Investor Attention and Global Stock Market Volatility: Evidence from COVID-19

This paper utilizes intraday five-minute stock market indices to investigate the causal relation between global stock market volatility and investor attention measured by the Google search volume index during the COVID-19 pandemic. Using the bi-power variation method proposed by Barndorff-Nielsen and Shephard (2004), we separate the realized volatility into two components: Continuous and Jump. Based on 5,583 stock indices-day observations, we find that investor attention is positively related to the realized volatility and its continuous component, but to a lesser extent to jumps. A growth in confirmed cases is positive to all measures of market volatility. Moreover, when the number of confirmed cases increases, more attentive investors reduce market volatility. Our findings are robust regarding various estimation approaches and are less likely to suffer from omitted variable biases and endogeneity concerns. Understanding the findings revealed in this paper is crucial to regulators and policymakers as warnings of additional risks facing retail investors around the globe over the extremely volatile periods.