Right after the September effect – the worst returns of the year, people say – comes the time of ghosts and lollipops, of werewolves, pumpkin and scary nights: the Halloween effect is upon us. But what does this have to do with finance?
The Halloween Effect: a definition
The Halloween effect or the Halloween strategy is a market timing strategy holding on to the belief that investments in stocks tend to have a higher return during a given period of the year.
Financial scholars noticed that, every year around May, stock owners sell and run away, as people believe that the month of May is the starting point of a bear market. People would better hold on to their cash, only to return in November and invest massively to benefit from the start of a bull market. In other terms, overall returns are higher during the winter than in the summer.
Historically speaking the strategy is closely tied with the “sell in May and go away” saying, which originated in England when wealthy families would leave London for their summer estate and come back in September for the winter – you can imagine it was quite stressful not to be able to look at your stocks for months, you might as well sell them when you know what they’re worth!
Does this theory hold empirically?
Funnily enough, it does. If we assume that stock markets are efficient (Efficient Market Hypothesis), this should mean that market anomalies resulting in an above average return should disappear by themselves as investors rush into them. While this is true for most market inefficiencies, Murphy’s law – “Anything that can go wrong, will go wrong” – does not seem to apply to the Halloween, as it remains statistically significant even today.
Although some scholars doubt the validity of the theory – arguing that it is nothing but a statistical aberration due to outliers –, and that we do not have a good explanation for what is causing it, the numbers don’t lie.
In a recent study, Zhang and Jacobsen (2021) showed evidence that the Halloween effect was a very robust and strong market anomaly, as 63 out of the 65 studied countries showed a higher return between November and April than during the rest of the year.
Plastun and al. (2020) gave us the number we were looking for: on average and across the world, returns were 4% higher during the winter season. Although the effect could vary from stock to stock, the statistical difference was robust in 18 out of 35 stocks, a data already proposed by Arendas et all. (2017).
Whatever the Efficient Market Theory says, it seems that we still have gaps in the market, with efficiencies to be exploited by investors. Although a study found that the effect was stronger in the 20th century, it is still here today and can provide a consistent strategy (Plastun et al., 2020).
How can I play such a strategy?
The more conservative scholars Dichtl and Drobetz (2014) were convinced that the classic buy and hold strategy was the way to go, and that no Halloween strategy could beat it.
For those who disagree and choose trust the strategy, it is quite simple: Swagerman and Novakovic (2010) argued that holding on to stock during the winter, and switching to t-bills – short term bonds – during the summer could achieve a higher return than a classic buy and hold strategy. This strategy proved to be a better option in 19 out of the 23 investigated stock markets, or 82% of the time, not bad.
Lloyd et. al. (2017) proposed an even more aggressive strategy, advising to go long in the winter and shorting the stock market during summertime. Between 2005 and 2017, this specific strategy was able to beat the classic Halloween effect strategy by 3.2 percentage points on average, and the buy and hold strategy by 4.77 percentage point per year on average (Arendas et al., 2017).
However, one must note that this strategy generally works 2 out of 3 years, which means that the adventurous investor should be ready to not win directly (Arendas et al., 2017).
According to the Halloween effect, stock returns are higher in the months from November to April. Therefore, the Halloween strategy is to invest in stocks starting November before selling everything in May.
This strategy has been observed for centuries now, getting more and more pronounced. The difference of performance between this strategy and the classic buy and hold was found to be of 4% on average.
While there is not satisfactory explanation for us to understand the how and the why of such a phenomena, empirical data remains a strong source of evidence, and the results are clear: we are facing a lasting market inefficiency. Unlike with ghosts, we now know that the Halloween effect exists, but thankfully, it’s a lot less scary.
Arendas, Peter; Malacka, Viera and Schwarzova, Maria. (2017) A Closer Look at the Halloween Effect: The Case of the Dow Jones Industrial Average. International Journal of Financial Studies, 6(42)
Haggard, K.; Witte, H. (2010) The Halloween effect: Trick or treat? International Review of Financial Analysis, 19(5), pp. 379-387
Plastun, Alex; Sibande, Xolani; Gupta, Rangan, Wohar, Mark. (2020) Halloween Effect in developped stock markets: A historical perspective. International Economics, 161, pp. 130-138
Zhang, Cherry, and Jacobsen, Ben. (2021) The Halloween indicator, "Sell in May and Go Away": Everywhere and all the time. Journal of International Money and Finance, 110
Swagerman, Dirk, and Ivan Novakovic. (2010) Multi-National Evidence on Calendar Patterns in Stock Returns: An Empirical Case Study on Investment Strategy and the Halloween Effect. The International Journal of Business and Finance Research 4: 23–42
Halloween Strategy, in Investopedia
Read our next article: Covered call versus naked call option strategies