AN INTERESTING THEORY, HOPE IT'S TRUE

This could be the touchdown pass the stock market needs: A legendary stock-market forecasting device based on the Super Bowl is predicting stocks will rise in 2010.

The Super Bowl predictor of the market has fumbled a few times in recent years, but still has a 79% accuracy rate—way better than NFL quarterbacks. It has predicted the direction of the market accurately after 34 of the 43 Super Bowls, including last year’s.
The quirky indicator is based on whether an “original” National Football League team wins the big game. If an original team wins, the market will rise for the year; it falls if the victory goes a team that joined the NFL because of the merger with the American Football League in 1970.
Last year it worked beautifully. The Pittsburgh Steelers, an original team, won the game, and the Dow Jones Industrial Average soared nearly 20% for the year. The predictor coughed up the ball miserably the previous year, when the “original” New York Giants won but the market sank.
This year, both the Indianapolis Colts and New Orleans Saints are original teams (the Colts still get that designation because of their Baltimore Colts roots), so maybe the stock market has a chance to pull out a win in the final three quarters.
“The predictor is going to point to another ‘up’ year,” says Robert Stovall, 83-year-old strategist for Wood Asset Management in Sarasota, Fla., who has long tracked the Super Bowl predictor. But football aside, he predicts, based on fundamental events, that 2010 “will be generally kind to investors” during a second year of market recovery.
Admittedly, all this has little science behind it, except that there are more teams linked to the “original” side, and the stock market tends to go up, recent history aside. “Even though it’s an anti-intellectual entertainment kind of indicator,” based on its accuracy you “cannot ignore it,” Mr. Stovall says.
But the predictor does continue to get respect, including in academic studies. The latest academic paper to discuss it comes from George W. Kester, a finance professor at Washington and Lee University in Virginia. Mr. Kester has written a forthcoming article in the Journal of Investing called “What Happened to the Super Bowl Stock Market Predictor?”
In the paper, the professor meticulously breaks down the tape on the predictor’s accuracy. Although the reliability of the indicator has diminished since its 90%-accuracy days, he backtested a market-timing strategy based solely on the outcome of Super Bowl over its history and found that it “would have resulted in more than twice the portfolio value of an investment in the S&P 500—a result that would be the envy of many portfolio managers.”
The professor also examined whether the traditional way of separating “original” and postmerger teams could be out of date, and whether it might be better to track an updated predictor based upon post-1970 conference affiliations. (In other words, National Conference teams are bullish, and American Conference is bearish.) But he concludes, after doing the math, that the prediction accuracy and investment performance of the traditional Predictor is “superior” to an updated one.