For the first time ever, the stock market could return 20% or more a year for four straight years.
With less than a month before the end of the year, the Standard & Poor's 500 stock index was up 21.96% for the year, the fourth consecutive year it has produced such a return.
"We've never had four consecutive years that had a return of 20% or more," said Jeffrey B. Schwartz, senior consultant, Ibbotson Associates Inc., Chicago.
"If it does hold up (through the year's end), it would be noteworthy from a historical perspective."
This four-year milestone would be just one of many superlatives of the extraordinary and continuing bull stock market of the past two decades.
Among other exceptional performance figures, according to data requested from Ibbotson:
* 1998 will mark the eighth-straight year the market has not had a loss for any one year. The only other time -- in 73 years of Ibbotson data the market -- has had eight consecutive years of positive yearly returns was 1982-'89. The 1991-'98 period, through the end of November, has produced a compound average annualized return of 19.98%. The 1982-'89 period returned 18.92% a year.
* The past 20 years, ending 1998(assuming December 1998's return has no impact), has been the best of any 20-year period ever, dating to 1926.
* The market has produced negative returns in only two years in the past 20, and both of those losses were in the low single digits, the smallest of any negative year in history, except for two.
* The spectacular returns of the 1990s and 1980s have boosted the historical long-term annual average return -- and the expectation for future returns -- on stocks by more than two percentage points.
* The equity risk premium -- the excess return of the S&P 500 over risk-free 20-year Treasury bonds, which some use as an expectation of future returns -- rose to 7.8%, up 4.4 percentage points since 1980.
"It's amazing, looking at these numbers," Mr. Schwartz added.
"Investors shouldn't let the numbers color their expectation that these kinds of returns will continue. But the numbers shouldn't leave investors expecting a prolonged loss going forward either," he said, based on the notion annual returns will revert to the historical mean average.
But there is definitely a new mean.
At the start of 1980, the 54-year compound annualized return on the S&P 500 was 9%. Now, adding 19 more years of returns, through the end of November, the compound annualized return on the S&P 500 over 73 years is 11.1%. That's 2.1 percentage points higher than what the historical compound average was.
"There is a concept of mean reversion, that if the market is exceedingly high, things will occur to bring the market back to historical average," Mr. Schwartz added.
"I'm not saying we endorse it."
But he said, "We now have a higher mean." And the higher mean is statistically more likely to occur over the long term in the future than the previous lower mean was.
"From a statistical standpoint, the greater the sample of data you have, the greater the statistical confidence you can have in the mean," he said.
Mr. Schwartz couldn't offer any clue how long the bull market will continue or whether it will end soon.
"From the late 1970s on, the S&P 500 has enjoyed phenomenal success without any major market correction," he said. "We've not had a true prolonged bear market in the past 20 years.
"In the past 20 years there have been only two years of losses," he added, "and they have been low." In 1981, the market was down 4.91%, and in 1990, it was down 3.17%.
Of the years the market was down, in only two years was it down less than 3.17%. That was in 1934, with a return of -1.44%, and in 1953, with -1%.
The four years ending 1998 so far average a compound annualized 29.6% return. That's the second best of any four-year period ever. The best was the four years ended 1936, with a 31.6% compound annualized return.
The 17.4% compound average annual return in the 20 years ending 1998, and assuming December 1998's return has no impact, translates into a cumulative return of 2,381%.
So, $1,000 invested in the S&P 500 on Dec. 31, 1978, would be worth $24,810 at the end of December 1998.
A $1 million investment, then, would now be worth about $25 million.
But, Mr. Schwartz said, "We cannot use the past 20 years to be indicative of the future. If we would use it as our guide, we'd have ridiculous expectations of returns greater than 17% a year with no prolonged bear markets."
The rising market has raised the actual historic risk premium on the S&P 500 tremendously.
Mr. Schwartz stresses the Ibbotson premium is based on historical data. In 1980, the premium was only 3.4%, based on 54 years of data. But big returns in the 1980s and 1990s have raised the historical annual average premium, based on 73 years of data through 1997, to 7.8%.
Mr. Schwartz acknowledged some academics and investment practitioners disagree with using such a high number for future expectations. Many prefer to use a more conservative risk premium in the 2% to 5% range, he said.
Indeed, at the Ford Foundation, New York, Laurence B. Siegel, director-policy research in the investment division, said the $10 billion fund uses 2.8% as its long-term projected equity risk premium of stocks over risk-free bonds.
"That's a lot lower than Ibbotson's," Mr. Siegel noted.
Mr. Schwartz agreed there are other methods of deriving the equity risk premium. In using any model, he said, investors need to ask if the model can be defended and if its expectations are reasonable.
There is an irony in using a lower expected risk premium, Mr. Schwartz said.
"The more conservative one's equity risk premium, the more aggressive one's portfolio needs to be in allocating to stocks to meet one's (return) objectives," he said.
"But the more aggressive (or higher) one's equity risk premium is, the more conservative one's portfolio can be in allocating to stocks to meet the same goal."