 |
The
Truth About "Selling in May and Going Away" |
|
"Sell
in May and go away." It's a clever, catchy cliché, isn't it?
It rhymes, AND it has a founded historical basis, sort of ... maybe
..... or maybe not really.
The Micro Cap
Press analytical staff haven't been indecisive about our bearish expectations
for stocks this summer. We're not looking for new bear market lows, but
we do believe stocks are overdue for a correction.
Does that
happen to coincide with the 'Sell in May' mantra? Yep. Are we making
that call because we assume averages and norms are meaningful enough to
bet on? Not in the least.
See, the 'Sell
in May & Go Away' advice is great - when it works. The problem
is, it doesn't work nearly as often as you think it might. And when
it doesn't work, it bucks the trend quite a bit, so much so
that blindly following the advice can put you at a serious disadvantage.
We'll show you
precisely what we mean below; the knockout punch doesn't come until the
very end of today's comments.
The truth is,
June through September - on average - is the worst of any four month
stretch for stocks. The idea is supported by historical evidence. And,
June, August, and September (but not July) are - on average - losers
for the market. (February is the only other typical loser.) Again, history
provides the factual evidence. That's why the cliché is now practically
common knowledge - because the idea somewhat held water.
 Funny
thing about averages though..... they in no way tell you about the important
exceptions
to the norm.
As an illustration,
take two high school students with an average test grade of a C (or a score
of 75%). Both students could consistently score 75% on all their tests,
in which case the average is a fair representation of each student's typical
success. Or, one student could earn an A+ (100%) on every test, and the
other could get a failing grade of 50% (or an F) on every test, and the
average score would still be 75%.... still a C average.
In the former
case, both students would be average; in the latter case, one
student would be an idiot while the other was a genius.
Point being,
the exceptions matter.
The same reality
applies to the average June, August, and September - the exceptions
are significant.
With that in
mind, and to provide some real (i.e. useful) perspective on the historical
odds of a weak summer, we've got some details to add to the common knowledge.
The nearby table
shows the average monthly performance for the S&P 500 going back to
1950, the number of wins versus the number of losses, the percentage of
the total months that were winners for each month of the year, the average
loss in a losing month, and the average gain in a winning month.
(Whew!)
 Anything
surprising stick out?
At first glance,
the overall results seem to be consistent with the idea of shedding stocks
in May and just sitting things out until after September. Take a closer
look though.
August is an
interesting month. It's a loser in terms of average returns, and the average
loss in August is bigger than the average gain in August. However, the
odds of actually taking a loss in August are pretty weak. More often
than not, August doles out gains... and pretty good ones at that (+3.0%)
July is another
curious and compelling month. We already know it's a small winner, on average.
However, those winning months are huge (+3.8%) when you get them....
which you would a little more than half the time.
June's barely
an average loser, yet loses ground less than half the time.
September's
the only real disaster, losing more often than not, and with the
typical loss being bigger than the typical gain. Of course, most all of
us know September's also the time to start wading back into the market's
waters.
On the other
side of the coin, the odds are very good that you'll experience a gain
of some size in January, August, October, and November. But, you better
be right about your gamble. On the off chance you're wrong and any of those
months brings disaster (and eventually they will), it will hurt - a lot.
Ahhh, sometimes
you almost didn't want to know the details, huh? Those standard
deviations can really wreak havoc with averages.
It gets even
more complicated than that though.
While the previous
data table is an eye opener, it's not necessarily complete. It doesn't
quite tell you about the overall and average performance of these entire
four-month
stretches.
 In
other words, an 8% gain in June would be remarkable, but it would be
irrelevant if July, August, and September each lost 4% in the same year
(creating
a net loss of 4% for the summer months).
So just to drive
home one more point, we've attached some charts of the most noteworthy
exceptions to the "Sell in May" norms. They'll really make you rethink
how married you want to be to the old rule of thumb.
The first scenario
where the bearish theory broke down only occurred three years ago. Between
the beginning of June and the end of September in 2006, the S&P 500
actually gained 4.97%. All four months were winners, even if a couple only
fostered tiny gains.
 In
2005, that same four month stretch produced a 3.35% gain. In the year 2000,
believe it or not, the market was up a modest 1.48% for the June-September
stretch, while in 1997 the S&P gained a whopping 11.74% during the
so-called 'weakest four months of the year'. Not bad.
And what
about losses for the four month period? Oh, we got plenty of those
too. In 2008, the S&P 500 gave up 17.8% between the end of May and
the end of September, and you don't even want to know how rough this period
was in 2001 and 2002. Yikes.
Nevertheless,
years like 2005, 2000, and 1997 put the validity of the cliché in
question.
But wait, we
haven't even gotten to the knockout punch yet ... an amazing and possibly
frustrating statistic.
As we promised,
a killer stat...
 Over
the last 59 years, only 24 times has the market lost ground between
the end of May and the end of September. That's right, in 35 of
the last 59 years, investors would have been better off NOT selling in
May and going away.
So how'd the
cliché ever get started? The average June-through-September loss
- when there was one - was a fairly hefty 8.3% dip. In only nine
years did the loss exceed double digits though; a few really bad apples
have skewed the averages. Those painful years were 1957, 1966, 1969, 1974,
1981, 1990, 2001, 2002, and 2008. If you take out those worst nine years,
that average four-month loss sinks to 3.1%.
The average
gain during the 35 winning June-through-September periods was 5.7%.
Point being,
the
June-September period can be horrendous, but it's a rarity.More
often than not, it's better to just ride it out.
The lesson to
be learned here is simple - take nothing at face value, and take everything
on a case by base basis. The norms and averages aren't consistent enough
to actually bet on.
Making the idea
more relevant for us, we're looking for the market to pull back over
the next several weeks, but it has nothing to do with the calendar.
|
Updated: 6:01 am PDT (13:01 GMT), September 9, 2010
