Overconfidence and disposition effect in the stock market : A micro world based setting

Modern financial theory relies on the rationality assumption of investors even though, evidence suggests that market investors are affected by behavioural biases such as overconfidence and disposition effect. Overconfident investors perceive situations better than what they actually are, while investors exhibiting disposition effect tend to dispose winner shares and keep loser ones. However there is not clear causal relationship between both biases. We contribute to the literature about overconfidence and its relationship with the disposition effect, using a simulation model often named micro world, representing an artificial financial stock market. We propose a methodology combining qualitative (QCA) and quantitative (Logistic Regression) techniques to correlate transactions' outcomes with investors' characteristics. Results suggest that overconfidence is explained by gender, career and education level, while age, nationality, and profits are not significant variables. We also confirm that investors exhibiting disposition effect are more prone to be overconfident.