I've been reading a lot in the media about the January five-day barometer. The idea being that if the first five days of the month are up, then one should buy and hold for the rest of the year. If the first five days of the month are down, then one should sell (or short) and hold that position for the rest of the year. At first glance, this "early warning system" appears to be superficially true. Since 1956, the first five days have been up on 31 occasions. Buying at the close of the fifth day and holding until the end of the year resulted in success in 23 out of 31 occasions, for a success rate of 74%. However, while it seems predictive, the question to ask is: "As compared to what?" Is it any more or less predictive than any other five-day period? In fact, since 1956, buying at the close of any positive five-day period, and holding for 245 trading days, has resulted in success on 71% of the 6,835 occurrences. So the predictability of the first five days is not really any more or less than that of any random five-day period, and I would not base a yearlong allocation on this so-called "early warning system."