Another Look at Why the "Sell in May" Strategy Fails

(MoneyWatch)  Yesterday, we looked at why the "Sell in May and Go Away" strategy doesn't hold up. Here's some further evidence on how the data can be manipulated to look like it works and why you still shouldn't follow such a strategy.

When we looked at the historical data yesterday, we examined the longest period available (as that's usually the most meaningful). We can also examine the results of the "Sell in May and Go Away" strategy for various other periods.

(As with yesterday, the Buy & Hold portfolio is invested in the S&P 500 Index throughout the time period. The Sell in May portfolio is invested in the S&P 500 from November through April and in 30-day Treasury bills May through October. Returns are annualized.)

January 1926-April 2010 Buy & Hold -- 9.86 percent
Sell in May -- 8.18 percent

January 1950-April 2010 Buy & Hold -- 11.11 percent
Sell in May -- 10.83 percent

January 1960-April 2010 Buy & Hold -- 9.54 percent
Sell in May -- 10.33 percent

January 1970-April 2010 Buy & Hold -- 9.97 percent
Sell in May -- 10.64 percent

January 1980-April 2010 Buy & Hold -- 11.36 percent
Sell in May -- 10.78 percent

January 1990-April 2010 Buy & Hold -- 8.43 percent
Sell in May -- 8.53 percent

January 2000-April 2010 Buy & Hold -- -0.26 percent Sell in May -- 2.85 percent

As the data demonstrates, there have been periods when the strategy "worked," at least before any considering implementation costs (trading costs or taxes). Consider the following:

From 1960 through 1979, the "Sell In May" strategy provided an annualized return of 9.65 percent, outperforming a buy and hold strategy by 2.82 percent per year. The problem is that the strategy had underperformed prior to that period. From 1926 through 1959, it underperformed the buy and hold strategy by 5.26 percent per year. Thus, there's no way any one would have known that the strategy would work beginning in 1960.

It's important to note that while strategies don't have costs, implementing them does. The "Sell in May and Go Away" strategy failed without even considering the impact of either trading costs or taxes. For investors who have their equities in taxable accounts (which is preferable to tax-advantaged accounts), such a trading strategy would be highly tax inefficient.

The above example of cherry picking a starting period and ignoring expenses is often the explanation for how myths arise. Randomly, we would expect almost any "system" (at least before expenses) to work ­just as there were times when the "Sell in May" strategy "worked." But there's an old adage about even blind squirrels finding an occasional acorn.