CEO Duality, Ownership, and Readability of Financial Statement Footnotes: Some Evidence from Indonesia

This study examines the association between CEO duality, CEO ownership and financial footnotes readability. The data were collected from 1,469 non-financial firms listed on the Indonesian Stock Exchange (IDX) between 2010 and 2018. Using ordinary least squares (OLS) regression, this study unveiled a negative and significant relationship between CEO duality and financial statement footnotes. In addition, the relationship between CEO ownership and financial statement footnotes is positive and significant. Furthermore, this indicates that CEO duality expands the embarrassment and provokes opportunistic behaviour from the CEO to prioritize personal interests by exploiting potential company resources to weaken the independence of the CEO. Hence, the readability of financial footnotes is undecipherable. On the other hand, CEO ownership is more responsible for improving corporate performance, so more financial statement footnotes are readable. Therefore


Introduction
The accounting scandals in corporations that have occurred recently have caused considerable losses to investors. The losses of those scandals are not only on the material side, but also the company's business continuity on a prolonged basis.
For example, Lehman Brothers (2008) caused a loss of 50 million dollars as a hidden loan asset that lowered the balance of income and expense accounts over the past few periods (Sorensen & Miller, 2017). On the contrary, in Indonesia, recently there has been an accounting scandal at PT. Garuda Indonesia Tbk has significant losses for its investors due to unpredictable revenues that have caused a scandal of fraudulent company financial statements (Aviantara, 2021). Many companies carry out governance reforms to determine what caused the scandal (Amiram et al., 2018). The scandal is inseparable from the role of the company's executives, in particular the company's CEO. In the Indonesian context, the CEO is analogous to the president director of the company (Harymawan et al., 2019). Also, the Chief Executive Officer (CEO) holds the top position in the management of the company and has responded to the company's operations and performance (Fu et al., 2017). Companies hire CEOs based on their expertise, experience, and ability to create value for shareholders, so the CEO's performance and responsibilities are critical to the company.
In the Indonesian setting, the position of a CEO is regulated in the financial services authority (OJK) regulation NO 33 / POJK.04 / 2014. This regulation states that a member of the board of directors/president director is allowed a maximum of 1 (one) issuer or other public company. However, this regulation does not explicitly regulate how the dual position of the CEO of a public company and its subsidiaries is linked, so this is an opportunity for opportunistic company CEOs to get more incentives. A number of the study state that CEO duality can reduce the effectiveness of monitoring managerial decisions and actions, thus allowing the CEO to make decisions that can increase his wealth instead of firm performance or value (Seifzadeh et al., 2021;Abernathy et al., 2019;Salehi et al., 2020). Ammari (2021) argues that CEO duality can reduce board independence and effectiveness and increase CEO power in making corporate decisions, thereby increasing agency conflict within the company (Antounian et al., 2021;Ginesti et al., 2017). According to several studies, CEO duality in corporate decision-making tends to be (Pavlopoulos et al., 2019;Al-ahdal et al., 2020). Tahir et al. (2020) also stated that CEO duality has less effective control in making effective decisions, indicating weak management performance. Ginesti et al. (2018), when managers have weak performance, tend to complicate the information they present annually, which negatively impacts the stock market because they respond to complex information. One of the critical information in the annual report is the financial statement footnotes. Seifzadeh et al. (2021) stated that corporate financial reports have always been among the most important information sources for decision-making, such as capital market practitioners (shareholders, creditors, and financial analysts), capital market legislators, and other stakeholders. For readers of financial statements, notes to financial statements are necessary because they help investors value companies in the stock market (Salehi et al., 2020). Financial statement footnotes are an integrated part of the company's financial statements. The increasing complexity of information in the financial statement footnotes makes readers have difficulty analyzing and processing existing information, which can cause readers or investors to find it challenging to determine what decision to choose (De Franco et al., 2020;Yen et al., 2021;Liu, 2021).
The readability of financial statement footnotes is critical because it communicates the numbers in the financial statements to every reader. With CEO duality positions in a subsidiary, it allows the CEO to have much room to make decisions that are not in line with shareholders' interests and can mislead the submission of information in a particular company's annual report in the financial statement footnotes. CEO's duality makes readability in the financial statement footnotes challenging to understand. However, Kim et al. (2021) argued that CEOs concurrently held by a subsidiary could reduce internal conflicts and inconsistencies in decision making. Likewise, if the multiple positions' role can utilize the CEO's knowledge, expertise, and information to increase the board's overall effectiveness, Kim et al. (2021), disclose information in financial statement footnotes that is understandable to readers.
Besides, companies often try to reduce conflicts of interest by giving their shares rights to their management, especially the CEO. A director who owns most of the company's shares bears the consequences and reaps the benefits of destructive and value-creating managerial actions; thus, CEO ownership can reduce agency costs (Kim et al., 2021). In addition, CEO ownership aligns with agents' and shareholders' interests, reducing the need for shareholder monitoring, thereby improving disclosure quality, especially the disclosure's readability. Bernini et al. (2021) argue that the need for more excellent monitoring and more transparent disclosure decreases with direct ownership of the company. Hence, director ownership is a corporate governance mechanism that acts as a substitute for exposure. CEO ownership expects to disclose information in the financial statement footnotes that the reader understands.
We found that CEOs who concurrently held subsidiaries' positions had readability of financial statement footnotes challenging to understand. Where CEO duality, indirectly, much power is placed in one person's hands, which will create agency problems (Melón-Izco et al., 2021;Carrillo et al., 2019), which can trigger opportunistic behaviour of boards, especially CEOs, to make a profit. Personal and at the expense of the interests of shareholders. García-Sánchez et al. (2022) also mention that CEOs with multiple positions tend to complicate their disclosure and be dishonest when evaluating their performance, thus leading to longterm organizational deviations. Also, CEOs who have shares in the companies they lead can reduce agency problems in the company so that the CEO acts in the shareholders' interests. CEO ownership makes more responsibility for their decisions because CEOs will receive the same consequences as other shareholders for their actions and decisions (Khatib et al., 2021) and can avoid negligence in a CEO (Moreno-Ureba & Bravo-Urquiza, 2019). So, CEO ownership will positively impact company performance and make the financial statement footnotes easy to understand. The remainder of this paper is organized as follows. The following section reviews relevant literature and develops research hypotheses. Section 3 presents our data and research design. Section 4 reports and discusses our results; before Section 5 concludes.

Literature review and hypothesis development
Readability is a person's ability to understand the context of sentences thrown through a writing instrument. Readability is initially a necessary process for a person to interpret the context of reading and then transform it into information for himself (Abshari & Rahman, 2020;Ajina et al., 2016). Gunning (1952) first coined the readability theory, which explained the ability to transfer information from obtained readings (Hassan et al., 2019). The readability of financial statement records is a process of understanding the context of financial statements aimed at digging for information on the company's segmentation and financial performance. According to Chall (1958), readability is the ability of a particular individual or subject to understand the context of reading influenced by a combination of the person's internal factors such as interests, willpower, and abilities. Some studies' readability of financial statement records indicates inconsistencies in results influenced by various characteristics of top management (Zhou, 2019;Uppal, 2020;Nguyen, 2020), CEO duality and CEO ownership as a dominant character.
In Indonesia, the CEO is a company organ that occupies the top position in company management and has responsibility for its operational activities. To create value for shareholders, companies usually will select and appoint CEOs based on their experience, expertise, and abilities (Harymawan et al., 2019). Managing and directing the company to achieve its goals is the CEO's full responsibility. The CEO also must plan and analyze the company's functional performance; compile company policies and procedures, and gather and determine the company's strategic decisions. Based on a regulation on Indonesia Financial Services Authority NO 75 /POJK.04/2017, a CEO is responsible for all the company's preparation and financial reporting, so disclosing information in financial reports is very important for stakeholders. The information might represent all of the activities and risks in the business process.
CEO duality, indirectly, much power is placed in the hands of one person, which in turn will create agency problems (Khatib et al., 2021;Zou et al., 2015) and affect company performance (Malik & Makhdoom, 2016) so that the quality of the disclosures made by the CEO is questionable, especially the readability of the financial statement footnotes. However, Goergen et al. (2020) argue that CEOs with multiple positions can improve company performance by increasing decisionmaking efficiency at the top management level and reducing conflict of opinion between board members. CEO duality is closely related to multiple controls, increasing the risk of incorrectly choosing less active board members, leading to poor company performance (Sehrawat et al., 2020).
Besides, in a weak and political organizational structure, the CEO may abuse his dual power and appoint a person who has a close relationship as a director who tends to work under the CEO's direction (Kao & Chen, 2020). Moreover, CEO duality can reduce the efficiency of organizational operations, such as control of the company by the CEO, which results in a waste of resources (Wijethilake & Ekanayake, 2020). Thus, CEOs with multiple positions can weaken the organizational monitoring structure and negatively affect company performance (Chatterjee, 2020). Uppal (2020) and Ozbek & Boyd (2020) state that CEOs with multiple positions can reduce CEO independence and worsen company performance. Chychyla et al. (2019) and Salehi et al. (2020) show that companies with a worse performance show that the company annual report is longer and more complicated than companies with a good performance. Furthermore, Chatterjee (2020) states that companies with weak performance present long financial reports that can decrease readability. CEO duality allows them to make their financial statement footnotes challenging to understand. Although CEO duality has a positive side in aligning parent and subsidiary companies' goals, it has not yet answered how open the CEOs are, given the high level of financial scandals among Indonesia's CEOs. Thus, the proposed hypothesis is: H1: CEO duality makes financial statement footnotes challenging to understand.
From the agency theory perspective, in general, every company mostly gives their stock ownership to CEOs to reduce conflicts of interest between agents and principals. The debate arose when CEOs held many shares in the public company over time. This reason comes true with the power of the CEO by improving the firm performance indicators (Tejerina-Gaite & Fernández-Temprano, 2021;Salma et al., 2003). Another consequence of a CEO's share ownership is that the CEO will get some punishment when they cannot give an action and decision managerial most appropriate for the corporate going concern (Habib & Monzur, 2019). Therefore, the CEO's share ownership gives more accountability for making an appropriate decision to reduce conflict between management and shareholders. Vu et al. (2018) suggest that CEO's share ownership will increase the firm performance, corporate governance, and firm value. Therefore, in the long term, the CEO's share ownership can encourage top management to focus more on determining business strategies rather than the negligence of using the firm's assets excessively (Abernathy et al., 2019). The CEO's share ownership is sometimes inseparable from the family ownership, where social ties between family members encourage loyalty and commitment to the family and company, thereby reducing agency costs.
On the other hand, the CEO's share ownership can increase the strength of a CEO in the company, so that the CEO tends to choose company personnel who place their credit on them, which can worsen company performance. Based on this evidence, poor company performance results in unreadable annual reports leading to reduced transparency or increased uncertainty (Nguyen, 2020). However, in general we think that with the ownership of shares in the CEO, it is hoped that the CEO will be more responsible for the company to present disclosures that are easy to understand, especially in financial statements and footnotes. Thus, the proposed hypothesis is: H2: CEO ownership makes financial statement footnotes easy to understand.

Research method
To test our hypothesis, we test the association between CEO Duality (Ownership) and the readability of financial statement footnotes by estimating the following model:

Readability indices
Readability is the dependent variable on our study. To measure readability, we use the content calculated from the financial statements' footnotes, which are part of the explanatory paragraph on accounting numbers in financial statements. Financial statements' footnotes could provide additional information on the risk of an audit engagement (Abernathy et al., 2019). The specific measurement of readability was measured by four readability indexes (Chan et al., 2020;Xu, 2020;Habib & Hasan, 2018) The complex words are determined based on words that have three syllables or more. To measure the readability score, we employ readability.exe, 1 an open-access textual analysis software. A higher score in a readability index means that the text is more complex (i.e. less readable) (Lo et al., 2017). In addition, the high and low scores depended more 1 https://trans4mind.com/personal_development/writing/Readability_ software/downloadReadability.htm on the complexity of words on financial statement footnotes.

CEO duality
CEO duality is the independent variable used to test (H1). A CEO with multiple positions is often associated with a one-tier system setting in corporate governance. A CEO is considering a double if the CEO is concurrently a commissioner in a company. However, this study brings new meaning in a two-tier system, where dummy variables measure CEO duality, with a value of 1 if the CEO serves as a director or commissioner in 1 subsidiary and 0 vice versa (Abels & Martelli, 2013).

CEO ownership
CEO ownership is an independent variable in this study that is used to test (H2). This variable is measured by a dummy variable, with a value of 1 if the CEO has ownership of the company and 0 if vice versa (Zona et al., 2018).

Control variables
Referring to previous research (Lee, 2019;Chan et al., 2020;Xu, 2020;Abernathy et al., 2019;Zona et al., 2018), this study uses Several control variables, including board size (BOARD SIZE), are measured by the natural logarithm of the number of members of the board of directors and the board of commissioners in the company. Independent commissioner measure (INDCOMSIZE), calculated by the number of independent commissioners in a company. Audit committee board size (AUCOMSIZE) is calculated by the number of audit committee boards in one company-return on Asset (ROA) is measured by dividing net income by total assets. Company losses (LOSS), measured using a dummy variable, have a value of 1 if the company received negative aftertax income in the current year and 0 if otherwise. Market Capitalization (MADCAP) was calculated using the natural logarithm of market capitalization (year-end closing price multiplied by total outstanding shares). Finally, total net sales (NET SALES) are calculated using the natural logarithm of the company's net sales. Besides, this study also uses several fixed effect variables to accommodate differences in observations' characteristics (Gujarati & Dawn, 2008), including year fixed effects and industry fixed effects. The year fixed effect variable controls differences in economic conditions during the research observation period. Furthermore, the fixed effect industry variable is used to control for differences in each industry's characteristics in the study. These two variables are nominal data types, which only explain the category, not the level, which is then transformed into several dummy variables at the time of analysis to capture the difference in intercept between research objects (Gujarati & Dawn, 2008

Sample selection
The population in this study are all companies listed on the IDX for the 2010-2018 period. Except for those included in the SIC (standard industrial classification), six codes cover depository institutions; non-depository credit institutions; securities and commodity brokers, dealers, exchanges and services; insurance holders; insurance agents, brokers, and services agents; holders and other investment offices. We also drop all observations with missing data. Finally, we drop observations for firms that do not provide a full-text annual report. Panel A Table 2 provides the sample selection process. The final sample consists of 1,469 firm-years observation. Panel B presents the number of observations by industry.  Furthermore, Panel B shows the distribution of the sample based on share ownership by the CEO. From the total observations, CEOs who have right of company shares are 423 observations or 28.80% of a total of 1469 observations, the construction industry (SIC2) is the industry that has the most CEOs who own company shares, namely 144 observations, then the mining industry (SIC1), as many as 74 observations, and the manufacturing industry (SIC4), as many as 66 observations.   Table 5 shows the Pearson correlation results, which indicates a positive relationship between the CEO DUALITY variable and the readability index (FLESCH, KINCAID, FOG, SMOG) at a significant level of 10%, 5%, 5%, and 10%. The relationship between the CEO OWNERSHIP variable and the readability index (FLESCH, KINCAID, FOG, SMOG) is negatively significant at the 1%, 5%, 10%, and 1% levels. It is shown CEO DUALITY will produce financial statement footnotes difficult to understand. In contrast, the relationship between CEO OWNERSHIP and the financial statements' footnotes shows that the CEO's share ownership makes the financial statements' footnotes easy to understand.    Table 7 shows the OLS regression results for CEO DUALITY and CEO OWNERSHIP with the readability of financial statement footnotes. The CEO DUALITY variable against FLESCH is 0.107 (t = 2.31) and is significant at the 5% level. The CEO DUALITY variable against KINCAID is 0.591 (t = 2.71) and is significant at the 1% level. The coefficient value of the CEO DUALITY variable against FOG is 0.124 (t = 2.51) and is significant at the 5% level. The CEO DUALITY variable against SMOG is 0.108 (t = 2.71) and is significant at the 5% level. The results of several regressions indicate that CEOs with multiple positions in subsidiaries have a significant positive relationship with the level of readability of the company's financial statements (FLESCH, KINCAID, FOG, SMOG).

CEO duality (ownership) and readability of financial statement footnotes
This result supports H1, where the CEO who concurrently holds the position makes the financial statement's footnotes difficult to understand. These results supported for study from Abernathy et al., (2019), where the readability of financial statements footnotes that is difficult to understand is closely related to increased audit risk and corporate litigation risk. Therefore, this study gives a short answer that one of the causes of the increased risk is the increase in agency problems. It is indicated that companies' weak supervision with a double role of CEO in subsidiary companies weakens the level of independence of a CEO (Grove et al., 2011).
In addition, the OLS regression result for CEO OWNERSHIP with readability of financial statement footnotes shows that the CEO OWNERSHIP variable against FLESCH is -0.144 (t = -2.84) and is significant at the 1% level. The CEO OWNERSHIP variable's coefficient value to KINCAID is -0.481 (t = -2.08) and is significant at the 5% level. The CEO OWNERSHIP variable against FOG is -0.093 with (t = -1.76) and is significant at the 10% level. The coefficient value of the CEO OWNERSHIP variable to SMOG is -0.165 (t = -2.96) and is significant at the 1% level. The results of several regressions indicate that the CEO's share ownership has a significant negative relationship with the level of readability of the company's financial statements (FLESCH, KINCAID, FOG, SMOG).
Thus, these results are consistent with the second research hypothesis (H2), which means that CEOs who own shares of companies tend to produce notes on financial statements that are easy to understand. CEO ownership of company shares gives a CEO more responsibility for their decisions because CEOs will receive the same consequences as other shareholders for their actions and decisions (Shleifer & Vishny, 1997). This is in line with Vu et al., (2018). With this increased sense of responsibility, a CEO will try hard to improve their performance and performance because they feel they own the company and make financial statement footnotes easy to read.

Robustness analysis
There is a possibility that the variable CEO DUALITY and CEO OWNERSHIP can be endogenous, where there is a possibility that there is a correlation between the CEO DUALITY treatment variable (CEO OWNERSHIP) and the observable variable. To overcome the endogeneity problem of the correlation between the independent variables and other variables in the observations in this study, in additional analysis, the coarsened exact matching (CEM) approach will be carried out. For this reason, this study will use CEM as an additional sensitivity analysis. In this study, the covariates were arranged into the same five strata, and using four covariates were included in the CEM model. Whereas Panel B shows a summary of CEM matching for testing CEO ownership where in this test sample there are 404 out of 423 observations of CEOs who own shares of the company, which are then matched with 924 out of 1046 observations of CEOs who do not own company shares so that a total sample is a number of 1328 of observations. This table reports a matching summary of all observations where the sample match with the duality and ownership of the CEO is treated as the number 1 while the sample mismatch of the two things is treated as 0. The result of the match of the sample is an observation that states the resolution of the problem of endogeneity of the sample. The table reported sample match for duality and CEO ownership of 1384 observations. Table 9 is the result of CEM regression on CEO duality and CEO ownership with the readability of financial statement footnotes. This table show that the relationship between the readability variables (FLESCH, KINCAID, FOG, SMOG) has a significant positive relationship with the independent variable CEO DUALITY and has a significant level variation of 5% (FLEACH), 1% (KINCAID), 5% (FOG), and 5% (SMOG), these results still support H1 and indicate that this model is free from endogeneity problems. In addition, this relationship also has a significant negative with CEO ownership on the significant level variation 1% (FLESCH), 5% (KINCAID), 10% (FOG), and 1% (SMOG). From our result, CEO ownership still supports H2 and the model also freely from endogeneity issues.

Discussion
The level of readability of the financial statement footnotes in this study is viewed from the perspective of duality and ownership from CEO; the duality of the CEO leads to a position controlled by him on several subsidiaries. The authority and discretion of the company will be compromised if the CEO has duality in every position he occupies. This will affect the vagueness of the information produced because there is a blurring of information from poor organizational performance. On the other hand, the fact shows that the CEO who has control of the company with shares ownership that has affects the operational decisions and policies of the company so that it will impact on the level of readability of the information was disclosed by the company (Grove et al., 2011). From our best knowledge, the study of readability level in Indonesia context from the duality and ownership of CEOs is still confined carried out. This study adds a bit to extent the literature for reviewing of the duality and ownership of CEOs that have the potential to affect the level of readability of financial statement footnotes.
The baseline test gives some sufficient evidence for the readability context, the critical information needs to be considered for the relationship between duality and ownership of the CEO at the readability of financial statement footnotes. Table 2 of panel B displaying in detail that the most prominent and smallest number of observations in terms of industrial scale is dominated by the mining and construction industry with a total of 240, the construction industry with 422, the manufacturing with 205, and transportation, communication, electricity, gas and cleaning services getting a total of 262 observations. In addition, the company with the smallest number of observations obtained health, legal, and educational services and consulting with a total of 29 companies, agriculture, forestry, and fishery with a total of 73 companies, and service industry with 85 companies. This shows that the average industry for our subject analysis is the construction and mining industry, where CEO duality dominates its influence on companies with many subsidiary companies under parent companies. In addition, from the table 3 shows that the distribution of samples based on CEO duality and CEO ownership, overall, the distribution of samples based on CEO duality shows that companies with CEOs with concurrent positions in subsidiary companies are as much as 46%. In comparison, those that do not have a position frame are 54%. This indicates that almost 50% of CEOs in Indonesia have concurrent positions in several positions in the company. Meanwhile, based on CEO ownership shows that CEOs who have share ownership in the company are as much as 29%, while those who do not own company shares are 71%. This result provides sufficient evidence that the average CEO in Indonesia rarely owns a stake in the company where he is the head of the company. Table 5 shows that the Pearson correlation between CEO duality and CEO ownership to the readability of financial statement footnotes. the result of Pearson correlation indicate that the CEO duality positively correlates to the readability of financial statement footnotes proxied with FLESCH, KINCAID, FOG, and SMOG. This provides information that the information disclosed by the company through financial statement footnotes with CEOs with concurrent positions has a positive relationship with the level of readability. In addition, CEO Ownership is negatively correlated with the readability rate. This evidence shows that ownership impacts increasing the level of readability of financial statement footnotes.
According the test results shows that a quite surprising fact about the influence of CEO duality and ownership on the level of readability. Concurrent CEOs have an impact on reducing the quality of financial information produced. When the CEO duplicates positions in several strategic positions in the company, the resulting decision has an information bias that can obscure the facts because the duality of the CEO can increase agency issues that give rise to interest debates between owners and management (Khatib et al., 2021;Zou et al., 2015). In addition, the duality of CEOs increases the dual control role, which causes an increased risk that CEOs fail to elect active board members so as to reduce the quality of company performance (Sehrawat et al., 2020). Based on the results of previous baseline tests, it can be concluded that the duality of the CEO makes the level of readability more challenging; this is indicated by reducing and weakening the monitoring structure of the organization and, in the end, can cause a negative influence on the performance of the organization. Thus, when the CEO has a position with multiple positions, there is a tendency that the CEO will do unfit work and produce incorrect information; it has an impact on reducing the quality of the information produced (Chatterjee, 2020). The decline indicates that management is reluctant to properly inform the company's condition so that the level of readability becomes more difficult to digest and challenge.
CEO ownership explained that the ownership or right of participation of shares in a company by the CEO could reduce conflicts of interest between principals and agents, where this happens because share ownership will increase the tension of the company's performance better than CEO do not a control related to the company (Habib & Monzur, 2019). In addition, while the they control the company, it will require the CEO to make performance following the company's long-term interests. This control from CEO can be proven by the increase in the company's performance for long time. Ownership of CEOs requires them to perform the best managerial actions and performance decisions that most appropriately follow the goals of the organization. Instead of taking actions that can bring penalties and consequences for a decrease in performance, causing a conflict of interest between the principal and management (Abernathy et al., 2019). In addition, the level of readability will be easier to understand because the information disclosed by the company reflects all available information. Furthermore, management seeks to increase information accountability for its responsibilities in the company by participating in shareholding and control. This statement reinforces evidence that CEO ownership in companies will bring and encourage those in the domain of top management to focus on determining valuable business strategies instead of misusing organizational assets (Nguyen, 2020). For this reason, the existence of an ownership CEO can increase his responsibility to improve the company's performance so that all the information presented reflects the conditions that are the objectivity of the CEO's performance without blurring the available information, and this will have an impact on the level of readability of the information more easily.
Finally, Ordinary Least Square (OLS) testing sometimes has endogeneity problems that cause the results presented to be biased. The bias of a test is caused because the resulting observations overlap each other, so a test tool is needed to analyze and prove that the resulting observation is biased. CEM or coarsened exact matching is a test tool that seeks to analyze and prove that the observations obtained are free from test bias. This tool works by equalizing the uniqueness of observation so that the observations will be produced with the same and unique characteristics. For this reason, CEM is an exciting test tool to overcome the endogeneity problem from regression testing. The problem of endogeneity is a problem that occurs because a test can negate or hold each other and will have an impact on the inaccuracy in the results that will be generated by the baseline regression. Based on tables 8 and 9, as cited in the previous section, presents a summary of the sample match and the test of the resulting sample match. The table display that there is consistency in results from CEM testing conducted with simple regression testing. This concludes that endogeneity was successfully overcome by the CEM test, which was proven to have a certain degree of significance. However, the results of the CEM test have become unproven. In other words, the endogeneity problem has not been resolved, while the test results do not show a significance level. Thus, the endogeneity problem has been resolved by the presence of CEM treatment and the results of such treatment.
The financial authority regulation No. 33/POJK. 04/2014 explained that a CEO has a limit on holding a position in the company, thus requiring them to act in accordance with their capacity on the board of directors and will affect the satisfaction of the CEO's interests in the company and have the potential to increase moral hazard behavior so that regulations are also needed that regulate the concurrent position of the CEO in the company to avoid interests that have the potential to clash with each other. CEO duality and ownership also can generate behaviors that lead to moral hazard and enrich themselves by acting outside the provisions of the company's strategic goals. They believe that their power is sufficient to remove traces of misbehavior over their interests and that such information is rarely disclosed in financial statements or information that is potentially convoluted and produces information that is difficult to understand.

Conclusions
The purpose of this study was to analyse the relationship between CEO duality in subsidiaries, CEO ownership, and the readability of financial statement footnotes in all companies listed on the IDX for the 2010-2018 period. CEO duality in subsidiaries has a significant positive relationship with the readability of financial statement footnotes. It indicates that companies with CEO duality tend to produce financial statement footnotes that are difficult to understand because a CEO duality can increase conflicts of interest and provoke opportunistic behaviour from the CEO to prioritize personal interests to weaken the independence of the CEO. The performance of the CEO is not optimal because he must hold responsibility for the two companies simultaneously. Meanwhile, CEO ownership has a significant negative relationship with the readability of financial statement footnotes. It shows that CEO ownership can increase responsibility to improve company performance and make the readability of financial statement footnotes easy to understand. In addition, share ownership by the CEO will increase the CEO's vigilance in leading the company because errors in the policies they take can impact their level of incentives in the form of company shares, which will be detrimental to the CEO.
This research enriches knowledge and implication to call for more ownership from CEO on the firm, especially in corporate governance and the readability of disclosure in an annual report, particularly in financial statement footnotes. Besides, this research provides an overview of the adverse effects and negative trends of multiple CEO positions in subsidiaries. In Indonesia, multiple roles are often misused in acts of corruption, collusion, or nepotism from a leader. This study also provides input on government regulations and financial authority regulation NO 33 / POJK.04 / 2014. This regulation only regulates public companies' concurrent positions with other public companies. Still, it is not clear how the relationship between simultaneous positions between public companies and subsidiaries. Therefore, it is essential to add to the new regulations to avoid CEOs' moral hazard of enriching themselves and ignoring the interests of the crowd.
This study has several limitations. First, it only focuses on concurrent positions of a CEO in one subsidiary company. However, it does not count how many positions a CEO holds in other subsidiaries because there is a possibility that a CEO will have multiple positions in many subsidiaries. It is hoped that in the future, this research can develop the impact of multiple positions in subsidiaries, especially in family companies, whether the results obtained are still consistent or vice versa.