Does one follow the other? A time-varying information spillover network of social media sentiment and news sentiment

Paper Overview: Investor sentiment has been studied extensively to have an impact on asset prices and volatilities. In this study, we propose the transfer entropy-based information spillover networks to examine how different sentiment sources can spillover from one asset to another. By capturing the directional and non-linear information flows dynamics, we find that the constructed sentiment spillover networks reveal some unique properties to explain the sentiment spillovers in the financial markets. This work is currently under review.