Wednesday, January 09, 2013

Torturing the Data

I get a kick out of so-called indicators that are supposed to predict how the stock market will do. I just read about one called the "First Five Days" indicator. This one supposedly predicts that stocks will rise for the full year if they are up during the first five trading days of that year. Since there are 252 trading days in 2013, I suppose this is better than the "First 250 Days" indicator.

The Super Bowl indicator predicts an up year if a team from the old NFL wins the Super Bowl. Some investors rely on astronomy to help them decide how to position their portfolios. Others rely on weather.

About 20 years ago one of the most prestigious academic journals in economics published a paper that supposedly proved that stocks are more likely to rise on days that are sunny in New York City than on days that are rainy. The theory is that weather affects the mood of traders. It isn't entirely clear why the weather in New York City would affect the mood of a trader in Dallas or San Francisco, but that's the theory. In any case, according to the author, the statistics supported the theory. However, the results could simply be a statistical anomaly. After all, stocks go up more often than they go down and, on average,  there more sunny days than rainy days in New York City.

These indicators remind me of what one of my favorite statistics professors in graduate school used to say, "If you torture the data long enough, it will confess."

Subscribe to MoneyMasters with Vahan Janjigian by Email

Enter your email address:

Delivered by FeedBurner