The Impact of COVID-19-Induced Sentiment on Firm Performance: The Moderating Impact of Sustainable ESG Activities

Abstract

This paper uses the data of nonfinancial firms from 49 countries to show that the benefits of improvements in COVID-19-induced sentiment accrue to firms that expend more resources on sustainable environmental, social, and governance (ESG) activities. The findings remain robust across various estimation strategies and across various subsamples. The findings also show that the social and environmental dimensions of ESG moderate the relationship between COVID-19-induced sentiment and firm performance. In contrast, the governance dimension has no significant impact. Our findings suggest that firms should prioritize the environmental and social dimensions of ESG to build resilience and protect themselves from adverse shocks.

1. Introduction

The COVID-19 outbreak that started in mid-January 2020 was considered a major threat to the entire world. It posed significant challenges to economies due to its adverse impact on most industries, including travel, healthcare, tourism, and construction. It also caused major shifts in decision making at individual and institutional levels. Quarantine policies reduced population mobility, which eventually undermined economic activities and resulted in recession across the world [1]. The COVID-19 pandemic also influenced financial markets in addition to actual markets. Makni [2]. demonstrates that the COVID-19 pandemic severely affected firms’ performance by lowering their overall income and investment levels, while Shen et al. [1] further show that the coronavirus crisis caused uncertainty in business decision making, which had a detrimental effect on firm operations.

This paper argues that investor sentiment had a significant impact on firm performance during the coronavirus pandemic. Sentiments are unrelated to fundamentals but can significantly affect expectations regarding the future profitability of firms. Deteriorating sentiments (pessimism) can make stock market participants lower their expectations about future cash flows and increase the discount rates used in valuations. Both factors can thus negatively affect firm performance. We maintain that one of the major ways in which firms protected themselves against deteriorating sentiments was by investing in environmental, social, and governance (ESG) practices. There are several reasons why ESG initiatives protect firms against negative sentiments. First, investing in ESG practices can act as a strategic approach during times of increased market turbulence [3], and firms that expend more resources on ESG can build their reputation as trustworthy firms [4]. Second, we believe that firms’ commitment to ESG practices sends positive signals to stakeholders about firms’ future prospects. When faith in economic recovery is low, spending resources on ESG practices indicates that managers have faith that their firms can weather the storm of uncertainty and maintain profitability. We argue that shareholders will allow sizable investments in ESG practices only if they believe that their own wealth will be unaffected by such investments. This paper maintains that these signals are valuable because investors reward such firms with positive returns. Third, by incorporating ESG considerations into their operations and business models, firms prioritize long-term value creation over short-term benefits. This can help firms build resilience and stability, which can mitigate the negative impact of market sentiment driven by short-term volatility.

In line with the aforementioned arguments, this study demonstrates that companies in 49 countries that committed to ESG practices benefited more from improved COVID-19-induced sentiment. Our results also show that that the social and environmental dimensions of ESG are the moderating factors of the relationship between sentiment induced by COVID-19 and corporate performance. Aside from the corporate social responsibility (CSR) strategy dimension, the governance dimension has no significant effect on this relationship. Our findings remain robust across various estimation strategies and across various subsamples.

Several studies have assessed how ESG practices affect stock price fluctuations and performance during financial crises [5,6]. As far as we are aware, none of the previous research has examined how ESG practices moderate the impact of COVID-19-induced sentiment on corporate performance. Our paper extends the existing literature by assessing whether firms’ engagement in ESG provides safe-haven opportunities for investors during turbulent market conditions. Therefore, our findings provide valuable insights for policymakers, investors, and corporations. Our study provides incentives for firms to adopt ESG activities in their strategic decision making, as this engagement can positively affect their financial performance. More specifically, our findings highlight that in order to build resilience and profitable strategies, businesses should prioritize social and environmental initiatives. Engaging in ESG activities signals that firms are dedicated to ethical and responsible business practices, even in periods of market turmoil. This can result in several positive externalities such as increased investor trust and more informed decision making, thereby improving efficiency and lowering information asymmetry in financial markets.

The remaining portions of this text are arranged as follows. The theory is developed in Section 2; the data and methodology are presented and explained in Section 3; Section 4 and Section 5 present the findings; and Section 6 concludes the paper.

2. Literature Review and Hypothesis Development

The economic downturn triggered by the coronavirus pandemic resulted in financial challenges for firms across industries. Shen et al. [1] argue that the coronavirus crisis exposed firms to heightened risks in the form of decreasing revenues and increasing costs. These risks increased the financial constraints and liquidity shortages faced by firms, which negatively affected firms’ performance [1,7]. An important factor driving the negative consequences of the COVID-19 pandemic on firm profitability was the excessive negative sentiments that prevailed in the markets during that time. These sentiments lowered the faith of investors in both the markets and the firms operating within them. The prior literature has suggested that lockdowns and the uncertainty surrounding the COVID-19 outbreak influenced investor sentiment and generated high levels of panic and hysteria among traders [7,8,9]. Regarding the influence of COVID-19-induced sentiment on firm performance, it is assumed that sentiment significantly affects asset pricing and stock market efficiency. This therefore casts doubt on conventional financial models, which assume that rational investors value prices at the discounted value of forecasted cash flows [10,11]. Under normal market conditions, it is assumed that rational investors make decisions based on their information-processing activities. Periods of market turbulence, such as the COVID-19 pandemic, may result in investors making fear-induced behavioral decisions and adopting biased asset allocation strategies. These behavioral biases may lead to inefficiencies in the market.

Although several research papers have shown that ESG practices positively influenced stock performance during the COVID-19 crisis period [5,12], some studies revealed that ESG practices alone did not provide a shield of immunization against this crisis [13]. The mixed evidence in the literature thus drives us to investigate the effects of ESG activities on firm performance during periods of high uncertainty, namely the COVID-19 crisis period. Therefore, we hypothesize that ESG activities can be instrumental in safeguarding corporations against the adversities caused by COVID-19-induced sentiment.

Other studies have investigated the impact of investor sentiment on green industry stocks. For instance, Wang et al. [14] show that environmental news has a pronounced influence on green industry stock performance in China. They also find that investor sentiment, as captured by online comments from individual investors, partially mediates this effect. Zhang and Zhang [15] study the effect of investor attention on ESG performance and indicate that it significantly improves the ESG standards of listed companies. Serafeim [4] assesses the effect of sentiment on the valuation of companies that engage in ESG-related activities. The author finds that ESG initiatives are undervalued in the presence of negative public sentiment, yielding positive future abnormal returns, while in contrast, the market valuation of ESG performance tends to be higher when positive sentiment increases. On the other hand, Dhasmana et al. [16] use NARDL and QNARDL models to evaluate the link between investor sentiment and high-ESG-scoring firms in India. The authors confirm the presence of an asymmetric and quantile-dependent relationship between investor sentiment and high-ESG-scoring firms. Conversely, Reboredo and Ugolini [17] find evidence that the Twitter Sentiment Index has no significant impact on renewable energy stocks within the United States market. With reference to other financial indices, López-Cabarcos et al. [18] find that social network sentiment influences the volatility of the ESG S&P Index more than the volatility of the S&P 500 itself. Focusing on the COVID-19 crisis, Naeem et al. [19] suggest an increased multifractality in global ESG equity markets when using implied volatility as a proxy for investor sentiment. Umar et al. [8] employ a wavelet coherence model to study the interdependence between the COVID-19-induced Panic Index and the volatility of five leading ESG indices across the world in the year 2020, with the results suggesting that ESG indices can serve as diversifiers particularly in times of extreme market conditions such as a pandemic.

In our paper, we argue that socially responsible firms have an edge over their peers, especially during periods of excessive turmoil and uncertainty. We suggest that engaging in ESG activities can be crucial, as it protects firms from the challenges posed by COVID-19-induced sentiment. We highlight three channels through which ESG activities generate advantages for firms during uncertain times. First, ESG activities improve the capacity of firms to absorb the shocks that arise from uncertain events. Borghesi et al. [20] highlight that ESG engagement can serve as insurance during periods of high uncertainty. Additionally, they contend that high-CSR companies benefited from easier access to capital and cheaper financing and transaction costs during the global financial crisis. The authors consider the emphasis on long-term investments and the enduring connections of these firms with internal and external stakeholders as the main reason behind these lower financing and transaction costs. Beloskar and Rao [21] support these arguments by showing that firms that engage in ESG activities exhibit fewer price declines and lower volatility levels compared to other firms during uncertain times. Along the same lines, Mattera et al. [22] demonstrate that a firm’s commitment to long-term CSR strategies enhances its resilience during crisis periods such as the COVID-19 pandemic.

Second, businesses that invest greater resources in ESG can generate social capital for themselves. Lins et al. [3] note that ESG-committed firms can outperform others during uncertain times due to the social capital associated with ESG activities. Beloskar and Rao [21] also confirm that traders may lower their expectations regarding future forecasted earnings when a worldwide crisis occurs; however, they are more confident in companies that have high ESG scores. According to Eccles et al. [23], stakeholder trust can be enhanced in high-CSR corporations that employ procedures that continuously engage with stakeholders over the long term. Along the same lines, Hoang and Phang [24] highlight that CSR engagement buffers against the unfavorable effects of negative events.

Third, the benefits of ESG activities may also be due to the signaling hypothesis and the influence of ESG in decreasing information asymmetry concerns. Cui et al. [25] show that firms’ ESG performance is linked to smaller variations in analysts’ forecasts and bid-ask spreads. Lys et al. [14] also demonstrate that firms invest in ESG in the current period when they anticipate future financial success. This conveys signals to investors about the firm’s future financial prospects. The authors argue that the positive relationship between ESG spending and firm performance is more likely to be caused by the signaling effect of ESG expenditure than profitable investment outcomes. Based on the abovementioned channels, we hypothesize that firms committed to ESG activities were better poised to navigate the challenges of the pandemic. Therefore, these firms were also more likely to benefit from any improvement in COVID-19-induced sentiment.

Hypothesis. 

Firms spending more on ESG activities are more likely to benefit from any improvement in COVID-19-induced sentiment than firms that spend fewer resources on ESG activities.

Several studies have also examined the individual impacts of the E, S, and G pillars on firm performance and practices. For instance, corporations that focus on the social dimensions of ESG are associated with better financial performance [26]. Aouadi and Marsat [27] argue that the benefits resulting from the social dimensions are particularly significant for high-attention (larger, better-performing, and more visible) firms. Additionally, robust corporate governance is also crucial for comprehensive ESG reporting and performance. Corporations with strong governance frameworks typically enjoy higher market value and financial stability [28]. Interesting findings are highlighted by Espinosa-Méndez et al. [29], who explore the ESG pillars’ impact on family firms’ values while accounting for the moderating effects of financial constraints and agency problems. The authors find that the environmental and social pillars exert significant positive influences on the performance of family firms, while the governance pillar shows no significant impact on firm value. Other studies, such as Engelhardt et al. [30], examine the impact of ESG on stock performance during crisis periods, particularly the COVID-19 pandemic, and find that firms with high ESG ratings exhibit higher returns and lower risk. The authors highlight that the main driver of their results is the social pillar of ESG. Yoo et al. [31] emphasize the significance of firms’ environmental scores in their sustainable activities, noting that these play a key role in increasing firms’ returns and decreasing risk during COVID-19.

In light of these findings, we aim to examine how each ESG dimension moderates the influence of sentiments on firm performance. Therefore, our sub-hypotheses are as follows:

Hypothesis 1a (H1a).

Firms with higher environmental pillar scores are more likely to benefit from any improvement in COVID-19-induced sentiment than firms with lower environmental pillar scores.

Hypothesis 1b (H1b).

Firms with higher social pillar scores are more likely to benefit from any improvement in COVID-19-induced sentiment than firms with lower social pillar scores.

Hypothesis 1c (H1c).

Firms with higher governance pillar scores are more likely to benefit from any improvement in COVID-19-induced sentiment than firms with lower governance pillar scores.

3. Data and Methodology

3.1. Data

This study makes use of information provided by Thomson Reuters Eikon, Worldscope, and Datastream. The sample consists of 7751 unique nonfinancial firms (13,102 firm-year observations) from 49 countries. Given that the COVID-19 pandemic was at its peak during 2020 and 2021, the analysis is based on these two years. The coronavirus sentiment indices data were obtained from the RavenPack database, which estimates a series of indices that measures the amount of news and data related to COVID-19-induced indicators including panic, hype, social, and fake news, among others.

3.2. Research Methodology

3.2.1. Effect of ESG on the Relationship between COVID-19-Induced Sentiment and Firm Performance

To evaluate how ESG activities influences the impact of COVID-19-induced sentiment on firm performance, we estimate different versions of the following regression equation:

$$\begin{array}{ll}\mathrm{Q}=\mathsf{\alpha }+{\mathsf{\beta }}_1\left(\mathrm{E}\mathrm{S}\mathrm{G}\right)& +{\mathsf{\beta }}_2\left(\mathrm{S}\mathrm{E}\mathrm{N}\mathrm{T}\mathrm{I}\mathrm{M}\mathrm{E}\mathrm{N}\mathrm{T}\right)+{\mathsf{\beta }}_3\left(\mathrm{E}\mathrm{S}\mathrm{G}\ast \mathrm{S}\mathrm{E}\mathrm{N}\mathrm{T}\mathrm{I}\mathrm{M}\mathrm{E}\mathrm{N}\mathrm{T}\right)+{\mathsf{\beta }}_4\left(\mathrm{S}\mathrm{I}\mathrm{Z}\mathrm{E}\right)+{\mathsf{\beta }}_5\left(\mathrm{L}\mathrm{E}\mathrm{V}\mathrm{E}\mathrm{R}\mathrm{A}\mathrm{G}\mathrm{E}\right)+{\mathsf{\beta }}_6\left(\mathrm{L}\mathrm{O}\mathrm{S}\mathrm{S}\right)\\ & +{\mathsf{\beta }}_7\left(\mathrm{G}\mathrm{R}\mathrm{O}\mathrm{W}\mathrm{T}\mathrm{H}\right)+{\mathsf{\beta }}_8\left(\mathrm{C}\mathrm{A}\mathrm{S}\mathrm{H}\right)+{\mathsf{\beta }}_9\left(\mathrm{D}\mathrm{I}\mathrm{V}\mathrm{I}\mathrm{D}\mathrm{E}\mathrm{N}\mathrm{D}\right)+{\mathsf{\beta }}_{10}\left(\mathrm{A}\mathrm{N}\mathrm{A}\mathrm{L}\mathrm{Y}\mathrm{S}\mathrm{T}\right)+{\mathsf{\beta }}_{11}\left(\mathrm{B}\mathrm{E}\mathrm{T}\mathrm{A}\right)\\ & +{\mathsf{\beta }}_{12}\left(\mathrm{R}\mathrm{E}\mathrm{S}\mathrm{E}\mathrm{A}\mathrm{R}\mathrm{C}\mathrm{H}\right)+{\mathsf{\beta }}_{14}\left(\mathrm{Y}\mathrm{D}\mathrm{U}\mathrm{M}\right)+{\displaystyle \sum _{\mathrm{C}=1}^{\mathrm{N}-1}}{\mathsf{\gamma }}_{\mathrm{C}}\left(\mathrm{C}\mathrm{D}\mathrm{U}\mathrm{M}\right)+{\displaystyle \sum _{\mathrm{I}=1}^{\mathrm{N}-1}}{\mathsf{\phi }}_{\mathrm{I}}\left(\mathrm{I}\mathrm{D}\mathrm{U}\mathrm{M}\right)+\mathsf{\epsilon }\end{array}$$

(1)

The dependent variable (Q) in the regression equation above gauges a firm’s performance. In line with prior literature, we use Tobin’s Q as an indicator of performance [32,33,34]. The main independent variables (ESG, SENTIMENT, and ESG*SENTIMENT) represent the ESG score of a firm, the COVID-19-induced sentiment for a country, and the interaction between the ESG score and COVID-19-induced sentiment. The ESG score indicates the degree of involvement of a firm in ESG activities (refer to

Table 1. Definition of variables.

Variables	Definition	Source
Q	${\displaystyle \frac{\mathrm{Market} \mathrm{value} \mathrm{of} \mathrm{equity}+\mathrm{Book} \mathrm{value} \mathrm{of} \mathrm{assets}-\mathrm{Book} \mathrm{value} \mathrm{of} \mathrm{equity}}{\mathrm{Book} \mathrm{value} \mathrm{of} \mathrm{assets}}}$	Worldscope
ESG	Environmental, social, and governance (ESG) score.	Refinitiv Eikon
SENTIMENT	The RavenPack Sentiment Index measures the level of sentiment across all news entities that is mentioned alongside coronavirus. Values range between −100 (most negative sentiment) and 100 (most positive sentiment), while 0 is neutral.	RavenPack
SIZE	Natural logarithm of market capitalization	Worldscope
LEVERAGE	${\displaystyle \frac{\mathrm{Total} \mathrm{debt}}{\mathrm{Book} \mathrm{value} \mathrm{of} \mathrm{assets}}}$	Worldscope
LOSS	$\left\{\begin{array}{c}1, \& \mathrm{i}\mathrm{f} \mathrm{e}\mathrm{a}\mathrm{r}\mathrm{n}\mathrm{i}\mathrm{n}\mathrm{g}\mathrm{s} \mathrm{b}\mathrm{e}\mathrm{f}\mathrm{o}\mathrm{r}\mathrm{e} \mathrm{i}\mathrm{n}\mathrm{t}\mathrm{e}\mathrm{r}\mathrm{e}\mathrm{s}\mathrm{t} \mathrm{a}\mathrm{n}\mathrm{d} \mathrm{t}\mathrm{a}\mathrm{x}\mathrm{e}\mathrm{s}<0\\ 0, \& \mathrm{i}\mathrm{f} \mathrm{e}\mathrm{a}\mathrm{r}\mathrm{n}\mathrm{i}\mathrm{n}\mathrm{g}\mathrm{s} \mathrm{b}\mathrm{e}\mathrm{f}\mathrm{o}\mathrm{r}\mathrm{e} \mathrm{i}\mathrm{n}\mathrm{t}\mathrm{e}\mathrm{r}\mathrm{e}\mathrm{s}\mathrm{t} \mathrm{a}\mathrm{n}\mathrm{d} \mathrm{t}\mathrm{a}\mathrm{x}\mathrm{e}\mathrm{s}\ge 0\end{array}\right.$	Worldscope
GROWTH	One-year growth in total sales	Worldscope
CASH	${\displaystyle \frac{\mathrm{Total} \mathrm{cash}}{\mathrm{Book} \mathrm{value} \mathrm{of} \mathrm{assets}}}$	Worldscope
DIVIDEND	$\left\{\begin{array}{c}1, \& \mathrm{i}\mathrm{f} \mathrm{c}\mathrm{a}\mathrm{s}\mathrm{h} \mathrm{d}\mathrm{i}\mathrm{v}\mathrm{i}\mathrm{d}\mathrm{e}\mathrm{n}\mathrm{d}\mathrm{s} \mathrm{p}\mathrm{a}\mathrm{i}\mathrm{d}>0\\ 0, \& \mathrm{i}\mathrm{f} \mathrm{c}\mathrm{a}\mathrm{s}\mathrm{h} \mathrm{d}\mathrm{i}\mathrm{v}\mathrm{i}\mathrm{d}\mathrm{e}\mathrm{n}\mathrm{d}\mathrm{s} \mathrm{p}\mathrm{a}\mathrm{i}\mathrm{d}\le 0\end{array}\right.$	Worldscope
ANALYST	Total number of analysts covering a firm.	Thomson Reuters Eikon
BETA	Sensitivity of stock returns to market returns (market risk).	Datastream
RESEARCH	${\displaystyle \frac{\mathrm{Research} \mathrm{and} \mathrm{development} \mathrm{expenditures}}{\mathrm{Total} \mathrm{sales}}}$	Worldscope
YDUM	Set of year dummies.	Worldscope
CDUM	Set of country dummies.	Worldscope
IDUM	Set of industry dummies based on ICB classification.	Worldscope
GOV	Governance pillar score.	Refinitiv Eikon
ENV	Environmental pillar score.	Refinitiv Eikon
SOC	Social pillar score.	Refinitiv Eikon
PANIC	The Coronavirus Panic Index measures the number of mentions of panic or hysteria and coronavirus. While values can range from 0 to 100, a value of n indicates that n% of news globally is talking about panic and COVID-19.	RavenPack
HYPE	The Coronavirus Hype Index measures the number of mentions of coronavirus. While values can range from 0 to 100, a value of n indicates that n% of news globally is talking about COVID-19.	RavenPack
FAKE	The Coronavirus Fake News Index measures the number of mentions of misinformation or fake news alongside coronavirus. While values can range from 0 to 100, a value of n indicates that n% of news globally is talking about fake news and COVID-19.	RavenPack
INFODEMIC	The Coronavirus Infodemic Index provides the percentage of all entities (places, companies, etc.) that would be in any way linked to COVID-19. While values can range from 0 to 100, a value of n indicates that n% of all media-covered entities are linked to COVID-19.	RavenPack
COVERAGE	The Coronavirus Media Coverage Index provides the percentage of all news sources that cover news about COVID-19. While values can range from 0 to 100, a value of n indicates that n% of all news providers are covering COVID-19-related stories.	RavenPack

). COVID-19-induced sentiment is a country-level indicator that measures the sentiment level in news related to the coronavirus pandemic.

We use the following variables to control the effects that various firm-specific characteristics may have on firm performance. The choice of these variables is driven by prior literature [34]. We use the total debt to total assets ratio (LEVERAGE) and market risk of a stock (BETA) to control the exposure to risk. The information environment of a firm is controlled by the number of analysts following a firm (ANALYST), log of market capitalization in dollars (SIZE), and research and development to total assets ratio (RESEARCH). Furthermore, variables indicating whether earnings before interest and taxes are negative (LOSS), one-year growth in total sales (GROWTH), the cash to total assets ratio (CASH), and whether a firm pays dividends (DIVIDEND) are used to control for various other aspects that may be important for the performance of a firm. In addition to these firm-specific control variables, we also include a set of country dummies (CDUM), a set of industry dummies (IDUM), and a year dummy that takes the value of 1 for 2020 and 0 for 2021 (YDUM) to account for the impact of industry, year, and nation-specific factors on business performance. Table 1 provides the definition of these variables.

3.2.2. Effect of ESG (Environmental Pillar Score) on the Relationship between COVID-19-Induced Sentiment and Firm Performance

The first component of the ESG score is related to the environment. This component (also referred to as the environmental pillar) incorporates three factors in the computation of the ESG score. These factors are related to resource use, emissions, and environmental innovation. The resource utilization factor assesses how well a company performs and how well it can reduce the usage of resources such as water, energy, and materials and find more environmentally friendly solutions by enhancing supply chain management. The emissions-related factor calculates a company’s effectiveness in cutting emissions. It gauges a company’s dedication to and effectiveness in cutting down on emissions from production and operations. Environmental innovation, the final factor, measures a company’s ability to lower environmental costs and burdens for its clients, opening new markets through eco-friendly products or new environmental processes and technologies. To gauge whether the environmental component of the ESG score affects the relationship between COVID-19-induced sentiment and firm performance, Equation (1) is modified as follows. In the following regression, ENV measures the environment-related activities in the ESG score.

$$\begin{array}{ll}\mathrm{Q}=\mathsf{\alpha }+{\mathsf{\beta }}_1\left(\mathrm{E}\mathrm{N}\mathrm{V}\right)& +{\mathsf{\beta }}_2\left(\mathrm{S}\mathrm{E}\mathrm{N}\mathrm{T}\mathrm{I}\mathrm{M}\mathrm{E}\mathrm{N}\mathrm{T}\right)+{\mathsf{\beta }}_3\left(\mathrm{E}\mathrm{N}\mathrm{V}\ast \mathrm{S}\mathrm{E}\mathrm{N}\mathrm{T}\mathrm{I}\mathrm{M}\mathrm{E}\mathrm{N}\mathrm{T}\right)+{\mathsf{\beta }}_4\left(\mathrm{S}\mathrm{I}\mathrm{Z}\mathrm{E}\right)+{\mathsf{\beta }}_5\left(\mathrm{L}\mathrm{E}\mathrm{V}\mathrm{E}\mathrm{R}\mathrm{A}\mathrm{G}\mathrm{E}\right)+{\mathsf{\beta }}_6\left(\mathrm{L}\mathrm{O}\mathrm{S}\mathrm{S}\right)\\ & +{\mathsf{\beta }}_7\left(\mathrm{G}\mathrm{R}\mathrm{O}\mathrm{W}\mathrm{T}\mathrm{H}\right)+{\mathsf{\beta }}_8\left(\mathrm{C}\mathrm{A}\mathrm{S}\mathrm{H}\right)+{\mathsf{\beta }}_9\left(\mathrm{D}\mathrm{I}\mathrm{V}\mathrm{I}\mathrm{D}\mathrm{E}\mathrm{N}\mathrm{D}\right)+{\mathsf{\beta }}_{10}\left(\mathrm{A}\mathrm{N}\mathrm{A}\mathrm{L}\mathrm{Y}\mathrm{S}\mathrm{T}\right)+{\mathsf{\beta }}_{11}\left(\mathrm{B}\mathrm{E}\mathrm{T}\mathrm{A}\right)\\ & +{\mathsf{\beta }}_{12}\left(\mathrm{R}\mathrm{E}\mathrm{S}\mathrm{E}\mathrm{A}\mathrm{R}\mathrm{C}\mathrm{H}\right)+{\mathsf{\beta }}_{14}\left(\mathrm{Y}\mathrm{D}\mathrm{U}\mathrm{M}\right)+{\displaystyle \sum _{\mathrm{C}=1}^{\mathrm{N}-1}}{\mathsf{\gamma }}_{\mathrm{C}}\left(\mathrm{C}\mathrm{D}\mathrm{U}\mathrm{M}\right)+{\displaystyle \sum _{\mathrm{I}=1}^{\mathrm{N}-1}}{\mathsf{\phi }}_{\mathrm{I}}\left(\mathrm{I}\mathrm{D}\mathrm{U}\mathrm{M}\right)+\mathsf{\epsilon }\end{array}$$

(2)

3.2.3. Effect of ESG (Social Pillar Score) on the Relationship between COVID-19-Induced Sentiment and Firm Performance

The social activities of a company are the subject of the second ESG score component. This component (also referred to as the social pillar) incorporates four factors in the computation of the ESG score. These factors are related to human rights, community, workforce, and product responsibility. A company’s efficacy in upholding core human rights conventions is gauged by the human rights factor. The community component gauges a company’s dedication to upholding good citizenship. It shows how dedicated a company is to upholding business ethics and preserving public health. The workforce factor gauges how well a company performs in terms of employee job satisfaction. Businesses that provide equal opportunities, diversity, and a safe and healthy work environment, as well as opportunities for staff development, receive high marks in this category. The final component, product responsibility, has to do with a company’s ability to manufacture high-quality goods and services while incorporating data privacy and integrity. To assess whether the social component of the ESG score affects the relationship between COVID-19-induced sentiment and firm performance, we modify Equation (1) as follows. In the following regression, SOC measures the social activities in the ESG score.

$$\begin{array}{ll}\mathrm{Q}=\mathsf{\alpha }+{\mathsf{\beta }}_1\left(\mathrm{S}\mathrm{O}\mathrm{C}\right)& +{\mathsf{\beta }}_2\left(\mathrm{S}\mathrm{E}\mathrm{N}\mathrm{T}\mathrm{I}\mathrm{M}\mathrm{E}\mathrm{N}\mathrm{T}\right)+{\mathsf{\beta }}_3\left(\mathrm{S}\mathrm{O}\mathrm{C}\ast \mathrm{S}\mathrm{E}\mathrm{N}\mathrm{T}\mathrm{I}\mathrm{M}\mathrm{E}\mathrm{N}\mathrm{T}\right)+{\mathsf{\beta }}_4\left(\mathrm{S}\mathrm{I}\mathrm{Z}\mathrm{E}\right)+{\mathsf{\beta }}_5\left(\mathrm{L}\mathrm{E}\mathrm{V}\mathrm{E}\mathrm{R}\mathrm{A}\mathrm{G}\mathrm{E}\right)+{\mathsf{\beta }}_6\left(\mathrm{L}\mathrm{O}\mathrm{S}\mathrm{S}\right)\\ & +{\mathsf{\beta }}_7\left(\mathrm{G}\mathrm{R}\mathrm{O}\mathrm{W}\mathrm{T}\mathrm{H}\right)+{\mathsf{\beta }}_8\left(\mathrm{C}\mathrm{A}\mathrm{S}\mathrm{H}\right)+{\mathsf{\beta }}_9\left(\mathrm{D}\mathrm{I}\mathrm{V}\mathrm{I}\mathrm{D}\mathrm{E}\mathrm{N}\mathrm{D}\right)+{\mathsf{\beta }}_{10}\left(\mathrm{A}\mathrm{N}\mathrm{A}\mathrm{L}\mathrm{Y}\mathrm{S}\mathrm{T}\right)+{\mathsf{\beta }}_{11}\left(\mathrm{B}\mathrm{E}\mathrm{T}\mathrm{A}\right)\\ & +{\mathsf{\beta }}_{12}\left(\mathrm{R}\mathrm{E}\mathrm{S}\mathrm{E}\mathrm{A}\mathrm{R}\mathrm{C}\mathrm{H}\right)+{\mathsf{\beta }}_{14}\left(\mathrm{Y}\mathrm{D}\mathrm{U}\mathrm{M}\right)+{\displaystyle \sum _{\mathrm{C}=1}^{\mathrm{N}-1}}{\mathsf{\gamma }}_{\mathrm{C}}\left(\mathrm{C}\mathrm{D}\mathrm{U}\mathrm{M}\right)+{\displaystyle \sum _{\mathrm{I}=1}^{\mathrm{N}-1}}{\mathsf{\phi }}_{\mathrm{I}}\left(\mathrm{I}\mathrm{D}\mathrm{U}\mathrm{M}\right)+\mathsf{\epsilon }\end{array}$$

(3)

3.2.4. Effect of ESG (Governance Pillar Score) on the Relationship between COVID-19-Induced Sentiment and Firm Performance

The last component of the ESG score is related to a firm’s governance structure. This component (also referred to as the governance pillar) incorporates three factors in the computation of the ESG score. These factors are related to management, shareholders, and CSR strategy. The management component assesses how well a company adheres to best practice corporate governance principles and how committed it is to doing so. The element pertaining to shareholders has to do with how well a company uses anti-takeover mechanisms and treats its shareholders equally. The final component, CSR strategy, represents how a company communicates the fact that it incorporates financial, social, and environmental aspects into its daily decision-making processes. To assess whether the governance component of the ESG score affects the relationship between COVID-19-induced sentiment and firm performance, we change Equation (1) as follows. In the following regression, GOV measures the governance pillar in the ESG score.

$$\begin{array}{ll}\mathrm{Q}=\mathsf{\alpha }+{\mathsf{\beta }}_1\left(\mathrm{G}\mathrm{O}\mathrm{V}\right)& +{\mathsf{\beta }}_2\left(\mathrm{S}\mathrm{E}\mathrm{N}\mathrm{T}\mathrm{I}\mathrm{M}\mathrm{E}\mathrm{N}\mathrm{T}\right)+{\mathsf{\beta }}_3\left(\mathrm{G}\mathrm{O}\mathrm{V}\ast \mathrm{S}\mathrm{E}\mathrm{N}\mathrm{T}\mathrm{I}\mathrm{M}\mathrm{E}\mathrm{N}\mathrm{T}\right)+{\mathsf{\beta }}_4\left(\mathrm{S}\mathrm{I}\mathrm{Z}\mathrm{E}\right)+{\mathsf{\beta }}_5\left(\mathrm{L}\mathrm{E}\mathrm{V}\mathrm{E}\mathrm{R}\mathrm{A}\mathrm{G}\mathrm{E}\right)+{\mathsf{\beta }}_6\left(\mathrm{L}\mathrm{O}\mathrm{S}\mathrm{S}\right)\\ & +{\mathsf{\beta }}_7\left(\mathrm{G}\mathrm{R}\mathrm{O}\mathrm{W}\mathrm{T}\mathrm{H}\right)+{\mathsf{\beta }}_8\left(\mathrm{C}\mathrm{A}\mathrm{S}\mathrm{H}\right)+{\mathsf{\beta }}_9\left(\mathrm{D}\mathrm{I}\mathrm{V}\mathrm{I}\mathrm{D}\mathrm{E}\mathrm{N}\mathrm{D}\right)+{\mathsf{\beta }}_{10}\left(\mathrm{A}\mathrm{N}\mathrm{A}\mathrm{L}\mathrm{Y}\mathrm{S}\mathrm{T}\right)+{\mathsf{\beta }}_{11}\left(\mathrm{B}\mathrm{E}\mathrm{T}\mathrm{A}\right)\\ & +{\mathsf{\beta }}_{12}\left(\mathrm{R}\mathrm{E}\mathrm{S}\mathrm{E}\mathrm{A}\mathrm{R}\mathrm{C}\mathrm{H}\right)+{\mathsf{\beta }}_{14}\left(\mathrm{Y}\mathrm{D}\mathrm{U}\mathrm{M}\right)+{\displaystyle \sum _{\mathrm{C}=1}^{\mathrm{N}-1}}{\mathsf{\gamma }}_{\mathrm{C}}\left(\mathrm{C}\mathrm{D}\mathrm{U}\mathrm{M}\right)+{\displaystyle \sum _{\mathrm{I}=1}^{\mathrm{N}-1}}{\mathsf{\phi }}_{\mathrm{I}}\left(\mathrm{I}\mathrm{D}\mathrm{U}\mathrm{M}\right)+\mathsf{\epsilon }\end{array}$$

(4)

3.2.5. Alternative Proxies

To better examine how ESG moderates the influence of sentiment driven by the COVID-19 pandemic on corporate performance, we use three alternative proxies that significantly influenced investor sentiment during this period. The first proxy, PANIC, quantifies the frequency of mentions of panic or hysteria in conjunction with the coronavirus, with a range from 0 to 100. A high value on the PANIC index is likely to reflect heightened fear and uncertainty among investors. Incorporating PANIC as an alternative proxy for sentiment is essential because traditional sentiment indicators often fail to capture extreme emotional reactions that drive market behavior during crises. PANIC thus reflects the heightened anxiety and fear among investors, which can lead to irrational decision-making and significant market volatility.

The second sentiment proxy, HYPE, tracks the frequency of coronavirus mentions in the news, with possible values ranging from 0 to 100. HYPE can provide more accurate insights into market dynamics during unprecedented events. During the coronavirus pandemic, social media and news platforms amplified negative news of rising infection rates, economic shutdowns, and the struggles of healthcare systems, contributing to a pervasive sense of pessimism. Conversely, positive news about potential treatments, vaccines, or economic recovery drove optimism in the markets. News significantly influenced market sentiment, leading to extreme volatility not fully explained by fundamental financial indicators.

The third proxy, FAKE, captures the amount of fake news related to the coronavirus crisis, with index values ranging from 0 to 100. Fake news circulating during this period had a significant impact on investor sentiment, often leading to misinformation-driven market movements. Misleading or false information about the severity of the pandemic, the effectiveness of treatments, or the timeline for economic recovery could create unrealistic optimism or undue pessimism among investors. Incorporating FAKE as an alternative proxy for sentiment during COVID is thus crucial since traditional sentiment measures may not fully capture the impact of misinformation and fake news on market behavior.

3.3. Summary Statistics

Table 2. Average values of main variables.

Country	Observations	Q	ESG	Sentiment
Argentina	47	1.1449	35.6017	−11.8003
Australia	592	2.2011	42.2187	−21.3294
Austria	53	1.2431	60.2664	−6.0719
Bahrain	6	1.2223	36.0764	−2.6515
Belgium	70	1.6151	56.2275	−7.4202
Brazil	124	1.6760	51.4253	−20.0186
Canada	613	1.8715	43.6666	−10.1636
Chile	58	1.0196	54.8048	−5.3498
China	1509	2.8026	36.3631	−23.8148
Colombia	21	0.9939	60.7101	−6.3020
Denmark	99	3.0990	55.5032	−5.6136
Finland	139	2.1047	53.0169	−4.8860
France	274	1.5864	62.0972	−7.8795
Germany	394	2.0940	54.3219	−19.4506
Greece	30	1.3485	50.6942	−11.8761
Hungary	10	1.1398	48.0607	−4.5836
India	292	3.2308	51.7419	−25.6103
Indonesia	51	2.1676	45.9904	−7.7992
Ireland	60	1.9743	59.3178	−12.5943
Israel	38	2.6999	38.8687	−15.8346
Italy	157	1.8148	57.4867	−20.4721
Japan	779	1.5670	52.7178	−16.4917
Kuwait	14	1.1123	34.5221	−2.2264
Malaysia	286	1.8076	43.9480	−8.9924
Mexico	88	1.4892	52.5764	−9.3302
Netherlands	108	2.6557	63.3214	−7.3514
New Zealand	97	1.9692	41.8000	−7.4709
Norway	108	2.0698	50.6784	−4.2916
Oman	8	0.8991	20.4912	−3.5977
Peru	36	1.0766	48.2114	−6.8424
Philippines	39	1.4885	51.3017	−12.5372
Poland	41	1.4520	46.4526	−6.3962
Portugal	17	1.2454	69.8274	−7.9063
Qatar	50	1.2445	21.6552	−2.1586
Russia	40	1.5479	57.0731	−14.4957
Saudi Arabia	52	2.2235	32.5627	−7.1205
Singapore	143	1.2065	51.2087	−17.2673
South Africa	185	1.2682	53.2112	−13.2762
South Korea	147	1.5957	50.9249	−25.3065
Spain	95	1.6491	68.3889	−20.1514
Sweden	493	3.0052	45.3152	−2.7029
Switzerland	274	2.6105	47.8726	1.7287
Taiwan	240	1.9929	57.6065	−4.3741
Thailand	252	2.0038	51.2693	−10.9671
Turkey	127	1.4985	62.2897	−4.6206
United States	3895	2.6517	43.8968	−3.9439
United Arab Emirates	35	1.5736	38.3202	−7.9913
United Kingdom	785	2.0112	49.7494	−10.2675
Vietnam	31	1.9518	27.3475	−0.7197

Note: The average firm performance (Q), ESG score (ESG), and COVID-19-induced sentiment index (SENTIMENT) values are shown in the table with the whole number of observations for every nation. The years 2020 and 2021 make up the sample. For the description of the variables, refer to Table 1.

lists the total number of observations as well as the average values of the three main variables (SENTIMENT, ESG, and Q) for each country. The table demonstrates that ESG activities differ significantly between nations. For example, firms headquartered in Oman, Qatar, and Vietnam have the lowest score for ESG activities. The average values of ESG scores for firms headquartered in these countries are less than 28. This contrasts with firms headquartered in the Netherlands, Spain, and Portugal, which score relatively high for ESG activities. The average values of ESG scores for firms headquartered in these countries are greater than 62. This table indicates an intriguing finding, namely that the majority of firms headquartered in emerging markets have low values for their ESG scores. A similar divergence is observed in the level of COVID-19-induced sentiment across countries. India, South Korea, and China have the most negative sentiments, with average values of less than −24, while Vietnam, Qatar, and Switzerland have the least negative sentiments, with average values of greater than −0.8. In the case of firm performance, the table shows that firms headquartered in Oman, Colombia, and Chile have the lowest performance. In contrast, firms headquartered in India, Denmark, and Sweden perform the best. The average value of Tobin’s Q in the worst-performing countries is less than 1.02, while it is greater than 3 in the best-performing countries.

Table 3. Descriptive statistics.

Variables	25th Percentile	Mean	Median	75th Percentile	Standard Deviation	Observations
Q	1.0742	2.2806	1.4951	2.5666	2.0909	13,102
ESG	29.8826	46.7051	45.8811	63.1595	20.8059	13,102
SENTIMENT	−17.9872	−10.9183	−7.4893	−4.5965	8.8678	13,102
SIZE	13.2599	14.4929	14.5638	15.6578	1.7913	13,102
LEVERAGE	10.7200	25.9350	25.1200	38.4400	18.4749	13,102
LOSS	0	0.2188	0	0	0.4135	13,102
GROWTH	−5.9700	10.5069	4.6400	19.1200	41.2647	13,102
CASH	0.0410	0.1417	0.0962	0.1857	0.1504	13,102
DIVIDEND	0	0.6504	1	1	0.4769	13,102
ANALYST	3	8.4911	6	12	7.7966	13,102
BETA	0.7193	1.1025	1.0471	1.4177	0.5863	13,102
RESEARCH	0	0.0823	0	0.0270	0.4876	13,102

Note: The descriptive statistics for the variables used in this paper are shown in the table. Table 1 shows the definitions of the variables.

documents the descriptive statistics of our study. The table indicates that the average value of the total debt to total assets ratio (LEVERAGE) is a little less than 26%, while it is around 1.1 for market risk (BETA). Both values indicate that the firms included in this analysis do not have excessive risk. Similarly, the average values of analyst coverage (ANALYST), propensity to pay dividends (DIVIDEND), and research and development expenditures (RESEARCH) show lower information asymmetries. These values indicate that an average firm is followed by 8.49 analysts and spends around 8% of sales on research. These values also indicate that 65.04% of firms pay dividends. Furthermore, the table shows that the average value of firm performance (Q), as measured by Tobin’s Q, is greater than 2.2. The average growth rate (GROWTH) is more than 10% and only 21.88% of firms generated losses. These characteristics are an indication that the firms included in this analysis are well-performing firms.

Table 4. Correlation matrix.

No.	Variables	1	2	3	4	5	6	7	8	9
1	SIZE	1.0000
2	LEVERAGE	0.0797	1.0000
3	LOSS	−0.3145	−0.0337	1.0000
4	GROWTH	0.0614	−0.0763	−0.0845	1.0000
5	CASH	−0.1372	−0.3293	0.2662	0.0924	1.0000
6	DIVIDEND	0.3512	0.0503	−0.4992	−0.0491	−0.2660	1.0000
7	ANALYST	0.7050	0.0410	−0.1102	0.0064	−0.0557	0.1715	1.0000
8	BETA	−0.0605	0.0121	0.1984	0.0380	0.0890	−0.2081	0.0234	1.0000
9	RESEARCH	−0.0693	−0.1176	0.2268	0.1021	0.2503	−0.1777	−0.0088	0.0690	1.0000

Note: The correlation between the variables utilized in this paper is shown in the table. The years 2020 and 2021 make up the sample.

documents the relationships among the variables used for the analysis. The table indicates that the highest correlation exists between analyst coverage and the size of a firm. This value is around 0.70. Although this value is relatively high, it does not increase VIF values to an extent that multicollinearity poses problems. The VIF values for all these variables are less than 10.

4. Results

4.1. Effect of ESG on the Relationship between COVID-19-Induced Sentiment and Firm Performance

The results derived from the estimation of Equation (1) are presented in

Table 5. Effect of ESG on the relationship between COVID-19-induced sentiment and firm performance.

Variables	Model (1)	Model (2)	Model (3)	Model (4)	Model (5)	Model (6)
ESG	−0.0199 ***	−0.0199 ***	−0.0017	−0.0209 ***	−0.0197 *	−0.0233 ***
	(−13.8929)	(−5.1634)	(−0.8029)	(−14.8534)	(−11.7302)	(−14.7995)
SENTIMENT	−0.0170 ***	−0.0170 **	−0.0162 ***	−0.0154 ***	−0.0548 **	−0.0180 ***
	(−2.8668)	(−2.2979)	(−3.7857)	(−3.7835)	(−2.6291)	(−2.9487)
ESG*SENTIMENT	0.0003 ***	0.0003 **	0.0003 ***	0.0002 ***	0.0003 *	0.0003 ***
	(3.5165)	(2.0182)	(3.9851)	(3.3657)	(8.0822)	(3.4055)
SIZE	0.5821 ***	0.5821 ***	1.4205 ***	0.7515 ***	0.5738 **	0.5993 ***
	(31.603)	(6.9847)	(26.4564)	(32.6895)	(25.0682)	(37.1261)
LEVERAGE	−0.0158 ***	−0.0158 ***	0.0001	−0.0159 ***	−0.0156 **	−0.0150 ***
	(−17.4148)	(−5.3777)	(0.0276)	(−13.0551)	(−19.6073)	(−15.9505)
LOSS	0.0153	0.0153	0.1039 ***	0.1489 ***	−0.0067	−0.0069
	(0.3523)	(0.2721)	(3.3799)	(5.1152)	(−0.0805)	(−0.1467)
GROWTH	0.0030 ***	0.0030 **	−0.0008 **	0.0009 **	0.0032	0.0024 ***
	(6.3259)	(2.4097)	(−2.1613)	(2.4400)	(2.5292)	(5.9191)
CASH	3.2775 ***	3.2775 ***	0.1385	2.0713 ***	3.2983 *	3.1766 ***
	(20.0549)	(12.7241)	(0.4609)	(11.1331)	(7.3506)	(25.5837)
DIVIDEND	−0.3024 ***	−0.3024 ***	−0.0545 **	−0.2392 ***	−0.2990	−0.3099 ***
	(−7.6054)	(−3.2342)	(−2.0018)	(−8.5619)	(−4.9163)	(−7.3146)
ANALYST	−0.0048	−0.0048	−0.0371 ***	−0.0308 ***	−0.0029	−0.0027
	(−1.236)	(−0.2286)	(−4.1390)	(−6.6118)	(−1.4068)	(−0.8362)
BETA	0.0038	0.0038	−0.2128 ***	0.0362	0.0029	0.0224
	(0.1218)	(0.0518)	(−2.6922)	(0.9758)	(0.1327)	(0.7450)
RESEARCH	−0.0775 **	−0.0775 *	−0.1460 ***	−0.0893 **	−0.0749	−0.0620 *
	(−2.4028)	(−1.7507)	(−3.1356)	(−2.3416)	(−4.7724)	(−1.8361)
Country Dummies	Yes	Yes	No	Yes	Yes	Yes
Industry Dummies	Yes	Yes	No	Yes	Yes	Yes
Year Dummy	Yes	Yes	Yes	Yes	No	Yes
Observations	13,102	13,102	13,102	13,102	13,102	11,615
R-Square	0.3544	0.3544	0.3160	0.2137	0.3585	0.3515
Durbin Score						96.7511 ***
Wu–Hausman						48.4585 ***
Eigen Value						25902.7

Note: The effect of ESG on the link between firm performance and COVID-19-induced sentiment is shown in this table. Parentheses surround the t-values. Firm performance (Q) is the outcome variable, while SENTIMENT (the COVID-19-induced sentiment index), ESG (the ESG scores), and ESG*SENTIMENT are the main independent variables. The years 2020 and 2021 make up the sample. Pooled OLS regression with robust standard errors is the foundation of Model (1); pooled OLS regression with standard errors clustered at the country level is the foundation of Model (2); panel regression with fixed effects and robust standard errors is the foundation of Model (3); panel regression with random effects and robust standard errors is the foundation of Model (4); Fama–MacBeth regression is the foundation of Model (5); and instrumental variable regression is the foundation of Model (6). p < 0.1, p < 0.05, and p < 0.01 are represented by the characters *, **, and ***, respectively.

. The main variable of interest in this table is ESG*SENTIMENT. Column (2) reports the results based on pooled ordinary least squares (OLS) regression. Based on strong standard errors, the t-values—as reported in parenthesis—are used to correct for heteroscedasticity. The results show that the coefficient estimate of ESG*SENTIMENT is significant and positive. This indicates that the benefits of improved COVID-19-induced sentiment accrue more to firms that rank high on ESG scores. More specifically, the findings show that the performance of firms that expend more resources on ESG activities benefits more from positive sentiments than that of firms less committed to ESG activities. These findings are consistent with the assumption that the need for ESG activities is more important during periods of uncertainty such as the COVID-19 pandemic. We believe that ESG activities can act as reputation-building mechanisms. Therefore, in uncertain times, these activities can help firms reap early benefits linked to changes in sentiments. Our findings are consistent with Lins et al. [3], who argue that the trust between investors and firms, built through the engagement of firms in ESG activities, pays off when firms face negative market shocks.

In addition to pooled OLS regression, we also use several alternate estimation procedures to estimate Equation (1). Column (3) uses pooled OLS regression with standard errors clustered at the country level. Such clustering of standard errors accounts for heteroscedasticity across countries. Column (4) and Column (5) use panel regressions with random and fixed effects, respectively, to account for time-series dependencies. Column (6) presents findings based on the Fama–MacBeth regression. As the first-step estimates have significantly milder autocorrelations than the autocorrelations (time trends) inside the observations, this regression avoids the false regression problem. Regression analysis of instrumental variables is used in Column (7) to address endogeneity issues with the ESG variable. The instruments used are the lagged values of ESG. The post-estimation findings suggest that the instruments are valid. The findings based on alternate estimation strategies are consistent with those presented in Column (2). The findings show a significantly positive coefficient estimate of ESG*SENTIMENT. These findings further support the hypothesis presented in this paper by indicating that the benefits of improved COVID-19-induced sentiment accrue more to firms that rank high on ESG scores. Hence, ESG investments can be considered as strategic decisions that can exert a significant influence on the performance of corporations during periods of turmoil.

To facilitate a better understanding of the findings observed in Table 5, Figure 1 shows the average marginal effect of COVID-19-induced sentiment (SENTIMENT) on firm performance (Q) at different levels of ESG scores. The figure shows that as ESG activities increase, the marginal impact of COVID-19-induced sentiment on firm performance also increases and becomes positive. Our findings imply that firms can use ESG activities as strategic decisions to protect themselves from rapidly changing market conditions.

4.2. Effect of ESG on the Relationship between COVID-19-Induced Sentiment and Firm Performance: Sample Reconstruction

Table 2 shows that more than 50% of our observations belong to four countries—the United States (3895 observations/29.72%), China (1509 observations/11.51%), the United Kingdom (785 observations/5.99%), and Japan (779 observations/5.94%). Such a marked clustering of observations in a few countries can drive the findings presented above. To address this concern, we removed each of these four countries one-by-one and then removed all four from our analysis.

Table 6. Effect of ESG on the relationship between COVID-19-induced sentiment and firm performance: sample reconstruction based on countries.

Variables	Excluding United States	Excluding China	Excluding Japan	Excluding United Kingdom	Excluding All
ESG	−0.0163 ***	−0.0176 ***	−0.0203 ***	−0.0196 ***	−0.0123 ***
	(−9.4919)	(−12.2135)	(−13.572)	(−13.392)	(−6.3555)
SENTIMENT	−0.0209 ***	−0.0135 **	−0.0181 ***	−0.0161 ***	−0.0129
	(−3.2741)	(−2.1422)	(−2.6981)	(−2.6835)	(−1.6078)
ESG*SENTIMENT	0.0004 ***	0.0002 **	0.0004 ***	0.0003 ***	0.0003 **
	(3.8824)	(2.3156)	(4.0533)	(3.4996)	(2.4556)
SIZE	0.4684 ***	0.6176 ***	0.5929 ***	0.5876 ***	0.5145 ***
	(23.7494)	(32.2332)	(31.0661)	(31.0664)	(22.7354)
LEVERAGE	−0.0164 ***	−0.0127 ***	−0.0164 ***	−0.0163 ***	−0.0138 ***
	(−15.5256)	(−13.7886)	(−17.3708)	(−17.3011)	(−11.1586)
LOSS	−0.0101	0.0512	0.0173	0.0265	0.0400
	(−0.2081)	(1.1359)	(0.3848)	(0.5741)	(0.6597)
GROWTH	0.0040 ***	0.0018 ***	0.0029 ***	0.0030 ***	0.0015 **
	(6.3288)	(3.9424)	(6.1955)	(6.1520)	(2.3447)
CASH	3.6921 ***	3.3054 ***	3.3211 ***	3.2492 ***	4.0217 ***
	(16.6568)	(19.7252)	(19.6711)	(19.2832)	(14.8984)
DIVIDEND	−0.2013 ***	−0.3670 ***	−0.2923 ***	−0.3268 ***	−0.2572 ***
	(−4.2176)	(−8.9762)	(−7.2578)	(−7.7861)	(−4.6376)
ANALYST	0.0081 *	−0.0278 ***	−0.0050	−0.0034	−0.0261 ***
	(1.7837)	(−6.8718)	(−1.2657)	(−0.8790)	(−5.0741)
BETA	−0.0403	−0.0676 **	0.0030	0.0150	−0.1912 ***
	(−0.9671)	(−2.2266)	(0.0952)	(0.4623)	(−4.6745)
RESEARCH	−0.1181 *	−0.0672 **	−0.0802 **	−0.0840 ***	−0.1653 ***
	(−1.8961)	(−2.0785)	(−2.4815)	(−2.6096)	(−2.8903)
Country Dummies	Yes	Yes	Yes	Yes	Yes
Industry Dummies	Yes	Yes	Yes	Yes	Yes
Year Dummy	Yes	Yes	Yes	Yes	Yes
Observations	9207	11,593	12,323	12,317	6134
R-Square	0.3497	0.3606	0.3531	0.3564	0.3543

Note: This table shows how ESG affects the link between COVID-19-induced sentiment and firm performance in various subsamples. Parentheses surround the t-values derived from the heteroscedasticity-robust standard errors. Firm performance (Q) is the outcome variable, while SENTIMENT (the COVID-19-induced sentiment index), ESG (the ESG scores), and ESG*SENTIMENT are the main independent variables. The years 2020 and 2021 make up the sample. Regression using pooled or OLS is applied. p < 0.1, p < 0.05, and p < 0.01 are represented by the characters *, **, and ***, respectively.

presents the analysis’s conclusions. We demonstrate that the coefficient estimate of ESG*SENTIMENT is significant and positive in each subsample, which is in line with earlier findings.

We perform a similar sample reconstruction based on industry classifications. In an unreported result, we find that around 50% of our observations belong to the Industrials (2392 observations/18.26%), Consumer Cyclicals (2076 observations/15.84%), and Technology (2041 observations/15.58%) sectors. We re-estimate Equation (1), as previously, by removing each of these three sectors individually in separate regressions as well as by removing them all at once.

Table 7. Effect of ESG on the relationship between COVID-19-induced sentiment and firm performance: sample reconstruction based on sectors.

Variables	Excluding Consumer Cyclicals	Excluding Industrials	Excluding Technology	Excluding All
ESG	−0.0204 ***	−0.0198 ***	−0.0185 ***	−0.0179 ***
	(−12.7987)	(−12.5758)	(−12.6499)	(−9.5647)
SENTIMENT	−0.0184 ***	−0.0188 ***	−0.0119 *	−0.0131 *
	(−2.8117)	(−2.8148)	(−1.9548)	(−1.6647)
ESG*SENTIMENT	0.0004 ***	0.0003 ***	0.0002 **	0.0003 ***
	(3.7328)	(3.3991)	(2.5596)	(2.5791)
SIZE	0.5918 ***	0.5935 ***	0.5312 ***	0.5353 ***
	(29.242)	(28.7621)	(27.2394)	(20.7987)
LEVERAGE	−0.0175 ***	−0.0153 ***	−0.0142 ***	−0.0149 ***
	(−16.8364)	(−15.0756)	(−15.4348)	(−11.7288)
LOSS	0.0209	0.0313	−0.0053	0.0446
	(0.4160)	(0.6280)	(−0.1210)	(0.6944)
GROWTH	0.0026 ***	0.0023 ***	0.0025 ***	0.0011 **
	(5.2874)	(4.8902)	(5.3174)	(2.2137)
CASH	3.1876 ***	3.2340 ***	3.0688 ***	2.7099 ***
	(18.3689)	(18.4422)	(17.1041)	(12.7386)
DIVIDEND	−0.3029 ***	−0.3809 ***	−0.2635 ***	−0.3774 ***
	(−6.5549)	(−8.2731)	(−6.5856)	(−6.2507)
ANALYST	−0.0023	−0.0056	−0.0069	−0.0064
	(−0.5092)	(−1.3314)	(−1.6240)	(−1.0583)
BETA	0.0029	−0.0010	−0.0222	−0.0433
	(0.0841)	(−0.0289)	(−0.6771)	(−1.0616)
RESEARCH	−0.0933 ***	−0.0851 ***	−0.0902 ***	−0.1095 ***
	(−2.8908)	(−2.673)	(−2.7717)	(−3.3934)
Country Dummies	Yes	Yes	Yes	Yes
Industry Dummies	Yes	Yes	Yes	Yes
Year Dummy	Yes	Yes	Yes	Yes
Observations	11,026	10,710	11,061	6593
R-Square	0.3633	0.3639	0.3376	0.3683

Note: This table shows how ESG affects the link between COVID-19-induced sentiment and firm performance. Parentheses surround the t-values derived from the heteroscedasticity-robust standard errors. Firm performance (Q) is the outcome variable, while SENTIMENT (the COVID-19-induced sentiment index), ESG (the ESG scores), and ESG*SENTIMENT are the main independent variables. The years 2020 and 2021 make up the sample. Regression using pooled OLS is employed. p < 0.1, p < 0.05, and p < 0.01 are represented by the symbols *, **, and ***, respectively. For a definition of the variables, see Table 1.

details the outcomes of these regressions. Our findings align with previous results and provide evidence that the coefficient estimate of ESG*SENTIMENT is significant and positive in all subsamples. This further highlights our main argument, namely that the positive effect of COVID-19-induced sentiment on business performance is more noticeable for firms that focus their efforts on maximizing their ESG-related outcomes.

4.3. Effect of ESG (Environmental Pillar Score) on the Relationship between COVID-19-Induced Sentiment and Firm Performance

Table 8. Effect of ESG (environmental pillar score).

Variables	ENV = Overall Score for Environment	ENV = Resource Use Score	ENV = Emissions Score	ENV = Environmental Innovation Score
ENV	−0.0138 ***	−0.0101 ***	−0.0093 ***	−0.0045 ***
	(−13.0354)	(−11.387)	(−9.7710)	(−5.2619)
SENTIMENT	−0.0147 ***	−0.0125 ***	−0.0159 ***	−0.0097 **
	(−2.8978)	(−2.6348)	(−3.1062)	(−2.2361)
ENV*SENTIMENT	0.0003 ***	0.0002 ***	0.0003 ***	0.0002 ***
	(4.1869)	(3.6907)	(4.2857)	(3.0669)
SIZE	0.5727 ***	0.5550 ***	0.5530 ***	0.4969 ***
	(31.5012)	(30.6646)	(30.6593)	(28.6477)
LEVERAGE	−0.0156 ***	−0.0159 ***	−0.0161 ***	−0.0173 ***
	(−17.1639)	(−17.4131)	(−17.5513)	(−18.6283)
LOSS	0.0303	0.0206	0.0199	0.0290
	(0.6955)	(0.4727)	(0.4539)	(0.6565)
GROWTH	0.0030 ***	0.0031 ***	0.0033 ***	0.0035 ***
	(6.3671)	(6.5143)	(6.8467)	(7.2575)
CASH	3.2462 ***	3.2752 ***	3.2901 ***	3.4084 ***
	(19.7079)	(19.862)	(19.8498)	(20.4497)
DIVIDEND	−0.3043 ***	−0.3173 ***	−0.3241 ***	−0.3518 ***
	(−7.6187)	(−7.9238)	(−8.0139)	(−8.6202)
ANALYST	−0.0063	−0.0071 *	−0.0084 **	−0.0149 ***
	(−1.6370)	(−1.8364)	(−2.1626)	(−3.8455)
BETA	−0.0079	−0.0161	−0.0167	−0.0174
	(−0.2505)	(−0.5111)	(−0.5301)	(−0.5485)
RESEARCH	−0.0880 ***	−0.0863 ***	−0.0901 ***	−0.0636 *
	(−2.7145)	(−2.6465)	(−2.7428)	(−1.8823)
Country Dummies	Yes	Yes	Yes	Yes
Industry Dummies	Yes	Yes	Yes	Yes
Year Dummy	Yes	Yes	Yes	Yes
Observations	13,102	13,102	13,102	13,102
R-Square	0.3529	0.3467	0.3453	0.3322

Note: This table shows the impact of ESG (environmental pillar score) on the relationship between COVID-19-induced sentiment and firm performance. Parentheses surround the t-values derived from the heteroscedasticity-robust standard errors. Firm performance (Q) is the outcome variable, while SENTIMENT (the COVID-19-induced sentiment index), ENV (the environmental pillar scores), and ENV*SENTIMENT are the main independent variables. The years 2020 and 2021 make up the sample. Regression using pooled OLS is employed. p < 0.1, p < 0.05, and p < 0.01 are represented by the symbols *, **, and ***, respectively.

presents the findings from the aforementioned analysis. Column (2) uses the overall score for the environmental pillar to define ENV, and Columns (3) to (5) use individual components of the environmental pillar (resource use, emissions, and environmental innovation) as proxies to define ENV. The findings indicate that the environmental pillar, as well as the individual components of the environmental pillar, affect the relationship between COVID-19-induced sentiment and firm performance. We report a significantly positive coefficient estimate of ENV*SENTIMENT. These results suggest that environmental considerations were crucial during the pandemic period for demonstrating sustainable practices as well as improving the financial performance of firms.

To gain a better understanding of the findings observed in Table 8, we also plot the average marginal impact of COVID-19-induced sentiment (SENTIMENT) on the performance of firms (Q) conditional on environmental activities undertaken by firms (ENV) in Figure 2. The figure shows that as environmental activities increase, the marginal effect of COVID-19-induced sentiment on firms’ performance also increases and becomes positive.

4.4. Effect of ESG (Social Pillar Score) on the Relationship between COVID-19-Induced Sentiment and Firm Performance

Table 9. Effect of ESG (social pillar score).

Variables	SOC = Overall Score for Social	SOC = Human Rights Score	SOC = Community Score	SOC = Workforce Score	SOC = Product Responsibility Score
SOC	−0.0142 ***	−0.0061 ***	−0.0111 ***	−0.0078 ***	−0.0045 ***
	(−11.3604)	(−7.6603)	(−10.8267)	(−7.4569)	(−4.8539)
SENTIMENT	−0.0155 ***	−0.0140 ***	−0.0067	−0.0157 ***	−0.0115 **
	(−2.8304)	(−3.2235)	(−1.3304)	(−2.806)	(−2.2776)
SOC*SENTIMENT	0.0003 ***	0.0003 ***	0.0001	0.0002 ***	0.0001 **
	(3.3773)	(4.8848)	(0.8793)	(3.1198)	(2.3469)
SIZE	0.5582 ***	0.5293 ***	0.5361 ***	0.5281 ***	0.4900 ***
	(30.5458)	(29.7466)	(29.8417)	(29.4601)	(28.2718)
LEVERAGE	−0.0163 ***	−0.0169 ***	−0.0167 ***	−0.0166 ***	−0.0172 ***
	(−17.8599)	(−18.2825)	(−18.1506)	(−18.0735)	(−18.4739)
LOSS	0.0244	0.0191	0.0246	0.0172	0.0128
	(0.5595)	(0.4344)	(0.5596)	(0.3897)	(0.2902)
GROWTH	0.0032 ***	0.0034 ***	0.0033 ***	0.0035 ***	0.0035 ***
	(6.7128)	(6.9901)	(6.9572)	(7.2749)	(7.2022)
CASH	3.3491 ***	3.3282 ***	3.3674 ***	3.3858 ***	3.4042 ***
	(20.4458)	(20.17)	(20.5143)	(20.3591)	(20.4604)
DIVIDEND	−0.3302 ***	−0.3488 ***	−0.3530 ***	−0.3485 ***	−0.3577 ***
	(−8.2505)	(−8.6066)	(−8.7383)	(−8.6138)	(−8.7921)
ANALYST	−0.0067 *	−0.0120 ***	−0.0109 ***	−0.0101 ***	−0.0138 ***
	(−1.7297)	(−3.0921)	(−2.8125)	(−2.6103)	(−3.5496)
BETA	−0.0079	−0.0097	−0.0107	−0.0289	−0.0280
	(−0.2524)	(−0.3052)	(−0.3394)	(−0.9183)	(−0.8806)
RESEARCH	−0.0659 **	−0.0879 ***	−0.0656 *	−0.0604 *	−0.0714 **
	(−1.994)	(−2.6515)	(−1.9465)	(−1.8297)	(−2.1033)
Country Dummies	Yes	Yes	Yes	Yes	Yes
Industry Dummies	Yes	Yes	Yes	Yes	Yes
Year Dummy	Yes	Yes	Yes	Yes	Yes
Observations	13,102	13,102	13,102	13,102	13,102
R-Square	0.3465	0.3384	0.3400	0.3366	0.3311

Note: The impact of ESG (social pillar score) on the link between firm performance and COVID-19-induced sentiment is shown in this table. Parentheses surround the t-values derived from the heteroscedasticity-robust standard errors. Firm performance (Q) is the outcome variable, and the main independent variables are SOC (social pillar scores), SENTIMENT (the COVID-19-induced sentiment index), and SOC*SENTIMENT. The years 2020 and 2021 make up the sample. Regression using pooled OLS is employed. p < 0.1, p < 0.05, and p < 0.01 are represented by the symbols *, **, and ***, respectively. For the definition of the variables, see Table 1.

presents the findings of the analysis. Column (2) uses the overall score for the social pillar to define SOC, and Columns (3) to (6) use individual components of the social pillar (human rights, community, workforce, and product responsibility) as proxies to define SOC. The findings indicate that the social pillar, as well as the individual components of the social pillar, affect the relationship between COVID-19-induced sentiment and company productivity. We report a significantly positive coefficient estimate of SOC*SENTIMENT, as anticipated. These results suggest that when accounting for COVID-19-induced investor sentiment, the benefits of improved sentiments accrue significantly to firms that spend resources on and engage with the social dimensions of ESG activities.

To better comprehend the results reported in Table 9, we also plot the average marginal impact of COVID-19-induced sentiment (SENTIMENT) on company performance (Q) conditional on the social activities undertaken by firms (SOC) in Figure 3. The figure shows that as social activities increase, the marginal impact of COVID-19-induced sentiment on firm performance also increases and becomes positive.

4.5. Effect of ESG (Governance Pillar Score) on the Relationship between COVID-19-Induced Sentiment and Firm Performance

Table 10. Effect of ESG (governance pillar score).

Variables	GOV = Overall Score for Governance	GOV = Management Score	GOV = Shareholder Score	GOV = CSR Strategy Score
GOV	−0.0104 ***	−0.0066 ***	−0.0029 ***	−0.0103 ***
	(−8.6349)	(−7.0127)	(−3.1565)	(−11.9678)
SENTIMENT	−0.0058	−0.0038	0.0003	−0.0138 ***
	(−0.9888)	(−0.7288)	(0.0569)	(−2.9018)
GOV*SENTIMENT	0.0001	−0.0001	−0.0001	0.0002 ***
	(0.4538)	(−0.0627)	(−1.4522)	(3.7612)
SIZE	0.5064 ***	0.4938 ***	0.4769 ***	0.5570 ***
	(28.8498)	(28.3446)	(27.7045)	(30.7600)
LEVERAGE	−0.0171 ***	−0.0174 ***	−0.0175 ***	−0.0155 ***
	(−18.5065)	(−18.7163)	(−18.7129)	(−17.023)
LOSS	−0.0014	0.0018	0.0149	0.0166
	(−0.0323)	(0.0419)	(0.3344)	(0.3811)
GROWTH	0.0034 ***	0.0035 ***	0.0037 ***	0.0032 ***
	(7.0618)	(7.2694)	(7.4989)	(6.6470)
CASH	3.3468 ***	3.3809 ***	3.4162 ***	3.3016 ***
	(20.1831)	(20.3579)	(20.4054)	(20.0503)
DIVIDEND	−0.3402 ***	−0.3503 ***	−0.3765 ***	−0.3115 ***
	(−8.3879)	(−8.5900)	(−9.1792)	(−7.7702)
ANALYST	−0.0139 ***	−0.0156 ***	−0.0176 ***	−0.0075 *
	(−3.5704)	(−4.0200)	(−4.5177)	(−1.9439)
BETA	−0.0105	−0.0162	−0.0311	−0.0278
	(−0.3326)	(−0.5086)	(−0.9801)	(−0.8913)
RESEARCH	−0.0664 **	−0.0614 *	−0.0717 **	−0.0830 **
	(−1.9773)	(−1.8163)	(−2.1033)	(−2.5069)
Country Dummies	Yes	Yes	Yes	Yes
Industry Dummies	Yes	Yes	Yes	Yes
Year Dummy	Yes	Yes	Yes	Yes
Observations	13,102	13,102	13,102	13,102
R-Square	0.3349	0.3309	0.3251	0.3492

Note: The influence of ESG (governance pillar score) on the relationship between corporate performance and sentiment induced by COVID-19 is shown in this table. Parentheses surround the t-values derived from the heteroscedasticity-robust standard errors. Firm performance (Q) is the outcome variable, and SENTIMENT (the COVID-19-induced sentiment index), GOV (the governance pillar scores), and GOV*SENTIMENT are the main independent variables. The years 2020 and 2021 make up the sample. Regression using pooled OLS is employed. p < 0.1, p < 0.05, and p < 0.01 are represented by the symbols *, **, and ***, respectively.

presents the findings of the analysis. Column (2) uses the overall score for the governance pillar to define GOV, and Columns (3) to (5) use the individual elements that make up the governance pillar (management, shareholders, and CSR strategy) as proxies to define GOV. The findings indicate that the governance pillar and two of the three individual components of the governance pillar do not affect the link between COVID-19-induced sentiment and corporate performance. The only factor that is shown to be significant is the CSR strategy. We report a significantly positive coefficient estimate of GOV*SENTIMENT when CSR strategy is used as a proxy for GOV.

To gain a better understanding of the findings observed in Table 10, we also plot the average marginal influence of COVID-19-induced sentiment (SENTIMENT) on company performance (Q) conditional on governance activities undertaken by firms (GOV) in Figure 4. In line with previous findings, the figure shows that as governance activities increase, the marginal impact of COVID-19-induced sentiment on firm performance does not change significantly.

5. Additional Tests

5.1. Effect of ESG on the Relationship between Alternative Proxies of COVID-19-Induced Sentiment and Firm Performance

As additional robustness checks, we use three alternative proxies of COVID-19-induced sentiment (SENTIMENT) to re-estimate Equation (1). Note that these different proxies capture a wide range of news related to the coronavirus pandemic that might affect investor sentiment. Among the significant topics covered by RavenPack sentiment proxies are ‘vaccines’ and ‘vaccine hesitancy’. Pharmaceutical and therapeutic interventions, particularly the development and distribution of vaccines, contribute to the sentiments surrounding COVID-19. This aligns with the understanding that successful vaccine deployment positively impacted investor sentiment by mitigating the pandemic’s effects.

The first proxy (PANIC) measures the proportion of news that creates panic or hysteria about the coronavirus pandemic. The second proxy (HYPE) measures the percentage of news that is related to the coronavirus crisis. The third proxy (FAKE) gauges the degree of news that refers to misinformation or fake news about coronavirus. Higher values of these proxies indicate deteriorating sentiment. The findings based on these alternative proxies of SENTIMENT are reported in

Table 11. Impact of ESG on the relationships among alternative proxies of COVID-19-induced sentiment and firm performance.

Variables	Sentiment = Panic	Sentiment = Hype	Sentiment = Fake
ESG	−0.0189 ***	−0.0175 ***	−0.0184 ***
	(−7.7225)	(−4.9036)	(−8.2114)
SENTIMENT	0.0759	0.0221 **	0.1871
	(1.4776)	(2.0943)	(1.1883)
ESG*SENTIMENT	−0.0011 **	−0.0002 *	−0.0047 **
	(−2.0757)	(−1.7466)	(−2.4849)
SIZE	0.5867 ***	0.5866 ***	0.5850 ***
	(31.8518)	(31.8363)	(31.6822)
LEVERAGE	−0.0158 ***	−0.0158 ***	−0.0158 ***
	(−17.4285)	(−17.3991)	(−17.4363)
LOSS	0.0153	0.0145	0.0146
	(0.3531)	(0.3343)	(0.3346)
GROWTH	0.0029 ***	0.0029 ***	0.0029 ***
	(6.2981)	(6.314)	(6.2963)
CASH	3.2806 ***	3.2798 ***	3.2843 ***
	(20.0649)	(20.0595)	(20.0901)
DIVIDEND	−0.3012 ***	−0.3020 ***	−0.3014 ***
	(−7.5572)	(−7.5757)	(−7.574)
ANALYST	−0.0053	−0.0053	−0.0052
	(−1.3816)	(−1.3775)	(−1.352)
BETA	0.0071	0.0060	0.0050
	(0.2262)	(0.1905)	(0.1607)
RESEARCH	−0.0785 **	−0.0779 **	−0.0786 **
	(−2.438)	(−2.4197)	(−2.4384)
Country Dummies	Yes	Yes	Yes
Industry Dummies	Yes	Yes	Yes
Year Dummy	Yes	Yes	Yes
Observations	13,102	13,102	13,102
R-Square	0.3540	0.3540	0.3541

Note: The impact of ESG on the link between various sentiment index types and company performance is shown in this table. Parentheses surround the t-values derived from the heteroscedasticity-robust standard errors. Firm performance (Q) is the outcome variable, and SENTIMENT (various COVID-19-related sentiment indices), ESG (ESG scores), and ESG*SENTIMENT are the main independent variables. The years 2020 and 2021 make up the sample. Regression using pooled or OLS is applied. p < 0.1, p < 0.05, and p < 0.01 are represented by the symbols *, **, and ***, respectively. For definitions of the variables, see Table 1.

. We report a significant influence of ESG incentives on the link between sentiments and firm performance. Our results again highlight that the trust between a corporation and its investors, developed through ESG investments, pays off during uncertain times.

5.2. Effect of Fiscal and Monetary Policies on the Relationship between COVID-19-Induced Sentiment and Firm Performance

Several macro confounding factors had a pivotal role in moderating the adverse impact of the COVID-19 crisis on the overall economy. For instance, different countries have implemented several monetary, fiscal, and macro-financial policies to counteract the impact of the COVID-19 crisis and ensure the stability of the economy [35]. Cortes et al. [36] show that the monetary policies implemented by countries as a response to the crisis possess positive spillovers to equity markets. Central banks often lower interest rates and purchase government securities, which reduces sovereign yields and makes bonds less attractive compared to equities, thus driving investors toward the stock market. Additionally, these policies increase market liquidity by injecting cash into the financial system, which enhances investors’ ability to buy stocks, thereby boosting equity prices. Moreover, unconventional measures like quantitative easing (QE) can improve market confidence and stability, encouraging investment and risk-taking in equity markets. These actions collectively lead to positive spillovers that benefit equity markets globally.

In addition, Benmelech and Tzur-Ilan [37] study the fiscal policies that were implemented during the pandemic to support households and businesses with the aim of maintaining economic relationships and stability. Governments around the world adopted measures such as direct financial assistance, tax deferrals, and business subsidies to mitigate economic turmoil. Fiscal policies in various countries included measures to support businesses directly through grants, loans, and tax reliefs, in addition to maintaining social stability by addressing inequality and providing targeted support to vulnerable populations. Moreover, fiscal policies were not limited to immediate financial support but also included strategic long-term investments aimed at stimulating economic recovery. These measures helped maintain consumer confidence, ensured liquidity, and prevented widespread bankruptcies, thereby sustaining economic stability and preserving critical economic relationships.

To confirm the robustness of our findings, we conduct our analysis again by first sorting the sample of firms into two groups based on macroeconomic stability, fiscal strength, and social capital variables, and by including country dummies. Our choice of variables is based on several reasons. First, macroeconomic stability reflects a nation’s growth sustainability and economic resilience to financial market turmoil. This measure covers the ability to maintain price stability, implement sustainable fiscal policies, and manage debt levels. Fiscal strength is the second macroeconomic variable employed in this manuscript, as it reflects the government’s capacity to maintain its spending and tax policies in the long run without the risk of insolvency or default. Our third proxy is the social capital factor, which measures the robustness of social relationships and networks within a society. This factor evaluates trust in institutions, civic participation, and community involvement. Macroeconomic stability, fiscal strength, and social capital variables are extracted from the Legatum Institute, capturing various monetary, fiscal, and macro-financial policies that might impact our analysis.

The findings presented in

Table 12. Effect of ESG on the relationship between COVID-19-induced sentiment and firm performance: the impact of macro confounding factors.

	Macroeconomic Stability		Fiscal Strength		Social Capital
Variables	Low	High	Low	High	Low	High
ESG	−0.0165 ***	−0.0227 ***	−0.0223 ***	−0.0179 ***	−0.0112 ***	−0.0223 ***
	(−8.7363)	(−10.2096)	(−11.242)	(−8.6197)	(−5.5294)	(−11.4399)
SENTIMENT	−0.013	−0.0195 *	0.0048	−0.0267 ***	−0.0328 ***	−0.0296 ***
	(−1.4066)	(−1.6879)	(0.5653)	(−3.3552)	(−4.3446)	(−2.6315)
ESG*SENTIMENT	0.0003 **	0.0003 **	−0.0001	0.0005 ***	0.0006 ***	0.0006 ***
	(2.3726)	(2.1807)	(−0.6629)	(4.3688)	(4.9644)	(3.5964)
SIZE	0.5806 ***	0.5894 ***	0.6633 ***	0.4896 ***	0.4214 ***	0.7445 ***
	(23.4479)	(21.2941)	(23.8657)	(20.5308)	(17.6548)	(26.8744)
LEVERAGE	−0.0113 ***	−0.0201 ***	−0.0135 ***	−0.0185 ***	−0.0172 ***	−0.0139 ***
	(−9.7261)	(−14.4489)	(−10.9047)	(−14.0064)	(−14.2126)	(−10.5654)
LOSS	−0.0744	0.1379 **	0.1274 **	−0.0426	−0.0132	0.0809
	(−1.2686)	(2.095)	(2.0976)	(−0.6939)	(−0.2579)	(1.2792)
GROWTH	0.0011 **	0.0055 ***	0.002 ***	0.0043 ***	0.0069 ***	0.0011 **
	(2.3698)	(6.0855)	(3.5598)	(5.5792)	(8.8132)	(2.0616)
CASH	3.1256 ***	3.4467 ***	2.8439 ***	4.0646 ***	2.9235 ***	3.4204 ***
	(15.1869)	(13.7505)	(14.3811)	(14.9273)	(10.7263)	(16.7894)
DIVIDEND	−0.2843 ***	−0.3651 ***	−0.4559 ***	−0.2329 ***	−0.2368 ***	−0.4055 ***
	(−5.3417)	(−6.16)	(−8.3991)	(−3.9803)	(−4.3552)	(−7.1005)
ANALYST	−0.0242 ***	0.01 *	−0.0231 ***	0.0153 ***	0.0183 ***	−0.0358 ***
	(−4.3176)	(1.9258)	(−4.2106)	(2.8472)	(3.5933)	(−6.074)
BETA	−0.0885 **	0.1241 **	−0.0236	0.002	−0.0148	0.0038
	(−2.546)	(2.2412)	(−0.6007)	(0.0378)	(−0.2805)	(0.0963)
RESEARCH	−0.0365	−0.1003 **	−0.0243	−0.1835 **	0.1534	−0.0939 ***
	(−0.8403)	(−2.1095)	(−0.6691)	(−2.5069)	(1.5791)	(−2.6752)
Country Dummies	Yes	Yes	No	Yes	Yes	Yes
Industry Dummies	Yes	Yes	No	Yes	Yes	Yes
Year Dummy	Yes	Yes	Yes	Yes	No	Yes
Observations	6551	6551	6551	6551	6551	6551
R-Square	0.3421	0.3839	0.3661	0.363	0.3425	0.372

Note: This table investigates the effect of ESG on the relationship between COVID-19-induced sentiment and firm performance when controlling for macroeconomic stability, fiscal strength, and social capital factors. These factors are extracted from Legatum Institute. We start by sorting our sample of firms into two groups based on macroeconomic stability, fiscal strength, and social capital variables. Firm performance (Q) is the outcome variable, and SENTIMENT, ESG, and ESG*SENTIMENT are the main independent variables. Parentheses surround the t-values derived from the heteroscedasticity-robust standard errors. Regression using pooled or OLS is applied. p < 0.1, p < 0.05, and p < 0.01 are represented by the symbols *, **, and ***, respectively. For definitions of the variables, see Table 1.

are consistent across different subsamples. Our results indicate that the coefficient for ESG × SENTIMENT continues to be positive and remains statistically significant in almost all groups, suggesting that macroeconomic stability, fiscal strength, and social capital are not the primary drivers of our results. This robust analysis supports our hypothesis that ESG activities play a moderating role in the relationship between SENTIMENT and firm performance, beyond these macro confounding factors.

6. Conclusions

Within the vast field of behavioral finance, studying the impact of investor sentiment during turbulent market conditions on market efficiency, on the one hand, and on firms’ decision making and performance, on the other, has become important for corporations and policymakers. Our research adds to the body of current literature by providing new crucial insights and understanding on this topic. When faced with investors’ panic, firms might engage in inefficient projects or forgo potential opportunities when their stocks are undervalued. Our study sheds light on the effectiveness of different types of ESG engagement activities on firms’ performance for firms highly impacted by COVID-19-induced investor sentiment. The findings suggest that the benefits of improvements in sentiment accrue to firms that engage in and spend more resources on ESG activities.

Using the data of nonfinancial firms from 49 countries, we show that the influence of COVID-19-induced sentiment on firm performance is more pronounced for firms that score high on ESG activities. The findings remain robust across various estimation strategies, across alternative COVID-19-induced investor sentiment, and across different subsamples. Our study suggests that corporations that integrated sustainable practices into their corporate culture and business plan were better equipped to face the coronavirus crisis, as they had already implemented solid ESG frameworks. All three ESG dimensions are important if firms are to implement effective crisis management and ensure business continuity during periods of market turbulence. However, in the context of the coronavirus pandemic, we show that some components of each dimension took on greater significance and had a key impact on the performance of firms that were highly affected by the COVID-19-induced fear that spread among investors. For instance, our additional analyses show that the social and environmental dimensions of ESG were the driving forces behind the moderating impact of ESG activities on the relationship between COVID-19-induced sentiment proxies and firm performance. We also demonstrate that various activities within the social (such as human rights, community, workforce, and product responsibility) and environmental (such as resource use, emissions, and environmental innovation) dimensions of ESG significantly affect the relationship between COVID-19-induced sentiment and company performance. In contrast to social and environmental activities, governance activities have no significant impact on this relationship, with the exception of CSR strategy.

The findings of this paper can assist corporations in their strategic decision making, as they show that the adoption of ESG initiatives is not only important for long-term sustainability; it can also provide financial benefits during periods of market turmoil. ESG incentives can thus be tailored to address specific industry challenges and opportunities. For instance, companies operating in the tech industry might focus on ethical AI and data privacy, thereby building customer confidence and enhancing brand loyalty, whereas firms operating in the manufacturing industry might prioritize reducing emissions and waste to demonstrate their environmental stewardship and attract stakeholders who value sustainability. Crucially, sustainable activities should be aligned with country-specific strategies, market conditions, and authorities’ expectations. For example, companies in developed countries might focus on stringent compliance and innovation with the aim of mitigating financial and operational risks, in addition to enhancing their efficiency and profitability. On the other hand, companies headquartered in emerging markets might emphasize social responsibility and community engagement to build trust and support, which will enable them to attract more foreign direct investment (FDI).

Our paper provides significant implications for academics, corporations, and policymakers. First, our study addresses a current research gap in the literature by examining the moderating impact of ESG activities on the relationship between COVID-19-induced sentiment and firm performance. Moreover, our results offer valuable insights for corporations by providing solid evidence of the importance of incorporating ESG incentives, particularly those within the environmental and social dimensions, into strategic managerial decision making. We emphasize that incorporating ESG activities can help mitigate the adverse effects of financial crises, as investors perceive firms that engage in sustainable activities as safe havens amid volatile market conditions. The commitment to ESG initiatives can thus generate several positive externalities, such as fostering investor trust, enabling more informed decision making, and reducing information asymmetry in financial markets. Our findings thus suggest that policymakers should promote regulations that encourage and support firms in their ESG initiatives, as these efforts play a crucial role in protecting corporations and, by extension, the economy during periods of financial turmoil.

Our study is limited due to the use of data solely from nonfinancial firms. Therefore, future research can explore the moderating impact of ESG activities on the relationship between COVID-19-induced sentiment and financial firms’ performance. This would enable researchers to determine whether similar patterns apply in financial firms or whether significant differences exist due to the unique characteristics of financial institutions. Moreover, future research can assess how shareholder activism linked to ESG incentives can impact investors’ sentiments during periods of market turbulence.