Friday, May 29, 2009

Data mining--or past performance is no guarantee of future performance

Calendar and other similar anomalies are generally very interesting and students love reading about them (well as much as they love reading anything in finance) but do not fit well with efficient markets.

For instance, why? Consider the January effect If stocks were predictably up in January and investors had calendars that showed December was the month before January, investors would buy in December. etc etc.

The following article form CNN addresses this and other anomalies.

Selling stocks in summer often isn't a smart thing to do - May. 29, 2009:
"Wall Street traders have a lot of funny ways to try and predict stock performance.

There's the hemline indicator, which tries to equate fashion trends to stocks. According to that maxim, the shorter women's skirts are, the better stocks should do. It's also known as the bull market, bare knees phenomenon. Classy.

Then there's the Super Bowl indicator. Stocks are supposed to go up in a year when an 'old' NFL team wins"

No comments: