Executive directors’ remuneration in South Africa: an examination of the effectiveness of corporate governance reforms Open Access

Navigating the aftermath of past corporate failures following the global financial crisis raises a crucial question: Have companies truly learned from their economic missteps, or are they still ensnared by flawed practices that have led to the downfall of corporate giants (Scholtz et al., 2022)? Such corporate failures have been blamed mainly on poor corporate governance practices, and specifically unethical excessive directors’ remuneration (EDR) practices (Padia and Callaghan, 2021). Steinhoff, a recent corporate scandal in South Africa, is one example where executive directors entered fraudulent transactions that allowed them to receive remuneration (i.e. pay) largely based on inflated firm performance (Cameron, 2018). Society’s tendency to reform regulatory measures whenever faced with new issues, such as corporate scandals, is evident in the increasing number of corporate governance codes issued in the last two decades (Romano, 2005; Dak-Adzaklo and Wong, 2024). The number of countries that have issued revised national codes of corporate governance, for example, increased from 24 countries in 1999 to 100 in 2015 (Chai-Aun et al., 2016). Research has however showed that such corporate governance reforms are often less effective in developing countries (Chai-Aun et al., 2016). Governance reforms are particularly important to study in developing countries like South Africa where costs associated with regulatory requirements are often cited as an important underlying reason for firms not to list or to delist. On average, 25 firms per year have, for example, delisted from the Johannesburg Stock Exchange (JSE) in South African between 2015 and 2022 (Larkin, 2022). The value of governance reforms, specifically in a developing country, is therefore not always that clear.

By examining recent remuneration governance reforms in South Africa, this study aimed to uncover the connection between EDR and firm performance, spotlighting its response to remuneration governance initiatives underscored by King IV. The reforms advocated by King IV as best practice recommendations align seamlessly with theoretical predictions of the optimal contracting theory. This study explored the relationship between EDR and firm performance, particularly examining potential shifts following the implementation of King IV. Surprisingly, despite its significance, no study has yet ventured into this specific effect of King IV, prompting our investigation into this uncharted territory.

South Africa provided an excellent setting for this study for several reasons. First, the King IV corporate governance regime has led to significant improvements in remuneration governance, the effects of which are yet to be explored. Second, excessive EDR has been identified as contributing to exacerbating income inequalities in South Africa (Lemma et al., 2020; Steenkamp et al., 2019) through concentrating wealth among a select few. This concentration of wealth at the top can result in economic disparities that extend beyond the corporate circles, affecting the broader societal well-being. The JSE, with its mix of established and smaller firms, along with its reputation as a leader in governance practices, provides an ideal empirical setting to examine whether the remuneration governance reforms brought about by King IV can help address the governance deficiencies and contribute to more equitable and performance-aligned pay structures. Third, despite the interest shown by several stock exchanges across Africa in promoting integrated reporting and aligning with global sustainability practices (Bananuka et al., 2019; Maama and Marimuthu, 2022), South Africa distinguishes itself as a country where integrated reporting has gained substantial traction. The JSE is particularly noteworthy in this regard, actively encouraging listed firms to adopt integrated reporting practices. This initiative not only promotes transparency but also ensures that financial data (including data on EDR) for public firms is readily accessible to stakeholders. Lastly, to the best of the authors’ knowledge, no empirical study has been conducted to assess how the relationship between EDR and firm performance has changed following King IV. This study sought to contribute to the literature by empirically exploring the link between EDR and firm performance among JSE-listed firms and documenting evidence on how this link changed after King IV came into effect.

South Africa is notable among African nations and emerging economies for its advanced corporate governance framework (Ntim et al., 2019; Sewpersadh, 2022). However, evidence on the alignment of EDR with firm performance is limited. Existing studies suggest a weak EDR-performance link in most emerging markets, reflecting a disconnect between executives’ and stakeholders’ interests (Ararat et al., 2021; Aslam et al., 2019; Bussin et al., 2023; Raithatha and Komera, 2016).

Theoretical and empirical literature highlights mechanisms for fostering long-term stakeholder value, including board monitoring (Agyemang Badu and Appiah, 2017; Elbadry et al., 2015; Jafeel et al., 2024), director shareholding (Martin et al., 2016), shareholder activism (Liu and Yin, 2023), remuneration disclosure (Clarkson et al., 2011), and performance-based EDR (Bussin et al., 2023; Groysberg et al., 2021; Jensen and Murphy, 1990). The King IV report underscores stakeholder capitalism, with remuneration governance as a key enabler (IoDSA, 2016). However, the implications of performance-based remuneration within South Africa’s governance landscape remain underexplored.

Consistent with agency and optimal contract theories, integrating incentive-based remuneration linked to firm performance alongside fixed remuneration is intended to align the interests of executive directors with shareholders and the strategic objectives of their firms (Bussin et al., 2023; Dias et al., 2020; Jensen and Murphy, 1990; Linder and Foss, 2015). Experienced, high-performing executives are in high demand, highlighting the need for boards to design remuneration contracts that attract and retain talent while ensuring sustainable stakeholder value and adherence to governance best practices (Alregab, 2015).

Using 2012–2022 data for firms listed on the JSE, this study, which examined the EDR-performance relationship has extended the remuneration governance literature in several ways. First, it adds to empirical evidence on the relevance of the agency, managerial power, and optimal contracting theories in the context of contemporary emerging markets. By demonstrating that incentive remuneration serves as a mechanism for aligning executives’ behaviour with the firm’s strategy, the research underscores the importance of remuneration governance as a developmental tool. Second, the study sheds light on the persistent concern surrounding the EDR-performance link within emerging markets. Despite advancements in remuneration governance policies and regulations, this relationship remains a topic of interest. Thirdly, the study provides new empirical evidence confirming the importance of governance reforms in a developing economy, where substantial reforms were made to ensure that firms in accordance with the King IV requirements remunerate fairly, responsibly and transparently (IoDSA, 2016) in a country that is known for having the highest income inequalities in the world.

The rest of this paper is structured as follows: Section 2 discusses the empirical literature related to remuneration governance, firm performance, and hypothesis development; Section 3 focuses on variables and research methods; Section 4 focuses on results presentation and discussion; and Section 5 concludes the study.

This section presents a detailed discussion of prior empirical literature on the relationship between EDR and firm performance, and the role of remuneration governance in influencing this relationship.

Agency theory posits that executives manage operations on behalf of shareholders, utilising their expertise to serve shareholders’ interests (Jensen and Meckling, 1976). Shareholders delegate authority due to limited time and expertise, forming the foundation of modern corporate governance (Linder and Foss, 2015). However, agency problems arise when executives use their informational advantage to prioritise personal interests (Attig et al., 2006; Elbadry et al., 2015).

To address agency problems, boards design remuneration contracts aligning executive behaviour with firm performance, adhering to optimal contract theory principles (Huu Nguyen et al., 2020; Jensen and Murphy, 1990). These contracts link remuneration components to performance, ensuring rewards are earned only after meeting targets (Defusco et al., 1991). Conversely, managerial power theory suggests powerful executives may influence remuneration structures to prioritise personal gain, resulting in higher EDR unrelated to firm performance (Bebchuk and Fried, 2003).

Empirical evidence on remuneration governance and the significance of linking EDR to firm performance, aimed at mitigate managerial opportunism and lower contracting costs, has been well-documented, primarily in developed markets (Benkraiem et al., 2017; Bhatia et al., 2024; Conyon and Murphy, 2002; Elmagrhi et al., 2020; Keasey et al., 2005; Murphy, 1985). In developed markets, governance reforms have established clearer measures for aligning EDR with firm performance, contributing to more accountable and performance-driven remuneration structures (Bebchuk and Fried, 2004; Brick et al., 2006; Cunat et al., 2016; Murphy, 2013) However, empirical evidence from emerging markets, including South Africa, presents a different picture. While the global focus has largely centred on developed markets, the applicability and effectiveness of these governance mechanisms in emerging economies remain less certain. Contemporary findings on the link between EDR and firm performance in emerging markets have been mixed and inconclusive, as some studies established a positive relationship (Aslam et al., 2019; Usman et al., 2019; Gulati et al., 2022), while other studies found an absence of a relationship (Luo, 2015; Raithatha and Komera, 2016; Saravanan et al., 2016).

These conflicting results are often attributed to the varying institutional dynamics and socio-economic conditions that characterise emerging markets (Bhatia et al., 2024; Fan et al., 2011). Factors such as weaker regulatory environments, concentrated ownership structures, and high levels of income inequality can complicate the EDR-performance alignment, as firms in these markets may face additional challenges in enforcing governance reforms. This study therefore seeks to contribute to this body of literature by examining the relationship between EDR and firm performance in the South African market, during the new era of enhanced remuneration governance mechanisms brought about by King IV. The next section details a review of prior studies on remuneration governance in emerging markets, South Africa in particular.

Corporate governance is increasingly recognised as a key driver of change (Ararat et al., 2021). Recent literature highlights its growing importance in emerging markets, shifting from addressing agency problems to broader reforms (Ararat et al., 2021; Clarke, 2015). A critical aspect of these reforms is remuneration governance, yet empirical evidence on the link between EDR and firm performance remains mixed (Kang and Nanda, 2017; Ndzi, 2016).

While some studies conducted in emerging countries, such as Unite et al. (2008) (and Aslam et al. (2019), found a positive EDR-performance relationship, others, like Raithatha and Komera (2016) and Watto et al. (2023), found no link. Similar inconsistencies appear in South African research on JSE-listed firms. Studies like Bussin et al. (2023), Lemma et al. (2020) and Naik et al. (2020) and support a positive link, while others, such as Padia and Callaghan (2021), report limited or no association. Furthermore, studies aligned with managerial power theory, including Bussin and Nel (2015), reveal negative links between fixed remuneration components and firm performance. Despite conflicting findings, there is consensus on the importance of designing optimal remuneration contracts that align executive and stakeholder interests. This study contributes by examining publicly listed firms during the King III and King IV eras, offering insights into evolving remuneration governance. Based on these perspectives, the following hypothesis was formulated:

The Institute of Directors in South Africa’s King Committee has been instrumental in driving corporate governance reforms in South Africa (IoDSA, 2016). South Africa has a long history of governance reforms since the publication of King I in 1994. While King I introduced recommendations for a remuneration report and total EDR disclosure, subsequent reports—King II (2002), King III (2009), and King IV (2016)—brought progressive reforms. King IV, effective for financial years starting on or after 1 April 2017, replaced King III entirely. It introduced a principles-and-outcomes approach, consolidating 75 principles into 17, and shifted from “apply or explain” to “apply and explain.” It retained the RAFT values from King III, adding integrity and competence to form ICRAFT values. King IV also emphasised transparency and accountability with a three-part remuneration report (background statement, policy, and implementation report) and encouraged a mix of variable and fixed remuneration for long-term value creation. Shareholder approval of the remuneration and implementation reports was also recommended.

Despite these reforms, developing countries like South Africa are often associated with ineffective governance reforms (Chai-Aun et al., 2016). For instance, in Pakistan, governance reforms weakened the link between governance practices and risk disclosures (Gull et al., 2022), while in Mauritius, they had little effect on the quality of corporate social responsibility disclosures (Soobaroyen et al., 2023). Studies using data from developed countries, such as Elsayed et al. (2022) and Banghøy et al. (2023), also support Chai-Aun et al.’s (2016) claim. Elsayed et al. (2022) found that corporate failures were less likely with lower executive remuneration levels. Similarly, Banghøy et al. (2023) noted that accounting reforms improved the remuneration-performance link only in Germany. To examine how King IV governance reforms in South Africa influenced the EDR-performance relationship, the study hypothesises the following:

This section presents the data sources, study variables, and research design used in this study. Following a positivist philosophy, the study adopted a deductive approach for a quantitative inquiry into the relationship between EDR and firm performance.

Following panel data tests recommended by Menard (2007) and Zulfikar and STp (2018), the F-test and Hausman test were conducted to determine the most suitable regression model. The F-test evaluates whether a fixed-effects (FE) model is preferable to a pooled OLS model by identifying unobserved differences across firms. The Hausman test was then applied to choose between FE and random-effects (RE) models by examining the correlation between unobserved heterogeneity and explanatory variables. In some regressions (i.e. 1, 3 presented in Tables 5 and 4–7 and 10 Presented in Table A1 included in  Annexure A), the test rejected the null hypothesis, confirming the FE model as more consistent and efficient. However, in others (i.e. 2 presented in Tables 5 and 8, 9, 11–15 in Table A1), the null hypothesis was not rejected, indicating the RE model as more appropriate in those cases.

In addition to panel regression, a quasi-natural experimental design with ANOVA regression was used to examine changes in the EDR-firm performance relationship following the implementation of King IV. This approach leverages natural policy variations to infer causality while controlling for confounding factors. ANOVA regression’s flexibility and clarity facilitate analysis of EDR changes over time, providing actionable insights for policymakers and stakeholders in corporate governance.

For this study, an initial sample of 100 JSE-listed firms was drawn for the period spanning 2012 to 2022. The sampling method employed was industry-based stratified random sampling, ensuring representation from both small and large firms, as elaborated in Section 4.1. To mitigate survivorship bias, all firms with a minimum listing period of three years were considered as part of the population e to allow for an adequate level of variation. After accounting for dropped observations where data were unavailable, the analysis was conducted using an unbalanced panel comprising 121 firms and 1 083 firm-year observations. Data for the study were sourced from the Iress database, and integrated annual reports (IAR), as detailed in Table 1. Table 1 further presents the variables used in the study, their brief descriptions, and the respective sources of data that were used.

In this study, the dependent variable that was adopted was EDR. The reason for this was that both the King III and King IV codes in South Africa emphasise the importance of aligning EDR with firm performance to promote fairness, accountability, and sustainability in remuneration practices (IoDSA, 2009, 2016). This alignment also facilitates comparison, as prior research commonly used EDR as the dependent variable (Elmagrhi et al., 2020; Luo, 2015; Raithatha and Komera, 2016; Usman et al., 2019). Given their distinct nature, this study relied on three proxies for the dependent variable (EDR): average short-term incentive remuneration (EDRS), average long-term remuneration (EDRL), and a composite variable (EDRI) that consists of EDRS plus EDRL. Basic remuneration was excluded since it is generally unresponsive to firm performance and EDR in South Africa primarily consists of EDRS and EDRL (Scholtz, 2024).

Firm performance variables were adopted as the independent variables. To align with the study objectives, variables were selected based on prior studies and relevant literature (Bussin et al., 2023; Elmagrhi et al., 2020; Lee and Isa, 2015; Ntim et al., 2019). We employed four proxy measures of firm performance, broadly categorised into accounting-based measures and market-based measures. The two accounting-based measures are EPS, which reflects the firm’s profitability, and ROA, which indicates the firm’s efficiency in generation profits from its assets. These proxies have been widely used in prior studies, for example, Aslam et al. (2019), Bussin et al. (2023), Kirsten and Du Toit (2018). The two market-based proxies used are TQ, which is a ratio of the market value of a firm’s assets to their replacement cost, serving as an indicator of the firm’s investment performance, and TSR which measures the total return received by shareholders through stock price appreciation and dividends over a specific period. These proxy measures have also been widely used in prior studies, for example, Aslam et al. (2019), Bussin et al. (2023), Wang et al. (2021). These measures of firm performance were selected based on their widespread use in the literature and their relevance to assessing both accounting and market-based aspects of firm performance. The inclusion of both accounting-based and market-based firm performance measures in one regression gives a comprehensive perspective of firm performance.

Consistent with prior EDR-performance studies, the current study incorporated a set of control variables in the regression models to address potential omitted variable bias. The following control variables were included in the model: leverage, which accounts for a firm’s debt levels and their impact on firm performance (Ortiz-Molina, 2007), revenue growth, which reflects the rate at which a firm’s sales are increasing over time and is considered an important determinant of firm performance (Park, 2023), firm size, measured by total assets, which controls for the scale of operations and its influence on firm performance (Blanes et al., 2020), and firm ownership, which captures the ownership structure of the firm. Firm ownership affects the firm’s decision-making processes and performance outcomes (Benamraoui et al., 2019). It is important to note that, consistent with the recommendation of Hünermund and Louw (2020), we did not include any specific expectations for the control variables or discuss their results (for example, level of significance and direction of relationship). Instead, these variables were included solely to control for potential confounding factors and ensure the validity of the analysis.

The King IV Report seeks to address stakeholders’ increasing concerns about EDR by promoting greater accountability and transparency (IoDSA, 2016). To achieve this, King IV introduced more explicit EDR disclosure requirements compared to its predecessors. King IV became effective for financial years starting on April 1, 2017. To provide firms with sufficient time for implementation, a dummy variable of 1 was assigned to all financial year-ends from 2019 to 2022.

Table 2 presents the descriptive statistics for the study. The natural logarithm transformation was applied to seven variables: EDRS, EDRL, EDRI, EPS, ROA, TSR, and SIZE, to reduce the skewness in distribution. Descriptive statistics for these variables are presented prior to the natural logarithmic transformations.

Table 2 displays significant cross-sectional variation for all variables, consistent with our sample selection strategy. The table reveals that the mean values of all incentive components of EDR are substantially higher than their respective median values, indicating a positively skewed distribution of EDR. Firms offering incentive remuneration tend to provide higher incentive packages, as demonstrated by the maximum EDRI value of R1 299.29 m and the sample mean EDRI of R9.71 m. Additionally, the results indicate that EDRL contributes relatively higher to EDRI relative to EDRS. This finding underscores the importance of considering the composition of remuneration packages in subsequent analyses. Furthermore, the mean values of all firm performance components are notably higher than their respective median values, suggesting that while few firms in the sample excel, the majority of the sampled firms exhibit a moderate or lower performance level.

Table 3 presents the Pearson correlation coefficients among firm performance measures, and incentive remuneration components. Multiple regression models necessitate the absence of multicollinearity between dependent, independent and control variables. This correlation analysis was particularly useful as it enabled a more precise examination of variations in the relationship between EDR and firm performance across the dataset than would have been feasible through regression equations alone. According to Arceneaux and Huber (2007), a serious multicollinearity issue arises if the correlation coefficient between variables is 0.7 or higher. N All correlation coefficients presented in Table 3 were lower than 0.7, except for the correlations between EDRS and EDRI. TIt should however be noted that the high correlation between EDRI and EDRS does not raise multicollinearity concerns, as the components are used in separate regressions as dependent variables. All relationships between the dependent and independent variables except for the EDRL-TSR and EDRI-TSR relationships were found to be significant and positive, as expected.

To test the first hypothesis (H1), the study employed panel linear regression analysis to determine the relationship between executive directors’ remuneration and firm performance by stating the following baseline model:

where $EDRit$ denotes executive remuneration, $PERFORMit$ represents performance variables, measured in terms of four proxies, namely EPS, TSR, and TQ. $CONTROLit$ refers to the set of control variables included in the study; i denotes firm; t denotes time in years. $β0$ is the intercept; $β1$ denotes the beta coefficient of focus; $δk$ is the regression coefficient associated with each of the 6 control variables, where k ranges from 2 to 7. $μi$ represents firm specific fixed or random effects and $єit$ denotes the residual of the model. The full definitions of variables are presented in Table 1. As discussed in Section 3.2.1, this study relied on three distinct proxies for the dependent variable, EDR. Table 4 therefore shows the empirical results for three separate regression analysis.

Table 4 provides the empirical findings on the relationship between EDR and firm performance.

The results provide evidence to support the first hypothesis (H1) as a significant association was found between EDR and various measures of firm performance. Firstly, the coefficient for the market-based firm performance measure (TSR) is positively and significantly associated with EDRS at the 5% level. Similarly, TQ shows a positive and statistically significant relationship at the 1% level with both EDRL and EDRI. Moreover, the accounting-based measure of firm performance, EPS, exhibits a positive and statistically significant association at the 1% level with EDRL. The results can be interpreted as evidence that performance-based remuneration paid to executive directors (i.e., EDRS, EDRL, and EDRI) is linked to both accounting-based and market-based firm performance. These results confirm that JSE-listed firms during the period under review implemented the practice of optimal contracting by linking EDR contracts to firm performance. The results align with the findings of two other South African studies, specifically Lemma et al. (2020) and Naik et al. (2020).

However, it is noteworthy that TSR is significantly and negatively associated with EDRL. This negative association may be due to TSR being influenced by share price performance and dividends, which are affected by external market conditions beyond executives’ control (Čupić and Todorović, 2011). The timing of stock or option grants may also play a role, as King IV recommends a three- to five-year vesting period for long-term share-based payments (IoDSA, 2016). Additionally, the finding may reflect a strategic response to income inequality concerns in South Africa, showing a commitment to responsible remuneration and pay-for-performance principles. It may also indicate the use of claw back mechanisms, enabling firms to adjust long-term incentives based on subsequent TSR levels.

On the other hand, ROA does not show a significant association with executive directors’ incentive remuneration. ROA measures how effectively management uses assets to generate profit (CFI, 2024). The insignificant association may stem from EDR being designed to reward short-term goals rather than promote long-term efficiency and asset management.

Given that the EDR variable was transformed using the natural logarithm, the significant coefficients of the respective components were back-transformed (exponentiated) to enhance interpretability. This transformation allows for expressing the effect of firm performance on EDR in terms of a percentage change in the original scale of EDR. Consequently, the results imply that, on average, every unit increase in EPS is associated with an approximately 8% increase in EDRL, while every unit increase in TSR corresponds to an approximately 6.2% increase in EDRS. Additionally, every unit increase in TQ is associated with an approximately 21% increase in EDRL and an approximately 15.8% increase in EDRI. Notably, a unit increase in TSR is associated with an approximately 4% decrease in EDRL. Overall, the results are partly consistent with the study hypothesis (H1) indicating a positive relationship between EDR and firm performance among JSE-listed firms. These findings align with theoretical predictions of optimal contracting (Cohen et al., 2023; Jensen and Meckling, 1976), and confirm prior research conducted in the South African context (Bussin and Nel, 2015; Lemma et al., 2020; Ntim et al., 2019; Scholtz and Smit, 2012).

To assess the causal impact of King IV on the EDR-firm performance relationship, we employed a mixed model ANOVA. The regression model is specified as follows:

where; $Yi$ is the dependent variable: short-term incentive remuneration (EDRS) for a firm i at time t; β₀ is the intercept; β₁ is the coefficient associated with King IV guidelines variable (dummy variable 0 for firm-years before King IV; 1 for firm-years after King IV); β₂ is the coefficient associated with the measure of performance (ROA or TQ), representing the baseline effect of firm performance on EDR. β₃ represents the coefficient associated with the interaction effect between King IV and firm performance measures, capturing how King IV modifies the relationship between EDR and firm performance. β_k represents the coefficient associated with each of the six control variables where k ranges from 4 to 9. $ui$ represents the error term capturing unobserved factors influencing EDR.

Tables 5 and Table A1 present the empirical findings on the impact of King IV on the EDRS–firm performance relationship. Table 5 includes only regressions with statistically significant results, while Table A1, provided in  Annexure for brevity, contains the results of all fitted regressions.

In the stated model (2), the variable of focus is the coefficient of the interaction term β₃, which captures the changes in EDR-firm performance link after the implementation date of the King IV remuneration guidelines. A positive or (negative) coefficient on β₃ denotes an increase / (decrease) in EDR-firm performance link after King IV implementation. Table 5, shows two ANOVA regression analysis (regressions (5) and (7) out of the 12 detailed results presented in Table A1. The negative coefficient (β₁) observed in both regressions suggests that the implementation of King IV had an overall negative independent effect on executive directors’ remuneration. However, the coefficients on the interaction terms (β₃) are significantly positive, indicating that following the implementation of King IV remuneration governance guidelines, JSE-listed firms experienced an improvement in the link between EDR and firm performance. These positive coefficients indicate that King IV played a significant role in enhancing the relationship between EDRS and firm performance, as measured by both accounting-based measures (ROA) and market-based measures (TQ). The results suggest that the King IV reforms encouraged firms to remunerate executive directors fairly, responsibly, and transparently. Furthermore, the King IV reforms promoted the practice of optimal contracting by linking EDR contracts to firm performance. The results also reveal that, without considering the presence of other variables in the model (marginal R²), approximately 10% (2) and 12% (4) of the changes in the EDR-firm performance link, measured by ROA and TQ respectively, can be explained by King IV. However, when considering the effects of King IV along with other variables in the model (conditional R²), approximately 34% (2) and 35% (4) of the change in the EDR-firm performance link is attributable to the variables included in the model. This indicates that the inclusion of other variables significantly contributed to the model’s overall goodness of fit. The results confirm that King IV had a substantial impact on improving the relationship between short-term incentive remuneration and both ROA and TQ among JSE-listed firms. Additionally, least squares (LS) mean graphs (Figures 1 and 2) were employed to provide further clarity and facilitate interpretation. Figures 1 and 2 demonstrate significant variation in the EDR-firm performance link between the pre-King IV and post-King IV implementation periods, relative to the variations within each period.

Diagnostic tests ensured the data met assumptions of normality, homoscedasticity, and multicollinearity. Due to slight non-normality, leverage and growth variables were winsorized at the 1% and 99% levels. The Breusch-Pagan test identified heteroskedasticity in all models, significant at 1%, which was addressed using the Huber-White robust covariance method (MacKinnon and White, 1985). Results adjusted for heteroskedasticity are shown in Table 4.

Multicollinearity was assessed using VIF, with all values below 2 and within acceptable tolerance limits (Weisburd and Britt, 2013). The Durbin-Watson test indicated no autocorrelation (values below 2) across all regressions (Table 4). The model specification test confirmed proper explanatory variable selection, and the significant Wald test (p = 0.000) verified the model’s fit and effectiveness in capturing the intended relationships.

Roberts and Whited (2013, p. 496) caution that many researchers fail to adequately address endogeneity. Endogeneity arises when an independent variable is correlated with the error term in a regression model (Wooldridge, 2009, p. 838). Endogeneity in EDR and firm performance studies can arise from simultaneity, where an independent variable is also influenced by the dependent variable, creating a two-way causal relationship (Wooldridge, 2009). The Durbin-Wu-Hausman test, frequently used in the literature (Ataay, 2018; El-Sayed, 2013), is designed to detect endogeneity. Conducting this test requires an instrumental variable (Roberts and Whited, 2013, p. 513). The Durbin-Wu-Hausman test is used to confirm or reject the suitability of the regression analyses used.

In this study endogeneity is evaluated by performing the Durbin-Wu-Hausman test and re-estimating the regression models using instrumental variables in the two-stage least-squares regression models. Following prior research (Ataay, 2018; El-Sayed, 2013), lagged performance variables were used as instrumental variables. The applicability of the panel data regressions was confirmed by the Durbin-Wu-Hausman test, which indicated no evidence of endogeneity.

Although EDR awarded in one year may depend on prior-year performance, this study measured dependent and independent variables concurrently. Under IAS 19, South African firms (and others using IFRS) must accrue EDR in the year it pertains to (Bussin et al., 2023; IASB, 2024). Most prior studies also use static agency models to examine EDR-performance relationships, relying on concurrent measures without lagged variables (Bhatia et al., 2024). These models assume EDR is based on performance achieved within the same year.

To enhance robustness, we conducted five additional regressions: three with lagged performance and control variables and two incorporating King IV variables from 2018. The EDRS-lagged performance relationships mirrored Table 4 findings, with TSR significant at the 10% level. However, no significant relationships were found between EDRL, EDRI, and lagged performance. Results with King IV variables were identical to those in Table 5.

This paper reports on a study that investigated the effectiveness of corporate governance reforms from a developing country perspective. To the best of the authors’ knowledge, this study is the first to explore how governance reforms association with King IV has influenced the relationship between executive directors’ remuneration and firm performance in the South African context.

The study employed panel linear regression analysis, with executive directors’ inventive remuneration as the dependent variable and firm performance proxied by accounting-based (EPS and ROA) and market-based (TSR and TQ) measures as the key independent variables. Contrary to the belief that EDR is not linked to performance, the empirical findings revealed a significant positive association between short-term executive directors’ incentive remuneration and TSR. Additionally, long-term executive director’s incentive remuneration, as well as total incentive remuneration, exhibited significant and positive associations with Tobin’s Q. Surprisingly, executive directors’ long-term incentive remuneration showed a significant negative relationship with TSR. This unexpected finding may reflect a strategic response to concerns over income inequality in South Africa, signalling a commitment to responsible remuneration practices and alignment with the pay-for-performance ideology. Overall, the study results indicate that executive directors’ incentive remuneration is significantly and positively associated with both accounting-based (EPS) and market-based (TSR and TQ) measures of firm performance, confirming the study hypothesis that there is a positive relationship between executive remuneration and firm performance among JSE-listed firms.

Moreover, the study employed a quasi-natural experimental design to investigate the effects of King IV on the relationship between executive directors’ remuneration and firm performance, treating the introduction of King IV remuneration governance guidelines as the treatment event. Through ANOVA regression analysis, the study rigorously examined how the EDR-firm performance relationship in publicly listed firms changed following the implementation of King IV.

The results provide evidence that King IV promoted the practice of optimal contracting theory as an improvement in the relationship between short-term incentive remuneration and both accounting-based (ROA) and market-based (TQ) measures of firm performance after the implementation of King IV was found. The results also suggest that King IV encouraged firms to provide fair, responsible, and transparent remuneration. While the study partially confirmed the hypothesis that the link between executive remuneration and firm performance improved after King IV came into effect, it also established that the relationship between executive directors’ long-term incentive remuneration and firm performance did not change significantly post-King IV.

The implications of this study are manifold. Firstly, the study provides valuable insights to the literature on the EDR-performance link in emerging markets, particularly in South Africa. Emerging markets often have unique economic, social, and regulatory environments that can influence corporate practices, including executive remuneration. Understanding EDR-performance dynamics in these contexts is crucial for policymakers, investors, and practitioners in such markets. Policymakers can use these findings to establish guidelines ensuring that EDR is linked to measurable performance metrics. Investors may use the results to identify firms to invest in that award optimal remuneration contracts linked to firm performance. Secondly, the study contributes to a greater understanding of incentive remuneration as a mechanism for behaviour and interest alignment. By understanding how incentives shape behaviour, firms can tailor remuneration packages to align more effectively with strategic objectives and performance goals. Thirdly, the study provides novel empirical evidence on the influence of King IV on pay-performance link in South Africa and it offers insights into the effectiveness of governance reforms in achieving their intended goals. In South Africa, a country with unique socio-economic challenges, governance reforms have been critical in promoting transparency and curbing excessive pay, particularly in light of the high-income inequality as evidenced by its Gini coefficient.

This information can be valuable for policymakers and regulators in evaluating the impact of policy changes and making informed decisions about future reforms. Lastly, the study underscores the importance of adhering to King IV recommendations, even for unlisted entities in South Africa. Therefore, South African firms yet to fully embrace King IV remuneration governance guidelines should consider incorporating both short-term and long-term incentive remuneration components into their executive remuneration contracts to align interests, mitigate agency problems, and ultimately improve firm value.

This study makes valuable contributions but has some limitations. It focuses only on popular accounting- and market-based performance variables to measure firm performance. Future research could include non-financial performance measures, such as environmental, social, and governance (ESG) factors, to further explore the EDR-performance relationship. Additionally, the impact of corporate governance factors or a governance index on the EDR-performance link was not considered. Future studies could examine the role of governance measures, such as compliance scores or board characteristics (e.g., board size and independence), in influencing this relationship. Lastly, this study does not address the effect of EDR on future firm performance. Future research could investigate both the EDR-performance and performance-pay relationships to understand how EDR impacts future outcomes post-governance reform.