Can future returns for the next 10 years really be predicted with CAPE?

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
Topic Author
YRT70
Posts: 193
Joined: Sat Apr 27, 2019 8:51 am

Can future returns for the next 10 years really be predicted with CAPE?

Post by YRT70 » Tue May 14, 2019 9:03 am

I was reading a text by ERN. I'm relatively new to investing. I'm about to invest a decent amount of money in stocks. I'm new to using CAPE for predicting returns so this left me puzzled.
That's right, today's CAPE ratio is pretty good at predicting future equity returns. Well, not perfectly but there seems to be a strong and statistically signicant inverse relationship between the CAPE and forward-looking equity returns, see chart below where we plot the CAPE ratio versus the subsequent 10-year annualized S&P500 return. For something as ostensibly unpredictable as stock returns, this is truly amazing. Equity returns are not exactly a random-walk! If we split the CAPE into four regions we get pretty different average equity returns by bin:
CAPE below 15 (below the median): Average equity return of 9% real (!) CAPE slightly elevated (15-20): Average equity return just under 6%, still very solid returns that will
likely support a 4% safe withdrawal rate.
CAPE moderately elevated (20-30): Only about 3% real return (!) going forward. Today's CAPE
falls into this range. The 9/30/2016 level was at just under 27, and after the recent rally, it's even a
bit above 27.
CAPE severely elevated (30+): A below -1% real return over the next ten years. Bummer! Good luck
starting your retirement in that environment!
Especially that last line worried me. Below -1% seems like a horrible outlook for the next ten years. I've got a feeling the guy who wrote it is overselling the predictive power.

Can someone educate me a bit? Maybe send me to higher quality articles? Thanks in advance.

Source of the quote: https://poseidon01.ssrn.com/delivery.ph ... 26&EXT=pdf
page 14

thx1138
Posts: 944
Joined: Fri Jul 12, 2013 2:14 pm

Re: Can future returns for the next 10 years really be predicted with CAPE?

Post by thx1138 » Tue May 14, 2019 9:19 am

YRT70 wrote:
Tue May 14, 2019 9:03 am
I've got a feeling the guy who wrote it is overselling the predictive power.
You'll do well in investing based on the conclusion you've drawn above ;)

The answer to the question in your title is somewhere between "No" and "A tiny bit".

Think of the market as a die and CAPE as having stuck a booger onto one side of it. Indeed the booger will somewhat affect the outcome and if you were able to place a bet on the distribution of values the die will roll over the next say 1000 rolls then indeed your knowledge of which side the booger was stuck on might do you some good.

Unfortunately for your retirement you get to roll the die once. It would be a bad idea to base all your retirement decisions on the knowledge of the booger's location.
Can someone educate me a bit? Maybe send me to higher quality articles? Thanks in advance.
This particular topic is discussed ad nauseam here on Bogleheads. In the upper right is a search box - type "CAPE" in there and you'll find endless articles referred to. Please note that none of the articles agree with each other at all. This should tell you something about the predictive power of CAPE. Also note that people can't even typically agree on exactly which definition of CAPE they are talking about...

EDIT: In general I should add the consensus though is that at the very least CAPE is useful as a warning to someone about to retire that they should perhaps consider that if CAPE is very high perhaps they should view their current portfolio value as inflated and plan their retirement spending (or the need to save more) as if that portfolio were smaller. As to using CAPE to change what you invest in you'll find few that advocate that here.
Last edited by thx1138 on Tue May 14, 2019 9:22 am, edited 1 time in total.

Living Free
Posts: 290
Joined: Thu Jul 19, 2018 7:31 pm

Re: Can future returns for the next 10 years really be predicted with CAPE?

Post by Living Free » Tue May 14, 2019 9:20 am

It's very hard to predict the future.

SovereignInvestor
Posts: 342
Joined: Mon Aug 20, 2018 4:41 pm

Re: Can future returns for the next 10 years really be predicted with CAPE?

Post by SovereignInvestor » Tue May 14, 2019 9:24 am

CAPE has a bias high IMO about 30% (roughly estimated in my article below) in recent years relative to older years bevause buybacks, CPI changes in mid 1990s and lower tax rate post 2017.

It is statistical invalid to compare recent readings to long term averages when the mean is signifcantly changing. It has understated returns in the last 2-3 decades.

https://seekingalpha.com/article/408638 ... d-buybacks

Shiller even acknowledged a major issue with buybacks and added a field to the public spreadsheet but IMO this doesn't fix the issue.

CAPE says SPX should be around 1700. Really, it should be at 10x forward earnings for a yield of 10%? That's just ridiculous, bevause even if ecpnomy never grew ever and therefore earnings never grew, long run returns would be around 10%, with 10Y yield under 2.5% for a risk premium of 7.5%. The historical average risk premium of stocks over 10Y note is barely 5%.

So that means if we take CAPE mean level and say market should be there, in the most bearish scenario of no economic growth or inflation in the future (no nominal GDP growth), the S&P likely still provides a much greater return premium over bonds than historically.

Seems like CAPE is off base. It had the bias issue which makes the current readings IMO 30% biased high compared to older years all else equal (constant prospective earnings power), and then aside from thst issue, it makes no distinguishment of interest rates. If it said stocks are unfavorable levels then it suggests buying bonds, but it's all relative, yet CAPE users implicitly make absolute comparisons.

dbr
Posts: 29533
Joined: Sun Mar 04, 2007 9:50 am

Re: Can future returns for the next 10 years really be predicted with CAPE?

Post by dbr » Tue May 14, 2019 9:27 am

thx1138 wrote:
Tue May 14, 2019 9:19 am

Think of the market as a die and CAPE as having stuck a booger onto one side of it. Indeed the booger will somewhat affect the outcome and if you were able to place a bet on the distribution of values the die will roll over the next say 1000 rolls then indeed your knowledge of which side the booger was stuck on might do you some good.

Unfortunately for your retirement you get to roll the die once. It would be a bad idea to base all your retirement decisions on the knowledge of the booger's location.

Well said. People constantly misunderstand what it means to predict or estimate returns. What is being estimated is the average of a very wide distribution of possible returns, aka the "expected return." But you can't expect to get the expected return.

megabad
Posts: 1858
Joined: Fri Jun 01, 2018 4:00 pm

Re: Can future returns for the next 10 years really be predicted with CAPE?

Post by megabad » Tue May 14, 2019 9:32 am

dbr wrote:
Tue May 14, 2019 9:27 am
thx1138 wrote:
Tue May 14, 2019 9:19 am

Think of the market as a die and CAPE as having stuck a booger onto one side of it. Indeed the booger will somewhat affect the outcome and if you were able to place a bet on the distribution of values the die will roll over the next say 1000 rolls then indeed your knowledge of which side the booger was stuck on might do you some good.

Unfortunately for your retirement you get to roll the die once. It would be a bad idea to base all your retirement decisions on the knowledge of the booger's location.

Well said. People constantly misunderstand what it means to predict or estimate returns. What is being estimated is the average of a very wide distribution of possible returns, aka the "expected return." But you can't expect to get the expected return.
This is great! Ill have to remember the die booger analogy in the future. But yes, statistically CAPE is a better predictor than many other predictors for future performance. Unfortunately it is still wrong a lot. I think the Vanguard report listed a 40 percent predictive “accuracy”. In that case, by using one of the best predictors we have, you would still be wrong most of the time.

User avatar
HomerJ
Posts: 13034
Joined: Fri Jun 06, 2008 12:50 pm

Re: Can future returns for the next 10 years really be predicted with CAPE?

Post by HomerJ » Tue May 14, 2019 10:58 am

Let's test those assertions for the past 27 years. Only 18 data points, since we're looking at 10-year returns.
  • CAPE below 15 (below the median): Average equity return of 9% real
  • CAPE slightly elevated (15-20): Average equity return just under 6%
  • CAPE moderately elevated (20-30): Only about 3% real return
  • CAPE severely elevated (30+): A below -1% real return over the next ten years.
CAPE crossed 20 in 1992, and has never dipped below it except for a brief period in 2009. Even then it just got back to 15, not below it.

So we only have 1 data point where CAPE was 15-20 in the past 27 years. (6% real predicted)
  • 2009 - 10-year returns: about 15% nominal (13% real)
That's a lot higher than 6% real...

We have 12 data points where CAPE was 20-30 in the past 27 years, and we have full 10-year returns. (3% real predicted)
  • 1992 - 11.5% nominal (8.5% real)
  • 1993 - 9.5% nominal (6.5% real)
  • 1994 - 11% nominal (8% real)
  • 1995 - 10% nominal (7% real)
  • 1996 - 8.5% nominal (5.5% real)
  • 1997 - 8.5% nominal (5.5% real)
  • 2003 - 8.5% nominal (6.5% real)
  • 2004 - 8.5% nominal (6.5% real)
  • 2005 - 9% nominal (7% real)
  • 2006 - 7% nominal (5% real)
  • 2007 - 7% nominal (5% real)
  • 2008 - 9.5% nominal (7.5% real)
Do any of those numbers look like a good match for the predicted value of 3% real? (Note I used 3% inflation for 90s, and 2% inflation for the 2000s).

2010 and 2011 had CAPEs in the 20-30 range, and even though we don't have 10-year returns for them yet, they are on track to get 10.5% real and 10% real

How can someone really say with a straight face that "today's CAPE ratio is pretty good at predicting future equity returns."?

We have 5 years where CAPE was above 30... 1998, 1999, 2000, 2001, 2002. (-1% real predicted)

1999 and 2000 were indeed very bad, negative real returns. The other years had very low, but positive real returns. I'm not sure how the average comes out to -1% real for those years (1999 was REALLY bad since the 10-year period ended in 2009, so maybe that dragged everything else down).

I think a big part of the those ten-year periods being poor was the coincidence of two bear markets 9 years apart.

But, in any case, we only have one set of data points above 30. So it's a pretty small sample.

As far as CAPE in the 20-30 range, CAPE has utterly failed over the past few decades to predict returns accurately in that range.

Remember, CAPE was DERIVED from the data from 1900-1988. The fact that it matches those years fairly well doesn't prove very much. Ever since it was discovered, it's done a poor job of predicting returns. That's how you test a model, and CAPE as a model has not done very well in the past 30 years.
The J stands for Jay

SovereignInvestor
Posts: 342
Joined: Mon Aug 20, 2018 4:41 pm

Re: Can future returns for the next 10 years really be predicted with CAPE?

Post by SovereignInvestor » Tue May 14, 2019 11:18 am

HomerJ wrote:
Tue May 14, 2019 10:58 am
Let's test those assertions for the past 27 years. Only 18 data points, since we're looking at 10-year returns.
  • CAPE below 15 (below the median): Average equity return of 9% real
  • CAPE slightly elevated (15-20): Average equity return just under 6%
  • CAPE moderately elevated (20-30): Only about 3% real return
  • CAPE severely elevated (30+): A below -1% real return over the next ten years.
CAPE crossed 20 in 1992, and has never dipped below it except for a brief period in 2009. Even then it just got back to 15, not below it.

So we only have 1 data point where CAPE was 15-20 in the past 27 years. (6% real predicted)
  • 2009 - 10-year returns: about 15% nominal (13% real)
That's a lot higher than 6% real...

We have 12 data points where CAPE was 20-30 in the past 27 years, and we have full 10-year returns. (3% real predicted)
  • 1992 - 11.5% nominal (8.5% real)
  • 1993 - 9.5% nominal (6.5% real)
  • 1994 - 11% nominal (8% real)
  • 1995 - 10% nominal (7% real)
  • 1996 - 8.5% nominal (5.5% real)
  • 1997 - 8.5% nominal (5.5% real)
  • 2003 - 8.5% nominal (6.5% real)
  • 2004 - 8.5% nominal (6.5% real)
  • 2005 - 9% nominal (7% real)
  • 2006 - 7% nominal (5% real)
  • 2007 - 7% nominal (5% real)
  • 2008 - 9.5% nominal (7.5% real)
Do any of those numbers look like a good match for the predicted value of 3% real? (Note I used 3% inflation for 90s, and 2% inflation for the 2000s).

2010 and 2011 had CAPEs in the 20-30 range, and even though we don't have 10-year returns for them yet, they are on track to get 10.5% real and 10% real

How can someone really say with a straight face that "today's CAPE ratio is pretty good at predicting future equity returns."?

We have 5 years where CAPE was above 30... 1998, 1999, 2000, 2001, 2002. (-1% real predicted)

1999 and 2000 were indeed very bad, negative real returns. The other years had very low, but positive real returns. I'm not sure how the average comes out to -1% real for those years (1999 was REALLY bad since the 10-year period ended in 2009, so maybe that dragged everything else down).

I think a big part of the those ten-year periods being poor was the coincidence of two bear markets 9 years apart.

But, in any case, we only have one set of data points above 30. So it's a pretty small sample.

As far as CAPE in the 20-30 range, CAPE has utterly failed over the past few decades to predict returns accurately in that range.

Remember, CAPE was DERIVED from the data from 1900-1988. The fact that it matches those years fairly well doesn't prove very much. Ever since it was discovered, it's done a poor job of predicting returns. That's how you test a model, and CAPE as a model has not done very well in the past 30 years.
Excellent points! You've made a strong case that CAPE has been a poor predictor and understating expected returns. Let me explain why:

It's not just a poor predictor in that is lacks precision and has high error.

It has a bias. It will understate returns after mid 1990s due to.CPI being changed and then especially severely after 2000 when buybacks really surged.

Right now the CAPE includes 2009 earnings when if they occurred in 2019 would be about 35% higher due to buybacks since then. This isn't an issue except that when deriving 1989 CAPE using data from 1979...there were not really buybacks then so the earnings per SHARE didn't have an on-level issue for buybacks/share count changes. Currently the EPS used aren't on the higher level prospectively reflecting lower share cpunts so CAPE will be greater now.

Not only is CAPE model calibrated on 1900-1980 data lack accuracy, but it will have a high bias of CAPE and low for predicted returns in that the expected return generated from model will all else equal on average be too low.

Using a model for data from 100 years when the data of last 20 has a substantially different mean is statistically invalid at best and malpractice at worst.

SovereignInvestor
Posts: 342
Joined: Mon Aug 20, 2018 4:41 pm

Re: Can future returns for the next 10 years really be predicted with CAPE?

Post by SovereignInvestor » Tue May 14, 2019 11:24 am

BY they way...on average CAPE understated returns about 4% per year for 10year periods that Homer listed. I estimated CAPE is biased high around 30% for periods after mid 1990s.

Cape being 30% too highky roughly translates into 4% bias error.

Very roughly speaking...30% adjustment over 10 years is about 2.75% a year in understating return from PE expansion or contraction. Then also if average shareholder yield is 5% roughly 30% difference is 1.5%. 2.75% + 1.5% roughly speaking is the 4% ballpark that CAPE has been off by.

IMO it will continue to understate real annual returns in the 4% ballpark.

ohai
Posts: 821
Joined: Wed Dec 27, 2017 2:10 pm

Re: Can future returns for the next 10 years really be predicted with CAPE?

Post by ohai » Tue May 14, 2019 11:45 am

I wonder how much this CAPE ratio is biased by companies like AMZN, which have essentially zero trailing earnings but nothing short of world domination priced into their future. I suspect that accounting rules might have changed as well, or companies have evolved in ways that lean towards understating earnings. On top of that, the macro environment has changed due to low interest rates. This would naturally cause valuation metrics to look more "expensive" than in the past.

Coltrane75
Posts: 64
Joined: Wed Feb 06, 2019 2:32 pm

Re: Can future returns for the next 10 years really be predicted with CAPE?

Post by Coltrane75 » Tue May 14, 2019 12:00 pm

I think using CAPE for estimating future returns is actually pretty decent. See the at the link below where expected and actual returns are plotted together. In this case, these are expected returns over 8 year periods.

https://www.gurufocus.com/shiller-PE.php

It uses the same formula Bogle used for determining what investors should expect.

There is also a study done using this formula, for 10 years, to predict ten year returns and it did reasonably well.

I don't think its very useful for market timing, but useful for confirming/adjusting your exposure to risk.

reformed.trader
Posts: 70
Joined: Sat Apr 15, 2017 11:14 pm

Re: Can future returns for the next 10 years really be predicted with CAPE?

Post by reformed.trader » Tue May 14, 2019 12:33 pm

It has shown to be good predicting returns cross sectionally. Saying the 10 returns of the US stock market at a cape of 30 will be X will most certainly be wrong, but for example, there is a great chance Emerging Markets as a whole will outperform the US over the next 10 years.

2018 was a decent year for low CAPE countries.

https://www.starcapital.de/fileadmin/us ... imling.pdf


And CAPE, as well as other value metrics, has been shown to have long term predictive power in countries across the globe.

https://www.starcapital.de/fileadmin/us ... imling.pdf

SovereignInvestor
Posts: 342
Joined: Mon Aug 20, 2018 4:41 pm

Re: Can future returns for the next 10 years really be predicted with CAPE?

Post by SovereignInvestor » Tue May 14, 2019 2:05 pm

ohai wrote:
Tue May 14, 2019 11:45 am
I wonder how much this CAPE ratio is biased by companies like AMZN, which have essentially zero trailing earnings but nothing short of world domination priced into their future. I suspect that accounting rules might have changed as well, or companies have evolved in ways that lean towards understating earnings. On top of that, the macro environment has changed due to low interest rates. This would naturally cause valuation metrics to look more "expensive" than in the past.
Amazon is taking earnings from somewhere else. So.for AMZN there is a counterpart that had massive earnings last 10 years and will have losses in future
In universe overall of.stocks this effect offsets and it shouldn't be an issue.

Low interest certainly need to be accounted for though.

The Wizard
Posts: 13355
Joined: Tue Mar 23, 2010 1:45 pm
Location: Reading, MA

Re: Can future returns for the next 10 years really be predicted with CAPE?

Post by The Wizard » Tue May 14, 2019 2:09 pm

Living Free wrote:
Tue May 14, 2019 9:20 am
It's very hard to predict the future.
Actually, it's not.
But predicting ACCURATELY and CONSISTENTLY? Yeah that's really tough.
And the same goes for weather forecasting lately...
Attempted new signature...

Living Free
Posts: 290
Joined: Thu Jul 19, 2018 7:31 pm

Re: Can future returns for the next 10 years really be predicted with CAPE?

Post by Living Free » Tue May 14, 2019 7:18 pm

The Wizard wrote:
Tue May 14, 2019 2:09 pm
Living Free wrote:
Tue May 14, 2019 9:20 am
It's very hard to predict the future.
Actually, it's not.
But predicting ACCURATELY and CONSISTENTLY? Yeah that's really tough.
And the same goes for weather forecasting lately...
touché

User avatar
Taylor Larimore
Advisory Board
Posts: 28400
Joined: Tue Feb 27, 2007 8:09 pm
Location: Miami FL

Re: Can future returns for the next 10 years really be predicted with CAPE?

Post by Taylor Larimore » Tue May 14, 2019 7:34 pm

Living Free:

In my opinion, no one can predict stock market returns tomorrow, next month, or the next ten years.

I started the Boglehead Contest to predict the S&P 500 Index in January 2001. Of 99 Diehard guesses that year, only 11 even guessed the direction of the stock market. Boglehead forecasts were worse in 2008. Only 2 out of 284 Bogleheads guessed how low the S&P 500 Index would plunge. Of 11 professional forecasters, each one thought the S&P would gain; it declined 38%."

We get what the market gives.

Best wishes.
Taylor
"Simplicity is the master key to financial success." -- Jack Bogle

Bacchus01
Posts: 2887
Joined: Mon Dec 24, 2012 9:35 pm

Re: Can future returns for the next 10 years really be predicted with CAPE?

Post by Bacchus01 » Tue May 14, 2019 9:10 pm

SovereignInvestor wrote:
Tue May 14, 2019 9:24 am
CAPE has a bias high IMO about 30% (roughly estimated in my article below) in recent years relative to older years bevause buybacks, CPI changes in mid 1990s and lower tax rate post 2017.

It is statistical invalid to compare recent readings to long term averages when the mean is signifcantly changing. It has understated returns in the last 2-3 decades.

https://seekingalpha.com/article/408638 ... d-buybacks

Shiller even acknowledged a major issue with buybacks and added a field to the public spreadsheet but IMO this doesn't fix the issue.

CAPE says SPX should be around 1700. Really, it should be at 10x forward earnings for a yield of 10%? That's just ridiculous, bevause even if ecpnomy never grew ever and therefore earnings never grew, long run returns would be around 10%, with 10Y yield under 2.5% for a risk premium of 7.5%. The historical average risk premium of stocks over 10Y note is barely 5%.

So that means if we take CAPE mean level and say market should be there, in the most bearish scenario of no economic growth or inflation in the future (no nominal GDP growth), the S&P likely still provides a much greater return premium over bonds than historically.

Seems like CAPE is off base. It had the bias issue which makes the current readings IMO 30% biased high compared to older years all else equal (constant prospective earnings power), and then aside from thst issue, it makes no distinguishment of interest rates. If it said stocks are unfavorable levels then it suggests buying bonds, but it's all relative, yet CAPE users implicitly make absolute comparisons.
With tariffs, is it really a lower tax rate environment?

SovereignInvestor
Posts: 342
Joined: Mon Aug 20, 2018 4:41 pm

Re: Can future returns for the next 10 years really be predicted with CAPE?

Post by SovereignInvestor » Tue May 14, 2019 9:51 pm

Bacchus01 wrote:
Tue May 14, 2019 9:10 pm
SovereignInvestor wrote:
Tue May 14, 2019 9:24 am
CAPE has a bias high IMO about 30% (roughly estimated in my article below) in recent years relative to older years bevause buybacks, CPI changes in mid 1990s and lower tax rate post 2017.

It is statistical invalid to compare recent readings to long term averages when the mean is signifcantly changing. It has understated returns in the last 2-3 decades.

https://seekingalpha.com/article/408638 ... d-buybacks

Shiller even acknowledged a major issue with buybacks and added a field to the public spreadsheet but IMO this doesn't fix the issue.

CAPE says SPX should be around 1700. Really, it should be at 10x forward earnings for a yield of 10%? That's just ridiculous, bevause even if ecpnomy never grew ever and therefore earnings never grew, long run returns would be around 10%, with 10Y yield under 2.5% for a risk premium of 7.5%. The historical average risk premium of stocks over 10Y note is barely 5%.

So that means if we take CAPE mean level and say market should be there, in the most bearish scenario of no economic growth or inflation in the future (no nominal GDP growth), the S&P likely still provides a much greater return premium over bonds than historically.

Seems like CAPE is off base. It had the bias issue which makes the current readings IMO 30% biased high compared to older years all else equal (constant prospective earnings power), and then aside from thst issue, it makes no distinguishment of interest rates. If it said stocks are unfavorable levels then it suggests buying bonds, but it's all relative, yet CAPE users implicitly make absolute comparisons.
With tariffs, is it really a lower tax rate environment?
Corporate income taxes.

The current CAPE averages 9 of the 10 older years that have a 35% tax rate. If those same earnings occurred in the present they'd be about 10% greater.

Bacchus01
Posts: 2887
Joined: Mon Dec 24, 2012 9:35 pm

Re: Can future returns for the next 10 years really be predicted with CAPE?

Post by Bacchus01 » Wed May 15, 2019 5:28 am

SovereignInvestor wrote:
Tue May 14, 2019 9:51 pm
Bacchus01 wrote:
Tue May 14, 2019 9:10 pm
SovereignInvestor wrote:
Tue May 14, 2019 9:24 am
CAPE has a bias high IMO about 30% (roughly estimated in my article below) in recent years relative to older years bevause buybacks, CPI changes in mid 1990s and lower tax rate post 2017.

It is statistical invalid to compare recent readings to long term averages when the mean is signifcantly changing. It has understated returns in the last 2-3 decades.

https://seekingalpha.com/article/408638 ... d-buybacks

Shiller even acknowledged a major issue with buybacks and added a field to the public spreadsheet but IMO this doesn't fix the issue.

CAPE says SPX should be around 1700. Really, it should be at 10x forward earnings for a yield of 10%? That's just ridiculous, bevause even if ecpnomy never grew ever and therefore earnings never grew, long run returns would be around 10%, with 10Y yield under 2.5% for a risk premium of 7.5%. The historical average risk premium of stocks over 10Y note is barely 5%.

So that means if we take CAPE mean level and say market should be there, in the most bearish scenario of no economic growth or inflation in the future (no nominal GDP growth), the S&P likely still provides a much greater return premium over bonds than historically.

Seems like CAPE is off base. It had the bias issue which makes the current readings IMO 30% biased high compared to older years all else equal (constant prospective earnings power), and then aside from thst issue, it makes no distinguishment of interest rates. If it said stocks are unfavorable levels then it suggests buying bonds, but it's all relative, yet CAPE users implicitly make absolute comparisons.
With tariffs, is it really a lower tax rate environment?
Corporate income taxes.

The current CAPE averages 9 of the 10 older years that have a 35% tax rate. If those same earnings occurred in the present they'd be about 10% greater.
I understand that. My point is that the lower corporate income tax rates have been replaced with very high tariffs. As a specific case in point, my $7B megacorp has had the lower corporate rate completely wiped out with the tariff impact. Net income is almost identical on similar revenue as a result.

3funder
Posts: 1007
Joined: Sun Oct 15, 2017 9:35 pm

Re: Can future returns for the next 10 years really be predicted with CAPE?

Post by 3funder » Wed May 15, 2019 5:31 am

No one person or tool can truly predict the future. That said, I don't think CAPE is entirely useless.

SovereignInvestor
Posts: 342
Joined: Mon Aug 20, 2018 4:41 pm

Re: Can future returns for the next 10 years really be predicted with CAPE?

Post by SovereignInvestor » Wed May 15, 2019 6:55 am

Bacchus01 wrote:
Wed May 15, 2019 5:28 am
SovereignInvestor wrote:
Tue May 14, 2019 9:51 pm
Bacchus01 wrote:
Tue May 14, 2019 9:10 pm
SovereignInvestor wrote:
Tue May 14, 2019 9:24 am
CAPE has a bias high IMO about 30% (roughly estimated in my article below) in recent years relative to older years bevause buybacks, CPI changes in mid 1990s and lower tax rate post 2017.

It is statistical invalid to compare recent readings to long term averages when the mean is signifcantly changing. It has understated returns in the last 2-3 decades.

https://seekingalpha.com/article/408638 ... d-buybacks

Shiller even acknowledged a major issue with buybacks and added a field to the public spreadsheet but IMO this doesn't fix the issue.

CAPE says SPX should be around 1700. Really, it should be at 10x forward earnings for a yield of 10%? That's just ridiculous, bevause even if ecpnomy never grew ever and therefore earnings never grew, long run returns would be around 10%, with 10Y yield under 2.5% for a risk premium of 7.5%. The historical average risk premium of stocks over 10Y note is barely 5%.

So that means if we take CAPE mean level and say market should be there, in the most bearish scenario of no economic growth or inflation in the future (no nominal GDP growth), the S&P likely still provides a much greater return premium over bonds than historically.

Seems like CAPE is off base. It had the bias issue which makes the current readings IMO 30% biased high compared to older years all else equal (constant prospective earnings power), and then aside from thst issue, it makes no distinguishment of interest rates. If it said stocks are unfavorable levels then it suggests buying bonds, but it's all relative, yet CAPE users implicitly make absolute comparisons.
With tariffs, is it really a lower tax rate environment?
Corporate income taxes.

The current CAPE averages 9 of the 10 older years that have a 35% tax rate. If those same earnings occurred in the present they'd be about 10% greater.
I understand that. My point is that the lower corporate income tax rates have been replaced with very high tariffs. As a specific case in point, my $7B megacorp has had the lower corporate rate completely wiped out with the tariff impact. Net income is almost identical on similar revenue as a result.
Yeah in theory the tariffs would understate earnings. But IMO market cared about near term forward EPS which would reflect tariffs, so market prices it in.

That said based on analyst estimates it doesn't have nearly the overall.impact as tax cuts. Certain companies will get hammered but many other ones especially ones solely domestic may see no impact.

It may shave like 2-3% off EPS which is tiny conpared tp 10% or so tax cut boost. Some stocks may see 10% like maybe yours as you describe but overall it's not as severe in magnitude the other way as tax cuts. Maybe if tariffs get to 50% or 100% it will be.

User avatar
CyclingDuo
Posts: 2358
Joined: Fri Jan 06, 2017 9:07 am

Re: Can future returns for the next 10 years really be predicted with CAPE?

Post by CyclingDuo » Wed May 15, 2019 7:11 am

ohai wrote:
Tue May 14, 2019 11:45 am
I wonder how much this CAPE ratio is biased by companies like AMZN, which have essentially zero trailing earnings but nothing short of world domination priced into their future. I suspect that accounting rules might have changed as well, or companies have evolved in ways that lean towards understating earnings. On top of that, the macro environment has changed due to low interest rates. This would naturally cause valuation metrics to look more "expensive" than in the past.
Although not as much as the top cash piles in tech yet (the top 5 are Apple, Microsoft, Google, Oracle and Cisco), cash on the balance sheet is impressive for AMZN and follows a general trend we are seeing on corporate balance sheets...

Amazon cash on hand for the quarter ending March 31, 2019 was $37.020B, a 48.3% increase year-over-year.

Cash growth on corporate balance sheets from 1945 up to the end of 2017...

Image

I believe end of 2018 showed $2.72T in cash on corporate balance sheets.
"Everywhere is within walking distance if you have the time." ~ Steven Wright

Bacchus01
Posts: 2887
Joined: Mon Dec 24, 2012 9:35 pm

Re: Can future returns for the next 10 years really be predicted with CAPE?

Post by Bacchus01 » Wed May 15, 2019 7:26 am

SovereignInvestor wrote:
Wed May 15, 2019 6:55 am
Bacchus01 wrote:
Wed May 15, 2019 5:28 am
SovereignInvestor wrote:
Tue May 14, 2019 9:51 pm
Bacchus01 wrote:
Tue May 14, 2019 9:10 pm
SovereignInvestor wrote:
Tue May 14, 2019 9:24 am
CAPE has a bias high IMO about 30% (roughly estimated in my article below) in recent years relative to older years bevause buybacks, CPI changes in mid 1990s and lower tax rate post 2017.

It is statistical invalid to compare recent readings to long term averages when the mean is signifcantly changing. It has understated returns in the last 2-3 decades.

https://seekingalpha.com/article/408638 ... d-buybacks

Shiller even acknowledged a major issue with buybacks and added a field to the public spreadsheet but IMO this doesn't fix the issue.

CAPE says SPX should be around 1700. Really, it should be at 10x forward earnings for a yield of 10%? That's just ridiculous, bevause even if ecpnomy never grew ever and therefore earnings never grew, long run returns would be around 10%, with 10Y yield under 2.5% for a risk premium of 7.5%. The historical average risk premium of stocks over 10Y note is barely 5%.

So that means if we take CAPE mean level and say market should be there, in the most bearish scenario of no economic growth or inflation in the future (no nominal GDP growth), the S&P likely still provides a much greater return premium over bonds than historically.

Seems like CAPE is off base. It had the bias issue which makes the current readings IMO 30% biased high compared to older years all else equal (constant prospective earnings power), and then aside from thst issue, it makes no distinguishment of interest rates. If it said stocks are unfavorable levels then it suggests buying bonds, but it's all relative, yet CAPE users implicitly make absolute comparisons.
With tariffs, is it really a lower tax rate environment?
Corporate income taxes.

The current CAPE averages 9 of the 10 older years that have a 35% tax rate. If those same earnings occurred in the present they'd be about 10% greater.
I understand that. My point is that the lower corporate income tax rates have been replaced with very high tariffs. As a specific case in point, my $7B megacorp has had the lower corporate rate completely wiped out with the tariff impact. Net income is almost identical on similar revenue as a result.
Yeah in theory the tariffs would understate earnings. But IMO market cared about near term forward EPS which would reflect tariffs, so market prices it in.

That said based on analyst estimates it doesn't have nearly the overall.impact as tax cuts. Certain companies will get hammered but many other ones especially ones solely domestic may see no impact.

It may shave like 2-3% off EPS which is tiny conpared tp 10% or so tax cut boost. Some stocks may see 10% like maybe yours as you describe but overall it's not as severe in magnitude the other way as tax cuts. Maybe if tariffs get to 50% or 100% it will be.
I see your point, and it is certainly very negligible impact on FANG and other service/financial industries (tariffs I mean). It certainly has the big impact on consumer goods and manufacturing, which is my reference point. The tax/tariff is very good for hospitality and resort, for example. Hmmmm.

Bacchus01
Posts: 2887
Joined: Mon Dec 24, 2012 9:35 pm

Re: Can future returns for the next 10 years really be predicted with CAPE?

Post by Bacchus01 » Wed May 15, 2019 7:28 am

CyclingDuo wrote:
Wed May 15, 2019 7:11 am
ohai wrote:
Tue May 14, 2019 11:45 am
I wonder how much this CAPE ratio is biased by companies like AMZN, which have essentially zero trailing earnings but nothing short of world domination priced into their future. I suspect that accounting rules might have changed as well, or companies have evolved in ways that lean towards understating earnings. On top of that, the macro environment has changed due to low interest rates. This would naturally cause valuation metrics to look more "expensive" than in the past.
Although not as much as the top cash piles in tech yet (the top 5 are Apple, Microsoft, Google, Oracle and Cisco), cash on the balance sheet is impressive for AMZN and follows a general trend we are seeing on corporate balance sheets...

Amazon cash on hand for the quarter ending March 31, 2019 was $37.020B, a 48.3% increase year-over-year.

Cash growth on corporate balance sheets from 1945 up to the end of 2017...

Image

I believe end of 2018 showed $2.72T in cash on corporate balance sheets.
I remember having cash discussions in 2005. Cash on the balance sheet was seen as a negative by many. It meant you couldn’t deploy for high returns so you were not seen as a well run company. High cash also was target for takeover as the PEs would leverage up and pull the cash out of the deal.

Now it seems like cash is just piling up with nowhere to go!

azanon
Posts: 2463
Joined: Mon Nov 07, 2011 10:34 am
Location: Little Rock, AR
Contact:

Re: Can future returns for the next 10 years really be predicted with CAPE?

Post by azanon » Wed May 15, 2019 7:41 am

It has less to do with prediction and more to do with probabilities. The value of using CAPE is probably the equivalent of being able to bet on the favorite horse at a racetrack (let's say a 2-1 horse), but getting the composite payout of all of the horses (let's say average 8-1) if you win. Point being, can the longer shot (read: overvalued US stock) still win? Of course, but it's not as probable.

Said another way, the actual results doesn't necessarily mean that CAPE was flawed. What it means is that occasionally a long shot wins the race. The US won the past 5 years, and Country House won the Kentucky Derby. That's probabilities for you....

bizkitgto
Posts: 48
Joined: Tue Mar 14, 2017 11:54 am

Re: Can future returns for the next 10 years really be predicted with CAPE?

Post by bizkitgto » Wed May 15, 2019 12:07 pm

Wasn't the Shiller PE high for the last 10 years do to the financial crisis? I would expect it to start dropping - any thoughts? Europe's low CAPE is interesting to me, historically Europe and US stocks track, so it's reasonable to expect Europe to outperform the next 10 years. Japan on the other hand, well, does CAPE even apply to them?
Keep it simple: 20% BND, 50% VTI and 30% VXUS

User avatar
HomerJ
Posts: 13034
Joined: Fri Jun 06, 2008 12:50 pm

Re: Can future returns for the next 10 years really be predicted with CAPE?

Post by HomerJ » Wed May 15, 2019 12:13 pm

azanon wrote:
Wed May 15, 2019 7:41 am
Said another way, the actual results doesn't necessarily mean that CAPE was flawed. What it means is that occasionally a long shot wins the race. The US won the past 5 years, and Country House won the Kentucky Derby. That's probabilities for you....
Sure, it's possible we just got lucky. That's absolutely true.

But it's certainly reasonable, after 27 years, to wonder if maybe, just maybe, the model has a flaw.
The J stands for Jay

Topic Author
YRT70
Posts: 193
Joined: Sat Apr 27, 2019 8:51 am

Re: Can future returns for the next 10 years really be predicted with CAPE?

Post by YRT70 » Wed May 15, 2019 2:03 pm

Thanks for the interesting discussion so far. I've been doing more reading on the topic and I was surprised that Michael Kitces found that dynamic asset allocation based on CAPE managed to outperform other asset allocations. This seems to support the use of CAPE as a prediction.

What do you guys think of it?

Image

The solid blue line is the dynamic asset allocation with bills.
Accordingly, a “valuation-based” tactical asset allocation approach was created, where the portfolio starts out at a “neutral” 45% in equity exposure, increasing to 60% in equities when markets are “cheap” and undervalued, decreasing to only 30% when markets are “expensive” and overvalued, and remaining at the neutral 45% in equities when markets are fairly valued in the middle (where the thresholds again are defined using the Graham and Dodd 2/3rds and 4/3rds thresholds). The chart below shows where these thresholds would have fallen relative to Shiller CAPE (P/E10) historically, and the asset allocation changes that would have occurred over the decades.


https://www.kitces.com/blog/valuation-b ... tGLsFsVBoE

azanon
Posts: 2463
Joined: Mon Nov 07, 2011 10:34 am
Location: Little Rock, AR
Contact:

Re: Can future returns for the next 10 years really be predicted with CAPE?

Post by azanon » Wed May 15, 2019 3:12 pm

HomerJ wrote:
Wed May 15, 2019 12:13 pm
azanon wrote:
Wed May 15, 2019 7:41 am
Said another way, the actual results doesn't necessarily mean that CAPE was flawed. What it means is that occasionally a long shot wins the race. The US won the past 5 years, and Country House won the Kentucky Derby. That's probabilities for you....
Sure, it's possible we just got lucky. That's absolutely true.

But it's certainly reasonable, after 27 years, to wonder if maybe, just maybe, the model has a flaw.
Why would anyone wonder that? Mebane Faber did a study of a couple dozen countries or so performance vs. CAPE (source: "Global Value" by Mebane Faber), and the US was one of the very few outliers not to follow the predictive power of CAPE. The vast majority performed better when their CAPE was lower, and vise versa. So, whew, so far it's reliable from a probability standpoint. But as I already described, selecting a more probable "horse" is no guarantee you're going to win.

User avatar
HomerJ
Posts: 13034
Joined: Fri Jun 06, 2008 12:50 pm

Re: Can future returns for the next 10 years really be predicted with CAPE?

Post by HomerJ » Wed May 15, 2019 5:01 pm

azanon wrote:
Wed May 15, 2019 3:12 pm
HomerJ wrote:
Wed May 15, 2019 12:13 pm
azanon wrote:
Wed May 15, 2019 7:41 am
Said another way, the actual results doesn't necessarily mean that CAPE was flawed. What it means is that occasionally a long shot wins the race. The US won the past 5 years, and Country House won the Kentucky Derby. That's probabilities for you....
Sure, it's possible we just got lucky. That's absolutely true.

But it's certainly reasonable, after 27 years, to wonder if maybe, just maybe, the model has a flaw.
Why would anyone wonder that? Mebane Faber did a study of a couple dozen countries or so performance vs. CAPE (source: "Global Value" by Mebane Faber), and the US was one of the very few outliers not to follow the predictive power of CAPE. The vast majority performed better when their CAPE was lower, and vise versa. So, whew, so far it's reliable from a probability standpoint. But as I already described, selecting a more probable "horse" is no guarantee you're going to win.
Why would anyone wonder that? For 27 years, almost from the very moment it was discovered, CAPE has failed to predict returns in the U.S. and no one should wonder why but instead make investment decisions going forward based on CAPE?

The full title of that book, by the way, is Global Value: How to Spot Bubbles, Avoid Market Crashes, and Earn Big Returns in the Stock Market. It came out in 2014, and so far it's failed to deliver on the "Earn Big Returns" part because anyone who followed that book and invested heavily in International and less in the U.S. has not done well.

Not done well YET, I will admit. Faber may indeed still proven to be correct. I'm not saying he is for sure wrong. The theory may be right, and one may indeed earn "Big Returns" going forward, investing based on CAPE values by country. But maybe not. It hasn't worked yet since he wrote the book. But corrections and booms happen quickly, and maybe it will all turn around going forward.

So I'm not saying he's wrong.

But he sure hasn't been proven right yet about international CAPE.

Of course it's reasonable to question the theory.

If one makes a prediction in 2014 like "International is far more likely to do well going forward, and the U.S. is far more likely to have very low returns" and then the exact opposite happens, one doesn't get to say "It's not reasonable to question this!"

One CAN say "Oh, it's all probabilities, and it's only been 5 years, and this reason and that reason." All legit. One can defend the theory with solid reasons. I'm NOT saying it's been proven wrong. I absolutely concede it could be right.

But one CAN'T say the theory is obviously correct. One can't say results have matched the theory.

And we see that all the time here... "CAPE predictions has pretty good results". But it hasn't. Not in the U.S. and not for nearly 3 decades. That doesn't mean it's wrong. But the results have not been good. One doesn't get to say they have been good.

One can explain the results. Fairly. One can bring up all the other variables involved and explain WHY results haven't matched. That's legit and correct. But that just proves that there are more variables involved and predicting returns on CAPE alone is not enough.
Last edited by HomerJ on Wed May 15, 2019 11:00 pm, edited 2 times in total.
The J stands for Jay

Thesaints
Posts: 2827
Joined: Tue Jun 20, 2017 12:25 am

Re: Can future returns for the next 10 years really be predicted with CAPE?

Post by Thesaints » Wed May 15, 2019 5:12 pm

A low CAPE can normalize in two ways. One will make you money and the other will do the opposite.

azanon
Posts: 2463
Joined: Mon Nov 07, 2011 10:34 am
Location: Little Rock, AR
Contact:

Re: Can future returns for the next 10 years really be predicted with CAPE?

Post by azanon » Thu May 16, 2019 7:23 am

HomerJ wrote:
Wed May 15, 2019 5:01 pm
azanon wrote:
Wed May 15, 2019 3:12 pm
HomerJ wrote:
Wed May 15, 2019 12:13 pm
azanon wrote:
Wed May 15, 2019 7:41 am
Said another way, the actual results doesn't necessarily mean that CAPE was flawed. What it means is that occasionally a long shot wins the race. The US won the past 5 years, and Country House won the Kentucky Derby. That's probabilities for you....
Sure, it's possible we just got lucky. That's absolutely true.

But it's certainly reasonable, after 27 years, to wonder if maybe, just maybe, the model has a flaw.
Why would anyone wonder that? Mebane Faber did a study of a couple dozen countries or so performance vs. CAPE (source: "Global Value" by Mebane Faber), and the US was one of the very few outliers not to follow the predictive power of CAPE. The vast majority performed better when their CAPE was lower, and vise versa. So, whew, so far it's reliable from a probability standpoint. But as I already described, selecting a more probable "horse" is no guarantee you're going to win.
Why would anyone wonder that? For 27 years, almost from the very moment it was discovered, CAPE has failed to predict returns in the U.S. and no one should wonder why but instead make investment decisions going forward based on CAPE?

The full title of that book, by the way, is Global Value: How to Spot Bubbles, Avoid Market Crashes, and Earn Big Returns in the Stock Market. It came out in 2014, and so far it's failed to deliver on the "Earn Big Returns" part because anyone who followed that book and invested heavily in International and less in the U.S. has not done well.

Not done well YET, I will admit. Faber may indeed still proven to be correct. I'm not saying he is for sure wrong. The theory may be right, and one may indeed earn "Big Returns" going forward, investing based on CAPE values by country. But maybe not. It hasn't worked yet since he wrote the book. But corrections and booms happen quickly, and maybe it will all turn around going forward.

So I'm not saying he's wrong.

But he sure hasn't been proven right yet about international CAPE.

Of course it's reasonable to question the theory.

If one makes a prediction in 2014 like "International is far more likely to do well going forward, and the U.S. is far more likely to have very low returns" and then the exact opposite happens, one doesn't get to say "It's not reasonable to question this!"

One CAN say "Oh, it's all probabilities, and it's only been 5 years, and this reason and that reason." All legit. One can defend the theory with solid reasons. I'm NOT saying it's been proven wrong. I absolutely concede it could be right.

But one CAN'T say the theory is obviously correct. One can't say results have matched the theory.

And we see that all the time here... "CAPE predictions has pretty good results". But it hasn't. Not in the U.S. and not for nearly 3 decades. That doesn't mean it's wrong. But the results have not been good. One doesn't get to say they have been good.

One can explain the results. Fairly. One can bring up all the other variables involved and explain WHY results haven't matched. That's legit and correct. But that just proves that there are more variables involved and predicting returns on CAPE alone is not enough.
> You still keep fixated on the US Market, when he researched 40 or so countries. I think I understand your frustration, so correct me if I'm wrong - you want CAPE to work specifically for the US market, and not just most countries in aggregate? I actually read the book, and what I recall from it was not any promise that any one country would do well based on CAPE, but in aggregate they could be expected to over time. He has specifically said that's why GVAL* has at least 10 countries included and not just one.

> So for the past, CAPE was predictive for all countries in aggregate (that's what's in the book), and nothing has changed since the publication of that book to change that conclusion - 5 years is noise and randomness compared to the timeframe studied. You're conclusion that it has failed to deliver, suggests to me that you have failed to even understand how CAPE works. What portion of the book gave you the impression that this method of stock selection works on a year-to-year basis, and again i'll add, always works for the US? It never claimed either of those things, I personally don't expect it to, but you seem to want to add this extra burden on Meb that he never claimed.

> Let me help your flawed argument; You don't need to cite the US's performance defying CAPE. Why not use Japan's 20 years of outperformance many years ago? Because you know how that ended? :D

The past 5 years is not unlike the late 90s - growth is king, and value is less than meaningless. The existence of a late 90s was not enough to change the overall observed outperformance of value. Don't take it from me - I'm just a Biologist - ask Larry S.

I'm glad that you managed to find the book despite my intentional shortening of the title.

* the ETF GVAL has several other flaws that I'm not going into that would cause me to not recommend it. I endorse quite a bit of Meb's work, but I personally think he should stop as a researcher and not go into the securities business. Some of them are discussed at www.etf.com, but I have other concerns as well....

Stormbringer
Posts: 808
Joined: Sun Jun 14, 2015 7:07 am

Re: Can future returns for the next 10 years really be predicted with CAPE?

Post by Stormbringer » Thu May 16, 2019 8:01 am

My hypothesis as to why CAPE has underestimated returns comes down to one thing: interest rates.

Rates have been falling for decades, and especially since 2008-2009. That has put sustained upward pressure on asset prices that CAPE does not adjust for.
"Compound interest is the most powerful force in the universe." - Albert Einstein

SovereignInvestor
Posts: 342
Joined: Mon Aug 20, 2018 4:41 pm

Re: Can future returns for the next 10 years really be predicted with CAPE?

Post by SovereignInvestor » Thu May 16, 2019 9:47 am

:sharebeer
Stormbringer wrote:
Thu May 16, 2019 8:01 am
My hypothesis as to why CAPE has underestimated returns comes down to one thing: interest rates.

Rates have been falling for decades, and especially since 2008-2009. That has put sustained upward pressure on asset prices that CAPE does not adjust for.
Rates have a big factor. But buybacks and CPI also are biasing the metric.

All else equal. Ignoring everything else...CAPE will be greatER if there are more buybacks in a 10 year period thsn if there are not since it creates larger gap between inflation adjusted older year EPS and current EPS. Since buybacks allow EPS to grow faster than inflstion.

The low rates make what is a FV PE multiple rise. But the CAPE itself as a measuring stick is biased high when buybacks are present. The 2000s saw it underestimate returns even if rates didn't drop because it is not on-leveling oldER year earnings to present share level.

The comparison of recent post 2000 CAPE which have massive buybacks to pre 2000 ones is biased.

Then it is made worse by the actual mean CAPE needing to be adusted for rates.

User avatar
HomerJ
Posts: 13034
Joined: Fri Jun 06, 2008 12:50 pm

Re: Can future returns for the next 10 years really be predicted with CAPE?

Post by HomerJ » Thu May 16, 2019 10:41 am

Stormbringer wrote:
Thu May 16, 2019 8:01 am
My hypothesis as to why CAPE has underestimated returns comes down to one thing: interest rates.

Rates have been falling for decades, and especially since 2008-2009. That has put sustained upward pressure on asset prices that CAPE does not adjust for.
There's all kind of good reasons. Legitimate solid reasons. Reasons I agree with and do not contest.

My issue is when people won't even admit that CAPE has underestimated returns in the U.S. People still actually say "CAPE has been a strong predictor of returns."

Mebane's Faber's book blurb says this.
Over 70 years ago, Benjamin Graham and David Dodd proposed valuing stocks with earnings smoothed across multiple years. Robert Shiller later popularized this method with his version of the cyclically adjusted price-to-earnings (CAPE) ratio in the late 1990s and correctly issued a timely warning of poor stock returns to follow in the coming years.
This is misleading.

Shiller predicted on July 21, 1996 that we would likely see 0% real returns over the next 10 years. This was the correct call based on the previous 100 years data. CAPE had hit 25 in 1996, which was ridiculously high at the time (Now CAPE of 25 is just considered "moderately elevated" - CAPE proponents keep changing the model, which is fine with new data, but they pretend today's model was the same back in 1996)

July 1996-2006 returns were actually 8.5% nominal (so 6.5% real?), very close to historical norms.

Even looking at periods ending in bear market years

July 1996-2002 returns were 6% nominal (so 4% real)
July 1996-2009 returns were 5% nominal (so 3% real)

Now, you can make a case that the last two are within the margin of error, but even the worst ending points weren't that terrible. All other ending points have higher returns.

July 1996-2016 returns were back to 8.5% nominal.

1996 had the second highest valuations in U.S. history. CAPE hadn't been that high for 70 years, not since the Great Depression.

And it was a pretty good time to buy stocks. At no point did a 1996 stock investor ever lose money, not even temporarily.

I find it hard to believe that someone can say that 0% real return call in 1996 was "correct and timely". It's one of the calls I point out to show how CAPE predictions haven't worked well. And here's Faber using it to prove that CAPE predictions work.

I'm flabbergasted.

Edit: I downloaded the book (it's a pretty decent little book - only 82 pages), and I'm reading it at lunch.

Turns out he's talking about Shiller's predictions in 1998 and 1999, not the original 1996 prediction.

Funny enough, Faber quotes Fama in the introduction as a counterpoint to himself.
I think most bubbles are twenty-twenty hindsight. Now after the fact you always find people who said before the fact that prices are too high. People are always saying prices are too high. When they turn out to be right, we anoint them. When they turn out to be wrong, we ignore them.
I feel like this is exactly what Faber has done. He's anointed Shiller for his 1999 predictions, but ignored his 1996 predictions.

Book isn't bad.. it does makes the mistake of looking at the past from today's point of view. For instance, 25 CAPE isn't consider super high anymore.

But it was considered super high at the time. And that's important.
Last edited by HomerJ on Thu May 16, 2019 11:38 am, edited 1 time in total.
The J stands for Jay

MnD
Posts: 4287
Joined: Mon Jan 14, 2008 12:41 pm

Re: Can future returns for the next 10 years really be predicted with CAPE?

Post by MnD » Thu May 16, 2019 10:53 am

Global market cap has a current PE of 15.
Another reason to buy and hold the haystack versus hand-wringing about single country rear-view mirror metrics.

Topic Author
YRT70
Posts: 193
Joined: Sat Apr 27, 2019 8:51 am

Re: Can future returns for the next 10 years really be predicted with CAPE?

Post by YRT70 » Thu May 16, 2019 10:54 am

I'm relatively new to this investing world so bear with me. If someone would have followed the CAPE based dynamic asset allocation strategy as proposed by Kitces & Pfau, they would have missed out on a lot of the growth in the last 5 years right?

So anyone who retired in 2014 and thought "oh CAPE is pretty high, I should better reduce my equity allocation" would have been wrong on hind sight, so far. I realise the next 5 years might be a different story of course.

Image

Post Reply