Sell in May and go away: myth or useful market pattern?

“Sell in May and go away” refers to the idea that stock market returns may be weaker during the May through October period compared with the November through April period.

Seasonal comparisons connected to the saying typically divide the calendar year into two periods:

• May through October: The period associated with the weaker seasonal performance described in the saying.

• November through April: The period that has sometimes shown stronger average returns in certain market datasets.

The phrase is often interpreted as suggesting that investors reduce equity exposure during the summer months and return later in the year. 

The expression is memorable because it presents a complex market observation in a simple way. It appears to summarize a recurring seasonal pattern and can make differences between calendar periods easier to understand.

However, a memorable market saying does not necessarily represent a reliable market rule. Seasonal patterns can vary across countries, market periods, and measurement methods.

The phrase is also sometimes connected with the longer expression, “Sell in May and go away, come back on St. Leger’s Day.” The saying has commonly been associated with historical market activity patterns in the United Kingdom, where financial activity was traditionally affected by seasonal social and business schedules.

Where did the market saying come from?

The expression “sell in May and go away” is commonly linked to historical British market culture. The saying developed during a period when financial markets operated differently from modern markets.

Earlier trading environments had different communication systems, market structures, and participation patterns compared with today’s globally connected financial markets. Business activity and social calendars could influence when market participants were more or less active.

The original phrase reflected both financial observations and cultural context. Seasonal patterns in markets have often been studied because economic activity, investor participation, and business conditions can change throughout the year.

Modern financial markets operate across regions, with electronic trading, global participation, and continuous access to information. These differences mean that the original environment in which the saying developed differs from current market conditions.

This does not determine whether seasonal effects can or cannot appear in modern markets. Instead, it highlights why seasonal patterns need to be evaluated using specific market data, time periods, and calculation methods.

For Canadian investors, the phrase also requires additional context because Canadian markets have their own characteristics, including different index compositions, sector exposure, and economic influences.

What do historical returns show?

Seasonal return comparisons examine whether average market performance has differed between specific periods of the calendar year. Researchers and analysts typically compare returns from May through October with returns from November through April using a selected benchmark.

Comparing May through October with November through April

A seasonal analysis may include several important details:

• Index: The benchmark being measured, such as the S&P 500 or a Canadian market index.

• Market: The country or region represented by the index.

• Time period: The years included in the calculation.

• Return type: Whether the calculation uses price returns or total returns.

• Dividend treatment: Whether reinvested distributions are included.

• Currency basis: The currency used to measure returns.

• Calculation method: Whether averages, medians, or other measurements are used.

These details matter because different methodologies can produce different results.

For example, a comparison based on a U.S. index may produce different findings from a comparison based on a Canadian index. Similarly, a total-return calculation that includes dividends may show different results from a price-return calculation.

A period with stronger average returns from November through April does not mean that the pattern occurred every year. Some years may show stronger performance during the May through October period, while other years may show the opposite.

Average results versus year-by-year consistency

Average seasonal returns can provide information about past performance, but averages do not describe every individual year.

Several factors can affect the interpretation of seasonal comparisons:

• A small number of unusually strong or weak years can influence the average.

• Median returns may differ from average returns.

• Results can change depending on the start and end dates selected.

• Different indexes can produce different outcomes.

• Positive returns can occur during periods that have historically shown weaker averages.

For this reason, seasonal patterns are generally examined as observations about past market behaviour rather than as fixed market rules.

Why the selected timeframe matters

The timeframe used in a seasonal analysis can influence the results. A comparison covering 10 years may show a different pattern from one covering 20, 30, or 50 years.

Longer periods may include different economic environments, market cycles, interest-rate conditions, and sector trends. Shorter periods may be influenced more by specific events that occurred during that timeframe.

For this reason, seasonal comparisons are generally reviewed alongside information about the benchmark, calculation method, and period selected. A single timeframe may provide one view of past market behaviour, but may not represent all possible outcomes.

Does “sell in May” apply to Canadian markets?

The “sell in May and go away” saying developed outside Canada, and results from other markets cannot automatically be applied to Canadian markets.

Canadian market benchmarks have different characteristics from many international indexes. For example, major Canadian indexes may have different levels of exposure to sectors such as financial services, energy, and materials compared with U.S. benchmarks.

These differences can influence how seasonal patterns appear.

Factors that may affect Canadian seasonal return comparisons include:

• Index composition: Different companies and sectors can contribute differently to index performance.

• Sector exposure: Canadian benchmarks may have different concentrations in industries that respond differently to economic conditions.

• Dividend contribution: Canadian companies have historically included many dividend-paying businesses, making the treatment of dividends relevant when measuring returns.

• Currency considerations: Comparisons involving international assets may produce different results depending on whether returns are measured in Canadian dollars or another currency.

• Economic conditions: Canadian markets can be influenced by domestic and global economic developments.

A seasonal pattern observed in one market does not necessarily apply in the same way to another market. Canadian results require separate analysis using Canadian benchmarks, time periods, and return calculations.

Seasonal comparisons can provide information about periods when market returns differed in the past, but the strength and consistency of any pattern can depend on the index, timeframe, and methodology used.

Why seasonal return comparisons have limitations

Seasonal patterns can be interesting to study, but several factors can affect how they are interpreted.

Historical patterns are not forecasts

Seasonal return comparisons describe previous market outcomes. They do not predict what may happen in a future year.

Market performance can be influenced by many factors, including economic conditions, company earnings, interest rates, investor expectations, and unexpected events. A seasonal pattern observed in one period may not appear in another period.

Missing strong market days can affect results

Market returns can sometimes be concentrated during a limited number of trading days. Some of those days may occur during periods that have historically shown weaker seasonal averages.

As a result, comparisons between calendar periods can be affected by which days are included or excluded from a measurement period. A seasonal return comparison may look different depending on the starting date, ending date, and measurement method used.

Trading costs, spreads, and taxes matter

Simplified seasonal comparisons often focus on index performance and may not include practical considerations associated with buying, selling, or changing investments.

Factors that may affect real-world results include:

• Trading commissions or account fees.

• Bid-ask spreads.

• Tax considerations in non-registered accounts.

• Currency conversion costs.

• Market timing (the timing of entering or exiting investments).

• Time spent outside of a market.

These factors can vary depending on the account type, investment, and individual circumstances.

Data selection can change the conclusion

The way seasonal data is selected and measured can influence the interpretation of results.

Important considerations can include:

• Choosing a specific start and end date.

• Using price returns instead of total returns.

• Comparing different indexes without accounting for differences.

• Excluding or including dividends.

• Focusing only on average returns.

• Not considering years that produced different outcomes.

Transparent methodology helps provide context when evaluating seasonal market patterns.

What the evidence does and does not establish

Financial research into the “sell in May and go away” pattern has identified periods where some markets produced different average returns between May through October and November through April.

The evidence may show that:

• Some markets have displayed seasonal return differences during certain periods.

• The November through April period has produced stronger averages in some datasets.

• Results can vary by market, index, timeframe, and calculation method.

• The pattern has not occurred consistently every year.

The evidence does not establish that:

• Selling in May reliably improves investment returns.

• The pattern will continue in future periods.

• A seasonal average identifies the correct time to enter or exit a market.

• Results from one market apply to all investors or asset classes.

• Trading costs, taxes, and other practical considerations can be ignored.

The bottom line on the “sell in May and go away” myth

The saying “sell in May and go away” reflects a seasonal market pattern that has been observed in some datasets and markets. However, the pattern has not been universal or consistent across all time periods.

The interpretation of seasonal returns depends on several factors, including the index examined, timeframe selected, dividend treatment, currency basis, and calculation method.

For Canadian investors, results from other markets may not provide a complete picture because Canadian benchmarks have different characteristics and sector exposure.

Seasonal patterns can offer context about previous market behaviour, but they do not represent a forecast or a dependable rule for future market performance.