Long term low returns predicted

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
User avatar
Beensabu
Posts: 5658
Joined: Sun Aug 14, 2016 3:22 pm

Re: Long term low returns predicted

Post by Beensabu »

HomerJ wrote: Mon Sep 27, 2021 10:46 am Not that I think the pendulum will ever stop swinging.
But what if it doesn't swing back up as far again for a really, really long time?
"The only thing that makes life possible is permanent, intolerable uncertainty; not knowing what comes next." ~Ursula LeGuin
User avatar
HomerJ
Posts: 21282
Joined: Fri Jun 06, 2008 12:50 pm

Re: Long term low returns predicted

Post by HomerJ »

Beensabu wrote: Mon Sep 27, 2021 2:39 pm
HomerJ wrote: Mon Sep 27, 2021 10:46 am Not that I think the pendulum will ever stop swinging.
But what if it doesn't swing back up as far again for a really, really long time?
Certainly possible. But can one confidently predict it won't swing back up again for a really really long time?

Because that is what these people with the super low long-term predictions are doing.

It's one thing to predict low short-term returns sometime in the future. That's pretty easy. But, so far, lower-than-average returns are followed at some point by higher-than-average returns, and that's how we get the long-term historical average.

It's possible the pattern will break or maybe the pendulum swings will get longer and we'll have 20-30 years between the bad years and good years.

That's possible.

But if someone claims it's likely to happen, I'm not sure I'm going to take them seriously. How could they possibly calculate the odds?
"The best tools available to us are shovels, not scalpels. Don't get carried away." - vanBogle59
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: Long term low returns predicted

Post by skierincolorado »

HomerJ wrote: Mon Sep 27, 2021 2:15 pm
skierincolorado wrote: Mon Sep 27, 2021 1:51 pm Periods of flat CAPE have seen lower returns.

Periods where CAPE has been flat have seen returns substantially below 6-7%.
Show some numbers. Find some long-term periods (around 20 years) where CAPE was flat.

1973-1992, CAPE of 18 in 1973 to 19 in 1992 - 4.5% real.
1989-2009, CAPE of 15 in 1989 to 15 in 2009 - 6.04% real
1993-2010, CAPE of 20 in 1993, to 20 in 2010 - 5.62% real
1996-2016 CAPE of 24 in 1996 to 24 in 2016 - 6.12% real
1999-2021 CAPE of 40 in 1999 to 38 today - 8.35% real

I'm using portfolio visualizer, do you have access to data farther back?

Anyway, you're right it's usually on the low end of the historical average when CAPE is flat.

But still not as far off the average as you claim. Not sure how you can confidently forecast 3.7% real going forward for the next 20-30 years with no change in valuations.

You knock off 1.5% from the historic average assuming no more rise in valuations, but it doesn't appear the historical data matches that number. We got pretty close to the historical average even when CAPE was flat over the long term. We don't appear to need a continued rise in valuations to get close to the average long-term returns.
For 1999-present I get 5.7% real, not 8.35%. Perhaps that was nominal?

2009 ended the year with a CAPE of 20.5, not 15.
2010 ended the year with a CAPE of 23, not 20.
2016 ended the year with a CAPE of 28 not 24.

If you account for the above, you will find all to be below 5.5%, and mostly below 5.0%, except 1999-present which was influenced by rapidly expanding corporate profit margins as the U.S. market business environment deteriorates and becomes less competitive.

For example, you list 1993-2010. The correct date range is 1993 (20.3) - 2009 (20.5). The real return over this period was 4.98%, not 5.62%.

Also using PV. Some of these periods are overlapping. The longest period is 1973-2009 for which we get 4.8%, despite a rise in CAPE from 18.7 to 20.5. This is 1.7% below 6-7% as a normal, despite some increase in CAPE. I think this is a reasonable base assumption. Of course this is constant but much lower CAPE than we had today - meaning the dividend/earning yield was higher. If you adjust for a lower earnings yield today you get down around 4%. One could also reasonably adjust downward for slowing population and productivity growth. One could also adjust down due to the unsustainable rise in corporate profits. These aren't just random variables. Profits and valuations are the two fundamental drivers of stock market returns. Profit margins and valuations can't increase indefinitely. Population and productivity growth have already slowed compared to history.

I'm not bashing on U.S. stocks. I hold a leveraged position in U.S. stocks. 3.5% real, 6% nominal returns are still quite good.
Last edited by skierincolorado on Mon Sep 27, 2021 3:46 pm, edited 7 times in total.
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: Long term low returns predicted

Post by skierincolorado »

vanbogle59 wrote: Mon Sep 27, 2021 2:27 pm
skierincolorado wrote: Mon Sep 27, 2021 1:47 pm
vanbogle59 wrote: Mon Sep 27, 2021 1:37 pm
skierincolorado wrote: Mon Sep 27, 2021 1:09 pm It's not even a "model" it's simply basic math.
Well, not really.
You are calling CAPE valuation, right?
And the C in CAPE is very modelly.
I mean 2 companies, with the exact same earnings last year can have drastically different CAPES based off of their history.
It's not math to say your current "valuation" depends on where you came from, is it?

As for the rest, I generally agree that the market looks "expensive" relative to history. It's been one hell of a bull market with one hell of an accommodating FED. :beer
I agree it is "modelly" to say CAPE=valuation.

There's nothing modelly about CAPE itself though. CAPE is a simple math calculation with a precise definition and means of calculation. What is modelly is the idea that CAPE means anything or can predict market returns. But that's not what I'm trying to do. I'm simply pointing out that in periods where CAPE has stayed constant, CAGR has been lower.

The only modeling I'm suggesting is that CAPE cannot increase indefinitely, so when deciding on a historical "norm" we should look at periods of flat CAPE.
hmmmm.....
Well, that takes out one moving part from the model, at least :D

1) "What is modelly is the idea that CAPE means anything..." (agreed)
2) "when deciding on a historical "norm" we should look at periods of flat CAPE."

If you really thought CAPE didn't mean anything, you wouldn't care about the second sentence at all.
That's why you don't advocate for "The market goes up with the hemlines in Paris" :D

So, still modelly for me :beer
I agree we'd have to asign at least some significance to CAPE. It doesn't have to be significant to the return of individual socks. It doesn't even have to be significant to the return of the whole market. All we have to agree is that, like hemlines, it can't go up forever.
User avatar
vanbogle59
Posts: 1314
Joined: Wed Mar 10, 2021 7:30 pm

Re: Long term low returns predicted

Post by vanbogle59 »

HomerJ wrote: Mon Sep 27, 2021 3:07 pm But, so far, lower-than-average returns are followed at some point by higher-than-average returns, and that's how we get the long-term historical average.
Now that's what I would call "just math"!
:beer
User avatar
vanbogle59
Posts: 1314
Joined: Wed Mar 10, 2021 7:30 pm

Re: Long term low returns predicted

Post by vanbogle59 »

skierincolorado wrote: Mon Sep 27, 2021 3:24 pm All we have to agree is that, like hemlines, it can't go up forever.
Kill joy.
:shock:
User avatar
HomerJ
Posts: 21282
Joined: Fri Jun 06, 2008 12:50 pm

Re: Long term low returns predicted

Post by HomerJ »

skierincolorado wrote: Mon Sep 27, 2021 3:17 pm
HomerJ wrote: Mon Sep 27, 2021 2:15 pm
skierincolorado wrote: Mon Sep 27, 2021 1:51 pm Periods of flat CAPE have seen lower returns.

Periods where CAPE has been flat have seen returns substantially below 6-7%.
Show some numbers. Find some long-term periods (around 20 years) where CAPE was flat.

1973-1992, CAPE of 18 in 1973 to 19 in 1992 - 4.5% 4.47% real.
1989-2009, CAPE of 15 in 1989 to 15 in 2009 - 6.04% 5.15% real
1993-2010, CAPE of 20 in 1993, to 20 in 2010 - 5.62% 5.07% real
1996-2016 CAPE of 24 in 1996 to 24 in 2016 - 6.12% 5.92%real
1999-2021 CAPE of 40 in 1999 to 38 today - 8.35% 5.90% real

I'm using portfolio visualizer, do you have access to data farther back?

Anyway, you're right it's usually on the low end of the historical average when CAPE is flat.

But still not as far off the average as you claim. Not sure how you can confidently forecast 3.7% real going forward for the next 20-30 years with no change in valuations.

You knock off 1.5% from the historic average assuming no more rise in valuations, but it doesn't appear the historical data matches that number. We got pretty close to the historical average even when CAPE was flat over the long term. We don't appear to need a continued rise in valuations to get close to the average long-term returns.
For 1999-present I get 5.7% real, not 8.35%. Perhaps that was nominal?
Whoops, I keep making typos... I put in 1989 to 2021, instead of 1999 to 2021... 1999 to 2021 was 5.90% real, basically historical average.
2009 ended the year with a CAPE of 20.5, not 15.
2010 ended the year with a CAPE of 23, not 20.
2016 ended the year with a CAPE of 28 not 24.
I'm just going with Jan 1st numbers from multpl.com - I tried to find any Jan 1st numbers that were close.

https://www.multpl.com/shiller-pe/table/by-year

And then back-testing using portfolio visualizer going year-to-year which does Jan 1st to Dec 31st

Sep 27, 2021 38.43
Jan 1, 2021 34.51
Jan 1, 2020 30.99
Jan 1, 2019 28.38
Jan 1, 2018 33.31
Jan 1, 2017 28.06
Jan 1, 2016 24.21
Jan 1, 2015 26.49
Jan 1, 2014 24.86
Jan 1, 2013 21.90
Jan 1, 2012 21.21
Jan 1, 2011 22.98
Jan 1, 2010 20.53
Jan 1, 2009 15.17
Jan 1, 2008 24.02
Jan 1, 2007 27.21
Jan 1, 2006 26.47
Jan 1, 2005 26.59
Jan 1, 2004 27.66
Jan 1, 2003 22.90
Jan 1, 2002 30.28
Jan 1, 2001 36.98
Jan 1, 2000 43.77
Jan 1, 1999 40.57
Jan 1, 1998 32.86
Jan 1, 1997 28.33
Jan 1, 1996 24.76
Jan 1, 1995 20.22
Jan 1, 1994 21.41
Jan 1, 1993 20.32
Jan 1, 1992 19.77
Jan 1, 1991 15.61
Jan 1, 1990 17.05
Jan 1, 1989 15.09
Jan 1, 1988 13.90
Jan 1, 1987 14.92
Jan 1, 1986 11.72
Jan 1, 1985 10.00
Jan 1, 1984 9.89
Jan 1, 1983 8.76
Jan 1, 1982 7.39
Jan 1, 1981 9.26
Jan 1, 1980 8.85
Jan 1, 1979 9.26
Jan 1, 1978 9.24
Jan 1, 1977 11.44
Jan 1, 1976 11.19
Jan 1, 1975 8.92
Jan 1, 1974 13.53
Jan 1, 1973 18.71


Using the January numbers I pulled from mutlpl, and then I used portfolio visualizer going from like 1996 to 2016... But you're right I made a mistake...

I should have gone from 1996 to 2015 in portfolio visualizer, because that's Jan 1st, 1996 to Dec 31st, 2015 (when I used 2016, I was going to Dec 31st 2016, not January 1st 2016)

Anyway corrections are above (and here below), and it does reduce the numbers... 4.5% to 5% in the past, 6% since the 1990s.

1973-1992, CAPE of 18 in 1973 to 19 in 1992 - 4.5% 4.47% real.
1989-2009, CAPE of 15 in 1989 to 15 in 2009 - 6.04% 5.15% real
1993-2010, CAPE of 20 in 1993, to 20 in 2010 - 5.62% 5.07% real
1996-2016 CAPE of 24 in 1996 to 24 in 2016 - 6.12% 5.92%real
1999-2021 CAPE of 40 in 1999 to 38 today - 8.35% 5.90% real

So you are correct that flat CAPE periods have been below historical average.
Also using PV. Some of these periods are overlapping. The longest period is 1972-2009 for which we get 5.2%. This is 1.3% below 6-7% as a normal. I think this is a reasonable base assumption.
Eh, I don't think it's reasonable base assumption to use one data point for anything.
If you adjust for a lower earnings yield today you get down around 4%. One could also reasonably adjust downward for slowing population and productivity growth. One could also adjust down due to the unsustainable rise in corporate profits.
All guesses. You are way too confident with a bunch of guesses layered over other guesses layered over a single old data point (while you ignore more recent data points from 1996 and 1999 that don't support your argument).

You may be right. But you don't know. It's not just "math" and indisputable.

"Predictions are hard, especially about the future."
- Yogi Berra

"An economist is an expert who will know tomorrow why the things he predicted yesterday didn't happen today. "
- Laurence J. Peter.
"The best tools available to us are shovels, not scalpels. Don't get carried away." - vanBogle59
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: Long term low returns predicted

Post by skierincolorado »

HomerJ wrote: Mon Sep 27, 2021 4:00 pm
skierincolorado wrote: Mon Sep 27, 2021 3:17 pm
HomerJ wrote: Mon Sep 27, 2021 2:15 pm
skierincolorado wrote: Mon Sep 27, 2021 1:51 pm Periods of flat CAPE have seen lower returns.

Periods where CAPE has been flat have seen returns substantially below 6-7%.
Show some numbers. Find some long-term periods (around 20 years) where CAPE was flat.

1973-1992, CAPE of 18 in 1973 to 19 in 1992 - 4.5% 4.47% real.
1989-2009, CAPE of 15 in 1989 to 15 in 2009 - 6.04% 5.15% real
1993-2010, CAPE of 20 in 1993, to 20 in 2010 - 5.62% 5.07% real
1996-2016 CAPE of 24 in 1996 to 24 in 2016 - 6.12% 5.92%real
1999-2021 CAPE of 40 in 1999 to 38 today - 8.35% 5.90% real

I'm using portfolio visualizer, do you have access to data farther back?

Anyway, you're right it's usually on the low end of the historical average when CAPE is flat.

But still not as far off the average as you claim. Not sure how you can confidently forecast 3.7% real going forward for the next 20-30 years with no change in valuations.

You knock off 1.5% from the historic average assuming no more rise in valuations, but it doesn't appear the historical data matches that number. We got pretty close to the historical average even when CAPE was flat over the long term. We don't appear to need a continued rise in valuations to get close to the average long-term returns.
For 1999-present I get 5.7% real, not 8.35%. Perhaps that was nominal?
Whoops, I keep making typos... I put in 1989 to 2021, instead of 1999 to 2021... 1999 to 2021 was 5.90% real, basically historical average.
2009 ended the year with a CAPE of 20.5, not 15.
2010 ended the year with a CAPE of 23, not 20.
2016 ended the year with a CAPE of 28 not 24.
I'm just going with Jan 1st numbers from multpl.com - I tried to find any Jan 1st numbers that were close.

https://www.multpl.com/shiller-pe/table/by-year

And then back-testing using portfolio visualizer going year-to-year which does Jan 1st to Dec 31st

Sep 27, 2021 38.43
Jan 1, 2021 34.51
Jan 1, 2020 30.99
Jan 1, 2019 28.38
Jan 1, 2018 33.31
Jan 1, 2017 28.06
Jan 1, 2016 24.21
Jan 1, 2015 26.49
Jan 1, 2014 24.86
Jan 1, 2013 21.90
Jan 1, 2012 21.21
Jan 1, 2011 22.98
Jan 1, 2010 20.53
Jan 1, 2009 15.17
Jan 1, 2008 24.02
Jan 1, 2007 27.21
Jan 1, 2006 26.47
Jan 1, 2005 26.59
Jan 1, 2004 27.66
Jan 1, 2003 22.90
Jan 1, 2002 30.28
Jan 1, 2001 36.98
Jan 1, 2000 43.77
Jan 1, 1999 40.57
Jan 1, 1998 32.86
Jan 1, 1997 28.33
Jan 1, 1996 24.76
Jan 1, 1995 20.22
Jan 1, 1994 21.41
Jan 1, 1993 20.32
Jan 1, 1992 19.77
Jan 1, 1991 15.61
Jan 1, 1990 17.05
Jan 1, 1989 15.09
Jan 1, 1988 13.90
Jan 1, 1987 14.92
Jan 1, 1986 11.72
Jan 1, 1985 10.00
Jan 1, 1984 9.89
Jan 1, 1983 8.76
Jan 1, 1982 7.39
Jan 1, 1981 9.26
Jan 1, 1980 8.85
Jan 1, 1979 9.26
Jan 1, 1978 9.24
Jan 1, 1977 11.44
Jan 1, 1976 11.19
Jan 1, 1975 8.92
Jan 1, 1974 13.53
Jan 1, 1973 18.71


Using the January numbers I pulled from mutlpl, and then I used portfolio visualizer going from like 1996 to 2016... But you're right I made a mistake...

I should have gone from 1996 to 2015 in portfolio visualizer, because that's Jan 1st, 1996 to Dec 31st, 2015 (when I used 2016, I was going to Dec 31st 2016, not January 1st 2016)

Anyway corrections are above (and here below), and it does reduce the numbers... 4.5% to 5% in the past, 6% since the 1990s.

1973-1992, CAPE of 18 in 1973 to 19 in 1992 - 4.5% 4.47% real.
1989-2009, CAPE of 15 in 1989 to 15 in 2009 - 6.04% 5.15% real
1993-2010, CAPE of 20 in 1993, to 20 in 2010 - 5.62% 5.07% real
1996-2016 CAPE of 24 in 1996 to 24 in 2016 - 6.12% 5.92%real
1999-2021 CAPE of 40 in 1999 to 38 today - 8.35% 5.90% real

So you are correct that flat CAPE periods have been below historical average.
Also using PV. Some of these periods are overlapping. The longest period is 1972-2009 for which we get 5.2%. This is 1.3% below 6-7% as a normal. I think this is a reasonable base assumption.
Eh, I don't think it's reasonable base assumption to use one data point for anything.
If you adjust for a lower earnings yield today you get down around 4%. One could also reasonably adjust downward for slowing population and productivity growth. One could also adjust down due to the unsustainable rise in corporate profits.
All guesses. You are way too confident with a bunch of guesses layered over other guesses layered over a single old data point (while you ignore more recent data points from 1996 and 1999 that don't support your argument).

You may be right. But you don't know. It's not just "math" and indisputable.

"Predictions are hard, especially about the future."
- Yogi Berra

"An economist is an expert who will know tomorrow why the things he predicted yesterday didn't happen today. "
- Laurence J. Peter.
For 1999-present I get 5.4%, not 5.9%, using VFINX an S&P500 index fund.

It's a 22.75 year period if that helps. The real return was 3.3x. 1.054^22.75=3.3.

So when we use the longest period available, we get under 5%. When we use the most recent, but shorter period which is dominated by an unsustainable explosion in corporate profit margins, we still only get 5.4%. None of these numbers are close to 6.5%. And throughout all of these periods the dividend yield was much much higher than it is today, productivity and population growth were also higher.

Your starting assumption seems to be the last 50 years of U.S. equity returns as if there is something "default" about this period. The reality is that this period is extremely unique and dominated by a few unsustainable trends 1) massive rises in equity valuations 2) massive rises in corporate profit margins 3) unusually high productivity and population growth 4) initially very low valuations with high earnings yields. Adjusting for any single one of these points is enough to bring expected returns below 5%. No major market player today expects real returns over 5%.
Last edited by skierincolorado on Mon Sep 27, 2021 4:25 pm, edited 1 time in total.
Somethingwitty92912
Posts: 490
Joined: Mon Sep 14, 2020 9:43 pm

Re: Long term low returns predicted

Post by Somethingwitty92912 »

alex123711 wrote: Wed Sep 22, 2021 8:50 pm Seeing a lot of predictions (including on this forum) about low returns for a prolonged period (30+ yrs) some predictions are only 1% - 2% above inflation/ CPI. If this does occur then I think index funds will be a bad investment for that time period as 2% over a 30 yr period would not even double your money. At what point does it become 'not worth it' and worth spending the money on other things/ investments/ real estate/ business?
What method did they use for diving the future? I’ve been using only the finest virgin cattle blood. I combine the astrological symbols in a counter clockwise variation using sub-par fractions and I am still coming up with 15% minimum return?
User avatar
HomerJ
Posts: 21282
Joined: Fri Jun 06, 2008 12:50 pm

Re: Long term low returns predicted

Post by HomerJ »

skierincolorado wrote: Mon Sep 27, 2021 4:13 pmFor 1999-present I get 5.4%, not 5.9%, using VFINX an S&P500 index fund.
I was using Asset class "U.S. Stock Market" for all my tests, since I didn't know any good funds to use in 1973.

When I use VFINX from January 1st 1999 to August 2021, I also get 5.41% real, so that matches what you see.

When I use VTSMX from January 1st 1999 to August 2021, I get 5.9% real, so I guess the Asset Class "U.S. stock market" is total market.

I usually use VTSAX (Vanguard Total Stock Market Admiral) when calculating "stock market" returns (But it doesn't start until Dec 2000), so I used VTSMX instead here.

Interesting that SP500 and Total Stock Market were so far apart.

See how hard this is? One can pick and choose the data that supports their argument, and technically be "correct". No one would fault you for using SP 500 from 1999-2021... But no one would fault me for using Total Stock Market Index either.

5.41% real is a decent chunk below historical average, while 5.9% is close enough to the historical average of 6%-7% that one could say we still got the long-term average even investing at CAPE of 40 in 1999.

I don't know how many times to say it, but economics is a messy business, and it's hard to even agree about the past, let alone the future.
Your starting assumption seems to be the last 50 years of U.S. equity returns as if there is something "default" about this period.
No, the last 120 years also had the same historical long-term averages.
"The best tools available to us are shovels, not scalpels. Don't get carried away." - vanBogle59
User avatar
Beensabu
Posts: 5658
Joined: Sun Aug 14, 2016 3:22 pm

Re: Long term low returns predicted

Post by Beensabu »

HomerJ wrote: Mon Sep 27, 2021 3:07 pm How could they possibly calculate the odds?
They can't. You can't calculate them either way, unfortunately. Everyone just weights different possibilities however makes most sense to them and then acts and plans accordingly. Every theory is just a theory, and every model is just a model.
"The only thing that makes life possible is permanent, intolerable uncertainty; not knowing what comes next." ~Ursula LeGuin
nobody special
Posts: 10
Joined: Sun Apr 23, 2017 12:33 pm

Re: Long term low returns predicted

Post by nobody special »

skierincolorado wrote: Mon Sep 27, 2021 3:24 pm I agree we'd have to asign at least some significance to CAPE. It doesn't have to be significant to the return of individual socks. It doesn't even have to be significant to the return of the whole market. All we have to agree is that, like hemlines, it can't go up forever.
From a position of relative ignorance, may I ask the following question?

Why do you assume that "constant CAPE" is actually a proper constant measurement? I recall discussions in this forum concerning the increase in CAPE, there were arguments that accounting changes can account for a ~ 5 point increase. If that is true then "constant" CAPE would not actually be constant.
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: Long term low returns predicted

Post by skierincolorado »

HomerJ wrote: Mon Sep 27, 2021 4:33 pm
skierincolorado wrote: Mon Sep 27, 2021 4:13 pmFor 1999-present I get 5.4%, not 5.9%, using VFINX an S&P500 index fund.
I was using Asset class "U.S. Stock Market" for all my tests, since I didn't know any good funds to use in 1973.

When I use VFINX from January 1st 1999 to August 2021, I also get 5.41% real, so that matches what you see.

When I use VTSMX from January 1st 1999 to August 2021, I get 5.9% real, so I guess the Asset Class "U.S. stock market" is total market.

I usually use VTSAX (Vanguard Total Stock Market Admiral) when calculating "stock market" returns (But it doesn't start until Dec 2000), so I used VTSMX instead here.

Interesting that SP500 and Total Stock Market were so far apart.

See how hard this is? One can pick and choose the data that supports their argument, and technically be "correct". No one would fault you for using SP 500 from 1999-2021... But no one would fault me for using Total Stock Market Index either.

5.41% real is a decent chunk below historical average, while 5.9% is close enough to the historical average of 6%-7% that one could say we still got the long-term average even investing at CAPE of 40 in 1999.

I don't know how many times to say it, but economics is a messy business, and it's hard to even agree about the past, let alone the future.
Your starting assumption seems to be the last 50 years of U.S. equity returns as if there is something "default" about this period.
No, the last 120 years also had the same historical long-term averages.
We used the PE ratios on the S&P500. It doesn't surprise me that the return of total market is different. It's also a short period with exploding corporate profit margins (5% rising to 13%). Do you expect that corporate profit margins will rise from 13% to 21% in the next 22 years? It's possible, but is this a reasonable base assumption?

The much longer and more meaningful period of 1973-2009 is below 5% as one would expect, and that's despite some increase in PE over this period. And there was some increase in corporate profit margins during this period as well, but not as overwhelming as from 1999-present.
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: Long term low returns predicted

Post by skierincolorado »

nobody special wrote: Mon Sep 27, 2021 4:55 pm
skierincolorado wrote: Mon Sep 27, 2021 3:24 pm I agree we'd have to asign at least some significance to CAPE. It doesn't have to be significant to the return of individual socks. It doesn't even have to be significant to the return of the whole market. All we have to agree is that, like hemlines, it can't go up forever.
From a position of relative ignorance, may I ask the following question?

Why do you assume that "constant CAPE" is actually a proper constant measurement? I recall discussions in this forum concerning the increase in CAPE, there were arguments that accounting changes can account for a ~ 5 point increase. If that is true then "constant" CAPE would not actually be constant.
This is a valid point, but matters less over longer time horizons. My understanding is that it amounts to more like a few points. CAPE rose from 18.7 to 20.5 from 1973-2009, and real returns were 4.8%. This is the longest most representative period I can find for for CAPE rising 2-3 points and is a good starting point for the real return of U.S. equities.

Other considerations would be corporate profit margins (which grew modestly over 1973-2009), slowing population growth, and slowing productivity growth.
User avatar
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 2:17 pm
Location: USA
Contact:

Re: Long term low returns predicted

Post by willthrill81 »

HomerJ wrote: Mon Sep 27, 2021 4:33 pm 5.41% real is a decent chunk below historical average, while 5.9% is close enough to the historical average of 6%-7% that one could say we still got the long-term average even investing at CAPE of 40 in 1999.

I don't know how many times to say it, but economics is a messy business, and it's hard to even agree about the past, let alone the future.
And it gets even messier when we throw in accumulating or withdrawing from the portfolio along the way, which is what nearly all investors do over a 20+ year period. 1999-2021 has been a fantastic period for accumulators (i.e., low initial returns followed by stellar returns) but not so good for retirees. But even so, the '4% rule' is virtually guaranteed to hold up for both 1999 and 2000 retirees.
The Sensible Steward
User avatar
HomerJ
Posts: 21282
Joined: Fri Jun 06, 2008 12:50 pm

Re: Long term low returns predicted

Post by HomerJ »

skierincolorado wrote: Mon Sep 27, 2021 5:38 pmIt's also a short period with exploding corporate profit margins (5% rising to 13%).
You can always find a variable that changed to back up your story.

But a thousand different variables changed. Some negative for returns, some positive.

A thousand different variables will change over the next 20 years. Some negative for returns, some positive.

If we end up with net positive, you will find one to focus on, and wave away the failure of valuations to predict the future. Again.

Every time someone says "Well, CAPE predictions didn't work because of THIS", I say "Exactly".

There are more variables than CAPE. That's the point.
"The best tools available to us are shovels, not scalpels. Don't get carried away." - vanBogle59
User avatar
vanbogle59
Posts: 1314
Joined: Wed Mar 10, 2021 7:30 pm

Re: Long term low returns predicted

Post by vanbogle59 »

HomerJ wrote: Mon Sep 27, 2021 6:54 pm
skierincolorado wrote: Mon Sep 27, 2021 5:38 pmIt's also a short period with exploding corporate profit margins (5% rising to 13%).
You can always find a variable that changed to back up your story.

But a thousand different variables changed. Some negative for returns, some positive.

A thousand different variables will change over the next 20 years. Some negative for returns, some positive.

If we end up with net positive, you will find one to focus on, and wave away the failure of valuations to predict the future. Again.

Every time someone says "Well, CAPE predictions didn't work because of THIS", I say "Exactly".

There are more variables than CAPE. That's the point.
Psychohistory is the answer.
I think there is a documentary on AppleTV.
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: Long term low returns predicted

Post by skierincolorado »

HomerJ wrote: Mon Sep 27, 2021 6:54 pm
skierincolorado wrote: Mon Sep 27, 2021 5:38 pmIt's also a short period with exploding corporate profit margins (5% rising to 13%).
You can always find a variable that changed to back up your story.

But a thousand different variables changed. Some negative for returns, some positive.

A thousand different variables will change over the next 20 years. Some negative for returns, some positive.

If we end up with net positive, you will find one to focus on, and wave away the failure of valuations to predict the future. Again.

Every time someone says "Well, CAPE predictions didn't work because of THIS", I say "Exactly".

There are more variables than CAPE. That's the point.
That's not what I said. The return 1999-present *was* low. You're saying it like it wasn't. It *was* low 1999-2021 because CAPE didn't rise. It would have been even *lower* if corporate profitability didn't blow the barn roof off.

Economies are complex. Intermediate predictions are near impossible. But the fundamentals of long term returns are quite simple. Economic production * corporate profit margins * valuations. Economic production is difficult to predict, but the latter two are simply constants that cannot increase indefinitely. You'd clearly rather cherry pick periods of history where these *constants* increased.
User avatar
HomerJ
Posts: 21282
Joined: Fri Jun 06, 2008 12:50 pm

Re: Long term low returns predicted

Post by HomerJ »

skierincolorado wrote: Tue Sep 28, 2021 12:06 am
HomerJ wrote: Mon Sep 27, 2021 6:54 pm
skierincolorado wrote: Mon Sep 27, 2021 5:38 pmIt's also a short period with exploding corporate profit margins (5% rising to 13%).
You can always find a variable that changed to back up your story.

But a thousand different variables changed. Some negative for returns, some positive.

A thousand different variables will change over the next 20 years. Some negative for returns, some positive.

If we end up with net positive, you will find one to focus on, and wave away the failure of valuations to predict the future. Again.

Every time someone says "Well, CAPE predictions didn't work because of THIS", I say "Exactly".

There are more variables than CAPE. That's the point.
That's not what I said. The return 1999-present *was* low. You're saying it like it wasn't. It *was* low 1999-2021 because CAPE didn't rise. It would have been even *lower* if corporate profitability didn't blow the barn roof off.
5.9% real is right on the low-end of the historical long-term average.

My point stands. Starting from a 40 CAPE, with no basically change in CAPE, we still got the historical average (the low-end, sure) over 21 years.

So starting from a 38 CAPE today doesn't automatically mean we're doomed over the next 20-30 years.
You'd clearly rather cherry pick periods of history where these *constants* increased.
You asked for periods where the constants didn't increase. Where the starting point and the end-point of CAPE were the same.

I provided those intervals in the last 50 years (I would have gone back further but portfolio visualizer only goes back 50 years).

There were slightly lower than the historical average. So you were indeed somewhat correct. But they weren't that far off from the historical average. So I also was somewhat correct. The stock market works in cycles.

Starting from a CAPE of 15, the market went way up, then crashed back to 15, and the good years and the bad years averaged out to 5.15% returns

Starting from a CAPE of 24, the market went up a little, then crashed down, then went back up to 24, and the good years and bad years averaged out to 5.92% real.

Starting from a CAPE of 40, the market crashed down, then went back up to 38, and the good years and bad years averaged out to 5.90% real.

The big mistake many people make about high valuations is they focus on the impending low returns, but so far, low returns have always been followed by high returns, and they average out to a decent return. So ignoring valuations for long-term investments has worked all these years.

I don't think any more evidence will change your opinion. My apologies for wasting your time.
Last edited by HomerJ on Tue Sep 28, 2021 12:29 am, edited 4 times in total.
"The best tools available to us are shovels, not scalpels. Don't get carried away." - vanBogle59
Thesaints
Posts: 5108
Joined: Tue Jun 20, 2017 12:25 am

Re: Long term low returns predicted

Post by Thesaints »

CAPE may have some relevance only in a stationary economy, i.e. an economy where the nature of businesses does not change much over time.
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: Long term low returns predicted

Post by skierincolorado »

HomerJ wrote: Tue Sep 28, 2021 12:21 am
skierincolorado wrote: Tue Sep 28, 2021 12:06 am
HomerJ wrote: Mon Sep 27, 2021 6:54 pm
skierincolorado wrote: Mon Sep 27, 2021 5:38 pmIt's also a short period with exploding corporate profit margins (5% rising to 13%).
You can always find a variable that changed to back up your story.

But a thousand different variables changed. Some negative for returns, some positive.

A thousand different variables will change over the next 20 years. Some negative for returns, some positive.

If we end up with net positive, you will find one to focus on, and wave away the failure of valuations to predict the future. Again.

Every time someone says "Well, CAPE predictions didn't work because of THIS", I say "Exactly".

There are more variables than CAPE. That's the point.
That's not what I said. The return 1999-present *was* low. You're saying it like it wasn't. It *was* low 1999-2021 because CAPE didn't rise. It would have been even *lower* if corporate profitability didn't blow the barn roof off.
5.9% real is right on the low-end of the historical long-term average.

My point stands. Starting from a 40 CAPE, with no basically change in CAPE, we still got the historical average (the low-end, sure) over 21 years.

So starting from a 38 CAPE today doesn't automatically mean we're doomed over the next 20-30 years.
You'd clearly rather cherry pick periods of history where these *constants* increased.
You asked for periods where the constants didn't increase. Where the starting point and the end-point of CAPE were the same.

I provided those intervals in the last 50 years (I would have gone back further but portfolio visualizer only goes back 50 years).

There were slightly lower than the historical average. So you were indeed somewhat correct. But they weren't that far off from the historical average. So I also was somewhat correct. The stock market works in cycles.

Starting from a CAPE of 15, the market went way up, then crashed back to 15, and the good years and the bad years averaged out to 5.15% returns

Starting from a CAPE of 24, the market went up a little, then crashed down, then went back up to 24, and the good years and bad years averaged out to 5.92% real.

Starting from a CAPE of 40, the market crashed down, then went back up to 38, and the good years and bad years averaged out to 5.90% real.

The big mistake many people make about high valuations is they focus on the impending low returns, but so far, low returns have always been followed by high returns, and they average out to a decent return. So ignoring valuations for long-term investments has worked all these years.

I don't think any more evidence will change your opinion. My apologies for wasting your time.
The most reasonable thing to do instead of focusing on the two shortest periods with the highest returns would be to take a duration weighted average of 73- dec 09 and 99-present. Also consider that CAPE rose from 18 to 20 from '73 to dec 31 '09. The two periods are 37 and 22.75 years respectively. The returns were 4.8% and 5.9% respectively.

The duration weighted average is a 5.2% return which is 1.3% lower than a 6.5% "normal" and very near the 1.5% difference I originally suggested.

And of course there are still 3 remaining reasons to adjust downwards

1) lower earnings yield. Flat 20 CAPE is not the same thing as flat 40 CAPE. The earnings yield is 2.5% lower (2.5% vs 5%)

2) corporate profit margins cannot increase forever

3) population and productivity growth have slowed


You might be willing to dismiss these fundamental drivers of stock returns with some hand waving and magic pendulums and dependent variables. But the fact is these are *the* fundamental drivers of stocks returns and predict previous returns well - including all of your "cycles". Your "cycles" are really just cycles in PE, corporate margins, and production. And right now we happen to be in a cycle at the top of all 3. I'm sure you'll just call me "over-confident" again without addressing the fundamental components of stock returns. I don't claim to know what the next 30 years will bring. Mabe a new technological invention will create powerful productivity growth. Maybe political instability will make returns even lower. But a good starting point, based on the historical fundamental drivers of stock returns, is 3-4% real return.
Post Reply