The adaptive market hypothesis and predictability of stock returns in Africa

Abstract

The weak-form Efficient Market Hypothesis postulates that stock price movement is random and future returns cannot be predicted based on past returns. On the other hand, the Adaptive Market Hypothesis proposed by Lo (2004) postulates that stock market efficiency is not an all-or-nothing condition but is a phenomenon that varies continuously over time depending on market conditions. The study examines the AMH and stock return predictability on eight African stock markets: Johannesburg Stock Exchange (JSE), Nigerian Stock Exchange (NSE), Ghana Stock Exchange (GSE), Stock Exchange of Mauritius (SEM), Namibian Stock Exchange (NSX), Botswana Stock Exchange (BSE), Egyptian Exchange (EGX), and the Casablanca Stock Exchange (CSE). The study uses autocorrelation, wild bootstrapped automatic variance ratio, and generalised spectral tests on price returns over the period 4 January 2011 to 30 November 2017. The results reveal that the full samples of only South Africa and Tunisia re not predictable and therefore deemed consistent with the weak-form efficient market hypothesis. Further tests on the yearly sub-samples suggest that all the markets have exhibited features of predictable markets at some point in line with the varying levels of return predictability. This finding provides evidence in favour of the postulation of the Adaptive Market Hypothesis. Keywords: Adaptive Market Hypothesis; African Stock Markets; Predictability; Martingale Difference Hypothesis