In the last decades, the studies developed in the field of techniques of early diagnosis of company antifunctionality conditions have been mainly directed towards the elaboration of increasingly faithful and accurate models. This has led the researchers to concentrate in primis on the improvement of the efficacy of forecasting models, putting in the background the aspects relevant to the enjoyability of these from the potential users.
Understandably, this behaviour has brought about inevitably a progressive reduction of the use of these significant instruments. In fact, it is important to specify that a greater diagnostic reliability often turned into a more than proportional degree of complexity to manage in the application time. Complexity meant not only in a technical sense, therefore the difficulty in a practical use, but also in an economical sense, such as the overall sacrifice sustained.
In the light of these observations, the present work aims to examine the forecastin models of insolvency studying conjointly with the dependability and efficiency features verifiable in them. Specifically, this study is inserted into the field of a hypothesis of public observation (on a local level) to prevention and control of socio-economic effects of company crisis. Therefore, it is easy to understand that in this context a precautionary policy assumes an essential role. The opportunity introduced by a public observation to have diagnosis models that are able to give primary information about the solvency status of a company, should allow to turn the available financial resources in a specific way, towards the companies that have been classified insolvent through predictivetechniques.
In particular, the aim of this analysis is to observe if some instruments, present in the literature and considered of easy use, offer sufficient guarantees of reliability regardless of adjustment interventions to the peculiarities of the companies examined.
Effectively, this provides substantial advantages in terms of inexpensiveness, rendering an employment on a \"large scale\" of these techniques. In more detail, the research has been led through an empirical analysis oriented to test the reliability degree of two well-known forecasting models (Altman's Z1 Score and Alberici's Z) to diagnose, in good time and with good reason, the solvency status of a sample of companies located in the territory of the province of Ferrara.
The analysis has been carried out using ex-post logic. In other words, the models have been applied to a sample of companies for which half of them had already manifested the crisis. The balance sheets of five years before the crisis have been analyzed, verifying if and in what measure the forecasts formulated through the predictive techniques have then been demonstrated to be correct.