Predicting Stock Market Dynamics with Google Trends

Abstract

We investigate investor attention measured by search volume index (SVI) data from Google Trends, and its impact on returns. Moreover, we examine the features of SVI as both an explanatory variable at time zero and its predictable powers for future returns. We collect the data from all companies on the S&P 500 between 2014-2018. This paper aims to contribute to existing studies on the subject by replicating previous research methods with current data. We provide evidence that SVI has a statistically significant negative impact on short-term returns. Furthermore, we argue that there is a relatively weak relationship between SVI and other traditional indirect proxies for investor attention.