The Month-of-the-Year Effect in the European, American, Australian and Asian Markets

Gualter Couto,Rui Alexandre Castanho,Pedro Pimentel,Catarina Barbosa

doi:10.3390/economies9040168

Gualter Couto, Rui Alexandre Castanho + Show 2 more

Open Access

https://doi.org/10.3390/economies9040168

Copy DOI

Abstract

This paper examines the existence of the month-of-the-year effects in four different continents, namely Europe, Asia, America, and Oceania. Nine indexes were analyzed in order to verify differences between monthly returns from January 1990 to December 2013, followed by an examination of the January effect, Halloween effect, and the October effect, testing for statistical significance using an OLS linear regression in order to verify whether those effects offer consistent opportunities for investors. Investors with globally diversified portfolios benefit from the Halloween effect, with a 1.2% average monthly excess return in winter and spring, while the pre-dotcom-bubble period had a better performance than the post-dotcom-bubble period. In the global post-dotcom-bubble period, there is statistical evidence for 1.60% and 1% lower average monthly returns in January (the January effect) and in months other than October (the October effect), respectively, contradicting the literature. The dotcom bubble seems to be responsible for the January effect differing from what might otherwise have been expected in the later period. There is no consistent and clear impact on continental incidence. The Halloween effect is revealed to be a fruitful strategy in the FTSE, DAX, Dow Jones, BOVESPA, and N225 indexes taken one-by-one. The January effect excess average return was only statistically significative for the pre-dotcom-bubble period for globally diversified portfolios. This paper contributes to a wider global and comparable view upon month-of-the-year effect.

Highlights

In recent years, certain abnormal behaviors were revealed to have an effect on the return on financial assets, thereby increasing research in this area
Renowned academics have investigated this topic, concluding that it breaks down into three anomalies: (i) the “January effect”, which maintains that the average return on shares tends to be higher in January compared to the other months of the year; (ii) the “May-to-October effect”, or the “Halloween effect”, which indicates that the average return on shares during the summer and autumn is lower compared to the months that make up the two missing seasons; and, (iii) the “October Effect”, which maintains that October is the month of the year with the lowest average shareholder returns
This paper examines the following research hypotheses, based on the literature review above, testing for: (i) The “January effect”, where average stock returns in January tend to be higher compared to the other months of the year; (ii) The “Halloween effect”, where average stock returns in summer and autumn tend to be lower compared to the spring and winter months; and (iii) The “October effect”, where average stock returns in October tend to be the lowest of the year

Summary

Introduction

Certain abnormal behaviors (calendar effects) were revealed to have an effect on the return on financial assets, thereby increasing research in this area. Our contribution to the literature comes from our analysis of the abnormal returns that may have resulted from these three month-of-the-year anomalies, allowing us to ascertain which anomalies, if any, had an impact on the capital market, and what the direction of that impact might have been, focusing upon the 1990–2013 period for European, American, Australian, and Asian markets. This research intends to contribute to our understanding of whether these calendar anomalies occur in the stock markets in a persistent and regular way, and whether they are fruitful strategies for investors. Our main findings point to a significant Halloween effect, as expected, and significant unexpected January and October effects, mainly in the post-dotcom-bubble period, for globally diversified portfolio investment strategies. In the pre-dotcom-bubble period do average returns in January seem to perform 1.4% higher than the remaining months for globally diversified portfolio investment. The Halloween effect appears to be a fruitful strategy for the FTSE, DAX, Dow Jones, BOVESPA, and N225 indexes taken one-by-one, proving to be an ongoing investment strategy for those indexes

Objectives

Methods

Results

Conclusion