Abstract
This paper aims to test two competing explanations for the contrarian profits in ten MENA stock markets: the risk-based rational explanation and the representativeness-based behavioral explanation. We found that the CAPM, the three-factor model of Fama and French (1996) and the five-factor model of Fama and French (2015) fail to capture the contrarian profits in most MENA stock markets, which indicates that these profits cannot be explained by the risk. Moreover, the hypothesis that the representativeness heuristic leads to investors' overreaction does not hold for all the stock markets. In fact, only Egypt, Jordan, Saudi Arabia, and Kuwait tend to overreact to a series of similar earnings surprises and the magnitude of the market overreaction increases with the length of such series. To explicitly examine whether representativeness explains the contrarian profits, we extended the five-factor model to include a factor based on similar past earnings surprises. Inconsistent with the representativeness heuristic, we found that the contrarian profits in most MENA markets are not captured by a zero-investment portfolio that is short on the portfolio with a series of positive earnings surprises and long on the portfolio with a series of negative earnings surprises. However, with a long series of earnings surprises the contrarian strategy in Kuwait and Oman seems to lose its profitability. These findings indicate that the effect of the representativeness heuristic may depend on the persistence of past performance.