Why 'Sell in May' doesn't work for investors

[April 30, 2012]  NEW YORK (AP) -- It's simple to understand, and has a nice ring to it. It's made some investors look like geniuses recently. Say it enough and you might even believe it.

Too bad "Sell in May and go away" doesn't work.

The idea behind the popular investing strategy is that you can make more money in stocks by avoiding the summer and early fall when prices tend to languish. Better to sell your holdings April 30 before they flatline, then buy them back November 1 before they rise.

The "Sell in May" folks are up big this past year, assuming the market doesn't crash Monday, April 30, when the high returns will supposedly end. Trading stocks this way has allowed them to ride a 16 percent gain in the past six months while avoiding the 7 percent loss that the "buy and hold" fools had to suffer the previous six.

Whether the strategy will produce big profits again this year is a matter of heated debate. Some money managers point out that "Sell in May" doesn't work in the last year of president's term when he's trying to stimulate the economy to get re-elected. Others argue that it doesn't work when the Federal Reserve is pumping money into markets like it's done recently.

One of the biggest fans of the strategy, Jeffrey Hirsch, editor-in-chief of the Stock Trader's Almanac, thinks the strategy will succeed again, and that you should sell. "I suspect we're not going to gain a lot of ground between now and the presidential election," he says.

The math is compelling, at least if you look at the averages. Since 1926, the S&P 500 rose an average 4.3 percent in the six months of May through October versus 7.1 percent in November through April.

But "Sell in May" has a lot of holes in it.

For starters, note that stocks still rose nicely in those warmer months. For the strategy to work, you need to find an alternative for your money that yields more while you're out of stocks. And this alternative needs to be safe -- 30-day Treasury bills, for instance -- so you're guaranteed it'll all be there in the fall when you have to buy back your stock.

Larry Swedroe, head of research at Buckingham Asset Management, ran the numbers a few years ago using 30-day Treasurys during the off months, and says the strategy is bunk.

He looked at returns through 2007 from six start dates since 1950. "Sell in May" beat "buy and hold" if you started investing in 1960, 1970 and 2000, but not if you started in 1950, 1980 or 1990.

"It's pure randomness," Swedroe says. "How would you ever know when to start?"

Then there is the problem with using averages to say anything meaningful about stocks. If you know the average temperature for a month stretching back decades, you can pretty much guess how hot or cold that month will be this year.

But that's not true with stocks. They move too widely above and below their averages in most years, and in most subsets of years, for those numbers to be used to predict where they're heading.

Ken Fisher of money manager Fisher Investments says "Sell in May" is "garbage" precisely for this reason. He gives the example of September, which has dropped an average 0.72 percent since 1926. But the month has had many big up and down moves over 85 years. He says if you remove just two of the worst Septembers, stocks break even for the month.

"The average is made up of extremes," says Fisher, who devotes a chapter skewering "Sell in May" in "Debunkery," his book on investing myths. "If you steer by averages, you're going to get thrown off."

All this is not to say the strategy is useless. It's great for Wall Street, for instance. It lures you into the false belief that if you'd just buy and sell a bit more, you can beat the stock indexes. And more trading means more brokerage commissions and various fund fees, and the industry is sorely in need of them now that ordinary investors are pulling out of the market and trading volumes are low.

It helps keep journalists employed, too. A Nexis search shows "Sell in May" showing up in 963 articles and blog posts over the past two years. Make that 964 when this one is published.

You'd also be doing Uncle Sam a favor. Your profits from selling stocks in May are considered short-term gains. And so instead of paying a capital gains tax of 15 percent on your winnings, you'll pay marginal income tax rates as high as 35 percent. And you get to do that again next year and all the rest of the years.

Which makes you wonder whether "Sell in May" would yield even a penny more after you'd paid your fees and taxes. In the two out of three periods when Swedroe says "Sell in May" worked, the returns were higher than "buy and hold" by less than half a percentage point.

For all the wasted money and newsprint, you can blame the British. Some London traders believed they could make more money if they sold their holdings in May, then bought them back in September, when a famous horse race was run on "St. Leger's Day." The original saying was "Sell in May and go away, come back on St. Leger's Day" and dates back at least to the 1930s, when England still ruled India.

The empire is gone, but the saying has somehow survived despite time and logic.

The fact is, it isn't clear why stocks should rise more in colder months than in warmer ones. Some say it's because traders are largely on vacation in summer so there are fewer buyers pushing stocks up. But there are also fewer sellers, too. Plus, traders return in September and October while the markets are supposedly still lagging, so that explanation doesn't make sense.

Others have pointed to the fact that people tend to get depressed as the days get shorter in the fall. That leads them to sell, a perfect set up for a rally in the winter months. But the days also get shorter in November and December yet stocks are supposedly rising then, too.

If you're still intent on following "Sell in May," why not try variations on the theme?

Fisher, the "Debunkery" author, says slicing up the year differently yields "winning" strategies that offer that same magic mix of compelling and misleading. For instance, the S&P 500 over the past 85 years has gained an average 7.2 percent in March through August versus 4.4 percent in September through February. He wonders why people aren't saying "Sell in September and go away."

The answer: It doesn't rhyme.

[Associated Press; By BERNARD CONDON]

Copyright 2012 The Associated Press. All rights reserved. This material may not be published, broadcast, rewritten or redistributed.