Are market moves purely random? It can be challenging to provide explanations for market price movements with so many factors driving one group of investors to buy and another to sell. Many find it easier just to accept dynamic moves as unpredictable. However, history contains many examples of human behavior influencing patterns in financial markets and sometimes creating investment opportunities.
Seasonal market trends may be one of the most cited examples. More specifically, the U.S. equity market has historically outperformed, on average, during the six-month period from Nov. 1 through April 30 compared with the other six months of the year.
Other well-known seasonal anomalies, such as the “Monday effect,” the “Friday effect,” the “turn of the month effect,” the “holiday effect,” and the “January effect,” often are mentioned, but these tend to lose strength when submitted to more rigorous examination.
However, the “sell in May” anomaly, as the November through April seasonality effect is sometimes called, has weathered a fair amount of academic scrutiny and appears to happen often. The historical pattern of springtime sell-offs is well-documented and can even be traced as far back in Britain as 1694.
The first comprehensive study on the topic, from Bouman and Jacobsen (2002), drew on extensive data to demonstrate the substantial difference between returns from May to October and those from November to April. Concluding that “the economic significance of this particular anomaly is considerable,” the authors went on to compare “a simple trading strategy based on the saying” with a buy-and-hold portfolio using past data in multiple countries. In many of these countries, the seasonality-inspired portfolio compared favorably.
The authors noted that that the scale of the sell in May effect is highly correlated to the length of a particular country's average summer vacation. While Bouman and Jacobson suggest that this trend relates to a decrease in risk aversion during one's vacation, others have assumed that the risk aversion has more to do with seasonal affective disorder (SAD) or that investor behavior is directly linked to temperature changes.
It's possible that other factors with more narrow calendar-related effects may contribute to the overall seasonal effect. For example, commentators sometimes speculate that changes in liquidity brought on by an influx of capital at year-end contribute to the January effect, which falls within the longer November-April timeframe. Another suggestion is that year-end bonuses and commissions may be distributed within a few months of Dec. 31. Others point to year-end earnings announcements. Another consideration is that many employer matches on retirement plans are invested at that time, or taxpayers may be receiving tax refunds and investing them in the early part of the year.
While definitive causes remain elusive, the persistence of stock market seasonality seems apparent in the overall November through April level. A follow-up study to Bouman and Jacobsen's research, this time conducted by Andrade, Chhaochharia and Fuerst in 2012, again found persistent benefits from following sell in May and go away as a market-timing strategy. The authors found “an economically large and statistically significant Sell in May effect in strategies exploiting the Value, Size, Credit Risk, FX Carry Trade, and Volatility risk premiums.”
The seasonality effect suggests investor actions may impact market moves that do not appear to be purely random. Tracking seasonal market factors may be one approach to taking advantage of that over time.
Isaac Braley is president of BTS Asset Management.