Author: Patrick Larkin
Covestor model: All Cap Value
Spring is here, and market timers' fancy has turned to thoughts of whether or not to "sell in May and go away." Not being content with proverbs when it comes to questions of wealth and poverty, I examined the data on value-weighted monthly U.S. stock market returns from January 1927 through December 2011 from Dartmouth professor Kenneth French's website.
The "sell in May and go away" rule takes the investor out of the market at the end of April and puts him back in at the beginning of November. Historically, U.S. stocks have returned 6.82% (14.11% annualized) in the November through April period, versus only 2.51% (5.07% annualized) in May through October.
While there is a big difference between 14.11% and 5.07%, before doing anything rash, investors might want to consider whether they can find a place to park their funds for the next sixth months that has an expected annualized return greater than 5%.
They might also want to consider whether they wish to liquidate and pay taxes on all of their winning positions, especially those that they have not held for a year and that do not qualify for the flat 15% rate on long-term capital gains.
An important question for investors is whether the lower historical returns from May through October are likely to be repeated in the future. If there is some good reason for stocks to under perform in May through October, then maybe the trend will continue. The only explanation that I've heard for "sell in May and go away" is the "Hamptons theory."
Maybe all of the buyers are away at their beach houses for six months a year, depressing demand for shares. One problem with this theory though is that prime vacation months July and August have actually been strong months for the U.S. market historically, with annualized returns of 13.02% and 9.50%, respectively.
Unsurprisingly to those familiar with market history, the only losing months for the U.S. market historically are September and October, with annualized returns of -14.61% and -2.78%, respectively. The best months have been December and January, with annualized returns of 18.73% and 14.26%, respectively.
If we exclude September and October from the May through October period, and exclude December and January from the November through April period, the difference in annualized returns is only 20 basis points (10.99% versus 10.79%).
I don't know exactly what to make of these figures, but I thought that they might be of interest to investors who are considering organizing their investment strategies around the Gregorian calendar.