Development and validation of bankruptcy prediction indicators of JSE listed companies

PhD (Accountancy), North-West University, Vanderbijlpark Campus, 2021The first attempt to identify bankruptcy as a result of financial distress dates back to the 1930s, when the first researchers Fitzpatrick (1931) and Merwin (1942) attempted to identify the potential of financial ratios as indicators of financial distress. Since then many studies have been conducted attempting to predict bankruptcy or financial distress. Bankruptcy has the ability to not only effect the bankrupt business itself but also their various stakeholders within the business. The ripple effect stretches to the corporate, social and political sphere surrounding businesses. Due to the potential bankruptcy has to cause significant damage, it is highly beneficial when bankruptcy can be predicted before if happens. The studies conducted in the field of bankruptcy prediction objective is to develop a model with the ability to accurately and consistently predict bankruptcy or business failure. A gap was indicated between theory and practice by the literature review performed in this thesis. Firstly, the majority bankruptcy prediction models are sector-based as no generic model exist. Secondly, South African bankruptcy prediction models developed are either based on American or European foundations and due to the diversity of South Africa these are not a perfect fit. The models are either too rigid or complexed or too flexible, which effect the consistency of the models. Based on these findings this thesis evaluated which financial ratios are able to adequately predict bankruptcy by applying an exploratory mixed-method research design. The financial ratios identified were nine ratios, return on equity, return on assets, total assets turnover, quick ratio, cash flow to debt, sales growth, earnings growth and interest cover. Five financial ratios were statistically significant: return on equity, return on assets, total assets turnover, cash flow to debt, and interest cover. In an attempt to meet the objectives this thesis sought to develop a generic bankruptcy prediction indicator as a tool that is easy to understand and apply by companies within any industry (excluding financial and mining). The quantitative findings of secondary data were supported by qualitative findings from the literature review. Both these findings support the newly develop bankruptcy prediction indicators. The newly developed bankruptcy prediction indicators were tested on Johannesburg Stock Exchange-listed companies, the study’s population. The study sample was acquired through the application of the following criteria: 1) Period of study – 2002–2016: period covering three different economic periods: • Pre-global crisis (2002–2006); • During global crisis (2007–2011); and • Post-global crisis (2012 – 2016). 2) Excluding financial and mining industry; 3) Including companies delisted (failed) due to bankruptcy; and 4) Available financial data five years prior to delisting resulting in a study sample of 17 failed companies, 46 non-failed companies and 5 out-of-sample companies. Three phase testing were done to validate the bankruptcy prediction indicators by comparing it to the renowned Altman’s Emerging market score model employed to study as a benchmarking measure. • Phase 1: Failed companies vs Non-failed companies – matched per sector • Phase 2: Failed companies – Different sectors • Phase 3: Out-of-sample data The study concluded that the bankruptcy prediction indicators are able to successfully predict bankruptcy or financial distress prior to business failure. Furthermore, it gives a good indication of bankruptcy or financial distress compared to the EMS in a South African context. The study proves that the newly developed bankruptcy prediction indicators can be understood and applied by all stakeholder within the business arena. Many studies attempted to predict failure, but none can ultimately be labelled as consistently accurate. The aim of this thesis was to examine whether the bankruptcy prediction indicators approach is able to predict bankruptcy or business failure early enough so that company management can take appropriate timeous remedial action.Doctora