There have been numerous discussions of whether to "Sell in May" this year. My personal impression is that it is probably better to play the odds, especially if the odds favor reducing risk.
I performed some analysis on the recent history of Sell in May using 158 components of the S&P 100 and NASDAQ 100. The analysis ignores dividends and looks at the adjusted close prices. For Sell in May, a given stock is bought on November 1st and sold on May 1st. If there is no trading that day, the first trade day prior to that is used.
I removed Apple (AAPL), Monster Beverage (MNST), Research In Motion (RIMM), and Wynn Resorts (WYNN) from the sample as their returns seemed to skew the numbers a bit. For example, AAPL had a return of 5,000% from 11/1/03, while MNST was over 15,000%. Any other omissions were for purely technical reasons, such as not enough price history.
From October 31, 2003 until May 1, 2007:
- Buy and Hold returned an average of 86%
- Sell in May (and buy in November) returned an average 137%
Extending the window from October 31, 2003 until April 30, 2012:
- Buy and Hold returned an average 108%
- Sell in May returned 211%
Of the 158 stocks, 25, or 16%, had a Buy and Hold gain versus Sell in May. Of these 25, 10 (6%) had better than a 100% gain versus Sell in May.
Contrast this with the Sell in May figures. 84% did better with Sell in May. 80 (51% of all stocks) had over a 100% gain versus Buy and Hold. 24 (15%) had a greater than 200% gain. We see that performance is substantially better for selling in May.
The lesson here is clear: Unless you are holding some monster performers, it has historically been significantly better to Sell in May. It may appear that very strongly performing stocks should be held; this deserves research. However, some pretty strong stocks, like Caterpillar (CAT) have done much better with Sell in May. CAT has returned 240% since 2003 but 484% with the Sell in May holding pattern.
Perhaps it is also worth noting that while 32 of the 158 (20%) have lost market value since 2003, there is not a single Sell in May loser.
Advice to hold this year may well be correct, but I don't see a compelling case to have the additional risk exposure.