Ed Yardeni has his neck out.

Way out.

The chief economist for Deutsche Bank in New York has shown data on his web site that stocks have been as much as 20 percent undervalued by in recent weeks--- but overpriced.

How can this be? How can stocks be undervalued and overpriced at the same time?

Simple. Dr. Yardeni is making estimates of future earnings that are far lower than the consensus of Wall Street analysts. If he is correct in his estimates for 1999 or 2000 the future looks grim for common stock investors. If the Street consensus is right, the pain and worry that abounds today will be forgotten in a wealthier tomorrow.

The starting point here is a fundamental question: How do you value common stocks?

Dr. Yardeni has used a model favored by Federal Reserve Chairman Alan Greenspan. This model states that stock prices are closely related to expected operating earnings of the S&P 500 companies and the yield on a 10 year Treasury. Using this model Dr. Yardeni notes that stocks were overvalued by more than 20 percent in the summers of 1997 and 1998. They were also overvalued by 34 percent in September 1987.

Expressed as a formula, the "fair" price for stocks would be equal to their expected earnings next year divided by the yield on a 10 year Treasury bond. The basic idea here is that stocks are always in competition with bond yields. If interest rates rise, stock prices will decline so that the earnings yield (earnings divided by price) on stocks is competitive with the yield on bonds. If interest rates fall, stock prices can rise. Since a small change in interest rates has a relatively large impact on valuation, stocks have yo-yoed between under and overvalued in the last three weeks.

In July, with the yield on a 10-year Treasury at 5.49 percent and the consensus earnings forecast for the S&P 500 at about $49.00 a share, the model suggested a fair value of $892, substantially less than the July peak of $1,186.

So the market was overvalued.

Since then interest rates have been in a free fall. The yield on a 10-year Treasury fell to 4.4 percent in early October, increasing the fair value of the S&P 500 index 1,114. That's above where the index was trading last week.

This may mean that the worst of the decline is over and that we can relax, right? Unfortunately, Dr. Yardeni says there is another shoe still to drop.

The problem, he says, is that "bottom up" estimates of corporate earnings--- the ones analysts make on individual companies--- tend to be optimistic. While they were relatively accurate in the recovery from the early 90's recession, they were far higher than actual earnings in 1991 and 1992.

Dr. Yardeni believes that consensus estimates are optimistic once again. While the consensus is $45.84 for 1998, he is estimating $44.00. Similarly, while the consensus estimate for 1999 is $53.91, a 17.6 percent increase, Dr. Yardeni is estimating $43.25, a small decline. Worse, while the consensus for 2000 is $60.08, Dr. Yardeni is estimating a major decline to $33.00 a share.

Looking forward, the difference is more than a gap. It's a chasm. With the consensus view, the worst is over and things are going to be better. With the Yardini view, we've only just entered a major bear market.

In a recent telephone interview I asked what supported his estimate.

"The year 2000 problem with computers will be extremely disruptive. It will exacerbate existing economic problems. Even if you get your computers fixed, others won't--- so we'll still have a recession. In addition, the worst isn't over for our current problems. The global crunch is still going on. Everyone's' tolerance for risk decreased at the same time. We're heading into a series of financial crises and business credit crunches," he said.

We can estimate the potential depth of this bear market by doing the math for the consensus estimate and Dr. Yardeni's estimate:

         1) If you assume no change in interest rates and consensus earnings, the S&P 500 would have a fair value of 1365 by the end of 1999, an increase from current levels that would surpass the July high nicely.

         2) Using Dr. Yardeni's earnings estimate the fair price will be 750, a decline of nearly 25 percent from current levels.

How much further would interest rates need to decline to support the current level of stock prices if earnings decline as much as Dr. Yardeni projects? Try a yield of 3.4 percent, 100 basis points lower than current rates.

Who's right?

It's the future, no one knows.

What can you do?

Watch the number of earnings surprises. If most are negative, big time, that's a powerful sign that Ed Yardeni is right and the consensus is wrong.

Want to learn more? Check out Dr. Yardeni's economics website. You can hear him explain the problem on RealAudio, visit his Online Chart Room, and download his and Federal Reserve papers on stock valuation.