A stock’s price is driven more by how much it 'stands out' to people than by any of its actual financial data.
March 19, 2026
Original Paper
Behavioral Factors in Asset Pricing: An Approach Through Machine Learning
SSRN · 6439412
The Takeaway
Investors often believe that stock prices are determined by complex risk metrics or profit forecasts. This study found that 'salience'—the simple psychological factor of how much a stock catches the eye—is a more dominant predictor of returns than nearly every traditional financial metric.
From the abstract
This study employs machine learning (ML) techniques to evaluate the importance and economic significance of behavioral factors among an extensive set of well-established pricing factors. Empirical findings reveal that ML improves return prediction accuracy and generates significantly larger positive return spreads compared to conventional Fama-MacBeth (FM) regressions. Behavioral factors-particularly the salience theory value-exhibit exceptional performance in both factor importance and economic