Applied Economics

ISSN: 0003-6846 (Print) 1466-4283 (Online) Journal homepage: [Link]

Firm performance, corporate governance and

executive compensation in Pakistan

Muhammad Fayyaz Sheikh, Syed Zulfiqar Ali Shah & Saeed Akbar

To cite this article: Muhammad Fayyaz Sheikh, Syed Zulfiqar Ali Shah & Saeed Akbar (2017):
Firm performance, corporate governance and executive compensation in Pakistan, Applied
Economics, DOI: 10.1080/00036846.2017.1386277

To link to this article: [Link]

Published online: 09 Oct 2017.

Submit your article to this journal

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

[Link]

Download by: [Southern Cross University] Date: 09 October 2017, At: 09:18
 APPLIED ECONOMICS, 2017
[Link]

Firm performance, corporate governance and executive compensation in

Pakistan
Muhammad Fayyaz Sheikha, Syed Zulfiqar Ali Shahb and Saeed Akbarc
a
Lyallpur Business School, G.C. University, Faisalabad, Pakistan; bWarwick Business School, The University of Warwick, Coventry, UK; cHull
University Business School, Hull, UK

ABSTRACT KEYWORDS
This study examines the effects of firm performance and corporate governance on chief executive Corporate governance;
officer (CEO) compensation in an emerging market, Pakistan. Using a more robust Generalized dynamic panel; emerging
Method of Moments (GMM) estimation approach for a sample of non-financial firms listed at markets; executive
Karachi Stock Exchange over the period 2005–2012, we find that both current- and previous-year compensation; firm
accounting performances has positive influence on CEO compensation. However, stock market performance; fixed effects
Downloaded by [Southern Cross University] at 09:18 09 October 2017

performance does not appear to have a positive impact on executive compensation. We further JEL CLASSIFICATION
find that ownership concentration is positively related with CEO compensation, indicating some G30; M52; G34; C13
kind of collusion between management and largest shareholder to get personal benefits.
Inconsistent with agency theory, CEO duality appears to have a negative influence, while board
size and board independence have no convincing relationship with CEO compensation, indicat-
ing board ineffectiveness in reducing CEO entrenchment. The results of dynamic GMM model
suggest that CEO pay is highly persistent and takes time to adjust to long-run equilibrium.

I. Introduction directors and where family or controlling share-

holders are more likely to exploit interests of min-
There has been an enormous growth in research on
ority shareholders. Specifically, this study examines
executive compensation over the last two decades
the role of firm performance, board structure and
with primary focus on compensation of chief execu-
concentrated/family ownership in designing CEO
tive officer (CEO). Much of this research focuses on
compensation contracts.
the question whether executive compensation con-
The Asian socio-economic and behavioural pecu-
tracts can be justified in terms of their contribution
liarities and institutional settings are different from
to the firm financial performance (Devers et al. 2007;
Western World and studies conducted in Western
Van Essen et al. 2012a). According to agency theory
World have limited implications for Asian countries
(Jensen and Meckling 1976), executives are self-
(Fan, Wei, and Xu 2011; Ghosh 2006; Gibson 2003;
interested and may behave opportunistically at the
Hofstede 1980; Sun, Zhao, and Yang 2010; Van
expense of shareholders’ interests. Therefore, corpo-
Essen et al. 2012a). While there is some evidence
rate boards are supposed to confine executive oppor-
on the link between firm performance, corporate
tunism and align the executives’ interests with that
governance and compensation from other Asian
of shareholders by better monitoring through effec-
countries, the Pakistani context is peculiar for a
tive corporate governance mechanisms and design-
number of reasons. First, concentrated and family
ing efficient pay contracts that typically link
ownership is more common in Pakistan than, for
executive compensation with firm performance.
instance, in Japan and Korea. Similarly, while
The objective of this study is to examine whether
Chinese firms have more ownership concentration
CEO compensation is influenced by firm perfor-
than in Pakistan, the nature of ownership concentra-
mance and corporate governance practices in an
tion in Chinese firms is different as the State usually
emerging market, Pakistan, where CEOs are pre-
holds high stakes in large firms (e.g. Bryson, Forth,
sumed to be more powerful than the boards of
and Zhou 2014). On the other hand, concentrated

CONTACT Syed Zulfiqar Ali Shah [Link]@[Link] Warwick Business School, The University of Warwick, Scarman Road, Coventry
CV4 7AL, UK
© 2017 Informa UK Limited, trading as Taylor & Francis Group
 2 M. F. SHEIKH ET AL.

ownership in Pakistan is maintained by non-govern- decisions (see, Fan, Wei, and Xu 2011; Sun, Zhao,
ment shareholders. Non-government ownership and Yang 2010; Van Essen et al. 2012a). Fan, Wei,
concentration makes firms like private-owned firms and Xu (2011), for instance, note that ‘Until now, we
which may have different implications for CEO still do not know much about how managers of
compensation. emerging market firms are paid and promoted and
Second, legal and political environment in factors that influence these decisions’ (p. 211).
Pakistan is weaker and the overall governance is Therefore, by analysing CEO compensation in an
poor (Rehman et al. 2012). The government effec- emerging market of Pakistan, we provide important
tiveness index and regulatory quality index esti- contribution to international literature on executive
mated by World Bank remained negative in the compensation.
last decade or so. In addition, there is more foreign Second, we find evidence that despite boards tend
influence on governance and corporate environ- to be weaker as compared to the management in
ment. Pakistan has been under the influence of Pakistan, CEO compensation is positively associated
International Monetary Fund (IMF) and other fund- with firms’ accounting performance. This is the first-
ing agencies for so long. Moreover, Pakistani econ- hand evidence in Pakistan.2 Further, we find that
Downloaded by [Southern Cross University] at 09:18 09 October 2017

omy is plagued with more corruption than many board size and presence of non-executive directors
other Asian countries. According to Transparency do not contribute towards CEO compensation in
International, the Corruption Perception Index any direction. This reflects the contextual settings
never cross 30 for Pakistan (100 shows no corrup- of Pakistan where non-executive directors are hired
tion). Therefore, people in Pakistan (including from within the family or they are proxies of con-
executives) are more prone to unethical and oppor- trolling shareholders, making board structure irrele-
tunistic behaviour (Mujtaba and Afza 2011). Third, vant. Contrary to agency theory arguments, we find
the disclosure requirement about CEO compensa- that ownership concentration is associated with
tion is stronger in Pakistan. Companies are required higher CEO compensation, indicating some sort of
to report all the components of CEO compensation. misappropriation of minority shareholders’ interests.
This is not the case for most of the other Asian Similarly, CEO duality appears to have a negative
countries (see, e.g. Basu et al. 2007; Conyon and relationship with CEO compensation. This particu-
He 2011; Kato, Kim, and Lee 2007). larly has important implication as separation of
Given above differences, Pakistani market pro- these positions has become mandatory requirement
vides a unique context to study how the boards see in the revised Code of corporate governance in
firm performance as a determinant of CEO compen- Pakistan. Overall, we highlight that corporate gov-
sation? What role concentrated/family ownership ernance variables do not seem to influence CEO
plays in setting CEO pay and how board structure compensation in the expected directions as sug-
affects CEO compensation decisions? These ques- gested by the agency theory.
tions are particularly interesting in countries like Third, our study considers persistence and adjust-
Pakistan as two seminal studies (Durnev and Kim ment of CEO pay by examining the effect of lagged
2005; Klapper and Love 2003) show that firm-level compensation on current compensation using
corporate governance practices matter more in dynamic panel model estimation. Further, bulk of
countries with weaker legal systems and investor the existing studies typically uses panel data estima-
protection.1 tion using fixed or random effects models. These
Our study contributes to the extant literature in a models, however, do not control for potential endo-
number of ways. First, our study can be considered geneity problems. Results based on these panel data
as a response to calls for more research on under- models are pruned to estimation problems. As a
standing how managers of emerging market firms consequence, we also employ more robust methodo-
are compensated and factors that influence these logical procedure such as GMM that simultaneously

1
Given the dominance of controlling shareholders in the developing markets, Jameson, Prevost, and Puthenpurackal (2014), for instance, call for more
country-level studies to better understand the influence of controlling shareholders on minority shareholders.
2
Our findings are in contrast to the results of two existing studies (Kashif and Mustafa 2012; Shah, Javed, and Abbas 2009) on CEO compensation in Pakistan.
These existing studies, however, are limited in scope and do not attempt to provide rigorous analysis.
 APPLIED ECONOMICS 3

accounts for unobserved heterogeneity, serial corre- shareholders’ interests (type-II agency problem). In
lation and endogeneity problems. addition, the expropriation of minority shareholders’
Rest of the article is organized as follows: interests increases when formal institutions such as
Literature review and hypotheses development is legal system and investor protection are weak.
presented in Section II. Section III describes data Therefore, given this context, we expect that CEO
and methodological procedures. Section IV provides compensation contracts are not based on arm’s
empirical results while conclusions are presented in length transaction between the board of directors
Section V. and CEOs; instead these are more likely to be
based on mutual interests of CEOs and controlling
shareholders. Hence, we state our hypothesis that
II. Literature review and hypotheses
H1: firm performance does not influence CEO
CEO compensation and firm performance
compensation.
In agency theory, the corporate boards, assuming the
power to look after the firm, involve in arm’s length
Downloaded by [Southern Cross University] at 09:18 09 October 2017

transaction with CEO and design such compensa-

CEO compensation and corporate governance
tion plans which provide CEO with efficient incen-
tives to maximize the shareholder value, and hence Corporate governance assumes the role of monitor-
reduce moral hazard problem arising from separa- ing and curbing managerial opportunism so that
tion of ownership from control (Bebchuk and Fried shareholder’s interests are protected (Fama and
2003). This predicts a positive link between CEO Jensen 1983). As an efficient corporate governance
compensation and firm performance. A number of structure provides close monitoring and oversight of
studies (e.g. Bayless 2009; Buck, Liu, and Skovoroda management, in the presence of strong corporate
2008; Conyon and He 2011; Conyon and He 2012; governance, the executive compensation would
Murphy 1999; Ozkan 2011) find evidence consistent be low.
with agency theory. However, there are other com- A number of corporate governance variables have
peting theories that emerge in response to the find- been reported to influence the CEO compensation;
ing that pay–performance relationship is modest however, their influence seems to be conditioned to
(see, Tosi et al. 2000; Van Essen et al. 2012a). The the context under examination (see, e.g. Core,
most prominent one is managerial power theory Holthausen, and Larcker 1999; Devers et al. 2007;
(MPT) (Bebchuk and Fried 2003). Sun, Zhao, and Yang 2010; Van Essen, Otten, and
According to MPT, if the balance of power shifts Carberry 2012b). Since concentrated and family
towards CEOs and they behave opportunistically, ownership is an important feature of corporate gov-
then there is likelihood that CEOs would involve in ernance environment in Pakistan and board compo-
rent extraction through setting their compensation sition has been one of the major focuses in the Code
high that is not in the interests of shareholders of corporate governance (2002; 2012) to mitigate the
(Bebchuk and Fried 2003). With increased power agency conflicts, we focus on concentrated/family
of CEO, the board of directors and compensation ownership and board structure (board size, board
committee, under the influence of CEO, compro- independence and CEO duality) in this study.
mise their fiduciary duties and settle upon excessive
CEO compensation possibly not linked to firm per-
Ownership concentration, family ownership and
formance (Bebchuk and Fried 2003).
CEO compensation
In Pakistani context, CEOs tend to be more
powerful than the boards because they are either Concentrated and family ownership can affect CEO
heads of the controlling families or have strong ties compensation contract in two competing ways: (1)
with controlling shareholders (Javid and Iqbal 2008; interest alignment effect and (2) entrenchment
Kamran and Shah 2014). The strong ties between effect. According to interest alignment effect, which
CEOs and family/controlling shareholders suggest relates to agency theory, large or family shareholders
greater possibility of expropriation of minority have strong incentives to oversee agents’ activities
 4 M. F. SHEIKH ET AL.

because of being insiders, strong commitment and compromise their monitoring role and hence
better firm-specific knowledge (Bertrand and Schoar weaken the internal governance structure.
2006; Harris and Raviv 2008; Jensen and Warner Consequently, executives gain more power over the
1988; Su, Li, and Li 2010). Therefore, concentrated internal control mechanisms, leading to more influ-
and family ownership generally suggests that share- ence on their own pay, resulting in higher executive
holders are better able to protect their interests in compensation. Many studies find that larger boards
their companies, leading to reduced managerial are related to higher executive compensation (e.g.
opportunism, higher interest alignment and lower Core, Holthausen, and Larcker 1999; Croci,
CEO compensation. Gonenc, and Ozkan 2012; Fahlenbrach 2009;
However, entrenchment effect suggests that Ozkan 2011; Shah, Javed, and Abbas 2009; Van
family or controlling shareholders can expropriate Essen, Otten, and Carberry 2012b).
minority shareholders’ interests through many ways In Pakistan, board size generally tends to be dri-
including excessive compensation packages (see, ven by directors appointed from the controlling
Croci, Gonenc, and Ozkan 2012; Su, Li, and Li families or by proxy directors working on behalf of
2010; Wang and Xiao 2011). CEOs in close relation controlling shareholders. Consequently, it is less
Downloaded by [Southern Cross University] at 09:18 09 October 2017

with controlling shareholders/family may set their likely that board size has any effective role in redu-
own pay opportunistically high, thereby expropriat- cing agency conflicts. In that case, board size is not
ing the minority shareholders’ wealth. Such expro- expected to influence CEO compensation. Therefore,
priation is very likely in emerging markets where we hypothesize that
formal institutions are weak to support mutually
beneficial impersonal exchange between economic H3: board size has no influence on the CEO
players (see, Jameson, Prevost, and Puthenpurackal compensation.
2014; Young et al. 2008). Since Pakistan is an emer-
ging market with weak legal systems, we hypothesize
that
Board independence and CEO compensation
H2a: Ownership concentration has a positive impact Agency theory suggests that independent directors
on CEO compensation. are likely to play important role in aligning share-
holder–manager interests by providing adequate
H2b: CEO compensation is higher in family firms monitoring. Independent outside directors are less
than in non-family firms. subject to collude with management and have repu-
tation to protect shareholders in the labour market
(Core, Holthausen, and Larcker 1999; Fama and
Jensen 1983). On the other hand, inside directors
Board size and CEO compensation
are more obligated to CEO and can be under greater
Board size is considered as an important determi- CEO influence, leading to compromised CEO mon-
nant of board effectiveness. It has significant contri- itoring to get personal benefits from CEO such as
bution towards quality of governance (Jensen 1993; career opportunities (see, Bebchuk and Fried 2003;
Lipton and Lorsch 1992). Larger boards are likely to Weisbach 2007). Nevertheless, external directors are
correlate with greater level of expertise and firm’s also prone to have negative impact on internal gov-
ability to extract critical resources (Dalton et al. ernance if they have some secret relationship with
1999; Provan 1980). However, they may become so management (Core, Holthausen, and Larcker 1999).
heavy, leading to ineffective executive monitoring Overall, board independence is expected to be
(Jensen 1993). Larger boards are less likely to func- related to less managerial opportunism, leading to
tion effectively and are easier to be controlled by lower executive compensation.
executives (Jensen 1993; Lipton and Lorsch 1992). Empirically, available evidence is mixed over the
In addition, they are likely to be plagued with com- relationship between board independence and
munication and coordination problem (Ozkan executive compensation. For example, Boyd (1994)
2007). Thus, larger boards are assumed to and Core, Holthausen, and Larcker (1999) find
 APPLIED ECONOMICS 5

positive association, while others (e.g. Byrd and Pakistan. Accordingly, we expect that CEO com-
Cooperman 2010; Conyon and He 2011; Conyon pensation is higher when CEO also holds the posi-
and He 2012) find no or negative relationship tion of chairman board of directors. We formally
between number of independent directors in board hypothesize that
and executive compensation. Given the Pakistani
context where non-executive directors are generally H5: CEO duality has a positive impact on CEO
hired from within the family or obligated to work on compensation.
behalf of controlling shareholders (Javid and Iqbal
2008; World Bank 2005), we may expect so-called
board independence to become irrelevant in corpo-
rate decision-making, leading to non-negative rela- III. Methodology
tion between CEO pay and board independence.
Data
Therefore our hypothesis is
We focus on all the non-financial firms listed at
H4: board independence has no impact on CEO Karachi Stock Exchange (KSE), Pakistan, for the
Downloaded by [Southern Cross University] at 09:18 09 October 2017

compensation. period 2005–2012. Out of 399 non-financial listed

companies classified in 12 industrial groups by State
Bank of Pakistan (SBP), 139 companies are dropped
because either they are declared as defaulted by KSE,
CEO duality and CEO compensation newly listed or merged/demerged (86 companies), or
their data on corporate governance and compensa-
CEO duality (CEO as chairman board of directors tion are not available (53 companies). All the data
at the same time) provides opportunities to ‘self- are extracted from companies’ annual reports col-
interested’ CEOs to influence major decisions in lected by hand from different sources including SBP,
order to maximize their own utilities instead of KSE and companies’ websites. As we go farther from
maximizing shareholders’ wealth (Core, 2012, the availability of annual reports decreases,
Holthausen, and Larcker 1999; Jensen 1993). therefore sample period is restricted to start from
CEO duality reduces the board independence 2005.
and increases the executive powers over control For the remaining 260 companies, we managed
decisions including designing executive compen- to collect data for at least 3 consecutive years,
sation contracts, leading to higher executive com- making an unbalanced panel data containing
pensation. Thus, CEO duality is considered to be 1836 firm-year observations. However, out of
a sign of inefficient corporate governance in both 1836 firm-year observations 328 observations are
agency theory and managerial power theory. dropped, because in those observations, CEOs are
Empirically, most of the studies (e.g. Boyd 1994; not paid any compensation. Final sample contains
Brick, Palmon, and Wald 2006; Conyon and He 1508 firm-year observations from 225 firms for the
2012; Core, Holthausen, and Larcker 1999; period 2005–2012.
Fahlenbrach 2009; Van Essen, Otten, and
Carberry 2012b) find positive association between
CEO duality and executive compensation, indicat-
Static CEO pay model
ing the likelihood of CEO entrenchment and
excessive payment to CEO when he/she is also a Following existing literature (e.g. Gallego and
chairman board of directors. In Pakistan, Code of Larrain 2012; Murphy 1999; Sun, Wei, and Huang
corporate governance encourages companies to 2013; Wang and Xiao 2011), the following linear
separate CEO position from chairman board of model is estimated to test the association of CEO
directors.3 Thus, CEO duality is considered to be compensation with firm performance and corporate
a potential cause of managerial entrenchment in governance in a static pay process:
3
In the revised version of the Code of corporate governance issued in March 2012, separation of position of chairman from CEO is mandatory with effect
from year 2013.
 6 M. F. SHEIKH ET AL.

Pit ¼ β1 ROAit þ β2 TRETit þ β3 OWNCONSit owned directly or indirectly and (2) two or more
þ β4 FAMOWNit þ β5 BDSIZEit family members sit on the board of directors.
þ β6 B INDit þ β7 DUALCEOit BDSIZE, board size, is measured as number of
directors sitting on the board as mentioned in the
þ β8 INSTOWNit þ β9 GROUPit
annual reports. Although Code of corporate govern-
þ β10 FIRMSIZEit þ β11 FIRMRSKit
ance in Pakistan encourages the representation of
þ β12 MTBit þ β13 FMAGEit independent directors on the board, this has not
þ β14 CEOCHNGit þ αi þ ωt þ εit (1) been a mandatory requirement until year 2013.
where LNCOMPit is log of compensation. In Further, disclosure regarding independent director
Pakistan, long-term incentive plans, stock options is very much inconsistent across the companies.
and restricted stocks are virtually non-existent. Therefore, we use ratio of non-executive directors
CEOs are paid in the form of base salary, cash to board size as a measure of board independence
bonuses, perks and benefits, and post-employment (B_IND). DUALCEO is CEO duality, and it is incor-
benefits. Companies use many accountings heads in porated as a dummy variable taking the value of 1 if
disclosure, and nomenclature of the accounting CEO is also the chairman of the board of directors
Downloaded by [Southern Cross University] at 09:18 09 October 2017

heads is not uniform across the firms. The predomi- and 0 otherwise.
nantly used accountings heads include managerial Institutional ownership and group affiliation are
remuneration, bonus, leave encashment, house important drivers of corporate governance in emer-
rent, utilities, medical expenses, gratuity, provident ging markets. According to agency theory, institu-
fund and retirement benefits. Consistent with exist- tional ownership serves a monitoring role and is
ing literature, we use two measures of CEO compen- related to lower CEO compensation (Hartzell and
sation, that is, cash compensation and total Starks 2003). Similarly, behaviour of business group
compensation. Cash compensation includes man- firms is known to be different from their counterparts
agerial remuneration and bonuses, while total com- (Siegel and Choudhury 2012). Ghosh (2006), for
pensation is the sum of all the components. ROAit, instance, reports that CEO compensation is lower in
return on assets, represents accounting performance, group-affiliated firms in India which have similar
and it is measured as ratio of income before interest socio-economic structure like Pakistan’s. We, there-
and taxes to total assets. TRETit, total return to fore, control for institutional ownership and group
shareholder, represents market performance, and it affiliation. Institutional ownership (INSTOWN) is
is measured as current market price of shares plus measured by percentage of shares held by institutions
dividend for the current year divided by previous- while Group (GROUP) is a dummy variable that takes
year market price. In Pakistan, it takes about 4 value 1 if a firm belongs to a group and 0 otherwise.
months until annual reports are published and dis- Other control variables include: firm size
tributed. Therefore, to avoid any inconsistencies, (FIRMSIZE), which is measured as log of total assets;
market price per share is taken Gibson, M. S. 2003. firm risk (FIRMRSK), which is measured as SD of
“Is Corporate Governanceon the date that is 4 monthly stock returns for the fiscal year; growth
months after the closing date. opportunities (MTB) proxied by market to book
OWNCONS represents ownership concentration. ratio as measured by market value per share divided
Consistent with existing literature (see, Holderness by book value per share; and firm age (FMAGE) as
2017; La Porta, Lopez-De-Silanes, and Shleifer 1999), indicated in the annual reports.
we use ownership of largest shareholder as a proxy Firm size is perhaps one of the most cited deter-
for ownership concentration. FAMOWN is a minants of CEO compensation across the world.
dummy variable taking value 1 for family firms and Countless studies report that firm size is positively
0 otherwise. Following Anderson and Reeb (2003), related to executive compensation (e.g. Devers et al.
Achleitner et al. (2014) and others, we define a 2007; Frydman and Jenter 2010). Managing a risky
family firm that fulfils one of the two conditions, company needs better managerial skills, leading to
(1) a person or family group holds at least 25% of higher compensations (Brick, Palmon, and Wald
voting right as measured by the percentage of shares 2006; Conyon and He 2011; Core, Holthausen, and
 APPLIED ECONOMICS 7

Larcker 1999). Firms with greater growth opportu- the employee from his/her subsequent real perfor-
nities are expected to hire the executives with better mance. The employer’s information and belief about
skills who can exploit the available growth opportu- the employee’s performance accumulate and are
nities to maximize the shareholder value. This leads serially correlated (Conyon and He 2012).
to a positive link between growth opportunity and Therefore, wages based on such expectations are
CEO compensation (e.g. Brick, Palmon, and Wald expected to correlate in adjacent years. This poten-
2006; Conyon and He 2012; Ho, Lam, and Sami tial correlation however is ignored in the majority of
2004). Similarly, aged firms are more likely to devise previous studies on executive compensation
more efficient compensation contracts (see, e.g. (Conyon and He 2012). Another viewpoint that is
Conyon and He 2012; Ho, Lam, and Sami 2004). relevant to adjacent year correlation in executive
Finally, if CEO is replaced in a firm, the new con- compensation is anchoring-and-adjustment heuristic
tract is unlikely to be the same as the previous one. (Bender 2003; Tversky and Kahneman 1974).
Further, if CEO is replaced in the middle of a year, Tversky and Kahneman (1974), for instance, suggest
there would be two persons drawing compensation that in many situations people make numerical esti-
in a year. Therefore, to account for such instances, mates by starting from an initial value (the anchor)
Downloaded by [Southern Cross University] at 09:18 09 October 2017

we include a variable (CEOCHNG) that represents and adjusting this to yield a final answer. However,
how many times a firm changes its CEO during the the adjustments made are usually insufficient there-
year, measured as an interaction term of dummy fore different starting points yield different estimates
variable taking value 1 if CEO is replaced during which are biased towards the anchor (Bender 2003;
the year and number of CEOs drawing remunera- Raithatha and Komera 2016). The board of direc-
tion during the year. tors/compensation committee often have a figure as
In order to account for unobserved heterogeneity, a starting point, that is, prior year’s compensation
we estimate fixed effect model after performing which has an influence on the current pay levels
Hausman (1978) test. In addition, both Breusch– (Bender 2003; Raithatha and Komera 2016).
Pagan/Cook–Weisberg heteroscedasticity test and As argued above, the current CEO compensation
White heteroscedasticity test indicate the presence may be a function of previous year’s CEO compen-
of heteroscedasticity. Similarly, Wooldridge (2002) sation because of incomplete information and
test for serial correlation in panel data suggests the dynamic learning process, and anchoring-and-
presence of serial correlation in error terms. adjustment heuristic. Therefore, we also estimate
Therefore, we use Huber–White robust SEs clustered dynamic panel model in addition to static panel
at firm level. This adjustment of SEs accounts for model to account for CEO pay persistence and
both heteroscedasticity and serial correlation dynamic adjustment.
(Greene 2011). The estimated model is as follows:

LNCOMPit ¼ α þ γLNCOMPit1 þ β1 ROAit

CEO pay persistence and dynamic panel model þ β2 TRETit þ β3 OWNCONSit
Agency theory assumes that executive compensation þ β4 FAMOWNit þ β5 BDSIZEit
contracts are static pay-for-performance contracts, þ β6 B INDit þ β7 DUALCEOit
where CEO compensation is in equilibrium in each þ β8 INSTOWNit þ β9 GROUPit
period. However, learning is an important factor that þ β10 FIRMSIZEit þ β11 FIRMRSKit
affects wage dynamics within a firm (Conyon and þ β12 MTBit þ β13 FMAGEit
He 2012). When an employee enters the labour þ β14 CEOCHNGit þ ωt þ εit
market, his/her ability is partially observable by the
(2)
employer. Education, experience and other charac-
teristics do reflect initial quality of the employee but The definitions of variables are the same as in
still this information is incomplete. Therefore, initial model (1).
wage decision has to be made by the employer on We estimate model (2) using instrumental vari-
the basis of employee’s expected performance. The able (IV) approach and GMM estimator. This strat-
employer then gradually learns about capabilities of egy controls for potential endogeneity problem in
 8 M. F. SHEIKH ET AL.

addition to unobserved heterogeneity. Before pro- in year 2009. This is possibly because of three rea-
ceeding, we first test the endogeneity of the regres- sons: (1) unrest due to political issues and general
sors using Durbin–Wu–Hausman test for elections, (2) start of energy crisis in Pakistan or (3)
endogeneity of all the regressors following Schultz, effects of financial crisis.
Tan, and Walsh (2010) and Nguyen, Locke, and Market return appears to be more fluctuating
Reddy (2015). than ROA. In year 2005, average market return is
We use system-GMM approach because it is more 27.70% which dramatically decrease to 3.49% in year
efficient (Blundell and Bond 1998; Roodman 2009). 2006 followed by an increase to 24.87% in year 2007.
System-GMM reduces the effect of high persistence Negative returns in year 2008 and 2009 seem to be
of corporate governance variables thereby improving indicating political unrest, energy crisis and financial
the power of estimations (Blundell and Bond 1998; crisis. Recovery seems to start after that with posi-
Nguyen, Locke, and Reddy 2015). In addition, sys- tive-market returns.
tem-GMM appears to be the best-performing esti- Average board size of pooled sample appears to be
mator for the data which is characterized by slightly above eight with SD of 1.57. Recently,
moderate length of time, low within firm variations Jameson, Prevost, and Puthenpurackal (2014) report
Downloaded by [Southern Cross University] at 09:18 09 October 2017

in corporate governance variables and the possibility similar average board size of eight for Indian firms.
of fixed-effects-driven dependent variable; some This is possibly due to resemblance in institutional
variables are endogenous and a dynamic relationship setting of concentrated and family ownership struc-
exists between variables (Filatotchev, Jackson, and ture in India and Pakistan. The average board size in
Nakajima 2013; Nguyen, Locke, and Reddy 2015; Pakistan is lower than the board size recently
Zhou, Faff, and Alpert 2014). Our data reasonably reported for China (mean 9.372 and median 9) and
possess similar properties. The choice of IVs is cru- US (mean 9.54 and median 9) where state has the
cial in GMM estimation techniques. We use Sargan major stake in the firms or ownership is widely held
(1958)/Hansen (1982) over-identification tests and (see, Conyon 2014; Huang and Wang 2015).
Arellano and Bond (1991) autocorrelation test for Boards are comprised of 63% non-executive
validity and suitability of the IVs.4 directors on average. However, percentage of non-
executive directors has slightly downward trend till
2011, seemingly because of decrease in board size
IV. Empirical results over time. The non-executive directors are more
likely to be an easy target when board size needs to
Descriptive statistics
be reduced.
Table 1 presents the descriptive statistics. CEO com- About 75% observations in our sample are from
pensation is deflated to the base year 2005 using family firms and, quite expectedly, this ratio is almost
Corruption Perception Index. During 2005–2007, stable over the sample period. The largest shareholder
CEO compensation increases first then appears to appears to be holding more than 30% average voting
be decreasing in financial crisis period till 2010. shares, indicating a highly concentrated ownership
After that again there is an upward trend. environment. Interestingly, ownership concentration
Consistently lower median value than mean value has slightly increasing trend over time which could
indicates that the distribution of compensation is have implications for CEO compensation contracts.
positively skewed suggesting that greater number of On average, about 34% CEOs appear to be hold-
CEOs are receiving pay that is less than overall ing the position of chairman board of directors also.
average pays. CEO duality shows maximum value, that is, 36% in
Mean ROA of pooled sample is 10.99% with SD of 2008, 2009 and 2010, the time characterized by poli-
13.39% while median is 9.51%. Average ROA does tical unrest, energy crisis and financial crisis.
not seem to vary abnormally across the years. ROA However, recent downward trend in 2011 and 2012
decreases from 12.52% in year 2006 to 10.17% in seems to be consistent with greater emphasis on
year 2007 followed by 9.78% in year 2008 and 9.26% separating the position of chairman and CEO in

4
See Roodman (2009) for details.
 APPLIED ECONOMICS 9

Table 1. Descriptive statistics.

Variables Statistic 2005 2006 2007 2008 2009 2010 2011 2012 Pooled Sample
Cash pay Mean 3353 3631 4092 3958 3784 3639 3720 4227 3810
Rs in ‘000’ Median 2000 2134 2338 2031 2165 1909 1979 2338 2104
base 2005 SD 4626 5412 6706 6636 5588 5362 5173 6656 5819
Total pay Mean 4817 5125 5609 5414 5292 5229 5255 5887 5341
Rs in ‘000’ Median 3156 3136 3577 3260 3151 3021 2997 3389 3210
base 2005 SD 5854 6515 7833 7896 7108 6927 6672 8986 7306
Return on Mean 11.93 12.52 10.17 9.78 9.26 12.28 11.74 10.54 10.99
assets % Median 9.14 9.58 7.86 7.96 9.53 11.20 10.84 10.32 9.51
SD 11.07 10.67 11.23 11.52 19.31 11.73 12.96 14.80 13.39
Market Mean 27.70 3.49 24.87 −14.35 −1.73 8.37 −0.47 86.46 16.74
return % Median 18.97 −1.01 7.38 −27.20 −18.04 −6.14 −9.99 47.84 −0.84
SD 52.18 39.08 71.94 80.74 67.19 62.89 72.12 114.89 80.04
Ownership Mean 32.72 33.01 33.07 33.23 33.36 33.61 34.39 35.51 33.67
concentration Median 26.46 26.61 26.71 26.25 26.13 26.13 27.45 29.40 26.66
% SD 21.01 21.22 20.78 20.48 20.44 20.30 20.74 21.32 20.74
Family Mean 0.75 0.75 0.76 0.75 0.74 0.75 0.75 0.74 0.75
ownership Median 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
SD 0.44 0.43 0.43 0.43 0.44 0.43 0.43 0.44 0.43
Board Mean 8.13 8.15 8.08 8.04 8.01 8.00 8.01 8.00 8.05
size Median 7.00 7.00 7.00 7.00 7.00 7.00 7.00 7.00 7.00
Downloaded by [Southern Cross University] at 09:18 09 October 2017

SD 1.68 1.68 1.62 1.57 1.53 1.51 1.51 1.51 1.57

Board Mean 64.32 64.31 63.90 63.65 62.85 63.23 62.58 63.91 63.54
independence Median 70.00 70.71 70.00 66.67 66.67 66.67 62.50 66.67 66.67
% SD 20.37 20.72 20.71 20.59 20.94 20.64 20.31 19.96 20.49
CEO duality Mean 0.30 0.33 0.33 0.36 0.36 0.36 0.35 0.33 0.34
Median 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
S.D 0.46 0.47 0.47 0.48 0.48 0.48 0.48 0.47 0.47
Institutional Mean 13.93 13.48 13.10 13.92 13.22 12.21 11.60 10.65 13.22
ownership Median 10.82 9.90 10.06 11.31 10.00 10.14 8.75 7.34 10.00
% SD 12.52 11.91 10.78 11.71 11.19 10.39 10.50 9.93 11.19
Group firm Mean 0.63 0.63 0.63 0.62 0.60 0.59 0.60 0.62 0.61
Median 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
SD 0.49 0.49 0.49 0.49 0.49 0.49 0.49 0.49 0.49
Total assets Mean 7173 8607 9827 10,898 11,636 12,675 14,490 17,215 11,833
Rs in Median 2385 2603 2970 3042 3034 3068 3544 3831 3061
millions SD 14,281 16,821 18,677 22,158 26,033 31,098 35,893 45,860 29,092
Firm risk % Mean 15.03 12.15 12.70 13.24 24.29 17.88 14.07 15.97 15.88
Median 13.54 11.20 12.13 11.66 19.32 13.18 11.15 13.29 13.01
SD 8.00 8.51 6.63 7.74 23.59 15.63 11.00 11.37 13.59
Market to Mean 1.79 2.35 2.17 1.74 1.56 1.31 1.30 1.84 1.73
book Median 1.19 0.99 1.13 0.86 0.62 0.52 0.43 0.65 0.75
SD 2.37 8.17 4.76 4.77 4.81 6.03 4.58 5.62 5.35
Firm age Mean 31.03 31.84 32.56 32.87 34.01 35.06 36.14 37.50 34.06
years Median 27.00 28.00 29.00 29.00 30.00 31.00 32.00 33.00 30.00
SD 16.65 16.75 16.48 16.43 16.61 16.56 16.46 16.59 16.65

Pakistan. This recent trend is similar to the UK and inflation factor (VIF) as none of the VIF is greater
China where emphasis is on separating the post of than the commonly used threshold level of 10
CEO from the chairman, and unlike the United (unreported).
States, where it is usual to combine these two posi- Both cash and total compensation have positive
tions (Conyon and He 2012). correlation with accounting performance (ROA), but
correlation with market performance (market
return) is not significant although sign is positive.
Correlation analysis Board structure variables board size and board inde-
Table 2 presents the correlation matrix. Log trans- pendence are also positively correlated with com-
formation is performed for compensation variables pensation, indicating potential ineffectiveness of
and firm size proxy, that is, total assets. board structure in monitoring and reducing CEO
Multicollinearity does not seem to be a problem in compensation. Family firms seem to pay lower com-
the data as none of the absolute values of correlation pensation to CEO as depicted by negative correla-
coefficients between explanatory variables is greater tion between family ownership and both measures of
than 0.70. This is further confirmed by variance CEO compensation.
 10 M. F. SHEIKH ET AL.

Table 2. Correlation matrix.

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15)
Total compensation (1) 1.00
Cash compensation (2) 0.97* 1.00
Return on assets (3) 0.31* 0.31* 1.00
Market return (4) 0.02 0.02 0.17* 1.00
Ownership (5) 0.36* 0.37* 0.22* 0.05 1.00
concentration
Family ownership (6) −0.38* −0.38* −0.22* 0.02 −0.54* 1.00
Board size (7) 0.32* 0.31* 0.13* −0.00 0.10* −0.45* 1.00
Board independence (8) 0.19* 0.14* 0.03 −0.02 0.06* −0.23* 0.35* 1.00
CEO duality (9) −0.29* −0.28* −0.17* 0.01 −0.08* 0.17* −0.26* −0.19* 1.00
Institutional ownership (10) 0.15* 0.12* −0.04 −0.06* −0.15* 0.00 0.14* 0.16* −0.06* 1.00
Group firm (11) 0.08* 0.06* 0.08* 0.00 −0.17* 0.28* −0.06* 0.04 −0.16* 0.07* 1.00
Firm size (12) 0.58* 0.56* 0.17* −0.04 0.25* −0.28* 0.38* 0.13* −0.22* −0.00 0.06* 1.00
Firm risk (13) −0.20* −0.20* −0.19* 0.20* −0.04 0.12* −0.13* −0.07* 0.20* 0.00 −0.10* −0.24* 1.00
Market to book (14) 0.09* 0.12* 0.18* 0.07* 0.28* −0.17* 0.08* −0.12* −0.10* −0.13* 0.04 0.08* −0.09* 1.00
Firm age (15) 0.25* 0.24* 0.07* 0.04 0.20* −0.22* 0.19* 0.05* −0.02 0.07* 0.01 0.12* −0.03 0.04 1.00
*p-Value <0.05.
Downloaded by [Southern Cross University] at 09:18 09 October 2017

Inconsistent with agency theory, ownership con- different model specifications. This confirms that
centration is positively correlated with compensation. despite CEOs are seemingly more powerful than
Similarly, CEO duality is negatively correlated with the boards, their compensation is still linked to
CEO compensation which is inconsistent with both firms’ accounting performance. This finding seems
agency theory and managerial power theory. Firm to be inconsistent with managerial power hypothesis
size, market to book and firm age appear to have and expropriation view as discussed earlier.
positive correlation with compensation. However, Market returns do not significantly contribute to
CEO compensation is reduced as firm risk increases, pay setting process as coefficient of market return is
which is inconsistent with the argument that risky not significantly different from zero in all models.
firms need to pay higher compensation to their CEOs. Since CEO compensation in Pakistan rarely includes
any restricted stocks, stock options and other stock-
based bonuses, weak link between CEO compensa-
Estimation results
tion and market performance is expected. Another
Table 3 presents the estimation results for total possible reason for an insignificant relationship
compensation and cash compensation as dependent between CEO compensation and market perfor-
variables. Robust SEs are reported for pooled and mance could be that Pakistani bourses are consid-
fixed effect models, while Wind-Meijer-corrected ered to be highly volatile (Sheikh and Riaz 2012);
SEs are reported for dynamic panel models. therefore using market performance as benchmark
Arellano–Bond serial correlation tests m1 and m2, for setting CEO compensation may not be a good
and instrument over-identification tests are also choice.
reported at the bottom of the table. Arellano–Bond We also find positive pay–performance link for
test for second-order (m2) validates the use of sec- past accounting performance when we replace cur-
ond and earlier lags of dependent variables as instru- rent firm performance variables with their lagged
ments. None of the values of m2 rejects the values in the models (unreported). However, sur-
hypothesis of no second-order correlation in error prisingly, lagged market performance appears to
terms. Similarly, p-values of over-identification tests, have negative association with both measures of
Sargan test and Hansen J test do not lead to rejecting CEO compensation. This negative association may
the hypothesis of joint validity of the instruments be interpreted as sign of cronyism which predict
used. negative association between excessive pay and firm
Consistent with the agency theory, both measures performance (Brick, Palmon, and Wald 2006).
of CEO compensation are positively related to cur- However, we believe that this negative association
rent firm accounting performance as measured by is more probable due to highly volatile bourses in
ROA. The results are consistent qualitatively over Pakistan. During the periods of bad-market
 APPLIED ECONOMICS 11

Table 3. CEO compensation and firm performance.

Log of total compensation Log of cash compensation
(1) (2) (3) (4) (5) (6)
Variables Pooled Fixed effects Dynamic panel (GMM) Pooled Fixed effects Dynamic panel (GMM)
Lagged compensation – – 0.7989*** – – 0.8921***
(0.0582) (0.0524)
Return on assets 1.2984*** 1.3044*** 0.5618*** 1.3316*** 1.3326*** 0.5025***
(0.3746) (0.3412) (0.1581) (0.3936) (0.3651) (0.1660)
Market return 0.0258 −0.0323 −0.0592 0.0362 −0.0225 −0.0622
(0.0267) (0.0247) (0.0631) (0.0273) (0.0260) (0.0749)
Ownership concentration 0.9356*** 0.7369** 0.1707* 1.0373*** 0.8581*** 0.0676
(0.3214) (0.2917) (0.0937) (0.3340) (0.3118) (0.0954)
Family ownership −0.3407** −0.0552 −0.0490 −0.3363* −0.0918 −0.0331
(0.1568) (0.1577) (0.0441) (0.1781) (0.1760) (0.0457)
Board size −0.0082 −0.0081 0.0066 0.0060 0.0141 0.0101
(0.0408) (0.0379) (0.0097) (0.0469) (0.0431) (0.0112)
Board independence 0.2134 0.1669 0.0193 −0.0128 −0.0372 0.0445
(0.2788) (0.2556) (0.0723) (0.2934) (0.2763) (0.0648)
CEO duality −0.2955** −0.1973* −0.0697* −0.2842** −0.2018* −0.0207
(0.1153) (0.1046) (0.0376) (0.1224) (0.1113) (0.0365)
Institutional ownership 1.6531*** 1.2969*** 0.2414* 1.5737*** 1.2684*** 0.1105
(0.4391) (0.3908) (0.1409) (0.4702) (0.4121) (0.1537)
Downloaded by [Southern Cross University] at 09:18 09 October 2017

Group firm 0.1784 0.1553 0.0299 0.1279 0.1085 0.0120

(0.1222) (0.1053) (0.0341) (0.1335) (0.1141) (0.0320)
Firm size 0.3516*** 0.3492*** 0.0671*** 0.3516*** 0.3540*** 0.0410**
(0.0335) (0.0351) (0.0205) (0.0357) (0.0377) (0.0178)
Firm risk −0.1904 −0.0907 −0.0645 −0.2956 −0.2104 0.0205
(0.1951) (0.1955) (0.1603) (0.2028) (0.1956) (0.1965)
Market to book ratio −0.0066 −0.0073 0.0016 −0.0024 −0.0046 0.0014
(0.0086) (0.0096) (0.0032) (0.0082) (0.0092) (0.0027)
Firm age 0.0084** 0.0063** 0.0006 0.0081** 0.0054 −0.0005
(0.0035) (0.0031) (0.0008) (0.0038) (0.0034) (0.0008)
Change of CEO −0.1156 −0.1116 0.0308 −0.1618** −0.1518** 0.0278
(0.0757) (0.0696) (0.0387) (0.0812) (0.0751) (0.0369)
Observations 1,508 1,508 1277 1,508 1,508 1,277
R-squared 0.4954 0.6216 – 0.4681 0.5830 –
Arellano–Bond
AR(1) in diff. (m1) p-value – – 0.000 – – 0.000
AR(2) in diff. (m2) p-value – – 0.764 – – 0.377
Over identification test
Sargan test p-value – – 0.149 – – 0.358
Hansen J statistic p-values – – 0.788 – – 0.489
Dependent variable is log of compensation, return on assets is measured by earnings before interest and taxes divided by total assets, market return is
measured by current market price plus dividend divided by previous year market price minus one, ownership concentration is measured by voting shares
held by the largest shareholder, family ownership is a dummy variable taking value one for family firms and zero otherwise, board size is measured by
number of directors on board, board independence is measured by ratio of non-executive directors to board size, CEO duality is a dummy variable taking
value one if CEO is also chairman board of directors, institutional ownership is measured by percentage of total shares held by institutions. Group firm is a
dummy variable taking value 1 if a firm is affiliated with a group and zero otherwise, Firm Size is measured by log of total assets, Firm Risk is measured by
SD of monthly stock returns over the fiscal year, market to book ratio is measured by market value divided by book value per share, firm age is number of
years of establishment of the firm as mentioned in annual reports, change of CEO is interaction term of dummy variable taking value 1 if CEO is replaced
during the year and number of CEOs drawing remuneration during the year.
Robust SEs in parentheses for pooled and fixed effect models
Windmeijer-corrected SEs in parentheses for dynamic panel model
***p < 0.01, **p < 0.05, *p < 0.1

performance, especially from 2008 to 2011, the com- previous-year CEO compensation as a reference
pensation may have increased due to positive- point while setting next-year compensation. This is
accounting performance, leading to negative rela- an important finding of this study as most of the
tionship between current compensation and pre- literature on CEO compensation ignores the
vious-year market performance. dynamic nature of CEO compensation and estimates
Consistent with pay persistence and dynamic static pay models, considering that pay is in
adjustment arguments as discussed above, the coef- equilibrium.
ficients of lagged CEO pay in dynamic panel models Ownership concentration appears to have positive
are positive and significant, indicating that CEO pay impact on both measures of CEO compensation which
is highly persistent and takes time to adjust to its is consistent with our hypothesis. The coefficients of
long-term equilibrium level. The boards consider ownership concentration are significantly positive. This
 12 M. F. SHEIKH ET AL.

is consistent with expropriation view which leads to expropriation through higher CEO compensation is
support Hypothesis 2a. Controlling shareholders might not supported in Pakistan.
be engaged in colluding with management to get per- Group affiliation is not creating any significant
sonal benefits at the expense of minority shareholders difference in level of CEO compensation as indi-
thus overlooking level of CEO compensation. cated by insignificant coefficients. However, insti-
In pooled regression, CEOs seem to receive lower tutional holding does have significant positive
compensation in family firms. However, in fixed influence on CEO compensation. This questions
effect and dynamic panel model, the coefficients of the monitoring role of institutional ownership as
family ownership do not appear to be significant. suggested by the agency theory (see Hartzell and
Thus, we find weak evidence that family ownership Starks 2003).
significantly influences the CEO pay setting process. Firm size as measured by log of total assets
The inconsistencies in results across different models appears to be significantly positively related to
need to be explored further using more variables on both measures of CEO compensation in all models.
family characteristics as discussed by Bertrand and This supports the argument that larger firms are
Schoar (2006). complex and difficult to run and hence require
Downloaded by [Southern Cross University] at 09:18 09 October 2017

Inconsistent with many studies (e.g. Core, quality CEOs with higher compensations. As
Holthausen, and Larcker 1999; Croci, Gonenc, expected, firm age is positively associated with
and Ozkan 2012; Fahlenbrach 2009; Ozkan 2011; CEO compensation while newly appointed CEOs
Shah, Javed, and Abbas 2009; Van Essen, Otten, are more likely to start with lower compensation
and Carberry 2012b), the coefficient of board size than the leaving CEOs as suggested by some sig-
is consistently insignificant in all models which nificant negative coefficients of CEO changed.
suggests that board size does not matter. Other control variables, firm risk and market to
Similarly, the coefficient of board independence book, do not appear to influence CEO compensa-
is insignificant in all models, indicating no influ- tion decisions.
ence of non-executive directors on CEO pay pro-
cess. These results are consistent with the general
view that boards in Pakistan are dominated by Robustness checks
non-executive directors who are hired from Although results presented are robust across differ-
within the family or they have close relation ent model specifications, we carry out some further
with controlling shareholders (Javid and Iqbal tests of robustness of our results. First, all the con-
2008). These are grey directors who do not have tinuous variables are winsorized using 1% level at
real influence on decisions made by family/con- both tails to eliminate potential outliers and all
trolling shareholders. The number of such direc- models are re-estimated. But, the results do not
tors does not matter hence the board size does change qualitatively; therefore, it is decided to
not matter. Thus, our Hypotheses 3 and 4 are report the original data results. Second, alternative
supported. measures of firm accounting performance, firm size
Surprisingly, in contrast to many existing studies and ownership concentration, such as, log of net
(e.g. Boyd 1994; Brick, Palmon, and Wald 2006; sales, earnings per share and voting shares held by
Conyon and He 2012; Core, Holthausen, and three largest shareholders respectively, are incorpo-
Larcker 1999; Fahlenbrach 2009; Van Essen, Otten, rated. Again, the results remain qualitatively similar
and Carberry 2012b), CEO duality appears to be to as reported above. Third, to control for endo-
significantly negatively related to both measures of geneity problem, following a number of studies
CEO compensation. The coefficients of CEO duality (e.g. Croci, Gonenc, and Ozkan 2012; Ozkan
are consistently negative in all model specifications. 2011) current values of all independent variables
Thus, the argument that more concentrated power except firm age and change of CEO in model (1)
in one person by combining the position of CEO are replaced with their lagged values treating them
and chairman board of directors leads to as potential cause of endogeneity. However, again,
 APPLIED ECONOMICS 13

results remain largely unaltered. We also estimate collusions between management and largest share-
models by incorporating group, family ownership, holder for rent extraction. In addition, we find weak
ownership concentration, institutional ownership evidence that CEOs in family firms receive lower
individually, but results remain largely the same as compensation than their counterparts. Thus, overall,
reported. Since correlation between these variables ownership structure does not affect CEO compensa-
and VIF are within acceptable range, we decided to tion as suggested by agency theory.
report them in one model. Board structure variables, board size and board
independence, have no convincing influence on
CEO compensation in any direction, indicating that
V. Conclusions
board size and the number of non-executive direc-
In Pakistan, investor protection and legal systems are tors do not matter when CEO compensation is set.
weak. Therefore, firms’ ownership is concentrated in Although agency theory suggest that CEO duality
few individuals or families, leading to more agency leads to higher CEO compensation, we find that
problems between controlling and minority share- CEO duality actually leads to lower CEO compensa-
holders. Further, Code of corporate governance issued tion in Pakistan.
Downloaded by [Southern Cross University] at 09:18 09 October 2017

to improve corporate governance practices in Pakistan Overall, our findings suggest that firms’ account-
has much emphasis on the board structure. Given the ing performance is an important determinant of
Pakistani context, we study how firm performance, CEO compensation. However, ownership concentra-
concentrated/family ownership and board structure tion and board structure variables are not affecting
contribute towards CEO pay setting process. CEO compensation in the way suggested by agency
Using different model specifications including a theory. Therefore, in order to understand CEO pay
dynamic panel model that also account for dynamic puzzle and corporate governance in emerging mar-
adjustment of dependent variable and control for kets, future research needs to account for differences
endogeneity problem, we find that current- and pre- in institutional context of the market under
vious-year firm accounting performances have sig- examination.
nificant positive influence on CEO compensation.
However, firm current market performance does
not have any influence on CEO compensation, sur- Disclosure statement
prisingly previous year’s market performance seems No potential conflict of interest was reported by the authors.
to be negatively influencing the CEO compensation.
We believe that this negative association is mainly
driven by highly volatile markets in Pakistan which References
are preventing boards to base CEO compensation on
Achleitner, A.-K., N. Günther, C. Kaserer, and G. Sicilianb.
market performance. 2014. “Real Earnings Management and Accrual Based
An important finding of this study is that CEO Earnings Management in Family Firms.” European
pay is highly persistent and takes time to adjust to Accounting Review 23: 431–461.
long-run equilibrium. Thus, current CEO compen- Anderson, R. C., and D. M. Reeb. 2003. “Founding Family
sation also depends upon their previous-year com- Ownership and Firm Performance: Evidence from the
S&P 500.” Journal of Finance 58: 1301–1329.
pensation and their innate time-invariant
Arellano, M., and S. Bond. 1991. “Some Tests of Specification
capabilities which are not fully observable initially for Panel Data: Monte Carlo Evidence and an Application
but gradually in subsequent periods through their to Employment Equations.” Review of Economic Studies
real outputs. Further, the boards use previous-year 58: 277–297.
CEO compensation as starting point to set next-year Basu, S., L.-S. Hwang, T. Mitsudome, and J. Weintrop. 2007.
compensation. “Corporate Governance, Top Executive Compensation
and Firm Performance in Japan.” Pacific-Basin Finance
Inconsistent with agency theory, we find that
Journal 15: 56–79.
ownership concentration has positive impact on Bayless, M. 2009. “The Myth of Executive Compensation: Do
CEO compensation. This finding supports the rent Shareholders Get What They Pay For?.” Applied Financial
extraction view. There may be some sort of Economics 19: 795–808.
 14 M. F. SHEIKH ET AL.

Bebchuk, L. A., and J. A. Fried. 2003. “Executive Durnev, A., and E. H. Kim. 2005. “To Steal or Not to Steal:
Compensation as an Agency Problem.” Journal of Firm Attributes, Legal Environment, and Valuation.”
Economic Perspectives 17: 71–92. Journal of Finance 60: 1461–1493.
Bender, R. 2003. “How Executive Directors’ Remuneration Is Fahlenbrach, R. 2009. “Shareholder Rights, Boards, and CEO
Determined in Two FTSE 350 Utilities.” Corporate Compensation.” Review of Finance 13: 81–113.
Governance 11: 206–217. Fama, E. F., and M. C. Jensen. 1983. “Separation of
Bertrand, M., and A. Schoar. 2006. “The Role of Family in Ownership and Control.” Journal of Law & Economics
Family Firms.” Journal of Economic Perspectives 20: 73–96. 26: 301–325.
Blundell, R., and S. Bond. 1998. “Initial Conditions and Fan, J. P. H., K. C. J. Wei, and X. Xu. 2011. “Corporate
Moment Restrictions in Dynamic Panel Data Modeis.” Finance and Governance in Emerging Markets: A
Journal of Econometrics 87: 115–143. Selective Review and an Agenda for Future Research.”
Boyd, B. K. 1994. “Board Control and CEO Compensation.” Journal of Corporate Finance 17: 207–214.
Strategic Management Journal 15: 335–344. Filatotchev, I., G. Jackson, and C. Nakajima. 2013.
Brick, I. E., O. Palmon, and J. K. Wald. 2006. “CEO “Corporate Governance and National Institutions: A
Compensation, Director Compensation, and Firm Review and Emerging Research Agenda.” Asia Pacific
Performance: Evidence of Cronyism?.” Journal of Journal of Management 30: 965–986.
Corporate Finance 12: 403–423. Frydman, C., and D. Jenter. 2010. “CEO Compensation.”
Bryson, A., J. Forth, and M. Zhou. 2014. “Same or Different? Annual Review of Financial Economics 2: 75–102.
Downloaded by [Southern Cross University] at 09:18 09 October 2017

the CEO Labour Market in China’s Publicly Listed Gallego, F., and B. Larrain. 2012. “CEO Compensation and
Companies.” Economic Journal 124: 90–108. Large Shareholders: Evidence from Emerging Markets.”
Buck, T., X. Liu, and R. Skovoroda. 2008. “Top Executive Pay Journal of Comparative Economics 40: 621–642.
and Firm Performance in China.” Journal of International Ghosh, A. 2006. “Determination of Executive Compensation
Business Studies 39: 833–850. in an Emerging Economy: Evidence from India.” Emerging
Byrd, J., and E. Cooperman. 2010. “Director Tenure and the Markets Finance and Trade 42: 66–90.
Compensation of Bank CEOs.” Managerial Finance 36: 86– Gibson, M. S. 2003. “Is Corporate Governance Ineffective in
102. Emerging Markets?.” Journal of Financial and
Code of Corporate Governance 2002. Securities Exchange Quantitative Analysis 38: 231–250.
Commission of Pakistan, [Link]. 2002. Code of Greene, W. H. 2011. Econometric Analysis. Harlow, UK:
Corporate Governance. Pearson Education.
Code of Corporate Governance (Revised) 2012. Securities Hansen, L. P. 1982. “Large Sample Properties of Generalized
Exchange Commission of Pakistan, [Link]. Method of Moments Estimators.” Econometrica 50: 1029–
2012. Code of Corporate Governance. 1054.
Conyon, M. J. 2014. “Executive Compensation and Board Harris, M., and A. Raviv. 2008. “A Theory of Board Control
Governance in US Firms.” The Economic Journal 124: and Size.” Review of Financial Studies 21: 1797–1832.
F60–F89. Hartzell, J. C., and L. T. Starks. 2003. “Institutional Investors
Conyon, M. J., and L. He. 2011. “Executive Compensation and Executive Compensation.” Journal of Finance 58:
and Corporate Governance in China.” Journal of 2351–2374.
Corporate Finance 17: 1158–1175. Hausman, J. 1978. “Specification Tests in Econometrics.”
Conyon, M. J., and L. He. 2012. “CEO Compensation and Econometrics 46: 1251–1273.
Corporate Governance in China.” Corporate Governance: Ho, S. S. M., K. C. K. Lam, and H. Sami. 2004. “The
An International Review 20: 575–592. Investment Opportunity Set, Director Ownership, and
Core, J. E., R. W. Holthausen, and D. F. Larcker. 1999. Corporate Policies: Evidence from an Emerging Market.”
“Corporate Governance, Chief Executive Officer Journal of Corporate Finance 10: 383–408.
Compensation, and Firm Performance.” Journal of Hofstede, G. H. 1980. Culture’s Consequences: International
Financial Economics 51: 371–406. Differences in Work-Related Values. Beverly Hills, CA:
Croci, E., H. Gonenc, and N. Ozkan. 2012. “CEO Sage.
Compensation, Family Control, and Institutional Holderness, C. G. 2017. “Culture and the Ownership
Investors in Continental Europe.” Journal of Banking & Concentration of Public Corporations around the
Finance 36: 3318–3335. World.” Journal of Corporate Finance 44: 469-486.
Dalton, D. R., C. M. Daily, J. L. Johnson, and A. E. Ellstrand. Huang, Y. S., and C.-J. Wang. 2015. “Corporate Governance
1999. “Number of Directors and Financial Performance: A and Risk-Taking of Chinese Firms: The Role of Board
Meta-Analysis.” The Academy of Management Journal 42: Size.” International Review of Economics and Finance 37:
674–686. 96–113.
Devers, C. E., A. A. Cannella, G. P. Reilly, and M. E. Yoder. Jameson, M., A. Prevost, and J. Puthenpurackal. 2014.
2007. “Executive Compensation: A Multidisciplinary “Controlling Shareholders, Board Structure, and Firm
Review of Recent Developments.” Journal of Performance: Evidence from India.” Journal of Corporate
Management 33: 1016–1072. Finance 27: 1–20.
 APPLIED ECONOMICS 15

Javid, A. Y., and R. Iqbal. 2008. “Ownership Concentration, Provan, K. G. 1980. “Board Power and Organizational
Corporate Governance and Firm Performance: Evidence Effectiveness among Human Service Agencies.” Academy
from Pakistan.” The Pakistan Development Review 47: of Management Journal 23: 221–236.
643–659. Raithatha, M., and S. Komera. 2016. “Executive
Jensen, M. 1993. “The Modern Industrial Revolution. Exit, Compensation and Firm Performance: Evidence from
and the Failure of Internal Control Systems.” Journal of Indian Firms.” IIMB Management Review 28: 160–169.
Finance 48: 831–880. Rehman, R., M. Hasan, I. Mangla, and N. Sultana. 2012.
Jensen, M., and W. Meckling. 1976. “Theory of the Firm: “Economic Reforms, Corporate Governance and
Managerial Behavior, Agency Costs, and Ownership Dividend Policy in Sectoral Economic Growth in
Structure.” Journal of Financial Economics 3: 305–360. Pakistan.” The Pakistan Development Review 51: 133–146.
Jensen, M. C., and J. B. Warner. 1988. “The Distribution Roodman, D. 2009. “How to Do Xtabond2: An Introduction
of Power among Corporate Managers, Shareholders, to Difference and System GMM in Stata.” The Stata
and Directors.” Journal of Financial Economics 20: Journal 9: 86–136.
3–24. Sargan, J. D. 1958. “The Estimation of Economic
Kamran, K., and A. Shah. 2014. “The Impact of Corporate Relationships Using Instrumental Variables.”
Governance and Ownership Structure on Earnings Econometrica 26: 393–415.
Management Practices: Evidence from Listed Schultz, E. L., D. T. Tan, and K. D. Walsh. 2010.
Companies in Pakistan.” The Lahore Journal of “Endogeneity and the Corporate Governance–
Downloaded by [Southern Cross University] at 09:18 09 October 2017

Economics 19: 27–70. Performance Relation.” Australian Journal of

Kashif, S., and K. Mustafa. 2012. “The Determinants of CEO Management 35: 145–163.
Compensation: Evidence from Family-Owned Listed Shah, S. Z. A., T. Javed, and M. Abbas. 2009. “Determinants
Corporations in Karachi Stock Exchange.” Middle East of CEO Compensation Empirical Evidence from Pakistani
Journal of Economics and Finance 5: 45–61. Listed Companies.” International Research Journal of
Kato, T., W. Kim, and J. H. Lee. 2007. “Executive Finance and Economics 32: 148–159.
Compensation, Firm Performance, and Chaebols in Sheikh, M. F., and K. Riaz. 2012. “Overconfidence Bias,
Korea: Evidence from New Panel Data.” Pacific-Basin Trading Volume and Returns Volatility: Evidence from
Finance Journal 15: 36–55. Pakistan.” World Applied Science Journal 18: 1737–1748.
Klapper, L. F., and I. Love. 2003. “Corporate Governance, Siegel, J., and P. Choudhury. 2012. “A Reexamination of
Investor Protection, and Performance in Emerging Tunneling and Business Groups: New Data and New
Markets.” Journal of Corporate Finance 10: 703–728. Methods.” The Review of Financial Studies 25: 1763–1798.
La Porta, R., F. Lopez-De-Silanes, and A. Shleifer. 1999. Su, Z., Y. Li, and L. Li. 2010. “Ownership Concentration and
“Corporate Ownership around the World.” The Journal Executive Compensation in Emerging Economies:
of Finance 54: 471–517. Evidence from China.” Corporate Governance: An
Lipton, M., and J. W. Lorsch. 1992. “A Modest Proposal for International Review 10: 223–233.
Improved Corporate Governance.” Business Lawyer 48: Sun, F., X. Wei, and X. Huang. 2013. “CEO Compensation
59–77. and Firm Performance Evidence from the US Property
Mujtaba, B. G., and T. Afza. 2011. “Business Ethics and Liability Insurance Industry.” Review of Accounting
Perceptions of Public and Private Sector Respondents in and Finance 12: 252–267.
Pakistan.” Far East Journal of Psychology and Business 3: Sun, S. L., X. Zhao, and H. Yang. 2010. “Executive
1–11. Compensation in Asia: A Critical Review and Outlook.”
Murphy, K. J. 1999. “Executive Compensation.” In Handbook Asia Pacific Journal of Management 27: 775–802.
of Labor Economics, eds. O. Ashenfelter and D. Cards. Tosi, H. L., S. Werner, J. P. Katz, and L. R. Gomez-Mejia.
New York: North Holland. 2000. “How Much Does Performance Matter? A Meta-
Nguyen, T., S. Locke, and K. Reddy. 2015. “Ownership Analysis of CEO Pay Studies.” Journal of Management
Concentration and Corporate Performance from a 26: 301–339.
Dynamic Perspective: Does National Governance Quality Tversky, A., and D. Kahneman. 1974. “Judgement under
Matter?.” International Review of Financial Analysis 41: Uncertainty: Heuristics and Biases.” Science 185: 1124–1131.
148–161. Van Essen, M., P. P. M. A. R. Heugens, J. Otten, and J. Van
Ozkan, N. 2007. “Do Corporate Governance Mechanisms Oosterhout. 2012a. “An Institution-Based View of
Influence CEO Compensation? An Empirical Executive Compensation: A Multilevel Meta-Analytic
Investigation of UK Companies.” Journal of Test.” Journal of International Business Studies 43: 396–423.
Multinational Financial Management 17: 349–364. Van Essen, M., J. Otten, and E. J. Carberry. 2012b. “Assessing
Ozkan, N. 2011. “CEO Compensation and Firm Managerial Power Theory: A Meta-Analytic Approach to
Performance: An Empirical Investigation of UK Panel Understanding the Determinants of CEO Compensation.”
Data.” European Financial Management 17: 260–285. Journal of Management 41: 164–202.
 16 M. F. SHEIKH ET AL.

Wang, K., and X. Xiao. 2011. “Controlling Shareholders’ World Bank. 2005. Pakistan - Report on the Observance of
Tunneling and Executive Compensation: Evidence from Standards and Codes (ROSC): Corporate Governance
China.” Journal of Accounting and Public Policy 30: 89– Country Assessment. Washington, DC: World Bank.
100. Young, M. N., M. W. Peng, D. Ahlstrom, G. D. Bruton, and
Weisbach, M. S. 2007. “Optimal Executive Compensation Y. Jiang. 2008. “Corporate Governance in Emerging
versus Managerial Power: A Review of Lucian Bebchuk Economies: A Review of the Principal–Principal
and Jesse Fried’s Pay without Performance: The Unfufilled Perspective.” Journal of Management Studies 45: 196–220.
Promise of Executive Compensation.” Journal of Economic Zhou, Q., R. Faff, and K. Alpert. 2014. “Bias Correction
Literature 45: 419–428. in the Estimation of Dynamic Panel Models in
Wooldridge, J. M. 2002. Econometric Analysis of Cross Corporate Finance.” Journal of Corporate Finance 25:
Section and Panel Data. Cambridge, MA.: MIT Press. 494–513.
Downloaded by [Southern Cross University] at 09:18 09 October 2017