This Is Where Sisyphus Starts Huffing and Puffing

Using data from Robert Schiller’s spreadsheet, it looks like the S&P 500 has a current PE of 15. The average PE for the 128 years Schiller has records is on the order of 16. So, for the momentum players we are below average, and that’s bullish.

Here’s my problem: it’s been 25 years since the PE for the S&P 500 has been in the single digits, and the important thing to remember about an average is sometimes the PE will go above the average and sometimes below, but usually there is regression to the mean.

Anyone looking at Schiller’s spreadsheet knows that the PE for the past 15 years has been well above 20, much less 16. And it hit an all time high in the 40s in 1999. Yet when the buzz ended in 2002, the hangover was not so horrible with the S&P500 sporting a PE above 21, with the PE above 25 for most of the rest of the decade.

There was another 25 year span where the PE stayed in single digits, from 1949 until 1974. I’m not saying that we can’t stay in the double digits longer, maybe we can. After all, from the point where Schiller starts computing the S&P’s PE in 1881, it stayed in double digits until 1917-- 36 years. And that includes the Great Depression of 1893, and the Panic of 1907.

But let’s review when the S&P fell into single digits:

• 1917-25, World War I and on into the twenties (8 years)
• 1932, the bottom of the Great Depression
• 1942, U.S. enters World War II
• 1949, U.S. enters Korean War
• 1974, the great recession
• 1977-84, the great inflation (7 years)

I don’t know what they will label the period we are living through, but GDP is falling, unemployment is rising, and housing starts are falling. GM and Chrysler are nearly out of business. It feels like a depression to me.

By the way, Sisyphus sounds like he would fit in with the CEO, trader crowd. Avaricious and deceitful, he was famous as the craftiest of men. He seduced his niece, stole his brother’s throne, and betrayed Zeus’s secrets. He felt he was a peer of the gods. When Zeus had had enough, he forced Sisyphus to roll a boulder up a steep hill, but before he could reach the top, it rolled down again. And to do this without end.

I’m thinking that the overhead resistance is going to get heavier and heavier the higher the averages take us.

Disclosure: I own shares in SDS.

Comments

[Larry House:]

Your conclusion seems right on, Thomas, but the market has a mind of its own. I am staying with positions a bit longer (probably too long). The trend is your friend until it isn't.

Jun 01 10:58 AM

[pslater:]

Once again P/E ratios are a function of interest rates. With the average non-financial company in the S&P 500 having a BAA credit rating, that rate (currently around 8%) implies a P/E of about 12.5 (100/8). It's no coincidence that in 1981 the P/E of the S&P was around 7 - because interest rates were around 13% (100/13 = 7.7).

If you want to know what an appropriate P/E ratio is use an appropriate interest rate. As for valuations, use your imagination as to what the earnings component should be....I would suggest that 10 year trailing earnings isn't it.

Jun 01 12:34 PM

[Fighting Yoda:]

What is the author talking about: "Using data from Robert Schiller’s spreadsheet, it looks like the S&P 500 has a current PE of 15".

As per S&P consensus data top down/operating earnings of $43 for '09 and $46 for '010; PE for '09 is 21 and 20 for '010. These are way way higher than historical averages of 16. An of course much higher than bear market PEs of around 10.

Jun 01 12:56 PM

[pslater:]

Not to mention that the 'As Reported' estimates for 2009 are about half of the 'Operating' estimates. From a valuation standpoint, it's really hard to make a case for anything much over 600....

<As per S&P consensus data top down/operating earnings of $43 for '09 and $46 for '010; PE for '09 is 21 and 20 for '010. These are way way higher than historical averages of 16. An of course much higher than bear market PEs of around 10.>

Jun 01 01:21 PM

[thiazole:]

I still think P/E is a worthless metric for valuation if any of the companies in the index are losing money. It means that the "value" of the index increases if you eliminate all the companies losing money. How does that make any sense? A stock can only go to zero, so it can't have a negative value to an investor.

Jun 01 03:02 PM

[David White:]

Interesting article. I am not sure where Robert Schiller got his data from. However, both Barron's and Ockham Research have recently come out with very different estimates of the S&P500's current PE. Barron's estimates it at 123x. Ockham Research estimates it at 46x. Given the fact that the S&P500 earnings were 30+% lower in Q1 2009 than in Q1 2008, I find it easier to believe the other groups than Robert Schiller.

Jun 01 09:24 PM

[Thomas Barnard:]

Schiller uses a 10 year average of earnings to smooth out the numbers. It would be reasonable to assume that if you used current data you might get 123x or 46x, but how meaningful is it? Everyone knows that those numbers are artificially high. I'm staying with Schiller


On Jun 01 09:24 PM David White wrote:

> Interesting article. I am not sure where Robert Schiller got his
> data from. However, both Barron's and Ockham Research have recently
> come out with very different estimates of the S&amp;P500's current
> PE. Barron's estimates it at 123x. Ockham Research estimates it at
> 46x. Given the fact that the S&amp;P500 earnings were 30+% lower
> in Q1 2009 than in Q1 2008, I find it easier to believe the other
> groups than Robert Schiller.

Jun 01 10:17 PM

[Thomas Barnard:]

Today certainly confirmed your decision. And it could go higher, the big Mo is with you. But I get a little nervous when guys like Art Cashin comes on saying that volume is slight.


On Jun 01 10:58 AM Larry House wrote:

> Your conclusion seems right on, Thomas, but the market has a mind
> of its own. I am staying with positions a bit longer (probably too
> long). The trend is your friend until it isn't.

Jun 01 10:19 PM

[Thomas Barnard:]

Yes, I agree. Schiller uses a 10 year average of earnings in his calculations to take this worthless stuff out that you are talking about.


On Jun 01 03:02 PM thiazole wrote:

> I still think P/E is a worthless metric for valuation if any of the
> companies in the index are losing money. It means that the "value"
> of the index increases if you eliminate all the companies losing
> money. How does that make any sense? A stock can only go to zero,
> so it can't have a negative value to an investor.

Jun 01 10:22 PM

[Thomas Barnard:]

Very interesting. I wondered if this held true. I had a devil of a time find historical interest rate data going back beyond 1960, but here's what I found with the 10 year Treasuries:

1919 4%
1932 3.5%
1942 1%
1949 1.5%
1974 8%
1980 11.5%

It looks like your relation holds for the periods 1974 and 1980, but otherwise, no.


On Jun 01 12:34 PM pslater wrote:

> Once again P/E ratios are a function of interest rates. With the
> average non-financial company in the S&amp;P 500 having a BAA credit
> rating, that rate (currently around 8%) implies a P/E of about 12.5
> (100/8). It's no coincidence that in 1981 the P/E of the S&amp;P
> was around 7 - because interest rates were around 13% (100/13 = 7.7).
>
>
> If you want to know what an appropriate P/E ratio is use an appropriate
> interest rate. As for valuations, use your imagination as to what
> the earnings component should be....I would suggest that 10 year
> trailing earnings isn't it.

Jun 01 10:52 PM