Good corporate governance, firm performance and COVID-19

Abstract

Purpose

This research is designed to analyze the effectiveness of the audit committee, nomination and remuneration committee, and ownership structure on company performance and how COVID-19 moderates the influence of these governance mechanisms on company performance.

Design/methodology/approach

437 annual reports of Indonesian manufacturing companies from 2018 to 2021 were used as research samples using multiple regression analysis and moderated regression analysis.

Findings

Good corporate governance plays a role in improving company performance. The presence of COVID-19 affects corporate governance, thereby reducing performance, but good corporate governance can limit this impact.

Practical implications

This research helps companies understand the effectiveness of the supervisory function in improving company performance. This research provides input for companies, regulators, and policymakers to pay attention to good corporate governance, especially when facing a crisis.

Originality/value

To my knowledge, research that examines corporate governance mechanisms and company performance related to COVID-19 and investigates whether COVID-19 moderates the influence of corporate governance mechanisms on company performance has never been conducted.

Citation

Putra, F. (2024), "Good corporate governance, firm performance and COVID-19", Asian Journal of Accounting Research, Vol. 9 No. 4, pp. 399-421. https://doi.org/10.1108/AJAR-07-2023-0227

1. Introduction

In Indonesia, there have been several accounting and financial scandals, such as the cases of Kimia Farma, Indofarma, Hanson International, Garuda Indonesia, and Envy Technologies Indonesia. To increase investor confidence in financial markets, the Indonesian government is trying to improve the corporate governance structure. The June 2018 edition of the rules requires companies listed on the Indonesia Stock Exchange (IDX) to have audit, nomination, and remuneration committees.

The effectiveness of audit committee monitoring depends on the size of the committee (Chaudhry et al., 2020), frequency of meetings (Musallam, 2020), and expertise (Salehi et al., 2018). According to resource dependency theory, firm performance improves when the audit committee is larger and exchanges of ideas occur during meetings. The importance of audit committee expertise in companies is reflected in various regulations, for example, the UK Corporate Governance Code (2016), the Sarbanes-Oxley Act (2002), and the Financial Services Authority Regulations (POJK) Number 55/POJK.04/2015, which confirm that at least one member of the audit committee has expertise in accounting and finance. Establishment of a nomination and remuneration committee based on POJK number 34/POJK.04/2014. If the nomination and remuneration committee finds prospective boards (the board of directors and the board of commissioners) who can align shareholder interests with an appropriate remuneration structure, then agency costs and information asymmetry can be reduced. Ownership structure is important because it aligns the interests of shareholders and company managers (Mardnly et al., 2018). This is in accordance with agency theory, which recognizes that ownership structure is related to monitoring.

COVID-19 is a global health and economic problem. This is due to the lockdown implemented by the government in various countries. Hsu and Liao (2022) found that good corporate governance (GCG) can reduce the impact of COVID-19 on stock price volatility and company trading volume but cannot help increase stock returns. COVID-19 affects company characteristics, including company performance (Khatib and Nour, 2021), while Thi Hoang and Nguyen (2023) found that disclosure related to COVID-19 in Vietnam was relatively low. Al-Maliki et al. (2023) found an increase in risk-taking, but they also found several opportunities by exploiting innovation. This was proven by Atayah et al. (2022), who found that the financial performance of logistics companies was much higher in 2020. For this reason, comprehensive policies are needed to overcome the negative impacts of the current and future crises (Boshnak et al., 2023). There have been many previous studies that have investigated the influence of corporate governance mechanisms on company performance, but not many have examined the impact during COVID-19. Previous research examining corporate governance mechanisms and COVID-19 only focused on them before and during COVID-19 (Anas et al., 2023; Atayah et al., 2022; Boshnak et al., 2023; Hsu and Liao, 2022; Khatib and Nour, 2021; Thi Hoang and Nguyen, 2023). Different from previous research, the motivation of this research is to determine corporate governance mechanisms (size, meetings, and expertise of the audit committee, nomination, and remuneration committee, as well as institutional and foreign ownership) with data before and during COVID-19 and to find out how COVID-19 moderates the influence of governance mechanisms on company performance.

This research uses manufacturing companies listed in Indonesia. Manufacturing is a type of company that is unique in the Indonesian stock exchange market because it has a very large number of issuers and a high market capitalization value on the IDX and provides the largest contribution to national gross domestic product (GDP), with an achievement of 17.34% in the second quarter of 2021. Indonesia has just been categorized as a developed country by the United States through the World Trade Organization (WTO), so GCG is needed to attract investors. When compared with developed countries in the world, corporate governance in Indonesia is still far behind, so the effectiveness of the audit committee, the existence of a nomination and remuneration committee, as well as the ownership structure (institutional and foreign), are important to research. Several previous studies on manufacturing companies found that corporate governance practices in Indonesia are still low, so they tend to compensate through higher dividend payments (Setiawan and Kee Phua, 2013). In line with this, Tandean and Winnie (2016) found that the audit committee had a positive effect, but institutional ownership had no effect on tax avoidance. Gill and Biger (2013) found that corporate governance plays a role in increasing the efficiency of working capital management in America. However, Wan Mohammad et al. (2016) found that the effectiveness of the board and audit committee improves earnings management in Malaysia. Large boards increase knowledge, thereby improving the quality of decision-making and performance in India (Arora and Sharma, 2016), but Palaniappan (2017) found board size reduces performance. Din et al. (2022) found that institutional ownership improves the performance of Pakistani companies, which is confirmed by the results of research by Gerged et al. (2023), which found board independence, gender diversity, audit committee independence, and institutional ownership reduce the likelihood of financial distress.

This research has several contributions to the corporate performance literature. First, expanding the corporate performance literature with aspects of the effectiveness of the audit committee, nomination and remuneration committee, and ownership structure Second, this research provides input to companies and regulators because there are still many companies that do not have nomination and remuneration committees. However, I provide evidence that these committees improve firm performance. Third, this research also provides input for regulators to pay attention to and formulate policies regarding the size and meetings of audit committees and majority share ownership in companies, especially institutional ones. Fourth, this research answers the effectiveness of the supervisory function during the crisis due to COVID-19, so this research contributes to all manufacturing companies and all stakeholders paying more attention to the supervisory function in companies because I found the expertise of the audit committee, nomination and remuneration committee, and foreign ownership have a significant positive effect on company performance, while audit committee size, audit committee meetings, and institutional ownership have no effect on company performance. The presence of COVID-19 negatively moderates audit committee meetings, nomination and remuneration committees, and foreign ownership but positively moderates audit committee expertise on company performance, and specifically for Tobin’s q, audit committee size is positively moderated, and when industrial control was carried out, audit committee meetings had a significant negative effect on Tobin’s q, and COVID-19 negatively moderates this relationship.

This research is further structured as follows: Section 2 contains the literature review and hypothesis development. Section 3 explains the research design. Section 4 presents the empirical results and discussion, and Section 5 contains the research summary and conclusions.

2. Literature review and hypothesis development

2.1 Characteristics of audit committees and firm performance

2.1.1 Audit committee size

POJK Number 55/POJK.04/2015 states that the audit committee must have a minimum of three members. A large audit committee has diverse abilities and ideas so as to increase knowledge (Anas et al., 2023). This opinion is in accordance with agency theory and resource dependency theory. Agency theory states that the separation of functions and the existence of contracts between principals (owners) and agents (managers) of a company give rise to information asymmetry, which opens up opportunities for opportunistic management behavior, so that supervision through an audit committee is needed. Meanwhile, resource dependency theory states that resources originating from outside the company will influence the administrative actions, behavior, and efficiency of the company. This relates to the audit committee, which is an independent committee. The larger the audit committee, the greater the independence and support for achieving company efficiency and performance. Several studies have found that audit committee size improves company performance (Chaudhry et al., 2020; Rahman and Ali, 2006). However, Fariha et al. (2022) found that audit committee size reduces Tobin’s q, while Salehi et al. (2018) and Reddy et al. (2010) found no effect. The audit committee provides input to the board of commissioners to anticipate and improve the impact of COVID-19 so that it will increase information disclosure related to COVID-19. Thi Hoang and Nguyen (2023) found that the presence of an audit committee in corporate governance had a positive impact on the extent of disclosure related to COVID-19, but Anas et al. (2023) found that audit committee size had no effect on company performance during COVID-19. Based on the theory and previous research that I have discussed, I have the following hypothesis:

H1a.

Audit committee size has a significant positive effect on firm performance.

H1b.

COVID-19 moderated the effect of the audit committee size on firm performance.

2.1.2 Audit committee meeting

Based on agency theory and resource dependency, meetings are a form of supervision and commitment by the audit committee to the company. OJK regulations state that the audit committee holds meetings at least once every three months. Several studies found that a high frequency of audit committee meetings led to increased company performance (Al Farooque et al., 2020; Musallam, 2020). However, Khatib and Nour (2021) found a decline in company performance, while Fariha et al. (2022) did not find any influence between the two. Audit committees tend to disclose more information regarding their initiatives related to COVID-19 and strategies that can increase transparency and accountability to stakeholders (Thi Hoang and Nguyen, 2023), which is obtained through meetings. In line with this, Anas et al. (2023) found that audit committee meetings during COVID-19 improved company performance. From the explanation and research that have been carried out, I have a hypothesis:

H2a.

Audit committee meetings have a significant positive effect on firm performance.

H2b.

COVID-19 moderated the effect of audit committee meetings on firm performance.

2.1.3 Expertise of audit committee

Based on agency theory, the audit committee’s accounting and financial expertise improves the monitoring and compliance function of financial report presentation. Meanwhile, based on resource dependency theory, diversity of expertise will increase insight. Research conducted by Musallam (2020) and Salehi et al. (2018) found that audit committee expertise improves company performance. Boshnak et al. (2023) found an increasing return on equity (ROE) before COVID-19, but research by Chaudhry et al. (2020) found no effect. Mousavi et al. (2022) found reduced financial reporting fraud and money laundering. Salehi et al. (2022) examine the relationship between corporate governance factors and financial reporting transparency before and after the crisis caused by the Islamic State of Iraq and Syria (ISIS). They found that audit committees that have capabilities in accounting and finance have a significant positive effect on financial reporting transparency, while the presence of ISIS causes a decrease in audit committee capabilities in accounting and finance on financial reporting transparency. Based on agency theory and resource dependency theory, as well as previous research I have presented, I hypothesize:

H3a.

Audit committee expertise has a significant positive effect on firm performance.

H3b.

COVID-19 moderated the effect of audit committee expertise on firm performance.

2.2 Nomination and remuneration committee, and company performance

Based on agency theory and resource dependency theory, this committee is tasked with supervising the performance of the boards and finding the best board members who will fill the position. Bansal and Singh (2022) found that remuneration nomination committees improve company performance, but Puni and Anlesinya (2020) found a decrease in the company’s financial performance. Jebran and Chen (2021) said that the remuneration committee was able to create an adjustment scheme that rewards executives who show extraordinary skills in running the firm during the COVID-19 period. Based on theory and previous research, I hypothesize as follows:

H4a.

The nomination and remuneration committee has a significant positive effect on company performance.

H4b.

COVID-19 moderates the influence of the nomination and remuneration committee on company performance.

2.3 Ownership structure and firm performance

2.3.1 Institutional Ownership

According to agency theory, the experience and ability to monitor institutional investors to protect their investments can improve company performance. This is proven by Yeh (2019), who found that institutional invesimprove company performance. Mishra and Kapil (2017) found an increase in Tobin’s q, while Salehi et al. (2018) found no effect. Jebran and Chen (2021) found that institutional ownership helped companies overcome the COVID-19 crisis. Salehi et al. (2022) find that corporate governance structures (including institutional ownership) played an important role in reducing the cost of equity during the ISIS crisis. Based on agency theory and the research that has been conducted, I hypothesize:

H5a.

Institutional ownership has a significant positive effect on firm performance.

H5b.

COVID-19 moderated the influence of institutional ownership on firm performance.

2.3.2 Foreign Ownership

Based on agency theory, foreign investors have large amounts of funds and have experience monitoring investment activities. Ferreira and Matos (2008) proved this by finding improvements in operational performance and company investment policies, as well as improving company performance (Mardnly et al., 2018), but Din et al. (2022) did not find any influence between the two. Jebran and Chen (2021) say that foreign investors are able to play a role in governance during a crisis because of their ability to monitor, have information about, and analyze the investments they make. Based on agency theory and previous research, I hypothesize:

H6a.

Foreign ownership has a significant positive effect on firm performance.

H6b.

COVID-19 moderated the effect of foreign ownership on firm performance.

3. Research design

3.1 Data, database, and sampling

This research is quantitative, using annual reports of manufacturing companies listed on the IDX in the 2018–2021 period to balance data before the crisis due to COVID-19 (2018–2019) and during the crisis due to COVID-19 (2020–2021) so that the research results are more accurate. I then processed the data with STATA. The calculation of the number of research samples can be seen in Table 1.

3.2 Data collection

The dependent variables in this research are company performance as measured by financial performance, namely return on assets (ROA), and market performance as measured by Tobin’s q. Meanwhile, the independent variables are Audit Committee Size (ACSIZE), Audit Committee Meetings (ACMEET), Audit Committee Expertise (ACEXPERTISE), Nomination and Remuneration Committee (NRC), Institutional Ownership (IO), and Foreign Ownership (FO). COVID-19 (COV) is used as a moderating variable. The control variables in this study use leverage (LEV), growth (GROWTH), and year control using COVID-19 (COV). All definitions and operational variables can be seen in Table 2.

3.3 Model specification

The research was conducted using multiple regression analysis and moderated regression analysis with the help of STATA. The assessment in the regression test is a sign of the regression coefficient value and significance value. The analysis in this research uses the following equational regression model:

(1)Performancei,t=α+β1ACSIZEi,t+β2ACMEETi,t+β3ACEXPERTISEi,t+β4NRCi,t+β5IOi,t+β6FOi,t+β7LEVi,t+β8GROWTHi,t+β9COVi,t+εi,t

(2)Performancei,t=α+β1ACSIZEi,t+β2ACMEETi,t+β3ACEXPERTISEi,t+β4NRCi,t+β5IOi,t+β6FOi,t+β7ACSIZEi,t × Covi,t+β8ACMEETi,t × Covi,t+β9ACEXPERTISEi,t × Covi,t+β10NRCi,t × Covi,t+β11IOi,t × Covi,t+β12FOi,t × Covi,t+β13LEVi,t+β14GROWTHi,t+β15COVi,t+εi,t

4. Result and analysis

4.1 Description statistics

Table 3 is descriptive statistics; the mean ROA and Tobin’s q are 0.024 and 1.313, respectively. The mean ROA has not shown promising financial performance. This is likely caused by the economic crisis resulting from the COVID-19 pandemic, so the minimum ROA value is −1.079. However, on the other hand, Tobin’s q is included in the expensive category because the mean value is above 1. The mean ACSIZE and ACMEET values are 3.038 and 6.670, with minimum values of 2 and 1. This is not in accordance with OJK regulations. The mean ACEXPERTISE value is 0.669, and the minimum value is 0.25, which means it complies with OJK regulations. The mean NRC value is 0.466, with a minimum value of 0 (does not have an NRC), which means that less than half of the sample companies have this committee. The mean value of IO is 0.854 and the mean value of FO is 0.275, while the maximum value of FO is 0.998, which means that there are shares that are majority owned by foreigners.

4.2 Correlation

ACMEET, NRC, FO, and GROWTH have a significant positive correlation with ROA, while LEV and COV have a significant negative correlation. Meanwhile, for Tobin’s q, the variables ACMEET, NRC, FO, LEV, and growth have a significant positive correlation; I show this in Tables 4 and 5.

4.3 Regression analysis

To select the regression model, this research uses the Chow test, the Lagrange multiplier (LM) test, and the Hausman test. After passing these tests, what meets the criteria for use in my research is a regression model using random effects. In Tables 6 and 7, it can be seen that ACSIZE has an insignificant positive effect on ROA and Tobin’s q. These results are in accordance with the research of Salehi et al. (2018) and Reddy et al. (2010). This is likely to happen because too many audit committees will cause coordination difficulties and incur costs for salaries, allowances, and official travel costs (related to agency costs). I show the results of my moderation in Tables 8 and 9, showing that the market gave a positive response, which means that COVID-19 strengthens the influence of ACSIZE on Tobin’s q. Meanwhile, COVID-19 strengthens the positive influence of ACSIZE on ROA, although it is not significant. This may be because the response to Tobin’s q is faster than ROA because the market believes that a larger number of audit committees will carry out a more effective supervisory function due to the social restrictions implemented by the government. These results also support agency theory, resource dependence theory, and OJK regulation because a large number of audit committees will strengthen supervision with various expertise and experience so as to reduce the impact of COVID-19 and improve company performance. Thus, H1a is rejected, but H1b is accepted.

ACMEET has an insignificant positive effect on ROA and Tobin’s q. These results are in accordance with several previous studies (Fariha et al., 2022; Rahman and Ali, 2006). This happens because audit committee meetings are only held to fulfill requirements without considering the quality of the meeting, and these meetings also create burdens for the company, such as travel, hotels, and time. The moderation results show that accounting and markets are responding to changes resulting from COVID-19 by weakening the function of committee meetings. These results do not support agency theory and resource dependency theory because holding meetings more frequently will incur large costs, thereby reducing company performance, while companies have to make savings due to the impact of COVID-19. My opinion is based on the meeting fees that the company has to pay. Therefore, in order to maintain coordination and communication, audit committee meetings can be held via teleconference (online meeting). Thus, H2a is rejected and H2b is accepted.

ACEXPERTISE has a significant positive effect on ROA and Tobin’s q. These results are in accordance with several studies (Mousavi et al., 2022; Musallam, 2020; Salehi et al., 2018). In addition, COVID-19 strengthens the influence of audit committee expertise on company performance, meaning that audit committee accounting and financial expertise are needed to monitor the impact of COVID-19 as a whole. These results support agency theory, resource dependency theory, and OJK regulation. Agency theory and OJK regulation state that audit committee accounting and financial expertise are needed to improve supervision related to financial presentation and reporting. while resource dependency theory states that the diversity of audit committee expertise will add insight and capabilities that can improve company performance. Thus, H3a and H3b are accepted.

NRC has a significant positive effect on ROA and Tobin’s q. The results of this study are in line with Bansal and Singh (2022). The inclusion of COVID-19 weakens the influence of the NRC on company performance. Restrictions on community activities are most likely the cause, so the NRC cannot carry out its supervisory function properly. These results support agency theory, resource dependency theory, and OJK regulations, which state that the nomination and remuneration committee is tasked with determining prospective boards who will serve and supervising the performance of members of the boards who currently serve in the company. Thus, H4a and H4b are accepted.

IO has an insignificant positive effect on ROA and Tobin’s q. These results support the research of Salehi et al. (2018) and the strategic alignment hypothesis, which says institutional investors only invest for short-term profits. Meanwhile, COVID-19 does not moderate the influence of IO on company performance, which means that the presence or absence of a crisis does not affect the influence of IO on company performance. This result does not support agency theory, which argues that institutional investors, with their ability and experience, can carry out intensive supervision to protect their investments. This likely happens because institutional investors keep reserve funds for their needs during COVID-19. So H5a and H5b are rejected.

FO has a significant positive effect on ROA and Tobin’s q. These results support previous research (Mardnly et al., 2018; Saidat et al., 2019). These results also support agency theory and resource dependency theory because foreign ownership can increase investment because the knowledge, experience, and technology used by foreign investors provide changes for the better in the company, thereby increasing company performance. These results also prove that the regulations issued by the Indonesian government increase the interest of foreign investors in investing their capital. COVID-19 weakens the influence of FO on ROA and Tobin’s q. This likely happened because the FO doubted the condition of the Indonesian economy, so the positive influence of the FO was weakened by COVID-19. The results of this research support the opinion of Jebran and Chen (2021), who say that foreign investors play a role in governance during a crisis because their monitoring ability reduces information asymmetry; thus, H6a and H6b are accepted. For the control variables, LEV and COVID have a significant negative effect on ROA and Tobin’s q, while growth has a significant positive effect.

4.4 Additional test

To confirm the research results, I carried out additional testing, namely using industrial controls. I did this because several industries experienced increased performance during COVID-19, such as the chemical, pharmaceutical, and medical equipment industries. The regression results show a slight difference, where COVID-19 does not weaken the influence of ACMEET on ROA. While for Tobin’s q, the majority of the results are the same, but ACMEET has a significant negative effect. I show these results in Tables 10 and 11.

4.5 Robustness test

To check the robustness of the results, I tested using return on equity (ROE) (Boshnak et al., 2023; Din et al., 2022). The majority of durability test results have similar results. In addition, to control the endogeneity problem, generalized method of moment (GMM) estimation is used with the first-differences GMM (FD-GMM) and system GMM (SYS-GMM) methods with the two-step GMM technique, which supports the basic regression results with random effects. I show these two robustness tests in Tables 1 and 12–14.

5. Conclusion

The GCG implemented by the company aims to improve performance so that it can provide benefits to all stakeholders. Previous research found that the COVID-19 pandemic changed the entire corporate governance system. Therefore, this research examines the effectiveness of the audit committee, the presence of a nomination and remuneration committee, and ownership structure on company performance by using COVID-19 as a moderating variable. Using a sample of manufacturing companies in Indonesia, this research finds that ACEXPERTISE, NRC, and FO have a significant positive effect on company performance, and COVID-19 weakens the influence of ACMEET, NRC, and FO, strengthens the influence of ACEXPERTISE on company performance, and strengthens ACSIZE on Tobin’s q. To confirm the results of this study, I have also performed additional tests and robustness tests. The majority of additional and durability test results are consistent with previous test results.

I recommend that companies be careful in determining the composition of the audit committee and ownership structure, as well as establishing a nomination and remuneration committee. Although several studies have found that audit committee size and audit committee meetings can improve company performance, this was not found in this study. The size of the audit committee and frequent audit committee meetings can cause agency problems and agency costs and make it difficult to coordinate with each other. Likewise with the ownership structure, companies must also be careful about the composition of the company’s shareholders, especially FO, because it can reduce company performance, especially during times of crisis. For policymakers, the results of this research provide input to improve corporate governance through laws and policies by determining the maximum number of ACSIZE, ACMEET, and shareholder composition and requiring the establishment of an NRC in companies. Even though this research was conducted in Indonesia and the crisis was caused by COVID-19, the results of this research can be applied to other developing countries, and the crisis was caused by other things.

This research opens up new research avenues in the future. Current research focuses on how companies survive in the face of economic crises. Therefore, I suggest that further research examine the period before, during, and after the COVID-19 pandemic so that researchers can explain the impact of improving corporate governance on increasing company performance. Further research can also increase the number of samples, not only for manufacturing companies, so that the research results can have a wider impact. Further research can also add other characteristics, such as experience in the number of audit committee industries and the length of time the audit committee has served in the company and industry, because this can influence the knowledge and ability of the audit committee to carry out its duties. This research only examines whether or not there is a nomination and remuneration committee. Therefore, for future research, I suggest examining committee characteristics such as expertise in human resources and experience in the field of nomination and remuneration. Finally, this research was conducted in Indonesia, which adheres to a two-level system. Future research could be conducted comparing the results with those of countries that adopted a one-tier system.

Table 1

Research sample

The number of manufacturing companies in Indonesia in 2021	207
Population of manufacturing company annual reports for the 2018–2021 period available	637
Incomplete data	137
Inaccessible annual report	55
The end of the financial reporting period is not December 31	8
Number of samples for annual observation reports	437

Source(s): Table created by author

Table 2

Definition and operational variables

Variable	Variable name	Acronym	Measurement
Dependent	Return On Asset	ROA	Calculated by comparing net profit and total assets (Chaudhry et al., 2020)
	Tobin’s q	Tobin’s q	Calculated by the number of firm stocks outstanding multiplied by the firm closing price plus the market value of debt divided by total assets (Bansal and Singh, 2022; Al Farooque et al., 2020)
Independent	Audit committee size	ACSIZE	The number of audit committees in the firm (Musallam, 2020)
	Audit committee meetings	ACMEET	The number of audit committee meetings in one year (Fariha et al., 2022)
	Audit committee expertise	ACEXPERTISE	The number of audit committees with accounting and finance expertise compared with the total number of audit committee members in the firm (Musallam, 2020)
	Nomination and remuneration committees	NRC	Dummy variable, given a value of 1 if there is a nomination and remuneration committee in the company and a value of 0 if there is no one (Bansal and Singh, 2022; Puni and Anlesinya, 2020)
	Institutional ownership	IO	Total shares owned by the institution divided by the number of outstanding shares (Mishra and Kapil, 2017; Salehi et al., 2018)
	Foreign Ownership	FO	Total shares owned by foreigners divided by the number of outstanding shares (Saidat et al., 2019)
Moderating	COVID-19	COV	The dummy variable, the period before COVID-19 (2018 and 2019), was given a score of 0 and a score of 1 for the COVID-19 crisis (2020 and 2021)
Control	Leverage	LEV	Measured by the DAR ratio, namely total debt divided by total assets
	Growth	GROWTH	Total assets at the end of the period minus the total assets at the beginning of the period divided by the total assets at the end of the period
	COVID-19	COV	Dummy year control is given a value of 1 if the year is the year that COVID-19 occurred (2020 and 2021) and is given a value of 0 if it is not the year that COVID-19 occurred (2018–2019)

Source(s): Table created by author

Table 3

Descriptive statistics

Variable	Obs	Mean	Std. Dev	Min	Max
ROA	437	0.024	0.133	−1.079	0.658
Tobin’s Q	437	1.313	0.906	0.000	4.995
ACSIZE	437	3.038	0.332	2	5
ACMEET	437	6.670	5.007	1	38
ACEXPERTISE	437	0.669	0.229	0.25	1
NRC	437	0.466	0.499	0	1
IO	437	0.854	4.568	0	0.958
FO	437	0.275	0.331	0.000	0.998
LEV	437	0.506	0.339	0.000	2.821
GROWTH	437	0.034	0.476	−8.532	2.076
COVID	437	0.501	0.500	0	1

Source(s): The data was processed by the STATA application, and the table were created by the author

Table 4

Correlation analysis ROA

Variables	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)
(1) ROA	1.000
(2) ACSIZE	0.069	1.000
	(0.152)
(3) ACMEET	0.087*	0.059	1.000
	(0.069)	(0.221)
(4) ACEXPERTISE	0.066	0.025	−0.179*	1.000
	(0.171)	(0.603)	(0.000)
(5) NRC	0.123*	0.180*	0.155*	−0.066	1.000
	(0.010)	(0.000)	(0.001)	(0.169)
(6) IO	0.045	−0.008	−0.004	−0.009	0.046	1.000
	(0.348)	(0.868)	(0.930)	(0.851)	(0.342)
(7) FO	0.139*	0.175*	−0.059	−0.012	0.103*	0.005	1.000
	(0.004)	(0.000)	(0.215)	(0.806)	(0.032)	(0.917)
(8) LEV	−0.423*	0.036	0.083*	0.031	−0.084*	−0.011	−0.125*	1.000
	(0.000)	(0.458)	(0.083)	(0.515)	(0.079)	(0.825)	(0.009)
(9) GROWTH	0.191*	0.021	−0.001	−0.031	0.087*	0.000	−0.080*	−0.018	1.000
	(0.000)	(0.663)	(0.986)	(0.515)	(0.068)	(0.994)	(0.094)	(0.714)
(10) COV	−0.121*	0.007	0.020	0.011	0.007	0.049	0.000	0.057	−0.092*	1.000
	(0.011)	(0.890)	(0.672)	(0.822)	(0.883)	(0.302)	(0.999)	(0.236)	(0.055)

Note(s): *p < 0.1

Source(s): Processed table of the STATA application

Table 5

Correlation analysis Tobin’s q

Variables	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)
(1) Tobin’s q	1.000
(2) ACSIZE	0.022	1.000
	(0.653)
(3) ACMEET	0.119*	0.059	1.000
	(0.013)	(0.221)
(4) ACEXPERTISE	0.030	0.025	−0.179*	1.000
	(0.535)	(0.603)	(0.000)
(5) NRC	0.208*	0.180*	0.155*	−0.066	1.000
	(0.000)	(0.000)	(0.001)	(0.169)
(6) IO	0.017	−0.008	−0.004	−0.009	0.046	1.000
	(0.719)	(0.868)	(0.930)	(0.851)	(0.342)
(7) FO	0.133*	0.175*	−0.059	−0.012	0.103*	0.005	1.000
	(0.005)	(0.000)	(0.215)	(0.806)	(0.032)	(0.917)
(8) LEV	0.092*	0.036	0.083*	0.031	−0.084*	−0.011	−0.125*	1.000
	(0.054)	(0.458)	(0.083)	(0.515)	(0.079)	(0.825)	(0.009)
(9) GROWTH	0.134*	0.021	−0.001	−0.031	0.087*	0.000	−0.080*	−0.018	1.000
	(0.005)	(0.663)	(0.986)	(0.515)	(0.068)	(0.994)	(0.094)	(0.714)
(10) COV	0.002	0.007	0.020	0.011	0.007	0.049	0.000	0.057	−0.092*	1.000
	(0.971)	(0.890)	(0.672)	(0.822)	(0.883)	(0.302)	(0.999)	(0.236)	(0.055)

Note(s): *p < 0.1

Source(s): Processed table of the STATA application

Table 6

ROA regression results

Independent variable	(1)	(2)	(3)	(4)	(5)	(6)	(7)
ACSIZE	0.018 (0.354)	0.023 (0.249)
ACMEET	0.000 (0.751)		0.000 (0.729)
ACEXPERTISE	0.045* (0.061)			0.048* (0.071)
NRC	0.019* (0.051)				0.021* (0.096)
IO	0.000 (0.402)					0.000 (0.317)
FO	0.033* (0.050)						0.038* (0.092)
LEV	−0.204*** (0.000)	−0.211*** (0.000)	−0.210*** (0.000)	−0.210*** (0.000)	−0.209*** (0.000)	−0.210*** (0.000)	−0.206*** (0.000)
GROWTH	0.033*** (0.000)	0.034*** (0.000)	0.034*** (0.000)	0.034*** (0.000)	0.033*** (0.000)	0.033*** (0.000)	0.034*** (0.000)
COV	−0.022*** (0.006)	−0.021*** (0.009)	−0.021*** (0.009)	−0.021*** (0.008)	−0.021*** (0.008)	−0.021*** (0.008)	−0.021*** (0.008)
Prob > chi2	0.000	0.000	0.000	0.000	0.000	0.000	0.000
Overall r-squared	0.238	0.218	0.213	0.219	0.218	0.214	0.22

Note(s): ***p < 0.01, **p < 0.05, *p < 0.1

Source(s): The data was processed by the STATA application, and the table were created by the author

Table 7

Tobin’s q regression results

Independent variable	(1)	(2)	(3)	(4)	(5)	(6)	(7)
ACSIZE	0.051 (0.650)	0.020 (0.856)
ACMEET	0.009 (0.262)		0.004 (0.601)
ACEXPERTISE	0.419* (0.094)			0.315* (0.098)
NRC	0.248** (0.036)				0.232** (0.040)
IO	0.003 (0.483)					0.002 (0.604)
FO	0.335** (0.023)						0.340** (0.020)
LEV	−0.736*** (0.000)	−0.707*** (0.000)	−0.710*** (0.000)	−0.698*** (0.000)	−0.701*** (0.000)	−0.709*** (0.000)	−0.730*** (0.000)
GROWTH	0.416*** (0.000)	0.411*** (0.000)	0.411*** (0.000)	0.411*** (0.000)	0.411*** (0.000)	0.411*** (0.000)	0.414*** (0.000)
COV	−0.059** (0.040)	−0.059** (0.059)	−0.058** (0.054)	−0.057** (0.052)	−0.060** (0.053)	−0.060** (0.053)	−0.060** (0.059)
Prob > chi2	0.000	0.000	0.000	0.000	0.000	0.000	0.000
Overall r-squared	0.060	0.024	0.021	0.025	0.052	0.024	0.037

Note(s): ***p < 0.01, **p < 0.05, *p < 0.1

Source(s): The data was processed by the STATA application, and the tables were created by the author

Table 8

Moderating the Covid-19 test using ROA

Independent variable	(1)	(2)	(3)	(4)	(5)	(6)
ACSIZE	0.025 (0.425)
ACMEET		0.000 (0.829)
ACEXPERTISE			0.039** (0.022)
NRC				0.021** (0.035)
IO					0.004 (0.832)
FO						0.024** (0.033)
ACSIZE × COV	0.037 (0.322)
ACMEET × COV		−0.001* (0.093)
ACEXPERTISE × COV			0.016* (0.051)
NRC × COV				−0.001** (0.037)
IO × COV					0.005 (0.796)
FO × COV						−0.031* (0.06)
LEV	−0.207*** (0.000)	−0.209*** (0.000)	−0.210*** (0.000)	−0.209*** (0.000)	−0.210*** (0.000)	−0.204*** (0.000)
GROWTH	0.033*** (0.000)	0.034*** (0.000)	0.034*** (0.000)	0.034*** (0.000)	0.034*** (0.000)	0.035*** (0.000)
COV	−0.009 (0.400)	−0.011 (0.403)	−0.032* (0.099)	−0.020* (0.063)	−0.025* (0.099)	−0.300*** (0.005)
Prob > chi2	0.000	0.000	0.000	0.000	0.000	0.000
Overall r-squared	0.208	0.214	0.219	0.218	0.215	0.215

Note(s): ***p < 0.01, **p < 0.05, *p < 0.1

Source(s): The data was processed by the STATA application, and the table were created by the author

Table 9

Moderating the COVID-19 test using Tobin’s q

Independent variable	(1)	(2)	(3)	(4)	(5)	(6)
ACSIZE	0.035 (0.750)
ACMEET		0.001 (0.855)
ACEXPERTISE			0.498* (0.056)
NRC				0.297** (0.013)
IO					−0.044 (0.665)
FO						0.374* (0.015)
ACSIZE × COV	0.015* (0.098)
ACMEET × COV		−0.009* (0.089)
ACEXPERTISE × COV			0.316** (0.048)
NRC × COV				−0.127* (0.087)
IO × COV					0.046 (0.648)
FO × COV						−0.082* (0.064)
LEV	−0.692*** (0.000)	−0.714*** (0.000)	−0.691*** (0.000)	−0.708*** (0.000)	−0.706*** (0.000)	−0.72*** (0.000)
GROWTH	0.402*** (0.000)	0.411*** (0.000)	0.404*** (0.000)	0.406*** (0.000)	0.410*** (0.000)	0.417*** (0.000)
COV	−0.257** (0.017)	−0.004 (0.947)	−0.268** (0.017)	−0.000 (0.991)	−0.089 (0.226)	−0.037 (0.433)
Prob > chi2	0.000	0.000	0.000	0.000	0.000	0.000
Overall r-squared	0.029	0.023	0.026	0.052	0.023	0.037

Note(s): ***p < 0.01, **p < 0.05, *p < 0.1

Source(s): The data was processed by the STATA application, and the table were created by the author

Table 10

Additional tests using ROA with industry control

Independent variable	(1)	(2)	(3)	(4)	(5)	(6)	(7)
ACSIZE	0.019 (0.346)	0.024 (0.237)
ACMEET	−0.000 (0.764)		−0.000 (0.891)
ACEXPERTISE	0.050* (0.072)			0.046* (0.059)
NRC	0.015* (0.080)				0.016* (0.074)
IO	0.000 (0.409)					−0.003 (0.858)
FO	0.031* (0.084)						0.022* (0.081)
ACSIZE × COV		−0.001 (0.309)
ACMEET × COV			−0.001 (0.387)
ACEXPERTISE × COV				0.016** (0.039)
NRC × COV					−0.072* (0.065)
IO × COV						0.004 (0.823)
FO × COV							−0.030* (0.076)
LEV	−0.206*** (0.000)	−0.210*** (0.000)	−0.210*** (0.000)	−0.211*** (0.000)	−0.211*** (0.000)	−0.211*** (0.000)	−0.206*** (0.000)
GROWTH	0.033*** (0.000)	0.034*** (0.000)	0.034*** (0.000)	0.034*** (0.000)	0.033*** (0.000)	0.034*** (0.000)	0.035*** (0.000)
COV	−0.022*** (0.007)	−0.011 (0.340)	−0.011 (0.412)	−0.032* (0.094)	−0.020* (0.061)	−0.024 (0.117)	−0.029*** (0.005)
Industry	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Prob > chi2	0.000	0.000	0.000	0.000	0.000	0.000	0.000
Overall r-squared	0.249	0.236	0.230	0.236	0.229	0.229	0.236

Note(s): ***p < 0.01, **p < 0.05, *p < 0.1

Source(s): The data was processed by the STATA application, and the table were created by the author

Table 11

Additional tests using Tobin’s q with industry control

Independent variable	(1)	(2)	(3)	(4)	(5)	(6)	(7)
ACSIZE	−0.035 (0.755)	−0.003 (0.976)
ACMEET	−0.009* (0.070)		−0.075** (0.037)
ACEXPERTISE	0.313** (0.018)			0.386** (0.041)
NRC	0.208* (0.081)				0.252** (0.036)
IO	0.002 (0.536)					0.038 (0.706)
FO	0.284* (0.055)						0.321** (0.036)
ACSIZE × COV		0.009* (0.056)
ACMEET × COV			−0.009** (0.029)
ACEXPERTISE × COV				0.321** (0.044)
NRC × COV					−0.123* (0.098)
IO × COV						0.040 (0.690)
FO × COV							−0.086** (0.044)
LEV	−0.731*** (0.000)	−0.709*** (0.000)	−0.708*** (0.000)	−0.686*** (0.000)	−0.704*** (0.000)	−0.697*** (0.000)	−0.713*** (0.000)
GROWTH	0.415*** (0.000)	0.410*** (0.000)	0.410*** (0.000)	0.403*** (0.000)	0.406*** (0.000)	0.409*** (0.000)	0.416*** (0.000)
COV	−0.059* (0.09)	−0.004 (0.943)	−0.006 (0.922)	−0.271** (0.016)	−0.001 (0.970)	−0.085 (0.249)	−0.036 (0.451)
Industry	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Prob > chi2	0.000	0.000	0.000	0.000	0.000	0.000	0.000
Overall r-squared	0.101	0.083	0.084	0.085	0.108	0.085	0.092

Note(s): ***p < 0.01, **p < 0.05, *p < 0.1

Source(s): The data was processed by the STATA application, and the table were created by the author

Table 12

Robustness test using ROE

Independent variable	(1)	(2)	(3)	(4)	(5)	(6)	(7)
ACSIZE	0.076 (0.725)	0.058 (0.435)
ACMEET	0.067 (0.640)		0.003 (0.851)
ACEXPERTISE	0.125* (0.085)			0.002* (0.096)
NRC	0.030* (0.083)				0.075* (0.072)
IO	0.002 (0.865)					−0.032 (0.912)
FO	0.226* (0.098)						0.288** (0.026)
ACSIZE × COV		0.025 (0.150)
ACMEET × COV			−0.008* (0.057)
ACEXPERTISE × COV				0.309* (0.061)
NRC × COV					−0.189* (0.097)
IO × COV						0.035 (0.905)
FO × COV							−0.100* (0.081)
LEV	−0.338* (0.086)	−0.350* (0.076)	−0.353* (0.086)	−0.353* (0.071)	−0.363* (0.071)	−0.366* (0.074)	−0.339* (0.097)
GROWTH	0.107* (0.069)	0.090* (0.051)	0.091* (0.053)	0.097* (0.055)	0.098* (0.055)	0.090* (0.053)	0.102* (0.086)
COV	−0.013* (0.082)	−0.062* (0.058)	−0.072* (0.054)	−0.193* (0.052)	−0.103* (0.059)	−0.103* (0.067)	−0.041* (0.081)
Prob > chi2	0.004	0.025	0.035	0.053	0.053	0.060	0.021
Overall r-squared	0.128	0.092	0.094	0.095	0.094	0.084	0.081

Note(s): ***p < 0.01, **p < 0.05, *p < 0.1

Source(s): The data was processed by the STATA application, and the table were created by the author

Table 13

Additional tests using ROE with industry control

Independent variable	(1)	(2)	(3)	(4)	(5)	(6)	(7)
ACSIZE	0.055 (0.802)	0.112 (0.598)
ACMEET	−0.006 (0.658)		−0.004 (0.831)
ACEXPERTISE	0.149** (0.042)			0.027* (0.050)
NRC	0.013** (0.028)				0.085* (0.075)
IO	0.003 (0.797)					−0.036 (0.903)
FO	0.259* (0.052)						0.331* (0.069)
ACSIZE × COV		−0.013 (0.484)
ACMEET × COV			−0.008 (0.768)
ACEXPERTISE × COV				0.309** (0.012)
NRC × COV					−0.193* (0.090)
IO × COV						0.039 (0.894)
FO × COV							−0.106* (0.082)
LEV	−0.353* (0.094)	−0.374* (0.071)	−0.370* (0.074)	−0.386* (0.062)	−0.378* (0.068)	−0.383 (0.064)	−0.355* (0.088)
GROWTH	0.093** (0.032)	0.079* (0.093)	0.008* (0.088)	0.086* (0.061)	0.086* (0.060)	0.079* (0.059)	0.091* (0.053)
COV	−0.012* (0.093)	−0.103* (0.058)	−0.069 (0.765)	−0.193 (0.653)	−0.104 (0.587)	−0.013 (0.955)	−0.042 (0.815)
Industry	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Prob > chi2	0.022	0.026	0.046	0.036	0.047	0.080	0.073
Overall r-squared	0.015	0.011	0.011	0.011	0.011	0.010	0.014

Note(s): ***p < 0.01, **p < 0.05, *p < 0.1

Source(s): The data was processed by the STATA application, and the table were created by the author

Table 14

Regression using GMM estimation

Independent variable	FD GMM			SYS GMM
	ROA	Tobin’s q	ROE	ROA	Tobin’s q	ROE
ROA L1	0.134* (0.065)			0.111* (0.086)
Tobin’s q L1		−0.032* (0.083)			0.129* (0.057)
ROE L1			0.658* (0.075)			0.819*** (0.000)
ACSIZE	0.073 (0.209)	−0.145 (0.327)	0.907 (0.399)	0.049 (0.286)	−0.121 (0.475)	0.541 (0.533)
ACMEET	0.002 (0.361)	−0.032** (0.045)	−0.035 (0.541)	0.001 (0.683)	−0.004* (0.079)	−0.031 (0.633)
ACEXPERTISE	0.179* (0.055)	0.271** (0.029)	0.360* (0.061)	0.007* (0.070)	0.215** (0.041)	0.881* (0.094)
NRC	0.030** (0.035)	0.363*** (0.004)	0.229** (0.022)	0.038* (0.061)	0.373** (0.046)	0.105** (0.018)
IO	0.041 (0.180)	0.043 (0.686)	0.199 (0.560)	0.021 (0.580)	0.034 (0.751)	0.144 (0.693)
FO	0.165* (0.057)	0.082** (0.033)	0.590* (0.080)	0.151* (0.094)	0.187* (0.061)	0.013* (0.091)
ACSIZE × COV	−0.056 (0.205)	0.092* (0.085)	−0.028 (0.235)	−0.075 (0.355)	0.098* (0.092)	−0.032 (0.352)
ACMEET × COV	−0.006 (0.751)	−0.003* (0.073)	−0.007 (0.765)	−0.003 (0.855)	−0.006* (0.053)	−0.005 (0.812)
ACEXPERTISE × COV	0.012* (0.076)	0.183** (0.041)	0.443* (0.074)	0.024* (0.081)	0.172** (0.021)	0.439* (0.094)
NRC × COV	−0.020** (0.041)	−0.040* (0.065)	−0.080** (0.025)	−0.009* (0.061)	−0.029** (0.046)	−0.039* (0.083)
IO × COV	0.040 (0.187)	0.047 (0.657)	0.202 (0.552)	0.020 (0.591)	0.038 (0.721)	0.136 (0.707)
FO × COV	−0.028* (0.074)	−0.013* (0.092)	−0.178* (0.084)	−0.004* (0.091)	−0.045* (0.050)	−0.019* (0.094)
LEV	−0.398*** (0.005)	−0.690*** (0.006)	−0.328 (0.791)	−0.234* (0.091)	−0.652** (0.011)	−0.263 (0.747)
GROWTH	0.045 (0.318)	0.488*** (0.000)	0.101 (0.578)	0.025 (0.566)	0.493*** (0.000)	0.085 (0.665)
COV	−0.018 (0.519)	−0.175 (0.184)	−0.474* (0.085)	−0.031 (0.315)	−0.175 (0.264)	−0.303* (0.047)
Industry	Yes	Yes	Yes	Yes	Yes	Yes
Prob > chi2	0.000	0.000	0.009	0.000	0.000	0.000
AR (1)	−1.7868 (0.174)	−1.172 (0.241)	0.674 (0.501)	−1.827 (0.167)	−1.730 (0.183)	0.582 (0.560)
AR (2)	−1.578 (0.157)	1.052 (0.224)	0.475 (0.445)	−1.678 (0.159	1.225 (0.255)	0.492 (0.483)
Number of obs	217	217	217	326	326	326

Note(s): ***p < 0.01, **p < 0.05, *p < 0.1 The two-step GMM technique uses robust standard errors to control for heteroscedasticity and autocorrelation

Source(s): The data was processed by the STATA application, and the table were created by the author