The Changes in the Economic Environment and Corporate Information Asymmetry—Focusing on the COVID-19 Pandemic

Abstract

The COVID-19 pandemic not only caused considerable disruptions in the capital markets, but also threatened firm sustainability. There are quite a few proxies that can measure a company’s sustainability, but information asymmetry is a representative one. Hence, we use information asymmetry to analyze whether the disruptions caused by changes in the corporate environment during the COVID-19 pandemic fostered significant differences in the sustainability of firms. The findings reveal the following: First, in the full sample, information asymmetry significantly increased during the pandemic. This suggests that the pandemic may have led to capital market disruptions, and thus, low sustainability of firms. Second, analyzing industry variations in information asymmetry during the pandemic reveals that information asymmetry does not significantly increase in the pharmaceutical and information technology industries. In contrast, it significantly increases in the samples belonging to the manufacturing and construction industries. This study provides the empirical evidence of the effects of COVID-19 on company’s sustainability. Based on this study, it should focus on finding way to reduce in-formation asymmetry in the capital market to increase the sustainability of firms in the long run.

1. Introduction

Despite being a relatively non-economic factor, the COVID-19 pandemic created tremendous uncertainty in the global economy [1,2,3]. According to [1], bond yields, oil prices, and stock prices plummeted after February 2020. Moreover, investments in safe assets (such as government bonds or gold) initially increased, but then declined, leading to record market instability surpassing even the 2008 global financial crisis. As the COVID-19 pandemic intensified, central banks stepped up their market interventions, not only injecting large amounts of public funds, but also dramatically lowering interest rates.

COVID-19 was first reported in December 2019 in Wuhan, China, and then spread across the globe, including to South Korea, the US, and Europe. On 11 March 2020, the World Health Organization declared COVID-19 a pandemic for the third time in history, following the 1968 Hong Kong flu and 2009 novel influenza. In the early days of the COVID-19 outbreak, it was predicted that only foreign firms with strong ties to China would experience supply chain problems or financial problems. However, the situation gradually worsened, with a huge impact not only in China but also globally. In South Korea, the number of COVID-19 cases began spiking after mid-February 2020. Many South Korean firms reported poor performance starting from the first half of 2020, showing increasing uncertainty about their future profits and even their viability. The end of the pandemic was declared in May 2023. However, the uncertainties continued thereafter.

COVID-19 significantly disrupted the life cycle and employment structure of firms worldwide. In particular, it significantly affected the survival and sustainability of firms. Bahaj et al. [4] used data from the UK to analyze the birth and survival of firms during the pandemic. The authors found that while firm entry increased in many countries during the pandemic, the firms that were created either disappeared or did not contribute to job creation. In particular, while the increase in firm entry contrasts with what happened in past economic downturns, new entrants were more likely to dissolve than firms established before the pandemic. This suggests that COVID-19 threatened the survival of firms. Next, KIEP [3] analyzed the closure rate of firms due to COVID-19. Specifically, analyzing the closure rate of start-ups that had been in existence for three years or less since their establishment revealed that the closure rate continuously increased after the currency and financial crises, then decreased in the early to mid-2010s, and increased again during the COVID-19 period. The authors also found that the closure rate significantly increased as the intensity of social distancing increased.

Clearly, COVID-19 is a direct factor that reduces firm sustainability. Hence, major countries used large-scale funding policies to increase the sustainability of firms. In the US, the USD 483 billion Paycheck Protection Program and Health Care Enhancement Act was enacted to provide guarantees, grants, and loans to small and medium-sized enterprises, as well as the Coronavirus Aid, Relief and Economy Security Act (CARES Act). In Germany, the emergency aid program (Soforthilfe) and bridging assistance (Überbrückungshilfe) were provided through the Economic Stabilization Fund (WSF) and Kreditanstalt für Wiederaufbau (KfW, “Credit Institute for Reconstruction”). The European Union (EU) provided a total of EUR 500 billion in grants through the Recovery Fund and later provided loans in a large scale through the Recovery Package. In the case of South Korea, public funds were injected into export and import companies, and companies operating in overseas markets. Furthermore, an Emergency Management Fund program was established to provide KRW 2 trillion in loans [3].

Overall, COVID-19 disrupted business operations, supply chains, and economies globally. Meanwhile, a significant amount of funds has been injected to overcome the COVID-19-induced economic crisis. Furthermore, firms tried their best to make decisions related to overall corporate operations, such as financing, logistics, human resources, and future strategy [5]. However, the changes in industrial structure and economic environment caused by COVID-19 have not easily recovered to previous levels. Therefore, understanding the impact of COVID-19 and developing measures to overcome the crisis caused by it is vital.

This study investigates the impact of the COVID-19 pandemic on information asymmetry in firms today, which is about one year after the end of the pandemic was declared. The pandemic, which resulted in more than 7 million deaths worldwide, had a much greater impact abroad than in South Korea. Nonetheless, in addition to the deaths, the outbreaks and quarantines affected the entire economy. Therefore, during the pandemic period, conflicts of interest between shareholders and agents surrounding firms, as well as disruptions in capital markets, would likely emerge. Therefore, we expect that the information asymmetry of firms increased due to the disruptions caused by changes in the environment of firms during COVID-19.

Information asymmetry is an economic term that refers to an unbalanced structure where each party to a transaction in a market has unequal access to information about the transaction counterpart. Information asymmetry related to firms occurs mainly in the shareholder–agent (manager) relationship. Here, the manager, as the shareholders’ agent, may not provide sufficient information to the shareholders to gain private benefits, thereby exacerbating the information asymmetry between the two parties. Consequently, it can lead to the adverse selection problem of shareholders who do not have an information advantage. This phenomenon is specifically addressed by agency theory [6].

Increased information asymmetry has diverse effects on capital markets. As suggested by agency theory, increased information asymmetry may exacerbate the agency problem, which in turn reduces the liquidity of stocks and increases the cost of capital [7,8]. Therefore, studies analyzed the factors that influence the mitigation of the agency problem. Bhattacharya et al. [9] reported that as the quality of accounting information decreases, adverse selection increases, and investors with information advantage are in a more favorable situation. In addition, Brown and Hillegeist [10] reported that the information asymmetry significantly decreases as the disclosure quality of accounting information increases.

Information asymmetry is highly correlated with uncertainty in capital markets. As uncertainty in capital markets increases, firms’ information asymmetry increases [11,12]. Moreover, the value relevance of accounting information decreases when an economic crisis occurs or capital market uncertainty increases [13,14]. In the case of South Korea too, capital market participants exhibited a lower reliance on accounting information during the COVID-19 period [15]. Therefore, the information asymmetry between shareholders and managers may have increased during the COVID-19 period. That is, COVID-19 acted as a crisis situation for firms, with a high likelihood of managers avoiding active disclosure (such as investor relation activities) to avoid the risk of not meeting market expectations or future business risks stemming from market uncertainty. Increasing information asymmetry can deepen a firm’s agent problem, which, when combined with market uncertainty conditions, can reduce the corporate sustainability. Therefore, the information asymmetry surrounding firms in capital markets may have increased during the COVID-19 period. This study uses the information asymmetry of firms to examine how the pandemic affected the sustainability of capital markets.

This study analyzed the impact of COVID-19 on the Korean capital market. Compared to other markets, South Korea’s capital market has unique characteristics in that it has chaebol groups and the proportion of foreign and individual investors among investors is quite high. As a result, the proportion of direct finance is high in the capital market and the volatility of the capital market is relatively high. Therefore, it is a suitable condition for analyzing the effects of information asymmetry caused by COVID-19.

The remainder of this article is organized as follows: We review the prior studies and derive the hypotheses in Section 2 and explain the study design in Section 3. We present the results in Section 4 and finally discuss the findings and conclusion in Section 5.

2. Literature Review and Hypothesis Development

The COVID-19 virus, which emerged in Wuhan, China in December 2019, spread horribly across the globe in just three to four months, starting in Asia in January 2020 [16]. This led to a downturn in the real economy and capital market crashes, with devastating consequences for the entire society and economy. Since then, scholars examined the short-term stock market response to COVID-19 in terms of stock returns and volatility in each country, finding that COVID-19 reduced stock returns and increased volatility [17]. Specifically, Zhang et al. [18] analyzed the changes in capital markets in different countries due to COVID-19. Europe, Asia, and North America were significantly affected by the pandemic, while Africa was less affected.

According to data published by the Coronavirus Resource Center at John Hopkins University, the UK had the highest number of patients and deaths among European countries at the end of 2021 [19]. Hence, Paterson et al. [20] analyzed the impact of COVID-19 on the financial markets in the UK. They reported that the FTSE 100 index and pound sterling exchange rate experienced unprecedented losses. Additionally, the simultaneous announcement of the Brexit withdrawal further worsened the outlook of the UK financial market. Meanwhile, Singh et al. [21] conducted an event study to investigate the impact of the COVID-19 outbreak on the stock markets of G-20 countries. The authors observed significant negative abnormal returns after the COVID-19 outbreak. In particular, the cumulative average abnormal returns ranged from −0.70% to −42.69% from the date of to 43 days after the outbreak. This suggested stock market panic after the outbreak.

In addition to disrupting capital markets, COVID-19 caused significant changes in the business environment of firms, and thus, in their financial state and business performance. Alhajjeah and Besim [22] analyzed whether the capital structure of UK firms changed during the COVID-19 pandemic, finding that firms reduced debt financing and favored internal financing during the pandemic period. By reducing debt, they reduced financial risk, suggesting that risk management is an important strategic choice in times of crisis.

Tax avoidance is a typical method of internal financing. Saptono et al. [23] analyzed whether tax avoidance changed during the COVID-19 period. They found that book-tax differences (BTD), a measure of tax avoidance, increased during the COVID-19 period. However, this was not due to an increase in temporary BTD, but rather an increase in permanent BTD. Furthermore, this increase in tax avoidance was more pronounced in industries more affected by COVID-19 than in those less affected.

KIEP [3] studied whether the business performance of South Korean firms changed during the COVID-19 crisis compared to the 2008 global financial crisis. The results for all firms showed a considerable decline in sales, although not as much as that in the 2008 global financial crisis. However, some industries (agriculture, forestry, fishing, education services, and leisure-related services) showed a greater decline in sales than that in the global financial crisis. Furthermore, dividing all firms into face-to-face and non-face-to-face industries revealed that the decline in sales was significantly greater for the former than for the latter.

Deterioration in financial condition and business performance may ultimately adversely affect the firm’s sustainability and increase its financial risk. Hence, Zhang et al. [24] analyzed the relationship between sustainability-related news disclosure and stock price crash risk using news data related to corporate sustainability for S&P 500 companies during the COVID-19 period. Their analysis showed that as the disclosure of sustainability-related keywords increases, the risk of a stock price crash in the following year increases. The authors argued that this was a result of investors’ lack of confidence in corporate news released during the pandemic period, which may have increased the risk of stock price crashes. That is, under normal circumstances, greater release of sustainability-related news may decrease information asymmetry, thereby reducing the risk of stock price crashes. However, under conditions of intensified agency problems or financial stress, sustainability-related news may be released as a tool to delay the disclosure of unfavorable conditions and manipulate managers’ legitimacy (i.e., information asymmetry may increase).

The crisis caused by COVID-19 also had various effects on firms’ accounting information. Specifically, in many industries, the collectability of accounts receivable decreased, the fair value of investment stocks and bonds declined, and inventory assets and tangible assets were impaired. These changes also led to increased audit risk for external audits [25]. Accounting supervisory authorities also raised concerns about corporate accounting practices in the unprecedented COVID-19 situation. The Financial Supervisory Service called for caution in accounting for impairment of financial assets, such as trade and loans receivables, and requested estimating reasonable use value when determining impairment of tangible and intangible assets [26].

The COVID-19 crisis differs from previous East Asian or global financial crises in that the impact is due to a pandemic rather than a decline in individual firm profitability or change in the macroeconomic environment. Nevertheless, the COVID-19 crisis increased economic policy uncertainty in the capital markets as an event of ongoing concern. As uncertainty in capital markets increases, firms’ information asymmetry increases [11,12]. Nagar et al. [11] analyzed the effect of uncertainty about government economic policies on information asymmetry. The authors reported that economic policy uncertainty not only increases information asymmetry, but also reduces stock price responses to earnings surprises. Moreover, Lei and Luo [12] conducted a literature review on the impact of political or policy uncertainty on financial analysts’ forecasts, firms’ disclosures, and information asymmetry, and the authors found that several studies reported that increasing uncertainty affects financial analysts’ forecasts and firms’ disclosures, which interact in terms of information asymmetry.

Information asymmetry refers to an unbalanced structure that occurs when each party to a transaction in a market possesses information about the transaction counterpart unequally. Information asymmetry has an important meaning not only to academics, but also to practitioners, regulatory agencies, corporate officials, and individual investors. The concept of information asymmetry is also widely used in economics and finance. Accounting information, including earnings information provided by firms, provides capital market participants with useful information for decision making. Thus, this fulfills the function of efficiently allocating scarce resources and reducing information asymmetry between managers and external investors. Such increased information asymmetry in capital markets reduces the liquidity of stocks and increases the cost of capital [7,8]. Therefore, reducing information asymmetry is important for the stability of capital markets.

As uncertainty in capital markets significantly increased during the COVID-19 period, the following studies analyzed the changes in information asymmetry due to COVID-19. Gofran et al. [27] analyzed the impact of COVID-19 on the stock markets in the US, UK, Brazil, China, Germany, and Spain. The results show that stock liquidity decreased in most markets, with a significant increase in bid–ask spreads. Gubareva [28] used bond market data from December 2019 to July 2020 to analyze whether the bid–ask spread changed during the pandemic period. The author reported that the bid–ask spread peaked in March 2020 due to changes in default risk caused by COVID-19. Furthermore, although liquidity improved since then, it did not return to pre-COVID levels. Yu et al. [29] analyzed the relationship between information asymmetry and cash holding levels using data on Chinese firms from 2014 to 2020. Their results show that firms’ cash holdings significantly increased due to COVID-19. Moreover, the impact of information asymmetry on cash holdings was more pronounced during the COVID-19 period.

Information asymmetry can also be measured by stock price volatility. Baek et al. [30] analyzed whether stock price volatility changed due to COVID-19 in the US market. The authors found that stock price volatility is sensitive to COVID-19 news. Specifically, the total stock price risk and firm-specific risk significantly increase due to COVID-19 news; however, systematic risk varies across industries. In particular, both negative and positive information significantly affect stock price volatility, but negative news has a stronger impact. Focusing on 15 countries, Kusumahadi and Permana [31] found significant stock price volatility due to COVID-19 in most countries. Khan et al. [32] also analyzed the impact of the COVID-19 pandemic on the stock markets of China, India, Pakistan, the UK, and the US using data from 2016 to 2021. Thus, the authors found that volatility significantly increased in the post-COVID period.

In South Korea, research on information asymmetry mainly focused on the effects of adopting K-IFRS. Many researchers predicted that accounting transparency and credibility of firms adopting K-IFRS would increase, which in turn would decrease the level of information asymmetry in the capital market. Shin and Choi [33] studied the changes in information asymmetry and stock price synchronization after K-IFRS adoption. Using a sample of South Korean listed firms from 2006 to 2012, they reported that bid–ask spreads and stock price synchronization decreased after the adoption of K-IFRS, thereby increasing the quality of information. Similarly, Kim and Cho [34] used the volatility of daily stock returns as a measure of information asymmetry and found that information asymmetry decreased after the adoption of K-IFRS.

The South Korean stock market has a high proportion of foreign and individual investors. Therefore, the impact of COVID-19 may be greater than in other countries. However, few studies analyzed the effect of COVID-19 on information asymmetry in the South Korean stock market. Based on the previous analysis, we estimate that the information asymmetry of South Korean firms increased during the COVID-19 period. Therefore, we formulate the following hypotheses:

Hypothesis 1. 

Information asymmetry increases during the COVID-19 period.

The impact of COVID-19 may not be the same for all firms. During the COVID-19 period, firms tried their best to make decisions related to overall business operations [5]. However, the characteristics of individual firms may influence these decisions.

Bai et al. [35] analyzed the impact of a firm’s IT capability on its agility through a survey. The analysis showed that IT capability increases the firm’s agility, which directly affects the firm’s business performance. Liu and Qi [36] also reported that the firm’s digital transformation contributes to its financial sustainability. Specifically, they reported that digital transformation improves the firm’s internal governance, increases analyst coverage, and eases financial constraints; that is, it acts as a factor that reduces corporate risks. When the firm has sufficient IT capability or digital transformation, it can be operated more efficiently, such as through telecommuting.

The impact of COVID-19 differs not only by firm characteristics, but also by industry characteristics. KIEP [3] categorized firms into face-to-face and non-face-to-face industries during the COVID-19 crisis period and analyzed whether the profitability of firms differed. The results show that the sales of face-to-face businesses were significantly lower than that of non-face-to-face businesses. Deloitte [1] analyzed the impact of COVID-19 in South Korea, China, the US, Europe, and Southeast Asia based on the year 2020. Their analysis showed that the whole world was significantly impacted, with little regional variation, and also showed that the intensity of the impact varied by industry. By industry, the telecommunications, media, and entertainment industries would experience limited impacts, while the demand for online businesses was rather expected to increase. Meanwhile, the automotive and auto parts industries were expected to experience significant production disruptions due to the collapse of supply chains and shutdown of production lines caused by the spread of COVID-19. In the case of consumer staples and consumer goods, the impact would be smaller compared to other industries. However, the demand was expected to be maintained online or through delivery. The oil, gas, and petrochemical industries were expected to be severely impacted by COVID-19 and lower oil prices, and the outlook for the industry was not promising. The financial industry was predicted to have a negative outlook due to a contraction in the real economy and decline in asset values.

KPMG [2] also published a report analyzing the impact of COVID-19 on each industry. It noted that the financial industry was expected to experience slowing growth and deterioration in its soundness, and the future outlook was not clear. Next, the automobile industry was also experiencing demand contraction, supply disruption, and employee infection risk. Moreover, the airline industry was directly affected by entry restrictions and a decline in travelers, with a worsening of this crisis in the future. In the retail industry, there were differences between online and offline retail, with online retail expected to benefit from COVID-19.

As such, the impact of COVID-19 on firms may vary depending on firm and industry characteristics. Accordingly, this study proposes the following hypothesis:

Hypothesis 2. 

Changes in information asymmetry during the COVID-19 period differ across industries.

3. Materials and Methods

3.1. Research Model and Definition of Variables

This study aims to analyze whether information asymmetry increases during the COVID-19 period and whether changes in information asymmetry during the COVID-19 period differ by industry. Accordingly, we set up the following regression model, which includes control variables with information asymmetry as the dependent variable and the COVID-19 period as the variable of interest. In Model 1, i refers to the firm and t refers to the year.

$$\begin{array}{cc}\mathrm{ASY}={\beta }_0& +{\beta }_1\mathrm{C}\mathrm{O}\mathrm{V}\mathrm{I}\mathrm{D}19+{\beta }_2\mathrm{M}\mathrm{T}\mathrm{T}+{\beta }_3\mathrm{L}\mathrm{n}\mathrm{M}\mathrm{C}\mathrm{V}+{\beta }_4\mathrm{F}\mathrm{O}\mathrm{R}\hfill \\ & +{\beta }_5\mathrm{S}\mathrm{i}\mathrm{z}\mathrm{e}+{\beta }_6\mathrm{L}\mathrm{e}\mathrm{v}\mathrm{e}\mathrm{r}\mathrm{a}\mathrm{g}\mathrm{e}+{\beta }_7\mathrm{R}\mathrm{O}\mathrm{A}+{\beta }_8\mathrm{L}\mathrm{o}\mathrm{s}\mathrm{s}\mathrm{d}\mathrm{u}\mathrm{m}\mathrm{m}\mathrm{y}\\ & +{\beta }_9\mathrm{G}\mathrm{R}\mathrm{W}+{\beta }_{10}\mathrm{B}\mathrm{I}\mathrm{G}4+\mathrm{M}\mathrm{K}\mathrm{T}+\mathrm{Y}\mathrm{e}\mathrm{a}\mathrm{r}+\mathrm{I}\mathrm{n}\mathrm{d}\mathrm{u}\mathrm{s}\mathrm{t}\mathrm{r}\mathrm{y}+\epsilon \end{array}$$

(1)

ASY	information asymmetry, standard deviation of daily returns for year
COVID-19	dummy variable is 1 if it falls under the COVID-19 period, or 0 otherwise
MTT	market transaction turnover, (value of traded value/market value)
LnMCV	natural logarithm of annual average of daily market closing value
FOR	share ratio of foreign investors
Size	natural logarithm of total assets
Leverage	total liabilities divided by total assets
ROA	net income divided by total assets
Lossdummy	dummy variable that is 1 if loss occurred in the previous year, or 0 otherwise
GRW	Sales growth, (current year sales–previous year sales)/previous year sales
BIG4	Big4 dummy, 1 if it was audited by a large auditor, or 0 otherwise
MTK	market dummy, equals 1 if the firm is in the KOSPI market and 0 otherwise
Year	year dummies
Industry	industry dummies

Information asymmetry (ASY) is measured as the standard deviation of daily stock returns. Leuz and Verrecchia [8] and Choi [37] reported that the variation in stock returns increases as the information asymmetry among investors increases; therefore, information asymmetry increases as the standard deviation of stock returns increases. Hence, we also use the standard deviation of daily returns by firm-year as a proxy for information asymmetry. The variable of interest in Model 1, COVID-19, is a dummy which equals 1 if it corresponds to the COVID-19 period and 0 if otherwise. That is, it has the value of 1 from 2020 to 2022 and 0 before 2020.

The control variables included are based on prior research. First, among the factors that affect the information asymmetry of individual firms, stock liquidity and visibility are typical factors. As a stock price increases, information asymmetry increases. Further, as trading volume increases, information asymmetry decreases. Therefore, we include the annual mean trading turnover (MTT) and annual mean closing price (lnMCV) as control variables. Furthermore, as the foreign ownership ratio increases, information advantage increases. This induces conservative accounting policies and reduces information asymmetry. Therefore, we include the foreign ownership ratio (FOR) of common stock at the end of the fiscal term as a control variable [38]. Information asymmetry is also determined by the characteristics of individual firms. Therefore, control variables related to firm characteristics are also included, such as firm size (SIZE), leverage ratio (LEV), return on assets (ROA), and sales growth rate (GRW) [37]. Firm size (SIZE) helps control for firm size. Leverage ratio (LEV) is a proxy variable for financial stability and is calculated by dividing total debt by total assets. Return on assets (ROA) refers to the firm’s profitability and is calculated by dividing net income by total assets. Sales growth rate (GRW) refers to the firm’s growth and is calculated by subtracting the previous year’s sales from the current year’s sales and then dividing it by the previous year’s sales. Large auditors provide better audit services. Hence, information asymmetry may be reduced by being audited by a large auditor [37]. Therefore, we include the Big Four auditor (BIG) variable as a control variable in this study. BIG has a value of 1 if the sample firm is audited by a large global accounting firm (Deloitte, KPMG, PwC, and E&Y) and 0 otherwise. Finally, we control for market, year, and industry effects in Model 1. The market dummy (MKT) variable equals 1 if the firm is in the KOSPI market and 0 if it is in the KOSDAQ market. Finally, we include dummy variables by year (Year) and industry (Industry).

3.2. Samples

This study uses a sample of firms listed on the KOSPI and KOSDAQ from 2016 to 2022. We extracted the data from the databases provided by KISVALUE and DataGuide. In the initial sample, we excluded firm-years that did not have a fiscal year end in December, and those which belonged to the financial industry. We also excluded firm-years that did not have the necessary data for our analysis. The final sample has 13,155 firm-years.

Table 1. Distribution of the sample.

Panel A. Distribution by Year or Market
Year	KOSPI	KOSDAQ		Total
2016	644	968		1612
2017	656	1044		1700
2018	678	1121		1799
2019	695	1184		1879
2020	703	1257		1960
2021	715	1336		2051
2022	724	1430		2154
Total	4815	8340		13,155
Panel B. Distribution by Industry
Industry			n		%
Agriculture, forestry, and fishing			35		0.27%
Mining			14		0.11%
Manufacturing			8688		66.04%
Electricity, gas, steam, and air conditioning supply			73		0.55%
Water supply; sewage, waste management			42		0.32%
Construction			360		2.74%
Wholesale and retail trade			986		7.50%
Transportation and storage			199		1.51%
Accommodation and food service activities			28		0.21%
Information and communication			1557		11.84%
Real estate activities			41		0.31%
Professional, scientific, and technical activities			903		6.86%
Business facilities management			103		0.78%
Education			61		0.46%
Arts, sports, and recreation related services			65		0.49%
Total			13,155		100.00%

shows the distribution of the sample.

Panel A in Table 1 shows the distribution by year and market. In total, 4815 and 8340 firm-years in the sample belong to the KOSPI and KOSDAQ markets. The stock market in South Korea is operated by KOSPI, KOSDAQ, and KONEX. Compared to the country’s main index KOSPI-listed companies, KOSDAQ-listed companies have a relatively low market capitalization, fewer issued stock, and higher corporate business risks. Although the sample size is larger for the KOSDAQ, the sample is evenly distributed across years. The distribution by industry in Panel B shows that the proportion of the manufacturing industry is high (66.11%). The manufacturing industry constitutes 66.04%, but if it is classified in more detail, it is distributed relatively evenly.

4. Results

4.1. Descriptive Statistics

Table 2. Descriptive statistics.

Variable	Mean	Std	Min	Median	Max
ASY	3.011	1.122	1.006	2.897	6.067
COVID-19	0.468	0.499	0.000	0.000	1.000
MTT	0.023	0.042	0.000	0.011	0.171
LnMCV	9.106	1.315	6.512	8.978	12.612
FOR	0.066	0.103	0.000	0.025	0.522
Size	26.020	1.359	23.754	25.777	30.567
Leverage	0.359	0.212	0.027	0.351	0.854
ROA	0.007	0.231	−0.514	0.023	0.365
Lossdummy	0.301	0.458	0.000	0.000	1.000
GRW	0.316	13.529	−0.641	0.047	2.888
BIG4	0.396	0.489	0.000	0.000	1.000
MKT	0.366	0.481	0.000	0.000	1.000
Obs.	3.011	1.122	1.006	2.897	6.067

Detailed definition of variables is follows. ASY = information asymmetry, standard deviation of daily returns for year; COVID-19 = dummy variable that is 1 if falls under the COVID-19 period, or 0 otherwise; MTT = market transaction turnover (value of traded value/market value); LnMCV = natural logarithm of annual average of daily market closing value; FOR = share ratio of foreign investors; Size = natural logarithm of total assets; Leverage = total liabilities divided by total assets; ROA = net income divided by total assets; Lossdummy = dummy variable that is 1 if loss occurred in the previous year, or 0 otherwise; GRW = Sales growth, (current year sales–previous year sales)/previous year sales; BIG4 = Big4 dummy, 1 if was audited by a large auditor, or 0 otherwise; and MKT = 1 if listed in KOSPI, else 0.

presents the descriptive statistics of the main variables. First, the mean information asymmetry (ASY) is 3.011, which is larger than the median of 2.897. Thus, the distribution is more spread out on the right-hand side. The mean of the COVID-19 dummy is 0.468, indicating that 46.8% of the sample falls within the COVID-19 period. The MTT variable has a mean of 0.023 and median of 0.011, while the lnMCV variable has a mean of 9.106 and median of 8.978. Again, a higher mean than the median shows that the distribution is skewed to the right. The mean of FOR is 0.066.

Next, among firm characteristics, firm size (SIZE) has a mean of 26.020 and median of 25.777. The leverage ratio (LEV) has a mean of 0.359 and median of 0.351. ROA, which indicates the firm’s profitability, has a mean of 0.007 and median of 0.023. Meanwhile, Lossdummy (indicating whether there was a loss in the previous year) has a mean of 0.301 and median of 0, indicating that approximately 30% of the sample had a loss in the previous year. GRW, which refers to the firm’s growth, has a mean of 0.316 and median of 0.047. The BIG variable has a mean of 0.396, indicating that 39.6% of the sample is audited by a big auditor. Meanwhile, MKT has a mean of 0.366, indicating that 36.6% of the sample belongs to the KOSPI market.

Table 3. Changes in information asymmetry by year.

Year	2016	2017	2018	2019	2020	2021	2022
ASY	2.814	2.413	3.086	2.783	3.792	3.032	3.042

and Figure 1 show the mean value of information asymmetry (i.e., ASY variable) by year. Before conducting the regression analysis, examining the mean of ASY for each year from 2016 to 2022 revealed that the information asymmetry is highest in 2020, and decreases in 2021 and 2022. Nevertheless, the annual mean information asymmetry is higher in 2021 and 2022 than in the pre-COVID years.

Table 4. Correlation matrix.

Variable	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)	(12)
(1) ASY	1.00 +
(2) COVID-19	0.22 +	1.00 +
(3) MTT	0.59 +	0.12 +	1.00 +
(4) LnMCV	−0.18 +	0.01	−0.14 +	1.00 +
(5) FOR	−0.22 +	−0.05 +	−0.15 +	0.40 +	1.00 +
(6) Size	−0.34 +	0.04 +	−0.23 +	0.52 +	0.46 +	1.00 +
(7) Leverage	0.08 +	0.00	0.01	−0.15 +	−0.06 +	0.15 +	1.00 +
(8) ROA	−0.15 +	0.00	−0.08 +	0.14 +	0.10 +	0.14 +	−0.13 +	1.00 +
(9) Lossdummy	0.26 +	0.06 +	0.08 +	−0.27 +	−0.16 +	−0.21 +	0.19 +	−0.24 +	1.00 +
(10) GRW	0.03 +	0.01	0.01	0.01	0.01	−0.01	−0.01	0.05 +	0.01	1.00 +
(11) BIG4	−0.20 +	−0.12 +	−0.12 +	0.29 +	0.27 +	0.41 +	0.04 +	0.04 +	−0.11 +	−0.01	1.00 +
(12) MKT	−0.26 +	−0.03 +	−0.16 +	0.26 +	0.23 +	0.57 +	0.11 +	0.05 +	−0.10 +	−0.01	0.30 +	1.00 +

⁺ denotes the significance at 1% level. A detailed definition of variables is in Table 2.

presents the results of the Pearson correlation analysis. First, the correlation between the dependent and main variables shows that COVID-19 has a significant positive (+) correlation with ASY. Next, examining the correlations between the main variable, COVID-19, and control variables revealed that MTT, SIZE, LossDum, and MKT have significant positive (+) correlations, while FOR and BIG have negative (−) correlations.

However, Pearson correlation simply indicates the relationship between two variables, it is necessary to conduct a multiple regression analysis considering the control variables to test the hypotheses. First, for testing multicollinearity, we measured the variance inflation factor (VIF) values. No VIF value was 5 or greater, indicating a low possibility of the occurrence of a multicollinearity problem.

4.2. Regression Results

Next, we present the multiple regression analysis results.

Table 5. Results of analysis.

	Dependent Variable = ASY
	(1) Full Samples	Market
		(2) KOSDAQ	(3) KOSPI
intercept	5.352 *** (28.37)	6.615 *** (21.57)	3.255 *** (13.79)
COVID-19	0.176 *** (6.86)	0.153 *** (4.58)	0.184 *** (5.14)
MTT	12.864 *** (78.52)	11.355 *** (62.92)	24.136 *** (55.60)
LnMCV	0.080 *** (11.69)	0.133 *** (13.48)	0.055 *** (6.23)
FOR	−0.315 *** (−3.98)	−0.490 *** (−4.25)	−0.265 *** (−2.70)
Size	−0.140 *** (−17.26)	−0.204 *** (−16.13)	−0.071 *** (−6.92)
Leverage	0.503 *** (14.14)	0.365 *** (8.21)	0.659 *** (11.68)
ROA	−0.186 *** (−6.11)	−0.210 *** (−6.32)	−0.084 (−0.92)
Lossdummy	0.347 *** (21.38)	0.363 *** (17.65)	0.254 *** (10.17)
GRW	0.001 *** (2.63)	0.001 ** (2.26)	0.002 (1.44)
BIG4	−0.048 *** (−3.03)	−0.066 *** (−3.23)	0.008 (0.34)
MKT	−0.168 *** (−9.31)
Year	included	included	included
Industry	included	included	included
VIF Range	1.00–3.65	1.00–3.71	1.00–3.60
F value	241.99 ***	142.44 ***	119.03 ***
Adj. R2	0.5318	0.4959	0.5871
N	13,155	8340	4815

**, and *** denote the significance at 5%, and 1% level, respectively. A detailed definition of variables is in Table 2.

shows the results for Hypothesis 1. Column 2 shows the analysis results for the full sample, while columns 3 and 4 show the results by classifying the markets. In column 2, the coefficient value of the COVID-19 variable is 0.176 (t = 6.86) and is significant at the 1% level. This means that information asymmetry significantly increased during the COVID-19 period.

In columns 3 and 4, the coefficient values are 0.153 (t = 4.58) and 0.184 (t = 5.14), respectively, which are also significant at the 1% level. Thus, the relationship is consistent across the two South Korean stock markets (KOSPI and KOSDAQ). Next, in the full sample results in column 2, the market dummy (MKT) has a significant negative (−) coefficient value. Thus, information asymmetry is less pronounced in the KOSPI market with larger firm sizes. However, since information asymmetry occurred in both markets during the COVID-19 period, firms indeed faced abnormal conditions during this period.

All models in Table 5 show that the F-value is significant at the 1% level. Therefore, information asymmetry, as measured by the standard deviation of daily stock returns, increased since 2020, when COVID-19 affected the South Korean stock markets. It can be interpreted that information asymmetry increased due to the opaque information environment created during COVID-19 in the South Korean stock markets.

Table 6. Results of analysis by industry.

	Dependent Variable = ASY
	(1) Pharmaceutical Industry	(2) IT Industry	(3) Manufacturing Industry	(4) Construction Industry
intercept	2.748 *** (3.11)	3.927 *** (4.02)	5.334 *** (21.01)	6.226 *** (6.49)
COVID-19	0.066 (0.67)	0.026 (0.22)	0.158 *** (4.70)	0.321 ** (2.45)
MTT	16.160 *** (23.12)	15.755 *** (17.20)	12.185 *** (61.41)	20.481 *** (14.65)
LnMCV	0.225 *** (7.33)	0.064 * (1.95)	0.077 *** (8.40)	0.106 *** (2.75)
FOR	−1.335 *** (−3.76)	−0.672 (−1.58)	−0.343 *** (−3.34)	0.359 (1.09)
Size	−0.088 ** (−2.24)	−0.080 * (−1.88)	−0.144 *** (−13.12)	−0.202 *** (−5.04)
Leverage	0.192 (1.31)	0.405 * (1.85)	0.690 *** (13.27)	0.473 ** (2.23)
ROA	−0.101 ** (−2.20)	−0.630 *** (−2.62)	−0.354 *** (−5.02)	−1.509 *** (−3.76)
Lossdummy	0.482 *** (8.33)	0.260 *** (3.16)	0.276 *** (12.45)	0.401 *** (4.15)
GRW	0.001 (1.39)	0.102 (1.26)	0.040 *** (3.16)	0.008 (0.14)
BIG4	−0.020 (−0.35)	−0.280 *** (−4.23)	−0.004 (−0.20)	−0.176 ** (−2.02)
MKT	−0.268 *** (−3.95)	0.009 (0.10)	−0.153 *** (−6.53)	0.212 ** (2.50)
Year	Included	Included	Included	Included
Industry	Included	Included	Included	Included
VIF Range	1.01–3.99	1.07–3.83	1.03–3.64	1.16–3.84
F value	90.29 ***	40.50 ***	255.89 ***	36.85 ***
Adj. R2	0.6014	0.6429	0.5131	0.6293
N	948	374	7740	360

*, **, and *** denote the significance at 10%, 5%, and 1% level, respectively. A detailed definition of variables is in Table 2.

presents the results by industry. During the COVID-19 period, firms actively used IT-based online sales. If they had the requisite IT capabilities or digital transformation in the process of sales activities, they used telecommuting to perform their operations more efficiently. In addition, social distancing was enforced due to COVID-19, which led to a difference in business performance between face-to-face and non-face-to-face businesses. Specifically, the sales of the former were significantly lower than those of the latter. Therefore, we categorized the pharmaceutical and IT industries, which are the beneficiaries of COVID-19, and the manufacturing (excluding pharmaceuticals) and construction industries, which are the victims, and tested whether information asymmetry differs between them.

The results show that the coefficient of the COVID-19 variable is insignificant for the pharmaceutical (column 2: 0.066; t = 0.67) and IT industries (column 3: 0.026; t = 0.22). Thus, information asymmetry did not significantly change during the COVID-19 period in the pharmaceutical and IT industries. However, in column 4, which shows the analysis results for the manufacturing industry sample excluding pharmaceuticals, the coefficient of the COVID-19 variable is 0.158 (t = 4.70) and is significant at the 1% level. The coefficient in column 5 for the construction industry sample is 0.321 (t = 2.45) at the 5% significance level. Thus, information asymmetry significantly increased during the COVID-19 period in the manufacturing and construction industries, except in the pharmaceutical industry.

Overall, based on the above empirical analysis results, it can be interpreted as follows: In the beneficiary industries (pharmaceutical and biosciences) and IT industries, where telecommuting was active during the COVID-19 period, information asymmetry was relatively less affected by the pandemic. Meanwhile, information asymmetry significantly increased in face-to-face-oriented industries such as manufacturing (excluding pharmaceuticals) and construction.

4.3. Robustness Test Results

Two additional analyses were performed to enhance the robustness of the results of this study. First, the results were reanalyzed using an alternative dependent variable, and then it was examined whether the results were different. Stock market transaction turnover (MTT) is included as one of the explanatory variables instead of an alternative measure of information asymmetry. Therefore, the analysis result is presented by setting the dependent variable as MTT instead of ASY.

Table 7. Results of additional analysis using the MTT variable.

	Dependent Variable = MTT
	(1) Full Samples	Market
		(2) KOSDAQ	(3) KOSPI
COVID-19	0.003 ** (2.34)	0.003 * (1.87)	0.003 *** (3.01)
Intercept	included	included	included
Control variables	included	included	included
Year	included	included	included
Industry	included	included	included
VIF range	1.00–3.65	1.00–3.74	1.00–3.59
F value	22.17 ***	11.46 ***	19.14 ***
Adj. R2	0.0894	0.0667	0.1768
N	13,155	8340	4815

*, **, and *** denote the significance at 10%, 5%, and 1% level, respectively. A detailed definition of variables is in Table 2.

and

Table 8. Results of additional analysis using MTT variable by industry.

	Dependent Variable = MTT
	(1) Pharmaceutical Industry	(2) IT Industry	(3) Manufacturing Industry	(4) Construction Industry
COVID-19	0.001 (0.12)	0.006 (0.88)	0.003 ** (2.04)	0.012 ** (2.46)
Intercept	included	included	included	included
Control Variables	included	included	included	included
Year	included	included	included	included
Industry	included	included	included	included
VIF Range	1.01–3.99	1.07–3.83	1.02–3.64	1.14–3.77
F value	15.49 ***	6.24 ***	19.85 ***	7.37 ***
Adj. R2	0.1867	0.1836	0.0702	0.2212
N	948	374	7740	360

**, and *** denote the significance at 5%, and 1% level, respectively. A detailed definition of variables is in Table 2.

present the results of analyzing Hypotheses 1 and 2 with the dependent variable as MTT.

Column 2 in Table 7 shows the analysis results for the full sample, while columns 3 and 4 show the results by classifying the markets. In column 2, the coefficient value of the COVID-19 variable is 0.003 (t = 2.34) and is significant at the 5% level. This means that information asymmetry significantly increased during the COVID-19 period. Additionally, in columns 3 and 4, the coefficient values are 0.003 (t = 1.87) and 0.003 (t = 3.01), respectively, which are also significant. This means that information asymmetry increased significantly during COVID, even if the dependent variable is used as MTT.

The results of Table 8, using MTT as a dependent variable, show that the coefficient of the COVID-19 variable is insignificant for the pharmaceutical (column 2: 0.001; t = 0.12) and IT industries (column 3: 0.006; t = 0.88). However, in column 4, which shows the analysis results for the manufacturing industry sample excluding pharmaceuticals, the coefficient of the COVID-19 variable is 0.003 (t = 2.04) and is significant at the 5% level. The coefficient in column 5 for the construction industry sample is 0.012 (t = 2.46) at the 5% significance level. This means that there were significant differences in information asymmetry by industry during the COVID period, even if the dependent variable was used as the MTT.

Next, since the main result of this study may be due to endogeneity, an additional analysis was performed by applying a methodology to control endogeneity. In this study, two-stage least squares (2SLS), using instrumental variables, was performed to control endogeneity. 2SLS using instrumental variables is a two-step model that estimates the effect on dependent variables after controlling the effect of endogenous variables using instrumental variables that are related to major endogenous variables and are not related to error terms. According to many previous studies, the lag variable is mainly used as an instrument variable to control the endogeneity of the dependent variable [39]. Therefore, in this study, the model was analyzed using the lag value of the dependent variable as an instrument variable.

Table 9. Results of additional analysis using 2SLS methodology.

	Dependent Variable = ASY
	(1) Full Samples	Market
		(2) KOSDAQ	(3) KOSPI
COVID-19	3.039 *** (15.23)	2.747 *** (12.05)	3.094 *** (9.30)
Intercept	included	included	included
Control Variables	included	included	included
Year	included	included	included
Industry	included	included	included
Sargan Test	1765.08 ***	873.25 ***	585.34 ***
F value	47.11 ***	32.95 ***	17.40 ***
Adj. R2	0.2046	0.2093	0.1879
N	11,111	7000	4111

*** denote the significance at 1% level. A detailed definition of variables is in Table 2.

and

Table 10. Results of additional analysis using 2SLS methodology by industry.

	Dependent Variable = ASY
	(1) Pharmaceutical Industry	(2) IT Industry	(3) Manufacturing Industry	(4) Construction Industry
COVID-19	−0.084 (−1.00)	−0.267 ** (−2.38)	2.952 *** (11.32)	0.228 * (1.81)
Intercept	included	included	included	included
Control Variables	included	included	included	included
Year	included	included	included	included
Industry	included	included	included	included
Sargan Test	103.54 ***	46.76 ***	1100.36 ***	54.17 ***
F value	99.47 ***	34.19 ***	51.25 ***	34.45 ***
Adj. R2	0.6495	0.6276	0.1972	0.6354
N	798	316	6546	308

*, **, and *** denote the significance at 10%, 5%, and 1% level, respectively. A detailed definition of variables is in Table 2.

present the results of the analysis by the 2SLS model. Column 2 in Table 9 shows the analysis results for the full sample, while columns 3 and 4 show the results by classifying the markets. In column 2, the coefficient value of the COVID-19 variable is 3.039 (t = 15.23) and is significant at the 1% level. This means that information asymmetry significantly increased during the COVID-19 period. Additionally, in columns 3 and 4, the coefficient values are 2.747 (t = 12.05) and 3.094 (t = 9.30), respectively, which are also significant. This means that information asymmetry increased significantly during COVID, even if 2SLS methodology was applied.

The results of Table 10, using 2SLS methodology, show that the coefficient of the COVID-19 variable is insignificant for the pharmaceutical industry (column 2: −0.084; t = −1.00), but in the case of the IT industry, the coefficient (column 3: −0.267; t = −2.38) value had a significant negative value. However, in column 4, which shows the analysis results for the manufacturing industry sample excluding pharmaceuticals, the coefficient of the COVID-19 variable is 2.952 (t = 11.32) and is significant at the 1% level. The coefficient in column 5 for the construction industry sample is 0.228 (t = 1.81) at the 10% significance level. This means that there were significant differences in information asymmetry by industry during the COVID period, even if the 2SLS methodology was applied.

The results of additional analysis are summarized as follows: Even if the dependent variable is used as MTT instead of ASY, information asymmetry increased significantly during the COVID-19 period, but information asymmetry differed by industry. In addition, the same result was found in the result of applying the 2SLS methodology. This can be interpreted as a result of securing robustness in the analysis results of this study.

5. Conclusions

We analyzed the impact of COVID-19 on firm sustainability using information asymmetry. Specifically, we analyzed whether information asymmetry increased during the pandemic and verified whether changes in information asymmetry during this period varied by industry. We used firm-year data from 2016 to 2022 for firms listed on the KOSPI and KOSDAQ markets, and measured information asymmetry as the standard deviation of daily returns. The findings are summarized below.

First, in the full sample, information asymmetry significantly increased during the COVID-19 period. Similar results were observed for the distinct samples for the KOSPI and KOSDAQ markets. Thus, all listed firms experienced an abnormal situation during this period. Second, analyzing the changes by industry revealed that information asymmetry did not significantly increase in the pharmaceutical and IT industries. In contrast, it significantly increased in the samples belonging to the manufacturing and construction industries during the same period. This may be interpreted as a result of differences in each industry’s business environment even during the same COVID-19 period, which resulted in different information asymmetry.

This study makes several contributions. First, it examined the impact of COVID-19 on the information environment of firms. This paper suggests that serious efforts should be made to reduce the information asymmetry in the capital market to enhance the sustainability of firms. Furthermore, differing from previous studies, this study examines the impact of contextual factors on the information environment of firms in the COVID-19 pandemic situation. We show that the sustainability of firms can be greatly affected not only by economic conditions or events, but also by natural phenomena such as disasters and diseases.

Finally, this study presents the results of an empirical analysis that the overall information asymmetry increased during COVID, but the impact was different for each industry. This suggests that even the same environmental factors may have different effects for each industry. Therefore, the characteristics of the industry should be considered in the process of finding ways to mitigate information asymmetry.

Clearly, we should recognize that uncertainties in the business environment, such as the COVID-19 outbreak, which no one could have predicted, can change significantly. Capital market participants need to understand not only this, but must also discuss various ways to respond to it.