Sell in May comes from a horse race, and the evidence is thinner than the slogan

Everything here earned its place by changing an outcome. Nothing about seasonal patterns in share prices is included to round the number up.

Where the saying comes from

The full version instructs the listener to sell in May, go away, and come back on a specific September race day in the English racing calendar. It describes the habits of a financial class that left London for the summer, when the market was thin and little business was done.

In an era when trading required physical presence, a market emptied of its participants genuinely behaved differently over the summer. The saying is therefore a description of a social calendar rather than a discovery about markets. That origin is worth holding on to, because it explains why the pattern might have existed without implying that it must persist.

What the studies find

Analyses across many national markets and long periods have reported that returns in the winter half of the year tend to exceed those in the summer half. The pattern appears in enough places and over enough decades that it is difficult to dismiss as pure coincidence. Explanations proposed include holiday effects on liquidity, the timing of institutional flows, and seasonal patterns in risk appetite.

Checked against the record, none of these explanations is settled, and an effect without an agreed mechanism should be held loosely. There is also a serious concern about data mining, since testing enough calendar rules will always produce some that look impressive.

Why it is hard to exploit

Acting on the rule means being out of the market for around half of every year, and missing a single strong summer can undo several good years of the strategy. Buying and selling twice a year incurs costs, and in taxable accounts it can trigger liabilities that a buy-and-hold investor never faces. The average difference reported is generally not large relative to the variation between individual years.

Somewhere in the retelling, a rule that works on average across a century may spend a decade being wrong, which is longer than most people's patience. The gap between a statistical pattern and a usable strategy is where most published market anomalies quietly die.

The general problem with calendar rules

Financial data offers an enormous number of possible calendar patterns, and testing them all guarantees that some will appear significant by chance. Effects that are published often shrink afterwards, either because they were artefacts or because participants adjust once the pattern is known. Any pattern reliable enough to profit from should attract enough trading to reduce it, which is an argument against durable calendar effects in general.

The primary source says otherwise: this does not prove the seasonal pattern is spurious, but it does explain why professionals treat such findings cautiously. A rule that survives publication and continues to work is unusual, and the burden of proof sits with the person claiming one.

What the folklore gets right

Summer trading volumes really are lower in many markets, and thin markets can move more sharply on less news. Institutional calendars are real, and reporting periods and fund year-ends genuinely cluster activity at particular times. So the underlying observation about seasonality is not nonsense, even if the trading rule attached to it is unreliable.

The saying survives partly because it is memorable and partly because it is repeated in the financial press every spring. Annual repetition in newspapers is not evidence, though it does a great deal for a phrase's apparent authority.

Printing a correction has a poor record of taking a myth out of circulation.

What this is not

Nothing here is a recommendation about anyone's money, and market timing decisions depend on circumstances that a general article cannot know. Costs, tax treatment and available products differ substantially between countries, and rules that work in one jurisdiction may be unusable in another.

Trace it back and anyone making decisions of this kind should seek regulated advice appropriate to where they live. The point of interest here is the origin of a piece of folklore and the gap between a statistical pattern and a usable rule. Both halves of that gap are worth understanding regardless of what anyone decides to do.

Everything above, in order of what to do first

1. Where the saying comes from. The full version instructs the listener to sell in May, go away, and come back on a specific September race day in the English racing calendar.
2. What the studies find. Analyses across many national markets and long periods have reported that returns in the winter half of the year tend to exceed those in the summer half.
3. Why it is hard to exploit. Acting on the rule means being out of the market for around half of every year, and missing a single strong summer can undo several good years of the strategy.
4. The general problem with calendar rules. Financial data offers an enormous number of possible calendar patterns, and testing them all guarantees that some will appear significant by chance.
5. What the folklore gets right. Summer trading volumes really are lower in many markets, and thin markets can move more sharply on less news.
6. What this is not. Nothing here is a recommendation about anyone's money, and market timing decisions depend on circumstances that a general article cannot know.

The takeaway

It began as a description of when bankers left London for the summer, which is not the same as a discovery about markets.

The satisfying version of a story is the one that travels, which is the whole problem.