Corporate Governance and Financial Statement Fraud during the COVID-19: Study of Companies under Special Monitoring in Indonesia

Abstract

The COVID-19 pandemic had a wide-ranging impact, resulting in a global recession due to weakened purchasing power. This circumstance necessitates business organizations adapting to developments and being more conscious of the risk of financial statement fraud. The intention of this research is to investigate the way corporate governance affected financial statement fraud during the COVID-19 pandemic. To acquire empirical data for examining corporate governance variables on financial statement fraud, the research was examined using quantitative methods. The study takes advantage of secondary data acquired from annual reports of companies under special monitoring listed on the Indonesia Stock Exchange of 2020–2021. The logistic regression method was used to evaluate 134 data sets, and financial statement fraud was measured using the Z-Score and F-Score models. The results indicate that when using the Z-score, only the board size has a negative effect on financial statement fraud during the COVID-19 pandemic. Meanwhile, using the F-Score, the corporate governance variables studied are not proven to have an influence on financial statement fraud during the COVID-19 pandemic.

1. Introduction

The COVID-19 pandemic affected all industrial sectors in various regions. Economic growth has stagnated as the purchasing power of people has weakened due to restrictions on people’s activities in various fields, leading to a global recession (Shen et al. 2020). Each entity tries to adapt to changes due to the pandemic and needs to increase awareness of fraud risks related to financial statements that may occur (Campanelli et al. 2020). The weakening economy is feared to have an impact on management performance assessments, especially those measured by financial targets. Efforts made to stabilize the financial statements can encourage management to commit fraud to cover up various deficiencies that occur in the company (Schilit et al. 2018). This condition gave the industry and companies the opportunity to commit financial statement fraud during the pandemic (Putra 2022).

Financial statement fraud is the intentional falsification of financial information in order to defraud investors, creditors, or other stakeholders. This can be done by both internal and external parties (Carmichael 2020). Personal gain, keeping the business viable, and maintaining a position as a leader in the organization are all motivations for perpetrating financial statement fraud. Fraudsters seek to inflate the perceived value of the firm in order to make the stock appear more appealing to investors, gain bank loan approvals, and/or justify huge salaries and bonuses when compensation is related to corporate performance (Beaver 2022).

Management fraud and the manipulation of financial statements can be discovered in the instance of PT Hanson International Tbk, which recognized revenue at the beginning using the full accrual technique and did not present the sale and purchase agreement in the 2016 financial statements. Due to this revenue recognition, the December 2016 financial statements were overstated by IDR 613 billion (Sandria 2021). Another case related to financial statement fraud also occurred at PT Tiga Pilar Sejahtera Food Tbk in 2018. In the 2017 financial statements, based on the auditor’s findings, there was an exaggeration of funds of IDR 4 trillion in the company’s receivable, inventory, and fixed assets accounts. The auditor also found an exaggeration in the sales item of IDR 622 billion and an exaggeration of earnings before interest, taxes, depreciation, and amortization of IDR 322 billion (Wareza 2019).

Financial statement fraud is an act of fraud committed intentionally to provide information that misleads users of financial statements because it contains errors and manipulations (ACFE 2020). This intentional negligence is material so that it can affect the decisions that will be made by interested parties. The motivation behind financial statement fraud is to maintain share prices so that investors feel their investments are secure. Another factor underlying financial statement fraud is the need to support bond and stock offerings in the capital market (Zimbelman 2014).

There have been only a few studies investigating financial statement fraud during the COVID-19 pandemic. Existing studies on the crisis or pandemic focus mostly on their impact on a company’s financial performance (El-Chaarani et al. 2022; Erkens et al. 2012; Gupta et al. 2013; Kowalewski 2016), or their impact on the quality of financial statements (Arthur et al. 2015; Cimini 2015; Filip and Raffournier 2014; Hsu and Yang 2022; Persakis and Iatridis 2015; Trombetta and Imperatore 2014; Xiao and Xi 2021). Moreover, studies related to financial statement fraud during a crisis or pandemic that have been conducted only examine factors in fraud theory such as pressure, opportunity, rationalization, and capability (Putra 2022; Wulanditya et al. 2022). Other factors that investigate financial statement fraud during a crisis or pandemic, such as corporate governance, have not been investigated so far.

Strengthening governance mechanisms during a crisis or pandemic is critical to mitigating risks. Corporate governance can reduce conflicts of interest because it reduces opportunistic attitudes and can inhibit fraud in a company’s financial statements (Razali and Arshad 2014). However, no research has been discovered that investigates the effectiveness of corporate governance in preventing financial statement fraud during a crisis or pandemic. Several studies conducted instead of the crisis or pandemic indicate that the existence of corporate governance can mitigate for the occurrence of financial statement fraud (Razali and Arshad 2014; Girau et al. 2019; Mulyadianto et al. 2020).

The originality of this study lies in how it investigates the existence of corporate governance mechanisms aimed at preventing financial statement fraud during the COVID-19 pandemic. In this study, financial statement fraud investigations employ two models: the Z-score and the F-score. The Altman z-score model is a strong indication of financial distress, which can trigger financial statement fraud (Aviantara 2023; Kukreja et al. 2020). Meanwhile, the Dechow F-score model is a strong indication of financial statement fraud (Aviantara 2023). The Altman Z-score and Dechow F-score models detect financial distress and fraud with different indicators of financial ratios. Combining these two models can thus provide a more comprehensive estimate of a company’s financial health and fraud risk (Kukreja et al. 2020; Saleh et al. 2021).

This study has both academic and practical implications. The first implication of this research is the way it complements the limited existing studies on corporate governance mitigation to detect financial statement fraud during the COVID-19 pandemic. The second implication is that the research contributes to the body of knowledge about the effectiveness of the supervisory function in corporate governance frameworks, particularly in crisis situations. The third implication is to provide guidelines for stakeholders to assess the hazards that may develop as a result of misleading financial statements, particularly during a crisis. The fourth implication is that the study’s findings can be used by regulatory bodies in Indonesia to review corporate governance regulations, particularly those associated with preventing financial statement fraud during a crisis or pandemic.

State of the Art

Agency theory explains the relationship between principals (shareholders) and agents (management), who have different interests (Jensen and Meckling 1976). The difference in information between agents and principals can lead to a condition known as information asymmetry, which can lead management to commit fraud. Agency theory suggests that conflicts of interest and information asymmetry can be reduced by appropriate monitoring mechanisms that align the interests of different parts of the company. Monitoring mechanisms in accordance with the objectives of agency theory can be implemented by using corporate governance mechanisms.

Corporate governance is a method to resolve conflicts of interest between principals and agents through the disclosure of financial information. Furthermore, corporate governance is an important practice to reduce the information asymmetry that exists in stock market transactions and can prevent opportunistic actions by insider investors. Financial reporting fraud can be effectively reduced by corporate governance structures. Each structure has a distinct role to play in strengthening governance in order to prevent financial statement fraud, earnings manipulation, and the likelihood of bankruptcy (Martins and Júnior 2020).

The boards play an important role in the structure of corporate governance by supervising to ensure the success of the organization. In the context of financial information, the boards are responsible for the transparency and credibility of financial statements because they have the highest level of control in a company (Alzoubi and Selamat 2012). A large board will effectively promote the supervisory function with overarching control, gathering numerous managers’ viewpoints and experiences (Fathi 2013). Large boards are correlated with outstanding performance on company reputation (Orozco et al. 2018) and minimized the likelihood of financial statement fraud (Kalbuana et al. 2022). However, several other research findings show that there is no effect of board size on financial statement fraud (Nguyen et al. 2022; Shan et al. 2013; Salleh and Othman 2016). Hence, the first hypothesis in this study is:

Hypothesis 1 (H1).

Board size affects financial statement fraud.

Hypothesis 1a (H1a).

Board size affects financial statement fraud measured by Z-score.

Hypothesis 1b (H1b).

Board size affects financial statement fraud measured by F-score.

Boards with international experience have valuable, rare, and inimitable characteristic features that can contribute to the company’s competitive advantage by using their experiences. International experience of the board members can be gained through international obligations in foreign companies that are accustomed to monitoring activities in organizations by foreign companies. The practice is likely to be influenced by the culture, rules, laws, and regulations in the country where the company operates. Such experience can assist board members in managing the complexities associated with earnings management practices. At the same time, with international experience being different from local experience, it is also believed that board members will assist in promoting and implementing more proactive earnings management prevention mechanisms within the organization (Razali and Arshad 2014). The inclusion of international board experience on supervisory boards may result in an improved financial reporting quality (Dobija and Puławska 2022). Several studies have discovered that board experience has an impact on financial statement fraud (Alzoubi and Selamat 2012; Mousavi et al. 2022). Hence, the second hypothesis in this research is:

Hypothesis 2 (H2).

International board experience affects financial statement fraud.

Hypothesis 2a (H2a).

International board experience affects financial statement fraud measured by Z-score.

Hypothesis 2b (H2b).

International board experience affects financial statement fraud measured by F-score.

The audit committee is an important element of the corporate governance structure because it reviews the independence and integrity of the company’s financial statements. A strong audit committee can encourage better and more effective assessment and monitoring to inhibit financial statement fraud (Razali and Arshad 2014). One of the valuable audit committee characteristics for effective monitoring is financial expertise. Audit committees with finance competence are related with lower levels of earnings management, which also reduces financial statement fraud (Badolato et al. 2014). Several studies found that the financial expertise of the audit committee improves monitoring capabilities, which in turn improves the quality of financial reporting (Alzoubi and Selamat 2012; Mousavi et al. 2022; Subair et al. 2020). Furthermore, audit committee members that are financially literate have a greater ability to detect and prevent fraudulent financial reporting (Kamarudin et al. 2014). However, the findings of the research conducted by Razali and Arshad (2014) show that international board experience has no effect on financial statement fraud. Based on this argument, the third hypothesis in this study is:

Hypothesis 3 (H3).

Audit committee financial expertise affects financial statement fraud.

Hypothesis 3a (H3a).

Audit committee financial expertise affects financial statement fraud measured by Z-score.

Hypothesis 3b (H3b).

Audit committee financial expertise affects financial statement fraud measured by F-score.

The internal audit function is one of the strongest mechanisms for monitoring and promoting a good governance system in an organization. Internal audit plays an essential role in reviewing control system activities, offering input for improvement, and supervising activities (Putra et al. 2022). An effective internal audit function will assist management to generate high-quality financial statements (Arum and Wahyudi 2020). Several studies have found that the internal audit function can prevent financial statement fraud (Abdullah et al. 2018; Jarah et al. 2022); (Onoja and Usman 2015); (Petraşcu and Tieanu 2014). One of the determinants of the effectiveness of the internal audit function is competence and sufficient training (Arens et al. 2020). Thus, the fourth hypothesis in this study is:

Hypothesis 4 (H4).

Internal auditor competence affects financial statement fraud.

Hypothesis 4a (H4a).

Internal auditor competence affects financial statement fraud measured by Z-score.

Hypothesis 4b (H4b).

Internal auditor competence affects financial statement fraud measured by F-score.

2. Materials and Methods

The type of research in this study is quantitative, using secondary data analyzed by the logistic regression method. Secondary data is obtained from the annual report of the company that is the subject of the study. Logistic regression analysis was applied using the SPSS 25 program.

The population in this study is made up of companies on the special monitoring list on the Indonesia Stock Exchange (IDX). The sampling technique used is total sampling, which means that all companies on the monitoring list on the IDX are sampled in this study if the variables studied are available. Observations were carried out for the company’s annual report data for 2020 and 2021. The research sample data amounted to 134. According to the IDX Director’s Decree on Regulation Number II-S on Trading Equity Securities under Special Monitoring, there are 11 criteria in assessing the stocks under special monitoring, which are:

• The average share price over the last 6 months in the regular market is less than IDR 51 per share.
• The last audited financial report received a disclaimer opinion.
• companies that have no revenue or no change in revenue in the audited financial statements and/or the latest interim financial statements compared to the previously submitted financial statements.
• For mineral and coal mining issuers that have carried out the production operation stage but have not yet reached the sales stage or that have not started the production operation stage at the end of the 4th fiscal year since being listed on the Exchange and have not obtained revenue from the main business activities (core business); b. for issuers that are holding companies that have controlled companies engaged in the mineral and coal sectors that have carried out the production operation stage but have not yet reached the sales stage or that have not yet started the production operation stage at the end of the 4th fiscal year since they were listed on the Exchange, has not obtained revenue from core business activities.
• Companies that have negative equity in the last financial report.
• Companies that do not meet the requirements set out in Regulation Number I-A concerning the Listing of Shares and Equity Securities Other Than Shares Issued by Listed Companies, for Listed Companies whose shares are listed on the Main Board or the Development Board; b. companies that fail to comply with the requirements to remain listed on the Exchange as stipulated in Regulation Number I-V concerning Special Provisions for the Listing of Shares and Equity Securities Other Than Shares on the Accelerated Board Issued by Listed Companies, for Listed Companies that have shares that are listed on the Accelerated Board.
• Companies with low liquidity based on a daily average share transaction value of less than IDR 5 million and a daily average share transaction volume of less than 10,000 shares on the main market during the last 6 months.
• The company has either petitioned for postponement of debt payment obligations or filed for bankruptcy.
• Companies with subsidiaries whose revenue contribution to the listed firm is material, and the subsidiaries are in the process of requesting the postponement of debt payment obligations or filing for bankruptcy.
• Companies that are subject to a temporary suspension of securities trading for more than 1 trading day caused by trading activities.
• Other conditions determined by the IDX after obtaining approval or orders from the Financial Services Authority (OJK).

Financial statement fraud in this study is measured by the Altman Z-score model (Altman 2000) and the Dechow F-score model (Dechow et al. 2011). The Z-score model is a proxy for bankruptcy risk, which can be an early warning sign for a potential collapse that will lead to fraud or manipulation in an organization. A Z-score of less than 1.81 is an indicator that the organization is in the “distress” zone; a score between 1.81 and 2.99 indicates that the organization is in the “gray” zone; and scores greater than 2.99 are an indicator that the organization is in the “safe” zone (Altman 2000). Meanwhile, the Dechow F-score model is optimized to estimate the likelihood of manipulation rather than bankruptcy (Dechow et al. 2011). If F-Score > 1, the risk of financial statement fraud is high, and if the F-Score > 1.85, the risk of financial statement fraud is low or normal. Z-score and F-score model equations are described in

Table 1. Measurement of Z-score.

Variables	Measurement	Scale
Financial Statement Fraud using Z-score (FSFZ)	$\begin{array}{c}Z-score=\left(1.2X1\right)+\left(1.4X2\right)+\\ \left(3.3X3\right)+\left(0.6X4\right)+\left(0.99X5\right)\end{array}$	Nominal
X1	$\frac{Working Capital}{Total Assets}$	Ratio
X2	$\frac{Retained Earnings}{Total Assets}$	Ratio
X3	$\frac{EBIT}{Total Assets}$	Ratio
X4	$\frac{Market Capitalization}{Total Liabilities}$	Ratio
X5	$\frac{Sales}{Total Assets}$	Ratio

Source: Altman (2000); Razali and Arshad (2014).

and

Table 2. Measurement of F-score.

Variables	Measurement	Scale
Financial Statement Fraud using F-score (FSFF)	$\begin{array}{c}F-score=-7.893+0.790RSST+\\ 2.518REC+1.191INV+1.979SOFTASSETS+\\ 0.171CASHSALES-0.932ROA+1.029ISSUE\end{array}$	Nominal
RSST	$\frac{WC+NCO+FIN}{Average Total Assets}$ WC = (Current Assets − Cash and Short-term Investments) − (Current Liabilities − Debt in Current Liabilities); NCO = (Total Assets − Current Assets − Investments and Advances) − (Total Liabilities − Current Liabilities − Long-term Debt); Fin = (Short-term Investments + Long-term Investments) − (Long-term Debt + Debt in Current Liabilities + Preferred Stock)	Ratio
REC	$\frac{Accounts Receivables}{Average Total Assets}$	Ratio
INV	$\frac{Inventory}{Average Total Assets}$	Ratio
SOFTASSETS	$\frac{(Total Assets-PPE-Cash and cash equivalents)}{Total Assets}$	Ratio
CASHSALES	Percentage change in cash sales $(Sales-Accounts Receivables)$	Ratio
ROA	$\frac{(Earnings t/Average total assets t)}{(Earnings t-1/Average total assets t-1)}$	Ratio
ISSUE	If a firm issued securities during the year-t, it is worth 1, else it is worth 0.	Nominal

Source: Dechow et al. (2011); Aghghaleh et al. (2016).

.

The corporate governance structure, which is an independent variable, is measured by the board size (BS), board members’ international experience (IBE), audit committee financial expertise (ACE), and internal audit competence (IAC). In addition to the independent variables, this study also uses two control variables to improve the relationship between the independent and dependent variables. The control variables used are company size (SIZE) and leverage ratio (LEV). The measurements of each variable are shown in

Table 3. Measurement of Independent Variables.

Variables	Measurement	Scale
Board Size (BS)	Total number of commissioners on the board (Martins and Júnior 2020; Razali and Arshad 2014)	Ratio
International Board Experience (IBE)	The percentage of members of the commissioners’ board with international experience to the total number of commissioners’ board members (Razali and Arshad 2014)	Ratio
Audit Committee Financial Expertise (ACE)	The percentage of audit committee members with accounting and finance backgrounds from the total number of audit committee members (Alzoubi and Selamat 2012)	Ratio
Internal Auditor Competence (IAC)	The internal auditor’s education, training, and experiences (Arens et al. 2020; Arum and Wahyudi 2020)	Ratio
Company Size (LNSIZE)	Logarithm of market value of equity (Razali and Arshad 2014)	Ratio
Leverage (LEV)	$\frac{Total Debt}{Total Assets}$	Ratio

Source: developed for this research from several sources (2022).

.

Two models were developed in this study to examine the hypotheses. The first model investigates the effect of BS, IBE, ACE, IAC, LNSIZE, and LEV on FSFZ with the Z-score Model. The second model examines the effect of BS, IBE, ACE, IAC, LNSIZE, and LEV on FSFF with the F-score Model. The logistic regression equation model in this study is as follows:

$${Ln\frac{FSFZ}{1-FSFZ}}_{it}={{\beta }_0}_{it}+{\beta }_1{BS}_{it}+{\beta }_2{IBE}_{it}+{\beta }_3{ACE}_{it}+{\beta }_4{IAC}_{it}+{\beta }_5{LNSIZE}_{it}+{\beta }_6{LEV}_{it}+\epsilon$$

(1)

$${Ln\frac{FSFF}{1-FSFF}}_{it}={{\beta }_0}_{it}+{\beta }_1{BS}_{it}+{\beta }_2{IBE}_{it}+{\beta }_3{ACE}_{it}+{\beta }_4{IAC}_{it}+{\beta }_5{LNSIZE}_{it}+{\beta }_6{LEV}_{it}+\epsilon$$

(2)

where:

• FSFZ is financial statement fraud, which measured by the Z-score model
• FSFF is financial statement fraud, which measured by the F-score model
• BS is the board size
• IBE is the board members’ international experience
• ACE is the audit committee’s financial expertise
• IAC is the internal auditor competence
• LNSIZE is the company size
• LEV is leverage

3. Results

The data analysis method employed in this study is the logistic regression analysis, which has four tests: assessing the overall model, determining the goodness of fit of the regression model, the coefficient of determination (Nagelkerke’s R Square), and the correlation matrix (Ghozali 2018).

The outcome of the statistical test results in

Table 4. Descriptive Statistics.

Variables	N	Min	Max	Mean	Std. Deviation
FSFZ	134	0.00	1.00	0.7761	0.41841
FSFF	134	0.00	1.00	0.0746	0.26377
BS	134	2.00	7.00	3.2985	1.34349
IBE	134	0.00	1.00	0.3768	0.34057
ACE	134	0.00	1.00	0.5482	0.32622
IAC	134	0.33	1.00	0.5820	0.23828
LNSIZE	134	4.25	18.44	13.3891	2.96147
LEV	134	0.00	5.53	0.7425	0.96431

Source: descriptive statistical output of processed data (2023).

reveals that the total sample size is 134.

According to the descriptive statistical testing results in Table 3, the mean value of FSF has a Z-Score is 0.7761 and with an F-Score is 0.0746. The mean values of the independent variables are as follows: BS = 3.2985; IBE = 0.3768; ACE = 0.5482; IAC = 0.5820; LNSIZE = 13.3891; and LEV = 0.7425.

3.1. Financial Statement Fraud Analysis Using Z-Score

The overall model is reflected in the significant value of the omnibus test, which must be below 0.05. Based on the research results in

Table 5. Omnibus Tests of Model Coefficients.

		Chi-Square	df	Sig.
Step 1	Step	38.554	6	0.000
	Block	38.554	6	0.000
	Model	38.554	6	0.000

Source: SPSS output of processed data (2023).

, the omnibus test value of the independent variables produces a significance value of 0.00, which is lower than 0.05. This indicates that there is a significant effect of the independent variables simultaneously affecting the dependent variable. The goodness of fit of the logistic regression model is reflected in the Nagelkerke R Square value in

Table 6. Model Summary.

Step	−2 Log Likelihood	Cox and Snell R Square	Nagelkerke R Square
1	103.962 a	0.250	0.382

^a. Estimation terminated at iteration number 6 because parameter estimates changed by less than 0.001. Source: SPSS output of processed data (2023).

, which shows that the independent variable can explain 38.2 percent of the dependent variable, as seen from the Nagelkerke Square value of 0.382. Meanwhile, the other 61.8 percent can be explained by other factors outside the independent variables in the logistic regression equation.

The model feasibility test is shown in the Hosmer and Lemeshow Test, with the category is considered good if the significant value is more than 0.05. Based on

Table 7. Hosmer and Lemeshow Test.

Step	Chi-Square	df	Sig.
1	15.041	8	0.068

Source: SPSS output of processed data (2023).

, the Hosmer and Lemeshow value is 0.068, which is greater than 0.05, indicating that it is a good model.

Table 8. Variables in the Equation.

		B	S.E.	Wald	df	Sig.	Exp(B)	95% C.I.for EXP(B)
								Lower	Upper
Step 1 a	BS	−0.676	0.195	12.030	1	0.001	0.509	0.347	0.745
	IBE	−0.544	0.931	0.341	1	0.559	0.580	0.094	3.601
	ACE	1.088	0.822	1.752	1	0.186	2.968	0.593	14.864
	IAC	1.361	1.128	1.456	1	0.228	3.901	0.428	35.581
	LNSIZE	0.231	0.088	6.870	1	0.009	1.260	1.060	1.497
	LEV	1.640	0.599	7.501	1	0.006	5.155	1.594	16.667
	Constant	−1.365	1.293	1.114	1	0.291	0.255

^a. Variable(s) entered on step 1: BS, IBE, ACE, IAC, LNSIZE, LEV. Source: SPSS output of processed data (2023).

describes the logistic regression analysis used in this study, while

Table 9. Correlation Matrix.

		Constant	BS	IBE	ACE	IAC	LNSIZE	LEV
Step 1	Constant	1.000	−0.102	0.069	−0.064	−0.459	−0.764	−0.110
	BS	−0.102	1.000	−0.265	−0.069	−0.128	−0.328	−0.027
	IBE	0.069	−0.265	1.000	−0.348	−0.070	0.028	−0.242
	ACE	−0.064	−0.069	−0.348	1.000	−0.255	−0.035	0.085
	IAC	−0.459	−0.128	−0.070	−0.255	1.000	0.161	0.031
	LNSIZE	−0.764	−0.328	0.028	−0.035	0.161	1.000	−0.027
	LEV	−0.110	−0.027	−0.242	0.085	0.031	−0.027	1.000

Source: SPSS output of processed data (2023).

describes the correlations matrix. According to Table 8, only BS, LNSIZE, and LEV have a value less than 0.05, indicating that they have a significant effect on financial statement fraud as measured by the Z-score. While IBE, ACE, and IAC have greater than 0.05, which indicates that they have no effect on financial statement fraud as determined by the Z-score.

3.2. Financial Statement Fraud Analysis Using F-Score

The overall model is reflected in the omnibus test’s significant value, which must be less than 0.05. According to the findings in

Table 10. Omnibus Tests of Model Coefficients.

		Chi-Square	df	Sig.
Step 1	Step	17.986	6	0.020
	Block	17.986	6	0.020
	Model	17.986	6	0.020

Source: SPSS output of processed data (2023).

, the omnibus test value of the independent variables provides a significant value of 0.02, which is less than 0.05. This indicates that there is simultaneously a significant effect of the independent variables on the dependent variable. The logistic regression model’s goodness of fit is shown in

Table 11. Model Summary.

Step	−2 Log Likelihood	Cox and Snell R Square	Nagelkerke R Square
1	63,154 a	0.058	0.140

^a. Estimation terminated at iteration number 7 because parameter estimates changed by less than 0.001. Source: SPSS output of processed data (2023).

, where the Nagelkerke R Square value of 0.140 indicates that the independent variable can explain 14.0 percent of the dependent variable. Meanwhile, other factors than the independent variables in the logistic regression equation can explain the remaining 86.0 percent.

The Hosmer and Lemeshow Test incorporates a model feasibility test, with the category considered good if the significant value is greater than 0.05. According to

Table 12. Hosmer and Lemeshow Test.

Step	Chi-Square	df	Sig.
1	5.188	8	0.737

Source: SPSS output of processed data (2023).

, the Hosmer and Lemeshow value is 0.737, which is greater than 0.05, indicating that it is a good model.

The logistic model employed in this investigation is shown in

Table 13. Variables in the Equation.

		B	S.E.	Wald	df	Sig.	Exp(B)	95% C.I.for EXP(B)
								Lower	Upper
Step 1 a	BS	0.040	0.256	0.025	1	0.875	1.041	0.630	1.721
	IBE	−1.061	1.373	0.597	1	0.440	0.346	0.023	5.106
	ACE	0.230	1.110	0.043	1	0.836	1.259	0.143	11.080
	IAC	1.161	1.498	0.600	1	0.438	3.193	0.169	60.204
	LNSIZE	0.260	0.196	1.760	1	0.185	1.296	0.883	1.903
	LEV	−1.614	1.048	2.373	1	0.123	0.199	0.026	1.552
	Constant	−6.030	2.881	4.381	1	0.036	0.002

^a. Variable(s) entered on step 1: BS, IBE, ACE, IAC, LNSIZE, LEV. Source: SPSS output of processed data (2023).

, while the correlations matrix is described in

Table 14. Correlation Matrix.

		Constant	BS	IBE	ACE	IAC	LNSIZE	LEV
Step 1	Constant	1.000	−0.092	0.250	0.039	−0.448	−0.893	0.047
	BS	−0.092	1.000	−0.276	0.028	−0.011	−0.216	0.227
	IBE	0.250	−0.276	1.000	−0.206	−0.203	−0.155	−0.330
	ACE	0.039	0.028	−0.206	1.000	−0.324	−0.152	0.053
	IAC	−0.448	−0.011	−0.203	−0.324	1.000	0.232	0.017
	LNSIZE	−0.893	−0.216	−0.155	−0.152	0.232	1.000	−0.210
	LEV	0.047	0.227	−0.330	0.053	0.017	−0.210	1.000

Source: SPSS output of processed data (2023).

. Table 13 shows that all factors have a significance level greater than 0.05, indicating that they have no effect on financial statement fraud as measured by the F-score.

4. Discussion

4.1. Board Size and Financial Statement Fraud Analysis

According to the Z-score test results, board size has a significant negative effect on financial statement fraud. Thus, H1a is accepted. This result indicates that the bigger the number of board members, the lower the possibility of financial statement fraud. Therefore, it may be stated that the more people involved in active oversight, the lower the likelihood of fraud. This finding supports the agency theory, in which the supervisory board is required to provide assurance to the principal regarding the performance of the agent. This result is consistent with the findings of study completed by (Fathi 2013; Orozco et al. 2018; Kalbuana et al. 2022) but contradicts several research works that show that there is no effect of board size on financial statement fraud (Nguyen et al. 2022; Shan et al. 2013; Salleh and Othman 2016).

In contrast to the previous findings, F-score testing shows that board size has no effect on financial statement fraud. Thus, H1b is rejected. The research data shows that the sample that is indicated to have committed fraud with the F-score size is relatively small, only 7.46%, so it can affect the output of statistical analysis. The sample selection of companies under special monitoring may also affect the research results, where there are many criteria that do not merely refer to fraud committed by the company. Observing data during the COVID-19 pandemic also slightly affects the results of the study because unusual business conditions require board oversight that can only be conducted remotely. The results of this study confirm the findings of Nguyen et al. (2022); Shan et al. (2013); (Salleh and Othman 2016).

4.2. International Board Experience and Financial Statement Fraud Analysis

The logistic regression on Z-score indicates that there is no effect of international board experience on financial statement fraud. Thus, H2a is rejected. Logistic regression on the Z-score indicates that there is no effect of international board experience on financial statement fraud. Thus, H2a is rejected. This finding indicates that international board experience does not contribute to its expertise in overseeing corporate governance. The results of this study contradict the research findings of Alzoubi and Selamat (2012) and Mousavi et al. (2022).

In a similar manner to the prior result, F-score testing revealed that board international experience has no effect on financial statement fraud. H2b is rejected as well. This finding is consistent with the research findings of Razali and Arshad (2014) and Subair et al. (2020). The results of this study have not supported agency theory where the board as a corporate governance structure should have extensive insight to carry out a good monitoring role in order to provide reasonable assurance regarding the performance of company management as an agent.

Observations of sample data during the COVID-19 pandemic reveal that the international board experience does not appear to improve its roles and duties in enhancing overall corporate governance performance, particularly those connected to financial statement fraud prevention.

4.3. Audit Committee Financial Expertise and Financial Statement Fraud Analysis

According to the output of logistic regression testing, there is no evidence that shows the effect of audit committee financial expertise on financial statement fraud, whether measured by the Z-score or F-score. Therefore, H3a and H3b are rejected. These study results are in accordance with the findings of research conducted by Razali and Arshad (2014). But contrary to the findings of research conducted by Alzoubi and Selamat (2012); Badolato et al. (2014); Kamarudin et al. (2014); Mousavi et al. (2022); Subair et al. (2020).

Observations of sample data collected during the COVID-19 pandemic show that audit committee financial expertise fails to improve its function and responsibilities in improving overall corporate governance performance, including ones related to financial statement fraud prevention. In addition, based on data from the research sample, some companies lack an audit committee with expertise in finance. These companies seem not to pay much attention to financial matters in terms of determining the members of the audit committee. Whereas based on the Financial Services Authority (OJK) Regulation No. 55 of 2015, the main task of the Audit Committee is to review the company’s internal control system, ensure the quality of financial statements, and improve the effectiveness of the audit function (OJK 2015). Therefore, it is appropriate for companies to consider expertise in finance as a requirement for the audit committee.

4.4. Internal Audit Competence and Financial Statement Fraud Analysis

There is no evidence that reveals the effect of internal audit competence on financial statement fraud, whether measured by the Z-score or the F-score, according to the results of the logistic regression testing. As a result, H4a and H4b are rejected. The results of this study are inconsistent with the findings of several previous researchers who found that internal audit can prevent and reduce financial statement fraud (Abdullah et al. 2018; Jarah et al. 2022; Onoja and Usman 2015; Petraşcu and Tieanu 2014).

Sample data observed during the COVID-19 pandemic suggests that internal audit competence has not contributed to reducing financial statement fraud. Uncertain situations require quick adaptation to enhance performance. Therefore, in addition to being competent, internal audit should also place more emphasis on increasing its involvement in risk management, primarily in uncertain situations.

5. Conclusions

The COVID-19 pandemic affected various aspects of society and eventually resulted in a global recession due to weakened purchasing power. Many companies are affected by this crisis condition. The decline in activity during the pandemic also had an impact on the decline in the company’s financial performance. This condition required business entities to adapt to changes and become more aware of the risk of financial statement fraud. Although activities may not run normally during a crisis or pandemic, effective corporate governance is expected to mitigate the occurrence of financial statement fraud.

The purpose of this study was to examine the impact of corporate governance, as measured by the board size, the percentage of board members with international experience, the percentage of audit committee members with a financial expertise, and the competence of internal auditors toward financial statement fraud in companies on the IDX special monitoring list during the COVID-19 pandemic. The logistic regression equation includes control factors such as company size and leverage ratio. Since financial statement fraud is measured by using the Z-score and the F-score, two logistic regression models are examined in this study.

The results showed that using the Z-Score, the board size proved to have a significant negative effect on financial statement fraud during the COVID-19 pandemic. The larger the number of boards, the smaller the chance of financial statement fraud. However, it turns out that testing using F-Score shows that the corporate governance variables under investigation have no proven effect on financial statement fraud during the COVID-19 pandemic. The Z-score and F-score models are used to assess the influence of corporate governance on financial statement fraud during the COVID-19 pandemic by examining financial ratios and detecting potential financial distress or fraudulent activities. The Z-score model predicts the possibility of a company getting bankrupt using five financial ratios, whereas the F-score model predicts the likelihood of financial statement fraud using eight financial ratios. These models can help investors and regulators make better decisions by identifying companies which may be at risk of financial fraud or insolvency. However, these models have several shortcomings, such as their failure to capture all aspects that lead to financial fraud, their reliance on financial statements, and an inability to account for changes in corporate governance frameworks over time, particularly in crisis or pandemic situations.

This study makes various academic and practical contributions. Academically, the first implication of this study is the addition of an empirical study that explores the mitigation of corporate governance in the prevention of financial reporting fraud during the COVID-19 pandemic. The findings contribute to the growing body of empirical evidence addressing the mitigation of corporate governance in preventing financial reporting fraud during the COVID-19 pandemic. Furthermore, this study supports the agency’s notion that board size, which involves increased oversight of a larger number of boards, is an essential factor in reducing the prevalence of misleading financial reporting during crises. In addition, the findings of this study can aid in the development of concepts and theories that provide a better understanding of the relationship between corporate governance and fraud prevention in a crisis. The practical implication of this study is to provide guidance for stakeholders to analyse the risks that may occur as a result of fraudulent financial statements, particularly during crisis. These findings could help companies and regulators strengthen oversight mechanisms to prevent fraudulent financial statements. In this context, Indonesian regulators are expected to review and strengthen corporate governance guidelines, particularly during crisis, while keeping the stakeholders’ interests in mind.

This study suffers from several limitations. First, this study investigates the corporate governance structure only from a few indicators, including board size, board members’ international experience, audit committee financial expertise, and internal auditor competence. Second, the observation subjects only consist of 134 public listed companies which are categorized under special monitoring. Third, this research focuses solely on quantitative techniques for measuring the corporate governance aspects. Fourth, only the Altman Z-score and Dechow F-score models are utilized to detect financial statement fraud. Given these limitations, future research ought to investigate into other indicators of the corporate governance structure to acquire a comprehensive view of the corporate governance structure’s role in minimizing financial statement fraud during the crisis. The sample size can be expanded beyond the public companies under special monitoring. Future research might investigate a combination of both qualitative and quantitative methods. Finally, additional investigations with different financial statement fraud models, such as the Sprigate S-score, Grover G-score, and Beneish M-score models, can be incorporated into the research. Overall, the lack of impact of corporate governance on financial statement fraud suggests that gaps or limitations in the current corporate governance structure must be addressed. Additional study is required to uncover these weaknesses and create more effective corporate governance procedures to avoid financial fraud.