Skip to main content

The U.S. suffered from a spasm of superstition and madness during the Salem Witch Trials of 1692. Thankfully the murderous episode was put to an end by May of 1693.

Strange behaviour can also afflict the stock market and, as it happens, a particularly persistent anomaly can be traced back to 1693. You likely know it by the adage to “Sell in May, and go away,” while others refer to it as the Halloween indicator.

The idea is to sell your stocks at the end of April each year, hide out in Treasuries (or bonds) and buy back into the market at the end of October.

Professors Cherry Y. Zhang and Ben Jacobsen discussed the technique in their paper The Halloween Indicator, “Sell in May and Go Away”: Everywhere and All the Time. They examined the efficacy of the indicator in 114 countries and used a treasure trove of data that goes back to 1693 in Britain and 1792 in the U.S.

I’ll start by highlighting the British results where the stock market is tracked by the U.K. FTSE All-Share Index. The index generated average capital gains (without dividends) of 1.6 per cent annually from February, 1693, through to April, 2017. (The returns herein are based on local currencies.)

The researchers observed that dividends represented almost the sole source of returns for stock investors into roughly the 1850s when capital gains started to become more important. More recently, dividends provided about 30 per cent of the market’s total returns in the 30-year period ending in 2017.

The British market enjoyed average annual total returns (including dividends) of 6.5 per cent from September, 1694, to April, 2017. The stock market outperformed the risk-free rate (three-month U.K. Treasuries) by an average of 2.2 percentage points annually over the same period.

The “Sell in May” method splits the year in half with the summer period running from May through October and the winter period from November through April. The total returns for both periods can then be compared to the risk-free interest rates (three-month Treasuries or their equivalent) in each country.

It turns out the British stock index provided a total return premium over its risk-free rate that averaged 2.4 percentage points in the winter months based on data from September, 1694, to April, 2017. But the shocker came in the summer months because the market’s average total return lagged Treasuries by an average of 0.2 percentage points.

The trend was similar in the U.S. with the S&P 500 index offering a total return premium to 90-day U.S. Treasuries that averaged three percentage points in the winter based on data from February, 1800, to April, 2017. In this case, the summer period was profitable with stocks paying an average total return premium of 1.2 percentage points.

The Canadian experience was similar to the British one with the S&P/TSX Composite Index sporting an average total return premium to three-month Canadian Treasuries of 5.3 percentage points in the winter and -0.4 percentage points in the summer based on data from March, 1934, to April, 2017.

The study also pooled data for 65 countries that had both total return and risk-free rate data to arrive at similar results. Over all, stocks provided a premium to risk-free rates that averaged 5.1 percentage points in the winter period and -1.1 percentage points in the summer.

I’m tempted to quip that holding stocks in the summer is a form of market madness.

Mind you, the market didn’t lag in the summer each and every year. Capital gains in the winter beat those in the summer 58 per cent of the time when data from all 114 countries was pooled together. In Britain, the win rate was 59 per cent, in the U.S. it was 56 per cent and it was 63 per cent in Canada.

While the “Sell in May” phenomenon appears to represent a significant market anomaly, it is important for investors to be mindful of its potential tax consequences and the extra trading costs it entails. In addition, I don’t think it’s an appropriate strategy for investors who are behaviourally better suited to the easier-to-implement buy-and-hold approach.

I hope to revisit the Halloween Indicator in the future with several possible extensions in mind. For instance, there may be some succour available to dividend investors who like to hold onto their stocks for the long term. But such musings will have to wait until we’re closer to Halloween when children chant, “Shell out! Shell out! The witches are out!”

Norman Rothery, PhD, CFA, is the founder of

Be smart with your money. Get the latest investing insights delivered right to your inbox three times a week, with the Globe Investor newsletter. Sign up today.

Follow related authors and topics

Authors and topics you follow will be added to your personal news feed in Following.

Interact with The Globe