Investor overconfidence under the adaptive markets hypothesis in selected African stock markets.

Abstract

Meticulous empirical research remains to determine whether the Adaptive Markets Hypothesis (AMH) or the more widely known Efficient Market Hypothesis (EMH) better explains investor overconfidence and stock return volatility behaviour. Investor overconfidence is vital in understanding why investment strategies are pursued so aggressively, leading to excessive market trading. It is often argued that the investor overconfidence bias makes markets less efficient because it creates pricing errors in extreme volatility and overestimates investors’ beliefs in the accuracy of their forecasts of their quotes on prices. This research analyses the effect of investor overconfidence on the volatility of stock market returns according to the AMH in seven African stock markets, including the Casablanca Stock Exchange, the Egyptian Exchange, the Johannesburg Stock Exchange, the Nigerian Stock Exchange, the Nairobi Stock Exchange, the Ghana Stock Exchange, and the Stock Exchange of Mauritius. The sample period includes secondary time series data from January 2005 to December 2019. The first goal was to develop and validate a measure of investor overconfidence. The second objective was to compare different levels of investor overconfidence in the selected African stock markets. The third objective was to evaluate the influence of investor overconfidence on the volatility of stock market return under changing market conditions, as described by the AMH. The estimation methods included the Generalised Methods of Moments dummy regression, regime-switching VAR models and rolling GARCH models, which are GARCH, EGARCH and TARCH. The results show that high investor overconfidence is more associated with bullish markets than periods of financial crises and bearish markets. The results also imply that it is not advisable to generalise the impact of market conditions on investor overconfidence across all the markets. Additionally, rolling GARCH estimates demonstrated that patterns of investor overconfidence evolve, consistent with the AMH. Assessing investor overconfidence under the AMH framework offers a stronger image of the adaptive behaviour of the Afri can equity markets. This research adds to existing knowledge in numerous ways. Foremost, it provides a standard measure of investor overconfidence in Africa’s equity markets. A measure that combines multiple proxies into a single index and neutralizes the disadvantages of each proxy when used separately to estimate investor overconfidence. Second, it provides a timely contribution to the effect of investor overconfidence on stock return volatility in African equity markets under the AMH paradigm. Third, according to the AMH, investor confidence is not vi static and can appear under specific market conditions and disappear under others. This bias occurs and disappears as market conditions change in the chosen African equity exchanges. This also shows that investor overconfidence is normal, changes over time and is adaptable in the African stock markets. Consequently, this study brings a new perspective regarding investor overconfidence and market efficiency in the face of the AMH paradigm. The results also have important implications for investors and brokers wishing to develop appropriate trading strategies. This study is also helpful for policymakers as they need to be wary about investor overconfidence impact on market momentum in periods of market expansion. This study argues that investor overconfidence in African stock markets conforms to the AMH than the EMH and the BF.