New and Improved CAPE is Higher!

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
Topic Author
grayfox
Posts: 5569
Joined: Sat Sep 15, 2007 4:30 am

New and Improved CAPE is Higher!

Post by grayfox »

Anyone see this? I was over at ONLINE DATA ROBERT SHILLER which is where you can download the stock market data that has all the CAPE stuff. I see that, starting in Sept-2018 he came up with an alternative version of CAPE. The alternate version accounts for changes in corporate payout policy, which I guess means share buybacks.

I was expecting that this would have the effect of reducing the calculated value of CAPE. So I downloaded the new version, and it shows the new CAPE is actually higher than the old CAPE! :shock:

For example, 2018.09.
Old-style CAPE is 33.18.
New-style CAPE is 36.15
User avatar
jeffyscott
Posts: 9150
Joined: Tue Feb 27, 2007 9:12 am
Location: Wisconsin

Re: New and Improved CAPE is Higher!

Post by jeffyscott »

But the historical CAPE becomes higher too. And it looks like there is a greater difference in the older data, which would effectively mean the current CAPE is lower than we think:

Image
The two greatest enemies of the equity fund investor are expenses and emotions. ― John C. Bogle
cjking
Posts: 1933
Joined: Mon Jun 30, 2008 4:30 am

Re: New and Improved CAPE is Higher!

Post by cjking »

I did notice this recently. I haven't quite got my head around the difference, other than that "price" in the new method is the price of a total-return index, rather than one where prices are reduced as a consequence of dividends having been paid. I suppose it makes sense that if you have a higher price and the same earnings, the PE ratio is going to be higher. (What I can't quite understand is the problem it is solving, but I'm happy to accept the new method is better. The two results are close enough that it doesn't make much practical difference.)
cjking
Posts: 1933
Joined: Mon Jun 30, 2008 4:30 am

Re: New and Improved CAPE is Higher!

Post by cjking »

jeffyscott wrote: Thu Feb 14, 2019 9:08 am But the historical CAPE becomes higher too. And it looks like there is a greater difference in the older data, which would effectively mean the current CAPE is lower than we think:
Though many CAPE users seem to think it's a good idea, I don't compare CAPE with it's history, as then you are implicitly assuming there's some reason why things should mean-revert. I prefer to just use it as an absolute measure, 1/36.15 implies 2.8% expected real return.
cjking
Posts: 1933
Joined: Mon Jun 30, 2008 4:30 am

Re: New and Improved CAPE is Higher!

Post by cjking »

This 2016 paper from Star capital references a 2014 paper by Shiller, where he apparently changed to using total return.

https://www.starcapital.de/fileadmin/us ... imling.pdf
2.1. CAPE criticism I: payout ratios
In the period 1881-1950, S&P 500 companies distributed
65.6% of their earnings in the form of dividends. Since
1990, it has been only 39.4%. The declining payout ratio
gives companies greater scope for investments and share
buybacks, which could increase EPS growth. Indeed,
corporate profits have grown by 2.7% annually since
1990, much more than the 1.0% from 1881 to 1950
(Figure 2).
This is not without consequences for the comparability
of CAPE: CAPE evaluates an equity market on the basis
of its average earnings during the previous 10 years. The
stronger the permanent earnings growth, the further the
current level of earnings moves away from the average,
which would lead to higher fair CAPE levels. As such, the
higher CAPE that we have witnessed since 1990 could
be partially explained by a modified dividend policy.
Therefore, Shiller and Bunn [2014] propose an adjusted
CAPE to take into account the modified payout ratios.
The authors calculate CAPE on the basis of (theoretical)
total return EPS, which presumes a payout ratio of 0%
— that is total share buybacks. Whether this adjustment
does indeed strengthen the position of CAPE in the
S&P 500 remains to be seen. Furthermore, the question
remains as to whether the adjustment increases the
comparability of CAPE among countries with different
payout ratios.
User avatar
jeffyscott
Posts: 9150
Joined: Tue Feb 27, 2007 9:12 am
Location: Wisconsin

Re: New and Improved CAPE is Higher!

Post by jeffyscott »

cjking wrote: Thu Feb 14, 2019 9:32 am
jeffyscott wrote: Thu Feb 14, 2019 9:08 am But the historical CAPE becomes higher too. And it looks like there is a greater difference in the older data, which would effectively mean the current CAPE is lower than we think:
Though many CAPE users seem to think it's a good idea, I don't compare CAPE with it's history, as then you are implicitly assuming there's some reason why things should mean-revert. I prefer to just use it as an absolute measure, 1/36.15 implies 2.8% expected real return.
But then had Shiller not come up with his new methodology, you would have expected return of 3% real. (not that this is much of a difference)

In any case, from his data, I get an average under the traditional CAPE of 16.9 and 20.3 under newCAPE. So instead of 1.96x it'd be 1.78x.
The two greatest enemies of the equity fund investor are expenses and emotions. ― John C. Bogle
columbia
Posts: 2933
Joined: Tue Aug 27, 2013 5:30 am

Re: New and Improved CAPE is Higher!

Post by columbia »

Was this the “needed” change pointed out by folks, who said that current CAPE was not comparable to measurements from the past?
Cody
Posts: 984
Joined: Sun Dec 02, 2007 9:19 am
Location: Stillwater, Mn

Re: New and Improved CAPE is Higher!

Post by Cody »

Could you provide a bit of background on how you personally use the CAPE in make investment decisions?

Thanks,
Cody
cjking
Posts: 1933
Joined: Mon Jun 30, 2008 4:30 am

Re: New and Improved CAPE is Higher!

Post by cjking »

jeffyscott wrote: Thu Feb 14, 2019 9:51 am But then had Shiller not come up with his new methodology, you would have expected return of 3% real. (not that this is much of a difference)
Yes.

(Edit: though I actually get my data from Star Capital web site rather than Shiller, as they give data for all countries. Don't know which method they use, and don't care if they suddenly change. It's unlikely to make much difference.)

In my spreadsheet I have a list of risky securities a I might invest in, geographic equity index funds that cover the world, plus some others. With each is an expected return, and as a diversification mechanism, maximum allowed exposure. So when buying, I work my way down the table to find the highest-ranking security I'm allowed to buy. Currently I'm invested in 12 out of 28 options in the table.

I'm allowed to be 100% invested in a US equity index tracker, but as that security is third from the bottom the table, I have zero exposure to it, nor to 13 securities above it with higher expected returns.

For other geographic equity index funds, I allow myself to hold double their share of world market capitalisation. My biggest holding is Europe-excluding-UK equities, at 30%, with an expected return of 4.8%. Second biggest is emerging markets, at 19% with expected return 5.9%. These expected returns are CAPE yields reduced by fund ERs and estimates of losses to withholding taxes on dividends.

I've just checked, and a 0.2% increase in expected return would make no difference to the position of a US tracker in the table.
stoxtrader
Posts: 14
Joined: Wed Jan 16, 2019 3:42 pm

Re: New and Improved CAPE is Higher!

Post by stoxtrader »

cjking -

Your spreadsheet sounds amazing, I would love to see a generic copy if you don't mind? How do you get CAPE info for all other markets? How do you pick best exposure to said options? etc....

very cool.
2015
Posts: 2906
Joined: Mon Feb 10, 2014 2:32 pm

Re: New and Improved CAPE is Higher!

Post by 2015 »

Where I live the Indians use to think inclement weather was some kind of sign. Now we have CAPE and investors driven by none other than ego, greed, and fear will use all kinds of narratives to divine patterns. CAPE also helps to continue to monetize Shiller's career.
software
Posts: 210
Joined: Tue Apr 18, 2017 2:02 pm

Re: New and Improved CAPE is Higher!

Post by software »

cjking wrote: Thu Feb 14, 2019 10:51 am
jeffyscott wrote: Thu Feb 14, 2019 9:51 am But then had Shiller not come up with his new methodology, you would have expected return of 3% real. (not that this is much of a difference)
Yes.

(Edit: though I actually get my data from Star Capital web site rather than Shiller, as they give data for all countries. Don't know which method they use, and don't care if they suddenly change. It's unlikely to make much difference.)

In my spreadsheet I have a list of risky securities a I might invest in, geographic equity index funds that cover the world, plus some others. With each is an expected return, and as a diversification mechanism, maximum allowed exposure. So when buying, I work my way down the table to find the highest-ranking security I'm allowed to buy. Currently I'm invested in 12 out of 28 options in the table.

I'm allowed to be 100% invested in a US equity index tracker, but as that security is third from the bottom the table, I have zero exposure to it, nor to 13 securities above it with higher expected returns.

For other geographic equity index funds, I allow myself to hold double their share of world market capitalisation. My biggest holding is Europe-excluding-UK equities, at 30%, with an expected return of 4.8%. Second biggest is emerging markets, at 19% with expected return 5.9%. These expected returns are CAPE yields reduced by fund ERs and estimates of losses to withholding taxes on dividends.

I've just checked, and a 0.2% increase in expected return would make no difference to the position of a US tracker in the table.
I mean...kudos to you for putting your money where your mouth is but it seems that you are deriving your entire investment strategy over the assumption that CAPE can accurately predict returns.

Personally, that assumption would not sit well with me given that this has been debunked repeatedly.
JackoC
Posts: 1738
Joined: Sun Aug 12, 2018 11:14 am

Re: New and Improved CAPE is Higher!

Post by JackoC »

software wrote: Fri Feb 15, 2019 2:21 pm
cjking wrote: Thu Feb 14, 2019 10:51 am
jeffyscott wrote: Thu Feb 14, 2019 9:51 am But then had Shiller not come up with his new methodology, you would have expected return of 3% real. (not that this is much of a difference)
Yes.

(Edit: though I actually get my data from Star Capital web site rather than Shiller, as they give data for all countries. Don't know which method they use, and don't care if they suddenly change. It's unlikely to make much difference.)
... These expected returns are CAPE yields reduced by fund ERs and estimates of losses to withholding taxes on dividends.

I've just checked, and a 0.2% increase in expected return would make no difference to the position of a US tracker in the table.
I mean...kudos to you for putting your money where your mouth is but it seems that you are deriving your entire investment strategy over the assumption that CAPE can accurately predict returns.

Personally, that assumption would not sit well with me given that this has been debunked repeatedly.
Estimating expected return is not, NOT, *NOT* predicting future realized return. The hands down most common misconception on this forum.

Nor is relating earnings yield, 1/PE[x], like correlating whether the east or west won the NBA finals to the return on stocks that year. The earnings yield has a fundamental theoretical relationship to real expected return via the dividend discount model. Whatever measure of PE you decide is most representative, given the vagaries of accounting and business cycle, if the PE goes up, expected return goes down. Like a bond's price and yield relationship.

The 'you decide is most representative' is tricky and subject to valid debate for stocks whereas obvious for 'riskless' bonds. But the idea that earning yield might really have no relationship to *expected* return makes no sense. That would mean I'd expect the exactly the same centroid for future unknown stock returns if earnings are completely stable but investors bid up the index to 200 as I would with the same earnings prospects but the price is 100.

Again, Shiller and others recasting of past PE's, and the relatively arbitrary choice of 10 yrs to even out the earnings cycle, those are subject to debate. Hence 1/PE=E[r], but 1/what number exactly?. But the idea that generally higher PE's don't mean generally lower expected returns is, sadly since we live in an age of generally high PE, groundless. We're paying more now in general for a $ of earnings. That means lower expected returns than in the past. Though that does not necessarily mean a lower future realized return. PE's could continue to rise, higher than expected return. Earnings could grow from their current multi-decade high % of GDP, higher than expected return. GDP could grow faster than expected, higher than expected return. Or the opposite for any one or all those things. The centroid of the distribution, the *expected* return, is still fundamentally tied to some concept of the earnings yield.
JackoC
Posts: 1738
Joined: Sun Aug 12, 2018 11:14 am

Re: New and Improved CAPE is Higher!

Post by JackoC »

duplicate
Last edited by JackoC on Fri Feb 15, 2019 5:24 pm, edited 1 time in total.
User avatar
gmaynardkrebs
Posts: 2128
Joined: Sun Feb 10, 2008 11:48 am

Re: New and Improved CAPE is Higher!

Post by gmaynardkrebs »

software wrote: Fri Feb 15, 2019 2:21 pm
cjking wrote: Thu Feb 14, 2019 10:51 am
jeffyscott wrote: Thu Feb 14, 2019 9:51 am But then had Shiller not come up with his new methodology, you would have expected return of 3% real. (not that this is much of a difference)
Yes.

(Edit: though I actually get my data from Star Capital web site rather than Shiller, as they give data for all countries. Don't know which method they use, and don't care if they suddenly change. It's unlikely to make much difference.)

In my spreadsheet I have a list of risky securities a I might invest in, geographic equity index funds that cover the world, plus some others. With each is an expected return, and as a diversification mechanism, maximum allowed exposure. So when buying, I work my way down the table to find the highest-ranking security I'm allowed to buy. Currently I'm invested in 12 out of 28 options in the table.

I'm allowed to be 100% invested in a US equity index tracker, but as that security is third from the bottom the table, I have zero exposure to it, nor to 13 securities above it with higher expected returns.

For other geographic equity index funds, I allow myself to hold double their share of world market capitalisation. My biggest holding is Europe-excluding-UK equities, at 30%, with an expected return of 4.8%. Second biggest is emerging markets, at 19% with expected return 5.9%. These expected returns are CAPE yields reduced by fund ERs and estimates of losses to withholding taxes on dividends.

I've just checked, and a 0.2% increase in expected return would make no difference to the position of a US tracker in the table.
I mean...kudos to you for putting your money where your mouth is but it seems that you are deriving your entire investment strategy over the assumption that CAPE can accurately predict returns.

Personally, that assumption would not sit well with me given that this has been debunked repeatedly.
Debunked repeatedly by whom? The geniuses on Seeking Alpha? Some posters on Bogleheads? It is still widely respected by professional investors and academics. Virtually all of the debunking I've seen is from (1) those who are unhappy about what it says about today's valuations and the range of future returns; (2) those who debunk some mis-characterization of it, quite often their own mis-characterization. Shiller has never claimed that CAPE can "accurately" predict future returns." In particular, it is not useful as a short term "market timing" tool; However, it is useful as one tool among many for building a long term plan that suits you best. I think that's what most of us are here for.
hifeng
Posts: 12
Joined: Fri Jul 27, 2018 6:28 pm

Re: New and Improved CAPE is Higher!

Post by hifeng »

I think most people would agree that current stock market is NOT cheap (e.g. CAPE). But expensive stock does not mean there will be a sink in the near future, maybe continue to soar. So predict market is almost impossible. Maybe you can slightly adjust your allocation, or not. Stay the course because you need to stay in the market. :sharebeer
JackoC
Posts: 1738
Joined: Sun Aug 12, 2018 11:14 am

Re: New and Improved CAPE is Higher!

Post by JackoC »

hifeng wrote: Sat Feb 16, 2019 1:04 am I think most people would agree that current stock market is NOT cheap (e.g. CAPE). But expensive stock does not mean there will be a sink in the near future, maybe continue to soar. So predict market is almost impossible. Maybe you can slightly adjust your allocation, or not. Stay the course because you need to stay in the market. :sharebeer
It does not mean the market will 'sink'. It does mean the *expected return* is lower than the historical realized return, which is also clear from risk free yields (around 2%-points lower than the avg. historical realized real return on long term bonds) and credit spreads (lower risk premium there than historical average). It's not as if high cyclically adjusted PE is an outlier, it's in line with almost every other measure in stocks, and in other asset classes, saying that expected return on capital is now lower than the historical average realized return on capital.

It has no definitive implication for asset allocation*. It is not a method of determining the specific future *realized return*, which is not predictable. It might however have a significant impact on determining the amount to be saved, because the midpoint expectation of future spending ability for a given amount saved should be taken as significantly lower than is implied by looking at historical outcomes.

*It's actually under the Pollyanna view that stock expected returns have not decreased compared to historical realized returns that there would be a reason a shift allocation, toward stocks. Since the Pollyanna view contends that stock expected returns aren't lower than historical realized or 'we can't say anything about future return [which 95% of the time in practice means: let's assume stock expected return=stock historical realized return]', and since bond expected returns obviously are lower (the yields are undeniably much lower than historical realized return on bonds), that implies stocks have become more attractive relative to bonds. Under the realistic recognition that stock expected returns have dropped below historical at least as much as bonds have, there isn't necessarily any reason to shift allocation either way. Stock expected return has apparently dropped below the historical realized return by more than bonds have since the risk premium has also compressed (as also seen in relatively lower credit spreads). But in an efficient market the compressed risk premia presumably represent actually lower risk, or anyway an efficient representation of more general appetite for risk. It does not imply that stocks are less attractive relative to other investments, just that it's unrealistic in general to plan based on historical realized returns.
Jeff Albertson
Posts: 832
Joined: Sat Apr 06, 2013 7:11 pm
Location: Springfield

Re: New and Improved CAPE is Higher!

Post by Jeff Albertson »

Vanguard published a paper on improving CAPE forecasting, summary here:
https://investornews.vanguard/valuing-t ... alue-cape/
Vanguard Investment Strategy Group’s ”fair-value” CAPE approach started with an observation: A long-term decline in interest rates and inflation depresses the discount rates used in asset-pricing models. The upshot is that investors are willing to pay a higher price for future earnings, thus inflating P/E ratios. The critical implication is that there is no reason for the CAPE ratio to revert to its long-term average. Instead, it should be expected to revert to a level that reflects current economic conditions.
Joe Davis briefly discusses this with Barry Ritholtz on his current podcast (starting at about 48 minutes).
https://ritholtz.com/2019/02/mib-joe-davis/
User avatar
gmaynardkrebs
Posts: 2128
Joined: Sun Feb 10, 2008 11:48 am

Re: New and Improved CAPE is Higher!

Post by gmaynardkrebs »

Jeff Albertson wrote: Sat Feb 16, 2019 6:47 pm Vanguard published a paper on improving CAPE forecasting, summary here:
Vanguard Investment Strategy Group’s ”fair-value” CAPE approach started with an observation: A long-term decline in interest rates and inflation depresses the discount rates used in asset-pricing models. The upshot is that investors are willing to pay a higher price for future earnings, thus inflating P/E ratios. The critical implication is that there is no reason for the CAPE ratio to revert to its long-term average. Instead, it should be expected to revert to a level that reflects current economic conditions.
Just to be clear, this is still not good news for equities. If it is the low discount rate driving a higher equilibrium CAPE ratio, that implies a lower future return from equities over the long run. It just means that investors are paying more than they used to for exactly the same earnings. That's even without assuming that valuations mean revert toward historical norms -- which would be even worse.
MNS CA
Posts: 104
Joined: Tue Dec 26, 2017 11:29 pm

Re: New and Improved CAPE is Higher!

Post by MNS CA »

cjking wrote: Thu Feb 14, 2019 9:32 am
jeffyscott wrote: Thu Feb 14, 2019 9:08 am But the historical CAPE becomes higher too. And it looks like there is a greater difference in the older data, which would effectively mean the current CAPE is lower than we think:
Though many CAPE users seem to think it's a good idea, I don't compare CAPE with it's history, as then you are implicitly assuming there's some reason why things should mean-revert. I prefer to just use it as an absolute measure, 1/36.15 implies 2.8% expected real return.
I don't think that can be right. This logic of divide by 1 over CAPE to get expected returns would mean that you could never expect negative returns. So if CAPE was 1000 you'd still expect a return of 0.001% instead of losing money because you massively overpaid for earnings.

I'd rather look at something like historical returns based on past values of CAPE, possibly with some adjustment for interest rates and inflation.
MoneyMarathon
Posts: 992
Joined: Sun Sep 30, 2012 3:38 am

Re: New and Improved CAPE is Higher!

Post by MoneyMarathon »

JackoC wrote: Sat Feb 16, 2019 10:21 amIt's actually under the Pollyanna view that stock expected returns have not decreased compared to historical realized returns that there would be a reason a shift allocation, toward stocks. Since the Pollyanna view contends that stock expected returns aren't lower than historical realized or 'we can't say anything about future return [which 95% of the time in practice means: let's assume stock expected return=stock historical realized return]', and since bond expected returns obviously are lower (the yields are undeniably much lower than historical realized return on bonds), that implies stocks have become more attractive relative to bonds.
I love it. Pollyanna certainly has a remarkable habit of being right in the long run. :wink:
kosomoto
Posts: 500
Joined: Tue Nov 24, 2015 8:51 pm

Re: New and Improved CAPE is Higher!

Post by kosomoto »

cjking wrote: Thu Feb 14, 2019 9:32 am
jeffyscott wrote: Thu Feb 14, 2019 9:08 am But the historical CAPE becomes higher too. And it looks like there is a greater difference in the older data, which would effectively mean the current CAPE is lower than we think:
Though many CAPE users seem to think it's a good idea, I don't compare CAPE with it's history, as then you are implicitly assuming there's some reason why things should mean-revert. I prefer to just use it as an absolute measure, 1/36.15 implies 2.8% expected real return.
How on Earth can the expected real rate of return of US equities be 2.8% when the earnings yield over the last 12 months is 5%?! Are you expecting earnings to not grow at all in the future? Not to mention earnings have significantly more room to grow as US companies enter emerging markets..

This cape nonsense is nonsense.
User avatar
gmaynardkrebs
Posts: 2128
Joined: Sun Feb 10, 2008 11:48 am

Re: New and Improved CAPE is Higher!

Post by gmaynardkrebs »

kosomoto wrote: Tue Aug 06, 2019 4:11 pm
cjking wrote: Thu Feb 14, 2019 9:32 am
jeffyscott wrote: Thu Feb 14, 2019 9:08 am But the historical CAPE becomes higher too. And it looks like there is a greater difference in the older data, which would effectively mean the current CAPE is lower than we think:
Though many CAPE users seem to think it's a good idea, I don't compare CAPE with it's history, as then you are implicitly assuming there's some reason why things should mean-revert. I prefer to just use it as an absolute measure, 1/36.15 implies 2.8% expected real return.
How on Earth can the expected real rate of return of US equities be 2.8% when the earnings yield over the last 12 months is 5%?! Are you expecting earnings to not grow at all in the future? Not to mention earnings have significantly more room to grow as US companies enter emerging markets..

This cape nonsense is nonsense.
CAPE is nominal yield, IIRC. Present S&P e-yield is 4.67% minus inflation, which puts the real yield close to 2.8%.
jdilla1107
Posts: 849
Joined: Sun Jun 24, 2012 8:31 pm

Re: New and Improved CAPE is Higher!

Post by jdilla1107 »

JackoC wrote: Fri Feb 15, 2019 3:41 pm
software wrote: Fri Feb 15, 2019 2:21 pm
cjking wrote: Thu Feb 14, 2019 10:51 am
jeffyscott wrote: Thu Feb 14, 2019 9:51 am But then had Shiller not come up with his new methodology, you would have expected return of 3% real. (not that this is much of a difference)
Yes.

(Edit: though I actually get my data from Star Capital web site rather than Shiller, as they give data for all countries. Don't know which method they use, and don't care if they suddenly change. It's unlikely to make much difference.)
... These expected returns are CAPE yields reduced by fund ERs and estimates of losses to withholding taxes on dividends.

I've just checked, and a 0.2% increase in expected return would make no difference to the position of a US tracker in the table.
I mean...kudos to you for putting your money where your mouth is but it seems that you are deriving your entire investment strategy over the assumption that CAPE can accurately predict returns.

Personally, that assumption would not sit well with me given that this has been debunked repeatedly.
Estimating expected return is not, NOT, *NOT* predicting future realized return. The hands down most common misconception on this forum.

Nor is relating earnings yield, 1/PE[x], like correlating whether the east or west won the NBA finals to the return on stocks that year. The earnings yield has a fundamental theoretical relationship to real expected return via the dividend discount model. Whatever measure of PE you decide is most representative, given the vagaries of accounting and business cycle, if the PE goes up, expected return goes down. Like a bond's price and yield relationship.
I don't think you are using the industry accepted definition of "expected return". This is because you are not accounting for earnings growth. To say that the "expected return" of stocks is just the PE ratio is not right. Stocks are different than a US treasury.

I submit this as an objective definition:

https://www.investopedia.com/terms/e/expectedreturn.asp

This matters because a high PE ratio can be brought lower JUST by earnings growth. (And by earnings growth that is higher than the historical average of the last X years.)
Last edited by jdilla1107 on Tue Aug 06, 2019 6:55 pm, edited 2 times in total.
lee1026
Posts: 395
Joined: Sat Jun 21, 2014 7:47 pm

Re: New and Improved CAPE is Higher!

Post by lee1026 »

Earnings yield is a real figure, since the corporations in question generally own real things.

Using REITs as an example because they have simple(r) balance sheets, rents on those apartments are going to go up with time.
User avatar
willthrill81
Posts: 20943
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: New and Improved CAPE is Higher!

Post by willthrill81 »

2015 wrote: Fri Feb 15, 2019 1:58 pm Where I live the Indians use to think inclement weather was some kind of sign. Now we have CAPE and investors driven by none other than ego, greed, and fear will use all kinds of narratives to divine patterns. CAPE also helps to continue to monetize Shiller's career.
I think that valuations do matter to some extent, but there doesn't seem to be anything particularly unique about CAPE10 or any other valuation metric. Valuations are not reliably mean reverting, and their predictive ability seems to be modest, at best. If we're going to use data mined metrics, the average investor equity allocation has been far more predictive of forward 10 year returns than CAPE and has held up well in this regard in the U.S. since it was put forth as well as in EAFE.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings
SovereignInvestor
Posts: 597
Joined: Mon Aug 20, 2018 4:41 pm

Re: New and Improved CAPE is Higher!

Post by SovereignInvestor »

columbia wrote: Thu Feb 14, 2019 9:54 am Was this the “needed” change pointed out by folks, who said that current CAPE was not comparable to measurements from the past?
Not at all. It doesn't fix the issue. It's not just buybacks, but also the CPI has changed to yield weaker inflation adjustments since about 1995.

It is not a like to like comparison.

This is why we have ridiculous output like saying fair value for the S&P is 1600
..that would mean it trades at a forward PE of 9. That's an earnings yield of 11%, while 10Y note is under 2%...meaning if earnings never grow then the S&P would provide a long run 9% premium to 10Y bonds even though the historical average premium has been 4-5% over last century.

The output from CAPE is nonsensical because comparing current readings to those from decades ago when the CPI is not the same and buybacks juice recent EPS more RELATIVE to 9-10 year ago earnings makes it provide poor results

The gap between recent Real EPS and 10 year ago Real EPS used in CAPE will be wider when there is buybacks...so naturally CAPE will be greater when there are buybacks.
https://seekingalpha.com/article/408638 ... d-buybacks

I recomend ysing forward PE instead. It is now 16. Long term historical average is around 13 but then again long term historical interest rates were much greater than now.
User avatar
firebirdparts
Posts: 1795
Joined: Thu Jun 13, 2019 4:21 pm

Re: New and Improved CAPE is Higher!

Post by firebirdparts »

I suddenly realized that cyclically adjusted numbers are much less useful than plain ol' numbers.

I do wonder whether the current fad of buybacks will continue. That may be a critical bit of information.
A fool and your money are soon partners
JackoC
Posts: 1738
Joined: Sun Aug 12, 2018 11:14 am

Re: New and Improved CAPE is Higher!

Post by JackoC »

jdilla1107 wrote: Tue Aug 06, 2019 6:52 pm
JackoC wrote: Fri Feb 15, 2019 3:41 pm
software wrote: Fri Feb 15, 2019 2:21 pm
cjking wrote: Thu Feb 14, 2019 10:51 am
jeffyscott wrote: Thu Feb 14, 2019 9:51 am But then had Shiller not come up with his new methodology, you would have expected return of 3% real. (not that this is much of a difference)
Yes.

(Edit: though I actually get my data from Star Capital web site rather than Shiller, as they give data for all countries. Don't know which method they use, and don't care if they suddenly change. It's unlikely to make much difference.)
... These expected returns are CAPE yields reduced by fund ERs and estimates of losses to withholding taxes on dividends.

I've just checked, and a 0.2% increase in expected return would make no difference to the position of a US tracker in the table.
I mean...kudos to you for putting your money where your mouth is but it seems that you are deriving your entire investment strategy over the assumption that CAPE can accurately predict returns.

Personally, that assumption would not sit well with me given that this has been debunked repeatedly.
Estimating expected return is not, NOT, *NOT* predicting future realized return. The hands down most common misconception on this forum.

Nor is relating earnings yield, 1/PE[x], like correlating whether the east or west won the NBA finals to the return on stocks that year. The earnings yield has a fundamental theoretical relationship to real expected return via the dividend discount model. Whatever measure of PE you decide is most representative, given the vagaries of accounting and business cycle, if the PE goes up, expected return goes down. Like a bond's price and yield relationship.
I don't think you are using the industry accepted definition of "expected return".
Sure I am. The earnings yield of stocks is related to the real expected return as follows:
The price of a stock S=Div/(r-g) S the stock price, r the return, and g the dividend growth rate, or IOW r=S/Div+g, the return the perpetual payout of the dividend plus dividend growth.

But dividend growth comes from earnings growth comes from return. Assume in equilibrium the real expected return on the stock is the return on capital of the company.
So restate the first equation as S= p*E/(r-(1-p)*r), where p is the payout ratio, ie Div=p*E, div growth rate is (1-p)*r
Simplify and you get r=E/S

Thus if you assume the real return is the reinvestment rate, 1/PE ratio is equal to the real return. Since r is a random variable, we say 1/PE ratio equals the expected return.

This is a wholly 'industry standard' inference.
Seasonal
Posts: 2094
Joined: Sun May 21, 2017 1:49 pm

Re: New and Improved CAPE is Higher!

Post by Seasonal »

Earnings yield (1/PE or E/P), whether based on current earnings or 10 year earnings, has been the best predictor of future returns, accounting for somewhere around 40% of the time variation of real returns. Vanguard has a paper on the subject. https://personal.vanguard.com/pdf/s338.pdf

Given the error bands surrounding earnings yield as a predictor, the details of the computation are not very important. The error bands are likely at least plus/minus 5 percentage points.

The changing definition of earnings as GAAP evolves is a big factor in comparing current numbers to prior numbers.
User avatar
Taylor Larimore
Advisory Board
Posts: 29967
Joined: Tue Feb 27, 2007 8:09 pm
Location: Miami FL

Re: New and Improved CAPE is Higher!

Post by Taylor Larimore »

grayfox wrote: Thu Feb 14, 2019 8:48 am Anyone see this? I was over at ONLINE DATA ROBERT SHILLER which is where you can download the stock market data that has all the CAPE stuff. I see that, starting in Sept-2018 he came up with an alternative version of CAPE. The alternate version accounts for changes in corporate payout policy, which I guess means share buybacks.
grayfox:

I stay-the-course and have no use for CAPE (Cyclically Adjusted Price-to-Earnings) ratio.

Best wishes.
Taylor
Jack Bogle's Words of Wisdom: "Stay the Course. No matter what happens, stick to your program. I've said "Stay the course" a thousand times, and I meant it every time. It is the most important single piece of investment wisdom I can give to you."
"Simplicity is the master key to financial success." -- Jack Bogle
jdilla1107
Posts: 849
Joined: Sun Jun 24, 2012 8:31 pm

Re: New and Improved CAPE is Higher!

Post by jdilla1107 »

JackoC wrote: Wed Aug 07, 2019 8:30 am Sure I am. The earnings yield of stocks is related to the real expected return as follows:
The price of a stock S=Div/(r-g) S the stock price, r the return, and g the dividend growth rate, or IOW r=S/Div+g, the return the perpetual payout of the dividend plus dividend growth.

But dividend growth comes from earnings growth comes from return. Assume in equilibrium the real expected return on the stock is the return on capital of the company.
So restate the first equation as S= p*E/(r-(1-p)*r), where p is the payout ratio, ie Div=p*E, div growth rate is (1-p)*r
Simplify and you get r=E/S

Thus if you assume the real return is the reinvestment rate, 1/PE ratio is equal to the real return. Since r is a random variable, we say 1/PE ratio equals the expected return.

This is a wholly 'industry standard' inference.
I have lost interest in the semantics, but the expected return definition I was going for was from CAPM: (Second paragraph... "The formula for calculating the expected return of an asset given its risk is as follows:")

https://www.investopedia.com/terms/c/capm.asp

You seem to be presenting a fixed income definition translation to stocks. An options trader would not agree with your definition.
JackoC wrote: Wed Aug 07, 2019 8:30 am we say 1/PE ratio equals the expected return.
1/PE is the current earnings yield.
kosomoto
Posts: 500
Joined: Tue Nov 24, 2015 8:51 pm

Re: New and Improved CAPE is Higher!

Post by kosomoto »

jdilla1107 wrote: Wed Aug 07, 2019 9:08 am
JackoC wrote: Wed Aug 07, 2019 8:30 am Sure I am. The earnings yield of stocks is related to the real expected return as follows:
The price of a stock S=Div/(r-g) S the stock price, r the return, and g the dividend growth rate, or IOW r=S/Div+g, the return the perpetual payout of the dividend plus dividend growth.

But dividend growth comes from earnings growth comes from return. Assume in equilibrium the real expected return on the stock is the return on capital of the company.
So restate the first equation as S= p*E/(r-(1-p)*r), where p is the payout ratio, ie Div=p*E, div growth rate is (1-p)*r
Simplify and you get r=E/S

Thus if you assume the real return is the reinvestment rate, 1/PE ratio is equal to the real return. Since r is a random variable, we say 1/PE ratio equals the expected return.

This is a wholly 'industry standard' inference.
I have lost interest in the semantics, but the expected return definition I was going for was from CAPM: (Second paragraph... "The formula for calculating the expected return of an asset given its risk is as follows:")

https://www.investopedia.com/terms/c/capm.asp

You seem to be presenting a fixed income definition translation to stocks. An options trader would not agree with your definition.
JackoC wrote: Wed Aug 07, 2019 8:30 am we say 1/PE ratio equals the expected return.
1/PE is the current earnings yield.
Right, 1/PE is not the expected return and definitely not the “industry standard” for expected return. You can’t calculate expected return without factoring growth.
JackoC
Posts: 1738
Joined: Sun Aug 12, 2018 11:14 am

Re: New and Improved CAPE is Higher!

Post by JackoC »

jdilla1107 wrote: Wed Aug 07, 2019 9:08 am
JackoC wrote: Wed Aug 07, 2019 8:30 am Sure I am. The earnings yield of stocks is related to the real expected return as follows:
The price of a stock S=Div/(r-g) S the stock price, r the return, and g the dividend growth rate, or IOW r=S/Div+g, the return the perpetual payout of the dividend plus dividend growth.

But dividend growth comes from earnings growth comes from return. Assume in equilibrium the real expected return on the stock is the return on capital of the company.
So restate the first equation as S= p*E/(r-(1-p)*r), where p is the payout ratio, ie Div=p*E, div growth rate is (1-p)*r
Simplify and you get r=E/S

Thus if you assume the real return is the reinvestment rate, 1/PE ratio is equal to the real return. Since r is a random variable, we say 1/PE ratio equals the expected return.

This is a wholly 'industry standard' inference.
I have lost interest in the semantics, but the expected return definition I was going for was from CAPM: (Second paragraph... "The formula for calculating the expected return of an asset given its risk is as follows:")
https://www.investopedia.com/terms/c/capm.asp

You seem to be presenting a fixed income definition translation to stocks. An options trader would not agree with your definition.
JackoC wrote: Wed Aug 07, 2019 8:30 am we say 1/PE ratio equals the expected return.
1/PE is the current earnings yield.
It's not about semantics, and definitely not options trading. And it's completely orthogonal to the CAPM definition of risk asset return, which relates risky return to riskless return. What we're talking about here is relating stock return to price and earnings.

A stock is a perpetuity. The value of a perpetuity is payout/(discount rate-payout growth rate), so in case of a stock S=Div/(r-g), r being rate of return, g the dividend growth rate, we take these terms to be real (ie inflation adjusted). We assume the total return of the stock and rate the company can earn on its investments is the same in equilibrium. Then we recognize that the dividend is the proportion of earnings which is paid out, p*E, and also that then only 1-p is of earnings is left to reinvest. Rearranging and simplifying, we get r=E/S=1/PE, IOW real expected return of stocks is approximated by the earnings yield. This is a standard derivation. If you need a written reference see Jeremy Siegel's "Stocks for the Long Run" among many others.

So back to my original point, CAPE isn't some random thing unrelated to return and then somebody found 'hey look it has some correlation to subsequent return!' It has a fundamental relationship to return under classic corporate finance assumptions. However still subject to real world questions.

One obvious issue for 1/PE as estimator of the real *expected* return is which 'E'? Jeremy Siegel (I believe fairly generally viewed as permabull or close to it) and Robert Shiller (typically more dour though not really a permabear) agree that E/S is a fundamentally based estimate of real expected return of stocks. But Siegel thinks Shiller's averaging of last 10 yrs inflation adjusted earnings for the E in CAPE is too pessimistic in the current condition. Shiller doesn't agree, fully at least.

The other assumption which can be challenged is the implicit one of zero net dilution. The whole proportion (1-p) of Earnings is available for new investments to grow earnings only if the firm's current earnings can be sustained for the cost of replacing its current assets given by the depreciation charge already subtracted out of cash flow to get earnings. That's generally not been true historically: firms have had to make new net investment just to sustain real earnings. This would make 1/PE too optimistic an estimator of E[r].

On the positive side potentially, the equation also would not account for 'payouts' in the form of buy backs, if Div=p*E is just the cash dividend. This has been discussed on several threads since I've posted here and I'm sure many before. But this paper has a pessimistic finding about net dilution in the age of buybacks, that it hasn't really gone down: accelerated dilution in old (to sustain earnings) and new (more senior executive compensation in the form of dilutive stock option grants) forms seem to have offset increased buybacks. As well, some of the increase in buybacks just recently is via changing firms' capital structures more toward debt, which does not increase earnings per share $ for $, since it increases the cost of debt service subtracted from future earnings. Other works have examined in general how debt and asset replacement simplifications impact the measure.
https://www.researchaffiliates.com/en_u ... irage.html

But again it's well accepted that earnings yield is an estimator of real expected return for the stock market under simple fundamental assumptions. Those rejecting this based on 'it doesn't include growth', please re read the derivation: yes it does.
Post Reply