Abstract
This research paper analyses the presence of calendar anomalies: month-of-the-year and day-of-the-week effects on the FTSE/JSE All Share Index (JALSH), Financial (Fini15), Industrial (Indi25) and Residual (Res10) sector indices. In order to investigate those calendar anomalies, the Ordinary Least Square regression (OLS) and Generalised autoregressive conditional heteroskedasticity (GARCH) model were employed on daily and monthly data from 1 January 2013 to 31 December 2021. The results for the day-of-the-week effects revealed a positive Monday effect for the JALSH, Indi25 and Res10, and a Tuesday effect for the Fini15. This research paper recommends that investors trade on Mondays and Tuesdays to earn the highest return during the week. On the other hand, the findings for the month-of-the-year effects displayed an April effect for all the indices (JALSH, Fini15, Indi25 and Res10). For monthly seasonality investors have the potential to earn excess returns by trading in April than in January. The presence of calendar anomalies in the South African market reveals that the market may be weak-form efficient. The originality of the study lies in the combined use of sectorial indices in assessing calendar anomalies in a developing stock market. Also, this study adds to the existing literature on month-of-the-year and day-of-the-week effects on the Johannesburg Stock Exchange indices. The results presented in this study indicate that the month-of-the-year and day-of-the-week effects are mostly a function of the statistical technique applied and answering the research objectives does not solve entirely the issue of calendar anomalies.
Keywords: Calendar anomalies, Day-of-the-week effect, GARCH, JSE indices, Month-of-the-year effect, OLS regression, Efficient Market Hypothesis (EMH)