Great article from Portfolio Charts on most efficient portfolios

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Nathan Drake
Posts: 2861
Joined: Mon Apr 11, 2011 12:28 am

Re: Great article from Portfolio Charts on most efficient portfolios

Post by Nathan Drake »

HanSolo wrote: Sat Dec 25, 2021 3:59 pm
Nathan Drake wrote: Sat Dec 25, 2021 3:42 pm
HanSolo wrote: Sat Dec 25, 2021 2:13 pm
Nathan Drake wrote: Sat Dec 25, 2021 12:00 pm
HanSolo wrote: Sat Dec 25, 2021 3:38 am As you can see from my sig, I think there's nothing wrong with 60/40. In addition, I have certain tilts in my IRA, some of which are discussed in this thread. I don't have a strictly Boglehead-compliant portfolio (like 3-fund), but I think that, on the whole, the Boglehead ideas are pretty good and aren't out of date. The ideas of Bogle and related authors were formulated after having experience with a lot of different market conditions. I think there's a reason why they never said, "well, you need to throw all that out if interest rates are lower than inflation". I think the reason why they never said that is because they didn't think that would be the right thing to do, based on their experience and insights.
Most of their experience was with a declining interest rate bond bull market and rising equity valuations. One particularly bad sequence in the late 60s through early 80s included similar metrics to where we are today
Do you really think Jack Bogle didn't know about the rising interest rate environment around the period 1945 to 1980? Or that he would just ignore it when formulating his investment recommendations?
That period coincided with reasonably cheap equity valuations.
Not really. The 1960s had higher than average equity valuations in comparison to history up to that point.

Bogle obviously knew about variability of valuations and interest rates. If your argument is that he didn't intend for his portfolio recommendations to be all-weather, then maybe you should suggest that there should be a revision to The Bogleheads Guide to Investing, saying that those recommendations can be thrown out in situations like the present.

Of course, I'd disagree with that.
Nathan Drake wrote: Sat Dec 25, 2021 3:47 pm We do know that the 40% portion is basically going to yield negative real returns.
No, we don't (other than for short-term bonds, perhaps).

But the more important question is what they'll yield relative to equities, not relative to inflation.
The mid 60s was the start of the terrible period I’m discussing where SWR failed or nearly failed.

If we look at long term secular bond markets, an era predominated by rising or even relatively flat interest rates tends to do less than inflation.

The real return is absolutely critical when it comes to SWR risk.

I do not believe Bogle’s beliefs were an all-weather portfolio. It was the one HE was comfortable with, but he was also extremely wealthy and probably did not need to analyze the details of a SWR scenario like most of us need to. He was not a financial advisor. And his major life achievement was the S&P 500 index fund, so through that lens it makes sense why he would be biased/partial to it.

Simplicity worked throughout most of his era because the environment generally featured reasonable asset prices for at least one of the two main components.

If you are retiring today, 2021, on 60/40, and don’t have much wiggle room to reduce expenses while aiming to achieve 4% inflation adjusted withdrawals, that’s a plan that just is simply far too risky to ignore.
20% VOO | 30% VXUS | 20% AVUV | 15% AVDV | 15% AVES
User avatar
Topic Author
willthrill81
Posts: 28078
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Great article from Portfolio Charts on most efficient portfolios

Post by willthrill81 »

GaryA505 wrote: Sat Dec 25, 2021 2:09 pm There may be much better alternatives to 60/40, or not, but we won't know what they are until the future.
Of course, but the fact that we don't know the future does not necessarily mean that all portfolios are equally likely to produce a desired outcome.
“Good and ill have not changed since yesteryear; nor are they one thing among Elves and Dwarves and another among Men.” J.R.R. Tolkien, The Lord of the Rings
GaryA505
Posts: 1190
Joined: Wed Feb 08, 2017 2:59 pm
Location: New Mexico

Re: Great article from Portfolio Charts on most efficient portfolios

Post by GaryA505 »

If bonds have expected negative real yield, wouldn't they be priced accordling?
"Get most of it right and don't make any big mistakes."
nigel_ht
Posts: 3458
Joined: Tue Jan 01, 2019 10:14 am

Re: Great article from Portfolio Charts on most efficient portfolios

Post by nigel_ht »

HanSolo wrote: Sat Dec 25, 2021 3:59 pm
Nathan Drake wrote: Sat Dec 25, 2021 3:42 pm
HanSolo wrote: Sat Dec 25, 2021 2:13 pm
Nathan Drake wrote: Sat Dec 25, 2021 12:00 pm
HanSolo wrote: Sat Dec 25, 2021 3:38 am As you can see from my sig, I think there's nothing wrong with 60/40. In addition, I have certain tilts in my IRA, some of which are discussed in this thread. I don't have a strictly Boglehead-compliant portfolio (like 3-fund), but I think that, on the whole, the Boglehead ideas are pretty good and aren't out of date. The ideas of Bogle and related authors were formulated after having experience with a lot of different market conditions. I think there's a reason why they never said, "well, you need to throw all that out if interest rates are lower than inflation". I think the reason why they never said that is because they didn't think that would be the right thing to do, based on their experience and insights.
Most of their experience was with a declining interest rate bond bull market and rising equity valuations. One particularly bad sequence in the late 60s through early 80s included similar metrics to where we are today
Do you really think Jack Bogle didn't know about the rising interest rate environment around the period 1945 to 1980? Or that he would just ignore it when formulating his investment recommendations?
That period coincided with reasonably cheap equity valuations.
Not really. The 1960s had higher than average equity valuations in comparison to history up to that point.

Bogle obviously knew about variability of valuations and interest rates. If your argument is that he didn't intend for his portfolio recommendations to be all-weather, then maybe you should suggest that there should be a revision to The Bogleheads Guide to Investing, saying that those recommendations can be thrown out in situations like the present.

Of course, I'd disagree with that.
Nathan Drake wrote: Sat Dec 25, 2021 3:47 pm We do know that the 40% portion is basically going to yield negative real returns.
No, we don't.

But the more important question is what they'll yield relative to equities, not relative to inflation.
Well if they are both negative real then 2 fund BH retirees will likely not be happy campers…

Even 3 fund given the high correlation between US and international large cap.

For folks still in early to mid accumulation it’s like “yay! A sale on stocks!” For late accumulators it will likely be a bit concerning if both are doing poorly.
User avatar
Topic Author
willthrill81
Posts: 28078
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Great article from Portfolio Charts on most efficient portfolios

Post by willthrill81 »

HanSolo wrote: Sat Dec 25, 2021 3:59 pm
Nathan Drake wrote: Sat Dec 25, 2021 3:47 pm We do know that the 40% portion is basically going to yield negative real returns.
No, we don't.
We don't know that for certain, but the likelihood of nominal bonds bought today producing negative real returns is very high. 10 year TIPS are currently yielding -1%. Buyers of those individual TIPS will experience negative real returns. There are more unknowns with nominal bonds, but the only ways that a zero or positive real return could materialize with them are (1) if inflation averages 1% below what the market has priced in and/or (2) if interest rates fall significantly below their near record lows.
“Good and ill have not changed since yesteryear; nor are they one thing among Elves and Dwarves and another among Men.” J.R.R. Tolkien, The Lord of the Rings
User avatar
Topic Author
willthrill81
Posts: 28078
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Great article from Portfolio Charts on most efficient portfolios

Post by willthrill81 »

GaryA505 wrote: Sat Dec 25, 2021 4:44 pm If bonds have expected negative real yield, wouldn't they be priced accordling?
Indeed they are. 10 year TIPS currently have a yield of -1.01%. This is why so many are rightly loading up on I bonds instead, which have a 0% real yield (but are currently offering a 7.12% nominal yield).
“Good and ill have not changed since yesteryear; nor are they one thing among Elves and Dwarves and another among Men.” J.R.R. Tolkien, The Lord of the Rings
nigel_ht
Posts: 3458
Joined: Tue Jan 01, 2019 10:14 am

Re: Great article from Portfolio Charts on most efficient portfolios

Post by nigel_ht »

Nathan Drake wrote: Sat Dec 25, 2021 4:09 pm
HanSolo wrote: Sat Dec 25, 2021 3:59 pm
Nathan Drake wrote: Sat Dec 25, 2021 3:42 pm
HanSolo wrote: Sat Dec 25, 2021 2:13 pm
Nathan Drake wrote: Sat Dec 25, 2021 12:00 pm

Most of their experience was with a declining interest rate bond bull market and rising equity valuations. One particularly bad sequence in the late 60s through early 80s included similar metrics to where we are today
Do you really think Jack Bogle didn't know about the rising interest rate environment around the period 1945 to 1980? Or that he would just ignore it when formulating his investment recommendations?
That period coincided with reasonably cheap equity valuations.
Not really. The 1960s had higher than average equity valuations in comparison to history up to that point.

Bogle obviously knew about variability of valuations and interest rates. If your argument is that he didn't intend for his portfolio recommendations to be all-weather, then maybe you should suggest that there should be a revision to The Bogleheads Guide to Investing, saying that those recommendations can be thrown out in situations like the present.

Of course, I'd disagree with that.
Nathan Drake wrote: Sat Dec 25, 2021 3:47 pm We do know that the 40% portion is basically going to yield negative real returns.
No, we don't (other than for short-term bonds, perhaps).

But the more important question is what they'll yield relative to equities, not relative to inflation.
The mid 60s was the start of the terrible period I’m discussing where SWR failed or nearly failed.

If we look at long term secular bond markets, an era predominated by rising or even relatively flat interest rates tends to do less than inflation.

The real return is absolutely critical when it comes to SWR risk.

I do not believe Bogle’s beliefs were an all-weather portfolio. It was the one HE was comfortable with, but he was also extremely wealthy and probably did not need to analyze the details of a SWR scenario like most of us need to. He was not a financial advisor. And his major life achievement was the S&P 500 index fund, so through that lens it makes sense why he would be biased/partial to it.

Simplicity worked throughout most of his era because the environment generally featured reasonable asset prices for at least one of the two main components.

If you are retiring today, 2021, on 60/40, and don’t have much wiggle room to reduce expenses while aiming to achieve 4% inflation adjusted withdrawals, that’s a plan that just is simply far too risky to ignore.
Counting $24K social security the SWR (100% success rate) is 4.4% on $1M which means 4% has a bit of padding vs the 1966 scenario.

That’s not perfect but it’s not “too risky to ignore”.
User avatar
Topic Author
willthrill81
Posts: 28078
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Great article from Portfolio Charts on most efficient portfolios

Post by willthrill81 »

nigel_ht wrote: Sat Dec 25, 2021 4:46 pm
HanSolo wrote: Sat Dec 25, 2021 3:59 pm
Nathan Drake wrote: Sat Dec 25, 2021 3:42 pm
HanSolo wrote: Sat Dec 25, 2021 2:13 pm
Nathan Drake wrote: Sat Dec 25, 2021 12:00 pm

Most of their experience was with a declining interest rate bond bull market and rising equity valuations. One particularly bad sequence in the late 60s through early 80s included similar metrics to where we are today
Do you really think Jack Bogle didn't know about the rising interest rate environment around the period 1945 to 1980? Or that he would just ignore it when formulating his investment recommendations?
That period coincided with reasonably cheap equity valuations.
Not really. The 1960s had higher than average equity valuations in comparison to history up to that point.

Bogle obviously knew about variability of valuations and interest rates. If your argument is that he didn't intend for his portfolio recommendations to be all-weather, then maybe you should suggest that there should be a revision to The Bogleheads Guide to Investing, saying that those recommendations can be thrown out in situations like the present.

Of course, I'd disagree with that.
Nathan Drake wrote: Sat Dec 25, 2021 3:47 pm We do know that the 40% portion is basically going to yield negative real returns.
No, we don't.

But the more important question is what they'll yield relative to equities, not relative to inflation.
Well if they are both negative real then 2 fund BH retirees will likely not be happy campers…

Even 3 fund given the high correlation between US and international large cap.

For folks still in early to mid accumulation it’s like “yay! A sale on stocks!” For late accumulators it will likely be a bit concerning if both are doing poorly.
Precisely. From 1981-2012, 10 year Treasuries returned 6% real. But returns have already fallen far off that mark. Since 2013, 10 year Treasuries have returned essentially 0% real. This year, they are likely to return around -8% real. Again, that doesn't mean that bonds are 'bad', but investors should not expect bonds to save the day if TSM does poorly.
“Good and ill have not changed since yesteryear; nor are they one thing among Elves and Dwarves and another among Men.” J.R.R. Tolkien, The Lord of the Rings
User avatar
vanbogle59
Posts: 1114
Joined: Wed Mar 10, 2021 8:30 pm

Re: Great article from Portfolio Charts on most efficient portfolios

Post by vanbogle59 »

willthrill81 wrote: Fri Dec 24, 2021 4:25 pm
vanbogle59 wrote: Fri Dec 24, 2021 4:15 pm
Tyler9000 wrote: Fri Dec 24, 2021 10:49 am
Triple digit golfer wrote: Fri Dec 24, 2021 9:58 am This is why I'm so leery of backtests. The conclusions drawn from one "long term" period rarely are the same as from another.
I agree. Which is why the methodology looks at every rolling 15-year timeframe and the metrics prioritize the most consistent options. :wink:

It's off-topic for this discussion, but you may find this article useful for thinking about backtesting: Open Your Eyes to the Power of Helpful History
Let me start by saying thank you for your website. I think it's wonderful.
However... :wink:

"By exploring how data is used well, perhaps we can learn a thing or two about how to also apply it effectively it in our personal portfolio decisions."
With all due respect, and I have a great deal of respect for what you have done here, that is precisely where I am having the gravest misgivings.
I have no idea how to "apply it effectively".

I reject categorically that anything resembling a "distribution" of future outcomes can be deduced from looking at historical data, especially when it comes to looking at things like decades-long portfolio returns.
I am firmly in the Kahneman camp: "The idea that the future is unpredictable is undermined every day by the ease with which the past is explained."
On point, compare your description of the job interview to Kahneman's excoriation of the Israeli military in predicting the future success of their own officers, about whom they know MUCH more than a simple resume.

OTOH, I do think there are broad lessons to be learned in the data. I am particularly intrigued by the "price" of gold as insurance, and the lack of "diversification" from exUS for US TSM.

Anyway, I haven't changed my ISP yet. But I might! :D
Thanks again. :sharebeer
I'm very confused by this. You say that you "think there are broad lessons to be learned in the data," but you also seem to reject using the past as a means of finding 'good' portfolios for the future.
Same reason I read history, but I don't try to calculate the start date of WWIII by running calculations based on WWI and WWII.

Yes, I'm convinced equities are a good long-term investment.
I'm further convinced that adding bonds to an equity portfolio can lower volatility.

I'm intrigued, but not convinced, that adding gold can add an element of "insurance" for some less-common economic events. That might even happen often enough to make it "worth it" in a financial sense. But even if that turns out to be true in general, it may still be irrelevant for constructing a personal portfolio where N=1.... I just don't know. Same goes for other asset classes with low or even negative correlations.

I also think there are market inefficiencies that really smart people can take advantage of. (Though I'm quite sure I'm not one of them.) That would be especially true of firms with information advantages. No way I can compete.

Finally, I know we don't know what we don't know about the future. And that could be the one ring that rules them all. Candidates include some unforeseen resource constraint or new technology, social change, the creation of new economic systems... IDK.

So, I am left with the haystack average, that I can get for free. And maybe one personal advantage (I only care about relatively long term results).
Beyond that??? :confused

Put me square in the camp that economic analysis is worthwhile (wonderful, even). I find it especially useful for "disproving" ideas.
But I still think it's a sure sign economists have a sense of humor when they start using decimal points.
:beer
nigel_ht
Posts: 3458
Joined: Tue Jan 01, 2019 10:14 am

Re: Great article from Portfolio Charts on most efficient portfolios

Post by nigel_ht »

vanbogle59 wrote: Sat Dec 25, 2021 5:16 pm
Same reason I read history, but I don't try to calculate the start date of WWIII by running calculations based on WWI and WWII.
No. But you can get a general rhythm for the rise and fall of empires. We aren't attempting to discern when WWIII will happen, just the general impact of it (assuming a non-nuclear WWIII) on our imperial portfolio...
nigel_ht
Posts: 3458
Joined: Tue Jan 01, 2019 10:14 am

Re: Great article from Portfolio Charts on most efficient portfolios

Post by nigel_ht »

willthrill81 wrote: Sat Dec 25, 2021 4:59 pm
nigel_ht wrote: Sat Dec 25, 2021 4:46 pm
HanSolo wrote: Sat Dec 25, 2021 3:59 pm
Nathan Drake wrote: Sat Dec 25, 2021 3:42 pm
HanSolo wrote: Sat Dec 25, 2021 2:13 pm

Do you really think Jack Bogle didn't know about the rising interest rate environment around the period 1945 to 1980? Or that he would just ignore it when formulating his investment recommendations?
That period coincided with reasonably cheap equity valuations.
Not really. The 1960s had higher than average equity valuations in comparison to history up to that point.

Bogle obviously knew about variability of valuations and interest rates. If your argument is that he didn't intend for his portfolio recommendations to be all-weather, then maybe you should suggest that there should be a revision to The Bogleheads Guide to Investing, saying that those recommendations can be thrown out in situations like the present.

Of course, I'd disagree with that.
Nathan Drake wrote: Sat Dec 25, 2021 3:47 pm We do know that the 40% portion is basically going to yield negative real returns.
No, we don't.

But the more important question is what they'll yield relative to equities, not relative to inflation.
Well if they are both negative real then 2 fund BH retirees will likely not be happy campers…

Even 3 fund given the high correlation between US and international large cap.

For folks still in early to mid accumulation it’s like “yay! A sale on stocks!” For late accumulators it will likely be a bit concerning if both are doing poorly.
Precisely. From 1981-2012, 10 year Treasuries returned 6% real. But returns have already fallen far off that mark. Since 2013, 10 year Treasuries have returned essentially 0% real. This year, they are likely to return around -8% real. Again, that doesn't mean that bonds are 'bad', but investors should not expect bonds to save the day if TSM does poorly.
As I commented, I think that just increases the cost of insurance. Not that treasuries will not still be protective when we get the next major TSM correction.
User avatar
Topic Author
willthrill81
Posts: 28078
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Great article from Portfolio Charts on most efficient portfolios

Post by willthrill81 »

nigel_ht wrote: Sat Dec 25, 2021 6:40 pm
willthrill81 wrote: Sat Dec 25, 2021 4:59 pmFrom 1981-2012, 10 year Treasuries returned 6% real. But returns have already fallen far off that mark. Since 2013, 10 year Treasuries have returned essentially 0% real. This year, they are likely to return around -8% real. Again, that doesn't mean that bonds are 'bad', but investors should not expect bonds to save the day if TSM does poorly.
As I commented, I think that just increases the cost of insurance. Not that treasuries will not still be protective when we get the next major TSM correction.
All else being equal, lower returns on bonds will result in lower SWRs.
“Good and ill have not changed since yesteryear; nor are they one thing among Elves and Dwarves and another among Men.” J.R.R. Tolkien, The Lord of the Rings
User avatar
vineviz
Posts: 10403
Joined: Tue May 15, 2018 1:55 pm

Re: Great article from Portfolio Charts on most efficient portfolios

Post by vineviz »

willthrill81 wrote: Sat Dec 25, 2021 10:41 am
nigel_ht wrote: Sat Dec 25, 2021 9:34 am It is possible to argue that once the SCV premium was discovered that it no longer exists but there’s not enough data yet to prove that.
An argument that the discovery of the SCV premium has/will eliminate it assumes the EMH to be quite false, not a fringe case.
And an argument that small cap value investing didn’t exist before 1990 is contra factual. Small cap value investors were alive and well in the 1960s and 1970s.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch
User avatar
vanbogle59
Posts: 1114
Joined: Wed Mar 10, 2021 8:30 pm

Re: Great article from Portfolio Charts on most efficient portfolios

Post by vanbogle59 »

nigel_ht wrote: Sat Dec 25, 2021 6:38 pm
vanbogle59 wrote: Sat Dec 25, 2021 5:16 pm
Same reason I read history, but I don't try to calculate the start date of WWIII by running calculations based on WWI and WWII.
No. But you can get a general rhythm for the rise and fall of empires. We aren't attempting to discern when WWIII will happen, just the general impact of it (assuming a non-nuclear WWIII) on our imperial portfolio...
LOL!
I feel like I'm eavesdropping on Anakin and Padme.
:beer
nigel_ht
Posts: 3458
Joined: Tue Jan 01, 2019 10:14 am

Re: Great article from Portfolio Charts on most efficient portfolios

Post by nigel_ht »

vanbogle59 wrote: Sat Dec 25, 2021 8:43 pm
nigel_ht wrote: Sat Dec 25, 2021 6:38 pm
vanbogle59 wrote: Sat Dec 25, 2021 5:16 pm
Same reason I read history, but I don't try to calculate the start date of WWIII by running calculations based on WWI and WWII.
No. But you can get a general rhythm for the rise and fall of empires. We aren't attempting to discern when WWIII will happen, just the general impact of it (assuming a non-nuclear WWIII) on our imperial portfolio...
LOL!
I feel like I'm eavesdropping on Anakin and Padme.
:beer
History doesn’t repeat but it does like to rhyme. The 20th century saw the fall of many empires arsonist wwi and wwii.

The most likely relevant to us is the way the British Empire fell with the loss of reserve currency status and the negation of its military might through economic blackmail - leave the Suez or we will crater the pound sterling.

We owe a lot of debt. Some to China…would the ability to crater the dollar and leverage their portion of the debt allow them to negate the USN around Taiwan without a shot being fired at us?

Not today. But if it comes to pass then Pax Americana will have ended and time to move completely away from VTI to VT and LTT likely no longer a good hedge against a market drop.
User avatar
steve r
Posts: 872
Joined: Mon Feb 13, 2012 8:34 pm

Re: Great article from Portfolio Charts on most efficient portfolios

Post by steve r »

vanbogle59 wrote: Sat Dec 25, 2021 5:16 pmBut I still think it's a sure sign economists have a sense of humor when they start using decimal points.
:beer
Classic!
“The closer you come to holding the entire market portfolio (all traded securities) the higher your expected return for the risk you take.” William F. Sharpe | VT, BND, TIAA RE and chill.
User avatar
steve r
Posts: 872
Joined: Mon Feb 13, 2012 8:34 pm

Re: Great article from Portfolio Charts on most efficient portfolios

Post by steve r »

Imagine it is 1970. You decide on a 60 /40 portfolio U.S. TSM and some bonds, but are not sure what type of bonds. Gold is still mostly on the gold standard.

You do some calculations on on a variety of bond holdings with your abacus (or $1000 adding machine) using data similar to the Simba data today on BH since 1871. Despite some serious reservations about the data, you think 100 years -- that ought to tell me something!

The results CAGR (s.d.):

a) 6.78 (11.39)
b) 6.78 (11.66)
c) 6.77 (11.39)
d) 6.93 (11.21)
e) 6.74 (11.43)

We have a winner -- d).

What are the portfolios? Again, 60 /40 U.S. TSM and either TBM, LTT, ITT, STT, and intermediate term govt / corp., in that order.

Yes, the winner would be Short Term Treasuries. :?: In fact, they would remain the back tested winner up through 2001! (and 1871 to 2001 is 131 years)

My point is picking the winning portfolio in 1970 (article data start date) would have been very difficult.

I do believe LTT should have a higher return that shorter term products in theory. That theory strikes me as sound. As does the idea they should provide a touch more diversification benefits. In fact if we extend the analysis though the full period (1871 to 2020), LTT based 60/40 portfolios do come out ahead in terms of CAGR. But not max drawdown, Sharpe, Sortino or Ulcer.

Today, the bond returns (only) mirror what I would have expected. LTT > ITT > STT and IT corp and govt > ITT for the full 151 year period. But this was not the pattern you would observe from 1871 to 2001. Some theories take a long time to pan out! (of course, there are other select shorter periods in history when the expectations did pan out and we are living in one today).
“The closer you come to holding the entire market portfolio (all traded securities) the higher your expected return for the risk you take.” William F. Sharpe | VT, BND, TIAA RE and chill.
Bam123Boo
Posts: 5
Joined: Wed Aug 18, 2021 10:15 am

Re: Great article from Portfolio Charts on most efficient portfolios

Post by Bam123Boo »

@Tyler9000

I'm curious as to why the article didn't mention the pinwheel portfolio for those who did not want as much allocation to gold. Did you switch from the pinwheel to golden butterfly now?
User avatar
Topic Author
willthrill81
Posts: 28078
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Great article from Portfolio Charts on most efficient portfolios

Post by willthrill81 »

Bam123Boo wrote: Tue Jan 04, 2022 2:22 pm @Tyler9000

I'm curious as to why the article didn't mention the pinwheel portfolio for those who did not want as much allocation to gold. Did you switch from the pinwheel to golden butterfly now?
The Pinwheel Portfolio was included in Tyler's analysis. He just didn't discuss it for whatever reason. It performed very well on the metrics he examined.
“Good and ill have not changed since yesteryear; nor are they one thing among Elves and Dwarves and another among Men.” J.R.R. Tolkien, The Lord of the Rings
User avatar
Tyler9000
Posts: 657
Joined: Fri Aug 21, 2015 11:57 am

Re: Great article from Portfolio Charts on most efficient portfolios

Post by Tyler9000 »

Bam123Boo wrote: Tue Jan 04, 2022 2:22 pm @Tyler9000

I'm curious as to why the article didn't mention the pinwheel portfolio for those who did not want as much allocation to gold. Did you switch from the pinwheel to golden butterfly now?
In the context of the article, I talked about the Golden Butterfly for its position in the cloud and relevance with the Frontier-3 assets. I also name-checked a few other portfolio options mainly to offer ideas for people who prefer to avoid those assets altogether. But I totally agree with your observation that the Pinwheel Portfolio is an excellent option for those who are open to smaller percentages of gold and SCV. You can see on the charts in the article that it performed quite well.
Bam123Boo
Posts: 5
Joined: Wed Aug 18, 2021 10:15 am

Re: Great article from Portfolio Charts on most efficient portfolios

Post by Bam123Boo »

Tyler9000 wrote: Tue Jan 04, 2022 3:32 pm
Bam123Boo wrote: Tue Jan 04, 2022 2:22 pm @Tyler9000

I'm curious as to why the article didn't mention the pinwheel portfolio for those who did not want as much allocation to gold. Did you switch from the pinwheel to golden butterfly now?
In the context of the article, I talked about the Golden Butterfly for its position in the cloud and relevance with the Frontier-3 assets. I also name-checked a few other portfolio options mainly to offer ideas for people who prefer to avoid those assets altogether. But I totally agree with your observation that the Pinwheel Portfolio is an excellent option for those who are open to smaller percentages of gold and SCV. You can see on the charts in the article that it performed quite well.
Thank you for responding. I saw the article and this thread and it took me quite some time to digest both. I did see the pinwheel in the charts and such, but there was no mention of the word pinwheel at all in the 3 Secret Ingredients article. In fact if you search for the word pinwheel, it would not show up in that article. I thought it would be neat to add some comments about the pinwheel since you articulated many other portfolios.

As I digest this thread and the article mentioned, it looks like the golden butterfly outperforms the pinwheel in terms of SWR/PWR/ULCER/Returns. I do understand past performance is not an expection of future performance. Have you switched to strictly golden butterfly instead of pinwheel?
User avatar
Tyler9000
Posts: 657
Joined: Fri Aug 21, 2015 11:57 am

Re: Great article from Portfolio Charts on most efficient portfolios

Post by Tyler9000 »

Bam123Boo wrote: Tue Jan 04, 2022 8:48 pm Have you switched to strictly golden butterfly instead of pinwheel?
My philosophy is that there are lots of excellent portfolios suitable for different types of people. And my goal is simply to use good data to illustrate how asset allocations operate in the real world to help investors make an educated choice. So I'm not strictly for any single portfolio.
Last edited by Tyler9000 on Tue Jan 04, 2022 11:47 pm, edited 1 time in total.
User avatar
Topic Author
willthrill81
Posts: 28078
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Great article from Portfolio Charts on most efficient portfolios

Post by willthrill81 »

Tyler9000 wrote: Tue Jan 04, 2022 10:15 pm
Bam123Boo wrote: Tue Jan 04, 2022 8:48 pm Have you switched to strictly golden butterfly instead of pinwheel?
My philosophy is that there are lots of excellent portfolios suitable for different types of people. And my goal is simply use good data to illustrate how asset allocations operate in the real world to help investors make an educated choice. So I'm not strictly for any single portfolio.
I really appreciate your stance on this issue Tyler. It's very refreshing to see someone who is open to many different possible paths and doesn't try to use a one-size-fits-all approach. There are clearly many roads to Dublin, and the 3-fund portfolio that many dote on here is certainly not the only good one.

And as always, thanks for your hard work on PC. It's one of the best sites out there. :sharebeer
“Good and ill have not changed since yesteryear; nor are they one thing among Elves and Dwarves and another among Men.” J.R.R. Tolkien, The Lord of the Rings
Bam123Boo
Posts: 5
Joined: Wed Aug 18, 2021 10:15 am

Re: Great article from Portfolio Charts on most efficient portfolios

Post by Bam123Boo »

As I take time to digest and ponder, my next question is in regards to baseline returns. Why exactly was the 15th percentile selected? Why not the 20th or 25th and how much does it change these numbers in regards to returns? I did read the average, beverages article but it did not say exactly why the 15th percentile was the magic number. Thanks for the discussion and thoughts as we put our minds together.
User avatar
Tyler9000
Posts: 657
Joined: Fri Aug 21, 2015 11:57 am

Re: Great article from Portfolio Charts on most efficient portfolios

Post by Tyler9000 »

Bam123Boo wrote: Wed Jan 05, 2022 12:03 pm my next question is in regards to baseline returns. Why exactly was the 15th percentile selected? Why not the 20th or 25th and how much does it change these numbers in regards to returns?
The baseline return is my effort to illustrate conservative real-world investing outcomes while excluding the handful of extremes that might be considered outliers. Note that I also track a mirror number called the stretch return. I chose the 15% cutoff for both the baseline and stretch returns so that a little over 2/3rds of the outcomes would fall between those metrics. I also didn't want to set the baseline so high that it excludes entire decades like the 1970s, as I believe understanding hard times with rising rates and high inflation is important. But play with the Rolling Returns chart and you can see the full range of outcomes for any portfolio and make your own judgment call.

Note that in the context of the cloud charts in the "Three Secret Ingredients" article, the baseline return only affects the Y-axis. Raising the cutoff to exclude more below-average outcomes would indeed make stock investors happy but would also make gold skeptics upset. ;) And it would also never move any portfolio to the left.

In any case, I plan to write future articles that use the same methodology to explore different metrics. And I'm always open to suggestions.
Last edited by Tyler9000 on Thu Jan 06, 2022 2:46 pm, edited 1 time in total.
MarkVH0518
Posts: 139
Joined: Tue Dec 13, 2016 2:14 pm

Re: Great article from Portfolio Charts on most efficient portfolios

Post by MarkVH0518 »

Tyler9000 wrote: Thu Jan 06, 2022 1:27 pm In any case, I plan to write future articles that use the same methodology to explore different metrics. And I'm always open to suggestions.
Tyler,
On the off chance you are still paying attention: are these axis independent?
Are you certain there is no correlation between the ulcer index and the 15-year baseline return?
I'm wondering if the reason some are so surprised by the results is because there is just a bit double counting going on.
I'm in no position to test my idea, but I thought you might.

Thanks for performing the study
Mark
User avatar
Tyler9000
Posts: 657
Joined: Fri Aug 21, 2015 11:57 am

Re: Great article from Portfolio Charts on most efficient portfolios

Post by Tyler9000 »

MarkVH0518 wrote: Thu Jan 06, 2022 2:44 pm Are you certain there is no correlation between the ulcer index and the 15-year baseline return?
The baseline return does not directly affect the ulcer index and vice versa. But since the former favors consistently high returns and the latter rewards consistently low drawdowns, it should be no surprise that highly volatile portfolios rank lower on both measures compared to more reliable options.

The fact that volatility can impact both risk and return on your own investing timeframe is a big reason why it's important to look beyond simple averages when choosing your own portfolio. The "risk" in the "risk premium" that people talk about when discussing volatile investments can manifest in more ways than you may realize.
User avatar
spdoublebass
Posts: 826
Joined: Thu Apr 27, 2017 10:04 pm
Location: NY

Re: Great article from Portfolio Charts on most efficient portfolios

Post by spdoublebass »

steve r wrote: Sun Dec 26, 2021 9:21 am Imagine it is 1970. You decide on a 60 /40 portfolio U.S. TSM and some bonds, but are not sure what type of bonds. Gold is still mostly on the gold standard.

You do some calculations on on a variety of bond holdings with your abacus (or $1000 adding machine) using data similar to the Simba data today on BH since 1871. Despite some serious reservations about the data, you think 100 years -- that ought to tell me something!

The results CAGR (s.d.):

a) 6.78 (11.39)
b) 6.78 (11.66)
c) 6.77 (11.39)
d) 6.93 (11.21)
e) 6.74 (11.43)

We have a winner -- d).

What are the portfolios? Again, 60 /40 U.S. TSM and either TBM, LTT, ITT, STT, and intermediate term govt / corp., in that order.

Yes, the winner would be Short Term Treasuries. :?: In fact, they would remain the back tested winner up through 2001! (and 1871 to 2001 is 131 years)

My point is picking the winning portfolio in 1970 (article data start date) would have been very difficult.

I do believe LTT should have a higher return that shorter term products in theory. That theory strikes me as sound. As does the idea they should provide a touch more diversification benefits. In fact if we extend the analysis though the full period (1871 to 2020), LTT based 60/40 portfolios do come out ahead in terms of CAGR. But not max drawdown, Sharpe, Sortino or Ulcer.

Today, the bond returns (only) mirror what I would have expected. LTT > ITT > STT and IT corp and govt > ITT for the full 151 year period. But this was not the pattern you would observe from 1871 to 2001. Some theories take a long time to pan out! (of course, there are other select shorter periods in history when the expectations did pan out and we are living in one today).
I’ve read this post like 10 times and have no idea what is being said.

Are you saying that in 1970 a 60/40 portfolio of TSM and ST treasury bonds would do better than with LT treasury bonds?

I’m not trying to be confrontational, I just can’t figure out what your trying to say.
I'm trying to think, but nothing happens
User avatar
Topic Author
willthrill81
Posts: 28078
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Great article from Portfolio Charts on most efficient portfolios

Post by willthrill81 »

spdoublebass wrote: Thu Jan 06, 2022 9:52 pm
steve r wrote: Sun Dec 26, 2021 9:21 am Imagine it is 1970. You decide on a 60 /40 portfolio U.S. TSM and some bonds, but are not sure what type of bonds. Gold is still mostly on the gold standard.

You do some calculations on on a variety of bond holdings with your abacus (or $1000 adding machine) using data similar to the Simba data today on BH since 1871. Despite some serious reservations about the data, you think 100 years -- that ought to tell me something!

The results CAGR (s.d.):

a) 6.78 (11.39)
b) 6.78 (11.66)
c) 6.77 (11.39)
d) 6.93 (11.21)
e) 6.74 (11.43)

We have a winner -- d).

What are the portfolios? Again, 60 /40 U.S. TSM and either TBM, LTT, ITT, STT, and intermediate term govt / corp., in that order.

Yes, the winner would be Short Term Treasuries. :?: In fact, they would remain the back tested winner up through 2001! (and 1871 to 2001 is 131 years)

My point is picking the winning portfolio in 1970 (article data start date) would have been very difficult.

I do believe LTT should have a higher return that shorter term products in theory. That theory strikes me as sound. As does the idea they should provide a touch more diversification benefits. In fact if we extend the analysis though the full period (1871 to 2020), LTT based 60/40 portfolios do come out ahead in terms of CAGR. But not max drawdown, Sharpe, Sortino or Ulcer.

Today, the bond returns (only) mirror what I would have expected. LTT > ITT > STT and IT corp and govt > ITT for the full 151 year period. But this was not the pattern you would observe from 1871 to 2001. Some theories take a long time to pan out! (of course, there are other select shorter periods in history when the expectations did pan out and we are living in one today).
I’ve read this post like 10 times and have no idea what is being said.

Are you saying that in 1970 a 60/40 portfolio of TSM and ST treasury bonds would do better than with LT treasury bonds?

I’m not trying to be confrontational, I just can’t figure out what your trying to say.
He's saying that had you analyzed the prior 100 years of history of these portfolios in 1970, you would have concluded that short-term Treasuries had been marginally superior to the other maturities and/or types of bonds in terms of both returns and volatility (as measured by standard deviation).
“Good and ill have not changed since yesteryear; nor are they one thing among Elves and Dwarves and another among Men.” J.R.R. Tolkien, The Lord of the Rings
User avatar
Topic Author
willthrill81
Posts: 28078
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Great article from Portfolio Charts on most efficient portfolios

Post by willthrill81 »

Tyler9000 wrote: Thu Jan 06, 2022 4:34 pm
MarkVH0518 wrote: Thu Jan 06, 2022 2:44 pm Are you certain there is no correlation between the ulcer index and the 15-year baseline return?
The baseline return does not directly affect the ulcer index and vice versa. But since the former favors consistently high returns and the latter rewards consistently low drawdowns, it should be no surprise that highly volatile portfolios rank lower on both measures compared to more reliable options.

The fact that volatility can impact both risk and return on your own investing timeframe is a big reason why it's important to look beyond simple averages when choosing your own portfolio. The "risk" in the "risk premium" that people talk about when discussing volatile investments can manifest in more ways than you may realize.
Your analysis also demonstrated that combining assets with comparatively high standalone volatility, which SCV, LTT, and gold all have had, that have not been positively correlated with each other when it really counts would have resulted in the kind of portfolio performance that many investors dream about but believe to be (or at least have been) impossible: great returns with shockingly small and brief drawdowns. In truth, it's a great illustration of modern portfolio theory in action. Sadly, many have been looking for years for assets with 'stock-like returns' that are uncorrelated with stocks, and the search for these assets has not been very fruitful. It seems that multiple good answers may have been under their noses the whole time.
“Good and ill have not changed since yesteryear; nor are they one thing among Elves and Dwarves and another among Men.” J.R.R. Tolkien, The Lord of the Rings
slicendice
Posts: 152
Joined: Tue Sep 22, 2020 12:08 am

Re: Great article from Portfolio Charts on most efficient portfolios

Post by slicendice »

steve r wrote: Sun Dec 26, 2021 9:21 am Imagine it is 1970. You decide on a 60 /40 portfolio U.S. TSM and some bonds, but are not sure what type of bonds. Gold is still mostly on the gold standard.

You do some calculations on on a variety of bond holdings with your abacus (or $1000 adding machine) using data similar to the Simba data today on BH since 1871. Despite some serious reservations about the data, you think 100 years -- that ought to tell me something!

The results CAGR (s.d.):

a) 6.78 (11.39)
b) 6.78 (11.66)
c) 6.77 (11.39)
d) 6.93 (11.21)
e) 6.74 (11.43)

We have a winner -- d).

What are the portfolios? Again, 60 /40 U.S. TSM and either TBM, LTT, ITT, STT, and intermediate term govt / corp., in that order.

Yes, the winner would be Short Term Treasuries. :?: In fact, they would remain the back tested winner up through 2001! (and 1871 to 2001 is 131 years)

My point is picking the winning portfolio in 1970 (article data start date) would have been very difficult.

I do believe LTT should have a higher return that shorter term products in theory. That theory strikes me as sound. As does the idea they should provide a touch more diversification benefits. In fact if we extend the analysis though the full period (1871 to 2020), LTT based 60/40 portfolios do come out ahead in terms of CAGR. But not max drawdown, Sharpe, Sortino or Ulcer.

Today, the bond returns (only) mirror what I would have expected. LTT > ITT > STT and IT corp and govt > ITT for the full 151 year period. But this was not the pattern you would observe from 1871 to 2001. Some theories take a long time to pan out! (of course, there are other select shorter periods in history when the expectations did pan out and we are living in one today).
In terms of pre-1970 treasury data-- how useful is it for backtesting? Maybe I'm mistaken, but for a long time (1940's and 1950's at least), weren't treasuries callable? At the long end of the duration spectrum in particular, that would negate the significant positive convexity which I think is a fairly important contributor to the diversification benefit to equities that Swenson alluded to when he talked about the diversification benefit of long-term, non-callable nominal treasuries to an equity portfolio.
User avatar
HomerJ
Posts: 18250
Joined: Fri Jun 06, 2008 12:50 pm

Re: Great article from Portfolio Charts on most efficient portfolios

Post by HomerJ »

willthrill81 wrote: Thu Dec 16, 2021 10:46 pm After examining 42,504 different combinations of 20 different assets using data going back to 1970, he found that the most efficient portfolios usually included small-cap value stocks, long-term Treasuries, and gold, though some portfolios that didn't include all or even any of these assets also performed well.
Missed this thread, and will read through it, but my first thought is that 1970 isn't far back enough.

Need a 30-year period where bonds did poorly at least. Long-term Treasuries were AMAZING from 1983 on. That's not likely to be true going forward.
"The best tools available to us are shovels, not scalpels. Don't get carried away." - vanBogle59
zie
Posts: 364
Joined: Sun Mar 22, 2020 4:35 pm

Re: Great article from Portfolio Charts on most efficient portfolios

Post by zie »

HomerJ wrote: Fri Jan 07, 2022 3:08 pm
willthrill81 wrote: Thu Dec 16, 2021 10:46 pm After examining 42,504 different combinations of 20 different assets using data going back to 1970, he found that the most efficient portfolios usually included small-cap value stocks, long-term Treasuries, and gold, though some portfolios that didn't include all or even any of these assets also performed well.
Missed this thread, and will read through it, but my first thought is that 1970 isn't far back enough.

Need a 30-year period where bonds did poorly at least. Long-term Treasuries were AMAZING from 1983 on. That's not likely to be true going forward.
Unfortunately we barely have 1 investor lifetime of good data so far(2021-1970 = 51 years) which is on the short side of one investment lifetime, especially as lifetimes keep increasing. I know people have 100+yrs of data, but I believe anything before 1970 was at best yearly and unlikely to be very accurate and almost wholly limited to USA. Even the 1970's onwards data is mostly US specific. In another 100 years, we will have 2 lifetimes worth of data, and then we can hopefully answer a lot of these questions more definitively... maybe!
Whether rich or poor, a young woman should know how a bank account works, understand the composition of mortgages and bonds, and know the value of interest and how it accumulates. -Hetty Green
User avatar
Topic Author
willthrill81
Posts: 28078
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Great article from Portfolio Charts on most efficient portfolios

Post by willthrill81 »

HomerJ wrote: Fri Jan 07, 2022 3:08 pm
willthrill81 wrote: Thu Dec 16, 2021 10:46 pm After examining 42,504 different combinations of 20 different assets using data going back to 1970, he found that the most efficient portfolios usually included small-cap value stocks, long-term Treasuries, and gold, though some portfolios that didn't include all or even any of these assets also performed well.
Missed this thread, and will read through it, but my first thought is that 1970 isn't far back enough.

Need a 30-year period where bonds did poorly at least. Long-term Treasuries were AMAZING from 1983 on. That's not likely to be true going forward.
I agree that we need more data, but as zie pointed out, it just doesn't exist. Heck, the S&P 500 was only formed in 1957, and its precursor with 90 stocks was only begun in 1923. So we don't even have 100 years of reliable data on U.S. large-caps, let alone small-caps and SCV.

We do have a long history of good data on U.S. Treasuries though going back to the 1800s.

Gold didn't even begin to free-float in the U.S. until about 1972, though it was never a problem in the U.K., and I believe that others have also found that a relatively small allocation to gold would have benefited U.K. investors over the long-term.
“Good and ill have not changed since yesteryear; nor are they one thing among Elves and Dwarves and another among Men.” J.R.R. Tolkien, The Lord of the Rings
User avatar
steve r
Posts: 872
Joined: Mon Feb 13, 2012 8:34 pm

Re: Great article from Portfolio Charts on most efficient portfolios

Post by steve r »

slicendice wrote: Fri Jan 07, 2022 1:41 pm
steve r wrote: Sun Dec 26, 2021 9:21 am Imagine it is 1970. You decide on a 60 /40 portfolio U.S. TSM and some bonds, but are not sure what type of bonds. Gold is still mostly on the gold standard.

You do some calculations on on a variety of bond holdings with your abacus (or $1000 adding machine) using data similar to the Simba data today on BH since 1871. Despite some serious reservations about the data, you think 100 years -- that ought to tell me something!

The results CAGR (s.d.):

a) 6.78 (11.39)
b) 6.78 (11.66)
c) 6.77 (11.39)
d) 6.93 (11.21)
e) 6.74 (11.43)

We have a winner -- d).

What are the portfolios? Again, 60 /40 U.S. TSM and either TBM, LTT, ITT, STT, and intermediate term govt / corp., in that order.

Yes, the winner would be Short Term Treasuries. :?: In fact, they would remain the back tested winner up through 2001! (and 1871 to 2001 is 131 years)

My point is picking the winning portfolio in 1970 (article data start date) would have been very difficult.

I do believe LTT should have a higher return that shorter term products in theory. That theory strikes me as sound. As does the idea they should provide a touch more diversification benefits. In fact if we extend the analysis though the full period (1871 to 2020), LTT based 60/40 portfolios do come out ahead in terms of CAGR. But not max drawdown, Sharpe, Sortino or Ulcer.

Today, the bond returns (only) mirror what I would have expected. LTT > ITT > STT and IT corp and govt > ITT for the full 151 year period. But this was not the pattern you would observe from 1871 to 2001. Some theories take a long time to pan out! (of course, there are other select shorter periods in history when the expectations did pan out and we are living in one today).
In terms of pre-1970 treasury data-- how useful is it for backtesting? Maybe I'm mistaken, but for a long time (1940's and 1950's at least), weren't treasuries callable? At the long end of the duration spectrum in particular, that would negate the significant positive convexity which I think is a fairly important contributor to the diversification benefit to equities that Swenson alluded to when he talked about the diversification benefit of long-term, non-callable nominal treasuries to an equity portfolio.
I think we can agree the data is the data. I am not sure what impact callable bonds would have had on returns of LTT versus STT. (Clearly, the buyer would demand higher returns in exchange for offer the seller this option.)

A common theme on this thread is how valid back testing is, particularly "way back" back testing. But even more recent back testing. Is anyone buying crypto currency "solely" because of back testing?
“The closer you come to holding the entire market portfolio (all traded securities) the higher your expected return for the risk you take.” William F. Sharpe | VT, BND, TIAA RE and chill.
User avatar
steve r
Posts: 872
Joined: Mon Feb 13, 2012 8:34 pm

Re: Great article from Portfolio Charts on most efficient portfolios

Post by steve r »

willthrill81 wrote: Thu Jan 06, 2022 10:35 pm
spdoublebass wrote: Thu Jan 06, 2022 9:52 pm
steve r wrote: Sun Dec 26, 2021 9:21 am Imagine it is 1970. You decide on a 60 /40 portfolio U.S. TSM and some bonds, but are not sure what type of bonds. Gold is still mostly on the gold standard.

You do some calculations on on a variety of bond holdings with your abacus (or $1000 adding machine) using data similar to the Simba data today on BH since 1871. Despite some serious reservations about the data, you think 100 years -- that ought to tell me something!

The results CAGR (s.d.):

a) 6.78 (11.39)
b) 6.78 (11.66)
c) 6.77 (11.39)
d) 6.93 (11.21)
e) 6.74 (11.43)

We have a winner -- d).

What are the portfolios? Again, 60 /40 U.S. TSM and either TBM, LTT, ITT, STT, and intermediate term govt / corp., in that order.

Yes, the winner would be Short Term Treasuries. :?: In fact, they would remain the back tested winner up through 2001! (and 1871 to 2001 is 131 years)

My point is picking the winning portfolio in 1970 (article data start date) would have been very difficult.

I do believe LTT should have a higher return that shorter term products in theory. That theory strikes me as sound. As does the idea they should provide a touch more diversification benefits. In fact if we extend the analysis though the full period (1871 to 2020), LTT based 60/40 portfolios do come out ahead in terms of CAGR. But not max drawdown, Sharpe, Sortino or Ulcer.

Today, the bond returns (only) mirror what I would have expected. LTT > ITT > STT and IT corp and govt > ITT for the full 151 year period. But this was not the pattern you would observe from 1871 to 2001. Some theories take a long time to pan out! (of course, there are other select shorter periods in history when the expectations did pan out and we are living in one today).
I’ve read this post like 10 times and have no idea what is being said.

Are you saying that in 1970 a 60/40 portfolio of TSM and ST treasury bonds would do better than with LT treasury bonds?

I’m not trying to be confrontational, I just can’t figure out what your trying to say.
He's saying that had you analyzed the prior 100 years of history of these portfolios in 1970, you would have concluded that short-term Treasuries had been marginally superior to the other maturities and/or types of bonds in terms of both returns and volatility (as measured by standard deviation).
Correct ...
Thanks for clarifying WillThrill81. In fact, you could have done a similar analysis in 2001 (using data back to 1871) and drawn the same conclusion.

FWIW, I am a firm believer in the theory that says LTT should have a higher rate of return than ITT or STT (particularly during a period of time when these things were always paid back).
“The closer you come to holding the entire market portfolio (all traded securities) the higher your expected return for the risk you take.” William F. Sharpe | VT, BND, TIAA RE and chill.
mesaverde
Posts: 520
Joined: Wed May 02, 2007 4:14 pm

Re: Great article from Portfolio Charts on most efficient portfolios

Post by mesaverde »

OP, that is a very well presented analysis, thank you for sharing it.
It's a reminder that the entire portfolio is what's important, not individual components.
Beyond that, its use seems negligible because none of us know what the future holds.
"Learn from the past, live in the present, plan for the future"
User avatar
Topic Author
willthrill81
Posts: 28078
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Great article from Portfolio Charts on most efficient portfolios

Post by willthrill81 »

mesaverde wrote: Fri Jan 07, 2022 7:47 pm OP, that is a very well presented analysis, thank you for sharing it.
It's a reminder that the entire portfolio is what's important, not individual components.
You're welcome. Yes, portfolios are more than the sum of their parts. This is the essence of modern portfolio theory, but at least some reject it because they appear to have a one or two dimensional view of the relationships between the performance of a portfolio and its components.
mesaverde wrote: Fri Jan 07, 2022 7:47 pm Beyond that, its use seems negligible because none of us know what the future holds.
I don't agree at all, but perhaps that's because I believe that the future will likely resemble the past in at least some ways. For instance, a portfolio that has provided very strong consistency in its returns for the last 50 years seems unlikely to me to suddenly become highly erratic.

If we ignore historic data, how would we go about determining what our investment strategy should be? How would we determine our withdrawal strategy in retirement?

It's very hard to completely divest one's strategy from the past, and I don't believe that it would be prudent to attempt to do so anyway.
“Good and ill have not changed since yesteryear; nor are they one thing among Elves and Dwarves and another among Men.” J.R.R. Tolkien, The Lord of the Rings
abc132
Posts: 777
Joined: Thu Oct 18, 2018 1:11 am

Re: Great article from Portfolio Charts on most efficient portfolios

Post by abc132 »

steve r wrote: Fri Jan 07, 2022 3:49 pm A common theme on this thread is how valid back testing is, particularly "way back" back testing. But even more recent back testing. Is anyone buying crypto currency "solely" because of back testing?
Backtesting is believed when it is convenient to one's own opinion.

That is really the only way to justify rejecting bond returns of 6% real while accepting gold returns of 6% real.

We should be skeptical of both.
User avatar
Topic Author
willthrill81
Posts: 28078
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Great article from Portfolio Charts on most efficient portfolios

Post by willthrill81 »

Tyler9000 wrote: Thu Jan 06, 2022 1:27 pm Note that in the context of the cloud charts in the "Three Secret Ingredients" article, the baseline return only affects the Y-axis. Raising the cutoff to exclude more below-average outcomes would indeed make stock investors happy but would also make gold skeptics upset. ;) And it would also never move any portfolio to the left.
This is a good point that I initially missed when you wrote this. To a very real extent, gold has been a sort of quasi-insurance against very poor stock performance in particular since the unwinding of Bretton Woods.

A very quick and dirty analysis seems to illustrate this fairly well. Since 1970, the 30 year SWR for a 60/40 portfolio of TSM and ITT was 4.4%. Moving the 40% completely to gold would have boosted that to 4.9%, more than a 10% improvement. And this came in spite of the fact that the time weighted returns of gold from 1980-1999 resulted in it losing almost 75% of its inflation-adjusted value over that specific period! It might seem almost impossible to many that an asset with such terrible performance for two decades managed to significantly boost the 30 year SWR over the entirety of the period analyzed, but it's true.

The reason for the above was simple: the 60/40 suffered badly in the 1970s, while gold soared during that period. But those who frequently opine that gold's performance during that decade was an aberration should observe that a somewhat similar event occurred around the years 1998-2002, when a 60/40 AA went 10-14 years with low real returns (averaging 0-3% real) but when a 60/40 with gold had significantly better returns.

Of course, I'm not recommending a 40% gold allocation to anyone. Rather, I say all of this to illustrate that gold has historically done a good job at counter-balancing the mid-term downside risks of stocks. Long-term Treasuries have as well with the notable exception of the 1970s, where LTT were crushed. And now that yields on LTT are below the market's expected inflation, it's very debatable how much counter-balancing they will provide going forward, notwithstanding your own excellent work on bond convexity.
“Good and ill have not changed since yesteryear; nor are they one thing among Elves and Dwarves and another among Men.” J.R.R. Tolkien, The Lord of the Rings
User avatar
HanSolo
Posts: 1068
Joined: Thu Jul 19, 2012 3:18 am

Re: Great article from Portfolio Charts on most efficient portfolios

Post by HanSolo »

willthrill81 wrote: Fri Jan 07, 2022 11:49 pm And now that yields on LTT are below the market's expected inflation, it's very debatable how much counter-balancing they will provide going forward,
I hope that those who take the side of the debate that bond valuations are actionable aren't also claiming that stock valuations aren't actionable (I think this may include some of the 100/0 people).
Strategic Macro Senior (top 1%, 2019 Bogleheads Contest)
User avatar
Topic Author
willthrill81
Posts: 28078
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Great article from Portfolio Charts on most efficient portfolios

Post by willthrill81 »

HanSolo wrote: Sat Jan 08, 2022 1:26 pm
willthrill81 wrote: Fri Jan 07, 2022 11:49 pm And now that yields on LTT are below the market's expected inflation, it's very debatable how much counter-balancing they will provide going forward,
I hope that those who take the side of the debate that bond valuations are actionable aren't also claiming that stock valuations aren't actionable (I think this may include some of the 100/0 people).
There does seem to be a logical inconsistency there.
“Good and ill have not changed since yesteryear; nor are they one thing among Elves and Dwarves and another among Men.” J.R.R. Tolkien, The Lord of the Rings
User avatar
HomerJ
Posts: 18250
Joined: Fri Jun 06, 2008 12:50 pm

Re: Great article from Portfolio Charts on most efficient portfolios

Post by HomerJ »

willthrill81 wrote: Fri Jan 07, 2022 7:54 pm
mesaverde wrote: Fri Jan 07, 2022 7:47 pm OP, that is a very well presented analysis, thank you for sharing it.
It's a reminder that the entire portfolio is what's important, not individual components.
You're welcome. Yes, portfolios are more than the sum of their parts. This is the essence of modern portfolio theory, but at least some reject it because they appear to have a one or two dimensional view of the relationships between the performance of a portfolio and its components.
mesaverde wrote: Fri Jan 07, 2022 7:47 pm Beyond that, its use seems negligible because none of us know what the future holds.
I don't agree at all, but perhaps that's because I believe that the future will likely resemble the past in at least some ways. For instance, a portfolio that has provided very strong consistency in its returns for the last 50 years seems unlikely to me to suddenly become highly erratic.
Long-term Treasuries (LTT) are likely to have very different returns compared to the last 40 years.
"The best tools available to us are shovels, not scalpels. Don't get carried away." - vanBogle59
User avatar
Topic Author
willthrill81
Posts: 28078
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Great article from Portfolio Charts on most efficient portfolios

Post by willthrill81 »

HomerJ wrote: Sun Jan 09, 2022 2:08 am
willthrill81 wrote: Fri Jan 07, 2022 7:54 pm
mesaverde wrote: Fri Jan 07, 2022 7:47 pm OP, that is a very well presented analysis, thank you for sharing it.
It's a reminder that the entire portfolio is what's important, not individual components.
You're welcome. Yes, portfolios are more than the sum of their parts. This is the essence of modern portfolio theory, but at least some reject it because they appear to have a one or two dimensional view of the relationships between the performance of a portfolio and its components.
mesaverde wrote: Fri Jan 07, 2022 7:47 pm Beyond that, its use seems negligible because none of us know what the future holds.
I don't agree at all, but perhaps that's because I believe that the future will likely resemble the past in at least some ways. For instance, a portfolio that has provided very strong consistency in its returns for the last 50 years seems unlikely to me to suddenly become highly erratic.
Long-term Treasuries (LTT) are likely to have very different returns compared to the last 40 years.
I entirely agree. But, like gold, that doesn't make them a 'bad' component of a larger portfolio. LTT might still do well in the event of a sudden stock decline. That might justify holding on to them. However, I would choose to hold long-term TIPS rather than nominal Treasuries.
“Good and ill have not changed since yesteryear; nor are they one thing among Elves and Dwarves and another among Men.” J.R.R. Tolkien, The Lord of the Rings
Fremdon Ferndock
Posts: 35
Joined: Fri Dec 24, 2021 12:26 pm

Re: Great article from Portfolio Charts on most efficient portfolios

Post by Fremdon Ferndock »

HomerJ wrote: Sun Jan 09, 2022 2:08 am
willthrill81 wrote: Fri Jan 07, 2022 7:54 pm
mesaverde wrote: Fri Jan 07, 2022 7:47 pm OP, that is a very well presented analysis, thank you for sharing it.
It's a reminder that the entire portfolio is what's important, not individual components.
You're welcome. Yes, portfolios are more than the sum of their parts. This is the essence of modern portfolio theory, but at least some reject it because they appear to have a one or two dimensional view of the relationships between the performance of a portfolio and its components.
mesaverde wrote: Fri Jan 07, 2022 7:47 pm Beyond that, its use seems negligible because none of us know what the future holds.
I don't agree at all, but perhaps that's because I believe that the future will likely resemble the past in at least some ways. For instance, a portfolio that has provided very strong consistency in its returns for the last 50 years seems unlikely to me to suddenly become highly erratic.
Long-term Treasuries (LTT) are likely to have very different returns compared to the last 40 years.
We can argue all day about this (and we do). The problem is that we don't know what will happen in the next couple of decades to a portfolio that we craft based on the last couple of decades (or even longer). So, I decided to jump into my time machine and head back to 2000 and run Portfolio Visualizer to come up with the best backtested portfolio allocation from 1978-2000 (22 years) using TSM, SCV, LTT, and Gold. It turned out to be 82% SCV, 18% LTT. It had an annualized return of 17.6%, and max drawdown of 25%, and a Sharpe of 0.72. So I went ahead and invested that way from 2001 - present (21 years). Over that period, the portfolio produced an annualized return of 9.7%, max drawdown of 46%, and Sharpe of 0.60. Not what I was expecting.

If I'd had the benefit of foresight, in 2001 I would have had an allocation of 20% TSM, 15% SCV, 48% LTT, and 17% Gold - based on backtesting from 2022. Shoot! TSM and Gold hadn't even shown up in my 1978 backtest. Now, if I could just get my Time Machine to fast-forward to 2042 so I can run the backtest for 2022-2042. Will Gold show up? Will LTT show up? What about SCV? Oh well, I'll probably be pushing up petunias by then anyway. :beer
elfoolio999
Posts: 12
Joined: Fri May 29, 2015 12:35 pm

Re: Great article from Portfolio Charts on most efficient portfolios

Post by elfoolio999 »

willthrill81 wrote: Sun Jan 09, 2022 10:48 am
HomerJ wrote: Sun Jan 09, 2022 2:08 am
willthrill81 wrote: Fri Jan 07, 2022 7:54 pm
mesaverde wrote: Fri Jan 07, 2022 7:47 pm OP, that is a very well presented analysis, thank you for sharing it.
It's a reminder that the entire portfolio is what's important, not individual components.
You're welcome. Yes, portfolios are more than the sum of their parts. This is the essence of modern portfolio theory, but at least some reject it because they appear to have a one or two dimensional view of the relationships between the performance of a portfolio and its components.
mesaverde wrote: Fri Jan 07, 2022 7:47 pm Beyond that, its use seems negligible because none of us know what the future holds.
I don't agree at all, but perhaps that's because I believe that the future will likely resemble the past in at least some ways. For instance, a portfolio that has provided very strong consistency in its returns for the last 50 years seems unlikely to me to suddenly become highly erratic.
Long-term Treasuries (LTT) are likely to have very different returns compared to the last 40 years.
I entirely agree. But, like gold, that doesn't make them a 'bad' component of a larger portfolio. LTT might still do well in the event of a sudden stock decline. That might justify holding on to them. However, I would choose to hold long-term TIPS rather than nominal Treasuries.
Aren't the Long Term Bonds protection against Deflation, which the TIPS would not provide? Maybe split 20% LTT allocation to include 50% TIPS? Another tweak to GB is to add some int'l.
User avatar
Topic Author
willthrill81
Posts: 28078
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Great article from Portfolio Charts on most efficient portfolios

Post by willthrill81 »

elfoolio999 wrote: Wed Jan 12, 2022 11:42 am
willthrill81 wrote: Sun Jan 09, 2022 10:48 am
HomerJ wrote: Sun Jan 09, 2022 2:08 am
willthrill81 wrote: Fri Jan 07, 2022 7:54 pm
mesaverde wrote: Fri Jan 07, 2022 7:47 pm OP, that is a very well presented analysis, thank you for sharing it.
It's a reminder that the entire portfolio is what's important, not individual components.
You're welcome. Yes, portfolios are more than the sum of their parts. This is the essence of modern portfolio theory, but at least some reject it because they appear to have a one or two dimensional view of the relationships between the performance of a portfolio and its components.
mesaverde wrote: Fri Jan 07, 2022 7:47 pm Beyond that, its use seems negligible because none of us know what the future holds.
I don't agree at all, but perhaps that's because I believe that the future will likely resemble the past in at least some ways. For instance, a portfolio that has provided very strong consistency in its returns for the last 50 years seems unlikely to me to suddenly become highly erratic.
Long-term Treasuries (LTT) are likely to have very different returns compared to the last 40 years.
I entirely agree. But, like gold, that doesn't make them a 'bad' component of a larger portfolio. LTT might still do well in the event of a sudden stock decline. That might justify holding on to them. However, I would choose to hold long-term TIPS rather than nominal Treasuries.
Aren't the Long Term Bonds protection against Deflation, which the TIPS would not provide? Maybe split 20% LTT allocation to include 50% TIPS? Another tweak to GB is to add some int'l.
LTT would protect somewhat against deflation, but TIPS are guaranteed to not lose nominal value, so they still provide some deflation protection. However, remember that Tyler didn't identify the SCV, LTT, and gold in a similar fashion to the way that Harry Browne constructed the Permanent Portfolio, where assets were specifically chosen based on their hypothesized performance in differing market conditions.
“Good and ill have not changed since yesteryear; nor are they one thing among Elves and Dwarves and another among Men.” J.R.R. Tolkien, The Lord of the Rings
Fremdon Ferndock
Posts: 35
Joined: Fri Dec 24, 2021 12:26 pm

Re: Great article from Portfolio Charts on most efficient portfolios

Post by Fremdon Ferndock »

Actually, I-Bonds provide more protection against deflation than TIPS as well as being a better inflation hedge (TIPS market price will go down if real interest rates increase). Of course, the problem with I-Bonds is that you are limited in the amount you can buy in any given year and you can't own them in tax-advantaged accounts. But ... if you've been accumulating I-Bonds you are looking really smart right now! Right, Mel?
User avatar
Mel Lindauer
Moderator
Posts: 32786
Joined: Mon Feb 19, 2007 8:49 pm
Location: Daytona Beach Shores, Florida
Contact:

Re: Great article from Portfolio Charts on most efficient portfolios

Post by Mel Lindauer »

Fremdon Ferndock wrote: Wed Jan 12, 2022 12:46 pm Actually, I-Bonds provide more protection against deflation than TIPS as well as being a better inflation hedge (TIPS market price will go down if real interest rates increase). Of course, the problem with I-Bonds is that you are limited in the amount you can buy in any given year and you can't own them in tax-advantaged accounts. But ... if you've been accumulating I-Bonds you are looking really smart right now! Right, Mel?
Amen to that, FF. Many of those older I Bonds with fixed rates as high as 3.6% are now paying more than 10% for something that some folks consider to be too "stodgy".
Best Regards - Mel | | Semper Fi
User avatar
Topic Author
willthrill81
Posts: 28078
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Great article from Portfolio Charts on most efficient portfolios

Post by willthrill81 »

Mel Lindauer wrote: Wed Jan 12, 2022 10:05 pm
Fremdon Ferndock wrote: Wed Jan 12, 2022 12:46 pm Actually, I-Bonds provide more protection against deflation than TIPS as well as being a better inflation hedge (TIPS market price will go down if real interest rates increase). Of course, the problem with I-Bonds is that you are limited in the amount you can buy in any given year and you can't own them in tax-advantaged accounts. But ... if you've been accumulating I-Bonds you are looking really smart right now! Right, Mel?
Amen to that, FF. Many of those older I Bonds with fixed rates as high as 3.6% are now paying more than 10% for something that some folks consider to be too "stodgy".
I really do hope that the current inflation creates lasting change in how investors view the impact of inflation on their bond holdings. For a long while now, I've advocated for inflation-linked bonds to be at least a major component (and often the sole component) of retail investors' bond holdings. The cost of their inflation 'insurance' has been so low as to arguably have been essentially zero for many years, and all it takes is a single event like current inflation for that insurance to really pay off.
“Good and ill have not changed since yesteryear; nor are they one thing among Elves and Dwarves and another among Men.” J.R.R. Tolkien, The Lord of the Rings
Post Reply