Analysis of Factors Affecting the Restatement of Financial Statements

The convergence of IFRS has brought an impact on the changes of financial accounting standards in Indonesia. The variation provide an opportunity for the restatement of financial statements. Therefore, this study aimed to examine CEO turnover, earnings management, and audit quality on the restatement of financial statements of manufacturing companies listed on the Indonesia Stock Exchange (IDX) during the period of2014 to 2018. In this study, purposive sampling method was carried out to consider samples with predetermined criteria. The obtained data were about 376 and analyzed using logistic regression with the SPPS 24.0 tool. As the results of data analysis, this study indicated that the CEO turnover had a positive effect on restatement; the earnings management had a positive effect on restatement; and the audit quality had no effect on restatement of financial statements.


Introduction
The restatement of a company's financial statements can lead to a variety of different perceptions from investors. There are two factors that cause restatement, namely internal factors and external factors of the company (Gleason et al., 2008). Internal factors can be in the form of errors that occur within the company such as an error made by the stakeholder in this case is the Chief Executive Officer (CEO), one of which is to increase the company's profits or assets so that the company's performance becomes better. External factors are audit quality produced by external auditors where restatement can be considered an indicator of audit failure (Liu et al., 2009). In general, restatements have a significant impact, namely reducing the quality of information contained in the financial statements and audit quality. In PSAK no. 25 mentioned the causes of the restatement of financial statements, namely: (1) Changes in Accounting Policies, (2) Changes in Accounting Estimates, and (3) Fundamental Errors.
The information presented in the financial statements must be of high quality so that it can be used in making decisions by investors, as stated first by (Staubus, 2003). Users of financial statements consider that the restatement of financial statements can be negative and positive, this is because restatements can improve the quality and integrity of financial information, but can also create opportunities for managers to manipulate data (Cai et al., 2014). Another opinion says that restatement is not like bankruptcy or company failure which reflects a financial crisis and is a violation of shareholder trust and is a signal that indicates that the financial statements are invalid (Chen et al., 2014).
In 2018, one of the SOEs made a restatement of the financial statements, namely PT. Garuda Indonesia Tbk (GIAA) quoted from economy.okezone.com (27/07/2019) in the financial report posted a net profit of USD 809.85 thousand where this figure increased sharply compared to the 2017 financial year which suffered a loss of USD 216.5 million. . In fact, this profit is partly due to the cooperation between Garuda and PT Mahata Aero Technology with a value of USD 239.94 million as stated in the cooperation agreement to provide in-flight connectivity services.
In this case, revenue is only recognized if there is a high probability that the royalties will be received, but until the end of the 2018 financial year, Mahata has not made any payments. The same thing also happened to PT Hanson Internasional Tbk where in the records of the Financial Services Authority (OJK) this company was proven to have manipulated the presentation of the annual financial report (LKT) for 2016. The directors and KAP of PT Garuda Indonesia Tbk (GIAA) and PT Hanson International Tbk is subject to material and social sanctions.
Restatements tend to be associated with complex transactions where there is intentional manipulation by error and managerial manipulation (Lin et al., 2017) and restatements are perceived as indicators of low audit quality (Christensen et al., 2016).
First, this research was motivated with the title "The Impact of Management Power on Financial Restoration" with 128 companies listed in Shanghai and Shenzhen. In addition, Agrawal & Cooper (2017) using a sample of 518 public companies in the US that announced restatements found strong evidence between CEO turnover and restatements. The same thing is also found in Rich & Zhang's (2016) research where management change has an important role in determining public policy.
Second, Earnings management is the issue that most often occurs in the restatement of financial statements. Khajavi & Arani (2015) find evidence that accrual earnings quality has a significant relationship with restatements. Wang & Zhang (2018) found something similar, where accrual earnings management showed a positive relationship with restatement of financial statements.
Third, restatement can also be interpreted as an audit failure, the company will replace their auditor. Audit quality shown by the timeliness of disclosure also shows that a longer auditor tenure does not have a significant impact on the restatement of financial statements. The economic consequences obtained from restatement are causing a bad image and reputation of the company in the future, increasing bad information and affecting investor confidence and the worst impact is causing the risk of financial difficulties for the company.
This study aims to 1) examine and analyze the effect of changing the Chief Executive Officer (CEO) on the restatement of financial statements; 2) examine and analyze the effect of earnings management on the restatement of financial statements; 3) examine and analyze the effect of audit quality on the restatement of financial statements.
The decision usefulness theory was first proposed in 1954 by (Staubus, 2003). This decision usefulness theory contains the requirements for an information to be said to be of high quality so that it can be used in decision making by information users. There are 3 important things in the theory of decision usefulness in providing financial information, namely from the accounting side, investors and investors in terms of owners and creditors.
Decision usefulness is information that is fairly presented from management assumptions that are used as the basis for decision making. A system used to formally present information will adapt to two general conditions. The first is the condition where the system is applied consistently. Second, the user aspect of the financial statements which requires the information produced to be relevant.
The usefulness of accounting information has components that can be considered by presenters of accounting information so that the existing coverage can meet the needs of decision makers. The presentation of accounting information needs to be considered and adjusted to the level of needs of the users of the financial statements. Financial statements are the end result of a process of recording financial transactions in a company that describes the financial condition of an accounting period and is also an overview of the company's performance. The definition of financial statements is financial information for parties outside the company which is Agency theory is defined as a contractual relationship within a company under one or more resource owners (principals) and agents are also referred to as managers or who control the use of resources (Jensen & Meckling, 1976). Management is a party that is contracted directly by shareholders, which can also be referred to as an agent who works in the interests of shareholders. Agency can cause problems in the company. First, information asymmetry here can be interpreted as the amount of information owned by the management rather than the principal which can cause different perceptions as agents and principals. Second, there is a conflict of interest between the management and the owner resulting from differences in objectives and the agent acting on his own volition.
The main focus of agency theory is that managers are motivated by their own personal gain in working to exploit their own personal interests rather than considering the interests of shareholders. Earnings management is a choice made by managers on accounting policies that are applied with specific objectives such as getting bonuses, tax savings, CEO turnover and initial public offerings. This causes management to commit fraud in processing financial statements so that they can achieve predetermined targets with the aim of getting incentives for themselves.
In this study, three hypotheses were developed. The first hypothesis, examines the change of Chief Executive Officer (CEO) to restatement of financial statements. Rich & Zhang (2016) stated that top management turnover has a positive effect on the restatement of financial statements. Agrawal & Cooper (2017) find strong evidence that CEO turnover has a positive effect on restatement. Li et al (2018), a positive relationship between top management turnover and restatement.
H1: The change of Chief Executive Officer (CEO) has a positive effect on restatement.
The second hypothesis predicts earnings management affects restatements. Hasnan et al (2020) conducted a study on executive compensation and leverage which as earnings management behavior has a significant positive relationship to the restatement of financial statements. The same finding is also found in the research of Wang & Zhang (2018) that securitization that leads to earnings management actions has a positive effect on restatement of financial statements. Khajavi & Arani (2015) find the same thing that the quality of earnings accruals and leverage has a significant positive relationship with restatement of financial statements.
H2: Earnings management has a positive effect on the restatement of financial statements.
The third hypothesis is to assess the effect of audit quality on presentation.
Positive relationship from audit quality to restatement of financial statements. Research also proves that audit quality has a significant positive relationship to the disclosure of financial statement restatements.
H3: Audit quality has a positive effect on the restatement of financial statements.

Methods
The data source used by the researcher is secondary data obtained from the Indonesia Stock Exchange, especially manufacturing companies during 2015-2018 with several conditions in the selection of samples. of 414 (four hundred and fourteen) data on manufacturing companies listed on the Indonesia Stock Exchange from the period 2015 -2018. Manufacturing companies that do not meet the criteria are companies for which no annual financial statements are found either on the Indonesia Stock Exchange website or on the company's official website and reports the financial statements are not issued in rupiah currency. Table 1 shows that the final sample used in this study was 376. The data needed is the annual report data, which can be found both on the IDX website and on the company's official website for the 2015-2018 period. The final total sample used is 376 manufacturing companies.
Logistic regression analysis in this study was used because the dependent variable was categorical or dummy data. The logistic regression model formulated is as follows: For the first and third hypotheses, the values 0 (zero) and 1 are used as dummy variables, while the second hypothesis uses the Dechow & Dichev (2002) model. Dechow & Dichev's model explains that the original characteristics of the accrual process suggest that the magnitude of the estimation error will be systematically related to company fundamentals such as the company's operating cycle and the company's operational variability. Furthermore, this model builds an accrual framework, where profit will be equal to cash flow plus accruals with the formula: From an accounting perspective, for every cash flow, there is a cash receipt or disbursement and the recognition of those cash flows in profit (as revenue or expense). CFt1 indicates that cash flows occur after an appropriate amount has been recognized in revenue (eg receivable receipts), CFtt refers to cash flows received or paid in the same period as the cash flows recognized in profit. Lastly, CFtt+1 refers to cash received or paid before revenue or expenses are recognized in profit. So the total cash flows for period t are: CFt = CFt t-1 + CFt t + CFt t+1 As long as the realized cash flows differ from the estimated accruals, the initial accruals (ACFt/t+1 o ) will contain estimation errors that are corrected by closing accruals (ACFt/t+1c). The initial accrual at time t, which reflects the expected cash flows t+1, is equal to the actual cash flows t+1 plus the error term (ℇ) which reflects the difference between the expected accruals and the realized cash flows. Assuming that all accruals are settled in one period so that the final accruals for the beginning of t-1. Using the equation for total cash flows for period t and defining total accruals as the sum of initial accruals and ending accruals allows us to express earnings as:

Et
= CFt + Accrualst Et = (CFt t-1 + CFt t + CFt t+1 ) + (ACFt+1/1 o + ACFt/t-1 c + (ACFt/t+1 o + (ACFt-1/t c )) Et = (CFt t-1 + CFt t + CFt t+1 ) + (CFt+1 t + ℇt+1 t -CFt t+1 -ℇt t-1 -CFt t+1 + CFt-1 t ) From the formula above, the accrual portion contained in profit (Et)is determined by the following formula: In calculating accrual at time t by resetting the accrual profit (At) portion of the equation above as follows: At = CFt-1 t -(CFt t+1 + CFt t-1 ) + CFt+1 t + ℇt+1 t -ℇt t-1 (4) The above equation states that (1) accruals are temporary adjustments that delay or anticipate the recognition of realized cash flows plus the term estimation error; (2) accruals are negatively related to current cash flows and positively related to past and future cash flows; and (3) the error term captures the extent to which accruals map to cash flow realization, and can be used as a measure of accruals and earnings quality. To obtain a practical measure of the quality of working capital using changes in working capital accruals (ΔWC) with the following formula: ΔWCt = b0 + (b1*CFOt-1) + (b2*CFOt) + (b3*CFOt+1) + ℇt With respect to the accrual of earnings equation, the measure of accruals is the change in working capital and the proxy for cash flows associated with accruals is cash flow from operations (CFO), so the above equation is used to calculate earnings management that occurs. The residual value of the regression reflects accruals that are not related to the realization of cash flows and the standard deviation of this residual is a measure of the quality of firm-level accruals, where a higher standard deviation indicates lower quality. The theoretical specification in the earnings accrual equation only uses the past, present and future cash flows associated with current accruals, because it cannot identify the components of these cash flows, the empirical version in the accrual working capital equation uses total CFO. Thus, the independent variable in the accrual working capital equation is measured by error, which implies that the regression coefficient tends towards 0 and R2 will decrease. Since the coefficient values of the earnings accrual equation are b1 = 1, b2 = -1, and b3 = 1, it is expected that 0 < b1 < 1, and -1 < b2 < 0, and 0 <b3 <1. To calculate changes in working capital and cash flow, researchers used financial statement data.  Table 2 describes the total manufacturing companies used as a sample of 376 sample data for the period 2015 -2018. Based on the table above, it is known that the restatement of financial statements (P.Kemb) has the lowest value of 0 (zero) and the highest value of 1 ( one). The value of the numbers 0 (zero) and 1 (one) is the value used as a dummy variable, where a value of 1 (one) is indicated for companies that revise their financial statements, in other words, a restatement of financial statements and a value of 0 (zero) is indicated for companies. who do not restate the financial statements? The average restatement of financial statements is 0.06 with a standard deviation of 0.23. This shows that the restatement variable data contains data that is heterogeneous or has the characteristics of sample data that tend to be different when not restating and when restating. Table 3, the overall model test with a value of -2 Log Likehood before the independent variable is entered into the model is 161,977 and block number 1 (one) which shows that the value of -2 Log Likehood after the independent variable is entered into the model becomes 146.754. From the value above, it shows that there is a decrease in the value of -2 Log Likelihood of 5.223 so it can be concluded that the logistic regression model in this study as a whole is in accordance with the data and is feasible to use.   The results of data processing to predict the restatement of financial statements can be seen in table 6. The definition of the logistic regression equation in table 6 is the probability of a company to get restatements and not to restate its financial statements can be calculated by this equation. The probability number of the company to get a restatement of financial statements and not to restate it with a value between 0 (zero) to 1 (one). If the value of the resulting equation is close to 1, it means that the possibility of the company to make a restatement is large. If the resulting value is close to 0, it means that the possibility of the company to make a restatement is small.
The results of testing the first hypothesis with a coefficient value of 1.361 and a significance of 0.018 prove that CEO turnover has a positive and significant effect on the restatement of financial statements in other words H1 is accepted. This indicates that the higher the CEO turnover rate, the more likely a restatement will occur and vice versa. The CEO is the top position in company management and is the person responsible for the operations and performance of a company. The results of the hypothesis also confirm the agency theory where management is a party contracted by shareholders and is given the authority to make the best decisions for the company. Agency theory also mentions the existence of conflicting information within the company, causing a conflict of interest itself. The new CEO will then have new strategies and policies in hopes of bringing about change for the company. So that the CEO will display his performance well in front of the principal. However, not all CEOs can work well, this causes the CEO's tendency to do anything to make their performance look good in the eyes of the principal. If the old CEO does not work as well as possible, the principal will fire the old CEO and replace him with a new CEO. Therefore the probability of restatement is higher with the change of CEO.
The results of this study are in line with research where CEO turnover has a positive relationship with restatement of financial statements. Li et al (2018) also found a similar relationship between CEO turnover and restatement of financial statements. Similar results were also found in the study of Rich & Zhang (2016). Therefore, the inaccuracy of the financial statements biases the public to evaluate how well the company's executives perform in fulfilling their responsibilities. So it can be concluded that CEO turnover can create opportunities to commit fraud that causes restatement of financial statements.
The second hypothesis shows that earnings management has a positive influence on the restatement of financial statements as indicated by the results of the hypothesis test in Table 6 with a significance value of 0.001 with a significant level of 0.05, which means the second hypothesis is accepted.
The results of this study confirm the theory proposed by Schipper (1989) about agents and principals, namely a condition when management intervenes in the process of preparing financial statements for external parties by taking actions to level, increase, and decrease earnings reporting for personal gain. Earnings management by management can affect cash flow, accounting policies adopted and applied by the company. The management can postpone or apply the chosen policy, change or change a certain accounting method with the aim of lowering costs or increasing revenues and ultimately affecting the profit recorded in the financial statements. The positive relationship between earnings management and restatement of financial statements can be seen from the explanation above.
In general, the principal will demand maximum performance which is usually measured by the distribution of profits or dividends received by the principal. The management will make estimates and arrangements in accounting estimates such as which items will be changed, either added or subtracted, adjusting the company's income statement so that it gets the desired profit. This earnings management behavior makes the financial statements biased and the information conveyed is not transparent. This results in incorrect decision making (decision usefulness theory) because the information presented in the financial statements is not valid (Staubus, 2003).
Earnings management practices may benefit firms when management applies earnings policies to increase discretionary earnings-related information. Earnings management actions that tend to maximize personal profits lead to incorrect financial reporting (Kempf et al., 2016). In the end, the earnings management carried out has an impact on the restatement of the financial statements. Management's purpose of restatement is to improve the behavior or actions they have taken.
The results of the study support the research of Khajavi & Arani (2015) where there is a positive relationship between the quality of accurate earnings and leverage with restatement of financial statements. The same thing was also found in the research of Wang & Zhang (2018) which showed a significant positive relationship between earnings management and restatement. Research from Hasnan et al (2020) also shows that the factors of executive compensation and corporate leverage are positively related to the restatement of financial statements. The above statement supports the possibility of misinformation or information asymmetry presented in the financial statements, causing the restatement of the financial statements to occur.
The third hypothesis in this study is that there is a positive effect of audit quality on the restatement of financial statements (restatement). In table 6 the significance value of the results of the third hypothesis test is 0.425 with a significance level of 0.05 and a B value of -0.624, which means H3 is rejected. The rejection of this third hypothesis shows that there is no influence between audit quality and restatement of financial statements.
In carrying out the assignment, the auditor is required to provide an opinion on the company's financial statements. The opinion given is a statement of fairness in all material respects, financial position and results of operations, and cash flows in accordance with financial accounting standards applicable in Indonesia. Audit quality is defined as the probability that an auditor will find and report violations in his client's accounting system (De Angelo, 1981 The restatement of the financial statements was due to a mathematical calculation error referring to PSAK No. 25.
In Indonesia, the restatement of financial statements can occur due to requests by regulators for audited financial statements. This shows that the independent auditor is not able to find any errors in the company's financial statements. so that for the next year's audit report the company will use the services of a new KAP. From the results of the research above, it can be said that audit quality is not the only thing that affects the restatement of financial statements in Indonesia. This is supported by the sample used only 2 companies that make changes in auditors. Based on these data, it can be interpreted that only 2 companies that make changes to the auditor make restatements of financial statements. In other words, there are other factors outside of auditor turnover that affect restatement. The results of the study that can be concluded are the change in auditors has no effect on the restatement of financial statements for manufacturing companies in Indonesia.
KAP does not determine whether to restate the financial statements. The restatement of financial statements needs to be done only because there are reports that need to be addressed, KAP is only limited to providing advice or input. For companies that already have financial statements that are properly disclosed, there is no need to restate the financial statements. On the other hand, if there are financial statements that need to be addressed, the company will automatically be restated.
The results of this study support previous research conducted by Akadiati (2018) where auditor turnover has no significant effect on the restatement of financial statements. Rachman (2019) also shows that longer audit engagements can reduce restatements. Research also found that auditor tenure has a negative effect on restatement of financial statements. Hasnan et al (2020) also did not find a significant relationship between audit quality and restatement of financial statements.

Conclusion
Based on the results of the discussion that has been described previously, it can be concluded that CEO turnover has a positive effect on the prediction of restatement of financial statements. This is due to the behavior of CEOs in determining the policies that are implemented so that their performance looks good from the principal's side, causing information asymmetry. Management behavior in intervening the process of preparing financial statements for external parties by taking actions to level, increase, and decrease profits with the aim of personal interest which has a positive influence on the restatement of financial statements. In general, audit quality has a negative effect on restatement. In other words, audit quality is not only one of the benchmarks used in restatement of financial statements.