• The Dynamics of Corporate Death

    John Walter Hawkins (see profile)
    Item Type:
    University of Bradford, School of Management
    Permanent URL:
    An investigation into the use of mathematical models in the prediction of corporate failure, and their utility to corporate interest groups. The principle objective of this study was to investigate whether a model might exist with which interest groups could help determine the likely future state of a company. It was suggested that ratio analysis might provide the basis for one such class of models. A review of previous studies on the use of such models indicated the potential success of this approach, although three drawbacks were noted. Firstly, the data from which the most promising models were prepared (those due to E.I. Altman and T.G. Townsend) is now more than ten years old. Secondly, the data for these models was collected on United States companies. Thirdly, it was implicitly assumed that a model developed on the basis of data relating to the year immediately preceding bankruptcy would perform well when applied to years prior to this. Data was collected for fifty-one pairs of failed and non-failed companies. The technique of multiple discriminant analysis was used to analyse twenty-four potentially discriminating ratios and develop a model ultimately utilising six of these ratios. This model could accurately classify 82 per-cent of companies in the year prior to bankruptcy, 88 per-cent two years prior, 76 per-cent three years prior, and 62 per-cent four years prior. Previously, the most accurate level of classification for companies four years prior to bankruptcy was less than 30 per-cent. The study was concluded with suggestions for further research, in particular the applicability of multiple discriminant analysis to the three-group situation where the possible scenarios would be continued solvency, bankruptcy and acquisition.
    Last Updated:
    4 years ago
    All Rights Reserved
    Share this:


    Item Name: pdf jwh-mba-dissertation.pdf
      Download View in browser
    Activity: Downloads: 42