Overconfidence vs. Herding: Analyzing Investor Behavior in the Moroccan Retail Trade Sector

Abstract

<jats:p>This study looks at overconfidence and herding in the Moroccan stock market, focusing on the retail trade sector. We use daily data from four major stocks; Autohall, Ennakl, Label Vie, and Total Maroc, covering March 2015 to January 2025. Investor behavior is analyzed with regression and rolling-window methods. Overconfidence is tested by seeing how volatility and past returns affect trading volume, while herding is measured using CSAD metrics for both trading volume and returns. The results provide strong evidence of overconfidence, particularly in Label Vie and Total Maroc, where volatility coefficients are significantly positive (ß = 1.52e+06 and ß = 6.17e+05, respectively). These findings confirm that higher volatility encourages excessive trading. Rolling regressions further show that overconfidence is time-varying and becomes more pronounced during volatile market conditions. By contrast, herding behavior appears largely absent. The squared market volume coefficient (ß = -1.512e-12, p = 0.747) and squared return coefficient (ß = 4.46, p < 0.01) indicate that dispersion increases rather than contracts during periods of elevated market activity, contradicting the expectations of herding models. Time-varying t-statistics also remain consistently above -2, confirming the lack of persistent herding across the sample period. Overall, the findings show that Moroccan retail investors are more influenced by their own confidence than by following the crowd.</jats:p>