CEO Turnover and Firm Performance In Indonesia

We investigated the effect of changes in CEO position on subsequent firm performance by studying 91 CEO turnovers in Indonesia. Our results show that firm performance decreases during the turnover year. Moreover, the incoming CEO does not increase firm performance in subsequent years. Indeed, there is evidence that firm performance decreases after such turnovers. We ultimately conclude that CEO turnovers in Indonesia do not have a positive effect on firm performance. Going further, we divided CEO turnovers into routine and non-routine turnovers on the basis of the turnover process. Both routine and non-routine CEO turnovers show similar results with all samples, in which the incoming CEO in a routine or non-routine turnover does not have a positive effect upon firm performance. Further evidence suggests that the incoming CEO tends to upsize firm assets rather than downsize them.


Introduction
A Chief Executive Officer (CEO) is a person who holds the highest executive position in a firm.The CEO plays a central role in formulating and implementing a firm's mission and vision, developing firm strategy to achieve shortterm and long-term objectives, allocating firm resources and investment decisions (Canals, 2010).Therefore, the CEO has significant influence on firm strategies (Miller, Le Breton-Miller, & Lester, 2011) and policies; financial, operating, and investment decisions; (Bertrand & Schoar, 2003;DeJong & Ling, 2013), accounting choices (DeJong & Ling, 2013); and ultimately firm performance (Mackey, 2008).The CEO is expected to deliver good performance.If a CEO fails to deliver good performance, the board may implement a disciplinary measure.
Several previous studies show that the probability of CEO turnover is negatively correlated with firm performance in the US (Chakraborty & Sheikh, 2008;Huson, Malatesta, & Parrino, 2004;Wang, Davidson III, & Wang, 2010).In emerging markets, some studies have also found a negative relationship between firm performance and the probability of a CEO turnover (for example, Fan, Lau, & Young, 2007;andKato &Long, 2006a and2006b for China;and Hou & Chuang, 2008 in Taiwan).CEOs face higher probability of being sacked when they cannot deliver good performance.The replacement of a CEO is an important event for a firm.Investors expect that incoming CEO would improve the firm's performance (Denis & Denis, 1995;Huson et al., 2004).Denis and Denis (1995) investigated the impact of new CEOs upon firm performance in the US.They showed that subsequent firm performance improved after the change, or in other words the replacement of the CEO has a positive effect on firm performance.Furthermore, they found that new CEOs tend to downsize firm assets and employee numbers.However, existing studies on CEO turnover and firm performance have mostly been conducted in developed countries.The difference in institutional settings may lead to the contention that empirical results from developed countries cannot be generalized to developing countries.Brown, Beekes, and Verhoeven (2011) suggested that studies should be carried out in various countries to understand the disparity in local contexts.
This present paper re-examines the impact of CEO turnover on firm performance in Indonesia under the influence of its distinct characteristics.First, Indonesia makes use of the two-tier board system.Second, equity ownership structures are mostly concentrated in family hands.Each company has two boards: the board of directors and the board of commissioners.The Board of directors (BOD) is responsible for the (executive) management of the firm, while the board of commissioners (BOC)1 performs the monitoring function and provides advice to the BOD.The CEO in Indonesia is the coordinator of the board of directors.The members of the BOD are not allowed to hold any position in the BOC, which means that the BOC must be independent from the BOD.This reduces the opportunity for a conflict of interest so that the BOC can perform its duties properly in enhancing the company's performance.If a CEO fails to show adequate performance, the BOC can suspend him/her and ask for him/her to be replaced the general meeting of shareholders.
Few studies have been done in the Indonesian context, such as Lindrianasari andHartono (2011), andSetiawan, Phua, andChee (2013).Lindrianasari and Hartono (2010) found a negative relationship between firm performance and the probability of CEO turnover.It could be concluded that Indonesian CEOs who do not perform well have a higher probability of dismissal.This result indicates that BOCs are properly carrying out their task in disciplining CEOs with poor performance.Setiawan et al. (2013) investigated how Indonesian investors react to the announcements of CEO turnover.They show that investors perceive CEO turnover as good news and react positively.However, these studies did not analyze the effect of the incoming CEO on firm performance.
Going deeper, we also consider the distinction between routine and non-routine turnovers.Arguably, incoming CEOs from non-routine turnovers are under greater pressure to deliver good performance for having replaced a poorlyperforming CEO.Kang and Shivdasani (1995) found that non-routine turnovers in Japan increase firm performance while routine turnovers do not.Denis and Denis (1995) revealed that both the turnover processes increase firm performance.On the other hand, Dahya, Lonie, and Power (1998) did not find any significant effect from routine turnover upon firm performance.Therefore, it is important to investigate how the turnover process affects subsequent firm performance.

Literature Review
The CEO is an important figure in a firm.Mackey (2008) found that CEOs have significant effect on firm performance.The CEO is expected to improve the firm's performance.CEOs work under a reward and punishment system where high-performing CEOs earn bonuses while poorly-performing CEOs are disciplined by shareholders through the boards.Therefore, CEOs have an incentive to achieve higher performance.
A CEO with a poor performance must face a discipline mechanism, and in the worst case he/she can be sacked from his/her position.This means that CEOs with poor performance have a higher probability of dismissal.Most empirical evidence confirms the negative re-lationship between performance and the probability of CEO turnover (for example, Huson et al., 2004;Wang et al., 2010).Wang et al. (2010) studied 670 CEO turnovers in 1999-2005 in the US to investigate the relationship between poor performance and the probability of forced CEO turnovers.They found a negative relationship between firm performance and probability of forced CEO turnover.This result is in line with Huson, et al. (2004), who found that firm performance tends to be lower before CEO turnovers.Weisbach (1988) also found an inverse relationship between prior performance and probability of CEO turnover.Engel, Hayes, and Wang (2003) revealed that the board of directors employ market-based and accounting-based information to evaluate CEO performance.Poor performance, whether market-based or accounting-based, increase the probability of CEO turnover.Lausten (2002) investigated the relationship between firm performance and CEO turnover in Denmark by studying 243 firms and 77 turnovers over the 1992-1995 period.This study found that firm performance is negatively associated with the probability of CEO turnover.This result is in line with Neumann and Voetmann (2005) who found and inverse relationship between firm performance and probability of CEO turnover.Pope and Florou (2008) investigated the link between firm performance and CEO turnover in the 300 largest UK firms from 1990 -1998.They showed that a CEO faces a higher probability of being fired if his/ her performance is poor.This result is consistent with Conyon and Florou (2002) who found that CEOs are highly likely to be dismissed if they performs poorly.Defond and Hung (2004) undertook a multicountry study that showed a negative relationship between firm performance and CEO turnover.However, they only found this negative relationship in countries with strong corporate governance practices, and not in countries with poor corporate governance.This result shows that CEOs in countries with poor corporate governance can entrench themselves in their positions.Gibson (2003) showed that CEOs with poor performance in nine emerging markets have a high probability of being fired, which means that CEOs also faces high probability of turnover for bad performance in emergingmarket firms.
In recent years many papers have discussed the relationship between firm turnovers and CEO turnovers in China such as Chi and Wang (2009), Fan et al. (2007), Firth, Fung andRui (2006), Kato andLong (2006a, 2006b), and Pi and Lowe (2011).Although China has different governance characteristics from western countries, most previous studies confirm the negative relationship between firm performance and CEO turnovers (e.g.Fan et al., 2007;Firth et al., 2006;Pi & Lowe, 2011), except Chi and Wang (2009) who did not find any significant effect from firm performance upon top executive turnovers.These results show that CEOs in China faces higher probability of dismissal if firm performance is poor.A negative relationship between firm performance and CEO turnover is also found in Indonesia (Lindrianasari & Hartono, 2011;Martani & Tarigan, 2009).Lindrianasari and Hartono (2011) found that poor performance increases the probability of CEO turnover.Martani and Tarigan (2009), using state-owned enterprises (SOE) as their sample, also documented that poor performance increases the probability of forced director turnover.
It has been explained above that incoming CEOs are expected to improve their firms' performance.Denis and Denis (1995) found evidence that incoming CEOs have a positive impact on firms.There is a significant improvement in firm operating income after the change of CEO.This results is confirmed by Kato and Long (2006a), who found that CEO turnover in China have a positive effect on firm performance.The incoming CEOs are often able to improve their firms' performance.Salomo and Leker (2000) found that CEO turnovers in Germany have a positive effect upon firm performance in the subsequent year.This evidences shows that CEO turnovers have a positive effect upon firm performance.However, Lindrianasari and Hartono (2011) found that CEO turnovers in Indonesia do not have any effect upon firm performance.There is no difference in ROAs before and after the CEO turnover.Previous studies show that CEO turnover improves firm performance (Denis & Denis, 1995;Kato & Long, 2006a;Salomo & Leker, 2000), except Lindrianasari and Hartono (2011).Thus, we would expect CEO turnovers in Indonesia to improve firm performance.H1: Firm performance is improved after CEO turnover event.
We also investigate the effect of whether the turnover is routine or non-routine upon firm performance.Denis and Denis (1995) found that both kinds of turnovers resulted in an improvement of firm performance.On the other hand, Lindrianasari and Hartono (2011) did not find any significant effect from either routine or non-routine turnovers upon firm performance.Salomo and Leker (2000) found that involuntary turnovers have a positive effect on firm performance.CEOs with poor performance have a higher probability of being fired, so incoming CEOs in non-routine turnovers face greater pressure to achieve higher performance.H2: Firm performance is improved after routine/non-routine CEO turnover event.
Weisbach (1995) argues that one of the reasons investors pay attention to CEO succession is because it signals a change in future corporate decisions.When incoming CEOs attempt to increase their firms' efficiency and effectiveness through restructuring, they may take radical action to reduce the number of employees and total assets.Incoming CEOs try to reduce employee numbers if employee costs are high enough to compromise their firms' competitiveness in the market.Christoph Mueller was hired as CEO in Malaysia Airlines (MAS) on May 1, 2015 when MAS was struggling for survival due to its poor financial condition and the unprecedented tragedy of losing two aircraft in the same year.Christoph Mueller dismissed 30% of MAS employees, restructured the firm's assets, and made other strategic decisions to improve the firm's performance.As a result, MAS turned to profitability in early 2017.This is a good signal for MAS.This case shows that incoming CEOs can conduct firm restructuring to improve firm value.Chiu et al. (2016) provided evidence that incoming CEOs tend to engage in corporate divestiture.Incoming CEOs from inside the firm focus on higher scale of divestiture while incoming CEOs from the outside focus on higher scope of divestiture.Therefore, it is important to investigate how incoming CEOs manage the number of employees and total assets.Weisbach (1995) found that incoming CEOs tend to review their predecessors' investment decisions.Incoming CEOs prefer to sell unprofitable investments to increase firm value, thus restructuring their firms after their appointment as the new leader.Denis and Denis (1995) provide evidence that incoming CEOs engage in downsizing of the firm.The incoming CEOs reduce the number of employees, total assets, and capital expenditure.This downsizing decision is expected to improve efficiency and effectively enhance firm performance.On the other hand, Huson et al. (2004) showed that company assets tend to grow three years after CEO turnovers compared to one year before CEO turnovers and that firms spend more on capital expenditure.This result contradicts Denis and Denis (1995) and Weisbach (1995).By studying Indonesian firms, Lindrianasari and Hartono (2011) found that total assets after CEO turnover are marginally greater than before CEO turnover.It means that incoming CEOs in Indonesia tend to upsize total assets rather than downsizing them.Dahya, Lonie, and Power (1998) revealed different behaviors between incoming CEOs from routine and non-routine turnovers.They found that incoming CEOs in routine turnovers tend to increase their firms' assets while new CEOs in non-routine turnovers reduce firm assets.On the other hand, Denis and Denis (1995) provided evidence that incoming CEOs from both routine and non-routine turnovers downsize firms, while Huson et al. (2004) showed that new CEOs from both routine and non-routine turnovers spend more on capital expenditures and increase total assets.Since previous studies have produced inconclusive results about the effect of CEO turnovers upon total assets, we propose a non-directional hypothesis.H3: Total assets after CEO turnover are differ-ent from total assets before CEO turnover.H4: The number of employees after CEO turnover is different from the number of employees before CEO turnover.

Research Methods
Our sample consists of CEO turnovers in all Indonesian publicly traded firms over the 1994 -2002 period.This study follows Denis and Denis (1995) and Kato andLong (2006a, 2006b) who used all publicly listed firms in the US and China respectively in their studies.They investigated the effect of CEO turnover on firm performance.As explained above, Indonesia has a two-tier board system: Board of Directors and Board of Commissioners.The CEO is equivalent to the chairman of the Board of Directors.In Indonesia, the CEO is also commonly known as the Direktur Utama (Dirut) or the Presiden Direktur (Presdir).We identify CEO turnovers by comparing the names of CEOs in two consecutive years according to the Indonesian Capital market Directory (ICMD).A CEO turnover is deemed to have taken place if the names are different.We also distinguish between routine and non-routine CEO turnovers by browsing relevant articles or news items in the authoritative Indonesian newspapers: Kompas and Neraca.When we cannot find any relevant news about CEO turnover, we follow the work of Kang andShivdasani (1995, 1996) where a turnover is deemed routine if the departing CEO still holds a position in the BOD.Since Indonesia has a two-tier board system, we also consider the turnover routine if the departing CEO still holds a position in the BOC.Otherwise, we define the CEO turnover event as non-routine.
Our analysis covers a five-year period around the turnover event (t-2 to t+2), t being the turnover year.Since publicly traded firms in Indonesia end their financial year on 31 December and they must publicize turnover events within 3 months of that date, we decided that 30 March is the cut-off point for determining t.The T-test is employed to test the hypotheses.If firms experience more than one turnover event in the 5-year period, the study will use the most recent one.This study compares firm performance before and after CEO turnovers.To measure firm performance, we use the accounting performance measures of Return on Assets and Return on Equity (Chen, Cheng, & Dai, 2013;Kato & Long, 2006a, 2006b).In addition, we use total assets and total number of employees to investigate the effect of CEO turnover on firm restructuring (Denis & Denis, 1995).

Descriptive Statistics
Our sample consists of 91 CEO turnovers in Indonesian publicly traded firms during the 1994 -2002 period.In terms of the turnover process, there are 47 routine CEO turnovers and 44 non-routine CEO turnovers.Table 1 presents the descriptive statistics of variables.
As shown in Table 1, ROA before turnover year tends to be positive on average.However, it tends to turn negative in the turnover year.This shows that firm performance decreases during the turnover year.Two years following the turnover, the ROA usually improves even Note: All = all CEO turnover, Routine = routine CEO turnover, Non-routine = non-routine CEO turnover N all sample = 91, N routine turnover = 47, N non-routine turnover = 44 ROA = return on assets, ROE = return on equity, TA = total assets, Empl = number of employees though it remains negative.This pattern holds for both routine and non-routine CEO turnovers.In the table of descriptive statistics, it is shown that firms with CEO turnover have positive ROE in t-2.However, it decreases in t-1 and t.ROE t-1 and ROE t are negative.Further, ROE increases in the year after the turnover but remains negative.This result is similar to the ROA trend.Firms that experience routine turnover have lower ROEs in the post-turnover years.However, firms with non-routine CEO turnover show a different pattern.There is positive ROE in the two years following turnover.This pattern shows that there is an improvement in firm performance after non-routine CEO turnovers.A detailed discussion on firm performance is provided in the next section.

Empirical Results
The results of hypothesis testing are presented in Table 2. Table 2 shows that the ROA in the turnover year is less than the ROA before the CEO turnover, which means that firm performance decreases in the turnover year.However, comparison between ROAs before the CEO turnover and after the CEO turnover shows that there is no significant difference between ROAs before and after the CEO turno-ver.The incoming CEOs do not have a positive effect on firm performance.Furthermore, the comparison between ROA t+1 , ROA t+2 and ROA t also produce no significant results.As shown in Table 2, there are significant differences between ROA t-2 and ROA t+1 , and between ROA t-2 and ROA t+2 .However, firm performance after the CEO turnover is lower than before the CEO turnover.This result shows that the incoming CEOs do not improve their firms' performance; the performance of incoming CEOs is worse compared to the departing CEOs.
We can see in Table 2 that the comparison between ROEs before and after the CEO turno-ver shows similar trends with the ROA.ROE t-2 and ROE t-1 are higher than ROE t .It can be seen that firm performance decreases in the turnover year.Further, the comparison between ROEs before after the CEO turnover does not produce any significant results.These results show that incoming CEOs do not have a positive effect on firm performance.The comparisons between ROE t and ROE t+1 and between ROE t and ROE t+2 do not exhibit any significant results either.Firm performance after CEO turnover does not differ from the turnover year and before CEO turnover.According to the statistical test as shown in turnover.We therefore reject H1.
Going deeper by distinguishing routine from non-routine CEO turnovers produces similar results.In firms that experienced routine turnover, their ROA t+1 and ROA t+2 are lower than ROA t-2 , ROA t-1 , and ROA t .These results show that incoming CEOs in routine turnovers do not have a positive effect on firm performance.Indeed, incoming CEOs from routine turnovers tend to have lower performance than departing CEOs.Table 1 shows that firms with routine turnovers do not have higher ROEs.The comparison between performance before turnovers (ROE t-2 , ROE t-1 ) and after turnovers (ROE t+2 , ROE t+1 ) exhibits that CEO turnovers do not produce any significant results.There is no improvement on firm performance after routine turnover.Therefore, H2 is rejected.
As seen in Table 2, total assets increased during the 5 years under observation.There is a significant increase in total assets in the turnover year compared to before the turnover.Further analysis shows that firms experiencing CEO turnover have more assets after the turnover year.This result shows that incoming CEOs upsize firm assets.A scrutiny of the effects of routine and non-routine CEO turnovers upon total assets also shows consistent findings.Both routine and non-routine CEOs turnovers lead to an increase in assets following the turnover.Thus, we conclude that H3 is supported.
As can be seen in Table 2, the number of employees after CEO turnovers does not differ from the number of employees before CEO turnovers.These results provide evidence that incoming CEOs seldom make any significant decision about the number of employees, whether laying off employees to reduce expenses or aggressively hiring more employees.Routine and non-routine CEOs turnovers exhibit similar trends in this regard; there is no significant difference in the number of employees before and after CEO turnovers in either case.However, non-routine CEO turnovers lead to reduced number of employees in the two years following the turnover.Thus, non-routine CEO turnovers tend to lead to significant decisions to reduce the number of employees.

Discussions
Our empirical results reveal that CEO turnovers do not have any positive impacts upon firm performance.Indeed, CEO turnovers tend to lead to a decrease in firm performance.This differs from the findings of Denis and Denis (1995), Huson, et al. (2004), and Kato and Long (2006a), who found that firm performance increases after turnover events.Our findings, however, are somewhat in line with Lindrianasari and Hartono (2011) who found that CEO turnovers do not have a positive outcome on post-turnover accounting performance.Likewise, Ahn, Bhattacharya, Jung, and Nam (2009) documented that firm performance in Japan is more likely to decline after turnover events.Our findings are also in line with Garcia, Gonzalez, and Ortega (2006) who found a negative impact from CEO turnover upon Venezuelan firms.
Going deeper, we found that CEO turnovers have a positive effect on firm assets.There is a significant increase in firm assets after CEO turnovers, which means that incoming CEOs tend to upsize firm assets.This is consistent with Huson et al. (2004) who found that incoming CEOs make significant decision in increasing firm assets.
Nevertheless, we have not confirmed that incoming CEOs make significant decisions to lay off or hire employees.There is no significant difference in the number of employees before and after CEO turnovers.This is not consistent with Denis and Denis (1995) who found that incoming CEOs reduce the number of employees to reduce firm expenses.This result may be due to the relatively strict employment regulations in Indonesia, which hinder incoming CEOs from aggressively upsizing or downsizing the number of employees.
The results of this study provide interesting results in the context of how incoming CEOs engage in firm restructuring.Instead of engaging in downsizing total assets, the incoming CEOs upsize firm assets.In other words, incoming CEOs prefer to make the firm larger.This result does not support previous research findings, such as Denis and Denis (1995) and Weisbach (1995), who find that incoming CEOs prefer to downsize firs.However, the result of this study is in line with Huson et al. (2004) who found that incoming CEOs use their discretion to increase firm performance.

Robustness Checks
We perform robustness checks to ensure that our findings are consistent.Instead of unadjusted ROA, ROE, total assets and total employee numbers, we use industry-adjusted ROA, ROE, total assets and total employees.We use industry-adjusted data to control the effect of industry trends on changes to firm profitability and firm assets.The results are consistent with regards to our main variables.

Conclusions
We have shown that CEO turnovers in Indonesia do not have any positive effects upon firm performance and some firms that experience CEO turnovers have lower performance than before the turnovers.These firms' performance declined after the turnover.This result is consistent for both routine and non-routine turnovers.Moreover, we found that incoming CEOs tend to upsize firm assets, given that our empirical results display a significant increase in total assets.However, no significant effect is found on the link between CEO turnovers and employee numbers.
The implications of this study are: first, investors need to give more attention on the succession events.CEO succession is an important event in any company since it will have significant effects upon the firm's future condition.This study shows that firm performance decreases after CEO turnover events.The incoming CEO does not have any positive impact on firm performance.However, as stated by Hillier and McColgan (2009), there is a difference in behavior between family CEOs and non-family CEOs.This study has not analyzed the effect of family ownership on the CEO turnover event.
Previous studies show the importance of family ownership in Indonesia on company decisions such as dividend payment (Setiawan et al., 2016).This is a suggested area of research for future studies.Second, the results of this study show that incoming CEOs tend to increase the firms' assets.Incoming CEOs prefer to use firm resources to increase firm assets rather than reducing them.
This paper uses the T-test to analyze differences in firm performance between the pre-and post-CEO turnover periods.This technique may fail to capture the possible endogeneity between turnover and performance limitation.However, we could not tackle this issue with a different technique due to limited availability of data and difficulties in defining the control group.Therefore, it would be appropriate for future studies to consider the endogeneity between firm performance and turnover.

Table 1 ,
firm performance (in terms of ROA and ROE) does not improve after CEO D.