Posted by: Aaron Pressman on August 29, 2007
As investors’ attention drifts and the end of summer draws near in the last week of August, analysts turn to that evergreen topic of analysis, calendar trading. What usually happens to the stock market around three-day weekends or even just Labor Day weekends? You can crunch all kinds of numbers and come up with smart-sounding answers. Sometimes, the answer changes depending on whether you run the data back to 1990, 1975 or 1253. I prefer the chart that the good folks at CXO Advisory Group concocted after looking at daily stock market returns back to 1950:
As you can see from the chart and their blog discussion, at first glance it appears that a useful pattern has been revealed with the market averaging great returns on the Friday before Labor Day and not so great returns on the couple of days following the holiday.
But what are the long green bars? They represent the standard deviation of returns, or the amount by which each year’s individual performance typically differed from the group’s total average. The big green bars are a lot bigger than the seeming pattern of average returns in the study. The amount of variation from year to year is far more significant that the average that’s been calculated. In other words, there’s a lot of random variation in the individual returns, noise as statisticians say, that pretty much outweighs whatever pattern seems to have been uncovered. Or other other words: nothing to see here, move along…