Oct. 18, 2021
Investors should not be spooked by Halloween. Now is a great time to buy equities.
Research shows there is a significant seasonality to stock markets and they consistently perform better in the winter months.
The old adage “sell in May and go away” has been around for decades. It recommends investors follow a strategy of selling equities at the start of May and not buying back into the market until the beginning of November.
Many have derided the theory over the years as an old wives’ tale and attempted to disprove it. But two new pieces of academic research have confirmed that the Halloween effect, as it is also known, is very much real and a highly dependable trading strategy to boot. Just don’t ask anyone to explain why it works.
In the biggest study of its kind, Ben Jacobsen, professor of finance at TIAS Business School in the Netherlands, and Cherry Zhang from Nottingham University Business School China analyzed the monthly returns from all 114 stock markets in the world, going back over 320 years.
The research found that on average stock markets rose 5.1%, or 0.85% a month, from November to April, compared with just 1.1%, or 0.18% a month, between May and October.
The higher mean returns were evident in 89 of the 114 countries and the difference statistically significant in 42 of those countries, compared to only one country, Mauritius, that had significantly higher returns for May-October.
The effect is truly global and most pronounced in Europe, North America, and Asia.
The reliability of the Halloween effect is also remarkable and Jacobsen and Zhang said it has been increasing over the past 50 years. It is such a powerful trend that it overrides shorter-term concerns, such as the outcome of a U.S. presidential election or dips in GDP growth.
The pair showed that in the U.K., for example, over a five-year time horizon, the buy in November and sell in April strategy beat the market 80% of the time. Over 10 years the probability of beating the stock market rose to 90%.
“Overall, our evidence suggests that the Halloween effect is a strong market anomaly that has strengthened rather than weakened in recent years. In sum what we can say for certain is that the Halloween effect is alive and kicking,” the pair added.
This research is not an outlier
Jacobsen and Zhang’s study, which is to be published in the Journal of International Money and Finance next February but can now be accessed online, is certainly eye-catching. The results were very similar to what the pair found in an earlier study they carried out in 2012.
That in turn built on research Jacobsen carried out with Sven Bouman, now CEO of Dutch fund manager Saemor Capital, in 2002. This found the trend present in 36 out of 37 countries they looked at, and this was supported by further work by Jacobsen & Visaltanachoti in 2009 and Grimbacher, Swinkels & van Vliet in 2010.
There have been plenty of naysayers over the years, however. Academic theory says that the Halloween effect should not work. The efficient market hypothesis, a cornerstone of finance textbooks, says that the pricing of assets reflects all known information. So anomalies like the Halloween effect, which are both known and highly predictable, should quickly disappear.
Many academics have tried to disprove the Halloween effect over the years. Jacobsen and Zhang said they updated and expanded their previous research to broaden the dataset and counter any accusations of sample selection bias distorting the findings.
A separate study this year by four academics hailing from universities in South Africa, Ukraine, and the U.K., published in International Economics, had set out to prove that statistical quirks were behind the findings.
What their tests actually found was that despite the trend being “impossible according to efficient market hypothesis”, it is real and “continues to provide opportunities to build a trading strategy that can beat the market”.
Similar studies aiming to explain away the effect by pinpointing problems with the methodology used, such as papers by Maberly & Pierce in 2003 and 2004, have also been dismissed. Jacobsen and Zhang were just too thorough and their dataset to comprehensive for their findings to be dismissed this way.
But for all of the academic study that has been carried out into the Halloween effect, no-one has ever been able to satisfactorily explain why it happens.
Past theories, such as thinner trading volumes over holidays, the timings of macro and corporate news announcements, or seasonal changes to our mood, have all been debunked.
Maybe it’s fitting that the Halloween effect retains an air of mystery.
© 2023 Forbes Media LLC. All Rights Reserved
This Forbes article was legally licensed through AdvisorStream.