It's almost May and talk about "sell in May, go away" is almost certain to emerge. Such market adages can make some sense for short-term traders, and portfolio managers might even be able to occasionally succeed in micro-timing stock market fluctuations, giving the adage some validity. Overall, however, history shows that paying too much attention to seasonal patterns comes with elevated performance risks and opportunity costs, especially in bull markets.
Just because something is repeated like a mantra over and over again, it does not have to be true. This certainly seems to apply to the old “sell in May, go away” adage, as well as its somewhat less-known addendum “…but remember to come back in September". Generally, this supposed market rule doesn’t really live up to the real world.
First, let us look at how the S&P 500 performed in the past 65 years in May. For simplicity, we focus on price movements, ignoring dividends payments. Since 1949, the US equity benchmark’s monthly performance has been negative in May a total of 28 times, or in 43% of the time. The median number for all months stands at 27, with the number ranging between 18 and 32. Meanwhile, the median 30-day performance after 1st May over the past 65 years has been 0.9%, which is only a little below the median for all months, which stands at 1.1%. Thus, on average, May is not much less profitable than most other months.
On the other hand, there is a grain of truth in the saying insofar as that market performance has been negative more frequently in the four months that follow, i.e. in June (30 times), July (28), August (31) and September (32). The median monthly performance for these four months is between 0% (for September) and 0.8% (for June). This shows that the stock market does go through a summer lull of some sort. Hence, investors with short-term time horizons who automatically sell all positions on 1st May are slightly more likely than not to avoid this period of potential weakness.
At the time, it is clear that the rule is not a good guide for long-term investors with a reasonable tolerance for market volatility. Even if transaction costs were very low, the probability of getting out in May and returning in September at ideal market levels is very low in reality - probably too low to forge a reliably investment strategy in the longer term. Thus, over time, opportunity losses (e.g. delayed later entry at a higher price) and transaction costs are likely to outweigh any theoretical performance optimization.
Moreover, even though the 1st May-30th September period tended to be comparatively weak historically, that weakness is more than offset by a strong recovery during the remainder of the year nearly 80% of the time. Since 1949, between 1st May and 31st December, the S&P 500 has posted gains 70% of the time, with the median gain standing at 5.8% (or 8.8% annualized), compared to 78% of the time and a median gain of 5% (or 21.9% p.a.) for the End-September to End-December period. Gains are slightly more frequent and also higher on an annualized basis in the latter period, which would seem to validate the “sell in May” saying.
Still, in practice, we could probably just as well rephrase the rule into “don’t sell in May, just remember to stay long through December," and be equally correct. That’s because those who sell their stocks at the start of May miss out on a median market gain of 2.5% through the end of September. Consequently, the long-only investor should actually outperform over the long term. Lastly, extra caution with regards potential seasonal market patterns is particularly advisable in a classic bull market - all the more so in an bullish environment in which equity valuations are not excessively high, corporate earnings continue to be consistently underestimated in most major markets, and when practically all advanced economies are experiencing a synchronous cyclical economic upswing.
To download the publication, click on "LGT Beacon" below. To subsribe, click on "Abo".