Research is inconclusive regarding which factors that determine Chief Executive Officers (CEO) remuneration. There is evidence of a positive link between the risky actions taken by CEOs, incentivised by their remuneration structure, which contributed to the financial crisis of 2008.
The purpose of this study was to determine whether organisation size and organisation performance were determinants of CEO remuneration and to what degree.
No consensus on a model, a set of variables, or a consistent view defines the principles of remuneration setting at the CEO level across organisations or industries. This study intended to provide further clarity on the matter.
The research employed a mono-method methodology, and the study was longitudinal in nature. Secondary data were collected over a 5-year period (2015–2019) using a homogenous purposive sampling method. Statistical analysis was performed to analyse the data.
Organisation size was not found to be a significant determinant of CEO remuneration in financial services organisations listed on the Johannesburg Stock Exchange (JSE). In contrast, organisation performance was found to be a significant determinant of CEO remuneration.
This study serves as a baseline for best practice, enabling remuneration committees to leverage when setting CEO remuneration, to ensure that the outcomes driven by remuneration are in line with the best interests of all stakeholders within the organisation.
The findings add to the body of knowledge on this topic and create an evidence base showing whether these exorbitant remuneration packages are performance driven or if they are merely driven by managerial power.
Chief Executive Officers (CEOs) are board-appointed leaders responsible for implementing an organisation’s strategic plans, as set out by the board of directors (Daft,
Given the nature and complexity of the role, CEOs are typically highly educated, skilled, and experienced individuals who have gained management prowess through their tenure in leadership roles (Acero & Alcalde,
Over time, two main schools of thought on the determinants of CEO remuneration have emerged. The optimal contracting approach theorises that CEO pay is driven by efficient bargaining between shareholders and CEOs to alleviate the agency-principal problem (Abdou, Ntim, Lindop, Thomas, & Opong,
This research addresses the relationship and relationship strength of organisational size, as measured by assets, revenue and the number of employees and organisational performance as measured by profit, return on equity (ROE) and earnings per share (EPS) on the components of CEO remuneration. Furthermore, it addresses how the abovementioned schools of thought play a role by testing two main assumptions:
Organisation size influences CEO remuneration.
Organisation performance influences CEO remuneration.
These assumptions were tested using six hypotheses where the fixed, variable and total remuneration of the CEO were contrasted against one another.
For the purposes of this study, only fixed remuneration and short-term incentives were included. Core, Holthausen and Larcker (
There has been isolated research conducted, where researchers have tested organisation size or organisation performance to establish the relationship that these variables have on CEO remuneration. However, there has been limited research conducted on testing the relationship of both variables (organisation size and organisation performance) on the elements of remuneration, in a single study. For the purposes of this study, proxy variables will be used to measure the direction and degree of influence that organisation size and organisation performance have, on the different components of CEO remuneration (fixed, variable and total), by testing the following research hypotheses:
H1 Organisation size is a significant determinant of fixed remuneration, for CEOs of financial services organisations, listed on the JSE
H2 Organisation size is a significant determinant of variable remuneration, for CEOs of financial services organisations, listed on the JSE
H3 Organisation size is a significant determinant of total remuneration, for CEOs of financial services organisations, listed on the JSE
H4 Organisation performance is a significant determinant of fixed remuneration, for CEOs of financial services organisations, listed on the JSE
H5 Organisation performance is a significant determinant of variable remuneration, for CEOs of financial services organisations, listed on the JSE
H6 Organisation performance is a significant determinant of total remuneration, for CEOs of financial services organisations, listed on the JSE
The outcomes of hypotheses one, two, and three will confirm whether organisation size is a significant determinant of fixed remuneration, variable remuneration and total remuneration of financial services CEOs and to what degree. Similarly, the outcomes of hypotheses four, five and six will confirm whether organisation performance is a significant determinant of fixed remuneration, variable remuneration and total remuneration of financial services CEOs and to what degree.
Jensen and Meckling (
Linder and Foss (
Jensen and Meckling (
The managerial power approach depicts CEOs as rent seekers who have a significant degree of influence over the board of directors and their remuneration. Bussin and Ncube (
Various authors (Acero & Alcalde,
Merhebi et al. (
Based on previous research, the view is that organisation size has an influence on CEO remuneration. These findings have been deduced using proxies for organisation size and confirm a positive relationship between the independent variables used to measure organisation size and CEO remuneration.
Zhou (
Ghazali and Taib (
Thus, there seems to be consensus on the positive relationship between organisation size and CEO remuneration, even in instances where different variables have been used as a proxy for organisation size. However, there are mixed views on whether organisation size has an influence on overall CEO remuneration or whether specific proxies for organisation size have an influence over a specific component of CEO remuneration. Sur, Magnan and Cordeiro (
Although organisation size has been established to have a positive and significant relationship with CEO remuneration, Sonenshine et al. (
Chief Executive Officers typically earn high returns for the effort they give, relative to those employees who rank below them in the organisation. However, of late, many organisations have started to adjust the way in which they reward their CEOs and a significant portion of their remuneration is tied to performance measures, making this portion of pay at-risk (Martin & Magnan,
Earlier studies on the topic found a negative relationship between CEO remuneration and organisation performance. In addition, Van Essen, Otten and Carberry (
Sonenshine et al. (
Based on the above evidence, it can be argued that organisation performance influences CEO remuneration. These findings were established by using proxies for organisation performance, and the findings confirm a positive relationship between the independent variables used to measure organisation performance and CEO remuneration.
Merhebi et al. (
Consequently, Sonenshine et al. (
In contrast to the findings mentioned above, various authors have found converse results when measuring the effect of organisation performance on CEO remuneration. Using profit and ROA as their proxies for organisation performance, Ghazali and Taib (
Having discussed the findings from previous research, the next section outlines the approach and methods that were adopted to conduct this research.
This section discusses the research approach and research method adopted in this study.
A deductive reasoning approach was used by testing the theoretical proposition, through a research strategy, designed to collect data for this purpose (Saunders & Lewis,
Secondary data were collected over a 5-year period (2015–2019) using a single data collection technique. This data collection technique lends itself to quantitative techniques on which statistical analysis was performed, and results were examined to determine the explanatory power of the independent variables on the dependant variable (Saunders & Lewis,
Given the research focus for this study, the population was limited to CEOs of financial services organisations, listed on the JSE within the context of South African financial service organisations.
As the research was focused on a specific population, namely CEOs in the financial services sector, a homogenous purposive sampling method was employed. The sample was limited to the top 15 financial services organisations on the JSE, based on their market capitalisation. The CEO of the selected company become the subject of this study and the period of observation spanned 5 years. Therefore, the dataset consisted of ~75 observations per variable. To qualify for inclusion, organisations had to meet the following criteria:
The organisation must have been in operation for the entire period under observation (2015–2019).
The organisation must have been listed on the JSE for the entire period of the study (2015–2019).
The data must have been available on the data sources specified for this study.
Given that this research was based on secondary data analysis, this section will discuss the unit of analysis and the sources used to retrieve the relative data.
The first unit of analysis was CEO remuneration, broken down into fixed remuneration and short-term incentives (bonuses). For this study, only fixed remuneration and short-term incentives were included. Core et al. (
The second unit of analysis was organisation size. The following variables were chosen as proxies: assets, revenue and number of employees. This information was sourced from I-NET Bureau for Financial Analysis (I-NET BFA), one of the leading sources of financial information in South Africa and validated against the annual financial statements for each organisation. Given that I-NET BFA is a trusted and widely used source of information, the information was deemed valid and reliable.
The third unit of analysis was organisation performance. The following variables were chosen as proxies: profit, ROE and EPS. This information was also sourced from I-NET BFA and validated against the annual financial statements for each organisation. The information was deemed valid and reliable.
Secondary data were collected and analysed using a logical structure to determine how the independent variables influence the dependent variables and to what extent. As such, research hypotheses were defined and tested, based on the premise outlined in the literature review. New theories have not been developed as an outcome of this research. Rather, existing theories were tested.
The focus for this research was confined to CEOs of financial services organisations listed on the JSE. Given the requirements of the JSE, all organisations are mandated to publish their CEOs’ remuneration in their annual financial reports. This made data collection standardised and accurate as financial statements for all JSE listed companies are subject to audit requirement.
After the secondary data were collected, the research hypotheses were tested to determine the degree of influence that the independent variables had on the dependent variable, CEO remuneration.
Throughout the research procedure, the researcher adhered to the ethical principles of respect, beneficence and justice.
The data were categorised into two distinct categories, namely (1) organisation size and (2) organisation performance.
All information relating to the organisation’s size and performance was extracted from the I-NET BFA database, which is South Africa’s leading provider of financial data and was validated against the annual financial statements of the organisation. The data were aggregated at a per-subject level after which, this data formed the basis for data analysis.
The dependent variable (CEO remuneration) was broken up into fixed remuneration and variable remuneration, where the former refers to guaranteed pay and the latter refers to bonus payments, made to the CEO, in the financial year under review. All other independent variables were presented as is. The data was prepared and run through Statistical Package for the Social Sciences (SPSS). Descriptive statistics were run to determine the mean and standard deviation for the variables.
Because of the small sample size, a
To test the strength of the relationship between the dependent variable (CEO remuneration) and the proxy variables for the two categories being tested (organisation size and organisation performance), a Pearson correlation test was used. Thereafter, inferential statistics were drawn using regression analysis to test the hypotheses and reach conclusions from the results.
The next section provides the results determined from the research analysis. The results convey the research hypotheses, contrasted against the findings of the study.
Regression analyses were used to test the hypotheses.
Research hypothesis one tested whether organisation size is a significant determinant of CEO fixed remuneration:
H0 Organisation size is not a significant determinant of CEO fixed remuneration
H1 Organisation size is a significant determinant of CEO fixed remuneration:
H1a Assets is a significant determinant of CEO fixed remuneration H1b Revenue is a significant determinant of CEO fixed remuneration H1c Number of employees is a significant determinant of CEO fixed remuneration
Assets had a non-significant positive effect on CEO fixed remuneration. This implies that assets do not influence CEO fixed remuneration, and therefore, the null hypothesis cannot be rejected.
Revenue had a non-significant positive effect on CEO fixed remuneration. This implies that revenue does not influence CEO fixed remuneration, and therefore, the null hypothesis cannot be rejected.
Number of employees had a non-significant positive effect on CEO fixed remuneration. This implies that the number of employees does not influence CEO fixed remuneration, and therefore, the null hypothesis cannot be rejected.
Multiple regression – Organisation size on chief executive officers’ fixed remuneration.
Organisation size | 2015 | 2016 | 2017 | 2018 | 2019 | 2015–2019 |
---|---|---|---|---|---|---|
0.587 | −0.043 | 0.043 | −0.037 | −0.622 | 0.040 | |
−0.017 | 0.248 | 0.175 | −0.063 | 0.190 | 0.096 | |
0.041 | 0.359 | 0.335 | 0.480 | 0.864 | 0.396 | |
0.375 | 0.296 | 0.271 | 0.156 | 0.216 | 0.263 | |
Adjusted |
0.205 | 0.104 | 0.072 | −0.074 | 0.003 | 0.062 |
Sample size | 15.00 | 15.00 | 15.00 | 15.00 | 15.00 | 15.00 |
The models adjusted
Research hypothesis two tested whether organisation size is a significant determinant of CEO variable remuneration:
H0 Organisation size is not a significant determinant of CEO variable remuneration
H2 Organisation size is a significant determinant of CEO variable remuneration:
H2a Assets is a significant determinant of CEO variable remuneration H2b Revenue is a significant determinant of CEO variable remuneration H2c Number of employees is a significant determinant of CEO variable remuneration
Assets had a significant negative effect on CEO variable remuneration in 2017 but when an aggregate regression was run, a non-significant positive effect on CEO variable remuneration was found for the entire 5-year period. This implies that assets do not influence CEO variable remuneration, and therefore, the null hypothesis cannot be rejected.
Revenue had a significant positive effect on CEO variable remuneration in 2017 and 2018 but a non-significant effect on CEO variable remuneration when an aggregate regression was run, for the entire 5-year period. This implies that revenue does not influence CEO variable remuneration and therefore, the null hypothesis cannot be rejected.
Number of employees had a non-significant positive effect on CEO variable remuneration. This implies that the number of employees does not influence CEO variable remuneration, and therefore, the null hypothesis cannot be rejected.
Multiple regression – Organisation size on chief executive officers’ variable remuneration.
Organisation size | 2015 | 2016 | 2017 | 2018 | 2019 | 2015–2019 |
---|---|---|---|---|---|---|
0.590 | 0.319 | −0.377 | 0.540 | 0.632 | 0.279 | |
−0.247 | −0.06 | 0.203 | 0.253 | 0.236 | 0.116 | |
0.214 | 0.42 | 0.95 | −0.075 | −0.255 | 0.331 | |
0.377 | 0.463 | 0.59 | 0.484 | 0.348 | 0.485 | |
Adjusted |
0.207 | 0.317 | 0.478 | 0.344 | 0.171 | 0.345 |
Sample size | 15.00 | 15.00 | 15.00 | 15.00 | 15.00 | 15.00 |
The models adjusted
Research hypothesis three tested whether organisation size is a significant determinant of CEO total remuneration:
H0 Organisation size is not a significant determinant of CEO total remuneration
H3 Organisation size is a significant determinant of CEO total remuneration:
H3a Assets is a significant determinant of CEO total remuneration H3b Revenue is a significant determinant of CEO total remuneration H3c Number of employees is a significant determinant of CEO total remuneration
Assets had a non-significant positive effect on CEO total remuneration. This implies that assets do not influence CEO total remuneration, and therefore, the null hypothesis cannot be rejected.
Revenue had a non-significant positive effect on CEO total remuneration. This implies that revenue does not influence CEO total remuneration, and therefore, the null hypothesis cannot be rejected.
Number of employees had a non-significant positive effect on CEO total remuneration. This implies that the number of employees does not influence CEO total remuneration, and therefore, the null hypothesis cannot be rejected.
Multiple regression – Organisation size on chief executive officers’ total remuneration.
Organisation size | 2015 | 2016 | 2017 | 2018 | 2019 | 2015–2019 |
---|---|---|---|---|---|---|
0.836 | 0.197 | −0.269 | 0.434 | 0.199 | 0.263 | |
−0.138 | 0.099 | 0.238 | 0.183 | 0.311 | 0.147 | |
−0.002 | 0.494 | 0.886 | 0.159 | 0.262 | 0.418 | |
0.534 | 0.59 | 0.689 | 0.559 | 0.497 | 0.628 | |
Adjusted |
0.408 | 0.479 | 0.604 | 0.438 | 0.359 | 0.527 |
Sample size | 15.00 | 15.00 | 15.00 | 15.00 | 15.00 | 15.00 |
The models adjusted
Research hypothesis four tested whether organisation performance is a significant determinant of CEO fixed remuneration:
H0 Organisation performance is not a significant determinant of CEO fixed remuneration
H4 Organisation performance is a significant determinant of CEO fixed remuneration:
H4a Profit is a significant determinant of CEO fixed remuneration H4b Return on equity is a significant determinant of CEO fixed remuneration H4c Earnings per share is a significant determinant of CEO fixed remuneration
Profit has a significant positive effect on CEO fixed remuneration. This implies that profit has an influence on CEO fixed remuneration, and therefore, the null hypothesis is rejected in favour of hypothesis H4a.
Return on equity has a significant negative effect on CEO fixed remuneration. This implies that ROE does influence CEO fixed remuneration, and therefore, the null hypothesis is rejected in favour of hypothesis H4b.
Earnings per share has a significant positive effect on CEO fixed remuneration. This implies that EPS does influence CEO fixed remuneration, and therefore, the null hypothesis is rejected in favour of hypothesis H4c.
Multiple regression – Organisation performance on chief executive officers’ fixed remuneration.
Organisation size | 2015 | 2016 | 2017 | 2018 | 2019 | 2015-2019 |
---|---|---|---|---|---|---|
0.576 | 0.450 | 0.402 | 0.272 | 0.184 | 0.386 | |
−0.328 | −0.323 | −0.375 | −0.517 | −0.545 | −0.465 | |
0.254 | 0.393 | 0.385 | 0.435 | 0.488 | 0.374 | |
0.732 | 0.666 | 0.67 | 0.689 | 0.667 | 0.737 | |
Adjusted |
0.652 | 0.566 | 0.579 | 0.604 | 0.576 | 0.658 |
Sample size | 15.00 | 15.00 | 15.00 | 15.00 | 15.00 | 15.00 |
ROE, return on equity; EPS, earnings per share.
The models adjusted
Research hypothesis five tested whether organisation performance is a significant determinant of CEO variable remuneration:
H0 Organisation performance is not a significant determinant of CEO variable remuneration
H5 Organisation performance is a significant determinant of CEO variable remuneration:
H5a Profit is a significant determinant of CEO variable remuneration H5b Return on equity is a significant determinant of CEO variable remuneration H5c Earnings per share is a significant determinant of CEO variable remuneration
Profit has a significant positive effect on CEO variable remuneration. This implies that profit has an influence on CEO variable remuneration. Therefore, the null hypothesis is rejected in favour of hypothesis H5a.
Return on equity has a non-significant positive effect on CEO variable remuneration. This implies that ROE does not influence CEO variable remuneration. Therefore, the null hypothesis cannot be rejected.
Earnings per share has a non-significant negative effect on CEO variable remuneration. This implies that EPS does influence CEO variable remuneration. Therefore, the null hypothesis cannot be rejected.
Multiple regression – Organisation performance on chief executive officers’ variable remuneration.
Organisation size | 2015 | 2016 | 2017 | 2018 | 2019 | 2015–2019 |
---|---|---|---|---|---|---|
β profit | 0.573 | 0.694 | 0.829 | 0.763 | 0.754 | 0.804 |
β ROE | 0.235 | 0.214 | 0.128 | 0.21 | 0.267 | 0.229 |
β EPS | −0.005 | 0.101 | −0.126 | −0.041 | −0.124 | −0.060 |
0.295 | 0.488 | 0.638 | 0.549 | 0.596 | 0.599 | |
Adjusted |
0.084 | 0.335 | 0.539 | 0.426 | 0.486 | 0.479 |
Sample size | 15.00 | 15.00 | 15.00 | 15.00 | 15.00 | 15.00 |
ROE, return on equity; EPS, earnings per share.
The models adjusted
Research hypothesis six tested whether organisation performance is a significant determinant of CEO total remuneration:
H0 Organisation performance is not a significant determinant of CEO total remuneration
H6 Organisation performance is a significant determinant of CEO total remuneration:
H6a Profit is a significant determinant of CEO total remuneration H6b Return on equity is a significant determinant of CEO total remuneration H6c Earnings per share is a significant determinant of CEO total remuneration
Profit has a significant positive effect on CEO total remuneration. This implies that profit has an influence on CEO total remuneration. Therefore, the null hypothesis is rejected in favour of hypothesis H6a.
Return on equity has a non-significant negative effect on CEO total remuneration. This implies that ROE does not influence CEO total remuneration. Therefore, the null hypothesis cannot be rejected.
Earnings per share has a non-significant positive effect on CEO total remuneration. This implies that EPS does not influence CEO total remuneration. Therefore, the null hypothesis cannot be rejected.
Multiple regression – Organisation performance on chief executive officers’ total remuneration.
Organisation size | 2015 | 2016 | 2017 | 2018 | 2019 | 2015–2019 |
---|---|---|---|---|---|---|
0.717 | 0.737 | 0.825 | 0.763 | 0.756 | 0.818 | |
−0.035 | −0.032 | −0.077 | −0.063 | −0.074 | −0.072 | |
0.157 | 0.291 | 0.084 | 0.166 | 0.166 | 0.146 | |
0.534 | 0.59 | 0.689 | 0.559 | 0.497 | 0.769 | |
Adjusted |
0.408 | 0.479 | 0.604 | 0.438 | 0.359 | 0.699 |
Sample size | 15.00 | 15.00 | 15.00 | 15.00 | 15.00 | 15.00 |
ROE, return on equity; EPS, earnings per share.
The model’s adjusted
All variables used in the models were tested for internal reliability and generated an adequate score. The skewness and kurtosis of the data were assessed, and the data were cleared for inferential statistics as all measures were deemed fit.
The literature review, which is supported by the findings of this study, shows that a remuneration-performance link is required to drive alignment between the agent and the principal. Linder and Foss (
As such, the results of this study have shown that financial services organisations have tied both elements of CEO remuneration to the overall performance of the organisation, which places the agent’s total remuneration at risk in the absence of satisfactory organisational performance in the South African context.
These findings are in line with King IV, which recommends that organisations in South Africa use performance measures that support positive outcomes within the organisation and that organisations should provide an account of the performance measures that have been used in CEO remuneration setting (Institute of Directors Southern Africa,
Therefore, when remuneration committees apply the guidelines of King IV, a higher degree of transparency and trust is created among organisational stakeholders. Applying these processes provides stakeholders with the information they need to hold the board of directors accountable for any poor decisions made and assists in neutralising managerial power to a certain degree. This is especially the case where CEOs have considerable influence over the board and attempt to use this influence to extract higher remuneration.
In addition to the measures that are driven from a Companies Act and King IV perspective, remuneration committees can introduce measures such as the ‘say-on-pay’ approach to the remuneration setting. This approach allows shareholders to have oversight and input on the remuneration paid to CEOs. In their research, Newman, Banning, Johnson and Newman (
The focus of this research was on the top 15 JSE listed financial services organisations, based on their market capitalisation. As such, these organisations can be deemed large, and the findings of this research may not lend itself to organisations that are classified as medium or small organisations. Davis, Batchelor and Kreiser (
This research focused primarily on the effects of organisation metrics, being size and performance. Finkelstein and Hambrick (
For this research, the sample was limited to financial service organisations that were listed on the JSE. Therefore, no comparatives can be drawn across industries to determine whether the findings are applicable to other JSE listed organisations that fall outside of the financial services industry. Based on the sample selection criteria, the findings of this research may not be applicable to private financial services organisations that are not listed on the JSE.
Future research should include a broader spectrum of industries. This will allow for comparatives to be drawn between the findings of each industry and to determine whether there is a convergence or divergence of remuneration setting practices over time. Future research should also focus on share schemes, particularly during the decline in share prices early in 2020.
Chief Executive Officers’ tenure should be included in the research to determine whether tenure plays a role in the influence of organisation size on CEO fixed remuneration.
The size of the organisations selected for the study should be more varied by not selecting organisations on the basis of market capitalisation. Rather, market capitalisation should be used to categorise these organisations into small, medium and large organisations to draw parallels based on the findings. This will allow for a broader degree of applicability of the results from the study.
Given that market trend shifts, a longer period of observation could show the shifts in remuneration setting trends, if any appear in the findings of the research. For example, a study that looked back prior to 2008 may have findings that show the shift in trends post the financial crisis of 2008, thereby adding more depth to the research and its findings.
This section outlines the key findings from this study. The key finding of this study is that organisation size, as measured by assets, revenue and number of employees is not a significant determinant of CEO fixed, variable or total remuneration. In contrast, organisation performance, as measured by profit, ROE and EPS are significant determinants of CEO fixed remuneration and, when measured by profit, are significant determinants of CEO variable and total remuneration.
The mixed views on the influence of organisation size on CEO remuneration highlighted the lack of consensus on the influence of organisation size on CEO remuneration or the proxies that should be used to accurately predict CEO remuneration. The findings of this study showed no significant relationship between organisation size and any of the components of CEO remuneration even after using universally accepted variables as proxies for organisation size. In addition, the models that were created showed moderate to low levels of explanatory power that indicated that organisation size, as measured by assets, revenue and number of employees, is not a good predictor of CEO fixed, variable and total remuneration, when assessing their influence over the remuneration earned by CEOs of financial services organisations listed on the JSE.
No consensus exists on the influence of organisation performance on CEO remuneration or the proxies that should be used to accurately predict CEO remuneration. The findings of this study indicated a significant relationship between organisation performance, as measured by profit, ROE and EPS and CEO fixed remuneration. In addition, organisation performance, as measured by profit, proved to be a significant determinant of CEO variable remuneration and CEO total remuneration. The models that were created showed moderate to high levels of explanatory power, thus indicating that organisation performance is a good predictor of CEO fixed, variable and total remuneration, when assessing their influence over the remuneration earned by CEOs of financial services organisations listed on the JSE.
The authors have declared that no competing interests exist.
All authors contributed equally to this work.
This article followed all ethical standards for a research without direct contact with human or animal subjects.
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
Data is available upon reasonable request.
The views and opinions expressed in this article are those of the authors and do not necessarily reflect the official policy or position of any affiliated agency of the authors.