argue with the theory. I guess I'm just pointing out the obvious... It's hard to tease out the precise effects of the MPT benefits with what @VanillaGorilla labeled a "particular bit of history"... like interest rates dropping.

Also, you should really be using treasuries if you are trying to reduce volatility - not total bond market.

Yeah, I agree. In real life, following Swensen's formula, I do use treasuries... FWIW, I use intermediate term...

Final comment: Appreciate your thoughtful constructive criticism of my comment. You obviously get this stuff.

[HPstache]

So I have a question about the gold portion of a portfolio when considered from the perspective of this article.  It almost seems like it has to be digital gold shares rather than actually owning physical gold to get the "ulcer index" reduction effect for your portfolio.  Is this correct?  I mean, if you would own physical gold instead of gold shares (let's ignore the high cost of actually buying & selling physical gold), it would be theoretically be doing all the same things under your mattress but your actual portfolio would not not get the smoothing affect that one would want to see by holding gold.  It almost seems like from an investment standpoint, it is counterintuitively superior to just hold gold shares than the real thing in your hand...

[SeattleCPA]

Consider this a comment from the peanut gallery...

My experience with gold is not as an investor but as an accountant who occasionally encounters Krugerrands in safe deposit boxes.

Others have pointed out the benefits of using something like IAU. Reasonable expense ratio. Easy rebalancing. Defer to the investors with real life experience for all of this stuff...

My gold coins based observations... There are benefits to having coins in a safe deposit box. Maybe, I guess, as long as the wrong people don't know about it.

The storage costs in that case are extremely low. You can store $500K in a safe deposit box that costs $100 or $200 a year. If you already have to have a safe deposit box, your storage costs may in effect be zero.

Also, the transportability and privacy may be useful. Not talking movie-script scenarios BTW. Think about a sick parent or grandparent who wants a way to easily and instantaneously give away a lot of wealth.

BTW, also this real life observation. It can also be very problematic for obvious reasons to have something like this in your portfolio at the end of the road.

[index]

So I have a question about the gold portion of a portfolio when considered from the perspective of this article.  It almost seems like it has to be digital gold shares rather than actually owning physical gold to get the "ulcer index" reduction effect for your portfolio.  Is this correct?  I mean, if you would own physical gold instead of gold shares (let's ignore the high cost of actually buying & selling physical gold), it would be theoretically be doing all the same things under your mattress but your actual portfolio would not not get the smoothing affect that one would want to see by holding gold.  It almost seems like from an investment standpoint, it is counterintuitively superior to just hold gold shares than the real thing in your hand...

I guess it depends on your definition of "portfolio". I you want to see the smoothing effect in your brokerage account, then holding paper gold is superior. You could manually update a net worth/investment spreadsheet or tool with the spot value of your gold in a safety deposit box everyday and see the same benefit. Just because the investment isn't marked to market by your brokerage doesn't mean its value hasn't changed. Physical real estate versus REITs are somewhat similar. Many real estate investors never mark their holdings to market on a regular basis so rent received and paper appreciation/losses are never accounted.

[index]

Your back test is showing lower performance due to the 60% vs. 55% equity component. Here is a back test with your same parameters - start with $1M and draw 40k/year with 60/40 stocks/bonds, 60/20/20 stocks/bonds/gold, and 60/40 stocks/gold.

So I don't think so... I think maybe the change in the date range produces the effect you're pointing to. But check my accuracy...

E.g., if I go 70% stocks, 10% gold and 20% intermediate treasuries, and look at just ten years (from 1/1/2010 to 12/31/2019), I get an almost unmeasurable bump in real return (6 basis points) but also a slightly noticeable bump in my variability.

BTW, I would have guessed if we swap in LT treasuries, that'll help. But it doesn't. In that MPT-y portfolio, the math doesn't work. At least over 10 years when looking at the average. BTW I would not be surprised if the baseline return Tyler likes to use shows differently.

My point in all this isn't to argue with the theory. I guess I'm just pointing out the obvious... It's hard to tease out the precise effects of the MPT benefits with what @VanillaGorilla labeled a "particular bit of history"... like interest rates dropping.

Also, you should really be using treasuries if you are trying to reduce volatility - not total bond market.

Yeah, I agree. In real life, following Swensen's formula, I do use treasuries... FWIW, I use intermediate term...

Final comment: Appreciate your thoughtful constructive criticism of my comment. You obviously get this stuff.

I am observing the same as you with those dates (first attachment). I have spent a lot of time following a risk parity strategy which can be summed up as investing for the four different prevailing market conditions as illustrated below:



There are a lot of posters on this forum that are 100% equities and have only invested in the disinflationary boom quadrant of the box. Tyler's research and the charts posted on this thread are making the case those without properly diversified portfolios are taking uncompensated risk and, in many cases, underperform diversified portfolios over long time periods. MPT and diversification work due to rebalancing and limiting variance. It is unintuitive to say the SP500 has returned 9% CAGR and gold has a 6% CAGR but a portfolio of 70/30 SP500 gold portfolio returned 8.9% CAGR with considerably less variability and allows for a higher SWR.

What gets really interesting is using a volatility target and leveraging a portfolio (attachment 3). These are poorly constructed portfolios, but exemplify how using a little leverage with poorly correlated assets can increase returns and actually decrease risk. The second portfolio is effectively 100% SP500, 25% gold, 25% LTT and the variability matches that of a 100% equity investor.     

[SeattleCPA]

... those without properly diversified portfolios are taking uncompensated risk and, in many cases, underperform diversified portfolios over long time periods. 

Totally agree. 100%.

[Blender Bender]

... those without properly diversified portfolios are taking uncompensated risk and, in many cases, underperform diversified portfolios over long time periods. 

Totally agree. 100%.

So many detailed charts from the past, ignoring the changed context.

Diversification is good for sure, unless it "ensures" underperformance. I'm assuming that we talk here about investing here, not SORR strategies.

I have used before phrases like "almost sure", "very likely". But here we have someone more bold :) : "Guaranteed TINA"
https://markets.businessinsider.com/news/stocks/stock-market-outlook-2022-fundstrat-tom-lee-upside-3-reasons-2021-12

I rather take advise from Warren Buffett. He shared his opinions very clearly about bonds for the next 10y, and gold.

[index]

... those without properly diversified portfolios are taking uncompensated risk and, in many cases, underperform diversified portfolios over long time periods. 

Totally agree. 100%.

So much detailed charts from the past... This is how advisors "manipulate" potential clients. :)

I rather take advise from Warren Buffett. He shared his opinions very clearly about bonds for the next 10y, and gold.

Good data to chew on.

However the future very likely (almost certain) will not look like the past:
1. Bonds are subject to face raising interest rates.
2. Gold in many minds got replaced by "crypto".

Buffet, Dalio, and Bill Gross have all expressed they do not see the past repeating itself with bonds. I agree with them and do not incorporate any bonds in my portfolio. I'm a big Buffett fan and I do have a 5% allocation to Berkshire.

I would disagree that gold got replaced by crypto - the market is a big place and the appetite for alternative investments has room for both. Gold has value because it can act as currency in times of stress. During WWII, the US took payment for goods in the form of gold, because the GBP might be worthless if the Germans won. I would bet if Brazil wanted to buy F-15s the US would accept payment in gold, but would tell Brazil to pound sand if they tried to pay with the Real. Crypto has the potential to be a hard currency like gold, but I don't know which crypto it is going to be in 10 years - so it is not investible to me right now. 

I hear you about relying too much on back testing. Going 100% equities is just hoping the past will repeat as well. In the case of all the 100% VTI posters on the forum, they are focusing on the past 10-12 years. That is why I really like what @Tyler is doing at portfolio charts. Over the last 40 years, diversification across asset classes has resulted in not only lower volatility, but superior returns. His newest article points out that portfolios with only two asset classes, equities and bonds, underperformed those that added alternatives. Gold is one of the easiest alternatives to add, but commodities, precious metals, REITs, MLPs, cyrpto etc. would also round out a allocation to alternatives.     

[Blender Bender]

I would disagree that gold got replaced by crypto - the market is a big place and the appetite for alternative investments has room for both. Gold has value because it can act as currency in times of stress. During WWII, the US took payment for goods in the form of gold, because the GBP might be worthless if the Germans won. I would bet if Brazil wanted to buy F-15s the US would accept payment in gold, but would tell Brazil to pound sand if they tried to pay with the Real. Crypto has the potential to be a hard currency like gold, but I don't know which crypto it is going to be in 10 years - so it is not investible to me right now. 

Personally I consider "crypto" a madness/religion. Perhaps crypto2.0, that is backed by governments, will take over at some moment (but would not be much speculatable/investable since should be stable, more like for wealth preservation than growth).
But if there is the religion/faith growing, then the sentiment affects negatively the gold future IMHO. Like kicking contest with a horse.

I hear you about relying too much on back testing. Going 100% equities is just hoping the past will repeat as well. In the case of all the 100% VTI posters on the forum, they are focusing on the past 10-12 years.

I don't think that static 100% VTI for very long term allocation is "fully efficient" (but is not bad per se). But for the next, i guess, few years it seems a solid bet. "Guaranteed TINA" https://markets.businessinsider.com/news/stocks/stock-market-outlook-2022-fundstrat-tom-lee-upside-3-reasons-2021-12

I think that some Gold might have its place as one of the AA for someone that does not have income anymore to protect against SORR. At this moment several additional AA (credit lines, REIT, Gold, some bonds etc) would make sense, but the primary reason to counter SORR. I would not call it investment per se.

I know that you don't like bonds, but this is a general statement:
Investing definition:
"Expend money with the expectation of achieving a profit or material result by putting it into financial plans, shares, or property, or by using it to develop a commercial venture."

Having bonds in general for the next foreseeable future does not fulfill the definition IMHO.

Knowing the history 100% is valuable, just extrapolating it could be very questionable.

UPDATE, adding anecdote:
A friend of mine, in the last 10 years, has put all his $ to hold physical gold. Some of it as jewelry...
He loved to talk about it every time. He really considered himself a (smart) investor.
Once i asked why they just not get VTI. His wife got upset. She said that once they bought a stock of a company a friend of theirs recommended. Too upset to share details. They lost almost all, and sweared never get involved in stock market ever since.

[index]

I know that you don't like bonds, but this is a general statement:
Investing definition:
"Expend money with the expectation of achieving a profit or material result by putting it into financial plans, shares, or property, or by using it to develop a commercial venture."

Having bonds in general for the next foreseeable future does not fulfill the definition IMHO.

When you invest in property or commercial business you buy insurance to protect yourself from black swan events.  Having a diversified portfolio is just insurance, and done correctly, it actually enhances your return on investment.

There is no where else is attempting to justify further equity inflation. There is always somewhere else.  The debt market is multiple times the size of the stock market and can quickly soak up all the money floating around if there is another shock of some sort. If double digits returns in equities are guaranteed the next few years then we are all idiots if we don't go 100% UPRO tomorrow!