Managerial Characteristics and Default Prediction Modeling for Small Firms. Evidence from Small Italian Firms. Preliminary Findings

From the 1960s onwards, a great number of studies focused on how to combine firm data with statistical classification models so as to predict business failure. This issue became more pressing in recent years after the Basel Capital Accords linked capital requirements to the banks' models for corporate failure prediction. However, most of the studies use mainly economic-financial ratios as independent variables, and they mostly focus on large and medium enterprises. In our study, logistic regression was applied to a sample of 760 Italian small enterprises (SEs) in order to test the accuracy of a company default prediction model based on both economic-financial ratios and managerial variables (such as propensities, personal convictions, experience, skills and education). We then compared the results obtained to those from a similar model, based on the same sample of firms, which used only economic-financial ratios. Our results show that: i) the model based on both financial ratios and managerial characteristics has a higher accuracy rate than the other one, and this is the case both when logistic regression is applied to the aggregate sample and when the decisional function is separately calculated according to size, geographical area and business sector; ii) when logistic regression is applied separately according to size, geographical area and business sector, both models are significantly more accurate than when the decisional function is calculated on the aggregate sample; and iii) the greatest increase in accuracy (in point ii) is obtained when the decisional function is calculated separately according to firm size. These results suggest that including managerial variables gives a marked improvement to a SE default prediction model's success rate; and they also suggest that SEs should be assessed with specifically designed models