The sum of all SCARES COVID-19 sentiment and asset return

Abstract

In this study, I constitute a search based COVID-19 sentiment index using Google search volume. I develop an alternative Scared COVID-19 Attitude Revealed by Eager Search (SCARES) index using the household search volume i.e. "coronavirus pandemic", "coronavirus epidemic", and "coronavirus outbreak" of United States (US) during the COVID-19 pandemic. Using daily data from May 1, 2020 to July 30, 2021, I find that SCARES index negatively explains stock market return and subsequent return reversals, implying that households' increased pandemic sentiment negatively affects equity market return. Furthermore, decile regressions on characteristics-sorted portfolio returns show that SCARES index predicts the return reversals of firms that are small, less profitable, and with low investment. I also report that COVID-19 search shocks of households do not significantly predict any of the Fama-French five-factors except SMB (small-minus-big). Moreover, I use two state Markov switching model and find that structural breaks associated with pandemic phases make SCARES positively related to indices i.e. twitter based uncertainty, volatility index, economic policy uncertainty, and business condition in high volatility regime. Finally, sub-period analysis reports that, in stock market context, people start to react slowly and become relatively less responsive to the COVID-19 search keywords. The findings of this paper can assist key stakeholders in the market to carefully analyze the asset return pattern during pandemic regimes.

Keywords: COVID-19; Google trend; Markov switching; SCARES; Sentiment.