The Sell in May Effect describes the tendency of the market over the six-month period Nov-Apr to outperform the market in the other six-month period (i.e. May-Oct).
The effect can be seen in the following chart, which plots the cumulative average daily returns of the market (i.e. it gives a representation of the market moves in an average year). More information on this chart can be found here.
As can be seen the market tends to be strong from November to April, and then flat for the six-month period May to October.
The problems of exploiting the Sell in May Effect
Although the effect is statistically significant, it is not an easy anomaly to exploit economically. In theory an investor might be long stocks Nov-Apr and then move to cash for May-Oct. But as can be seen in the above chart, the market doesn’t necessarily fall in the summer period (except possibly the short May-Jun period), rather it is flat. And by moving to cash the investor would forego dividends paid in the May-Oct period.
It may make sense moving to cash if interest rates were high (i.e. to benefit from high returns on cash for the summer period) – but that is not the case currently. And in any case that has to be balanced with the fact that when interest rates are high expected growth rates in equities tend to be high as well (i.e. not a time to be out of the market).
One significant reason why it may make sense to be out of the market over the summer period is that volatility is much higher then than in the Winter period (as shown here). For example, eight of the ten largest one day falls in the FTSE 100 Index happened in the Summer period. Hence, not only are returns lower in the Summer, but also risk-adjusted returns are significantly lower.
But generally, this is a little frustrating: the Sell in May Effect is a significant market anomaly, but tricky to exploit.
So, what to do?
Exploit the sector rotation
One idea is to stay in the market throughout the year but to re-balance a stock portfolio according to which sectors perform the best in the two six-month periods as defined by the Sell in May Effect.
The following two tables show the performance of the FTSE 350 sectors in the respective summer and winter periods since 1999. The tables have been ranked by average returns of the respective sectors over the 17-year period.
Sector performance in the summer period since 1999
Sector performance in the winter period since 1999
From these tables two portfolios of sectors can be constructed that have historically performed strongly in the respective summer and winter periods.
A few filters were applied:
- Sectors with less than 4 component stocks were not considered
- Sectors must have a minimum 13-year track record
- Standard deviation (i.e. volatility) of a sector’s returns must be below the average standard deviation
- Positive returns must be over 50%
The portfolios selected were-
|Summer Portfolio||Winter Portfolio|
|Gas, Water & Multiutilities||Construction & Materials|
|Health Care Equipment & Services||Chemicals|
So, the Sell in May Sector Strategy (SIMSS) is
- in the summer period: long sectors Gas, Water & Multiutilities, Beverages, and Health Care Equipment & Services
- in the winter period: long sectors Construction & Materials, Industrial Engineering, and Chemicals
Performance of SIMSS
The following chart shows the simulated performance of the Sell in May Sector Strategy backdated to 1999 compared to the FTSE 100 Index.
After 17 years the SIMSS portfolio would have grown in value to 1021 (from a starting value of 100). While the FTSE 100 (buy and hold) portfolio would have grown to 111.
This simulation does not include transaction costs, but as the strategy only trades twice a year these would not significantly change the above results.
More articles about the Sell in May Effect.