Using Shiller PE to Time the Market

[Texas Proud]

That sounds a lot like "it's going to be different this time". It might be different, but it seems like the rules of economics are still basically in place. Balance sheet assets are certainly different than intellectual capital, and widgets, different than services, but either way, they've got to be making a profit in the long run or they will fail.

I actually think it is different now than before... companies are coming out with valuations over $50 billion with few assets and low earnings... back when I was young, a market cap that big was a BIG company...... and had profits to back it up or paid the price...

Amazon has a market cap of $236 billion and does not make money....
Tesla has a market cap higher (almost $32b) than the regular auto companies and does not make a profit... heck, the are burning through cash like a dot com company....

There are many other companies that can be put into this category... they are hoping for a payday in the future like Google did...

It used to be hard for a company to grow so fast... now it seems like a company can go from nothing to $20B, $50B or even $100B valuation in less than a decade...

I heard that the total S&P 500 PE was around 15... that is not out of line... so even with these high flyers, the market does not seem overpriced...

Now, it does not mean there will not be a drop of 10% to 20%.... that can happen due to the Fed raising rates... but, since a timing method has to have two right calls, a sell and a buy or a buy and a sell, I will just live with what I know works for me....


OH, I also heard that the vast majority of gains came from 90 trading days over the past 10 years.... if you were out of the market during those days you were basically flat... another reason to just hold....

 

[34rlsa]

If anyone is interested... the links below provide current PE Ratios

S&P 500 PE Ratio

Current S&P 500 PE Ratio: 19.66 +0.02 (0.12%)
4:05 pm EDT, Thu Sep 3

Shiller PE Ratio

Current Shiller PE Ratio: 24.74 +0.03 (0.12%)
4:05 pm EDT, Thu Sep 3

 

[sengsational]

Amazon has a market cap of $236 billion and does not make money....
Tesla has a market cap higher (almost $32b) than the regular auto companies and does not make a profit... heck, the are burning through cash like a dot com company....

As you say, people are hoping that some day the ship will come in for these guys. It might. People are more patient than I am with these kinds of companies. By default, I own some of them as part of the indexes, but if I had my druthers, I'd not own them....too much of a gamble. I'd rather buy something that has a scrap value something close to the market cap, but I suppose I'm being old-school.

OH, I also heard that the vast majority of gains came from 90 trading days over the past 10 years.... if you were out of the market during those days you were basically flat... another reason to just hold....

I think that's true. Probably the majority of the losses came in a relative few days too, though. If, by some magic, someone told me with 100% certainty that this PE 10 trading scheme was not overfit to the data, I'd use it for sure, and risk being out of the market on a few big up days. The scheme, in the past 134 years, has you in the market 82% of the time, so a good chance to catch most of the ups.

 

[Totoro]

If, by some magic, someone told me with 100% certainty that this PE 10 trading scheme was not overfit to the data, I'd use it for sure, and risk being out of the market on a few big up days. The scheme, in the past 134 years, has you in the market 82% of the time, so a good chance to catch most of the ups.

Posted this elsewhere, but I did an analysis based on the real (CAPE-10)/P. so CAPE-10 inverted - inflation (yoy). Basically it bugged me that inflation is never taken into account and I look at equities as an infinite-life junior bond (with interest = earnings yield).

What I noticed: virtually anytime it drops below 0%, big trouble happened. The indicator stays fully invested roughly 90% of the time.

What it did (little bit simplified)
* It avoided a big part of the 2008 meltdown
* went out of the market in a flat 2005
* completely sidestepped the late 1999 meltdown, to re-enter mid 2001
* got out mid 1979, back in early 1981. Missed out on a 30% rise (bonds yielded 10% or so, so net 20% 'loss')- so a false flag
* got out mid 1973, back in mid 1975. Missed a 20% drop
* Avoided a small drop 15% from late 1968 to early 1970
* Got out first half of 1951, but nothing really happened (flat market)
* Ran away mid 1946, came back in mid 1948. Avoided a 10% drop.
* Something similar late 1941 to mid 1942

What it didn't do is avoid the 1929 meltdown because of deflation.

Depending on the periods you look at (e.g. 10y or 20y return, only look at more recent decades or everything since 1925) the net effect of all that is a 2% to 4% p.a. higher return. With less volatility (since it avoids major drops).

Not bad.

Is this overfitting? an artifact of chance? I don't know.

What I do know is that real earnings yield as a concept makes sense to me, just like it does in bonds. And that it's not so wise to invest in instruments with a real negative yield when there are other options.

I've been thinking to redo this analysis for other stock markets, see if it holds up there. Japan for instance would be interesting. Europe too.

[Edit]: since inflation right now is close to flat, the indicator stays in with a rather safe margin.

 

[gcgang]

What is the Shiller PE now? I just viewed an interesting presentation showing Shiller PE ranges for the last 100 or so years. 1999 was the highest reading ever, at 44. I think it was down to 26 at the beginning of the year. Btw, 26 is still higher than most previous market tops.

The presenter also showed market yield. The YE 2014 1.9% was the lowest on record, I think.


Sent from my iPhone using Early Retirement Forum

 

[samclem]

What is the Shiller PE now?

See post 127.

 

[gcgang]

See post 127.

Duh. Thanks. Don't know how I missed that.


Sent from my iPhone using Early Retirement Forum

 

[Running_Man]

As you say, people are hoping that some day the ship will come in for these guys. It might. People are more patient than I am with these kinds of companies. By default, I own some of them as part of the indexes, but if I had my druthers, I'd not own them....too much of a gamble. I'd rather buy something that has a scrap value something close to the market cap, but I suppose I'm being old-school.

I think that's true. Probably the majority of the losses came in a relative few days too, though. If, by some magic, someone told me with 100% certainty that this PE 10 trading scheme was not overfit to the data, I'd use it for sure, and risk being out of the market on a few big up days. The scheme, in the past 134 years, has you in the market 82% of the time, so a good chance to catch most of the ups.

Yes it may be TECHNICALLY TRUE but not true in any sort of logical way. In 2008 if you missed October 13th and October 28 you missed 22.37% gains in the S&P500. Now the opportunity that may have occurred is that by December the stock was 20% lower than it was on October 12th so if you invested on December 31st you would have gained 20% more than anyone that was in the market for the 2 largest point gains in the history of the S&P500.

As a matter of fact 2008 saw 8 of the highest 20 single gains in the history of the S&P500 and if you missed those days you missed a total of 54.28% gains, but 2008 actually was a loss of 35% so anything that uses single days as a basis for being fully invested at all times is a red herring.

 

[Totoro]

The presenter also showed market yield. The YE 2014 1.9% was the lowest on record, I think.

Assuming you are referring to dividend yield. Late 1999 was lowest: 1.2%.

1.8% is actually "middle of the road" if you look at the past 20 years. And given the low inflation, as well as increased buybacks, it's actually better than that.

 

[sengsational]

... so anything that uses single days as a basis for being fully invested at all times is a red herring.

Agreed. If the big up day was preceded by a big down day, a long-term reallocation scheme would have been out for both, or in for both, so a wash. A short-term market timer can be killed by that effect, but for a scheme that was out of the market 5 times in 135 years (counting this latest sell) can only get caught with that pitfall a handful of times, and they're small potatoes in the big picture.

 

[target2019]

What is the Shiller PE now? I just viewed an interesting presentation showing Shiller PE ranges for the last 100 or so years. 1999 was the highest reading ever, at 44. I think it was down to 26 at the beginning of the year. Btw, 26 is still higher than most previous market tops.

The presenter also showed market yield. The YE 2014 1.9% was the lowest on record, I think.

This link has the data you want.
Shiller PE Ratio

 

[Mulligan]

If anyone is interested... the links below provide current PE Ratios



S&P 500 PE Ratio



Current S&P 500 PE Ratio: 19.66 +0.02 (0.12%)

4:05 pm EDT, Thu Sep 3



Shiller PE Ratio



Current Shiller PE Ratio: 24.74 +0.03 (0.12%)

4:05 pm EDT, Thu Sep 3

Kind of sobering stat isn't it. We have had a recent nice market drop off and it is still comfortably above mean average. And to think it is only up what about 400 points since 2000 peak?


Sent from my iPad using Tapatalk

 

[sengsational]

Kind of sobering stat isn't it. We have had a recent nice market drop off and it is still comfortably above mean average. And to think it is only up what about 400 points since 2000 peak?

Sobering indeed. 400 points up, but if you factor inflation, we basically bounced off of the 2000 peak earlier this year. That means we're down about 8% from the 2000 peak, yet still in 'overvalued' territory according to this metric.

 

[target2019]

The 30's were really rough. I got involved with this planet in the early 50's. Our investing started in the 80's. Something tells me we'll be alright until the next meteor hit.

https://public.dreyfus.com/documents/manual/sales-ideas/skyandex.pdf

 

[sengsational]

The 30's were really rough. I got involved with this planet in the early 50's. Our investing started in the 80's. Something tells me we'll be alright until the next meteor hit.

https://public.dreyfus.com/documents/manual/sales-ideas/skyandex.pdf

A busy chart, that one. If I were in accumulation phase, I don't know if I'd welcome a fall that put the PE 10 down to the mid teens, but I'd not be as concerned since I'd have an opportunity to buy low. Rebalancing is ok, but one must live long enough to benefit from the eventual recovery.

 

[Totoro]

I'd only be concerned if earnings dropped off on a structural basis as well.

If they don't dividend yield will shoot up (in % terms) if P/E falls that dramatically. And strong inflation should be contained as well, so don't see an issue there.

Then again, who knows.

 

[gcgang]

Surprisingly, Earnings were not as important as PE or yield in the presentation I viewed

http://youtu.be/q5UOo9xyvVY

Here's the link. An hour long, but I like this stuff, and found it very worthwhile.


Sent from my iPhone using Early Retirement Forum

 

[bmcgonig]

Interesting video of Shiller last week. He's not optimistic.

http://www.cnbc.com/2015/09/03/risk-of-big-stock-drops-grows-robert-shiller.html


Sent from my iPad using Early Retirement Forum

 

[Fred123]

Interesting video of Shiller last week. He's not optimistic.

Risk of big stock drops grows: Robert Shiller


Sent from my iPad using Early Retirement Forum

Schiller has been promising for nearly 20 years that the PE10 is going to drop below its long term median value of about 16. In the last 24 years, it's been below that level once -- in 2009. It probably will happen again some day, but being correct once in 25 years doesn't give me a lot of confidence that Schiller's PE10 is a good market timing tool.

 

[bmcgonig]

Schiller has been promising for nearly 20 years that the PE10 is going to drop below its long term median value of about 16. In the last 24 years, it's been below that level once -- in 2009. It probably will happen again some day, but being correct once in 25 years doesn't give me a lot of confidence that Schiller's PE10 is a good market timing tool.

In fairness to Shiller he emphasizes that it's not a market timing tool at all.


Sent from my iPad using Early Retirement Forum

 

[Totoro]

Here's the link. An hour long, but I like this stuff, and found it very worthwhile.

Thank you, just watched it this morning.

Not disagreeing with you or the presenter. Multiples have fluctuated wildly and are dominant on a 10 to 20 year scale. No reason to assume that will change as well.

As long as earnings stay intact however, and inflation doesn't go on a rampage (stays below 4% or so) we'll be fine, as dividend yield will go up.

Call me naïve, but I do trust most of the central banks (USA, Europe, Japan, China) to keep inflation within a reasonable range (-2% to +4%).

Unless a big asteroid hits. A real one.

 

[photoguy]

I'm not a market timer but valuation methods seem like the only sensible approach to me. However, given all the shifts in accounting rules, revenue recognition, earnings management, and increased restatements, I don't feel like I can trust the underlying data and trading rules based on that data.

My biggest fear would be that these factors cause a structural shift in valuations that cause one to permanently miss a runup.



Sent from my Nexus 5 using Early Retirement Forum mobile app

 

[samclem]

My biggest fear would be that these factors cause a structural shift in valuations that cause one to permanently miss a runup.

That's one reason I wouldn't use valuation to make a binary "all in" or "all out" decision. But using it to vary between high and low edges of one's equity "comfort window" (e.g. between allocating 40% to equities or 70% to equities) would seem appropriate and a good fit to the imprecise (esp regarding timing) nature of these PE10 indicators. It's possible that such an approach would allow many people to stay more fully invested than they'd otherwise be--rather than missing a runup, they'd participate more fully.

Regarding the usefulness of comparing PEs computed under today's rules to those of long ago--I don't know, but that's a good point. Maybe we should be making decisions based on a PE10 moving average of some more recent time period: long enough to avoid ratcheting up (too much) with "bubbles", but short enough to incorporate changes in accounting, etc that make earlier PEs less useful/appropriate for making decisions today.

 

[Texas Proud]

In fairness to Shiller he emphasizes that it's not a market timing tool at all.


Sent from my iPad using Early Retirement Forum

Then what good is it And why are people talking like it IS a market timing tool?


Finally, what does it mean.... if it means the market is overvalued, then why not market time...

 

[bmcgonig]

Then what good is it And why are people talking like it IS a market timing tool?


Finally, what does it mean.... if it means the market is overvalued, then why not market time...

Should have said "short term" timing tool. Seemingly good long term.


Sent from my iPad using Early Retirement Forum