THE ODDS ARE AGAINST THE SUPER BOWL THEORY IN 2002
By DOLORES KONG

01/20/2002
The Boston Globe

With Super Bowl Sunday just two weeks away, and the
end of January around the corner, now is the time to
beware the lure of stock market myths.

The myth known as the Super Bowl Predictor - or
Indicator or Theory, depending on whom you ask - says
if a team from the original National Football League
wins the championship, the stock market is likely to
go up. But if a team from the original American
Football League wins, the market is headed for a dive.

Then there's the January Effect. If the market is up
at the end of the first month of the year, it's likely
to be up for the whole year. And if it's down at the
end of January, it'll be down for the whole year. Or
so that theory goes.

The trouble with these and other theories, such as the
Hemline Indicator or the Triple Crown Curse: There's
no cause and effect between the theory and the stock
market, even if statistics suggest there's been close
to a 90 percent correlation at times.

Why should the market go up if an old NFL team wins,
or down if an old AFL team takes the title? And why
should the yearly market performance depend on January
results?

And as for the Hemline Indicator or the Triple Crown
Curse, they go like this: Falling hemlines suggest a
falling stock market, and a year in which a horse wins
the Triple Crown will likely be a bearish one.

I don't know which way hemlines are going these days,
and I don't follow the races, but the stock market has
certainly been down so far this month. If it closes in
negative territory at the end of January, are we
really to believe the market will be in the red once
again at the end of 2002?

That would mean three down years in a row for the
market, which hasn't happened since the 1930s.

As unlikely as these theories may seem, they actually
have been tracked by market analysts, studied by
economists and mathematicians, and bandied about by
the financial press.

Robert H. Stovall, currently senior vice president and
market strategist for Prudential, has been the keeper
of the Super Bowl Predictor, doing it as a lark and
not as an investment recommendation, although he's
acknowledged he may have created a "financial
Frankenstein."

But with the predictor wrong the last few years, it
appears Stovall may be losing interest.

A December 2000 item on TheStreet.com said he'd given
it up after the predictor was wrong three years in a
row. But a few weeks later, just before the
Ravens-Giants Super Bowl matchup last year, Stovall
was still being quoted by the media, saying that
whichever team won, it was a bullish sign.

That's because the Baltimore Ravens, while an American
Football Conference team, has its roots in the
original National Football League, as the Cleveland
Browns. So even if the AFC team won, which it did, it
should suggest an up market. Or so the theory goes.
But we all know the market headed south in 2001. The
predictor is now 0-4 since 1998.

If that's not enough to disprove the Super Bowl theory
to you, maybe a recent study by economics professor
Paul M. Sommers at Middlebury College in Vermont will.


Applying a statistical test to the Super Bowl results
and the annual change in the Dow Jones industrial
average over three decades, Sommers found no evidence
to support the theory. In fact, the longer the period
of analysis, the weaker the theory, so that it's no
more telling than the flip of a coin.

"While the Super Bowl Theory still enjoys popular
support, its predictive power today is much less than
what it was 12 or even six years ago," Sommers wrote
in The College Mathematics Journal.

"Despite the Super Bowl Theory's surprisingly strong
early record, reading the sports page now is not
making it any easier to read economic tea leaves," he
concluded.

In an interview last week, Sommers added, "I don't
think it works. It doesn't hold as fast as it did
several years ago."

He's also studied the January Effect and the "Santa
Claus Rally." In an analysis of the January Effect
between 1970 and 2001, Sommers found that in the 20
times that the Dow closed up in January, the market
was up 18 times for the year, and in the 12 times it
closed down in January, it was down seven times for
the year.

"It's curious," said Sommers, adding that it seems
stronger than the Super Bowl Theory.

But with the bearish markets in 2000 and 2001 (despite
an up month in January last year), and with the Dow
down about 3 percent so far this month, Sommers is not
hoping for the January Effect to hold true in 2002 -
for the sake of his and everyone else's investments.
To further investigate these and other theories, check
out the "Anomalies" link on www.inves torhome.com.

For a lark, and to show how ridiculous these theories
can be, I thought I'd invent a new one based on 2002
being a rare palindromic year. The idea came to mind a
few weeks ago when my brother-in-law, a London banker
with a mathematics degree from Cambridge University,
made a chance comment.

Well, someone's already beat me to the punch. Chet
Currier, a Bloomberg News columnist, calls it the
"palindromic precedent," suggesting that if 1991, the
last palindromic year, is a sign, 2002 should be a
recovery year.

The beauty of this theory is it's hard to prove or
disprove, with so many years between palindromes. The
last before 1991 was 1881, and difficult to find stock
market data for. But I tracked down historical results
from Yale economist Robert J. Shiller. Between January
1881 and January 1882, the Standard & Poor's Composite
Stock Price index fell 4.36 percent.

Well, based on that, I'd say the Palindromic
Prognosticator suggests a 50-50 chance that 2002 could
be a down year.