« Earnings from Moog and Becton, Dickinson | Main | See a Pattern? »

From Gary Alexander:

One of the most widely-touted market calendar myths is that you should “sell in May and go away.” This system apparently worked well from 1950 until 2004. The theory runs like this: Brokers love to enjoy the summer market doldrums at the beach and then come back after Labor Day to sell stocks, causing a scary decline. Then, in November, they buy back for year-end “window dressing” in a “Santa Claus Rally.”

According to the Stock Trader’s Almanac (2006 edition), the 55-year Dow gain (1950 to 2004) from November 1 to April 30 was nearly 50-fold (or +7.9% per year, on average), while you actually lost money (-5%) in the May 1 to October 31 period. Final tally: +4,900% in cold months vs. -5% in warm months. Practically, this means investing in stocks November 1 and switching to fixed income on May 1.

Sounds convincing! Alas, this hoary old theory didn’t work in 2009: The S&P 500 gained 18.7% from May to October, even including the disappointing 250-point Dow drop last Friday. In the previous six months – in the supposedly superior November 1 to April 30 period – the S&P fell 9.9%. Over the past seven years, the November-April gains have averaged 1%, while the May-October gains averaged +3.3%.

So, ask yourself: Do you really want to throw away the 3.3% average recent gains from May 1 to October 31, based on a theory that worked last century? And do bank CDs or Treasury bonds on the “sidelines” offer you anything better than a 6.6% annual rate? And don’t forget the tax consequences of short-term trading. If you’re trading in a taxable account, six-month switches can decimate your after-tax gains.

My answer: Don’t sell in May, and don’t go away. Switching sectors might be prudent, but not leaving stocks altogether. However, if you’re looking for a good time to buy, November is still #1, historically.

Posted by edelfenbein at November 5, 2009 12:50 PM