There has been a marked increase in executive pay, and there is recognition that remuneration arrangements are no longer aligned with the interests of the shareholder (Frydman & Jenter, 2010). This may be because of flaws in the governance processes that dictate the arrangements (Bebchuk & Fried, 2005). The compensation policy of the company is considered as one of the most important factors that contribute to its success and is a controversial topic (Jensen & Murphy, 1990; O’Reilly & Main, 2010).

Erick, Kefah and Nyaoga (2014) raised an interesting point: if executives are not adequately paid, then there may be a situation of inadequate incentive to perform. Furthermore, Deysel and Kruger (2015) indicated that there needs to be a balance between long- and short-term pay, as well as keeping the market value of the executive role in mind. Murphy (1998) noted that if executives were only paid a fixed salary, there would be little incentive to improve company performance because they do not get paid in relation to the performance. More recent literature indicates that there appears to be a trend towards the pay strategy of executives moving away from performance-related elements (Bussin & Modau, 2015). Again, there seems to be a conflict between the patterns regarding the structuring of executive pay.

There has been a significant amount of research that has been conducted on the relationship between executive pay, considering the chief executive officer (CEO) only, and the associated company performance. A summary of this research is detailed in Table 1. The existing research has covered a wide range of countries and industries, and there was a high level of variance in the observed results. It was noted that 56.5% of previous studies reviewed showed a significant relationship of varying strength between executive pay and company performance. The remaining 43.5% found no significant relationship at all. Hence, it can be seen that there is currently no decisive conclusion regarding this relationship. Further to this, there are limited studies applicable to the South African market. Of these studies, only two considered the state-owned entity (SOE) environment, namely, studies by Bezuidenhout (2016) and Ngwenya and Khumalo (2012).

The problem is that executive pay (otherwise known as remuneration or compensation) is on the increase, whilst associated company performance is stagnating or even declining. This perception of a lack of performance is noted by Marshall and Lee (2016) where the pay of the CEO – as the lead executive of the company – does not reflect the long-term performance of the company. On the other hand, Schumpeter (2012) captured the argument that executives are remunerated for improving their company’s performance. According to Jensen and Murphy (1990) – contrary to the literature at the time – executive pay was worsening with respect to company performance. Therefore, the primary objective of this study was to evaluate whether there was a relationship between executive pay and associated company financial performance, within the context of South African SOEs.

The findings of this study contribute to the existing body of literature surrounding the pay–performance relationship, specifically in the SOE environment.

There is no single definition of company performance, and it can consist of various inputs and measures. Typically, performance is considered on accounting-based and market-based measures (Attaway, 2000). Murphy (1998) is of the opinion that accounting measures should be used to describe company performance. These accounting measures should include absolute financial metrics, as well as financial performance ratios. Other researchers believe that the absolute performance measures (such as revenue and profit) are ideal because they are observable measures from audited annual financial statements (Tosi, Werner, Katz, & Gomez-Mejia, 2000). Previous studies have included return on equity (ROE) as the predominant definition of performance (Attaway, 2000; Bussin & Modau, 2015; Bussin & Nel, 2015; De Wet, 2012). Ngwenya and Khumalo (2012) argued that ROE can be manipulated to improve the ratio; hence, using return on assets (ROA) is a more reliable measure.

State-owned entities are autonomous companies that play an essential role in the South African economy. They are owned (at least in part) by the South African government. These companies fulfil specific functions and are expected to operate within the governing legislation to promote effective and efficient service delivery (Wendy Ovens & Associates, 2013).

State-owned entities differ from a typical company as a result of their positioning between the public and corporate environments. An implication of this is that the South African government is empowered as both the supervisory body and a majority (or sometimes sole) stakeholder (Arries, 2014). It does not preclude the SOE from performing profit-making activities to generate returns for their stakeholders. The source of funds for SOEs, outside of normal business operations, is the state.

Odainkey and Simpson (2013) proposed that SOE performance is a measure of how well the problem of agency theory has been addressed. The performance of SOEs is supported by the implementation of good corporate governance. Arries (2014) highlighted that SOEs need to comply with more legislation and laws than regular companies, inclusive of the Constitution of the Republic of South Africa, the Companies Act, the Public Finance Mismanagement Act, as well as the King Report. Based on this, it is apparent that there is no fully inclusive SOE legislation framework in South Africa (Presidential Review Commission, 2012).

Of the 23 studies reviewed (as detailed in Table 1), 10 were from South Africa, three from Kenya and the remaining 10 from across the northern hemisphere (North America, Europe and Asia). The studies covered a wide range of business sectors, of which only two focussed on SOEs. The period of research varied across the studies, where the average research period was 7 years.

From the existing literature reviewed, there was no consensus as to whether there was a relationship between executive pay and company financial performance. Of the 23 studies considered, 13 showed that there was a significant relationship of various strengths, and 10 showed no significant relationship. The metrics used varied from study to study, whilst the most common were ROE (16 studies), ROA (11 studies), revenue (seven studies), earnings before interest, taxation, depreciation and amortisation (EBITDA – six studies) and solvency ratio (four studies).

This research entailed a desktop study of an archival nature using secondary data to define the dependent and independent variables (Saunders, Lewis, & Thornhill, 2012). This study was ex-post facto where the characteristics of the variables were reported rather than manipulated (Cooper & Schindler, 2014).

This study considered panel data, and therefore a longer period of study allows for a more reliable set of results to be achieved (Hsiao, 2007). The research period was chosen to be 8 years from 2009 to 2016. This period was not extended further into the past, as the researchers did not want to introduce potential instabilities in the data as a result of the global financial crisis in 2008 (Campello, Graham, & Harvey, 2010).

The pay–performance relationship was considered for each company included in the study as well as the complete dataset.

Data analysis typically involves the cleaning and organising of data, describing the data and finally answering the research questions using inferential statistics (Trochim, 2006). This was achieved using a three-stage approach.

This article followed all ethical standards for carrying out research without direct contact with human or animal subjects.

This study defined executive pay as the total pay (the sum of fixed pay and variable pay).

Interestingly, in 2013, there was a marked decline in executive pay (17.67% average year-on-year decrease from 2012), with half of the companies in the sample exhibiting this declining trend. Aside from the exception to the trend in 2013, there was still an average growth in total pay of 28.68% over the 8 year research period, or a year-on-year average increase of 5.00%.

The financial performance measures adopted for this research included revenue, operating profit or loss, net profit or loss, ROE, ROA and the solvency ratio.

The aim of this research was to determine whether there was a relationship between executive pay and company financial performance. The sections that follow detail the results of the correlation analysis performed. In the results tables, the significant correlation coefficients have an indicator whether it was significant at the p < 0.05 level or the p < 0.01 level. The sign in front of the correlation coefficient indicates whether the relationship was positive or negative. The strength provides the interpretation of the correlation coefficient, using the ranges in Table 2 as a reference. The total pay for the CEOs of each of the 14 companies was evaluated against the six financial metrics identified, implying that there were 84 potential relationships investigated.

From the results presented in Table 3, the total pay and associated company financial performance was not consistent across the different companies. However, there were a greater number of strong or better correlations (57.14%) in the sample, with 20 of the 84 relationships (23.81%) showing a statistically significant coefficient. These significant relationships were all very strong, with nine related to operating profit or loss, six related to revenue and three related to net profit or loss. There were no significant results observed for ROA.

Only three companies showed no significant relationship between total pay and company performance, although some of the correlation coefficients were strong.

This study sought to determine whether there was a relationship between executive pay and company performance in South African Schedule 2 SOEs. When considering the overall dataset results, as seen in Table 4, there was one significant correlation between total pay and net profit or loss. Whilst the relationship was observed to be significant, the actual relationship was weak. Hence, it can be concluded that there is a weak relationship between total pay and company performance in South African Schedule 2 SOEs.

When considering the companies included in this study separately, six of the 14 showed a significant, very strong relationship. However, when considering the total dataset, there was a non-significant, negligible relationship. This non-significant result supports the findings of Ngwenya and Khumalo (2012) and Gigliotti (2013). However, there were a number of previous studies that found results contrary to this observation, which showed significant relationships of various strengths (Bezuidenhout, 2016; Scholtz & Smit, 2012; Theku, 2014; Van der Laan et al., 2010; Zhou, 2010).

Of the studies reviewed, it appears that the results of this research contradict the majority of the studies that included revenue as a measure of performance.

Interestingly, the operating profit or loss metric showed the greatest number of significant correlations (nine of 14 SOEs) when considering the individual companies associated with total executive pay. When considering the whole dataset, it was found that there was a non-significant, negligible positive relationship with total pay. This set of non-significant results supports the findings of four other authors (Bezuidenhout, 2016; Deysel & Kruger, 2015; Scholtz & Smit, 2012; Van der Laan et al., 2010), whereas two other studies disagreed with the results of this research (Shaw, 2011; Theku, 2014).

It appears that the majority of research tends to agree that operating profit or loss exhibits no significant relationship with executive pay.

When considering the net profit or loss, only three of 14 SOEs yielded significant, very strong relationships. There was a different result observed when considering the total dataset where a significant weak positive relationship was seen. Shaw (2011) exhibited similar results as observed in this; however, the strength of the correlation was stronger. Similarly, Bezuidenhout (2016) reported a significant weak to moderate relationship, but the relationship was negative in comparison to the positive relationship observed in this study.

As net profit or loss is not one of the more common performance metrics used in research on this topic, the conclusion is not as robust. It appears that there could be a weak relationship between executive pay and net profit or loss.

The ROE was one of the most popular metrics used in the previous studies reviewed, with 16 of 23 studies using ROE as a metric. The results of this study showed that there was no significant relationship observed between total executive pay and ROE in the total dataset. The same (or similar) results were observed by Aduda (2011), Bezuidenhout (2016), Busaule (2014), Deysel and Kruger (2015), Gigliotti (2013), Lam et al. (2013), Shaw (2011), Theku (2014), Uyanik (2017) and Van Blerck (2012).

In comparison to this, six authors observed results that differed from those presented in this research. Zhou (2010), Sigler (2011) and De Wet (2012), as well as Raithatha and Komera (2016), showed a significant positive relationship of different strengths. Bussin and Nel (2015) found that there was a negative relationship. Finally, Bussin and Modau (2015) noted a significant relationship that appeared to be declining over the research period.

Whilst ROE is one of the most common metrics used to define company performance, the evidence from previous studies, as well as this study, points towards ROE not exhibiting a relationship with executive pay.

The ROA was the only financial performance metric used in this research that did not have any significant observations when considering the individual SOEs. It is also important to note that all observations for the total dataset were negligible. This finding supports the findings of five other studies (Aduda, 2011; Gigliotti, 2013; Lam et al., 2013; Luo, 2015; Ngwenya & Khumalo, 2012).

There were a number of other researchers who found significant positive relationships between executive pay and ROA. These included De Wet (2012) – strong relationship; Rahman (2017) – negligible relationship; Raithatha and Komera (2016) – moderate relationship; and Theku (2014) and Uyanik (2017) – weak relationship.

The evidence is not conclusive regarding the relationship between executive pay and ROA. However, the results of this study improve the validity of the non-significant relationship observed.

There was a majority of non-significant results observed for this performance metric. An interesting result was the number of negative relationships (of varying strength). For the overall dataset, a non-significant, positive, negligible relationship was observed. This result supported all three of the previous studies that used the solvency ratio as a performance metric (Bezuidenhout, 2016; Lam et al., 2013; Shaw, 2011).

From the evidence observed in previous studies and this research, it can be concluded that there is no significant relationship between executive pay and solvency.

It was observed in this study that there was limited evidence regarding the pay–performance relationship of SOEs. Executive pay was considered using total pay only, and company financial performance was defined using six metrics (revenue, operating profit or loss, net profit or loss, ROE, return on assets and solvency ratio). The pay–performance relationship was considered for each of the 14 SOEs included in the study as well as the dataset as a whole.

Whilst there were significant relationships observed, executive pay in the South African Schedule 2 SOE environment showed no conclusive relationship with company financial performance. Considering the total executive pay, there was only one significant relationship observed with net profit or loss which was weak. These results were similar to Shaw (2011). Whilst there were only weak relationships observed for some of the absolute financial measures, there were no significant relationships observed for the financial ratios. This supports the findings of Tosi et al. (2000).

These results support the proposition that there is an agency problem existent in South African SOEs.

Based on the results observed in this research, several considerations need to be catered for regarding the structuring of executive pay. The most important consideration needs to be addressed at a government level – the development of an all-encompassing SOE legislation framework as proposed by the Presidential Review Commission (2012). This will standardise the manner in which SOEs structure the pay of executives, as well as the reporting requirements. This would improve the quality and reliability of the data, as well as improve the relationship between executive pay and company performance.

Further to this, executive performance needs to be evaluated accurately to determine its impact on the performance of the company. Jensen, Murphy and Wruck (2004) indicated that the incorrect performance measures lead to inappropriate pay. Gopalan, Horn and Milbourn (2017) recommended a number of ways in which the attitude of the CEO can become more performance driven. This could include the use of multiple performance metrics, increasing pay based on the risk they are exposed to, reward relative to competitors and the inclusion of non-financial targets to ensure that the bottom line does not become the sole focus. If executive performance can be aligned to the expectations of the shareholder, then the agency problem that exists can be alleviated.

An agency problem appears to exist in South African Schedule 2 SOEs, where the CEOs are being paid in a manner than does not reflect the performance of their companies. This research showed that the pay–performance relationship was tenuous at best, with only weak significant correlations considering executive total pay and company financial performance. Hence, there was no conclusive relationship between executive pay and company financial performance in South African Schedule 2 SOEs.