"Sell in May and Go Away," or "Stay and Play?"
by Don Vialoux, EquityClock.com
We want to explore a topic of discussion that we've seen with greater frequency over the past few years. It pertains to the use of geometric means and arithmetic averages in the calculation of seasonal studies. Coming into the unfavourable period for stocks, I'm sure you've heard by now that the broad market, as gauged by the S&P 500 Index, has produced a flat result (+0.14%), on average, between May 5th and October 27th since 1950. Positive results were realized in 41 of the past 66 periods, or a 62.1% frequency of success. However, looking at a geometric mean of this unfavourable period, the result shifts negative, coming in at -0.40%, impacted by some significant drawdowns that have occurred over the past 66 years. This analysis does not take into account the impact of dividends, which can add considerable value to portfolios during a volatile summer period. Depending on volatility, the difference between the geometric and arithmetic mean calculations can often be insignificant. For ease of understanding, we tend to focus on the arithmetic mean in our publications and database, although we do look at both as part of our regular analysis.
Considering the geometric mean alone, the well known adage "sell in May and go away" seems to be justified as holding the S&P 500 Index strictly during the period of volatility for stocks would have resulted in a decline for equity portfolios since 1950. But what if we take each year as being independent of one another, rather than a series where investors just blindly invest during this period of seasonal weakness. The largest declines during this summer period have occurred in and around recessionary periods, the most notable of which was the last economic downturn that saw a May to October decline in 2008 of 39.7%. There has been 13 May to October periods with declines of 5% or more since 1950, the worst of which were the result of recessionary downturns. This is where fundamentals come into play. Generally, it is best to avoid equities when the economy is contracting, whether it be within the seasonally favourable or unfavourable period for stocks. So while it is easy to paint the summer as being a negative period, taking an unbiased, independent look at each period may be appropriate to assure that you're not selling stocks during a season that has shown more successes (positive years) than failures.
S&P 500 Index Returns between May and October | ||
Start Date | End Date | Return |
May 5, 2015 | October 27, 2015 | -1.13% |
May 5, 2014 | October 27, 2014 | 4.08% |
May 5, 2013 | October 27, 2013 | 9.00% |
May 5, 2012 | October 27, 2012 | 3.13% |
May 5, 2011 | October 27, 2011 | -3.78% |
May 5, 2010 | October 27, 2010 | 1.42% |
May 5, 2009 | October 27, 2009 | 17.66% |
May 5, 2008 | October 27, 2008 | -39.69% |
May 5, 2007 | October 27, 2007 | 1.97% |
May 5, 2006 | October 27, 2006 | 3.89% |
May 5, 2005 | October 27, 2005 | 0.53% |
May 5, 2004 | October 27, 2004 | 0.35% |
May 5, 2003 | October 27, 2003 | 11.29% |
May 5, 2002 | October 27, 2002 | -16.38% |
May 5, 2001 | October 27, 2001 | -12.79% |
May 5, 2000 | October 27, 2000 | -3.70% |
May 5, 1999 | October 27, 1999 | -3.76% |
May 5, 1998 | October 27, 1998 | -4.50% |
May 5, 1997 | October 27, 1997 | 5.62% |
May 5, 1996 | October 27, 1996 | 9.24% |
May 5, 1995 | October 27, 1995 | 11.46% |
May 5, 1994 | October 27, 1994 | 3.21% |
May 5, 1993 | October 27, 1993 | 4.52% |
May 5, 1992 | October 27, 1992 | 0.40% |
May 5, 1991 | October 27, 1991 | 0.89% |
May 5, 1990 | October 27, 1990 | -9.95% |
May 5, 1989 | October 27, 1989 | 8.92% |
May 5, 1988 | October 27, 1988 | 7.14% |
May 5, 1987 | October 27, 1987 | -21.04% |
May 5, 1986 | October 27, 1986 | 0.44% |
May 5, 1985 | October 27, 1985 | 4.13% |
May 5, 1984 | October 27, 1984 | 3.88% |
May 5, 1983 | October 27, 1983 | 0.34% |
May 5, 1982 | October 27, 1982 | 14.97% |
May 5, 1981 | October 27, 1981 | -8.46% |
May 5, 1980 | October 27, 1980 | 20.21% |
May 5, 1979 | October 27, 1979 | -0.12% |
May 5, 1978 | October 27, 1978 | -2.01% |
May 5, 1977 | October 27, 1977 | -7.76% |
May 5, 1976 | October 27, 1976 | 0.87% |
May 5, 1975 | October 27, 1975 | -0.39% |
May 5, 1974 | October 27, 1974 | -23.19% |
May 5, 1973 | October 27, 1973 | 0.34% |
May 5, 1972 | October 27, 1972 | 3.74% |
May 5, 1971 | October 27, 1971 | -9.63% |
May 5, 1970 | October 27, 1970 | 5.75% |
May 5, 1969 | October 27, 1969 | -6.13% |
May 5, 1968 | October 27, 1968 | 5.62% |
May 5, 1967 | October 27, 1967 | 0.55% |
May 5, 1966 | October 27, 1966 | -8.76% |
May 5, 1965 | October 27, 1965 | 3.12% |
May 5, 1964 | October 27, 1964 | 5.09% |
May 5, 1963 | October 27, 1963 | 5.68% |
May 5, 1962 | October 27, 1962 | -17.66% |
May 5, 1961 | October 27, 1961 | 2.74% |
May 5, 1960 | October 27, 1960 | -2.26% |
May 5, 1959 | October 27, 1959 | -0.57% |
May 5, 1958 | October 27, 1958 | 15.14% |
May 5, 1957 | October 27, 1957 | -12.41% |
May 5, 1956 | October 27, 1956 | -4.62% |
May 5, 1955 | October 27, 1955 | 11.95% |
May 5, 1954 | October 27, 1954 | 13.18% |
May 5, 1953 | October 27, 1953 | -3.08% |
May 5, 1952 | October 27, 1952 | 1.82% |
May 5, 1951 | October 27, 1951 | 0.18% |
May 5, 1950 | October 27, 1950 | 8.51% |
Arithmetic Mean: | 0.14% | |
Geometric Mean: | -0.40% | |
Gain Frequency: | 62.12% |
The tendency of the benchmark to perform in the manner that it does comes back to the constituents within it. Looking prior to 1950, the seasonal profile of the US equity market had a significantly different appearance than what it does today. The period of strength for stocks was actually realized in the summer, between May and September, while the September through to May period would see sluggish returns. This was the result of a primarily agriculture driven economy causing equity market returns to peak around the time of the annual harvest. Today, with the pickup in consumer spending in the fall and spring and increased manufacturing activity in the first half of the year, stocks have shown their best gains of the year between October and May, while returns during the off-season are rather lacklustre by comparison. The sectors that tend to draw on the major averages during the summer are those that benefitted from the upside catalysts derived by consumer spending and manufacturing in the favourable season, namely the consumer discretionary, industrial, and material sectors. Looking for opportunities outside of these areas can allow investors to weather the storm during the summer. Higher yielding areas of the market, such as health care, consumer staples, utilities, and REITs have historically provided value over and above the market return between May and October. More cyclical, energy and agriculture have seen strength at certain points of the summer months. And if volatility becomes particularly prominent, gold and bonds provide an ideal hedge through the third quarter. So while stocks are entering a six-month period where returns have historically been less than the six months that preceded it, on average, there always remains opportunities out there. "Sell in May and go away" has become a very broad statement that will inevitably garner criticism from a market that is dominated by fundamental analysts, but, if you look beyond the headline print, you may find ways to "stay and play" in seasonally favoured opportunities.
Copyright (c) EquityClock.com