Don't be Fooled

Exaggerated Earnings and P/E Pipe Dreams

(February 4, 2009)  I am not a fundamental investor, which means I do not make investment decisions based on corporate earnings forecasts. Stock prices go up or down based on whether investors are primarily buying or selling stocks. Period. Sometimes a company loses money but the stock price goes up because people buy it anyways, and sometimes a highly profitable company sees its stock price plunge because of irrational panic. Predicting the stock market is more about understanding human behavior than reading balance sheets.

That being said, corporate earnings are not entirely irrelevant. As I noted in my previous post, the real "as reported" price-to-earnings ratio (P/E) of the S&P 500 typically varies between 10 and 20, meaning the grand average price of all S&P 500 companies is usually between 10 and 20 times their annual earnings, or about 15 times earnings on average. Therefore the S&P 500 is considered overpriced when the P/E is more than 15, and under-priced when it's below 15. One could imagine a simple market-timing strategy based on the P/E ratio which invests 100% in stocks at a P/E of 10, but cashes out completely if the P/E reaches 20.

Unfortunately, the majority of websites, newsletters, and media outlets don't use the real P/E ratio of the S&P 500. Instead, they usually provide the P/E ratio based on "operating earnings." Bob Brinker, for instance, uses operating earnings for his forecasts, and it's one of the reasons his portfolios have been decimated by this bear market.

Operating earnings are optimistic, idealized calculations that ignore one-time losses that companies don't consider to be part of their normal operating budgets. Right now, for instance, banks which have lost billions of dollars in defaulted loan payments are not necessarily subtracting them from their operating budget. Future projections of these earnings assume that their budgets will go right back to what they were before the losses, which is a highly dubious assumption. Real "as-reported" earnings, on the other hand, don't hide anything, and include every blemish in the budget.

The following table, using data from the Standard and Poor's website, demonstrates just how much of a difference there can be between the two earnings calculations:

Price-to-Earnings Ratio for S&P 500 Companies (Feb. 2009)
QuarterReal P/E*Operating P/E*
2008 Q32518
2008 Q42414
2009 Q1**2815
2009 Q2**3217
2009 Q3**3017
*Earnings (E) are the total earnings of the previous 12 months (4 quarters).
**The 2009 forecasts predict what the P/E ratio would be if the S&P 500 index remained at its current level of 840. If the S&P 500 index falls, that would be a reduction in "P," and thus it would lower the P/E ratio.

It is curious that operating earnings remain 40% to 90% higher (P/E ratios lower) than real earnings for five straight quarters even though one-time losses supposedly constitute the only difference. To my eye, large differences between the two columns seem to be happening on a regular and consistent basis. Does anyone else smell something fishy? To make matters worse, earnings forecasts have been steadily decreasing for the past few months, which has raised P/E ratio forecasts to ever more over-valued levels. These forecast revisions may continue for a while still.

If you're not already baffled that anyone uses operating earnings, then take a look at what can be found on Yahoo Finance. There are two ETFs that mimic the S&P 500 index, SPY and IVV, and Yahoo reports today that their P/E ratios are between 10 and 11, which is even more optimistic than the best P/E ratio in the above table. I submit that the P/E numbers reported by Yahoo are pure fantasy, because there are no earnings numbers - past or future - high enough to give such low P/Es.

So the P/E ratio of the U.S. stock market is somewhere between 10 and 28 depending on whom you ask, and these numbers are being revised every week. What's an investor to do? Well, I mostly ignore earnings, because even the real earnings numbers can hide certain details.

The only fundamental number that affects my investment decisions is the dividend yield, because it can't be exaggerated. The dividend yield is the sum of all cash payments made to stockholders in the past 12 months divided by the stock price. That's it! There's nothing to fudge, hide, or imagine. Check various websites for the dividend yield of the S&P 500 index, (yahoo, google, Standard and Poor's, Wall Street Journal) and suddenly you find agreement where before there were different universes of earnings. The dividend yield of the S&P 500 has historically oscillated between 3% (overvalued) and 6% (undervalued), meaning a simple form of market-timing could be based on this one number.  However, one should remember that the dividend yield is still not a perfect predictor of the stock market.  If investors have the desire to buy stocks and the money to do it, then prices will rise regardless of the fundamentals, just like they did in 1998-2000.