“Sell in May and go away” is an often repeated saying that most people assume has some truth to it. After diving deeper into the theory, it might be a dangerous saying for people to follow.
I received a call the other day from a client that caught me off-guard and made me laugh. Amidst the current corona-virus climate, the vast majority of calls have been people panicking and needing to be talked some sense into, and those seeing a huge buying opportunity and tempering those looking to go 100% equity.
This call was different, though. The client asked if he should sell everything now and get back in after the summer. “Sell in May and go away”. I laughed. And then I realized he was serious. I explained to him that he just stood tall against one of the most turbulent times ever in the stock market and his account balances have started to bounce back. Why would he want to sell now? No, stay invested for the long-term and don’t get involved in market-timing. He was good with that advice and went off to continue enjoying his daily life.
The conversation had me thinking, though. I had heard of the phrase before and had an idea of where it came from, but did people still follow its advice? And more importantly – was there anything to it? Did it lead to an investing edge?
Like anything else that I can’t get out of my head, I started doing research. First, where did the phrase come from?
Next, was there ever an investing advantage to it? Maybe when Wall Streeters escaped the city during the summer but not now, right?
Finally, does anyone still follow the old expression and move in & out of equities based on the seasons?
Where did the phrase “Sell in May and go away” come from, and what does it mean?
The full saying is “Sell in May and go away, don’t come back till St. Leger’s Day”. It’s a reference to the English upper-class and bankers who would escape London during the hot summer months. St. Leger’s is the third leg of the British triple crown and is held in September.
The basic premise behind the saying is that not much will take place in the equity markets during the summer because all of the big movers and shakers will be out of the city. It’s better not to waste your time or energy during those months and concentrate on the more important time when everyone is participating in the markets.
In recent years, the UK has committed to holding general elections on the first Thursday of May. Could that further cement May as a good jumping-off point for the London Stock Exchange?
The phrase made sense in America as well since the Manhattan bankers would also ‘summer’ outside of New York City. Memorial Weekend in May and Labor Day Weekend in September still mark the unofficial beginning and end of summer. Traders would sell their equity holdings in May and keep their investments in much safer holdings like cash and bonds while they enjoy their summer. They would then reinvest in equities after returning to the city in the fall.
At least that’s the story behind the phrase. Did they follow that plan back then? I’m afraid there’s not enough data on trading volume that far back to see if there was a big drop during the summer. Either way, it stands to logic that the theory wouldn’t hold up in this day and age where most investors don’t ‘summer in the Hamptons’ and computer trading never slows down. Even if there was a correlation back in the day, that correlation wouldn’t still hold now, right? Let’s find out.
The Halloween Indicator
A lot of studies into the ‘Sell in May’ theory have used November to April and May to October, so they have even 6-month periods to work with. This made Halloween a big date in the studies – hence the nickname: Halloween Indicator. You invest during the busier winter months (November-April) and stay on the sidelines during the summer months (May-October).
There are two parties on opposite sides of the issue. On one side, you have those that believe efficient market theory would eliminate any kind of deviation this strategy could expose. The other side will look at technicals, returns, different periods or even alter the seasonal dates being used to show there is some sort of correlation.
What do all of the studies say?
In general, there’s not much to the phrase and it shouldn’t be followed. Even back when everyone would leave the city, the correlation between bad returns during the summer and great returns in the winter just didn’t hold up. Several studies did find winter months to have higher returns than the summer months, but since the summer month returns were still positive you would be better off staying invested rather than selling in May and going away.
With so many variables to explain during these long timeframes, it’s hard to pinpoint exactly what causes the higher winter returns. Most believe it has to do with how the holidays stack up during the year. Companies have higher sales and report better earnings in the winter due to Thanksgiving, Black Friday sales, Hanukkah, Christmas, New Years, and Valentines. There are also a couple of events this time of year that put extra money in people’s pockets: tax refunds and annual bonuses.
On the flip side, summer will have lower trading volume while vacation time gets used up. While not proven, lower trading volume is believed to correlate with lower returns. I’m guessing that correlation is about as weak as the correlation is for selling in May!
The ‘Sell in May’ correlation could be caused by just a few outliers.
One of the more celebrated studies was published in 2002. ‘Sell in May’ was found to have statistically significant excess returns versus a buy-and-hold strategy. A couple of years later, Maberly & Pierce examined the results of that study and discovered that almost all of the statistical significance was due to a couple of outliers. The crash of October 1987 and the collapse of hedge fund Long-Term Capital Management in August 1998 both dragged down the summer month returns by large amounts. Excluding those two events revealed all of the other differences in returns to be insignificant statistically.
So, most studies didn’t find much of a correlation, but I did find one that was pretty interesting…
The President’s Third Year Theory
Here’s a pretty cool study that discovered the seasonal trade only works during the third year of a President’s term. The first, second, and fourth years had about the same returns between summer and winter. The third year had an excess gain of over 10% each winter. Their theory is that the mid-term election gives the most certainty of what the President’s final two years will be like. Will he have the Senate and House of Representatives working with him to further policies or will they work against him shutting down any progress? It doesn’t matter which occurs. The certainty of knowing is what matters. The winter period of the third year takes place right after that certainty is made known.
The authors take it one step further and say that nearly all of the equity premium over the last 200 years is accounted for during this period after midterm elections. That is, 98% of the excess returns for stocks above the risk-free rate occurred between December and April every four years. For their theory, you would sell in May (of the current President’s third year) and go away (until December of the next presidential term’s third year). Not as catchy of a slogan, is it?
Did their theory hold up during President Trump’s third year?
My first thought is there’s no way it holds up since we had that huge drop at the end of 2018, right in the middle of his third winter in office. Let’s see what the market data looks like:
The S&P 500 index opened on November 1, 2018, at 2717.58 and closed on April 30, 2019, at 2945.83. That is a return of 8.4% for those six months (16.8% annualized). That’s amazing to me considering the huge drop we had in December of 2018. Maybe there is something to Chan and Marsh’s study!
One last check – was the winter return significantly higher than the ensuing summer’s return? Sticking with the S&P 500 index, it opened at 2952.33 on May 1, 2019, and closed at 3037.56 on October 31, 2019. The summer months had a return of 2.89% or 5.77% annualized. Yes, the winter had clear outperformance versus the summer. Like most other studies concluded, though, you would’ve been better off just staying invested. Otherwise, you missed out on an extra 2.89% return.
Does anyone in modern-day times still follow the ‘Sell in May’ adage?
Based on the number of posts created in various forums all asking the same basic question, “does ‘Sell in May and Go Away’ work?”, there seems to still be a lot of interest in this strategy. People have put a lot of effort into justifying their opinions, as well. One even says the real saying should be ‘Sell in August and come back mid-October’ based on his research. There’s no way to tell if any of the commenters have put the strategy into play with real money, though.
Unfortunately, there also seems to be a lot of confusion on whether it’s a solid way to invest. Let’s clear that up now:
No, it is not a solid way to invest.
- First, it’s based on market timing, which is never a good foundation for investing.
- Second, the market timing is based on correlations that barely exist. Yes, winters may have higher returns than summers, but you’ll still be missing out on positive average returns each summer.
- Third, you will pay the highest amount of taxes possible because you will be paying short-term capital gains every year on all of your gains. That’s the same tax rate you pay on your last dollar of earned income, rather than the lower long-term capital gains rates.
To reiterate: Yes, the Halloween effect is real and winter months show higher returns than the summer months. The summer months are still positive, though, so staying invested will lead to higher long-term returns. The buy-and-hold advantage is even greater when transaction costs and taxes are taken into account.
My recommendation: Instead of fully selling everything and sitting in cash or bonds for half the year, why don’t you use the old maxim as a reminder to rebalance your portfolio? Once in the spring, again in the fall.
As Reddit commenter IGiveHoots so eloquently quipped: “Trade due to rhymes and you’ll lose all your dimes”