Stock Market Valuations: Too High, Too Low or Just Right?

Internet
Stock Market Valuations: Too High, Too Low or Just Right?
By David Edwards
Special to TheStreet.com
4/24/2001 6:08 PM EDT
URL: http://www.thestreet.com/p/rmoney/internet/1402993.html

A recent column on valuations by Peter Eavis reminded me that determining fair value for the stock market is more complicated than just checking the S&P 500's price-to-earnings ratio vs. its historical average. The P/E ratio should expand when interest rates, inflation rates and investor uncertainty is low, contract as these parameters rise. So how do I determine whether the stock market is fairly valued? Unfortunately, there's no magic touchstone, but here are several models that can provide guidance.

The following chart shows yields on the 10-year treasury bond, inflation rates (year-over-year CPI) and the earnings/price ratio of the S&P 500 since 1953. (The E/P is the inverse of the P/E ratio; the conversion allows pictorial representation of spread relationships.)

Source: Robert Shiller (http://www.econ.yale.edu/~shiller/), author of "Irrational Exuberance."

In the 1950's, fear of the 1929 stock market crash was still so pervasive that investors would accept a coupon yield of 3% on Treasuries but not less than 6% earnings yield from the S&P 500. Interest in equities increased through the 1960's, even as yields and inflation rates rose steadily. The terrible bear market of 1973-74 put investors off equities for a decade, as inflation ravaged returns on stocks and bonds.

From 1980 on, investors steadily returned to equities to the point where the earnings yield fell to just about 3% (equivalent to a P/E of 34), less than the yield on the 10-year Treasury and almost par with inflation. More recently, the earnings yield has increased to 4.4% as P/E's have fallen to 22.8. Do we have to wait until the earnings yield rises to 6.25%, which would correspond to a P/E of 16 before we can buy stocks again? Not unless we're expecting inflation to rise above 6%.

So how can we determine fair value on the S&P 500? A simple model might assume that the S&P 500 earnings yield should trade as a spread over inflation or over the 10-year Treasury. Using the average spreads since 1980 (22 years of roughly equivalent economic conditions) we get:

	E/P - CPI	10 yr yield - E/P	S&P 500
Average Since 1980	2.51%	1.72%
Current	1.26	0.51	1165
Implies S&P 500 @	907	1588

The range of 907-1588 (plus/minus 30%) is not very useful.

Economists at www.Dismal.Com have analyzed the same data using regression, a technique that allows for multi-variable analysis, and prepared a stock market valuation calculator. Assuming that S&P 500 earnings fall 5% over the next year (First Call estimates are for a decline of 5%) and the 10-year Treasury yield stays around 5%, the regression formula predicts a price target of 1257, a gain of around 4%. If corporate earnings growth is in fact closer to 0%, but treasury yields rise to 5.5%, the model projects the S&P 500 at 1197, which is right around its current levels. If the Fed continues to cut aggressively, and the 10-year yield drops to 4.5%, the price target pops 19% to 1393. Given likely scenarios of the next year, this model implies stocks are fairly valued to somewhat undervalued.

Alston Boyd of Martin Capital Advisors has prepared a different model of related valuations of the S&P 500 earnings and dividends to yields on T-Bills and short-term commercial paper. You can view the chart here.

By this model, the S&P 500 is fairly valued. In recent years investors' focus on future earnings allowed the model to swing way into overvalued territory. The slide of the last year has addressed that issue.

But what if trailing earnings information is in fact irrelevant to stock market valuation? After all, investors are looking forward, not backward. Ed Yardeni at Deutsche Bank Alex Brown built a model based on Federal Reserve Bank research which derives valuation from 12-month forward consensus earnings from First Call and 10-year Treasury yields.

By this model the stock market has fallen from 70% overvalued in March 2000 to about 12% overvalued as of April 20. The methodology is carefully explained at this Web page.

Lastly, Jim Bianco of Bianco Research has prepared a model relating total stock market capitalization to U.S. Gross Domestic Product (GDP). Logically speaking, the value of stocks should be related to the overall productivity of the economy.

Stock Market Capitalization As
A Percentage Of Nominal GDP

Until 1996, the value of the stock market never exceeded that of GDP (in fact, the previous high-water mark was 81%, set in August 1929. By March 2000, the stock market cap was 181% of GDP. As of April 2001, Bianco estimates that the ratio is around 135% to 140%. The average since 1925 is 53%. Since stock market earnings now include a healthy component of international earnings, I don't see this ratio falling back below 100. Still, from current levels, this model implies that the S&P 500 is overvalued by about 28%.

How do I use these models in my work? They certainly cannot be used for daytrading. However, they are useful in evaluating whether risk in the stock market is increasing or decreasing.

In January 2000, I began to worry more about protecting clients' gains rather than stretching for higher returns. I cut my technology allocations from 40% to 25%, decreased my overall equity exposure from 90% to 80% and within the equity component, increased my clients' exposure to what I call non-correlated assets (i.e., stocks such as REITS and energy stocks with price movements generally not in synch with the overall S&P 500).

Our clients gained 4.6% in 2000 after gaining 43.6% in 1999. At this point in time, I see that the risk in the stock market is substantially reduced and am increasing my equity exposures, particularly to technology stocks such as Sun Microsystems (SUNW) , Oracle (ORCL) , EMC (EMC) and Cisco, which I perceive as bargains after the selloff of the past six months.

David Edwards is a portfolio manager and president of Heron Capital Management, a New York managment firm. At the time of publication, his firm was long General Electric and Cisco, though positions may change at any time. Under no circumstances does the information in this column represent a recommendation to buy or sell stocks. Edwards appreciates your feedback and invites you to send it to David Edwards.