A Reappraisal of the Causal Relationship between Sentiment Proxies and Stock Returns

Abstract

This article investigates the suitability of 13 investor sentiment proxies as causal explanations for monthly stock returns of the European S&P 350 constituents over 45 years. We analyzed a sample of 362 companies covering 16 European countries. Analyses incorporate multiple categories of investor sentiment arising from market or survey data, as well as technical analysis, risk measures, company fundamentals and macroeconomic variables. In addition, we provide an extended review of sentiment proxy measures based on market data. This work applied general model of moments (GMM) to dynamic panel data to estimate short-run and long-run influences along with Granger causality. Our findings demonstrate the role of several investor sentiment measures in predicting stock returns even after controlling for variables such as fundamentals, macroeconomic, market and technical analysis. Together, sentiment measures of implied volatility in stock options, such as VIX and VSTOXX, put and call ratios, gold, government bond yield spreads, mispricing along with economic and confidence sentiment indicators can significantly predict how irrational behaviors of investors can determine stock returns. Co-movements between markets provide further evidence of contagion. Considering the unobservable nature of sentiment, we provide a set of sentiment proxies, and reveal new measures such as gold, government yields spread and a mispricing ratio, that serve to predict European market returns. Furthermore, to our knowledge this study is one of the few studies to apply time dynamic panel data estimation to a large set of sentiment proxies and a set of complete control variables in a long-term framework.

Keywords:

• Investor sentiment
• Return causality
• Stock returns
• Behavior
• Optimism
• Pessimism

Additional information

Funding

This work is financed by national funds through FCT - Fundação para a Ciência e Tecnologia, I.P., under the project UIDB/05583/2020. Furthermore, we would like to thank the Instituto Politécnico de Viseu and CISeD for their support.