Skip to main content
Message Font: Serif | Sans-Serif
No. of Recommendations: 9
This column for was published last week. It also includes a chart I created, which unfortunately doesn't paste. So, you will have to go to Prof. Shiller's website, which is here For RM subscribers the link is here:

Despite the Selloff, the Market Still Isn't Cheap

By Hewitt Heiserman Contributor
3/20/2008 10:40 AM EDT

With Tuesday’s 4% gain in the S&P 500, do stocks still offer long-term investors compelling prospective returns?

One way to answer this question is to look at the market’s price-earnings ratio. P/E’s tell us how much investors are paying for earnings. The lower the P/E, the more value you get in return—provided, of course, these earnings hold steady or improve.

The S&P now sells for 19 times earnings, based on yesterday’s close of 1331 and $71.71 worth of trailing twelve month’s earnings (source: Barrons’s). With the benchmark 10-year Treasury selling at 29 times earnings, large-cap stocks are more attractive than bonds, on this basis. (A bond’s multiple is the reciprocal of its yield; e.g., 1/3.46 = 29x.)

Market P/Es are partly influenced by the business cycle, however. If, for example, top-line sales are roaring and profit margins are at a peak, then stocks may look cheap even if we are on the cusp of a recession. By the same token, the market will look expensive if corporate America is hurting, even if the economy is about to strengthen.

To mitigate these distortions some analysts use as their denominator the average of the last several years’ worth of earnings. Rolling-period earnings have the advantage of smoothing out the economy’s year-to-year vagaries.

Yale Professor Robert J. Shiller, for example, uses the average of the last ten years worth of inflation-adjusted, or “real,” earnings to estimate a market P/E. The S&P fetches 23 times earnings on this basis, as the dark blue line in the chart below shows. The pale blue line is the market’s TTM multiple, which is 19x as mentioned above. I want to thank Professor Shiller for permission to use this chart, which I updated to reflect yesterday’s close. To learn more, go here:

The lesson of this chart is that while the S&P is a lot less expensive than the Himalayan levels of 2000, at 23x it is still not cheap. Indeed, 127 years of stock market history shows just three other times when real 10-year PE’s breached current levels: 1901, 1929, and 1966. In each instance, the multiple later contracted to below 10x. Mean regression can hurt!

One other item to think about. Long-term rates, which are depicted in red, have trended down since Sept. 1981, when they hit 15.3%. The decline in the cost of capital was a tonic for stocks, igniting an 18-year bull market.

But with the 10-year Treasury now down to 3.46%, further gains from these levels seem unlikely, barring a depression. Indeed, with the exception of a few months in 2003, long bond yields haven’t been this low since the Eisenhower Administration. If rates climb back to 4.7%, which is the median yield since 1881, then bonds will offer more value than stocks.

My interpretation of Prof. Shiller’s research is that the stock market will tread water for several years until earnings catch up—with plenty of white-knuckle episodes along the way. Thus, if you are a long-term investor, buy companies that 1) possess authentic earnings power, 2) enjoy a durable competitive advantage, and 3) trade at a discount to intrinsic value, as I describe in my book It’s Earnings That Count. Also, update your portfolio mix, if necessary. Stocks as an asset class are not always the best investment.
Print the post Back To Top
No. of Recommendations: 0
I'm not sure where the chart you're talking about can be found. Is it in the stock data spreadsheet?
Print the post Back To Top
No. of Recommendations: 2
Click here

then click the link in this paragraph:

One can access an Excel file with the data set (used and described in the book) on stock prices, earnings, dividends and interest rates since 1871, updated.

Prof. Shiller last updated this spreadsheet in 2006. To create the chart in my Real Money article, I downloaded Prof. Shiller's spreadsheet onto my computer, then updated his data through March 2008.

The key point in my column is that the S&P 500's real 10-year P/E is 23x. This is an extreme valuation, based on 130 years of market history. Indeed, there are just three other times when the market was this expensive, not including the current period:

In June 1901 the market peaked at 239 and the real 10-year PE was 25x. 19.5 years later, in Dec. 1920, the market troughed at 74. The new PE was just 5x, and the market's CAGR during this period was -6%.

In Sept. 1929 the market peaked at 383 and the real 10-year PE was 33x. Almost four years later, in Jun. 1932, the market troughed at 74 (same as Dec. 1920, above). The new PE was 6x, and the market's CAGR during this period was -36%.

In Dec. 1968 the market peaked at 635 and the real 10-year PE was 22x. Almost 14years later, in Jul. 1982, the market troughed at 238. The new PE was 7x, and the market's CAGR during this period was -7%.

The stock market is going to be a tough place to make money for the next several years, if history is a guide.

Print the post Back To Top
No. of Recommendations: 0
How are dividends figured into this analysis? I believe Jeremy Siegel argued that higher P/E ratios could be justified in the future because dividend payout rates are historically low (and continue to fall).

(I only had a second to glance at the spreadsheet and type this up...gotta get back to work! :)
Print the post Back To Top
No. of Recommendations: 0
Unfortunately more than half of my current savings are locked up in a 401K with fairly limited choices for investment. I'm not sure there is much of a choice... I don't think bonds are very good right now either. Maybe an income oriented fund :(

-- Dennis
Print the post Back To Top
No. of Recommendations: 0
Prof. Shiller's 10-year PE excludes dividends. Browsing through the index to his book Irrational Exuberance, Shiller address dividends. However, I have not read this far yet.

I recommend his book to anyone who wants to learn more about this topic.

Print the post Back To Top