Siegel on PE10 CAPE

[Lsbcal]

This article has links to a Financial Times article by Jeremy Siegel and also a longish video talk he gave: Jeremy Siegel: Stocks Are the Most Stable Asset Class in the Long Run | Enterprising Investor

Many have focused on the CAPE ratio to argue that stocks are richly valued, but the S&P 500 (SPX) earnings data suffer from an aggregation bias in which large losses from a few companies can offset the earnings of the rest of the index. The S&P 500 treats companies as if they were a single corporation with 500 divisions. Of course, this approach to constructing the S&P earnings is wrong, particularly when there are large losses generated by a small number of firms. Importantly, P/E ratios computed from the S&P earnings data set are problematic, and these problems are amplified when one uses CAPE to examine cycles.

There is more detail in the links. He also discusses some reasons he thinks arguments of high historical profit margins are overstating the case.

Anyway for those using PE10 as a compass, this sort of thinking has to be examined before rejecting it.

 

[Onward]

Touche!

 

[NW-Bound]

I watched the entire video. It's nice to see Siegel making his bullish case.

Regarding the "aggregation bias", it still bothers me a bit. It is true that a few companies' large losses such as AIG in 2008 knocked down the entire S&P earning quite a bit, but aren't these losses real? In other worlds, the indexers did lose that money, did they not?

I guess one can say that it was a one time event, and since the shenanigan associated with the housing bubble and subprime mortgages is not likely to be repeated and that the past is gone, we should not let the bad years earnings influenced the projection going forward.

But if we "edited" out the bad years where AIG and the likes of Bear Stearns went belly up, should we not also edit out the prior years when they made a lot of money, the money that they did not really deserve?

 

[Lsbcal]

I think Siegel did discuss that point about selective removal of AIG earnings and why he thought that it was OK to exclude it. But I cannot remember his argument. The video is 1 hour so it's not easy to find the quote.

He also mentions the changes in accounting in recent years which biased earnings in one direction. He showed the spike down in earnings for 2008 which was much worse the the 1930's Depression. When the FASB accounting standard was more equalized, that 2008 earnings appeared more in reasonable. At least that's how I recall it.

 

[karluk]

I'm not a PE10 disciple, but I don't disregard it either. I suppose it's as good a tool as any in mankind's never ending quest to know the unknowable, in this case the future direction of the stock market.

What I really believe is that all of these prognostications have limited predictive power and that, with the exception of extreme cases (such as the astronomically high valuations at the end of the tech bubble), they should all be taken with a grain of salt.

Take Jeremy Siegel in particular. The following link from February, 2012, quotes his hilariously wrong prediction from 2008, but then uses that bad prediction to question his 2012 forecast of DJIA 15k. As we all know with the benefit of 20/20 hindsight, Siegel was wrong in 2012 - by being too pessimistic! The DJIA as of today stands above 16k, NOT 15k.

So should Siegel's opinion of PE10 be taken seriously? I don't see why. He was dead wrong in 2008, but has been mostly right since then. I don't see a trend from all of this that allows us to evaluate the accuracy of his future predictions.

Kass: The 'Wisdom' of Jeremy Siegel - TheStreet

 

[NW-Bound]

...
He also mentions the changes in accounting in recent years which biased earnings in one direction. He showed the spike down in earnings for 2008 which was much worse the the 1930's Depression. When the FASB accounting standard was more equalized, that 2008 earnings appeared more in reasonable. At least that's how I recall it.

Yes, Siegel did make a point about FASB accounting rule changes making it harder to book gains, while losses had to be accounted for immediately.

I'm not a PE10 disciple, but I don't disregard it either. I suppose it's as good a tool as any in mankind's never ending quest to know the unknowable, in this case the future direction of the stock market.

What I really believe is that all of these prognostications have limited predictive power and that, with the exception of extreme cases (such as the astronomically high valuations at the end of the tech bubble), they should all be taken with a grain of salt.

Take Jeremy Siegel in particular. The following link from February, 2012, quotes his hilariously wrong prediction from 2008, but then uses that bad prediction to question his 2012 forecast of DJIA 15k. As we all know with the benefit of 20/20 hindsight, Siegel was wrong in 2012 - by being too pessimistic! The DJIA as of today stands above 16k, NOT 15k.

So should Siegel's opinion of PE10 be taken seriously? I don't see why. He was dead wrong in 2008, but has been mostly right since then. I don't see a trend from all of this that allows us to evaluate the accuracy of his future predictions.

Kass: The 'Wisdom' of Jeremy Siegel - TheStreet

On the other hand, Shiller did admit in one interview that he was too bearish, and missed out on getting more invested, in the recovery of 2009-2010 I believe.

 

[Lsbcal]

...(snip)...
So should Siegel's opinion of PE10 be taken seriously? I don't see why. He was dead wrong in 2008, but has been mostly right since then. I don't see a trend from all of this that allows us to evaluate the accuracy of his future predictions.
...http://www.thestreet.com/story/11417253/1/kass-the-wisdom-of-jeremy-siegel.html

We all know that predictions on the future direction of equities are fraught with danger. So we could easily ignore the predictions business.

My interest was in the data discrepancies Siegel pointed out in PE10. Those are issues more of fact and calculations.

 

[haha]

IMO Jeremy Siegal is a fraud and a hack. But the public will always love bulls, especially when markets are high.

The best quote from Kass' article:

"Dr. Siegel comes off as a very nice person, but he is an academic who has been bullish at some very wrong times. Importantly, his theories regarding equities for the long term have been wildly off, as bonds have outperformed stocks for one, five, 10, 30 and 40 years, which, according to his investment thesis, is impossible. "

 

[Texas Proud]

I watched the entire video. It's nice to see Siegel making his bullish case.

Regarding the "aggregation bias", it still bothers me a bit. It is true that a few companies' large losses such as AIG in 2008 knocked down the entire S&P earning quite a bit, but aren't these losses real? In other worlds, the indexers did lose that money, did they not?

I guess one can say that it was a one time event, and since the shenanigan associated with the housing bubble and subprime mortgages is not likely to be repeated and that the past is gone, we should not let the bad years earnings influenced the projection going forward.

But if we "edited" out the bad years where AIG and the likes of Bear Stearns went belly up, should we not also edit out the prior years when they made a lot of money, the money that they did not really deserve?

I did not watch the video, so do not know why they would edit out a bad year like and AIG.... but I would because as an investor there is only so much that you can lose...

IOW, once the stock goes to zero, it really does not matter how much more loss there is... so if a company can go BK with a $4 billion loss, does it make any difference if they have a $20 billion loss

But, attributing that $20B loss to all 500 would make you think there is a problem over a good number of them when in reality it is a few... and once those few are excluded (which they will be if they go belly up) then you can look at the remaining ones in aggregate....

 

[Lsbcal]

Yes, Siegel did make a point about FASB accounting rule changes making it harder to book gains, while losses had to be accounted for immediately. ...

Another thing I think Siegel is saying is that earnings should be normalized for the capitalization of the company in the SP500 index. It is a cap weighted index but if, for example, we take a big firm with somewhat lower earnings and a small firm with big negative earnings, we could get a negative number. This is the aggregation bias he mentions.

He did not mention this but maybe the calculation should be by individual company PE10 ratios. So if we have 2 companies the PE10 would be:
PE10 = a*PE10(a) + b*PE10(b)
where a and b are the cap percentages
and PE10(a) is the PE10 of company "a", PE10(b) is the PE10 of company "b"

Maybe this cannot easily be done because the 10 year earnings of the individual companies are not easily obtained? Just a guess.

 

[audreyh1]

I think Siegel did discuss that point about selective removal of AIG earnings and why he thought that it was OK to exclude it. But I cannot remember his argument. The video is 1 hour so it's not easy to find the quote.

He also mentions the changes in accounting in recent years which biased earnings in one direction. He showed the spike down in earnings for 2008 which was much worse the the 1930's Depression. When the FASB accounting standard was more equalized, that 2008 earnings appeared more in reasonable. At least that's how I recall it.

I think his argument was something like "if the S&P 500 were one corporation, that's not how it would do it's accounting, so it's not OK to aggregate the losses like that" but I don't really understand why that would be different than the earnings.

 

[Lsbcal]

IMO Jeremy Siegal is a fraud and a hack. But the public will always love bulls, especially when markets are high.

The best quote from Kass' article:

"Dr. Siegel comes off as a very nice person, but he is an academic who has been bullish at some very wrong times. Importantly, his theories regarding equities for the long term have been wildly off, as bonds have outperformed stocks for one, five, 10, 30 and 40 years, which, according to his investment thesis, is impossible. "

I think you are reacting to the predictions which get widely quoted.

I'm more interested in the calculation issues on PE10.

 

[karluk]

I think you are reacting to the predictions which get widely quoted.

I'm more interested in the calculation issues on PE10.

I'll let the academics argue about the merits of how PE10 is calculated. From an investor's point of view the real question is whether any of these supposed anomalies lessen its rather impressive history of correlation with long term stock market performance. The answer to me is a clear cut "maybe, mabe not". Ask me again in ten years.

I note, for example, that the Wilcox article in the CFA Digest appears to be data mining various alternatives to PE10 in an attempt to explain away the discrepancy from July, 2011, when PE10 said that stocks were overvalued, whereas S&P 500 earnings indicated that stocks were undervalued. Wilcox says that maybe we should be using PE6 instead of PE10, because it reduces the valuation discrepancy by 25%.

I personally don't have any problem reconciling the two measures from 2011. Why can't stocks be a good short term buy, but a bad (or at least not so great) long term hold?

 

[NW-Bound]

I did not watch the video, so do not know why they would edit out a bad year like and AIG.... but I would because as an investor there is only so much that you can lose...

IOW, once the stock goes to zero, it really does not matter how much more loss there is... so if a company can go BK with a $4 billion loss, does it make any difference if they have a $20 billion loss

But, attributing that $20B loss to all 500 would make you think there is a problem over a good number of them when in reality it is a few... and once those few are excluded (which they will be if they go belly up) then you can look at the remaining ones in aggregate....

The idea of CAPE is that you take the earnings over the past 10 years to average out bad years against good years, as CA stands for "Cyclically Adjusted".

So, if you throw out the loss due to a shenanigan like AIG in a bad year, you should also throw out the somewhat illicit gains it made in the prior years. That would bring down the average earning in the denominator of the PE10, which makes the PE10 value higher than if you let the good and unsustainable years stay.

Another thing I think Siegel is saying is that earnings should be normalized for the capitalization of the company in the SP500 index. It is a cap weighted index but if, for example, we take a big firm with somewhat lower earnings and a small firm with big negative earnings, we could get a negative number. This is the aggregation bias he mentions.

He did not mention this but maybe the calculation should be by individual company PE10 ratios. So if we have 2 companies the PE10 would be:
PE10 = a*PE10(a) + b*PE10(b)
where a and b are the cap percentages
and PE10(a) is the PE10 of company "a", PE10(b) is the PE10 of company "b"

Maybe this cannot easily be done because the 10 year earnings of the individual companies are not easily obtained? Just a guess.

I do not know how exactly how the S&P index and its earning is constructed from the individual companies, but the above would make sense.

 

[bmcgonig]

Meb Faber has a nice chart comparing CAPEs to CAPDs. Basically says that They both agree

http://www.mebanefaber.com/2013/11/07/are-stocks-overvalued-cape-vs-dividends/

 

[NW-Bound]

... Why can't stocks be a good short term buy, but a bad (or at least not so great) long term hold?

It can be, depending on whom you ask.

The problem is that if a person says yes, then what should he do? He would be buying when he perceives that stocks are cheap, then unloads when he thinks they are overvalued. That would be called market timing and frowned upon in some circles.

 

[MooreBonds]

IMO Jeremy Siegal is a fraud and a hack. But the public will always love bulls, especially when markets are high.

The best quote from Kass' article:

"Dr. Siegel comes off as a very nice person, but he is an academic who has been bullish at some very wrong times. Importantly, his theories regarding equities for the long term have been wildly off, as bonds have outperformed stocks for one, five, 10, 30 and 40 years, which, according to his investment thesis, is impossible. "

Does anyone really expect the bond rate anomaly that existed in the early 80s to repeat itself? When you go from 16% long-term yields and - in a nearly linear path - go down to about 3.5% today, of course you'll have kickass returns that are difficult to match. However, I would be willing to bet that you will have a far easier time (relatively speaking) of beating bonds in any given period with stocks in future periods, given that the sky-high rates of the early 80s stand a good chance of never being seen again.

So if that result of bonds beating stocks on every time period up to 40 years due to the constant interest rate decline is a given, then surely supporters of that should all agree that stocks are nearly guaranteed to beat bonds for the next 40 years, since there's a pretty good chance of the exact inverse interest rate movement going forward as we saw from 1980 to now (or, hopefully, just from 1990 to now)

Plus, don't forget the MAJOR difference in net after-tax returns of bond interest vs equity qualified dividends and capital gains (starting around 2000). Paying twice as much in taxes on interest compared to qualified dividends/cap gains can really impact your net return.

 

[Lsbcal]

This chart shows sometimes bonds keep up with stocks. But gaps do seem to grow. Interestingly bills seem to have gone flattish since the 1930's.

 

[MasterBlaster]

I can't help but remember the Glassman and Hassett book "Dow 36000" from back in 1999.

you can buy it now for a penny on Amazon.

Dow 36, 000: The New Strategy for Profiting from the Coming Rise in the Stock Market: James K. Glassman, Kevin A. Hassett: 9780609806999: Amazon.com: Books

"This time it's different"

 

[NW-Bound]

This chart shows sometimes bonds keep up with stocks. But gaps do seem to grow. Interestingly bills seem to have gone flattish since the 1930's.

This chart makes a good point that stocks beat everything else in the long run.

But, but, but it is really in the looong run. As an ER who has perhaps 20 years left, 30 if I am really lucky, I worry more about the immediate decades.

Look at 1980-2000, where both stocks and bonds shone. No wonder many of us could afford to ER. Was that due to smart or luck?

On the other hand, if one squints and looks closely at the interval of 1960-1980, he will see that both stock and bond were either negative or nearly flat over that period. We would be all dead meat if that condition repeats. There's not much help in trying to balance between two lousy investments.

But look at gold during the above time period. No wonder people had been piling into gold the last few years. Gold has been retreating, and that perhaps is bearer of good news.

 

[HFWR]

Past performance is not indicative of future results...

 

[NW-Bound]

Which past performance will we see in the future?

I'd say all of it. Surely, stocks and bonds will fluctuate as they have always done.

 

[Running_Man]

Does anyone really expect the bond rate anomaly that existed in the early 80s to repeat itself? When you go from 16% long-term yields and - in a nearly linear path - go down to about 3.5% today, of course you'll have kickass returns that are difficult to match. However, I would be willing to bet that you will have a far easier time (relatively speaking) of beating bonds in any given period with stocks in future periods, given that the sky-high rates of the early 80s stand a good chance of never being seen again.

So if that result of bonds beating stocks on every time period up to 40 years due to the constant interest rate decline is a given, then surely supporters of that should all agree that stocks are nearly guaranteed to beat bonds for the next 40 years, since there's a pretty good chance of the exact inverse interest rate movement going forward as we saw from 1980 to now (or, hopefully, just from 1990 to now)

Plus, don't forget the MAJOR difference in net after-tax returns of bond interest vs equity qualified dividends and capital gains (starting around 2000). Paying twice as much in taxes on interest compared to qualified dividends/cap gains can really impact your net return.

I agree with HA that Jeremy Siegel is merely someone who will promote a bull case by utilitzing and changing any data necessary to prove his case. The illustration of how stock market data was created for 200 years for his book and analyzed as to the process is very informative. Notice how the removal of bad performing stocks was allowed both in the 19th century as well as the last decade to prove better performance than what really occured.

In a WSJ article this morning, Jason Zweig puts together a pretty compelling critique Does Stock-Market Data Really Go Back 200 Years? :

“There is just one problem with tracing stock performance all the way back to 1802: It isn’t really valid.
Prof. Siegel based his early numbers on data first gathered decades ago by two economists, Walter Buckingham Smith and Arthur Harrison Cole.
For the years 1802 through 1820, Profs. Smith and Cole collected prices on three dozen banking, insurance, transportation and other stocks — but ended up including only seven, all banks, in their stock-market index. Through 1845, they tracked 19 insurance stocks, but rejected 95% of them, adding only one to their index. For 1834 onward, they added a maximum of 27 railroad stocks.
To be a good measure of stock returns, an index should be comprehensive (by including many stocks) and representative (by including the stocks commonly held by investors). The Smith and Cole indexes are neither, as the professors signaled in their 1935 book, “Fluctuations in American Business.” They cherry-picked their indexes by throwing out any stock that didn’t survive for the whole period, whose share prices were too hard to find or whose returns seemed “inflexible,” “erratic,” or “non-typical.”
​

Thus, Siegel’s basis for Stocks for the Long Run exclude 97% of all the stocks in the early history of the US market by cherry picking winners, ignoring survivorship bias, and engaging in data smoothing.

 

[Lsbcal]

I really don't worry about the pre-1920's data. The economy seems to get the less like ours the further we go back. So the Siegel data going back that far isn't very interesting to me.

What is interesting in the Siegel graph is the behavior of bonds after 1930. But still this doesn't help me guess much at the future.

Can we even say the 1930's world economy is relevant to the current one? Maybe, I don't really know the answer to that.