Environmental Risk Concern and Short-Term IPO Performance of Green Stocks During the COVID-19 Crisis Period

Abstract

This study examines the effect of firms’ greenness on IPO underpricing and subsequent short-term performance during the COVID-19 crisis period. Using 173 U.S. IPOs, we find that IPO underpricing is more pronounced for brown firms (i.e., firms have higher carbon footprints or operate in pollution-intensive industries) than for green firms (i.e., firms are engaged in environmentally sustainable practices). However, when we account for the exogenous change in environmental concerns, we find that an increase in environmental concerns causes lower initial day returns for brown firms. Later, we examine the post-IPO 3-month (6-month) holding period returns and find that brown firms earn higher returns than green firms when environmental concerns increase. Additionally, cross-sectional regressions indicate that firm-level characteristics, such as offer price and Hi-Tech, are positively associated, while R&D, leverage, and profitability are negatively associated with IPO.

1. Introduction

In response to the pandemic, federal, state, and municipal policymakers have issued waivers and announced plans to stop enforcing key environmental laws and regulations.1 Government agencies cited different justifications for these changes, including reduced health concerns. Do investors lose sight of environmental concerns during the COVID-19 period? The unprecedented disruption caused by the pandemic has precipitously slowed down economic and social activities. They have also been associated with a significant reduction in air and water pollution and greenhouse gas (GHG) emissions (Khan et al., 2021). Air quality levels during the COVID-19 period improved dramatically, largely due to the reduced production in factories and traffic emissions (Kapparashetty, 2020). As a matter of fact, global airline traffic dropped by 60%, which also helped to reduce the emission of carbon dioxide. On the other hand, Singh et al. (2020) argued that the disposal of medical personal protective equipment (PPE) and other hospital waste caused adverse effects on the environment and created a new environmental concern at the same time. As a result, exogenous environment concerns during the COVID-19 period have changed significantly from time to time. How an investor incorporates their attention in the investment decision regarding the climate risk during the relaxed environmental laws is an important and unexplained question, more specifically, for investment decisions where investors are subject to more information inaccessibility, such as IPO firms.

IPO pricing has been a prominent focus for many financial economists for the last several decades (Ibbotson, 1975; Booth & Smith, 1986; Rock, 1986; Michaely & Shaw, 1994; Loughran & Ritter, 2004; X. Liu & Ritter, 2011; and many others). In a relevant study, Chan and Walter (2014) found an insignificant association between firms’ greenness and underpricing; however, they did not consider the pandemic period and the exogenous change of environmental concerns as a potential source of underpricing. Although the pandemic had caused significant economic disruptions worldwide, U.S. IPO underpricing increased by 23.60% during the year 2000 (1 January 2020 to 31 December 2020) compared to the previous four decades (Ritter, 2014).2 Existing studies have partially explained IPO underpricing during the COVID-19 crisis period (Baig & Chen, 2021; Kuswanto, 2021; Mazumder & Saha, 2021; among others). The driving force of the IPO underpricing during the pandemic begs more academic attention. In this study, we investigated firms’ environmental scores on IPO underpricing and subsequent short-term performance during the pandemic when an exogenous environmental concern exists. Our motivation to study these questions stemmed from mainly two reasons. We considered IPO firms as a laboratory test because investors do not have a lot of prior information of these firms. First, information about IPO firms is mostly limited, as they are not subject to similar level information disclosures like existing public firms. As a result, investors’ perceptions about policy changes and new environment concerns can be reflected better in an IPO firm than in an existing company. Second, we chose the COVID-19 period because it causes an exogenous shock on the environment in both directions.

Green firms are considered more environmentally sustainable or climate-friendly firms; thus, they are subject to lower ${CO}_2$ emissions. On the other hand, brown firms are less climate-friendly ones. The substantial rise of Environmental, Social, and Governance (ESG) practices has lately fueled research on the relationship between ESG and financial decision-making. To date, this line of research has produced mixed findings. On the one hand, higher ESG scores contain value-relevant information, such as higher quality disclosures (Lopez-de-Silanes et al., 2020) and lower information asymmetry (El Ghoul et al., 2011), which may lead to lower underpricing.3 Consistent with this notion, E. D. Baker et al. (2021) reported a negative association between ESG government ratings and firm-level IPO underpricing. On the other hand, Yamashita et al. (1999) found that companies with the worst environment scores have lower-than-average returns. The findings are also validated by Klassen and McLaughlin (1996) and Derwall et al. (2005). Higher ESG scores may drive investors toward those firms (Da et al., 2011; L. X. Liu et al., 2014). Consequently, higher ESG scores could lead to higher initial day returns. The findings of our study reconcile the mixed findings of those prior studies.

ESG ratings can affect firms’ subsequent performance via several mechanisms. Prior research suggests that higher ESG ratings can help reduce information asymmetry for IPO firms. Dyck et al. (2019) supported the notion in the framework of institutional investors. They argued that institutional investors from countries with strong ESG disclosure regulations have a positive impact on companies’ ESG policies. Since institutional investors are the larger IPO stakeholders and communicate more with the company during the IPO process, they can influence firms to become more involved in ESG activities. In the capital market equilibrium with an incomplete information model, Merton (1987) argued that investors hold the securities in the portfolio that they know well. Consistent with Merton’s model, El Ghoul et al. (2011) found that information asymmetry is less severe for firms with higher ESG scores due to these firms’ higher analyst attention. At the same time, ESG ratings disclose firms’ proximity to climate risk. Some prior work provides evidence that firms should earn risk premiums for climate risk (Bolton & Kacperczyk, 2021; Hsu et al., 2023; Balvers et al., 2017). In another vein, the underperformance hypothesis suggests that the risk-adjusted returns for socially responsible funds are lower due to the higher cost structure and restricted investment opportunity. In contrast, the overperformance hypothesis proposes companies with higher environmental standards create shareholders’ wealth in the long run. More specifically, the hypothesis argues that companies with higher social responsibility can avoid the potential cost of corporate social crises as well as an environmental hazard. In line of this notion, we argued that investors perceive lower environment-friendly firms as risky investments (Bolton & Kacperczyk, 2021), especially when the media sentiment on environmental concerns increases. Our findings reconciled this prior work.

Using 173 U.S. IPO data from February 2020 to December 2020, we initially found that brown firms have more underpricing than green firms. Next, we examined the IPO underpricing of green and brown firms when there was an exogenous change in environmental concerns. Prior studies on media coverage and sentiment found an association between media sentiment and IPO underpricing (Y. Chen et al., 2020b; Derrien, 2005; Wang & Wu, 2015; Lee et al., 1991; Bhattacharya et al., 2009; Guldiken et al., 2017; Pollock & Rindova, 2003; Bajo & Raimondo, 2017; Zou et al., 2020; Liew & Wang, 2016; Mazumder & Saha, 2021). Investors’ concerns during the COVID-19 period changed from time to time. When we accounted for the exogenous change in the environmental concern on IPO underpricing, we found that an increase in the concern for the climate drove investors toward green firms and led to a lower initial return for brown firms. We found that IPO initial returns decreased by 13.02% for the brown firms when the lagged environment concern increased by one standard deviation.4 The association was robust when we used market-adjusted returns.

Next, we examined how the subsequent performance of firms, separated by their median g-scores, was affected when there was an exogenous change in the climate concern. Using Pástor et al. (2022, hereafter PST (2022)) model, we found that post-IPO 3-month and 6-month holding period returns (HPR) were positively associated with brown firms if environmental concern increased. Precisely, the 3-month (6-month) returns of brown IPO stocks increased by 19.3% (23.4%) if contemporaneous climate change concerns increased by one standard deviation. It suggests that investors require more returns to be compensated for that risk (Bolton & Kacperczyk, 2021), especially when the environmental concerns are higher. At last, we examined the firm-level characteristics that explained the post-IPO performance in a sub-sample analysis of green and brown firms. Firm size, offer price, and Hi-Tech explained post-IPO performance positively and significantly for green IPO firms. On the other hand, R&D explained positively, and leverage and being traded in NASDAQ explained negatively for brown firms’ post-IPO performance. Finally, we examined the association between firms’ greenness and stock liquidity. The results showed that brown firms were less liquid and the association was less pronounced when the environmental concerns increased.

This study contributed to two streams of literature. We contributed to the growing COVID-19 literature in the strand of IPO underpricing and also climate risk literature in asset pricing. The study by E. D. Baker et al. (2021) was the closest to our own, analyzing the effects of ESG government risk on international IPO underpricing. Our findings suggest that investors demand risk premium for investing in brown firms, consistent with Bolton and Kacperczyk (2021). However, investors were willing to pay a premium for green stocks when the exogenous environment concerns increased. Existing IPO studies of the pandemic focused on how investors’ fear sentiment (Mazumder & Saha, 2021) and uncertainty (Baig & Chen, 2021) explained IPO underpricing. We extended the literature by analyzing the effect of firms’ greenness on IPO underperformance. Second, we further studied how the change of environment concern explained IPO underpricing for green and brown stocks. Third, this paper extended the investigation of post-IPO short-term performance when an exogenous risk, climate change, increased. Finally, the cross-sectional regressions for green and brown sub-samples revealed the firm characteristics that explained post-IPO short-term underperformance.

The remainder of the paper is organized as follows. Section 2 briefly reviews the existing literature and develops hypotheses. Section 3 describes the sample and data. Section 4 presents the empirical results, followed by discussion and conclusion in Section 5.

2. Literature Review and Hypothesis Development

2.1. Underpricing IPO and ESG Disclosure

IPO literature explains the first-day return of IPO firms mostly from two different perspectives. One school of thought explained the first-day return based on the firm-specific attributes, such as information asymmetry between investors (Rock, 1986) and the reputation of underwriters (Megginson & Weiss, 1991), signaling by qualitative firms (Grinblatt & Hwang, 1989; Welch, 1989). Several other firm-level attributes also explained underpricing, e.g., ex-ante uncertainty of issuing firms (Beatty & Zajac, 1994), uncertainty about future growth opportunity, firm age (Ritter, 1984; Loughran & Ritter, 2004), higher P/E ratio (G. Chen et al., 2004; Engelen, 2003), the proportion of insider shareholding (Habib & Ljungqvist, 2001), and so on.

However, most of the prior studies overlooked an important internal attribute, which was firms’ greenness or environment-friendly nature as a firm-specific attribute for underpricing. The environment factor is one of the major components of the ESG rating, and it contains value-relevant information. As private firms are not subject to higher disclosure requirements, information about IPO firms is often limited. Therefore, information asymmetry between issuers and underwriters (Baron, 1982), issuers and investors (Welch, 1989), and among different groups of investors (Rock, 1986) exists and influences the underpricing of IPO firms. Existing research suggests that a higher ESG rating is associated with lower information asymmetry and can have two different impacts on IPO underpricing.

El Ghoul et al. (2011) argued that information asymmetry was less severe for firms with higher ESG ratings. Moreover, Feng et al. (2018) argued that higher ESG firms were more ethical and transparent, which led to lower underpricing for their seasoned equity offering. These studies supported the notion that higher ESG or environmental friendliness of the firms was associated with lower information asymmetry among the different stakeholders of the firm and would lower information uncertainty. Consequently, IPOs with higher ESG ratings should have lower underpricing. We were motivated to examine the role of a firm-specific attribute, such as g-score, in explaining the IPO underpricing during the COVID-19 period. Following Pástor et al. (2022), we used g-score as a proxy for the ESG score that could determine the level of information asymmetry for the firm. Thus, we hypothesized:

H1: 

The underpricing will be more for brown firms during the COVID-19 period.

Similarly, it is also possible that higher ESG or environmentally friendliness could drive investors toward those firms and lead to higher initial day returns for green firms (Da et al., 2011; L. X. Liu et al., 2014) because firms’ greenness attributes attract investors. Dimson et al. (2015) showed that institutional ownership increased in the year following the adoption of the improved ESG practices. These findings support the notion that investors were attracted by the higher ESG rating and were more willing to buy those stocks on the first day.

2.2. IPO Underpricing and Exogenous Climate Concern

Another school explained the IPO’s initial day return regarding outside factors or exogenous shock. A growing stream of research advocated investor sentiment as a critical factor for IPO initial return (Y. Chen et al., 2020b; Derrien, 2005; Wang & Wu, 2015). Usually, issuing firms tend to time the market, and more firms used to go public when the investors’ sentiment is high (Lee et al., 1991) to coincide with periods of excessive market valuation (M. Baker & Wurgler, 2002). Zhao et al. (2018) found that investors’ attention positively explained the IPO underpricing. Media coverage (Bhattacharya et al., 2009; Y. Chen et al., 2020a; Guldiken et al., 2017; Pollock & Rindova, 2003) and media tone (Bajo & Raimondo, 2017; Zou et al., 2020) for the IPO also influence the initial day return. Twitter sentiment can also explain the IPO underpricing, especially the pre-IPO dates tweets (Liew & Wang, 2016). Mazumder and Saha (2021) found that the initial return of IPOs was negatively affected by the pandemic fear during the first half of 2020.

Motivated by these studies, we combined an exogenous shock, the change in climate concern, with firm attributes, such as the firm’s greenness (g-score), to explain the IPO underpricing during the COVID-19 period. COVID-19 had a significant positive impact on the environment as the pollution reduced drastically due to restricted movement and reduced production in factories and traffic emissions. However, media coverage of huge medical waste disposal created a new environmental concern. Therefore, we hypothesized that investors moved toward investing more in green firms when there was an increase in climate concern, which drove the lower initial returns for brown firms.

H2: 

The underpricing for brown firms is lower if the exogenous climate concern increases during the COVID-19 period.

2.3. Post-IPO Short-Term Performance and Exogenous Climate Concern

Next, we examined whether the investment in green IPO firms on the first trading day was good for maximizing wealth because green firms have different investment opportunity sets than brown firms in the long run. Bolton and Kacperczyk (2021) argued that investors assume brown firms are risky firms and thus require more returns for those risks. Balvers et al. (2017) found a risk premium for temperature-related environmental concerns. Hsu et al. (2023) also showed a pollution premium for the stocks. Thus, investors demand higher returns to hold environmentally unfriendly firms in their portfolios. In addition, the underperformance hypothesis states that environment-friendly firms may face constraints in their investment opportunities due to their environment-friendly practices, which could lead to a lower risk-adjusted return (Renneboog et al., 2008; Revelli & Viviani, 2015). Moreover, environmental concerns during the COVID-19 crisis period increased investors’ risk perception about brown firms. Considering higher environmental concern as a risk factor, we formalized:

H3: 

The post-IPO short-term performance is higher (lower) for brown (green) firms if the exogenous climate concern increases during the COVID-19 period.

There could be alternative explanations for hypothesis three. The overperformance hypothesis proposes companies with higher environmental standards create shareholders’ wealth in the long run.

3. Methodology

In this section, we explained how we collected the data for empirical analysis and constructed samples consistent with the standard literature. We also reported the results and subsequent discussion.

3.1. Data and Sample Construction

This section explains how we constructed the sample for our empirical analysis. In the United States, the first COVID-19 case was confirmed on 20 January 2020, although China reported to the World Health Organization (WHO) a string of pneumonia-like cases in Wuhan for the very first time on 31 December 2019. As a result, we chose our sample period ranging from February 2020 to December 2020.

Our primary dataset came from several different sources. We collected IPO offer price and IPO firm’s financial data from COMPUSTAT, SEC EDGAR, and Professor Jay Ritter’s website. Firm’s g-score came from PST (2022), which covers until December 2020.5 PST (2022) reported the g-score at the MSCI industry level. We mapped the industry-level score to the firm level. The climate concern data was the aggregate climate concern data as published in the Media and Climate Change Observatory Data by Daly et al. (2021). They report the climate change and global warming news from the top five U.S. newspapers by month.6 We constructed climate concern proxy, $C_m$, as the aggregate news covered by month. Change of climate concern, ${Delta}_m=C_m-C_{m-1}$, proxies an increase or decrease of climate concern. To construct a proxy for market return, we used the return data of the S&P 500. Consistent with the IPO-pricing literature, we excluded IPOs that have an offer price of less than 5; are in financial and utility sectors; are not traded in NYSE, NASDAQ, and AMEX; and are ADRs. This left us with a final sample of 173 firms for both initial and subsequent period holding return analysis.

3.2. Variables Construction

Here, we explained how we calculated our variables and measures. Following the convention, we calculated the initial return, adjusted initial return, and 3- and 6-month holding period raw (adjusted) return for each firm as follows:

$${InitialReturn}_i={\displaystyle \frac{{CP}_i-{OP}_i}{{CP}_i}}$$

(1)

where ${CP}_i$ is the closing price on the first day of trading and ${OP}_i$ is the offer price.

$${AdjInitialReturn}_{id}={InitialReturn}_{id}-R_{md}$$

(2)

where ${InitialReturn}_{id}$ is the first-day return and $R_{md}$ is the market return on that day.

$${HPR}_i={\prod }_{t=1}^n(1+r_{it})$$

(3)

where $r_{it}$ is the monthly holding return of IPO firms and n represents 3- or 6-month holding periods.

$${Adj-HPR}_i={HPR}_i-R_m$$

(4)

where ${HPR}_i$ is the 3- and 6-month holding period return and $R_m$ is the market return of similar length period.

3.3. Study Methods

Our empirical analysis used OLS regression with fixed effects. Initially, we studied the impact of firms’ greenness on IPO returns. For the initial performance of IPOs, we estimated the following regression:

$${IR}_i={\beta }_0+{\beta }_1{LowAVGG}_{t-1}+X_{i,t-1}+d_m+{ϵ}_i$$

(5)

where ${IR}_i$ was the initial returns or adjusted initial returns of IPOs. ${LowAVGG}_{t-1}$ consisted of firms with a g-score below the median cutoff based on firms’ greenness in 2019. $X_{i,t-1}$ was a vector of control variables in the year 2019. We controlled IPO offer size, firm size, offer price, NASDAQ dummy, Hi-Tech dummy, market capitalization, R&D, volume, leverage, and ROA (following Brav & Gompers, 1997; Chahine et al., 2020; Krishnan et al., 2011; Vong & Trigueiros, 2010; Zhou & Sadeghi, 2019).7 $d_m$ was the month fixed effect.8 ${ϵ}_{i,t}$ was the white noise when standard errors are heteroscedasticity-consistent robust.

We further extended the analysis to see whether the negative association was unique during the crisis periods or was common to the pre-crisis periods. To address this issue, we adopted a difference-in-difference (DID) model with month fixed effects. In this analysis, our sample started in January 2019 (one year before the crisis began) and ended in December 2020. We estimate the following model:

$${IR}_i={\beta }_0+{\beta }_1{LowAVGG}_{t-1}+{\beta }_2\mathit{COVID-}19+{\beta }_3{LowAVGG}_{t-1}\ast \mathit{COVID-}19+X_{i,t-1}+d_m+{ϵ}_i$$

(6)

Here, $\mathit{COVID-}19$ is a dummy variable set to 1 if the data lies between 2 January 2020 and 31 December 2020 and 0 otherwise.

Then, we analyzed the impact of an exogenous increase in climate concern on IPO firms’ initial returns when firms were green or brown. Specifically, we regressed initial returns on the firms’ g-score, investors’ change of environment concern in the last month, and control variables by using the following PST (2022) model:

$${IR}_i={\beta }_0+{\beta }_1{LowAVGG}_{t-1}+{\beta }_2{Delta}_m+{\beta }_3{LowAVGG}_{t-1}\ast {Delta}_m+{{\beta }_4{LowAVGG}_{t-1}\ast {Delta}_{m-1}+\beta }_5^{\prime }{X}_{i,t-1}+{ϵ}_i$$

(7)

Here, ${Delta}_m$ is the change in the aggregate climate concern for that month (see Daly et al., 2021), whereas ${Delta}_{m-1}$ is the lagged change in the aggregate climate concern. $X_{i,t-1}$ is a vector of control variables that we take in Equation (5). ${ϵ}_{i,t}$ is the white noise when standard errors are heteroscedasticity-consistent robust.9

Next, we examined the impact of the firm’s greenness and investors’ sentiment of environment risk on the IPO firm’s subsequent short-term holding period returns. Here, our dependent variable was the subsequent 3- and 6-month holding period returns by using the following PST (2022) model:

$${HPR}_i=={\beta }_0+{\beta }_1{LowAVGG}_{t-1}+{\beta }_2{Delta}_m+{\beta }_3{LowAVGG}_{t-1}\ast {Delta}_m+{{\beta }_4{LowAVGG}_{t-1}\ast {Delta}_{m-1}+\beta }_5^{\prime }{X}_{i,t-1}+{ϵ}_i$$

(8)

The main difference in Equation (8) from Equation (7) is the dependent variable. The dependent variable, ${HPR}_i$, captures subsequent 3- and 6-month HPR (adjusted HPR) returns, which are calculated using Equations (3) and (4), respectively. We considered four proxies (e.g., 3- and 6-month actual or adjusted buy-and-hold returns) for the subsequent performance analysis.

4. Empirical Results

4.1. Descriptive Statistics

Table 1. Summary statistics: IPO initial returns. This table presents the summary statistics for the initial return and adjusted initial return of initial public offerings from February 2020 to December 2020. The sample consists of 173 firms. Initial return is the ratio of the difference between the closing price on the first day of trading and the offer price divided by the offer price.11 Adjusted initial return represents a market-adjusted return, which is the difference between the initial return and the S&P 500 return on the same day. Panel A shows the return distribution of IPO firms for the entire sample. Panel B shows the mean of initial and adjusted initial returns of IPO firms by their year and month of IPO.

Panel A: Summary Statistics of IPO Returns
	Mean		Std. Dev	p5	p25		Median	p75		p95	N
Initial Return	0.341		0.531	−0.189	0.000		0.150	0.540		1.579	173
Adj. Initial Return	0.343		0.529	−0.195	0.002		0.145	0.546		1.575	173
Panel B: Monthly IPO Mean Returns
Year: 2020
Month		Initial Return				Adj. Initial Return			#IPOs
February		0.139				0.138			15
March		−0.006				0.048			3
April		0.195				0.185			6
May		0.177				0.178			10
June		0.283				0.284			21
July		0.434				0.434			18
August		0.444				0.443			20
September		0.275				0.282			24
October		0.177				0.179			27
November		0.355				0.353			7
December		0.794				0.793			22

presents the summary statistics of the initial and adjusted initial returns for IPOs from February 2020 to December 2020.10 In Panel A, we have shown the distribution of initial and adjusted initial returns for our entire sample. The mean first-day return for the IPO firms is 34.1%, while the median first-day return is 15.0%. The total number of IPOs in our sample is 173. In Panel B, we have shown the distribution of initial and adjusted initial returns of the sample period by month. The results have displayed that higher underpricing was mostly concentrated in July, August, and December. The first-day returns for all other months except March were positive, indicating the presence of underpricing. The number of IPOs was also concentrated in the second half of the sample period. Figure 1 displays the bar chart of the change in investors’ environment sentiment, initial returns, and adjusted initial returns by month.

Table 2. Descriptive statistics. This table presents the descriptive statistics for variables used in the regression. AVG G is the firm’s greenness (see Pástor et al., 2022). Delta is the change in the aggregate climate concern for a month (see Daly et al., 2021). Lag_Delta is the lagged change in the aggregate climate concern. Asset is the total asset size in millions. Price is the IPO offer price. Offer Size is the natural log of total proceeds of IPOs. NASDAQ, a dummy variable, takes 1 if the firm is traded in NASDAQ and 0 otherwise. Hi-Tech is a dummy variable equal to 1, if a firm has a certain SIC code (see Loughran & Ritter, 2004), and 0 otherwise. Market capitalization is calculated based on post-IPO shares and first-day closing price. R&D represents R&D expenditures scaled by total assets. Log volume is the log of the first-day trading volume. Leverage is the sum of long-term and short-term debt scaled by total assets. ROA represents net income scaled by total assets. HPR3 (Adj. HPR3) is the 3-month buy-hold return (market-adjusted return) of the IPO firm starting on the first day of trading. HPR6 (Adj. HPR6) is the 6-month buy-hold return (market-adjusted return) of the IPO firm starting on the first day of trading.

	Mean	Std. Dev	p5	p25	Median	p75	p95	N
AVG G	0.337	0.657	−1.040	0.489	0.567	0.732	0.835	173
Delta	−2.925	137.682	−164.000	−115.000	−18.000	106.000	275.000	173
Lag_Delta	15.740	144.710	−164.000	−115.000	−18.000	106.000	275.000	173
Asset	562.368	1872.729	0.027	2.572	48.412	225.424	2981.829	173
Price	17.696	12.934	5.500	10.000	17.000	20.000	28.000	173
OfferSize	18.887	1.272	16.341	18.293	19.063	19.519	20.822	173
NASDAQ	0.832	0.375	0.000	1.000	1.000	1.000	1.000	173
Hi-tech	0.064	0.245	0.000	0.000	0.000	0.000	1.000	173
Market Capitalization	1999.310	4997.380	43.435	227.932	639.520	1409.600	8652.740	173
R&D	29.528	84.288	0.000	0.000	10.632	36.102	102.052	173
Log(Volume)	11.490	6.266	1.064	4.961	14.635	15.954	17.923	173
Leverage	0.517	1.349	0.000	0.000	0.166	0.597	1.838	173
ROA	−2.627	15.468	−3.484	−0.738	−0.205	0.000	0.104	173
HPR3	0.243	0.675	−0.399	−0.082	0.087	0.390	1.145	173
HPR6	0.184	0.715	−0.577	−0.188	0.055	0.363	1.384	173
Adj. HPR3	0.170	0.660	−0.446	−0.142	0.039	0.323	1.074	173
Adj. HPR6	0.020	0.708	−0.720	−0.344	−0.078	0.163	1.222	173

presents descriptive statistics for our variables of interest and the other control variables used in the baseline regression. Appendix A provides variables’ descriptions. AVG G is the average MSCI industry g-score for the firms. The mean g-score is 0.337, and the median value is 0.567. ${Delta}_m$ (${Delta}_{m-1}$) is the change in the aggregate monthly (lagged monthly) environmental concern. The mean value of ${Delta}_m$ (${Delta}_{m-1}$) is −2.93 (15.74). The average asset size of the IPO in our sample is 562.37 million, while the 50th percentile value is 48.41 million. The mean offer price for IPO firms is $17.70.12 The mean log of proceeds is 18.89, while the median value is 19.06. Around 83.2% of the IPO firms in our sample are traded in NASDAQ. The mean market capitalization of IPO firms is $1999.31 million. The mean leverage ratio of the sample firms is 0.52, whereas the mean ROA is −2.63. The post-performance of IPO day returns is calculated as holding period return (HPR) using Equation (3). The average 3-month HPR for the IPO firms is 24.3%, while the 6-month average HPR is 18.4%.

Table 3. Industry greenness and average IPO returns. This table presents the average initial return, adjusted initial return, and greenness by industry from February 2020–December 2020. Industries are defined as Fama–French 49 industries. The sample consists of 173 firms.

	Initial Return	Adj. Initial Return	AVG G
Agriculture	0.150	0.145	−2.019
Almost Nothing	0.044	0.047	0.677
Business Services	0.498	0.498	0.478
Business Supplies	−0.219	−0.211	−3.783
Computer Hardware	−0.162	−0.138	−0.391
Computer Software	0.521	0.522	0.078
Construction	0.250	0.243	−0.116
Construction Materials	0.087	0.080	−0.444
Consumer Goods	0.125	0.122	−0.116
Electrical Equipment	0.660	0.660	−0.444
Electronic Equipment	0.110	0.116	−0.524
Entertainment	0.286	0.283	−0.542
Food Products	0.855	0.878	−2.019
Healthcare	0.146	0.156	0.761
Machinery	1.600	1.601	−1.040
Measuring and Control Equipment	1.713	1.713	0.199
Medical Equipment	0.843	0.844	0.777
Personal Services	0.066	0.068	0.667
Pharmaceutical Products	0.338	0.339	0.560
Precious Metals	−0.057	−0.022	−2.193
Real Estate	0.003	−0.005	−0.548
Restaraunts, Hotels, Motels	0.450	0.452	−0.629
Retail	0.540	0.539	0.215
Shipbuilding, Railroad Equipment	0.444	0.428	−0.173
Trading	−0.016	−0.008	0.197

presents the average initial and adjusted initial returns, as well as the g-score, by industry. A higher g-score indicates greater environmental friendliness. The medical equipment and personal services industries have higher average g-scores, suggesting they are more environmentally friendly. In contrast, firms in the precious metals and agriculture industries tend to be less environmentally friendly. A careful analysis reveals that higher average initial and adjusted initial returns are primarily concentrated in various equipment-related industries. In contrast, firms in the business supply, computer hardware, precious metals, and trading industries exhibit negative first-day returns.

4.2. Statistical Test Results

4.2.1. Green Score and Initial Returns

Table 4. Firm’s greenness on IPO initial return. This table presents regression results of the firm’s greenness on initial returns and adjusted initial returns of IPOs from February 2020 to December 2020. Initial return is the ratio of the difference between the closing price on the first day of trading and the offer price divided by the offer price. Adjusted initial return represents a market-adjusted return, where S&P 500 is the proxy of the market. AVG G is the firm’s greenness (see Pástor et al., 2022). We define ${LowAVGG}_{t-1}$ firms as the firms whose greenness is below the median cutoff. Log (Asset) is the natural log of total assets. Price is the IPO offer price. Offer Size is the natural log of total proceeds of IPOs. NASDAQ, a dummy variable, takes 1 if the firm is traded in NASDAQ and 0 otherwise. Hi-Tech is a dummy variable equal to 1, if a firm has a certain SIC code, and 0 otherwise. Market capitalization is calculated based on post-IPO shares and first-day closing price. R&D represents R&D expenditures scaled by total assets. Log volume is the log of the first-day trading volume. Leverage is the sum of long-term and short-term debt scaled by total assets. ROA represents net income scaled by total assets. Month fixed effects are included in all specifications. Standard errors are heteroskedasticity-consistent standard errors. t-statistics are displayed in parentheses. Significance levels are denoted by *, **, and ***, which correspond to the 10%, 5%, and 1% levels, respectively.

	(1)	(2)	(3)	(4)
VARIABLES	Initial Return	Adj. Initial Return	Initial Return	Adj. Initial Return
${\mathrm{L}\mathrm{o}\mathrm{w}\mathrm{A}\mathrm{V}\mathrm{G}\mathrm{G}}_{\mathrm{t}-1}$	0.1938 **	0.1925 **	0.1917 *	0.1811 *
	(2.5629)	(2.5383)	(1.8012)	(1.7563)
${\mathrm{L}\mathrm{o}\mathrm{g}\left(\mathrm{A}\mathrm{s}\mathrm{s}\mathrm{e}\mathrm{t}\right)}_{\mathrm{t}-1}$			−0.0031	−0.0030
			(−0.2427)	(−0.2420)
${\mathrm{P}\mathrm{r}\mathrm{i}\mathrm{c}\mathrm{e}}_{\mathrm{t}-1}$			0.0046	0.0055
			(1.0848)	(1.4301)
${\mathrm{O}\mathrm{f}\mathrm{f}\mathrm{e}\mathrm{r}\mathrm{S}\mathrm{i}\mathrm{z}\mathrm{e}}_{\mathrm{t}-1}$			0.0541	0.0514
			(1.0993)	(1.0477)
${\mathrm{N}\mathrm{A}\mathrm{S}\mathrm{D}\mathrm{A}\mathrm{Q}}_{\mathrm{t}-1}$			0.1066	0.1030
			(1.2905)	(1.2636)
Hi-Tech			0.5411 **	0.5425 **
			(2.5228)	(2.5627)
Market Capitalization			0.0000	0.0000
			(0.5924)	(0.5108)
${\mathrm{R}\&\mathrm{D}}_{\mathrm{t}-1}$			−0.0003	−0.0003
			(−0.8672)	(−0.9259)
Log(Volume)			−0.0012	−0.0017
			(−0.1074)	(−0.1549)
${\mathrm{L}\mathrm{e}\mathrm{v}\mathrm{e}\mathrm{r}\mathrm{a}\mathrm{g}\mathrm{e}}_{\mathrm{t}-1}$			−0.0401	−0.0404
			(−1.6211)	(−1.6423)
${\mathrm{R}\mathrm{O}\mathrm{A}}_{\mathrm{t}-1}$			0.0018	0.0017
			(1.0178)	(0.9849)
Constant	0.2087 ***	0.2112 ***	−1.1166	−1.0652
	(3.7834)	(3.8510)	(−1.2682)	(−1.2124)
Observations	173	173	173	173
R-squared	0.0291	0.0290	0.2399	0.2507
Month FE	NO	NO	YES	YES

presents the baseline regression results for the initial returns of 173 IPOs using Equation (5). Columns 1 and 2 report the univariate results for the g-score, while columns 3 and 4 present the regression results of the g-score on IPO underpricing after controlling for all control variables. We consider two proxies for the initial returns following Equations (1) and (2): initial returns or adjusted initial returns. Consistent with the information asymmetry hypothesis, Table 4 provides evidence that IPO firms with lower g-scores tend to exhibit higher underpricing. In column 1, the coefficient of ${LowAVGG}_{t-1}$ is positive and statistically significant at 1% level. It implies that brown firms have more underpricing on average than green firms during the COVID-19 period. Our finding is consistent with the notion that higher ESG scores contain value-relevant information, such as higher quality disclosures (Lopez-de-Silanes et al., 2020) and lower information asymmetry (El Ghoul et al., 2011), which leads to more underpricing for brown firms. Firms with lower g-scores have more information asymmetry (El Ghoul et al., 2011). Consequently, it leads to more underpricing. We found robust evidence in column 2 when we used adjusted initial returns. In columns 3 and 4, we performed the same regression taking standard controls and found the evidence robust. The results are economically significant, as brown firms are 19.17% (18.11%) more underpriced than green firms during the COVID-19 crisis period when we consider initial IPO returns (adjusted initial IPO returns).

4.2.2. COVID-19 as an Exogenous Shock: Firm’s Greenness on IPO Initial Return

The findings so far provided evidence that the ex-ante firm’s greenness negatively affected the IPO underpricing during the COVID-19 crisis period when the overall pollution decreased significantly. We further extended the analysis to see whether the negative association was unique during the crisis periods or was common to the pre-crisis periods. To address this issue, we adopted a difference-in-difference (DID) model with month fixed effects. In this analysis, our sample started in January 2019 (one year before the crisis began) and ended in December 2020.

Table 5. COVID-19 as exogenous shock: firm’s greenness on IPO initial return. This table presents regression results of the firm’s greenness on initial returns and adjusted initial returns of IPOs considering COVID-19 as an exogenous shock. The sample starts from January 2019 to December 2020. COVID-19 is a dummy variable if IPO date is in 2020 and 0 otherwise. Initial return is the ratio of the difference between the closing price on the first day of trading and the offer price divided by the offer price. Adjusted initial return represents a market-adjusted return, where S&P 500 is the proxy of the market. AVG G is the firm’s greenness (see Pástor et al., 2022). We define ${LowAVGG}_{t-1}$ firms as the firms whose greenness is below the median cutoff. Log (Asset) is the natural log of total assets. Price is the IPO offer price. Offer Size is the natural log of total proceeds of IPOs. NASDAQ, a dummy variable, takes 1 if the firm is traded in NASDAQ and 0 otherwise. Hi-Tech is a dummy variable equal to 1, if a firm has a certain SIC code, and 0 otherwise. Market capitalization is calculated based on post-IPO shares and first-day closing price. R&D represents R&D expenditures scaled by total assets. Log volume is the log of the first-day trading volume. Leverage is the sum of long-term and short-term debt scaled by total assets. ROA represents net income scaled by total assets. Standard errors are heteroskedasticity-consistent standard errors. t-statistics are displayed in parentheses. Significance levels are denoted by *, **, and ***, which correspond to the 10%, 5%, and 1% levels, respectively.

	(1)	(2)
VARIABLES	Initial Return	Adj. Initial Return
${\mathrm{L}\mathrm{o}\mathrm{w}\mathrm{A}\mathrm{V}\mathrm{G}\mathrm{G}}_{\mathrm{t}-1}$	−0.0097	−0.0084
	(−0.1356)	(−0.1167)
$\mathrm{COVID-}19$	0.6047 ***	0.6066 ***
	(4.0575)	(4.1274)
${\mathrm{L}\mathrm{o}\mathrm{w}\mathrm{A}\mathrm{V}\mathrm{G}\mathrm{G}}_{\mathrm{t}-1}\ast \mathrm{COVID-}19$	0.1939 *	0.1824 *
	(1.8087)	(1.7485)
${\mathrm{L}\mathrm{o}\mathrm{g}\left(\mathrm{A}\mathrm{s}\mathrm{s}\mathrm{e}\mathrm{t}\right)}_{\mathrm{t}-1}$	0.0024	0.0026
	(0.2555)	(0.2853)
${\mathrm{P}\mathrm{r}\mathrm{i}\mathrm{c}\mathrm{e}}_{\mathrm{t}-1}$	0.0075 **	0.0079 ***
	(2.3780)	(2.6192)
${\mathrm{O}\mathrm{f}\mathrm{f}\mathrm{e}\mathrm{r}\mathrm{S}\mathrm{i}\mathrm{z}\mathrm{e}}_{\mathrm{t}-1}$	0.0147	0.0148
	(0.4411)	(0.4513)
${\mathrm{N}\mathrm{A}\mathrm{S}\mathrm{D}\mathrm{A}\mathrm{Q}}_{\mathrm{t}-1}$	0.0236	0.0181
	(0.4125)	(0.3255)
Hi-Tech	0.2705 **	0.2726 **
	(2.4525)	(2.4967)
Market Capitalization	0.0000	0.0000
	(0.5850)	(0.5029)
${\mathrm{R}\&\mathrm{D}}_{\mathrm{t}-1}$	−0.0005	−0.0005
	(−1.3338)	(−1.3485)
Log(Volume)	0.0017	0.0009
	(0.2028)	(0.1107)
${\mathrm{L}\mathrm{e}\mathrm{v}\mathrm{e}\mathrm{r}\mathrm{a}\mathrm{g}\mathrm{e}}_{\mathrm{t}-1}$	−0.0344 *	−0.0338 *
	(−1.7295)	(−1.7219)
${\mathrm{R}\mathrm{O}\mathrm{A}}_{\mathrm{t}-1}$	0.0022	0.0021
	(1.3355)	(1.2958)
Constant	−0.3844	−0.3712
	(−0.6959)	(−0.6816)
Observations	307	307
R-squared	0.1871	0.1971

presents the regression results examining the relationship between firms’ greenness and IPO returns. In both specifications, we found that the interaction between firms’ greenness and COVID-19 is negatively associated with IPO underpricing, indicating that brown firms experienced higher IPO underpricing during the COVID-19 crisis period. This finding further supports our conclusion that brown firms’ IPO underpricing was more pronounced during the crisis period. It is generally assumed that information asymmetry is higher during crises. Since brown firms tend to have greater information asymmetry (El Ghoul et al., 2011), the significant interaction effect reinforces the robustness of this association.

4.2.3. Green Score, Climate Concern, and Initial Returns

In the previous section, we found that underpricing was higher for brown firms, possibly due to lower information disclosure leading to greater information asymmetry. However, besides the positive news of pollution reduction or waivers of environmental regulations, the media coverage of environmental concerns also increased in some of the months due to the increased disposal of medical waste during the COVID-19 period. Thus, it is important to understand the impact of changing climate concerns on brown firms’ IPO underpricing when the overall belief about environmental pollution was low. In this subsection, we analyzed the impact of an exogenous increase of climate concern on IPO firms’ initial returns when firms were green or brown using Equation (7).

Table 6. Firm’s greenness and climate concern on IPO initial return. This table presents regression results of the firm’s greenness and climate concern on initial returns and adjusted initial returns of IPOs from February 2020 to December 2020. Initial return is the ratio of the difference between the closing price on the first day of trading and the offer price divided by the offer price. Adjusted initial return represents a market-adjusted return, where S&P 500 is the proxy of the market. AVG G is the firm’s greenness (see Pástor et al., 2022). We define ${LowAVGG}_{t-1}$ firms as the firms whose greenness is below the median cutoff. ${Delta}_m$ is the change in the aggregate climate concern for that month (see Daly et al., 2021). ${Delta}_{m-1}$ is the lagged change in the aggregate climate concern. Log (Asset) is the natural log of total assets. Price is the IPO offer price. Offer Size is the natural log of total proceeds of IPOs. NASDAQ, a dummy variable, takes 1 if the firm is traded in NASDAQ and 0 otherwise. Hi-Tech is a dummy variable equal to 1, if a firm has a certain SIC code, and 0 otherwise. Market capitalization is calculated based on post-IPO shares and first-day closing price. R&D represents R&D expenditures scaled by total assets. Log volume is the log of the first-day trading volume. Leverage is the sum of long-term and short-term debt scaled by total assets. ROA represents net income scaled by total assets. Standard errors are heteroskedasticity-consistent standard errors. t-statistics are displayed in parentheses. Significance levels are denoted by *, **, and ***, which correspond to the 10%, 5%, and 1% levels, respectively.

	(1)	(2)
VARIABLES	Initial Return	Adj. Initial Return
${\mathrm{L}\mathrm{o}\mathrm{w}\mathrm{A}\mathrm{V}\mathrm{G}\mathrm{G}}_{\mathrm{t}-1}$	0.1475	0.1451
	(1.4873)	(1.4732)
${\mathrm{D}\mathrm{e}\mathrm{l}\mathrm{t}\mathrm{a}}_{\mathrm{m}}$	0.0000	0.0001
	(0.0952)	(0.1660)
${\mathrm{L}\mathrm{o}\mathrm{w}\mathrm{A}\mathrm{V}\mathrm{G}\mathrm{G}}_{\mathrm{t}-1}\ast {\mathrm{D}\mathrm{e}\mathrm{l}\mathrm{t}\mathrm{a}}_{\mathrm{m}}$	−0.0007 *	−0.0007 *
	(−1.6771)	(−1.6661)
${\mathrm{L}\mathrm{o}\mathrm{w}\mathrm{A}\mathrm{V}\mathrm{G}\mathrm{G}}_{\mathrm{t}-1}\ast {\mathrm{D}\mathrm{e}\mathrm{l}\mathrm{t}\mathrm{a}}_{\mathrm{m}-1}$	−0.0009 ***	−0.0009 ***
	(−3.3490)	(−3.3347)
${\mathrm{L}\mathrm{o}\mathrm{g}\left(\mathrm{A}\mathrm{s}\mathrm{s}\mathrm{e}\mathrm{t}\right)}_{\mathrm{t}-1}$	−0.0044	−0.0045
	(−0.3632)	(−0.3728)
${\mathrm{P}\mathrm{r}\mathrm{i}\mathrm{c}\mathrm{e}}_{\mathrm{t}-1}$	0.0072 *	0.0074 *
	(1.7701)	(1.8348)
${\mathrm{O}\mathrm{f}\mathrm{f}\mathrm{e}\mathrm{r}\mathrm{S}\mathrm{i}\mathrm{z}\mathrm{e}}_{\mathrm{t}-1}$	−0.0171	−0.0172
	(−0.3836)	(−0.3854)
${\mathrm{N}\mathrm{A}\mathrm{S}\mathrm{D}\mathrm{A}\mathrm{Q}}_{\mathrm{t}-1}$	0.1361 *	0.1379 *
	(1.7856)	(1.8062)
Hi-Tech	0.4601 **	0.4588 **
	(2.0810)	(2.0929)
Market Capitalization	0.0000	0.0000
	(0.8670)	(0.8540)
${\mathrm{R}\&\mathrm{D}}_{\mathrm{t}-1}$	0.0000	0.0000
	(0.0641)	(0.0443)
Log(Volume)	0.0189 **	0.0184 **
	(2.3318)	(2.2872)
${\mathrm{L}\mathrm{e}\mathrm{v}\mathrm{e}\mathrm{r}\mathrm{a}\mathrm{g}\mathrm{e}}_{\mathrm{t}-1}$	−0.0302	−0.0304
	(−1.1628)	(−1.1748)
${\mathrm{R}\mathrm{O}\mathrm{A}}_{\mathrm{t}-1}$	0.0025	0.0026
	(1.4405)	(1.4630)
Constant	0.0973	0.1024
	(0.1175)	(0.1237)
Observations	173	173
R-squared	0.3436	0.2722

presents the regression results of IPO initial returns, on g-score, and change of climate concern after controlling for all the control variables. In column 1, we interacted brown firms with contemporaneous and lagged changes in the climate concern. We found that both of the interactions are negative and statistically significant. It implies that when there is an increase in climate concern, investors move toward investing more in green firms, which drives the lower initial returns for brown firms (Da et al., 2011; L. X. Liu et al., 2014). The results also indicate that investors tend to invest in safer stocks (flight to safety) during higher exogenous uncertainty (Bekaert et al., 2009). Economically, a one standard deviation increase in the change in the lagged climate concern was associated with a 13.0% decline in the initial returns for the brown firms. Our results are robust and consistent if we use adjusted initial returns. Results in Table 4, Table 5 and Table 6 imply that brown firms have more underpricing than green firms when we did not account for the exogenous change in the climate shock. However, when we accounted for the change in the climate concern, we found that heightened climate concern drives investors toward green firms and causes lower initial returns for brown firms.

4.2.4. Green Score, Climate Concern, and Initial Returns Before and During the COVID-19 Crisis Period

So far, the paper established a link between IPO underpricing and firms’ greenness during the COVID-19 crisis period. We argued that the IPO underpricing became low for the brown firms if environment concerns increased during the COVID-19 crisis period, as investors have a tendency to fly to safety toward green stocks. One may be curious to find the cause of the association, whether it stems from environment concern or COVID-19 crisis periods. To distinguish the association, we interacted COVID-19 with environment risk concern. In

Table 7. Climate concern and IPO initial return before and during the COVID-19 crisis period. This table presents regression results of the climate concern on initial returns and adjusted initial returns of IPOs before and during the COVID-19 crisis period. COVID-19 is a dummy variable if IPO date is in 2020 and 0 otherwise. Initial return is the ratio of the difference between the closing price on the first day of trading and the offer price divided by the offer price. Adjusted initial return represents a market-adjusted return, where S&P 500 is the proxy of the market. ${Delta}_{m-1}$ is the lagged change in the aggregate climate concern (see Daly et al., 2021). Log (Asset) is the natural log of total assets. Price is the IPO offer price. Offer Size is the natural log of total proceeds of IPOs. NASDAQ, a dummy variable, takes 1 if the firm is traded in NASDAQ and 0 otherwise. Hi-Tech is a dummy variable equal to 1, if a firm has a certain SIC code, and 0 otherwise. Market capitalization is calculated based on post-IPO shares and first-day closing price. R&D represents R&D expenditures scaled by total assets. Log volume is the log of the first-day trading volume. Leverage is the sum of long-term and short-term debt scaled by total assets. ROA represents net income scaled by total assets. Standard errors are heteroskedasticity-consistent standard errors. t-statistics are displayed in parentheses. Significance levels are denoted by *, **, and ***, which correspond to the 10%, 5%, and 1% levels, respectively.

	(1)	(2)
VARIABLES	Initial Return	Adj. Initial Return
$\mathrm{COVID-}19$	0.1851 ***	0.1831 ***
	(3.0729)	(3.1016)
${\mathrm{D}\mathrm{e}\mathrm{l}\mathrm{t}\mathrm{a}}_{\mathrm{m}-1}$	0.0002	0.0002
	(0.6816)	(0.5936)
$\mathrm{COVID-}19\ast {\mathrm{D}\mathrm{e}\mathrm{l}\mathrm{t}\mathrm{a}}_{\mathrm{m}-1}$	−0.0008 **	−0.0008 **
	(−2.0386)	(−2.0051)
${\mathrm{L}\mathrm{o}\mathrm{g}\left(\mathrm{A}\mathrm{s}\mathrm{s}\mathrm{e}\mathrm{t}\right)}_{\mathrm{t}-1}$	0.0152 **	0.0149 **
	(2.2398)	(2.2467)
${\mathrm{P}\mathrm{r}\mathrm{i}\mathrm{c}\mathrm{e}}_{\mathrm{t}-1}$	0.0094 ***	0.0098 ***
	(2.6314)	(2.8129)
${\mathrm{O}\mathrm{f}\mathrm{f}\mathrm{e}\mathrm{r}\mathrm{S}\mathrm{i}\mathrm{z}\mathrm{e}}_{\mathrm{t}-1}$	−0.0267	−0.0243
	(−0.8449)	(−0.8146)
${\mathrm{N}\mathrm{A}\mathrm{S}\mathrm{D}\mathrm{A}\mathrm{Q}}_{\mathrm{t}-1}$	0.0289	0.0265
	(0.5562)	(0.5145)
Hi-Tech	0.2062 *	0.2104 *
	(1.8430)	(1.9034)
Market Capitalization	0.0000	0.0000
	(0.6732)	(0.5975)
${\mathrm{R}\&\mathrm{D}}_{\mathrm{t}-1}$	−0.0005	−0.0005
	(−1.1442)	(−1.1492)
Log(Volume)	0.0111 *	0.0102
	(1.7194)	(1.6037)
${\mathrm{L}\mathrm{e}\mathrm{v}\mathrm{e}\mathrm{r}\mathrm{a}\mathrm{g}\mathrm{e}}_{\mathrm{t}-1}$	−0.0191	−0.0185
	(−0.7819)	(−0.7646)
${\mathrm{R}\mathrm{O}\mathrm{A}}_{\mathrm{t}-1}$	0.0014	0.0013
	(0.7841)	(0.7231)
Constant	0.3043	0.2690
	(0.5379)	(0.5046)
Observations	307	307
R-squared	0.1288	0.1398

, we found that the interaction between COVID-19 and lag difference of environment concern negatively associated with IPO underpricing and the association was statistically significant at 1% level. The results indicate that environment risk concern reduces IPO underpricing during the COVID-19 crisis period, which is consistent with the uncertainty hypothesis of Mazumder and Saha (2021). Behavioral finance studies reveal that investors’ negative sentiments and mood affect financial decision-making and asset pricing (Da et al., 2015).

Economically, a one standard deviation increase in the change in the lagged climate concern was associated with a 10.12% decline in the initial returns during the COVID-19 crisis period. Our results were robust and consistent if we used adjusted initial returns. Results in Table 5 and Table 7 imply that IPO underpricing is higher in the COVID-19 crisis period consistent with Mazumder and Saha (2021). However, when we accounted for the change in the climate concern, we found that heightened climate concern led to lower IPO underpricing during the COVID-19 crisis period.

4.2.5. Green Score, Climate Concern, and Subsequent Holding Period Returns

Table 8. Firm’s greenness and climate concern on IPO subsequent holding period return. This table presents regression results of the firm’s greenness and climate concern on subsequent holding period returns of IPOs from February 2020 to December 2020. HPR3 (Adj. HPR3) is the 3-month buy-hold return (market-adjusted return) of the IPO firm starting on the first day of trading. HPR6 (Adj. HPR6) is the 6-month buy-hold return (market-adjusted return) of the IPO firm starting on the first day of trading. AVG G is the firm’s greenness (see Pástor et al., 2022). We define ${LowAVGG}_{t-1}$ firms as the firms whose greenness is below the median cutoff. ${Delta}_m$ is the change in the aggregate climate concern for that month (see Daly et al., 2021). ${Delta}_{m-1}$ is the lagged change in the aggregate climate concern. Log (Asset) is the natural log of total assets. Price is the IPO offer price. Offer Size is the natural log of total proceeds of IPOs. NASDAQ, a dummy variable, takes 1 if the firm is traded in NASDAQ and 0 otherwise. Hi-Tech is a dummy variable equal to 1, if a firm has a certain SIC code, and 0 otherwise. Market capitalization is calculated based on post-IPO shares and first-day closing price. R&D represents R&D expenditures scaled by total assets. Log volume is the log of the first-day trading volume. Leverage is the sum of long-term and short-term debt scaled by total assets. ROA represents net income scaled by total assets. Standard errors are heteroskedasticity-consistent standard errors. t-statistics are displayed in parentheses. Significance levels are denoted by *, **, and ***, which correspond to the 10%, 5%, and 1% levels, respectively.

	(1)	(2)	(3)	(4)
VARIABLES	HPR3	Adj. HPR 3	HPR 6	Adj. HPR 6
${\mathrm{L}\mathrm{o}\mathrm{w}\mathrm{A}\mathrm{V}\mathrm{G}\mathrm{G}}_{\mathrm{t}-1}$	0.1059	0.0928	−0.0439	−0.0693
	(1.0930)	(0.9656)	(−0.3736)	(−0.5794)
${\mathrm{D}\mathrm{e}\mathrm{l}\mathrm{t}\mathrm{a}}_{\mathrm{m}}$	−0.0000	−0.0002	−0.0003	−0.0001
	(−0.0063)	(−0.6333)	(−0.7047)	(−0.1967)
${\mathrm{L}\mathrm{o}\mathrm{w}\mathrm{A}\mathrm{V}\mathrm{G}\mathrm{G}}_{\mathrm{t}-1}\ast {\mathrm{D}\mathrm{e}\mathrm{l}\mathrm{t}\mathrm{a}}_{\mathrm{m}}$	0.0014 **	0.0014 **	0.0017 ***	0.0014 **
	(2.2327)	(2.2701)	(2.6445)	(2.3454)
${\mathrm{L}\mathrm{o}\mathrm{w}\mathrm{A}\mathrm{V}\mathrm{G}\mathrm{G}}_{\mathrm{t}-1}\ast {\mathrm{D}\mathrm{e}\mathrm{l}\mathrm{t}\mathrm{a}}_{\mathrm{m}-1}$	0.0006 *	0.0006 *	0.0000	−0.0000
	(1.8292)	(1.7998)	(0.0087)	(−0.1102)
${\mathrm{L}\mathrm{o}\mathrm{g}\left(\mathrm{A}\mathrm{s}\mathrm{s}\mathrm{e}\mathrm{t}\right)}_{\mathrm{t}-1}$	0.0067	0.0076	0.0328 *	0.0238
	(0.4828)	(0.5454)	(1.7353)	(1.3052)
${\mathrm{P}\mathrm{r}\mathrm{i}\mathrm{c}\mathrm{e}}_{\mathrm{t}-1}$	−0.0013	−0.0014	−0.0127 ***	−0.0127 ***
	(−0.4055)	(−0.4556)	(−2.9965)	(−2.7999)
${\mathrm{O}\mathrm{f}\mathrm{f}\mathrm{e}\mathrm{r}\mathrm{S}\mathrm{i}\mathrm{z}\mathrm{e}}_{\mathrm{t}-1}$	0.0001	−0.0029	0.0719	0.0696
	(0.0018)	(−0.0641)	(1.5442)	(1.5228)
${\mathrm{N}\mathrm{A}\mathrm{S}\mathrm{D}\mathrm{A}\mathrm{Q}}_{\mathrm{t}-1}$	−0.0230	−0.0170	−0.1708	−0.1394
	(−0.2504)	(−0.1944)	(−1.0734)	(−0.8994)
Hi-Tech	0.1456	0.1428	0.0499	0.0660
	(0.8205)	(0.8147)	(0.3205)	(0.3913)
Market Capitalization	−0.0000	−0.0000	0.0000	0.0000
	(−1.3096)	(−1.2614)	(0.1195)	(0.2515)
${\mathrm{R}\&\mathrm{D}}_{\mathrm{t}-1}$	0.0002	0.0002	0.0000	0.0001
	(0.8585)	(0.8005)	(0.0493)	(0.2528)
Log(Volume)	0.0058	0.0057	−0.0147	−0.0124
	(0.7468)	(0.7582)	(−1.4839)	(−1.2464)
${\mathrm{L}\mathrm{e}\mathrm{v}\mathrm{e}\mathrm{r}\mathrm{a}\mathrm{g}\mathrm{e}}_{\mathrm{t}-1}$	−0.1811 **	−0.1774 **	−0.0872	−0.0879
	(−1.9972)	(−1.9804)	(−1.1294)	(−1.1520)
${\mathrm{R}\mathrm{O}\mathrm{A}}_{\mathrm{t}-1}$	−0.0221	−0.0219	−0.0160	−0.0154
	(−1.5993)	(−1.5784)	(−1.2715)	(−1.2292)
Constant	0.1478	0.1326	−0.7283	−0.8546
	(0.1671)	(0.1552)	(−0.8378)	(−0.9995)
Observations	173	173	173	173
R-squared	0.3363	0.3277	0.1729	0.1609

presents the regression results of 173 IPOs’ subsequent returns on the g-score and climate concern, where columns are different in terms of different proxies for buy-and-hold returns. In column 1, our dependent variable is 3-month holding period returns. Notably, the coefficient of ${LowAVGG}_{t-1}$ is statistically insignificant. It signifies no difference in the subsequent holding period return between green and brown stocks when there is no exogenous change in the climate concern. When we interacted the firms’ g-score with the change in climate concern, we found that the coefficient was positive and statistically significant. It implies that brown firms’ subsequent return is significantly more than their green counterpart when there is an exogenous increase in the climate concern. Our findings are consistent with Bolton and Kacperczyk (2021) that investors consider brown firms as risky and require more returns for holding those firms, especially when the media coverage on environmental concerns is higher. Our results are robust when we used the adjusted 3-month holding period return. Moreover, we found similar evidence across 6-month (adjusted) holding period returns. Economically, a 1% increase in the standard deviation of the change in current climate concern was associated with a 19.3% (23.4%) increase in a brown firm’s 3-month (6-month) holding period returns.

4.3. Additional Analysis

4.3.1. Firm-Level Characteristics and IPO Underpricing and Subsequent Returns

In this section, we take a closer look at the firm-level factors that explain IPO underpricing during the COVID-19 period in a sub-sample setting. The cross-sectional analysis helped us to disentangle the association between the firm-level factors and first-day returns. In

Table 9. Firm characteristics and IPO initial return. This table presents regression results of firm characteristics that explain IPO initial returns from February 2020 to December 2020. Columns are different in terms of the firm’s greenness. Log (Asset) is the natural log of total assets. Price is the IPO offer price. Offer Size is the natural log of total proceeds of IPOs. NASDAQ, a dummy variable, takes 1 if the firm is traded in NASDAQ and 0 otherwise. Hi-Tech is a dummy variable equal to 1, if a firm has a certain SIC code, and 0 otherwise. Market capitalization is calculated based on post-IPO shares and the first-day closing price. R&D represents R&D expenditures scaled by total assets. Log volume is the log of the first-day trading volume. Leverage is the sum of long-term and short-term debt scaled by total assets. ROA represents net income scaled by total assets. Standard errors are heteroskedasticity-consistent standard errors. t-statistics are displayed in parentheses. Significance levels are denoted by *, **, and ***, which correspond to the 10%, 5%, and 1% levels, respectively.

	(1)	(2)
VARIABLES	Initial Return	Initial Return
	High AVG G	Low AVG G
${\mathrm{L}\mathrm{o}\mathrm{g}\left(\mathrm{A}\mathrm{s}\mathrm{s}\mathrm{e}\mathrm{t}\right)}_{\mathrm{t}-1}$	−0.0196	−0.0416
	(−0.7170)	(−0.7675)
${\mathrm{P}\mathrm{r}\mathrm{i}\mathrm{c}\mathrm{e}}_{\mathrm{t}-1}$	0.0201 *	0.0059
	(1.8096)	(1.5017)
${\mathrm{O}\mathrm{f}\mathrm{f}\mathrm{e}\mathrm{r}\mathrm{S}\mathrm{i}\mathrm{z}\mathrm{e}}_{\mathrm{t}-1}$	−0.0115	0.1485 **
	(−0.1547)	(2.0858)
${\mathrm{N}\mathrm{A}\mathrm{S}\mathrm{D}\mathrm{A}\mathrm{Q}}_{\mathrm{t}-1}$	0.0428	0.2993
	(0.4701)	(1.2016)
Hi-Tech	0.3284 *	0.5504
	(1.9854)	(1.2985)
Market Capitalization	0.0000	0.0001 **
	(0.0380)	(2.6150)
${\mathrm{R}\&\mathrm{D}}_{\mathrm{t}-1}$	−0.0184 ***	−0.0006
	(−3.8666)	(−1.4557)
Log(Volume)	0.0108	−0.0366 *
	(0.8671)	(−1.7417)
${\mathrm{L}\mathrm{e}\mathrm{v}\mathrm{e}\mathrm{r}\mathrm{a}\mathrm{g}\mathrm{e}}_{\mathrm{t}-1}$	−0.3060 ***	−0.0285
	(−2.9198)	(−0.7686)
${\mathrm{R}\mathrm{O}\mathrm{A}}_{\mathrm{t}-1}$	−0.1852 **	0.0014
	(−2.1235)	(0.6635)
Constant	0.1168	−2.5050 **
	(0.0858)	(−2.2498)
Observations	55	118
R-squared	0.7589	0.3676
Ind. FE	YES	YES
Month FE	YES	YES

, we split our sample into green and brown stocks based on the median g-score. Then, we regressed initial returns on all the control variables used in our baseline regression. Column 1 reports the regression results for the green IPO firms. We found that offer price and Hi-Tech firms explain underpricing positively, while R&D, leverage, and profitability are associated negatively and significantly for green IPO firms. On the other hand, offer size and market capitalization explained underpricing positively, and first-day trading volume explained negatively for brown IPO firms.

Next, we examined the key firm-level factors that explain the subsequent short-term performance of IPO firms during the COVID-19 period in a sub-sample setting. In

Table 10. Firm characteristics and post-IPO holding period return. This table presents regression results of firm characteristics that explain subsequent holding period returns from February 2020 to December 2020. Columns are different in terms of the firm’s greenness. Log (Asset) is the natural log of total assets. Price is the IPO offer price. Offer Size is the natural log of total proceeds of IPOs. NASDAQ, a dummy variable, takes 1 if the firm is traded in NASDAQ and 0 otherwise. Hi-Tech is a dummy variable equal to 1, if a firm has a certain SIC code, and 0 otherwise. Market capitalization is calculated based on post-IPO shares and the first-day closing price. R&D represents R&D expenditures scaled by total assets. Log volume is the log of the first-day trading volume. Leverage is the sum of long-term and short-term debt scaled by total assets. ROA represents net income scaled by total assets. Standard errors are heteroskedasticity-consistent standard errors. t-statistics are displayed in parentheses. Significance levels are denoted by *, **, and ***, which correspond to the 10%, 5%, and 1% levels, respectively.

	(1)	(2)	(3)	(4)
VARIABLES	HPR3	HPR3	HPR6	HPR6
	High AVG G	Low AVG G	High AVG G	Low AVG G
${\mathrm{L}\mathrm{o}\mathrm{g}\left(\mathrm{A}\mathrm{s}\mathrm{s}\mathrm{e}\mathrm{t}\right)}_{\mathrm{t}-1}$	0.0251	−0.0203	0.1158 **	0.0143
	(1.5246)	(−0.3266)	(2.4672)	(0.2780)
${\mathrm{P}\mathrm{r}\mathrm{i}\mathrm{c}\mathrm{e}}_{\mathrm{t}-1}$	0.0269 ***	0.0004	0.0075	−0.0068
	(2.9657)	(0.0744)	(0.4896)	(−1.3544)
${\mathrm{O}\mathrm{f}\mathrm{f}\mathrm{e}\mathrm{r}\mathrm{S}\mathrm{i}\mathrm{z}\mathrm{e}}_{\mathrm{t}-1}$	−0.1933 ***	−0.0528	−0.0006	0.0104
	(−3.5623)	(−0.4484)	(−0.0053)	(0.1088)
${\mathrm{N}\mathrm{A}\mathrm{S}\mathrm{D}\mathrm{A}\mathrm{Q}}_{\mathrm{t}-1}$	−0.1215 *	−0.4388 *	−0.1124	−0.0536
	(−1.8684)	(−1.7795)	(−0.5817)	(−0.2773)
Hi-Tech	0.7482 ***	0.2045	0.8859 **	−0.0647
	(3.5041)	(0.6759)	(2.2332)	(−0.2983)
Market Capitalization	−0.0000 **	−0.0000	−0.0000	−0.0000
	(−2.3721)	(−0.9831)	(−0.1867)	(−0.2631)
${\mathrm{R}\&\mathrm{D}}_{\mathrm{t}-1}$	−0.0065 **	0.0009 *	−0.0121 **	0.0010 *
	(−2.2639)	(1.8485)	(−2.0552)	(1.9824)
Log(Volume)	0.0255 ***	0.0267 *	−0.0194	−0.0165
	(2.7730)	(1.7174)	(−0.9285)	(−1.2081)
${\mathrm{L}\mathrm{e}\mathrm{v}\mathrm{e}\mathrm{r}\mathrm{a}\mathrm{g}\mathrm{e}}_{\mathrm{t}-1}$	−0.2883 ***	−0.2397 *	−0.3047 ***	−0.1453
	(−4.8590)	(−1.7479)	(−2.8612)	(−1.2260)
${\mathrm{R}\mathrm{O}\mathrm{A}}_{\mathrm{t}-1}$	−0.1453 ***	−0.0207	−0.3646 ***	−0.0169
	(−3.1405)	(−1.2752)	(−4.0628)	(−1.1486)
Constant	3.5181 ***	1.2934	0.5639	0.5447
	(3.6963)	(0.6534)	(0.2925)	(0.3477)
Observations	55	118	55	118
R-squared	0.6101	0.1617	0.7727	0.1794
Ind. FE	YES	YES	YES	YES

, we reported the regression results in the same split sample by green and brown stocks based on the median g-score. Then, we regressed 3- and 6-month holding period returns on all the control variables used in the baseline regression. Column 1 reports the regression results for the 3-month holding period returns of green IPO firms. We found that offer price, being Hi-Tech and first-day trading volume, explained positively, and offer size, R&D, leverage, and profitability explained negatively for 3-month holding period return for green IPO firms. On the other hand, being traded in NASDAQ and leverage explained negatively, while R&D and first-day trading volume explained positively to the 3-month holding period return for brown IPO firms. When we turned to 6-month holding period returns, we found firm size and Hi-Tech explained positively, and R&D, leverage, and profitability were negatively associated with green firms’ short-term post-IPO performance.

4.3.2. Firms’ Greenness and IPO Stocks’ Liquidity and Information Asymmetry

Finally, we examined the association between firms’ greenness and the market characteristics of IPO stocks during the COVID-19 crisis period. Specifically, we analyzed stock liquidity using Quoted Spread and Effective Spread and considered Price Impact as a measure of information asymmetry.13 

Table 11. Firm’s greenness and IPO stocks’ liquidity. This table presents regression results of the firm’s greenness on IPO stocks’ liquidity from February 2020 to December 2020. Liquidity measures are Quoted Spread, Effective Spread, Price Impact, and Inverse Price. Price is the logarithm of the mean stock price of an equity. Log Dollar Trade Volume is the logarithm of the average daily dollar trading volume. We define ${LowAVGG}_{t-1}$ firms as the firms whose greenness is below the median cutoff. Log (Asset) is the natural log of total assets. Price is the IPO offer price. Offer Size is the natural log of total proceeds of IPOs. NASDAQ, a dummy variable, takes 1 if the firm is traded in NASDAQ and 0 otherwise. Hi-Tech is a dummy variable equal to 1, if a firm has a certain SIC code, and 0 otherwise. Market capitalization is calculated based on post-IPO shares and first-day closing price. R&D represents R&D expenditures scaled by total assets. Log volume is the log of the first-day trading volume. Leverage is the sum of long-term and short-term debt scaled by total assets. ROA represents net income scaled by total assets. Month fixed effects are included in all specifications. Standard errors are heteroskedasticity-consistent standard errors. t-statistics are displayed in parentheses. Significance levels are denoted by *, **, and ***, which correspond to the 10%, 5%, and 1% levels, respectively.

	(1)	(2)	(3)
VARIABLES	Quoted Spread	Effective Spread	Price Impact
${\mathrm{L}\mathrm{o}\mathrm{w}\mathrm{A}\mathrm{V}\mathrm{G}\mathrm{G}}_{\mathrm{t}-1}$	0.1136 ***	0.0357 ***	0.0342 ***
	(3.8623)	(2.6201)	(4.8376)
${\mathrm{L}\mathrm{o}\mathrm{g}\left(\mathrm{A}\mathrm{s}\mathrm{s}\mathrm{e}\mathrm{t}\right)}_{\mathrm{t}-1}$	0.0005	−0.0005	−0.0003
	(0.1392)	(−0.3132)	(−0.3326)
${\mathrm{P}\mathrm{r}\mathrm{i}\mathrm{c}\mathrm{e}}_{\mathrm{t}-1}$	−0.0101 ***	−0.0047 ***	−0.0028 ***
	(−3.4037)	(−3.3236)	(−3.7644)
${\mathrm{O}\mathrm{f}\mathrm{f}\mathrm{e}\mathrm{r}\mathrm{S}\mathrm{i}\mathrm{z}\mathrm{e}}_{\mathrm{t}-1}$	−0.0081	−0.0076	0.0007
	(−0.5258)	(−1.0840)	(0.1817)
${\mathrm{N}\mathrm{A}\mathrm{S}\mathrm{D}\mathrm{A}\mathrm{Q}}_{\mathrm{t}-1}$	0.0777 ***	0.0298 ***	0.0169 **
	(3.0101)	(2.9337)	(2.3115)
Hi-Tech	−0.0550	−0.0287	−0.0107
	(−1.2415)	(−1.5901)	(−0.8140)
Market Capitalization	−0.0000	−0.0000	−0.0000
	(−0.1882)	(−0.0210)	(−0.2081)
${\mathrm{R}\&\mathrm{D}}_{\mathrm{t}-1}$	−0.0003 **	−0.0001 **	−0.0000 *
	(−2.3561)	(−2.1736)	(−1.8318)
Log(Volume)	−0.0070 ***	−0.0030 ***	−0.0020 ***
	(−2.6929)	(−2.7748)	(−3.0728)
${\mathrm{L}\mathrm{e}\mathrm{v}\mathrm{e}\mathrm{r}\mathrm{a}\mathrm{g}\mathrm{e}}_{\mathrm{t}-1}$	−0.0208 ***	−0.0092 ***	−0.0042 **
	(−2.7107)	(−3.0034)	(−2.0600)
${\mathrm{R}\mathrm{O}\mathrm{A}}_{\mathrm{t}-1}$	−0.0005	−0.0003	−0.0001
	(−0.7482)	(−0.9879)	(−0.6022)
Price	0.0069 ***	0.0034 ***	0.0021 ***
	(5.7140)	(5.9839)	(6.7354)
Log Dollar Trade Volume	−0.0459 ***	−0.0259 ***	−0.0093 ***
	(−3.7500)	(−4.8713)	(−3.1892)
Inverse Price	−1.5258 ***	−0.6984 ***	−0.3478 ***
	(−4.3174)	(−4.7974)	(−4.2571)
Constant	1.1050 ***	0.6820 ***	0.1953 ***
	(3.5211)	(5.0630)	(2.6235)
Observations	173	173	173
R-squared	0.6005	0.6462	0.6406
Month FE	YES	YES	YES

presents the regression results on the impact of firms’ greenness on the liquidity and information asymmetry of 173 IPO stocks during the COVID-19 period. Existing literature suggests that firms’ greenness enhances stock liquidity (Z. Chen et al., 2023). Firms with lower g-scores exhibit greater information asymmetry (El Ghoul et al., 2011), which consequently results in lower liquidity. Consistent with the existing literature, our findings indicate that firms’ greenness improves stock liquidity. Table 11 presents the results, revealing that green firms’ stocks have better liquidity, whereas brown firms’ stocks have worse liquidity. We found robust evidence supporting both liquidity and information asymmetry measures. These results align with our preliminary hypothesis that higher ESG scores convey value-relevant information, such as higher-quality disclosures (Lopez-de-Silanes et al., 2020) and lower information asymmetry (El Ghoul et al., 2011), which ultimately leads to better liquidity for green firms. The findings highlight the broader implications of corporate greenness on IPO market dynamics, underscoring its potential significance in investment decision-making and financial market stability.

Next, we conducted whether exogenous environment concerns could enhance or worsen liquidity. Consistent with Z. Liu et al. (2024), we found that exogenous climate risk can increase stock liquidity perhaps due to higher operational risk. In

Table 12. Firm’s greenness, environmental concerns, and IPO stocks’ liquidity. This table presents regression results of the firm’s greenness and environmental concerns on IPO stocks’ liquidity from February 2020 to December 2020. Liquidity measures are Quoted Spread, Effective Spread, Price Impact, and Inverse Price. Price is the logarithm of the mean stock price of an equity. All variables are described in Appendix A. Month fixed effects are included in all specifications. Standard errors are heteroskedasticity-consistent standard errors. t-statistics are displayed in parentheses. Significance levels are denoted by *, **, and ***, which correspond to the 10%, 5%, and 1% levels, respectively.

	(1)	(2)	(3)
VARIABLES	Quoted Spread	Effective Spread	Price Impact
${\mathrm{L}\mathrm{o}\mathrm{w}\mathrm{A}\mathrm{V}\mathrm{G}\mathrm{G}}_{\mathrm{t}-1}$	0.1481 ***	0.0509 ***	0.0420 ***
	(4.7366)	(3.4133)	(5.6268)
${\mathrm{D}\mathrm{e}\mathrm{l}\mathrm{t}\mathrm{a}}_{\mathrm{m}}$	0.0140 ***	0.0069 ***	0.0041 ***
	(4.8708)	(4.6115)	(5.7580)
${\mathrm{L}\mathrm{o}\mathrm{w}\mathrm{A}\mathrm{V}\mathrm{G}\mathrm{G}}_{\mathrm{t}-1}\ast {\mathrm{D}\mathrm{e}\mathrm{l}\mathrm{t}\mathrm{a}}_{\mathrm{m}}$	−0.0002	−0.0001	−0.0001
	(−1.1548)	(−0.8276)	(−1.5816)
${\mathrm{L}\mathrm{o}\mathrm{w}\mathrm{A}\mathrm{V}\mathrm{G}\mathrm{G}}_{\mathrm{t}-1}\ast {\mathrm{D}\mathrm{e}\mathrm{l}\mathrm{t}\mathrm{a}}_{\mathrm{m}-1}$	−0.0004 **	−0.0001 *	−0.0001 *
	(−2.3491)	(−1.6772)	(−1.8600)
${\mathrm{L}\mathrm{o}\mathrm{g}\left(\mathrm{A}\mathrm{s}\mathrm{s}\mathrm{e}\mathrm{t}\right)}_{\mathrm{t}-1}$	0.0027	0.0005	0.0002
	(0.6028)	(0.2869)	(0.1619)
${\mathrm{P}\mathrm{r}\mathrm{i}\mathrm{c}\mathrm{e}}_{\mathrm{t}-1}$	−0.0128 ***	−0.0060 ***	−0.0034 ***
	(−3.4345)	(−3.3559)	(−3.8688)
${\mathrm{O}\mathrm{f}\mathrm{f}\mathrm{e}\mathrm{r}\mathrm{S}\mathrm{i}\mathrm{z}\mathrm{e}}_{\mathrm{t}-1}$	0.0354 **	0.0132 *	0.0106 ***
	(2.2627)	(1.9376)	(2.8037)
${\mathrm{N}\mathrm{A}\mathrm{S}\mathrm{D}\mathrm{A}\mathrm{Q}}_{\mathrm{t}-1}$	0.1046 ***	0.0429 ***	0.0231 **
	(3.2089)	(3.1558)	(2.6079)
Hi-Tech	−0.0649	−0.0312	−0.0130
	(−1.3445)	(−1.5462)	(−0.9788)
Market Capitalization	−0.0000	−0.0000	−0.0000
	(−0.3623)	(−0.3164)	(−0.3551)
${\mathrm{R}\&\mathrm{D}}_{\mathrm{t}-1}$	−0.0004 **	−0.0001 **	−0.0001 **
	(−2.4412)	(−2.4015)	(−2.1433)
Log(Volume)	−0.0064 **	−0.0028 **	−0.0019 **
	(−2.0922)	(−2.2611)	(−2.4227)
${\mathrm{L}\mathrm{e}\mathrm{v}\mathrm{e}\mathrm{r}\mathrm{a}\mathrm{g}\mathrm{e}}_{\mathrm{t}-1}$	−0.0298 ***	−0.0131 ***	−0.0063 ***
	(−4.3917)	(−4.3440)	(−3.8229)
${\mathrm{R}\mathrm{O}\mathrm{A}}_{\mathrm{t}-1}$	−0.0007	−0.0004	−0.0001
	(−1.1546)	(−1.3038)	(−0.8469)
Price	0.0087 ***	0.0042 ***	0.0025 ***
	(6.5165)	(6.3669)	(8.0130)
Log Dollar Trade Volume	−0.0502 ***	−0.0278 ***	−0.0103 ***
	(−4.1251)	(−5.1750)	(−3.4007)
Constant	1.8907 ***	1.0788 ***	0.4800 ***
	(4.8499)	(5.3613)	(5.1801)
Observations	173	173	173
R-squared	0.5186	0.5513	0.5854
Month FE	YES	YES	YES

, we interacted the exogenous shock with brown firms and found that even though the brown firms had lower liquidity, the association became less weak if the exogenous lagged environment risk increased. The results are consistent with the notion of higher investors’ attention to brown IPO stocks if environment concerns increase. This may lead to better liquidity for the brown stocks.

5. Discussion and Conclusions

This study analyzed the first-day IPO returns during the COVID-19 crisis period for green and brown firms. The previous study by Chan and Walter (2014) found indistinguishable underpricing for green and brown stocks. However, no other studies examined the first-day returns of green IPOs during the COVID-19 crisis when environmental concerns shift from time to time. Our analysis found that brown firms had more underpricing than green firms when we do not consider the exogenous change in the climate concern, which was consistent with the notion that firms with lower ESG scores had more information asymmetry and higher underpricing (El Ghoul et al., 2011). However, we showed that heightened climate concern drove investors toward green firms, which consequently led to lower initial returns for brown firms.

Next, we examined how the subsequent performance of green and brown firms is affected when there is an exogenous change in the climate concern. Using the PST (2022) model, we found that brown IPO firms’ following 3-month and 6-month returns were positive and statistically significant if environmental concern increases. We argued that investors perceive brown firms as risky firms and require more returns to be compensated for that risk (Bolton & Kacperczyk, 2021). At last, we examined the firm-level characteristics that explained the underpricing in a sub-sample analysis of green and brown firms. Offer price and Hi-Tech explained underpricing positively and significantly for green IPO firms. On the other hand, offer size and market capitalization explained underpricing positively, and first-day trading volume explained negatively for brown IPO firms. This study extended the huge IPO literature in a new perspective. While previous studies analyzed the underlying reasons for IPO underpricing, in this study, we analyzed IPO underpricing as a laboratory experiment of underperformance theory. Our results support the underperformance hypothesis, which proposes that companies with higher environmental standards perform poorly in the short run while brown firms outperform green firms (Bolton & Kacperczyk, 2021) when environmental concern increases because investors consider a brown firm as a risky investment. Hence, a brown premium exists in the short run. Later, we examined the stock liquidity and information asymmetry of green IPOs during the COVID-19 crisis period. Consistent with prior research, our findings showed that green IPOs exhibited higher stock liquidity and lower information asymmetry.

Finally, our study has policy implications as well. Regulatory authorities and activist shareholders would be well advised to keep in mind the consequences of waivers of key environmental laws and regulations. For example, during and after the pandemic, in connection with the crisis, several agencies have issued waivers or announced plans to stop enforcing key environmental laws and regulations. While federal, state, and municipal policymakers adopted a number of temporary policies that suspended, delayed, and relaxed various environmental safeguards in response to the COVID-19 pandemic, investors’ short-term and long-term responses to the waivers will enhance our understanding of the effectiveness of the new policies. Our study demonstrates the investors are more willing to invest in green stocks when the environmental concern increases even though the regulations are relaxed temporarily. Given the growing investor interest in ESG considerations, further research could explore the long-term effects of firms’ greenness on post-IPO performance and liquidity persistence beyond the COVID-19 period.