Calendar Anomalies, Much Ado About Nothing

An anomaly is a pattern in stock returns that deviates from what is expected based on established financial theories or models. These patterns can sometimes present opportunities for abnormal returns. However, they are often inconsistent and challenging to exploit. Many anomalies have achieved consensus and, thus, have been incorporated into factor-based models, including size, value, momentum, profitability, and investment. And mutual funds and ETFs have been developed that incorporate these factors. An exception is that despite calendar-based anomalies (such as day of the month, month of the year, “Sell in May and Go Away,” and many more) having achieved much attention from academic researchers.

For example, see the following articles:

The Halloween Indicator, ‘Sell in May and Go Away’: Another Puzzle

Are Seasonal Anomalies Real?

An Anatomy of Calendar Effects

Equity Returns at the Turn of the Month

Do Seasonal Anomalies Still Work?

There are no models that incorporate them, nor are there even any funds that attempt to exploit them.

A plausible explanation for this failure is that calendar anomalies could result from statistical data snooping—they suffer from look‑ahead and survivorship biases. Another is that just because a statistical significance of a calendar anomaly exists does not mean that after implementation costs it will lead to abnormal portfolio performance—statistical studies often ignore transactions and other implementation costs, let alone taxes. In other words, are calendar anomalies just an ado about nothing?

New Research

To answer that question, Samveg Patel, author of the study, “Calendar Anomalies: A Portfolio Approach,” published in the September 2024 issue of The Journal of Wealth Management, evaluated the calendar‑anomalies‑based active portfolio strategies using daily closing prices of the S&P 500 Total Return Index (TRI) and MSCI All Country World Index (ACWI). To test the calendar anomalies found in the literature from a statistical perspective, Patel applied a time‑series regression model with dummy variables for each day, date, and month and constructed portfolios based on equal amounts of investments on a particular day, date, or month. His sample covered the period from inception (1988 for S&P 500 TRI and 2001 for MSCI ACWI) to 2023. Following are his key findings:

• The days, dates, and months that provided the highest and lowest returns were inconsistent across decades and indices.
• The percentage of up‑days and down‑days were very similar across days, dates, and months.
• He could not find consistent higher or lower returns for a particular day, date, or month across decades and indices.
• The findings were robust when dividing the full sample into decades.
• The negligible differences in the annualized returns for day, date, and month portfolios suggests that calendar‑anomalies‑based portfolio strategies do not generate superior returns for a particular day, date, or month portfolio in either US or global portfolios.

His findings led Patel to conclude:

“Contrary to popular belief, calendar anomalies do not exist in the US and world market indices.” He added: “Calendar anomalies are just statistical gibberish and not ‘real anomalies,’ at least in the context of portfolio management, and are just an ado about nothing…. The findings of this study are expected to settle the debate on exploiting calendar anomalies to generate abnormal returns in portfolio management.”

Investor Takeaways

In our book “Your Complete Guide to Factor-Based Investing” Andrew Berkin and I provided investors with six criteria that should exist (all six must exist) before an investors should consider allocating to a strategy—there must be a premium that has been persistent (across time and economic regimes), pervasive (across sectors, countries, regions and if applicable asset classes), robust (to various definitions), implementable (survives transactions costs), and have risk- or behavioral-based explanations for why you should believe they should persist. Calendar anomalies (including “Sell in May and Go Away”) fail this test as there is no rational argument as to why a particular average daily or monthly return should be higher.

For investors the key takeaway from this study is that although statistical analysis provides evidence about the existence of calendar anomalies (like day of the week, date of the month, and month of the year) in the US and world market indices, there is no significant difference in the annualized returns for calendar‑anomalies‑based portfolio strategies. Thus, the selection of a day or date does not make any significant difference in a portfolio’s returns—investors can choose any day or date for a systematic investment plan based on what is convenient for them, and their portfolio returns will not differ across the selection of day or date. The bottom line is that the timing of stock purchases and sells does not matter to investors in the long run. And that means you should invest whenever you have the cash to do so.

Perhaps the most important takeaway is that the efficient market hypothesis suggests that the stock markets are efficient and that it is extremely difficult to find anomalies that can generate consistent abnormal returns. Thus, investors are best served by avoiding active management strategies that seek to exploit mispricings.

Larry Swedroe is the author or co-author of 18 books on investing, including his latest Enrich Your Future.