Valuations Matter....So Do Earnings

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
Topic Author
Alchemist
Posts: 638
Joined: Sat Aug 30, 2014 6:35 am

Valuations Matter....So Do Earnings

Post by Alchemist »

In December 2020 the P/E ratio of the S&P 500 was at a shockingly high 39.26. The 10 year treasury was yielding 0.93%. Many posters on Bogleheads were warning of an epic bubble. We were at valuations not seen since dotcom irrational exuberance days, surely the market could go no further.

Since then the PE ratio of the S&P 500 tumbled 29% to 27.96 today (14 Oct 2021). The 10 year treasury is now over 1.5%.

Surely that means a big tumble in the stock market right? Quite the contrary, Vanguard Total Stock Market Index Fund (VTSAX) is up 18.95% year to date and the Vanguard 500 Index Fund (VFIAX) is up 19.48%.

What happened? As myself and some other Bogleheads pointed out time and again, the P/E ratio has two variables. The "E" can also go up and that is what happened. The 12 month forward P/E estimates from December 2020 were in the mid-20's thanks to what was expected to be a strong economic turn around. If earnings came in around expectation, the market was appropriately valued. Thus far earnings have been exceeding expectations. Therefore prices have actually gone up while valuations have come down.

My point is not to say I told you so (well not entirely :wink: ). Rather it is to provide another example of how dangerous it can be to rely on squiggly lines in a chart to predict the future. Valuations really do matter but they exist within a broader context. A P/E of 40 while bonds are yielding 5-7% is probably not a good sign. A P/E of 39 with trillions of dollars of pent up demand ready to be deployed as a solution to a global crisis suddenly becomes available may not be anything to worry about at all.

If you predict the past you are likely to be very surprised by the future.

Sources:

Valuations: https://www.multpl.com/s-p-500-pe-ratio/table/by-month

Interest Rates: https://www.multpl.com/10-year-treasury ... e/by-month

VTSAX Performance: https://www.morningstar.com/funds/xnas/vtsax/quote

VFIAX Performance: https://www.morningstar.com/funds/xnas/vfiax/quote
User avatar
JoMoney
Posts: 16260
Joined: Tue Jul 23, 2013 5:31 am

Re: Valuations Matter....So Do Earnings

Post by JoMoney »

Back as of August 31, 2011, a portfolio of S&P 500 Growth Index stocks had a Price/Book value of 3.1x (source: Vanguard fund annual report)

It was last reported as 10.0x (Vanguard's fund website)
Over 10 years, that's a +12% annualized/CAGR valuation change just in the multiple... that the underlying fundamentals did not keep up with.

Given the extremely low, and falling, interest rate environment, a business that is able to grow it's fundamental value intrinsically should demand a higher price... But call me skeptical that the underlying has the ability to maintain that multiple and fundamentals grow into it :annoyed
Last edited by JoMoney on Fri Oct 15, 2021 7:37 am, edited 1 time in total.
"To achieve satisfactory investment results is easier than most people realize; to achieve superior results is harder than it looks." - Benjamin Graham
asif408
Posts: 2616
Joined: Sun Mar 02, 2014 7:34 am
Location: Florida

Re: Valuations Matter....So Do Earnings

Post by asif408 »

All you've shown is that using one year of trailing 12 month reported earnings is a poor measuring for valuing a stock market, especially after a recession. At the end of 2008/ beginning of 2009 and toward the end of 2001 beginning of 2002, the P/E shot up and showed stocks were massively overvalued. Obviously neither one was true, so the measure failed when you needed it most. At least it's better than the forward P/E, though (shudder).

The Shiller P/E, which is as you probably are aware of, is based on average inflation-adjusted earnings from the previous 10 years, and doesn't show almost any drop: https://www.multpl.com/shiller-pe. And it more correctly reflected valuations levels in the 2000-2002 and 2007-2009 bear markets. I don't want to get into a discussion about flaws in the Shiller P/E, but its more than obvious that it's a better measure of valuation that TTM P/E, particularly at peaks and troughs. The other shorter term P/E measures fail miserably at peaks and troughs.
Topic Author
Alchemist
Posts: 638
Joined: Sat Aug 30, 2014 6:35 am

Re: Valuations Matter....So Do Earnings

Post by Alchemist »

My point is not that valuations, including the Shiller P/E, are meaningless. The point is that they are not some crystal ball and must be considered with a lot of other context. Anyone trying to make real world investing decisions based on valuations have lost out on a lot more gains than they have saved themselves from market drops.
Da5id
Posts: 5066
Joined: Fri Feb 26, 2016 7:20 am

Re: Valuations Matter....So Do Earnings

Post by Da5id »

Alchemist wrote: Fri Oct 15, 2021 7:12 am In December 2020 the P/E ratio of the S&P 500 was at a shockingly high 39.26. The 10 year treasury was yielding 0.93%. Many posters on Bogleheads were warning of an epic bubble. We were at valuations not seen since dotcom irrational exuberance days, surely the market could go no further.
I don't change my allocation in response to predictions or valuations personally. And think using valuations for any sort of market timing is not historically a good idea.

But I think it is premature to say that those predicting that valuations are too high are wrong. I think even for most who believe in using valuations for something, high valuations don't result in them predicting a *short term* outcome either way. Because even if they are correct about "Overly high valuations", for whatever that means, the result can manifest in a crash, in slower future growth over the next 10 year, etc.
asif408
Posts: 2616
Joined: Sun Mar 02, 2014 7:34 am
Location: Florida

Re: Valuations Matter....So Do Earnings

Post by asif408 »

Alchemist wrote: Fri Oct 15, 2021 8:20 am My point is not that valuations, including the Shiller P/E, are meaningless. The point is that they are not some crystal ball and must be considered with a lot of other context. Anyone trying to make real world investing decisions based on valuations have lost out on a lot more gains than they have saved themselves from market drops.
For short term timing I agree, they are poor, valuations don't tell you anything about the next few years. That's why your example doesn't really tell us anything, it's been less than a year.

The bigger picture is that it depends on how the valuations are being used. If you're using valuations to switch between stocks and some other stocks or bonds and make short term timing based on weeks, months, or a couple of years that's a bad idea. If you use it to make occasional adjustments among your stock allocation that you plan to stick with for the next decade or two that seems like a perfectly valid option. You just have to have patience and settle for possibly several years of underperformance and/or potentially never seeing outperformance show up. No risk, no reward.

Certainly valuations of international and value stocks have looked better than domestic US and growth stocks for the last few years, and I haven't seen anyone here suggesting international value and emerging market value stocks are in a bubble. Those markets sell for P/E ratios in the low teens to single digits, if you are interested in P/Es. And interest rates in many of those countries are even lower than in the US. So if the US market is fairly valued adjusted for interest rates UK and Japanese stocks must be a screaming buy.

Of course, international and value stock have underperformed for several years and may continue (though in the last year international value and emerging market value funds have beaten the US Total Stock Market fund: https://www.portfoliovisualizer.com/bac ... ion3_3=100) , but they certainly aren't overvalued. Of course my example is like yours and only the past year, so who know if this continues.
Nathan Drake
Posts: 6235
Joined: Mon Apr 11, 2011 12:28 am

Re: Valuations Matter....So Do Earnings

Post by Nathan Drake »

Alchemist wrote: Fri Oct 15, 2021 8:20 am My point is not that valuations, including the Shiller P/E, are meaningless. The point is that they are not some crystal ball and must be considered with a lot of other context. Anyone trying to make real world investing decisions based on valuations have lost out on a lot more gains than they have saved themselves from market drops.
Yes, context matters

Like US corporate profit margins being at the extremes (does this seem sustainable?)

Like higher than anticipated inflation and labor shortages eroding earnings in the future

Like preferential corporate tax treatment potentially being curtailed to fund ballooning debt.

Extreme levels of leverage permeate the market

Lots of “extremes” baked in to the incredibly optimistic earnings projections and past earnings.

What has happened this year does not tell us what will happen in 5 or 10 years. The valuations could mean revert if any of these risks start showing up in a meaningful way
20% VOO | 20% VXUS | 20% AVUV | 20% AVDV | 20% AVES
atdharris
Posts: 2091
Joined: Wed Jan 02, 2019 2:18 pm

Re: Valuations Matter....So Do Earnings

Post by atdharris »

Forward PE of the market is 20. While that is still higher than the historical 15x earnings, with rates around 1.5%, that is not an epic bubble valuation.
MarkRoulo
Posts: 1198
Joined: Mon Jun 22, 2015 10:25 am

Re: Valuations Matter....So Do Earnings

Post by MarkRoulo »

JoMoney wrote: Fri Oct 15, 2021 7:35 am Back as of August 31, 2011, a portfolio of S&P 500 Growth Index stocks had a Price/Book value of 3.1x (source: Vanguard fund annual report)

It was last reported as 10.0x (Vanguard's fund website)
Over 10 years, that's a +12% annualized/CAGR valuation change just in the multiple... that the underlying fundamentals did not keep up with.

Given the extremely low, and falling, interest rate environment, a business that is able to grow it's fundamental value intrinsically should demand a higher price... But call me skeptical that the underlying has the ability to maintain that multiple and fundamentals grow into it :annoyed
I don't think it is obvious at all that the book value of a company such as Microsoft or Google or Apple has a lot to do with the company's ability to make and grow profits.

Both Apple and Microsoft could double (or more) their product sales without needing to increase their factory footprint. Because the actual manufacturing is outsourced. iPhones are assembled by FoxConn, iPhone SoC chips are manufactured by TSMC, iPhone screens are manufactured by Samsung and LG and whoever it is now. Microsoft can crank out DVDs to install MS-Office without factories -- this, too, is outsourced. Or Microsoft can allow folks to download the software, but the servers needed are such a small part of the product development that Microsoft won't need to double their server farm footprint to sell x2 as much software.

Back when industry was making steel (or canning soup) you pretty much needed to double your factory to double your output, so price/book really mattered. If one can replicate the factory for ~1x or buy the factories of a company (via stock) for 3x one of these was a much better choice.

But Apple and Microsoft and Google (and Facebook, and ...) are not cranking out commodity widgets (like steel and soup). You can't look at one of these trading a high price/book ratio and go off and build an equivalent company for book value. The value isn't in the assets that are traditionally counted as book value.

200 years ago, *land* was what was valuable. Specifically agricultural land.

I'll note that the balance sheet for 1860s industrial companies tended to be very light in agricultural land, but the companies were valuable none the less. The accounting needed to change to reflect this.

I'm pretty sure that industrial age accounting (book value, etc.) doesn't capture much of the true value of modern tech companies (Microsoft, Google, Facebook ... but also smaller ones such as Applied Materials) and modern high-end service companies. Bain isn't valuable because of the buildings.

Which doesn't mean that these companies can't be overvalued. I really don't understand Tesla's valuation.

But pointing to book value for a company like Apple may well be missing where most of the accumulated value lies.
azanon
Posts: 3142
Joined: Mon Nov 07, 2011 9:34 am

Re: Valuations Matter....So Do Earnings

Post by azanon »

The OP has the wrong link for Valuation. Shiller PE is actually here: http://www.econ.yale.edu/~shiller/data.htm Shiller PE in Dec 2020 was 33.77 and now (2021.09) it's 38.34, so definitely higher than Dec 2020. I say wrong because P/E 1 is next to useless.

And of course "earnings also matter" - it's one of the two variables in Shiller PE, granted it's smoothed 10 year earnings, and I'll not go into why 10 years is used instead of the (near useless) 1 yr earnings.

Is 38 and change a bubble for Shiller PE? I think so considering it's only been higher than that for 1.5 years or so of 150 years of history.
afan
Posts: 8195
Joined: Sun Jul 25, 2010 4:01 pm

Re: Valuations Matter....So Do Earnings

Post by afan »

Serious people do not claim that either PE ratio predict stock returns over the following 10 months. They do not claim that anything does this.

The claim is that high PE's predict low returns over 10 and 20 year periods. The data supporting that proposition are quite strong.

The predictions are not nearly good enough to use for timing, however. This makes for fun casual conversation and perhaps useful for projecting funds available from investment over a long time. But the limited predictive value means one should be cautious in making, for example, retirement plans based on valuations.

"More interesting than useful."
We don't know how to beat the market on a risk-adjusted basis, and we don't know anyone that does know either | --Swedroe | We assume that markets are efficient, that prices are right | --Fama
randomguy
Posts: 11295
Joined: Wed Sep 17, 2014 9:00 am

Re: Valuations Matter....So Do Earnings

Post by randomguy »

azanon wrote: Fri Oct 15, 2021 10:43 am The OP has the wrong link for Valuation. Shiller PE is actually here: http://www.econ.yale.edu/~shiller/data.htm Shiller PE in Dec 2020 was 33.77 and now (2021.09) it's 38.34, so definitely higher than Dec 2020. I say wrong because P/E 1 is next to useless.

And of course "earnings also matter" - it's one of the two variables in Shiller PE, granted it's smoothed 10 year earnings, and I'll not go into why 10 years is used instead of the (near useless) 1 yr earnings.

Is 38 and change a bubble for Shiller PE? I think so considering it's only been higher than that for 1.5 years or so of 150 years of history.
And of those 150 years, how many times was E calculated using the same rules that apply in 2020? How much different is E today versus in say in 1990?

At a certain level we have a problem where we are either in like a 30 year stock market bubble (The average PE10 before 1990 was like 12. It is up to 15 these days AND we have barely touched that for more than short period in 2009) or something has changed to support higher PEs. Or a mixture of both.
User avatar
burritoLover
Posts: 4097
Joined: Sun Jul 05, 2020 12:13 pm

Re: Valuations Matter....So Do Earnings

Post by burritoLover »

Alchemist wrote: Fri Oct 15, 2021 7:12 am In December 2020 the P/E ratio of the S&P 500 was at a shockingly high 39.26. The 10 year treasury was yielding 0.93%. Many posters on Bogleheads were warning of an epic bubble. We were at valuations not seen since dotcom irrational exuberance days, surely the market could go no further.

Since then the PE ratio of the S&P 500 tumbled 29% to 27.96 today (14 Oct 2021). The 10 year treasury is now over 1.5%.

Surely that means a big tumble in the stock market right? Quite the contrary, Vanguard Total Stock Market Index Fund (VTSAX) is up 18.95% year to date and the Vanguard 500 Index Fund (VFIAX) is up 19.48%.

What happened? As myself and some other Bogleheads pointed out time and again, the P/E ratio has two variables. The "E" can also go up and that is what happened. The 12 month forward P/E estimates from December 2020 were in the mid-20's thanks to what was expected to be a strong economic turn around. If earnings came in around expectation, the market was appropriately valued. Thus far earnings have been exceeding expectations. Therefore prices have actually gone up while valuations have come down.

My point is not to say I told you so (well not entirely :wink: ). Rather it is to provide another example of how dangerous it can be to rely on squiggly lines in a chart to predict the future. Valuations really do matter but they exist within a broader context. A P/E of 40 while bonds are yielding 5-7% is probably not a good sign. A P/E of 39 with trillions of dollars of pent up demand ready to be deployed as a solution to a global crisis suddenly becomes available may not be anything to worry about at all.

If you predict the past you are likely to be very surprised by the future.
So basically you are claiming that you know that the U.S. is NOT in bubble (which is a prediction of the future) because bond yields are low relative to P/E of the broad market? So, you are essentially saying that you can predict the future based on current P/E and bond yields.
Marseille07
Posts: 16054
Joined: Fri Nov 06, 2020 12:41 pm

Re: Valuations Matter....So Do Earnings

Post by Marseille07 »

burritoLover wrote: Fri Oct 15, 2021 11:29 am So basically you are claiming that you know that the U.S. is NOT in bubble (which is a prediction of the future) because bond yields are low relative to P/E of the broad market? So, you are essentially saying that you can predict the future based on current P/E and bond yields.
Sure, and the same goes for those claiming the US PE is too high and poised for a crash.
azanon
Posts: 3142
Joined: Mon Nov 07, 2011 9:34 am

Re: Valuations Matter....So Do Earnings

Post by azanon »

randomguy wrote: Fri Oct 15, 2021 11:29 am
azanon wrote: Fri Oct 15, 2021 10:43 am The OP has the wrong link for Valuation. Shiller PE is actually here: http://www.econ.yale.edu/~shiller/data.htm Shiller PE in Dec 2020 was 33.77 and now (2021.09) it's 38.34, so definitely higher than Dec 2020. I say wrong because P/E 1 is next to useless.

And of course "earnings also matter" - it's one of the two variables in Shiller PE, granted it's smoothed 10 year earnings, and I'll not go into why 10 years is used instead of the (near useless) 1 yr earnings.

Is 38 and change a bubble for Shiller PE? I think so considering it's only been higher than that for 1.5 years or so of 150 years of history.
And of those 150 years, how many times was E calculated using the same rules that apply in 2020? How much different is E today versus in say in 1990?

At a certain level we have a problem where we are either in like a 30 year stock market bubble (The average PE10 before 1990 was like 12. It is up to 15 these days AND we have barely touched that for more than short period in 2009) or something has changed to support higher PEs. Or a mixture of both.
It could be you know something more than I do. I assumed professor Shiller used the same formula for the entire set of data which starts in 1871.

Even if there's a higher mean, I hope I can safely presume you wouldn't estimate it to be 38. Maybe instead of the actual mean of ~ 17, it's now 21 or 22. Still, 38's a long ways away from either.
marcopolo
Posts: 8446
Joined: Sat Dec 03, 2016 9:22 am

Re: Valuations Matter....So Do Earnings

Post by marcopolo »

azanon wrote: Fri Oct 15, 2021 11:43 am
randomguy wrote: Fri Oct 15, 2021 11:29 am
azanon wrote: Fri Oct 15, 2021 10:43 am The OP has the wrong link for Valuation. Shiller PE is actually here: http://www.econ.yale.edu/~shiller/data.htm Shiller PE in Dec 2020 was 33.77 and now (2021.09) it's 38.34, so definitely higher than Dec 2020. I say wrong because P/E 1 is next to useless.

And of course "earnings also matter" - it's one of the two variables in Shiller PE, granted it's smoothed 10 year earnings, and I'll not go into why 10 years is used instead of the (near useless) 1 yr earnings.

Is 38 and change a bubble for Shiller PE? I think so considering it's only been higher than that for 1.5 years or so of 150 years of history.
And of those 150 years, how many times was E calculated using the same rules that apply in 2020? How much different is E today versus in say in 1990?

At a certain level we have a problem where we are either in like a 30 year stock market bubble (The average PE10 before 1990 was like 12. It is up to 15 these days AND we have barely touched that for more than short period in 2009) or something has changed to support higher PEs. Or a mixture of both.
It could be you know something more than I do. I assumed professor Shiller used the same formula for the entire set of data which starts in 1871.

Even if there's a higher mean, I hope I can safely presume you wouldn't estimate it to be 38. Maybe instead of the actual mean of ~ 17, it's now 21 or 22. Still, 38's a long ways away from either.
Prof. Shiller does not calculate E, it is an input. It is calculated by the companies, following a set of rules. Those rules have been changed multiple times over the years, mostly resulting in lower E than would have been reported under older rules.

Regarding why 10 years, I know you don't want to go into it, but it is not anything magical, it's just what produced the best results (best matched returns) back when Prof Shiller performed his data mining.
Once in a while you get shown the light, in the strangest of places if you look at it right.
Thesaints
Posts: 5108
Joined: Tue Jun 20, 2017 12:25 am

Re: Valuations Matter....So Do Earnings

Post by Thesaints »

azanon wrote: Fri Oct 15, 2021 10:43 am The OP has the wrong link for Valuation. Shiller PE is actually here: http://www.econ.yale.edu/~shiller/data.htm Shiller PE in Dec 2020 was 33.77 and now (2021.09) it's 38.34, so definitely higher than Dec 2020. I say wrong because P/E 1 is next to useless.
Actually, it is all the opposite. P/E10 is only useful in a stationary economy, where all companies continue to do the very same activities and nothing new happens. In that case, there is a constant average to which deviations are bound to revert.

In real life economies, that average is ever changing and P/E10 is not worth the cost of the bandwidth necessary to upload it on your PC.

And of course "earnings also matter" - it's one of the two variables in Shiller PE, granted it's smoothed 10 year earnings, and I'll not go into why 10 years is used instead of the (near useless) 1 yr earnings.
The most useful thing is of course the 10-year earnings, but it is 10 years in the future, not in the past.
Since we can't get those, proj-P/E-proj or ttm-P/E will have to do.
Again, unless you think the last 10 years of earnings are somewhat representative of the next 10 years of earnings, P/E10 is quite useless.


As the OP is concerned, everybody knows (or should know) that there are two ways an abnormal P/E can become normal.
azanon
Posts: 3142
Joined: Mon Nov 07, 2011 9:34 am

Re: Valuations Matter....So Do Earnings

Post by azanon »

marcopolo wrote: Fri Oct 15, 2021 11:56 am
azanon wrote: Fri Oct 15, 2021 11:43 am
randomguy wrote: Fri Oct 15, 2021 11:29 am
azanon wrote: Fri Oct 15, 2021 10:43 am The OP has the wrong link for Valuation. Shiller PE is actually here: http://www.econ.yale.edu/~shiller/data.htm Shiller PE in Dec 2020 was 33.77 and now (2021.09) it's 38.34, so definitely higher than Dec 2020. I say wrong because P/E 1 is next to useless.

And of course "earnings also matter" - it's one of the two variables in Shiller PE, granted it's smoothed 10 year earnings, and I'll not go into why 10 years is used instead of the (near useless) 1 yr earnings.

Is 38 and change a bubble for Shiller PE? I think so considering it's only been higher than that for 1.5 years or so of 150 years of history.
And of those 150 years, how many times was E calculated using the same rules that apply in 2020? How much different is E today versus in say in 1990?

At a certain level we have a problem where we are either in like a 30 year stock market bubble (The average PE10 before 1990 was like 12. It is up to 15 these days AND we have barely touched that for more than short period in 2009) or something has changed to support higher PEs. Or a mixture of both.
Even if there's a higher mean, I hope I can safely presume you wouldn't estimate it to be 38. Maybe instead of the actual mean of ~ 17, it's now 21 or 22. Still, 38's a long ways away from either.
Regarding why 10 years, I know you don't want to go into it, but it is not anything magical, it's just what produced the best results (best matched returns) back when Prof Shiller performed his data mining.
What produced his "best results"? To clarify for the rest, I presume you mean 10 years was selected because it seemed to best smooth out the outliers. That's a good thing, and makes the data useful, and indirectly explains why P/E 1 is useless. P/E 1's are quite often.... outliers, relative to the entire data set.

Point me to your support for your claim that the changes in Earnings calculations would result in a lower value with each successive change. I trust you aren't going to think I'm just going to take your word for it.
azanon
Posts: 3142
Joined: Mon Nov 07, 2011 9:34 am

Re: Valuations Matter....So Do Earnings

Post by azanon »

Thesaints wrote: Fri Oct 15, 2021 12:12 pm
azanon wrote: Fri Oct 15, 2021 10:43 am The OP has the wrong link for Valuation. Shiller PE is actually here: http://www.econ.yale.edu/~shiller/data.htm Shiller PE in Dec 2020 was 33.77 and now (2021.09) it's 38.34, so definitely higher than Dec 2020. I say wrong because P/E 1 is next to useless.
Actually, it is all the opposite. P/E10 is only useful in a stationary economy, where all companies continue to do the very same activities and nothing new happens. In that case, there is a constant average to which deviations are bound to revert.

In real life economies, that average is ever changing and P/E10 is not worth the cost of the bandwidth necessary to upload it on your PC.

And of course "earnings also matter" - it's one of the two variables in Shiller PE, granted it's smoothed 10 year earnings, and I'll not go into why 10 years is used instead of the (near useless) 1 yr earnings.
The most useful thing is of course the 10-year earnings, but it is 10 years in the future, not in the past.
Since we can't get those, proj-P/E-proj or ttm-P/E will have to do.
Again, unless you think the last 10 years of earnings are somewhat representative of the next 10 years of earnings, P/E10 is quite useless.


As the OP is concerned, everybody knows (or should know) that there are two ways an abnormal P/E can become normal.
The superiority of P/E 10 as a valuation metric relative to P/E 1 is so widely documented, it's not worth my time to reinvent the wheel. It's beyond the level of debate.
afan
Posts: 8195
Joined: Sun Jul 25, 2010 4:01 pm

Re: Valuations Matter....So Do Earnings

Post by afan »

burritoLover wrote: Fri Oct 15, 2021 11:29 am ...you are essentially saying that you can predict the future based on current P/E and bond yields.
There have been studies that suggest this to be true. Bond yields and PE combined reportedly better than PE alone in predicting long term stock returns. But, again, not good enough for timing.

The changes in earnings definitions have also been discussed many times. Here the problem is that it is extremely difficult to construct what past earning would have been if reported under current rules. Companies reported earning based on the rules at the time. There is a consensus that earnings would have been lower if current rules had been used but no one knows how much lower.

One would like a study that went back a hundred years and determined what the current-rules E would have been for every company. Then calculate historical PEs comparable to current. Since the rules were different companies did not report and maybe did not even know the information needed to do this.

The low expected returns based on high valuations persist. It is just that no one knows what number to put on "high" valuations. Another reason not to worry about it.
We don't know how to beat the market on a risk-adjusted basis, and we don't know anyone that does know either | --Swedroe | We assume that markets are efficient, that prices are right | --Fama
JackoC
Posts: 4714
Joined: Sun Aug 12, 2018 11:14 am

Re: Valuations Matter....So Do Earnings

Post by JackoC »

asif408 wrote: Fri Oct 15, 2021 8:43 am
Alchemist wrote: Fri Oct 15, 2021 8:20 am My point is not that valuations, including the Shiller P/E, are meaningless. The point is that they are not some crystal ball and must be considered with a lot of other context. Anyone trying to make real world investing decisions based on valuations have lost out on a lot more gains than they have saved themselves from market drops.
For short term timing I agree, they are poor, valuations don't tell you anything about the next few years. That's why your example doesn't really tell us anything, it's been less than a year.

The bigger picture is that it depends on how the valuations are being used. If you're using valuations to switch between stocks and some other stocks or bonds and make short term timing based on weeks, months, or a couple of years that's a bad idea. If you use it to make occasional adjustments among your stock allocation that you plan to stick with for the next decade or two that seems like a perfectly valid option. You just have to have patience and settle for possibly several years of underperformance and/or potentially never seeing outperformance show up. No risk, no reward.

Certainly valuations of international and value stocks have looked better than domestic US and growth stocks for the last few years, and I haven't seen anyone here suggesting international value and emerging market value stocks are in a bubble. Those markets sell for P/E ratios in the low teens to single digits, if you are interested in P/Es. And interest rates in many of those countries are even lower than in the US. So if the US market is fairly valued adjusted for interest rates UK and Japanese stocks must be a screaming buy.
Using valuation to allocate is tricky at best. It's a worthless exercise using 'spot' P/E to make short term tactical changes, in which respect I think the whole thread is something of a straw man. Who actually says to do that?

In some other contexts valuation might actually be an input to allocation, for example in the case of US v international stocks. Not that US and international stocks have to have the same PE's (smoothed as per Schiller or otherwise) in 'the long run'. Not if different types of company predominate in various indexes, and they remain different. Right now that explains a good deal (not necessarily all) of CAPE differences across countries. If an index has a lot more *types of company* the market gives lower valuations to (commercial banks, traditional energy etc) then the index does. OTOH indirectly it's still relevant IMO. The fact that CAPE is so historically high in the US relative to peers does tell you, even with a 100% assumption 'the market always knows best', you're buying something different in the US v non-US indexes. And if you can easily buy more types of different things, that's more diversification, and diversification is hard to argue against generally.

And, valuation is absolutely relevant to estimating expected return for basic planning purposes. People who project the expected return of US stocks as the average of past returns are kidding themselves IMO. A good deal of recent decades' return is explained by valuations increasing. If valuations stop increasing (not even decrease) return will be lower. That's not directly a matter of allocation though, because allocation is about alternatives. A low expected return on US stocks doesn't mean to abandon them for bonds if the expected return on bonds is also much lower than historical realized (which it is). And as in 2nd paragraph US 'vs' international is most logically considered as a question of why would you not diversify? If the idea is to ditch US stocks altogether in favor of foreign ones that's much more questionable IMO. But expected return is basic to financial planning, and valuation is basic to expected return.
marcopolo
Posts: 8446
Joined: Sat Dec 03, 2016 9:22 am

Re: Valuations Matter....So Do Earnings

Post by marcopolo »

azanon wrote: Fri Oct 15, 2021 12:18 pm
marcopolo wrote: Fri Oct 15, 2021 11:56 am
azanon wrote: Fri Oct 15, 2021 11:43 am
randomguy wrote: Fri Oct 15, 2021 11:29 am
azanon wrote: Fri Oct 15, 2021 10:43 am The OP has the wrong link for Valuation. Shiller PE is actually here: http://www.econ.yale.edu/~shiller/data.htm Shiller PE in Dec 2020 was 33.77 and now (2021.09) it's 38.34, so definitely higher than Dec 2020. I say wrong because P/E 1 is next to useless.

And of course "earnings also matter" - it's one of the two variables in Shiller PE, granted it's smoothed 10 year earnings, and I'll not go into why 10 years is used instead of the (near useless) 1 yr earnings.

Is 38 and change a bubble for Shiller PE? I think so considering it's only been higher than that for 1.5 years or so of 150 years of history.
And of those 150 years, how many times was E calculated using the same rules that apply in 2020? How much different is E today versus in say in 1990?

At a certain level we have a problem where we are either in like a 30 year stock market bubble (The average PE10 before 1990 was like 12. It is up to 15 these days AND we have barely touched that for more than short period in 2009) or something has changed to support higher PEs. Or a mixture of both.
Even if there's a higher mean, I hope I can safely presume you wouldn't estimate it to be 38. Maybe instead of the actual mean of ~ 17, it's now 21 or 22. Still, 38's a long ways away from either.
Regarding why 10 years, I know you don't want to go into it, but it is not anything magical, it's just what produced the best results (best matched returns) back when Prof Shiller performed his data mining.
What produced his "best results"? To clarify for the rest, I presume you mean 10 years was selected because it seemed to best smooth out the outliers. That's a good thing, and makes the data useful, and indirectly explains why P/E 1 is useless. P/E 1's are quite often.... outliers, relative to the entire data set.

Point me to your support for your claim that the changes in Earnings calculations would result in a lower value with each successive change. I trust you aren't going to think I'm just going to take your word for it.
1) You might want to read a little about data mining to understand why it is not a good thing. Here is what Prof Shiller wrote at the end of one of his early papers describing the CAPE research:

"Our search over economic relations that us to study the price divided by 30-year moving average of earnings may have stumbled upon a chance relation with no significance. In other words, the relation studied here might be a spurious relation, the result of data mining. Neither the statistical tests nor the monte carlo experiments take account of the search over other possible relations.

It is also dangerous to assume that historical relations are necessarily applicable to the future. There could be fundamental structural changes occurring now that mean that the past of the stock market is no longer a guide to the future."


That quote is from a paper posted on his web site:
http://www.econ.yale.edu/~shiller/data/peratio.html (see last section at bottom of the page)


2) Here is an article that does a reasonably good job explaining the more recent changes:
https://www.philosophicaleconomics.com/2013/12/shiller/
Once in a while you get shown the light, in the strangest of places if you look at it right.
User avatar
HomerJ
Posts: 21282
Joined: Fri Jun 06, 2008 12:50 pm

Re: Valuations Matter....So Do Earnings

Post by HomerJ »

azanon wrote: Fri Oct 15, 2021 10:43 am Is 38 and change a bubble for Shiller PE? I think so considering it's only been higher than that for 1.5 years or so of 150 years of history.
Someone could have said the same thing in 1996 when Shiller PE went above 25. It had only been higher for 1 year or so of 120 years of history.

Yep, someone could have called it a bubble back then too.

Oh wait, Shiller himself did. He predicted 0% real return for the next 10 years, and instead we got like 6% real from 1996-2006, almost the historical average.

The problem is that the definition of "high" valuations keeps changing. 25 is nearly considered normal nowadays.

In any case, valuations don't matter because over the long term, it all evens out.

If high valuations predict lower returns going forward in short-term, and we actually get lower returns, then valuations will go down, and start predicting higher returns in the short-term, and in the long run, it all evens out to the historical average.

That's where the average comes from.
Last edited by HomerJ on Fri Oct 15, 2021 1:51 pm, edited 1 time in total.
"The best tools available to us are shovels, not scalpels. Don't get carried away." - vanBogle59
garlandwhizzer
Posts: 3565
Joined: Fri Aug 06, 2010 3:42 pm

Re: Valuations Matter....So Do Earnings

Post by garlandwhizzer »

I believe the most important things for investors are simply to stay invested in a well-thought-out portfolio through thick and thin for many years, to reduce market timing portfolio changes to the minimum achievable, to keep investing regularly like clockwork paying as little attention as possible to the talking heads, market forecasts, or other inputs that pretend to know where the market is going in the short or intermediate term. Most of the time, it is reasonable to assume that the market knows more about asset pricing than you do as an individual. The thing that leads to long term investing success is not being an oracle who consistently foresees the market's future--no such individuals have ever lived--but instead being a consistent plodder who develops a sound strategy and sticks to it regardless of which way the current market wind is blowing.

The key is finding the portfolio that is right for the individual, a portfolio that takes as much risk as you can realistically tolerate but does not exceed that level. In the past you could gain significant long term real inflation adjusted returns risk-free with quality bonds. Those days appear to be behind us. Going forward for expected positive inflation adjusted returns, there is no alternative to taking on risk. The traditional 60/40 may not be the optimal solution for all.

Garland Whizzer
Thesaints
Posts: 5108
Joined: Tue Jun 20, 2017 12:25 am

Re: Valuations Matter....So Do Earnings

Post by Thesaints »

azanon wrote: Fri Oct 15, 2021 12:19 pm The superiority of P/E 10 as a valuation metric relative to P/E 1 is so widely documented, it's not worth my time to reinvent the wheel. It's beyond the level of debate.
That's ridiculous. Just look on a plot of P/E10 and P/E1, which of the two has more peaks corresponding to more crashes and recessions and check how far the peaks are from the beginning of the recession.
User avatar
Forester
Posts: 2400
Joined: Sat Jan 19, 2019 1:50 pm
Location: UK

Re: Valuations Matter....So Do Earnings

Post by Forester »

Nothing has changed since December 2020, the S&P 500 was due a subdued 2020s then, and it is due a subdued 2020s now.
Amateur Self-Taught Senior Macro Strategist
marcopolo
Posts: 8446
Joined: Sat Dec 03, 2016 9:22 am

Re: Valuations Matter....So Do Earnings

Post by marcopolo »

Forester wrote: Fri Oct 15, 2021 4:05 pm Nothing has changed since December 2020, the S&P 500 was due a subdued 2020s then, and it is due a subdued 2020s now.
Well, except that the market will have to have a 20% worse outcome to be at the same "subdued" level than it would have a year ago.
Once in a while you get shown the light, in the strangest of places if you look at it right.
User avatar
Vulcan
Posts: 2996
Joined: Sat Apr 05, 2014 11:43 pm

Re: Valuations Matter....So Do Earnings

Post by Vulcan »

Eight years ago I wrote this note to future me after reading some thread on bogleheads about how the sky was about to fall because of PE something or other.

Then I simply stayed the course :beer

Image
If you torture the data long enough, it will confess to anything. ~Ronald Coase
User avatar
Vulcan
Posts: 2996
Joined: Sat Apr 05, 2014 11:43 pm

Re: Valuations Matter....So Do Earnings

Post by Vulcan »

marcopolo wrote: Fri Oct 15, 2021 12:56 pm "Our search over economic relations that us to study the price divided by 30-year moving average of earnings may have stumbled upon a chance relation with no significance. In other words, the relation studied here might be a spurious relation, the result of data mining.
In other words... see signature :sharebeer
If you torture the data long enough, it will confess to anything. ~Ronald Coase
docL
Posts: 25
Joined: Sat Jul 10, 2021 8:25 am

Re: Valuations Matter....So Do Earnings

Post by docL »

MarkRoulo wrote: Fri Oct 15, 2021 10:37 am
JoMoney wrote: Fri Oct 15, 2021 7:35 am Back as of August 31, 2011, a portfolio of S&P 500 Growth Index stocks had a Price/Book value of 3.1x (source: Vanguard fund annual report)

It was last reported as 10.0x (Vanguard's fund website)
Over 10 years, that's a +12% annualized/CAGR valuation change just in the multiple... that the underlying fundamentals did not keep up with.

Given the extremely low, and falling, interest rate environment, a business that is able to grow it's fundamental value intrinsically should demand a higher price... But call me skeptical that the underlying has the ability to maintain that multiple and fundamentals grow into it :annoyed
I don't think it is obvious at all that the book value of a company such as Microsoft or Google or Apple has a lot to do with the company's ability to make and grow profits.

Both Apple and Microsoft could double (or more) their product sales without needing to increase their factory footprint. Because the actual manufacturing is outsourced. iPhones are assembled by FoxConn, iPhone SoC chips are manufactured by TSMC, iPhone screens are manufactured by Samsung and LG and whoever it is now. Microsoft can crank out DVDs to install MS-Office without factories -- this, too, is outsourced. Or Microsoft can allow folks to download the software, but the servers needed are such a small part of the product development that Microsoft won't need to double their server farm footprint to sell x2 as much software.

Back when industry was making steel (or canning soup) you pretty much needed to double your factory to double your output, so price/book really mattered. If one can replicate the factory for ~1x or buy the factories of a company (via stock) for 3x one of these was a much better choice.

But Apple and Microsoft and Google (and Facebook, and ...) are not cranking out commodity widgets (like steel and soup). You can't look at one of these trading a high price/book ratio and go off and build an equivalent company for book value. The value isn't in the assets that are traditionally counted as book value.

200 years ago, *land* was what was valuable. Specifically agricultural land.

I'll note that the balance sheet for 1860s industrial companies tended to be very light in agricultural land, but the companies were valuable none the less. The accounting needed to change to reflect this.

I'm pretty sure that industrial age accounting (book value, etc.) doesn't capture much of the true value of modern tech companies (Microsoft, Google, Facebook ... but also smaller ones such as Applied Materials) and modern high-end service companies. Bain isn't valuable because of the buildings.

Which doesn't mean that these companies can't be overvalued. I really don't understand Tesla's valuation.

But pointing to book value for a company like Apple may well be missing where most of the accumulated value lies.

as you note, book value is nearly useless in for some companies/industries as it relates to any attempt at valuation. And it is less useful now than perhaps ever. Looking at how book value for an index has changed over time is not useful.
Topic Author
Alchemist
Posts: 638
Joined: Sat Aug 30, 2014 6:35 am

Re: Valuations Matter....So Do Earnings

Post by Alchemist »

azanon wrote: Fri Oct 15, 2021 10:43 am The OP has the wrong link for Valuation. Shiller PE is actually here: http://www.econ.yale.edu/~shiller/data.htm Shiller PE in Dec 2020 was 33.77 and now (2021.09) it's 38.34, so definitely higher than Dec 2020. I say wrong because P/E 1 is next to useless.

And of course "earnings also matter" - it's one of the two variables in Shiller PE, granted it's smoothed 10 year earnings, and I'll not go into why 10 years is used instead of the (near useless) 1 yr earnings.

Is 38 and change a bubble for Shiller PE? I think so considering it's only been higher than that for 1.5 years or so of 150 years of history.
I do not mention Shiller PE in the OP and neither did I intend to. I referenced the 12 month trailing PE explicitly and linked to that number.

In anycase I find Shiller PE/CAPE10 to be as useless as CAPE1 for making real world decisions.
asif408 wrote: Fri Oct 15, 2021 8:43 am Certainly valuations of international and value stocks have looked better than domestic US and growth stocks for the last few years, and I haven't seen anyone here suggesting international value and emerging market value stocks are in a bubble. Those markets sell for P/E ratios in the low teens to single digits, if you are interested in P/Es. And interest rates in many of those countries are even lower than in the US. So if the US market is fairly valued adjusted for interest rates UK and Japanese stocks must be a screaming buy.
My argument is that looking at valuations in isolation is useless and horribly misleading. Concluding that UK and Japanese stocks have great return prospects because of valuations is just the flip side of saying US returns look poor because of valuations. Much of the developed world's stock markets have lower valuations than the US market....and there are good structural reasons for that (demography, geopolitical risk, geography, access to energy markets, etc) better discussed in other threads. But none of those other reasons will show up in P/E whether it is a 12 month or 12 decade look back.
User avatar
JoMoney
Posts: 16260
Joined: Tue Jul 23, 2013 5:31 am

Re: Valuations Matter....So Do Earnings

Post by JoMoney »

docL wrote: Fri Oct 15, 2021 6:33 pm
MarkRoulo wrote: Fri Oct 15, 2021 10:37 am
JoMoney wrote: Fri Oct 15, 2021 7:35 am Back as of August 31, 2011, a portfolio of S&P 500 Growth Index stocks had a Price/Book value of 3.1x (source: Vanguard fund annual report)

It was last reported as 10.0x (Vanguard's fund website)
Over 10 years, that's a +12% annualized/CAGR valuation change just in the multiple... that the underlying fundamentals did not keep up with.

Given the extremely low, and falling, interest rate environment, a business that is able to grow it's fundamental value intrinsically should demand a higher price... But call me skeptical that the underlying has the ability to maintain that multiple and fundamentals grow into it :annoyed
I don't think it is obvious at all that the book value of a company such as Microsoft or Google or Apple has a lot to do with the company's ability to make and grow profits.
[...snipped...]
But pointing to book value for a company like Apple may well be missing where most of the accumulated value lies.
as you note, book value is nearly useless in for some companies/industries as it relates to any attempt at valuation. And it is less useful now than perhaps ever. Looking at how book value for an index has changed over time is not useful.
I was saying the same thing when the "growth" indices were in the 5-6x book value range. The multiple has expanded quite a bit since then. I don't disagree with the idea of it not being a good way to value something like an IT business with intellectual property and things that are less tangible... but when comparing the same businesses to themselves at a different point in time, the multiple has gone up quite a bit.
"To achieve satisfactory investment results is easier than most people realize; to achieve superior results is harder than it looks." - Benjamin Graham
marcopolo
Posts: 8446
Joined: Sat Dec 03, 2016 9:22 am

Re: Valuations Matter....So Do Earnings

Post by marcopolo »

JoMoney wrote: Fri Oct 15, 2021 8:15 pm
docL wrote: Fri Oct 15, 2021 6:33 pm
MarkRoulo wrote: Fri Oct 15, 2021 10:37 am
JoMoney wrote: Fri Oct 15, 2021 7:35 am Back as of August 31, 2011, a portfolio of S&P 500 Growth Index stocks had a Price/Book value of 3.1x (source: Vanguard fund annual report)

It was last reported as 10.0x (Vanguard's fund website)
Over 10 years, that's a +12% annualized/CAGR valuation change just in the multiple... that the underlying fundamentals did not keep up with.

Given the extremely low, and falling, interest rate environment, a business that is able to grow it's fundamental value intrinsically should demand a higher price... But call me skeptical that the underlying has the ability to maintain that multiple and fundamentals grow into it :annoyed
I don't think it is obvious at all that the book value of a company such as Microsoft or Google or Apple has a lot to do with the company's ability to make and grow profits.
[...snipped...]
But pointing to book value for a company like Apple may well be missing where most of the accumulated value lies.
as you note, book value is nearly useless in for some companies/industries as it relates to any attempt at valuation. And it is less useful now than perhaps ever. Looking at how book value for an index has changed over time is not useful.
I was saying the same thing when the "growth" indices were in the 5-6x book value range. The multiple has expanded quite a bit since then. I don't disagree with the idea of it not being a good way to value something like an IT business with intellectual property and things that are less tangible... but when comparing the same businesses to themselves at a different point in time, the multiple has gone up quite a bit.
Isn't that what you would expect from any business that derives it's growth from non-tangible assets like intellectual property, software, etc.? Profits and market capitalization grows dramatically, while tangible "book value" only growth a little.
Once in a while you get shown the light, in the strangest of places if you look at it right.
User avatar
Random Musings
Posts: 6771
Joined: Thu Feb 22, 2007 3:24 pm
Location: Pennsylvania

Re: Valuations Matter....So Do Earnings

Post by Random Musings »

Speculation and leverage matters too. So does the Federal Reserve experiment of greatly expanding their balance sheet, which has helped fuel these activities. With rates so low, it will be interesting to see the impact of even, what he have considered in the past to be, small interest rate increases.

RM
I figure the odds be fifty-fifty I just might have something to say. FZ
User avatar
HomerJ
Posts: 21282
Joined: Fri Jun 06, 2008 12:50 pm

Re: Valuations Matter....So Do Earnings

Post by HomerJ »

marcopolo wrote: Fri Oct 15, 2021 4:38 pm
Forester wrote: Fri Oct 15, 2021 4:05 pm Nothing has changed since December 2020, the S&P 500 was due a subdued 2020s then, and it is due a subdued 2020s now.
Well, except that the market will have to have a 20% worse outcome to be at the same "subdued" level than it would have a year ago.
Except that I sold off a good chunk of that 20% gain, and I'll never lose it.

And even those who didn't sell off are still in a higher spot and can handle a downturn easier than before. Because a 20% buffer is a 20% buffer.

So when Forester says nothing has changed since December 2020, as usual, he is wrong.
"The best tools available to us are shovels, not scalpels. Don't get carried away." - vanBogle59
nzahir
Posts: 226
Joined: Mon Jul 27, 2020 11:44 pm

Re: Valuations Matter....So Do Earnings

Post by nzahir »

HomerJ wrote: Fri Oct 15, 2021 1:05 pm
azanon wrote: Fri Oct 15, 2021 10:43 am Is 38 and change a bubble for Shiller PE? I think so considering it's only been higher than that for 1.5 years or so of 150 years of history.
Someone could have said the same thing in 1996 when Shiller PE went above 25. It had only been higher for 1 year or so of 120 years of history.

Yep, someone could have called it a bubble back then too.

Oh wait, Shiller himself did. He predicted 0% real return for the next 10 years, and instead we got like 6% real from 1996-2006, almost the historical average.

The problem is that the definition of "high" valuations keeps changing. 25 is nearly considered normal nowadays.

In any case, valuations don't matter because over the long term, it all evens out.

If high valuations predict lower returns going forward in short-term, and we actually get lower returns, then valuations will go down, and start predicting higher returns in the short-term, and in the long run, it all evens out to the historical average.

That's where the average comes from.
And what were the earnings from 1996-2003?

Or what happens when you extend this to 2008?

When prices are high in CAPE there will be periods when you can capture them when they are low

The same reason why Small Cap Value has crushed Large Cap Growth throughout history. Being cheaper and undervalued makes it easier to grow
User avatar
HomerJ
Posts: 21282
Joined: Fri Jun 06, 2008 12:50 pm

Re: Valuations Matter....So Do Earnings

Post by HomerJ »

nzahir wrote: Fri Oct 15, 2021 11:35 pm
HomerJ wrote: Fri Oct 15, 2021 1:05 pm
azanon wrote: Fri Oct 15, 2021 10:43 am Is 38 and change a bubble for Shiller PE? I think so considering it's only been higher than that for 1.5 years or so of 150 years of history.
Someone could have said the same thing in 1996 when Shiller PE went above 25. It had only been higher for 1 year or so of 120 years of history.

Yep, someone could have called it a bubble back then too.

Oh wait, Shiller himself did. He predicted 0% real return for the next 10 years, and instead we got like 6% real from 1996-2006, almost the historical average.

The problem is that the definition of "high" valuations keeps changing. 25 is nearly considered normal nowadays.

In any case, valuations don't matter because over the long term, it all evens out.

If high valuations predict lower returns going forward in short-term, and we actually get lower returns, then valuations will go down, and start predicting higher returns in the short-term, and in the long run, it all evens out to the historical average.

That's where the average comes from.
And what were the earnings from 1996-2003?

Or what happens when you extend this to 2008?
Returns from Jan 1996 - Jan 2006 for Total Stock Market was 9.07% nominal (6.40% real)

Returns from Jan 1996 - Jan 2003 for Total Stock Market was 6.18% nominal (3.72% real)

Returns from Jan 1996 - Jan 2009 for Total Stock Market was 4.75% nominal (2.25% real)

Returns from Jan 1996 - Jan 2016 for Total Stock Market was 8.23% nominal (5.92% real)

1996, despite having the highest valuations in U.S. history for the past 70 years, was still a good time to buy. Over the long-run, even buying at the highest valuations since 1929, right before the Great Depression, you still got very close to the historical average investing from 1996-2016.
When prices are high in CAPE there will be periods when you can capture them when they are low
The problem is that the definition of "high" CAPE keeps changing. 25 absolutely was considered "sky-high insane" in 1996. Now it's just a bit above normal. CAPE hasn't been "low" since it was discovered by Shiller.

33 years and it still haven't mean reverted. And it's had plenty of chances. Multiple recessions, two stock market crashes, a financial panic.

Now, you can say it's because the "Fed intervened" or because "accounting standards changed", and that's fine. Those things DID happen, so maybe that's why CAPE has been bad at predicting the future.

But all that is saying is that there are more variables than just Price and Earnings. Which is my point. There are far too many variables out there to think you can predict the future on CAPE. Especially since it's done such a terrible job since it was invented at predicting future returns.
"The best tools available to us are shovels, not scalpels. Don't get carried away." - vanBogle59
nzahir
Posts: 226
Joined: Mon Jul 27, 2020 11:44 pm

Re: Valuations Matter....So Do Earnings

Post by nzahir »

HomerJ wrote: Sat Oct 16, 2021 11:45 am
nzahir wrote: Fri Oct 15, 2021 11:35 pm
HomerJ wrote: Fri Oct 15, 2021 1:05 pm
azanon wrote: Fri Oct 15, 2021 10:43 am Is 38 and change a bubble for Shiller PE? I think so considering it's only been higher than that for 1.5 years or so of 150 years of history.
Someone could have said the same thing in 1996 when Shiller PE went above 25. It had only been higher for 1 year or so of 120 years of history.

Yep, someone could have called it a bubble back then too.

Oh wait, Shiller himself did. He predicted 0% real return for the next 10 years, and instead we got like 6% real from 1996-2006, almost the historical average.

The problem is that the definition of "high" valuations keeps changing. 25 is nearly considered normal nowadays.

In any case, valuations don't matter because over the long term, it all evens out.

If high valuations predict lower returns going forward in short-term, and we actually get lower returns, then valuations will go down, and start predicting higher returns in the short-term, and in the long run, it all evens out to the historical average.

That's where the average comes from.
And what were the earnings from 1996-2003?

Or what happens when you extend this to 2008?
Returns from Jan 1996 - Jan 2006 for Total Stock Market was 9.07% nominal (6.40% real)

Returns from Jan 1996 - Jan 2003 for Total Stock Market was 6.18% nominal (3.72% real)

Returns from Jan 1996 - Jan 2009 for Total Stock Market was 4.75% nominal (2.25% real)

Returns from Jan 1996 - Jan 2016 for Total Stock Market was 8.23% nominal (5.92% real)

1996, despite having the highest valuations in U.S. history for the past 70 years, was still a good time to buy. Over the long-run, even buying at the highest valuations since 1929, right before the Great Depression, you still got very close to the historical average investing from 1996-2016.
When prices are high in CAPE there will be periods when you can capture them when they are low
The problem is that the definition of "high" CAPE keeps changing. 25 absolutely was considered "sky-high insane" in 1996. Now it's just a bit above normal. CAPE hasn't been "low" since it was discovered by Shiller.

33 years and it still haven't mean reverted. And it's had plenty of chances. Multiple recessions, two stock market crashes, a financial panic.

Now, you can say it's because the "Fed intervened" or because "accounting standards changed", and that's fine. Those things DID happen, so maybe that's why CAPE has been bad at predicting the future.

But all that is saying is that there are more variables than just Price and Earnings. Which is my point. There are far too many variables out there to think you can predict the future on CAPE. Especially since it's done such a terrible job since it was invented at predicting future returns.
Thanks

This is all a sort of cherry picking data though as if you got 2008 or early 09, the returns are around 4% or less even

You are right that the CAPE has continued to move up and 25 was high back then. However we had seen around 30 or so during 1929

Sadly we won't know the answer for another few years at least or even longer

I believe that having a diversified portfolio US, Intl, EM, Small Cap value for some, good alts if possible (not available to most retail), and bonds for some and adding consistently is probably the best way to go

You are also right that the FED has continued to intervene and be accommodative. This has crushed the middle class though and brought high real inflation when you look at housing prices (not included), college, health care, food, entertainment, and eating at restaurants
Post Reply