FALLIBILITY OF THE ROUGH SET METHOD IN THE FORMULATION OF THE FAILURE PREDICTION INDEX MODEL OF DYNAMIC RISK

Bankruptcy is one of the most important entrepreneurial problems studied by financial theory. Despite this great effort, there is not a significant progress in order to predict the economic failure. In this way, the evidence suggests that this problem, related to the experimental design, is still present because of two main reasons: ignorance about bankruptcy process and the use of the accounting information as the unique input to construct the predictive models. In order to solve those problems, the RPV Model included both qualitative and accounting information with excellent results. So, the Earning Power Theory – upon which the RPV is based – could cause problems of specification and structure in the model. Empirical results not only verify those suspicions, but they made a stronger model possible by introducing to the equation ERC values adjusted to the risk.