By E.S. BROWNING
Staff Reporter of THE WALL STREET JOURNAL
Yet another classic investment principle is under attack.
Bulls say they have debunked an investment superstition, fueling their view that stocks are on their way to more levels once thought impossible. Bears see the trampling of another basic rule as just one more sign of a stock market gone mad and in for trouble.
The rule in question is a simple one. It holds that stocks don't offer the safe, guaranteed returns of government bonds, and hence investors need to be offered extra returns to make it worth their while to put money into stocks instead of bonds.
Stock gains, after all, come mainly from the unpredictable advances of the market; stock prices can go up 20% one year and down 20% the next. So investors will put money into stocks only if they expect that stock returns, over time, will outpace bond returns by some amount that compensates them for the added risk of owning stocks.
This extra return from stocks long has been called the "risk premium" -- literally, the premium you receive in exchange for owning a riskier, more volatile instrument.
The idea that stocks should return significantly more than bonds is "a long-term principle" of finance, says Robert Bissell, president of Wells Capital Management, Wells Fargo's money management arm.
Ivo Welch, a professor at the Anderson School of Management of the University of California at Los Angeles who has been studying the phenomenon, goes further. "It is the single most important number for anybody in finance, either academic or practitioner," he maintains.
Trouble is, by some interpretations, this fundamental rule is beginning to flash a big red warning light about the future direction of the stock market. Big stocks have gained so much since 1994 that the typical forecast of additional stock-market gains, based on expected gains in company earnings, now indicates that investors could get almost the same return from bonds. The risk premium that can be forecast today for stocks has shrunk to two percentage points or less -- far below the four to eight points that analysts typically use as a benchmark. That, according to classic thinking, heralds bad news for the stock market.
William Dudley, director of U.S. economic research at Goldman Sachs, worries that some investors, who now consider 20% annual stock-market gains normal, don't understand that mathematical limits may be closing in. Those investors could be in for a shock.
"Investors may already be too optimistic in their assessment of future expected returns," he cautions in a report this month.
But revisionists scoff at such worries; the old math, they say, no longer applies. James Glassman, a resident fellow at the American Enterprise Institute in Washington, maintains that, based on studies of long-term investing over the past century, stocks actually have been no riskier than bonds. Over long periods, he says, their returns have been more steadily up than those of bonds.
"If you define risk as volatility," which is what most analysts do, Mr. Glassman says, "history has shown that stocks are no more volatile than bonds" over the long term. Stocks can be more volatile over short periods, however.
In an opinion article published in The Wall Street Journal in March, he and Kevin Hassett, an American Enterprise Institute resident scholar, argued that long-term investors should demand no extra premium for holding stocks. And if investors didn't expect any extra premium, the two authors add, stocks still would look like highly attractive investments today as many forecasts show them likely to gain more than bonds in the future.
In fact, Messrs. Glassman and Hassett calculate that, if you eliminate the expectation of a risk premium and forecast low inflation and strong earnings growth, the Dow Jones Industrial Average still is far short of where it should be. In theory, at least, they figure that its price-earnings ratio could hit 100, four times the current level, although they acknowledge that it might in practice stay lower.
In an interview, Mr. Glassman says traditional-style market analysts didn't foresee the stock market's recent gains in part because they thought investors require a higher risk premium than they actually expect.
But to the traditionalists, the Glassman analysis amounts to nothing less than changing the rules to fit the game, which they see as a recipe for disaster.
Who is right? One reason for the disagreement is that the experts all agree that no one has the tools to forecast what the risk premium will be in the future.
There is some agreement on what it has been in the past. Since about 1926, stocks have returned about five percentage points to eight percentage points more than Treasurys, depending on which Treasury instruments you use and how you compare them. In the past year, the gap has been enormous -- a 32% gain in the Standard & Poor's 500-stock index, not counting dividends, has dwarfed Treasury yields (for buy-and-hold investors) of less than 7%.
Another way to look at the risk premium is to calculate how volatile stocks have been compared with bonds over recent periods. The more volatile stocks are, the higher the premium should be because the risk of a sudden change in fortune for a stockholder is greater.
But looking backward doesn't necessarily help analysts forecast how much -- if any -- stocks will beat bonds in the future -- or what kind of expectations investors will have for the risk premium.
Most analysts try to do that through a formula based on the consensus earnings forecasts for the S&P 500. The idea is that a stock's price is simply the present value of its expected future earnings.
One measure of the expected return on a stock's price is to take the per-share forecast earnings for the current or coming year, and divide by the stock price. That figure is called the "earnings yield," and for the S&P 500 it currently is less than 5%. That is barely above the inflation-adjusted Treasury yield, which is 3% to 4%, depending on your estimate of coming inflation.
Little wonder that Morgan Stanley Dean Witter strategist Byron Wien recently wrote an article entitled "Risk Premium-R.I.P." He figures that, even using a risk premium of 2%, his models show the S&P 500 17% overvalued compared with bonds. He would have to cut the risk premium to 1% in order to make the S&P 500 look slightly below fair value, he says.
Messrs. Wien and Dudley worry that stock investors could be in for disappointments if they expect continuing double-digit stock gains; so does Prof. Welch, who himself thinks, based in part on a survey he has done of other academics, that the risk premium probably ought to be calculated at something closer to 4%. Many other academics think it should be 5% or 6%.
Much of the basic research suggesting that the risk premium should be zero was done by Jeremy Siegel of the University of Pennsylvania's Wharton School of Business. But Prof. Siegel took the trouble to write a letter to the editor of The Wall Street Journal contesting the Glassman-Hassett thesis that the price of the Dow industrials could hit 100 times earnings. Prof. Siegel says that stocks have gained so much that they already are priced at fair value.
"The biggest danger is this expectation of many shareholders that 15%, 20%, 25% gains are going to become the norm in the future, because that will push share prices to unreasonable levels," he says.
Bulls aren't interested in that view.
Too much worry about things like the risk premium and earnings yields would have kept people out of recent bull markets, says Richard Bernstein, Merrill Lynch's chief quantitative strategist. "Those things haven't worked in the past 18 years," he says.
Adds Paul Lesutis, a portfolio manager at Brandywine Asset Management in Wilmington, Del.: "I'm not sure there should be a risk premium on equities over bonds, because over my career equities have done so much better than bonds. Over the past 15 years equities actually have been less volatile than bonds."