"Sell In May And Go Away"

THE THEORY
The theory itself is simple: You sell stocks during May and you do not buy again until November. This theory is based on the belief that the market performs best between November and the end of April. The selling part seems to make sense during the summer, when most traders and brokers on Wall Street are on vacation. The theory itself is based on the Memorial Day to Labor Day period in the markets when the returns are the lowest. The truth is that mid-October seems to be when the market starts to move higher. Can such a simple saying be something we should heed? The only way to know for sure is to crunch the numbers and look at past performances.

The NYSE Composite Index is an interesting case study. From the start of May to the end of October the NYSE gained 2,096.72 points. From May 3 to October 30, 1998, the NYSE lost 408 points. The very next year during those same months it was down 198 points. So what about November to April over 10 years? From November to the end of April, the NYSE gained 5,106.87 points. There was only one down year and that was November 2000 to April 2001, when the NYSE was down 314 points. Clearly, there seems to be some merit to this "Sell in May and go away" business.

I decided to test the results of this saying on the Standard & Poor's 500. During May to the end of October, the S&P 500 was down a total of 26.74 points. From May 1 to October 31, 2001, the S&P 500 lost 207 points. The next year it was down another 199 points.

Taking a look at the colder months tells a completely different story. Starting in November and ending in April the S&P 500 was up a total of 500 points. From November 2, 1998, to April 30, 1999, the S&P 500 was up 224 points. There are a few more big-point gains like that in 1997, 2005, and 2006. The biggest loss suffered by the S&P 500 during November to the end of April was 172 points in 2000 into 2001. There is no question that if you had held from May to October you would be down. The saying seems to hold true when it came to the S&P 500.

The NASDAQ was also used to test out this theory. The results here are interesting as well. Starting in May and ending in October the NASDAQ gained 966 points from 1996 to 2006. This was aided by a roller-coaster ride in those months. From May 1 to October 31, 1997, the NASDAQ gained 323 points. The next year it lost 102 points. Starting on May 3 and ending on October 29, 1999, the NASDAQ gained a whopping 431 points.

As in true roller-coaster fashion there is always a big dip. In this case, May 1 to October 31, 2000, was the big dip. The NASDAQ was down 588 points. This pattern would continue for the next two years, as the NASDAQ was up 478 points followed by a loss of 350 points the next year. Those are the summer months.

The NASDAQ was no less spectacular during the winter months. From November to end of April the NASDAQ gained 1,034 points. From November 2, 1998, to April 30, 1999, the NASDAQ was up 741 points. The next year saw an even bigger gain of 893 points. From November 1, 2000, to April 31, 2001, the NASDAQ lost 1,217 points. The theory of selling in May did not work out so well when tested on the NASDAQ. The gains from November to April are indeed larger than those gains during May to October. The gap, however, is not that big, and you would have missed out on some nice returns.

The Dow Jones Industrial Average (DJIA) is the most fascinating of any of the indexes I tested this theory on. May to the end of October proved to be bad times for the DJIA. During those months it was down a total of 1,360.34 points from 1996 to 2006. From May 1 to October 30, 1998, the DJIA was down 555 points. This was followed the next year by another loss of 285 points. The biggest losses were yet to come. From May 1 to October 31, 2001, the DJIA was down 1,823 points. The next year it was down another 1761 -- huge losses, to say the least. From May 1 to October 31, 2003, the DJIA gained 1763 points.

As bad as the summer months were for the DJIA, they were that much better during the colder months. The period from November to the end of April proved to be the DJIA's strongest time. During those months it gained 7,519 points. During the period of November 1, 1996, to April 30, 1997, the DJIA was up 987 points.

The following year was even better, as the DJIA had a gain of 1,389 points. And the party did not stop there. Starting November 2, 1998, and continuing till April 30, 1999, the DJIA was up 2,083 points. From November 1, 2001, to April 30, 2002, the DJIA gained 682 points. The period of November 1, 2005, to April 28, 2006, the DJIA gained 960 points. The most recent period of November 1, 2006, to April 30, 2007, the DJIA gained 1,029 points.

To be fair, the DJIA has been up all year. From May 1 to October 31, 2006, the DJIA gained 737 points. The last two years have seen spectacular gains for the DJIA, regardless of the months. Until recently, this theory worked perfectly on the DJIA.

Recent gains have brought this theory into question. If you had exited during the May to October months, you would have missed out on a continuing rally. History does show that selling in May and then going away isn't such a bad idea.

When it comes to the DJIA and the S&P 500, though, the theory cannot be dismissed entirely. Clearly, it has been wrong the last two years, as the DJIA has been making tremendous gains despite the climate or the months.

In the end, it seems like a good idea to remember that the warmer months may be hazardous to your portfolio. For the investor who will not be able to keep an eye out on his or her portfolio during the summer months, making use of this theory could be the way to go. Those summer months can lead many of us to not keep as close an eye on our portfolios as we would normally do, and no one wants to end summer with a big loss in our portfolios.