The Non-Linear and Linear Impact of Investor Sentiment on Stock Returns: An Empirical Analysis of the US Market

Gebka, B

doi:10.1007/978-1-4614-8060-0

The Non-Linear and Linear Impact of Investor Sentiment on Stock Returns: An Empirical Analysis of the US Market

Lookup NU author(s): Professor Bartosz Gebka ORCiD

Downloads

Full text for this publication is not currently held within this repository. Alternative links are provided below where available.

Abstract

In this study, we unify the tests of the existence of linear and non-linear relationships between sentiment and future returns. In addition, we investigate whether the existing contradictory findings could be due to the fact that results reported elsewhere were not robust but affected by sample-specific features, quantile regressions are employed to analyse the robustness of unveiled causality patterns across the conditional distribution of stock returns and volatilities. Our findings show a significant and robust positive causality from sentiment to future returns, especially from high sentiment, and a non-robust negative causal impact of positive lagged sentiment volatility to returns. In addition, return volatility is driven by volatility of positive lagged sentiment, albeit this relationship is not robust as it only exists in a narrow range of quantiles. Hence, we unveil the existence of both linear and non-linear causality between sentiment and future stock returns. In light of the DSSW (1990) model, these findings suggest that the hold more effect dominates the price pressure effect and the create space effect is stronger than the Friedman effect.