Experiments and Results by Modeling the Financial Domain with UML

The current research illustrates a framework for modeling the bankruptcy prediction models using the Unified Modeling Language. The developed framework uses the Anghel bankruptcy prediction model which is ranked by the local specialists as one of the most representative and actual autochthonous model. Also, our decision regarding the modeling of Anghel bankruptcy prediction function was based on the high accuracy rate of the model which is higher that 95%. Thus, through the UML modeling of the Anghel function it was intended to highlight the extent to which current economic transactions, carried out within the company (sales, purchasing, commissioning), submit through accounting operations (which supplies the balance of accounts) items in the annual financial statements information that determines the final score of Anghel function. Due to such application, any company could be able to find out at any moment the Anghel score value and hence the company's financial status (which can be either financial stability or bankruptcy risk). Owning such information could be vital for the company because, in the case of financial instability warning, the top management will have a sufficient margin of time to bring the activity to a healthy state and thus to prevent the bankruptcy settlement. Furthermore, the current UML developed model can be acquired by a add-on computer application and integrated into an ERP system within a Business Intelligence Component or a Scorecard component/module. Through such integration, the application will benefit from the computerization of other ERP modules from which will be able to retrieve information regarding the account balances. Such information is necessary to determine the cumulative, multi-level hierarchical Z-score value function as presented in our detailed UML model.