The Interconnectedness between COVID-19 Uncertainty and Stock Market Returns in Selected ASEAN Countries

ABSTRACT

This paper examines the interconnectedness between the COVID-19 uncertainty index and stock returns in selected ASEAN countries. Results from the study, which uses the dynamic connectedness approach, show that on average, 47.06% of a shock on one index spills over to all other indices. This indicates that stock market returns are highly interconnected to the COVID-19 uncertainty index in ASEAN countries. However, the COVID-19 uncertainty index is not a predictor of stock returns in the ASEAN countries chosen for the study.

KEYWORDS:

• COVID-19 uncertainty
• volatility spillover
• dynamic connectedness
• stock returns
• Westerlund-Narayan predictability test
• ASEAN

JEL CLASSIFICATION:

• G11
• G15
• G18

Acknowledgments

We are thankful to the editor-in-chief and anonymous reviewers for their valuable comments and suggestions. The earlier version of this paper was presented in the 15^th Bulletin of Monetary Economics and Banking (BMEB) conference on September 2-3, 2021, and the authors would also like to thank the discussants and participants for their valuable comments and suggestions.

Disclosure Statement

No potential conflict of interest was reported by the author(s).

Notes

1. Using the traditional ADF and PP unit root tests to check the stationary property of stock returns is not the best means of addressing the presence of structural break dates in return series. The ideal way to test the unit root is by employing the two structural breaks tests, proposed by Narayan and Popp (2010), which is the most popular unit root test in recent literature (see Rath and Akram, 2021). However, this study does not use the Narayan and Popp (2010) unit root test because the period considered is relatively short (during the COVID-19 pandemic) and the study assumes that the return series will follow a stationary path irrespective of structural breaks. The other reason for not using the Narayan and Popp (2010) test is that the structural break–based predictive regression model may not perform well (Sharma 2016).