Blogs | April 19, 2013

What’s the idea behind the share market saying “sell in May”?

At last, spring has arrived. You might think that while this fact is pleasant enough, it has nothing to do with economics. You’d be wrong.

Although many people work in air-conditioned offices and factories and agriculture now accounts for only a tiny fraction of economic activity, the coming and going of the seasons still matters enormously.

Is there a season to buy shares?
For one thing, changing seasons tell us the time is approaching to sell shares. Ben Jacobsen and Cherry Zhang of Massey University in New Zealand have established this. They studied as many national stock markets as they could find, over as long a period as data were available; this comprises 108 national markets, some dating back three centuries. They found that, on average, shares fall by 0.18% between May Day and Halloween, but rise by 4.35% on average between Halloween and May Day.

In 81 of 108 countries with stock markets, returns are worse in the summer than in the winter. Of course, this doesn’t mean prices always fall between May Day and Halloween. But it does mean the odds are against the market doing well. “Sell in May” is as reliable a share market rule as we’re likely to get.

Brighter mood, bigger bubbles
How can this be? An experiment by Yaron Lahav and Shireen Meer, two Israeli economists, suggests an answer. They got people to trade some artificial shares under laboratory conditions. Before trading, some people were shown a film of a Jerry Seinfeld comedy routine and some weren’t. They found that people shown the film traded the same asset at higher prices. “Positive mood leads to bigger bubbles” they concluded.

A similar thing happens in spring. The warmer weather and lighter nights brighten our mood, and this may lead to greater optimism about the economy and increased willingness to take risk. This can push share prices up to unsustainable levels.

Halloween shock?
The opposite happens in the autumn; darker nights might increase our anxiety and depress us, which forces share prices down too far. It’s no coincidence that traditional May Day celebrations are cheerful whilst Halloween is a time of dread and fear. The fact that it’s spring has another implication. It means the German economy is now almost certainly recovering, while the UK’s is in recession.

You won’t know this from the figures that get reported in the news, because official statisticians tweak the GDP numbers to remove seasonal fluctuations. But such fluctuations are huge. Since 1992, the German economy has shrunk every year in the first quarter by an average of 2.5%; that’s almost 10% on an annualised basis. The UK economy has also shrunk in all but one of those periods, by an average of 2.8%.

Yes, that’s right. These two countries have had a "recession" almost every year for the last 21 years, in the first quarter.

We prepare for holidays
However, the German economy has almost always bounced back in the second quarter, growing by an average of two percent. The UK economy, however, has not; it has shrunk by an average of 1.5% in Q2. It’s in the third quarter that growth gets going, with Germany expanding by an average of 2.7% and the UK by 3.8%. Yes, people tend to take time off in the summer, but work hard before and after the break.

In Q4, the economies diverge. Germany shrinks on average by 0.9 per cent, while the UK grows three percent on average. This could be because some are more lavish spenders on Christmas food and gifts. In other words, all of Germany’s growth since reunification has come in the middle six months of the year. All of the UK’s growth has come in the last six months.

Seasonal recessions matter
Yi Wen, an economist at the St Louis Federal Reserve in the USA, has shown that seasonal recessions can sometimes lead to longer-term downturns. For example, if firms facing a “seasonal” fall in demand cut employment, that can spill over into weaker consumer spending later.

But they matter another way. Media and politicians obsess about whether a country is in recession or not, as if a tenth of a percentage point in unreliable data matters. But they completely ignore much larger falls in economic activity every year. Which poses a question: why should recessions matter so much after seasonal adjustment but not before? It is, after all, what happens before that statistical manipulation that is people’s actual experience.

InvestingEmotionSharesHappinessCalendar effects

Chris Dillow
Chris Dillow

Investors Chronicle writer and economist