The Rest of the ‘Sell in May’ Story

Posted by

May 4, 2010

It’s that time of year again - time to “Sell in May and go away”, since the market’s average monthly returns suggest the beginning of May through the end of September is not only a weak patch, but usually so weak that this five month period is a net loser.

Are you actually going to take action based on this historical data? If you’re like most investors, you’ll at least make some sort of adjustment based on the notion, having experienced it to some degree on your own with real money. Funny thing about averages though…

Check the Fine Print

They say the market’s average annual return is about 10%. That’s reasonable. Think back though…. think back to the last time the market actually gained anywhere actually close to 10% in a calendar year. The 30%+ and 30%- moves are plentiful and easy to recollect, but an actual 10% gain? The last time we even came close was 2006’s 13.6% gain. In 2004, we saw a 9.0% rally from the S&P 500. Other than those two cases, we have to go back to the 80’s to find an annual gain anywhere close to 10%.

Point being, the long-term averages are probably mathematically correct, but the standard deviations are wildly high….too high to try and turn the data into actionable information.

The same is true for the ‘May through September’ theory. So, here’s the rest of the story you don’t hear about the data, based on the S&P 500’s results going back to 1971.

By the Numbers

While it is true that the months between May and September are usually - on average - tepid to the point where it feels like it’s (almost) not worth being in the market, it’s actually rare for all five months for any given year to be weak. More often than not, you’ll see one or two of the five months come in as losers, but they’re usually offset by strong gains or rebound in two or three of the other months in the five month span.

Reality - the average gain between May and September for any given year since 1971 is actually a 0.8% loss. BUT, the market actually made gains 2/3 of that time over the last 39 years (26 winners to 13 losers).

Removing the worst four May/September spans from the last 39 (1974’s 29% dip, 2002’s 24% loss, 2001’s 17% decline, and 2008’s 16% loss), and you actually get an average May/September return of 3.4%.

Even taking out the top four May/September returns - to counter the removal of the bottom four - you’re still going to see an average gain of 1.6% for t he May/September span. (Those best four are the 21% gain from last year, the 18% rally from 1997, the 18% advance from 1980, and the 13% gain from 1995.)

Conclusions? A few really, really bad May/September time periods are bringing down the averages. The fact is though, the odds of making at least some sort of gain - even small ones - are actually in your favor.

Slice It Any Way You Want It

Even when we break it down in a month-by-month basis, the reality of ‘misleading averages’ still hits home.

May: Despite the warning, the average ‘May’ since 1971 has returned 0.91%; 24 of the last 39 Mays have been positive ones.
June: The average June over the last 39 years has netted gains of 0.48%. Twenty-three of the last 39 were winners.
July: The month of July has barely averaged out to be a winner over the last 39 years, with a typical gain of only 0.06%. Still it’s better than a loss. If you take out the worst five Julys though (about 15% of the total), the average return ratchets up to 1.04%. That’s not bad, considering only 16 of the last 39 Julys have been winners.
August: August has proven to be only a little better than June since 1971, with an average gain of only 0.25%. Still, given that 23 of the last 39 Augusts were positive ones, you still have to like your odds. If you take out the worst one (the 14.5% dip from 1998), the average gain jumps to 0.64%.
September: The last month of the so-called ‘weak period’ is indeed the biggest disappointment, typically losing 1.0%; 22 of the last 39 have left the market in the red (17 winners). If you remove the worst four though, the average return turns positive again.

Conclusions? Again, even on a month-by-month basis, the ‘Sell in May and go away’ axiom proves neither accurate or all that helpful.

The lesson learned is simple…. don’t believe everything you hear. Sometimes selling May can be a brilliant decision. Most of the time though, history says you’re better off staying in the market.

Share and Enjoy: