The stock market boom of the 1990s was no bubble built by wild speculation, but propelled largely from economic productivity factors, including corporate earnings, say Roger G. Ibbotson and Peng Chen.
Going forward, they say, pension fund executives and other investors should continue to look to equities as the favored place to invest, at least over bonds.
The expected equity risk premium - the return advantage of equities over risk-free long-term bonds - is alive and well and is expected to deliver close to its historical returns, according to the two researchers. Mr. Ibbotson is professor in the practice of finance at Yale University's School of Management in New Haven, Conn., and chairman of Ibbotson Associates Inc., Chicago; Mr. Chen is vice president-research at Ibbotson.
Their view is contrary to recent predictions of other analysts, who contend the expected equity risk premium will be zero or negative at least over the next few years.
A key element in the Ibbotson and Chen study is the integration of economic and fundamental factors to support the conclusion of their research.
"What happens in the stock market is probably going to be related to what happens in the economy," Mr. Ibbotson said in an interview. "Some think the stock market marches to a different drummer," he added. But it hasn't overall, and he and Mr. Chen don't expect it to in the future.
"Investors should not expect a much higher or much lower return than that produced by the companies in the real economy," they wrote in a draft of a paper they are near completing. "This model estimates the stock market return expectation (is) justified by the macroeconomic productivity."
The findings have important implications for asset allocation decisions of pension sponsors and other investors, especially "what proportion to invest in stocks vs. bonds," said Mr. Ibbotson. "What are the payoffs for risk?"
Messrs. Ibbotson and Chen say in their paper, "Contrary to several recent studies that declare the forward-looking equity risk premium is close to zero or negative, we find the long-term (expected) equity risk premium is only slightly lower than the pure historical estimate." They don't name any of the studies
Voices of dissent
Among recent research taking a pessimistic outlook for the equity risk premium is a study by Robert D. Arnott, managing partner of First Quadrant LP, Pasadena, Calif., and Ronald J. Ryan, president, Ryan Labs Inc., New York. They predict the death of the equity risk premium (Pensions & Investments, Nov. 13). They believe it will be at best zero and likely negative. That is, they expect the total, inflation-adjusted return on stocks will be between 3% and 4% a year, either the same as or below the expected 4% bond return.
Overall, Messrs. Ibbotson and Chen expect the stock market to grow at a 10.7% compound annual return in the long term, meaning over 25 years. That 10.7% includes inflation, the risk-free return on bonds and their expected equity risk premium.
They estimate the equity risk premium will be 3.96 to 4.25 percentage points over bonds. That is about one percentage point less than the historical risk premium of investing in domestic common stocks, which averaged a compound annual 5.24 percentage points over the risk-free rate of long-term U.S. government bonds from 1926 through 2000.
That means "investors were compensated 5.24 (percentage points) per year (more) for investing in common stocks rather than long-term risk-free assets like" U.S. government bonds, they wrote.
This forecast is "consistent with the historical supply (or sources) of U.S. capital market earning and (gross domestic product) per capita growth over the period 1926 to 2000," they wrote.
The 5.24 figure comes from a GDP model they use, while the 3.96 figures comes from an earnings model they also use. Both figures are in line with their findings, but "we are mostly stressing the earnings forecast," giving the 3.96 as the expected equity risk premium, Mr. Ibbotson said.
"The key to understanding a forecast is to understand what's been" the source of returns in history, Mr. Ibbotson said.
"We never say buy stocks, per se," Mr. Ibbotson said. "There is risk in the stock market and there is this payoff, the expected equity risk premium, for being in stocks. We are saying over the long run, we certainly expect stocks to beat bonds by a pretty wide margin."
The two use several methodologies in their paper, all arriving at the same conclusion and the same projected equity risk premium. They break down the return on equities into essential components. They show roughly half of the long-term return on equities comes from the equity risk premium and the other half from inflation and the risk-free rate on long-term government bonds. The equity return portion includes income or dividends, capital gains and price-earnings growth.
All the components are derived from the economic productivity of the company, except for the price/earnings ratio, which is determined by investors' willingness to pay for earnings, sometimes leading to wild speculation.
In their projection, growth in p/e accounts for only a small portion of the expected equity risk premium, as it has historically. The p/e of the overall market has grown about 2.5 times since the beginning of 1926, from 10.22 to 25.96 at the end of 2000, they note. They keep the p/e constant, not projecting a continuation of that growth rate.
"We don't predict another change" in the growth in p/e over the next 25 years, Mr. Ibbotson said. "The p/e has gone up so high because the market forecast a good economy.
Selective tech bubble
While Messrs. Ibbotson and Chen discuss only historical returns in their paper and not the 1990s or any market bubbles, in the interview they said the general equities boom of the 1990s was driven largely by economic and corporate productivity factors. They acknowledge there was a bubble in some technology segments of the market. "I thought there was an Internet bubble to be sure," Mr. Ibbotson said.
Some analysts talk about returns of the 1990s being a bubble. "But the returns really have been economic returns," Mr. Ibbotson said. "You get wild-eyed bulls and pessimists of various stripes in history. Our forecast is in line with what the economy has done for 75 years," in terms of corporate earnings and GDP per capita.
"This model implies the long-run supply of stock returns should be in line with the productivity of the overall economy," which they find true historically and which, they assert, will continue to produce the expected advantage in return of investing in equities over bonds.
"The idea behind the forward-looking GDP per capital model is that equity returns are related to overall economic productivity," the write in their paper. "This model estimates the stock market return expectation is justified by the macroeconomic performance."
In another method, they break down equity returns into four components: inflation; the real growth rate of overall economic productivity, or the GDP per capita; the increase of the equity market relative to the overall economic productivity; and dividend yield.
"Overall, GDP per capita slightly outgrew earnings and dividends," they write. "In other words, overall economic productivity increases faster than corporate earnings and dividends," historically.
Messrs. Ibbotson and Chen contend other researchers err in their use of dividends in their analysis.
"(W)e believe that dividend growth does not accurately reflect the growth of companies' productivity, especially during the past decade," they contend.
The dividend yield is at is lowest point since 1926, 1.1% at the end of 2000, they note. By contrast, in 1980 it was close to 5%. "(T)here has been a paradigm shift in corporations' dividend policy," they contend in their research. "As investors favor more rapid earnings and price growth, corporations are paying fewer and fewer dividends recently. Instead of paying dividends, many companies reinvest earnings, buy back shares" or acquire other companies.
A corporation's "dividend decision will affect only the form in which shareholders receive their returns, i.e. as dividends or capital gains, but not the total return," their paper points out.