Uncertainty, Disagreement and Sentiment effect_data

This study proposes a sentiment asset pricing model featuring parameter uncertainty and disagreement within a framework of Bayesian learning. In the model, the stock-level sentiment is constructed as the cross-sectional wealth-weighted average of investors’ sentiments, and disagreement is constructed as their variance. We document that stock price is positively correlated with parameter uncertainty and disagreement when sentiment is optimistic but negatively when sentiment is pessimistic. Accordingly, parameter uncertainty and disagreement amplify sentiment-induced mispricing. Moreover, empirical evidence aligns with the model predictions, and further analysis confirms that the amplifying effects of parameter uncertainty and disagreement cannot be explained by existing explanations.