Monetary Policies and Financial Stress During the COVID-19 Pandemic: An Event Study Analysis

ABSTRACT

Using an event study design, this paper examines the effects of announcements of financial policies, especially monetary policies, on a measure of financial stress in some advanced and emerging economies during the COVID-19 pandemic period. We construct a daily financial stress index for 15 countries during the period from April 1, 2019 to September 30, 2021 . Our results show that announcing financial policies of any type increased financial stress on the day the policy was announced but the effect faded away rather quickly. Moreover, different types of financial policy announcements had different effects on the financial stress subindices. We also find that each component of financial stress responds to the announcement of financial policies differently and announcements of financial policies affect financial stress in most of the countries in our sample, but to different degrees.

KEYWORDS:

• Monetary policy
• unconventional monetary policy
• financial stress
• financial instability
• COVID-19

JEL:

• E44
• E52
• E61
• G01

Disclosure statement

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

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/1540496X.2022.2148462.

Notes

1. Event study approach has been used to study the effects of various events on the financial market such as the US’s government shutdown (Nguyen et al. 2021), The PRC-US trade war 2016–2018 and recently, announcement of the COVID-19 pandemic and governments’ announcements of stimulus packages (e.g., Harjoto et al. 2021; He et al. 2020; Naidu and Ranjeeni 2021; Rahman, Amin, and Al Mamun 2021).

2. We also use different event windows [−3,3] and [−10; 10] as robustness checks.

3. For robustness check, we also estimate the restricted approach using the window of [−5;+5] instead of window [−2;+2]. That is, the event days are days that only one type of instrument was implemented and there were no other policies implemented five days before and five days after the event day.

4. There are two reasons for us to use the BIS database as the main database. First, we focus more on the effects of monetary policy on financial stress. Second, the BIS database, in our view, is more comparable across countries.

5. We carefully checked the two databases to avoid duplication. We find that, for each type of financial policy that we are focusing on, information on the announcements from the two databases is quite consistent.

6. The variance-equal weighting approach is widely used in the literature and has proven to be a very efficient method in constructing financial indices due to the simplicity of calculations and its accuracy in representing and signaling financial stress and episodes of turbulence (e.g., Cardarelli, Elekdag, and Lall 2011).

7. For the case of Indonesia and Malaysia, we use 3-year Indonesian and Malaysian government bond since the Indonesian and Malaysian governments do not issue 2-year bonds.

8. Descriptive statistics for our empirical results are presented in Appendix Table A1.

9. Appendix Table A2 presents the number of policies implemented when we use the restricted methods. Adopting the restricted method reduces the number of policies implemented from 607 to 315.

10. As mentioned earlier, we also use a more restricted approach with the window of [−5; +5]. The number of policies implemented reduced from 315 with window [−2; +2] to 275 with window [−5;5]. The estimation results are presented in Online Appendix 2.

11. Please refer to Online Appendix 3 for more detail on the Diebold-Yilmaz (2012) approach. Online Appendix Table 5 presents the connectedness table, obtained from estimating a system VAR equation using the Diebold-Yilmaz (2012) approach.

12. That is, after removing the contributions of other countries’ FSI in one country’s FSI variation.