Busting the Sell in May and Go Away Myth

MythBusters is a popular Discovery channel television program. It uncovers the truth behind widely held myths and legends from whether it’s possible to beat police radar detectors to whether a bull really causes destruction in a china shop. The curious hosts mix scientific method with plain old-fashioned ingenuity to separate fact from fiction among many popular beliefs.

The investment industry also has a Mythbuster of sorts. It’s CXO Advisory Group of Manassas, Virginia. The firm separates fact from fiction among noisy investment strategies, analyst predictions and trader beliefs. CXO’s mission is to challenge any and all conventional market wisdom with analytical skepticism.

There is a belief among many investors that excess returns can be earned in the market by owning stocks for only 6 months out of the year and sitting the rest of the time in interest-bearing Treasury securities. This market-timing strategy is commonly known as Sell in May and Go Away because the exit period is May.  CXO Advisory took a look at “Sell in May” over the long-term to determine if this belief was fact or fiction.

Figure 1 compares on a logarithmic scale the cumulative values of a $1.00 initial investment for three portfolio strategies over 142 years ending in 2012. The three strategies are:

  1. Buy and hold stocks over the entire 142-year period (green line).
  2. Hold stocks from May-October and Treasuries for the remaining 6 months (blue line).
  3. Hold stocks from November-April and Treasuries for the remaining 6 months (red line – the sell in May strategy).

Stock returns are courtesy of Robert Shiller’s database at Yale University. This dataset consists of monthly stock index prices and dividends paid since 1871. Dividends were accrued and reinvested in each strategy while the portfolio was in stocks.

The “cash” return while out of stocks is represented by a risk-free rate using combined Treasury data. A 1.45 percent term-spread between long-term rates and short-term rates was used over the period (see the CXO website for an explanation of the interest rate used in the calculations).

Certain assumptions were made about frictional trading costs in each portfolio. CXO used a 1 percent one-way trading cost when a strategy moved into or out of the stock market. They also report results for no frictional trading in Figure 2, which brings the strategy results closer together.

Figure 1: The Long-term Results of Three Portfolio Strategies

Source: CXO Advisory Group, 2013

The clear winner in this three horse race was a buy and hold investment strategy (green line). $1 grew to nearly $200,000 over the entire 142 year period. Sell in May and go away (red line) did outperform the opposite strategy (buy in May – the blue line), although the returns of both market-timing strategies fell far short of buy and hold.

CXO tested several trading cost, dividend payment and interest rate assumptions in their model. Figure 2 illustrates the results of five scenarios: baseline as represented in Figure 1; no trading costs; no dividends paid or reinvested; no trading costs and no dividends; and no interest earned on cash.

Figure 2: Three Strategies under Various Trading Cost, Dividend and Interest Assumptions

Source: CXO Advisory Group, 2013

Sell in May and Go Away achieved superior results only in the 4th scenario from the left where there was no trading cost and dividends were ignored. Neither of these assumptions is realistic. There is always a cost to trade and stocks do pay dividends even in the worst of times.

At best, we can make the assumption that a 1 percent frictional cost to trade is too high. This might place the returns for the strategies between the baseline (1st on the left) and frictionless (2nd from the left). The return for buy and hold was still considerably higher.

Market timing does not work, at least not using Sell in May and Go Away. As with most market-timing strategies, a little bit of analysis often reveals the devil in the details.

A better strategy is to own a diversified portfolio of asset classes based on your needs, and then rebalance occasionally. This strategy is highlighted in All About Asset Allocation. It’s a portfolio management method that’s sure to capture your fair share of market returns.

I wish to thank Steve LeCompte of CXO Advisory for providing data used in this article. He is CXO’s founder and is solely responsible for web content. LeCompte was a project officer on the staff of Admiral H.G. Rickover at Naval Reactors; engineering and program management for complex systems with IBM; and a senior consultant and executive manager with International Data Corporation. He holds a B.S. in physics, summa cum laude, from Miami University, an M.S. in Physics from the University of Michigan, and is a graduate of the Program Management Course of the Defense Systems Management College.