Stock returns in the time of COVID-19 pandemic

ABSTRACT

This paper investigates the dependence between COVID-19 shocks and stock market returns for seven countries: Canada, China, France, Germany, South Korea, the UK and the US. Using a quantile-on-quantile (QQ) regression analysis, we find that all stock markets responded to COVID-19 shocks mostly negatively. However, we also observe instances where an increase in the number of COVID-19 confirmed cases resulted in higher stock returns. This unexpected positive dependence captures the much debated stock market-real economy disconnect at the time of the onset of the COVID-19 pandemic. Furthermore, although stock markets’ responses to both the global and the country-specific COVID-19 shocks are similar across countries, substantial differences also exist. In particular, stock markets in the US and China differed in their responses to the global COVID-19 shocks, with a higher degree of heterogeneity observed for the Chinese market. Finally, Asian markets reacted differently to country-specific COVID-19 shocks when compared to the US market. This finding could be explained by the fact that investors in the US stock markets may form their expectations mostly from the US COVID-19 shock, while investors in Asian countries, including China, may value more the information transmitted through the global COVID-19 shock.

KEYWORDS:

• COVID-19 shocks
• epidemiological growth models
• stock returns
• QQ regression analysis

JEL CLASSIFICATION:

• C14
• C22
• C53
• C58
• G15.

Acknowledgments

We are grateful to the Editor Professor Mark Taylor and two anonymous referees for their constructive comments and suggestions. We also thank Sean Parry for his support with the R code. Associate Professor Doko Tchatoka acknowledges the financial support from the Australian Research Council (ARC) through the Discovery Project Grant DP200101498.

Notes

¹ Numbers obtained from the https://covid19.who.int/ WHO Coronavirus (COVID-19) Dashboard on 5 July 2021.

² Here, a negative stock market refers to a relatively poor performing market, while a negative COVID shock is defined as a relative increase in confirmed cases.

³ See, e.g. Louis-Philippe, Brodeur, and Taylor (2020); Baker et al. (2020); Hassan et al. (2020); Baker et al. (2020).

⁴ In our regression, we use the growth rate of the policy variables. As such, the economic support index – which records measures such as income support and debt relief – was excluded because its growth rate was mostly zero during the sample period covered.

⁵ A similar grid is used in Sim and Zhou (2015) and Doko Tchatoka, Masson, and Parry (2019).

⁶ See, e.g. Gormsen and Koijen (2020) who analyse changes in the aggregate dividend growth rates and market discount rates due to COVID-19.

⁷ See https://hbr.org/2020/03/understanding-the-economic-shock-of-coronavirusCarlsson-Szlezak, Reeves and Swartz.

Additional information

Funding

This work was supported by the Australian Research Council [Discovery Grant DP200101498].