The influence of audit committee characteristics on impression management in chairman statement: Evidence from Malaysia

Abstract After the emergence of financial scams like Enron, WorldCom and Toshiba, the researchers have emphasized the significance of the audit committee (AC) in curtailing the earnings management (EM). The role of the AC in financial reporting has received significant attention in accounting research because it offers information to shareholders for investment decision making. Researchers have examined whether AC characteristics mitigate managers’ EM practices. Whereas there is a scarcity of research that investigates the association among ACs and impression management (IM). The objective of this research is to examine the relationship between the AC characteristics and IM practices in the chairman statement of the Malaysian companies We used IM as the measure for financial reporting quality (FRQ). Secondary data is utilized that is obtained from the annual report and Thomson Reuters Database. The sample of this study comprises companies listed on the Main Market of Bursa Malaysia in 2017–2018. Interestingly, this research findings identify that AC independence has a significantly negative relationship with the level of IM measured on the basis of qualitative scores (IMSC1) and quantitative scores (IMSC2). The findings of this research supported the agency theory because it argues that strong internal governance monitoring mechanisms improve the FRQ and decrease the IM. AC meeting has a significant positive association with the level of IM based on qualitative scores (IMSC1). This result does not support the argument that increasing number of ACs meetings results in reduced IM and enhanced FRQ. Also, the AC meeting has no relationship with the level of IM in the quantitative scores (IMSC2), whereas IM and AC size has no association with AC financial experts. This result may not support the argument of the agency theory that AC size, meeting and financial experts significantly enhances the FRQ. Findings of the study may enable regulatory bodies and policymakers to devise policies and strategies to improve the credibility of financial statements in Malaysia. Future studies may consider the effect of other AC characteristics, like AC quality, AC members’ ownership and AC members’ remuneration on IM.


PUBLIC INTEREST STATEMENT
This paper examines the effect of audit committee characteristics on impression management in the Malaysian market. Audit committee characteristics is a unique feature of the Malaysian market and impression management is more pervasive. Inconsistent empirical results regarding the effect of audit committee control on financial reporting quality motivate conducting this research in Malaysia. The study uses secondary panel data that were hand collected from the annual report and Thomson Reuters Datastream. The results revealed that audit committee independence is negatively and significantly associated with IM, supporting the signalling and the agency theories. The findings of the current study may benefit regulators and policymakers as well as investors interesting in audit committee in Malaysia market.

Introduction
The published financial report is one of the primary ways of disclosure which is considered as a source of information for outside users (Botosan, 1997). However, the financial failure of several large firms, for example, Enron and WorldCom, among other firms in Asia, has led to the decline of investor's confidence in the integrity of corporate disclosure and management acts (Hasnan et al., 2013;Johari et al., 2009;Norwani et al., 2011). Previous research associated with FRQ has emphasized largely on earnings management (EM) (Merkl-Davies, 2007). As a reason that earnings management (EM) is considered as an important trait of financial reporting quality (Ghaleb et al., 2020). Nevertheless, businesses may also use a more hidden form of influencing outsiders' impressions regarding firm performance and prospects, particularly through manipulating the presentation and content of narrative annual report's sections to distort readers' perceptions of corporate achievements (Godfrey et al., 2003). The collapse of high-profile businesses has highlighted the need for an effective corporate governance system worldwide (Norwani et al., 2011). Corporate governance is forming a balance between socioeconomic, individual, and communal goals while encouraging the efficient use of resources, accountability, the use of power, and stewardship at the same time, aligning the interests of individuals, corporations, and society (Wakaisuka-Isingoma et al., 2016). Corporate governance can play a vital role in ensuring that firms report quality earnings (Asogwa et al., 2019). The primary attention has been placed by the main players such as corporate governance bodies on the board of directors (BOD), audit committee (AC), managers, and auditors (Norwani et al., 2011). The activities of the audit committee in a company is of exceptional importance for the good practice of corporate governance (Tušek, 2015).
With respect to Malaysia, researchers have examined that how corporate governance influence the earnings quality (EQ) and EM (Abdullah, 1999;Saleh et al., 2005Saleh et al., , 2007Hashim & Devi, 2008;Ismail et al., 2010;Johl et al., 2013). AC has a significant role to monitor FRQ and accountability (Zhang et al., 2007). However, the quality of AC is dependent on the effectivity of its members. As reported in previous studies, the existence of AC is important but not adequate to ensure strong monitoring and FRQ alone (Bajra & Cadez, 2018;Menon & Williams, 1994). Financial reporting process of the firm is monitored by the AC. Krishnamoorthy et al. (2002) state that the presence of AC enables the board to ensure high audit quality and monitoring the financial reporting process. Lin and Hwang (2010) believe that an effective corporate governance mechanism enables the management to utilise the business resources to protect the interests of shareholders. Furthermore, it is argued that the board of directors and the committees have the responsibility to control the management practices.
In fact, AC has several characteristics that may improve its monitoring functions and result in better financial performance. Academic researchers argue that individual measures of AC features (e.g., AC size, expert, meetings and independence) are not sufficient in examining the quality of AC as these characteristics complement each other; considering one and ignoring another will render the AC ineffective (Al-dhamari et al., 2018;Connelly et al., 2012). Saleh et al. (2007) conclude that AC independency alone could not deter accrual-based earnings management (AEM) in Malaysian companies unless the AC meets frequently. Bajra and Cadez (2018) claim that the existence of ACs is important but not sufficient alone to ensure high-level monitoring and FRQ. However, few researchers have focused on the vital role of the AC and its influence on the IM. Therefore, this research is aimed to examine the association among the AC characteristics and IM in the chairman statement in Malaysia. With respect to various aspects, this research has a significant contribution to the literature. Firstly, the agency and signalling theories are utilised to explain the association among CG mechanisms, namely, the AC, and IM particularly, in the Malaysian context. Secondly, this research is aimed to develop a better understanding the problems associated to impression management (IM), whereby in the context of corporate governance the studies have claimed that corporate governance is not much effective due to the power of the controlling shareholders (Abdul Rahman & Ali 2006;Abdullah & Nasir, 2004;Park & Shin, 2004). Lastly, these findings could enrich the literature on the impact of AC on the IM.

Impression management
The conflict of interest between management and shareholders exists in all organisation. The agency theory states that agents have more information concerning firms as compared to principals and that where the asymmetry information negatively impacts the principal's ability to successfully monitor whether their stakes are effectively fulfilled by the agents or vice versa. The development of agency problems is attributed to the agents hiding information or taking action for their self-serving interests (Mirza et al., 2019). This motivates principals to invest in monitoring and in providing incentives to agents in order to align their interest with principals. Besides agency theory, the signalling theory also provides a framework to understand how two parties (i.e. an agent and a principal) address information asymmetries in a contractual exchange or relationship (Spence, 1973). According to Connelly et al. (2011), signalling theory is suitable to describe the behaviour of two parties (organisations or individuals) when they are able to obtain various information. Strategic signalling refers to the actions conducted by the companies or individuals and which influence the behaviour of different types of stakeholders such as suppliers, competitors, investment analysts, customers, employees, investors and partners (Zmud et al., 2010). Signalling theory has been described in the management, finance, information systems, marketing, and accounting literature (Boulding & Kirmani, 1993;Certo, 2003;Connelly et al., 2011). Most researchers have used it to examine the characteristics of executives in signalling company quality to the financial markets (Cohen & Dean, 2005;Higgins & Gulati, 2006;Zhang & Wiersema, 2009). The signalling theory concentrates on the problems created by information asymmetry in the market. The signalling theory explains how this dissimilarity can be decreased with additional information signalling to others. Information asymmetry is the cornerstone for signalling which is a common phenomenon that can be applied in any market.
Managers may engage in IM actions to mislead users about firm performance. IM is an action taken by the firm to manage its positive image in the public eyes. Generally, executive managers use IM practices to hide the real economic and financial information of firms, which may influence the quality of financial reporting and thereafter mislead the shareholders. IM was defined as the field of study in social psychology on how individuals present themselves to others to be viewed positively (Hooghiemstra, 2000). According to Clatworthy and Jones (2006) and Merkl-Davies and Brennan (2007), IM is defined as trying to control and handle the impression that is transmitted to users of accounting information. It can be seen from many dimensions, even as an attempt by rational management to the influence decision-making process for the investment by the users of financial statements or as a social phenomenon linked to self-presentation and responsibility (Merkl-Davies et al., 2011).
Based on agency and signalling theories, the conflict of interests between principals and agents called for the need to establish governance regulations to monitor the actions of agents in the companies (Ismail et al., 2010). Yue-Duan. et al. (2007) found that to protect the rights of investors and improve the transparency of information, regulators and information intermediaries have made great efforts to strengthen the corporate governance, thereby reducing the emergence of negative selection and agency problems due to information inconsistency. Past researchers have comprehensively investigated how do CG mechanisms improve the FRQ. However, the empirical results have examined the relation between CG mechanisms and IM (Abrahamson & Park, 1994;Garcia-Osma & Guillamon-Saorin, 2011;Ingram & Frazier, 1983;Lewellen et al., 1996;Mather & Ramsay, 2007). Ingram and Frazier (1983) assert that discretionary disclosure depends on the conflict of interest among the managers and shareholders and the degree of managerial control on the firm. Lewellen et al. (1996) focus their attention on performance comparison of manipulation techniques. These authors' findings highlight the bias in the benchmark industry and peer group stock returns reported, and how the magnitude of this bias is lower for highly concentrated and management-owned firms. Abrahamson and Park (1994) examine how negative outcomes are concealed through the narratives in the president's letters. The results show difficulties in interpretation: ownership concentration does not influence concealment. On the contrary, outside directors and large institutional investors appear to be effective in limiting concealment, whereas institutional investors who have little shareholding and outside directors who are also shareholders serve to encourage concealment. Based on these results, the authors recommend more studies to clarify the role of ownership structure further.
IM is considered as a concept which has EM under it (Clatworthy & Jones, 2003). Davidson et al. (2005) posit that attempting to manage impressions may lead to excessive EM. Thus, IM constitutes the foundation for EM, or, in different words, EM constitutes one manifestation of IM. This means that the classic definition of EM by Healy and Wahlen (1999) can be applied to IM. The conflict of interest arises when management uses their decision-making powers to manipulate the financial statements to misguide users about the financial position of the company.
With respect to various proxies of financial reporting, previous researchers have focused largely on EM (Merkl-Davies, 2007). Nevertheless, businesses may also use a more hidden form of influencing outsiders' impressions of firm performance and prospects, namely by means of manipulating the presentation and content of narrative annual report sections with the purpose of distorting readers' perceptions of corporate achievements (Godfrey et al., 2003). Based on the above, there are studies that examine the relationship between corporate governance such as AC characteristics and IM. However, Mather and Ramsay (2007) and Garcia-Osma and Guillamon-Saorin (2011) investigate the role of a strong corporate governance structure in reducing opportunistic disclosure practices. Mather and Ramsay (2007) focus on selective strategies in annual report disclosure in the year a firm changes its CEO. Similarly, Garcia-Osma and Guillamon-Saorin (2011) investigated narratives in the press releases of Spanish firms and found that strong structures of governance reduce the impact of detection practices that are compatible with IM.

Audit committee independence
AC Independence has been recognised as one of the main factors that improve AC effectiveness (Ghafran & O'Sullivan, 2013;Klein, 2002). Malaysian Code on Corporate Governance Malaysian (MCCG) 2017 attempted to reinforce the role of ACs by recommending that an AC should comprise completely of non-executive directors (NEDs). Bursa Malaysia Listing Requirements (BMLRs, 2018) stated that all members of the AC must be NEDs, and the majority of members must be independent directors.
For the AC to be effective, it should comprise of NEDs who are not related to and are independent of management (Ika & Ghazali, 2012). Several empirical studies have supported the argument that independent members improve AC's effectiveness in overseeing financial statements. Among them, Sharma and Kuang (2014) found that AC independence leads to reduced EM. Meanwhile, Garcia et al. (2012) reported that the association among EM and the independence of AC is insignificant. Also, the study by Katmon and Al Farooque (2017) found the insignificant relationship between DA and independence of AC.
In the Malaysian context, Yunos (2011) provided evidence that a significant relationship exists between independent members serving on the AC and discretionary accruals. Salleh and Che Haat (2014) examine the impact of AC independence on EM for pre and post MCCG 2007. They found that AC independence in the post-revised MCCG 2007 period was more effective in limiting EM compared to pre-revised MCCG 2007 period. Recently, Al-Rassas and Kamardin (2016) found that independent AC had a significant and direct association with accruals earning management (AEM). However, Haji-Abdullah and Wan-Hussin (2015) found no significant relationship among the independence of AC and real EM.
The AC is essential to enhance the financial reporting, whilst managers engage in IM to mislead users of the financial reporting by giving an uncertainty of information. It is expected that a high percentage of independent AC members can enhance FRQ, which in turn, mitigate the IM practices in the financial report. Based on the agency theory, it is claimed that AC with independent directors can improve corporate governance because they can find solutions to disagreement among internal managers (Fama & Jensen, 1983). This will help in alleviating some of the agency issues that aid to decrease FRQ. With respect to agency theory and other empirical researches, the present study assumes that independent directors on AC enhance the ability of the committee to perform its functions better and have a negative influence on the IM practices.
Based on the above discussion, there is a scarcity of literature that examines the association among AC independence and IM. Prior studies highlight the importance of AC independence as being part of the governing bodies to ensure the effectiveness of the supervisory function of the committee and the credibility of the reporting process. AC independence can challenge management's action by giving independent verdicts to enhance the process of internal control, thus reducing information inconsistency between external users and managers. Thus, based on the arguments above, this study develops the following hypotheses: H1: There is a negative association between AC independence and IM.

Audit committee size
The size of the AC has been addressed in several corporate governance guidelines published in the Cadbury Report (Cadbury, 1992) and Smith Report (2003). These reports show that AC should not be less than three members. Likewise, the Sarbanes-Oxley Act (2002) concludes that there should be at least 3 members in the AC. In the context of these reports, it is suggested that in accordance with the firm size and sector (Buchalter & Yokomoto, 2003). MCCG (2017) stipulated that it is necessary for the effectiveness of the AC that it must have 3 independent members. BMLRs (2018) states that an AC must comprise of a minimum of three members, all of them should be nonexecutive directors, and the majority of the members must be independent directors of the BOD.
According to Ismail et al. (2010), who reported that the AC size has an important impact on its decisions, and AC with the lower number of directors has better coordination as a part of CG. Bédard and Gendron (2010) conducted a meta-analysis to highlight the role that AC size plays on the effectiveness of the committee and found that most studies supported the major role of the size of the AC on effective monitoring.
Previous studies have found an association among the size of the AC and earning management actions. Ghosh et al. (2010) found that discretionary accruals are significant in firms with small ACs, suggesting that an AC with a large number of members possessed sufficient skills and knowledge and is more effective in the financial reporting monitoring. Salihi and Jibril (2015) documented that an AC with a large size had a significant connection with the magnitude of EM. However, Sun et al. (2014) have not provided sufficient proof to support the role of the size of the AC in constraining real earning management.
With reference to studies in Malaysia, Ismail et al. (2010) argued that contended that the proportion of AC members are associated with the level of EQ. Besides, Mansor et al. (2013) identified the association among corporate governance and earnings management in the familyoriented and non-family businesses in Malaysia. Contrarily, Salleh and Che Haat (2014) concluded that there is no significant association among the AC size and EQ. Additionally, Haji-Abdullah and Wan-Hussin (2015) documented an insignificant association between the size of the AC and real EM. Based on the above discussion, there is a lack of studies that examine the relationship between AC independent and IM.
The notion is that with many AC directors, more diverse skills and knowledge can be utilised by the committee to improve its monitoring role, thus decreasing the likelihood of experiencing IM behaviour. This is consistent with the agency theory which states that large boards could include more independence members, which in turn, lead to better control of management and an increase in the quality of financial reporting. This study expects that AC with bigger size can contribute more towards FRQ and decrease in IM. Thus, the following hypotheses tested: H2: There is a negative association among AC size and IM.

Audit committee meetings
The frequency of meetings is one of the most important attributes of AC (Beasley, 1996;McMullen & Raghunandan, 1996;Menon & Williams, 2004;Song & Windram, 2004). Bedard and Johnstone (2004) argued that the activity level of the committee meetings is measured by two main features: (i) the responsibilities it has to perform and (ii) the number of meetings.
Prior studies reported the significance of meeting frequency. Among them, Xie et al. (2003) in the US, Garcia et al. (2012) in Spain, and Gonzalez and Garcia-Meca (2014) in Latin America reported that frequency of AC meetings is associated with lower EM. On the other hand, Katmon and Al Farooque (2017) found AC meetings are related to high DA. However, studies by Davidson et al. (2005) and Baxter and Cotter (2009) in Australia, and Soliman and Ragab (2014) in Egypt reported that there is no significant relationship among AC meetings and EQ. Also, Shawtari et al. (2015), Abdul Rahman and Ali (2006) Chandrasegaram et al. (2013) provided evidence that there is a weak relationship among the frequency of AC meetings and earnings manipulation as measured by DA. Haji-Abdullah and Wan-Hussin (2015) found an insignificant association between the number of AC meetings and real EM. Based on the above discussion, it can be observed that only a few researchers have examined the association among AC meetings and IM.
AC with more meetings will consume additional time to take on effectively overseeing the process of the financial reporting of the business. The frequency of meetings signals the efficiency of the oversight committee's role and the credibility of the information provided. The AC meeting's frequency indicates that the committee intends to remain cautious and informed. This shows that outside users perceive less risk to make an investment in companies that organize more AC meetings and may require fewer practices of IM in corporate reporting.
Similarly, the frequency of AC meetings can help to send a signal for external users, because high frequency refers to increased company disclosure's credibility. Therefore, based on the argument of agency theory and the signalling theory, this study develops the following hypotheses: H3: There is a negative association among the frequency of AC meetings and IM.

Audit committee financial expert
Financial expert of AC members is a vital characteristic and all the AC members should pertain the reasonable knowledge of accounts and finance to effectively control the financial reporting process for the improvement of FRQ. The MCCG (2017) has a requirement that members of the AC should have a good understanding of finance and among them, one must have experience in accounting firms. Bursa Malaysia (2013) also adopts the same condition in its listing requirements (BMLRs). Badolato et al. (2013) examine the effect of interactions between an AC member with both financial expertise and status of EM. A financial expert serving on the AC with clear authority supported by sufficient regulations may constrain the earnings manipulative actions of the firm's managers. Krishnan et al. (2011) examine whether an AC with legal expert members enhanced FRQ. Hassan and Ibrahim (2014) found that the financial literacy of AC is effective in limiting real EM actions. However, Sun et al. (2014) found an ineffective role of the financial expert of AC members in constraining real EM. Garcia-Sanchez et al. (2017) explore whether financial experts on an AC improved EQ and found an effective role for the financial experts on an AC in enhancing EQ.
These studies propose that financial experts have a good understanding of how financial reports are manipulated, therefore they may be able to mitigate the level of IM and enhance the FRQ. Consistent with agency theory, it is suggested that financially knowledgeable AC members who possess financial and accounting qualifications are more likely to emphasize on the FRQ and reduce IM practices. Previous studies indicate that financially experienced members improve the effectiveness of the AC. Based on the theory of signalling, the above discussion indicates that AC members with an expert in finance and accounting are a good way of sending a signal about the credibility of the board. Hence, the following hypotheses is proposed: H4: There is a negative association among AC financial expert and IM.

Research methodology
This study utilised secondary data and it was obtained from the financial statements of the firm registered in the capital markets of Malaysia. The sample of this study comprises non-financial firms registered in the capital market of Bursa Malaysia for the years 2017-2018. The sampling data was obtained by utilizing the random sampling method. The sample firms were selected from the 11 different sectors of Bursa Malaysia and sample data represents all the sectors (Ahmad, 2016;Salleh, 2009). There are 934 companies registered on the Bursa Malaysia (total number of companies registered in Bursa Malaysia as at the end of December 2018). The proposed sample size was 274 companies for this study based on random sampling methods (Krejcie & Morgan, 1970;Sekaran & Bougie, 2016). We excluded 140 companies because these companies do not have the chairman statement in their annual report, and we chose the 134 that have the chairman statement. So, the total observations are 268 firm from the year 2017 and 2018 of the company listed on the Bursa Malaysia. As the aim of this research is to identify the association among IM and AC characteristics in the chairman statement. The dependent variable is IM which was collected from the chairman statement. While, for independent variables data (independence, financial expert, frequency of meetings and size of AC) were collected from corporate governance and directors' profile. The control variables items (return on assets, sales growth, Market Capitalization) were gathered from Thomson Reuters DataStream.
The following model examines the relationship between the AC characteristics and IM. The model is adapted based on past studies (Brennan et al., 2009;Garcia-Osma & Guillamon-Saorin, 2011).
Following Brennan et al. (2009), IM was analysed in quantitative and qualitative information. Three techniques (emphasis, the tone of disclosure and comparisons of performance) were used to examine IM. In order to calculate the composite quantitative score, it was based on (i) amounts repetition (negative and positive), (ii) amounts (negative and positive), (iii) comparisons of performance (negative and positive) and (iv) information locations (least, next-most and most emphasized part of the chairman statement section). Likewise, the composite of the qualitative score was calculated by coding: (i) statements repetition (negative and positive), (ii) statements and keywords (negative and positive), (iii) keywords promotion (reinforcement) (negative and positive) and (iv) information locations (least, next-most and most emphasized part in the Chairman's statement section).
The quantitative and qualitative composite results are weighted to capture all the detection practices described above. The weights are subjective, and the application system of these weights are devised by Brennan et al. (2009) and experimentally tested Garcia-Osma and Guillamon-Saorin (2011). The IMSC was calculated as the sum of composite point (scores) for all the quantitative/qualitative positive amounts or words minus the sum of composite point for all the quantitative/qualitative negatives amounts, divided by the sum of the compound points for all quantitative/qualitative numbers or words and amounts which is in line with (Brennan et al., 2009;Garcia-Osma & Guillamon-Saorin, 2011). So, the IM measurement was divided into qualitative scores (IMSC1) and quantitative scores (IMSC2). (2006), Klein (2002), Saleh et al. (2007), and Xie et al. (2003), the AC independence (ACIND) examined by the percentile of the number of independent non-executive members divided by the total number of AC members. AC size (ACSIZE) was the total number of AC members (Ismail et al., 2010;Salihi & Jibril, 2015;Xie et al., 2003). Following prior literature by Sun et al. (2014), andAl-Rassas andKamardin (2015), AC financial expert (ACFE) is defined as the ratio of the AC financial and accounting acquaintance to the total number of AC members. However, the AC meetings (ACMEET) calculated by the frequency of AC meetings per annum (Goh, 2009;Saleh et al., 2007;Xie et al., 2003;Yusof, 2010).

Descriptive analysis
The descriptive results of IM, four characteristics of the ACCs and firm attributes are presented in Table 1. The average IMSC1 for industrial and services firms in Malaysia is .543, while, the average IMSC2 is .339. The highest value of IMSC1 is 1 and the lowest value is −1. Meanwhile, the maximum value of IMSC2 is also 1 and the minimum value is −1.
For AC independence, Table 1 shows that 90.4% as the mean of the non-executive directors in the AC. The maximum and minimum values range from 0.5 and 1. Moreover, the AC size ranges from three to six. On average, the companies appointed three individuals as a mean number of the AC members. The maximum number for meetings is 18, a minimum of 4 meetings and an average of 5 AC meetings in an accounting period. The mean of the AC financial expert is .421 as shown in Table 1, with minimum members of 0 and a maximum of 1. However, the Code of Corporate Governance (2012) in Malaysia stipulates that listed companies should include in the AC at least one member with accounting certification or financial expert.
In relation to the control variables, Table 1 demonstrates that the average ROA is 2.282, while, the lowest value is −104.43 and the highest value is 35.05. The average sales growth is 23.873, with the minimum value of −87.53 and the maximum value of 2172.05. However, the mean of the market capitalization, as measured by natural log, is 12.63 with lowest and highest of 8.763 and 18.275 respectively. Table 2 displays the coefficients of the Pearson correlation between the variables. The multicollinearity test was conducted using Pearson correlation to measure correlation among independent variables (Gujarati & Porter, 2009). The result shows that the values of all correlation coefficients are below 0.90. There are no multi-collinearity problems for this study.

Regression analysis and hypotheses testing
As evidenced in Table 3, the regression analysis reveals that the R 2 for the model is 19.0%; which indicates that the independent variables demonstrate 19.0% of the variance in IM as qualitative composite score (IMSC1). Moreover, the model is significant (p < 0.000), indicating that the model significantly explains the difference in IM among Malaysian listed firms. Findings of Table 3 further reveal a negative significant association among AC independence and IM for Malaysian listed companies (t-value = −1.56, p-value = 0.0605). Based on this finding, H1 is supported. This   empirical result concurs with the previous study by Azzoz et al. (2016) and Alzoubi (2016). This finding is similar to the agency theory proposition, suggesting that the autonomous members of the AC might decrease the practices of IM. The implication is that the autonomy of the members of the AC and their activities are being separated from management influence. It can assist in preventing the issue of manipulation of financial statements and reporting of such statements in companies. Therefore, the company will be able to have efficient and effective internal control to offer sound financial reporting. The results suggest that independent members on the AC are effective in mitigating the IM level. Companies may want to send a good signal about the quality of the financial statement by appointing a higher number of ACI. With reference to Table 3, the outcome of the statistical result indicates that there is an insignificant positive relationship among IM and AC size (t-value = 0.52; p-value = 0.302). Thus, H2 is not supported. This result indicates that a large number of directors serving on in AC in Malaysian companies does not contribute to enhancing the monitoring role of AC to reduce the IM level.
This finding indicates that IM is not influenced or affected by the AC size for listed firms in Malaysia. The study's finding contradicts with the findings by Vafeas (2005) that found AC size will not be able to perform its functions excellently and efficiently. The finding of this study is supporting the results of Hamdan et al. (2013) which similarly identified insignificant relationship among AC size and discretionary accruals (DA) measured by the modified Jones model. Table 3, there is significantly positive association among AC meeting and IM (t-value = 1.60; p-value = 0.0555), so H3 is not supported. Research results support the findings of past researches conducted by Al-Rassas and Kamardin (2015) and Habbash et al. (2012) that AC meetings and EM are positively associated (measured by DAs) in the capital markets of Malaysia and UK respectively. This unexpected result could be due to the ineffectiveness of meetings, as it led to routines that make members become uncritical and consequently perform only a ceremonial function (Habbash et al., 2012;Spira, 1999). Table 3 shows that AC financial expert has insignificant relationship with IM (t = −0.39; p-value = 0.35), so H4 is not supported. The AC members are equipped with accounting and financial knowledge, and their accounting skills makes it very difficult for financial statements to be manipulated. This is because the monitoring power of the AC is strong. It is in line with Al-Rassas and Kamardin (2016) study in Malaysia and concluded that there is no significant association among ACFE and EM. Also consistent with Sun et al. (2014) that find an insignificant relationship between ACFE and real EM in the US.

The regression result in
For control variables, the findings in Table 3 reveal a positive significant relationship between IM and return on assets, sales growth and market capitalization (t-value = 2.32; p-value = 0.0105), (t-value = 4.52; p-value = 0.000) and (t-value = 4.16; p-value = 0.000) respectively. This implies that the return on assets, sales growth and market capitalization have an influence on the ability to manipulate financial reporting. Table 4 displays the regression analysis with R2 for the model is of 21.9%; which indicates that the independent variables demonstrate 21.9% of the variance in IM. Moreover, the model is significant (p < 0.000), indicating that the model significantly explains the difference in IM among Malaysian listed firms. Results reported in Table 4 show that the independent AC has a negative significant association with IM (t-value = −1.96, p-value = 0.0255), which support the H1 that predicted a negative association among them. These results suggest that independent members on the AC are effective in mitigating the IM level. These findings are supporting recent studies that report a significant association among independent AC and IM (Alzoubi, 2016;Azzoz et al., 2016). In turn, the independent of the AC could signal to shareholders that the IM has decreased. This result indicates that increased AC independent leads to an increased level of IM and decreased quality of financial reporting. The hypotheses related to AC size predicts insignificantly negative association with and IM as (t-value = 0.51; p-value = 0.3055), so H2 is not supported. These findings are in agreement with Haji-Abdullah and Wan-Hussin (2015), Al- Rassas and Kamardin (2016), and Saleh et al. (2007) who showed an insignificant relationship of ACSZ with EQ (measured by different proxies) with respect to Malaysia. These findings indicate that a large number of directors serving on in AC in Malaysian companies does not contribute to enhancing the monitoring role of AC to reduce the IM level. Table 4 show that the frequency of AC meetings is insignificantly associated with IM (t-value = −0.12; p-value = 0.451). Thus, the result fails to support the H3, it is supporting this research results of Abdul Rahman and Ali (2006), Davidson et al. (2005), Baxter and Cotter (2009), and Soliman and Ragab (2014) who suggested that there is no association among the frequency of AC meetings and EM (measured by discretionary accruals).

Results reported in
Hypotheses 4 predicts that the percentage of AC members with the experience in finance and accounting has negatively associated with the level of IM. However, results in Table 4 show that there is no significant positive association among ACFE and IM ((t = 0.69; p-value = 0.2465), which means ACFE failed to curb IM. Thus, H4 is not supported. The finding is similar to the previous Malaysian empirical study conducted by Al-Rassas and Kamardin (2016) which reported an insignificant relationship between ACFE and EM. Findings of this research support the study results of Sun et al. (2014) who find no significant association among ACFE and REM in the US. However, the result is inconsistent with Haji-Abdullah and Wan-Hussin (2015) findings that report a significantly negative relationship among ACFE and REM in Malaysia. It is also contrary to Yusof's (2010) findings which suggest that an AC with a higher proportion of financial experts would lead to credible financial reporting. The result evidences the premise that the existence of financial experts on the AC has a key role in enhancing the monitoring role of the committee and in lessening the practice of IM.
Results related to control variables reported in Table 4 reveal a significantly positive association among return on assets, sales growth and market capitalization and IM (t-value = 3.25; p-value = 0.0005), (t-value = 5.57; p-value = 0.000) and (t-value = 2.99; p-value = 0.0015) respectively. This implies that the control variables have an effect on the ability to manipulate financial reporting.

Conclusion
Influence of AC characteristics (size, independence, meetings and expert) on IM is investigated in this study. A sample size of 268 firm-year observations was used in this research. The observations were from the year 2017 and 2018 of the company listed on the Bursa Malaysia. The key purpose of the study was to identify the association among IM and characteristics of AC in the chairman statement. This relationship is examined in the context of Malaysia. The agency theory and signalling theory are utilized as the main theories of the research. Findings of this research have revealed that AC meeting has a significantly negative association with the level of IM in the qualitative scores (IMSC1) and no relationship with the level of IM in the quantitative scores (IMSC2). While independent AC has a negatively significant negative association with IM (in both). As per the regression analysis results there exists no relationship between IM and AC size and AC financial expert.
Findings of the current study are important for policymakers and regulators for policy and strategy formulation in terms of the FRQ in the Malaysia context. The regulators must devise a strategy regarding IM practices in the financial statements of the firms. Therefore, Malaysian regulators or Bursa Malaysia should consider these results and provide AC members with information regarding the significance of their role in enhancing financial reporting quality. By controlling IM practices, it helps to increase the FRQ in the publicly listed companies. The findings will also help regulators to improve the managerial activities in these firms through the reinforcement of accountability transparency. The difference of the findings in the current study and previous studies might be due to sample size or period of the study (Zahid et al., 2019). The results are consistent for both measures, albeit slightly weaker for the quantitative model. The evidence is in line with prior work on impression management that suggests that qualitative data is vague and easily managed (Garcia-Osma & Guillamon-Saorin, 2011).