CEO pay ratios and company performance : a study of JSE-listed consumer goods and services companies

The disparity in remuneration between company CEOs and other employees is a topical and highly controversial issue globally. Theoretically, there are two explanations for this pay disparity - tournament theory and behavioural theory. Tournament theory says that employees are more motivated to compete with a larger pay gap, while the behavioural theories say that employees feel inadequate and thus demotivated in the presence of a larger pay gap, resulting in poorer performance. In response to growing concerns about the pay gap, new legislation in the USA has required companies to disclose their pay ratios1 in their financial statements, which is also likely to come to South Africa. As a means to explore CEO pay ratios in a South African context, a study of the determinants and performance effects of companies' CEO pay ratios was conducted in the Consumer Goods and Consumer Services subsectors on the JSE. Data was collected on companies for the period 2006 to 2014 and pay ratios were estimated for each company where the data allowed. Due to the complexity of CEO remuneration, three different pay ratios were calculated, which differed in how long-term incentive payments were treated in each case. Using the same method as Shin, Kang, Hyun, & Kim (2015) used in their South Korean study, three different analyses were conducted. Firstly, the factors determining pay ratios were analysed in a regression analysis, which found CEO tenure, companies' future investment opportunities and company size to be key determinants of pay ratios. Secondly, the deviations from companies' expected pay ratios were regressed against subsequent company performance to see whether CEOs being paid the, "wrong," amount relative to employees affects company performance. It was found that deviations from the expected pay ratio negatively affected company performance, and there was no difference in performance between under- and over-paying CEOs relative to employees. Finally, as a means to test whether tournament theory or behavioural theories better explain the CEO pay ratio in South Africa, subsequent company performance was regressed against the three different pay ratios calculated. It was found that there is little evidence of a relationship between subsequent company performance and the pay ratio, except in the case where performance is measured by return on assets, and the pay ratio is measured such that it excludes long-term incentives completely. The relationship in this case was found to be positive, indicating that tournament theory better explained the relationship between pay ratios and company performance. One of the limitations of this study was the limited availability of data, which gives rise to self-selection bias