The Impact of Ownership Structure on Earnings Management: Evidence from the Indonesian Stock Exchange

Business development encourages business people to improve company performance to survive in their environment. Company managers play a big role in keeping or attracting investors as a way to strengthen the company's finances. For this reason, quality financial reports are needed. Financial reports are useful for companies, both external and internal parties, as means of assessing a company and helping management achieve company goals. Information from financial statements is also useful for making economic decisions (Felicya & Sutrisno, 2020). This situation makes it possible to motivate managers to show results of questionable quality to ensure the stability of the company, as well as to ensure the necessary funds for the company's investments. Therefore, managers apply earnings management to influence investors' perceptions of managers' discretion in financial reporting. Earnings management can take a variety of forms, for example compiling specific revenues, expenses, and transactions; changing the size of the accounting; and accruals management.

In terms of legality, the practice of earnings management is said to be legal if the disclosure of a company's earnings is in accordance with GAAP guidelines, for example changing the valuation of depreciation or inventory on current assets. Contrastingly, earnings management practices can also be illegal if the activity is not in accordance with GAAP standards, for example by advancing revenue recognition and decelerating expense recognition (Aygun et al., 2014). Several shady earnings management practices have occurred in large companies in Indonesia, for example at Bank BRI in 2011. The head of the branch of Bank BRI, Tapung Raya, who held a top management position, engineered the financial statements in the cash section to benefit his interests. Another case occurred in 2014 by Inovisi Infracom. In the financial statements for the September 2014 period, indications of misstatement were found, where eight items in the financial statements should have been corrected. The company's earnings per share appear to be higher as the company recognizes its net income per share based on current period earnings, whereas the company should have used the profits attributable to the owners of the parent entity.

Another case appeared in 2019 by the state-owned company PT Garuda Indonesia. This company was found to have used the profits by recognizing unearned revenues of $239.94 million in the 2018 fiscal year. This manipulation was deliberately carried out by managers in order to show better financial performance than the previous year. Thus, substantial supervision of earnings management practices in a company is necessary, especially companies listed on the stock market because the market value of public companies is very important to investors' perceptions (Reyna, 2018). Previous studies have proven that ownership structure and company characteristics have a significant effect on earnings management. Alzoubi (2016) studying Jordanian companies reveals that family ownership, institutional ownership, blockholder ownership, foreign ownership, and managerial ownership have a bigger influence on the quality of financial reporting, because more broadly, they can potentially limit earnings management. The study agrees with Reyna (2018) using a sample of companies in Mexico whose earnings management practices can be reduced by the presence of family and institutional ownership. This however depends on the size of the company as well because it can provide different impacts. However, several studies consider that ownership control has a positive impact to earnings management (Anwar & Buvanendra, 2019; Bao & Lewellyn, 2017; Mardianto, 2020). Anwar and Buvanendra (2019) found a significant positive relationship between foreign ownership and earnings management in Sri Lanka companies. The reason may be that foreign investors expect short-term financial results, so they cannot limit earnings management discretion by managers.

Most of the earnings management research has been conducted in developed countries but little attention has been paid to how earnings management activities vary in countries with very different corporate ownership structures and national institutional environments (Bao & Lewellyn, 2017). The characteristics of earnings management in each country are different. Based on the research by Wardani and Kusuma (2012) who examined the characteristics of earnings management in Southeast Asian countries, accrual-based earnings management is more aggressive than real earnings management in Indonesia. This study also finds that earnings management is opportunistic in Indonesia, which means that managers use information asymmetry between external and internal companies to maximize their utility in relation to compensation contracts, debt contracts and regulations.

The previous research described above focuses on the relationship between corporate governance and earnings management and does not use all types of ownership as variables. This study develops previous research but only focuses on the ownership structure as an independent variable to determine its effect on earnings management as the dependent variable. Specifically, this study adds variables by including all types of share ownership, namely family ownership, institutional ownership, blockholder ownership, managerial ownership, foreign ownership, individual ownership, and government ownership. There is interest in examining the influence of each of these variables in the Indonesian context, where companies are still in their infancy and the influence of the decision-making process on companies by some shareholders is still limited. Therefore, this study is conducted to test whether ownership structure has an influence on earnings management inIndonesia.

Agency theory shows a contractual relationship between an owner and his interests and the manager, and whose relationship can lead to conflict because each party acts to maximize their respective advantage (Jensen & Meckling, 1976). In some situations, the risk appetite of the manager and the shareholders is not aligned. This is because the manager has more information about the conditions and prospects of the company in the future than the shareholders. The condition of information asymmetry is an opportunity for management to carry out earnings management, so management monitoring is needed to ensure the protection of shareholder interests and produce reliable financial reports. A structured corporate governance mechanism is expected to reduce earnings management because it results in active management monitoring, especially in the financial reporting process. Corporate governance can facilitate the reduction of agency costs that arise due to agency conflicts. Corporate governance also plays a role in confirming approval of the financial accounting system and enforcing the credibility of the financial statements (Alzoubi, 2016). Many studies suggest that corporate ownership structure offers a significant monitoring mechanism for managers; hence, may have a monitoring role in limiting earnings management activities.

Klein (2002) states that earnings management is a practice of deviations from the company's actual financial performance. This occurs when managers compile transactions in financial statements, with the intent to misguide stakeholders' views on the actual financial performance of companies or to influence transactions that rely on reported accounting values (Bao & Lewellyn, 2017).

Subramanyam and John (2010) reveal several reasons that lead to earnings management practices, such as avoiding debt requirements, influencing stock prices, and fulfilling analysis expectations. The most common earnings management practice is the changing of accounting methods and policies (Silviana et al., 2013).

The dependent variable of this study is earnings management, where the calculation uses discretionary accruals of the Modified Jones Model with the formula (Reyna, 2018):

$\frac{T A_{i t}}{A_{i t-1}}=\alpha_1\left(\frac{1}{A_{i t-1}}\right)+\alpha_2\left(\frac{\Delta R E V_{i t}-\Delta R E C_{i t}}{A_{i t-1}}\right)+\alpha_3\left(\frac{P P E_{i t}}{A_{i t-1}}\right)$

Total accruals for period t are expressed in the equation:

$T A_{i t}=N I_{i t}-O C F_{i t}$

Non-Discretionary Accruals (NDA) can be determined by the equation:

$N D A_{i t}=\alpha_1\left(\frac{1}{A_{i t-1}}\right)+\alpha_2\left(\frac{\Delta R E V_{i t}-\Delta R E C_{i t}}{A_{i t-1}}\right)+\alpha_3\left(\frac{P P E_{i t}}{A_{i t-1}}\right)$

After regression, the company's DA can be calculated by the equation:

$D A_{i t}=\frac{T A_{i t}}{A_{i t-1}}-[ \alpha_1\left(\frac{1}{A_{i t-1}}\right)+\alpha_2\left(\frac{\Delta R E V_{i t}-\Delta R E C_{i t}}{A_{i t-1}}\right)+\alpha_3\left(\frac{P P E_{i t}}{A_{i t-1}}\right) ]$

where:

$T A_{i t}=$ Total accruals of firm $i$ in the yearly period $t$.

$N I_{i t}=$ Net income of firm $i$ in the yearly period $\mathrm{t}$.

$O C F_{i t}=$ Cash flow from operating activities.

$A_{i t-1}=$ Total assets of firm $i$ in the yearly period $t$ - 1 .

$\Delta R E V_{i t}=$ Change in revenue of firm $i$ in the year $t$ compared to year $t-1$.

$\Delta R E C_{i t}=$ Change in account receivable of firm $i$ in the year $t$ compared to year $t-1$.

$P P E_{i t}=$ Fixed assets of firm $i$ in the yearly period $t$.

$N D A_{i t}=$ Non-discretionary accrual of firm $i$ in the yearly period $t$.

$D A_{i t}=$ Discretionary accrual of firm $i$ in the yearly period $t$.

It is mandatory that the internal control system is owned by the company, and this for various reasons, both for the smooth operation of the company and the level of security. Internal control is a system created by a company or organization in managing all its activities in it to achieve the company's goals. The ownership structure is an internal control component that refers to the manner in which representative rights reallocate company capital in one or more individuals or legal entities.

Family owners are family members who carry out various operational activities in the company and are shareholders in the company. The measurement of family ownership according to Darmadi (2016) is the sum of family ownership shares divided by the total outstanding shares.

The ownership of shares by companies and financial institutions in a company is called institutional ownership (Paramitha & Firnanti, 2018). Following Reyna (2018), this study considers institutions only when their ownership represents 5 percent and above of the company's share capital. The measurement of institutional ownership according to Reyna (2018) is the sum of institutional ownership shares divided by the total outstanding shares.

Blockholder ownership is defined as the number of shares owned by individuals who are not family members. Following Nguyen et al. (2020), this study considers blockholders only when their holdings represent 5 percent or above of the company's equity share capital. Measurement of blockholder ownership according to Nguyen et al. (2020) is the sum of ownership shares by blockholders divided by the total outstanding shares.

Company shares owned by foreign individuals, foreign legal entities, foreign governments are included in foreign ownership (Meilita & Rokhmawati, 2017). The measurement of foreign ownership according to Kablan (2020) is the sum of foreign ownership shares divided by the total outstanding shares.

This happens when the management of the company owns shares in the company (Meilita & Rokhmawati, 2017). Measurement of managerial ownership according to Tran et al. (2020) is the sum of managerial ownership shares divided by the total outstanding shares.

Individual ownership is a part of shares owned by outside parties such as the general public who do not have a special relationship with the company (Yusnita, 2019). Measurement of individual ownership according to Hendi and Lisniati (2020) is the sum of individual shares whose ownership is below 5% divided by the total outstanding shares.

Government ownership is a situation where the government owns the company's shares. The measurement of government ownership according to Attia (2019) is the sum of shares owned by the government divided by the total outstanding shares.

The source of operational funds used by the company is described as leverage. Measurement of leverage according to Reyna (2018) is total liabilities divided by total assets.

Company scale is usually divided into three, namely small-scale enterprises, medium-scale enterprises, and large-scale enterprises. Company size is determined in terms of company assets, company sales, company share value, and others (Yuliana & Trisnawati, 2015). The measurement of company size according to Reyna (2018) is the natural logarithm of total assets.

Profitability is measured by return on equity (ROE), which is a measure that shows the amount of profits obtained from shareholders on investments made by the company. The measurement of return on equity according to Lai and Tam (2017) is net income divided by equity.

Company growth is the ability of a company to increase its size (Paramitha & Firnanti, 2018). Growth opportunities are measured as the annual sales growth rate (Reyna, 2018). The measurement of company growth according to Reyna (2018) is the difference between current sales and previous year's sales divided by previous year's sales multiplied by 100%.

Al-Duais et al. (2019) stated that companies with family ownership tend not to allow earnings management because generally families invest a lot of personal assets for their own interests and reputation. For this reason, family owners are highly motivated to focus on monitoring management. Mardianto (2020) found that family ownership had a significant effect on earnings management in Indonesian manufacturing companies. The increase in earnings management practices in family companies occurs because the shares are owned by family members, who have wider access to said financial statements and can manipulate them to make the company look good. Family-owned companies will be more effective in terms of maintaining the quality of financial reports and lowering agency costs when compared to public-owned companies.

H1 = Family ownership has significant positive effect on earnings management.

Many studies conclude that institutional ownership has a significant negative effect to earnings management (Alhadab et al., 2020; Alzoubi, 2016; Anwar & Buvanendra, 2019; Aygun et al., 2014; Liu & Tsai, 2015; Nugroho, 2017; Paramitha & Firnanti, 2018; Reyna, 2018). Lin et al. (2014) who conducted research in Taiwan, noted the fact that in recent years, the major shareholders in Taiwanese companies have been institutional investors. And in nearly 80% of the recognized daily transactions, investment decisions of institutional investors are considered to be informed and knowledgeable. Based on the literature, Liu and Tsai (2015) concluded that institutional ownership and earnings management were significantly negatively related. With evidence from Taiwanese companies, earnings management behavior will decrease if the company has high institutional ownership because its monitoring role is maximized.

Reyna (2018) similarly found that there is a clear relationship between high institutional investor participation and decreased earnings management. This finding is in line with previous studies which revealed that ownership control and involvement are very effective in tightening earnings management practices. Great commitment by long-term oriented investors to the company can be seen as a result of declining earnings management.

H2 = Institutional ownership has significant negative effect on earnings management.

Previous research (Kablan, 2020; Nguyen et al., 2020; Saona et al., 2020; Yahaya et al., 2019) found that majority shareholders have more interference power than minority shareholders in pushing managers to convey financial information, so this motivates managers to reduce earnings management, especially when the company's financial performance is decreasing (Reyna, 2018).

H3 = Blockholder ownership has significant negative effect on earnings management.

One of the many ways that are expected to tighten profit manipulation activities in a company is foreign investment. Aggarwal et al. (2005) argue that companies with good quality accounting information will attract foreign investors to invest their capital in these companies. Many researchers from various countries have found that highly knowledgeable foreign investors can hold back on earnings management (Alzoubi, 2016; Lai & Tam, 2017; Nia et al., 2017; Osemene et al., 2018; Widyaningsih, 2017). Alzoubi (2016) who examined this variable with a sample of Jordanian companies found that earnings management would be smaller with the supervision of foreign ownership. This research is in line with Chung et al. (2004) who studied companies in Japan and determined that the role of foreign ownership can reduce earnings management because they are less involved in cutting discretionary costs but play an active role in deterring managerial opportunistic behavior.

H4 = Foreign ownership has significant negative effect on earnings management.

Widyaningsih (2017) investigated whether high managerial ownership could encourage earnings management practices. Based on agency theory, management does not always act in the interests of the owner, because of differences in interests, thus managers will tend to implement earnings management. The misalignment of the goals of management and owners will cause agency costs. The higher the managerial ownership, the higher the related party transactions carried out for earnings management purposes. The higher share ownership by management will increase the incentive for managers to conduct business transactions with related parties, both for the purpose of earnings management and to achieve certain targets (Nugroho, 2017).

H5 = Managerial ownership has significant positive effect on earnings management.

The more shares the public owns, the more information will be disclosed in the annual report. Individual ownership is one way to reduce agency conflicts. Agency theory states that individual ownership has the ability to monitor managers well. Agency conflict will decrease because of the high supervision carried out by individual ownership, thus creating transparency which is a principle of good corporate governance. Public ownership determines the amount of personal information that must be distributed by the manager to the public (Kusumaningtyas & Farida, 2016; Meilita & Rokhmawati, 2017). According to Claessens et al. (2000), public ownership is one of the largest blockholder groups in developing countries.

H6 = Individual ownership has significant negative effect on earnings management.

Research on the relationship between ownership characteristics and earnings management conducted by Guo and Ma (2015) who studied companies in Shanghai between 2004 and 2010 showed that earnings management is determined by the motivation of various types of ownership. The researcher specifically states that firms are less likely to engage in earnings management when the state is the largest shareholder. However, a recent study by Nguyen et al. (2020) with a sample of Vietnamese companies found that state ownership has a significant positive effect on earnings management. The reasons for the impetus for earnings management are profit maximization and political goals. The results of this study are in line with studies with the same variables in different places (Attia, 2019; Ben-Nasr et al., 2015; Ding et al., 2018; Lai & Tam, 2017; Mardianto, 2020).

H7 = Government ownership has significant positive effect on earnings management.

This study uses a type of comparative causal research, which is a type of research that seeks to identify a causal relationship between the independent variables represented by the ownership structure and the dependent variable represented by earnings management. This study uses secondary data in that it only uses data on the number of shareholdings and financial statements contained in the annual report. For companies to be included in the sample, they would need to have an annual report issued by the IDX for four consecutive years between 2016 and 2019. Whilst this period is close to the current research year, it is still before the Corona virus pandemic which might have had a significant effect on company income from previous years. Another criteria which needed to be met by a company to be included in the sample was that its share ownership had to be divided into several types of share ownership. The number of companies used in the research sample was 391 and their financial statements were observed for 4 years. So, the total number of entries tested was 1564.

The descriptive statistical table above demonstrates that the value of discretionary accrual earnings management has an average value that is close to 0.1 and is negative. This output means that the non-financial Indonesian companies in the sample are engaging in "income decreasing" earnings management , as in other studies (Felicya & Sutrisno, 2020; Hendi & Lisniati, 2020; Kusumaningtyas & Farida, 2016; Mardianto, 2020; Widyaningsih, 2017). The minimum and maximum values state that there are several companies that do not have several types of ownership which are variables in this study, and it also states that all companies in this sample have been listed on the Indonesia Stock Exchange. When looking at different types of ownership, family ownership is only present in 3.6% of the companies sampled, institutional ownership is present 61.7%, blockholder ownership is present 66%, managerial ownership is present 5%, foreign ownership is present 22.6%, individual ownership is present 25%, and government ownership is only present 2.9%. This shows that the dominant shareholders in manufacturing companies listed on IDX are institutional investors and blockholder investors. The company's average debt in the analyzed period is 118%. The average results of the calculation of firm size are quite varied, which indicates that there is no problem of bias in terms of size. Profitability represented by the return on equity variable shows an average of -7.1 percent, and the average value of the company's growth in terms of annual sales is 32 percent during the 2016-2019period.

The table above shows that the probability values on Chow Test and the Hausman test method. Both probability results are 0.0000. Therefore the model to be used is the Fixed Effect Model because both results are below 0.05.

The probability result from the F test states that the ownership structure variable simultaneously has a significant effect on earnings management because the probability value is below 0,05. So, this regression model can be used to predict the factors that affect earnings management.

The results of the t test show that all independent variables have a probability value of > 0.05, which concludes that all ownership structures have no significant effect on earnings management. The negative coefficient values are present in blockholder ownership, managerial ownership and government ownership variables, the rest have positive coefficient values. Only the variables of leverage, size, and company growth have a significant effect on earnings management.

The test results show that H₁ is not proven but is consistent with previous researchers (Abdullah & Ismail, 2016; Alhadab et al., 2020; González & Garcia- Meca, 2014). The results of this study suggest that in Indonesia, agency problems do not apply and family members who own shares and control the company do not pursue their own interests.

The test results show that H₂ is not proven but is consistent with previous researchers (Farouk & Bashir, 2017; Felicya & Sutrisno, 2020; González & Garcia- Meca, 2014; Lestari & Murtanto, 2018; Purnama, 2017; Widyaningsih, 2017; Yasser et al., 2017). This happens in accordance with the theory put forward by Porter (1992) which states that institutional shareholders are shareholders who focus more on short-term profits. Therefore, managers take earnings management actions that can increase the company's short-term profits. In addition, from the research data sample, one notes that several companies have an unstable proportion of share ownership.

The test results show that H₃ is not proven but is consistent with the results of previous studies (Al-Fayoumi et al., 2010; Anwar & Buvanendra, 2019; Kablan, 2020; Reyna, 2018). It is similar to institutional ownership, considering that the type of blockholder ownership consists of institutions and individuals who own the majority shares and blockholder shareholders only focus on short-term profits. This may also be the case, as Pound (1988) argues, that they are either ineffective in their supervisory role due to possible lack of expertise or suffering from free- riding issues among themselves or strategically allied with management.

The test results show that H₄ is not proven but is consistent with the results of Hossain’s (2020) research. There are factors that affect the inability of foreign shareholders to carry out the supervisory function, namely the lack of participation of foreign parties in the implementation of corporate governance, regulatory differences, and cultural diversity, and foreign shareholders only focus on company profitability for investment.

The test results show that H₅ is not proven but is consistent with the results of Liu and Tsai (2015). If the committee appointed by the company has done its job well in the sense that the shareholders' goals have been achieved, then the supervisory role by managerial shareholders does not affect the reduction of agency conflicts (Warfield et al., 1995).

The test results show that H₆ test is not proven but is consistent with previous research (Al-Fayoumi et al., 2010; Hendi & Lisniati, 2020; Moslemany & Nathan, 2019; Yusnita, 2019) where individual ownership has no effect on earnings management. Individuals generally only want a high return from the chosen investment without considering other factors, so the supervisory function of management is not so influential (Al-Fayoumi et al., 2010).

The test results show that H₇ is not proven but is consistent with the results of Guo’s and Ma’s (2015) research. Managers who work in government-owned companies generally receive compensation in the form of political promotions and remuneration, not bonus compensation from a percentage of profits like commercial companies. This clearly proves that the supervisory function of the government is not a measure of the existence of earnings management practices in a company.

The conclusion of this study is that all types of ownership structures as independent variables have no significant effect on earnings management in the sample of Indonesian companies for the period 2016-2019. Leverage, company size, and company growth are control variables that have a significant positive effect on earnings management. Companies that have a high level of leverage are more likely to exceed debt agreements where this has an effect on improving earnings management (Reyna, 2018). The pressure to carry out earnings management will be even greater if the size of the company is large. Large companies have a tendency to manage earnings to attract market attention and meet the expectations of financial observers (Barton & Simko, 2002). Companies with good growth usually have a tendency to increase their profits in order to maintain investor confidence to keep investing their funds in the company (Felicya & Sutrisno, 2020).

This study may have managerial implications where our findings show that shareholders have a little role in the supervision of earnings management in Indonesia, but internal company factors can influence the occurrence of earnings management. The supervisory role of shareholders may not have an effect because there may be other committees that have been formed by the company that have worked well in achieving shareholder goals. However, the larger the company's debt, size and growth, the more the role of supervision is needed. Therefore, we suggest to companies that the role of the ownership structure needs to be improved and strengthened to make it clearer and more beneficial to the company. Some of the limitations of this study are that many companies listed on the IDX do not publish complete financial statements consistently throughout the years so the number of the sample is reduced. Many companies do not distribute shareholders according to the type of ownership structure. This study also focuses only on companies listed on the IDX and uses only public financial data, so we cannot deal with increased control structures. The researchers in this study believe that ownership structure cannot be used solely to measure the level of earnings management in the case of developing economies such as Indonesia. There are other factors to be considered, such as the political situation of the country, the level of economic development, and cultural issues related to the country. Future research should be able to take samples of primary data sourced from questionnaires or interviews and secondary data from all sectors of the company. Future research is also expected to include more independent variables to analyze their effect on earnings management and increase the period of research.