A Season for Every Stock Strategy?

Sell in May and then walk away is an old Wall Street rhyme that just might have an element of reason in it

By Sam Stovall

The names on Standard & Poor's list haven't materially changed since late October, with just a few rotating in and out of the top 10. So this week, I thought I would take a different tack and dissect the "Sell in May and then walk away" adage, which derives from seasonal trends in which stock-price movements are substantially stronger in the November-April period than in May-October.

The old Wall Street rhyme means that investors should move into cash in May -- thus avoiding the coming weak period for stocks -- and back into stocks in November, when share prices typically begin to recover. Is this a valid strategy, or merely an old adage that has made its way into history books?


  Looking at the price action for the S&P 500-stock index during the past 30 years, it appears that the old saw has some merit. While the S&P 500 advanced an average of 4.3% during all six-month periods since 1972, the average increase for the S&P 500 in November-April was 7.1%, vs. an average gain of only 1% during the May-October periods. What's more, the November-April stretch outperformed the May-October period 70% of the time.

Why? History shows that the S&P's worst month is September and that the worst three-month period is the third quarter. What's more, October is historically a bottoming month, so the S&P 500 enters November at a fairly low level compared to other months, giving the November-April period the advantage of starting at a low base.

It also includes two periods of large cash infusions: January, when pension funds typically put a lot of money to work, and April, when many individuals add to their IRAs. And it's around this time of year that analysts begin looking ahead by five quarters, rather than just focusing on the final one or two.


  Have sectors experienced a similar skewing of performances? During the past 12 years (the most available for S&P 500 sector data), it appears that they've also seen a favorable and unfavorable pattern.

The accompanying table shows price changes (without dividends reinvested) during the two six-month periods for sectors in the S&P 500 that have been around since 1991. As with the overall market, it shows that most sectors and industries have also seen their best results in the November-April period. What's more, on average, none of the groups posted negative results during the November-April timeframe, while half of the sectors in the 500 posted average declines May-October. Only Health Care showed superior performances in the May-October timeframe, while Consumer Staples had a relatively similar performance in both.

Percent Changes (w/o dividends) 1991-2002
Sector May-Oct. Nov.-Apr.
Consumer Discretionary -1.3 14.0
Financials 4.9 13.3
Information Technology 2.5 13.2
Materials -4.4 11.4
Industrials -0.2 11.1
S&P 500 1.5 8.7
Energy -0.2 6.8
Health Care 7.4 6.3
Consumer Staples 5.5 5.8
Telecommunication Services 1.6 2.3
Utilities -0.2 1.8

A market timer who was long on stocks from November through April but then moved into cash from May through October would have outperformed the broad market only marginally during the entire year. And after taxes and commission costs, it would hardly be worth the effort.

The most valuable lesson this study offers is how to invest during both periods: Had an investor taken an aggressive stance by rotating into the S&P 500 from November through April and then a defensive approach by embracing either health care or consumer staples from May through October, he or she would have significantly outperformed a mere buy-and-hold investor. The 12-year (October, 1990, through October, 2002) compound annual growth rate for the S&P 500 was 6.4% (without dividends reinvested), while the approach of owning the S&P 500 from November through April and then rotating into the S&P Health Care sector would have returned 11.5% per year.

A similar rotation approach that substituted consumer staples for health care would have returned 8.8% on average. Of course, no one can guarantee that what worked in the past will continue to work in the future.

S&P Relative Strength Rankings

These industries carry 12-month relative strength rankings of "5" as of Dec. 6, 2002 -- meaning they're in the top 10% of the 116 industries in the S&P Super 1500 (the combined S&P 500, S&P MidCap 400, and S&P SmallCap 600) based on prior 12-month price performance.

Industry/Sector Largest Company (Market Cap.) S&P STARS* Rank
Apparel, Accessories & Luxury Goods/Consumer Discretionary Quiksilver (ZQK) 5 STARS
Brewers/Consumer Staples Anheuser-Busch (BUD) 4 STARS
Casinos & Gaming/Consumer Discretionary Harrah's Entertainment (HET) 4 STARS
Consumer Electronics/Consumer Discretionary Harman International (HAR) Not Ranked
Distillers & Vintners/Consumer Staples Constellation Brands (STZ) 5 STARS
Gold/Materials Newmont Mining (NEM) 4 STARS
Housewares & Specialties/Consumer Discretionary Newell Rubbermaid (NWL) 5 STARS
Metal & Glass Containers/Materials Pactiv (PTV) 4 STARS
Photographic Products/Consumer Discretionary Eastman Kodak (EK) 1 STAR
Publishing & Printing/Consumer Discretionary Gannett (GCI) 5 STARS
Trucking/Industrials Yellow Corp. (YELL) 5 STARS

*S&P's ranking system for the appreciation potential of stocks over a 6- to 12-month period: 5 STARS (buy), 4 STARS (accumulate), 3 STARS (hold), 2 STARS (avoid), 1 STAR (sell).

Stovall is chief investment strategist for Standard & Poor's

Before it's here, it's on the Bloomberg Terminal.