Author Topic: Swedroe: CAPE 10 Ratio In Need Of Context  (Read 1847 times)

Classical_Liberal

  • Handlebar Stache
  • *****
  • Posts: 1151
  • Age: 43
Swedroe: CAPE 10 Ratio In Need Of Context
« on: May 20, 2016, 03:31:44 PM »
http://www.etf.com/sections/index-investor-corner/swedroe-ratio-need-context?nopaging=1

A relative Newbie to intelligent investing here, but have absorbed tons of information over the past year.  While I'm totally sold on indexing and have actively moved my portfolio to a 3 fund (as close as I can get with my 401K tax deferred portion), the high CAPE valuation has been an ongoing splinter in my mind.  I have to actively fight the urge to keep too much cash on the sidelines.

This Swedroe article appears to provide some legitimate points, however, it also has the "feel" of data mining to paint a rosier picture of future returns. I'm curious to see thoughts from some of the more seasoned investment minds of this forum. 

The article is more than a month old, but I didn't see it posted anywhere else, sorry if this is a repeat.

Thanks!

Seppia

  • Pencil Stache
  • ****
  • Posts: 616
  • Age: 39
  • Location: NYC
Re: Swedroe: CAPE 10 Ratio In Need Of Context
« Reply #1 on: May 20, 2016, 03:50:11 PM »
All this sounds very reasonable.
So did all the justifications of hi values in the past.
Also: why are Europe and emerging markets around 15 in terms of Shiller PE?

MustacheAndaHalf

  • Handlebar Stache
  • *****
  • Posts: 1805
Re: Swedroe: CAPE 10 Ratio In Need Of Context
« Reply #2 on: May 20, 2016, 10:24:28 PM »
You might want to look at Vanguard's white paper on forecasting stock returns.
https://personal.vanguard.com/pdf/s338.pdf

For example, P/E is the best signal for future returns in that paper... but it takes decades, and the correlation is only 0.40.  So guessing about a crash next year from P/E isn't a good idea.  Even Schiller (who co-authored the paper on CAPE 10) cautions timing the market based on CAPE 10.

Classical_Liberal

  • Handlebar Stache
  • *****
  • Posts: 1151
  • Age: 43
Re: Swedroe: CAPE 10 Ratio In Need Of Context
« Reply #3 on: May 20, 2016, 11:19:12 PM »
You might want to look at Vanguard's white paper on forecasting stock returns.
https://personal.vanguard.com/pdf/s338.pdf

For example, P/E is the best signal for future returns in that paper... but it takes decades, and the correlation is only 0.40.  So guessing about a crash next year from P/E isn't a good idea.  Even Schiller (who co-authored the paper on CAPE 10) cautions timing the market based on CAPE 10.

Thanks for the info, I've read it before.  Not trying to market time.  The only measure to show any correlation to long term real returns is the CAPE.  Unfortunately the CAPE is sitting more than a standard deviation above historical mean.  I'm still buying, I would just like to think (hope) we can get more than 3-4% real annual returns over the next decade.  The Swedroe article makes the case with several points (some of which I've read before, others not) perhaps the CAPE mean has been reset higher.  This makes me feel better, I like to feel better.   His arguments (and others not presented in this article from other sources) seem logically sound.  However, I wonder if someone in 1981 could have made a persuasive case that the CAPE was reset to a lower mean and argued returns would suck through the next decade despite CAPE being a standard deviation below historical mean. 

Seppia

  • Pencil Stache
  • ****
  • Posts: 616
  • Age: 39
  • Location: NYC
Re: Swedroe: CAPE 10 Ratio In Need Of Context
« Reply #4 on: May 21, 2016, 12:42:48 AM »
You might want to look at Vanguard's white paper on forecasting stock returns.
https://personal.vanguard.com/pdf/s338.pdf

For example, P/E is the best signal for future returns in that paper... but it takes decades, and the correlation is only 0.40.  So guessing about a crash next year from P/E isn't a good idea.  Even Schiller (who co-authored the paper on CAPE 10) cautions timing the market based on CAPE 10.

Great paper, thanks a ton for the link.
Just to specify: of course my post above referred to long term returns, there is no metric out there that has any predictive power over short term returns.

forummm

  • Walrus Stache
  • *******
  • Posts: 7396
  • Senior Mustachian
Re: Swedroe: CAPE 10 Ratio In Need Of Context
« Reply #5 on: May 21, 2016, 07:14:05 AM »
You might want to look at Vanguard's white paper on forecasting stock returns.
https://personal.vanguard.com/pdf/s338.pdf

For example, P/E is the best signal for future returns in that paper... but it takes decades, and the correlation is only 0.40.  So guessing about a crash next year from P/E isn't a good idea.  Even Schiller (who co-authored the paper on CAPE 10) cautions timing the market based on CAPE 10.

Thanks for the info, I've read it before.  Not trying to market time.  The only measure to show any correlation to long term real returns is the CAPE.  Unfortunately the CAPE is sitting more than a standard deviation above historical mean.  I'm still buying, I would just like to think (hope) we can get more than 3-4% real annual returns over the next decade.  The Swedroe article makes the case with several points (some of which I've read before, others not) perhaps the CAPE mean has been reset higher.  This makes me feel better, I like to feel better.   His arguments (and others not presented in this article from other sources) seem logically sound.  However, I wonder if someone in 1981 could have made a persuasive case that the CAPE was reset to a lower mean and argued returns would suck through the next decade despite CAPE being a standard deviation below historical mean. 

I also have thought the CAPE mean should be adjusted higher. Also, instead of looking at a 140+ year period, you should only look at the last 75 years or perhaps even less.

Average monthly PE (not CAPE) was
15.57 from 1871-2015,
17.78 from 1950-2015,
18.85 from 1960-2015,
19.07 from 1970-2015,
21.01 from 1980-2015,
24.59 from 1990-2015.

Average monthly CAPE was
16.6 from 1871-2015,
18.92 from 1950-2015,
19.7 from 1960-2015,
19.49 from 1970-2015,
21.4 from 1980-2015,
25.3 from 1990-2015.

If you believe that the world and financial markets are the same now as they were in 1885 (when the stock market was mostly railroads), then your expectation of a reversion to the mean of 16.6 would make sense to you. But in addition to things like the Fed, the SEC, there is stuff like no longer being on the gold standard, dramatically increased participation of the population in the equity markets, the creation and widespread adoption of index funds, electronic trading, immediate availability of information globally, moving from 1/8 of a dollar ticks to 1 penny, etc. Stocks are no longer seen only as gambling for rich people. They are part of the retirement portfolios for most (?) people. A higher PE probably means lower expected return (you are paying more for future profits). But perhaps there's less risk that returns will be even lower due to a dramatic downward shift in the multiplier.

Classical_Liberal

  • Handlebar Stache
  • *****
  • Posts: 1151
  • Age: 43
Re: Swedroe: CAPE 10 Ratio In Need Of Context
« Reply #6 on: May 21, 2016, 04:04:20 PM »

Average monthly PE (not CAPE) was
15.57 from 1871-2015,
17.78 from 1950-2015,
18.85 from 1960-2015,
19.07 from 1970-2015,
21.01 from 1980-2015,
24.59 from 1990-2015.

Average monthly CAPE was
16.6 from 1871-2015,
18.92 from 1950-2015,
19.7 from 1960-2015,
19.49 from 1970-2015,
21.4 from 1980-2015,
25.3 from 1990-2015.

If you believe that the world and financial markets are the same now as they were in 1885 (when the stock market was mostly railroads), then your expectation of a reversion to the mean of 16.6 would make sense to you. But in addition to things like the Fed, the SEC, there is stuff like no longer being on the gold standard, dramatically increased participation of the population in the equity markets, the creation and widespread adoption of index funds, electronic trading, immediate availability of information globally, moving from 1/8 of a dollar ticks to 1 penny, etc. Stocks are no longer seen only as gambling for rich people. They are part of the retirement portfolios for most (?) p`eople. A higher PE probably means lower expected return (you are paying more for future profits). But perhaps there's less risk that returns will be even lower due to a dramatic downward shift in the multiplier.

All good points!  Your means are very telling, I don't know why I hadn't thought to do that before. 

Also interesting: Real average annual dividend reinvest S&P returns
1871 - 2015  6.83%
1950 - 2015  7.38%
1960 - 2015  5.86%
1970 - 2015  6.01%
1980 - 2015  8.16%
1990 - 2015  7.04%

Data from http://dqydj.com/sp-500-return-calculator/  January to January