Potential Big Bath Accounting Practice in CEO Changes (Study on Manufacturing Companies Listed in Indonesia Stock Exchange)

This Research aims to compare the earnings management which is big bath accounting model while CEO Changes in Indonesia. This research is using Secondary data which is Financial Statement from the Indonesian Stock Exchange. CEO change is classified either as routine or non-routine based on RUPS (General Shareholders Meeting) and RUPSLB (Extraordinary General Shareholders Meeting) information. The purposive sampling was used in this research by sampling 14 listed company of CEO Change non-routine and 34 listed company of CEO Change routine. These samples are observed from 2004 to 2014. To identify the big bath accounting practice. Although CEO Change non-routine made a high correlation in this study, the study provides there is no difference in earnings management big bath accounting model while CEO Changes between routine and non-routine changes.

Sometimes managers applying accounting policies by lowering revenue. It is called the big bath. Bath or also called big bath, and this may occur during of the organizational pressure at the turn of the new management is to acknowledge failure or deficit because the old management and new management wanted to avoid such failures. This technique can also recognize the costs in future periods and the loss of the current period when the potential loss cannot be avoided in the current period. Consequently, the management undertakes "big bath" by charging the cost estimates later and do clear the desks. As a result, the next period earnings will be higher than it should be.
Many empirical studies have documented the big bath done by managers (Walsh et al., 1991;Beattie et al., 1994;Hwang & Ryan, 2000;Christensen et al., 2008;Riedl & Srinivasan, 2010). However, the one of interesting empirical research found that the big bath behavior is observed with the arrival of a new CEO, examples Pourciau (1993) and Murphy and Zimmerman (1993) that resulted is CEO turnover will affect the CEO motivation to manipulate earnings reports.
The phenomenon of management changes can be divided into routine and non-routine. Routine means that management changes performed regularly or periodically until the term ends. Non-routine management changes are sporadic or are dismissed before the term ends. In the management turnover rival, the new management appointed will perform a big bath in the first year because of the potential losses, the reputation of which is owned by the new management will not bear the debt to shift the responsibility on a compensation system that occurs at that time or others (Obinata, 2007;Tokuga & Yamashita, 2011). On the other hand, in the case of routine management changes, big bath may occur when (during the year) the resignation of the management such as the manager who resigned have the authority to propose the name of the manager for a period of the next year (Otomasa, 1997;Yamaguchi, 2009), the possibility exists that the manager who resigned will do big bath to allow his successor to make a fresh start in his first year in office. Based on these explanations, the formulation of the research questions are as follows: (1) Is the period before the routine change of management, the old management practice significant bath accounting? (2) Is the period before the non-routine change of management, old management practice big bath accounting? (3) Is the period after the routine change of management, new management practice big bath accounting? (4) Is the period after the non-routine change of management, new management practice big bath accounting?

Agency Theory
The agency theory describes that the agency relationship arises when one or more persons (the principal) employs another person (the agent) to provide a service and then delegate decision-making authority to the agent. The relationship between the principal and the agent can lead to imbalance condition information (asymmetrical information) for the agent in a position to have more information about the company than the principal. Assuming that individuals act to maximize their self-interest, then the information asymmetry that has will encourage agents to hide some information. In the www.scholink.org/ojs/index.php/ijafs conditions of asymmetry information, the agent can influence the accounting numbers presented in the financial statements caused by earnings management practice. This information asymmetry led to the possibility of conflict between the principal and the agent .
Regarding Healy and Wahlen (1999) and Leuz et al. (2003), the motivation of earnings management is as followed: "One study found that earning management occurs for a variety of reasons, including influencing the stock market, increasing management compensation, reducing the likelihood of violating lending agreements, and avoiding intervention by government regulators". The first and the second motivation relates to agency theory which states opportunistic behavior of the agent (manager) that leads to information asymmetry between managers and investors. This information asymmetry revealed managers in communicating misleading information to investors and other users of financial statements (as lender) to achieve their own goals (Utami & Nugroho, 2017;Soekapdjo et al., 2018).

Earnings Management
Earnings management is intervening management in the process of preparation of financial statements for external companies to achieve a certain level of income with the aim to benefit himself or his own business (Schipper, 1989;Kothari, 2001). According to Scott (2000) and Ahmed and Duellman (2007), "Earnings management is the choice by a manager of accounting policies, or actions affecting earnings, to achieve some specific reported earnings objective".
The reason and purpose of a manager involved in earnings management practice are followed Managers may attempt to manage earnings because they believe reported earning influences investor and creditor decisions (Schroeder, 2011). In most cases, earnings management techniques are designed to improve reported income effects and to lower the company's cost of capital. On the other hand, in a move to increase future profits, management may take the opportunity to report more bad news in periods when performance is low.
According to Scott (2000) and Cheng and Warfield (2005)  Regarding Baldwin et al. (2007), the strategies used in managing earnings: (1) Managers were increasing current period income to make the company better seen; (2) The manager did a big bath through a reduction in profit of this period. Selected periods usually a period with poor performance (often at times of recession where another company also reported earnings that bad) or current events going on an unusual occurrence such as change of management, mergers or restructuring; (3) Income smoothing is a common form of earnings management. In this strategy, manager increase or decrease in reported earnings to reduce the fluctuation. Smoothing earnings also include not report the share of profit in the period either by creating a backup or "bank" profits and then report this income when a bad period.

Big Bath Accounting
In 1998, Arthur Levitt, an SEC official, said one of the earnings management techniques that could threaten the integrity of the financial statements is Taking a bath. The one-time overstatement of restructuring charges to reduce assets, which reduces future expenses. The expectation is that the one-time loss is discounted in the marketplace by analyst and investors who will focus on future earnings (Schroeder, 2011).
"Big bath accounting profit large as reducing write-offs or income-decreasing discretionary accruals in income statements especially when a new Chief Executive Officer (CEO) is appointed, so new CEO Cleans the accounts to be able to use it in the future to smooth earnings" (Stolowy & Breton, 2004).
One type of strategy in earnings management according to Subramanyam andWild (2009) andBhandari andIyer (2013), Big Bath conducted through the removal (Write-Off) as much as possible in one period.
The selected period is usually a period of poor performance (often at times of recession where another company also reported earnings that bad) or upon a change of management, merger or restructuring.

CEO Changes
Chief Executive Officer (Top Executive Officers) is the top level of a company whose determine the direction of the company. The term CEO himself is a word derived from English as an international standard. However, in Indonesia, this may designation CEO is unusual sound, the reason is that in Directors, it will be useful as the closing of the RUPS (General Shareholders Meeting).

CEO Change Routine
CEO change routine is the CEO changeover process is performed regularly and planned. CEOs will be replaced when his term has expired or resigning before his term expires for reasons of illness or death.
Old CEO who finished his tenure will usually serve as a member of the Board of Commissioners while the new CEO usually recruited from within the company itself so automatically the old and new CEO already know each other. Routine CEO turnover data is usually obtained from the decision of the RUPST (General Shareholders Meeting Annually).

CEO Change Non-Routine
CEO change non-routine is the CEO changeover process conducted irregularly and suddenly impressed.
CEO will be replaced even though his term has not expired and is not promoted by the company. Several possibilities, for example, old CEO cause this change had a poor performance and or conduct earnings management. Old CEO who dismissed will be out of the company and will not be served as a member of the Board of Commissioners, while the new CEO usually recruited from outside of the company.
Non-routine Data CEO turnover is usually obtained from the decision of the RUPSLB (Extraordinary General Shareholders Meeting). Although in the decision of the RUPS or RUPSLB never mentioned reasons dishonorable discharge, researchers must be able to assume that if dismissed before his term ends, not promoted and was not given a new position in the internal company there must be something that is not good, or there has been hostility.

Hypothesis
In turn, hostile management, the new management appointed will perform big bath in the first year because of the hope that is despite record substantial losses, the reputation of which is owned by the management will not bear the debt to shift responsibility (Obinata, 2007). Then Pourciau (1993), which examines the level of earnings management when a change of executive non-routine found evidence that the new executive implements policies to lower revenue in the first year to increase revenue better in the next year. Regarding Jones (1991) and Dechow and Dichev (2002), an indication of earnings management is done by the new CEO with the type of earnings management that is done is a decrease in profit. Furthermore, related to Xie et al. (2003) and Peasnell et al. (2005), the new CEO served to earn management using discretionary accruals to lower earnings in turn.
In the case of management changes as well (routine), the big bath may occur when (during) the resignation of the old management as certain conditions such as the manager who resigned have the authority to propose the name of the manager for the next year (Otomasa, 1997). The possibility exists that managers who resign will do a big bath to allow his successor to make a fresh start in its first year and the new management will not conduct earnings management (Weisbach, 1995).

Figure 1. Description of Event Study Period after the Change of Non-Routine CEO
Hypothesis 1 states in the period after a change of management because of the hostility (non-routine) new management practices big bath accounting. Furthermore, the decision of taking action of the big bath because in the first year of management want to show the performance at the beginning look bad so that impact on performance in the next years will look better. In Figure 1. Notated t2, earnings management is used to reduce reported earnings. Through hypothesis testing are expected in the years after the CEO Change non-routine occurs of earnings management that is used to lower profits.

Figure 2. Description of Event Study Period before the Change of Routine CEO
Hypothesis 2 states the period before the change of management routine that the old management practice big bath accounting because of the change is done amicably so that management hopes that his successor can begin the next performance look better. In Figure 2, the year before his term ends notated t1. Through routine testing of CEO turnover table, t1 expected their profits to decline.

Figure 3. Description of Event Study Period after CEO Turnover Routine
Hypothesis 3 states that in the period after a change of management amicably (routine) new management does not practice big bath accounting that reflects on Figure 3. Therefore, hypothesis 3 provide proof that the CEO turnover amicably so that the new management would not do earnings management (big bath) to disfigure the previous management so that the third hypothesis testing is expected in the period after a change of routine will not be detected earnings management.  Hypothesis 4 states that in the period before the hostile management changes (non-routine), the previous management did not practice big bath accounting that reflects on Figure 4. Therefore, hypothesis 4 want to provide proof of the cause of CEO turnover non-routine, whether because of poor performance or caused by another factor so that the hypothetical test 4, if the CEO is replaced because of poor performance or caused by another factor, it is expected that in the period before non-routine change of CEO will not be detected earnings management.

Research Design and Methods
This type of research used in this study is a comparative descriptive, which aims to explain or describe  The company has a complete data related to total assets, cash, part of long-term debt maturing within one year, depreciation expense, current assets, fixed assets (property, plant, equipment), current liabilities, receivables, sales, and net income.

Variable Manipulation Accrual
Accrual basis is the basis used by the accounting to measure the performance of companies in the form of profit. One of the models that are used to detect the earnings management in the accounting literature is the total ratio accrual of sales (Dechow et al., 1995;Leuz et al., 2003). The reason for using the accrual of working capital is more appropriate as Peasnell et al. (2005) has reviewed that the error residuals did not estimate discretionary accrual because the technique is considered to be relatively complex, and therefore it is used ratio accrual of working capital to sales. The reason sales as working capital accruals deflator is due to earnings management occur in many sales accounts as revealed by Nelson et al. (2002).
The steps in calculating working capital accruals are as follows: 1) Calculating Total Accrual = Δ Current Assets -Δ Debt Current -Δ Cash + Δ section Long Term Debt maturing within one year -Depreciation.
2) Earnings Management is the value of total accruals divided by Sales

Methods of Analysis and Testing Hypotheses
This research used quantitative descriptive analysis with the aim to describe the phenomenon of earnings management around the turn of the CEO. The author uses a variable earnings management models to compare earnings management practices before and after the turnover of the CEO of both routine and non-routine.

Hypothesis Testing Different Test
After determining the sample, counting the total accruals and earnings management, the next step is to test the hypothesis by using Paired Sample t-Test. Hypothesis 1 is stated be supported if the results of the t count for the model in the period t2 in routine CEO turnover greater than t table and significant.
Hypothesis 2 is stated be supported if the results of the t count for the model in the period t1 non-routine CEO turnover greater than t table and significant. Hypothesis 3 is stated be supported if the results of the t count for the model in the period t2 non-routine CEO turnover greater than t table and significant.
Hypothesis 4 is stated be supported if the results of the t count for the model in the period t1 non-routine CEO turnover greater than t table and significant.

Results and Discussion
The authors calculated the value of TA in one year before the change of the CEO (TA t-1 ) and one year after CEO change (TA t+1 ) either the change of routine and non-routine using the Microsoft Excel program. Then, to test the hypothesis using IBM SPSS Statistics 20. The focus of this research is the accrual ratio of working capital to sales for representing earnings management because a lot of earnings management occurs on account of sales.
In general, accrual, which are accounting products, can be considered to have the number of the "relatively fixed" from year to year because the accounting rules also related unchanged. Accrual changes that occur, therefore, be regarded as abnormal. This change is the result of use policy (discretion) management excessive, and when at the same time management also have a motive to manipulate earnings accrual then the change that occurs is considered as a form of earnings manipulation that did by the management. The average total accruals before non-routine CEO turnover is higher than after the non-routine change of CEO. The average value of total accruals for non-routine before the change 64,015 and after the change is -49,335. Minimum value before the turn of the non-routine at -978,246 and a maximum value of 1.619,000, while the value of non-routine minimum after the turn of -302,673 and to a maximum of 257,000. Negative accruals derived from the increasing changes in current liabilities and the amount of depreciation thus, the ratio of the change in current assets is smaller than the change in current liabilities.
The amount of change in current liabilities arises due to the result of the purchase of goods for investment in the short term which would increase the value of current assets. Current liabilities also arise as a result of the recognition of the number of loads increases to minimize earnings. Negative accruals are also sourced from the depreciation expense is so large compared to changes in long-term debt presented in current liabilities.
In the period after the management changes because of the feud (non-recurring) new management has the potential of big bath accounting practices, seen from the ratio of the total accrual to the company's sales Karya Unggul Tbk, which has total assets on average the smallest among the sample of about Rp49 billion, and PT Astra International, Tbk, which made the turn because its CEO dies. Regarding the analysis, the author can be told that this has been done in the first year of management wanted to show a poor performance for the performance in subsequent years look better and earnings management techniques used is by lower the earnings through the recognition of the increasing burdens.
In the period before the management changes because of the feud (non-routine), the old management did not have the potential to make a big bath accounting practice. The table above shows that of the 14 companies there are no companies that potentially dominant big bath accounting practices. One of them turns because there is CEO who died as PT Astra International Tbk. However, in Total Accrual before the turn of the non-routine average of positive Rp 64 billion greater than after the turn of the average minus Rp 49 Billion. This means that the larger the company's management recognizes the component of accruals, the higher the potential for earnings management may also have been a factor the company carries out the RUPSLB to replace the management who have been earning management by increasing profit to make it look good performance to get the bonus.
The average total accruals before CEO change routine greater than after CEO change routine. The average value of total accruals before the turn of the routine of -2,717 and after the turn of -189,931.
Minimum value before the turn of the routine of -2,883,304 and a maximum value of 3,556,777, while the minimum value after a change of routine for -5,248,772 and to a maximum of 1,975,249.
In the period before the turn of the peaceful management (routine) the old management has the potential of big bath accounting practices. Of the 34 samples of the company being taken, 20 companies showed view was a decline in profit before a change of routine management. According to the authors, this is because of the change is done amicably so that the old management hopes that his successor can begin the next performance look better with the potential to help the big bath accounting practices at the end of his term.
In the period after the turn of the peaceful management (routine) new management is not potentially big bath accounting practices. From the table above that, the new management is doing a potential decline in profits through increased burdens but not as big as before the change. Before the change of -0.066 and after the change into -0.036. Through each worth minus but the new management continued to show a better performance than the previous management. Researchers considered that the new management would demonstrate its performance is getting better over time slowly to get the bonus. .000

Hypothesis Testing with Different Test
.673 According to the Table 1 above it can be seen that the sample of 14 companies, the average earnings management before non-routine CEO turnover amounted to 0.153 with a standard deviation of 0.688 and standard error of the mean 0.183. While non-routine after CEO turnover average earnings management increased by 0.518 with a standard deviation of 2.122 and standard error of the mean is 0.567. Based on these results it appears that the average earnings management before the non-routine change of CEO is smaller than after the non-routine change of CEO and both are positive.
In routine CEO turnover obtained results mean earnings management before the change of -0.066 with a standard deviation of 0.236 and standard error of the mean 0.040. Then after a routine CEO turnover obtained results mean -0.036 with a standard deviation of 0.193 and standard error of the mean 0.033.
Based on these results it appears that both earnings management is negative before and after the change of CEO routine. However, after the turn of the CEO routine, the mean of earnings management looks better because the negative value was getting smaller than before the change of CEO. Correlations table shows how strong and the relationship between pairs of variables. Paired Samples Correlations based on the results above it can be seen that the CEO turnover nonroutine has a healthy relationship between earnings management before the change of the CEO with earnings management after the change of CEO. It can be seen from the correlation of 0.878 and a significant level of 0.000. That is real, and there is a healthy relationship between earnings management before the non-routine change of CEO with the earnings management after non-routine change of CEO.

Results of Paired Samples
While the results of Paired Samples Correlations on CEO turnover routine there is no substantial relationship between earnings management before the change of CEO and earnings management after the change of CEO. It can be seen from the weak correlation of 0.075 with a significant level of 0.673.
This means there is no healthy relationship between earnings management before CEO turnover routine with earnings management after CEO turnover routine and not real. before CEO turnover routine with earnings management after CEO turnover routine. Then, from the value of significance (2-tailed) of 0.563 and this is greater than 0.05, it indicates that Ho is acceptable so that it can be concluded that there was no significant difference between earnings management before CEO turnover routine with earnings management after CEO turnover routine.

Conclusion
Based on the calculation above, it can be concluded that there is potential for earnings management of big bath accounting model around the change of the CEO, but there was no significant difference between earnings management of big bath accounting model before the change of the CEO with earnings management of big bath accounting model after change of the CEO both conducted routinely and non-routine.

Suggestion
Furthermore, in this study showed a potential means most likely to occur, especially at the turn of the CEO of non-routine, but less strong evidence found in this study probably due to the different characteristics of each company as well as the condition of the succession is not the same so that it reflects conditions general economic differently. Then that becomes the other reason is the limited sample of companies lacking a lot due to the events of CEO turnover is not many done by the company in the period of observation and limited to manufacturing companies. The result is expected to be a reference for further research, especially to reveal better their potential earnings management practices focus on change of CEO non-routine by extending the scope of the study sample area that is not only limited to the manufacturing company.