A Refutation of the Existence of the Other January Effect^*

Abstract

Cooper et al. report US evidence of the 'other January effect,' where returns in January are shown to have predictive power for returns over the subsequent 11 months. We re-examine the latest sub-period that they examine and find that the results using excess returns are not unique to January and that the effect for January is not apparent for raw returns. Further, using excess (raw) return data for 38 (44) other countries, limited support is found for the other January effect, with eight (five) of the remaining 11 months demonstrating a statistically significant effect in at least as many countries as exhibited the 'other January effect.' Further, there is no evidence to suggest that different tax-year ends across countries can explain the result. n We thank Bruce Grundy, Richard Heaney, seminar participants at the University of Melbourne and conference participants at the 12th Finsia-Melbourne Centre for Financial Studies Banking and Finance Conference for helpful comments, and acknowledge the financial support provided by a Melbourne Centre for Financial Studies research grant.1 Extensive sensitivity analysis was performed where multiple proxies for the 1-month risk-free rate of return were available. The results were not sensitive. 2 The methodology used here follows that of Cooper et al. However, as shown by Richardson and Stock (1989), among others, care should be taken when drawing inferences from multi-year returns.

The Other Month Effect: Some Evidence from the Central and Eastern European Markets

“…Panel C for Hungary indicates the absence of the OJE similar to the results of Easton -Pinder (2007). Instead, 6-month holding period returns following positive January returns are, on average, statistically lower than those following negative January returns.…”

“…However, they find that the OJE cannot be attributed to these possible variables. Easton -Pinder (2007) find no evidence that different tax-year ends across countries can explain the OJE. Stivers et al (2009) indicate three possible explanations for the OJE, including an internationally priced risk factor, a ubiquitous behavioural bias and a temporary anomaly.…”

“…Stivers et al (2009) show that the predictive power of the OJE in many countries (except the US) has declined over time. For many markets outside the US, Easton -Pinder (2007) and Marshall -Visaltanachoti (2010) do not support the conclusion of Cooper et al (2006). Even for the US market, Darrat et al (2013) report results contrary to Cooper et al (2006) and further show stronger anomalies for non-January months, particularly February and September.…”

“…Similarly to the result for Hungary, the OJE does not exist for Poland either as shown in Panel D. Easton -Pinder (2007) also fail to find the OJE for Poland. The 1-and 3-month holding period returns following positive March returns are larger than those following negative March returns, but only at the 10% level of significance.…”

“…First, we contribute to the literature on the other January/month effect and focus on transition economies. Easton -Pinder (2007) examine the other month effect in international markets including the Czech Republic, Hungary, and Poland. However, their paper only uses 11-month holding period returns, while we explore the case for 1-, 3-, and 6-month holding period returns.…”