Thursday, April 20, 2023

Sell In May And Go Away(?)

May Day is coming, and how apt it is to pen down something which sounded like a nursery rhyme for investors and traders.

You may have heard of the jingle “sell in May and go away”, but where does this adage come from? Accordingly, the phrase originated from London’s financial district, but somehow made it to Wall Street and to a certain extent our local Singapore market. 


There is a second part to the phrase, and it goes like “come back on St. Leger’s Day”. St Leger’s Day, or known as St. Leger’s Stakes, is a horseracing event that takes place in September. Thus “sell in May and go away, come back on St. Leger’s day” implies that we should exit the market or reduce our investible stakes during the period of May to September (to add, a modern version of the adage goes something like “sell in May and go away, and wake me up when September ends”).


The ”sell in May and go away” also has another name: Halloween Effect. The main difference between the two is the market reentry period, for the Halloween Effect the time to buy back into the markets is in beginning of November, after the Halloween holidays.


When I first heard about them, they sounded like one of those traditional rules that does not seem to make any sense. The very first thought that came to my mind was: is it a form of self-fulfilling prophecy? And is it just affecting the United States and United Kingdom markets, since it was deemed to be prevalent there?


According to a 2001 research paper on the study of the Halloween Effect1 on 37 financial markets (Singapore included) between the years 1973 and 1996, an investor adopting the Halloween Strategy (i.e., sell in the beginning of May to park in a risk-free asset, and then buying back into the market in end October) would fare better in returns in most of the markets than an investor who buy-and-hold throughout (that included Singapore). A related follow-up 2018 paper2 sampled the period between Oct 1998 and Apr 2017 of the same countries still showed roughly the same results.


Based on the research numbers, this prophecy holds true, most of the time (2020 was a glaring exception, and we knew what happened back then). And since the frequency of it happening is high, shouldn’t we just ditch buy-and-hold and follow this to get higher returns? With the number of low-cost brokerages around nowadays and that meant lower transaction costs, it makes more sense to adopt this approach.


The Bedokian’s Take


Past performance is not indicative of future results, but I admit the “sell in May”/Halloween Effect defied known logic, at least for me. Statistically, the average smoothed-out returns over the November-April period do look better, so I guess there will be people who declare this as empirical proof that it works.


In my opinion the results are not empirical enough. Yes, it works most of the time, but if I want to adopt this, I must make sure it works all the time. However, we all know economics and the stock market are not hard sciences like chemistry or physics, so if it could be wrong on one occasion, I would rather be prudent and stick to my preferred investment strategy and methodology to reduce uncertainty and risk, than to go by the time of the year to make buy-sell decisions.



1 – Jacobsen, Ben and Bouman, Sven. The Halloween Indicator, ‘Sell in May and Go Away’: Another Puzzle. 1 Jul 2001. (accessed 20 Apr 2023)


2 – Jacobsen, Ben and Zhang, Cherry Yi. The Halloween Indicator, ‘Sell in May and Go Away’: Everywhere and All the Time. 1 Oct 2018. (accessed 20 Apr 2023)

No comments:

Post a Comment