Investor Sentiment and Stock Market Return Predictability Evidence from the Fear and Greed Index, VIX, and AAII Investor Sentiment Survey

Sammanfattning

This thesis examines whether investor sentiment can predict future stock market returns and if such predictability is economically meaningful, consistent across markets, and stable over time. Three sentiment proxies are evaluated, the CNN Fear and Greed Index, the Volatility Index (VIX), and the AAII Investor Sentiment Survey. The analysis covers six equity indices which are the S&P 500, Dow Jones Industrial Average, NASDAQ Composite, Russell 2000, OMX Stockholm 30, and OMX Stockholm Small Cap. Across OLS regressions, cumulative forward return regressions, VAR models with Granger causality tests, sub-period analysis, and a sentiment-based trading strategy, the results show mixed evidence of predictability. The VIX exhibits the strongest and most consistent predictive power, although its relationship with future returns is also consistent with a rational volatility premium. The Fear and Greed Index provides weaker but economically meaningful evidence, concentrated in OMX Stockholm Small Cap and supported by a trading strategy that outperforms a buy-and-hold benchmark on a risk-adjusted basis in the out-of-sample period. The AAII survey shows limited predictive ability across all methods. Overall, the findings suggest that investor sentiment can contain predictive information, but its usefulness depends on the sentiment measure, market segment, and prevailing market conditions.