Analyzing Settlement Agreements: A Data Driven Study of Corporate Financial Dynamics in Crisis

Descrizione: his doctoral thesis analyzes the determinants and dynamics of corporate financial distress with a specific focus on Italian firms. Financial distress is examined as a complex and evolving process that affects firms’ performance, capital structure, and survival prospects. The study investigates how key financial variables such as profitability, measured by EBITDA, and leverage influence the likelihood of distress and the ability of firms to recover through restructuring processes. From a methodological perspective, the analysis combines traditional econometric techniques with machine learning approaches. In particular, logistic regression models are employed to estimate the probability of financial distress, while Random Forest and SHAP values are used to capture non-linear relationships and assess the relative importance of explanatory variables. The empirical results show that higher profitability significantly reduces the probability of distress, while excessive leverage increases financial vulnerability. These findings are robust to several checks, including alternative performance measures, exclusion of outliers, and controls for regional heterogeneity. Overall, the thesis contributes to the literature on corporate finance and financial distress by providing new empirical evidence on Italian firms and by integrating econometric and machine learning techniques to better understand crisis dynamics and restructuring outcomes.