The Impact of Employee Compensation Disclosure on Corporate Governance Structure

Chiao Da Jvfanagement Review I令l. 30 No. 1.2010

pp.81-119

員工獎酬資訊揭露對公司治理結構的

影響

The Impact ofEmployee Compensation Disclosure

on Corporate Governance Structure

1

林靜香2 Ching-Hsiang Lin

國立高雄第一科技大學財務管理學率

Depa此mentof Finance, National Kaohsiung First University of Science and Technology 王萬成 WanncherngWang

國立中山大學企業管理學率

Department ofBusiness Administration, National Sun Yat-Sen University

摘要:本文主要探討證期局強制員工獎酬資訊揭露，對公司治理機制中的股 權結構、會計資訊在獎酬與績效關聯性監督角色的影響。預期在獎酬資訊揭 露前，會計資訊治理功能不彰，股權結構扮演績效與獎酬關聯性重要監督的 角色;獎酬資訊揭露後，提升會計資訊治理的功能，此時股權結構監督績效 與獎酬關聯性的重要程度降低。 實證結果發現: 1 強制獎酬資訊揭露前，大股東持股對績效與獎酬關聯性存 在顯著正向的影響;獎酬資訊揭露後，大股東持股監督績效與獎酬關聯性重 要程度降低 。 2 獎酬資訊揭露前，外資機構持股對績效與獎酬關聯性扮演重 要監督的角色;獎酬資訊揭露後，外資機構持股監督績效與獎酬關聯性的重 要程度降低 。 3 獎酬資訊揭露前，管理者持股對獎酬政策的影響支持利益掠 奪假說，即管理者持股對績效與獎酬關聯性存在負向影響;但強制獎酬資訊 揭露無法顯著改善管理者持股對績效與獎酬問負向的影響。進一步發現管理 者持股外，若公司同時存在大股東持股或外資機構持肢，獎酬資訊揭露使大 股東或外資投資機構更容易觀察獎酬決策的制定，提升資訊揭露在降低管理

1 The authors gratefully acknowledge the helpful comments and suggestions provided by two 個onymous_revlewers.

Corresponding author: Department of Finance, National Kaohsiung First University of Science and Technology, Kaohsiung City, Taiw<m. E-mail: ching66.1in@rnsa.hinet.net

82 The !mpact of Employee COl7lpel1satiol1 Disclosure

011 Corporate Governance Structure

者持股對績效與獎酬關聯性負向影響的治理功能 。

關鍵詞:員工獎酬;資訊揭露;公司治理;股權結構

Abstract: Mandatory compensation discIosure can have great impact on the structure of corporate governance. Our empirical results show that mandatory compensation disclosure decreases major stockholders' and foreign investment institutions' importance in overseeing performance-contingent rewards. A1so, consistent with the entrenchment hypothesis, our findings indicate 由at managerial stock ownership exerts a negative effect on the link between employee compensation and firm performance and that mandatory information disclosure does not mitigate this negative effect. Compared with regular shareholders, major stockholders and foreign investment institutions are more capable of and motivated in monitoring managers' self-interested behavior. Mandatory disclosure facilitates compensation oversight and contributes to the improvement of co叩 orategovemance

Keywords: Employee compensation; Information disclosure; Corporate govemance; Ownershi p structure

1.

Introduction

During the 2009 global financial crisis, many banks in the United States asked for federal financial assistance as they were on the verge of bankruptcy However, while the Obama administration released billions of dollars to bail them out, the failing banks were found to engage in abusive distribution of bonuses to their employees. The compensation scandals, criticized as highly irresponsible behavior by President Obama, have brought executive compensation plans to the forefront of public attention and debate3. Mter surveying 162 directors and 72 institutional investors in the US, Watson Wyatt4, concluded that companies

3 Clùna Daily News. (2009), Obama Denounced the Financial Industry, High-dividend Shameless, Available at: h即://www.cdnnews.com.tw/2009013 1.(In Chinese)

Chiao Da Management Review f01. 30 No. J, 2010 83

should strengthen the link between managers' compensation and their performance

Compensation plans usuaIIy incIude three components: cash bonus, stocks bonus and stock options. Prior to the amendment of Commercial Accounting Law in May 2006, Taiwan companies used to recognize granted stock bonus at par value instead of market value and treat employee bonus as an earnings distribution rather than an expense item. An ideal compensation plan should be

cIosely tied to firm performance and should motivate managers to work toward the goal of maximizing shareholders' wealth. However, this accounting method underestimates the cost of employee compensation, potentiaIIy encouraging companies to pay employee bonuses while their businesses are suffering great losses

Corporate ownership structure has become widely diffused nowadays Information asymmetry between shareholders and management causes the principal-agency problem, which in turn affects the performance-compensation relationship. DiscIosing more useful information and enhancing information transparency is the most direct solution to this agency problem. The Securities and Futures Bureau (SFB) requires th剖 public companies discIose compensation information, effective on January 31, 2003. With the mandatory discIosure of compensation information, the governance mechanism of accounting information can help investors protect their interests by effectively monitoring employee compensations plans

Corporate governance structure comprises various mechanisms among which substitution effect exists. Depending on its characteristics, a firm can adopt different mechanisms to optimize its governance structure. Studying the interaction between accounting information and other governance mechanisms,

La Porta et al (1998) find th剖 ownership concentration across countries is inverseIy related to the extent of a country's accounting discIosures. AIso, good accounting standards and shareholder protection law are associated with lower concentration of ownership. This suggests that ownership concentration is an

84 The Jmpact 0/ Employee Compel1satiol1 Disclosure 011 Corporate GOVernal1Ce Structure

outcome of poor investor protection. When accounting information and regulations fail to protect investors' interests, major shareholders will play a more important role in overseeing compensation plans. Bushman et αl. (2000) assert that the less information the accounting system provides, the higher cost the shareholders bear to collect data and monitor the link between firm performance and employee compensation. Young (2003) investigates how c。中orate governance structures va可 with the timeliness of accounting earnings. His empirical results point out a significant negative relation between the timeliness of earnings and the equity-based incentives of all officers and directors, and the equity-based incentives of outside shareholders

A11 the above studies demonstrate how critical a firm's accounting information is in its governance structure. The regulatory change on compensation disclosure in Taiwan provides us a unique opportunity to examine the c。中orate governance role of accounting disclosure. This study particularly focuses on the moderating effect of corporate compensation information on the governance role of major shareholder ownership structure and foreign investment ownership under the convergence-of-interest and entrenchment hypotheses, which to the best of our knowledge, have not been examined by prior studies

When compensation information disclosure is not mandatory and employee bonus is recognized as earnings distribution, accounting reports provide investors limited information for judging the performance-compensation association Information insufficiency forces m句 or shareholders to spend extra efforts in colIecting information to evaluate the performance-compensation relationship Mandatory disclosure of compensation information alIows regular shareholders to monitor and evaluate compensation plans and thus reduces the governing roles of major shareholders and foreign investment institutions. Existing literature provides two hypotheses that can be extended to examine the effect of managerial ownership on compensation policy: convergence-of-interest hypothesis and entrenchment hypothesis (Jensen and Meckling, 1976; Jensen and Ruback, 1983) The convergence-of-interest hypothesis asserts that shareholding aligns 由e interests of managers and shareholders and therefore will induce managers' efforts to maximize shareh

ChiaoDa λ1anagemeJ7l Review Tlól. 30 No. 1,2010 85

exerts a positive effect on the perforrnance-compensation relationship. On the

other hand, the entrenchment hypothesi s asserts that managerial ownershi p

protects the incumbent managers from displacement. Their positions in the

company being consolidated, managers, in spite of their ownership stakes, are

tempted to adopt a compensation plan in their own interest that decreases the firrn

value and shareholders' wealth. This implies a negative

perforrnance-compensation relationship.

As proposed by the convergence-of-interest hypothesis, managers in pursuit

of their self-interests should be motivated to adopt a proper compensation policy

Therefore, inforrnation disclosure should reduce the positive effect of managerial

ownership on the performance-compensation relationship. However, the

entrenchment hypothesis suggests that stock-holding managers may adopt a

compensation policy that hurts shareholders' wealth. Inforrnation disclosure will

prevent managers from adopting a compensation plan that decrease shareholders'

wealth. In other words, based on the entrenchment hypothesis, information

disclosure should alleviate the negative effect that managerial ownership exerts on the perforrnance-compensation link

To find evidence for the above propositions, we study information

technology companies listed in the stock exchange market during the years

1998-2005. Our empirical results indicate that before compensation inforrnation

disclosure is made mandatory, accounting repo此s fail to provide timely

inforrnation. Major shareholders and foreign investment institutions play a

dominant role in overseeing the performance-compensation relationship. After

mandatory inforrnation disclosure becomes effective, accounting reports provide

more useful inforrnation to facilitate the monitoring function of regular

shareholders, which then reduces the oversight role of major shareholders and

foreign institutional investors. We also find that the monitoring function of major

shareholders and foreign investment institutions encourages managers to adopt perforrnance-contingent rewards and prevent them from pursuing their

self-interests at the expense of shareholders' wealth. Furtherrnore, as proposed by

the entrenchment hypothesis, mandatory inforrnation disclosure mitigates the

86 The lmpact 0/ El71ployee Compensation Disc/osure

011 Corporate Governal1ce Structure

relationship. Mandatory compensation information disclosure enhances the transparency of information and acts as an effective corporate governance mechanism

2. Literature Review

2.1.

Substitution Effects of Corporate Governance Mechanisms

Substitution effects exist among corporate governance mechanisms, which can be categorized into two types: internal and externaL Internal governance mechanisms include oversight by the board of directors, incentive schemes, ownershi p structu悶 and accounting information. External governance mechanisms include regulations and laws, oversight by shareholders and creditors,

capital and managerial labor market, as well as threat of takeover. The findings of Aggarwal and Samwick (1999) show that in a competitive indust旬， a firm's incentive schemes are sensitive to and positively related to the rival firm's performance. Numerous studies demonstrate the substitution effects between accounting information and ownership structure. Verrecchia (1982) asserts that capital market participants will gather private information at a higher expense when the quality of the disclosed accounting information deteriorates. If the benefits of private information gathering exceed its costs, stakes in stock ownership motivate shareholders to collect private information to monitor managers' activities. Warfield et al. (1995) examine how the level of managerial ownership impacts the informativeness of earnings information. Their empirical results show that managerial ownership positively moderates the association between the earnings and the stock price ofthe company. When the accounting regulations and laws are unable to protect investors, major shareholders wiU monitor management activities (La Porta et al. 1998). Bushman et al. (2000) claim that the less a firm discloses its accounting information, the higher costs its shareholders pay to collect information and monitor 由e activity of the management. Fan and Wong (2002) investigate the relation between c。中orate

Chiao Da Management Rel'iew r'ól. 30 No. 1, 2010 87

ownership structure and the informativeness of eamings in East Asia. Their

research shows that the informativeness of earnings is negatively related to the

level of an u1timate owner' s voting control and to the discrepancy between the

voting rights and cash tlow rights of the ultimate controlling owners. Young

(2003) conducts a cross-sectional analysis on how the timeliness of eamings

information impacts co中orate govemance structure. Y oung (2003) finds that

there is a significant negative relation between the timeliness of eamings and the

equity-based incentives of officers and directors, and the equity-based incentives

of outside shareholders respectively. When the eamings repo此 provides an

inefficient forecast, co中orate structure will substitute the extemal higher-cost

investor-monitoring mechanism for the intemal accounting information mechanism

2.2. The Effect of Compensation Disclosure on the Performance-Compensation Relationship

On October 15, 1992, the US Securities and Exchange Commission (SEC)

approves new compensation disclosure rules and requires more compensation

disclosures in the annual proxy statements. Murphy (1996) examines the impact

of the 1992 proxy disclosure rules on company compensation. He finds that

managers bear nonpecuniary cost of reporting high level of compensation, and

wilI adopt reporting methodologies that reduce compensation cost. A lower

level of reported employee compensation alleviates managers' pressure from

politics and shareholders. Vafeas and Afxentiou (1998) also investigate how the

1992 SEC regulation affected the pay-for-performance relationship. The results

suggest that accounting and market performance measures following the new rule explain more ofthe cross-sectional variation in CEO pay compared to the pre-rule period

A paper by Ke et al. (1999) indicates the relation between CEO

compensation and accounting performance measures as a function of ownership

structure. They compare the use of accounting-based incentive pay contracts

across widely held firms and closely held firms. Ke et al. find closely held

88 The Jl7lpact 01 Employee Compel1satiol1 Disclosure

011 Corporate Governal1ce Structure

compensation less often than do widely held insurance finns. Park et al. (2001) also find that the advent of mandatory executive compensation disclosure encourages the use of perfonnance-contingent compensation. Craighead et al.

(2004) inves世gate how mandatory compensation disclosure atTects the CEO compensation practices in widely-held finns versus in closely-held finns. The results show 出at in the absence of mandatory disclosure, CEO cash compensation is less performance-contingent in widely held finns than in closely held firms.

With the advent of mandatory disclosure, performance-contingent cash compensation increases more in widely held firms than in closely held firms.

Compensation is less responsive to accounting performance infonnation in closely held finns than in widely held firms. The above research suggests that mandatory compensation disclosure increase the use of perfonnance-contingent compensation; however, the firm's ownership structure could sway the increase.

This paper studies the governance interaction between ownership structure and accounting information in detennining performance-compensation sensitivity.

3. Research Design

3.

1. Research Hypothesis

Infonnation asymmetry between managers and shareholders forms the main cause of principal-agency problems. The pu中ose of corporate governance is to prevent agency problems and protect the interests of small shareholders, who often are unable to oversee the management. Incentive compensation aligns the interests of the managers with those of the shareholders. An etTective compensation program should make executive pay sensitive to firm performance The better the firm perfonns, the more the executives are compensated, and vice versa. Governance mechanisms such as ownership structure and accounting information help monitor the performance-compensation link and increase the finn value. As finns often possess characteristics of their own, different govemance mechanisms evolve to perfonn the oversight function. Also,

Chiao Da A1anagemenl Rel叩w Vol. 30No. 1,2010 89

substitution effect exists among the various governance mechanisms. For instance,

ownership structure could take the governance role of accounting information if the latter fails to perforrn its oversight function

The following sections discuss the rnonitoring functions of inforrnation disc\osure, major shareholders, foreign investment institutions, and managerial ownership as well as the hypotheses we propose

3.1.1 Major Shareholders versus Information Disclosure

When accounting reports provides insufficient information, rnarket participants will try to collect cost1y private information (Verrecch悶， 1982). Since the cost of information collection often exceeds its benefit, srnall shareholders usually do not have strong incentives to acquire private information thernse\ves They rely on the rnajor shareholders to perform the rnonitoring 旭sk. Berle and Means (1932) asse付出at diffuse ownership structure lowers shareholders' incentives to rnonitor managerial perquisite-consurnption so performance-based compensations should be adopted to reduce agency costs. Schleifer and Vishny ( 1986) c\aim that cornpared with srnall shareholders, rnajor shareholders of a widely-held company have lower marginal cost of information collection and greater incentives to rnonitor rnanagers' performance. Managers under the c\ose monitoring of rnajor shareholders wil1 thus work toward the goal of maximizing shareholders' wealth. Agrawal and Mandelker (1990) find evidence supporting the active monitoring hypothesis that shareholders owning a large stake of the company will play a more active role in rnonitoring the management to enhance firm value. In other words, the existence of large shareholders contributes to the rnonitoring of firm activities, which then orients the managers toward maximizing firm value

As stated by La Porta et al. (1998), the soundness of a financial accounting system has great impact on the implementation of investor protection regulations When the regulators of a count可 provide poor protection for its investors, the governance rnechanism wil1 shift frorn legal protection to rnajor shareholders' overseeing. Bushrnan and Smith (2001) also find 伽t the less information provided by the financial accounting system, the more rnonitoring needed frorn

90 刃'1eJmpact of Employee Compensation Disclosure on Corporate Governance Structure

the maj or shareholders. Y oung (2003) regards the timeliness of eamings information as an important major determinant of the corporate govemance structure. The govemance importance of eamings information decreases when the information provided lacks timeliness. Lin and Hu (2003) find that as major shareholders' ownership increases, board members are more likely to adopt incentive contracts that are contingent upon performance. As indicated by the results of Chang's empirical research (2005), managers of a widely-held company tend to grant more stock option compensations in their own interests because the shareholders cannot efficiently monitor the activities of the management. Liao (2007) claims that the higher percentage of stocks the major shareholders possess,

the more attention they pay to the relationship between managerial compensation and eamings quality.

The above studies indicate that major shareholders serve an etTective monitoring role because they have lower marginal costs of acqumng and disseminating information, and receive a bigger share of the monitoring benefits owing to their large shareholdings. Prior to January 2003, when the FSB had not mandated the disclosure of bonus information, accounting repo此s provided insufficient information for investors to evaluate the perforrnance-compensation association. Major shareholders were motivated to collect information and evaluate how managers' compensations were aligned with their performance Before the disclosure of compensation information is made mandatory, co中orate govemance relies on the monitoring mechanism of major shareholders.

Accordingly, we proposeHl.

Hl: Prior 的 the mandatoηJ compensation disclosure

,

major shareholders' ownership has a positil'e effect on the pe枷'mance-compensation

relationship.

The main criterion of co叩orate govemance is to provide reliable, timely,

and transparent information. However, unless requested by laws or regulations,

companies usually are reluctant to fully disclose their important information and decisions. Morck et αl. (2000) find that stock prices do not efficiently reflect firm value in countries whose regulations and laws provide poor investor protection

Chiao Da Afanagement Rel'iew /lól. 30 No. 1, 2010 91

Ball et al. (2000) conclude th叫 establishing higher standards of common-law reduces the agency cost to monitor the management. In the wake of the Enron and WorldCom financial scandals, US Congress enacted a new law known as Sarbanes-Oxley Act (SOX) in July of 2002. SOX Section 404 mandates information disclosure, monitoring responsibilities, internal controls, and external

auditing. Increasing stringency of procedures and requirements for financial reporting is expected to improve information transparency and reduce agency problems

Performance-contingent compensation plans motivate employees to work toward enhancing firm value. Given disclosure on compensation, shareholders can evaluate whether compensation plans are designed to enhance the firm value Managers under the oversight of shareholders will implement a compensation plan that is tied to the performance of the firm. Vafeas and Afxentiou (1998) find that compensation disclosure mandated by the SEC strengthens the correlation between performance and compensation. This result upholds the new disclosure rule that aims to improve the governance of public companies. Ke et al

(1 999) 的 sertthat the association between cash bonus and accounting performance is stronger in widely-held companies that disclose more significant information than in closely-held companies that disclose less information. Their results show that within c1osely-held firms, CEO compensation is based less on objective measures such as accounting information and more on subjective measures Craighead et al. (2004) find 由肘， in the absence of mandatory disclosure, CEO

cash compensation is less performance-contingent in widely-held firms than in closely-held firms. Also, with the enforcement of mandatory disclosure,

performance-contingent cash compensation plans become more popular in widely-held firms than in closely-held firms

Ownership structure and information disclosure may substitute each other for governing the performance-compensation link. When compensation information is not disclosed and accounting reports provide insufficient information, small shareholders do not have much incentive to conduct cost-ineffective information collection. Major shareholders, on the other hand, have greater incentive to oversee the management and 出us play a critical role in

92 盯'1elmpact 01 Employee COmpel1satiol1 Disclosure

011 Corporate GOVernal1Ce Structure

monitoring the performance-compensation link. Mter compensation disclosure is mandatory and the govemance function of accounting information improves, extemal investors such as small shareholders are better able to monitor the alignment between compensation and performance. Different govemance mechanisms involve different costs. The cost of information disclosure by companies is generally lower than that of information collection by shareholders Hence, it is expected that mandatory information disclosure should improve the governing function of accounting information and reduce major shareholders' monitoring role. That is, compensation information disclosure substitutes monitoring mechanism of major shareholders for govemance mechanism of accounting information. ConsequentIy, we propose H2.

H2: After mandatory compensation disclosure

,

the positil'e effect 01 n呵。r

shareholders' ownership on the pe可ormance-con可pensation relation is

decreased.

3.1.2 Foreign Investment Institutions Versus Information Disclosure

Institutional investors also play an important monitoring role in corporation governance. Schleifer and Vishny( 1986 )point out that insti仙tional investors have more abilities and incentives to monitor managers and to enhance the relationship between compensation and performance. Pound (1988) advocates the efficient monitoring hypothesis 出at institutional investors have lower monitoring costs because of their professional knowledge and expertise. Denis (2001) asserts that institutional investors can monitor and restrain managers' self-interested behavior by either private recommendation or negotiation. Therefore, compared with regular shareholders, institutional investors are more efficient in monitoring the management to increase company value. Hartzell and Starks (2003) find that institutional investor ownership has a significant negative impact on managerial compensation. Yeh at a1. (2002) indicate th剖 compared with small shareholders, institutional investors, holding relatively more shares, are more motivated to protect their interests by monitoring management activities. Song (2006) finds that active institutional investors are more capable of preventing directors'

Chiao Da A1anagement Review 拘 1. 30 No. 1, 2010 93

self-interested behavior than are passive institutional investors

AlI taken together, the more shares institutional investors own, the more they are motivated to monitor the activities of the management. In Taiwan,

employee bonus under the regulation of commercial accounting law used to be regarded as earning distribution before 2006. It wasn't until 1998 when accounting treatment of employee bonus in Taiwan was highly criticized by foreign inst!1'Jtional investors, had the capital market started to regard employee bonus as company expenses. When compensation disclosure is not mandatory and compensation reports provide insufficient information, foreign institutional investors with their professional expertise act as the main corporate govemance mechanism. Formally,

H3: Prior to mandatoη compensation disclosure

,

institutional inl'estor

ownership has a positil'e 彷èct on the peφrnumce-compensation

relationship.

Foreign investment institution ownership and information disclosure interact with each other to impact the performance-compensation relationship When compensation disclosure is not mandatory and accounting reports provide insufficient information, foreign institutional investors have greater incentives to utilize their professional knowledge for monitoring firm activities. After mandatory compensation disclosure, accounting information augments its monitoring role and accordingly decreases the positive effect of foreign investment institution ownership on the performance-compensation relationship ConsequentIy, we propose

H4: After mandatory compensation disclosure

,

the positil'e effect of foreign

in附tment institution ownership on the pe.φrmance-compensation relationship decreases.

3.1.3 Managerial Ownership versus Information Disclosure

The effect of managerial ownership on the performance-compensation relationship is opposite to the convergence-of-interest hypothesis and the

94

entrenchment hypothesis

The Jmpact 01 Employee Compensation Disclosure

011 Corporate Governance Structure

Agency the。可 argues that the separation of

ownership and management results in managers' perquisite-consumption behavior

in the pursuit of self interests and thus harms the firm value. An increase in

managerial ownership aligns managers' interests with shareholders' interests and prevents the losses resulting from managers' perquisite-consumption behaviors

(Jensen and Meckling, 1976). Core and Guay (1999) find that incentive-based

compensation such as stock option, a c。中orate govemance mechanism, reduces

the agency problems between shareholders and managers. Watson Wyatt ( 2008 ) surveys S&P 1500 companies and finds that total shareholder retum is about 30% higher in companies with more managerial ownership than in those with less

managerial ownership 5. In order to improve the correlation between top

executive's compensation and shareholders' equi句， many American firms have

established Executive Share Ownership Guidelines. Based on the Executive

Share Ownership Guideline, top executives should not sell their company stocks

when their shareholding does not meet the lowest required level. Tsai (2007)

finds that the positive relation between the timeliness of eamings information and measures of manager compensation-eamings sensitivity increases when

integration degree between board incentive and shareholder incentive improved.

The higher the managerial ownership, the more the shareholders' and the

managers' interests converge. Managerial ownership motivates managers to adopt

an effective compensation system and strengthens the performance-compensation

relationship. Moreover, when board members make compensation decisions,

managers with higher ownership have more power to decide their own

compensation package. With the alignment of shareholder and manager

interests, managers tend to adopt a compensation policy 血atincreases firm value

Hence, we conclude that managerial ownership has a positive effect on the

performance-compensation relationship (Finkelstein and Hambrick, 1989)

Nevertheless, Holmstrom (1979) suggests that while

performance-contingent compensation contracts, such as variable compensation,

align the interests of shareholders and managers, they also expose managers to

5 Clú, M. (2008), From Four Large View Look Advanced Charge Reward Desig

n,

Watsoo Wyatt,

Chiao Da A1anagement Review 1毛主1. 30入10. 1，2010 95

risks. Risk-averse managers will choose a fixed compensation system over a

perforrnance-contingent one unless the board members impose pressure them not

to (Gomez-Mejia et al., 1987). The more shares the managers own, the more

power they possess to act against the monitor of the board members. In the pursuit

of their self-interests, managers may increase perquisite-consumption or choose

policies that hurt shareholders' weaIth. Jensen and Ruback (1983) find that

managers controlling the majority of the shares may maintain their power and

self-interests by choosing aIternatives that are less beneficial to the shareholders

Yermack (1997) find that when ac仙al level of compensation is greater than

expected, managers willlower the value of stock option on a grant day. Core and

Guay (1999) suggest that managers, being opportunistic, decide the quantity of

option in order to increase their compensation. Hung (2004) finds th剖 anmcrease

in top managers' ownership decreases the positive relation between stock

compensation and accounting performance. Tsai (2006) asserts that an increase in

director ownership decreases the relation between director compensation and firm

performance. Therefore, as proposed by the entrenchment hypothesis, when

managerial ownership gets higher, the board has less power in constraining

managers' decision making. Managers are likely to establish compensation

policies such as increasing fixed compensation to promote their own interests

This leads to a decrease in the positive effect of managerial ownership on the

performance-compensation re1ationship

No previous studies conclude whether the effect of managerial ownership on the performance-compensation relationship is consistent with that proposed by either the convergence-of-interest hypothesis or the entrenchment hypothesis This paper aims to seek evidence for the following competitive hypothesis

H5: Under the conl'ergence-o

f-

interest hypothesis

,

prior to n間ndatoη

conψensation disclosure

,

numagerial ownership has a positil'e 吧。ect on

the performance-conψensation relationship.

H6: Under the entrenchment hypothesis

,

prior to m.andatory compensation d

必'iscl彷I扣os捌ur，閃e角 翩m.11削z閻ana句~ge伊伊~ger臼臼叮en吋'戶吋叫"'

96 The lmpact 01 Employee Compensatiol1 Disclosure

011 Corporate Governal1ce Structure

Mandatory compensation disclosure makes it possible and easy for outside shareholders to gain compensation-related information. Shareholders are thus empowered to monitor whether managers adopt compensation plans that are contingent upon the firm performance. Zeckhauser and Pound [1990] suppo此 th剖

mandatory compensation disclosure strengthens co中or剖e govemance by helping shareholders exert pressure on the board if necessary. Ball et al. (2000) find that common law disclosure guidance reduces the agency costs of monitoring the managers. The govemance mechanisms of managerial ownership and information disclosure are interrelated. Increasing transparency of compensation information can lead to the decreasing importance of managerial ownership in c。中orate

govemance. Therefore, we expect that under the convergence of interest hypothesis, mandatory compensation disclosure reduces the positive effect of managerial ownership on the relationship between performance and compensatlOn

Agency problems occur when shareholders and managers have asymmetric information. Mandatory information disclosure reduces 伽 information

asymmetry problems and protects shareholders' interests. Bushman and Smith (2001) point out that financial accounting information serves the monitoring function and helps avoid manager opportunism. Lobo and Zhou (2001) and Hunton et al. (2006) suggest that information disclosed or transparent information limits managers' manipulation of eamings, and reduces the profits getting from eaming management, which in tum discourages managers to manage eamings Chang and Fang (2006) assert that manipulations of eamings information substantially reduce after the enforcement of the

“

Information Disclosure Evaluation System". DiscIosing compensation information, such as the relationship between compensation and performance, reduces information asymmetry between managers and shareholders, and prevents managers from increasing their personal wealth through excessive compensation. Mandatory compensation discIosure increases managers' risks of engaging in abusive compensation plans. Cou1ton et al. (2003) find a negative association between the transparency of compensation information and the monetary amount of compensation. The govemment authority regulates the compensation information

Chiao Da Alanagement Review /'ól. 30 No. 1, 2010 97

disclosure based on persuasion effect. According to the entrenchment hypothesis, mandatory compensation disclosure can increase information transparency, monitor managers' decision making of compensations, and thus consolidate the compensation-performance relationship. H7 and H8 are accordingly established based on the above analysis

H7: Under the conl'ergence-o

f-

interest hypothesis

,

mandatory co仰的sation

disclosure leads 的 a reduction in the positi阿 effect of managerial

ownership on the performance-compensation relationsh伊.

H8: Under the entrenchment hypothesis

,

n的ndatory compensation di.sclosure

I伽eωad品s ω a re吧edJ伽'uctio仰ni仿n t.伽h如e ne啥'ga仙d的l'e effec叫ctofm翩1l附仰帥uana傅'ge.伊伊'ger臼臼er.伊叫. p

戶eφ枷枷rm翩1l削削仰z閻仰仰仰aω仰nt缸仰ce-c吋叫Cωω0ω，呻'咿qψ仰仰rJ附仰eelnsaαd伽d伽伽伽仰0ωn rel，蚓'ati伽0ω側ns的sh均智.

4. Empirical Resu

It

s

4.1

Descriptive Statistics and Correlation Analysis

Table 1 presents the descriptive statistics ofthe variables used in this sωdy.

The means of STOCK, ROE, and ROA are $67,547(in thousands), 16 .2~1o， and 13.48%, respectively. The average values of BIG, FO

R,

and CEO are 14.66%, 8.27%, and 2.90%, respectively. The means of the controlled variables (SIZE = 15.27 and MVBV = 2.22) signify that the sample is composed of high-growth

companies. The BETA of 0.98 is used as the measurement variable of firm risk.

Table 2 presents the Pearson correlation matrix for dependent and independent variables. As indicated by the univariate analysis, STOCK is positively related to ROE, ROA, SIZE, and 恥1VBV， implying that employee stock bonus increases with the firm size and perfo口nance.

4.2 Empirical Results

4.2.1 ROE as Proxy for Company Performance

98 The Jmpact 01 Employee Compel1satiol1 Disc/osure 011 Corporate GOl'ernal1ce Structure explanatory power because the F-value of 53.15 has a significance level of 0.01

2.

and the adjusted R" is 0.21. As predicted, before mandatory disclosure of bonus information, the coefficient of ROE*BIG (172.86) is positive and significant (t-stat = 2.26). This means that when information governance is ineffective, major shareholders will monitor the performance-compensation relationship. Thus, we find suppo吋 forHJ. Furthermore, the coefficient of ROE*BIG*DIS (-143.16) is negative and significant (t-st的= -2.95). This means that mandatory compensation disclosure strengthens inforrnation governance function and reduces the positive effect of major shareholders on the relation between compensation and performance. This finding supports H2. The control variable SIZE has a coefficient of 72,614, which is positive and significant (t-stat = 15.93), meaning that more stock bonus is distributed as the size of the company increases

Table 4 presents the results of regression model (2), which also has strong explanatory power because the F-value of 80.72 has a significance level of O.Oland the adjusted R2is 0.29. The coefficient ofROE*FOR (500.04) is positive and significant (t-stat 7.80). As stated in H3, before the mandatory compensation information disclosure, ownership by foreign investment institutions exerts positive effect on the performance-compensation relationship The coefficient of ROE*FOR*DIS (-290.14) is negative and significant (t-stat = -6.07), implying that information disclosure reduces foreign investment institutions' importance in compensation oversight Accordingly, H4 is sustained

Table 5 presents the resuIts of regression Model (3), which has strong explanatory power because the F-value of 52.24 has a significance level of O.Oland the adjusted R2 is 0.20. The coefficient of ROE*CEO (-276.27) is negative and significant (t-sta t= -2.09), meaning that managers make compensation decisions to seek self-interests rather than to maximize shareholder weaIth. Managerial stock ownership exerts negative effect on the perforrnance-compensation reIationship; hence, H6 is supported.

Based on the entrenchment hypothesis, information disclosure facilitates the compensation oversight job of regular shareholders, which then leads managers to adopt perforrnance-contingent rewards. The coefficient of ROE*CEO*DIS (61.92) is positive and non-significant (的tat = 0.5 月It shows

Chiao Da Management Review ~ól. 30 No. 1.2010 99

that infOImation disclosure fails to reduce the negative effect of managerial

ownership on the relationship between performance and compensation. This

finding does not support H8.

Table 1

Descriptive Statistics Standard

Variables Mean Max Min

Deviation STOCK 67547.00 242315.00 4674426 0.00 ROE 16.29 11.31 79.10 0.02 ROA 13.48 8.40 56.85 -7.24 B1G 14.66 8.65 55.95 0.01 FOR 8.27 11.41 71.16 0.01 CEO 2.90 3.49 25.16 0.01 丘IZE 15.27 1.42 20.33 11.90 λ.fI;BV 2.22 1.63 22.53 0.33 BETA 0.98 0.29 1.98 -Ll5

Variable Defmitions: STOCK = employee stock bonus; PER = measurement of firm performan白，

defmed as ROA and ROE ; ROE= retum on equi紗， defined as net income divided by

average shareholders equity; ROA = retum on 品sets，defined as income before interest and tax divided by average total assets; BIG=major shareholder owner吉凶p， defmed as major shaI芯holder stockholding divided by outstanding shares; FOR = ownerslùp of foreign investment institution, defmed as foreign investment institution stockholding divided by outstanding shares; CEO = managerial ownerslùp, defmed as managerial stockholding divided by outstanding shares; SIZE=fmn size, defmed 品 natural logarithm of assets;

此1VBV = ratio of market -to-book, defmed as market value of common equity divided by the book value of common equity; BET A = fum ri址，defined as firm systematic risk.

100 刃leJmpact of Employee Compensation Disclosure on Corporate Governance Structure

Table 2

Pearson Correlation Analysis

Variables STOCK ROE ROA BIG FOR CEO SIZE MVBV BETA STOCK 1.00 ROE 0.10*** 1.00 ROA 0.17*** 0.88*** 1.00 BIG 0.02 0.06** 0.07*** 1.00 FOR 0.38*** 0.08*** 0.10*** 。 13*** 1.00 CEO -0.10*** 。 15*** 。 13*** -0.11*** -0.13*** 1.00 SIZE 。 44*** 。 12*** 0.02 -0.06** 0.38*** -0.13*** 1.00 MVBV 。 12*** 0.63*** 。.58*** -0.01 0.14*** 。 15*** 0.11 *** 1.00 BETA 。 19*** -0.09*** -0.07** -0.25*** 0.05* -0.13*** 0.41 *** -0.06** 1.00 Variable definitions are given in Table 2 什)， 什竹， and (***) represent being statisticalIy

significant at (0.1), (0.05), and (0.01) levels, respectively Table 3

The Effect ofMajor Shareholders Ownership and Employee Bonus information Disclosure on the Performance-Compensation Relationship: Using

Return on Equi句 (ROE) 的 proxyfor Firm Performance

STOCK_it = α。 +α\ROE_it + α2B1Gjt +α3ROEjt * BJGit + α4ROEjt * BJGit * DJS

+αsSJZE_it + α6MVBVit + α7BETAit + G_a Variables JNTERCEPT ROE BJG ROE*BJG ROE*BJG*DJS SJZE MVBV BET.A Adj R2 F-value Sign

+

Coefficient t-statistic -1101695.00*** -16.85 -344.68 -0.31 782.55 0.64 172.86** 2.26 -143.16*** -2.95 72614.00*** 15.93 7557.22 1.60 24478.00 1.07 0.21 53.15***

ChiaoDa λ1anagement Review J'ÓI. 30 No. 1, 2010 101

Variable defmitions are given in Table 2. DIS, an indicator variable, equals 0 if observations are

from 1998-2001 (pre-mandatory disclosure) and equals 1 if from 2002-2005

(post-mandatory disclosure). (*), (艸)， and (*料) represent being statistically significant at (0.1), (0.05), and (0.01) levels, respectively 甘le t-statistics are based on White (1980) standard errors

Table 4

The Effect of Foreign Institutional Ownership and Employee Bonus

information Disclosure on the Performance-Compensation Relationship:Using

Return on Equity (ROE) 的 Proxyfor Firm Performance

STOCK_il 二_β

。

+β'IPER_il + β'2_ROEil +β'3ROEu * FOR_il + β'4ROEil * FOR_u * DIS

+

_ßsSIZEu + ß6MVB~1 + β~BETAu + 丸

Variables Sign Coefficient t-statiståc

lNTERCEPT -836133.00*** -l3.14 ROE -1077.l3 -1.47 FOR 720.02 0.72 ROE*FOR + 500.04*** 7.80 ROE*FOR*DIS -290.14*** -6.07 SlZE 55232.00*** 11.81 MVBV 1293.80 0.29 BETA 29327.00 1.38 AdjR2 _0.29 F-value 80.72***

Variable defmitions are given in Table 2. D時，an indicator variable, equals 0 if observations are

from 1998-2001 (pre-mandatory disclosure) and equals 1 if from 2002-2005

(post-mandaωry disclosure). (*), (**), and (***) represent being statistically sigrúficant at (0.1), (0.05), and (0.01) levels, respectively. The (-statistics are based on White (1980) standard errors

102 The Jmpact of Employee Compensation Disclosure

011 Corporate Governal1ce Structure

Table 5

The Effect ofManager Ownership and Employee Bonus information Disclosure on the Performance-Compensation Relationship: Using Return on Equity (ROE) 的 Proxyfor Firm Performance

STOCKit = yo + y}ROEit + Y2 CEOit + 几 ROEit *CEOit + 几ROEit *CEOit

*

DIS + ys_SIZEit + Y6MVB~t + Y7 BETAit + 6it

Variables Sign Coefficient t-statistíc

JNTERCEPT -1088234 ∞*** -16.71 ROE 1055.33 1.33 CEO 1034.67 0.35 ROE*CEO ? -276.27** -2.09 ROE*CEO*DJS ? 61.92 0.55 SJZE 72284.00*** 15.73 MVBV 12895 ∞*** 2.72 BETA 16096.00 0.72 AdjK 0.20 F-value 52.24***

Variable defiIÙtions aI芯 givenin Table 2. D時，a indicator variable, equals 0 if observations are

from 1998-2001 (p記-mandatory disclosure) and eq閏Is 1 iffrom 2002-2005 (post-mandatory disclosure). (*), (**), and (*抖) represent being statistically significant at (0.1), (0.05), and

(0.01) levels, respectively. The t-statistics are based on White (198θ) standard errors

4.2.2 ROA as Proxy for Firm Performance

Table 6 presents the results of model (1) with Retum on Assets (ROA) 的

the proxy for firm performance. The coefficient ofROA*BIG (451.01) is positive and significant (t-st的= 4.78). This finding is consistent with H 1 that before mandatory disclosure of compensation information, major shareholders ownership has positive effect on the performance-compensation relationship. The

Chiao Da A1anagement Rel'iew Vol. 30λlo. 1, 2010

Table 6

The Effect of Major Shareholders Ownership and Employee Bonus

Information Disclosure on the Performance-Compensation

103

Relationship:Using Return on Assets (ROA) 的 Proxyfor Firm Performance

STOCKil = α。+αjROAil +α2BIGit +α3ROAit * BIGiI + α4ROAit * BIGil * DIS

+αsSIZE_i/ + α6MVB~/ + α7BETA_{i/ +&i/}

Variables 1NTERCEPT ROA B1G ROA*B1G ROA *B1G*D1S SIZE MVBV BETA Adj R2 F-value Sign

+

Coefficient t-statistic -111741.00*** -17.17 2376.24 1.61 -1696.03 -1.34 451.01 *** 4.78 -274.97*** -4.81 74015.00*** 16.66 -8512.06* -1.90 20729.00 0.93 0.24 64.16***

Variable defmitions are given in Table 2. DIS, an indicator variable, equals 0 if observations aI芯

from 1998-2001 (pre-mandatory disclosure) and equals 1 if from 2002-2005 (post-mandatory disclosure). (*), (料)， and (*料) represent being statistically significant at (0.1), (0.05), and (0.01) levels, respectively. The t-statistics are based on White (1980) standard errors

coefficient ofROA*BIG*DIS (-274.97) is negative and significant (t-stat = -4.81),

implying that infonnation disclosure decreases major shareholders' importance in

overseeing the performance-compensation relationship. This finding also suppo此S

H2

Table 7 shows the resuIts of model (2) with ROA as the proxy for firm

perfonnance. The coefficient of ROA *FOR (874.89) is positive and significant

(t-stat = 15.97). The coefficient of ROA*FOR*DIS (-3109.48) is negative and

significant (/-stat = -3.92). Accordingly, Hypotheses 3 and 4 are sustained

Table 8 displays the results of regressing the performance-compensation

104 The Jmpact 01 Employee Compel1satiol1 Disclosure

011 Corporate GOVernal1Ce Structure

when using ROA as the proxy for firm performance. The coefficient of

ROA *CEO (-688.80) is negative and significant (t-stat = -3.95); the coefficient of

ROA *CEO*DIS (-74.90) is negative but non-significant (t-stat = -0.52). The

results suppo此 the entrenchment hypothesis. Mandatory information disclosure

does not mitigate the negative impact of managerial stock ownership on the performance-compensation relationship

Table 7

The Effect of Foreign Institutional Ownership and Employee Bonus

Information Disclosure.on the Performance-Compensation Relationship:

Using Return on Assets (ROA) 的 Proxy for Firm Performance

STOCKit = ßo + β'lROA;t + β2FORit + ß3ROAit * FORit + ß4ROA;t * FORit * DIS

+ ßsSIZEit + ß6MVB

V;

t + ß7 BETAit + sit Variables JNTERCEPT ROA FOR ROA*FOR ROA *FOR *DJS SJZE MVBV BETA A哼;If F-value Sign

+

Coefficient t-statistic -788037.00*** -13.21 625.55 0.54 -7767.99*** -8.17 874.89*** 15.97 -3109.48*** -3.92 54588.00*** 12.73 -10121 ∞抖 -2.49 26933 ∞ 1.38 0.39 130.92***

V扭曲le defmitions are given in Table 2. DIS, an indicator variable, equals 0 if observations are

from 1998-2001 (pre-mandatory disclosure) and equals 1 iffrom 2002-20的。ost-man曲的ry

disclosure). (*), (**), and (料*) rep時sentbeing statistically significant at (0.1), (0.05)，訂“ (0.01) levels, respectively. The t-statistics are based on Wlùte (1980) standard errors.

Chiao Da A1anagement Rel'iew I,ól. 30 入fo. 1,2010 105

Table 8

The Effect of Managerial Ownership and Employee Bonus information Disclosure on the Performance-Compensation Relationship:

Using Return on Assets (ROA) 的 Proxyfor Firm Performance

STOCKit

=

Yo + y1ROAit + Y2CEOit + Y3 ROA;t *CEOit + Y4 ROAit *CEOit

*

DIS + YsSIZEit + y6M間几 +y7BEl月it + εit

Variables Sign Coefficient t-statistíc

INTERCEPT -1164127.00*** -18 日 ROA 8222.60*** 8.07 CEO 7100.23** 2.46 ROA*CEO

?

-688.80*** -3.95 ROA *CEO*DIS ? -74.90 -0.52 丘lZE 73240.00*** 16.38 λ1VBV -1918.30 -0.42 BETA 18640.00 0.86 AdjR2 0.24 F.-value 63.35***

Variable defmitions are given in Table 2. D芯， an indicator variable, equals 0 if observations are

from 1998-2∞ 1 (pre-mandatory disclosure) and equals 1 if from 2002-2005 (post-mandatory

discIosure). (*), (抖)，and (抖*) represent being statistically significant at (0.1), (0.05), and

(0.01) levels，時spectively. 甘let-statistics are based on White (1980) standard errors

4.3 Additional Analysis

4.3.1 The Roles of Major Shareholders and Foreign Investment Institution

Ownership in Overseeing Managerial Entrenchment Behavior

This paper aims to examine the effect of mandatory compensation

disclosure on the perfonnance-compensation relationship. Offering employee

stock bonus increases the level of managerial ownership. Managers used to have a

great deal of power over compensation policy making before compensation

106 The Jmpact 01 Employee Compensation Disclosure on Corporate Governance S，的lcture

managerial ownership exerts a negative effect on the performance-compensation relationship. Unexpectedly, this negative effect does not decrease after compensation information disc10sure is made mandatory even though the

governance ability of information disc10sure is supposed to increase. We then investigate whether managerial ownership, in the presence of ownersl叩 by major shareholders or foreign investment institutions, reduces its negative effect on the performance-compensation relationshi p6. Model (4) is formed by adding major

shareholders ownership to Model (3), and Model (5) by adding foreign investment institution ownership

STOC4

=.Â.v

+

À,.P E

l?t

+ 也Blq

+

Â.3CEQ + λ'4PEl?t *Blq +在PEI?， 牢 Blq 牢 DIS

+λ'6PERit _*CEOit + λ7PERit *CEO_it 牢

DlS+ λgSlZE_it +~B V;t

+λ10BETA iI +ε (4)

STOC4

=

00 +~P E

1fr

+02FO

I\

+B3CEQ +禹PE~*FO~+冉PE~

*

FO~

*

DlS

+θ'6PERit *CEO_it + θ'7PERit

*

_CEOit

*

DlS + θ'gSlZEit + θ'cþ1VBVit

+θ10 BETA iI

+

& i (5)

The regression results of model (4) with ROA as proxy for firm performance is displayed in Table 9. The coefficient of ROA旬的 (435.03) is positive and significant (t-stat 4.44). The coefficient of ROA *BIG*DIS (-348.44) is negative and significant (t-stat = -5.43). The results are similar to those of Model (1). Before compensation information disclosure is made mandatory, managerial ownership has a negative effect on the performance-compensation relationship (coefficient

=

-892 .3 8 ，的tat = -4.90 ),

6 Multi-collinearity problem was found among the variables of major shareholder宮，ownership, foreign investment institutions' ownership, and managerial ownership in 血emodel.

Chiao Da Management Review Vol. 30 No. 1, 2010 107

which supports the entrenchment hypothesis. After mandatory compensation information discIosure, managerial ownership exerts a positive effect on the performance-compensation relationship (coefficient二325.87， t-stat = 2.02). The results, while supporting H8, are divergent from the empirical result ofModel (3). We suspect that information discIosure can reduce managers' self-interested behavior because major shareholders possess great incentive to monitor compensation plans 弘1andatory disclosure of bonus information facilitates major shareholders' job of overseeing compensation plans and thus improves the governance function of information disclosure. Similar empirical results are achieved either by using ROE or by using ROA as the proxy for firm performance.

Table 0 presents the regression results of model (5) with ROA as the proxy for firm performance. Before mandatory compensation information disclosure, foreign investment institution ownership has a positive effect on the performance-compensation relationship (ROA可OR coefficient = 849.87， 的tat= 15.48). The mandatory compensation information disclosure reduces the importance of foreign investment institution in monitoring the performance-compensation relationship (ROA *FOR *DIS coefficient = -4500.85, t-stat = -4.41). The results are similar to those of model (2). Moreover, managerial

ownership has a negative effect on the performance-compensation relationship (ROA *CEO coefficient = -632.46, t-stat = -3.72) before mandatory compensation

information discIosure and a positive effect (ROA *CEO*DIS coefficient = 306.07,

t-stat =1.8月 after mandatory discIosure. The results of model (5) are different from those of model (3). We suspect that foreign investment institutions have greater incentive to monitor managers' self-interested behavior. Mandatory discIosure makes the job of compensation oversight easier for foreign investment institutions. Mandatory disclosure thus improves the governance function of compensation information disclosure and decreases the negative effect of managerial ownership on the performance-compensation relationship

108 η'1e Jmpact of l!.ìnployee Compel1satiol1 Disclosure 011 Corporate GOl'ernal1Ce Structure

Table 9

The Effect of Major Shareholders Ownership

,

Manager Ownership and

Employee Bonus information Disclosure 00 the Performance-Compensatioo

Relationship:Using Return on Assets (ROA) 的 Proxyfor Firm Performance

STOCKit = 九+人RO~t +~BIGit +À-;CEO_it +À.4RO~t _*BIGit +À-;RO~t _{*BIGit *DIS}

+λ6ROA íI *CEO it + λ7ROA íI *CEO it

*

DIS + λ8SIZE íI + λ9W官VíI +λABETA. + G.

Variables Sigo Coefficient t-statistíc

JNTERCEPT -1144997.00*** -17.14 ROA 5892.31 3.50 BIG -990.14 -0.78 CEO 6324.79** 2.19 ROA*BJG

₊

435.03*** 4.44 ROA*BJG 呵)JS -348.44*** -5.43 ROA*CEO ? -892.38*** -4.90

ROA *CEO *DIS ? 325.87** 2.02

SJZE 73098.00*** 16.51

λ1VBV -4358.14 _{。 96}

BETA 20376.00 0.91

Adjk 0.25

F-value 48.66***

Variable defmitions 剖-egiven in Table 2. DIS，個 indicatorvariable, equals 0 if observations are

from 1998-2001 (p時-mandatorydisclosure) and equals 1 iffrom 2002-2005 (post-mandatory

disclosure). (*), (**), and (***) represent being statistically significant at (0.1), (0.05), and

Chiao Da Management Review Vol. 30 No. 1, 2010 109

Table 10

The Effect of Foreign Institutional Ownership

,

Manager Ownership

,

and

Employee Bonus information Disclosure on the Performance-Compensation

Relationship:Using Return on Assets (ROA) 的 Proxyfor Firm Performance

STOCK_i, = θ。 +θlROAiI + θ2FORi' + B)CEOil + θ'4ROAi' * FORi, + θ5ROAsr*FOR'， *DIS

+θ6ROA il * CEO il + θ7 ROA it * CEO il * DIS it + θgSIZE it + θgW官 Vit

+θloBETA il + ε

Variables Sign Coefficient t-statistic

JNTERCEPT -813024.00*** -13.36 ROA -3429 日** 2.50 FOR -7610.34*** -8.01 CEO 4257.59* 1.65 ROA*FOR

+

849.87*** 15.48 ROA *FOR *DJS -4500.85*** -4.41 ROA*CEO ? _-632.46*** -3.72 ROA *CEO*D1S ? 306.07* 1.85 SJZE 54763.00*** 12.80 MVBV -7658.27* -1.87 BETA 27169.00 1.39 AdjK 0.40 F-value 94.06***

Variable defmitions are given in Table 2. D時，an indicator variable, equals 0 if observations 3I它

from 1998-2∞ 1 (pre-mandatory disclosure) and equals 1 iffrom 2002-2005 (post-mandatory disclosure). (*), (**), and (***) represent being statisticaUy significant at (0.1), (0.05), and (0.01) levels, respectively. The t-statistics are based on White (1980) standard errors.

4.3.2 Adding Employee Bonus Information Disclosure and Firm Performance of Previous Year as Independent Variables

1n Models (1), (2) and (3), we use the dummy variable of employee bonus infonnation disclosure and its interaction term with ownership structure to test the

110 口1elmpact of Employee Compensation Discfosure 011 Corporate Governance Structure

effect of bonus information disclosure on c。中orate governance. The dummy variable of employee bonus information disclosure is not treated as an independent variable in any of the three regression models. We then add the dummy variable of employee bonus information disclosure and firm performance of the previous year into models (l), (2) and (3) because firm performance of the previous year can impact the amount and type of employee bonus of the current year. We use ROA as the proxy for firm performance and present the regression resuIts in Table 11. The F-value of 48.68 has a significance level of 0.01 and the adjusted R2 is 0.24. The coefficient of ROA*BIG (514.29) is positive and significant (的tat 4.68). After the mandatory compensation information disclosure, the coefficient of ROA *BIG*DIS (-340.55) is negative and significant (t-stat = -4.22). These resuIts are similar those resuIts of modeJ (1), providing further support for Hypotheses 1 and 2.

Table 12 displays the effect of foreign investment institution and compensation information disclosure on the performance-compensation relationship. The F-value of 105.86 has a significance level of 0.01 and the adjusted R2 is 0.41. The coefficient of ROA*FOR (1096.92) is positive and significant (的tat 16.20). After the mandatory compensation information disclosure, the coefficient of ROA *FOR *DIS (-265.95) is negative and significant (t-stat = -4.72). These resuIts are similar those of Model (2), also supporting Hypotheses 3 and 4

Table 13 shows the effect of managerial ownership and compensation information disclosure on the performance-compensation relationship. The F-value of 48.77 in this model has a significance level ofO.Ol and the adjusted R2 is 0.24. The coefficient of ROA *CEO (-764.34) is negative and significant (t-stat

= -4.6) after the mandatory compensation information disclosure. The coefficient

of ROA *CEO*DIS (59.05) is positive but non-significant (t-stat = 0.36). These resuIts are also similar to those ofModel (3), supporting H5 but not H6

As indicated by the above analysis, adding the dummy variable of employee bonus information disclosure and co叩oration performance of the previous year as independent variables does not change the regression resuIts Major shareholders and foreign institutional investors can supervise the

Cσ'hiωaoDσ Maωnagement Rel'叩w~均台1. 30No. 1,2010 111

perfonnance-compensation relationship before infonnation disclosure is mandatory. Mandatory information disclosure helps improve the governance function of accounting information, which then reduces the governing roles of major shareholders and foreign investment institutions. Before compensation infonnation disclosure is made mandatory, the negative effect of managerial ownership on the performance-compensation relationship suppo此s the entrenchment hypothesis. Mter mandatory, information disclosure still fails to perfonn its governance function to reduce managers' self-interested behavior

Table 11

The EfTect of Major Shareholders Ownership and Employee Bonus Information Disclosure on the Performance-Compensation Relationship:

Adding Employee Bonus Information Disclosure and Corporation Performance ofPrevious Year as Independent Variables

STOCKjl = α。+αjROAiI+α2B1Git +α3D1SiI + α4ROAit * BIGiI + αsROAit * BIGjl * DIS

+α6ROA_{it_l + α7}SIZE_{it 十 αgMVB V;t + α9}BETA_{it + ε}

+

Coefficient t-statistic -1141560*** -16.60 3035.28* 1.84 -1962.77 -1.49 20847 1.03 514.29*** 4.68 -340.55*** -4.22 -994.97 -1.16 75268*** 16.42 -8045.45* -1.69 17346 0.74 0.24 Variables lNTERCEPT RO.4 B1G DlS RO.4*BIG RO.4 *BIG*DlS Sign

rc-M

F:"value 48.68***

Variable defiIÙtions are given in Table 2. D店， an indicator variable, equals 0 if observations are

from 1998-2001 (pre-mandatory disclosure) and equals 1 if from 2002-泊的(post-mandatory

112 The Jmpact

01

Employee Compensation Disclosure on Corporate GOl'ernance Structure

(0.01) levels, respectively. The l-statistics are based on White (1980) standard errors

Table 12

The Effect of Foreign Institutional Investors Ownership and Employee Bonus Information Disclosure on the Performance-Compensation Relationship:

Adding Employee Bonus Information Disclosure and Corporation

Performance of Previous Year as Independent Variables

STOCK;r = α。+α， ROA ;r +α2FORit +α3D1S;r + α4ROAit * FOR;r +αsROA_it * FOR;r * DIS

+α6 ROAir _1 +α7SIZEit + αgM間只r+α9BE1月it + ε

Variables JNTERCEPT ROA FOR DJS ROA*FOR ROA *FOR *DJS Sign

+

Coefficient t-statistic -778233*** -12.39 -1100.63 -1.04 -7644.96*** -7.91 -7313.32 -0.5 1096.92*** 16.20 -265.95*** -4.72 -1167.84 -1.54 55110*** 12.57 -9223.47** -2.21 19688 0.98 0.41 ROA 品l ' ， r 弓 ， 必 古 mM-R 日 HUE-dd pdAB-A F.-value 105.86 艸*

Variable definitions aJ芯 givenin Table 2. D時，an indicator variable, equals 0 if observations are from 1998-2001 (pre-mandatoI)' disclosure) aod equals 1 if from 2002-2∞5 (post-mandatoI)'

disclosu記). (*), (料)，and (林*)represent being statistically significant at (0.1), (0.05), and (0.01)

Chiao Da A1anagement Review r令1. 30 No. 1, 2010 113

Table 13

The Effect of Manager Ownership and Employee Bonus Information Disclosure on the Performance-Compensation Relationship:Adding Employee Bonus Information Disclosure and Corporation Performance of Previous Year

as Independent Variables

STOCKu = α。 +αlROA_i/+α2CEOit +α3D1Su + α4ROAit *CEOi/ +αjROA_it *CEOi/ * DIS +α ROÆÎt-. .+ α SIZE. l ' '-^'71V....~A....J it + α MVB V. + α BETÆ. ' """"'8 +&

Variables Sign Coefficient t-statistic

1NTERCEPT -1144630*** -16.92 ROA 9160.93*** 7.37 CEO 6641.27** 2.20 D1S -35561 ** -2.22 ROA*CEO ? -764.34*** -4.6 ROA *CEO *D1S ? 59.05 0.36 ROA ιI -710.31 _{。 82} S1ZE 74603*** 16.20 MVBV -4859.29 -1.01 BETA 10981 0.48 Adj J?1 0.24 F-value 48.77***

Variable def1lÚtions are given in Table 2. D店，an indicator variable, equals 0 if observations

are from 1998-2∞ 1 (pre-mandatory disclosure) and equals 1 江 from2002-泊的(post-mandatory

disclosure). (*), (艸)， and (***) represent being statisticaUy significant at (0.1), (0.05), and (0.01)