The lunar moon festival and the dark side of the moon

Abstract

We propose and adduce evidence for a new seasonal anomaly associated with the Lunar Moon Festival (LMF) in East Asian economies. While the LMF effect bears some resemblance to the festivity and vacation anomalies, it is mainly driven by nostalgia, historically negative associations, the full moon and uncertainty about future harvest prospects. This negative sentiment and associated increase in risk and loss aversion are responsible for reducing share turnover, return volatility and stock returns over a 2-week period. The LMF effect is the strongest for China, Taiwan and South Korea where it is not only celebrated as a public or cultural holiday but it also impacts on neighbouring stock markets where overseas Chinese investors possess significant resources. Robustness checks demonstrate that it has a distinctive and more pronounced impact than competing seasonal effects associated with lunar phases and the summer vacations.

Acknowledgements

Comments and suggestions by Xiaoquan Liu, Timothy J. Rakow and members in the Essex Finance Centre are gratefully acknowledged. Earlier versions of this article were presented at the 17th European Financial Management Association (EFMA) Annual Meetings, Athens, Greece, June 2008 and at the Applied Economics 40-year Jubilee Meeting, Clare College, Cambridge, September 2008. We are grateful to the participants for their insightful comments.

Notes

¹ Intercalation is the insertion of a leap day, week or month into some calendar years to make the calendar follow the seasons or Moon Phases (MP). For example, the LMF took place on 25 September 2007 and then on 14 September 2008 in the Gregorian calendar. Because there is an intercalary month in the fifth month in the Chinese Lunar calendar, there are two lunar fifth months in 2009. Therefore, the LMF took place on 3 October rather than 3 September in 2009 for the Gregorian calendar. For more details, please see Miller et al. (2009).

² See Isen et al. (1978), Zajonc (1980), Johanson and Tversky (1983), Schwarz and Clore (1983), Schwarz (1990), Bechara et al. (1997), Luce et al. (1999), Loewenstein et al. (2001), Solvic et al. (2002), Verma et al. (2008) and Durand et al. (2009).

³ Edmans et al. (2007) found that stock returns in the losing country are significantly lower on the first trading day after the international soccer matches. Nofsinger (2003) also concludes that higher (lower) stock returns may be caused by positive (negative) public emotion.

⁴ Although the LMF only became a public holiday since 2008, China is included in the group where the LMF is a public holiday as it is a cultural holiday for the Chinese and thus may have a strong impact.

⁵ For more cultural and historical detail including the history of moon cake and legendary tales, see Eberhard (1958), Gai (2009) and Miller et al. (2009).

⁶ We record missing values for the financial data when the markets were closed to avoid a downward bias in the results.

⁷ For brevity, the results are not reported but are available from the authors upon request.

⁸ See Criss and Marcum (1981); Hicks-Caskey and Potter (1991); Sands and Miller (1991) and De Castro and Pearcey (1995).

⁹ The authors use the dummy of Lunardummy and set it to one for full moon period but zero for new moon period.

¹⁰ Here, August is in terms of Chinese lunar calendar since the LMF is scheduled at the 15th of lunar August.

¹¹ These are similar to the results in Yuan et al. (2006) that only stock returns in the full moon periods are significantly lower than those in the new moon periods, but this pattern does not apply for trading volume and return volatility.