COVID-19: A pandemic for financial
markets?
Folkert te Velde
S3142426
University of Groningen
Faculty of Economics and Business
MSc. Finance
Supervisor: Dr. I. Souropanis
January 2021
Poortstraat 12A
9716JH Groningen
+31611822044
F.te.Velde@student.rug.nl
Abstract
This paper studies the impact of the COVID-19 outbreak on stock returns of the 75 largest firms listed on the Euronext Amsterdam. We obtain the results by using a Prais-Winsten panel regression with panel-corrected standard errors over the period of 1 January 2020 to 21 August 2020. Firstly, we report that stock returns are negatively affected by daily new confirmed COVID-19 cases. However, the relation between stock returns and daily new confirmed COVID-19 deaths is insignificant. Secondly, we observe that the impact of COVID-19 is more pronounced for the largest firms in the sample. Thirdly, the results show that the impact of COVID-19 is different for the various sectors studied in this paper. Fourthly, we note that financial markets react negatively to national press conferences. Although, when the distinction is made between good and bad press conferences, we see that investors react strongly to bad press conferences and weakly to good press conferences. Lastly, a robustness test shows that the reaction to COVID-19 is strong for the first wave, lasting until 24th of June, and weak for
the second wave.
1
1. Introduction
SARS-CoV-2, more commonly known as the coronavirus or COVID-19, will be remembered in the history books for years to come.1 The virus originated in Wuhan City, capital of the
Chinese province Hubei, on the 31st of December 2019. Nearly a month after its inception and
seven days after the lockdown of Wuhan City, the World Health Organisation (WHO) declared the disease a Public Health Emergency of International Concern (WHO, 2020A). COVID-19 proved to be extremely infectious as the rapid spread of the disease caused it to be declared as a pandemic on 11 March (WHO, 2020B).2
The pandemic impacts nearly every facet of the daily life of people around the globe. To combat the disease, many countries adopted quarantines, lockdowns, curfews, stimulus packages and other poignant measures. Restaurants, cafes and bars were forced to close, most sports activities were no longer possible and citizens were urged to work from home. Consequentially, the virus also impacts the world economy. The Euro area’s GDP decreased 11.80% in the first quarter of 2020 whereas the GDP of the United States decreased 8.99%.3 Additionally, as of June 2020, global economic growth is projected at -4.9% (IMF, 2020).4
Financial markets are not spared from the pandemic either. At the beginning of the year, indices all over the world reported record drops as fear of the economic impact of the virus rose.5 A vast body of literature has already studied the impacts of this new pandemic on different aspects of financial markets. These include the impact on stock returns (Al-Awadhi et al., 2020; Ashraf, 2020; Khan et al, 2020; Chia et al., 2020; Salisu and Vo, 2020; Davis and Hansen, 2020), the effect on stock market volatility (Ali et al, 2020; Devpura and Narayan, 2020; Albulescu, 2020; Baker et al., 2020) and the effect on other financial indicators such as gold, cryptocurrencies and oil prices (Corbet et al, 2020; Mensi et al, 2020; Mnif et al., 2020; Conlon and McGee, 2020; Prabheesh et al, 2020; Iyke, 2020).
1 The virus was named COVID-19 by the WHO at the media briefing of 11 February 2020. Accessed
on December 7, 2020 from: https://www.who.int/dg/speeches/detail/who-director-general-s-remarks-at-the-media-briefing-on-2019-ncov-on-11-february-2020
2 As of 16 August 2020, the total number of worldwide confirmed COVID-19 cases was 21.2 million
and the total number of deaths was 761,000 (WHO, 2020C). For a concise overview of the origin and progression of the COVID-19 outbreak, see Appendix A.
3 Data of Real GDP is collected from the website of the Federal Bank of St. Louis. The series of the
Real GDP of the Euro-area is: CLVMEURSCAB1GQEA19, real GDP of the United States is: GDPC1.
4 The current situation is in sharp contrast to the Global Risks Report published by the World
Economic Forum on 15 January 2020. The top five risks to the world economy were all environmental in nature. Infectious diseases were ranked at the 10th place and considered to be quite unlikely (WEF, 2020).
2 In this study, we expand on the first strand of the aforementioned literature. We will analyse the impact of the novel COVID-19 virus on stock returns of the 75 largest firms listed on the Euronext Amsterdam in the Netherlands using a panel regression analysis. As well as the rest of the world, the Netherlands is greatly affected by COVID-19. The first detected case in the Netherlands dates back to 27 February.6 Soon after, the number of Dutch COVID-19 cases
spiked, leading to the initiation of the so-called “intelligent lockdown” on the 23rd of March.7
The intelligent lockdown prohibited most of the daily-life activities of Dutch citizens. This, in turn, greatly affected the Dutch economy, as well as financial markets. At the end of February and the beginning of March, stock markets experienced large drops and tremendous losses. All in all, this provides us enough incentives to study Dutch stock market behaviour during the COVID-19 outbreak.
In this paper, we analyse the impact of COVID-19 on financial markets in the Netherlands. Firstly, we study the overall effect of COVID-19 and report that it is significantly negative for new confirmed cases and insignificant for new deaths. Secondly, we provide evidence that the impact of COVID-19 is different depending on firm size and economic sector of the firm. Thirdly, we show that press conferences regarding COVID-19 have a significant negative impact on stock returns, in particular those relating to a worsening of the situation. Finally, we report that the effect of COVID-19 is significant during the first wave and insignificant during the second wave.
This research contributes to the existing body of literature in several ways. Firstly, to the best of our knowledge, this is the first study to research the effect of COVID-19 on the stock markets in the Netherlands. To this date, several COVID-19 related studies have been performed on individual countries. Al-Awadhi et al. (2020) study China, Chia et al. (2020) focus on Malaysia and Davis and Hansen (2020) analyse the United States. It is important to study individual countries, because countries responded differently to the pandemic. For instance, China initiated a complete lockdown of Wuhan City, the Malaysian government instated a movement control order and the Netherlands introduced an “intelligent lockdown”. COVID-19 related measures differ greatly around the globe. Hence, it is only logical that investors respond differently to the pandemic. Secondly, this study is among the first to study the impact of national press conferences. Inspired by the studies regarding asymmetrical reactions to news (e.g. Akhtar, 2013; Soroka, 2006; Giner and Rees, 2001), we study the impact of national press conferences, hosted by prime minister Mark Rutte, on stock returns of that day. A study closely
6A tourist, who recently travelled to the heavily contaminated North Italy, was the first to bring the
virus back to the Netherlands. (RIVM, 2020)
7https://www.government.nl/latest/news/2020/03/23/stricter-measures-to-control-coronavirus
3 related to this, is the paper of Baker et al. (2020). The authors use newspaper articles to explain the extraordinarily large reaction regarding the pandemic.
The remainder of this paper is structured in the following way. Section 2 provides a literature review, divided in two parts. The first part discusses previous literature regarding the effect of diseases and other real-life events on stock markets. The second part provides the hypotheses that will be tested later on. Section 3 reports the method of data collection and a preliminary analysis. Section 4 discusses the methodology used in the study. Section 5 reports the empirical results, along with theoretical explanations. Finally, section 6 concludes the paper with a summary and a discussion.
2. Literature Review
2.1 Previous literaturePrevious literature has found that pandemics affect financial markets. A pandemic can be defined as “an epidemic occurring worldwide, or over a very wide area, crossing international boundaries and usually affecting a large number of people” (John, 2001). However, before the COVID-19 outbreak, scant research has been performed on the impact of diseases on financial markets. Chen et al. (2009) studied the impact of the SARS outbreak in Taiwan and found that the outbreak had a positive effect on some industries and a negative effect on others. Additionally, Chen et al. (2007) reported that stock returns of Taiwanese hotels declined heavily after the SARS-outbreak. The Ebola outbreak of 2014 was also researched sparsely. Ichev and Marinč (2018) studied the effect of the Ebola outbreak on American firms, while controlling for geographic proximity. They showed that the impact of Ebola was the strongest for firms with operational exposure to the outbreak area (West African Countries and the U.S.). Lastly, Donedelli et al. (2017) researched the relation between disease-related news (DRNs) and stock returns. The authors report that there is a significant positive relation between DRNs and stock returns of pharmaceutical companies.
Unlike previous pandemics, there is already a large amount of literature regarding the economic impact of COVID-19 worldwide. Financial markets reacted more to the COVID-19 outbreak than previous pandemics and diseases. The explanation for this reaction is twofold: the economic impact of the pandemic and the role of investor sentiment in financial markets.
4 consumer spending, and the cancelation of many entertainment services. Thirdly, overall uncertainty will lead to an increase in the cost of capital. These effects are already discussed in recently published research. Barua (2020) notes that the global supply chain faces distress due to two main channels; production shocks and shocks to trade flows. Production shocks are caused by the closure of factories due to lockdowns. Country interdependence causes an economic ripple effect of these closures throughout the world. Shocks to trade flows are mainly caused by the closure of airports around the world, which greatly hinders transportation of labourers as well as materials. Both these shocks resulted in a record low for China’s Purchasing Managers’ Index in February, indicating a significant contraction of China’s production activities.8 The impact on the demand side is also described in the literature. Baker et al. (2020) state that the initiated quarantines, lockdown and voluntary social distancing caused a dramatic decrease in revenues of businesses and tourism worldwide. Moreover, the cancellation of the majority of events had a great negative effect on the services sector. Coibion et al. (2020) report that income and wealth losses, due to lockdowns, contribute significantly to the decrease in consumer spending, with the largest drops in the tourism and clothing industries. Andersen et al. (2020) report that aggregate consumer spending was 27% below the level of spending without COVID-19.
Secondly, investor sentiment, influenced by the pandemic, has an impact on financial markets. The field of behavioural finance has provided a vast amount of literature regarding the impact of investor sentiment. Studies have written about the impact of the weather (Saunders, 1993; Hirshleifer and Shumway, 2003; Kamstra et al. (2003), sporting games (Edmans et al., 2007; Kaplanski and Levy, 2010A;2014) the weekend effect (Lakonishok and Smidt, 1988; French, 1980; Keim and Stambaugh, 1984; Olson et al., 2015; Steeley, 2001), the Monday effect (Cross, 1973; Dubois and Louvet, 1996; Alt et al., 2011), the January effect (Thaler, 1987; Haug and Hirschey, 2006; Ciccone, 2011) and the holiday effect (Ariel, 1990; Kim and Park, 1994; Marrett and Worthington, 2009; Gama and Vieira, 2013) on financial markets. Furthermore, studies have provided evidence of the effects of real-world disasters and incidental events on stock returns. These include aviation disasters (Kaplanski and Levy, 2010B; Chance and Ferris, 1987), terrorist attacks (Drakos, 2010; Arin et al., 2008) and environmental disasters (Shan and Gong, 2012; Shelor et al., 1990; Aiuppa et al., 1993; Ewing et al., 2006; Weiderman and Bacon; 2008; Lee et al., 2007; Wang and Kutan, 2013; Capelle-Blancard and Laguna, 2010; Ferstl et al., 2012).
Previous research relates investor sentiment to stock market performance during these real-world events. Pandemics and diseases also fall in this category. News about pandemics can
8
5 spread fear and anxiety among individuals, resulting in a negative investor sentiment. Ichev and Marinč (2018) connect pessimistic investor moods to bad stock market performance during the Ebola outbreak. Additionally, Donadelli et al. (2017) mention investor sentiment as an explanation for the reaction to disease-related news of financial markets. Current research on COVID-19 often relates pessimistic investor sentiment to bad stock market performance and increased volatility (Chia et al., 2020; Chen et al., 2020; Baig et al., 2020; Mamaysky, 2020; Salisu and Akanni, 2020; Liu et al., 2020).
Investor sentiment can be linked to the media coverage of events. Blendon et al. (2004) report, after studying media coverage during the SARS outbreak, that the media disproportionately covers rare events, new events and dramatic events. Additionally, Klibanoff et al. (1998) show that investors assign more weight to news that was more extensively covered in the media, even though news with less media coverage might be just as important. Not only is regular news an important factor for COVID-19, the impact of social media is essential as well. Previous studies show that the general mood on social media platforms such as Twitter can predict financial markets (Bollen et al., 2011; Zhang et al., 2011; Luo et al., 2013). COVID-19 is a heavily debated topic on social media. Depoux et al. (2020) state that due to the massive dissemination of information and misinformation on social media platforms, the fear and panic caused by social media spread faster than the virus itself. Additionally, Gao et al. (2020) report that high values of social media exposure during the COVID-19 period is significantly positively related to mental health problems. Therefore, a high exposure to social media can further strengthen fear and pessimism among investors.
2.2 Testable hypotheses
In the following section, we will outline the hypotheses that are going to be tested in the study. Firstly, news about the number of new COVID-19 cases and deaths may deteriorate investor sentiment, which in turn might result in a decrease of stock returns. An increase of COVID-19 cases or deaths can cause anxiety, fear and pessimism by the general population and consequently stock market investors. Previous studies find that a positive or negative investor mood, caused by real-world events, can have a significant impact on stock market performance. Pandemics also fall in this category of real-world events. Ichev and Marinč (2018) find that the Ebola disease negatively impacts stock returns and Chen et al. (2009) find that the effects of the SARS disease can be both positive and negative. Moreover, the economic impact of the pandemic leads to pessimistic investor expectations, resulting in a downwards pressure on asset prices. This leads to the first testable hypothesis:
Hypothesis 1:
6 Secondly, investors may respond differently to COVID-19 cases relative to COVID-19 deaths. Ashraf (2020) finds, using a sample of 64 countries, that investors react stronger to daily new confirmed cases than daily new confirmed deaths. This finding is further supported in the COVID-19 literature (e.g. Al-Awadhi et al., 2020; Chia et al., 2020). One possible explanation is that investors pre-emptively incorporate the future growth of COVID-19 related deaths in current asset prices. When investors observe a large growth in COVID-19 cases, they can predict the growth of COVID-19 deaths a few days later. Hence, they already incorporate the expected negative consequences of COVID-19 in current asset prices. Another explanation is that investors react more to new cases because this is generally a higher number than new deaths. As is shown in Hartzmark and Sussman (2019), investors react strongly to easy-to-read figures and numbers. The higher numbers of COVID-19 cases could therefore lead to a stronger investor reaction. This leads to the second testable hypothesis:
Hypothesis 2:
The impact of the COVID-19 pandemic on stock returns is more pronounced for daily new confirmed cases than daily new confirmed deaths.
Thirdly, the impact of COVID-19 might differ depending on firm size. Brown and Cliff (2005) find that investor sentiment has a larger impact on small firms relative to large firms. This finding is supported by Donadelli et al. (2017) who find that disease-related news has the largest impact on small firms. However, Baker and Wurgler (2006) find that when investor sentiment is low (high), that small, young and hard-to-arbitrage stocks generate high (low) returns in subsequent periods. Moreover, the literature of the impact of diseases on stock returns reports ambiguous findings concerning the small-firm effect. On one hand, Ichev and Marinč (2018) find that the effect of Ebola is more pronounced for small firms. Furthermore, Chia et al. (2020) find that small firms experience larger losses due to COVID-19 than big firms. On the other hand, Al-Awadhi et al. (2020) find that the impact of COVID-19 is significantly more negative for large firms relative to small firms. To conclude, we expect that COVID-19 has a different impact on firms, depending on size. This leads to the third testable hypothesis:
Hypothesis 3:
The impact of the COVID-19 pandemic on stock returns differs depending on firm size.
7 it was negative for the beverages and transportation industries. Lastly, Mazur et al. (2020) report differences in the impact on various industries in the United States. The authors report that the natural gas, food, healthcare and software industries earned positive returns whereas petroleum, real estate, entertainment and hospitality sectors earned negative returns in the COVID-19 period. This leads to the fourth testable hypothesis:
Hypothesis 4:
The impact of the COVID-19 pandemic on stock returns differs between economic sectors.
Fifthly, national press conferences regarding COVID-19, hosted by the Dutch prime minister, might affect stock returns. During the COVID-19 period, numerous press conferences aired on national television. These ranged from good press conferences, where a decrease in COVID-19 cases or a relaxation of COVID-19 related measures was announced. There were also numerous bad press conferences, where an increase in COVID-19 cases or more severe measures were announced. Previous literature notes that financial markets react asymmetrically to news. Soroka (2006) and Akthar (2013) report that investors react strongly to negative news and weakly to positive news. Hence, we expect that bad press conferences have a negative impact on stock returns, whereas we expect no significant relation between good press conferences and stock returns. This translates to the fifth testable hypothesis:
Hypothesis 5:
National press conferences regarding COVID-19 have a moderating negative effect on the impact of COVID-19 on stock returns.
3. Data and preliminary analysis
3.1 Data collection8 AEX) exists because the weights of the indices are reviewed on an annual basis.9 Index prices are determined by way of a weighted average of the closing values of constituent stocks.10
Data is collected over the period of 1 January 2020 until 21 August 2020, resulting in a total of 12,062 observations.11 The cross-section of closing stock values, daily company market
capitalization and price-to-book ratio is gathered from Thomson Reuters’ Eikon. Daily stock prices were transformed into returns using the following formula:
𝑅𝑖,𝑡 =(𝑃𝑖,𝑡−𝑃𝑖,𝑡−1)
𝑃𝑖,𝑡−1 ∗ 100%, (1)
where 𝑅𝑖,𝑡 is the return of the stock i at day t, 𝑃𝑖,𝑡 and 𝑃𝑖,𝑡−1 are the closing value of stock i on
day t and t-1, respectively. Daily company market capitalization is transformed to the natural logarithm. These values are all denominated in Euros.
Additionally, the daily TED-spread is included in the model to capture macroeconomic effects. The TED-spread is calculated as the difference between the month LIBOR and the three-month T-bill interest rate. The time-series of the TED-spread, over the period of 1 January 2020 to 21 August 2020, is collected from the website of the Federal Reserve Bank of St. Louis.12
The data for daily new confirmed COVID-19 cases and deaths in the Netherlands, is collected from www.data.europa.eu, also covering the period of 1 January 2020 until 21 August 2020. However, due to the exchange closing on weekends and holidays, observations with missing stock data values had to be removed on days where no trading occurred. Furthermore, the number of daily new cases and deaths were transformed into growth rates using the cumulative number of daily new cases and deaths, respectively. In order to account for outliers, daily growth rates were winsorized at the 1% and 99% level.
In order to study sector-specific differences, firms were categorized in economic sectors using firm level information that was provided by https://www.investing.com. The following sectors are included in the sample: Basic Materials, Capital Goods, Consumer Cyclical, Consumer/Non-Cyclical, Energy, Financial, Healthcare, Services, Technology and Transportation.
9 For more information about regulations and admission rules, see https://live.euronext.com 10 For a list of all the firms included in the study, see Appendix B.
9 Finally, to analyse the impact of the news, information about the dates and content of national press conferences is collected from the website of the Dutch government.13 Furthermore, a distinction was made between good and bad press conferences. A good press conference announces a decrease in COVID-19 and/or COVID-19 related measures, whereas a bad press conference mentions an increase in COVID-19 and/or COVID-19 related measures. An overview of the press conferences is presented in Table 1, including the date and main takeaway of each conference. The total number of press conferences included in the sample is 17, including 6 good press conferences and 11 bad press conferences.
Table 1
Dates and takeaways of COVID-19 press conferences Date Important Takeaways
09/03/2020 Stop shaking hands, work at home when possible.
12/03/2020 Stay at home with light symptoms, first public events are cancelled. 19/03/2020 Minister of medical care resigns, more events are cancelled.
23/03/2020 Stricter COVID-19 related measures
31/03/2020 Current COVID-19 measures are extended until 28 April.
07/04/2020 Further updates about COVID-19, no new measures were announced. 15/04/2020 No relaxation of COVID-19 measures.
21/04/2020 Schools reopen, most measures are extended until 20 May.
06/05/2020 Several COVID-19 measures are relaxed due to the decrease of new cases. 19/05/2020 Relaxation of COVID-19 measures is confirmed for 1 June.
20/05/2020 Extension of financial support for local entrepreneurs for another three months. 27/05/2020 Gyms and saunas are allowed to reopen sooner then planned.
03/06/2020 Traveling to most European countries is allowed.
24/06/2020 Provisional final press conference, almost everything is allowed considering the 1,5-meter society.
22/07/2020 Number of COVID-19 cases is back on the rise.
06/08/2020 Call to adolescents to consider COVID-19 measures due to increase in new cases. 18/08/2020 New COVID-19 measures for certain regions.
Note: This table reports the dates and important takeaways from the national press conferences held by the Dutch prime minister regarding news about COVID-19. The data was collected from the website of the Dutch government over the period of 1 January 2020 until 21 August 2020.
3.2 Data description
The next section provides a preliminary analysis of the collected data, as described in the previous section. Figure 1 shows the time series graphs of the cumulative average returns, over the full sample period. As can be seen in the figure, stock returns drop significantly to around -0.15% on 12 March. Immediately after the drop, returns steeply increase, followed by a period where returns float around zero percent. Figures 2 and 3 report the time-series graphs of daily new confirmed COVID-19 cases and deaths, respectively. Panel A reports the absolute numbers and Panel B reports the growth rates. When looking at the number of daily new confirmed cases, we observe that the graphs are opposite to Figure 1. Figures 2 and 3 show that both new cases and deaths increase in the period after 12 March. Figure 2 clearly illustrates that the peak
13
10 of 19 is in March and April. Furthermore, we note that there is a clear second COVID-19 wave that starts around July. Moreover, when comparing the progression of new cases and deaths, we see that the peak of new deaths is later on in the period. The peak of new COVID-19 deaths is at the beginning of April, compared to the middle of March for new cases. This is in agreement with the study of Wang et al. (2020). The authors report that the median number of days from the first symptom until death is 14 days. Fourthly, Figure 4 illustrates the movement of the spread over the sample period. As can be seen in the figure, the TED-spread peaks around the beginning of March, reaching up to almost 1.50%. This is similar to the graphs of the daily new COVID-19 cases and deaths.
Figure 1
Time-series graph of the cumulative average returns
Note this figure reports the daily cumulative average returns of the stocks of the 75 largest firms on the Euronext Amsterdam over the period of 1 January 2020 until 21 August 2020.
Figure 2
Time-series graphs of daily new confirmed COVID-19 cases
Panel A: Daily New confirmed cases Panel B: Daily growth in new confirmed cases
11
Figure 3
Time-series graph of daily new confirmed COVID-19 deaths
Panel A: Daily new confirmed deaths Panel B: Daily growth in new confirmed deaths
Note: This figure reports the daily new confirmed COVID-19 deaths (Panel A) and the daily growth in confirmed COVID-19 deaths (=DGTDC) (Panel B) over the period of 1 January until 21 August 2020 in the Netherlands.
Figure 4
Time-series graph of the TED-spread
Note: This figure reports the TED-spread, published by the Federal Reserve Bank of St. Louis, over the period of 1 January 2020 until 21 August 2020 (163 observations).
12 5.47% with a maximum of 118.75%. Furthermore, the table illustrates that returns are negatively skewed, while the growth rates of daily new cases and deaths are positively skewed. This implies that large negative returns are more likely to occur than positive returns.
On the other hand, it is more likely that there are large positive COVID-19 growth rates than negative growth rates. Finally, the large value of Kurtosis implies that daily returns are not normally distributed, which is a common characteristic of daily stock returns (Brown and Warner, 1985).
Table 2
Descriptive Statistics
Statistics Returns lnMC PTB TED Cases Deaths DGTCC DGTDC
Mean -0.13% 21.42 3.16 0.37% 258.42 24.23 7.36% 5.47% Median 0% 21.29 1.79 0.20% 144 2 0.62% 0.05% St. Dev. 3.57 1.88 10.390 0.36 309.46 46.60 20.42 16.51 Minimum -30.43% 17.77 -113.56 0.11 0 0 0% 0% Maximum 27.06% 26.30 71.89 1.42 1224 234 137.50% 118.75% Skewness -0.43 0.40 -1.43 1.67 1.34 2.30 4.44 4.72 Kurtosis 11.91 2.59 40.06 4.39 3.74 7.69 25.05 29.19 Observations 12,062 12,062 12,062 12,062 12,062 12,062 12,062 12,062
Note: This table reports the descriptive statistics of the 75 largest firms listed on the Euronext Amsterdam in terms of market capitalization over the period of 1 January 2020 until 21 August 2020. Returns are the daily growth rates of the closing values of constituent stocks, lnMC is the natural logarithm of daily company market capitalization in Euros, PTB is the daily price-to-book ratio in Euros, TED is the daily TED-spread, Cases are daily new confirmed COVID-19 cases, Deaths are daily new confirmed COVID-19 deaths, DGTCC and DGTDC are the daily growth rates of the cumulative number of cases and deaths respectively.
Table 3 reports descriptive statistics per index. At the first glance, we do not detect significant differences. The mean returns are within reasonable range of the other indices and the means are all negative (-0.11% for AEX, -0.16% for AMX and -0.12% for ASCX). From the table, we observe that the AMX index has the largest spread and variance, while the AEX and ASCX are similar in spread and variance.
Table 3
Descriptive Statistics of the returns per index
Indices Mean (%) Median (%) St. Dev. Minimum (%) Maximum (%) Observations
AEX -0.11 0 3.38 -27.90 20.26 3912
AMX -0.16 -0.07 3.71 -30.43 27.06 4075
ASCX -0.12 0 3.60 -26.63 25.20 4075
Note: This table reports the descriptive statistics of the returns of the 75 largest firms listed on the Euronext Amsterdam over
13 Table 4 reports summary statistics for each sector that is represented on one of the aforementioned indices. The table shows that most of the sectors have a negative mean return over the full sample period. Notably, the energy sector performed the worst (-0.40%) while the Technology sector is the only sector with a positive mean (0.08%). Furthermore, we observe that the standard deviations of the sectors are similar. However, there is one outlier. The standard deviation of the energy sector is considerably higher than the other sectors (4.64).
Table 4
Descriptive Statistics of the returns per sector
Sectors Mean (%) Median (%) St. Dev. Minimum (%) Maximum (%) Observations Basic Materials -0.11 -0.05 3.38 -23.59 17.57 1793 Capital Goods -0.15 -0.08 3.61 -22.23 13.81 1141 Consumer Cyclical -0.15 0 3.78 -21.53 16.91 489 Consumer/ Non-Cyclical -0.12 0 2.16 -15.26 11.07 815 Energy -0.40 -0.38 4.64 -30.43 21.62 489 Financial -0.16 -0.02 3.86 -26.63 24.74 1304 Healthcare -0.04 .011 3.94 -27.90 24.15 978 Services -0.21 -0.07 3.65 -25.78 27.06 2934 Technology 0.08 0.14 3.29 -22.07 25.20 1793 Transportati on -0.32 -0.13 3.41 -13.58 12.87 326
Note: This table reports the descriptive statistics of the returns of economic sectors that constitute firms listed on one of the
main indices on the Euronext Amsterdam (AEX, AMX, ASCX) over the period of 1 January 2020 until 21 August 2020.
14 Table 5 Correlation Matrix Returns lnMC PTB DGTCC DGTDC TED Returns 1.00 lnMC 0.01 1.00 PTB 0.02 0.14 1.00 DGTCC -0.16 -0.01 0.00 1.00 DGTDC -0.16 -0.03 -0.02 0.27 1.00 TED 0.04 -0.05 -0.02 0.25 0.59 1.00
Note: This table reports the correlation matrix of the variables used in the sample. Returns are the daily growth rates of the closing values of constituent stocks, lnMC is the natural logarithm of daily market capitalization in Euros, PTB is the daily price-to-book ratio in Euros, DGTCC and DGTDC are the daily growth rates of the cumulative number of cases and deaths, respectively and TED is the TED-spread.
4. Methodology
4.1 Model specificationTo analyse the impact of new COVID-19 cases and deaths on stock returns, we employ a panel data regression. There are a number of arguments in favour of using panel data analysis. Firstly, COVID-19 is not a one-time event. The peak of the virus is not at the beginning of the period and its impact on society grows over time. Secondly, panel data analysis is able to identify the time-varying relationship between the dependent and independent variables (Baltagi, 2008). Thirdly, panel data deals with problems relating to estimation bias, multicollinearity and heteroscedasticity. Additionally, it controls for firm heterogeneity (Hsiao, 2014; Wooldridge, 2010). The following model will be used to estimate stock returns:
𝑅𝑖,𝑡 = 𝛼 + 𝛽1𝐶19𝑡−1+ 𝛽2𝑙𝑛𝑀𝐶𝑖,𝑡+ 𝛽3𝑃𝑇𝐵𝑖,𝑡+ 𝛽4𝑇𝐸𝐷𝑡+ 𝜀𝑖,𝑡 , (2) where 𝑅𝑖,𝑡 is the daily return of stock i on day t, 𝛼 is the regression intercept, 𝐶19𝑡−1 is either
(1) the growth in daily new confirmed COVID-19 cases or (2) the growth in daily new confirmed COVID-19 deaths of the previous day, 𝑙𝑛𝑀𝐶𝑖,𝑡 is the natural logarithm of daily
market capitalization of firm i on day t, 𝑃𝑇𝐵𝑖,𝑡 is the daily price-to-book ratio of firm i on day t, 𝑇𝐸𝐷𝑡 is the TED-spread on day t and 𝜀𝑖,𝑡 is an error term. In the model, market
capitalization and the daily price-to-book ratio are used to control for firm-specific
characteristics. The TED-spread as used as a proxy for systematic risk (Boyson, 2010; Boudt, 2017).
4.2 Parameter expectations
15 a fixed number of shares outstanding, an increase in market capitalization is paired with an increase in the stock price. An increase in the stock price leads to an increase in stock return. Hence, it is expected that 𝛽2 > 0. Furthermore, since the book values of the firm’s shares are mostly fixed, an increase in the price-to-book value comes from an increase in the stock price. Again, an increase in the stock price is paired with an increase in stock return. Therefore, we predict that 𝛽3 > 0. Finally, the TED-spread is used as a proxy for global systematic risk, where an increase in the TED-spread is related to periods of low investor confidence (due to increasing risk premia) and a decrease is related to high investor confidence. Hence, we expect that 𝛽4 < 0.
5. Empirical analysis
The following section will discuss the empirical results acquired by estimating Equation (2). In the first section, we will discuss the model specification. In the second section, the main regression will be analysed. In the third and fourth section, the model including index-specific and sector-specific dummies will be discussed. The fifth section will analyse the impact of national press conferences. Finally, the sixth section analyses whether there are differences in the results when separating the sample in the first and second COVID-19 wave.
5.1 Regression specification
In this section we will discuss the various tests that were performed in order to determine the most appropriate estimation method. Table 6 reports the values of these diagnostic tests, using Equation (2) for both the model with daily confirmed COVID-19 cases and deaths.
16 standard errors are lower and the R2 is higher than their true values. The table reports a significant test statistic for every level. Hence, we conclude that the idiosyncratic errors in the model are serially correlated.
Taking these tests into account, we have chosen to estimate Equation (2) using a Prais-Winsten regression with panel-corrected standard errors (PCSEs), while assuming AR1 autocorrelation within panels and controlling for firm fixed-effects. This model is the most suitable for the sample because it assumes that errors are, by default, heteroskedastic and cross-sectionally dependent (Beck and Katz, 1995).
Table 6
Diagnostic tests
Test Value (DGTCC) Value (DGTDC) Conclusion
Hausman test (Χ2) 83.24*** 80.34*** Fixed-effects preferred over
random-effects Modified Wald test for
groupwise heteroskedasticity (Χ2)
1756.71*** 1841.05*** Errors are heteroskedastic
Breusch-Pagan LM test of independence (Χ2)
81200.52*** 87197.503*** Cross-sectional dependence/contemporaneous
relationship in the errors Wooldridge test for
autocorrelation (F-test)
1335.73*** 1614.655*** Serial correlation in the errors
Note this table reports the test statistics regarding diagnostic tests performed on Equation (2) for both the daily growth in confirmed COVID-19 cases and deaths. Where the first column presents the test that was run, the second and third column present the values of their respective tests and column 4 reports the conclusion that is drawn from the test; *,**,*** report statistical significance at the 10%, 5% and 1% levels, respectively.
5.2 Main analysis
17
Table 7
Main Regression
Note: This table reports the results of the panel data regression estimated with Equation (2) using a sample of the 75 largest firms listed on the Euronext Amsterdam over the period of 1 January 2020 until 21 August 2020. Panel A reports the results when the daily growth in confirmed COVID-19 cases (DGTCC) is used as independent variable. Panel B reports the results when the daily growth in confirmed COVID-19 deaths (DGTDC is used as independent variable. The independent variable of the regression is Ri.t, which denotes the returns of stock i on day t. α is the intercept, lnMC is the natural logarithm of daily
firm market capitalization, PTB is the price-to-book ratio and TED denotes the TED-spread retrieved from the Federal Reserve Bank of St. Louis. The Wald Χ2-test is conducted to check for joint significance of the variables. Panel-corrected standard
errors are in parentheses; *,**,*** denote statistical significance at the 10%, 5% and 1% level, respectively.
Taking the previously mentioned results into account, we conclude that investors react strongly to new confirmed COVID-19 cases and weakly or not at all to COVID-19 deaths. This provides support to our first and second hypothesis. Moreover, these findings are in line with the current body of COVID-19 literature (Al-Awadhi et al., 2020; Ashraf, 2020; Chia, 2020). Therefore, we conclude that pandemics fall in the category of real-life events that drastically impact financial markets (e.g. aviation disasters, earthquakes, terrorist attacks), where investor pessimism and fear, due to the ongoing pandemic, negatively impacts stock returns.
Panel A: Daily Growth in confirmed Cases
(1) (2) (3) (4) α 0.05 -15.74 -16.04 -34.51** (0.29) (16.04) (16.04) (16.20) DGTCC -0.02*** -0.02*** -0.02*** -0.03*** (0.01) (0.01) (0.01) (0.01) lnMC 0.72 0.73 1.56** (0.73) (0.73) (0.74) PTB -0.01 -0.01 (0.01) (0.01) TED 1.26** (0.52) Observations 11,988 11,988 11,988 11,988 Wald Χ2 8.25*** 9.13** 10.14** 17.89*** R-squared 0.02 0.03 0.03 0.04
Panel B: Daily Growth in Confirmed Deaths
18
5.3 Firms size analysis
In the following section, we analyse whether the impact of COVID-19 differs when a distinction is made between firm sizes. Table 8 reports the results for the estimated equation with index-specific dummy variables. Column (1) presents the results when the AEX is used as a reference variable and column (2) presents the results with the AMX as reference variable. The table illustrates that small firms performed the best in the COVID-19 period, followed by mid-cap firms and large firms. We observe that firms on the ASCX significantly outperformed the AEX and AMX by 8.34% and 5.25%, respectively. Furthermore, the AMX exceeded the returns of the AEX by 6.42%.
Table 8
Regression with index-specific dummy variables
(1) (2) α -43.15** -36.73** (17.07) (17.07) lnMC 1.66** 1.66** (0.78) (0.78) PTB -0.01 -0.01 (0.01) (0.01) TED 0.80 0.80 (0.53) (0.53) AEX -1.13** (0.50) AMX 6.42** (2.87) ASCX 8.34** 5.25** (3.73) (2.59) Observations 12,062 12,062 Wald Χ2 7.86 7.59 R-squared 0.01 0.01
Note: This table reports the results of the panel data regression using a sample of the 75 largest firms listed on the Euronext Amsterdam over the period of 1 January 2020 until 21 August 2020 considering specific indices. Firms are placed in each index based on market capitalization, where there are 25 constituent firms on each index. The largest 25 firms are placed on the AEX, firms 26-50 are placed on the AMX and firms 51-75 are placed on the ASCX. AEX denotes the Amsterdam Exchange Index, AMX denotes Amsterdam Midcap Index and ASCX stands for Amsterdam Small Cap Index. Column 1 reports the results when the AEX is the reference variable and column 2 when the AMX is the reference variable. The independent variable of the regression is Ri.t, which denotes the return of stock i on day t. α is the intercept, lnMC is the natural logarithm of daily firm
market capitalization, PTB is the price-to-book ratio and TED denotes the TED-spread retrieved from the Federal Reserve Bank of St. Louis. The Wald Χ2-test is conducted to check for joint significance of the variables. Panel-corrected standard
errors are in parentheses; *,**,*** denote statistical significance at the 10%, 5% and 1% level, respectively.
19 explanation resides in the paper of Brown and Cliff (2005). The authors state that investor sentiment can sometimes be less pronounced for smaller firms due to the fact that the used sentiment variable applies to the market as a whole, whereas investor sentiment affects smaller firms more when it is more specific to those firms. This also seems a likely explanation for the impact of COVID-19. News of COVID-19 is important for every firm on the market, it could be possible that the effects of investor sentiment are felt more by the large firms due to the simple fact that they are traded more frequently and receive more attention of investors than smaller firms.
5.4 Sector analysis
In this section, we report whether the impact of COVID-19 differs between sectors. Table 9 reports the estimated regression with sector-specific dummy variables. Dummies for the following sectors were added to the regression: Basic Materials, Capital Goods, Consumer Cyclical, Consumer/Non-Cyclical, Energy, Financial, Healthcare, Services, Technology and Transportation. As can be seen from the table, the Basic Materials, Capital Goods, Consumer-Non/Cyclical and Technology sectors significantly outperformed the market. In particular the Technology and Capital Goods sectors performed significantly better than the market with coefficients of 6.00 and 5.32, respectively. On the contrary, the Energy, Financial, Healthcare and Services sectors performed significantly worse than the market. The energy sector performed notably poorly with a coefficient of -6.42. Finally, the coefficients of Consumer Cyclical and Transportation are insignificant.
From this we conclude that the impact of COVID-19 was felt different for the studied sectors in the sample. This is in agreement with our fourth hypothesis. However, it is not the case that certain sectors performed well and others performed weakly. As can be seen in the descriptive statistics presented in Table 4, every sector in the sample, excluding Technology, had a negative average return over the COVID-19 period.
20
Table 9
Regression with sector-specific dummy variables
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) α -39.67** -36.73** -36.73** -36.73** -36.73** -36.73** -36.73** -36.73** -36.73** -36.73** (18.79) (17.07) (17.07) (17.07) (17.07) (17.07) (17.07) (17.07) (17.07) (17.07) lnMC 1.66** 1.66** 1.66** 1.66** 1.66** 1.66** 1.66** 1.66** 1.66** 1.66** (0.78) (0.78) (0.78) (0.78) (0.78) (0.78) (0.78) (0.78) (0.78) (0.78) PTB -0.01 -0.01 -0.01 -0.01 -0.01 -0.01 -0.01 -0.01 -0.01 -0.01 (0.01) (0.01) (0.01) (0.01) (0.01) (0.01) (0.01) (0.01) (0.01) (0.01) TED 0.80 0.80 0.80 0.80 0.80 0.80 0.80 0.80 0.80 0.80 (0.53) (0.53) (0.53) (0.53) (0.53) (0.53) (0.53) (0.53) (0.53) (0.53) Basic Materials 2.94* (1.77) Capital Goods 5.32** (2.46) Consumer 0.33 Cyclical (0.28) Consumer/Non- 5.25** Cyclical (2.59) Energy -6.42** (2.87) Financial -3.64** (1.64) Healthcare -3.87** (1.90) Services -4.04** (1.97) Technology 6.00** (2.81) Transportation -0.93 (0.59) Observations 12,062 12,062 12,062 12,062 12,062 12,062 12,062 12,062 12,062 12,062 Wald Χ2 9.02** 5.53 5.65 6.44* 6.55* 5.88 5.75 5.77 5.41 5.56 R-squared 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01
Note: This table reports the results of the panel data regression using a sample of the 75 largest firms listed on the Euronext Amsterdam over the period of 1 January 2020 until 21 August 2020 considering specific sectors. The independent variable of the regression is Ri.t, which denotes the return of stock i on day t. α is the intercept, lnMC is the natural logarithm of daily firm
market capitalization, PTB is the price-to-book ratio and TED denotes the TED-spread retrieved from the Federal Reserve Bank of St. Louis. Basic Materials, Capital Goods, Consumer Cyclical, Consumer/Non-cyclical, Energy, Financial, Healthcare, Services, Technology and Transportation are dummy variables that take on the value one if a firm belongs to that respective sector and zero otherwise. The Wald Χ2-test is conducted to check for joint significance of the variables.
Panel-corrected standard errors are in parentheses; *,**,*** denote statistical significance at the 10%, 5% and 1% level, respectively.
21 decrease in cash flows, causing asset prices to drop. Fourthly, as can be seen in the table, COVID-19 had the largest negative impact on the energy sector. One of the main reasons is the drop of the oil prices at the beginning of 2020. COVID-19 led to great macro-economic uncertainties regarding the demand and the price of oil.14 Lastly, one sector has benefitted from
COVID-19. The Technology sector significantly outperformed the market and was able to generate positive average returns over the sample period. The COVID-19 crisis provided many opportunities for tech-companies. People were more at home due to the lockdown. The need for technical services to work from home and communicate with people from a distance, provided opportunities for tech-companies to innovate and to deliver new services.
5.5 Impact of press conferences
In this section, we will discuss whether the national press conferences, hosted by the Dutch prime minister, had an impact on stock returns in the COVID-19 period. We discuss the results for all press conferences as well as when the distinction is made between good and bad press conferences. Table 10 reports the estimated regression including dummy variables of the press conferences. The dummy variable PC takes on the value one if there was a COVID-19 related press conference on that day and zero otherwise. We have chosen for the day of the press conference instead of a day after the press conference because the contents of the conference were published by the press earlier on the day, giving investors enough time to react before markets close. Column (1) reports the regression with a dummy variable included for all press conferences and columns (2) and (3) report the regression when the distinction is made between press conferences that present good news (GPC) and bad news (BPC). Good press conferences relate to a decrease of COVID-19 or a relaxation of COVID-19 related measures. On the other hand, a bad press conference discusses the increase of COVID-19 or stricter measures. The distinction between good and bad press conferences is made in order to observe whether there was an asymmetric reaction to news, which is found in the literature (e.g. Soroka, 2006; Akhtar, 2013). Previous studies reported that investors react strongly to bad news and weakly, or not at all, to good news.
As can be seen from Table 10, the overall effect of the COVID-19 related press conferences is significantly negative. On days where there was a press conference, stocks performed 1.70% worse compared to days when there was no press conference. This effect is even stronger when the distinction is made between good and bad press conferences. Column (3) shows that the coefficient of bad press conferences is significantly negative with a coefficient of -2.52 while the coefficient of good press conferences is insignificant. Taking these results into account, we can conclude that there is an asymmetrical reaction to the press conferences. Investors react strongly negative to press conferences that bring bad news and weakly to positive press
22 conferences. This confirms that the asymmetric reaction to news was present in the COVID-19 period, which provides evidence for our fifth hypothesis.
Table 10
Regression with press conference dummy variables
(1) (2) (3) α -31.13* -37.18** -31.33* (17.04) (16.95) (16.62) lnMC 1.41* 1.68** 1.41* (0.78) (0.77) (0.76) PTB -0.01 -0.01 -0.00 (0.01) (0.01) (0.01) TED 0.99* 0.82 1.21** (0.53) (0.53) (0.53) PC -1.70*** (0.56) GPC 0.43 (1.02) BPC -2.52*** (0.66) Observations 12,062 12,062 12,062 Wald Χ2 14.55*** 6.49 19.80*** R-squared 0.03 0.01 0.05
Note: This table reports the results of the panel data regression using a sample of the 75 largest firms listed on the Euronext Amsterdam over the period of 1 January 2020 until 21 August 2020 while considering national press conferences. The independent variable of the regression is Ri.t, which denotes the return of stock i on day t. α is the intercept, lnMC is the natural
logarithm of daily firm market capitalization, PTB is the price-to-book ratio and TED denotes the TED-spread retrieved from the Federal Reserve Bank of St. Louis. PC is dummy variable that takes on the value one if there is a press conference on that day and zero otherwise. GPC is a dummy variable that takes on the value one if there is a press conference that presents good news regarding COVID-19 on that day and zero otherwise. BPC is a dummy variable that takes on the value one if there is a press conference that presents bad news regarding COVID-19 on that day and zero otherwise. The Wald Χ2-test is conducted
to check for joint significance of the variables. Panel-corrected standard errors are in parentheses; *,**,*** denote statistical significance at the 10%, 5% and 1% level, respectively.
5.6 Subsample analysis
23
Table 11
Regression of the first and second COVID-19 wave
Note: This table reports the results of the panel data regression estimated with Equation (2) using a sample of the 75 largest firms listed on the Euronext Amsterdam, split in the first and second COVID-19 wave. The first wave runs from 1 January 2020 until 24 June 2020 and the second wave runs from 25 June 2020 until 21 August 2020. Panel A reports the results when the daily growth in confirmed COVID-19 cases (DGTCC) is used as independent variable. Panel B reports the results when the daily growth in confirmed COVID-19 deaths (DGTDC is used as independent variable. The independent variable of the regression is Ri.t, which denotes the returns of stock i on day t. α is the intercept, lnMC is the natural logarithm of daily firm
market capitalization, PTB is the price-to-book ratio and TED denotes the TED-spread retrieved from the Federal Reserve Bank of St. Louis. The Wald Χ2-test is conducted to check for joint significance of the variables. Panel-corrected standard
errors are in parentheses; *,**,*** denote statistical significance at the 10%, 5% and 1% level, respectively.
to the second wave. This is the case for both new cases and new COVID-19 related deaths. There are a number of possible reasons for why investors react less severe to the second wave relative to the first wave. Firstly, the absolute COVID-19 numbers are higher in the first wave. The maximum number of new cases in the first wave is 1224, compared to 779 in the second wave. Secondly, investors might have grown used to the COVID-19 situation. Negative reactions in the first wave can be caused by general panic and fear. Investors might have already anticipated the second wave, causing them to react less strongly.
Panel A: Daily Growth in Confirmed Cases
First Wave Second Wave
α -49.28** -174.00*** (24.06) (28.13) DGTCC -0.03*** 0.07 (0.01) (0.27) lnMC 2.22** 8.01*** (1.09) (1.28) PTB 0.01 -0.02*** (0.01) (0.01) TED 1.64** -10.35 (0.65) (9.26) Observations 8,880 3,033 Wald Χ2 14.91*** 41.66*** R-squared 0.05 0.07
Panel B: Daily Growth in Confirmed Deaths
First Wave Second Wave
24
6. Conclusions
Diseases and pandemics affect financial markets. Previous literature has found that SARS (Chen et al., 2009; Chen et al., 2007), Ebola (Ichev and Marinč, 2018) and disease-related news (Donadelli et al., 2017) significantly impact stock returns. In this study, we analysed whether the novel COVID-19 pandemic of 2020 had a significant impact on Dutch stock returns.
Using panel data analysis on a sample of the 75 largest firms listed on the Euronext Amsterdam, over the period of 1 January 2020 until 21 August 2020, we obtained a number of results. Firstly, we found that COVID-19 had a significant negative impact on stock returns. Secondly, the growth in new confirmed COVID-19 cases had a significant negative effect, whereas the growth in new confirmed COVID-19 related deaths had no impact. Thirdly, we showed that the impact of the pandemic was more negatively pronounced for the largest firms in the sample, while less pronounced for the smallest firms. Fourthly, we found evidence that the effect of COVID-19 was different for various sectors in the sample. We reported that the Energy, Financial, Healthcare and Services sectors significantly underperformed compared to the market. The Technology sector was the only sector in the sample that benefitted from the virus. Fifthly, we showed that the national press conferences, regarding the pandemic, had a significant negative effect on stock returns. More precisely, when categorized in good and bad press conferences, we found evidence of an asymmetrical reaction. Bad press conferences had a significant negative impact and good press conferences had no significant effect. Finally, as a robustness check, we reported evidence that the first wave of COVID-19, lasting until 24 June, had a negative impact, whereas the second wave had no significant relation to stock returns.
25
References
Abiad, A., Arao, R. M., & Dagli, S. (2020). The economic impact of the COVID-19 outbreak on developing Asia. Asian Development Bank.
Aiuppa, T. A., Carney, R. J., & Krueger, T. M. (1993). An examination of insurance stock prices following the 1989 Loma Prieta Earthquake. Journal of Insurance Issues, 1-14.
Akhtar, S., Faff, R., Oliver, B., & Subrahmanyam, A. (2013). Reprint of: Stock salience and the asymmetric market effect of consumer sentiment news. Journal of Banking & Finance, 37(11), 4488-4500.
Al-Awadhi, A. M., Al-Saifi, K., Al-Awadhi, A., & Alhamadi, S. (2020). Death and contagious infectious diseases: Impact of the COVID-19 virus on stock market returns. Journal of Behavioral and Experimental Finance, 100326.
Albulescu, C. T. (2020). COVID-19 and the United States financial markets’ volatility. Finance Research Letters, 101699.
Ali, M., Alam, N., & Rizvi, S. A. R. (2020). Coronavirus (COVID-19)–An epidemic or pandemic for financial markets. Journal of Behavioral and Experimental Finance, 100341.
Alt, R., Fortin, I., & Weinberger, S. (2011). The Monday effect revisited: An alternative testing approach. Journal of Empirical Finance, 18(3), 447-460.
Andersen, A. L., Hansen, E. T., Johannesen, N., & Sheridan, A. (2020). Consumer responses to the COVID-19 crisis: Evidence from bank account transaction data. Working Paper. Available at SSRN 3609814.
Ariel, R. A. (1990). High stock returns before holidays: Existence and evidence on possible causes. The Journal of Finance, 45(5), 1611-1626.
Arin, K. P., Ciferri, D., & Spagnolo, N. (2008). The price of terror: The effects of terrorism on stock market returns and volatility. Economics Letters, 101(3), 164-167.
Ashraf, B. N. (2020). Stock markets’ reaction to COVID-19: cases or fatalities?. Research in International Business and Finance, 101249.
Baig, A., Butt, H. A., Haroon, O., & Rizvi, S. A. R. (2020). Deaths, Panic, Lockdowns and US Equity Markets: The Case of COVID-19 Pandemic. Working Paper. Available at SSRN 3584947.
Baker, M., & Wurgler, J. (2006). Investor sentiment and the cross‐section of stock returns. The Journal of Finance, 61(4), 1645-1680.
Baker, M., & Wurgler, J. (2007). Investor sentiment in the stock market. Journal of Economic Perspectives, 21(2), 129-152.
Baker, S. R., Bloom, N., Davis, S. J., Kost, K. J., Sammon, M. C., & Viratyosin, T. (2020). The unprecedented stock market impact of COVID-19 (No. w26945). National Bureau of Economic Research.
26 Beck, N., & Katz, J. N. (1995). What to do (and not to do) with time-series cross-section data. American Political Science Review, 634-647.
Blendon, R. J., Benson, J. M., DesRoches, C. M., Raleigh, E., & Taylor-Clark, K. (2004). The public's response to severe acute respiratory syndrome in Toronto and the United States. Clinical Infectious Diseases, 38(7), 925-931.
Bollen, J., Mao, H., & Zeng, X. (2011). Twitter mood predicts the stock market. Journal of Computational Science, 2(1), 1-8.
Barua, S. (2020). Understanding Coronanomics: The economic implications of the coronavirus (COVID-19) pandemic. Working paper. Available at SSRN 3566477
Boone, L., Haugh, D., Pain, N., & Salins, V. (2020). Tackling the fallout from COVID-19. Economics in the Time of COVID-19, 37.
Boudt, K., Paulus, E. C., & Rosenthal, D. W. (2017). Funding liquidity, market liquidity and TED spread: A two-regime model. Journal of Empirical Finance, 43, 143-158.
Boyson, N. M., Stahel, C. W., & Stulz, R. M. (2010). Hedge fund contagion and liquidity shocks. The Journal of Finance, 65(5), 1789-1816.
Brown, G. W., & Cliff, M. T. (2005). Investor sentiment and asset valuation. The Journal of Business, 78(2), 405-440.
Brown, S. J., & Warner, J. B. (1985). Using daily stock returns: The case of event studies. Journal of Financial Economics, 14(1), 3-31.
Capelle-Blancard, G., & Laguna, M. A. (2010). How does the stock market respond to chemical disasters?. Journal of Environmental Economics and Management, 59(2), 192-205.
Chance, D. M., & Ferris, S. P. (1987). The effect of aviation disasters on the air transport industry: a financial market perspective. Journal of Transport Economics and Policy, 151-165.
Chen, C. D., Chen, C. C., Tang, W. W., & Huang, B. Y. (2009). The positive and negative impacts of the SARS outbreak: A case of the Taiwan industries. The Journal of Developing Areas, 281-293.
Chen, M. H., Jang, S. S., & Kim, W. G. (2007). The impact of the SARS outbreak on Taiwanese hotel stock performance: an event-study approach. International Journal of Hospitality Management, 26(1), 200-212.
Chen, C., Liu, L., & Zhao, N. (2020). Fear sentiment, uncertainty, and bitcoin price dynamics: The case of COVID-19. Emerging Markets Finance and Trade, 56(10), 2298-2309.
Chia, R. C. J., Liew, V. K. S., & Rowland, R. (2020). Daily new Covid-19 cases, the Movement Control Order, and Malaysian stock market returns. International Journal of Business and Society, 21(2), 553-568.
27 Coibion, O., Gorodnichenko, Y., & Weber, M. (2020). The cost of the covid-19 crisis: Lockdowns, macroeconomic expectations, and consumer spending (No. w27141). National Bureau of Economic Research.
Conlon, T., & McGee, R. (2020). Safe haven or risky hazard? Bitcoin during the COVID-19 bear market. Finance Research Letters, 101607.
Corbet, S., Larkin, C., & Lucey, B. (2020). The contagion effects of the covid-19 pandemic: Evidence from gold and cryptocurrencies. Finance Research Letters, 101554.
Cross, F. (1973). The behavior of stock prices on Fridays and Mondays. Financial Analysts Journal, 29(6), 67-69.
Davis, S. J., Hansen, S., & Seminario-Amez, C. (2020). Firm-Level Risk Exposures and Stock Returns in the Wake of COVID-19 (No. w27867). National Bureau of Economic Research.
Depoux, A., Martin, S., Karafillakis, E., Preet, R., Wilder-Smith, A., & Larson, H. The pandemic of social media panic travels faster than the COVID-19 outbreak. Journal of Travel Medicine, 27(3), taaa031.
Devpura, N., & Narayan, P. K. (2020). Hourly oil price volatility: The role of COVID-19. Energy Research Letters, 1(2), 13683.
Donadelli, M., Kizys, R., & Riedel, M. (2017). Dangerous infectious diseases: Bad news for Main Street, good news for Wall Street?. Journal of Financial Markets, 35, 84-103.
Dong, E., Du, H., & Gardner, L. (2020). An interactive web-based dashboard to track COVID-19 in real time. The Lancet Infectious Diseases, 20(5), 533-534.
Drakos, K. (2010). Terrorism activity, investor sentiment, and stock returns. Review of Financial Economics, 19(3), 128-135.
Dubois, M., & Louvet, P. (1996). The day-of-the-week effect: The international evidence. Journal of Banking & Finance, 20(9), 1463-1484.
Edmans, A., Garcia, D., & Norli, Ø. (2007). Sports sentiment and stock returns. The Journal of Finance, 62(4), 1967-1998.
Ewing, B. T., Hein, S. E., & Kruse, J. B. (2006). Insurer stock price responses to Hurricane Floyd: An event study analysis using storm characteristics. Weather and Forecasting, 21(3), 395-407.
Ferstl, R., Utz, S., & Wimmer, M. (2012). The effect of the Japan 2011 disaster on nuclear and alternative energy stocks worldwide: An event study. Business Research, 5(1), 25-41.
French, K. R. (1980). Stock returns and the weekend effect. Journal of Financial Economics, 8(1), 55-69.
28 Gao, J., Zheng, P., Jia, Y., Chen, H., Mao, Y., Chen, S., ... & Dai, J. (2020). Mental health problems and social media exposure during COVID-19 outbreak. Plos One, 15(4), e0231924.
Giner, B., & Rees, W. (2001). On the asymmetric recognition of good and bad news in France, Germany and the United Kingdom. Journal of Business Finance & Accounting, 28(9‐10), 1285-1331.
Greene, W. H. (2003). Econometric analysis. Pearson Education India.
Hartzmark, S. M., & Sussman, A. B. (2019). Do investors value sustainability? A natural experiment examining ranking and fund flows. The Journal of Finance, 74(6), 2789-2837.
Haug, M., & Hirschey, M. (2006). The january effect. Financial Analysts Journal, 62(5), 78-88.
Hausman, J. A. (1978). Specification Tests in Econometrics. Econometrica, 46, 1251–1271.
Hirshleifer, D., & Shumway, T. (2003). Good day sunshine: Stock returns and the weather. The Journal of Finance, 58(3), 1009-1032.
Hsiao, C. (2014). Analysis of panel data (No. 54). Cambridge University Press.
Ichev, R., & Marinč, M. (2018). Stock prices and geographic proximity of information: Evidence from the Ebola outbreak. International Review of Financial Analysis, 56, 153-166.
IMF. (2020). World Economic Outlook Update, June 2020.
Iyke, B. N. (2020). COVID-19: The reaction of US oil and gas producers to the pandemic. Energy Research Letters, 1(2), 13912.
John, M. L. (2001). A dictionary of epidemiology. Oxford University Press.
Kamstra, M. J., Kramer, L. A., & Levi, M. D. (2003). Winter blues: A SAD stock market cycle. American Economic Review, 93(1), 324-343.
Kaplanski, G., & Levy, H. (2010A). Exploitable predictable irrationality: The FIFA World Cup effect on the US stock market. Journal of Financial and Quantitative Analysis, 45, 535-553.
Kaplanski, G., & Levy, H. (2010B). Sentiment and stock prices: The case of aviation disasters. Journal of Financial Economics, 95(2), 174-201.
Kaplanski, G., & Levy, H. (2014). Sentiment, irrationality and market efficiency: The case of the 2010 FIFA World Cup. Journal of Behavioral and Experimental Economics, 49, 35-43.
Keim, D. B., & Stambaugh, R. F. (1984). A further investigation of the weekend effect in stock returns. The Journal of Finance, 39(3), 819-835.
Khan, K., Zhao, H., Zhang, H., Yang, H., Shah, M. H., & Jahanger, A. (2020). The impact of COVID-19 pandemic on stock markets: An empirical analysis of world major stock indices. The Journal of Asian Finance, Economics, and Business, 7(7), 463-474.
29 Klibanoff, P., Lamont, O., & Wizman, T. A. (1998). Investor reaction to salient news in closed‐ end country funds. The Journal of Finance, 53(2), 673-699.
Lakonishok, J., & Smidt, S. (1988). Are seasonal anomalies real? A ninety-year perspective. The Review of Financial Studies, 1(4), 403-425.
Lee, H. Y., Wu, H. C., & Wang, Y. J. (2007). Contagion effect in financial markets after the South-East Asia Tsunami. Research in International Business and Finance, 21(2), 281-296.
Liu, H., Manzoor, A., Wang, C., Zhang, L., & Manzoor, Z. (2020). The COVID-19 outbreak and affected countries stock markets response. International Journal of Environmental Research and Public Health, 17(8), 2800.
Lu, H., Stratton, C. W., & Tang, Y. W. (2020). Outbreak of pneumonia of unknown etiology in Wuhan, China: The mystery and the miracle. Journal of Medical Virology, 92(4), 401-402.
Luo, X., Zhang, J., & Duan, W. (2013). Social media and firm equity value. Information Systems Research, 24(1), 146-163.
Mamaysky, H. (2020). Financial Markets and News about the Coronavirus. Working Paper. Available at SSRN 3565597.
Marrett, G. J., & Worthington, A. C. (2009). An empirical note on the holiday effect in the Australian stock market, 1996–2006. Applied Economics Letters, 16(17), 1769-1772.
Mazur, M., Dang, M., & Vega, M. (2020). COVID-19 and the march 2020 stock market crash. Evidence from S&P1500. Finance Research Letters, 101690.
Mensi, W., Sensoy, A., Vo, X. V., & Kang, S. H. (2020). Impact of COVID-19 outbreak on asymmetric multifractality of gold and oil prices. Resources Policy, 69, 101829.
Mnif, E., Jarboui, A., & Mouakhar, K. (2020). How the cryptocurrency market has performed during COVID 19? A multifractal analysis. Finance Research Letters, 36, 101647.
Olson, D., Mossman, C., & Chou, N. T. (2015). The evolution of the weekend effect in US markets. The Quarterly Review of Economics and Finance, 58, 56-63.
Prabheesh, K. P., Padhan, R., & Garg, B. (2020). COVID-19 and the oil price–stock market nexus: Evidence from net oil-importing countries. Energy Research Letters, 1(2), 13745.
RIVM. (2020). Patient with novel coronavirus COVID-19 in the Netherlands.
Salisu, A. A., & Akanni, L. O. (2020). Constructing a global fear index for the COVID-19 pandemic. Emerging Markets Finance and Trade, 56(10), 2310-2331.
Salisu, A. A., & Vo, X. V. (2020). Predicting stock returns in the presence of COVID-19 pandemic: The role of health news. International Review of Financial Analysis, 71, 101546.
30 Shan, L., & Gong, S. X. (2012). Investor sentiment and stock returns: Wenchuan Earthquake. Finance Research Letters, 9(1), 36-47.
Shelor, R., Anderson, D., & Cross, M. (1990). The impact of the California earthquake on real estate firms' stock value. Journal of Real Estate Research, 5(3), 335-340.
Sohrabi, C., Alsafi, Z., O’Neill, N., Khan, M., Kerwan, A., Al-Jabir, A., Iosifidis, C., & Agha, R. (2020). World Health Organization declares global emergency: A review of the 2019 novel coronavirus (COVID-19). International Journal of Surgery, 76, 71-76.
Soroka, S. N. (2006). Good news and bad news: Asymmetric responses to economic information. The Journal of Politics, 68(2), 372-385.
Steeley, J. M. (2001). A note on information seasonality and the disappearance of the weekend effect in the UK stock market. Journal of Banking & Finance, 25(10), 1941-1956.
Thaler, R. H. (1987). Anomalies: the January effect. Journal of Economic Perspectives, 1(1), 197-201.
Wang, L., & Kutan, A. M. (2013). The impact of natural disasters on stock markets: Evidence from Japan and the US. Comparative Economic Studies, 55(4), 672-686.
Wang, W., Tang, J., & Wei, F. (2020). Updated understanding of the outbreak of 2019 novel coronavirus (2019‐nCoV) in Wuhan, China. Journal of Medical Virology, 92(4), 441-447. WEF. (2020). The Global Risks Report 2020.
Weiderman, I., & Bacon, F. (2008). Hurricane Katrina's effect on oil companies stock prices. Academy of Strategic Management Journal, 7, 11-16.
WHO. (2020A). Statement on the second meeting of the International Health Regulations (2005) Emergency Committee regarding the outbreak of novel coronavirus (2019-nCoV).
WHO. (2020B). Director-General’s opening remarks at the media briefing on COVID-19 – 11 March 2020.
WHO. (2020C). Coronavirus disease (COVID-19): weekly epidemiological, update 1.
WHO. (2020D). Pneumonia of unknown cause – China.
Wooldridge, J. M. (2002). Econometric analysis of cross section and panel data MIT Press. Cambridge, MA, 108.
Wooldridge, J. M. (2010). Econometric analysis of cross section and panel data. MIT Press.
Yang, Y., Peng, F., Wang, R., Guan, K., Jiang, T., Xu, G., ... & Chang, C. (2020). The deadly coronaviruses: The 2003 SARS pandemic and the 2020 novel coronavirus epidemic in China. Journal of Autoimmunity, 102434.
