Historically, September has been a sorry month for the markets -- the worst, in fact. And, sure enough, it looks like we'll close it out about 11% poorer in the S&P 500 index. But there's some good news for those who believe in historical trends and the seasonality of stock performance.

Dan Sullivan, publisher of The Chartist newsletter, ran some numbers and found what can only be called an amazing historical performance gap between two six-month periods: May through October and November through April. Over the past half century, money invested in the S&P 500 index at the beginning of May and sold at the end of October -- repeating the process every year -- would have returned only a total of 35%. The November through April time period, however, yielded 3,180%.

Put another way, it seems almost all of the market's average annual 10% returns occurred during the latter time period, while investors were just running in place during the former.

It's hard to say why things have worked out this way. The Washington Post's James Glassman says it might be tax-related, or just a quirk. We tend to go with the "quirk" explanation, even though the overall performance gap is so large. If there were a solid reason for the differences, it would likely be apparent by now. One thing is for certain, though. If there is a reason behind the seasonality, arbitrageurs will jump in and erase any advantage that might have been gained.

Our take: Rather than playing games and trying to time the market, we still prefer to stick with our Foolish principles of finding good businesses at good prices, and buying with the intention of holding for the long term.