It was my birthday recently. It appears with reassuring regularity. In fact, most of my life has a pattern, a routine, a seasonality. But it’s not just me – people in general have habits, and these can explain the lunchtime queue at Tesco or the quiet office in the summer. These patterns mean that we could opt to avoid the holiday motorway rush or buy our sarnies earlier. By recognising the seasonality of human behaviour, perhaps we can make better decisions.
There are patterns in the stock market too. The most famous annual pattern even has its own rhyme: "sell in May and go away", with an additional "and come back on St Leger Day" for the Brits. According to the finance website Investopedia, "this phrase refers to the custom of aristocrats, merchants and bankers who liked to leave the City of London and go to the country to escape the heat during the summer months. St Leger Day refers to the St Leger Stakes, a thoroughbred horse race in mid-September." Surprisingly, there is some statistical evidence to support this.
In the first academic analysis of this pattern in 2002, Bouman and Jacobsen found higher returns in 35 of the 37 global equity markets they studied from November to April than in May to October. They found that the "sell in May" effect tended to be particularly strong in Europe.
Could this really be down to City types going on holiday, or does it reflect a seasonality of the underlying economy or earnings results? It is difficult to establish causation, but whatever the reason the statistics say that summer/autumn is a weaker time in markets than winter/spring. What should we do with that information? Should we sell in May and buy back in November? I’m not convinced.
First, the statistics don’t actually say that markets in general go down over the summer. Data from the US equity market since 1950 shows it increasing on average by 1 per cent from May to October. September is the weakest month, generating a negative return in 38 years out of the 68 reviewed.
Second, the statistics aren’t reliable for any single month. Look at the September statistic again. The worst month gives you a negative return in 38 out of 68 years, or 56 per cent of the time. That means it gives you a positive return in each of the 30 other years, or 44 per cent of the time. In other words, if you sold you’d be wrong nearly half of the time. As an example of the perils of short-term market timing, in 2010 September was the best month for the US equity market, up by almost 9 per cent. And, assuming you are a long-term investor, you’d need to buy back into the market in October, incurring transaction costs along the way.
There are other cycles: some shorter, such as the best day to buy (Monday); others longer – there is some evidence that the second half of any decade is better than the first half, with the best-performing years ending in a five.
But although these seasonal factors are interesting, none of them is statistically powerful enough to influence investment decisions in a meaningful way. They speak of short-term patterns and aren’t based on fundamental analysis. For long-term charity investors, remaining invested and watching the seasons come and go is the most robust strategy.
Kate Rogers is head of policy at Cazenove Charities