Mark Twain believed there were three kinds of lies: Lies, damned lies and statistics.
The author would have had a field day with the way statistics are used to “prove” dubious investment theories. A perfect case in point is “Sell in May and go away.”
The idea behind this persistent old idea is that stocks have traditionally had higher returns from November through April, and weaker returns from May through October. At first glance this makes sense: Since 1926, stocks have returned an average of 1.16% each month from November through April, and just .72% the rest of the year. The course of action is obvious: Earn your higher returns during the traditionally “strong” half of the year, and then get out so that the “weak” half of the year doesn’t dilute those returns.
But sometimes a great idea just doesn’t survive the trip from the paper it’s written on into the real world. “Sell in May and go away” is a case in point. The first problem that this legendary piece of advice runs into is a practical one: Once you’ve pulled your money out of the market in May, where do you put it?
If you keep it in cash, you lose ground to inflation, which was 1.5% last year and will probably trend higher in coming years. How about putting that cash in a “risk-free” investment of 30-day U.S. Treasury bills? History shows that isn’t such a great idea either: Between January 1926 and October 2012, the average monthly return for 30-day U.S. Treasury bills was just .29%. Would you rather earn .72% for half of the year (in the stock market) or .29% (in 30-day T-bills)?
Another fundamental problem with dumping all your stocks and rebuying them every six months is that it gets expensive. Not only are you racking up brokerage fees for all that buying and selling, but since you’re holding all of your stocks for less than a year, you’ll pay the short-term capital-gains tax rate rather than the lower, long-term rate. Do you see how the neat, simple calculus of “sell in May” doesn’t look as great in the real world?
Now let’s turn back to the statistical underpinning of the sell-in-May crowd. The approach is based on accurate historical data, which appears to lend it legitimacy. But the data are selectively chosen. Larry Swedroe, an author and principal of BAM Advisor Services, has back-tested the "sell in May” theory for the U.S. market and found that its success depends on the period you’re looking at. From 1960 through 1979, the strategy returned an average of 9.7% a year, 2.8 percentage points better than staying in the market. But during the previous 33 years, holding your stocks year-round earned 10.3% a year—more than 5 percentage points better than the sell-in-May approach.
Cherry picking historical data can make “sell in May” look pretty compelling, but you don’t want to follow advice based on cherry-picked information. We’re reminded of an old joke: A fellow who’s convinced that he’s indestructible jumps off the Empire State Building; as he passes the 50th floor he shouts, “So far so good!”
Our advice? Don’t jump. Keep your feet firmly planted in the market—all year round.