Effects of COVID-19 Pandemic on Investor Attention in Global Stock Markets: An Empirical Assessment

Abstract

Using the data from 50 global stock markets, this article examines the impact of the COVID-19 pandemic on investor attention. Using Google search volume as a proxy for country-specific and worldwide investor attention, this article provides initial evidence on the relationship between the spread of the pandemic and investor attention. Our results suggest that the increasing number of daily deaths and confirmed cases significantly negatively impact investor attention. Results are robust to the country characteristics like the nature of the financial system, financial stability and investor attention risk pricing nature. Our results indicate a positive and significant impact of government responses for restoring investor attention in the market.

JEL Codes: G10, G14, G15, C23

Keywords

Investor attention COVID-19 stock returns Coronavirus pandemic

I. Introduction

The ongoing COVID-19 pandemic has resulted in sharp volatility and frequent crashes in the global equity indices (Baker et al., 2020a, b; Baldwin & di Mauro, 2020; Schell et al., 2020). The number of COVID-19 confirmed cases had surpassed 79 million worldwide, with the death toll exceeding 1,738,168 during May 2020. The number of confirmed cases and deaths is escalating every day. According to recent estimates from the International Monetary Fund (IMF), most countries’ economies will be at least 5% smaller after recovery. Since mid-January 2020, the global economy has plunged into a severe contraction due to the massive shock of the coronavirus pandemic and associated socio-economic shutdown measures to contain it. Consistent with this view, IMF chief economist cogently mentions that ‘this is a deep recession. It is a recession that involves solvency issues, and unemployment going up substantially and these leave scars’. No previous infectious disease outbreak has wedged the stock market as powerfully as the COVID-19 pandemic (Baker et al., 2020a, b; McKibbin & Fernando, 2020).

Motivated by this series of destructive evidence amid the early COVID-19 spread, this article aims to shed light on the coronavirus pandemic that has impacted investor attention on the global equity market. The growing body of recent literature focuses on the COVID-19 impact on stock return behaviour (Al-Awadhi et al., 2020; Smales, 2020, 2021), alternative asset selection for safe haven properties (Corbet et al., 2020; Ji et al., 2020), information flow (Haroon & Rizvi, 2020; Schell et al., 2020) and market volatility (Ali et al., 2020; Zaremba et al., 2020). However, the impact of the COVID-19 pandemic on investor attention has been largely ignored. Therefore, it becomes critical to comprehend how a global pandemic event like COVID-19 could have influenced the equity market investor’s attention worldwide. As the COVID-19 crisis’ damage continues to unfold, this article hypothesises that coronavirus (COVID-19) influences country-specific and worldwide investor attention. Our article using daily data from 50 international stock markets during the early spread of the pandemic (January–May 2020), examines the following two research questions. First, the effects of COVID-19 pandemic on investor attention in global stock markets. Second, the possible moderating role of country-specific characteristics like financial market development and financial stability on the COVID-19 and investor attention relationship.

Several recent studies suggest that investor attention may play an essential role in determining asset prices (Andrei & Hasler, 2015; Chen, 2017; Da et al., 2011; Smales, 2020, 2021). Since cognitive attention is scarce (Kahneman, 1973), investors succumb to limited attention bias for information processing in their investment decisions (Peng & Xiong, 2006). To resolve the uncertainty over identifying precise information among the available large amount of data, investors may choose to pay more attention to specific information and neglect others while updating their economic expectations. Moreover, greater investor attention can exacerbate the effect of investor behavioural biases (such as sentiment) on asset prices (Smales, 2021). In empirical research, the early study by Da et al. (2011) proposes a new and direct measure of retail investor attention using search frequency in Google. A large and growing body of research (Andrei & Hasler, 2015; Chen, 2017; Da et al., 2011; Smales, 2020, 2021) suggests a significant negative effect of investor attention on the expected returns. The economic rationale for the observed negative impact of investor attention on stock return is supported by the limited attention bias (Kahneman, 1973; Peng & Xiong, 2006) and price pressure hypothesis. That is, investors create price pressure on attention-grabbing financial assets due to limited attention bias (Kahneman, 1973) and their inability to devote time and resources to analyse the entire universe of feasible investments (Smales, 2020; Peng & Xiong, 2006). Attention-induced price pressure results in overvalued financial assets in the market. Hence, a higher (positive) internet search attention should predict higher stock prices in the short term and price reversals in the long run (Chen, 2017; Da et al., 2011). In our study, we hypothesise three possible channels through which pandemic information can influence investor attention. For simplicity of their discussion, we categorise them as policy uncertainty channel, information channel and flight-to-quality or flight-to-liquidity channel.

First, the policy uncertainty channel could have influenced the investor’s attention due to the high-end economic uncertainty. The global COVID-19 pandemic spread lead to unprecedented stringent policy responses like social distancing, self-isolation, travel restrictions, declarations of state of emergency and localised or nationwide lockdown restrictions (Baldwin & di Mauro, 2020; Dasgupta et al., 2021; Goel & Dash, 2021; Narayan et al., 2021a; Schell et al., 2020). Given this crisis’ nature, it is a combination of multiple economic setbacks such as demand shock, supply shock and financial shock (Ali et al., 2020; Baker et al., 2020a, b; Ji et al., 2020; Schell et al., 2020). Mishkin (1999) conjectures that uncertainty may result from a recession, policies instability or collapse of financial institutions, and hence can lead to information asymmetry and financial instability. Furthermore, greater information uncertainty leads to lower expected returns following bad news (Zhang, 2006), and the fall in expected return is more severe for high uncertainty which coincides with a pessimistic view of the macro economy (Segal et al., 2015). Several other studies document the adverse impact of increasing macro-economic uncertainty on asset prices (Mishkin, 1999; Phan et al., 2020). Given the unprecedented nature of the COVID-19 pandemic, there is enormous uncertainty as to how severe the decline in global output will be and the magnitude of the deterrent effect on economic expectations of market participants (Baldwin & di Mauro, 2020; OECD, 2020; World Bank, 2020). For instance, Altig et al. (2020), Baker et al. (2020a, b), and Narayan et al. in 2020 reveal that economic policy uncertainty indicators reached their highest point as recorded during the COVID-19 pandemic. It is therefore reasonable to argue that the recent pandemic with heightened economic uncertainty will have a psychological influence on investor behaviour and thus on investor attention. For simplicity, we call this the economic and policy uncertainty channel.

Second, the flow of information and information demand could also affect investor attention. For instance, Garcia (2013) support the conjecture that media content can indeed predict asset prices and their trading volume. In a recent study, Haroon and Rizvi (2020) highlight that recent evidence of increasing volatility in the equity markets is the outcome of overwhelming panic generated by coronavirus-related news. Zhang (2006) documents that more significant information uncertainty following bad news can play a critical role as it leads to lower expected returns. The social amplification of risk framework features a crucial role of news media in risk communication by interacting with other agents in intensifying risk (Young et al., 2013). Donadelli et al. (2017) observe that low-probability, high-consequence events, such as potential outbreaks of rare diseases and associated health risks, are overemphasised in media-generated news waves. On the other hand, the socio-economic hypothesis conjectures social mood (i.e., the collective mood of individuals), as an essential causal variable in financial and social trends (Olson, 2006). Considering the social mood hypothesis, and contextualising it with the impact of information flow, it is expected that the recent COVID-19 pandemic events could translate significant shock wave to affect investor attention. We consider this effect as an information channel. The third possible channel which may influence investor attention during pandemic events such as the spread of the novel coronavirus is investor behaviour for asset selection popularly known as flight-to-quality or flight-to-liquidity phenomenon. In highly uncertain environments, investors care about the liquidity preference and shy away from assets characterised by high volatility like stocks (Ji et al., 2020). This supports the conjecture of Huang and Liu (2007). During the recent pandemic outbreak, the prevailing pessimism among investors due to higher volatility in the stock markets can be described as flight-to-liquidity. Based on the earlier arguments, we conjecture a negative and significant effect of COVID-19 pandemic information such as counts of daily deaths and daily confirmed cases on investor attention.

To analyse the role of daily death cases and daily confirmed cases of COVID-19 (pandemic information) on investor stock market attention, we use daily Google search volume information (GSVI) as a proxy for investor attention for 50 countries during the most recent period of COVID-19 virus spread: January–May 2020. In our empirical analysis, we incorporate investor attention in terms of country-specific attention and worldwide attention. Using panel regressions, we explore the potential role of pandemic information on investor attention. Furthermore, we control government policy response variables encompassing a wide range of economic and health measures. We find a strong negative impact of the coronavirus pandemic on investor attention around the world. Our robustness tests extend the discussion by making a sub-sample analysis of countries based on their financial market development, financial stability, and risk pricing of investor attention. Our findings complement the related strand of literature, which documents a negative and significant effect of COVID-19 on asset prices and financial market volatility (Al-Awadhi et al., 2020; Ji et al., 2020; Schell et al., 2020; Smales, 2020, 2021; Zaremba et al., 2020). Overall, we observe a consistently positive impact of government policy responses on investor attention irrespective of the nature of the financial market development and economic stability in the economy. The positive and significant effects of the overall government response index and containment and health index help ascertain that, effective policy response alleviates investor attention in the market.

We extend the related literature in several ways. First, using GSVI as a proxy of investor attention, we develop scant research on apprehending the impact of the COVID-19 pandemic on stock market behaviour. Admittedly our article is not the first to use the GSVI data to capture investor attention in the financial market during the COVID-19 crisis. For instance, Smales (2020, 2021), using GSVI information for the ‘coronavirus’ keyword as a proxy for pandemic information demand, find that COVID-19 pandemic attention is negatively associated with stock index returns across G7 and G20 countries. However, Smales’ (2020, 2021) findings do not answer what happened to the stock market-specific investor attention or information demand amid the COVID-19 pandemic. Our approach differs from Smales (2020, 2021) because the country-specific stock market GSVI measures investor attention for a particular stock market at the local and global levels. In simple words, while Smales (2020, 2021) focuses on pandemic attention, we focus on the stock market attention. To our knowledge, this is perhaps the first paper that explores the effects of a pandemic event like the most recent novel coronavirus on investor attention. Second, using a sample of 50 stock markets across geographies, we also provide detailed insights on the role of country-level financial market characteristics in moderating pandemic effects on stock market attention. Third, our article also adds to the currently scant literature on the effectiveness of government policy interventions in reducing negative economic shock due to infectious pandemics. Our article provides some initial evidence on the extent to which the government responses in terms of economic support and social restrictions around the world affect investor attention. Fourth, our findings also contribute to a large body of growing literature on the role of investor attention on asset pricing by extending further empirical evidence during a health crisis pandemic.

The rest of this article proceeds as follows. The second section describes the sample, data, and summary statistics. The third section discusses the empirical approach. The fourth section elaborates on empirical results. The fifth section concludes the article.

II. Data and Variables

Sample Selection and Control Variables

This study uses daily stock return data of 50 countries. Moreover, countries from five regions that had been tragically impacted by COVID-19, that is, Europe, Asia, South America, North America, and Africa are well represented through our sample. Selected equity markets also constitute a higher percentile rank in terms of the market breadth (market capitalisation to GDP) and market depth (turnover to GDP) of the World Bank development indicators. The sample period runs from 3 January 2020 till 18 May 2020 encompassing 94 daily observations each. Our selected sample countries are Argentina, Australia, Austria, Belgium, Brazil, Canada, Chile, China, Colombia, Denmark, Ecuador, Egypt, Finland, France, Germany, Greece, Hungary, India, Indonesia, Ireland, Israel, Italy, Japan, Kenya, Malaysia, Mexico, Netherlands, Nigeria, Norway, Peru, Philippines, Poland, Portugal, Qatar, Romania, Russia, South Korea, Saudi Arabia, Serbia, Singapore, South Africa, Spain, Sweden, Switzerland, Taiwan, Thailand, Turkey, United Arab Emirates, the UK and the USA. We use three market related control variables—daily return (RET), short term interest rates (SIR) and realised return volatility (VOL). Further details are given in Table 1.

Table 1.

List of Variables and their Description.

Variables	Definition	Source
Stock market variables
1. RET	Natural logarithmic first difference of the daily closing price.	Thomson Reuters
2. SIR	Three-month short term deposit rates.	Thomson Reuters and investing.com
I. VOL	Realised return volatility of past 20 days.	Authors’ computation
Pandemic variables
II. LDCC	Natural logarithm of daily confirmed COVID-19 cases.	ourworldindata.org
III. LDCD	Natural logarithm of daily confirmed COVID-19 deaths.	ourworldindata.org
IV. LGRI	Natural logarithm of Government response index. It shows how strong or weak has been the government response during the time of the virus outbreak.	bsg.ox.ac.uk
V. LCHI	Natural logarithm of containment and health index. It indicates lockdown restrictions with closure measures, such as school closures, ban on travelling, testing policy and contract tracing, implemented by the governments across the world.	bsg.ox.ac.uk
Investor attention variables
VI. MAT_C	Country-specific investor attention.	trends.google.com
VII. MAT_W	Worldwide investor attention.	trends.google.com
Country-characteristics variables
VIII. FMDI	Financial Market Development Index is a comprehensive measure constituted through the combination of three sub-indices encompassing the depth, access and efficiency of financial markets.	International Monetary Fund
IX. Z-score	Bank Z-score indicator captures the probability of default of a country’s commercial banking system.	The World Bank

Source: The authors.

Coronavirus (COVID-19) Pandemic Variables

Consistent with Al-Awadhi et al. (2020) and Zaremba et al. (2020) we use the natural logarithm of daily confirmed COVID-19 cases (LDCC) and daily confirmed COVID-19 deaths (LDCD) as our measures for the Coronavirus pandemic. Data for COVID-19 confirmed and death cases are collected from Our World in Data website (https://ourworldindata.org/coronavirus) COVID-19 dataset provided by Roser et al. (2020) at the University of Oxford¹. We rely on Thomson Reuters for financial data; IMF and World Bank for country-level financial market characteristics, Google Trends for constructing investor attention measure, and investing.com to collect information inaccessible on Thomson Reuters. Apart from the above-mentioned market and pandemic related control variables, we have taken data for overall government response index (LGRI) and containment and health index (LCHI) from the Oxford COVID-19 Government Response Tracker (OxCGRT) website. Variable description is provided in Table 1.

Investor Attention Measure

We collect GSVI data using Google Trend (http://www.Google.com/trends) as a proxy for investor attention. The increase (decrease) in the GSVI indicates a rise (fall) in the attractiveness of the search query rather than an absolute search number. Consistent with existing literature (Andrei & Hasler, 2014; Chen, 2017; Da et al., 2011), we use the following four steps to compute investor market attention (MAT). The logarithm of GSVI, for each search term j for day t, that is, ln(GSVI _jt ), and the daily change in search term j was defined as ∇GSVI_jt = ln(GSVI _jt ) – ln(GSVI _jt ₋₁). In this case, our search term (j) is a country specific stock index. We winsorise raw search volume for all terms at 2.5% in each tail. We extract residuals after regressing ΔGSVI _jt on week dummies. Each search term residuals are standardised to have zero mean and unit-variance. To accommodate the search requirement for stock market information within the specific country and around the world, we have constructed two measures of investor attention by restricting our search region to the specific country/geography (MAT_C) and then worldwide (MAT_W). We utilise several keyword combinations for each country to reassure that the only pertinent information is captured through the search intensity. Regarding the search terms, we either use the Bloomberg tickers for country specific benchmark indices or the index name or other equivalent abbreviation.

Tables 2 and 3 provide descriptive statistics and a correlation matrix of our variables. Table 2 reveals that the mean value of MAT_W (−0.0011) and MAT_C (−0.0034) during the sample period is consistent with the prevailing pessimism around the globe. Table 2 reports the Levin–Lin–Chu and Im–Pesaran–Shine panel unit root tests and safely rejects a unit root’s null hypothesis in our sample. In Table 3, we observe a negative correlation between market attention (MAT_C, MAT_W) and stock market return. The negative correlation between investor attention and stock return is consistent with the notion that a higher level of investor attention results in a lower stock return (Chen, 2017). We observe a negative but not statistically significant correlation between market attention (MAT_W, MAT_C) and pandemic effect (LDCD, LDCC). However, we observe a significant positive correlation between LDCD and LDCC with volatility. This is in line with the notion that early information campaigns for COVID-19 and public event cancellations are essential contributors to the growth of volatility in international stock markets (Haroon & Rizvi, 2020; Zaremba et al., 2020).

Table 2.

Descriptive Statistics.

Variables	N	Mean	Median	SD	Min	Max	Levin-Lin-Chu Unit Root Test	Im-Pesaran-Shine Unit Root Test
MAT_W	4,700	−0.001	−0.039	0.906	−4.857	5.372	−49.80***{0.00}	−62.60*** {0.00}
MAT_C	4,700	−0.003	−0.025	0.915	−5.173	6.875	−50.18***{0.00}	−60.51*** {0.00}
LDCD	4,700	4.747	4.477	4.348	0.000	11.403	−39.45***{0.00}	−52.78*** {0.00}
LDCC	4,700	2.498	0.110	3.154	0.000	14.212	−33.82***{0.00}	−45.29*** {0.00}
RET	4,700	−0.176	0.001	2.514	−16.397	13.908	−31.17***{0.00}	−55.34*** {0.00}
VOL	4,700	1.915	1.523	1.544	−3.804	8.048	−7.95*** {0.00}	−16.92** {0.02}
SIR	4,700	3.030	1.253	6.365	−1.200	60.510	−6.01*** {0.00}	−9.91*** {0.00}
LGRI	4,700	2.753	3.269	1.563	0.000	5.000	−7.48*** {0.00}	−5.52*** {0.00}
LCHI	4,700	2.788	3.436	1.604	0.000	5.000	−7.66*** {0.00}	−5.43*** {0.00}

Source: The authors.

Notes: This table reports the descriptive statistics and unit root test of all variables. Variable description is Table 1. p-Values are given in curly bracket. *** indicates the significance at 1% level.

Table 3.

Correlation Matrix.

	MAT_W	MAT_C	LDCD	LDCC	RET	VOL	SIR	LGRI	LCHI
MAT_W	1.00
MAT_C	0.28**	1.00
LDCD	−0.00	−0.02	1.00
LDCC	−0.01	−0.01	0.91**	1.00
RET	−0.14**	−0.13**	0.12**	0.09	1.00
VOL	0.04	0.01	0.39**	0.52**	0.05	1.00
SIR	0.01	0.01	−0.11**	−0.13**	0.00	0.06	1.00
LGRI	0.04	0.02	−0.16**	−0.12**	0.02	0.02	−0.05	1.00
LCHI	0.04	0.02	−0.13**	−0.18**	0.02	0.01	−0.04	0.94**	1.00

Source: The authors.

Notes: ** indicates the significance at the 5% level.

Country Characteristics

To analyse the impact of country characteristics on the relationship between the COVID-19 pandemic and investor attention, we resort to three essential aspects of the financial market. First, we consider the financial market development index (FMDI) provided by the IMF. We consider the 2017 median value of FMDI to split the sample countries into high and low financial market development categories. Second, we consider the financial stability of the countries measured by their 2017 bank Z-score (Phan et al., 2020). These characteristics are described in detail in Table 1.

Third, we use the pricing coefficients of investor attention for explaining stock return behaviour from Chen (2017, p. 363) study. We use the median value of investor attention coefficients (Chen, 2017, p. 363, model 1) to split our sample into high and low investor attention pricing countries.

III. Empirical Approach

In our study, we have used a panel data set. The use of the fixed-effect model allows the intercept to vary over the individual country, while the slope coefficients remain constant. This study employs the following panel-regression models to investigate the effect of COVID-19 pandemic on daily (t) market attention (MAT) of each country (j).

{M A T}_{j t} = c + α_{1} {M A T}_{j t - 1} + α_{2} {P a n d e m i c}_{j t - 1} + α_{3} {R E T}_{j t - 1} + α_{4} {V O L}_{j t - 1} + α_{5} {S I R}_{j t - 1} + c_{j} + ε_{j t}

(1)

{M A T}_{j t} = c + α_{1} {M A T}_{j t - 1} + α_{2} {P a n d e m i c}_{j t - 1} + α_{3} {R E T}_{j t - 1} + α_{4} {V O L}_{j t - 1} + α_{5} {S I R}_{j t - 1} + α_{6} {L G R I}_{j t - 1} + {c_{j} + ε}_{j t}

(2)

{M A T}_{j t} = c + α_{1} {M A T}_{j t - 1} + α_{2} {P a n d e m i c}_{j t - 1} + α_{3} {R E T}_{j t - 1} + α_{4} {V O L}_{j t - 1} + α_{5} {S I R}_{j t - 1} + α_{6} {L C H I}_{j t - 1} + c_{j} {+ ε}_{j t}

(3)

where the dependent variable ${M A T}_{j t}$ represents the country-specific and worldwide market attention variables (MAT_C and MAT_W). To account for autocorrelation induced by dynamic relationship in market attention, we use one-period lag (t−1) market attention variable as a regressor. ${P a n d e m i c}_{j t - 1}$ stands for one-period lag (t−1) of daily coronavirus confirmed cases (LDCC) and confirmed deaths (LDCD). Our control variables are lag daily market return ( ${R E T}_{j t - 1})$ , volatility ( ${V O L}_{j t - 1}$ ), short-term interest rate ( ${S I R}_{j t - 1}$ ), the logarithm of overall government responsive index ( ${L G R I}_{j t - 1}$ ) and logarithm of government containment and health index ( ${L C H I}_{j t - 1}$ ). c is the intercept term. Slope coefficients are represented by $α_{1}, α_{2}, α_{3}, α_{4}, α_{5}, a n d α_{6} .$ $ε_{j t}$ is the error term. We hypothesise a significant negative coefficient for $α_{2}$ , and a significant positive coefficient for the estimated $α_{6}$ parameters across our specifications.

We use the Lagrange Multiplier (LM) test to check the acceptability of panel data models over classical regression models. Hausman test has been conducted to choose a suitable panel data model such as fixed-effects model or random-effects model. In addition, Wald test has been conducted to account for group-wise heteroscedasticity. To address possible heteroscedasticity, robust standard errors are used.

IV. Results Discussion

Table 4 reports the estimation results of Equations (1), (2) and (3) for the full sample. Panel (A) and (B) document the impact of daily death cases (LDCD) and confirmed cases (LDCC) on country specific and worldwide market attention (MAT_C, MAT_W), respectively. Reported results in Panel (A) suggests that LDCD and LDCC has a significant negative impact on the MAT_C in all our specification. Moreover, the negative impact of LDCD and LDCC is robust to the inclusion of government economic and heath response variables like LGRI and LCHI. As shown in column (1) of Panel (A), LDCD has a significantly negative coefficient, −0.011 (t = −3.17), which suggests that a 1% increase in logarithm of daily confirmed deaths leads to a decrease in the country MAT_C by 0.01%. This is economically significant because the median MAT_C is around 0.025 (Table 2). Similarly, in column (5) and (6) of Panel (A) we notice that a 1% increase in LDCC hampers investor attention approximately 0.02% (Coefficient = −0.0169, t = −3.37; Coefficient = −0.015, t = −3.08). In the Panel (B) of Table 4, we observe a similar negative significant impact of LDCC and LDCD on the MAT_W. The magnitude of the impact of LDCC and LDCD on worldwide investor attention (MAT_W) is also equally comparable with the country specific investor attention (MAT_C).

Table 4.

Market Attention and Coronavirus (COVID-19) Pandemic.

Panel (A): Estimation Results of Market Attention (Country)
	(1)	(2)	(3)	(4)	(5)	(6)
${M A T_C}_{j t - 1}$	−0.311***(−15.29)	−0.312***(−15.41)	−0.3122***(−15.43)	−0.309***(−15.22)	−0.310***(−15.34)	−0.311***(−15.35)
${L D C D}_{j t - 1}$	−0.011***(−3.17)	−0.020***(−5.09)	−0.020***(−4.73)
${L D C C}_{j t - 1}$				−0.005*(−1.81)	−0.0169***(−3.37)	−0.015***(−3.08)
${R E T}_{j t - 1}$	−0.008*(−1.76)	−0.006(−1.46)	−0.0065(−1.45)	−0.009**(−1.95)	−0.0078*(−1.69)	−0.0076*(−1.69)
${V O L}_{j t - 1}$	0.0313***(2.92)	0.015(1.47)	0.015(1.40)	0.0308**(2.75)	0.0216*(1.96)	0.022**(1.99)
${S I R}_{j t - 1}$	0.016***(7.12)	0.018***(6.69)	0.019***(6.78)	0.017***(6.70)	0.0185***(6.44)	0.0187***(6.46)
${L G R I}_{j t - 1}$		0.035***(3.73)			0.0434***(3.02)
${L C H I}_{j t - 1}$			0.0344***(3.38)			0.0412**(2.77)
LM test	38.22 {0.00}	44.00 {0.00}	68.00 {0.00}	55.22 {0.00}	130.00 {0.00}	80.00 {0.00}
Hausman test	110 {0.00}	89 {0.00}	121 {0.00}	188 {0.00}	211 {0.00}	166 {0.00}
F test	73 {0.00}	101 {0.00}	162 {0.00}	77 {0.00}	90 {0.00}	122{0.00}
Adj. R²	0.09	0.11	0.10	0.14	0.12	0.12
No. of Obs.	4,650	4,650	4,650	4,650	4,650	4,650
Panel (B): Estimation results of market attention (World)
${M A T_W}_{j t - 1}$	−0.277***(−14.66)	−0.279***(−14.93)	−0.28***(−14.97)	−0.275***(14.60)	−0.277***(−14.88)	−0.278***(−14.90)
${L D C D}_{t - 1}$	−0.0108**(−2.04)	−0.022***(−3.71)	−0.0217***(−3.78)
${L D C C}_{j t - 1}$				−0.0056(−1.24)	−0.0201***(−3.34)	−0.0189***(−3.39)
${R E T}_{j t - 1}$	−0.007*(−1.86)	−0.0059(−1.45)	−0.0059(−1.45)	−0.0087**(−2.01)	−0.0068(−1.63)	−0.0069*(−1.65)
${V O L}_{j t - 1}$	0.050***(4.11)	0.031**(2.47)	0.0308**(2.56)	0.051***(4.55)	0.0389***(3.02)	0.0388***(3.08)
${S I R}_{j t - 1}$	0.0123***(3.85)	0.014***(4.28)	0.0144***(4.29)	0.0129***(4.09)	0.0143***(4.51)	0.0144***(4.51)
${L G R I}_{j t - 1}$		0.0422***(3.20)			0.0548***(3.34)
${L C H I}_{j t - 1}$			0.043***(3.17)			0.0529***(3.31)
LM test	50.00 {0.00}	28.60 {0.00}	33.00 {0.00}	44.33 {0.00}	59.30 {0.00}	88.00 {0.00}
Hausman test	220 {0.00}	270 {0.00}	121 {0.00}	150{0.00}	199{0.00}	202{0.00}
F test	111 {0.00}	85 {0.00}	55{0.00}	144 {0.00}	140 {0.00}	212{0.00}
Adj. R²	0.07	0.08	0.08	0.09	0.08	0.06
No. of Obs.	4,650	4,650	4,650	4,650	4,650	4,650

Source: The authors.

Notes: This table reports the panel fixed effect estimation results in two panels. Panel (A) and Panel (B) present the estimated results considering country-wise market attention (MAT_C) and world market attention (MAT_W), respectively. The values in parenthesis are t-statistics, and p-values are given in curly brackets. All specifications were estimated using fixed effects for accounting country heterogeneity. To address possible heteroscedasticity, robust standard errors are used. *, ** and *** indicate statistical significance at 10%, 5% and 1%, respectively.

During the ongoing COVID-19 pandemic; there has been an increasing consensus among policy makers and market participants regarding its multi-dimensional economic impact as it is much more than just a health crisis (United Nations, 2020). Our results support the notion that the COVID-19 has quickly evolved from a provincial health scare to an unprecedented panic shock for the financial market participants, and thus investor attention in the international stock markets declined due to massive drops and high volatility (Al-Awadhi et al., 2020; Baker et al., 2020a, b; Haroon & Rizvi, 2020; Smales, 2020, 2021). The behavioural finance literature suggests that psychological biases among investors could be more pronounced when there is more uncertainty (Donadelli et al., 2017; Garcia, 2013). The realisation preference of investors behaves differently for the prospects of gain compared to losses (Tversky & Kahneman, 1979), human emotions affect information processing and investment decision-making. As a cognitive artefact, human emotion exerts a considerable influence on financial decision-making (Nofsinger, 2005; Olson, 2006). Therefore, human behaviour is significantly different in times of anxiety and fear versus periods of prosperity and tranquillity (Akerlof & Shiller, 2010; Garcia, 2013; Segal et al., 2015). Therefore, the spillover effects of the wave of coronavirus infections and its accompanying global industrial activity downturn could have affected investor information demand in the financial market. The available literature on investor attention and information demand suggests that when investors are more risk-averse, they prefer to have less risky assets and reduce the frequency of information updates (Huang & Liu, 2007). Andrei and Hasler (2015) posit that investors’ active attention is state-dependent and related to time-varying risk aversion to extreme downturns. Given the downside effect of the current pandemic spread, it is reasonable to expect that investor expectations will have a comparable benchmark with risk-averse behaviour. For instance, recent study by Smales (2021) concludes that during the COVID-19 crisis, probably, retail investors are searching for non-fundamental information to resolve heightened uncertainty, rather than examining fundamental information. Due to the COVID-19 pandemic shock and associated economic uncertainty, investors could be paying more attention to the pandemic information search than the fundamentals due to ‘category-learning behaviour². Therefore, a combination of policy uncertainty and information channels may have caused investors to move from risk assets like equity to haven assets (flight-to-safety), hence an adverse effect of the COVID-19 on investor equity market attention.

We also find a positive and significant impact of government policy response variables like LGRI and LCHI on investor attention. In Panel (A), a 1% increase in the overall government response (LGRI) leads to approximately 0.035% (Coefficient = 0.035), and 0.043% (Coefficient = 0.0434) increase in the MAT_C. Similarly, a 1% increase in the health-related measures (LCHI) directs around 0.034% (Coefficient = 0.0344) and 0.041% (Coefficient = 0.0412). In panel (B) of Table 4, we find a consistently significant and positive impact of LGRI and LCHI on MAT_W. Our results suggest that governments prioritise economic stimulus through fiscal policy measures and invest in public health to help countries recover faster from the COVID-19 pandemic. These findings support the notion that lockdowns, travel bans and economic stimulus packages, all had a positive effect on the financial markets (Baldwin & di Mauro, 2020; Dasgupta et al., 2021; Goel & Dash, 2021; Narayan et al., 2021b). The positive (negative) relationship between volatility (return) and the market attention is consistent with the findings of existing market attention literature (Chen, 2017; Da et al., 2011). The positive impact of SIR on MAT_C and MAT_W can be attributed to the lower interest rate policy response environment undertaken by most governments to bring resiliency in the corporate sector.

Robustness Tests

In this section, we carry out three subsample analyses as further robustness tests. Tables 5, 6 and 7 report the subsample analysis based on financial market development, financial stability and pricing of investor attention, respectively. We follow similar model specifications (Equations 1, 2 and 3) in all subsample analyses. Our primary interest is to analyse whether several financial market-related characteristics influence pandemic effects on investor attention.

Table 5.

Market Attention and Coronavirus Pandemic: Role of Financial Market.

Variables	High Financial Market Development Index				Low Financial Market Development Index
Variables	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
Panel (A): Country specific investor attention
${M A T_C}_{j t - 1}$	−0.283***(−10.12)	−0.277***(−10.14)	−0.275***(−9.96)	−0.275***(−9.98)	−0.350***(−12.19)	−0.35***(−12.16)	−0.350***(−12.28)	−0.350***(−12.21)
${L D C D}_{j t - 1}$	−0.023***(−3.99)	−0.025***(−4.19)			−0.0237***(−3.48)	−0.020**(−2.79)
${L D C C}_{j t - 1}$			−0.0166**(−2.03)	−0.018**(−2.22)			−0.022***(−3.49)	−0.0164**(−2.34)
${L G R I}_{j t - 1}$	0.04** (2.51)		0.04*(1.74)		0.04***(3.15)		0.05***(3.30)
${L C H I}_{j t - 1}$		0.05**(2.79)		0.05**(2.07)		0.03**(2.07)		0.04**(1.97)
Controls	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
LM &HMtest	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
F test	99{0.00}	200{0.00}	111{0.00}	108{0.00}	156{0.00}	183{0.00}	149{0.00}	244 {0.00}
Adj. R²	0.12	0.10	0.11	0.10	0.07	0.07	0.10	0.09
No. of Obs.	2,511	2,511	2,511	2,511	2,139	2,139	2,139	2,139
Panel (B): World-wide investor attention
${M A T_W}_{j t - 1}$	−0.246***(−10.59)	−0.247***(−10.72)	−0.245***(−10.38)	−0.246***(−10.53)	−0.316***(−11.36)	−0.315***(−11.32)	−0.316***(−11.44)	0.315***(−11.35)
${L D C D}_{j t - 1}$	−0.025***(−2.99)	−0.025***(−3.29)			−0.022**(−2.51)	−0.019**(−2.08)
${L D C C}_{j t - 1}$			−0.022**(−2.41)	−0.023**(−2.79)			−0.020**(−2.54)	−0.014**(−2.04)
LGRI _jt ₋₁	0.05**(2.70)		0.06**(2.69)		0.04**(2.38)		0.06**(2.50)
LCHI _jt ₋₁		0.06***(2.96)		0.07***(3.05)		0.04*(1.79)		0.04*(1.87)
Controls	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
LM & HMtest	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
F test	153 {0.00}	89{0.00}	209{0.00}	181 {0.00}	201 {0.00}	68{0.00}	208{0.00}	77 {0.00}
Adj. R²	0.11	0.14	0.11	0.13	0.10	0.12	0.07	0.09
No. of Obs.	2,511	2,511	2,511	2,511	2,139	2,139	2,139	2,139

Source: The authors.

Notes: This table reports the fixed effects estimation results of Equations (1), (2) and (3) for higher and lower financial market development country category. Panel (A) and Panel (B) present the estimated results considering MAT_C and MAT_W, respectively. We use the 2017 median value of IMF financial market development index value (FMDI) to split the sample countries into high and low financial market development category. We have not reported the coefficients of the market related control variables (RET, VOL, SIR), Hausman test (HM) and Lagrange Multiplier (LM) test results for brevity. The values in parenthesis are t-statistics, and p-values are given in curly brackets. All specifications were estimated using fixed effects for accounting country heterogeneity. *, ** and *** indicate statistical significance at 10%, 5% and 1%, respectively.

Table 6.

Investor Attention and Coronavirus Pandemic: Role of Financial Stability.

Variables	High Financial Stability (Bank Z-Score)				Low Financial Stability (Bank Z-Score)
Variables	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
Panel (A): Country specific investor attention
${M A T_C}_{j t - 1}$	−0.335***(−12.47)	−0.3353***(−12.48)	−0.3344***(−12.36)	−0.3344***(−12.37)	−0.286***(−9.43)	−0.287***(−9.51)	−0.287***(−9.53)	−0.288***(−9.53)
${L D C D}_{j t - 1}$	−0.0156**(−2.60)	−0.0144**(−2.26)			−0.0263***(−4.09)	−0.0258***(−3.99)
${L D C C}_{j t - 1}$			−0.0069(−1.07)	−0.0054(−0.78)			−0.026***(−3.55)	−0.023***(−3.11)
LGRI _jt ₋₁	0.03**(2.25)		0.02(1.37)		0.03*(1.52)		0.03*(1.57)
LCHI _jt ₋₁		0.03*(1.72)		0.02*(1.69)		0.03*(1.46)		0.03*(1.48)
Controls	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
LM & HM test	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
F test	310{0.00}	360{0.00}	202{0.00}	175{0.00}	300{0.00}	112{0.00}	157{0.00}	297 {0.00}
Adj. R²	0.10	0.11	0.09	0.14	0.17	0.19	0.09	0.12
No. of Obs.	2,325	2,325	2,325	2,325	2,325	2,325	2,325	2,325
Panel (B): World-wide investor attention
${M A T_W}_{j t - 1}$	−0.284***(−11.73)	−0.284***(−11.75)	−0.283***(−11.56)	−0.283***(−11.58)	−0.274***(−9.45)	−0.275***(−9.48)	−0.275***(−9.53)	−0.275***(−9.52)
${L D C D}_{j t - 1}$	−0.024***(−3.21)	−0.023***(−3.05)			−0.020**(−2.27)	−0.020**(−2.48)
${L D C C}_{j t - 1}$			−0.017**(−2.33)	−0.016**(−2.11)			−0.022**(−2.51)	−0.020**(−2.65)
LGRI _jt ₋₁	0.03(1.52)		0.04(1.57)		0.05***(2.91)		0.07***(3.12)
LCHI _jt ₋₁		0.03(1.46)		0.03(1.48)		0.06***(2.95)		0.07***(3.12)
Controls	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
LM & HM test	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
F test	177{0.00}	222 {0.00}	210{0.00}	189 {0.00}	393 {0.00}	254{0.00}	170{0.00}	289{0.00}
Adj. R²	0.08	0.13	0.12	0.12	0.14	0.16	0.12	0.13
No. of Obs.	2,325	2,325	2,325	2,325	2,325	2,325	2,325	2,325

Source: The authors.

Notes: This table reports the fixed effects estimation results of Equations (1), (2) and (3) for high and low financial stability (bank Z-score) country category. Panel (A) and Panel (B) present the estimated results for MAT_C and MAT_W, respectively. We use the 2017 median value of bank Z-score to split the sample countries into high and low financial market development category. The values in parenthesis are t-statistics, and p-values are given in curly brackets. All specifications were estimated using fixed effects for accounting for country heterogeneity. To address possible heteroscedasticity, robust standard errors are used. *, **, and *** indicate statistical significance at 10%, 5% and 1%, respectively.

Table 7.

Investor Attention and Coronavirus Pandemic: Role of Investor Attention Pricing.

Variables	High Investor Attention Pricing				Low Investor Attention Pricing
Variables	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
Panel (A): Country specific investor attention
${M A T_C}_{j t - 1}$	−0.292***(−8.91)	−0.291***(−8.91)	−0.29(−8.95)	−0.288***(−8.86)	−0.318***(−12.80)	−0.319***(−12.82)	−0.318***(−12.75)	−0.317***(−12.71)
${L D C D}_{j t - 1}$	−0.024***(−3.42)	−0.024***(−3.14)			−0.019***(−3.95)	−0.019***(−3.63)
${L D C C}_{j t - 1}$			−0.017**(−2.22)	−0.019**(−2.21)			−0.016**(−2.66)	−0.014**(−2.34)
${L G R I}_{j t - 1}$	0.03**(2.48)		0.039*(1.90)		0.04***(3.06)		0.04**(2.55)
${L C H I}_{j t - 1}$		0.04**(2.53)		0.05**(2.10)		0.04**(2.63)		0.04**(2.20)
Controls	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
LM & HM test	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
F test	283{0.00}	189{0.00}	212{0.00}	98{0.00}	233{0.00}	120 {0.00}	101{0.00}	212 {0.00}
Adj. R²	0.20	0.13	0.22	0.24	0.11	0.19	0.13	0.21
No. of Obs.	1,209	1,209	1,209	1,209	3,441	3,441	3,441	3,441
Panel (B): World-wide investor attention
${M A T_W}_{j t - 1}$	−0.290***(−7.74)	−0.290***(−7.78)	−0.289***(−7.60)	−0.290***(−7.65)	−0.275***(−12.56)	−0.275***(−12.59)	−0.275***(−12.57)	−0.275***(−12.58)
${L D C D}_{j t - 1}$	−0.026**(−2.51)	−0.027**(−2.81)			−0.020***(−3.02)	−0.019***(−3.02)
${L D C C}_{j t - 1}$			−0.020*(−1.71)	−0.022*(−2.08)			−0.019***(−2.82)	−0.017***(−2.73)
LGRI _jt ₋₁	0.02(0.97)		0.04(1.06)		0.05***(3.12)		0.05***(3.21)
LCHI _jt ₋₁		0.03(1.17)		0.04(1.38)		0.05***(2.96)		0.06***(3.01)
Controls	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
LM & HM test	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
F test	303{0.00}	211{0.00}	100{0.00}	182{0.00}	188{0.00}	388 {0.00}	286{0.00}	255{0.00}
Adj. R²	0.20	0.13	0.22	0.24	0.19	0.23	0.15	0.17
No. of Obs.	1,209	1,209	1,209	1,209	3,441	3,441	3,441	3,441

Source: The authors.

Notes: This table reports the fixed effects estimation results of Equations (1), (2) and (3) for high and low investor attention risk pricing country category. For our country classification, we rely on the Chen (2017) study and use the investor attention risk pricing coefficients to split the sample countries into two sub-samples. The values in parenthesis are t-statistics, and p-values are given in curly brackets. All specifications were estimated using fixed effects for accounting the country heterogeneity. *, ** and *** indicate statistical significance at 10%, 5% and 1%, respectively.

Table 5 reports the results for Equations (1), (2) and (3) for higher and lower values of IMF financial market development index (FMDI). Panel (A) and Panel (B) focuses on MAT_C and MAT_W, respectively. To the extent that depth, access and efficiency of financial markets help to reduce informational asymmetries, lessen financial constraints and promote risk sharing (Svirydzenka, 2016), the deterrent effect of COVID-19 on the investor attention for countries with higher (lower) values of FMDI will be less (more) pronounced. Consistent with our earlier discussion, we find a significant negative impact of pandemic effect variables (LDCD, LDCC) on investor attention (MAT_C, MAT_W). Moreover, the magnitude of the impact has no substantial difference across the two categories of FMDI.

Thus, irrespective of the nature of the financial system, our results underscore the unprecedented impact of COVID-19 pandemic on investor attention. We notice that government response (LGRI, LCHI) has a positive and significant effect on MAT_C and MAT_W, and the magnitude of the impact is qualitatively similar across the broad financial market development categorisation.

Table 6 emphasises on the sub-sample analysis based on financial stability, measured through bank Z-score. Existing research focusing on the extensive sample multi-country evidence suggests that higher value of bank Z-score measures financial stability, and better capital regulations (Phan et al., 2020). In a recent study, Phan et al. (2020) suggest that a stable financial system can absorb both internal and external shocks (economic and non-economic), thereby protecting the economy and other financial systems from failure. Therefore, it is interesting to analyse how the countries with distinctive financial stability scenario behave for quantifying the impact of pandemic information and government response on country-specific and world-wide investor attention. Reported results in Panel (A) and Panel (B) of Table 6 indicate that irrespective of the nature of financial stability. The impact of LDCD and LDCC on investor attention (MAT_C, MAT_W) is negative and significant. The overall impact of LGRI and LCHI is positive and significant. It is interesting to note that in Panel (B) countries with lower Z-score (i.e., lower financial stability) resort to government measures (LGRI and LCHI) in a higher magnitude to boost investor attention. For instance, we notice that a 1% increase in the LGRI and LCHI increases worldwide investor attention (MAT_W) by 0.07% (Coefficient = 0.07, t = 3.12), as compared to the countries with higher Z-score. Thus, the impact of government intervention (LGRI and LCHI) is more noticeable in the low Z-score countries.

Table 7 presents the results for investor attention pricing categorisation. In Panel (A) and (B), we find that LDCD and LDCC coefficients are negative and significant for both high and low investor attention pricing. Moreover, MAT_C and MAT_W negatively affect the stock index returns, which aligns with Chen (2017). The overall impact of government response indices (LGRI and LCHI) is positive and significant across both the panels except for Panel (B). The heightened economic uncertainty due to COVID-19 (Baker et al., 2020a, b) may be the reason for the positive and significant impact of government responses on world-wide investor attention for countries with inherently high attention pricing.

V. Conclusion and Policy Implications

This article hypothesises a negative impact of the COVID-19 pandemic on country-specific and worldwide investor attention. For our empirical research, we use a panel model specification for daily data of 50 countries’ stock market samples ranging from January 2020 to May 2020. Furthermore, our robustness tests help shed more light on whether the country-specific characteristics like financial market development, financial stability and investor attention risk pricing influence the relationship between pandemic information and investor attention.

We find a robust negative relationship between the COVID-19 pandemic information variables like daily confirmed cases and death cases on investor attention in the global stock market. Our results support the existing empirical evidence of the negative impact of the coronavirus pandemic on the expected returns and increasing volatility across global stock markets (Al-Awadhi et al., 2020; Ali et al., 2020; Ji et al., 2020; Zaremba et al., 2020). Our robustness tests indicate that the magnitude of the negative impact does not vary a lot between countries characterised by their financial market development, financial stability and asset pricing implication of investor attention. Our results support the information channel (Andrei & Hasler, 2015; Huang & Liu, 2007; Zhang, 2006), and heightened uncertainty concern (Baker et al., 2020a, b; Mishkin, 1999; Segal et al., 2015) view for the market participants behaviour with lower investor attention. In the current scenario of pandemic spread, governments and central banks worldwide have announced unprecedented health measures and fiscal and monetary stimulus to support their economies. Our empirical analysis uses the overall government response index and containment and health index to find a positive impact on investor attention. Results suggest that increasing policy response from the government helps to minimise the uncertainty hovering around the large-scale global financial crisis. Various regulatory policy announcements can restore investor confidence in the financial markets. For instance—Oxford COVID-19 Government Response Tracker (OxCGRT) maintains a record of multiple non-pharmaceutical interventions governments undertake worldwide. Recent studies validate such measures’ effectiveness in trimming down coronavirus’s devastating impact on stock markets (Goel & Dash, 2021; Salisu & Akanni, 2020).

The number of countries limits our study in our sample as our research uses the benchmark market index. Moreover, the firm-specific effects might be very different depending on the variation in the firm characteristics. It will be more insightful to examine the COVID-19 impact on the firm and industry-specific investor attention. The firm characteristics effect to mitigate the COVID-19 effect could also be a potential research question. There could be a scope to extend our empirical setting with a longer time horizon, including a second wave and recovery after this crisis. Our findings also provide a scope to extend the analysis with other financial assets like commodities, foreign exchange, cryptocurrencies, equity traded funds, hedge funds, etc. along with stocks. We retain these limitations as the motivation for our future research endeavours.

Footnotes

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Funding

The authors received no financial support for the research, authorship and/or publication of this article.

Notes

ORCID iD

Garima Goel

References

Akerlof

G. A.

, & Shiller

R. J.

(2010). Animal spirits: How human psychology drives the economy, and why it matters for global capitalism . Princeton University Press.

Al-Awadhi

A. M.

, Al-Saifi

, Al-Awadhi

, & Alhamadi

(2020). Death and contagious infectious diseases: Impact of the COVID-19 virus on stock market returns. Journal of Behavioral and Experimental Finance , 27, 100326.

Ali

, Alam

, & Rizvi

S. A. R.

(2020). Coronavirus (COVID-19)–An epidemic or pandemic for financial markets. Journal of Behavioral and Experimental Finance , 27, 100341.

Altig

, Baker

, Barrero

J. M.

, Bloom

, Bunn

, Chen

, Davis

S. J.

, Leather

, Meyer

, Mihaylov

, Mizen

, Parker

, Renault

, Smietanka

, & Thwaites

(2020). Economic uncertainty before and during the COVID-19 pandemic. Journal of Public Economics , 191, 104274.

Andrei

, & Hasler

(2015). Investor attention and stock market volatility. The Review of Financial Studies , 28(1), 33–72.

Baker

S. R.

, Bloom

, Davis

S. J.

, Kost

K. J.

, Sammon

M. C.

, & Viratyosin

(2020a). The unprecedented stock market impact of COVID-19 (No. w26945). National Bureau of Economic Research (NBER).

Baker

S. R.

, Bloom

, Davis

S. J.

, & Terry

S. J.

(2020b). COVID-induced economic uncertainty (Working Paper No. w26983). National Bureau of Economic Research (NBER).

Baldwin

, & di Mauro

B. W.

(2020). Mitigating the COVID economic crisis: Act fast and do whatever it takes . The Centre for Economic Policy Research (CEPR) Press.

Chen

(2017). Investor attention and global stock returns. Journal of Behavioral Finance , 18(3), 358–372.

10.

Corbet

, Larkin

, & Lucey

(2020). The contagion effects of the COVID-19 pandemic: Evidence from gold and cryptocurrencies. Finance Research Letters , 35, 101554.

11.

, Engelberg

, & Gao

(2011). In search of attention. The Journal of Finance , 66(5), 1461–1499.

12.

Dasgupta

, Dierckxsens

, & Verick

(2021). The impact of the COVID-19 crisis on middle-income countries. The Indian Economic Journal , 69(3). https://journals-sagepub-com-s.web.bisu.edu.cn/doi/full/10.1177/00194662211023847

13.

Donadelli

, Kizys

, & Riedel

(2017). Dangerous infectious diseases: Bad news for main street, good news for Wall Street? Journal of Financial Market , 35, 84–103.

14.

Garcia

(2013). Sentiment during recessions. The Journal of Finance , 68(3), 1267–1300.

15.

Goel

, & Dash

S. R.

(2021). Investor sentiment and government policy interventions: Evidence from COVID-19 spread. Journal of Financial Economic Policy , 14(2), 242–267.

16.

Haroon

, & Rizvi

S. A. R.

(2020). COVID-19: Media coverage and financial markets behavior—A sectoral inquiry. Journal of Behavioral and Experimental Finance , 27, 100343.

17.

Huang

, & Liu

(2007). Rational inattention and portfolio selection. The Journal of Finance , 62(4), 1999–2040.

18.

, Zhang

, & Zhao

(2020). Searching for safe-haven assets during the COVID-19 pandemic. International Review of Financial Analysis , 71, 101526.

19.

Kahneman

(1973). Attention and effort . Prentice-Hall.

20.

McKibbin

W. J.

, & Fernando

(2020, March 2). The global macroeconomic impacts of COVID-19: Seven scenarios (CAMA Working Paper No. 19/2020). https://ssrn.com/abstract=3547729

21.

Mishkin

F. S.

(1999). Anatomy of a financial crisis. Journal of Evolutionary Economics , 2(2), 115–130.

22.

Narayan

P. K.

, Phan

D. H. B.

, & Liu

(2021a). COVID-19 lockdowns, stimulus packages, travel bans, and stock returns. Finance Research Letters , 38, 101732.

23.

Narayan

, Purnaningrum

, & Khawari

(2021b). Assessing the financial implications of COVID-19 within the SVAR framework for some Asian Countries. The Indian Economic Journal , 69(4). https://journals-sagepub-com-s.web.bisu.edu.cn/doi/full/10.1177/00194662211036097

24.

Nofsinger

J. R.

(2005). Social mood and financial economics. The Journal of Behavioral Finance , 6(3), 144–160.

25.

OECD. (2020). OECD economic outlook (Vol. 2020, Issue 1: Preliminary version, No. 107). OECD Publishing. https://doi.org/10.1787/0d1d1e2e-en

26.

Olson

K. R.

(2006). A literature review of social mood. The Journal of Behavioral Finance , 7(4), 193–203.

27.

Peng

, & Xiong

(2006). Investor attention, overconfidence and category learning. Journal of Financial Economics , 80(3), 563–602.

28.

Phan

D. H. B.

, Iyke

B. N.

, Sharma

S. S.

, & Affandi

(2020). Economic policy uncertainty and the financial stability—Is there a relation? Economic Modelling , 94, 1018–1029. https://doi.org/10.1016/j.econmod.2020.02.042

29.

Roser

, Ritchie

, Ortiz-Ospina

, & Hasell

(2020). Coronavirus Pandemic (COVID-19). https://ourworldindata.org/coronavirus

30.

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.

31.

Schell

, Wang

, & Huynh

T. L. D.

(2020). This time is indeed different: A study on global market reactions to public health crisis. Journal of Behavioral and Experimental Finance , 27, 100349.

32.

Segal

, Shaliastovich

, & Yaron

(2015). Good and bad uncertainty: Macroeconomic and financial market implications. Journal of Financial Economics , 117(2), 369–397.

33.

Smales

L. A.

(2020). Investor attention and the response of US stock market sectors to the COVID-19 crisis. Review of Behavioral Finance , 13(1), 1940–5979.

34.

Smales

L. A.

(2021). Investor attention and global market returns during the COVID-19 crisis. International Review of Financial Analysis , 73, 101616.

35.

Svirydzenka

(2016). Introducing a new broad-based index of financial development (Working Papers, WP/16/5). International Monetary Fund (IMF).

36.

Tversky

, & Kahneman

(1979). Prospect theory: An analysis of decision under risk. Econometrica , 47(2), 263–291.

37.

United Nations. (2020). A UN framework for the immediate socio-economic response to COVID-19 . https://unsdg.un.org/sites/default/files/2020-04/UN-framework-for-the-immediate-socio-economic-response-to-COVID-19.pdf

38.

World Bank. (2020). The global economic outlook during the COVID-19 Pandemic: A changed world . https://www.worldbank.org/en/news/feature/2020/06/08/the-global-economic-outlook-during-the-covid-19-pandemic-a-changed-world

39.

Young

M. E.

, King

, Harper

, & Humphreys

K. R.

(2013). The influence of popular media on perceptions of personal and population risk in possible disease outbreaks. Health Risk Society , 15(1), 103–114.

40.

Zaremba

, Kizys

, Aharon

D. Y.

, & Demir

(2020). Infected markets: Novel coronavirus, government interventions, and stock return volatility around the globe. Finance Research Letters , 35, 101597.

41.

Zhang

X. F.

(2006). Information uncertainty and stock returns. The Journal of Finance , 61(1), 105–137.