Few market experts are ready to abandon the January effect theory - which says small-cap stocks outperform large-cap stocks that month - even though the theory has not rung true for four consecutive years.
Look at the historical evidence, these market pundits say:
The January effect has been remarkably consistent since 1926.
So has a corollary of the effect - the January "signal" - which says that if small caps outperform in January, they will outperform for the year, and vice versa.
There have been only 13 years from 1926 to the end of 1995 when small caps underperformed large caps in January. In 10 of those 13 years, small caps lagged large caps for the full year.
But in recent periods, the January outperformance has not been as consistent or as strong.
In January 1995, the Russell 2000 index, a small-cap index, had a total return of -0.75% as of Jan. 30, while the Russell 1000, a large-cap index, gained 2.32%.
Since 1926, small caps have outperformed large caps in January 86% of the time. But in the period from 1983 - when the effect was first documented and therefore made known to investors - to Dec. 31, 1995, small caps have outperformed large caps in January 75% of the time.
While the outperformance since 1926 averaged 5.14 percentage points; in the period since 1983, small stocks outperformed large caps by an average of 2 percentage points, according to Ibbotson Associates Inc., Chicago, which looks at the smaller end of the small-cap market.
Still, even though the January effect hasn't been in force in a few years, the January signal has proven an accurate predictor.
In 1993, 1994 and 1995 - when there was no January effect - small caps lagged large caps for the full year, according to Claudia Mott, director of small-cap research of Prudential Securities, New York. She uses CRISP 9-10 data from the Center for Research in Securities Prices of the University of Chicago, which tracks mainly companies in the ninth and 10th decile of the New York Stock Exchange.
Does that mean small caps will underperform large caps in 1996?
It seems likely. The last time small caps underperformed in January but outperformed for the year was in 1982, said Mark Stumpp, chief investment officer of PDI Strategies, a money manager in Short Hills, N.J.
"If this hypothesis holds true on average, I would say this would be a bad year for small caps - but anything can happen," said Derek Sasveld, a consultant with Ibbotson.
"Once you do a study, it's all over. That's market efficiency at work. People will exploit an inefficiency (like the January effect) till it no longer exists."
Said Prudential's Ms. Mott: "This year's January effect is being held hostage by the budget talks and the proposed capital gains tax cut embedded in there."
She said if there is a capital gains tax cut later this year, the small-cap market will sell off sharply (it fell 18% both in the fourth quarter of 1978 and the third quarter of 1981 when capital gains tax cuts were announced).
"If the market sells off sharply, it bounces right back again, which is basically what the January effect is," she said.
So we might see a January effect in, say, June, this year.
Abby Joseph Cohen, an analyst at Goldman, Sachs & Co., argues the January effect has been occurring earlier and earlier in recent years.
In a report titled "The January Shudder," Ms. Cohen said the recent market weakness is attributable to the faltering budget process and early company reports for the fourth quarter, which suggest further earnings disappointments.
"The so-called 'January' rally appears to have begun during November. If so, the usual spunk offered by this seasonal phenomenon is now missing," she said in the report.
Ms. Cohen said the institutionalization of U.S. equity markets has pushed ahead the January rally to around Thanksgiving. Individual stock market ownership is increasingly through mutual funds, whose fiscal years typically end Oct. 31; and institutional investors typically position their portfolios for the coming year when they prepare economic and earnings forecasts late in the preceding year, she noted.
"Previously, the end of the tax year on Dec. 31 encouraged individual investors to clean up their portfolios prior to that date, often selling, only to re-establish positions early in the new year. Smaller-cap stocks were often the most volatile during this period due to their ownership by individual investors."
Noted Mariko Gordon, president of CastleRock Capital Management, New York, a small-cap manager, said: "All money managers in the small-cap area do look for candidates that were hurt unduly by tax selling."
She has had trouble finding such stocks this year.
Others say the January effect has little to do with taxes.
Ibbotson's Mr. Sasveld said: "The January effect has held in virtually every country regardless of tax treatment and capital gains loss selling."
If taxes were responsible, the January effect should occur in April in Australia. But it doesn't. And there would have been no January effect in Canada before the introduction of a capital gains tax around 1970. In actuality, the January effect has been documented in Canada since 1954, he said.
Mr. Sasveld also questions the theory that the January effect stems from year-end window dressing by money managers, which often involves unloading volatile small-cap stocks in favor of conservative names. If window dressing were to blame, "there should be a semiannual or quarterly effect - but there isn't."
Steven Leuthold, chairman of The Leuthold Group, Minneapolis, said the January effect has been absent the past three years because "it has become so well known that people started buying small caps in late December...When people quit trying to take advantage of it, it will probably come back."
Mr. Stumpp believes the January effect is still valid: "I think it's a really important effect."
He points out that from 1926 through 1994, the Standard & Poor's 500 logged an average annual return of 10.7%, while small-cap stocks logged an average annual return of 12.7%. But if January is excluded, small caps, as measured by companies in the bottom 10% of the New York Stock Exchange, underperformed large caps by a wide margin, returning 5.7% vs. 9% for S&P stocks.
More recently, from 1960 through 1994, the story is the same, he said.
"If you're going to buy small stocks, the only time to have done it is January," Mr. Stumpp said.
"The January effect is not a done deal. It may evaporate in any one year. But over extended periods, the January effect really dominates."