Governance with Multiple Objectives Evidence from Top Executive Turnover in China By Eric C. Chang And Sonia Wong.

Governance with Multiple Objectives Evidence from Top Executive

Turnover in China

By

Eric C. ChangAnd

Sonia Wong

1. Introduction1. Introduction

Purpose of this Study

Examine the relations between CEO turnover and firm performance in China’s listed firms.

Motivations

The turnover-performance relation serves as an indicator of the effectiveness of corporate control over managers.

With China’s accession to the WTO and its expansion of capital market, the motivations and incentives of shareholders of listed companies to exercise effective corporate control should be of great interest to policy makers and global investors.

1. Introduction1. Introduction

Past Studies on Turnover-Performance Issues

Financial performances and the likelihood of turnovers

• Jensen and Warner(1988): Top management turnover is inversely related to stock price performance.

• Huson, et al.(2001): The nature of CEO turnover has changed over time, but the relation between the likelihood of forced CEO turnover and firm performance did not changed significantly.

• Kaplan(1994): Consistent with US evidence, fortunes of Japanese top executives are positively correlated with stock performance and cash flows.

1. Introduction1. Introduction

Past Studies on Turnover-Performance Issues

Comparisons of performances prior to and after turnovers

• David J. Denis & Diane K. Denis (1995): Forced resignations of top managers are preceded by significant performance declines and followed by large performance improvements.

• Parrino(2003): Aggregate institutional ownership declines in the year prior to forced CEO turnover.

• Huson, et al.(2004):Turnover announcements are associated with positive abnormal stock returns, which are in turn significantly positively related to subsequent performance.

1. Introduction1. Introduction

Private Shareholders vs. State Shareholders Past studies focus on firms controlled by private

shareholders This paper examine China’s listed firms controlled by state-

owned shareholders Specialties of State Shareholders

• not real owners but bureaucrats

• tend to use state-owned firms to serve multiple social and political objectives

• are plagued by agency problems

• may have incentives different from those of private shareholders in maximizing profit and disciplining managers

1. Introduction1. Introduction

Objectives Integration

Jensen(2001): It is impossible to maximize more than one objective when tradeoffs exist.

State shareholders need to assign weights to various objectives and integrate them.

Expectations on Firms with Multiple Objectives

Current firm performance affects performance weight.

Current firm performance affects discipline incentive.

Relative importance of performance in objective functions affects post-turnover performance changes.

1. Introduction1. Introduction

An Overview of Results

There is a significant negative relation between pre-turnover profitability level and CEO turnover in loss-making firms, but not in profit-making firms.

Post-turnover profitability improves significantly in loss-making firms, but not in profit-making firms.

Implication of the Results

Turnover-performance relations are different in loss-making firms and in profit-making firms.

State shareholders decide their incentives to discipline CEOs according to firms’ performance.

Objective functions are time-varying.

1. Introduction1. Introduction

Additional Contributions

offers empirical evidence on the monitoring activities of state shareholders, based on a large sample of partially privatized public listed firms.

provides evidence that shareholders rely more on average performance over a CEO’s tenure rather than annul performance to evaluate their CEOs.

shows that loss-making firms and profit-making firms have different extents of post-turnover performance improvements.

2. Research Background2. Research Background

Governance Structures in China’s Listed Firms

Nearly all listed firms in China’s stock market are spin-offs of large state-owned enterprises (SOEs).

State asset management agencies and parent SOEs retain dominant control and act as state shareholders.

Although listed firms have formal modern governance structures, local governments maintain control in them.

CEO turnovers are controlled by governments through their personnel appointing authority.

State shareholders’ incentive structures influence CEO turnovers in China’s listed firms.

2. Research Background2. Research Background

State Shareholders’ Behaviors Under Different Incentive Views

Stewardship View

• State shareholders direct listed firms to promote the government’s social and political goals, and therefore has weak profit-maximizing incentives.

• Loss-making firms would put higher weight on performance to avoid eroding resources available for the government’s social and political plans.

• Stewardship view implies that state shareholders of loss-making firms would have a higher incentive to discipline managers on the basis of performance.

2. Research Background2. Research Background

Self-dealing View

• State shareholders use state-owned firm resources to promote their own interests, and therefore also has weak profit-maximizing incentives.

• State shareholders of loss-making firms may place a higher weight on firm performance to increase resources available for their self-serving purposes.

• Self-dealing view implies that shareholders of loss-making firms have higher incentive to discipline CEOs on the basis of firm performance than shareholders of profit-making firms.

2. Research Background2. Research Background

Bureaucrat View

• State shareholders concern themselves ultimately on their job security and prospects.

• Bureaucratic decisions tend to be slow and pressure-driven to maintain prudence, conformity and consensus-building. (Merton,1940; Fligstein,1987)

• Being a lagging and comprehensive indicator of a CEO’s performance, average performance is more in line with state shareholders’ desire for prudence and conformity.

• Bureaucrat view implies that shareholders of loss-making firms rely more on average rather than annual performance to evaluate their CEOs.

2. Research Background2. Research Background

Determinants of Performance Changes Following Turnover

State shareholders’ profit-maximizing incentives would affect post-turnover performance changes.

Turnovers of loss-making firms are more likely to be followed by large improvements, as shareholders have higher incentives to discipline their managers on firm performance basis.

The story is on the opposite side for profit-making firms, as shareholders of such firms enjoy greater latitude to serve social and personal goals.

3. Data, Sample Selection, 3. Data, Sample Selection, and Research Methodand Research Method

Data Sources

Based on all firms listed by the Shanghai and Shenzhen Stock Exchanges from 1995 to 2000.

Obtain data on CEO turnovers from CCGRD.

Stated reasons for these turnovers are provided by CCGRD and can be classified into 11 categories.

3. Data, Sample Selection, 3. Data, Sample Selection, and Research Methodand Research Method

1995 1996 1997 1998 1999 2000 1995-2000

Total number of CEO turnovers 47 81 136 210 273 330 1077

Number of CEO turnovers afterconsolidation 44 80 130 196 254 301 1005

Annual turnover rate afterconsolidation 14% 16% 18% 24% 27% 29% 23%

25% 30% 32% 25%Annual turnover rate 15% 16% 19%

826 924 1033 4328Number of listed firms 311 514 720

Table 1

Annual CEO Turnover Rate and Performance in China's Listed Firms: 1995-2000

3. Data, Sample Selection, 3. Data, Sample Selection, and Research Methodand Research Method

NumberPercentageof Sample Number

Percentageof Sample

1. Change of job 336 31.20% 305 30.40%

2. Retirement 33 3.10% 31 3.10%

3. Contract expiration 209 19.40% 205 20.40%

4. Change in controllingshareholders

76 7.10% 74 7.40%

5. Resignation 110 10.20% 103 10.30%

6. Dismissal 51 4.70% 42 4.20%

7. Health 34 3.20% 32 3.20%

8. Personal reasons 3 0.30% 3 0.30%

9. Corporate governance reform 178 16.50% 169 16.80%

10. Legal disputes 10 0.90% 9 0.90%

11. No reason given 24 2.20% 21 2.10%

12. Completion of acting duties 13 1.20% 11 1.10%

Total number of observations 1077 100.00% 1005 100.00%

Full Sample Consolidated Sample

Table 2

Stated Reasons for CEO Turnover in China’s Listed Enterprises

3. Data, Sample Selection, 3. Data, Sample Selection, and Research Methodand Research Method

Sample Selection: Forced vs. Non-Forced Turnovers Distinguish forced turnovers from non-forced ones:

• exclude turnovers where the stated reasons were retirement, health, corporate governance reform, change of controlling shareholders, and legal disputes;

• for the remaining turnovers, trace the destinations of the departing managers and exclude cases where the departing managers took better positions after the turnover;

• further exclude cases where the managers’ tenure was less than 1 year;

• add 2 cases where the stated reason is retirement but the age of the departing mangers is less than 55;

• end up with 287 cases, representing 28.6% of all turnovers.

3. Data, Sample Selection, 3. Data, Sample Selection, and Research Methodand Research Method

Sample Selection: Nature of the Firm

Base of a total of 4246 firm-year observations, we:

• exclude 122 observations involving firms in the finance industry or listed only by the B-share market;

• exclude 284 observations where the firms have private shareholders as the ultimate controlling shareholders;

• exclude 151 observations involving firms with negative equity;

• exclude 163 observations involving turnovers where CEO’s tenure in the turnover year is less than 1 year;

• exclude 416 observations with missing values;

• end up with 3106 observations in the final sample.

3. Data, Sample Selection, 3. Data, Sample Selection, and Research Methodand Research Method

Research Method

Adopt four performance measures

• ROA

• IROA (industry-adjusted ROA)

• MROA (moving average of ROA)

• MIROA (moving average of IROA)

Set up four logit models to estimate turnover-performance links, each model with one performance measure

4. Regression Results4. Regression Results

Determinants of Turnovers (Table 7)

Turnover probability and control variables:

• coefficients of Age are significantly positive

• coefficients of Tenure and Duality are significantly negative

Turnover probability and performance measures:

• negative related to all four performance measures

• turnover rates are more sensitive to average performance than to annual performance

• models using average performance have higher explanatory powers

4. Regression Results4. Regression Results

Table 8. Logit Regression Estimation of Turnover-Performance Links in China’s

Profit-making Firms (Selected Part)

Table 9. Logit Regression Estimation of Turnover-Performance Links in China’s

Loss-making Firms (Selected Part)

1 2 3 4 1 2 3 4

ROA 0.163 ROA -3.571

(0.076) (1.801)*

IROA -0.537 IROA -2.997

(0.240) (1.523)

MROA -3.011 MROA -11.721

(1.190) (4.446)***

MIROA -2.407 MIROA -10.332

(0.969) (4.150)***

Constant -1.566 -1.536 -1.299 -1.39 Constant -6.064 -5.729 -8.451 -7.864

(0.760) (0.746) (0.634) (0.682) (1.538) (1.477) (2.170)** (2.069)**

Observations 2639 2639 2591 2591 Observations 467 467 460 460

Pseudo R-squared 0.065 0.065 0.068 0.067 Pseudo R-squared 0.056 0.054 0.115 0.102

4. Regression Results4. Regression Results

Turnover Sensitivities to Firm Performance (Table 8 and 9)

Profit-making firms

• all performance measures are statistically insignificant

• coefficients of Duality and Tenure are significantly negative

Loss-making firms

• all performance measures are significant, at different levels

• coefficients of Duality and Tenure are insignificant

Conclusions:

• consistent with stewardship and self-dealing views

• CEO’s power is able to reduce the probability of turnovers in profit-making firms, but not in loss-making firms

4. Regression Results4. Regression ResultsTable 11 (Panel A & B)

Changes in Post-turnover Performance in China’s Listed Firms

AllTurnover

s

Turnoversin Profit-Marking

Firms

Turnoversin Loss-makingFirms

AllTurnovers

Turnoversin Profit-Marking

Firms

Turnoversin Loss-makingFirms

1 2 3 4 7 8

+1 to –1 -0.0081** -0.0165*** 0.0116 0.0019 -0.0027* 0.0251**

+2 to –1 -0.0112** -0.0234*** 0.0238* 0.002 -0.0068* 0.0292**

+3 to –1 -0.0136* -0.0286*** 0.0258*** 0.002 -0.008 0.0498***

+1 to 0 -0.0011* -0.0056*** 0.0203*** 0.0083 0.0001* 0.0431***

+2 to 0 -0.0042 -0.0125*** 0.0325*** 0.0084 -0.004* 0.0472***

+3 to 0 -0.0066 -0.0177*** 0.0345*** 0.0084** -0.0051 0.0678***

YearRelative toTurnover

Panel A: Median Changes inUnadjusted Return on Asset

Panel B: Median Changes inIndustry-adjusted Return on Asset

4. Regression Results4. Regression ResultsTable 11 (Panel C & D)

Changes in Post-turnover Performance in China’s Listed Firms

Turnoversin Profit-Marking

Firms

Turnoversin Loss-makingFirms

AllTurnovers

Turnoversin Profit-Marking

Firms

Turnoversin Loss-makingFirms

5 6 7 8 9

+1 to –1 0.0067* 0.0593*** 0.0164*** 0.0026 0.0542***

+2 to –1 0.0051* 0.0385* 0.0115** 0.0001 0.0339

+3 to –1 0.0095* 0.0669*** 0.0167*** 0.0001 0.0649***

+1 to 0 0.0012 0.0238** 0.004* 0.0013 0.0193**

+2 to 0 -0.0004 0.003 -0.0009 -0.0012 -0.0009

+3 to 0 0.004 0.0315*** 0.0043* -0.0012 0.0301***0.0083**

0.0095**

0.0171***

0.0043

0.0007

YearRelative toTurnover

All Turnovers

9

0.0132***

Panel C: Median Changes in Control-group adjusted Return on Asset

Panel D: Median Changes in Control-group and industry adjusted Return on

Asset

4. Regression Results4. Regression Results

Performance Changes Surrounding Turnover (Table 11) Use either the turnover year (year 0) or the preceding year (year

-1) as the base year and trace performances from year +1 to +3 Findings:

• Changes in ROA and IROA:declined in profit-making firms, more significant in ROAsignificantly improved in loss-making firms, especially in

IROA• Median changes in CROA and ICROA:

no significant changes in profit-making firms significantly improved in loss-making firms, except in

year 2 Conclusions: significant post-turnover improvements occur in

loss-making firms, not in profit-making firms

5. Concluding Remarks5. Concluding Remarks

This study examines turnover-performance relations in China’s listed firms where shareholders have multiple objectives.

We provide evidence on:

the existence of different turnover-performance sensitivities in profit-making and loss-making firms

the existence of different post-turnover performance in profit-making and loss-making firms

Our study suggests that when firms are experiencing losses, state shareholders tend to attach a higher weight to firm performance and have higher incentive to discipline managers.

Generalization of our results to private shareholders should be taken with caution.

Table 8 Table 8 Table 8. Logit Regression Estimation of Turnover-Performance

Links in China’s Profit-making Firms 1 2 3 4 1 2 3 4

force force force force force force force force

List year 0.116 0.113 0.092 0.097 State 0.069 0.067 0.046 0.043

(2.920)*** (2.884)*** (2.306)** (2.480)** (0.301) (0.290) (0.198) (0.189)

Age 0.062 0.061 0.061 0.061 ROA 0.163

(5.532)*** (5.518)*** (5.447)*** (5.463)*** (0.076)

Tenure -0.25 -0.25 -0.238 -0.24 IROA -0.537

(4.383)*** (4.396)*** (4.203)*** (4.239)*** (0.240)

Duality -1 -0.995 -1.012 -1.014 MROA -3.011

(4.519)*** (4.505)*** (4.574)*** (4.592)*** (1.190)

StockReturn

-0.351 -0.34 -0.334 -0.334 MIROA -2.407

(1.415) (1.369) (1.350) (1.353) (0.969)

Leverage 0.055 0.054 0.047 0.049 Constant -1.566 -1.536 -1.299 -1.39

(2.662)*** (2.721)*** (2.790)*** (2.817)*** (0.760) (0.746) (0.634) (0.682)

Size -0.223 -0.223 -0.223 -0.227 Observations 2639 2639 2591 2591

(2.330)** (2.317)** (2.338)** (2.375)** Pseudo R-squared 0.065 0.065 0.068 0.067

Table 9Table 9

1 2 3 4 1 2 3 4

force force force force force force force force

List year 0.088 0.093 0.067 0.08 State 0.233 0.233 0.143 0.163

(0.972) (1.026) (0.689) (0.824) (0.539) (0.535) (0.326) (0.371)

Age 0.057 0.056 0.062 0.059 ROA -3.571

(2.835)*** (2.774)*** (2.973)*** (2.875)*** (1.801)*

Tenure -0.044 -0.046 0.015 0.007 IROA -2.997

(0.474) (0.493) (0.148) (0.066) (1.523)

Duality -0.025 -0.014 -0.016 0.035 MROA -11.721

(0.071) (0.039) (0.044) (0.100) (4.446)***

StockReturn

1.232 1.209 1.342 1.286 MIROA -10.332

(2.858)*** (2.797)*** (3.237)*** (3.040)*** (4.150)***

Leverage -0.011 -0.009 -0.054 -0.045 Constant -6.064 -5.729 -8.451 -7.864

(0.565) (0.466) (1.786)* (1.653)* (1.538) (1.477) (2.170)** (2.069)**

Size 0.044 0.026 0.148 0.102 Observations 467 467 460 460

(0.247) (0.147) (0.841) (0.594) Pseudo R-squared 0.056 0.054 0.115 0.102

Table 9. Logit Regression Estimation of Turnover-Performance Links in China’s Loss-making Firms