BUNK 25
SELL IN MAY
Everybody’s heard, “Sell in May and go away.” Hopefully, most ignore it. It’s supposed to mean summer is a bad time for stocks, so you can safely sell in May, let stocks drift lower, and buy back in the fall. It’s also very wrong and financially bad advice.
Amazingly, every May, particularly if stocks are down, this gets echoed endlessly in headlines. (It’s very similar to “so goes January” from Bunk 24—folks who are bearishly biased seem to love this.) I hear from some clients and Forbes readers if stocks have a tough week or two during the summer months, “Everybody knows to ‘sell in May!’”
But what does “sell in May” really mean? Sell when in May? May 1? The 31? The 12? April 30? And if everyone knows to sell in May (if we ever figure out exactly when), if you’re smart, don’t you sell in April before everyone else starts selling? And if you could figure that out, why not sell even earlier to beat that early crowd? When does it end? (People who like automatic sell rules like these never think through that logic.)

Sell in May . . . When? And Then What?

There’s more. If I sell in May, when the heck do I buy back? People say it has to do with the summer being bad—so do I buy back on September 1? The autumnal equinox? When? “Sell in May” has myriad permutations, but none hold water. Check history. Vary the date however you want—it doesn’t work.
Table 25.1 shows average S&P 500 returns by month. Since 1926, stocks have averaged 0.38 percent in May. Not huge, but not negative. They’ve averaged 1.19 percent in June and 1.83 percent in July. July has been historically best—smack dab in midsummer, which some say is so dreadful (and others say it’s when you start a “summer rally”—another piece of statistical nonsense). August has averaged fine too. So many stats. So little meaning! Welcome to this little bunkhouse.
Table 25.1 Average Stock Returns by Month—Why Would You Sell in May?
Source: Global Financial Data, Inc., S&P 500 total return from 12/31/1925 to 12/31/2009.
MonthAverage S&P 500 Return
January1.49%
February0.06%
March0.72%
April1.67%
May 0.38%
June1.19%
July1.83%
August1.30%
September-0.75%
October0.43%
November1.07%
December1.74%
Now, some will say “sell in May” doesn’t mean May, but that the summer months overall are bad for stocks. Fine—June, July, and August together average 4.51 percent.1 Together, they beat the average of any other three consecutive months! (How the heck did this myth get started?) And it’s likely much better than what you could get from a 90-day T-bill, and definitely better than cash, never mind adding in transaction costs and taxes from inning-and-outing. Then some will say, “No, it’s the summer half that’s worse—May to October.” Sheesh! This simple saying has a lot of small print. Since 1926, May 1 through October 31 has averaged 4.26 percent, while November 1 through April 30 averaged 7.07 percent.2 Yes! The “winter” half is better. Big deal. (Again, the summer “half” is still positive and still much better than the average returns of cash or six-month T-bills.) We can slice and dice this anyway we want. March through August averages 7.16%, but September through February just 4.38%.3 So why not “sell in September?” (Probably because it doesn’t rhyme with the “and go away” bit.) These are all just statistical quirks.
Plus, these are all averages. There’s a lot of wild variability cooked into them—big ups and big downs (Bunk 5). Look at September in Table 25.1 again—the only negative average. Should you sell in September? Not because of history alone. Take out the two worst Septembers (down 29.6 percent in 1931 and down 13.8 percent in 1937—both in the Great Depression),4 and the remaining Septembers don’t look so bad. How much faith do you want to put into those two terrible Septembers out of 84 Septembers somehow being representative? Or into any average return from any group of months? Averages often mislead. (Basic debunkery: Look beneath the averages to what comprises them.) And just because something has had a negative average doesn’t mean it will be more negative than not going forward. There’s nothing inherently more risky about September—except maybe kids are going back to school and you have to be careful driving. And inning-and-outing around one month or a group of months because they have lower average returns historically is silly, particularly since you can’t guarantee the alternatives net you anything better.
That should kill “sell in May” forever. Debunkenasia it! But in addition to “so goes January,” there are many other day-of-the-week, month, seasonal, and holiday myths—all equally silly. They often feel appealing and sometimes work out ok, but overall, over time, simply don’t work and end up costing those who follow them money. Santa Claus rallies. The January effect. Triple witching. Sell on Monday. Since 1980, every time the New York Giants won a Super Bowl (1986, 1990, and 2007), we had a bear market soon after. The Pittsburgh Steelers are supposed to be bullish—silly coincidences as fundamental as “sell in May.”
Ignore “sell in May” and any other saying about selling automatically based on dates, months, pro football teams, or anything else not somehow grounded in sound, fundamental economics or portfolio theory. (Though there is the rumor that Harry Markowitz is a Steelers fan.)
..................Content has been hidden....................

You can't read the all page of ebook, please click here login for view all page.
Reset
3.238.116.201