The effect of earnings per share categories on the share prices of the top 40 JSE listed companies

Abstract

Investment decisions can be influenced by different factors, including press coverage of an entity’s, recommendations made by advisors, tax consequences of the investment, the financial needs of the investor and published accounting information. Public entities publish annual financial statements, which investors collect and study in order to compare the performance of different entities before making investment decisions. Earnings per share (EPS) is considered as an important accounting indicator of risk, entity performance and corporate success. It is used to forecast potential growth in future share prices, because changes in EPS are often reflected in share price behaviour. The relationship between EPS and share prices can, ultimately, be attributed to the law of demand. Should investors analyse the EPS of a firm and find it satisfactory, the share of such an entity is acquired. This leads to an increase in the demand for this type of share and, consequently, in the share price. Companies listed on the Johannesburg Stock Exchange (JSE) are required to publish three different categories of EPS: basic, diluted and headline EPS. It has become apparent that there is no indication as to which category explains share price behaviour best. The study strived to answer the research question: Which category of EPS has the greatest effect, if any, on the share price behaviour of the top 40 JSE listed companies? No previous South African study has attempted to answer this question. The top 40 JSE listed companies were selected and the relationship between different categories of EPS and share prices was analysed empirically for the period 2005 to 2013. This study demonstrated that basic EPS correlated best with the changing behaviour of share prices of the top 40 companies. Furthermore, the study established that headline EPS proved to deliver lower correlation coefficients than other EPS categories. This study recommends that basic EPS should, ultimately, be considered by investors and business managers of the top 40 JSE listed companies. Headline EPS proved to be less useful; therefore, it can be posited that the measurement requirements of IAS 33 are sufficient in excluding headline EPS from its measurements. Limitations of the study included that the sample was limited to the top 40 JSE listed companies and that findings cannot be generalised. Furthermore, multicollinearity was present among independent variables, which eliminated the possibility of prediction modelling. Despite these limitations, the study was able to fulfil its primary objective by indicating that basic EPS and stock prices share the greatest interdependence.