Calendar Anomalies, Market Regimes, and the Adaptive Market Hypothesis in African Stock Markets

Abstract

Purpose: This paper examines the changing behavior of two calendar anomalies in African stock returns-the month-of-the-year and the intra-month effects-and their implications for the adaptive market hypothesis (AMH). Methodology: We applied two-stage Markov switching models (MSMs) instead of the conventional single state regression model. The sample period includes the daily index return of Nigerian, South African, Mauritian, Moroccan, and Tunisian stock exchanges from January 1998 to February 2018. Findings: We found that (i) all the markets except for the Johannesburg Stock Exchange (JSE) have a higher tendency to be in bearish state than bullish state, (ii) month-of-the-year and intra-month effects appear in one regime and disappear in another regime, and (iii) the behavior of calendar anomalies is affected by market conditions and conforms to AMH rather than the efficient market hypothesis (EMH). Practical Implications: We present that (i) calendar anomaly is a characteristic that changes under different regimes or market conditions in African stock markets, (ii) active investment management may yield profits for market participants, depending on the market conditions and the anomaly in question, and (iii) the right approach would be for investors to consider each market with its own peculiarity even when they are in the same continent. Originality/Value: The sensitivity of the month-of-the-year and the intra-month effects to market conditions has not been documented in African stock markets, especially with the use of regime-switching models.