Decoding market psychology: a multi-frequency, cross-market framework for return predictability and market reversals

This study constructs a novel country-specific composite investor sentiment index comprising of implied volatility, trading volume, bitcoin, crude oil and FEARS, to investigate the dynamic sentiment return relationship across seven major economies namely, the United States, Canada, the United Kingdom, Germany, France, Italy, and Japan. Using high-frequency data at daily, weekly, and monthly intervals, the study examines three key aspects: the co-movement between sentiment and stock returns, the predictive power of sentiment on future market performance, and the tendency of sentiment to persist or reverse over time. Findings indicate a statistically significant and positive co-movement between sentiment and returns across most markets and time frequencies, with stronger effects observed at higher frequencies, highlighting the behavioral foundations of market fluctuations. Moreover, investor sentiment demonstrates robust predictive power for future returns, particularly at weekly and monthly horizons, lending empirical support to behavioral finance theories. Additionally, patterns of sentiment reversal are more prominent at short term and long-term intervals, whereas medium term sentiment tends to exhibit persistence. The study offers implications for policymakers and market participants, suggesting that the integration of sentiment indicators into forecasting models and risk management frameworks enhance the effectiveness of investment strategies and contribute to improved market stability.