Master Thesis
Does CEO compensation affect corporate risk-taking? A
closer look at the technology industry
By
Yibo Wang
Supervisor: Dr. H. Gonenc
Co-Assessor: Dr. W. Westerman
Submitted for the degree of
MSc International Financial Management
Double Degree with Uppsala University
Faculty of Economics and Business
Rijksuniversiteit Groningen
Uppsala Universitet
y.wang.79@student.rug.nl
Abstract
This thesis examines the effect of executives compensation on corporate risk-taking using
19,026 firm-year observations over the period of 2004 to 2015. The study especially
emphasizes the characteristics of the technology industry and its inherently high levels of
uncertainty. The individual level of corporate governance is developed as a firm-level
moderator, and levels of investor protection, uncertainty avoidance and individualism as
country-level moderators for corporate governance and cultural dimensions. It is concluded
that there is a positive relationship between executive compensation and risk-taking and
investor protection positively moderates this relation. Besides, uncertainty avoidance negatively moderates the relation between compensation and risk-taking. Both effects are
found to be stronger in the technology sector. I find country-level variables but not firm-level
variables to be determinants of the compensation-risk relationship.
Abstract 1
1. Introduction 3
2. Literature review and hypotheses development 8
2.1 Literature review 8
2.1.1 CEO compensation and agency theory 10
2.1.2 CEO compensation in technological sector 12
2.2 Hypotheses development 14
2.2.1 Relationship between CEO compensation and corporate risk-taking 14 2.2.2 Firm-level Governance as a moderator of the Compensation-Risk relationship 16 2.2.3 Country-level Governance as a moderator of the Compensation-Risk relationship 20 2.2.4 Cultural Dimensions as moderators of the Compensation-Risk relationship 23
Figure 1. Conceptual model 26
3. Methodology, Data and Sample 26
3.1 Methodology 26
3.1.1 Dependent variable 26
3.1.2 Independent variable 27
3.1.3 Firm-level corporate governance variables 28
3.1.4 Country-level variables 28
3.1.5 Control variables 29
3.1.6 Regression models 30
3.2 Data and sample 32
4. Results 33
4.1 Descriptive statistic and correlations 33
4.1.1 Descriptive statistics 34
4.1.2 Correlations and multicollinearity 37
4.2 Regression results 38
5. Conclusion 45
Acknowledgement 49
1. Introduction
Not least because of the financial crisis in 2008 and the still perceptible consequences up to
this point, corporate risk-taking is in the focus of public attention. Excessive corporate
risk-taking was said to be one of the key triggers for the last financial crisis that expanded to a
whole economic crisis. A balanced level of corporate risk-taking is therefore not only crucial
for the success of each individual firm, but also for the stability of the economy as a whole.
The excessive levels of risk that are against laws and regulations, executives can also damage
their firms with excessive levels of risk-taking within the boundaries of law. The levels of
risk-taking are a key determinant of firms successes or failures and therefore present a very
interesting field of study. Especially the technology sector is characterized by enormously
high levels of uncertainty in the market. The pace of technological change has been growing
for years in is expected to continue to grow at an even higher pace. In this highly risky
environment one might wonder how firms and shareholders in particular manage to keep the
level of risk in their corporates’ decision-making procedures within acceptable boundaries.
Not only high risks can lead to firm failure, but also too low levels of risk-taking can harm a
firm’s financial success because the firm misses out the technological developments of the
future.
The former CEO of Ford in the 1960s, Lee Iacocca, made this famous statement about risk in
the technology sector:
“The greatest hindrance of business success in our industry is managerial overthinking. At some point, you just need to embark on a journey to the unknown. The best decision can be
wrong if it takes too long.” (Lee Iacocca, 1972)
One tool for shareholders to direct executive risk-taking is executive remuneration. The
design of compensation packages is an effective tool to overcome managers’ individual risk
tendencies and direct their level of risk towards the shareholders’ expectations. The balance
between fixed pay elements and incentive pay elements is crucial in this respect. This paper
aims to contribute to the existing body of literature as it develops hypotheses about
and thereby takes a closer look at the technology sector. I will now shortly review relevant literature and derive my research question.
Numerous studies have been done about the relationship between CEO compensation and
organizational performance (Murphy, 1985; Kren & Kerr, 1997; Carpenter et al., 2003; Brick
et al., 2006; Firth et al., 2006; Kwon & Yin, 2006; Makri et al., 2006; Lin et al., 2011; Banker
et al., 2012). Most of the researchers suggest that there is a positive relation between these
two variables (Murphy, 1985; Carpenter et al., 2003; Brick et al., 2006; Firth et al., 2006;
Banker et al., 2012), some however also find a negative relationship. Banker et al. (2012) find
a negative relation between future Return on Equity (ROE) and executives bonus. No direct
association between pay and financial performance is found by Kren & Kerr (1997).
Given that CEO pay and performance are mostly considered to positively correlated, CEOs
will be inspired to make risky decisions and take higher levels of risk when they receive more
compensation, leading to a better performance. Therefore, higher levels of CEO
compensation will motivate CEOs to take higher risk, indicating that corporate risk taking is
positively affected by CEO compensation. The importance of executives compensation
packages has attracted much attention lately and it has developed into a part of corporate
governance policy.
Many researchers have examined the relationship between executives compensation and
corporate risk-taking (Aggarwal & Samwick, 1998; Carpenter et al., 2003; Dee et al., 2005;
Brick et al., 2006; Coles et al., 2006; Fernandes et al., 2011). Prendergast (2000) finds that
there is a positive relationship between executives incentive compensation and uncertainty.
Additionally, Dee et al. (2005) state that firms ought to make tradeoff decisions between
executives compensation and corporate risk-taking, as managers will request higher pay for
the greater risk they bear, especially in a dynamic fast-changing environment, such as the
technology sector. Jensen and Meckling (1976) restate agency theory in a way that managers
will make an effort to enhance their private wealth and interest, in order to ensure personal
income and the stability of their job. Coles et al. (2006) maintain that there is a propensity that
managers need to decide from different investments, which will consequently harm shareholder value.
Technology has been playing a critical role in the growth and development of the society and
economy in the 21st century (Makri, et al, 2006). The rapid growth in technology improved
social development to a great extent, and innovation in production and advancement in
production methods play a significant role in economic development (Barney, 1991; Balkin et
al, 2000). Rogers (2001) shows that the number of high-tech firms is booming since
technology has been a driving factor for social sustainable development, and the large number
of emerging high-tech firms is one of the primary driving forces in modern economic growth.
In addition, huge changes in competitive markets provides enormous space for the
improvement of innovation and creativity (Rogers, 2001). Ryan et al. (2002) state that
executives compensation packages are different from those of other industry since there is
higher R&D intensity in technological industry. R&D projects are highly risky themselves, so
firms need to establish different managers compensation structures and higher level of pay to
encourage decision makers to invest in those projects even though they are risky (Rogers,
2001).
A large number of studies have been done about the relationship between executives
compensation and corporate risk-taking, but not many of them focus on technological firms.
Therefore, it is interesting to study the relationship between executives compensation and
risk-taking in technological industry. This thesis contributes to the existing research by
studying the compensation-risk relationship in technological industry.
Besides, firm-level corporate governance characteristics have an effect on the correlation
between executives compensation and corporate risk-taking as well according to previous
study (Jain et al., 2014; John et al., 2008). A higher level of corporate governance can fixate
corporate executive’s remuneration at a level where it is best in order to serve shareholder’s
interest and prevent or at least minimize the extraction of private benefits by corporate
executives. If firms have high levels of corporate governance, they can more efficiently
overcome their internal agency problems and therefore display better performance measures
their monitoring capabilities with higher degree of corporate governance and will reach a more efficient level of risk-taking and higher quality of corporate performance, which
suggests the influence of corporate governance on compensation-risk relationship. Therefore,
it is also interesting to research the effect of firm-level corporate governance characteristics on
compensation-risk relationship in this study.
Further, some country-level characteristics like investor protection also have an impact on
executives compensation and corporate risk-taking besides firm-level corporate governance
characteristics and industry characteristics. According to Zheng et al. (2016), it is suggested
that there is a positive relationship between investor protection and executives compensation.
In countries with stronger investor protection, agents will be less encouraged to act in their
own private interests and have more motivation to generate more profits for shareholders
(Zheng et al., 2016). John et al. (2008) point out that the alignment of interests between
shareholders and managers can be accomplished with better investor protection, leading
managers to bear higher risk. On the other hand, stakeholders will have bigger influence on
companies when firms locate in countries with poor investor protection. Agents will
undertake lower level of corporate risk as a result of the pressure from these stakeholders
(John et al, 2008). Stronger investor protection will encourage managers to bear greater risk,
and invest in riskier projects with higher profits (Houston et al., 2010).
Moreover, previous studies show that national culture also has an effect on management control systems and corporate governance practices, such as executives compensation
structure (e.g. Hofstede et al., 1991; Chow et al., 1999; Van der Stede, 2003; Jansen et al.,
2009). Jansen et al. (2009) indicate that uncertainty avoidance and individualism as cultural
dimensions have the most critical influence on the design of executives compensation packages.
Considering these firm and country-level moderators of the compensation-risk relationship
and the the special characteristics of the technology sector, the research question of this thesis
How do firm and country-level characteristics affect the relationship between CEO compensation and risk-taking? Does the higher level of uncertainty cause these effects to be different in the technology sector?
Over a sample period from 2004 to 2015, I construct a dataset of 19,026 firm-year
observations and conduct statistical analysis to answer the posed question. I subdivide my
sample into a technology and non-technology cluster to highlight the special characteristics of
the technology sector. I find country-level variables to be significant determinants of the
compensation-risk relationship opposed to firm-level variables that don’t have an effect.
Investor protection positively moderates the relationship, whereas uncertainty avoidance
moderates it negatively. Both effects are found to be stronger in the technology sector due to
its high level of inherent uncertainty and higher required levels of corporate risk-taking.
The rest of the thesis is structured as follows. Section 2 provides a brief review and discussion
about previous research over the relationship between executives compensation and corporate
risk-taking, and hypotheses development. Section 3 presents data collection and regression
models. This thesis uses Linear regression method to test the hypotheses developed in section
2. Literature review and hypotheses development
The following section will proceed as follows: First, I will summarize the relationship
between CEO compensation and corporate risk-taking from previous literature. In a next step,
I will highlight the specific characteristics of the technology sector and their effect on the
relationship of CEO remuneration and corporate risk-taking. I will then proceed to the
development of my hypothesis.
2.1 Literature review
Numerous studies have been conducted in order to discuss the impact of executive
remuneration on organizational performance and risk-taking (e.g. Aggarwal & Samwick,
1998; Cui & Mak, 2002; Carpenter et al., 2003; Dee et al., 2005; Brick et al., 2006; Coles et
al., 2006; Finkelstein et al., 2009; Fernandes et al., 2011). Aggarwal and Samwick (1998)
indicate that there is a positive relationship between executives incentive compensation and
risk-taking. Coles et al. (2012) show, that CEOs should get convex compensation to reduce
the propensity that risk-averse managers avoid risky investments. They emphasize the importance of CEO motivation through specific incentive structures. These incentive
structures mainly refer to a heightened use of performance-based compensation components
(e.g. bonuses) compared to fixed compensation components (e.g. base salary). Scholars further conclude, that pay-performance sensitivity is negatively associated with risk
(Aggarwal & Samwick, 1998; Jin, 2002; Dee et al., 2005). Therefore, performance-based pay
is positively evaluated by many scholars in order to induce managers to make right investment
decisions and choose positive Net Present Value projects and consequently create more value
(Prendergast, 2000; Dee et al., 2005; Coles et al., 2006). The study of Ayadi et al. (2012)
further suggests, that the implementation of equity-based compensation will encourage
managers to take more risk. It can therefore be concluded, that risk-taking is positively related
to the use of long-term incentives. Ayadi et al. (2012) highlights the crucial importance of
CEO compensation structures for the economy as a whole by referring to the development and
evolution of the financial crisis in 2008. Excessive corporate risk-taking gave rise to the credit
crisis because CEOs were encouraged to take extremely high risks to get high bonuses in the
financial sector (Ayadi et al., 2012). Despite a broad consensus in the literature, some scholars
states, that managers will make an effort to improve the diversification of the portfolio they
have in order to decrease the risk of the portfolio in a performance-based compensation
environment. He argues that this could be bad news for investors due to the negative effect of
insider sale, and consequently harm firm value because investors will demand compensation
for this negative effect. It is therefore critical for firms to establish an appropriate
compensation system and keep a balance between CEOs’ fixed salary and variable pay. This
will motivate decision-makers not to neglect risky but profitable projects, but still control risk
within safe levels (Dee et al., 2005).
Many studies suggest high cross-country differences amongst CEO compensation. Academics
observe the fact, that on average US firms pay a lot more to their CEOs than other countries.
Especially in the banking industry, which even lead to the bankruptcy of investment banks
during the financial crisis in 2008, as this kind of compensation practice incentivised
corporate executives to focus mainly on short-term benefits instead of long-term objectives
(Friedman & Friedman, 2009; Adams, 2012). However, Fernandes et al (2012) claim that
American CEOs compensation is appropriate, after controlling for firm size, industry, country
characteristics, board structure and ownership, and that US CEO compensation structures just
comply with the requirements of the board and shareholders about performance-based pay.
The compensation structure differs amongst the portions of fixed and performance-based pay
and the total level of CEO compensation amongst countries, characteristic for the US is a high
level of performance based bay and a high total compensation. Germany and its very high
proportion of fixed pay components and a comparably low level of total compensation
presents a sharp difference to that.
According to Fernandes et al (2012), not only industry, firm size and leverage have an effect
on the CEO compensation structure, but also country characteristics, institutional ownership and corporate governance practices will affect CEO remuneration. Similarly, Bertrand and
Mullainathan (2003) find, that a higher proportion of institutional shareholders implies a
higher pressure on management activities, this requires higher managerial abilities to cope
with these expectations and subsequently a closer relationship between CEOs compensation
and organizational performance can be observed. Fernandes et al (2012) further note, that
firms have a higher proportion of institutional shareholders. It can be further observed that
shareholders will use the corporate executives compensation structure as a tool in order to
push their preferred risk-profile in the firm’s actions. In contrast to economic intuitions, CEO
pay will be higher with better supervising and monitoring of shareholders and a higher
firm-level corporate governance (Fernandes, et al, 2012). Risk-profiles of shareholders and
firm’s levels of corporate governance are shown to substantially differ amongst countries,
further highlighting the importance of cross-country variables when analyzing CEO compensation structures. Furthermore, these characteristics even differ amongst industries and
require a consideration of industry-level characteristics as well. The technology sector and its
characteristically dynamic and therefore also risky environment are an interesting object of
study in this context since the riskier nature of business is likely to have an effect on the
discussed relationship between CEO remuneration and corporate risk-taking. Corporate
executives in the technology sector are likely to get a high proportion of variable
compensation components in order to motivate them to invest in risky projects with higher
expected returns and subsequently create more shareholder value (Kwon & Yin, 2006;
Yanadori & Marler, 2006). The next section will outline the implications of agency theory on
the effect of compensation structures on corporate risk-taking in order to understand the
general relationship in more detail.
2.1.1 CEO compensation and agency theory
According to previous research, financial incentives can induce CEOs to manage and control
firms with responsibility, so managers are paid bonuses or other forms of rewards besides
their base salary in order to be motivated (Jansen et al., 2009). As mentioned, fixed pay and
performance-based pay are the two main types of executives compensation. Reference to
Hillier et al. (2012) reveals that performance-based pay can be divided into equity-based
compensation, such as restricted stocks and stock options, so-called long-term incentives and
earnings-based compensation, for example bonuses based on net profits or cash flow,
so-called short-term incentives. These two types of performance-based pay have their own
advantages and disadvantages. In terms of equity-based compensation, it encourages agents to
promote stock price, which is in line with shareholders’ expectations but makes stock prices
highly volatile on a daily basis (Hillier et al., 2012). Elaborating on earnings-based
likely to obtain this kind of compensation (e.g. bonus on net profit), but also corporate
executives from private companies can get it since it’s not based on publicly-traded stock
prices. But according to the work of Hillier et al. (2012), the values of financial indicators,
such as net profit or cash flow, could be manipulated by changing accounting policies or
applying other manipulation techniques to improve managers’ financial rewards.
Earning-based compensation therefore offers corporate executives opportunities to commit
financial information fraud. These opportunities were exploited by executives and resulted in
large scandals, such as Worldcom and Arthur Anderson. Legislative bodies tried to prevent
earning manipulation through the Sarbanes-Oxley act in a response to these scandals
(Rockness & Rockness, 2005). Since the occurrence of earning manipulations and thereby the
destruction of shareholder value however happened time and again, firms now largely abandoned earnings-based pay structures. The mentioned scandals highlight the crucial
importance of corporate remuneration structures for the firm’s risk-taking through agency
conflict.
The work of Jensen and Meckling (1976) indicates, that the separation of management and
ownership will cause a conflict of interests between managers and shareholders. The
separation causes information asymmetries between managers and owners, as well as
managers’ efforts to extract their own benefits from firm resources. According to agency
theory, managers will try hard to maximize their own wealth and interest, to guarantee
personal income and the stability of their job (Jensen & Meckling, 1976). In addition, the
work of Jensen and Meckling (1976) indicates that the separation of management and
ownership gives rise to a firm’s competitive advantages as well as agency problems. These
agency issues are opposed to organizational objectives of shareholders value maximization
and carry the possibility of opportunistic management behaviour. Rogers (2001) argues, that
in order to decrease managers’ opportunistic behaviour, firms need to establish efficient
corporate governance practices and effective monitoring mechanisms to force managers to act
in the firm’s collective interest. Moreover, Jensen and Murphy (1990) emphasize the
importance of executive compensation structures to overcome these agency issues. The design
of executive compensation is supposed to create incentives and motivation for
decision-makers to increase firm value and most importantly act in line with the shareholder
risky projects, leading to managers’ risk-aversion, which means decision-makers tend to
choose less risky projects with lower expected returns, instead of risky projects with higher
expected returns projects when faced with different options. This managerial risk-aversion is a
form of agency problem which has negative influence on shareholder value. Ryan et al.
(2002) conclude, that the primary function of equity-based compensation is to give managers
motivation to act in shareholders’ benefits by raising corporate risk-taking. Therefore the
mitigation and elimination of the agency problem brought by risk-averse managers is
especially important to firms’ sustainable development and long-term objectives and firms
can individually choose specific compensation structures to overcome this problem. (Coles et
al., 2006). Jensen and Meckling (1976) show, that shareholders will search for and implement
certain types of compensation structure to align the interests of managers and owners,
encouraging managers to take higher risk and subsequently create more shareholders value
and decreasing the possibility of managers’ opportunistic behaviour. Rajgopal and Shevlin
(2002) show, that the conflict of interests between managers and owners could even increase
if managers are offered too much equity-based compensation components.
I conclude, that CEO compensation structures are a crucial determinant of corporate
risk-taking. This relationship is affected by cross-country and cross-industry variables, I will
now further elaborate on the individual industry sector and the sector characteristics as a
cross-industry variable. Since the technology sector is widely considered the most dynamic
and uncertain sector in the economy and therefore imposes a high level of risk simply by the
nature of the sector, this sector is especially interesting to look at.
2.1.2 CEO compensation in technological sector
Makri et al. (2006) argue that technology has been playing an important role in social
development these years. The fast improvement in science and technology promoted the
development of human society, in the meantime innovation in production technology and
changes in production methods play a critical role in economic development (Barney, 1991;
Balkin et al, 2000). The work of Rogers (2001) reveals that the number of high-tech firms is
booming due to technology being the main driving factor for sustainable development. The
rapid emergence and growth of high-tech firms is one of the primary driving forces of social
space for the improvement of innovation and creativity (Rogers, 2001). The fact that
innovative technology and achievements are exploited in high-tech firms in order to improve
the efficiency of production and management through services or products, has been one of
highlights in the new era of economic growth in the 21st century (Rogers, 2001). Firms in the
technological sector are characterized by a high quantity and ideally also quality of R&D
investments, plenty of intangible property, such as patents, employees with high education,
and strong creative capabilities. (Kwon & Yin, 2006). The growth of high-tech firms is
therefore the inevitable outcome of sustainable development of innovation and technology.
Scholars show, that corporate executives prefer incentive pay, such as stock options compared
to fixed salary items in high-tech firms, a very risky and fast-changing environment
(Prendergast, 2000). The work of Dee et al. (2005) argues, that firms need to make tradeoff
decisions between executives remuneration and risk, as managers will request higher pay for
the greater risk they bear, especially in the highly developing and highly volatile industries,
such as high-tech firms. Furthermore, there are substantial differences in CEO compensation
practices between high-tech and low-tech firms according to the study of Kwon and Yin
(2006). They state, that corporate executives get higher total compensation in high-tech firms
compared to low-tech firms. High-tech firms provide managers with higher proportions of
equity-based compensation, such as stock options, yet their base salary or cash rewards are
not necessarily higher than those of managers in low-tech firms (Kwon & Yin, 2006). The
authors provide different explanations to explain these differences amongst the
technology-level of a firm. One explanation is, that high-tech firms need to spend more
money on R&D projects to maintain their innovative capabilities in the
dynamic-technological market, which means managers bear higher risk in high-tech firms
than in low-tech firms. The naturally high level of uncertainty is one of the key attributes of
R&D investments, which make them more difficult to monitor and assess (Kwon & Yin,
2006). Thus, those high-tech firms are more willing to provide CEOs higher pay to undertake
greater level of risk (Cui & Mak, 2002). At the same time, CEOs will request for higher pay
to compensate the higher risk they undertake in high-tech firms (Cui & Mak, 2002). Kwon
and Yin (2006) further argue, that the nature of business in the technology sector and the
induced level of uncertainty make it very hard for shareholders to efficiently monitor and
create a stronger bond between shareholder motives and the motives of the individual
executive. If high-tech firms use high levels of earnings-based compensation, such as bonuses
basing on total cash flow, it will discourage managers to invest in positive-NPV projects with
high risk (Kwon & Yin, 2006). The implementation of stock options as incentive
compensation is more common in technological firms in order to motivate managers to act in
shareholders’ interests and align the interests of stockholders and corporate executives as their
investment decisions will have an effect on their private benefits through stock prices (Rogers,
2001).
It can be concluded, that due to the fast-changing and risky environment, CEO compensation
structures are different in technological firms compared to those in other industries. (e.g. Coy,
2000; Nesheim, 2000; Rogers, 2001; Cui & Mak, 2002; Carpenter et al., 2003; Dee et al.,
2005; Kwon & Yin, 2006; Makri et al., 2006; Yanadori & Marler, 2006; Lin et al., 2011).
Managers have to undertake higher risks in the technological industry caused by the high level
of uncertainty in the sector, shareholders will consequently provide more incentive
compensation elements to corporate executives to make up for the higher risk they bear. All
these studies suggest a stronger relationship of CEO compensation (as a measure for good
corporate governance) and corporate risk taking.
2.2 Hypotheses development
2.2.1 Relationship between CEO compensation and corporate risk-taking
As Coles et al. (2006) have indicated, executive compensation systems have a significant
impact on corporate risk taking and risk-taking is positively related to the sensitivity of
executives compensation to the volatility of the firm stock price. For example, riskier
practices like higher R&D spending, lower investment in tangible assets or more debt (higher
leverage) are more likely to be implemented in the organization if the corporate remuneration
packages are efficient in optimizing corporate risk-taking. Coles et al. (2006) further state,
that CEOs should get convex compensation to reduce the propensity that risk-averse managers
avoid risky investments. These convex compensation packages imply an increasing degree of
CEO compensation participation in firm success with increasing profits. This will increase
their motivation and incentives to bear risks (Coles et al., 2006). The higher coupling of CEO
to increase their risk-tolerance in order to maximize their own remuneration. The work of Dee et al. (2005) argues, that firms need to make tradeoff decisions between executive
remuneration and risk, as managers will request higher pay for the greater risk they bear.
Previous studies emphasize, that pay-performance sensitivity is negatively associated with
risk (e.g. Aggarwal & Samwick, 1998; Jin, 2002; Dee et al., 2005). The work of Ryan et al.
(2002) suggests that there is a negative relationship between the use of equity-based
compensation (such as stock options) and leverage, which could be a proxy for risk-taking.
This implies, that not all elements of CEO compensation are suitable in the generation of
incentives for executives to increase their risk-taking practices. In the financial crisis, one of
the main driving forces of the excessive risk-taking in the economy were flawed and
enormously high CEO compensation packages who encourages executives to take too much
risk. This recent evidence however supports the hypothesis of a positive relationship of CEO
compensation and corporate risk-taking as this effect could be observed in the economy. Dee
et al. (2005) consequently find, that equity-based compensations will induce managers’
risk-taking behaviour. The financial crisis proves this assumption in a drastic manner, as the
positive relationship between corporate compensation and corporate risk-taking seems to be
characterized by exponential or at best linear growth.
All these studies suggest a positive relationship between CEO compensation and corporate
risk-taking because CEO compensation packages can overcome the initial risk-averse
tendencies in corporate executive’s decision-making processes. This hypothesis is further
supported by the aforementioned aspects in agency-theory, incentive-alignment theory and
managerial power theory.
As mentioned in the literature review, I expect this relationship to be even stronger in the
technological sector. A study by Prendergast (2000) shows, that there is a positive relationship
between executive compensation and corporate-risk taking in R&D intensive firms. The
individual characteristics of the technology sector (e.g. short product lifecycles, high required
innovative capacity, high level of uncertainty about the future) suggest, that CEO
compensation packages and corporate risk-taking must have a stronger relationship in these
industries. As Firms use remuneration packages to encourage executives to overcome their
relatively high-level of corporate risk-taking is necessary for them to survive. One popular
example for this phenomenon is Google, who is undoubtedly part of the technology industry.
Google holds massive levels of excess cash in their balance sheet to be able to quickly react
on the market and purchase promising patents or start-up companies. This high level of cash
poses a risk to the company since it looses possible returns from other projects where this
money could be invested (even bank interests). Managers will typically hesitate to take this
kind of risk unless their compensation packages give them incentives to do so. The
technology sector requires firms to take all kinds of risks, like hold high levels of excess cash,
high investments in R&D and for companies with lower possibilities of internal financing
than the big one like Google or Apple, high leverages are also part of this high required level
of risk-taking (Prendergast, 2000; Dee et al., 2005; Coles et al., 2006). To sum up, technology
firms have higher needs for an appropriate level of risk-taking in their corporate
decision-making processes and therefore need to implement better (higher) compensation
packages to meet this need. The sensitivity of corporate risk-taking for CEO compensation is
therefore hypothesized to be higher in the technology sector compared to other sectors. Firms
and executives will both be aware of the highly volatile and risky environment in the
technology sector and the effect of compensation on risk-taking will therefore be higher in
this sector.
H1a:
CEO compensation has a positive effect on corporate risk-taking.
H1b: The positive effect of CEO compensation on corporate risk-taking is higher for firms in the technology industry compared to other industries.
2.2.2 Firm-level Governance as a moderator of the Compensation-Risk relationship
Firm-level corporate governance measures have to be considered in the examination of the
compensation-risk relationship as they are a primary determinant of a firm’s ability to use
governance tools in order to achieve its goals. A higher level of corporate governance can
fixate corporate executive’s remuneration at a level where it is best in order to serve
shareholder’s interest and prevent or at least minimize the extraction of private benefits by
corporate executives. If firms have high levels of corporate governance, they can more
measures (Coles et al., 2006). The same study suggests, that the extraction of private benefits
by corporate executives, even so far as executives will manipulate firm statements, is much
more likely in the absence of high corporate governance standards. Cyert et al. (2002) find,
that firms can improve their monitoring capabilities with higher levels of corporate
governance and will subsequently arrive at more efficient levels of risk-taking and a better
corporate performance. This relationship is however found to be questionable, as Larcker et
al. (2012) find a positive effects of weaker corporate governance standards of firm
performance. Their research suggests, that the higher monitoring costs that higher corporate governance standards incur will surpass the managerial entrenchment costs that the firm
would be subject to otherwise. This finding is labeled managerial-power hypothesis in the
literature. Derived from these findings, it seems likely that better corporate governance
standards can also lead to a ‘worse-off’-effect in terms of risk-taking as high corporate
governance standards and thereby higher compensation packages to encourage risk-taking
may be more costly for firms than the abandonment of risky, positive NPV-projects by
risk-averse corporate executives. Information asymmetries also play a crucial role in the
effectiveness of corporate governance on the compensation-risk relationship. Managers that
are subject to high corporate governance standards and thereby a high level of monitoring
might chose to withhold valuable information about the firm in order to avoid higher
monitoring levels. Higher corporate governance might therefore lead to higher levels of
information asymmetries. These asymmetries will negatively affect the relationship of
compensation and risk-taking because the design and scope of remuneration packages will
become less optimal. Board members will have less available information in the process of
designing efficient compensation packages. A study by Chan et al. (2008) finds similar results
as they conclude, that managers will react badly if they perceive to be subject to too high
monitoring procedures and they feel forced to satisfy shareholder’s interests. Managers who
feel intimidated by this are likely to engage in earnings manipulation and other managerial
entrenchment practices. Firms with high levels of corporate governance standards allow investors to closely monitor managers and make their voices heard in corporate
decision-making procedures which could lead managers to engage in the aforementioned
practices. Acharya et al. (2013) find, that well-governed firms will be more likely to
efficiently use compensation packages to direct executive’s behavior in a
corporate governance standards do not always lead to efficient levels of CEO-pay as they find
that a considerable proportion of US-firms still overpay their CEOs even under a high level of
corporate governance. This fact is found to be caused by the “market of talent” in the human
resource markets. Managers with high managerial skills are a scarce resource in the market
and this makes their remuneration subject to market-price inflation. Firms will be forced to
pay too high compensations in order to hire high-skilled managers for their company. These
market forces may negatively affect the compensation-risk relationship as firms have to
consider them in the process of pay-determination and cannot simply consider optimal risk
levels for their compensation packages. Motivation effects may also play a role in this respect
as a high level of pay will higher the probability that managers deploy their full set of
capabilities to achieve the shareholder’s long-term goals. A study by Randøy and Nielsen
(2002) reveals a positive mediating effects of corporate governance on the ability of CEO
compensation to overcome agency problems. They find that firm-level corporate governance
measures play a mediating role in the agency problems - CEO power relationship. Differences
in firm-level corporate governance are found to have a significant effect on variations in
corporate executive remuneration packages. Corporate Governance is further found to have a
significant positive effect on a firm’s level of risk-taking. As a study by John et al. (2008)
reveals, corporate risk-taking is closer to the firm’s optimal risk-levels in the presence of high
corporate governance standards. The mentioned studies suggest positive and negative effects
of corporate governance standards on the compensation-risk relationship. In order to evaluate
which one prevails, I hypothesize:
H2a: Firm-level corporate governance positively moderates the relationship between CEO compensation and corporate risk-taking.
H2b: Firm-level corporate governance negatively moderates the relationship between CEO compensation and corporate risk-taking.
The technology sector further poses an interesting subject of study in this respect. The special
characteristics of the technology sector make corporate governance even more important for
firms there. The high levels of uncertainty in the sector and the high level of information
environment makes it very hard for shareholders to efficiently monitor CEO behaviour.
Executives must be able to act very fast and take quick decisions in order to cope with the
high development speed of the market. Corporate governance mechanisms and the implied
higher level of monitoring and more sophisticated and long-lasting decision-making
procedures pose a risk for the company as quick reactions to market-developments are
hampered and long-term success and a growth of shareholder wealth is in danger. A study by
Cheung et al. (2005) finds, that Chinese companies in the technology sector are more likely to
show better corporate performance if they show lower levels of corporate governance. Higher
corporate governance standards will have a negative effect on the efficiency of remuneration
packages in optimizing a firm’s risk level according to these findings, because they will
increase managerial entrenchment and decrease efficiency levels in the firm. Especially in the
dynamic technology environment, quick reactions to market developments and technological
breakthroughs are essential, and high governance will hamper them and subsequently
negatively influence the efficiency of remuneration packages to optimize and increase
corporate levels of risk-taking. Executives will feel intimidated by the high level of scrutiny
and will therefore be more reluctant to take the high levels of risk that are especially needed in
the technology sector.
However, Balkin et al. (2000) find, that high levels of corporate governance are essential for
technology firms in order to achieve shareholder goals. As the “DotCom-bubble” and the
destruction of massive amounts of shareholder wealth show, technology firms and their
success can be very short-lasting and unsustainable. High levels of corporate governance are
therefore required to ensure a minimization of information asymmetries in the sector.
Shareholders need close monitoring mechanisms to ensure they obtain the required
informations to evaluate the corporate executive’s performance and subsequently the outlooks
for success of their own investments. Higher corporate governance is therefore an efficient
tool to optimize the efficiency of remuneration packages to derive at optimal levels of
corporate risk.
H2c: The moderating effect of firm-level corporate governance is more positive for firms in the technology industry compared to other industries.
H2d: The moderating effect of firm-level corporate governance is more negative for firms in the technology industry compared to other industries.
2.2.3 Country-level Governance as a moderator of the Compensation-Risk relationship
Not only firm-level governance has an influence on the relationship of compensation and
corporate risk-taking, but also country-level governance indicators (Hansen et al., 2012).
Randoy and Nielsen (2012) conducted a study amongst scandinavian firms and found, that
country-level variables play a crucial role in the determination of CEO compensation
packages. They show that several country characteristics like investor protection are good
predictors for the level of CEO compensation in a country. Houston et al. (2010) further find
an effect of country-level variables on the average level of corporate risk-taking across countries. They conclude that the pay-performance sensitivity is largely affected by
cross-country variables. This study will extend this definition and examine the effect of
country-level characteristics on the pay-risk sensitivity. Klapper and Love (2004) examine the
effect of various country characteristics on CEO pay and find positive effects of law
enforcement and investor protection regimes. A study by Albuquerque (2013) shows that
country-level corporate governance is one of the main determinants of firm-level governance.
He further finds complementary effects of country-level governance on firm-level governance
and thereby further suggests interlinkage-effects between the two levels of governance.
A study by Hansen et al. (2012) emphasizes the special importance of investor protection for
the mentioned relationship. Investor protection is defined as the degree of minority shareholders and creditors’ rights and interests being protected by legislation from insiders’
expropriation, such as mandatory disclosure of accounting information, gaining dividends,
and voting rights etc. (La Porta et al., 2000). Many researchers have emphasized the
importance of investor protection, such as La Porta et al. (2000), since the lack of minority
shareholders’ protection, caused by poor corporate governance, would lead to a high cost of
capital. The work of La Porta et al. (2000) shows that there are stronger investor protection
protection systems are found in code law countries, for example, France, Germany and
Scandinavian countries. In a study of Morck et al. (2000), they find that there is a relationship
between the quality of investor protection and the level of informed risk arbitrage.
Furthermore, inefficient production and low growth rate would be caused by a low level of
informed risk arbitrage in bad investor protection countries (Durnev et al., 2004). John et al.
(2008) note that not only management behavior and executives compensation have an impact
on corporate risk-taking, but also investor protection, and a positive relationship between
corporate risk-taking and the quality of investor protection was found in their research.
According to agency theory, managers would reject some positive-NPV projects to protect
their own interests, but this does harm to investors’ benefits, especially minority shareholders
(Jensen & Meckling, 1976). Besides, John et al. (2008) point out that the alignment of
interests between shareholders and managers could be achieved under the environment of
strong investor protection, causing higher risk-taking by managers. The literature defines this
phenomenon as the incentive-alignment theory. Furthermore, stakeholders have bigger
impacts on firms when they operate in countries with poor investor protection. So managers
will take less risk due to the pressure from these stakeholders (John et al., 2008). Investor
protection will encourage decision-makers to take higher risk, and invest in riskier projects
with higher expected returns (Stulz, 1999; Klapper & Love, 2004; John et al., 2008; Houston
et al., 2010). However, there are also some studies suggesting a negative relation between
investor protection and corporate risk taking. For instance, the work of Paligorova (2010)
states that the number of dominant shareholders will be less in good investor protection
countries, which offers risk-averse agents more opportunities to avoid corporate risk, since
ownership concentration and investor protection complement each other. Additionally, shareholders’ concerns about the expropriation by managers will decline when those
shareholders’ rights are defended in countries with stronger investor protection (Paligorova,
2010).
Additionally, some researchers note that the quality of investor protection will also affect
CEO compensation (e.g. Hartzell & Starks, 2003; Arye Bebchuk & Fried, 2004; Frieder &
Subrahmanyam, 2006; Albuquerque & Miao, 2013; Zheng et al., 2016). As Zheng et al.
(2016) point out, there is a significant positive relationship between investor protection and
private interests will decrease and they will receive more explicit incentives to induce them to
create more profits under the environment of stronger investor protection (Zheng et al., 2016).
The work of Arye et al. (2003) claims, that managers have no choice but to waive some of
their private benefits under better investor protection, so improving executives pay will make
up for the loss of their private interests and it will lead to higher efficiency of the firm under
stronger investor protection. Moreover, La Porta et al. (2000) indicate that the quality of
investor protection is related to corporate governance, which includes executives
compensation system. This paper argues that investor protection has a moderating effect on
the relationship between executives compensation and corporate risk-taking. Considering the
effect of the quality of investor protection on executives compensation and corporate risk
taking and the ambiguous views of the literature on this relationship, the third hypothesis is
developed as follows:
H3a: Country-level governance and specifically the level of investor-protection positively moderates the relationship between CEO compensation and corporate risk-taking.
H3b: Country-level governance and specifically the level of investor-protection negatively moderates the relationship between CEO compensation and corporate risk-taking.
In line with the preceding chapters, this study will take a closer look on the special
characteristics of the technology industry and their impact on the hypothesis. As mentioned
before, the technology sector is characterized by high levels of uncertainty and a high
competition dynamics within the industry. The moderating effect of country-level governance
and specifically investor-protection should either be stronger due to higher agency problems
or weaker due to the mentioned effects of increasing information asymmetries and the
executives’ perceived pressures of shareholders that force them to act in their interest. The
same argumentation as for firm-level corporate governance indicators applies here since
country-level governance is the main determinant of shareholder’s ability to implement
H3c: The positive moderating effect of country-level governance and specifically the level of investor protection is stronger for firms in the technology industry compared to other industries.
H3d: The negative moderating effect of country-level governance and specifically the level of investor protection is stronger for firms in the technology industry compared to other industries.
2.2.4 Cultural Dimensions as moderators of the Compensation-Risk relationship
Previous research suggest that national culture has an influence on management control
systems and corporate governance practices, such as executives compensation structure (e.g.
Hofstede et al., 1991; Chow et al., 1999; Van der Stede, 2003; Jansen et al., 2009;
Greckhamer, 2011). Hofstede et al. (1991) identified five cultural dimensions, including
power distance, individualism, uncertainty avoidance, masculinity, and long-term orientation.
Managers may prefer performance-based pay and focus on personal achievement in countries
with higher levels of individualism, while those managers have higher possibility to prefer a
more stable fixed pay in countries with lower level of individualism (Chow et al., 1999). In
some Anglo-Saxon countries like the U.S., CEOs pay more attention to production efficiency
and organizational performance due to the high level of individualism and masculinity,
whereas people from Continental-European countries, Dutch managers, for example, receive
less incentive income than American managers due to long-term oriented national culture
(Jansen et al., 2009). As the work of Jansen et al. (2009) notes, even when Dutch firms
implement incentive compensation, they didn’t get satisfying feedback or positive effects on
firm value. Besides the fact, that the U.S. has a higher level of masculinity but lower levels of
long-term orientation according to Hofstede (1980), it is explained why American managers
prefer stock options and other equity-based compensation than managers from other countries.
Jansen et al. (2009) argue that uncertainty and individualism cultural dimensions have the
most critical influence on the design of executives compensation. According to Van der Stede
(2003), uncertainty avoidance is defined as how much people hate, or feel frustrated by an
in countries with high level of uncertainty avoidance, as performance-based pay is more
unstable and risky (Van der Stede, 2003). In terms of the individualism dimension, Van der
Stede (2003) notes that it represents the trend of a person to regard himself as an individual
instead of a member of a community. Chow et al. (1999) propose, that people would rather
get performance-based compensation than clearly formula-based salary since
performance-based incentives offer them a sense of achievement and people focus on personal
value in individualistic societies.
In the meantime, national culture will also affect corporate risk-taking according to Schuler
and Rogovsky (1998). Managers are less likely to invest in highly risky projects in countries
with high level of uncertainty avoidance, while those managers would like to take more
corporate risk in countries with lower level of uncertainty avoidance. Slocum and Lei (1993)
observe that American CEOs would take higher corporate risk when managing firms since the
U.S. has lower scores in uncertainty avoidance. In countries with higher level of uncertainty
avoidance, managers are less likely to prefer performance-based compensation due to its
riskiness (Harrison et al., 1994). According to the argument discussed above, considering
cross-culture variables have influence both on CEO compensation and risk taking, this thesis
argues that national culture also has an effect on the relation between CEO compensation and
risk taking, especially uncertainty avoidance and individualism dimensions. So the fourth hypothesis is developed as follows:
H4a: The level of uncertainty avoidance moderates the relationship between CEO compensation and corporate risk-taking.
H4b: The level of individualism moderates the relationship between CEO compensation and corporate risk-taking.
The technology sector and its special risk characteristics further justify a special look at the
sector in terms of culture. As the word already indicates, uncertainty avoidance is highly
associated with risks and the avoidance of them. It can therefore be hypothesized, that the
effect of uncertainty avoidance on the relationship between executive compensation and
of managers and shareholders make it even more important to achieve efficient risk-taking through specific compensation packages. One can therefore assume that the effect of
compensation on corporate risk-taking is higher in uncertainty avoiding environments.
Technology firms in these environments have a higher need for an efficient direction of
corporate risk-taking by compensation packages. On the other hand, one can imagine a natural
selection process for industry sectors and the executives working there. Since the higher levels
of risk in this industry are widely known amongst managers, one can assume that risk-averse
managers will avoid positions in the technology sector and prefer other industries. This
re-allocation of risk-averse managers to other industry sectors would undermine and minimize
the effects of uncertainty avoidance on the compensation-risk relationship and drive them up
in other industries since risk-averse managers will gather there. Consequently, I hypothesize
H4c: The moderating effect of uncertainty avoidance is stronger for firms in the technology industry compared to other industries.
H4d: The moderating effect of uncertainty avoidance is weaker for firms in the technology industry compared to other industries.
Figure 1. Conceptual model
3. Methodology, Data and Sample
3.1 MethodologyTo examine the hypothesis above, the linear regression method is used to implement the
estimation of the unbalanced panel dataset. All relevant variables are presented in the
following subsections.This thesis examines the relationship between CEO compensation and
corporate risk-taking in technological industry. Further, industry-effects, country-level effects
are also to be studied in this thesis. The sample period is from 2002-2015. And panel data
analysis would be used in order to test the hypothesis discussed above. All variables examined
in this thesis to research the relation between CEO pay and risk-taking will be reported in this
section.
3.1.1 Dependent variable
According to previous research, there are several way to measure corporate risk-taking, such
as leverage, R&D intensity, and stock return volatility, etc. (An et al., 2014; Coles et al.,
2006). The dependent variable in this thesis is the level of corporate risk-taking. And this
risk-taking, following the way of An et al. (2014). To be specific, SROA is calculated as the
standard deviation of EBITDA to total assets over past 3 years, which is commonly used as a
proxy for risk-taking in previous research (e.g., An et al., 2014). And R&D intensity is
measured as the ratio of R&D expenditure to total assets (Coles et al., 2006).
SROA = natural logarithm of (standard deviation of (EBITDA / total assets) for past 3 years )
R&D spending is another common proxy for corporate risk-taking since R&D projects are
highly risky investments (Balkin et al., 2000). But the amount of R&D spending varies a lot
from firm to firm, depending on many conditions of firms, such as their size, profitability,
industry characteristics, etc. Hence this thesis uses the percentage of R&D spending to total
assets to make it comparable, as shown in the following formula:
RD = natural logarithm of ( R&D expenditures / total assets )
3.1.2 Independent variable
Previous research have used many ways to measure CEO compensation, such as cash
compensation, total CEO compensation, the value of stock options compensation, the ratio of
equity-based compensation to total CEO compensation, etc. (Aggarwal & Samwick, 1998;
Coles et al., 2006; Dee et al., 2005).Since studies about executives compensation packages
have attracted much attention, the number and detail of compensation data have improved
greatly (Anderson et al., 2000). This thesis uses CEO total compensation (including base
salary, bonus, the value of stock options and restricted shares ) and total stock options
compensation as the independent variables, and regression models will explore the influence
of CEO compensation on corporate risk-taking. The data for these independent variables
could be collected from ASSET4 ESG database. The ASSET4 ESG database include data in
terms of 4 perspectives: environment, economy, society, and corporate governance. This
database includes almost 750 individual assessment score for each listed company, which
generates more than 250 important performance index (Gonenc & Scholtens, 2017). And
those data were collected from many public sources, including financial reports and corporate
websites (Gonenc & Scholtens, 2017). I collected the data for executives compensation from
corporate governance (Cyert et al., 2002). The score and value for companies’ operation and
governance are gathered in corporate governance pillar, which evaluates the quality of board,
managers compensation, etc. Considering the availability of data in ASSET4 ESG of
datastream, I collected total executives compensation and total stock options compensation as
independent variables and define them as following formula since the value of compensation
is quite big compared to dependent variables (SROA and RD):
COMP = natural logarithm of total executives compensation
OPTION = natural logarithm of total stock options compensation
3.1.3 Firm-level corporate governance variables
This thesis also includes corporate governance score as a proxy for firm-level corporate
governance characteristics to evaluate the quality of corporate governance in firms and the
data can be collected from ASSET4 ESG database. And corporate governance score offers a
tool to evaluate the degree of protecting stakeholders’ benefits. Therefore, this study will
research the moderating effect of firm-level corporate governance CGS on
Compensation-Risk relationship using the following formula:
CGS = Corporate Governance Score
3.1.4 Country-level variables
This thesis uses the level of investor protection as a proxy for country-level characteristics,
which means the level of protection for investors against resources expropriated by agents It is
measured by the anti-self-dealing index, which is gathered from the paper of Djankov et al.
(2008). This paper provides the data for anti-self-dealing index for 72 countries, which is
established on the work of La Porta et al. (2000) and the index of anti-director rights.
However, anti-self-dealing index is superior to anti-director rights index according to Djankov
et al. (2008), as it evaluates the difficulties that happen to investors when starting a new
transaction gathered from questionnaire. Therefore, this study uses anti-self-dealing index to
In terms of national culture dimensions, Hofstede developed various index for different
countries and he established his own website to present the outcome of his study. Thus the
data for individualism and uncertainty avoidance index is collected from Geert Hofstede’s
website https://geert-hofstede.com/cultural-dimensions.html. I adopt these index to measure
national culture dimensions of individualism and uncertainty avoidance in this thesis. And I
use UA as uncertainty avoidance, IND for individualism.
3.1.5 Control variables
This thesis uses some control variables determining financial positions and investment
opportunities, which are established on previous literature (Fernandes et al., 2011; De Cesari
et al., 2016). Those control variables act for the driving forces which link CEO compensation
with financial positions and investment opportunities. This thesis uses the natural logarithm of
total assets to control firm size, which is a common proxy for firm size. According to
Fernandes et al. (2011), there is less possibility for large firms to take higher risk and firm size
also affects the level of CEO compensation. Additionally, Market-to-Book ratio is frequently
used as a proxy for firm growth and investment opportunities (Coles et al., 2006). And this
proxy is measured by the ratio of market value of assets to book value of assets in this thesis.
To control country-level characteristics, GDP per capita is also included as one of the control
variables in this study, calculating by natural logarithm of GDP per capita in each country.
According to Croci et al. (2012), executives will get a greater level of compensation when
there is a bigger board size in firms, which means that executives compensation is positively
affected by board size. This thesis defines Board size as the number of executive directors and
the number of non-executive directors following the method of Croci et al. (2012).
Additionally, Fernandes et al. (2011) suggest that the level of Board independence has an
impact on managers’ compensation and there is a positive relation between the level of
executives compensation and the percentage of independent board members. And those data
could be gathered from ASSET4 ESG database in Datastream. It is measured as the ratio of
the number of independent board members to the total number of board members. The
following formula are the way this thesis defines those control variables: