One of the most popular sayings on Wall Street is “sell in May and go away” because, proponents argue, you could achieve 100% of the market gains and avoid the biggest risks. This was certainly true in years like 1929, 1987 and 2008 when market crashes in the fall led to large losses. But is it always a good idea to sell in May? Rather than looking at a few years when the strategy worked, let’s put the idea to a test using a longer history.
We can test the idea using the Dow Jones Industrial Average from January 1900 through the end of April 2016, a 100-year test period. The idea we are testing is that you should own stocks only for the six months beginning in November and ending in April, a time period we’ll call “the best six months.” The rest of the year will be considered “the worst six months.”
For the test we need a baseline. Over the entire time period, the Dow gained an average of 5.2% a year. When the Dow is calculated dividends are ignored and they will be ignored in this article so that test results are directly comparable to the construction of the index.
For the first test, let’s assume you held stocks only during the best six months. Your average annual gain would be 5.2%, equal to all of the Dow’s gains. Now let’s assume you held stocks only for the worst six months. This time period shows an average annual loss of 0.3%. Based solely on this quick test, it does look like you should be out of stocks for half the year.
Before analyzing the results in detail, I would like to note you could use any index for these tests and you’d get similar results. I’m using the Dow because it has the longest history but results for the S&P 500 since 1928 look almost exactly the same. So do results on the Nasdaq Composite Index using data from 1984. The same effect is seen in international markets which I confirmed by testing indexes including Japan’s Nikkei benchmark index and Poland’s Warsaw WIG index.
The best six months do exist but the question is will the pattern hold in the future. We can’t know for sure but we can determine if the best six months are a statistical fluke.
Digging deeper into the results shows the winning percentage is similar for each time period. We see winning trades 69.3% of the time for the best six months and 60% of the stocks deliver gains in the worst six months. This indicates there is a high likelihood the results are being skewed by just a few big winners or losers.
Looking at the best six months, we learn the five biggest winners account for half of the overall gains and the seventeen biggest winners account for all of the gains. So the best six months is entirely explained by bull markets in 14% of the 116 years.
For the worst six months, a similar pattern emerges. The worst year, 2008, explains why this six-month period shows a loss. Excluding that year, we have an average annual gain of 3.6% in the worst six months. This is still lower than the best six months but we are seeing that the idea of there being a best or worst six months is driven by the results of just a few years.
This is confirmed by looking at the percent of time the index rises in any given month. The chart below shows that since 1900, there are only two months when the index fell more than 50% of the time – May and September, which are part of the worst six months.
Most of the time, stocks go up and the fact that a best and worst six months exist, is driven by the results in just a few years. This means there is really no reason to expect the best six months to deliver a gain in any given year and there is no reason to expect a loss, or even subpar performance, in the worst six months. In fact, quite often (60% of the time) the worst six months deliver gains. Being in cash means you miss out on these gains.
Even worse, there is no guarantee crashes that occur in the future will be in the fall. It’s entirely possible crashes can occur during the best six months, as they did in the 2008 and 2009 bear market.
Despite the fact that the best and worst six months are really just a statistical artifact, the idea has drawn a lot of attention and many researchers have studied the question of how to trade the best and worst six months of the year. One of the most interesting studies was done by Sam Stovall, a highly respected researcher at Standard & Poor’s.
Stovall confirmed you could earn all of the market gains in the six best months of the year, while avoiding much of the market risk. He also found some reasons the market should rise more from November through May than in does in the summer and fall – vacations and retirement accounts. Stovall noted many Wall Street pros take some time off in the summer and this reduces trading volume. He also noted Wall Street pays large bonuses, billions of dollars a year, early in the year and as pros invest those bonuses into their own funds before April, the market goes up.
After explaining exactly why it might be a good idea to sell in May, Stovall dug deeper into the data and discovered a simple way to turn the summer and fall into a profitable time of year. He found defensive sectors, stocks investors turn to when they are concerned about risk, outperform during the worst six months. In particular, he discovered investing in consumer staple and healthcare ETFs during the worst six months could deliver market-beating results, an average gain of about 10% a year, instead of the average gain of 5.2% a year obtained by switching to cash.
Stovall’s strategy is simple:
- Buy SPDR S&P 500 ETF (NYSE: SPY).
- Sell half of your SPY holdings at the close on April 30 every year.
- Split the proceeds from that sell evenly between Consumer Staples Select Sector SPDR ETF (NYSE: XLP) and Health Care Select Sector SPDR ETF (NYSE: XLV) and buy these two ETFs at the close on April 30.
- Sell XLV and XLP at the close on September 30 every year.
- Reinvest the proceeds from selling XLV and XLP into SPY by buying that day.
Repeating this process every year doubled the performance of a simple sell in May and go away strategy. Stovall noted this was likely due to the fact that most years, stocks go up in the worst six months and these ETFs provide exposure to those gains. In the rare down years, these ETFs tend to lose less than the market.
Of course it’s possible to improve on Stovall’s strategy. Instead of buying the ETFs, we can buy the stocks in the ETF that have the highest probability of beating the market over the next six months. This can be done with a simple relative strength (RS) strategy.
To implement an RS strategy, you select a group of stocks, calculate their returns over the past six months and buy the top performers. Studies have shown the top performers tend to beat the market over the next six months.
Right now, this strategy points to six stocks as buys:
- Johnson & Johnson (NYSE: JNJ)
- UnitedHealth Group Incorporated (NYSE: UNH)
- Bristol-Myers Squibb Company (NYSE: BMY)
- Philip Morris International, Inc. (NYSE: PM)
- Wal-Mart Stores Inc. (NYSE: WMT)
- Colgate-Palmolive Co. (NYSE: CL)
These stocks are buys because they have high RS and follow the rules Stovall provided for trading the worst six months. Buying these six conservative stocks helps generate income and reduces downside risks in the unlikely event of a market crash.