Brett Arends’s ROI: Should you sell your stocks in May?

The rhyme actually stems from England, where St. Leger’s Day refers to a classic horse race that traditionally marked the end of the summer social season and the beginning of autumn. The St. Leger’s day race has been run in September every year since the U.S. Declaration of Independence (there is no connection). This year it will be run on Sept. 17.

Here in the U.S., analysts prefer to talk in terms of staying out of the market for a further month and a half, until the end of October. Hence it is called “the Hallowe’en Effect.” Selling your stocks at the start of May and buying them back at the end of October would mean a full six months out of the market.

The argument is that, traditionally or often, the summer months are really lousy for stocks. The chances of a crash or a bad slump are high. The chances of big returns are slim. And overall the average reward for owning stocks during the summer months doesn’t match the risks.

It sounds crazy—like using astrology or tarot cards to manage your money—but, weirdly, there is something to the theory. Some exhaustive research looking at the performance of all the stock markets in the world over all the periods for which there is data, in some cases going back centuries, has found that there really has been a “Halloween Effect” everywhere except Mauritius (don’t ask).

In a nutshell, over the long term effectively all the superior returns from the world’s stock markets have come thanks to the winter months, from Halloween through the end of April, researchers found. And on average stocks over the winter six months beat the summer six months by 4% a year.

Think: The Crash of 1929 (September and October)

Think: The crash of 1987 (October).

Think: Lehman Brothers in 2008 (September and October).

Think: The bear market of 2000-2, which began in earnest in September 2000 and bottomed…in October 2002.

Remarkable, and defying good rational explanation. During nearly 30 years in this business I’ve heard people offer all sorts of rationales, up to and including the idea that the shorter days after June 21 makes investors gloomier and want to sell their stocks. (But as the research shows, the Hallowe’en Effect has also worked in the Southern Hemisphere—where that should operate in reverse.)

A personal disclosure: Every year around this time I think about taking the advice and selling in May. Every year people who are much smarter and better informed than me tell me not to.

Every year I take their advice instead.

And most of the time, I end up kicking myself and sending them sarcastic emails at some point during the subsequent months, as the market tanks.

Those who argue against selling in May offer some compelling logic. For instance, there is no particular reason to think this bizarre anomaly will keep working. And actually it’s worked less this century than it did in the past. And: If you sell your stocks you may trigger trading costs and taxes. And: If you are investing long term, so what if your stocks go nowhere for six months a year? And: Will it actually end up saving you anything? And: Are you going to buy stocks back on St. Leger’s Day, or Halloween, or on some other day—and why?

Against this I can only offer one observation, which is the reason I feel the urge to sell: It’s not just about the returns, but the volatility.

The summer months almost always seem really volatile, and there is often a nice, juicy selloff when you can buy your stocks back cheaper—or, more to the point, buy back more stocks than you sold for the same amount of money.

I ran the numbers—using the Dow Jones Industrial Average DJIA, -0.09% —going back to 1900. On 63 occasions, or 51% of the time, the market fell by 5% or more at some point during the summer months. That’s comparing the summer low with the price at the end of April.

On 40 occasions, or almost exactly one-third the time, it fell by 10% or more—a full blown slump.

And in most years the market actually declined at some point, even if only by a couple of percentage points.

So someone who sold at the end of April usually had an opportunity to buy their stocks back cheaper at some point over the next six months. But the question then becomes: When should you buy them back? At 5% down? 10% down? After six months? If you’re only going to buy them back once they have fallen by, say, 2% you might save yourself the time and effort.

My own take is that if I’m going to rebalance my portfolio once or twice a year, between stocks, bonds and commodities, I might as well do it at the end of April and again at the end of October as at any time. If there’s a Hallowe’en Effect I’ll benefit, and if there isn’t…well, I needed to rebalance anyway.