ABSTRACT
This paper describes some of the problems which confront the practitioner and researcher when trying to use a company failure prediction model. It shows that the strict mathematical standards required by the methodology offer little chance of a modeI being applied consistently in practice.
The final section introduces Natural Selection—a new methodology for research into financial ratio analysis. It is a simple methodology which will allow the practitioner to 'pick-and-mix' his own selection of ratios. The chosen ratios could then be used individually, or combined to produce a model to measure changes in the financial health of companies.
