The Relative Performance of Auditors’ Going-Concern Opinions and Statistical Failure Prediction Models in Jordan

There are several tools that can help investors and other stakeholders in the assessment of the going-concern status of the company. Among other statistical failure prediction models and audit reports of auditors can be of benefit. In this article, we evaluate how valid and accurate statistical failure prediction models are in a geographic area that is different from the context in which they were originally developed or tested (the United States and European countries), and we compare their predictive power with going-concern opinions (GCOs) of auditors. We examine the relative performance of the statistical failure prediction models and GCOs in Jordan, as an example of a developing country in the Arab region. We use the Altman Z-score model (1968) and two specially developed logit (Alareeni, 2011), as examples of statistical failure prediction models. We believe that the failure prediction statistical models generally give much better results than the GCOs. Our findings may help investors and other stakeholders to prevent or reduce losses when they are likely to be facing bankruptcy.