There is, quite frankly, a whole lot of garbage written and said about seasonality in the stock market, especially around this time of year. If you follow the financial news and opinion sites, I am sure you have come across articles proclaiming that one month or another is the best or worst for stocks, but the fact that many of them differ as to what those months are is an indication of the basic problem.
Some say January is the best month to buy, some say December. Some say September is the worst month, some say August. Some say that there is only one month that is on average a loser, some that there are three or four.
The problem here is not unique to monthly stock return averages. It is an inherent problem with averages themselves. When looking at average performance for each month in the stock market, what results you get depend on how far back you go. One would assume that the more data you include the more accurate and reliable a picture you get, but with the stock market that is not necessarily the case.
The NYSE is 200 years old this year, and over that time there have been massive changes to the nature of the economy and the way traders and investors behave. How relevant are 100-plus-year-old data to the modern market? Common sense would say not very.
If you don’t go back far enough, however, you run the risk of giving too much weight to single events that had nothing to do with what month it was. The Lehman Brothers bankruptcy, for example, came on September 15, 2008, and the SP 500 lost around ten percent that month and over sixteen percent the next. The bankruptcy did not come because it was September, but if the sample size is small it heavily skews average returns for that month and the next.
Those same two months have been exceptionally bad in other years too. September of 2002 was one of the worst months of the so-called tech wreck and the SP lost eleven percent. October bore the brunt of the 1987 crash and the same index dropped over 20%. There is no logical connection between those occurrences and the month in which they occurred. Major moves in the stock markets are driven by events, not timing.
Many people fail to understand the nature of averages, especially when used to assess probability. September may, depending on how far back you go, be the the worst of the twelve months for stocks historically, but that doesn’t mean that every September will be bad. Over the last ten years there have been five Septembers with positive returns on the SP 500 and five with losses, and selling on the basis that the average is bad can make you miss out on some great profits.
In 2010, for example, stocks gained nearly nine percent in September, accounting for well over half of that year’s total return.
Basing trading and investing decisions on what calendar month we are in is essentially the same as basing them on astrology. You will be right some of the time and, if you are a skilled position manager, that will be enough to make the strategy profitable. That does not, however, make it wise.
No matter what history says, September is just another month.
If the White House and Congress can agree on a tax plan that looks like it can pass and that offers what the market is looking for, chances are it will be a good one. If not and we see a repeat of the health care debacle, chances are it won’t. That, the Fed’s actions, the geopolitical situation, the fate of the dollar, the price of oil, the economic data for August, and a whole host of other factors are what will move the market over the next thirty days. So, if you want to be scared of something, focus on those things, not what month is coming.