Sell in May and go away?
Where the claim comes from
It’s an old saying in the European financial press, with an ending about coming back for the St Leger horse race in September. In the US it’s called the “Halloween indicator”: be in stocks from 31 October to 30 April and out for the other half of the year. Most market sayings are noise. This one turned out to be harder to dismiss.
What’s actually true
- The pattern is real in the data. In 2002 Sven Bouman and Ben Jacobsen tested it in the American Economic Review. In their words, they found “this inherited wisdom to be true in 36 of the 37 developed and emerging markets studied”. It was especially strong in Europe, and in the UK they traced it back to 1694. They found no convincing explanation in risk, the January effect or links between markets.
- It kept working after it was published. Most anomalies fade once academics publish them. A 2013 Financial Analysts Journal paper by Andrade, Chhaochharia and Fuerst tested the same 37 markets out of sample, from 1998 to 2012. They found November–April returns still averaged about 10 percentage points higher than May–October.
- But “lower” isn’t “negative”. Bouman and Jacobsen found May–October returns “close to zero in many countries”, not reliably below zero. Sitting in cash means giving up whatever the summer does return, and it brings trading costs and, in many countries, taxes. A seasonal average says nothing about whether this summer will be a bad one.
The truth
The truth: This is one of the rare pieces of market folklore with serious evidence behind it. As a fact about averages, it holds up. As a trading rule, it’s a bet on an average, made once a year with real costs. Respect the pattern, but don’t treat it as a law.