CAPE with tapered weights

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
Topic Author
Beliavsky
Posts: 932
Joined: Sun Jun 29, 2014 10:21 am

CAPE with tapered weights

Post by Beliavsky » Thu Jan 10, 2019 2:30 pm

An analyst note

Is this popular measure of stock market value giving a misleading “sell” signal?
Here's why scare stories about the current CAPE being a predictor of doom may prove to be wide of the mark.
10 JANUARY 2019
Schroders
by Duncan Lamont

says that the 10-year CAPE for the U.S. stock market will fall substantially once the low earnings of 2009 roll off. To reduce this echo effect, one could computed a weighted average earnings with declining weights, for example

avg(t) = (3*x(t) + 2*x(t-1) + x(t-2))/6

for a 3-term average. Then the impact of prior-year earnings would decline gradually. I wonder if anyone who has backtested CAPE has looked at this. A weighted CAPE gives more weight to recent earnings, and to compensate for this one could extend the lookback period from 10 to say 15 years.

Topic Author
Beliavsky
Posts: 932
Joined: Sun Jun 29, 2014 10:21 am

Re: CAPE with tapered weights

Post by Beliavsky » Wed May 29, 2019 7:22 am

It does not make sense that in 2020, when 2008 losses roll off, that one's estimate of fair value of the the S&P 500 will suddenly jump. This echo effect could be reduced by computing average earnings with linearly declining weights for prior years, as I suggested in the post above.

Are Stocks as Expensive as Shiller’s Measure Says?
A valuation tool popularized by Robert Shiller is affected by corporate losses that occurred more than a decade ago
By Justin Lahart
Wall Street Journal
May 29, 2019 6:33 a.m. ET

...

The CAPE stands at a very high 28.9 according to the model Mr. Shiller provides on his website, which compares with a 50-year average of about 20. That is worrisome because very high CAPE levels have tended to be associated with subpar medium-term market returns.

But it wouldn’t be quite so elevated if it weren’t for the steep losses companies in the S&P 500 experienced in the final quarter of 2008. That is especially odd since the fourth quarter of 2008 was more than 10 years ago. Yet, in the way Mr. Shiller has put together the spreadsheet he maintains on his website, the CAPE isn’t actually an average of the past 10 years earnings but a 10-year average of the previous year’s earnings.

Mr. Shiller did this to maintain consistency with historical data that extends beyond the 83 years’ worth of earnings figures available from S&P Dow Jones Indices. As a result, though, his CAPE figure residually reflects those late-2008 losses. Recalculating the CAPE to remove this residual effect brings it down to 27.2. That is still high, but it is about 6% lower than Mr. Shiller’s CAPE.

Of course, the fact that Mr. Shiller’s CAPE calculation reflects 11, rather than 10, years’ worth of data is true across the entire history of his data, so his CAPE always has tended to be a bit higher than it might otherwise be. The difference now is that, because of the distortions introduced by the financial crisis, the gap between his figure and the true 10-year CAPE is far wider than it has been in at least the past 73 years.

...

Stormbringer
Posts: 822
Joined: Sun Jun 14, 2015 7:07 am

Re: CAPE with tapered weights

Post by Stormbringer » Wed May 29, 2019 8:03 am

The problem with comparing CAPE ratios over time is that it doesn't account for interest rate differences over time. In order to judge something as expensive, you need to ask the question "compared to what?". Sure, CAPE might be 28 or 29, but last I checked the 10-year treasury was selling for 44x earnings. Back when CAPE was 14 or 15, the 10-yr was probably selling for 20x earnings.
"Compound interest is the most powerful force in the universe." - Albert Einstein

larryswedroe
Posts: 15955
Joined: Thu Feb 22, 2007 8:28 am
Location: St Louis MO

Re: CAPE with tapered weights

Post by larryswedroe » Wed May 29, 2019 8:21 am

While not irrelevant dropping 2008 moves the CAPE about say 3. So if move from 29 which gives real return of 3.4 to 26 which gives real return of 3.8. Still well below historical average. And I would note that profit margins at historically very high levels and if anything that has tendency to RTM, and earnings have been helped massively by the large corporate tax cut (which could easily be reversed with a change in administration, should that occur). Of course my crystal ball remains cloudy
And re interest rates, be careful to avoid falling for the money illusion problem. Lower rates don't necessarily justify higher P/E ratios (as lower rates can reflect lower future earnings due to low growth.)

asif408
Posts: 1733
Joined: Sun Mar 02, 2014 8:34 am
Location: Florida

Re: CAPE with tapered weights

Post by asif408 » Wed May 29, 2019 9:06 am

Stormbringer wrote:
Wed May 29, 2019 8:03 am
The problem with comparing CAPE ratios over time is that it doesn't account for interest rate differences over time. In order to judge something as expensive, you need to ask the question "compared to what?". Sure, CAPE might be 28 or 29, but last I checked the 10-year treasury was selling for 44x earnings. Back when CAPE was 14 or 15, the 10-yr was probably selling for 20x earnings.
So why are the CAPE ratios for foreign countries much lower than the US, even though interest rates on those countries equivalent to 10 year Treasuries lower that the US? If what you say is true, CAPE ratios for the UK, Japan, Spain, Australia, Hong Kong, etc. should be much higher, but they aren't. In many cases they are half or less.

marcopolo
Posts: 2166
Joined: Sat Dec 03, 2016 10:22 am

Re: CAPE with tapered weights

Post by marcopolo » Wed May 29, 2019 2:36 pm

larryswedroe wrote:
Wed May 29, 2019 8:21 am
While not irrelevant dropping 2008 moves the CAPE about say 3. So if move from 29 which gives real return of 3.4 to 26 which gives real return of 3.8. Still well below historical average. And I would note that profit margins at historically very high levels and if anything that has tendency to RTM, and earnings have been helped massively by the large corporate tax cut (which could easily be reversed with a change in administration, should that occur). Of course my crystal ball remains cloudy
And re interest rates, be careful to avoid falling for the money illusion problem. Lower rates don't necessarily justify higher P/E ratios (as lower rates can reflect lower future earnings due to low growth.)
Larry,

Your numbers above seem to indicate that you are basing forward expected real returns on 1/CAPE.
Do you think that applies across time given the accounting changes that have somewhat re-defined how earning are reported (the E in CAPE).
So, if 25 years ago 1/CAPE was a good predictor of future real returns, can the same be true today without some adjustment for the changing nature of the denominator in the CAPE calculations?

I would think that is one of (several) reasons why CAPE seems to be persistently higher today than it was 25 years ago. If one is to use 1/CAPE to predict future real returns, how to adjust for these changes?
Once in a while you get shown the light, in the strangest of places if you look at it right.

marcopolo
Posts: 2166
Joined: Sat Dec 03, 2016 10:22 am

Re: CAPE with tapered weights

Post by marcopolo » Wed May 29, 2019 2:39 pm

asif408 wrote:
Wed May 29, 2019 9:06 am
Stormbringer wrote:
Wed May 29, 2019 8:03 am
The problem with comparing CAPE ratios over time is that it doesn't account for interest rate differences over time. In order to judge something as expensive, you need to ask the question "compared to what?". Sure, CAPE might be 28 or 29, but last I checked the 10-year treasury was selling for 44x earnings. Back when CAPE was 14 or 15, the 10-yr was probably selling for 20x earnings.
So why are the CAPE ratios for foreign countries much lower than the US, even though interest rates on those countries equivalent to 10 year Treasuries lower that the US? If what you say is true, CAPE ratios for the UK, Japan, Spain, Australia, Hong Kong, etc. should be much higher, but they aren't. In many cases they are half or less.
One possible contributor is accounting changes that affected how earning are reported. The changes (GAAP) lowered reported earning in the US. Not sure if there have been similar rules put in place in other countries. Inflated earnings numbers (relative to US), would make those countries CAPE seem lower.
Once in a while you get shown the light, in the strangest of places if you look at it right.

larryswedroe
Posts: 15955
Joined: Thu Feb 22, 2007 8:28 am
Location: St Louis MO

Re: CAPE with tapered weights

Post by larryswedroe » Wed May 29, 2019 3:00 pm

Marco, the changes would adjust it about 3 downward, so doesn't change expected returns all that much and again I urge caution on conservative side due to extremely high margins today while we have very low UE---note that margins have come down quite a bit already and my bet would be they are going lower (also were helped by corporate tax cut)

SovereignInvestor
Posts: 365
Joined: Mon Aug 20, 2018 4:41 pm

Re: CAPE with tapered weights

Post by SovereignInvestor » Wed May 29, 2019 3:52 pm

asif408 wrote:
Wed May 29, 2019 9:06 am
Stormbringer wrote:
Wed May 29, 2019 8:03 am
The problem with comparing CAPE ratios over time is that it doesn't account for interest rate differences over time. In order to judge something as expensive, you need to ask the question "compared to what?". Sure, CAPE might be 28 or 29, but last I checked the 10-year treasury was selling for 44x earnings. Back when CAPE was 14 or 15, the 10-yr was probably selling for 20x earnings.
So why are the CAPE ratios for foreign countries much lower than the US, even though interest rates on those countries equivalent to 10 year Treasuries lower that the US? If what you say is true, CAPE ratios for the UK, Japan, Spain, Australia, Hong Kong, etc. should be much higher, but they aren't. In many cases they are half or less.
Current CAPE will be biased high for the US for the following

1) Buybacks lead to greater CAPE than historical sicentre they are relative new to after year 2000. They make EPS grow much faster than inflation and a mere CPI adjustment to.older years doesn't adjust the earnings on a PER SHARE basis to current earnings power. This wasn't problem ore 1990s when buybacks didn't occur so the issue is when one compares current reading to long term average.


2) CPI adjustments have become less after 1990d when CPI was changed. Boskin commission estimates impact is 0.6% per year less.

3) US had massive corporate tax cuts.in 2018 so EPS before 2018 which is 9 of the last 10 years in data reflect EPS that would need to be juiced 10% higher if it occurred in present with new corporate tax rates so the EPS in Cape is understated by about 9% by not reflecting pro forma tax rate.


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

This doesn't include lower interest rates which affects all countries.

CAPE is ridiculously suggesting fair value is about 1600. Why should SPX trade at under 10x forward earnings for a yield of over 10%, when the 10Y is under 2.5%? That would Mean if the economy totally stagnated total earnings flatlined and valuations never grew from that low level the S&P would put out 10%+ returns or a 7.5%+ risk premium to 10Y note..which is way above the historical average in the 4.5% area and this is assuing a super bearish scenario of total stagnation. CAPE is just totally off base.

Amadis_of_Gaul
Posts: 102
Joined: Tue Dec 18, 2018 5:57 pm

Re: CAPE with tapered weights

Post by Amadis_of_Gaul » Wed May 29, 2019 6:18 pm

It seems to me that when inflation is low, CAPE should be higher. The discount rate is less, so future returns should be worth more.

I am not convinced, btw, that low international price-earnings ratios indicate anything more than grim prospects for a bunch of countries with unfavorable demographics. International CAPE ratios may be elevated too, but the baseline is so low they don't look like it.

User avatar
willthrill81
Posts: 11803
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: CAPE with tapered weights

Post by willthrill81 » Wed May 29, 2019 6:31 pm

As we've seen Larry Swedroe do in this thread, many now believe that CAPE's real usefulness is not as a tool for predicting some kind of mean reversion, which is how Shiller was using it when he made his 1996 prediction that stocks would have negative real returns over the next decade, but rather that 1/CAPE is a good estimate of the next 10 year's returns. Still, the point remains that earnings from nearly a decade ago should seemingly not have as big of an impact on today's prices as last year's earning, for instance.
“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

Post Reply