Overvalued? Stocks' Price Is Finally Right

By Michael Edesess, chief economist at Lockwood Advisors.

WSJ 99-01-12:

As the Dow Jones Industrial Average soars above 9500, market prices are being compared unfavorably to their historic levels. Today's market, the argument goes, is "too high" if its price/earnings ratio is higher than the historic average, or its dividend yield is lower.

But isn't it possible that historic market levels were too low? In the past, the information investors needed to assess the value of an investment was far less readily available than it is today. Investments may, as a result, have seemed more risky than they really were. Investors may have been too cautious to anticipate the high corporate earnings growth rates that subsequently came to pass.

Recently, market levels--as measured by P/E or price-to-dividend ratios--are higher than they were for much of market history. Technology stocks in particular have reached unheard-of P/E ratios. Stock-market returns have also increased markedly. Returns for cumulative periods ending in 1997 grow higher and higher as the period gets shorter and more recent. The cumulative annual return is 12.3% for the past 40 years, 16.6% for the past 20, 18.0% for the past 10, and 20.2% for the past five.

What accounts for this increase? Return can be viewed as the sum of dividend yield, corporate-earnings growth, and growth in the P/E ratio. Over the past 15 to 25 years, dividend yield has been declining while corporate earnings growth rates have shown a definite increase. Even more marked over that period has been the increased growth in P/E ratios, which are now around 30 as against their historic average of 14.5.

High P/E ratios are a sign that the market believes future earnings growth will be high. How reliable are the market's expectations of long-term corporate earnings growth, as reflected in the P/E ratio? An analysis by Lockwood Advisors compared P/E ratios from 1933 to 1997 with the subsequent 10-year earnings growth. The study shows that when P/E ratios were forecasting above-average earnings growth, they were right 77% of the time. So what can we say about the current high P/E ratio? Only that history suggests it is likely to be correct in its implicit earnings growth forecast.

A forward-looking measure of the appropriate stock-market level can be found in virtually any finance text. Price can be estimated as the present value of future dividends, discounted at the required rate of return (that is, at the rate of return the investor thinks appropriate for the risk being assumed). A bit of mathematics shows that this measure is roughly equal to one divided by the difference between the required rate of return and the anticipated dividend growth rate.

The long-term historical growth rate of corporate dividends is about 8%, but over the past 20 years the rate has been about 10%. Assuming future growth of 9% and a required return of 11%, the two-percentage-point difference between 9% and 11% is 1/50th. Thus, the appropriate ratio of price to dividends is 50--just about what it is now.

Now let's look at history. In 1945 the market's P/E ratio was 14.4 and its price-to-dividend ratio was 24--both near the long-term average. Suppose the market had been less "undervalued" in 1945, at 50 times dividends instead of 24. Given the higher stock prices at the beginning of the period, the subsequent rate of return over the years 1946 to 1997 would have been a respectable 10% (comprising 2% yield and 8% growth)--or it might have been higher if the doubling of the front-end investment proved to be a stimulant to growth.

So, was the market priced correctly in 1945 at 24 times dividends? Or would it have been better priced at 50 times dividends? Had investors known what we now know about earnings and dividend growth, and had they been less skittish about the unknown, the market then would probably have looked much as it does now.