THE ASSOCIATIONS of BOARD OF DIRECTORS’ CHARACTERISTICS WITH MODIFIED AUDIT OPINION

This study aims to find a relationship between board of directors’ characteristics and the likelihood that a company receives a modified audit opinion (as a measure of the external reporting of companies) in Malaysia. To test our hypotheses, we use the pooled cross-sectional logistic regression analysis for 136 firm-year observations listed on Bursa Malaysia over the period 2009-2011.The evidence we uncover is consistent with the hypotheses that companies with large board size and greater financial expertise of the board of directors are less likely to receive a modified audit opinion. The results obtained in this study are consistent with the listing rules of the Malaysian-Corporate-Governance Code and the requirements of the Bursa Malaysia Corporate-Governance-Guide, which consider and the requirements of the Bursa Malaysia Corporate-Governance-Guide, which consider the significance of the board of directors as an aspect of good corporate governance and its critical role in the Malaysian financial reporting process.


INTRODUCTION
In recent times, a series of giant companies from around the world collapsed, including such firms as Enron and WorldCom in the US, Parmalat, Ahold, Gescartera and BBVA in Europe (to name a few). It has been shown that most corporate failures can be caused by the lack of internal monitoring of management in the financial reporting process, arising from poor corporate governance of companies (Agrawal & Chadha, 2005;Deng & Wang, 2006;Charitou et al., 2007;Che Haat et al., 2008;Lin & Hwang, 2010;Darus & Mohamad, 2011). Malaysia is not an exception, albeit the Malaysian Code on Corporate Governance (MCCG) and Bursa Malaysian corporate governance Guide strengthens, incrementally, the efforts towards enhancing corporate governance. There are still many questionable marks on the financial reporting quality since pervasive cases of fraudulent financial reporting and failures (e.g., Megan Media Hodlings Berhad, NasionCom Hodlings Berhad, United U-LI Corporation Berhad and Transmile Group Berhad) have emerged in the marketplace. Under such circumstances, several of these cases were given a modified audit opinion, as the result of lower financial reporting quality. Therefore, these cases of corporate financial reporting failures have raised questionable marks on the financial reporting quality, and the effectiveness of the board of directors, as one of vital internal corporate monitoring mechanisms on management in Malaysian setting. As such, this can enhance the financial reporting quality (Johari et al., 2008;Darus & Mohamad, 2011). Thereby, this study attempts to answer the following question: Does the board of directors' characteristics in individual and aggregated tests (i.e., independence, size, meetings and financial expertise) decrease the probability of Malaysian publicly listed companies in receiving a modified audit opinion?
There are several studies that have examined board of directors' characteristics and the possibility of a company receiving a modified audit opinion (Sanchez-Ballesta & Garcia-Meca, 2005;Firth et al., 2007;Rusmin et al., 2009;Farinha & Viana, 2009). These studies have uncovered mixed and inconclusive results, due to the omission of important determinants (Vafeas & Theodorou, 1998). Notably, it is well documented by the extant research in other disciplines that board of directors' financial expertise is considered one of the crucial factors in increasing the board of directors' effectiveness and monitoring the role that it plays. Consequently, a positive impact of the financial expertise of those sitting on the board may enhance reporting quality (Carcello et al., 2002;George, 2003;Lin et al., 2003;Park & Shin, 2004;Bedard et al., 2004;Cohen et al., 2008). In this regard, board members with financial expertise would be able to rectify potential financial problems before they become bigger. As such, the directors who have sufficient expertise in financial matters, greater understanding in accounting principles, and are able to ask the right questions and understand the answers, will have a positive influence on the quality of the financial reporting. Ultimately, the probability of obtaining an unmodified audit opinion is increased.
To the best of the researchers' knowledge, there are several studies that have examined the relationship between the board of directors' characteristics and the possibility of a company receiving a modified audit opinion. The Malaysian context is not an exceptional one, due to the fact that most Malaysian publicly listed companies that have received modified audit opinions, correspond with a lower financial reporting quality during our period of study. This circumstance raises a questionable mark on the role of the board of directors on the probability of Malaysian publicly listed companies receiving a modified audit opinion. A study investigating the association of the characteristics of the board of directors with a modified audit opinion in the Malaysian setting does not exist.
In this aspect, the current study is motivated by an effort to fill the gap left by the previous audit opinion literature (e.g., Sanchez-Ballesta & Garcia-Meca, 2005;Firth et al., 2007;Rusmin et al., 2009;Farinha & Viana, 2009). As to the best of the researchers' knowledge, previous research in the field of audit opinion ignored the influence of financial expertise of the board of directors on the probability of receiving a modified audit opinion. Therefore, this study seeks to extend the work of Farinha and Viana (2009) by empirically investigating how the effectiveness of the board of directors' influence on the probability of receiving a modified audit opinion. This is the ultimate measure of financial reporting quality in the economies of Asian countries, especially in the Malaysian context.
In the particular case of the Malaysian context there are few studies that have examined the relationship between the board of directors' characteristics and the modified audit opinion (e.g., Alkilani et al., 2019a;2019b). Therefore, a significant aspect of the present study is that of contributing to the understanding of the effect of key characteristics of the board of directors on a modified audit opinion in Malaysian publicly listed companies. As such, this should fill a gap in the existing literature of Asian economies, especially in the Malaysian context, by highlighting a specific dimension, namely the quality of financial reports. Additionally, the presence of a modified audit opinion can be observed as a direct measure of financial reporting quality. Using such a dimension of financial reporting quality, we document that, in accordance with our monitoring hypothesis, the financial expertise of the board of directors contributes towards improving the effectiveness of its monitoring role. Further, by enhancing the quality of financial reporting, we uncover that the board of directors with the greatest degree of financial expertise is significantly related to companies being able to avoid obtaining a modified audit opinion. Moreover, we also find that companies are less likely to receive a modified audit opinion when they increase the number of the board members. However, we fail to report significant evidence regarding the association of the independent non-executive directors and the frequency of board meetings with the modified audit opinion. Overall, our results confirm that the financial expertise of the board of directors, and the size of the board of directors contribute to increases of the 8 AD-MINISTER effectiveness of the board of directors in its monitoring role, related to enhancing the quality of the financial reporting which, consequently, leads to a decrease in the probability of receiving a modified audit opinion.
The data used in this study are drawn from the annual reports of a sample of non-financial Malaysian publicly listed companies which received a modified audit opinion from the period 2009 to 2011, as a test sample, with companies receiving unmodified audit opinions for the same period such as a matched-pair sample.
The results of this study should be of interest in providing for a basis for regulatory actions, namely those aiming to influence the structure of the board of directors.

Modified Audit Opinion and the Quality of Financial Reporting
The quality of financial reporting indicates the extent to which the financial information disseminated through the financial statements gives a true and fair view of the financial position, the results of its operations and its cash flows for the reporting period (Watkins et al., 2004). Arguably, it is the quality of financial reports that is critical to the users of financial reports in so much as they can make the right decisions related to investment and finance, as they base their decisions on such reports. However, the absence of such quality will eventually lead to the inefficient allocation of economic resources, and in turn, poor economic growth, as a result of incorrect decisions being made based on misleading financial statements (Gwilliam & Macve, 1982;Elliot, 1995;Wallman, 1995Wallman, , 1996Sridharan et al., 2002).
In this aspect, financial reporting quality can be judged through the type of audit opinion issued by the external auditor (Cohen et al., 2004). The auditor report is the final product of the audit examination that auditors use to communicate to users of the company's financial statements (Porter et al., 2003). Thereby, it has a major role to play in informing financial statement users of the impending problems with the company's financial reporting or controls of internal processes (Hope & Langli, 2010;Bennouri et al., 2015;Chen, 2020). The auditor's report may be classified into two types: unmodified and modified audit opinions (including qualified, disclaimed, and adverse opinions), as mentioned in the International Standards on Auditing (ISA, 700).
As a measure of financial reporting quality (Pucheta-Martínez & de Fuentes, 2007;Farinha & Viana, 2009), the type of audit opinion can enhance or decrease the reliability of financial information and the integrity of the financial reporting process (Watts & Zimmerman, 1986;Adiloğlu & Vuran, 2011). This is because the audit opinion is designed to report the detection of material misstatements in audited financial statements (Elfouzi & Zarai, 2009;Stanley et al., 2009;Adiloğlu & Vuran, 2011). Thereby, the audit opinion appears through agency theory in providing indicators about the company's quality of financial reports to users of the financial statements (e.g., Carcello & Neal, 2000;Hope & Langli, 2010). Specifically, the decision to receive or avoid a modified audit opinion is considered as primafacie evidence of the quality of the financial statements (Chen et al., 2001;Pucheta-Martínez & de Fuentes, 2007;Farinha & Viana, 2009). Whereby, "the modified audit report includes a remark related to the quality and presentation of financial statements" (Sundgren, 2009). Thus, it can be argued that the modified audit opinion can be interpreted as a signal of lower financial reporting quality (Farinha & Viana, 2009), and may indicate the presence of an agency problem, because the presence of a modification; a concern that is central to the user of the financial statements (Sanchez-Ballesta & Garcia-Meca, 2005).

Board of Directors Characteristics
The main corporate governance mechanism responsible for monitoring the management, is the board of directors (Fama & Jensen, 1983). As a result of several corporate failures, boards have become a subject of attention for regulators (Adams et al., 2010), for governance reform, and a center of concern for the attention of many researchers (Omer & Al-Qadasi, 2020;Puni & Anlesinya, 2020). The board of directors acts as one of the important internal control mechanisms used in increasing public confidence in the credibility and the objectivity of financial reports, and protecting shareholder rights (Almutairi & Quttainah, 2020;Almujamed & Alfraih, 2020) by enhancing the quality of monitoring process while decreasing the probability of receiving a modified audit opinion. The following part discusses the development of four hypotheses which meet the objectives of the study, and are related to the impact of the effectiveness of the board of directors on the probability of receiving a modified audit opinion.

Board Independent Non-Executive Directors
A common contractual means for the mitigating of agency problems and reducing its cost is the presence of an independent party to the board for monitoring the management and control of company activities (Fama, 1980;Fama & Jensen, 1983;Cohen et al., 2008;Zainal Abidin et al., 2009;Fernández-Temprano & Tejerina-Gaite, 2020). This is because boards are less likely to exert control over management when they lack independence from the management (Carpenter & Westphal, 2001). According to Farinha and Viana (2009) and Zhang et al. (2007), the presence of independent, non-executive directors on the board is required to enhance its effectiveness in overseeing and reviewing the financial reporting process, and to introduce better corporate governance practices. (Lagos-Cortés et al., 2017;Thenmozhi & Sasidharan, 2020). In line with these arguments, The Malaysian Code on Corporate Governance (MCCG, 2007) requirement mandates that one third of the board members must be independent non-executive directors. In tandem, Bursa Malaysia listing requirements stipulate that at least two directors or one third of the board, whichever is higher, must be independent (paragraph 15.02 listing requirements).
Empirically, previous studies on audit opinion provide some support for the notion that independent, non-executive directors will do a better job of oversight of financial reporting processes and thus reduce the probability of a company receiving a modified audit report. For example : Firth et al. (2007) uncover that the existence of independent non-executive directors of the board in a Chinese setting have a negative association with a modified audit opinion. Likewise, Farinha and Viana (2009) argue that a larger percentage of independent, non-executive directors of board members increases the quality of financial reporting, whereby, they uncover that the percentage of Portuguese, independent non-executive directors is a negative significantly related to a modified audit opinion. Conversely, Rusmin et al. (2009) report a negative but insignificant association between the independent, non-executive directors and a modified audit opinion in an Australian context. The possible explanation of this contradictory finding may be due to the differing sample design and the time period of the study test.
In Malaysia, there is only one study (Iskandar et al., 2011) that links the role of the board independent, non-executive directors and going concern problems, using data from 2004. This study reports a positive insignificant association. It does not seem sensible to develop the hypothesis of the present study dependent upon one study that has used somewhat old data because of the fact that after 2004, many changes have been put in place in the MCCG (MCCG, 2007and Revised MCCG, 2007 and, Bursa Malaysia listing requirements, have focused on showing greater clarity of the role of the board of directors in the monitoring processes, (among others). This study, however, uses recent data from 2009-2011 to examine the impact of the percentage of independent, non-executive directors on the board on the modified audit opinion. Therefore, this study develops its hypothesis based of the conjuncture of the agency theory and the direction of the studies that are consistent with the theory to add one more debate to the proposed association. Thus, the following hypothesis is proposed: H 1 : There is a negative association between board independent non-executive directors and the probability that a company receives a modified audit opinion, all else constant.

Board Size
The revised MCCG states that "every board should examine its size, with a view to determining the impact of the number upon its effectiveness" (revised MCCG, P, 12, 2007). In this regard, based on the conjuncture of the agency theory, Kiel and Nicholson (2003) argue that a larger board is more likely to be vigilant in addressing agency problems. This is because increased board size may increase the monitoring strength of the board of directors, due to greater oversight over the company's management (Pearce & Zahra, 1992;Borokhovich et al., 2005;Miller, 2009;Marlin & Geiger, 2011). In this regard, Al-Najjar (2011) suggests that a large board provides for better internal monitoring activities. Hence, board of directors' size is viewed as another important element in board characteristics that may affect board of directors' effectiveness in monitoring activities, oversighting financial reporting process and internal control, which in turn enhances the quality of financial reports (Bradbury et al., 2006;Abdul Rahman & Mohamed Ali, 2006;Zhang et al., 2007;Firth et al., 2007;Rusmin et al., 2009). It also reduces the probability of receiving a modified audit opinion (Sanchez-Ballesta & Garcia-Meca, 2005;Firth et al., 2007;Rusmin et al., 2009).
In the audit opinion context, empirical evidence on the impact of the board of directors' size on a modified audit opinion is conflicting and inconclusive. Rusmin et al. (2009) uncover a negative and significant association between board of directors' size and modified audit opinion. Implicit in this result is that a large number of members on the board enables the board to function effectively and leads to high quality of financial reporting, and consequently less likelihood of receiving a modified audit opinion. However, this result is in the opposite line with Farina and Viana (2009); Sanchez-Ballesta and Garcia-Meca (2005) and Firth et al. (2007) whereby they find that board directors size is unrelated to the probability that a company receives a modified audit opinion. One possible explanation for the conflicting results reported is the that listed companies in environments of Spanish, Chinese and Portuguese settings have high ownership concentration where a single shareholder often has effective control of the company (e.g., Sanchez-Ballesta & Garcia-Meca, 2005;Firth et al., 2007;Farinha & Viana, 2009), compared those with companies listed in an Australian setting (Rusmin et al., 2009). This may lead to the impact of the effectiveness of board of directors' size as a monitoring device on the corporate financial reporting process, and its capacity to monitor management.
Due to the above arguments and the limited evidence, this study attempts to examine the impact of the board size on the modified audit opinion in Malaysian context. Therefore, the study develops it hypothesis based on the suggestion of the agency theory and the direction of the studies that are consistent with the theory to add one more debate to the proposed relationship. Thus the following hypothesis is proposed: There is a negative association between board of directors' size and the probability that a company receives a modified audit opinion, all else constant.

Board Meetings
The board of directors' meetings is considered as a resource that leads to board diligence and in turn enhances board effectiveness (Conger et al., 1998;Vafeas, 1999;Xie et al., 2003;Garcı´a Lara et al., 2009). Overall, the intensity of board activities is likely to contribute to the effectiveness of its oversighting function particularly in matters concerning the financial reporting process, resulting in improved financial reporting quality (Xie et al., 2003;Yatim et al., 2006;Firth et al., 2007;Alshirah et al., 2020). In turn, there is a decrease in the probability of receiving a modified audit opinion (Farinha & Viana, 2009). Hence, the meetings of the board of directors 12 AD-MINISTER appears to have the capability to minimizes agency problems and reduce agency cost, as a result from lower information asymmetry (Foo & Mat Zain, 2010;Lorca et al., 2011). Consistent with these perspectives, the latest Guide on Corporate Governance by Bursa Malaysia highlights that a typical board of directors would hold a minimum of 6 to 8 board meetings annually.
Regarding the influences of the board of directors' meetings on a modified audit opinion, prior empirical studies are still small in number and provide an unclear result to support theoretical arguments. For example, Farinha and Viana (2009) report that the increase of the board of directors' meetings increases the quality of financial reporting and reduces the probability of receiving a modified audit opinion. However, this result is contrary with Firth et al. (2007) whereby they suggest that the relationship between board diligence (proxy by the board of directors' meetings) and the issuance of a modified audit opinion is not statistically significant. The potential justification of these paradoxical results might be due to the different corporate governance structure. Chinese listed companies have a twotier internal corporate governance structure made up of a supervisory board and a board of directors (e.g., Firth et al., 2007). The supervisory board oversees the board of directors and this governance structure differs from the Portuguese corporate governance structure (e.g., Farinha & Viana, 2009).
Thus, based on the standpoint of the agency theory, it can be argued that the increase of the board of directors' meetings will likely improve the monitoring process of corporate financial reporting and internal control; thereby, an increase in the probability of an unmodified audit opinion (Farinha & Viana, 2009). Therefore, the hypothesis is developed based on the proposition of the agency theory and the direction of the studies that are in line with the theory to add one more related debate. Thus the following hypothesis is proposed: H 3 : There is a negative association between board of directors' meetings and the probability that a company receives a modified audit opinion, all else constant. Cohen et al. (2008) suggest that the primary characteristic for a board member in the agency perspective is expertise in monitoring and control. In this regard, previous studies argue that board of directors' financial expertise has an important impact on its effectiveness in its monitoring role and how this, consequently, can influence the quality of financial reporting (Lin et al., 2003;George, 2003;Xie et al., 2003;Bedard et al., 2004;Park & Shin, 2004;Agrawal & Chadha, 2005). Therefore, the board of directors, which has more experts is more likely to discharge its oversight responsibility effectively in the internal control and financial reporting process, thus sustaining the quality of financial reporting, and eventually increasing the probability of receiving an unmodified audit opinion.

Board of Directors' Financial Expertise
To the best of the researchers' knowledge, there is no empirical study that has linked board of directors' financial expertise and modified audit opinion. Prior literature highlights the importance of addressing different board of directors' characteristics as the crucial factors in increasing board of directors' effectiveness and the monitoring role it plays (Xie et al., 2003;Lin et al., 2003;Agrawal & Chadha, 2005;Goh, 2009). It is therefore logical to think that the increase in expertise of the board of directors on financial matters leads to an increase its understanding of the accounting numbers. This enables them to ask the right questions and to be abler to understand the answers, thus increasing the board's ability to monitor management while enhancing internal control and financial report quality. This, consequently, reduces the probability of receiving a modified audit opinion, and thereby contributes to protecting the interest of all shareholders.
This study develops it hypothesis based on the suggestion of the agency theory and the direction of the studies that are consistent with the theory to add one more debate to the proposed association. Thus the following hypothesis is proposed: H 4 : There is a negative association between board of director expertise and the probability that a company receives a modified audit opinion, all else constant.

METHODOLOGY Data Collection and Sample Description
Data used in this study are collected from two separate sources: annual reports and Datastream. Any missing financial figures from Datastream were acquired from the annual reports. The annual reports were retrieved from the Bursa Malaysia website at www.bursamalaysia.com.my. All relevant data was collected from 2009 to 2011. The sample companies were extracted from the population of publicly-available information mainly annual reports of publicly listed companies on the Bursa Malaysia, which received a modified audit opinion in the period of 2009-2011, as a test sample.
Each firm year of the test sample is matched with a company which received an unmodified audit opinion as a control test (matched-pair sample) in the event year. In this study, the matched-pair sample is identified using the following criteria: fiscal year, industry, same market and nearest total asset amount. The matched-pair sample method has been used in prior studies of audit opinion such as Sanchez-Ballesta and Garcia-Meca (2005). In this study, the financial companies were excluded because they possess unique characteristics, and since they are not required to adopt some of the corporate governance requirements imposed on public companies, and most are regulated by different regulations and standards they would thus require a separate study, (Sanchez-Ballesta & Garcia-Meca, 2005;Chu & Cheah, 2006;Yatim et al., 2006;Farinha & Viana, 2009). Given the above, our sample is composed of 136 companyyears observations relating to 96 companies from 7 different industries sectors. This can be more pronounced as will show later.

Model Specification and Measurement
To examine the relationship between independent variables and modified audit opinion, we test our hypothesis by using the following cross-sectional logistic regression model which is proposed as: MAO = β 0 + β 1 BINEDR + β 2 BSIZE + β 3 BMEET + β 4 BEXPERT + β 5 AUDSIZE + β 6 logAUDLAG + β 7 LOSS + β 8 NCFGOAR + β 9 LogEAF + e Where: Modified audit opinion (MAO) will take the value of "1" if the opinion is a modified one, i.e., when a qualified, disclaimed, and adverse opinions has been issued by the auditor, and "0" otherwise.

Control variables
There are many factors that affect a modified audit opinion. In our model we have introduced several of these factors as controlling variables, among others, which have been reported in the literature, which can impact the probability that a company receives a modified audit opinion. Hence, this study seeks to analyze the relationship between a modified audit opinion and a comprehensive set of control variables. The discussions of other controlled variables are offered in the following: Most prior archival studies have documented that the audit firm size (AUDSIZE), audit report lag (AUDLAG), the losses (LOSS), the operating cash flow (NCFGOAR) and the statutory external audit fees (EAF) are significantly associated with modified audit opinion. Hence, these variables are included as control because this is uncovered to exert a significant influence on modified audit opinion. In this study, with regard to the audit firm size and the operating cash flow it is expected to have a negative association with a modified audit opinion (Pucheta-Martínez & de Fuentes, 2007;Firth, et al., 2007;Goh, 2009;Farinha & Viana, 2009;Rahmat et al., 2009;Kaplan & Williams, 2013;Tsipouridou & Spathis, 2014;Moalla & Baili 2019). In respect to the audit report lag (AUDLAG), the losses (LOSS), and the statutory external audit fees (EAF) are expected to have a positive relationship, with a modified audit opinion (Francis, 1984;Chen & Church, 1992;Bell et al., 2001;Behn et al., 2001;DeFond et al., 2002;Geiger & Rama, 2003;Pucheta-Martínez & de Fuentes, 2007;Firth et al., 2007;Johl et al., 2007;Basioudis et al., 2008;Farinha & Viana, 2009 . Therefore, the companies are expected to be received an unmodified audit opinion when the companies have high quality auditors, shorter audit report lag, strong financial health, appropriate net cash flow generated from operating activities, and suitable audit services fee paid by the company to its statutory external auditor on audit engagement.

Sample Descriptive
After the screening process, 68 company-years observations relating to 48 companies that have received a modified audit opinion were eligible to be included in the analysis as a test sample group. The test sample group companies were matched with the same number of companies that have received an unmodified audit opinion, resulting in a final sample of 136 company-years observations relating to 96 companies.

RESULTS AND DISCUSSIONS
Descriptive Statistics and Univariate Analyses Table 1: provides descriptive statistics and univariate test results for variables measured as continuous metrics. In this table, summary statistics for the total groups of sample, modified audit opinion (test sample) and unmodified audit opinion (the matched-pair sample) as a control test sample are tabled in separate columns.
Regarding the independent, non-executive directors on the board, the mean (median) of BINEDR is .464 (.43); with a minimum of .29 and a maximum of .80, suggesting that independent non-executive directors on the board are very much closer to half of the board of directors' members in total groups of the sample. This composition follows the suggestion made by the MCCG which requires at least one-third (1/3) of the board members to be independent non-executive. In audit opinion studies, these results are similar to the mean (median) of .36 (.40) reported in Portuguese for 2002-2005 by Farinha and Viana (2009). Comparing the mean (median) of BINEDR in modified audit opinion sample is .492 (.50); which is significantly higher when compared with the mean (median) .435 (.43) of an unmodified audit opinion (the matched-pair sample).
In contrast, we can observe that the mean (median) board of directors' size BSIZE is 6.856 (7) members; with a minimum of 3 and maximum of 13, in total groups of the sample. In comparing terms with audit opinion studies, Farinha and Viana (2009) uncover that the mean (median) of board of directors' size BSIZE in their study is 7.51 (7). In the comparative, the mean (median) of BSIZE in a modified audit opinion sample is 6.33 (6); which is significantly lower when compared with the mean (median) 7.38 (7) of an unmodified audit opinion (the matched-pair sample).
Concerning the frequency of meetings of the board of directors, we can observe that the mean (median) number of board of directors' meetings BMEET is 6.053 (5) (minimum of 3 and maximum of 25), in total groups of the sample. These values contrast to those obtained from Farinha and Viana (2009), as they uncover that the mean (median) number of boards of director meetings is 13.78 (12) with a minimum of 3 and maximum of 58 meetings. One can observe that the mean (median) of BMEET in a modified audit opinion sample is 7.27 (6); which is significantly higher when compared with the mean (median) 4.83 (5) of an unmodified audit opinion (the matched-pair sample).
In terms of the board of directors' financial expertise, we find that the mean (median) of the board of directors' financial expertise BEXPERT is .303 (.29) (minimum of 0 and maximum of .80), in the total groups of the sample. The mean (median) of BEXPERT in modified audit opinion sample is .304 (.29) which is quite similar when compared with the mean (median) of an unmodified audit opinion (the matched-pair sample) .302 (.29) and the total groups of samples .303 (.29). Comparing the BEXPERT in the two groups of audit opinion (modified audit opinion with that of unmodified audit opinion) it shows a statistically insignificant difference between the mean of the board of directors' financial expertise BEXPERT in the two groups (t = .096 p = .924). This preliminary result does not provide directional support for the association of the board of directors' financial expertise with a modified audit opinion. Table 2 presents the descriptive statistics and univariate test results for (dummy) variables by audit opinion. Like the previous t-test, the analysis in this table is related to the relationship between modified audit opinion and unmodified audit opinion in order to be consistent with the multivariate model. The results of the Mann-Whitney U-test for the distribution differences between modified audit opinion and unmodified audit opinion reveals that audit firm size AUDSIZE, is statistically insignificant. In this case, we can argue that these variables are statistically independent from the type of opinion being granted by the auditor (at the 10% level of significance). This result indicates that no evidence of significant differences exists between the two groups of audit opinion samples for audit firm size AUDSIZE, (modified audit opinion group: .43; an unmodified audit opinion group: .57).
Based on our observation, a majority of our sample was audited by the Non-BIG 4 audit firms (63%), which reported that the modified audit opinion group audited by (BIG 4: .32; Non-BIG 4: .68) and an unmodified audit opinion group audited by (BIG 4: .42;. This is inconsistent with previous studies such as Farinha and Viana (2009) who reported that the Big 4 auditors are responsible, directly or through subsidiaries, for about 70% of the audit reports in Portugal.
Meanwhile, the differences in loss in previous years LOSS samples was found to be highly significant at p < .000. This result is consistent with prior audit reporting research (Johl et al., 2007;Firth et al., 2007;Pucheta-Martínez & de Fuentes, 2007;Farinha & Viana, 2009), the modified audit opinion group were more likely to have a loss and more likely to have received a modified audit opinion than the unmodified audit opinion group. Regarding the existence of multicollinearity problems among the independent variables, several examinations have been identified to check for the possible existence of multicollinearity. These include: the correlation matrix, variance inflation factor (VIF) and tolerance (1/VIF).
In terms of the correlation matrix, in this study, the correlation matrixes confirm that no multicollinearity exists among the variables in the study model, the modified audit opinion model as none of the variables correlates equal or above 0.80, which is the first indication to absence of a substantial collinearity problem between the independent variables (Hair et al., 1995;Greene, 1999;Cooper & Schindler, 2003). Among independent variables, the highest correlation is equal to -0.501, that found for BSIZE and BINEDR in modified audit opinion; All the variables have a correlation that are below the critical limits of 0.80. As shown in Table 3. This provides for indications that multicollinearity is not a problem in the study model estimations. With regard to the (VIF) and the tolerance (1/VIF), in general, the existence of multicollinearity problems is detected if the variance inflation factors (VIF) is more than 10 and tolerance value is less than 0.10 (Hair et al., 2006;Pallant, 2007). Based on the analyses, variance inflation factors (VIF) and the tolerance (1/VIF) for the study model report in this study do not exceed (10) and (0.10), respectively. As shown in Table 4. This indicates that multicollinearity is not a problem to model the explanation in this study.

MULTIVARIATE LOGISTIC REGRESSION RESULTS
The results of multivariate logistic regression for testing the hypotheses are illustrated in Table 5. We uncover that the model correctly classifies approximately 86.4 of the observations. In this respect, the Hosmer-Lemeshow test statistics is greater than 0.05 for the study model (0.948) indicating that study model fit is acceptable. Moreover, the relationship between dependent and independent variables is significant (X2 = 106.40, p < 0.000). The pseudo X2 = (0.5815) implies a relatively strong association between the dependent and independent variables. In terms of the Cox and Snell R2 value for the study model is (0.553) indicating that the model fit. The Nagelkerke R2 values for the study model is (0.738). This is indicating that the independent variables in the model explain 73.8% of the change in the dependent variable, implying a reasonably explanatory model. The presence some of outliers in the dataset amongst independent variables may mislead the results significantly obtained through the logistic regression and thus must be considered for treatment (Menard, 2002). In this respect, there are several ways to identify outliers. In this study, to ascertain whether the results are driven by unusual sample observations, outliers were identified using the standardized residual values greater than +/− about 3.3 (Tabachnick & Fidell, 2013;Pallant, 2007). Accordingly, in this study, there is some outliers, only a few (two) of them. Thus, such observations had been excluded from the main analysis. Eliminating this observation together with its unmodified audit opinion matched group enhances the regression results as presented in the following Table 5.
This is demonstrated even though The Malaysian Code of Corporate Governance (2000,2007) emphasized the importance of independent, non-executive directors on the board to bring an independent judgment to the board and enhance its effectiveness (Ponnu & Karthigeyan, 2010). The results show that the relationship between the independent, non-executive directors and modified audit opinion is a negative, but not statistically significant even at the 10% level. The result implies that the Malaysian independent non-executive directors may not be enough to act as an effective monitoring mechanism (Darus & Mohamad, 2011), in order to avoid companies from obtaining a modified audit opinion. Hence, it can be inferred that hypothesis H1 is rejected. This result is consistent with the finding of Rusmin et al. (2009) who report a negative insignificant association between the independent nonexecutive directors and a modified audit opinion. Likewise, in a local context, Iskandar et al. (2011) suggest that the association between the Malaysian independent, nonexecutive directors and going concern problems is a positive insignificant.
A plausible explanation for this result, perhaps, is that these independent nonexecutive directors as a part of the board of directors, might have a lack of time and insufficient efforts to carry out their duties to monitor financial reporting quality. This is because the board members may be holding more directorships; consequently, they are associated with ineffective monitoring of management (e.g., Fich & Shivdasani 2006). Specifically, directors with multiple directorships may have limited dedicated time to one company, thus directors with multiple directorships become so busy, and may have less time to closely scrutinize the internal control system, which leads to less effectiveness in monitoring of management (Yatim, 2011;Kamardin et al., 2012).
Regarding the influence of the board of directors' size on the likelihood of receiving a modified audit opinion, our results show a significant negative relationship between the large directors on the board and a modified audit opinion at the 5% level. Implicit in this result suggests that a large number of members on the board of directors enable them to function effectively and provide better internal monitoring activities; and consequently, decrease the probability of Malaysian publicly listed companies in obtaining a modified audit opinion. We can therefore conclude that our evidence expected H2. This result is consistent with Rusmin et al. (2009) who find that the board of directors' size is a negative and is related to the probability that a company is receiving a modified audit opinion.
Contrary to the stated hypothesis, the study finds a statistically significant relationship at the 1% level, but with the positively associated direction between board of director meetings with a modified audit opinion. This result indicates that the frequent meetings of board of directors do not help reduce the likelihood of Malaysian publicly listed companies receiving a modified audit opinion. Hence, it can be inferred that hypothesis H3 is unsupported. This result is opposing with the suggestion of empirically indicated research for a modified audit opinion; for example,, Farinha and Viana (2009) report that the board of director meetings has a negative impact on a modified audit opinion.
Thus, a plausible interpretation from this result may be based on the fact that the increase in the board of directors meetings meant that holding additional meetings and spending more time with dealing with the company problems, discussing strategy and legal issues, does not mean spending more time towards discussing the financial reporting issues or provide for a better monitoring role for the financial reporting processes. In addition, as a result of a lower quality of financial reporting there is therefore an increase in the propensity of Malaysian publicly listed companies in receiving a modified audit opinion. Therefore, the company needs to ensure that board of directors' members raised and resolved issues with management during these meetings, and as a result improve the quality of reporting (Hashim & Abdul Rahman, 2011).
In terms of the board of directors' financial expertise, the results show that the board of directors' financial expertise has a negative statistically significant relationship at the 10% level with the probability that companies receive a modified audit opinion. This result indicates that the increase of financial expertise on the board of directors leads to an increase in its understanding of accounting matters, which enables them to contribute towards improving its effective monitoring role, and in turn, enhance the quality of financial reporting and thereby, increase the probability of Malaysian publicly listed companies receiving an unmodified audit opinion. This is consistent with the theoretical model and the stated hypothesis. Hence, it can be inferred that hypothesis H4 is supported.
In regard to control variables, as expected in the study model all the control variables are significant in the predicted directions except for audit firm size AUDSIZE and external audit fees LogEAF, which are insignificant with the modified audit opinion. More specifically, there is an insignificant positive relationship between audit firm size AUDSIZE and the modified audit opinion even at the 10% level. As unpredicted, this finding reveals that the audit firm size does not bring any sort improvement in the monitoring processes, towards enhancing the quality of financial reporting, and in turn impact the likelihood of Malaysian publicly listed companies receiving an unmodified audit opinion. This result is consistent with numerous empirical previous studies on modified audit opinion; for instance, Firth et al. (2007), Pucheta-Martínez andde Fuentes (2007) and Farinha and Viana (2009) which uncover a positive insignificant association between the audit firm size and the modified audit opinion. Likewise, in the local context, Malek and Che Ahmad (2011) find that the size of an audit firm is negatively but not significantly related to an unqualified audit opinion.
The same applies as well to the external audit fees, as unexpected; there is an insignificant positive relationship between external audit fees LogEAF and the modified audit opinion even at the 10% level. This result reveals that the external audit fees do not bring any sort improvement in the monitoring processes, towards enhancing the quality of financial reporting, and in turn impact the probability of Malaysian publicly listed companies receiving an unmodified audit opinion. This result reinforces the findings of prior local studies; for illustration, Johl et al. (2012) uncover that the qualified audit report is a negative unrelated to natural logarithm of audit fees. Likewise, Malek and Che Ahmad (2011) find that the external audit fee is a positively but insignificant related to an unqualified audit opinion.
Nevertheless, we uncover a positive and highly significant (at p < 0.01) association between both audit report lag logAUDLAG and the losses LOSS and the likelihood probability of receiving a modified audit opinion. Also, in accordance with expectations of the influence operating cash flow, our evidence finds that the percentage of net cash flow generated from operating activities NCFGOAR has a negative significantly impact on the probability of receiving a modified audit opinion (In the study Model: p < 0.01).

DISCUSSION AND CONCLUSION
We draw on the argument that board of directors' acts are one of the important internal control mechanisms used in increasing public confidence in the credibility and the objectivity of financial reports through its effectiveness in overseeing of the corporate financial reporting process, and contribute to the protection of the interests of all shareholders. This study seeks to investigate the association between the effect of the characteristics of the board of directors and the probability that a company receives a modified audit opinion, as a measure of the quality of companies external financial reporting. Using a logistic regression with a matched pair sample design, we analyze the data of the sample of 136 observations relating to 96 non-financial companies listed on Bursa Malaysia over the period 2009-2011. In this study several important findings emerge. Firstly, based on the study Model relating to an individual test for independent variables, our results provide evidence that Malaysian directors on the board with large size and greater expertise in financial matters are more effective in monitoring the financial accounting process, and ensuring the reliability of financial reports that, in turn, decrease the probability of Malaysian public listed companies receiving a modified audit opinion. In so doing, this contributes to protecting the interests of all shareholders. Therefore, the results indicate that board of directors' size and financial expertise are two of the most important determinants of a modified audit opinion, and reflect good internal control practices in a Malaysian setting.
However, the insignificant relationship between the percentage of the independent, non-executive directors and the probability of Malaysian publicly listed companies obtaining a modified audit opinion indicates that the independence of the Malaysian directors may not be sufficient to protect the interest of all shareholders. This is because of the fact that the independent, non-executive directors as a part of the board of directors, might have a lack of time and insufficient efforts to carry out their duties to monitor financial reporting quality, due to the board members being busier. Arguably, directors' busyness reduces their effectiveness in their monitoring duties as they relate to executive duties and the monitoring of financial reporting quality (Song & Windram, 2004;Fich & Shivdasani, 2006;Chang, 2009;Yatim, 2011;Kamardin et al., 2012).
Moreover, while it is expected that board of directors' meetings will be able to reduce the likelihood that a company receives a modified audit opinion as documented in Farinha and Viana (2009), this study uncovers a controversial and positive direction between the frequency of meetings of board of directors with the modified audit opinion. As such, this may reflect some concerns that more frequent meetings of board of directors as an internal corporate governance mechanism may not be able to necessarily provide a better monitoring role to the financial reporting processes in companies. Thus, the company needs to ensure board of directors' members raise and resolve issues with management during such meetings, and as a result, improve the quality of reporting (Hashim & Abdul Rahman, 2011).
The results of this study provide comprehensive evidence to support the agency theory. In this aspect, it can be argued that the significantly negative associations reported between that the board of directors' financial expertise and board of directors' size with a modified audit opinion are proportionate to the suggestion of the agency theory. Because of a significant negative association between the size and financial expertise of the board of directors with a modified audit opinion, it indicates that an increase in the effectiveness of the board of directors enhances the quality of companies external financial reporting. In turn, this mitigates the agency problems between the financial statement users and safeguards the shareholders' interests by decreasing the prospect of Malaysian public listed companies receiving a modified audit opinion.
The results of this study could be valuable to management, shareholders, auditors, creditors and investors who are concerned about improving the practices of corporate governance, and the quality of financial reporting in the Malaysian market. This is because they have a better understanding of how the board of directors' characteristics impact the modified audit opinion.
Likewise, as with other studies, this study is subject to a number of limitations. Foremost, the results obtained in this study based on the data collected from annual reports of the sample of 136 observations relating to 96 non-financial companies listed on Bursa Malaysia, which only covers a three-year period from 2009 until 20011. Thereby, our study limits the ability to generalize the results across companies (i.e., in other sectors), years (a longer year data period would show the trend on a modified audit opinion) and other countries (due to a different structure of ownership and regulations). In addition, providing a strong statistical power in explaining the test results depends on the sample size in which a small sample size may fail to provide such power. Secondly, we focused on a few aspects of the board of directors' characteristics. Clearly there may be other characteristics of the board of directors that could affect the results, such as the legal expertise, ethnicity (indigenous vs non-indigenous) and business knowledge (industry expertise), which may have an influence on a modified audit opinion. Thirdly, there might be other variables that impact the probability of receiving a modified audit opinion that are not included in this study, such as the characteristics of the audit committee and the ownership structure. Future studies possibly will address these issues. Additionally, currently, as a result of the prevailing global conditions resulting from the pandemic (Covid-19), which has had a major impact on the global economy, extending this study by using data on companies in these exceptional circumstances in future studies is important to determine if there are any differences in the results.
Despite these limitations, the results uncovered in this study provides some preliminary evidence that a modified audit opinion is associated with the financial expertise of the board of directors and the size of the board of directors. This study confirms that the board of directors' characteristics are considered as a good internal monitoring mechanism in mitigating the agency conflicts, and in protecting the interests of shareholders among Malaysian companies. In doing so, the results of this study provide strong evidence to support the current debate on the effectiveness of the board of directors' characteristics as one of an internal corporate governance mechanism in the context of Malaysia, and constitute a further contribution to modified audit opinion studies while narrowing the gap in respect to the accountancy literature.