Are Those Crazy Stock Market Prediction Theories Actually Right?

Most investors know about the (un)conventional wisdom that the winner of the Super Bowl can forecast the stock market. The theory is that at least two of three major market indices will rise in a year when an original NFL team wins.

If a team from the original AFL wins, at least two indices from among the Dow Jones industrial average, the Standard & Poor's 500 index and the New York Stock Exchange composite index will go down in that year.

This year both teams had their roots in the old NFL, which guaranteed an old NFL victory. Good news for the bulls, if the theory proves to be true. For what it’s worth, the Super Bowl effect has proven accurate about 80 percent of the time overall.

Then there is the January effect in which stock prices supposedly increase during the month of January. This year the Dow was up 6.2 percent in January, the S&P was up 5.3 percent and the NASDAQ, 4 percent.

According to Lauren Rudd at the Pittsburgh Post-Gazette, the January Effect has an accuracy rate of almost 90 percent.

On the other hand, Rudd also recounts that 20 years ago David Leinweber, a visiting economist at Caltech, determined that butter production in Bangladesh had a statistically significant correlation (an r-squared of 99 percent) with the S&P 500 index.

Daily Finance, lists several other “theories” including the Sports Illustrated Swimsuit Cover theory that says if an American graces the cover of this popular issue, it’s good news for the S&P. Also mentioned is the Hemline Index, which ties the stock market to the length of skirt hemlines in women. (The higher the hemline, the higher the market will go.)

As for the results, the S&P 500 had an average return of 14.3 percent during the 17 years since 1978 that featured an American, with positive returns 88 percent of the time, according to the Bespoke Investment Group.

The Hemline Index, unfortunately, does not hold up to serious research. Marjolein van Baardwijk and Philip Hans Franses of the Econometric Institute Erasmus School of Economics conducted the most complete recent study of hemlines and the economy was done.

They could find no evidence that hemlines predict economic performance, but did find that economics predicts hemlines with a three-year to four-year delay. In other words, 3 years after the economy goes into recession, hemlines plunge.

The problem with all these so-called predictors is that there is no discernible cause and effect connection between the two events. All researchers can say at this point is that when A happens, B happens much of the time.

Even when only considering so-called “valid” stock market predictors, Reuters says they will probably be wrong much of the time as history shows equity index forecasts are usually wrong.

"When you ask people for their predictions, that's driving them towards the most likely outcome and that's removing them from the ability to think about extreme scenarios," said Greg Davies, head of behavioral and quantitative finance at Barclays Wealth.

Instead of relying on possible cause and effect relationships that probably do not exist, investors might want, instead, to take a tip from Davies.

"What the industry should be doing, rather than more accurate forecasting is constantly trying to make the point that the world is uncertain and the right way to think about it is over a five to 10 years' horizon."