Pick your diversifiers wisely - rant against bonds

[MustacheAndaHalf]

At the risk of feeding the troll...

<snip because irrelevant>

<back on ignore>

ditto.

You replaced my comments about gold's price from 1972-1974 with "snip because irrelevant", but notice you did not refute anything I said.  The U.S. left the gold standard in 1971, before which the price of gold was fixed.  Gold rose to its actual value from 1972-1974, which can't happen again.  If 1972-1974 is "irrelevant", why does vand always include those years in gold's performance?

[AnotherEngineer]

Please take this as a lighthearted step back out of the weeds. I love a good dataset as much as the next guy and running the numbers in ever increasingly complex ways is a common pastime.

The more I dig into SWGs with BigERN, Tyler, folks on the ChooseFI Facebook group, etc, the more wisdom I see in MMM/JLCollins/MadFientist and others not overcomplicating SWR. If your FIRE date depends on the start date of the backtest of your drawdown allocation relative to the end of the gold standard, you are either cutting it too close, suffering analysis paralysis, or misusing your early retirement (go outside or something)!

Your SWR has so many assumptions and uncertainties in it, chiefly how you define the "S" given the huge possible range of market performance but also unknown possible future changes to the ACA, capital gains taxes, FAFSA rules and college costs, tax brackets, climate-change induced insurance spikes, AI and automation, etc.  plus personal factors such as health, travel, family obligations, etc. Any single one of these could break a 50 year forecast.

MMM's approach to consider 4% but really be flexible and make some side income and have a healthy dose of optimism seems the way to do it, even if one can argue that 5% is achievable with the golden butterfly allocation.

No one knows the future and the whole SWR process is throwing a confidence interval over a wide range of past performances with many assumptions thrown in, which we all know don't guarantee future returns. Do your own math and choose an allocation that works for you. I think there is a good reason that JLCollins book and MMM's best post both have Simple in the title.

[dandarc]

@MustacheAndaHalf  Ditto was for <back on ignore>. And if you don't know what that means, it means I needed to do an extra click to see any of your posts.

[ChpBstrd]

Please take this as a lighthearted step back out of the weeds. I love a good dataset as much as the next guy and running the numbers in ever increasingly complex ways is a common pastime.

The more I dig into SWGs with BigERN, Tyler, folks on the ChooseFI Facebook group, etc, the more wisdom I see in MMM/JLCollins/MadFientist and others not overcomplicating SWR. If your FIRE date depends on the start date of the backtest of your drawdown allocation relative to the end of the gold standard, you are either cutting it too close, suffering analysis paralysis, or misusing your early retirement (go outside or something)!

Your SWR has so many assumptions and uncertainties in it, chiefly how you define the "S" given the huge possible range of market performance but also unknown possible future changes to the ACA, capital gains taxes, FAFSA rules and college costs, tax brackets, climate-change induced insurance spikes, AI and automation, etc.  plus personal factors such as health, travel, family obligations, etc. Any single one of these could break a 50 year forecast.

MMM's approach to consider 4% but really be flexible and make some side income and have a healthy dose of optimism seems the way to do it, even if one can argue that 5% is achievable with the golden butterfly allocation.

No one knows the future and the whole SWR process is throwing a confidence interval over a wide range of past performances with many assumptions thrown in, which we all know don't guarantee future returns. Do your own math and choose an allocation that works for you. I think there is a good reason that JLCollins book and MMM's best post both have Simple in the title.

I keep being drawn back to the idea that all these detailed backtests describe what would happen if the legal, economic, cultural, valuation, and foreign affairs situations were even remotely like they were in the past.

However, compared to 50 or 60 years ago, the legal environment is radically different (IRA's, tax brackets and loopholes, regulations), the economy functions in a very different way (electronic trading and payments, different debt levels, QE/QT, inflation targeting, dominance of information vs. railroad industries), the culture is radically different (women as independent workers, technological changes, longer lifespans, higher literacy), valuations are much higher (CAPE=37 vs. a long-term median of 16), and the foreign situation being much different (a plethora of developed capitalistic and industrialized economies, reduced manufacturing dominance in developed countries, much lower military spending compared to GDP, massive international trade).

Yes, these factors changed throughout the history of the U.S. stock market, and one could say that at any given year during the past we were in uncharted territory. But throughout history things happened which had never happened before. The Great Depression was unprecedented. So were the boom years of rapid change during the 50s and 60's. So was the 1970's stagflation. So was Paul Volker in the 80s. So was the introduction of the personal computer in the 90's, and so on.

Now that everything is configured differently than in the past, why shouldn't we expect a continuation of the trend of all past trends being broken?

It's like economic theories / SWR studies are trying to nail down with scientific precision how a few billion humans are going to behave in response to hundreds of factors which constantly change in ways they've never been before. We already know the actual results that occur in the future will be a unique combination of influences, each of which have never been seen in history. Yet we're trying to invent rules and formulas that would have the same predictive value in 2025 as they would in 1925.

I suspect the failure of economics has something to do with this theory-based approach, and the future might look something like a massive multiplayer simulation with AI's making economic choices to simulate outcomes.

[AnotherEngineer]

It's like economic theories / SWR studies are trying to nail down with scientific precision how a few billion humans are going to behave in response to hundreds of factors which constantly change in ways they've never been before. We already know the actual results that occur in the future will be a unique combination of influences, each of which have never been seen in history. Yet we're trying to invent rules and formulas that would have the same predictive value in 2025 as they would in 1925.

I suspect the failure of economics has something to do with this theory-based approach, and the future might look something like a massive multiplayer simulation with AI's making economic choices to simulate outcomes.

By coincidence, I'm reading Issac Asimov's Foundation series right now where the protaganist/hero is tasked with making just this sort of prediction of everything.

[ATtiny85]

At the risk of feeding the troll...

<snip because irrelevant>

<back on ignore>

ditto.

You replaced my comments about gold's price from 1972-1974 with "snip because irrelevant", but notice you did not refute anything I said.  The U.S. left the gold standard in 1971, before which the price of gold was fixed.  Gold rose to its actual value from 1972-1974, which can't happen again.  If 1972-1974 is "irrelevant", why does vand always include those years in gold's performance?

I keep checking back here hoping to hear a response to your query. The 71 move really makes the following years something unique.

[Telecaster]

By coincidence, I'm reading Issac Asimov's Foundation series right now where the protaganist/hero is tasked with making just this sort of prediction of everything.

Funny, I was thinking of Foundation when I read that as well.  Asimov was a PhD chemist.  There is a branch of chemistry and physics call statistical mechanics.   Basically, the idea is that you can't predict what any one atom or molecule will do, but based on the properties of a single atom, you can predict how the entire system will behave.   In Foundation, Asimov took this concept and applied it to people, which is how Hari Seldon was able to predict the collapse of the Galactic Empire.

[GuitarStv]

By coincidence, I'm reading Issac Asimov's Foundation series right now where the protaganist/hero is tasked with making just this sort of prediction of everything.

Funny, I was thinking of Foundation when I read that as well.  Asimov was a PhD chemist.  There is a branch of chemistry and physics call statistical mechanics.   Basically, the idea is that you can't predict what any one atom or molecule will do, but based on the properties of a single atom, you can predict how the entire system will behave.   In Foundation, Asimov took this concept and applied it to people, which is how Hari Seldon was able to predict the collapse of the Galactic Empire.

I liked how his predictions got all messed up when the mutant came into the picture.  The foundation series was one of my favourite reads in elementary school.

[MustacheAndaHalf]

At the risk of feeding the troll...

<snip because irrelevant>

<back on ignore>

ditto.

You replaced my comments about gold's price from 1972-1974 with "snip because irrelevant", but notice you did not refute anything I said.  The U.S. left the gold standard in 1971, before which the price of gold was fixed.  Gold rose to its actual value from 1972-1974, which can't happen again.  If 1972-1974 is "irrelevant", why does vand always include those years in gold's performance?

I keep checking back here hoping to hear a response to your query. The 71 move really makes the following years something unique.

It's unfortunate there was no constructive discussion.  I suppose the hit to gold's performance is so dramatic, that it becomes harder to promote it without the years right after the gold standard ended in the U.S.

I left out the all-time best year for gold, 1978, when its price more than doubled.  That debate could be more nuanced, with inflation + oil crisis + speculative frenzy all playing a role.  Removing that decade cuts gold's performance by more than half, so its an important discussion for those considering gold in their retirement portfolio.

1972 - 2024: gold 7.9%/year
1975 - 2024: gold 5.3%/year
1980 - 2024: gold 3.5%/year

[ChpBstrd]

Gold is a weird asset with price moves that seem not to fit the paradigms often used to justify it.

If anyone thinks gold is an inflation hedge, please explain the 100% jump in GLD prices between the deflationary year 2008 and the near-zero-inflation 2011. Then explain how GLD puttered around within a tight range during the high inflation years of 2021 to 2023, only to skyrocket once it became clear by late 2023 that inflation was getting under control.

The counterintuitive slogan "gold goes up when inflation goes DOWN" fits the data as well as the "gold is an inflation hedge" narrative. The historical trends are all over the place.

As a theoretical diversifier or rebalancing tool, what do we say about GLD's 15% drop between March 2022 and October 2022 - in sync with a historic meltdown in bonds and a bear market in stocks?

I'd be a lot more comfortable with gold if we had a working theory of action explaining what it does, or how it plays a role in a portfolio going forward. It's not an inflation or interest rates hedge. Even the theory that gold is a random returns generator is contradicted by recent pro-cyclical action.

Could it be that gold's price is driven by the same sorts of market flows and media discussions that drive the price of other non-earning, greater-fool-theory investments like Dogecoin, NFTs, baseball cards, or fine art? Intelligent investors tend to ignore such asset classes, except for some reason when they are named gold.

[Scandium]

Gold is a weird asset with price moves that seem not to fit the paradigms often used to justify it.

If anyone thinks gold is an inflation hedge, please explain the 100% jump in GLD prices between the deflationary year 2008 and the near-zero-inflation 2011. Then explain how GLD puttered around within a tight range during the high inflation years of 2021 to 2023, only to skyrocket once it became clear by late 2023 that inflation was getting under control.

The counterintuitive slogan "gold goes up when inflation goes DOWN" fits the data as well as the "gold is an inflation hedge" narrative. The historical trends are all over the place.

As a theoretical diversifier or rebalancing tool, what do we say about GLD's 15% drop between March 2022 and October 2022 - in sync with a historic meltdown in bonds and a bear market in stocks?

I'd be a lot more comfortable with gold if we had a working theory of action explaining what it does, or how it plays a role in a portfolio going forward. It's not an inflation or interest rates hedge. Even the theory that gold is a random returns generator is contradicted by recent pro-cyclical action.

Could it be that gold's price is driven by the same sorts of market flows and media discussions that drive the price of other non-earning, greater-fool-theory investments like Dogecoin, NFTs, baseball cards, or fine art? Intelligent investors tend to ignore such asset classes, except for some reason when they are named gold.

Thank you, said better than I could what my general feelings are about gold. The market moves of gold seems to me to just be as a speculative asset; crypto for boomers... And occasional "gold goes up when uncertainty", simply because everyone knows gold goes up with uncertainty.. The theory is sustained simply by it own self-fulfilling nature. And as you point out, that's not even always the case.
Looking from 90s till today, I don't see gold doing particularly well during recessions?   
https://www.macrotrends.net/2608/gold-price-vs-stock-market-100-year-chart

I'm still curious which timeframe a 50% gold portfolio sustains a 5%+ SWR?? (that's not early 1970s..)

[tooqk4u22]

Gold is a weird asset with price moves that seem not to fit the paradigms often used to justify it.

If anyone thinks gold is an inflation hedge, please explain the 100% jump in GLD prices between the deflationary year 2008 and the near-zero-inflation 2011. Then explain how GLD puttered around within a tight range during the high inflation years of 2021 to 2023, only to skyrocket once it became clear by late 2023 that inflation was getting under control.

The counterintuitive slogan "gold goes up when inflation goes DOWN" fits the data as well as the "gold is an inflation hedge" narrative. The historical trends are all over the place.

As a theoretical diversifier or rebalancing tool, what do we say about GLD's 15% drop between March 2022 and October 2022 - in sync with a historic meltdown in bonds and a bear market in stocks?

I'd be a lot more comfortable with gold if we had a working theory of action explaining what it does, or how it plays a role in a portfolio going forward. It's not an inflation or interest rates hedge. Even the theory that gold is a random returns generator is contradicted by recent pro-cyclical action.

Could it be that gold's price is driven by the same sorts of market flows and media discussions that drive the price of other non-earning, greater-fool-theory investments like Dogecoin, NFTs, baseball cards, or fine art? Intelligent investors tend to ignore such asset classes, except for some reason when they are named gold.

I have never been a gold enthusiast mostly because it has no real purpose other than being a a shiny object......its only real use is for jewelry, no cash flow, no real industrial/commercial use.....just a big powerful psychological history that goes back before any form of monetary system.   But now that more data is being brought to the forefront as it is the belle of the ball right now is quite interesting regarding returns.

So a couple of thoughts or points about your comment:

2008 - Financial Crisis/Great Recession.
It was all about systemic risk, where capital markets overall and specifically bond markets were potentially imploding followed by what was at the time record stimulus (it looks like a rounding error compared to the pandemic stimulus).  That trade was not about inflation, it was about fear and moving to a perceived safer physical asset, which is why it almost entirely reversed once markets stabilized, stimulus stopped and some got repaid, and growth came back.

2021-2023 - Flat period
- yes it was flat during that period but a lot of the damage had already been done as you are ignoring the 24% rise in gold from 2/2020 (pre-pandemic) to 8/2020, which was a combination of fear (world had stopped) and stimulus (so much money).
- It was flat after that because fear was leaving the system (world didn't end and was actually getting better) but it was offset by continued excessive and unnecessary stimulus - if not for the additional and ongoing stimulus it is likely gold would have reverted mostly back to the pre-pandemic level (not all the way because of the initial stimulus). 
- Basically the initial 24% run up matches what ended up to be about the same inflation over that time. 

2023 to Now - a lot of shit contributing to the massive run up (+/-38%)
- China, and secondarily India, reduced treasuries and bolstered gold reserves
        -China alone increased gold reserves (tonnes not value) by 16% from end of Q3 2022 to Q1 2024 (flat in Q2 2024)
- fear trade from increased geopolitical risk (Russia/Ukraine, Israel vs the rest of middle-east)
- inflation may be coming down but is persisting combined with small political race in the US where both parties are espousing policies that have significant inflationary/deficit pressures. 
- There is also a bit of a momentum trade going on.

Also there is just way too much liquidity in the system. 

I honestly have no idea where it goes from here, but I would expect gold to recede with conflicts being de-escalated or ended, a divided government so policies can't be approved (or at the very least the parties have to work together), and possibly if China gets going again (while inflationary it could force them to have to sell gold and go back into US treasuries).   

[AnotherEngineer]

I have never been a gold enthusiast mostly because it has no real purpose other than being a a shiny object......its only real use is for jewelry, no cash flow, no real industrial/commercial use

I might as well build my reputation for pedantry, but I was curious and it is worth sharing. Gold is used in electronics and aerospace because it conducts and doesn't corrode, but only tiny amounts in connectors. Maybe that is what you meant by "no real purpose". Nearly half of gold is used in jewelry, one quarter is bought for investments, one fifth is acquired by banks, and only 6% is used in industry. Such a weird mix for one material!

The interesting thing about the 6% is that most of that is not recovered due to poor electronics recycling, so that amount is lost when the rest is maintained (minus lost jewelry).

https://www.statista.com/statistics/299609/gold-demand-by-industry-sector-share/

[ChpBstrd]

That is interesting @AnotherEngineer and may explain why the price of gold seems not to have much to do with anything. There are just too many factors all changing at any given time for that particular commodity.

It makes me wonder whether silver, platinum, palladium or other expensive, invest-able metals might be more correlated with inflation. If a thing has more industrial uses, and relatively less interest from goldbug investor culture, jewelers, central banks, etc, then maybe its price reflects the constantly-rising costs of mining, which might be a good or bad proxy for inflation at the consumer level.

Maybe if I find the time I'll correlate these prices with the Producer Price Index or Consumer Price Index, or see if someone else has done it.

[tooqk4u22]

I have never been a gold enthusiast mostly because it has no real purpose other than being a a shiny object......its only real use is for jewelry, no cash flow, no real industrial/commercial use

I might as well build my reputation for pedantry, but I was curious and it is worth sharing. Gold is used in electronics and aerospace because it conducts and doesn't corrode, but only tiny amounts in connectors. Maybe that is what you meant by "no real purpose". Nearly half of gold is used in jewelry, one quarter is bought for investments, one fifth is acquired by banks, and only 6% is used in industry. Such a weird mix for one material!

The interesting thing about the 6% is that most of that is not recovered due to poor electronics recycling, so that amount is lost when the rest is maintained (minus lost jewelry).

https://www.statista.com/statistics/299609/gold-demand-by-industry-sector-share/

Yes, that is what I meant - 6% is basically nothing and the half that is used in jewelry falls under the shiny object category.

Although it is interesting that the statista data indicated that the  global production of gold is 3,000 metric tons and  jewelry industry uses 2,168 metric tons, which means that in reality the jewelry accounts for 72% of gold production.

That is interesting @AnotherEngineer and may explain why the price of gold seems not to have much to do with anything. There are just too many factors all changing at any given time for that particular commodity.

It makes me wonder whether silver, platinum, palladium or other expensive, invest-able metals might be more correlated with inflation. If a thing has more industrial uses, and relatively less interest from goldbug investor culture, jewelers, central banks, etc, then maybe its price reflects the constantly-rising costs of mining, which might be a good or bad proxy for inflation at the consumer level.

Maybe if I find the time I'll correlate these prices with the Producer Price Index or Consumer Price Index, or see if someone else has done it.

Gold is an inflation hedge only because of history and treating it as a store of value that goes back thousands of years......its the lizard brain in us.  It's a shiny object that is just rare enough.

Those other precious metals have more productive uses but act more like other commodities like copper - so inflation impacts them but so does supply/demand.   

[MustacheAndaHalf]

2021-2023 - Flat period
- yes it was flat during that period but a lot of the damage had already been done as you are ignoring the 24% rise in gold from 2/2020 (pre-pandemic) to 8/2020, which was a combination of fear (world had stopped) and stimulus (so much money).
- It was flat after that because fear was leaving the system (world didn't end and was actually getting better) but it was offset by continued excessive and unnecessary stimulus - if not for the additional and ongoing stimulus it is likely gold would have reverted mostly back to the pre-pandemic level (not all the way because of the initial stimulus). 
- Basically the initial 24% run up matches what ended up to be about the same inflation over that time. 

I suppose each of us look at this differently.

The Consumer Price Index (CPI) is a common measure of inflation that the Fed watches.  In 2020, CPI rose 1.2% while gold rose 24.8%.  A very small level of inflation, and a very big jump for gold.  And yet in 2022, CPI rose +8% while gold fell 4%.  A big jump in inflation, and gold took losses.  What I see is individual years where inflation was unrelated to gold's moves.

As to fear, perhaps consumers felt it differently than investors.  The S&P 500 lost 34% from Feb 19 to Mar 23 2020.  During that time, gold fell 3.6%.  During the point of maximum fear, gold did not rise.  Similarly, when the market surged from Mar 23 to Dec 31 2020, gold surged as well (from the Mar 23 bottom: +67% market, +21% gold).  Gold moved in the same direction as the market (by smaller amounts).  Relative to investors, gold didn't do much to diversify the fear of Covid-19.


(Historical CPI, S&P 500 and GLD ETF prices below)
https://www.minneapolisfed.org/about-us/monetary-policy/inflation-calculator/consumer-price-index-1913-
https://finance.yahoo.com/quote/SPY/history/?period1=1577836800&period2=1609372800
https://finance.yahoo.com/quote/GLD/history/?period1=1577836800&period2=1609372800

[Rockies]

I've had a 60/40 vanguard fund for the past 15 years. Its made a lot of gains since I purchased it and its in an unregistered account.

I kindof want to get out of it because I dont beleive in bonds. Of course the bond percentage slowly shrinks because I purchase other investments over the years.

So far I havn't sold it because I dont want to realize all those capital gains taxes to convert the 60/40 fund to a 100% stock index fund. How would others approach this? Just leave it and let the returns drag for the next 15 years, or bite the bullet and sell it for something that will likely do better?

[markbike528CBX]

I've had a 60/40 vanguard fund for the past 15 years. Its made a lot of gains since I purchased it and its in an unregistered account.

I kindof want to get out of it because I dont beleive in bonds. Of course the bond percentage slowly shrinks because I purchase other investments over the years.

So far I havn't sold it because I dont want to realize all those capital gains taxes to convert the 60/40 fund to a 100% stock index fund. How would others approach this? Just leave it and let the returns drag for the next 15 years, or bite the bullet and sell it for something that will likely do better?

So you are avoiding capital gains taxes now in the hope of having more capital gains/taxes in the future?
Your capital gains taxes might not be that bad, either now or in the future.
https://engaging-data.com/tax-brackets/

[vand]

Returning to the conundrum of what is the best portfolio mix when you consider multiple different diversifying assets, I used the PortfolioCharts WithDrawal Rates page (https://portfoliocharts.com/charts/withdrawal-rates/#chart) tested combinations of 4 assets:

US LCB
Gold
REITs
10y bonds

Theoretically working in 5% allocation step increments, 2 different assets gives us 21 different combinations, 3 assets gives us 231 possible combinations and 4 assets gives us a total of... 1771 possible combinations. That's a lof of combinations to test!

Fortunately we don't need to test every one of these , as we already know what the best combination with just 2 assets, (65/35 stocks gold) we can take the equity allocation of that as our base scenario and introduce a 3ed asset, find the best combination for 3 assets, then use that as a base case to test the effects of the 4th asset...

In short, here's what I came up with..




Key takeaways

- If considering just 2 assets (stocks & gold) the best combination is 65/35, which supports a SWR of 5.7%. No other asset by itself will diversify a stock portfolio as well as gold. (from opening post).

- Adding in the next best asset (REITS) is very worthwhile. While REITS by themselves don't diversify as well as gold, a combination of REIT+Gold is hugely beneficial to the portfolio, and in fact the optimal portfolio composed of these 3 assets will hold more REITS than gold in a 45/25/30 stock/gold/REIT mix.

- the SWR continues to improve as you reduce equity expose from 65% (and by inference from 100%) and allocate it amongst diversifiing assets. This is true until we get all the way down to a 45% equity allocation, at which point the starts to go the other way again, so 45% is the optimum equity allocation when considering 3 or more assets.

- Taking our 45/25/30 best 3-asset portfolio and modifying it with an allocation to 10yr bonds fails to improve portfolio. With a small 5% allocation (taken from stocks or REITs) the SWR remains at 6.4% but the perpetual withdrawal rate (PWR - which I look at as a sort of tie-breaker) falls by 0.1-0.2%.  A 10% allocation sees a more deleterious reduction on the SWR.  The reduction in portfolio efficiency is more pronounced if you take it from the gold allocation than from the REIT or Stock allocation.

Baseline 100% US LCB: 3.8% SWR
Best Stock/Gold combination: 65/35, 5.7% SWR
Best Stock/REIT combination: 40/60, 5.1% SWR
Best Stock/Gold/REIT combination: 45/25/30, 6.4% SWR
Best Stock/Gold/REIT/10YUST combination: 45/25/30/0, 6.4% SWR


Personally, these findings were revealing to me, not for the reaffirmation that it's difficult to make a case for bonds on the testing numbers presented here, but how well the combination of REITS and Gold works (considerably better than either by themselves).  I do hold some REITs already.. and will definitely be increasing my exposure to them going forward.

[BringFuturamaBack]

I understand that the portfolio generator shows gold is incredibly valuable for diversification but in a world where it's at an all time high (accounting for inflation) and it produces nothing, I don't think it's a great time to add. I get not timing the market as it should continually improve and people become more efficient....I'm not convinced it's the same for gold.

[vand]

I understand that the portfolio generator shows gold is incredibly valuable for diversification but in a world where it's at an all time high (accounting for inflation) and it produces nothing, I don't think it's a great time to add. I get not timing the market as it should continually improve and people become more efficient....I'm not convinced it's the same for gold.

Gold has been inconveniently making new highs while producing nothing for 5000 years at least, so I think other points of view should be carefully considered :)

And if "making new highs" is a reason to avoid, then gee whiz, I guess you better dump all your stocks as well

[ChpBstrd]

I understand that the portfolio generator shows gold is incredibly valuable for diversification but in a world where it's at an all time high (accounting for inflation) and it produces nothing, I don't think it's a great time to add. I get not timing the market as it should continually improve and people become more efficient....I'm not convinced it's the same for gold.

I always try to think about the specifics of what happened during the historical Sequence of Returns Risk events that determine our SWR's. I'd be willing to bet REITs contributed positively to SWR during the inflationary 1970s, when they were a relatively new market product that was priced cheaply.

Today, REITs are relatively expensive when we compare their yields to treasuries. VNQ's yield is 3.85% compared to 4.425% on the 7 year treasury note. One buys extra business risk in exchange for getting less yield, and the only rationale I can imagine is the expectation for faster earnings growth.

Gold is a completely different investment than it was in the early 70's, with easy-to-trade funds like IAU or GLD available for free trading at the push of a button. Gold only became available to retail investors in the 1970s and you had to find a way to buy it, transport it safely, insure it, and rent a safe deposit box for it (these real-world costs are never included in the backtesting return numbers). To have actually purchased physical gold back then would have been as weird as if today you kept a couple tons of copper or tin on pallets in the back yard. The convenience of buying gold ETFs may have run up the price a bit, but it's not a factor that is expected to change.

So we have two investment classes that became available in the late 60's / early 70's and did well right around the time stocks did poorly. What is the case for that happening again in a similarly stagflationary late 2020's? I could make a case that it won't happen that way again, because (1) REITs and gold already have inflated values, (2) they are no longer new or esoteric portfolio options that most people are unaware of, and (3) high and rising interest rates might make 1-5 year treasury notes or TIPS (another invention of recent decades) an attractive alternative.

Then there is the issue that it's easy to use options to directly hedge a portfolio against the type of SORR events we are worried about. This is a direct hedge, not some theoretical link between inflation and gold (there isn't one), or theoretical outperformance of REITs during inflationary/recessionary times. Theoretical relationships can fail, but options are firm contracts, backed by exchanges.

Yet we cannot put, for example, a +20%/-15% collar strategy into a backtest because there is no dataset, and because like gold or real estate, the really old data would require a caveat for the lack of realism. People in 1970 could not whip out their smartphones and buy puts. In fact, they'd need a Wall Street connection, a massive lot size, and be willing to pay a four-figure commission on top of ridiculous bid-ask spreads. LEAPS didn't even exist.

A lack of data is why we don't talk about options strategies to mitigate SORR years, but that doesn't mean they aren't the optimal strategies.

[AnotherEngineer]

I understand that the portfolio generator shows gold is incredibly valuable for diversification but in a world where it's at an all time high (accounting for inflation) and it produces nothing, I don't think it's a great time to add. I get not timing the market as it should continually improve and people become more efficient....I'm not convinced it's the same for gold.

I always try to think about the specifics of what happened during the historical Sequence of Returns Risk events that determine our SWR's. I'd be willing to bet REITs contributed positively to SWR during the inflationary 1970s, when they were a relatively new market product that was priced cheaply.

...

A lack of data is why we don't talk about options strategies to mitigate SORR years, but that doesn't mean they aren't the optimal strategies.

Hear! Hear! Moving forward into unprecedented times (as every decade has been before this one too), backtesting has an important but limited value as does even the Trinity study. I'm as data-driven as anyone (see my handle), but there is so much subtlety and context to consider. Portfolio Charts is fascinating, but you have to consider so much (e.g., that it shows 15th percentile values). All of this SWR talk is magnified by SORR as noted, which is largely due to unknowable factors even in the short term. It is valuable to know the range of potential futures through something like cfiresim, but as for me, I'm crafting my FIRE plan on first principles of what stocks, bonds, and commodities like gold fundamentally are with some optimism and flexibility baked in. Others should do whatever makes them able to both retire and sleep at night, but I would suggest that endless AA simulation and even reading too much ERN, God bless him, may have diminishing returns if not be counterproductive.

Edit: I don't have much to say about options as I have learned enough to know I have no interest in them. My point is about how backtesting is fraught due to a very short timeframe of current AA choices and the larger economic trends of those times.

[vand]

I understand that the portfolio generator shows gold is incredibly valuable for diversification but in a world where it's at an all time high (accounting for inflation) and it produces nothing, I don't think it's a great time to add. I get not timing the market as it should continually improve and people become more efficient....I'm not convinced it's the same for gold.

I always try to think about the specifics of what happened during the historical Sequence of Returns Risk events that determine our SWR's. I'd be willing to bet REITs contributed positively to SWR during the inflationary 1970s, when they were a relatively new market product that was priced cheaply.

...

A lack of data is why we don't talk about options strategies to mitigate SORR years, but that doesn't mean they aren't the optimal strategies.

Eh? Bullshit I say, bullshit.

Why? rewind back to first principals:

A well diversified portfolio of options has a total expected nominal return of exactly 0.0%.  And this is not really debatable if you understand what an option is - it's a contract where one party's loss (the option buyer) is another party's gain (the option writer's).  So it's HIGHLY debateable how an asset class that loses purchasing power over the long term to inflation can enhance the long term risk/reward characteristics or improve the survivability of a drawdown portfolio.

I know some people love their options trading like its some dark art magically enhanching your portfolios and beyond the understanding of us plebs, but if you can't historically backtest or prove it, then it's just a bullshit theory. Conversely, it's demonstrable - even unarguable - that gold and other assets can diversify a stock portfolio to improve risk/return in a measurable, testable and demonstrable way... which is what gives it validity and moves it beyond being just a theory and towards something that is actually useful.

[ChpBstrd]

I understand that the portfolio generator shows gold is incredibly valuable for diversification but in a world where it's at an all time high (accounting for inflation) and it produces nothing, I don't think it's a great time to add. I get not timing the market as it should continually improve and people become more efficient....I'm not convinced it's the same for gold.

I always try to think about the specifics of what happened during the historical Sequence of Returns Risk events that determine our SWR's. I'd be willing to bet REITs contributed positively to SWR during the inflationary 1970s, when they were a relatively new market product that was priced cheaply.

...

A lack of data is why we don't talk about options strategies to mitigate SORR years, but that doesn't mean they aren't the optimal strategies.

Hear! Hear! Moving forward into unprecedented times (as every decade has been before this one too), backtesting has an important but limited value as does even the Trinity study. I'm as data-driven as anyone (see my handle), but there is so much subtlety and context to consider. Portfolio Charts is fascinating, but you have to consider so much (e.g., that it shows 15th percentile values). All of this SWR talk is magnified by SORR as noted, which is largely due to unknowable factors even in the short term. It is valuable to know the range of potential futures through something like cfiresim, but as for me, I'm crafting my FIRE plan on first principles of what stocks, bonds, and commodities like gold fundamentally are with some optimism and flexibility baked in. Others should do whatever makes them able to both retire and sleep at night, but I would suggest that endless AA simulation and even reading too much ERN, God bless him, may have diminishing returns if not be counterproductive.

Eh? Bullshit I say, bullshit.

Please rewind back to first principals:

A well diversified portfolio of options has a total expected nominal return of exactly 0.0%.  And this is perfectly arguable if you understand what an option is - one party's loss (the option bearer) is another party's gain (the option writer's).  It's HIGHLY debateable how an asset class that loses purchasing power over the long term to inflation can enhance the long term risk/reward characteristics or improve the survivability of a drawdown portfolio.

I know some people love their options trading like its some dark art magically enhanching your portfolios and beyond the understanding of us plebs, but if you can't historically backtest or prove it, then it's just a bullshit theory.

Are you thinking people are writing an IPS and saying "I will invest xx% of my money in 'options'."? Pretty sure the only people doing that are on Wall Street Bets, dropping their life savings into 0DTE calls on Palantir.

With the long-term collars I advocate, my asset allocation to options is approximately a net 0%. But I obtain lower portfolio volatility plus ironclad insurance against SORR. Unless the market goes waaay up or waaay down, my outcome upon expiration will be the same as someone who just put 100% into the index fund.

You are correct that derivatives are a zero-sum game. That's why I used shares+options which is not a zero sum game. The combination allows investors to customize their range of possible returns, which cannot be done with any other AA. 

[AJDZee]

Just a quick thank you to the OP for starting this thread.
Your arguments convinced me, and at the end of last year my asset rebalancing (after 2 consecutive years of crazy returns for stocks) included adding a large gold position, and it's worked out wonderfully :)

[vand]

Just a quick thank you to the OP for starting this thread.
Your arguments convinced me, and at the end of last year my asset rebalancing (after 2 consecutive years of crazy returns for stocks) included adding a large gold position, and it's worked out wonderfully :)

Everyone's gotta find a balance that works for them. Just to show that it's ok not to take your own advice too literally so long as you have taken all sides of the debate seriously,  I have actually been buying bonds again more recently, having had zero exposure since 2021. 

Hoping to take my exposure up to an initial 5% or so of total portfolio value - 2/3rds in long duration gilts currently yielding 4.5% and 1/3 in EM bonds that currently yielding about 6.5%.

Scaling back slightly in my global equity and gold miners exposure.  At the end of the day it's probably not going to amount to much more than a rounding error in the overall annualised numbers.. but narrowing the range is the whole point..

[BicycleB]

A thought-provoking thread! Thanks to all contributors.

[blue_green_sparks]

Last time I painted our two-story home, I used a ladder. This time I bought scaffolding. If there is a next time, I will shell out the big bucks and hire professional painters.

Post FIRE, there is a certain value in having a chunk of income that is at least as reliable as my paycheck used to be, so I appreciate quality bonds. When calculating net worth, I disregard the market value fluctuations of my bonds because I plan hold to maturity.

Of course, now we have the added stress of increased odds that the republic may fail. At least there are many who believe this could happen. In that case, I doubt stocks would fare much better. I do think it is more likely that the US will persist as a democracy. A valuable lesson will be learned. I am not "praying" for my country because irrational belief in magic has much to do with the state of things right now. It does give me pause that a plague transformed the country into a wannabe theocracy wallowing in misinformation, authoritarianism, nationalism, isolationism and lawlessness.

[VanillaGorilla]

I've already noted my qualms with the data sources cited in this thread, so I won't reiterate those. However, in the spirit of making a comprehensive overview, when backtested to 1871, bonds do in fact raise the absolute failsafe withdrawal rate according to Karsten Jeske, see his chart. He shows that particularly at lower withdrawal rates, a 25% bond allocation has historically saved a portfolio for one or two historical events.

That's not hugely relevant for me personally, but it's worth noting, lest people get carried away considering returns since only 1972.

In agreement with the thesis of this thread, he finds that a 60/25/15 equity/bond/gold allocation outperforms 75/25 over 30 years.

[vand]

I've already noted my qualms with the data sources cited in this thread, so I won't reiterate those. However, in the spirit of making a comprehensive overview, when backtested to 1871, bonds do in fact raise the absolute failsafe withdrawal rate according to Karsten Jeske, see his chart. He shows that particularly at lower withdrawal rates, a 25% bond allocation has historically saved a portfolio for one or two historical events.

That's not hugely relevant for me personally, but it's worth noting, lest people get carried away considering returns since only 1972.

In agreement with the thesis of this thread, he finds that a 60/25/15 equity/bond/gold allocation outperforms 75/25 over 30 years.

I gave my thoughts on ERN's missive into gold allocation earlier in the thread:

https://forum.mrmoneymustache.com/investor-alley/pick-your-diversifiers-wisely-rant-against-bonds/msg3302893/#msg3302893

"I know he's considered adding gold to portfolios with the conclusion that it does improve SWRs slightly, but it was still from the baseline of "gold as the 3rd option".  What if he just removes the bond component and works from first principals, using gold as the go-to diversifer?"

Question is still valid.

However -- and this point really is crucial to understand about gold  and the reason I take umbrage with ERN/cfiresim/anyone else's conclusions -- if they talk about being able to holding gold during Bretton Woods era and draw any conclusions from it then it is not just invalid, it's misleading. 

If you are well versed in the history of money and gold, you will understand why all modern timeseries datasets that prices gold for the 1944-1971 period are essentially useless.   All these dataseries do is take the $35 price from 1944 and increase it with inflation.  In reality gold didn't trade during this period, therefore there was no mark to market price. 

The backfilled prices you can get on modern dataseries are just for convenience so that the likes of cfiresim can rather conveniently calculate over SWRs over 100yrs of history without crashing due to a div/0 error. 

[ChpBstrd]

There is a case to be made that this administration's policies tilt the calculus even more toward gold. Specifically I'm thinking of:

• Tariffs which could drive up inflation, reduce foreign demand for treasuries, and potentially reduce the USD's role as reserve currency.
• Much higher deficits after the coming tax bill, which could result in a bond market seizure or drive debt/GDP over 150%.
• Possible selective defaults of treasuries.
• Adding to national debt to fund cryptocurrency reserves (i.e. pump and dumps) or to supplant the U.S. currency with crypto. I do not see this increasing the confidence of foreign investors or traders.
• A reduction in the independence of the Federal Reserve, which could raise fears of a Turkish scenario where the currency is sabotaged for political gain.
• A reduction in rule-of-law, investor protections, and democracy, which could reduce perceptions of the USD as the safest currency to store wealth.

So the downside of bonds is that you could be paid back in depreciated dollars. And all the policies listed above have an effect of reducing the appeal of storing wealth in dollars/treasuries instead of Euros/eurobonds or Yen or Aussie dollars, etc.

The present value of a series of real cash flows would be strongly affected if you added some probabilistic risk of an unplanned sudden devaluation, such as a 20% chance of a 30% devaluation, for example. Such a risk, if investors started assigning it, could more than wipe out the real yield available now, and would increase the attractiveness of other sovereign debts.

Different investors will assign different odds and damage. Bonds (and equities) become a lot less appealing even if you apply even small odds of a small currency devaluation. As I scan for "intriguing investments of the day" I will be looking at non-dollar denominated assets to diversify currency risk.

[tooqk4u22]

There is a case to be made that this administration's policies tilt the calculus even more toward gold. Specifically I'm thinking of:

• Tariffs which could drive up inflation, reduce foreign demand for treasuries, and potentially reduce the USD's role as reserve currency.
• Much higher deficits after the coming tax bill, which could result in a bond market seizure or drive debt/GDP over 150%.
• Possible selective defaults of treasuries.
• Adding to national debt to fund cryptocurrency reserves (i.e. pump and dumps) or to supplant the U.S. currency with crypto. I do not see this increasing the confidence of foreign investors or traders.
• A reduction in the independence of the Federal Reserve, which could raise fears of a Turkish scenario where the currency is sabotaged for political gain.
• A reduction in rule-of-law, investor protections, and democracy, which could reduce perceptions of the USD as the safest currency to store wealth.

So the downside of bonds is that you could be paid back in depreciated dollars. And all the policies listed above have an effect of reducing the appeal of storing wealth in dollars/treasuries instead of Euros/eurobonds or Yen or Aussie dollars, etc.

The present value of a series of real cash flows would be strongly affected if you added some probabilistic risk of an unplanned sudden devaluation, such as a 20% chance of a 30% devaluation, for example. Such a risk, if investors started assigning it, could more than wipe out the real yield available now, and would increase the attractiveness of other sovereign debts.

Different investors will assign different odds and damage. Bonds (and equities) become a lot less appealing even if you apply even small odds of a small currency devaluation. As I scan for "intriguing investments of the day" I will be looking at non-dollar denominated assets to diversify currency risk.

I agree with all of this except that the equities, bonds, economies, currencies, and such of other countries aren't and haven't been that great for a long time, so yes it could cause significant decline in all of the US versions but there is still likely a floor and in reality while the US gets worse the other countries probably don't get better.   

But as it relates to gold, central banks all around the globe have been adding to gold reserves over last years or more - notably and not surprisingly Russia, China, Turkey and India and I don't thing that is ending anytime soon.

[BicycleB]

I agree that negative events in US politics, global politics and global trade could reduce the value of dollar-denominated assets without increasing the value of foreign-denominated assets - in terms of purchasable goods, a global currency basket, or gold!

And with that, gold's role as something that can be traded between currencies all over the world strengthens its role as a store of value. Despite the rising price and the possibility that BRICS' gold buying could lessen, which are risks that gold could be overpriced, it seems possible that gold is still a good value.

In particular, it probably does have a "keeps at least some value under a very wide range of circumstances" value that could help keep a floor under a portfolio in uncertain times.

[baconschteam]

I’m a naive and lazy investor and generally stick to a 90% VTI 10% VFMX mix. I’d like at some point to develop a more sophisticated strategy, and I’m always very interested to hear what those who’ve been around here for a while have to say. @ChpBstrd, I always love your logical and well laid out responses to things, and I’ll be looking into this collar strategy.

Naive perspective on gold: based on what ya’ll are saying, it appears to me that gold could be a valuable part of one’s portfolio because it really doesn’t correlate to anything else, it hasn’t followed a discernible pattern, and even if you can dispute whether its average growth has been anywhere from 2-6% after inflation, it adds the element of chaos to a logically assembled portfolio, in a chaotic world. Oh no, it’s non-productive and the value is just in people’s heads? Well, that’s how value is made in this world we live in. Enough people agree on it. I don’t think its value has that much to do with economics, it’s more immutable, historical, emotional, mystic.

[Telecaster]

Naive perspective on gold: based on what ya’ll are saying, it appears to me that gold could be a valuable part of one’s portfolio because it really doesn’t correlate to anything else, it hasn’t followed a discernible pattern, and even if you can dispute whether its average growth has been anywhere from 2-6% after inflation, it adds the element of chaos to a logically assembled portfolio, in a chaotic world. Oh no, it’s non-productive and the value is just in people’s heads? Well, that’s how value is made in this world we live in. Enough people agree on it. I don’t think its value has that much to do with economics, it’s more immutable, historical, emotional, mystic.

And historically it has sucked as an investment.   The inflation-adjusted return of gold for the last 45 years has been zero.  And before you point out those points are cherry picked, they absolutely are cherry picked--which is my whole point.  Entry and exits point matter a lot for gold returns.   That's true for stocks, but with stocks over any reasonably long period the entry and exit points become less and less important. 

I also echo @vand's caution that the data prior to 1972 aren't very good to the point of potentially not being useable for this purpose. 

[vand]

Naive perspective on gold: based on what ya’ll are saying, it appears to me that gold could be a valuable part of one’s portfolio because it really doesn’t correlate to anything else, it hasn’t followed a discernible pattern, and even if you can dispute whether its average growth has been anywhere from 2-6% after inflation, it adds the element of chaos to a logically assembled portfolio, in a chaotic world. Oh no, it’s non-productive and the value is just in people’s heads? Well, that’s how value is made in this world we live in. Enough people agree on it. I don’t think its value has that much to do with economics, it’s more immutable, historical, emotional, mystic.

And historically it has sucked as an investment.   The inflation-adjusted return of gold for the last 45 years has been zero.  And before you point out those points are cherry picked, they absolutely are cherry picked--which is my whole point.  Entry and exits point matter a lot for gold returns.   That's true for stocks, but with stocks over any reasonably long period the entry and exit points become less and less important. 

Any single asset can have long drawdowns.

Longest real drawdown for US stocks is about 26 years starting from 1929 peak.
Longest drawdown for US 10yr bonds is about 54 years between 1939 - 1992

This is why you should never have all your wealth in just one assset class. I suppose if you can stomach 26 years of going from 0% to -80% and clawing back to 0% then by all means put all your eggs in the best least worst option.   

But otherwise, you'd do well to take seriously how best to combine assets to give you the best balance -for which we return to top of the thread for a reasonable crack at the answer..