Sell in May and go away? No way

With the U.S. equity market reaching record highs this summer, the old adage of “Sell in May and go away” doesn't seem to be holding true. But proponents of the seasonal theory, also known as the “Halloween indicator,” say not to fear. For one thing, there are several months of trading left before the fall. For another, the seasonal factors that seem to make make May through October the diciest months to own stocks don't come into play every year. “Whenever the market doesn't go into correction in May, people always start talking about how seasonal timing doesn't work,” said Sy Harding, publisher of the Street Smart Report newsletter, which offers a model portfolio based on the strategy. The seasonal theory “does not say that the market is going to go into a correction on May 1,” he said. “All it says is that it will probably be lower by October or November.” Whether that happens this year remains to be seen, said Jeff Hirsch, editor-in-chief of the Stock Trader's Almanac. Since May 1, the S&P 500 is up 5.9%. The Stock Trader's Almanac offers a Best Six Months seasonal strategy to subscribers. Mr. Hirsch's father, Almanac founder Yale Hirsch, first published the trading model in 1986. “Other forces are weighing against the normal seasonal forces” that contribute to a typical summer doldrum period, Mr. Hirsch said. “There's a whole lot of quantitative easing [by the Federal Reserve] going on.” Even if the seasonal effect doesn't appear this year, it won't be a disaster for followers, Mr. Harding said. “It just means you don't make money in the summer” while sitting in cash, he said. Plenty of skeptics think that the seasonal theory is based on a statistical fluke. “If you look at [market] statistics over a long period of time, from May to November, and the market does something 75% of time, that's an impressive number,” said Greg Morris, chairman of the investment committee for Stadion Money Management LLC. But “you can't make an investment decision from that,” he said. “What if you start following that [pattern], and you have four bad periods in a row?” Mr. Morris asked. Unlike other market-timing theories, though, “the historical evidence in favor of this [seasonal] one is surprisingly strong,” Mark Hulbert, publisher of The Hulbert Financial Digest, wrote in his April MarketWatch column. Mr. Hulbert, who tracks investment newsletters, figures that over the past 50 years, the Dow Jones Industrial Average has gained an average 7.5% during the winter months and lost an average 0.1% during the summer months. Although researchers have documented the Halloween indicator in markets around the world, the causes are unknown. “Because many things tend to be correlated with the seasons, it is hard to distinguish between these causes when trying to link stock returns to … potential explanations” for the anomaly, according to a 2005 paper by academics at Massey University in New Zealand. MONEY FLOWS Mr. Harding thinks that seasonal patterns of market performance are driven by flows of money. “There are big chunks of money that come into investors' hands” in the fall through spring, he said, including year-end profit and bonus payments, company 401(k) contributions, dividend payouts and tax refunds. “Once those extra chunks of money dry up in the summer, it doesn't leave any money available to buy the dips, and it leaves the market much more vulnerable to any pressure,” Mr. Harding said. Mr. Hirsch and other researchers theorize that the summer doldrums in the markets coincide with the summer vacation season. The seasonal pattern has ap¬peared only since 1950, once the U.S. had moved away from an economy dominated by agriculture, he said. In fact, from 1900 to 1951, August was the best month for stocks, due to cash earned from farm harvests, he said. Since 1987, though, August has been the second-worst month of the year for U.S. stocks.