Sell in May and Go Away?

We've all heard it before. It gets talked about every year right around this time. But what does it mean exactly? Where does this saying come from?

Well, legend has it that it derived from the British. The original phrase goes, “Sell in May and go away. Stay away till St. Leger’s Day”. So basically, in a similar way that market moving new york fund managers go out to the Hamptons for the summer, slowing down the stock buying, out in England they don't get back to business until Horse Racing season is over in the Fall. The last leg of the English Triple Crown, the St. Leger Stakes, is a huge celebration that goes back to the late 1700s. We Americans like to call that Football Season.

So the annual cycle is very simple: you buy stocks in November and sell them at the end of April. "The worst six months", as my friend Jeff Hirsch likes to put it, begins next week (See my Dow Chart here). The numbers historically are incredible. Let's say you had invested $10,000 in the Dow Jones Industrial Average in November of 1950, sold it all at the end of April 1951, and did that every year since, you would have made $674,073 by 2011. Now, had you done the opposite with that $10,000 and bought in May and sold on Halloween every year since 1950, you would have somehow lost $1,024 during that same time. So in other words, every single dollar made during that 60+ year period in the Dow was made during "The Best six months" of the year. That is just unbelievable to me. I've been aware of these stats for many years and it still impresses me every time I read it or say it out loud.

Now, as we all know and has been well documented, post-election years are typically some of the worst of the four-year Presidential cycle. And as it turns out, the best/worst six month trading strategy gets affected as well. From the Stock Trader's Almanac:

"Post-presidential-election years, such as this year, have the worst performance record out of the cycle. Since 1833, post-election years on average have gained just 2.0% (up 20, down 24). In comparison, top performing pre-election years have gained 10.4% on average (up 34, down 11). As May nears, the following historical look at how DJIA has performed during the Worst Six Months of post-election years since 1905 should prove useful while trying to decide whether or not to “Sell in May” this year"

"Two sets of averages are presented. The first is the DJIA’s includes all data since 1905 while the second starts at 1953. Prior to 1950 the U.S. economy (and the global economy) was substantially different than now. From 1901 to 1951 farming made August the best performing month of the year. This is no longer the case and August is now the second worst month of the year.

Largely due to the fact that post-election years generally perform poorly, there is little difference in the overall outcome whether or not years before 1953 are included or not. The ratio of advancing/declining “Worst Six Months” is nearly the same around 2:1, although the average gain during the “Worst Six Months” since 1953 is significantly smaller, virtually flat. Also of note is there has been just one double digit gain (2009) during the “Worst” months of post-election years since 1953. While there have been four (five if rounded) double-digit declines. DJIA’s performance from 1985 to 1997 is also impressive, but those results are from the last secular bull market that lasted from 1982-2000, not the secular bear of the last decade-plus."

So what have we learned? Number 1: "the worst six months of the year" really are the worst. And number 2: in post-election years, they're somehow even more vulnerable to correct. So let's keep these stats in mind as we make all-time highs in some of the major US averages.

We'll continue to let price dictate our actions, that will never change. But I think it's prudent, as always, to be aware of the seasonality in the stock market. Nothing wrong with that. In fact, to ignore it I think would be irresponsible of us.

Source:

Sell in May, Post-Election Year Edition (StockTradersAlmanac)

Tags: $DJIA $DIA $DJI